Protection of Space Materials From The Space Environment

PROTECTION OF SPACE MATERIALS
FROM THE SPACE ENVIRONMENT

Space Technology Proceedings
VOLUME 4
The titles published in this series are listed at the end of this volume.
PROTECTION OF SPACE
MATERIALS FROM
THE SPACE
ENVIRONMENT
Proceedings of ICPMSE-4,
Fourth International Space Conference,
held in Toronto, Canada, April 23-24, 1998
Edited by
JACOB I. K L E I M A N
Integrity Testing Laboratory Inc.,
Toronto, Canada
and
RODERICK C. T E N N Y S O N
Institute of Aerospace Studies,
University of Toronto,
Toronto, Canada
If
SPRTNGER-SCIENCE+BUSINESS MEDIA, B.V.
A C L P . Catalogue record for this book is available from the Library of Congress.
ISBN 978-94-010-3838-6 ISBN 978-94-010-0714-6 (eBook)

DOI 10.1007/978-94-010-0714-6
Printed on acid-free paper
A l l Rights Reserved
© 2001 Springer Science+Business Media Dordrecht
Originally published by Kluwer Academic Publishers in 2001
Softcover reprint of the hardcover 1st edition 2001
No part of the material protected by this copyright notice may be reproduced or
utilized in any form or by any means, electronic or mechanical,
including photocopying, recording or by any information storage and
retrieval system, without written permission from the copyright owner.
Table of Contents
Acknowledgements ix
Introduction xi
Organizing Committee xiii
Section A: AONUV Effects, Evaluation and Modeling
1. Monte Carlo Computational Modeling of the Energy

Dependence of Atomic Oxygen Undercutting of Protected Polymers
BruceA. Banks, Thomas 1. Stueber, MaryJo Norris
2. Dynamics of Atomic-Oxygen-Induced Degradation of Materials 15

T. K. Minton, J. W. Seale, D. 1. Garton, and A. K. Frandsen
3. Generalization of the Predictive Model of Erosion of Carbon-Based 33

Materials by Atomic Oxygen
1. I. Kleiman, Z. Iskanderova, Y. Gudimenko, R. C. Tennyson
4. Spectroscopic Ellipsometry Characterization of Polymers Modified 51

by Atomic Oxygen and Ultraviolet Radiation
Corey L. Bungay, Thomas E. Tiwald, Michael J. Devries,
John A. Woollam, Kim K. De Groh
5. Effect of Atomic Oxygen Exposures on the Tribological Properties 73

of Molybdenum Disulfide Lubricants
M Tagawa, 1. Ikeda, H. Kinoshita, M Umeno, N. Ohmae
6. Influence of Shielding Upon the Structure and Properties 85

of Polymer Composite Materials, Exposed on the Surface of
the Mir Orbital Complex
0. V. Startsev, v. v. Isupov, E.F. Nikishin
7. Cleanliness Support of Vehicle Fairings for Spacecraft 95

V. Sitalo, V. Tikhiy
8. In situ XPS Studies of Kapton Exposed to 5 eV Atomic Oxygen 103

E. Grossman, Y. Lifshitz, 1. T. Wolan, c.x. Mount, G.B. Hojlund
vi
Section B: Instrumentation, Sensors, New Technologies
9. A Low Cost, Lightweight Reusable Sensor for Atomic Oxygen Detection 115
1.1. Osborne, G. T. Roberts, S.B. Gabrieland A.R. Chambers
10. Large-Scale ECR-CVD Preparation ofIntegrated Thin-Film Structures 125

for Space Applications
Roman V. Kruzelecky, Asoke K. Ghosh, Ethel Poire and Darius Nikanpour
11. Non-Silicone Inorganic Polymer-Based Coatings for Atomic Oxygen 137

Protection in LEO - Part 2
E. Poire, B.G. Sellars and D. Nguyen
12. Research Aspects of Scaling Up the Implantox Technology 145

for Protection of Polymers in Space
ZA . Iskanderova, 1. I. Kleiman, Y. Gudimenko, R. C. Tennyson,
1. G. Brown, O.R. Monteiro, B.A. Banks, M1. Norris
Section C: Thermal Control in LEO
13. A Comparison of Space and Ground-Based Facility 165

Environmental Effects for FEP Teflon
Sharon K. Rutledge, BruceA. Banks, Michael Kitral
14. Evaluation of Low Earth Orbit Environmental Effects on International 181

Space Station Thermal Control Materials
Joyce A. Dever. Sharon K. Rutledge, Mark MHasegawa, Charles K.Reed
15. Some Aspects of the Degradation ofFEP Teflon Material 197

Under the Action ofVUV, the Sun and Ground Test VUV Facilities
A. Miltntchouk , M Van Eesbeek
16. Evaluation ofCOF MDPS Chromic Acid Anodization 211

in LEO Environment
C. Acket, 1. Marco, A. Pail/ous, R. Reuletand V. Viel
17. A Database for the Selection of Surfaces and Materials for Space 229
and Ground-based Applications
S. H. McCall, A.1. Clark, R.A. Ellis, R.A. Henderson, A.E. Piotrowski,
L.A. Piotrowski, MA . Reid, 1. W. Rodney, R. P. Breault, ML . McCall
18. Thermal Control Materials and Coatings Currently Being Developed 235
in France for Use in Low Earth Orbit
F. Guerardand1. C. Guillaumon
vii
19. Modification of Thermal Control Paintsby Photosil' Technology 243
1.1. Kleiman, Y. Gudimenko, Z. A. Iskanderova, R.C. Tennyson,
W.D. Morison
Section D: Charged Particles/Soft X-Rays/Meteoroidsffhermal Cycling
20. Simulated Solar FlareX-Rayand Thermal Cycling Durability 253

Evaluation of HSTThermal Control Candidate Replacement Materials
Kim. K. de Groh, B. A. Banks, E. A. Sechkar, D. A. Scheiman
21. Radiation-Induced Effects in Si0 2 Protective Coatings on 281

Polymeric Spacecraft Materials
C. Coia, A.C. Fozza, MR. Wertheimer, G. Czeremuszkin and A. Houdayer
22. Influence of Electron and ProtonFluxes on Optical Properties 291

of Protective Enamel
S. N. Skovorod'ko, A. V. Kurilovich, V. Ya. Mendeleyev, V. N. Vasil'yev
23. Meteoroid Hazard in the LEO SpaceEnvironment 295

Ya.S. Yatskiv, V.G. Kruchinenko and Yu.l. Voloshchuk
24. Characterizing the Annual Meteoroid Streams 305

Martin Beech
25. Oxidation Kinetics of VB Group Metals by Atomic and 331

Molecular Oxygen
A.G. Gusakov, S.A. Raspopov, A.G. Voropayev,
ML. Zheludkevich. and A.A. Vecher
26. The New Approach in Developing Methods for Better Protection 339
of Materials and Structures from LEO SpaceEnvironment
A.K. Galytsky, A.I. Stegnyy, V.P. Gusynin and B.M Shevchuk
Author Index 351

SubjectIndex 353
Acknowledgements
We would like to acknowled ge the following for their generous support of ICPMSE-4 ,
the Fourth International Conference on Protection of Materials and Structures from the
Low Earth Orbit Space Environment;
• The Canadian Space Agency; the Space Station Program Office

and the Directorate of Space Mechanics, part of the Directorate General
of Space Technology;
• Materials and Manufacturing Ontario (MMO)
• Com Dev Ltd.
• CRESTech ,
• SPAR Aerospace Ltd.
• ASM International , Ontario Chapter;
.The University of Toronto Institute for Aerospace Studies
As well, we would like to acknowledge the efforts of text editor Daphne Lavers, Delta
Blue Communications, Toronto and Dr. Z. Iskanderova from the Integrity Testing
Laboratory Inc. for her very enthusiastic and thorough review of the edited material.
Jacob Kleiman, Integrity Testing Laboratory Inc.

Conference Chairman
ix
Introduction
Th is publication presents the proceedings of ICPMSE-4, the fourth

international conference on Protection of Materials and Structures from the Low Earth
Orbit Space Environment, held in Toronto April 23-24, 1998.
The conference was hosted and organized by Integrity Testing Laboratory Inc.
(ITL) , and held at the University of Toronto's Institute for Aerospace Studies (UTIAS).
Twenty two industrial companies, six universities and fourteen government agencies
from Canada, USA , United Kingdom, France, Israel, Russia, Ukraine and the
Netherlands were represented by over 75 participants indicating increasing
international co-operation in this critical arena of protection of materials in space .
Twenty-seven speakers, world experts in their fields, delivered talks on a wide variety
of topics on various aspects of material protection in space. Representatives from the
Canadian, American, European and Israeli space agencies as well as from leading
space research laboratories of major aerospace industries gathered at UTIAS to discuss
the latest developments in the field of material and structure protection from the harsh
space environment.
We are very proud that this meeting is held in Canada. With the launch of its
first satellite, Alouette I, in 1962, Canada firmly established itself among the leading
nations in space exploration programs. With its seventh astronaut, Dr. Dave Williams,
in space on board the U.S. shuttle Columbia as the conference began , Canada continues
to contribute significant effort to the exploration and conquest of space. The Canadian
Space Agency , through its numerous programs such as the Strategic Technologies for
Automation and Robotics (STEAR), promotes the advancement of strategic space
technologies, including those contributing to the longevity of materials and structures
in the space environment.
This forum , as in previous years , brought to Toronto a substantial and
distinguished representation of the scientific communities of North America, Europe
and Asia, working in different fields of protection of materials and structures from the
harsh conditions of LEO space environment.
A brief glance at the achievements in space exploration in the last two years
are impressive. The recent findings by Hubble, the breath-taking journey of the
Pathfinder on Martian terrain, the stunning pictures from the Galileo mission, those are
just a few of the achievements in the last years. It is, without doubt, however, the
International Space Station that presently attracts the most effort and attention around
the world.
Countries from all over the world are involved in the construction of the
International Space Station (ISS) Alpha . The main structure of the 28-foot-long, 14-
foot-wide U.S. laboratory module and the nearly 80,000 lb (36,300 kg) of hardware
were constructed in 1995 . This includes solar array panels , rack structures, and hatch
assemblies. The signing of an agreement between NASA and the Russian Space
Agency (RKA ) in 1996 led to a $5.63 billion design and development contract with the
Boeing Corporation. Space assembl y began in June, 1998, with the Proton launch of
xi
xii
the Functional Cargo Block (FGB). Canada's contribution is the Mobile Service
System (MSS) to provide external station robotics. Japan's contribution is the Japanese
Experiment Module (JEM), and in Europe, the European Space Agency developed
both a pressurized laboratory called Columbus Orbital Facility (COF) and the
Automated Transfer Vehicle (ATV) for supplying logistics and propulsion .
A very important milestone in the development of the ISS was achieved in
January of 1998 in Washington when senior government officials from 15 countries
signed agreements to establish the framework for cooperation among the partners on
the design, development , operation and utilization of the ISS. It is also remarkable that
the first four international crews consisting of American astronauts and Russian
cosmonauts were named by the U.S. and the Russian space agencies.
For the ISS and other future space exploration projects , the safety of the crew
and the soundness of the structures will be the major concern. Questions about thermal
stability, resistance to soft and hard radiation sources, and combined effects of vacuum
ultraviolet, atomic oxygen and micrometeoroids will continue to accumulate with the
development of new materials and the increased use of polymers, plastics and
composite materials. For these reasons, the ICPMSE-4 meeting initiated in 1991
remains timely and necessary .
These proceedings are organized into four sections:
a) AOIVUV and Radiation Effects, Evaluation and Modelling;

b) Instrumentation, Sensors, new Technologies;
c) Thermal Control in LEO;
d) Charged particles/Soft X-Rays/Meteoroids/Thennal Cycling
Jacob Kleiman, Chairman/Organizing Committee/ICPMSE-4

Integrity Testing Laboratory Inc.,
April 1998
Organizing Committee
Chair:
Jacob I. Kleiman,
Integrity Testing Laboratory Inc.
Members:
R. C. Tennyson,
University of Toronto Institute/or Aerospace Studies, Canada
Bruce A. Banks,
NASA Lewis Research Centre, U.S.A.
Session Chairs
Session A: AO\VUV Effects, Evaluation and Modelling

Don Morison,
UTIAS. Toronto, Canada
Session B: Instrumentation, Sensors, New Technologies

Claude Brunet,
Canadian Space Agency
Session C: Thermal Control in LEO

Marc Van Eesbeek,
European Space Agency, Netherlands
Session D: Charged Particles/Soft X-rayslMeteoroids/

Thermal Cycling
Zelina Iskanderova,
Integrity Testing Laboratory Inc./UTIAS, Toronto, Canada
xiii
MONTE CARLO COMPUTATIONAL MODELING OF
THE ENERGY DEPENDENCE OF ATOMIC OXYGEN
UNDERCUTTING OF PROTECTED POLYMERS
BRUCE A. BANKS
NASA Lewis Research Center
Cleveland, Ohio
THO~SJ.STUEBER
NYMA, Inc.
NASA Lewis Research Center Group
Brook Park, Ohio
and
~YJONORRIS
Cleveland State Univ ersity
Cleveland, Ohio
Abstract
A Monte Carlo computational model has been developed which simulates atomic
oxygen attack of protected polymers at defect sites in the protective coatings. The
parameters defining how atomic oxygen interacts with polymers and protective
coatings, as well as the scattering processes which occur, have been optimized to
replicate experimental results observed from protected polyimide Kapton on the Long
Duration Exposure Facility (LDEF) mission. Computational prediction of atomic
oxygen undercutting at defect sites in protective coatings for various arrival energies
was investigated. The atomic oxygen undercutting energy dependence predictions
enable one to predict mass loss that would occur in low Earth orbit, based on lower
energy ground laboratory atomic oxygen beam systems. Results of computational
model prediction of undercut cavity size as a function of energy and defect size will be
presented to provide insight into expected in-space mass loss of protected polymers
with protective coating defects based on lower energy ground laboratory testing.
1.0 Introduction
Atomic oxygen in low Earth orbit (LEO) has been found to readily react with most
hydrocarbon and fluoropolymer materials [Ref. 1-5]. The rates of oxidation of most
Ll. Kleiman and R.C. Tennyson (eds.),

Protection ofSpace Materials from the Space Environment, 1-14.
© 2001 Kluwer Academic Publishers.
2
polymers have been sufficiently high that concerns about long term durability have
prompted efforts to develop polymers which are resistant to atomic oxygen attack
(Ref. 6), or atomic oxygen durable protective coatings which would prevent the
underlying material from being attacked [Refs. 2, 7,8-10].
Efforts to evaluate the effectiveness of atomic oxygen protective coatings on
polymers have indicated in both ground laboratory and in-space testing that atomic
oxygen is capable of entering defect sites (pin windows and cracks) in protective
coatings to produce undercutting through oxidation much greater in extent than the
initial size of the defect[Refs. 11-14]. Even though pin window and scratch defects in
protective coatings may represent a small fractional area of a polymer surface in LEO,
undercutting oxidation may result in structural failure of the polymer. The extent of
oxidation can ultimately be sufficiently large to allow interconnection of the undercut
oxidation sites. Although ground laboratory facilities can be used to simulate the
oxidizing effects of LEO atomic oxygen exposure, differences in the arriving atomic
oxygen energies between ground facilities and LEO cause some concern about
applying the durability information derived from the ground laboratory to the space
environment. Because atomic oxygen reaction probability is strongly dependent upon
its impact energy, differences between the ground laboratory impact energies and in-
space energies must be taken into account to reliably predict in-space durability based
on ground laboratory testing [Ref. 15].
To a great extent, the energy differences have been compensated for by using
effective fluence when measuring effects of atomic oxygen erosion. The effective
fluence of atomic oxygen is the number of impinging atoms per area that would be
required in LEO to produce the amount of erosion observed in an arbitrary energy
ground laboratory system. Thus, for ground laboratory systems which produce atomic
oxygen energies below those observed in LEO, many more actual atoms are required
to produce the same amount of erosion that would occur in LEO.
Confidence in prediction of in-space durability based on ground laboratory
testing can be improved by quantifying the dependence of undercutting oxidation on
arriving oxygen atom energy. This investigation uses a Monte Carlo computational
modelling technique to compare the differences in undercutting oxidation resulting
from atomic oxygen arrival at various energies. By comparison of the volume loss or
mass loss (by multiplying by the density) as a function of atomic oxygen arrival
energy for the same size defect in the protective coating, mass loss in space can be
projected at whatever appropriate LEO energy exists, based on known energy ground
laboratory data. Because a realistic protective coating has a distribution of defect sizes,
undercutting dependence upon defect size must be taken into account.
This paper addresses the prediction of in-space mass loss under the
conditions of fixed direction atomic oxygen attack compared to ground laboratory
mass loss at different energies for thick (polymer thickness exceeds maximum erosion
depth) polymers protected by thin film coatings which contain defects, that allow
atomic oxygen exposure of the underlying polymer.
3
2.0 Procedure
A Monte Carlo computational model was used to simulate the erosion effects of
atomic oxygen interacting with protected polymers at various energies for this
investigation [Ref. 16-18]. The computational model simulates a cross-section of a
portion of polymer being exposed to simulated atomic oxygen impact at the site of a
defect in the protective coating. Because the model is a two-dimensional model, each
defect is considered a crack or scratch type defect as opposed to a pin-window. The
polymer is simulated by an orthogonal array of up to 500x500 cells. The cells
represent groups of polymer atoms which can potentially be oxidized. The model is
capable of simulating mono-energetic or distributed energy atomic oxygen arrival as
well as angularly-distributed atomic oxygen arrival. Such angularly-distributed arrival
occurs as a result of the combined effects of Maxwellian distribution of thermal search
atoms and velocity components associated with orbital inclination [Ref. 18). The
computational model simulates the effect of atomic oxygen attack by allowing model
atomic oxygen atoms to enter the defect site and impinge upon the underlying
polymer.
Three possible events can occur as a result of the atomic oxygen impact on a
representative polymer model cell. An atomic oxygen reaction can occur which would
result in the removal of the cell; the oxygen can scatter from the cell without reaction;
or the oxygen can recombine to form non-reactive O2• The probability of atomic
oxygen reacting with the polymer cell is dependent upon the impact energy of the
atom as well as local angle of attack. Atomic oxygen which does not react and then
scatters from the polymer surface can do so in a manner which ranges from specular
to diffuse. The atomic oxygen which leaves the surface can also partially thermally
accommodate to the polymer surface such that it leaves with a lower energy than the
initial impact energy. Scattered unreacted atomic oxygen atoms are free to potentially
react at other locations within the undercut cavities or be ejected from the initial
defect entrance . The interaction parameters used for this investigation are listed in
Table I. Quantification of the parameters is based on the results of an effort to
optimize the Monte Carlo modelling interaction parameters, in order to produce
computational results in agreement with experimental observations of atomic oxygen
undercutting observed on protected polyimide Kapton samples returned from the Long
Duration Exposure Facility.
The actual parameters used for the simulation of ground laboratory atomic
oxygen exposure would differ from Table 1 in that the initial impact reaction
probability (0.11) may be different if the initial energy of the atomic oxygen was not
4.5 eY. The initial reaction probability is based on the actual atom energy and the
activation energy. For ground laboratory atomic oxygen exposures, the initial impact
energy assumed for the model is whatever energy is being considered as representative
of the ground laboratory facility.
4
Table I. Computational modelparametersand reference valuesfor LEO atomicoxygeninteraction with Kapton.
Initial impact reaction probability 0.11

Activation energyEo, eV, for energydependent reaction probability form 0.26
Atomic oxygen reaction probability angle of impactdependence 0.5
exponent, n, in cos" 8 angular dependence of atomicoxygen erosionyield
where e is the angle between the arrival direction and the local surface
normal
Probability of atomic oxygen recombination upon impactwith protective 0.13
coating
Probability of atomic oxygen recombination with polymer 0.24
Fractionalenergy loss upon impactwith polymer 0.28
Degreeof specularity as opposed to diffuse scatteringof atomicoxygen 0.4
upon non-reactive impactwith protective coating where 1 = fully specular
scattering and 0 = fully diffuse scattering
Degree of specularity as opposed to diffuse scatteringof atomicoxygen 0.035
upon non-reactive impact with polymer where 1 = fully specular
scatteringand 0 = fully diffuse scattering
Temperature for thermally accommodated atom (K) 300
Bounce limit 25
Thermally accommodated energy/actual atom energyfor atoms to be 0.9
assumed thermally accommodated
Atomic oxygen averagearrival direction with respect to initial surface 0
normal, dezrees
Initial oxvzenatomic energy, eV 4.5
Thermosphere temperature for atomicoxygen, 1< 1000
Atomic oxygen arrival plane relative to Earth for a Maxwell-Boltzmann Horizo
atomicoxygen temperature distribution and an orbital inclination of ntal
28.5°.
Because each Monte Carlo atom represents many actual atoms and each
Monte Carlo cell represents a measurable cross-sectional square of polymer, one must
calibrate the Monte Carlo simulated exposure to draw meaningful conclusions from
the predicted undercutting erosion.
The computation of how manyMonte Carlo atoms should enter through the
defect to representan actual in-space fluence was accomplished using Equation 1:
[1]
Where: N = Number of Monte Carlo atoms needed to simulatean effective

fluence ofF (atoms)
5
M = Number of atoms for wide defect (atoms)

Width of defect» thickness of polyimide
E = Space erosion yield (cmvatom)
F = Fluence in space (atoms/cur)
W = Narrow crack width (cell)
H = Wide crack width (cell)
D = Monte Carlo thickness lost (cell)
L = Length of the edge of a cell (ern/cell)
This calibration technique is shown schematically in Figure 1. Because the

Monte Carlo model requires the presence of a protective coating with a crack defect,
the computational calibration technique uses a very wide defect (shown on the right
side of Figure 1) to simulate the erosion which would occur as representative of an
unprotected material. Also, for an unprotected material , we know how much polymer
should be eroded based on the atomic oxygen fluence and in-space erosion yield (3.0
x 10'24 cnr'zatom), That information then is used to determine the proportionally
correct number of Monte Carlo atoms that should be brought into the model defect to
produce an undercut cavity in which each cell is the length (L) cm of an edge. Using
this technique with the assumptions listed in Table I, a Monte Carlo computational
model undercut cavity was produced (see Figure 2) which represents the consequence
of an atomic oxygen fluence of 7.15 x 1021 atoms/em/ exposure on protected polyimide
Kapton with a defect crack width of 4.16 x 10'4 cm. This undercut cavity and exposure
is representative of samples returned from the Long Duration Exposure Facility. The
thickness of the protective coating was assumed to be 2.45 x 10'5 em.
To measure the ratio of in-space mass loss to ground laboratory mass loss,
the number of oxygen atoms that reacted was measured using Monte Carlo
calculations both as a function of energy and defect width. Based on information from
these results, the ratio of in-space mass loss (at 4.5 eV) to ground laboratory mass loss
(at arbitrary energies) was calculated as a function of ground laboratory facility
directed beam energy.
3.0 Results And Discussion
Based on the results of the Monte Carlo computational model, the reaction
parameters optimization is summarized in Table 1. The probability of atomic oxygen
reacting with a polymer upon the first impact is given in Equation 2:
Probability of Reaction Upon Initial Impact = k exp -(EJE» [2]
Where k = 0.1165
Eo= activation energy = O.26eV
E = energy of atomic oxygen impact, eV
6
The consequence of this dependence of probability of reaction on initial

impact energy is that the number of Monte Carlo atoms required to be submitted at
low energies is much greater than the number needed to produce the same amount of
erosion as would occur with the same unprotected material at the high energies (4.5
eV) of LEO as shown in Figure 3.
The shape of the curve in Figure 3 is obviously dependent on more than only
the energy dependence of reaction probability because of the potential for scattered
atom reactions. Thus , a variety of Monte Carlo computational model interaction
parameters, as shown in Table I, play a role in quantification of this calibration.
Based on all the known interactions, it appears that directed beams at near room
temperature thermal energies require absolute fluences two to three orders of
magnitude higher than would be required if they were typical LEO 4.5 eVenergy.
Examination of the number of Monte Carlo computational model oxygen
atoms that react as a function of energy for a constant effective fluence results in a plot
such as shown in Figure 4. As can be seen from Figure 4, use of the calibration
technique causes the number of reacted atoms to be approximately constant with
energy for any specific width crack defect. This is in spite of the fact that the number
of atoms required to be submitted at low energies is orders of magnitude higher than
that at LEO energy.
The slight variation in the number of reacted atoms with energy below 1 eV
is not understood at the present time. It may be due to the particular characteristics of
interaction parameters at low energy where multiple collisions are encountered. Also
shown in Figure 4, for almost any size crack, the number of atoms reacted appears to
be proportional with the crack width . This issue is more comprehensively explored in
Figure 5, which shows there is very little influence of protective coating defect width
upon the fraction of Monte Carlo atoms that react. However, at low energy, orders of
magnitude fewer atoms react than at orbital energies . The characteristic that narrow
cracks behave similarly to wide cracks in terms of fraction of Monte Carlo atoms
reacted, allows simplifying assumptions to be made concerning the projection of in-
space mass loss based on ground laboratory mass loss data. It is not necessary to know
the size distribution of protective coating defects because they behave similarly in
terms of fraction of atoms reacted. Thus, the calibration from ground laboratory to in-
space is dominated only by the energy dependence for the calibration of effective
fluence which encompasses all the interaction parameters shown in Table I.
Figure 6 is a plot of the in-space mass loss relative to ground laboratory mass
loss as a function of energy for a fixed effective fluence of 7.15 x ]021 atoms/em",
based on Monte Carlo modelling data using the calibration techniques previously
described . This plot has a shape similar to that in Figure 4. A specific width of 4.16 x
10'4 em width defect in the protective coating was used for this simulation based on
the results in Figure 5, however, any width defect could have been used. The reference
energy for the in-space mass loss was 4.5 eY. The term "mass loss" is used as opposed
to "atoms reacted" because mass or volume is considered a matter of prime concern,
and mass loss or volume is proportional to the number of oxygen atoms reacted.
As can be seen from Figure 6, the Monte Carlo model predicted ratio of in-
space mass loss to ground laboratory mass loss has only small variations from unity if
7
the Kapton H effective fluence calibrations for operating directed beam ground
laboratory facilities are used. The predicted in-space mass loss is, at most, 10 per cent
higher and not more than 30 per cent lower than ground laboratory mass loss test
results would indicate. For highly elliptical LEO missions, energies above 4.5 eV
would exist which is the purpose of gathering predictions up to 8 eV as shown in
Figure 6. It is expected that the shape of this in-space to ground laboratory mass loss
curve shown in Figure 6 would be sensitive to material differences whose atomic
oxygen interaction characteristics differ from those listed in Table 1. However, atomic
oxygen interactions with Kapton H have been found to be quite similar to many other
hydrocarbon materials. Thus , one would be inclined to believe that using effective
fluence measurements with whatever specific material is in question, coupled with in-
space data from that material , would allow reasonable confidence that the in-space
mass loss would quite closely replicate the results of ground laboratory testing to the
same effective fluence .
4.0 Summary
A Monte Carlo computational model was used to explore the energy dependence of
atomic oxygen reacting with polyimide Kapton H at sites of crack defects in protective
coatings . The model was used in a calibrated form based on effective atomic oxygen
fluence. Using Monte Carlo model atomic oxygen interaction parameters based on
replication of undercut cavities experimentally measured from the Long Duration
Exposure Facility, atomic oxygen reaction as a function of energy and crack width was
explored. Because the fraction of reacted oxygen atoms relative to submitted atoms is
essentially constant with defect width, and only dependent upon impact atom energy,
the ratio of in-space mass loss to ground laboratory mass loss as a function of energy
for any size distribution of protective coating defects on polyimide Kapton H can be
calculated. Results indicate that in-space mass loss should be at most 10 per cent
higher and not less than 30 per cent lower than ground laboratory testing would
predict, based on tests conducted to the same effective fluence as in space.
8
5.0 References
1. Leger, L. 1., "Oxygen Atom Reaction of Shuttle Materials at Orbital Altitudes," NASA Technical
Memorandum , TM 8246 , 1982 .
2. Banks, BA, Mirtich, M. 1., Rutledge, S. K., and Swec, D. M., "Sputtered Coatings for
Protection of Spacecraft Polymers," NASA Technical Memorandum TM 83706,1984.
3. Coulter, D. R., Liang, R. H., Chung, S. Y., Smith, K. 0 ., and Gupta, A, "O-atom Degradation
Mechani sms of Materials," Proceedings of the NASA Workshop on Atomic Oxygen Effects,
David E. Brinza, editor, JPLPublication 87-14, Nov . 1986 .
4. Silverman, E., "Space Environmental Effects on Spacecraft: LEO Materials Selection Guide," NASA
Contractor Report 4661 , Aug . 1995 .
5. Banks, B. A , "The Use of Fluoropolymers in Space Applications," Modern Fluoropolymers, edited
by 101m Schiers, 101m Wiley & Sons, 1997 .
6. Connell, J. W., Hergenrother, P.M., and Smith , J.G., "Oxygen Plasma -Resistant Phenylphosphene Oxide-
containing Polyimides, " Polymer, v. 36, 1995. p. 5.
7. Rutledge, S. K., and Mihelcic, 1. M., "The Effect of Atomic Oxygen on Altered and Coated Kapton
Surfaces for Spacecraft Applications in Low Earth Orbit ," Proceedings ofthe Materials Degradation in
Low Earth Orbit symposium at the 1 19th annual meeting ofthe TMS, Anaheim, CA, Feb. 17-22, 1990
(V. Srinivasan and B. Banks, eds.).
8. Rutledge, S. K., and Olle , R. M., "Space Station Freedom Solar Array Blanket Coverlay Atomic Oxygen
Durability," Proceedings of the 38th 1nt'l. SAMPE Symposium , May 10-13, 1993 , pp. 679-693 .
9. Jaworske, D. A , de Groh, K. K., Podijil, G. M., McCollum, T. A, and Anzic, 1., "Leveling Coatings for
Reducing Atomic Oxygen Defect Density in Graphite Fiber Epoxy Composite s," J . of1ES, May/June
1994, pp. 26-31.
10. Forkapa, M. 1., Stidham, C. R., Banks, B. A , Rutledge , S. K., Ma, D. H., and Sechkar, E. A , "Atomic
Oxygen Durability Testing ofInternational Space Station Solar Array Validation Coupon," NASA TM
107212, presented at the 3rd International Space Conference on Protection of Materials and Structures
from the LEO Space Environment," University of Toronto, Toronto, Canada, April 25-26, 1996 .
11. Rutledge, S. K., and Milhelcic , J. A. "Undercutting of Defects in Thin Film Protective Coatings on Polymer
Surfaces Exposed to Atomic Oxygen," NASA Technical Memorandum, TM 101986, 1989.
12. Banks, B. A , Rutledge, S. K., de Groh , K. K., Auer, B. M., Hill, C. M., "Atomic Oxygen Protective
Coatings," Proceedings of the NATO Advanced Study Institute Conference, Pitlochry, Scotland,
July 7-19, 1991.
13. Banks, B. A, Rutledge, S. K., de Groh, K.K., Auer, B. M., Mirtich, M. 1., Gebauer, L., Hill, C. M.,
Lebed, R. F., "LDEF Spacecraft, Ground Laboratory and Computational Modelling Implications on Space
Station Freedom Solar Array Materials and Surfaces Durability," Proceedings ofthe IEEE Photovoltaic
Specialists Conference, Las Vegas, Nevada, October 7-11,1991.
14. De Groh, K. K., and Banks, B. A, "Atomic Oxygen Undercutting of Long Duration Exposure Facility
Aluminized Kapton Multi-layer Insulation," 1. Spacecraft & Rockets , 31 :45, pp. 656-664, Aug. 1994 .
15. Banks, B. A. Rutledge, S. K., de Groh, K. K., Stidham, C. R., Gebauer, L., and Lalvloureaux, C., "Atomic
Oxygen Durability Evaluation of Protected Polymers using Therm al Energy Plasma Systems," NASA
TM 106855, Proceedings of the International Confe rence on Plasma Synthe sis and Proces sing of
Materials, sponsored by the Metallurgical Society, Denver, Co., Feb. 21-25, 1993 .
16. Banks, B. A, Stueber, T. J., "Monte Carlo Computational Techniques for Prediction of Atomic Oxygen
Erosion of Materials," Proceedings ofthe NATO Advanced Research Workshop on Computer Modell ing
ofElectronic and Atomic Processes in Solids (Space and Electronic Materials) , Wroclaw, Poland, May
1996 .
17. Banks, B. A , de Groh, K. K., Rutledge, S. K., DiFillipo, F. 1., "Prediction of In-Space Durability of
Protected Polymers Based on Ground Laboratory Thermal Energy Atomic Oxygen," NASA TM 107209,
Proceedings ofthe 3rd International Space Conference on Protection ofMaterials and Structures from
the LEO Space Environment, "University ofToronto, Toronto, Canada, April 25-26, 1996 .
18. Banks, B. A., Stueber, T. 1., Snyder, S. A., Rutledge , S. K., and Norris , M. J., "Atomic Oxygen Erosion
Phenomena," presented at the American Institute of Aeronautics and Astronautics. Spac e Programs
Conference. Huntsville, AL, Sept. 23-25. 1997 .
9
111
III
br:::=J
_1_ . .-
Figure I. Monte Carlo computational model calibration.

10
111
H
500j.lm
4.5 eV
Figure 2. Monte Carlo computation model predictedundercut cavity for a fluence of

7.15 x 1021 atoms/em/ exposure at a defect width of 4.16 x 10-4 em
in a protectivecoating on Kapton H polyimide.
11
100
~
It)
•
.f
....l'I1
III
E
0
a;
o
-
10
:::E
W-
....co
III
E
....0co
o
:::E
0.1 10
Energy (eV)
Figure 3. Ratio of number of Monte Carlo (MC) computational model atoms required
at energy E; to those at 4.5 eV to cause the same amount of erosion for
unprotected Kapton H.
12
SOOOO-r-----------------------------,
Crack width em:
45000
40000 \ r:
6.36 x 10"
.
35000 .~
~ 30000 4.16x10"
~ 25000 l »->:
i
~ 20000
V-
2.20 x 10"
V
15000
10000
5000
O-+-----t----+-----t----+-----t----+----+------1
o 2 3 4 5 6 7 8
Energy (eV)
Figure 4. Number of Monte Carlo computation model atoms that react as a function of
energy for 3 widths of defects in the protective coating and a constant
effective fluence of 7.15 x 1021 atoms/em ' .
13
----------._---
'Q-- • . _ • •_ .•_ ••_ • • -<;l- . . _
---
.. -<;l- "---9 ' - " - " - " - " _ " - ' i l
~-~------~--~-~---------~
0.01
~ ~ ~ ~ 0
....
1e-4 1e-3 1e-2
Defect Width (em)
-+- 0.038eV
·· 0 ·· 0.05eV
--<p- 0.1 eV
-9 " 0.25 eV
___ 0.5eV
-0 . . 1.0eV
-+. 2.0eV
-0- 4.5eV
.. ... 8.0 eV
Figure 5. Ratio of reacted Monte Carlo atoms to submitted Monte Carlo atoms as a
function of protective coating crack width for various energy for a constant
effective fluence of7.15 x 1021 atoms/em'.
14
1.2
.
....
0
..J
::Ii
..
.e
..J
0.8
"C
C
::l
!! 0.6
.
Cl
iii
..
0
oJ
0.4
..
::Ii
u
0.
In
0.2
1:
0
0 2 8
Energy (eV)
Figure 6. In-space mass loss relat ive to ground laboratory mass loss as a function of
energy for a fixed effect ive fluence of 7.15 x 1021 atoms/em', based on
Monte Carlo modeling data using the calibrat ion techniques .
DYNAMICS OF ATOMIC-OXYGEN-INDUCED DEGRADATION OF
MATERIALS
TIMOTHY K. MINTON, JAMES W. SEALE, DONNA J. GARTON,

AND ANGELA K. FRANDSEN
Department ofChemistry and Biochemistry
Montana State University
Bozeman, MT 59717 USA
Abstract
We have investigated the mechanisms that lead to degradation of hydrocarbon-based

materials that are subject to attack by atomic oxygen. Beams of energetic oxygen atoms
were directed at liquid (squalane) and solid (Kapton) surfaces , and the reactive and
nonreactive products that scattered from these surfaces were detected with a rotatable mass
spectrometer detector. Angularly resolved flux and energy distributions of the products
revealed that OH and H20 are the initial products of ground-state OeP) reactions .
Subsequent reactions that become important on a static surface under continuous O-atom
bombardment ultimately produce CO and CO 2 , All observed products exited the surface
via thermal and nonthermal mechanisms, and the balance between these mechanisms was
dependent on incident O-atom translational energy. Preliminary results suggest that
concomitant surface bombardment by energetic atomic or molecular species can enhance
the removal rate of CO and CO 2 , Protection of a polymer surface with a coating
dramatically reduced the reactive signal that was detected and thus suggests an approach
to materials testing in atomic oxygen environments that involves in situ monitoring of the
process .
1. Introduction
The energetic collisions between spacecraft and atomic oxygen in low Earth orbit promote
oxidation and erosion of exposed polymeric materials. I Mitigation of atomic-oxygen-
induced degradation and assessment of materials durability in space have been major
challenges for more than a decade . Phenomenological data reveal the end result of an
atomic oxygen exposure under a particular set of test conditions in space or in a ground-
based facility, and models have been developed to explain the observed trends in materials
degradation.' However, the detailed nature of the interaction of oxygen atoms with a
IS
J./. Kleiman and R.C. Tennyson (eds.),
Protection ofSpace Materials from the Space Environment. IS-31.
16
surface have remained elusive . We have instituted a program aimed at uncovering the
mechanisms ofpolymer degradation under atomic oxygen attack. All poss ible interactions
are considered, including nonreactive collisions , initial reactive events , and material
removal processes . We hope that a detailed picture of atomic oxygen interactions with
polymer surfaces, as well as the experimental techniques employed to gain such a deep
understanding, will guide new and improved approaches to mitigate the problem of
materials degradation in space (e.g., better defmition of test parameters, more realistic
theoretical models, design of more durable materials).
Atomic oxygen reactions with hydrocarbon surfaces may resemble gas-phase
reactions , or they may be unique to the surface. The initial reactions are most likely to
involve abstraction of a hydrogen atom or insertion into a C- H bond. Such reactions are
known to occur in the gas phase . More complex reactions may also occur on a surface,
especially when it is under continuous bombardment. For example, initial radical products
may undergo secondary reactions with the surface, while radical sites generated by initial
and secondary reactions would be particularly susceptible to further reaction with incoming
o atoms. It is even conceivable that an energetic atom or molecule striking the surface may
not react directly, but instead promote a surface reaction by depositing its energy in the
surface . Every step in the process of material oxidation and degradation, from initial
reactive and nonreactive events to scission of the hydrocarbon backbone and removal of
carbon , will be governed by the atom-surface interaction dynamics, which in tum will
depend sensitively on exposure parameters , such as incident flux, collision energy,
approach angle, and surface temperature.
The approach that we have taken to understand the interactions of atomic oxygen
with a material is to direct atomic and molecular beams at a surface and monitor the
reactive and nonreactive products that scatter from a surface with a rotatable mass
spectrometer detector. Angularly resolved flux and energy distributions allow us to infer
the qualitative features ofthe interaction mechanisms at the surface. Such data should also
provide a sensitive test for theoretical models, which would yield a more quantitative
understanding of the interactions at a surface subjected to atomic oxygen attack.
The experimental results presented here give an overview of various experiments that
we have conducted in an effort to build a picture of what happens when oxygen atoms
bombard a hydrocarbon surface. A continuously-refreshed liquid surface has enabled us
to uncover the initial reactions between incident 0 atoms and a surface, while more
preliminary experiments with solid surfaces are aiding our understanding of the
mechanisms by which material is ultimately removed in the form of volatile carbon-
containing products, such as CO and CO 2, An outgrowth ofthe more fundamental studies
is the possibility of applying a beam-surface scattering technique to materials testing . This
application is discussed in the context of a preliminary test of a polymer that was protected
with a coating .
17
2. Experimental Details
The experiments were performed with two dist inct, yet similar, crossed molecular beam
apparatuses.' The details of the experiments conducted with a relatively low energy ,
continuous RF discharge source for atomic oxygen have been presented previously.v' A
schematic diagram of the apparatus used at Montana State University to study the
interaction of higher energy oxygen atoms with surfaces is shown in Fig. 1. A laser
detonation source" (operating at 1.9 Hz) provides a pulsed beam of energetic atomic and
molecular oxygen that is directed onto a target surface placed 92 ern from the apex of the
conical nozzle. The beam diameter at the target position was defmed by a collimating slit
to be 1.5 mm. The target and a mass spectrometer detector can be rotated about the same
axis; therefore, scattered species can be detected as a function of incident and fmal
scattering angles ()j and ()o with respect to the surface normal, The detector can also be
positioned such that the detection axis coincides with the beam axis in order to determine
the species in the beam and their energy distributions.
TO ION COUNTING
SYSTEM
QUADRUPOLE
MASS ALTER
PULSED
VALVE
SOURCE CHAMBER
/
MIRROR
• CO, L A S E R /
-CHOPPER
WHEEL
MAIN SCATTERING CHAMBER
I----l
10cm
Figure 1. Schemati c diagram of the hyperthermal beam source, the differentially pumped
scattering region containing the surface, and the rotatable mass spectrometer detector.
Figure 2 shows translational energy distributions ofthe atomic and molecular oxygen
components in the beam pulse for various operating conditions of the source . The top
panel in the figure shows the energy distributions ofthe two components in the entire beam
pulse. The bottom panel contains plots of narrowed distribut ions that were obtained with
the use of a synchronized chopper wheel that allowed the selection of a relatively small
portion of the overall beam pulse . Two sets of distributions are shown, corresponding to
18
6.1 11 «v A
1
--w
Q.
0
3.1 5.3 6.5 "10.8 eV

1 !
B
O~I,
'\
I, t
\
\
II
t \
II I \
,I
0
0 5 10 15 20 25
Energy, eV
Figure 2. Translational energy distributions of atomicand molecularoxygen components in the

hyperthermal atomicoxygen beamsused. The total beamdistributions (A) werenarrowed with
the use of a chopperwheel, allowing the selection of the beamenergy conditions in (B).
19
average O-atom energies of 3.1 and 5.3 eV. The beam always contains a significant
fraction of molecularoxygen. For the studies reportedhere,the ratio between atomic and
molecular oxygen varied from 0.25 to 1.0.
Atomic oxygen beams were directed onto two surfaces, a solid polymer film of
Kapton polyimide and a liquid film of squalane (C30H62) . A continuously-refreshed film
of squalane was producedby forminga filmon a stainlesssteel wheel that rotatedthrough
a liquidreservoir, maintained at 20°C. The wheel rotationrate was 0.25 Hz.
Products that scattered from the surfaces were detected with the rotatable mass
spectrometerdetector. The raw dataconsisted of time-of-flight(TOF) distributionsbased
on the flight time of neutralspecies from some modulated source (a chopper wheel or the
pulsed O-atom source itself) to the electron impact ionizer in the detector. TOF
distributions were collectedat a varietyof final angles for three different incidentangles.
The TOF distributions N(t) are number densitydistributions, and they may be converted
intotranslational energydistributions prE) (proportional to flux)by takinginto accountthe
proportionality prE) oc fN(t) , where t is the relevantflight time.
3. Initial Reactions with a Liquid Hydrocarbon Surface
Theinitialreactionsbetweenthe beamand a hydrocarbon surfacewere probedwiththe use

of the continuously-refreshed squalane surface. The surface was refreshed at such a rate
that multiplecollisions at a single site on the surface could be neglected.
The dominantreaction productdetected when an atomicoxygenbeam impinged on
a squalane surface was OH. Representative TOF distributions for OH are shown in Fig.
3. The panelson the left side of the figurecontaindatathat werecollectedwith a relatively
low energy RF discharge source for atomic oxygen.v' while the panels on the right side
show data collectedwith the use of the pulsedhyperthermal source. In general,the shapes
of the TOF distributions depended on the incident and final angles; however, all TOF
distributions appeared to be bimodal. We deconvoluted the bimodal distributions by
assuming that the slow component corresponded to products that were in thermal
equilibriumwith the surface and therefore exited the surface with a Maxwell-Boltzmann
(M-B) distributionof velocities. The differencebetweenthe overall TOF distributionand
the M-B component was taken to be the TOF distribution of the second, hyperthermal,
component, which corresponds to products that exited the surface before thermal
equilibrium was established.
The onlyother significantproductinthe reactionof 0 atomswithsqualanewas H20 .
Figure 4 shows representative TOF distributions for the H20 product. Unlike the OH
signals,the H20 signalsare dominated by a thermalcomponent. The hyperthermal signal
becomes much more pronouncedwith higher incidentenergies.
The possibility that the O2 component in the beam could be contributing to the
reactions leading to these product signals was checked by varyingthe 0/0 2 ratio in the
incident beam while maintaining a constant incidentenergy. Figure5 shows OH and Hp
TOF distributions collectedwith two different incidentbeams, one containing 50% 0
20
mlz =17 (OH+)

~-----------. .------------------,
1.0
<E j > = 0.22 eV
A
<E j > =3.1 eV
c
<Ef > = 0.25 eV <Ef > = 1.3 eV
0.5
\
'iii"
~
c 0.0
~
.d
---
'-
ftl
1.0 B D
:z <E j > =0.50 eV <E j > =5.3 eV.
<Ef > = 0.34 eV <E,> = 2.0 eV
0.5
o 400 800 1200 0 400 800 1200
Flight time, us
Figure3. Representative time-of-flight distributions ofOH products following reactionofoxygen

atomswithasqualane surfaceat fourincidentenergies and initialand finalangles of60· and 45°,
respectively. Signals in (A) and (8) wereproducedin an experiment that utilizeda low energy,
continuousbeamsource (see Refs. 4 and 5). Signalsin (C) and (0) wereproducedin a different
experimentthat employed the beamswhoseenergy distributions are shown in Fig. 28. Average
final energies <E> are given in each panel.
atoms and the other containing 20% 0 atoms.' The shapes of the TOF distributions are
identical, indicating that molecular oxygen was not contributing to the initial reactions
leading to OH and Hp products; they must therefore be the result ofO-atom reactions .
The dynamical properties of the hyperthermal OH product indicate that it is formed
through a direct-abstraction (Eley-Rideal) mechanism. Rettner and Auerbach have pointed
out that direct atom-surface reactions yield products with hyperthermal translational
energies that are proportional to incident energy and with angular distributions that are
asymmetric with respect to the surface normal." The fmal energies for the hyperthermal
OH component do increase with incident energy, as seen in the example shown in Fig. 3.
In addition, the angular distributions are peaked away from normal, toward the specular
direction and are much narrower than a cosine distribution (see Refs . 4 and 5).
21
o c
iii
:= 0.0
c
:::J
...
.d
--...
I'll
Z
1.0 '"
<Ej> = 5.3 eV
0
<Ef > = 1.8 eV
0.6
o 400 800 1200 0 400 800 1200
Flight time, J.1s

Figure 4. Representative time-of-flight distributions of H20 products following reaction of
oxygen atoms with a squalane surface at four incident energies and incident and final angles of
60 and 45°, respectively. The experimental conditions are analogous to those described in
0
Fig. 3.
The hyperthennal components in the TOF distributions for H20 also exhibit
dynamics indicative of a direct reaction , although the hyperthermal HzO product cannot
correspond to a single-collision mechanism. The sum of two direct reactions, initial
formation ofOH and subsequent abstraction of another H atom, is the simplest explanation
for the nonthermal H20 products.
The reactions that lead to hyperthennal OH products are reminiscent of gas-phase
abstraction reactions; however, the thermal products must be the result of unique surface
processes. We observed TOF distributions of inelastically-scattered atoms (not shown), °
which also show both thermal and hyperthermal components. These two components may
be described by the two limiting cases of inelastic scattering and trapping desorption,
respectively." The existence of a trapping desorption component implies that atomic
22
+
1 A m1z = 17 (OH )
[OJ
= 0.25
(0)
<Ej(O» =5.3 eV
<E 1(02» = 10.8 eV
- [0]
[oJ = 1.0
60°
=
fI)
c
::::)
0
-e<I:
--
....::. B m1z = 18 (H20 )
+
1
Z
o 500 1000 1500 2000
Flight time, J1s
Figure 5. Representative time-of-flight distributions (normalizedto 1.0)ofOH andH,O products

following reactionof a squaianesurfacewiththe narrowed hyperthermal oxygenbeam(Fig. 28)
at an incidentO-atomenergyof5 .3 eV (10.8 eV forO,) and incident and final anglesof 60°and
300, respectively. The incident energy was held constant while altering the fraction of 0 ,
dissociatedin the beam. The [0)/[0,] ratios were0.25 and 1.0 as indicated.
23
oxygen can be trapped on the surface long enough to reach thermal equilibrium. These
thermalized ground-state Oe P) atoms cannot react directly with the hydrocarbon surface
because the activation energy is too high. However, they may react through a mechanism
that involves triplet-singlet crossing of a long-lived collision complex." In the latter case
(or if an excited-state OeD) atom strikes the surface), an insertion product (excited
alcohol) would be formed. While such reactions may indeed be occurring, they are not
expected to yield volatile OH products, as the alcohol should be stabilized on the surface.
The more likely explanation for the thermal OH is that direct abstraction is followed by
thermal accommodation of the OH product on the surface. Once the OH becomes
thermall y accommodated on the surface, it may desorb thermally or abstract another atom
to form HP, which could also desorb thermally and account for the large thermal
component in the TOF distributions .
The dynamical information that we have been able to infer from our scattering data
have led to a qualitative picture (Fig . 6) of the initial interactions between hyperthermal
oxygen atoms and a hydrocarbon surface. The balance between the various possible
reaction pathways is certainly dependent on incident energy and incident and final angles .
Note especially that the relat ive importance of the nonthermal processes increases
dramatically at high incident energies.
Summary of Reaction Mechanisms
@\ HYPERTHERMAL
P ~
abstraction
~
HYPERTHERMAL
P P @-p E5 THERMAL
abstraction abstraction
@
o, desorption
<r
p p ~ p ~ p/ THERMAL
abstraction trapping
@
""y
C?P reaction desorption
G\ o/'
~"'lir i
trapping
p 0('0)
Insertion stabilized alcohol
Figure 6. Summary of initial reactionmechanismsinferred from theexperimentwiththe liquid

squalane surface.
24
4. Loss of Carbon as CO and CO 2
Time-of-flight distributions of products following reaction of the O-atom beams with a

solidKaptonsurfaceweresimilarto thoseobserved fromthe reaction withsqualane except
for additional large signalsat m/z = 28(CO+) and 44(COz+). Figure7 shows that reactive
CO and COz productsignalsfromsqualaneweresmallcompared withthosefromKapton.
Like all TOF distributions observed at other masses, those detected at m/z = 28 and 44
appearto containat leasttwocomponents, onethermal andtheotherhyperthermal. Similar
trends were seen previously in the reaction of lower-energy O-atom beams with a
polypropylene surface."
Kapton Squalane
<Ei > = 6 eV =
<E? 6 eV
1400 A m/z =44 (C0 2 )

+ B m/z= 4
700
~
Olo.6w----------l
1400
c m/z =28 (CO· D m/z= 2
700
01W'~----------.;-l
o 1000 2000 3000 0 1000 2000 3000

Flight time, us
Figure 7. Representative time-of-flight distributions of CO and CO, products following the

interaction of the hyperthermal beam(Fig. 2A)withKapton andsqualane surfaces at incident and
final anglesof 60° and 300, respectively.
25
The sequence of reactions that occur under continuous a-atom bombardment is
undoubtedly complex, and yet they are central to material loss from the surface. Oxides
may form on the surface through initial insertion reactions . Furthermore, once a radical
site is created, it will be susceptible to subsequent a-atom reactions. The reaction
sequences ultimately lead to CO and CO 2, as evidenced by our observations, but
understanding the mechanisms of their formation and ejection from the surface requires
a more in depth investigation. We have begun such an investigation, and the preliminary
results are presented here.
The question that is most easy to address first is how CO and CO 2 are removed from
the surface . Do oxygen atoms react directly with carbon or an oxidized carbon site on the
surface to form hyperthermal products? Or can the hyperthermal CO and CO 2 be explained
by collisional events, involving the incident a atoms, that remove these products as they
are formed through thermal reaction sequences on the surface? If products are formed
thermally, then is there a competition between thermal desorption and collisionally-assisted
processes?
Hyperthermal Effusive Oxygen

Beam Source RF Plasma Source
Figure 8. Schematic diagramshowingthe position of the RF plasmasourcewith respectto the

hyperthermal beam sourceand targetsurface.
26
Unraveling the processes that lead to carbon loss is complicated in our experiment
by the presence of molecular oxygen with high translational energies that can exceed
10 eV. Although the 02 did not appear to playa role in the initialreactions, it might react
at radical sites, or perhaps it can leaddirectlyto reactions at high energies by dissociating
on impact. Even if O2 does not react at all, it mightbe ableto enhancethe material removal
rate through collision-induceddesorptionof weakly-bound reactionproductsthat build up
on the surface.
We added a thermal oxygen atom source to our experiment, which was used in
conjunctionwiththe hyperthermal sourceto investigate the possibilityof collisionaleffects
(Fig. 8). The thermal source is an inductively-coupled RF plasma with a 0.5 rom orifice
on the end of the quartz plasma tube, which was placed 2.5 em from the target surface.
The plasma was operated in a continuous mode with an O2 pressure of 0.100 Torr and a
power of 50 W. Time-of-flight measurements on the plasmasourceshowedthat the atomic
and molecular oxygen had a Maxwell-Boltzmann distribution of velocities with a
temperature of375 K. The atomic oxygen fraction of the "thermal" beam was 20%, and
no evidence of ions in the beam was detected. Experiments were conducted with the
simultaneousexposure of the surface to the hyperthermal and thermal beams and to each
of the beams alone. The thermal beam alone did react with polymersurfaces to produce
CO and CO2, and the product energy distributions were essentiallythermal and reflective
of the surface temperature. The fact that we observed product signals indicates that
significant oxidation of the surface was taking place even withthe thermal O-atom beam.
Figure 9 shows the results of an experimentdesignedto test the role of collisions in
removingCO2 fromthe surface. A hyperthermal argonbeamwas generatedwiththe laser-
detonation source, and this beam was directed at a Kapton surface that was under
°
continuous bombardmentwith atoms from the thermal source. The top panel in Fig. 9
shows the energydistributionof the entire beam pulse (energycoveringthe range 2.5 - 17
eV) and the distribution of a beam whose high-energy tail was truncated with the use of a
synchronized chopper wheel (energy range 2.5 - 9.5 eV). The bottom panel shows TOF
distributions of the CO2 signals that were modulated with the Ar beam (the constant
background arising from the CO2 produced from the thermal beam has been subtracted).
As can be seen, the truncated Ar beam produced almostno CO2 signal, while the total Ar
beam produced a large CO2 signal. The only difference between the beams is the high-
energy tail. It thus appears that there is a process involving collision-induced production
of CO2 on a polymer surface that is under attack from atomic oxygen, and this process
appears to have a threshold energy of approximately 10 eV. Because of the threshold
behavior, a small fraction of high-energyAr atoms leads to a relatively large signal.
The existence of such a collision-induced process suggests that the O2 component'
in the hyperthermal atomic oxygen beam may be playing a large role in the observed CO2
(and possibly, CO) signal. Figure 10shows CO2 product TOF distributions from Kapton
collected with the hyperthermal O-atombeam (Fig. 2A) and with simultaneousexposure
to the hyperthermal O-atom and thermal beams. The addition of the thermal beam leads
to increasedsignal, but the shape of the distribution does not change. In fact, the shape of
the TOF distributionis very similar to the shape of the TOF distribution obtained with the
27
1
Energy Distributions
of Hyperthermal
Argon Beams
--
W
Q.
,,
-,,
" ....,
0
.............
0 5 10 15 20
Energy, eV
- Entire Beam
- Trun cated Beam
100
--z
o
o 500 1000 1500 2000 2500 3000
Flight time, J.1s
Figure 9. Top panel: Translational energy distribution of hypertherrnal Ar-atorn beam

(dashed) and truncatedAr-atom beam(solid). Bottom panel: RepresentativeTOF distributions
ofe0 2 (Or = 30°) resulting from the impactof Ar atoms (OJ = 60°) on a Kapton surfacethat was
simultaneouslyexposedto oxygenatomsfromthe RF plasmasource.
28
Ar beam. The similarity in shapes suggests the possibility that the observed signals from
the hyperthermal atomic oxygen beam come mainly from a collisional process involving
the O2 component of the beam . We have not yet investigated the effect of O, translational
energy on the observed CO 2 signals.
The collisional production of CO 2 seems unlikely to be simple collision-induced
desorption. Once a CO 2 molecule is formed on the surface , it would be expected to desorb
rapidly (in less than a nanosecond) because of its low binding energy to the surface. Even
if it did linger long enough to be struck by an incoming atom or molecule , the threshold for
knocking it off the surface should be much less that 10 eV. It thus seems plausible that
high-energy collisions above 10 eV may transfer enough energy to the surface to promote
a surface reaction that involves oxygen already present on the surface. This idea is
cons istent with the increased CO 2 signal when the thermal beam is added to the
hyperthermal beam exposure becaus e there should be an increased steady-state
concentration of oxygen on the surface.
rn'z = 44 (Co. ')

500
...e
~
U)
400
::::s 300
0
--....
z
CJ
200
100
o 500 1000 1500 2000 2500 3000 3500 4000

Flight time, J.Ls
Figure 10. Representative time-of-flight distributions of CO, following the reaction of the
hyperthermal oxygen beam (Fig. 2A) with a Kapton surface. Simultaneous exposure with the
thermal beam leadsto increased signal over the hyperthermal beam alone.
It is unclear at this time how important a collisional process is in the production of

CO 2 in our beam experiments. In low Earth orbit, polymers erode when subjected to 5 eV
collis ions with atomic oxygen. The molecular oxygen fraction in space is low compared
with our beam , so the mechanism of carbon removal would not be dominated by a
collis ional process involving 0 2. However , the number density of molecular nitrogen can
be as much as one half the atom ic oxygen number density. II The collision energy of
29
molecular nitrogen with a spacecraft surface at an orbital velocity of 7.8 km/s is - 9 eV
(with a distribution about this value), so it is possible that N2 could be assisting in carbon
removal from a spacecraft surface through a collisional process that has a threshold of
10eV. The nature of such a collisionalinteraction (includingthe mechanisms behind the
apparent thermal and nonthermal processes), awaits further study. Future experiments
should also identify the dominantmaterial removal mechanisms at different incident 0-
atom energies, with different secondaryspecies present. Such experiments could have a
profound effect on the definition of parameters for ground-based testing of candidate
spacecraftmaterials.
5. Test of Boron Nitride as a Protective Coating
The ability to monitor reaction products, such as CO and CO2, that are indicative of the
erosion of a materialunder atomic oxygenattackraises the possibility of a new approach
to testing. If the productsignalcould somehow be relatedto the reactionefficiency, either
as an absolutemeasurement or relativeto a known standard(e.g., Kapton), then materials
could be tested quickly. A test lasting one or two hours could conceivablyreplace a test
lastingoneor two days,and nopost-exposure masslossmeasurements wouldbe necessary.
We illustratethe concept here with a test of a candidate protective coating.
1500
.!c 1000
:::J
o
U
-""'..."
z 500
BN-coated
Kapton
o
o 1000 2000 3000 4000
Flight time, ~s
Figure 11. Representative time-of-flight distributions for CO 2 (0,= 30°) following the reaction
of the hyperthennal oxygen beam (Fig. 2A, OJ = 60°) with boron-nitride-coated Kapton and
uncoated Kapton.
30
We received two samples of Kapton from the Space Vacuum Epitaxy Center at the
University of Houston, TX. One was a control sample, and the other had a protective
coating ofboron nitride. This coating was 100 nm thick and was prepared by ECR-assisted
electron beam evaporation. Figure II shows the CO 2 signals observed from the protected
and control Kapton surfaces after steady-state reaction conditions had been reached. The
signal from the control surface was enormous compared with the signal from the BN-
coated surface, indicating that the coating effectively prevented the atomic oxygen beam
from reacting with the Kapton substrate. The small signal that does come from the BN-
coated surface might reflect limitations in the ability of this particular film to stop
completely atomic oxygen reactions with the underlying Kapton . Steady-state hydrocarbon
contamination might also give rise to the small signal from the coated surface. The
nominal contamination load from our source and chamber must be fully quantified before
drawing conclusions regarding such a small signal as that seen in Fig . II . Nevertheless,
one could, in principle, quantify the erosion rate of the protected Kapton relative to the
control Kapton by integrating the respective product fluxes. Standard exposure tests would
have to be done initially in order to validate and calibrate the method.
6. Conclusion
Beam-surface scattering dynamics experiments on a continuously-refreshed liquid

surface have allowed us to infer that many mechanisms, involving insertion and abstraction
with and without thermal accommodation at the surface, initiate degradation of a polymer
surface by incident 0 atoms . Analogous experiments on static polymer surfaces reveal that
carbon is removed from the surface in the form of CO and CO 2, The bimodal time-of-
flight distributions for these products indicate that thermal and non thermal mechanisms
lead to their ejection from the surface. Important additional information was gathered from
simultaneous exposure ofthe surface to hyperthermal Ar-atom and thermal O-atom beams,
which showed evidence of CO 2 production via a collisional process with a threshold near
10 eV in the incident Ar translational energy. This result raises questions about the nature
ofthe dominant material removal processes in ground-based test facilities and in low Earth
orbit. Energetic species other than atomic oxygen may be important in determining the rate
of erosion. A new approach for atomic oxygen testing of space environmental effects on
materials was suggested. In situ monitoring of reaction products (C0 2) from a Kapton
control surface and a BN-coated Kapton surface showed clearly that the BN coating
dramatically reduced the reaction rate ofan atomic oxygen beam with the Kapton substrate.
More work must be done in order to validate and calibrate the method for general use. The
study of the dynamical processes occurring when a hyperthermal atomic oxygen beam is
allowed to interact with a hydrocarbon surface is beginning to uncover some ofthe myriad
physical and chemical interactions that lead to the degradation ofmaterials in space. New
mechanisms that were previously not cons idered may in fact dominate the degradation rate
under some exposure conditions. The reaction conditions will undoubtedly govern the
31
balance between the various possible mechanisms . An understanding of the dominant

mechanisms as a function of the exposure parameters should ultimately lead to improved
guidelines for the design and testing of candidate polymers that must be durable on
spacecraft in low Earth orbit.
Acknowledgments
This work was supported by a grant from the Department of Defense Experimental
Program for the Stimulation ofCompetitive Research (DEPSCoR), administered by the Air
Force Office of Scientific Research (Grant No. F49620-96-0276) and by a contract from
Ionwerks, of Houston, TX. D. J.G. appreciates the support ofthe NASA Graduate Student
Researchers Program. We are grateful for the assistance of J. Zhang and G. LeBihan in
the collection of data and for discussions with Drs. J. Albert Schultz and A. Bensaoula.
We thank A. Tempez of SVEC at the University of Houston for preparing the BN-coated
and control Kapton samples .
References
1. L. J. Leger,NASA TM-58246 (1982); L. J. Leger and J. T. Visentine, AerospaceAmerica 24, 32 (1986);

Y. Haruvy,ESA Joumal14 , 109 (1990); L. E. Murrand W. H. Kinard, AmericanScientist81, 152 (1993).
2. Z. A. Iskanderova, J. I. Kleiman, Yu. Gudimenko, and R. C. Tennyson, J. Spacecraft and Rockets 32,5
(1995).
3. Y. T. Lee, 1. D. McDonald,P. R. LeBreton, and D. R. Herschbach,Rev. Sci.lnstrum . 40,1402 (1969).
4. D. 1. Garton, T. K. Minton, M. Alagia,N. Balucani,P. Casavecchia, and G. G. Volpi,Discuss. FaradaySoc.

108, in press (1998).
5. D.J. Garton,T. K. Minton, M. Alagia, N. Balucani,P. Casavecchia, and G. G. VOlpi, in Proceedings ofthe
/997 American Institute for Aeronautics and Astronautics Conference and Exhibit, Huntsville, AL,
September23-25,1997 ; paper no. 97-3947.
6. G. E. Caledonia, R. H. Krech, and D. B. Green, AlAA J. 25, 59 (1987); U.S. Patent No.4 894 511 (16
January 1990).
7. The calculation of the relative fraction of atomic oxygen is based on a value of 1.9 for the ratio of total
ionizationcross sections for O2 and 0 . L. 1. Kiefferand G. H. Dunn, Rev. Mod. Phys. 38, I (\966).
8. C. T. Rettner and D. J. Auerbach, Science, 263, 365 (1994).
9. M. E. Saecker and G. M. Nathanson, J. Chem. Phys. 99, 7056 (1993).
10. Y. Rudich, Y. Hurwitz,G. J. Frost, V. Vaida, and R. Naaman, 1. Chem. Phys. 99, 4500 (1993).
II. D. R. Peplinski, G. S. Arnold, and E. N. Borson, 13th Space Simulation Conference, NASA CP-2340
(Orlando,FL, 1984), pp. 133-45.
GENERALIZAnON OF THE PREDICTIVE MODEL OF EROSION OF
CARBON-BASED MATERIALS BY ATOMIC OXYGEN
Z. A. ISKANDEROVA, J. I. KLEIMAN, YU. GUDIMENKO

Integrity Testing Laboratory Inc.
4925 Dufferin Street, Toronto, Ontario , Canada, M3H 5T6
R. C. TENNYSON
University of Toronto Institute for Aerospace Studie s
4925 Dufferin Street, Toronto, Ontario, Canada, M3H 5T6
Abstract
Prediction and evaluation of erosion effects in polymeric materials and carbon fibre
reinforced plastic composites in low Earth orbit (LEO) due to interaction with
atomic oxygen are very important issues for these materials' durability evaluation
and expectation of performance in space missions. A recently- developed predictive
model and a useful correlation, introduced to explain the results of hydrocarbon
polymer erosion in LEO, have been extended to include fluorinated polymers. The
"screening" effects associated with fluorine atoms and fluorine-containing
functional groups in polymer structures have been proposed as playing the main
role in the resistance of these materials to atomic oxygen erosion in space. Special
attention was also paid to carbon fibre reinforced plastic composites. Analysis of the
erosion rate data for these materials has been performed for LEO flights of different
duration. It was demonstrated that for composites, due to their complex structure,
the erosion yield can be significantly different for short- and long-duration
missions, and in both cases, it can be predicted based on the proposed model. Using
the proposed concepts, estimates and predictions of erosion behavior can be made
for various polymers and polymer-based composites, including untested and new
materials.
1.0 Introduction
Polymeric materials and carbon-based composites have increased importance in

recent years due to their many unique advantages over other materials . For
example, polymers are lightweight and corrosion-resistant, have versatile electronic
properties and low manufacturing costs. However, due to their inherent softness and
susceptibility to degradation in oxidizing and/or ultraviolet (UV) environments,
their use has generally been limited to relatively mild service applications . These
33
Protection ofSpaceMaterials from the SpaceEnvironment, 33-50.
© 2001 KluwerAcademic Publishers.
34
materials also often have low resistance to photo-oxidative degradation, and are
rapidly eroded or etched in extremely oxidative environments, such as plasma or
atomic oxygen (AO) fluxes . Carbon-based composite materials, for example carbon
fibre-reinforced plastic (CFRP) composites and carbon fibre-carbon composites, are
lightweight, tough and rigid, making them very attractive materials for spacecraft
components. However , like polymers, composites are also eroded in extremely
oxidative environments, such as LEO ; and under fast atomic oxygen beams (FAa),
with energy E::::: (2-5) eV, that are used in ground-based space simulators.
This paper presents an extension and generalization of the recently-
developed predictive model of LEO erosion of hydrocarbon polymers on two other
families of polymeric materials, i.e., linear-chain fluorinated polymers and fluoro-
chloropolymers, that are extensively used in space applications. An approach was
found that quantitatively takes into account the protective shielding or "screening"
effect of fluorine atoms and fluorine-containing functional groups . The calculated
erosion yield values are in good agreement with flight data.
For carbon fibre reinforced plastic composites, significantly different
erosion rates have been found in short - and long-duration missions, and, for the last
case, at significantly different locations. Analysis of this data , with regard to the
composites' macrostructure and the chemical composition of the matrix polymeric
materials, allowed calculating and predicting both erosion yield values, using the
extended predictive model.
2.0 Erosion of hydrocarbon and fluorinated polymers by atomic oxygen in

LEO - generalization of a predictive model
Recently, a model was developed for interaction of AO in LEO with hydrocarbon

polymers [1,2]. Bombardment-enhanced chemical etching had been proposed as the
main general mechanism of accelerated erosion of these materials in space . The
composition and chemical structure of the polymeric material has been shown to
have a significant influence on the erosion rate . This model allows not only an
explanation but also a prediction of the erosion yield of these types of materials in
space. It was shown that the erosion yield R, LEO could be expected to be inversely
proportional to the mass density of the "effective" carbon atoms. The content of
intramolecular oxygen and the nature of oxygen-containing functional groups in a
polymer repeat unit affects this parameter. In the first-order approximation,
decarbonilation was assumed to be the main oxidative degradation reaction,
consuming intermolecular oxygen . It was shown that it is important to take into
account both decarbonilation and decarboxilation reactions in the pathway of
oxidative degradation of some hydrocarbon polymers by AO in LEO to achieve the
best agreement of the calculated or predicted erosion yield values with space flight
data . Finally, the following equation was introduced in [1,2], in order to estimate
the erosion yield of hydrocarbon polymers in LEO environment:
35
(1)
Here NT is the total number of atoms in a polymer repeat unit (PRU), Nc is the total
number of carbon atoms in PRU, Nco is the number of carbon atoms that can be
"volatilized" from the PRU due to erosion processes by both decarbonilation and
decarboxilation processes. M is the average atomic weight in PRU, and p is the
density of the material.
For many hydrocarbon polymers MJp "" const., that is brought to the last
expression in Eq. 1. This approach was extended in [1,2] for materials containing
sulfone and other functional groups that can form additional volatile products at
oxidative degradation and erosion by atomic oxygen in LEO.
It was demonstrated in many studies that, in general, fluorinated polymers
are more resistant to AO, than hydrocarbon materials [3,4], with the erosion
resistance in LEO (in comparatively short flights), increasing with the degree of
fluorination [5]. It is a common belief that fluorinated polymers are, in general,
resistant to atomic oxygen attack because oxidation of the C-F bond is an
endothermic process, and the high resistance of Teflons to AO, associated with the
high intrinsic resistance of CF2 group, was mentioned many times [6].
LEO
Perfluorinated polymers, Teflon TFE and Teflon FEP, have shown the Ro .in
some of the space flights almost two orders of magnitude less than for hydrocarbon
polymers. At this point, the possible role of FAO/VUV synergism [7], or the
influence of soft X-rays [8], or other factors [4] on the accelerated erosion and
degradation of Teflons in some LEO missions is not considered.
The high stability of Teflons in FAa and other chemical reactive
environments are often explained by the following. The two types of valence forces
binding the PTFE structure are the carbon-carbon bonds, which form the backbone
of the polymer chain, and the carbon-fluorine bonds. These chemical bonds are
extremely strong and are the key contributors in imparting an outstanding
combination of properties to PTFE. If the atoms attached to the carbon-chain
backbone were smaller or larger than fluorine, the sheath would not form a regular,
uniform cover. This sheath shields the carbon chain from attack and confers
chemical inertness and stability. It also reduces the surface energy, resulting in low
coefficient of friction and non-stick properties.
Extensive studies of etching processes on hydrocarbon and fluorinated
polymers have been performed earlier using various oxygen or oxygen-based
plasmas [9-15]. Both "in the glow" and "out of the glow" etching configurations,
i.e., downstream of the plasma, had been used. The results were compared with
flight data in [12,14,15], and some of this comparison will be discussed below.
36
Figure 1 shows the effect of structure on etch rate downstream of RF 0 -

plasma reactor for a particular family of ethylene-type polymers with different
degrees of fluorine substitution [12]. The sharp increase in etch rate is going from
HDPE to PYF (with one F atom per monomer unit, or 25 per cent F in 'H(F)'
composition), followed by a sharp decrease going to PVDF and TFEIET (both with
50 per cent F in ' H(F)' composition), and by an even lower rate for PTFE (with 100
per cent in 'H(F)' .
"C F4 JN~
0 20 40 60 80 100
PE: -cHrCH;rCH2-eHr .5
.3 15000~
PVF: -cHz-CHF-cH:z-CHF- <
PVDF: -CHr CF;rCH;rCFr Z
TFElET: -cHrCHz-CF2-cFr (alt.) 0
l-
i
N
PTFE: -cF2-cF2-cF2-c F2-
- ~
g ~
}.2
u;
l0000 ci
I-
...>:
<
II:
G
I-
.
0
II.
0
III
III
---
5000 I-
<
........... ~ II:
%
.... ~
0 20 40
TFE/ET
60
---
80
--
PTFE
"f IN " HI F)" COMPOSITION OF -CH,,Fy-CHXFy-
Figure 1
Etch rate as a fun ction of fluorine content in a series of ethy lene-type polymers, exposed downstr eam from an
oxygen plasma (- )[12), and etch rate of Kapton polyim ide [11) as a function of'Cf', concentration in the
oxygen plasm a gas feed (. _• •) (from [12 )).
37
The sharp increase and the following decrease in etch rate almost parallel
the dependence of etch rate of Kapton on the CF4 concentration in the oxygen
plasma gas feed, reported by Egitto et all [11], and also presented in Figure I. It was
shown [11] that the position of the maximum in the plot of Kapton etch rate vs. CF4
concentration depends on the optimum ratio of a to F atoms in the plasma, the
"atomic fluorine [enhancing] etching by surface reactions which make etching sites
available."
Based on the above data Ill], it was suggested [12] that the higher etch
rate for PVF vis-a-vis HDPE is associated in some way with the formation of F
atoms in the course of Oep) interaction with PVF, that can give rise to a greatly
increased etch rate by promoting the O2 ~ 20 conversion, or by further reacting
with PVF. These processes, of course, could not occur in LEO, since any F atoms
produced would be rapidly "flashed off' and not react further with the PVF.
This view is consistent with almost similar etch rates for HDPE and PVF
obtained from the space flight experiments. In PVDF and TFEIET, on the other
hand, the CF2 groups are presumed to be much more resistant to a-atom attack than
are either the CFH groups in PVF or the CH2 groups in PE; hence, PVDF and
TFEIET have very low etch rates. The similarity in etch rates for PVDF and
TFEIET can be attributed [12] to the fact that in both PVDF and TFEIET, every
CH2 group is attached to a CF2 and is thereby ' protected' to some extent against
Oep) attack. The very high intrinsic resistance of CF2 finds its maximum
effectiveness in the extremely low etch rate of PTFE, both in RF oxygen plasma
"out of the glow" exposure [12] and in LEO exposure [4, 5,12].
Figure 2 represents the data for erosion yield in LEO of a few fluorinated
polyolefins, with an increasing degree of fluorination, including also perfluorinated
materials, for a comparatively short duration space flight [5]. All of these
fluoropolymers are known to be important space materials [16].
It is clear from Figure I and Figure 2 that, opposite to etching downstream
from an a-plasma reactor [12], the erosion yield in LEO decreases with the increase
of degree of fluorination . For olefinic chlorofluoropolymers such as Aclar (or Kel-F)
chlorothreefluoroethylene, and Halar, a copolymer of ethylene and chlorothree-
fluoroethylene, R,LEO is less than for the majority of hydrocarbon polymers, and
close to some of fluorinated ones. For these materials ReLEO also decreases with the
increase of fluorine content [5].
It was recognized starting from the early 80s [10], that inclusion of one or
more chlorine atoms in carbonaceous polymers causes a strong enhancement of the
etching rate in oxygen plasma. These effects were attributed to the catalytic role that
the CI radicals and, perhaps, those radicals in exited states, are playing in a-plasma
during the process of enhanced polymers etching [10]. These processes, of course,
would not occur in space, in LEO conditions, since any F or Cl atoms produced
would be rapidly "flashed off' at the polymer films, and would not react with them
any further.
38
HOPE
_4 - / PVF (Tedlar)
--
E
o /
as
CO) 3
E
o
~
C\I
b
.,..2
x
o
---- / PVOF (Kynar)
UJ
-l
(1) 1 TFElET (Tefzel)
a: TFETefion
/ Teflon FEP
o 25 50 75 100
% F in "H{F)" Composition of - CH F - CH F -
u v x y
Figure 2
Erosion yield of polyethylene and selected fluorinatedpolymersin LEO environment [5] as a
function of fluorinecontent in the polymerrepeatingunit.
The protective "shielding" or "screening" effect of fluorine in interaction

of fluorinated materials with atomic oxygen can be illustrated by experiments with
low-energy oxygen ion beams [17] or OzlCF4 plasmas in reactive etching (RIE)
mode [13], where etching of PTFE heavily depended on ion bombardment. At
etching with an oxygen ion beam, for instance, with E = 30-40 eV or E = (40-90)
eV [17], physical sputtering of the structure started, making Teflon FEP even less
resistant than Kapton, or both Kapton and polyethylene [17].
To account for the interaction of fluorinated polymers with AO in LEO, the
concept of a "screening factor" was introduced, and associated with the presence of
fluorine atoms in different functional groups. This approach is consistent with the
one already developed for hydrocarbon polymers, and can be viewed as an extension
or generalization of the developed model and the predictive correlations , including
other families of important polymers. The results were obtained mostly for linear
chain chlorofluoropolymers and fluorinated polyolefines, that can be described, in
general form, as follows: -CHuFv-CHxFy - . A similar approach can be established for
39
fluorinated aromatic polymers, as more data will become available for these
materials.
We found, that a "screening factor" i5 can be proposed for these types of
fluorinated polymers, so that Eq. (1) can be presented in the following form:
s;LEO -M-'y'(l-o)
I
p (2)
In the "screening factor" 8, screening coefficient I was assigned to the

single asymmetric F atoms, such as those in the CHF groups. A screening
coefficient 2 was assigned to the F atoms in the CF2 group, to take into account the
special strength of this group. And finally, assuming that the hydrogen-carbon
bonds, as always, are vulnerable to AO attack, the screening factor 8 was presented
in the form:
(3)
where Nl 2 is the number of fluorine atoms from -CF2 _ groups in the PRU, NF is
the number of other (single) fluorine atoms in a PRU, and N, is the sum of all
single F, all H and all Cl atoms in PRU of the material.
In both families of polymers mentioned above, MJp is no longer a constant,
as in hydrocarbon polymers, and deviates significantly, mostly increasing with an
increase in the degree of fluorination, a fact that obviously should be taken into
account.
Finally, using Eq. (2) and Eq. (3), we arrive at an expression
(4)
The flight data for erosion yield R,LEO for fluorinated materials and
fluorochloropolymers [5], plotted versus y'mod, as given by Eq. (4), is presented in
Figure 3, and shows a well-defined linear correlation.
40
4..----------------
Tedlar
ow
-iu1
a:
10 15 20
I
- [2NF2 N1
y mod = M. 1- F + F I
P Y 2NF2+N
F P
Figure ;}
Erosion yield in LEO for selected fluorinat ed and fluorinated-chlorin ated polymers [5), presented
versu s y lmt>d
The "screening factor" should be taken into account only for the above-
mentioned two polymer families, and does not affect any results of the model
regarding hydrocarbon polymers presented in [1,2].
Figure 4 presents a collection of data on erosion yields from a number of
LEO flights for a variety of hydrocarbon polymers (see [1,2] for details), combined
with the data for fluorinated polymers [5], presented in Figure 3 and plotted
altogether versus y'mod.
The results for fluorinated and fluorinated-chlorinated polymers, presented
in Figure 4, fit well the linear trend and are in good agreement with the regression
line that was calculated earlier (1,21 only for hydrocarbon polymers using Eq . (I).
The final regression line includes its extension to the area of lower erosion yield
values that are characteristic of fluorinated polymers, as can be seen from Figure 4.
The predictable capability of the modified correlation R, LEO (y'mod) is extended now
for two more important polymer families.
41
-E 7
- 0
.J!! 6
C')
E
..,.
0
5
b.....
0
w
~
- x 4
3
CD
a:
2 .~LEO
Data for AnalysIs
'\9 caJculated Data
·1 • LEO flight Data
0
0 15 20 25 30 35 40 45 50
F2+N
- l' p._2NF F ]
l' mod =M.
P 2NFF2+NP
Figure 4
Erosion yield in LEO for a variety of hydrocarbon polymers [1.2] and fluorin ated polymers (5). presented
versus y'mod. Th e solid line represents the extended least squares fit to the experimental flight data .
An experimental ground-based study of erosion behavior by fast atomic

oxygen (FAO) for some of these materials (Teflon, Tefzel, Aclar, and Halar) was
undertaken in the UTIAS Space Simulator. The facility and testing procedures
remained the same as for hydrocarbon polymers and were described in [7). The
experiment involved five to eight hours ofFAO exposure. i.e., to FAO fluences (jlFAO
:::: (1.2 - 1.8) x 1020 at . em", with following mass loss measurements and scanni ng
42
electron microscopy (SEM) evaluation of the materials surface morphology.

Preliminary results from these experiments confirmed a linear trend, similar to the
one in Figure 3. The erosion yield values for Aclar, Tefzel, and Halar have been
almost 35-50 per cent less than the space data. A special study of combined
exposure to FAD + VUV should be done to evaluate a possible influence of VUV
radiation on erosion of these materials.
For all of these polymers except Teflon, matte surface appearance of the
FAO exposed area due to morphology change, was obvious even to the naked eye,
but it was less than for the control Kapton sample. Some results of SEM analysis of
surface morphology change for Halar (blue) are presented in Figure 5a. Results of
SEM analysis of the eroded surface for a Kapton witness sample, exposed to FAD in
the same run, are presented for comparison in Figure 5b at the same magnification .
The comparison of surface morphologies at high magnification (Figure 5a
and Figure 5b) indicates a similar general phenomenon of "rag-like" or "carpet-
like" surface morphology development under FAD exposure for an aromatic
hydrocarbon (Kapton), and for a fluorinated-chlorinated polymer (Halar). The
difference is mostly in the shape and height of the specific features, developed on
every material that is easy to distinguish at high magnification. The surface
morphology of Halar is quite close to the one that was developed on this polymer by
LEO exposure to a slightly higher fluence [l8J.
3. 0 Carbon Fibre Reinforced Plastic Composites Erosion In Space
Advanced composites are finding an increasing role in the development of space

hardware, i.e., in spacecraft structural, payload, power and thermal control
subsystem applications.
Composites of interest for spacecraft applications are mainly polymer
matrix composites, because of their lightweight, dimensional stability, high
structural rigidity and low thermal expansion. Depending on the selection of fibres
and the design of the lay-up, the coefficient of thermal expansion (CTE) can be set
to a low positive value, zero, or even a negative value. When combined with low or
zero CTE optical materials, an entire optical system can be almost completely
desensitized to temperature changes.
Polymer resins, mostly used in CFRP composites, include a variety of
epoxies, polyimides, polysulphones and phenolics. The resin determines the
interfibre mechanical properties. It was shown in many shuttle flights and space
missions, that the high-energy fast atomic oxygen in LEO interacts with both
carbon/graphite fibres and the resin matrix, causing surface degradation, erosion
and change in surface morphology (Fig. 6). Due to the erosion phenomena, polymer
matrix composites exposed to LEO environment showed thickness loss and optical,
mechanical and thermal property changes [I 9J. For these reasons, the evaluation
and prediction of erosion yield of composites in space missions of various durations
is one of the important issues.
43
For multi-component space materials and real structural components such

as weaved impregnated materials, or parts and structures of carbon fibre reinforced
plastic composites, not only the difference in erosion yield of the components plays
a significant role, but also the macro-structural organization and location of the
material. This general statement was confirmed in many short- and long-duration
space missions . The main trends in the erosion behavior of advanced CFRP
composites in LEO flights can be explained using an extension or generalization of
our predictive model to such materials. Here, the predictive capability of the model
for various hydrocarbon polymers, as well as for graphite, can be used extensively.
The numerous results of erosion phenomena in LEO and the data for measured
erosion yield for various composites in short- and long-duration space missions [19-
26] have been compiled and carefully analyzed.
From the analysis conducted, it can be suggested that the "normal" erosion
yield as defined in [26], would be better to use for the comparison of the erosion
yield of various CFRP composites in different flights and for those at different
locations . "Normal" erosion yield is the value that one would measure for a material
placed normally to the incident atomic oxygen flux.
All different erosion yield results collected in various flights of bare CFRP
composites with advanced high -performance hydrocarbon polymers , used as matrix
materials, can be divided conveniently into short-term and long-term LEO exposure
results. These materials are mostly cured epoxy resins, polyimides, polysulfone, etc.;
structural formulae for some of them are presented in Figure 7. Similar trends were
found for all these composites , i.e., a significant difference in the erosion yield has
been shown in short- and long-term missions [19-26]. Despite the more obvious
effects of erosion for bare composites in missions of longer duration, the average
erosion yield was mostly 2-3 times less than for similar materials in short-term
missions . To explain this difference , both the microstructure and the macrostructure
of the composites should be taken into account.
It is well known that for CFRP composites , a resin layer 10-20 urn thick
appears on the surface. This layer is the first one to be affected by erosion,
depending on the total AO fluence and materials location . It plays the dominant role
in surface degradation at low fluences, i.e. in short-term missions .
For "high" fluences (long-duration missions) more and more carbon fibres
will appear on the surface when the AO fluence increases , and the erosion will
finally be the predominant limiting factor at high doses of AO exposure . This
means that the erosion yield of the polymer matrix itself should be mainly
considered for short missions , and can be estimated by using our predictive model.
In long-duration missions, the expected or predicted erosion yield is supposed to be
limited mostly by the erosion of carbon fibres, i.e., it should be close to the erosion
yield of carbon .
The erosion yield results from space flight experiments of different
duration are presented in Table I. The erosion yield values were also calculated for
both "extremes" using our predictive model, and are also shown in the appropriate
columns . As can be seen from Table I, the ReLEO values as measured for short- and
44
long-duration missions are in very good agreement with those calculated using the
predictive model approach.
4.0 Conclusions
A recently-developed model of interaction of one of the major factors in the fast

atomic oxygen LEO environment with hydrocarbon polymers has been generalized
and extended to a number of important families of polymeric materials and to
composites. The effective carbon content, introduced in the originally-developed
model, was shown also to be one of the limiting factors for erosion of fluorinated
polymers in space in LEO conditions, and by FAa in ground-based testing. A
concept of fluorine-based "screening effects" was introduced in this study for
fluorinated polymers and fluorochloropolymers, and a quantitative expression was
developed for the screening factor. A good agreement of calculated results and
flight data has been proven for polymeric materials of such types. The predictive
capability of the modified or generalized model is extended now to two other
important space-related polymer families.
An extensive analysis of all available flight data from various space
missions was performed for carbon fibre reinforced plastic composites with
advanced high-performance polymers, used as matrix materials. Those materials
represent high-performance, thermo-stable cured epoxy resins, polyimides,
polysulfones, etc. A significant difference in erosion yield of the composites was
found for short- and long-duration missions. This phenomenon can be explained by
the influence of the composite's macrostructure on erosion processes. The higher
erosion yield values in short missions can be associated with the erosion yield of the
matrix resin, which appears at the surface of a composite material, and the lower
erosion yield is mostly limited by the carbon fibres. The erosion yield values for
both situations, predicted by the generalized model, are in good agreement with
flight data.
45
(a)
(b)
Figure 5
Scanning electron microscopy images of polymer film surfacesafter 7 hours ofFAO exposure; with
atomic oxygen fluence <P FAO = 1.7 x 1()20 at . cm-2. (a) Halar (blue), x 20,000; (b) Kapton H (wit-
ness sample), x 20,000
46
Figure 6
Cross-sectional scanning electron microscopy image of an epoxy-carbon fiber compos ite after pro-
longed LEO exposure on LDEF mission . The upper ply of the composite exposed to the space envi-
ronment is seen strongly eroded. Notice the directionality of the erison pattern . Magnification x1,000
47
~-o ~~~
~.I ~
q.. '-'
~ MONOMER RESIN &'n.,
:N-@- C~ -@-N:
~ q.
, ~h
,~
,'-' \ ,-"c
~
d'~<:) CURE AGENT 0 ...;r a)
~NOS
-@-? -@-
0 N,H
H" II 'H
o
Ul
CaHtcNA
o
-(Q)-o~-
c)
d)
e)
Figure 7
Structural formulae for the matrix mat erials of advanced carbon fibre reinforced plastic composites: a) Epoxy
resin 5208; b) Idealized stru cture of PMR- I 5 polyimide [27); c) PEEK polyetheretherketone; d) PI 700
polysulfone (Udel); e) Victrex PES polyethersulfone.
48
Table I. Erosion Yield Determination for CFRP Composites in LEO
MATERIALS EROSION YIELD RELEO, crrr'zat
Composite Material "Short-term" (low dose) "Long duration "(high

(Polymeric Matrix) exposure, dose) exposure,
20 21
q> '" (2-3)' 10 at-ern" q> '" (6.9-7.7)' 10 at-ern"
Flight data Predicted Flight data Predicted
value value
Graphite/epoxy 5208 2.9 [3,19] 0.9-1 [20] 1-1 .2
(Epoxy resin 5208) 2.7
Graphite/epoxy 934 (2-2.8) 0.9-1 [21] 1-1.2

(Epoxy resin 934) [3,19,26 ]
2.6
Graphite/polyimide 3.1 [22] 1.3 [23] 1-1.2

(Polyimide PMR-15)
2.8
Graphite/PEEK 2.2-3 (22) ---- 1-1.2

(PEEK)
2.2
I -1.2
Carbon/Carboncomposites 1.1-1.2 0.9-1.1 (20)
(Carbon) [22,24] 1. -1.2
Graphite/P170 2.2 -2.4 (0.92 -1) 1-1.2

[3,25] [26]
(Polysulfone) 2.2
Graphite/ polyethersulfone ---- (0.89-1) 1-1 .2

(Polyethersulfone) [26]
2.2
49
5. 0 REFERENCES
l. Iskanderova, Z. A, Kleiman , J. I., Gudimenko, Y. I., and Tennyson, R. C., (1995), Influence of Content
and Structure of Hydrocarbon Polymers on Erosion by Atomic Oxygen, Journal of Spacecraft and
Rockets, 32, 878-884.
2. Kleiman, J. I., Iskanderova, Z. A., Gudimenko, Y. I., Lemberg, V., Talas, D., and Tennyson, R. C.,
(1997), Predictive Models of Erosion Processes in LEO Space Environment: A Basis for Development
of an Engineering Software in Computer Modeling of Electronic and Atomic Processes in Solids,
Kluwer Acad. Publ., 277-287.
3. Banks , 8. , (1990) Atomic Oxygen, LDEF Materials Data Analysis Workshop, NASA CP 10046 , 191-
216 .
4. Banks, B., (1997), The Use of Fluoropolymers in Space Applications , in Modem Fluoropolymers, Ed.
T. Scheirs, 103-113.
5. Koontz, S. L., Leger, L. 1., Vizentine, J. T., Hunton, D. E., Cross, 1. 8., and Hakes, C. L., (1995),
EOIM-3 Mass Spectrometry and Polymer Chemistry : STS-46, Journal of Spacecraft and Rockets, 32,
483-495.
6. Koontz, S. L., Leger, L. 1., Albin, K., and Cross, J, (1990), Vacuum Ultraviolet Radiation/Atomic
Oxygen Synergism in Materials Reactivity, Journal ofSpacecraft and Rockets, 27, 346-348.
7. Tennyson, R. C., (1991), Atomic Oxygen Effects on Polymer-Based Materials, Canadian Journal of
Physics, 69, 1190-1208.
8. Milinitchouk, A, Van Eesbeek, M., Levadou , F., (1996), Soft X-Ray Radiation as a Factor in the
Degradation of Spacecraft Materials, Proceedings of Third International Space Conference ICPMSE-3,
Toronto, Canada" 57 .
9. Hansen, R. H., Pascale, 1. V., De Benedictis, J., and Rentzepis, P. M., (1965), Effects of Atomic Oxygen
on Polymers, Journal ofPolymer Science, 3, 2205-2214.
10. Taylor, G. N., and Wolf, T. M., (1980), Oxygen Plasma Removal of Thin Polymer Films, Polymer
Eng. and Science, 20,1087-1092.
I I. Egito, F. D., Emmi, F., Hormath, R. S., and Vukanovic , V., (1985), Plasma Etching of Organic
Materials,J. Vac. Sci. Tech. 3 (3) , 893-904.
12. Golub, M. A., Wydeven , T., (1988), Reactions of Atomic Oxygen (Oep» with Various Polymer Films,
Polymer Degradation and Stability , 22, 325-338.
13. Egito, F. D., Matienzo, L. J., and Schreyer, H. 8., (1992), Reactive Ion Etching of
Poly(tetrafluoroethylene) in O 2 - CF 4 plasmas, J. Vac. Sci. Techno!., AIO, 3060-3064.
14. Koontz, S. L., Keith, A., and Leger, 1., (1992), Atomic Oxygen Testing with Thermal Atom Systems : A
Critical Evaluation, J. Spacecraft and Rockets, 28, 315-323.
15. Whitaker, A. F., Jang, B. Z., (1994), Oxygen Plasma Environment: Its Effect on Polymers, SAMPE
Journal, 30, 30-4 I.
16. Gilmore , D. G., (1994), Satellite Thermal Control Handbook. The Aerospace Corp. Press, California.
17. Banks, 8. A., Rutledge , S. K., Paulsen, P. E., and Stueber, T. J., (1991), Simulation of the Low Earth
Orbital Atomic Oxygen Interaction with Material by Means of an Oxygen Ion Beam, NASA TM
101971 ; Vered, R., Matlis, S., Grossman, G., Marorn, A., and Lifshitz, Y., (1994), Degradation of
Polymers by Hyperthermal Atomic Oxygen, Surface and Interface Analysis .
18. Kamenetzky, R. R., Linton, R. C., Finchkenor , M., Vaughn, J., (1993), Effects of Atomic Oxygen on
Polymeric Materials Flown on EOIM-3, in Proceedings of LDEF, Third Post-Retrieved Symposium,
NASA CP 3275, Ed. A. Levin, 1011-1023.
19. Reddy, R., (1995), Review : Effects of Low Earth Orbit Atomic Oxygen on Spacecraft Materials, J. of
Mat. Sci., 30,281-307.
20. LDEF Materials Workshop'91, (1992), Ed. Stein, B. A , and Young, P. R., NASACP-3162 (2) .
21. George , P. E., Dursch, H. W., Hill, S. G., (1993), Space Environment Effects on LDEF Composites: A
Leading Edge Coated Graphite-Epoxy Panel, in Proceedings of the 2nd LDEF Post-Retrieved
Symposium, NASA CP 3194, 923-939.
22. Gillespie , T., (1993), EOIM-3 Data Analysis Workshop , JPL, Arcadia, California.
23 . George, P. E., (1992), LDEF Materials Workshop '91. Ed. Stein, B. A., and Young, P. R., NASA CP-
3162 (2), 543-552.
24. Spady , B. R., Synowski , R. A, Hale, J. S., DeVries, M. 1., Woollam , 1. A., Moore, A. W., and Lake,
M., (1995), LEO Degradation of Graphite and Carbon-Based Aboard Space Shuttle Flight STS-46, in
Proceedings ofthe Third LDEF Post-Retrieved Symposium, NASA CP 3275, 989-997.
50
25. Whitaker, A F., Kamenetzky, R. R., Finchkenor, M. M., Norwood, J. K., (1993), Atomic Oxygen
Effects on LDEF Experiment AOI71, in Proceedings of Second LDEF Post-Retrieved Symposium,
NASA CP 3194 , 1l25-1l35.
26. Tennyson, R. C., Cool, G. R., and Zimcik, D. G., (1993), Space Environmental Effects on Polymer
Matrix Composites as a Function of Sample Location on LDEF, LDEF-69 Month in Space, Third Post-
Retrieval Symposium, NASA Conf. Publ. 3275 , 613-630 .
27. Engineered Materials Handbook, V. I , Composites, (1987) , Ed. Dostal, C. A, ASM International.
SPECTROSCOPIC ELLIPSOMETRY CHARACTERIZATION OF
POLYMERS MODIFIED BY ATOMIC OXYGEN AND ULTRAVIOLET
RADIATION*
COREY L. BUNGAY, THOMAS E. TrWALD, MICHAEL 1. DEVRIES,

AND JOHN A. WOOLLAM
Center for Microelectronic and Optical Materials Research,
Dep t. ofElectrical Engin eering,
University ofNebraska-Lincoln, Lincoln, NE, 68588-0511
KIM K. DE GROH
Nation Aeronautics and Space Administration
Lewis Research Center
Cleveland, Ohio '
Abstract
Atomic oxygen (AO) and ultraviolet (UV) radiation-induced surface modifications of

several space application polymers are investigated with spectroscopic ellipsometry
(SE). The polymers studied include Kapton" polyimide, FEP Teflon" , CV-1144-0
silicone and polyarylene ether benzimidazole (PAEBI). Using in situ and ex situ SE
covering a large spectral range from 190 nm to 14 urn, various parameters and trends
of the polymers are measured as they are exposed to AO and/or UV light.
UV light greater than 200 nm wavelength is used to modify the surface of
PAEBI and Kapton. Thickness and index of refraction as a function of depth of the
UV-modified surface layers is quantified. The modified layers are optically-modelled
assuming an exponentially graded layer, as opposed to standard models that often use
either a single homogenous layer or linearly-graded layer. Measurements show that
UV light causes a decrease in the index of refraction in both materials. The damaged
layer in Kapton is approximately 500 nm into the polymer and approximately 200 nm
in the PAEBI.
Two oxide forming polymers, CV-1144-0 silicone and PAEBI, are exposed to
AO in an oxygen plasma. Protective oxide layers that form on the polymers are studied
"Research supported by NASA Lewis Grant NAG3-1802

51
J.l . Kleiman and R.C. Tennyson (eds. },
52
using SE. Oxide formation trends, rates of oxidation, amount of porosity, volume
percentage of water, and change in optical properties as a function of AO exposure are
quantified. Furthermore, studying the change in molecular bond vibrations in the
infrared allowed a quasi-chemical analysis of the silicone.
Single wavelength ellipsometry over a large range of incident angles is used
to measure changes in optical density of aluminized FEP Teflon after being exposed to
the Low Earth Orbit environment aboard the Hubble Space Telescope. The purpose of
the study is to help support physical density measurements acquired by NASA Lewis
Research Center. The optical density measurements are in good agreement with
physical density measurements . These measurements help support the theory that
excessive heating was a major contributor to increases in density of the FEP returned
from the second servicing mission.
1. Introduction
Orbital spacecraft like the International Space Station (ISS) and Hubble Space
Telescope (HST) are subject to the harsh Low Earth Orbit (LEO) environment.
Neutral species of atomic oxygen (AO), UV radiation, thermal cycling, and micro-
meteroid impacts all contribute to degradation of space-application materials [1, 2, 3].
Space system designs demand consideration of the LEO environment to ensure proper
function , reliability, and lifetime.
In this study, spectroscopic ellipsometry (SE) is used to monitor changes in
polymers induced by AO and UV radiation. The basic concepts of ellipsometry are
illustrated in Figure 1. An incoming light beam of known polarization state encounters
a material under study. Upon reflection off the material surface, the polarization state
is converted to some elliptical polarization dependant on the physical properties of the
material. Ellipsometry measures the change in polarization state of the probing light
beam. The measured values are expressed as an amplitude ratio (\jI) and a phase
difference (~) . These values are related to the ratio of Fresnel reflection coefficients, R,
and R. for p- and s- polarized light, respectively. The Fresnel coefficients are derived
from solutions to Maxwell's equations. The "fundamental equation of ellipsometry" is
given as:
R
P = tan('II)e "~ = -L. =
Rs
-:%p,
p
e;
E
i
(I)
Esi
where " i " and "r" refer to incident and reflected, respectively. Because ellipsometry
measures the ratio of two values, it can be highly accurate and very reproducible. The
ratio is a complex number, thus it contains "phase" information (~) , which makes the
measurement very sensitive to ultra thin layers. [4, 5, 6].
53
1. linearly polarized light ...

E p-plane
3. elliptically polarized light!
2. reflect off sample ...

plane of incidence
Figure 1.
Description of polarization state change due to reflection.
To obtain material properties from ellipsometric data, the sample structure is

modelled to determine the "theoretical" results expected. Unknown parameters, such as
film thickness or optical constants (n and k) can be determined by "fitting" the model-
generated data to experimental data through a non-linear regression . The quality of the
"fit" is interpreted by a weighted mean squared error (MSE) [7]. It is also important to
consider correlation between "fit" parameters. If parameters become strongl y corre-
lated, a unique solution can not be determined. Therefore , a "best fit" is one that
achieves the lowest MSE with low parameter correlation.
The optical properties of an isotropic material can be described using two
parameters. These parameters characterize how a material will respond to excitation by
light of a given wavelength . One representation of the optical constants is the complex
index of refraction , given as:
Ii = n +ik , (2)
where n is the real part of the index of refraction and k is the extinction coefficient.
The index n, describes the speed light travels through a material compared to
propagation in a vacuum . It can also be considered the optical density of a material.
The absorption of light in a material is governed by the extinction coefficient, k.
Alternately, the optical constants can be described using the complex dielectric
constant, given as :
e = &\ +i&2 (3)
The complex dielectric function is related to the complex index of refraction of a
material by
i=li 2 ~
Both types of optical constants are commonly encountered in the literature, and the
choice of optical constant type is generally determined by the application.
54
2. Characterization of UV Irradiated Polymers
2.1 EXPERIMENT
Bulk optical constants of each polymer (each approximately 127 um thick) were
obtained with a variable angle spectroscopic ellipsometer (VASE) over a spectral range
from 190 to 1700 om. The polymers were modelled as optically isotropic, however they
are potentially anisotropic. To avoid complex anisotropic modelling, care was taken to
orient each sample in the same direction for all measurements. Also, roughening the
backside of the films with sandpaper eliminated backside reflections. By only
collecting light reflected from the sample surface, the effects of anisotropy were
insignificant.
Each polymer was irradiated by UV light from a 150 watt xenon lamp in
normal room ambience. In situ SE monitored the sample during irradiation, as seen in
Figure 12. Kapton" and PAEBI are known to absorb water in air. Since the experiment
was performed in air over a long enough period of time to absorb water, the samples
were exposed to room ambience for a long period of time to ensure saturation . In this
way, the samples would not be absorbing water during the experiment and changing
properties.
To xenon lamp
Ellipsometer
detector unit
Figure 2.
Experimental setupto acquirespectroscopic ellipsometry data during irradiation of polymers.
The lamp power supply current and voltage were fixed at 8.9 amps and 17
volts, respectively, which coincide with settings during calibration. The calibrated
spectral output of the lamp, measured by Opto-Cal. Inc., is given in Figure 3. The
polymers studied were significantly absorbing only in the UV region of the spectrum
(>3.1 eV), thus any modification of the polymers was assumed due to photons at
energies greater than 3.1 eV. Photon energies required to dissociate many molecular
bonds found in space application polymers are given by Joyce Dever [2].
55
The UV light was transmitted through a fiber optic bundle onto the polymers
with 25 mm from the end of the bundle to the sample. The in situ ellipsometer
acquired data at 88 wavelengths simultaneously, spanning a spectral range from 280 to
760 nm. After long-time UV irradiance, the optical data saturated (changed little with
further irradiance) . The exposure to UV light was stopped and ex situ VASE
measurements were again taken over a spectral range from 190 to 1700 nm.
200
180
160
El40
c
N
E 120
.!:!
~100
III
.
u
e
'0
80
~ 60
40
20
300 400 500 600 700 800 900 1000 1100

Wavelength (nm)
Figure 3.
Spectral output of xenon lamp used to irradiate polymer films.
2.2 OPTICAL MODELLING OF ELLIPSOMETRIC DATA
Optical constants of bulk and heavily UV modified layers were obtained from analysis
of ex situ VASE data. Both sets of optical constants (pristine and heavily modified
material) were parameterized using three Gaussian oscillators. Each Gaussian
oscillator is described in the data analysis software by an amplitude, a center energy,
and a broadening term, plus an additive constant, I>ro, to account for absorptions outside
the spectral range [8].
Pristine polymer optical constants (those determined before irradiation)
remained fixed as "substrates" in irradiated polymer optical models. The UV radiation
damage was initially modelled as a single homogeneous layer on top of the pristine
polymer substrate. This single-layer model did not provide a satisfactory "fit."
Therefore, "graded layer" models that vary the optical constants throughout the UV
modified layer were used. In this approach, the modified layer is subdivided into N
sublayers (N=15) of equal thickness. The effective dielectric function of each sublayer
is determined from a linearly-varying volume fraction of two constituent materials, in
56
this case pristine (bulk) and "heavily UV modified" materials. The effective dielectric
function, B, at a given depth from the surface, d, is given as:
e» fa (d)&a +(1- la(d»)&b, (5)
where Ga.b are the dielectric functions of the two constituent materials and fa is the
relative volume fraction of material a.
Linearly-graded and exponentially-graded layers were investigated. Both
graded layer models use Equation 5 to obtain the dielectric function. For a layer
thickness, t, the normalized depth from the surface is given as:
o=d (6)
t
The volume fraction for linear grading is directly related to the depth:
la(8) = 8 (7)
Whereas, the volume fraction for an exponentially-graded layer is given as:
l a(8) = Ase- B8 (8)
where As is the percent conversion at the surface and B is a constant. After long
exposures As will equal 100.
Model simplicity, lowest MSE, fit parameter 90 per cent confidence limits,
parameter correlation, and physical realism of all three models were considered in
determining the best optical model to use. The MSE was greatly reduced by using a
graded layer over a single layer model in all cases. The MSE was only slightly reduced
from a linearly-graded to an exponentially-graded model. This small reduction in MSE
alone probably was not enough to warrant use of the more complex exponentially
graded model. However, by looking at the 90 per cent confidence limits, as well as at
how physical the results were, we concluded that the exponential grading was the best
model for this analysis. Based on the above discussion, results are now presented in
detail using the exponentially-graded model.
The VASE data taken before irradiation were parameterized, as discussed
above, using three Gaussian oscillators. To obtain the optical constants at the top
surface after heavy UV irradiation , the exponentially-graded layer model was used, in
which the damaged material was also modeled with three Gaussian oscillators. The
Gaussian parameter values (amplitude, center energy, broadening, and E:oo) for bulk
polymer optical constants were used as starting values in the regression for the UV
modified layer parameters. Only the amplitudes of the first two longer wavelength
oscillators, Boo, and the exponential B parameter were allowed to vary in the regression
fit to the VASE data. Thus, center energy and broadening values for all three
oscillators, and the shortest wavelength oscillator amplitude value were fixed at the
bulk polymer values. This resulted in good fits to the data and minimal correlation.
For in situ SE data analysis, the exponential grading parameters (As and B)
were fit to the spectroscopic data taken at a series of exposure times. The optical
constants for the pristine substrate and "heavily UV modified" materials were obtained
from ex situ VASE data, as discussed above, and were not allowed to vary.
57
2.3 DISCUSSION OF UV INTERACTION WITH KAFTON AND PABBI RESULTS
2.3.1. PAEBI
As seen from Figure 4, the index of refraction and extinction coefficient decreased
when PABBI was irradiated. Data were fit using three oscillators centered at 190, 280,
and 330 nm. The 190 nm Gaussian oscillator amplitude of the modified layer was fixed
at the bulk PABBI value because there was no significant change with irradiance.
Regression analysis results showed significant decreases in oscillator amplitudes at 280
and 330 nm. Furthermore, the regression fit was significantly improved by varying the
f:<XJ model parameter. Allowing f:<XJ to vary causes a broad band index shift down from
the pristine PABBI index starting value.
2.2 1.00
:c: 1.9
c:
0
0.86
0.71
--o·
m
X
S·
0
U 1.5 ::I
--e
mL..
0
1.2
0.57
0.43
o
0
<D
::Ji
0
0.29 iii·
X
Q)
"C
c:
0.8 I"\(\
0/ \
\
\
0.14 -
::I
~
0.5
0 300 600 900 1200 1500
Wavelength (nm)
Figure 4.
Optical constantsof bulk and UY modifiedPAEBI obtainedfrom YASE"' measurements
Data analysis suggests a broad band decrease in the index (f: <XJ decreased). This
is likely due to the material becoming less dense and/or the polymer heating up and
water evaporating out, leaving small pockets of void. Photons of high enough energy to
dissociate molecular bonds within the PABBI molecule (some dissociation energies
given in reference 2) were striking the PABBI, thus altering the chemistry and
changing the materials ' optical density. In addition, the PABBI is known to absorb
water, and high intensity light incidence on the sample is likely to have some heating
effect. However, the contributing factor of each to the change in the ellipsometric data
was not determined. Separating the effect of each of these factors is worthwhile for
further experiments.
Amplitude decreases in UV absorption bands (seen in Figure 4) are partially
due to higher energy radiation dissociating molecular bonds. Molecular unit structures
absorb at selective wavelengths, so small molecular structural changes can bring about
perceptible changes in the UV absorption spectra. The most important electronic
58
energy changes occurring when polymers absorb UV radiation include (1) transitions
between bonding orbitals and antibonding orbitals or (2) promotion of non-bonding
electrons (unshared electrons) into antibonding sigma or antibonding pi orbitals [9].
Low Photon fluxes below 300 nm suggests most effects are probably due to excitation
of pi-electrons.
Analysis of in situ ellipsometry data provided refractive index depth profiles
for increasing UV exposure. The index depth profiles (determined from regression fits
to equation 8) for various exposure times at a wavelength of 500 nm are given in
Figure 5. The index exhibits a steady decrease until reaching saturation, with the
saturation index at the surface being 1.71 compared to the bulk index of 1.81 at 500
nm wavelength. The damage depth, at saturation, is approximately 200 nm from the
surface.
182
1.8
..... 10 min. UV Exposure

c 1.78 -e- 50 min. UV Exposure
0
:;:: --&-1 00 min. UV Exposure
U
.......CII
III -M- 200 min. UV Exposure
1.76 -&-300 min. UV Exposure
a:
.... ___ 375 min. UV Exposure
0
><
CII 1.74
"D
.E
1.72 UV Damage Lay er BulkPAEBI
1.7
Surtace.O 100 200 300 400 500
Depth from Bulk/UV Affected PAEBI Interface (nm)
Figure 5.
Index of refraction depth profile showing the depth and amount ofUV induced modification into PAEBI film
2.3.2. KAPTO~ H POLYIMIDE

Kapton" was irradiated for approximately 33 hours, with in situ ellipsometry data
acquired during the first 762 minutes only. Figure 6 compares pristine Kapton and UV
modified Kapton optical constants. Similar to the PAEBI, UV radiation appeared to
cause decreases in the index of refraction over all the measured spectrum (6.,
decreased), as well as the amplitudes of the UV absorption bands (Gaussian oscillators)
at 290 and 220 nm.
The index of refraction depth profile at a wavelength of 500 nm, seen in
Figure 7, shows little change until approximately 100 minutes of irradiation (there was
no measurable change before that). At 100 minutes of irradiation , the index only
59
decreased by approximately 0.008. At saturation, the index at the surface decreased

by approximately 0.066, and there were measurable effects nearly 500 nm into the
film.
2.2 1.5
~
1.2 "E
Q)
'0
0.9 IE
Q)
Figure6.
- KaptoniD>-Gaussian, n o Optical constantsfor Kapton"
- - - uv Affected KaptonGil-Gaussian, n 0.6 o
-KaptonGil-Gaussian, I<. 'H c:
and UV modifiedKaptonobtained
-- -UV Affected KaptonGil-Gaussian, o from VASEGil measurements.
k 0.3 n
.5
Jj
300 600 900 1200 1500
Wavelen~th nrn)
1.83 F;::::;::;::;:~iiijiiiiiiii;;o;;;;;;;:;;;::;:::;;::;:::;;::;:::;9
1.825
1.62
1.61
:c 1.61 -1~1 min
-201m;n
-301 min
Figure 7. .~ 1.80 -401 min
Indexof refraction depth profile ~ 1.6 -501 min
-631 min
Showingthe depth and amount &1.79 -761 min
ofUV inducedmodification
into Kapton'" film.
.
'0
~ 1.78
1.79
UV Affected Layer Bulk Kaplon"

.5 1.7
1 .775~====:::;::====:;=~==::;::::===~--J
o 200 400 600 800 1000
Surface
Depth from BulkIUV Affected Kapton" Interface (nm)
Kapton" took longer to reach saturation than PAEBI. The times required
were 1970 minutes and 375 minutes, respectively. A possible reason is that PAEBI
has an absorption band in the region correspond ing to the largest UV photon flux in
the xenon lamp spectrum (300 to 400 nm). Figure 8 is a plot of the extinction
coefficient and photon penetration depths (calculated form the measured extinction
coefficient) in the UV spectral region for both Kapton" and PAEBI. From this figure
it is seen that UV photons will penetrate deeper into the Kapton", thus dissociation
of more molecular bonds is required to reach saturation. This is also why the
damaged layer is nearly twice as deep in the Kapton" compared to the PAEBI.
1200
'U
:T
1000
o
g o. 8'
::J
c:
<1l 600 ~
'0 o. ::J
~ <b
41l
o 600 ~
U s::J
l§ o.
400 0
U
c:
<b
'E.
;n 200 :T
~ ~ ~ ~::2::::':~ 0
220 240 260 280 300 320 340 36 0 380 400
Wav e le ngth (nm)
60
3. Study of AO interactions witb oxide forming polymers
3.1 EXPERIMENTAL HARDWARE
In much of this, work an infrared spectroscopic ellipsometer (IR-SE) was used. This
instrument acquires data in the infrared, covering a spectral range from approximately
4000 to 700 em" (2.5 to 14 11m) by 8 em" . The IR-SE utilizes a rotating polarizer,
rotating compensator optical configuration. The IR light source, for the initial studies,
is a commercially-available infrared spectrometer (BOMEM MB102). However, for
some of the later measurements, the light source was upgraded to an argon plasma arc
source. The IR-SE optical setup is shown in Figure 9.
Additionally, two ellipsometers in the UV-visible spectrum were used. First,
for fast data acquisition, an ellipsometer that acquires 88 wavelengths simultaneously
over a spectral range from 280 to 760 nm was used. A large spectral range from 190 to
1700 nm was measured with a standard UV-VIS-NIR, monochrometer based, rotating
analyzer ellipsometer.
COMPU TER
CO NTROL
c~~--ill .. IR MICHELSON
SOURCE
20 :1 HeN e LASER
.,
MOTOR
CONTROL
....- - --c:J1;J ROTATING POLARIZER
THE TA STAGE (SA MPL E )

2-THE TA STAGE (ANALYZER,
DETECTOR)
Figure9
Instrumentsetup for the infraredvariable angle spectroscopic ellipsometer
For all experiments , the RF plasma asher quartz vacuum chamber was
pumped down to a base pressure of 20 mTorr and then back-filled with high purity O2
gas to approximately 70 mTorr. The power was set at 100 watts (13.56 MHz). At these
20
operating conditions , the "effective AO flux" was approximately 1.45x10 atoms/em"
61
hr. The "effective AO flux" is determined by mass loss of Kapton polyimide witness
coupons presuming erosion yields (3xlO-24 cnr'zatom) from in-space tests [10].
3.2. EXPERIMENTAL PROCEDURE AND ANALYSIS
3.2.1. CV-1144-0 Silicone

Study of AO interac tions with CV-1l44-0 silicone were performed exclusively using
the IR-SE. Th e silicone and oxide (formed on the silicone surface by AO) are very
closely index-matched in the visible region of the spectrum . Thus, standard UV-Vis-
NIR ellipsometry has little sensitivity to this material structure. However, in the IR
there is large optical contrast because the molecular bond spectra of the silicone and
oxide are largel y different. Thus , IR ellipsometry is very sensitive to the oxide on
silicone structure.
Thin films of CV-1l44-0 silicone were spin-coated onto optically-thick gold
films on silicon wafers. Opaque gold was used as the substrate because it is highly
reflective and relatively featureless in the IR. IR-SE data were acquired on the
optically-thick gold films prior to applying the silicone films. The index of refraction
(n) and extinction coefficient (k) of the gold substrates were obtained from a direct
"inversion" of the ellipsomet ric parameters ' 'fI ' and ' I.!.'. Prior to exposing the samples
to oxygen plasma, IR-SE data were acquired and analyzed to obtain silicone layer
thickness and optical constants. The silicone optical constants in the IR were described
using a summation of Gaussian oscillators. The Gaussian oscillators describe
molecular bond vibrational absorptions present in the IR. A total of seven Gaussian
oscillators were used. The resulting optical constants are given in Figure 10.
2.0 I I I
1.3
m
-c: ~
1.71- --n
c:: -------k - 1.0 ::::l
Q.
0
o'
+::
--
o 1.4 f-
cu J/ v ::::l
()
i
~
Q)
0.7 0
~
~ ,
1.1 f- ,I
0 "
"
'I ,~
"'l, o
roO
X " :", 0.3
,, ,,
"
Q)
"C 0.81-
, I
",I I
, .....
::::l
c:: " , ,,
I
"
,'
"
"
,
I
I ~ :. ::s:;:
:1 '.. ~~_ ... , _ ... __ __ _-l __ _ __ __ _r ___ __ _ J \~ -- - --- - ____ __
0.5 \1 \ •.
0.0
a 1000 2000 3000 4000
Wave Number (em -1)
Figure 10
Measu red CV-1 144-0 silicone optical constants covering a spectral range from 700 to 4000 em"
(2.5 to 14.2 urn)
62
The silicone samples were exposed to AO for increasing lengths of time. After
each exposure, IR ellipsometry data were acquired. The AO treated surface layer was
optically-modelled as a three constituent Bruggeman effective medium approximation
(BEMA) [11]. In a three material EMA, the wavelength dependent "effective dielectric
function", 8(11.), is given by:
e ( /L1) = fia 8 a(IL ) - e(IL) + fb eb(IL)-e(IL) + fic ee(IL)-e(IL) , (9)

ea(IL) + 2e(IL) eb(IL)+2e(IL) ee(IL) + 2e(IL)
where A is the wavelength, 8", b, c(A) audla. b. c are the wavelength dependent dielectric
functions and volume fractions (fa + fi, +j., = 1) of materials a, b, and c, respectively. In
this case Si02 was material a, H20 was material b, and void was material c. Void is
used to account for any porosity in the film.
The parameters allowed to vary in the optical model included the EMA layer
thickness, silicone layer thickness, volume percentage of water, and volume percentage
of void (film porosity). The same model was fit independently to each data set from the
different AO exposure times.
3.2.2 . PAEBJ
PAEBI optical constants were acquired on samples of bulk PAEBI film (approximately
127 microns thick) over a spectral range from 0.19 to 14 urn. The back-surface of the
film was roughen to eliminate any back-surface reflections and appear as optically
"bulk" material. The measured optical constants in spectral range from 190 to 1700
nm are given in Figure 4 and the measured optical constants in the IR are shown in
Figures 11. Since the PAEBI samples were essentially bulk materials , optical constants
in the IR were obtained from a direct inversion of the ellipsometric parameters.
1.90,-----,---,---,-----,---,---.,...--,------,---,---, 0.80
x
m
......
S·
c 1.73 0.60 ......
0
o
+:; o'
o ::J
--n
i'5 1.55
.'.'
I -------k
0.40 00
(1)
3i
.'.' 0
(D'
x
<I> 1.38 .'
I' ,.:'
"
II 0.20 ::J
"0 ......
C :: :~: :
- ,I
Ill'.',
II
::
"
..,I,'
~
:1,' ~l ~ J ~)~, __ "i:
1.20'------'-------'----'-----"-=c:.===.L.;;'O"""'--'---'----'------' 0.00
1000 1500 2000 2500 3000 3500
Wave Number (em -1)
Figure II
PAEBI optical constants covering a spectral range from 700 to 4000 em" (2.5 to 14. um)
63
The samples were exposed to AO for increasing lengths of time. After each
exposure, data were acquired with both an 88 wavelength ellipsometer (UV-Vis) and
the IR-SE. The oxide thickness was determined from analysis of the IR data. The oxide
thickness was obtained by modelling the oxide layer in the transparent region of the IR
(1700 to 4000 em") with a Cauchy dispersion relationship given as:
n( A) = An + ~~ + ~; , (10)
where n(It} is the wavelength dependent index of refraction; An, Bn, and e n are the
Cauchy parameters; and It is the wavelength in microns. The oxide thickness obtained
from IR ellipsometry data analysis was then applied to the UV-VlS model and was not
allowed to vary. With the thickness of the oxide known, nand k of the oxide were "fit"
at each UV-Visible wavelength (280-760 nm) .
3.3 DISCUSSION OF MEASUREMENT RESULTS: AO INTERACTION WITH

SILICONE AND PAEBI.
3.3.1. CV-JJ44-0 Silicone

Two samples of silicone on gold were studied. The silicone layer on the first sample
was quite thin (104 nm) with the second sample being much thicker (241 nm). The
first sample was exposed to oxygen plasma for a total exposure time of 250 minutes
and did not appear to completely reach saturation, likely due to its thickness and the
poor quality of the film (large number of "pin holes" present). The oxygen apparently
was able to reach the silicone without the formation of a continuous protective oxide
layer. The second sample formed appeared to reach saturation after an AO fluenee of
approximately 3.6 x 1020 atoms/em' . The general trends that the layer thicknesses and
volume percentages followed were similar for both samples.
As seen in Figure 12, the conversion of silicone to oxide follows an
exponential trend to saturation . The reaction rate in space may differ due to the fact
that the mean energy of oxygen atoms is different for an RF plasma (0.1 eV) and LEO
(4.5 eV), however, the general trend should be similar as long as the silicone layer is
thick enough. Once the oxide reached a saturated thickness of 56 nm it appeared to
become fairly stable. Presumably, the saturated oxide thickness in LEO would be
thicker than 56 nm because the oxygen atoms have much more energy and will
penetrate deeper. The large reduction in CV-1l44-0 thickness (- 240 to 100 nm), seen
in Figure 12, is a result of AO dissociating carbon and hydrogen pendent bonds in the
film, and forming volatile species (CO, CO2 , H2 and H20) (12, 13]. Thus, a large
majority of the molecule is lost.
64
300 .------ - - -- - - - - - - - - -- - - - --,
250 ~ Silicone Layer Thickness (nm)

La ~
-a- Oxide Layer Thickness (nm)
ye
200 I~
r
Th
ick150 \
~
ne
ss I
(n 100 -v
m) I
0 0 0
0
o 246 8
Effective AO Fluence (X 10"20)(atoms/cm2)
Figure 12
Measured silicone and oxide layer thicknesses with increased exposure to oxygen plasma
As seen in Figure 13, the ellipsometric ' '¥ ' parameter has absorption peaks at
1650 and 3400 em" whch indicated the presence of H20 in the oxide layer. From this
figure, it is seen that IR-SE is very sensitive to the constituents of the oxide layer. As
seen in a plot of the measured percentage of H20 in the film (Figure 14) also follows
an exponential trend to saturation. At saturation, the measured volume percent of
water trapped in the film was approximatel y 25 per cent.
50r----~---r--~--r--"'T""""----r"-____r----,
11l
Q)
Q)
L..
C)
Q)
"0 Hp absorptions -Unashed CV-1144
C
..... Ashed 10 min.
- - -Ashed 60 min.
IJ - -Ashed 300 min.
~ Si0 2 absorption
1000 2000 3000 4000

Wave Number (ern")
Figure 13. Ellipsometric data showing presence of water in the oxide surface layer
65
According to this optical model, at lower oxygen fluences the film has a large
volume percentage of void (seen in Figure 14), indicating that the film is quite porous.
As the fluence increased, the volume percentage of void in the film decreased to zero
in an exponential trend. This can be explained by hypothesizing the dynamics of the
film conversion. The AO initially dissociates the weaker pendent bonds of the methyl
and phenyl groups very rapidly, producing the discussed volatile oxides. This leaves
chains of Si-O with large areas of void around them. As the exposure to AO is
increased the Si-O bonds dissociate and reconfigure into a more dense Si02 film. This
is shown graphically in Figure IS.
60
~
50
I ~ Volume Percentage of Wate r (%)
I -B- Volume Percentag e of Void , (Film Porosity) (%),

l-
~
~40
Cl
co
'E
~30 ~
:u
a..
Ql
§20
\/ -v
~
10
A
o ~
o Effectfve AD Fluence 6<10"20) (atoms?cm"2)
8
Figure 14
Measu red silicone and oxide layer thicknesses with increased exposure to oxygen plasma
CH 3
a
\
I
COz, Hz, HzO , CO
I
Si - ° ---=> {si- 0-Si} =:::;> HzOand
SiOz with some
SiOH
I n
n
Figure 15
Graphica l representation of suggested dynam ics at CV-1144-0 silicone surface as AO converts it to an oxide
66
3.3.2. PAEBI
The measured oxide thicknesses obtained from analysis of the IR ellipsometry data
were plotted as a function of oxygen plasma exposure. PAEBI could not be studied to
saturation because the surface became too rough after a certain amount of exposure to
the oxygen plasma. The probing light beam became so scattered upon reflection that
there was not high enough intensity reaching the detector to acquire data. However,
Figure 16 shows the oxide layer growth trend to a point at which the sample becomes
unmeasureable.
180
160
140
.______ 1
....
I
I
'
»>
"E
c
120
;; 100 »: I
IJl
(])
~ 80
/ I -+- Oxide Thickness (nm)
.2
~ 60
./
40
I
/
20 /
~
o
o 0.1 0.2 0.3 0.4 0.5
Effective AO Fluence (X101\20){atoms/cmI\2)
Figure 16
Measured PAEBI surface oxide layer growth trend
In Figures 17a and 17b, a comparison is made of PAEBI oxide optical

constants in the UV and visible spectral regions at increasing times of AO exposure.
The optical constants cover the spectral region that the ellipsometer measures (280 to
760 nm). This comparison shows how dramaticall y the surface layer optical properties
change as a function of AO exposure. It was observed that both the index of refraction
and the extinction coefficient changed as the PAEBI was converted to an oxide. It is
seen in Figure 17b that a PAEBI absorption band at 340 nm is significantly reduced in
magnitude and at the same time a new absorption band appears to form at 380 nm. The
magnitude of the extinction coefficient at 340 nm before exposure to AO is
approximately 0.65, whereas, after 5 minutes of exposure it is reduced to less than 0.2.
67
3.0
-t:; - - Unashed PAEBI, n
2.7 ---- ·Oxide After 1 min. AO Exposure, n
c::: - - .Oxide After 3 min. AO Exposure, n
0 - - Oxide After 5 min . AO Exposure, n
+:l
2.3 - - Oxide After 7 min. AO Exposure, n
o
ro
I-
- - Oxide After 10 min. AO Exposure, n a)
-'Q) 2.0
- - Oxide After 15 min. AO Exposure, n
I-
0 1.7 -. --.. _--

X
Q)
"'C
c::: 1.3
1.0
200 300 400 500 600 800
Wavelength (nm)
0.60
~
-
c:::
Q)
'0
0.50
0.40
- - Unashed PAEBI, k
---- ·Oxide after 1 min. AO Exposure, k
- _ . Oxide after 3 min. AO Exposure, k
- - Oxide after 5 min. AO Exposure, k
lE
Q) 0.30 - - Oxide after 7 min. AO Exposure, k
- - Oxide after 10 min. AO Exposure , k b)
0
o 0.20
- - Oxide after 15 min. AO Exposure , k
c:::
0
0.10
~
c:::
~ 0.00
W
-0.10
200 300 400 500 600 700 800
Wavelength (nm)
Figures 17a and b
Comparison of the measured PAEBI oxide index of refraction (a) and extinction coefficient (b) for increased
exposure to oxygen plasma
The index of refraction at 600 nm decreased from approximately 1.8 to a

value of approximately 1.2. For this large of a decrease in the index, the surface oxide
is likely quite porous. The exact optical constants of the surface oxide are not known,
but even if they were lower than PAEBI, it is not likely that they would be as low as
1.2 without having a fairly large percentage of porosity. Furthermore, the oxide is
likely porous because of the manner in which it is fanned. Presumably, when AO
interacts with PAEBI it dissociates the majority of the atoms in the molecule leaving
68
only the phosphorous oxygen double bond (P=O). Thus, there is likely a large amount
of void surrounding the phosphorous oxygen molecule that remains.
4. Optical density measurement of FEP Teflon (pristine and returned from

Hubble Telescope)
4.1 EXPERI1v1ENT
Three samples of aluminized FEP Teflon retrieved from the Hubble Space Telescope
(HST) and one pristine FEP sample were characterized with variable angle ellipso-
metry to measure the optical density (n). Two samples, HST-SMI and HSTSMI-200,
were returned from the first servicing mission (December 1993) and the third ,
HSTSM2, was returned from the second servicing mission (February 1997). Sample
HSTSMI was studied as retrieved and Sample HSTSMI-200 was heated to 200°C in
the lab prior to the study. The sample was heated by Kim DeGrogh at NASA Lewis
Research Center. The goal of this work was to measure the optical density (index of
refraction) of the FEP and correlate that to the physical density measured by NASA
Lewis Research Center.
The aluminum layers on the backside of the samples were first removed with
a sodium hydroxide solution and then the backsides of the samples were roughened
with a small "sand blaster" . This was done to eliminate backside reflections and make
the samples appear as bulk material. Reflection ellipsometry data were acquired over a
range of angles (30 0 to 800 by 0.5 0 ) at a single wavelength (500 nm). An ellipsometric
angle scan provides the most acquire absolute measurement of the index of refraction
at that particular wavelength for bulk transparent materials . Thus, the index (optical
density) of each sample at 500 nm was measured and compared to determine if
exposure to LEO changed the density, hopefully showing the same trend seen with
physical density measurements at NASA Lewis.
4.2 RESULTS AND DISCUSSION
The pristine sample index of refraction (at 500 nm wavelength) was measured to be
1.3503 ± 0.0006. The data also indicated that there was a thin layer on the surface with
a lower index than the FEP, presumably surface roughness. The surface roughness was
modelled using a 50150 Bruggeman effective medium approximation (BEMA). The
BEMA mixes the FEP optical constants with void (n= I) to produce a layer with a
lower index of refraction than the FEP. The thickness of the roughness layer was
measured to be 19 A. The ellipsometric ' /).' data shown in Figure 18 have rounded
edges as it passes through the Brewster angle (/). at 900 ) indicating there is a thin layer
of lower index on the surface. If /). goes above 1800 or below 00 then there is a layer
with a higher index of refraction on the surface. The one data point in Figure 18 that
goes below 0 0 is likely noise.
69
200 I
150 "' -Model Fit

---- -Exp E 500nm
-
en
Ql
Ql 1001- -
'-
Cl
Ql
-0
c 50f- -
<l
~
.,
Of- .'
j
-50 I I I
30 40 50 60 70 80
Angle of Incidence (0)
Figure 18
Experimental and corresponding model data from pristine FEP showing rounding of edges due to surface
roughness
Sample HSTSMI index was measured to be 1.3507 ± 0.0009. This is same as

the pristine FEP within the error of the measurement. However, if the numbers are
correct there is a 0.03 per cent increase in the optical density. Furthermore, t1 data
appears to go above 180° and below 0° near the Brewster angle, suggesting there is a
thin layer on the surface with a higher index than the FEP. The surface layer is so thin
that it is impossible to obtain its index of refraction, so an index had to be assumed.
Therefore, optical constants of Si02 were assumed (n = 1.462 at 500 nm).
There are several possiblities as to what the surface layer is: 1) a thin layer at
the surface of the FEP becoming more dense than the rest of the bulk film due to
environmental degradation : or 2) a contamination layer could be present. Si02
contamination is often present in spacecraft in LEO. The thickness of the layer is
approximately 14 A.
Sample HSTSMl-200 (sample heated to 200°C) had a significant increase in
its index of refraction over the two previous samples. The index of sample HSTSMl-
200 was measured to be 1.3703 ± 0.0008. This is a 1.46 per cent increase over the
pristine sample. Furthermore , there appeared to be a surface layer with a higher index.
Again, the surface layer was assumed to have Si02 optical constants. The measured
thickness of this layer is approximately 30 A. The t1 data from sample HSTSMl-200 is
given in Figure 19. This data is a good example of how t1 will go above 180° and
below 0° if there is a surface layer of higher index.
70
Generated and Experimental

250 I I
,I
200 ,J" -Model Fit

- - - _.ExpE 500nm
150
en
Ql
...
Ql
Ol
100
Ql
-0
c 50
<l
0
-501-- -
-100 I I I
30 40 50 60 70 80
Angle of Incidence n
Figure 19
Experimental and corresponding model data from sample HST-FEPI-200 showing effects due to a thin surfa ce
layer of higher index than the heated FEP
The absolute index of refraction for sample HSTSM2 (returned from the
second servicing mission) could not be obtained . The extremely small size of the
sample made it very difficult to acquire adequate data . Also, the sample does not lie
flat and there is very little specular reflection making it nearly impossible to ensure
proper alignment, which is critical for ellipsometry.
The results from this study are in very good agreement with and support
physical density measurements acquired at NASA Lewis. The physical density results
were not known prior to ellipsometry data analysis , thus they did not sway the results
of this study. The results from the physical and optical density measurements are
summarized in Table I .
Table 1 Comparison of measured optical and physical density

Sample Optical % change Physical % change
Density (n) from pristine Density from pristine
3
(2Jcm )
Pristine 1.350310.0006 N/A 2.138 N/A
HSTSMI 1.350710.0009 0.03% 2.14 0.09%
HSTSMl-200 1.370310.0008 1.48% 2.171 1.52%

HSTSM2 No data 2.183 2.06%
Sample HSTSM2 was found to have the largest increase in physical density,
but we were not able to verify that result. However, all other results are in good
agreement and thus support the theory presented by NASA Lewis researchers. The
aluminum backside of sample HSTSM2 was exposed to direct sunlight on the Hubble
71
due to the FEP blanket cracking and curling over. The direct exposure to the aluminum
side caused maximum temperatures to reach 200°C, significantly higher than normal.
Therefore, because a significant increase in density was measured only when sample
HSTSMI was heated, it is suggested that the increase in density of sample HSTSM2 is
due to elevated temperatures.
5. Conclusion
Spectroscopic ellipsometry was used to study environmental degradation of space-

application polymers in both ground-based simulations and returned from long
duration space flights . The degradation induced by both UV and atomic oxygen were
studied.
Ellipsometric data taken on samples of Kapton® and PABBI after UV
exposure were fit to a layer model that graded the index of refraction exponentially .
From this analysis, the degree of degradation of these polymers due to photons greater
than 200 nm wavelength was quantified. Dynamic ellipsometric measurements, taken
during exposure, provided information about the degree, depth and time dependence of
degradation . UV radiation appears to cause decreases in the index of refraction and
extinction coefficient in both PAEBI and Kapton" . The damage appears to reach
saturation within approximately 370 and 1970 minutes for PAEBI and Kapton" ,
respectively. These exposure times are much lower than for long duration space flights .
However, once the materials reach saturation , further significant damage due to the
UV radiation is not expected. Uv-induced (photon energies < 6.2 eV) changes
observed in these polymers are not likely significant enough to threaten the lifetime of
a spacecraft.
Analysis of oxide forming polymers with spectroscopic ellipsometry proved to
be a valuable means of determining the dynamics of how these materials interact with
AO. Trends in how the surface was converted to an oxide were obtained from an IR
ellipsometry study of AO treated CV-1144-0 silicone and PAEBI. The surface oxide of
the silicone appeared initially to be quite porous, and as the AO fluence was increased ,
the oxide becomes a more dense Si02 layer. Absorption peaks at 1650 and 3400 em"
indicated that some H20 is formed and trapped in the film when AO interacts with
CV-1l44-0. At saturation, the measured volume percentage of H20 was approximately
25 per cent. Changes in CV-1144-0 optical properties appear to saturate at an AO
fluence of approximately 3.6 x 1020 atoms/em'. Analyses of ellipsometry data acquired
on AO-treated PABBI revealed the oxide index of refraction was significantly lower
than bulk PAEBI. The surface oxide index at 600 nm wavelength was approximately
1.2, whereas bulk PABBI index was approximately 1.8. It is in the authors experience
that for the oxide index to be this low, it is likely quite porous. In the UV-visible
spectral region, significant changes in PABBI absorption were observed. An absorption
band at 340 nm was significantly reduced and a new absorption at 380 nm appears to
form.
72
The index of refraction (optical density) of aluminized FEP Teflon®from the

exterior thermal blanket of the Hubble Telescope was measured to back physical
density measurements acquired at NASA Lewis research center. The optical density
measurements were in good agreement with physical density measurements. The
density of the FEP only significantly changed when the material was heated to around
200°C. These measurements suggest that the significant increase in density of the FEP
returned from the second servicing mission was due to the sample being exposed to
elevated temperatures (approximately 200°C). The FEP from the second servicing
mission was heated to approximately 200°C because the thermal blanket cracked and
curled over, exposing the aluminum backside to direct sunlight [14, 15, 16].
References
1. "US Standard Atmosph ere, 1976," NASA TM-X-74335, (1976).

2. 1. A Dever, "Low Earth Orbital Atomic Oxygen and Ultraviolet Radiation Effects on Polymers," NASA Tech.
Mem o. 103711, (1991).
3. B. A Banks, K. K. De Groh, S. K. Rutledge, and Frank 1. DiFilippo , "Prediction of In-Space Durability of
Protected Polymers Based on Ground Laboratory Thermal Energy Atomic Oxygen," NASA Tech. Memo .
107209, (1996) .
4. R.MA Azzam, and N.M. Bashara, Ellip sometry and Polarized Light, North Holland Puhl ishing, New York,
(1977).
5. 1. A Woollam, B. Johs, W. A McGah an, P. G. Snyder, 1. Hale, and H. W. Y ao,Mat. Res. Soc. Sy mp. Proc.
324, IS (1994).
6. J. A Woollam , and P. G. Snyder, in Encyclopedia ofMaterial Characterization, Butterworth-Heinemann,
Publishers, Boston, 40 I (1992).
7. J. Jellison, Proceedings ofthe First International Conference on Spectroscopic Ellipsometry, Thin Solid
Films, 234, 416 (1993).
8. F. Wooten, Optical Properties ofSolids, Academic Press, New York , (1972).
9. S.F. Dyke, A1. Floyd, M. Sainbury and R.S. Theobald, Organi c Spectroscopy , Longman Group Limited,
London, 13 (1971).
10. S. K. Rutledge , B. A Banks, et al., "An Evaluation of Candidate Oxidation Resistant Materi als for Space
Applications in LEO," NASA Tech. Mem o. 100122 (1986).
11. D. E. Aspnes, Phy s. Rev. B, 33, 677 (19 85) .
12. B. A Banks, K. K. de Groh, S. K. Rutledge, and F. 1. DiFilippo, "Prediction of In-Space Durability of
Protected Polymers Based on Ground Laboratory Thermal Energy Atomic Oxygen," NASA Tech. Memo.
J07209 (1996) .
13. B. A Banks, 1. A Dever, L. Gebauer, and C. M. Hill. " Atomic oxygen interactions with FEP Teflon and
silicones on LDEF, " Proceeding s ofthe LDEF First Post-Retrieval Symposium, Kissimmee, FL, June 2-8
1991, NASA CP 3134, Part 2, pg. 801-815.
14. T . M. Zuby, K. K. de Groh, and D. C. Smith, "Degradation ofFEP Thermal Control Materials Returned
from the Hubble Space Telescope ," NASA Tech . Memo. 104627, (1995).
15. K. K. de Groh and D. C. Smith, " Investigation of Teflon FEP Embrittlement on Spacecraft in Low Earth
Orbit," Proc. 7th Int. Symp. on Mat . in Sp ace Env ., Toulouse, FR, 255 (1997).
16. B. A Banks , T. Stueber, S. K. Rutledge, D. Jaworske , and W. Peters, "Th ermal Cycling-caused Degradation
of Hubble Space Telescope Aluminized FEP Thermal Insulation," AIM #98-0896, ( 1998)
EFFECT OF ATOMIC OXYGEN EXPOSURES ON THE TRIBOLOGICAL
PROPERTIES OF MOLYBDENUM DISULFIDE LUBRICANTS
M. TAGAWA, J. IKEDA, H. KINOSHITA, M. UMENO

Department ofMaterial and Life Science. Osaka University
2-1 Yamada-oka, Suita, Osaka 565-0871 Japan
N.OHMAE
Department ofMechanical Engineering. Kobe University
I-I Rokko-dai, Nada, Kobe. 657-8501 Japan
Abstract
In this paper, the effect of atomic oxygen exposures on the tribological properties of
molybdenum disulfide (MoSz) is studied with a laser-induced breakdown type
atomic oxygen facility. McS, single crystals as well as sputtered films are used as
specimens to clarify both fundamental reaction and change in tribological properties
ofMoS z.
X-ray photoelectron spectroscopy and residual gas analysis show that SO
and Mo0 3 are formed at the MoSz (000 I) surfaces. A significant loss of S is found to
be due to volatilization of SO. The atomic oxygen exposure on the MoSz induces
high friction (six times higher than normal value) at the beginning of any sliding
action on the sputtered films, but no significant effect on a single crystal. It is
demonstrated that high friction with sputtered films is being retained when sufficient
atomic oxygen is provided while sliding.
1. Introduction
Energetic atomic oxygen (AO) reaction with materials used in the exterior of a
spacecraft is one of the unsolved problems in the space environmental effect
community. Much attention has been paid to polymer degradation due to AO attack,
however little has been made clear in the last decade. Lubricants as well as
polymers are at hazard from AO. A lubrication malfunction in a space system
influences the potential for serious damage to the whole system. In many cases,
there exists no back-up system for such tribological failure. Therefore, it should be
stressed that the influence of AO exposure on the lubricants used on the exterior of
the space system must be evaluated before the mission.
Molybdenum disulfide (MoSz) is one of the widely-used lubricants in space
systems. Its low friction coefficient (typically 0.02-0.05), and durability in vacuum,
are suitable for space application. Thin MoSz lubricant is usually prepared by sput-
73
Protection ofSpace Materialsfrom the Space Environment. 73-84.
© 200! KluwerAcademicPublishers.
74
tering or spray methods. The films formed by these methods show a wide variety of
tribological properties depending on their deposition/spray conditions. Therefore,
AO durability of the lubricant film may be evaluated on the ground by specimens
manufactured in the same lot, or at least, in the same deposition condition of the
flight model. On the other hand, from a scientific point of view, MoS 2 has a large
anisotropic nature in crystalstructure; i.e. MoS2 has a layered structure. It is natural
to consider that AO reactivity of basal and prism planes could be different. This
crystallographic anisotropy may reflect the macroscopic AO reactivity of the MoS2
lubricant prepared in various forms (the degree of crystallization of the film depends
strongly on the deposition conditions). Also, spray-bonded film includes polymeric
binders as well as MoS2 pigment, and thus the effect of AO exposure on the binder
and to the MoS2 pigment may be different and must be evaluated separately. From
the viewpoints mentioned above, we need to know the AO reactivity with a well-
characterized MoS 2 surface in order to understand AO-induced degradation of the
MoS2 lubricant.
In this study, AO reaction with the MoS2 single crystal and sputtered film
are studied. Auger electron spectroscopy (AES), X-ray photoelectron spectroscopy
(XPS) and residual gas analysis (RGA) data are interpreted to understand the AO
reaction with a MoS2 single crystal. Changes in the tribological properties of MoS2
single crystal and sputtered film due to AO exposure are also examined.
2. Experiment
The specimens used in this study are single crystal and sputtered MoS2 . The single
crystal MoS2 specimens are naturally-produced crystals, whereas the sputtered films
are prepared by the radio frequency (RF) sputtering technique at the National
Aerospace Laboratory (NAL), Japan. Sputtering conditions of the RF sputtered
specimens are listed in Table 1[1]. In the case of single crystal specimen, MoS 2
(000 I) surfaces are always prepared by a cleavage before testing.
The AO source used in this study is a replica of that at Montana State
University [2]. This type of source was originally invented by Physical Sciences
Incorporation (PSI), and uses laser-induced breakdown of the oxygen gas [3]. The
AO source is attached to the AO testing facility developed in our laboratory (Figure
I) [4]. The AO beam is always monitored by the time-of-flight (TOF) measurement
system consisting of a quadrupole mass spectrometer (QMS) and a multi-channel
scalar. Typical TOF spectra of the beam are shown in Figure 2. In the m/q=16
spectrum, cracking of the molecular oxygen (MO) in the ionizer is subtracted. We
see hyperthermal AO as well as thermal and hyperthermal MO components.
Table 1. RF sputtering conditi on of the MoS2 film
Sputtering gas Argon

Pressur e 6.5 Pa
Sputtering power 220W
Substrate Temperature 600C
Thickness I fUll
Substr ates SUS440C
75
Figure I
The AO-testing facility with the PSI type AO source. The PSI source (right) is attached to the original
YAG laser type AO source. The UHV friction tester is located at the top of the AO source chamber. The
reaction chamber with the RGA for gas analysis (centre), TOF chamber with QMS (front left) and analysis
chamber with AES (back left) are evident
Mean translational energies of these components are 4.7, 0.1 and 5.8 eV,
respectively. Relative intensities of these components are estimated from the peak
areas. The relative ionization cross section of AO and MO is assumed to be 0.6 in the
calculation [5]. From this calculation, we find that 45 per cent of the beam is AO, 24
per cent thermal MO, and 31 per cent hyperthermal MO. The flux of the beam is
calculated from the mass gain of a silver film deposited on a temperature-controlled
quartz crystal microbalance (QCM) [6].
m/q = 16 mlq= 32
o o
o 1000 2000 3000 4000 5000 o 1000 2000 3000 4000 5000
Flight Time (J1S) Flight Time (us)
Figure 2
TOF spectra ofthe beam; mlq = 16 (left) and mlq =32 (right). The hyperthennal AO component is
obvious at 350 IJS in the mlq=16 spectrum, the hyperthennal and thermal MO components are at 400 IJS
and at 2500 IJS in the mlq=32 spectrum
76
Table 2. Sliding conditions for the tribological testing
Pin Ti-6Al-4V (r=2mm)

Disk Single crystal or sputtered MoS z
Load 2N
Sampling speed 20 Hz
Track length 8 mm
Sliding speed 2.5 mmls
Tribological testing of the specimen is carried out in the AO source chamber. The
UHV-friction tester used in this study is especially designed for this facility [7]. This
UHV-friction tester is based upon the conventional pin-on-disk layout. However,
movement of the pin and detection of the forces are both performed at the pin ann in
order to make the transfer of the disk to the AES chamber easier. The testing
conditions are summarized in Table 2. The specimen can be transferred to the AES
to measure the atomic concentrations at the surface . XPS measurements are
performed by a stand-alone XPS system, so that the AO-exposed surface is exposed
to ambient air before measurement.
3. Results and Discussion
3.1. XPS ANALYSIS
Figure 3 indicates M03d XPS spectra of the MoS 2 (0001) surfaces before and after
AO exposure. The AO fluence is 6.7 x 10\7 atoms/em' (10,000 AO shots) . The XPS
spectra obtained from the control surface are also indicated in Figure 3. In the M03d
spectrum of the control surface, we see two XPS peaks from MoS 2 at 229.3 and
232.2 eV. These two peaks are due to spin-orbital splitting of M03d orbit ofMo (IV)
atoms in the MoS 2 structure . An additional XPS peak at 235.3 eV is obvious in the
XPS spectrum of the AO-exposed MoS 2. Also, the peak at 232.2 eV becomes higher
than that at 229.3 eY.
These spectral changes are explained by the oxidation of Mo. Namely , Mo
(IV) is converted to Mo (VI), which shows +3.5 eV of chemical shift, by means of
the formation of Mo0 3. From the peak deconvolution of these spectra, we find that
17
52 per cent ofMo is oxidized to Mo03 by the AO exposure at 6.7 x 10 atoms/em'.
The amount of Mo (VI) increases with AO fluence and it reaches 64 per cent at 1.3 x
1018 atoms/em" (20,000 AO shots) . Surface composition of the AO-exposed surfaces
examined by the XPS is summarized in Table 3.
From Table 3, the loss of S (60 to 21 per cent) and increase of 0 (4 to 52
per cent) is remarkable. The amount of Mo does not change significantly (36 to 27
per cent), however, the oxidation ofMo (Mo (IV) 100 to 36 per cent, and Mo (VI) 0
to 64 per cent) is notable . The atomic concentration of the AO-exposed MoS 2
surfaces is also examined by AES.
77
I Control AO.Exposed
\Iu
240 236 232 226 224 220 240 236 232 226 224 220
Binding Energy (ev) Binding Energy leV)
Figure 3
Mo3d XPS spectra ofthe MoS2 (0001) surfaces before (left) and after (right) AO exposure with a fluence
of6.7 x 1017 atoms/em'. The additional peak at 235 eV becomes obvious after AO exposure.
Table 3. Surface composition of the MoS2 surfaces analyzed by XPS
Specimens Mo Mo(IV) ; Mo(VI) S 0

Control 36 100 : 0 60 4
1,000 AO shots l ) 24 87 : \3 37 39
10,000 AO shots 2) 28 48 : 52 18 54
20.000 AO shots 3) 27 36 : 64 2\ 52
16 17 18
AO fluences : \) 6.7 x 10 ,2) 6.7 X 10 ,3) 1.3 X 10 atoms/em"
The results of AES analysis do not agree with the XPS data shown in Table 3, i.e.,
only 3 per cent of oxygen is detected in the 10,000 AO shots sample. Similar results
are observed not only in the other MoS 2 specimens, but also in the AOIHOPG
system [8). Since the high energy primary electron beam of AES may cause electron
stimulated desorption (ESD) of the weakly-bonded gaseous atoms/molecules on the
surface, we believe XPS data is trustworthy even though it is affected by exposure to
ambient air [9].
The XPS measurements are carried out with tile sputtered MoS 2 surfaces.
Even at similar AO fluence (2 x 1018 atoms/cur), the sputtered MoS 2 surfaces show
complete oxidation of Mo (Mo (IV) : Mo (VI) = 0 : 100) and loss of S (13 per cent).
This would be explained by the characteristics of the sputtered MoS 2 : (1) the high
diffusion rate of AO into the MoS2 film, and (2) dangling bonds of Mo existing in
tile film.
3.2. RGA ANALYSIS
Volatile species formed on the MoS 2 (0001) surfaces during AO exposure are
detected by the RGA installed in the reaction chamber (see Figure 1). In this system,
tile TOF spectrum contains the reactive species bounced at the chamber wall as well
as those directly from tile specimen surface. Therefore, it is not possible to measure
the translational energy of the reactive products, but the species can be identified .
For this measurement, the AO beam is collinlated by the 4mm 2 orifice located
between the AO source and the reaction chamber to maintain the vacuum of the main
chamber low enough to operate the QMS. The QMS output of the designated mass
is detected with an MCS to synchronize the pulsed beam operation. The QMS signal
at m/q=48 and 64 thus measured has a very broad peak at around 30 ms after the
78
laser firing. Peak intensity of this signal at 3 1.5 ms is recorded every 30 shots and
plotted in Figure 4.
- - - AO SHOT vs SO
··0·· AO SHOT vs S02
_ 300 · . . .
.. O' ~ • • • • O'O' O' O' :- O' O' • O'O' ~ O' _ : ••
··
• O'O' O'O' • • • ~
.
O'O' • • • O' O' • • •: _ O'
.
~ - • • • - - O'.
..
"
~
C
::J
-8 200
>-
..-
'iii
·
····
.
... .. ......... . . ....... .. ... . ..........
....
... ...
..
..
..
c
~
c · ········:· ······· ·· i·, ··· ······:·········
;.
.........·: . ···{·· . .
100 ·· .. · .
·;······ -,
. .,
... .,
o • • , ,
··· ... : m/q=p4
o~~~~~~~~~
o 200 400 600 800 1000 1200 1400 1600
AD Fluence (shots)
Figure 4
RGA signal of mlq = 48 and mlq = 64 during AO exposure. It is obvious that SO (mlq = 48) is
the dominant reactive specie of the AO/MoS2 system
As is shown in Figure 4, SO (m1q=48 ) is the dominant reactive species.

Also nonlinear growth of the signal is observed. The breakpoint at 1,000 shots
clearly indicates that the rate-limiting process of SO formation is changed at around
1,000 shots.
3.3. TRIBOLOGICAL PROPERTY
Tribological properties of the MoS2 single crystal is examined with the UHV friction
tester attached in the AO source chamber (see Figure 1). AO exposure is temporally
stopped when the friction force is measured in order to reduce electrical noise from
the laser-induced plasma. Figure 5 indicates a typical load and friction force traces of
the AO-exposed MoS2 single crystal. The MoS2 surface is exposed to an AO beam of
2 x 1017 atoms/em ' prior to sliding. As is often reported in the literature [10, 11),
high friction at the initial stage of sliding is observed. However, such high friction at
the initial stage in sliding is not characteristic of the AO-exposcd MoS2 . A similar
effect is obvious even in the non-AO-exposed (virgin ) MoS2 surfaces as is shown in
Figure 6. The same effect is also observed in the sputtered MoS2 specimens . It is
therefore concluded that the initial high friction observed in the AO-exposed MoS2
79
single crystal is due mainly to the resistance to movement of the pin due to the wear
track formation ,
2.0
'"'
z 1.0
.......
B
...
0
u.
0.5
0.0
-0.5
0 50 100 150 200
Time (sec)
Figure 5
17
Load and friction force traces measured on the AO-exposed MoS 2 single crystal (AO fluence of2 x 10
atoms /ern"). The friction force becomes large at the beginning of the sliding action
2.0
1.5
z
'-'
~ 0.5
o
u.
0.0
-0.5 l-..-<........- ' - - ' - - . . L -...........---'----'-.....L..."""'---..........."-"'--L--'-"""""--"-.o-.I

o 50 100 150 200
Time (sec)
Figure 6
Load and friction force traces measured on the virgin MoS1 single crystal. The friction force
becomes large at the beginning of the sliding action
80
To eliminate the effect of wear track formation, AD exposure was carried

out on the MoSz surface where wear track already exists. Specifically, the first
sliding test was performed before AD exposure, then the wear track was exposed to
AD beam and the second sliding test was done after AD exposure. During a series of
tests, the load of the pin was not released so that the misalignment of the pin/wear
track in the sliding tests was negligible. The specimen surface was kept in a vacuum
of at least 10-6 Torr during the test. The result of this test is indicated in Figure 7. A
very small increase of friction force is detected, but only in the beginning of the
sliding action in the second test. From the result obtained in this test we concluded
that the effect of AD exposure on the tribological properties of McS, (000 1) surface
is negligible.
The same tests are carried out with the MoSz sputtered films. The result is
shown in Figure 8. It is observed that the increase of friction force in the sputtered
film is much higher than that in the single crystal: i.e., the friction coefficient of the
initial stage of friction reaches twice as high as the steady-state sliding condition .
The friction force measurement was in the following sequence with these friction
force traces:
( I) Since hysteresis of the friction force by the direction of movement is obvious in
this system, the data during sliding in one direction is used. This is because the arm,
to which the pin is attached, is bent by the applied nonnalload.
(2) The average output after sliding (when pin is stopping) is set as zero-point.
(3) Difference between the output when the pin is moving in steady-state friction and
the zero-point set in (2) is measured as a steady-state friction force (Fsteady-state) .
(4) Difference between the output at the first friction and the zero-point set in (2) is
measured as an initial friction force (Finitiai).
2.5 ,--,.--r--..-....,......~r--I---.---.--r-...,.-.,.......,........,.......--r-.......-....--,....-,
2.0
.- 1.5
-z~
o
u,
0.5
Friction
0,0
-0.5 L...-L-........................L...-...........--'---'-....L-""""----'--...........--L--'--'---'--.........,
o 50 100 150 200

Time (sec)
Figure 7
Load and friction force traces measured on the AO-exposed MoS2 single crystal (AO tlu ence of
2 x 10" atoms/em'). Note that the wear track has alrea dy been formed before the test. High friction at
the beginning of the sliding action is not observed
81
The AO fluence dependence of the FinitiallFsteady-stateratio is shown in Figure

9. Figure 9 clearly shows that the friction force increases with the AO exposure at
16
the fluences higher than 100 AO shots (2 x 10 atoms/em'). The Finitial reaches 6
18
times larger than the Fsteady-state at the AO fluence of 10,000 AO shots (2 x 10
atoms/em'). In contrast, thermal MO beam exposure does not cause any influence on
the friction force. XPS spectrum of the AO-exposed MoS z sputtered film (AO
fluence of 2 x 1018 atoms/em') shows complete oxidation of Mo and loss of S (see
Section 3.1).
It is, therefore , concluded that severe oxidation of the lubricant surface and
loss of S results in initial high friction . The XPS data show the amount of S and Mo
(IV) in the wear track increases with lubrication, and this will recover its tribological
property.
Figure 8 also indicates that it takes several slides to recover the friction
force from Finitial to Fsteady_state.This fact suggests that the tribological system, which is
operated with intervals in LEO, may be affected by high friction in all cases. Figure
10 demonstrates the change in friction force in such situation . Four AO exposures (2
x 1017 atoms/em') have been carried out at 0, 50, 100, 160 sec in the abscissa . In the
sliding action after AO exposures, high frictions are obvious, but are followed by the
recovery to a steady state level. Therefore , it is necessary to consider that the
tribological property of a sputtered MoSz lubricant used in LEO environment is
always affected by AO exposure .
2.5
2.0
-
Z
~ 1.5
o
LL 0.0
-0.1
-0.2
o 50 100 150 200

Time (sec)
Figure 8
Load and friction force traces measured on the AO-exposed MoS! sputtered film (AO fluence of
2 x IOt 7 atoms/em'), Increaseof the friction forceat the beginning of the sliding action (Fin;';,I) is clearly
observed
82
6 • Hyperthemal AO beam
o Thermal 02 beam
2 ....;..:..:..;. :.:.;.;.: : ;..;.:.;.;_::: ;" 0;'.;";";"::: :,.;__;" _;.;.;.; .
s 5
III •
o
•
•
• • •• •• •
• • • • • • •
•
•
•
•
• • • • • ••
• • • • • • •
•
•
•
•
• • •
• • •
. 0.
• 0'
•
••
"
•
•• • •
• • • •
I • • • . • • • •• • • • • • • • •• • • • • • ' 0' •• • • • ••
··· ·f··]··[·r jnr····]···r·[·rrTf·····r· [·rn r "r'f'[ r~T ' "

>. • • • , • • • • , • •• " • • • • , • • • • " 0 •• • •• ••
'C · • • • ••• • • • •• • • • • • • • • • • • '0 ' • •• • •• •
(lJ
4 • • • • • • • • • . . , • • • • • • . , • • • • 0 . • • • • • • .
*
IJ..
-. 3
~
...-:..:..~~. : ~~~~ .....:..-:-.:.;~.:- ::: .... ~...: . ~. :-;i ····..~. -; .~ ~-7:-:··- ·
o • • • • • • • • ••• • • • •• • • • • • • • 0' • " • •• •
~ ~ ~ ~ ~ ~~~~ ~~:~~ ~~ ~~ ~~ t!:~~ ~: :~~ ~ :

'c
u:..- 2
_H j : -: :tt -- j : :-: : l~ : : ::i-- inn :-- .:
1 · ···f··, j·-1 -·i· i1-H· ·· ··j··. ·j,· ·j·fH~;; ·· ·· ·.!· ··f·~·H p. . .."1."}')'·H·+···
· . ...
o
. . ..
• • •• • • • •
. .
.
• •• •
,
•
. . 0.
• • • •
. .
• • • • •
.
•
.
0
1 10 100 1000 10000
AO Fluence (shots)
Figure 9
Increase of initial friction force compared with a steady state friction. In this graph. AO flux is
14
l x 10 AO/shot. Th e data where one AO shot means control surface. The empty circle indicates the result
of thennal MO beam
2
...-.
Z
.......
8 Load
~
0
IJ.. 0 Friction
-0.25 L-.I...-'--'..............L....I..--'--"--'-....L-L-.l...-'--'---I--'--'--'--.......L......I...-.............~
o 100 200 300 400 500

Time (sec) Figure 10
Friction force of a MoS 2 sputtered film surface during AO exposures. AO exposures are carried out four
times in this test at 0, 50. 100, 160 sec in the abscissa
83
4. Conclusions
We have studied the surface characteristics and the tribological properties of the
energetic AO beam exposed MoS2 (0001) surfaces. A combination of the PSI type
high flux AO source and UHV-friction tester attached to the source chamber is a
powerful tool for evaluating the effect of AO exposures to the lubrication materials.
The tribological testing performed in this study clearly shows that the effect of AO
exposure on the MoS 2 single crystal is almost negligible, whereas that on sputtered
film is notable. XPS spectra clearly show that AO reactivity with MoS 2 has a crystal
plane dependence; MoS 2 (0001) surface has high anti-oxidation capability against
AO attacks compared with sputtered films. It is, however, demonstrated that the wear
track on the MoS 2 sputtered film exposed to the AO beam with fluences higher than
2 x 1016 atoms/em' shows increase of the initial friction force, and it reaches 6 times
higher than the normal value. The AO flux at 250 km in LEO reaches this fluence
within a few minutes. This result suggests that the lubrication system, which is
operated at intervals in the LEO environment, may show high friction in all
operations due to AO-induced oxidation of the materials.
5. Acknowledgments
A part of this work was supported by the Grant-in-Aid of Scientific Research and
Encouragement of Young Scientists from the Ministry of Education, Science, Sports
and Culture Japan; by the Special Coordination Funds for Promoting Science and
Technology from the Science and Technology Agency, Japan; and by the Space
Utilization Research Promotion from the Japan Space Forum . The authors are
grateful to M. Suzuki and K. Matsumoto for providing the sputtered MoS 2
specimens .
M. T. expresses his gratitude to T. K. Minton for many helpful comments
about the operation of the source, and to A. Paillous for safety information about the
usage of the cryo pump in the source.
84
6.0 References
l. Matsumoto, K., Suzuki, M., Imagawa, K., Okada, Y ., and Tagashira , M.: Evaluation of tribologi cal
characteristics of sputtered MoSz films exposed to LEO environment, Proceedings of the 2i st
International Sy mposium on Space Technology and Science, (1998) paper #98-b-30.
2. Minton, T.K., and Moore, T.A : Molecular beam scattering from l3C-enriched Kapton and correlation
with the EOlM-3 carousel experiment, Proceedings of 3rd W EF Post Retrieval Meeti ng,
Williamsburg, AL, November 1994, 11l3-1117.
3. Caledonia, G.E., Krech, R.H., and Green, B.D.: High flux source of energetic oxygen atoms for
material degradation studies, AIM}., 25 (1987 ) 59-63; U.S. Patent No.48945I I (16, January, 1990).
4. Kinoshita, H., Ikeda, J., Tagawa, M., Umeno, M., and Ohrnae, N.: A fast atomic oxygen beam facility
with in-situ testin g/analysis capabilities, Review of Scientific instruments, (1998) in press.
5. Paillous, A., Cazaubon, 8. , Barassin, A , and Thomas, R.: Utilization of mass spectrometry for
characterization of the interaction of hyperthermal O-atom beams with materials, Proceedings of 7th
international Symposium on Materials in a Space Environment, Toulouse, France, June 1997, 185-193.
6. Tagawa, M., Matsushita. M.. Umeno, M., and Ohmae, N.: In-situ experiments on the influence of
atomic oxygen on space tribology, Proceedings ofthe International Tribology Confe rence, Yokohama
Japan, October 1995, 1865-1869.
7. Matijasevic, V., Garwin, E.L., Hammond, R.H.: Atomic oxygen detection by silver-coated quartz
deposition monitor, Review of Scientific Instruments, 61 (1990) 1747-1749.
8. Kinoshita. H., Ikeda, J., Tagawa, M.• Umeno, M.. and Ohmae N.: Hyperthermal Atomic Oxygen
Interaction with Highly Oriented Pyrolytic Graphite Studied by Scanning Tunneling Microscopy,
Proceedings of the 7th International Symposium on Materials in a Space Environment, Toulouse
France, June 1997, ESA-SP399, 211-216.
9. Effect of ambient air exposure has been evaluated by AES. No significant change in oxygen
concentration is observed even after air exposures.
10. Arita, M., Yasuda, Y ., and Ohmae, N.: Investigation of lubricating characteristics of M oSz films in
space environment, Proceedings of 4th European Symposium on Space Mechanics and Tribology,
Cannoe France, November 1992, ESA-SP299, 67-7 l.
11. Ohmae, N., Tagawa, M.. Akiyama, M., Gyougi, T., Kishi, K., Arita, M.: Atomic oxygen effects on
tribology of solid lubricants. Proceedings of the 6th interna tional Symposium on Materials in a Space
Environment, Noordwijk The Netherlands. September 1994. ESA-SP368, 327-332.
INFLUENCE OF SHIELDING UPON THE STRUCTURE AND
PROPERTIES OF POLYMER COMPOSITE MATERIALS, EXPOSED ON
THE SURFACE OF THE MIR ORBITAL COMPLEX
O.V. STARTSEV, V.V. ISUPOV, E.F. NIKlSHIN

A ltai State University
66, Dimitrov St., Barnau/, 656099, Russia
Abstract
An important task in the field of materials for space applications is the study of the
influence of outer space factors on polymer composite materials (PCM). During the
last two decades, the stability of these materials in outer space conditions (by
exposing them on the outside of the Salyut and Mir orbital stations) under a long
term program at the M. V. Khrunitchev State Space Scientific Production Center
("Salyut "Design Bureau) is studied after returning them to the Earth .
Over years of flight experiments, the exposure of a wide class of PCM for
102, 304, 382, 456, 686, 1024 and 150I days at "Salyut" and "Mir" orbital stations
was carried out. Objects of the research were KMU-2, KMU-31, KMU-4l carbon
fiber reinforced plastics, VPS-7V glass fiber reinforced plastic and also their
compounds with different variants of surface protection from the direct influence of
natural factors: aluminium foil, aluminium alloy, glass netting, lacquer and paint
coating.
With the help of various physical methods the influence of coatings and
screens upon the complex of mechanical, diffusion, heat-physical parameters of
monoplastics and hybrid systems was investigated. From the electronic microscopy
and diffusion analysis data the destruction and microerosion effects of materials'
surfaces exposured to the space without protection were estimated. It has been
shown that a consequence of damage of PCM's surfaces is the rough rise in diffusion
coefficient, decrease of the microhardness and dynamic shear modulus parameters
of the unprotected samples. The mentioned effects are significantly weakened under
shielding of open plastics' surfaces by aluminium foil or by a layer of another
plastic.
Special attention was paid to the research of KMU-31 carbon fiber
reinforced plastic with a thin layer of glass fiber cloth used as a covering for panels
for micrometeoroid protection of the Mir orbital complex, exposed for 1024 days
under the layer of same material. It has been shown that the protected plastic is
completely shielded from destruction. At that, changes in physical properties are
determined by postcuring processes and improvement of structural homogeneity of
85
Protection ofSpace Materials from the Space Environment, 85--93.
© 2001 Kluwer Academic Publishers. .
86
epoxy binders. The consequence is increase in dynamic shear modulus at raised

temperatures and also in heat-resistance of the composites after 300-500 days of
exposure to the space.
Basis on the obtained results, the substantiation of service stability for
composite materials used for micrometeoroid protection of the "Mir" orbital
complex under vacuum skin was given. Models of physical aging processes of PCM
with different variants of protection were worked out and a forecast in the change of
basic operation parameters for 15 years in the space was given.
1.0 Introduction
An actual task in the field of materials for spacecraft equipment is the study of the
influence of outer space factors on polymer composite materials (PCM). For two
decades, the stability of these materials under conditions found in space have been
studied in Russia through direct natural exposure of experimental samples on the
external surfaces of both the Salyut and Mir orbital stations. Following their return
to Earth, detailed research of the properties of these samples has been conducted at
the M.V. Khrunitchev State Space Scientific Production Center (the Salyut Design
Bureau).
The results of changes in the properties of these materials in ground-based
tests are used to evaluate and forecast a set of operational parameters which allow
conclusions about the serviceability of structures containing PCM. This paper deals
with the evaluation of the effectiveness of protective screens and shields applied in
PCM, used in the construction of micrometeoroid protection on the Mir orbital
station, using the criteria of dynamic mechanical analysis (DMA), sorption analysis,
microhardness and other indicators.
2.0 Experimental
Panels of two types are used in the construction of the rnicrometeoroid protection on
the Mir space station: VPS-7v glass fibre-reinforced plastic and a three-layer
honeycomb system with two coverings made of KMU-31 carbon fibre-reinforced
plastic CSlocarbon') with the thickness of 0.46 mm and glass honeycomb plastic
GHP-O.l with a thickness of 10 mm on VK-36 adhesive. In this work, we used
separately-exposed samples of KMU-31, KMU-41 carbon fibre-reinforced plastics
(66 x 50x0.46 mm and 58 x 8 x 1.2 mm) and VPS-7v glass fibre-reinforced plastic
(58 x 8 x 1.2 mm). Samples of KMU-3. coated on one side with a layer of glass
cloth, were exposed in the holders in such a way that one sample (No.1) was
exposed with the layer facing the space environment while the other one (No.2)
was beneath it with the coated side up. Thus, it was obscured by Sample 1. Two
other samples of KMU-31 PCM (No.3 and No.4) were oriented with the layer of
glass fibre-reinforced plastic to the space environment. In addition, samples of
87
adhesive compounds with different types of surface screening (A/-foil, alloy AMG-
6m, lacquer and paint coating LKP 40-1-114-87) were examined under direct
exposure to the space environment. The exposure ofPCM experimental samples was
carried out for 102, 304, 382,456,686, 1024 and 1501 days.
To obtain diffusion parameters - diffusion coefficient D and moisture count w -
the samples were dried in thermostat at 343±3 K during 16-24 days. The changes of
mass were controlled using the analytical balance with resolution of 10-7 kg.
Microhardness H measurements were obtained on PMT-3M by Vicker's method [4].
The method of dynamic mechanical analysis was employed to obtain data on stress-
strain properties of PCM specimens and adhesive compounds in their initial state
and after exposure - temperature dependencies of dynamic shear modulus G',
mechanical loss tangent tg (5 and sound velocity Ct in the temperature interval
ranged between 77-573 K. Measurements of the above parameters were performed
with the experimental error of not more than 3-5 per cent.
3.0 Results and Discussion
The effect of exposure to space on PCM properties of the materials under

consideration, mentioned above in [1-4] is determined by the epoxy binder post-
curing process and by micro-erosion on the exposed surfaces of the samples . The
processes of etching and micro-erosion in polymer composites are caused by atomic
oxygen, microparticles and space junk.
Figure I shows the dependence of the moisture diffusion coefficient D of two
couples of exposed KMU-3/ samples as compared to the control sample.
Comparison of values D proves that the glass cloth layer completely screens the
carbon fibre-reinforced plastic from the destructive influence of the space
environment, while on the unscreened surface of Sample I microscopic cracks can
be observed, stripping of the surface layer and loosening of the structure of PCM
(Fig. 2). Thus , the acknowledged fact of a drastic increase of moisture diffusion
coefficient of samples with unscreened carbon fibre-reinforced plastic layer, is
corroborated [3, 4].
There were no essential changes discovered on the surface of the glass cloth
layer of this series of materials, except for some changes in color and loss of shine .
The same method was employed to explore the whole series of materials exposed
on the Mir space station during 1024 days (Table I). Moisture count of the samples
constitutes 0.2-0.6 per cent. The value of moisture count of the materials with A/-
foil coating is approximately 0.2 per cent higher, compared to that of the materials
with lacquer and paint coating.
When the PCM surface is screened with an A/-alloy AMG-6m layer, the
moisture count of the sample is sufficiently reduced (by 0.3 per cent). Moreover, this
effect is more pronounced in the materials when their external exposed surface was
screened. The analogy is observed in hybrid composites KMU-4/IVPS-7v and VPS-
7v/KMU-4/, the latter having the most essential maximum moisture count
88
difference which amounts to 0.1 per cent. In hybrid KMU-411KMU-41 the moisture
count value was the same as compared to that of monoplastic KMU-41.
2.5
C 2.0
.su
....
<..::
<I
0 1.5
U
..
0
' ;;
1.0
~ 0.5
Q
0.0
I:] absolute values D ·IO,IO, 1cm2/s

o relative values D/DiniL
Figure 1
Diffusion coefficients D of KMU-31samples exposed in two layers as compared to the initial sample
Table 1. Diffusion coefficient ofPCM with different variants of screening after exposure to space
environment
W«, %, W«,%. D . 10.1°, D . 10.1°,

Material DID....
init. expo cm2/s , init. cm2/s , expo
KMU-41 0.407tO.014 0.474 t O.004 0.4 1to.06 4.90t 1.52 12.0
KM U-4I1KMU -41 0.407 t O.014 0.545 t o.008 0.41 t o.06 2.85±0.31 7.0
KMU-4il AMG-6m 0.166tO.008 0.162± 0.002 0.71t O.12 6.38±0.91 8.9
AMG-6mIKMU-4i 0.166 ± 0.008 O.l 53± 0.005 0.71±0.12 2.14±0.39 3.0
KMU-4iNPS-7v 0.407±0.014 0.450 ±0 .004 0.41 to.06 6.86 t 1.01 16.7
VPS-7v/KMlJ-41 0.407± 0.0 14 0.383 ±0 .007 0.41±0.06 2.69tO.40 6.6
KMlJ-41 with LKP 0.407±0.014 0.413 ±0 .009 0.41 to.06 2.83±0.50 6.9
KMU -4i with Ai-foi l 0.407±0.014 0.59 1±0 .009 0.41 ±0.06 1.42±0.22 3.4
It is significant that out of 12 materials explored, KMU-31 carbon fibre-

reinforced plastic has the minimum increase of diffusion coefficient after long-term
exposure in LEO (by 2.3 times in the case of unscreened Sample 1). It is possible to
achieve minimum changes in the DlD init value of KMU-41 carbon fibre-reinforced
plastic only when its surface is screened with A I-alloy AMG-6m. Thus , during a
multi-year exploitation in space environment , KMU-31 is supposed to stay most
resistant to the micro-erosive influence of plasma and microparticles as compared to
the other materials.
89
. •
..-
Figure 2
Orientation ofKMU-3/ samples during exposure to outer space
..
The analysis of physical properties of PCM after several series of exposures to
the space environment demonstrates that thermal cycling is the main factor which
can cause changes of mechanical properties of materials. With the help of some
physical methods (DMA linear dilatometry, optical and scanning electron
microscopy, fracture analysis) it has been ascertained that during exposure to the
space environment, the epoxy binder post-curing process predominates [1-4].
Because of the fact that exposure to the space environment may introduce a
gradient of mechanical properties inside composite material [3, 4], it was decided to
estimate values of micro-hardness of the binder across the thickness of a sample.
The measuring of micro-hardness of all samples was done under the same regime
load (50 g) and loading time (30 sec). This is dictated by the fact that further
increase in load and exposure time does not lead to a change in microhardness H by
more than 1 per cent, which is within the experimental error. The samples of both
exposed and initial variants of PCM were prepared in individual layers. Their
surfaces were polished on a felt weel with the purpose of improving the visibility of
the resulting imprint. To evaluate the extent of micro-hardness spread, 10-20
measurements were made on each layer. As a result, the diagonal size of imprint,
micro-hardness H, dispersion CiH and a relative error EH were calculated. This
method allowed the achievement of good stability of the experiment with the
minimum experimental error of not more than 7 per cent.
Table 2 shows maximum effects of micro-hardness index changes in the volume
of the binder of PCM samples after 1024 days' exposure to space. The increase of
micro-hardness in composite materials based on binder 5-211B (KMU-3/) and
ENFB (KMU-4/) is observed after exposure to the space environment. This is,
obviously, a result of a long-term thermal cycling, which determines binder post-
curing processes during exposure. Polymeric composite material VPS-7v based on
binder EDT-lOP is characterized by a micro-hardness reduction (-7.7 per cent).
During layer-by-layer scanning of the microhardness index in carbon fibre-
reinforced plastics, it was found that post-curing processes predominate in the
extemallayers. Relative increase of micro-hardness in KMU-41 binder is higher if
the surface of carbon fibre-reinforced plastic is screened with AI-foil or a layer of
other material (VPS-7v, AI-alloy AMG-6m). This is caused by a considerable
90
temperature gradient during exposure to the space environment [3,4) and the
reduction of the influence of destructive factors on the binders in the volume of the
sample.
Table 2. Micro-hardness in the volume of a PCM binder with different
variants of screening after 1024 days exposureto the space environment
kgs kgs Lev el

Material H init., - - H exp., --
2 across EH , %
2
mm mm thickness,
mm
KMU-41 34 .9 36.0 0.54 3 .2
KMU-41 wi th LKP 34 .9 35 .3 0 .46 1.1
KMU-41 with Al- 32 .7 37 .3 0 .38 14.1
foil
KMU-41/ AMG- 34 .9 37 .0 0 .39 6.0
6m
AMG-6m/KMU- 34 .9 37 .7 0 .58 8 .0
41
KMU-41 in
34 .9 37 .5 0 .23 7 .4
hy brid
KMU-41/VPS-7lJ
KMU-41 in
34 .9 43 . 1 0 .20 2 3 .5
hybri d
VPS -7lJ/ KMU-41
KMU-3 1 28 .6 37 .6 31.4
VPS-7 lJ in h ybrid
19.6 18 .1 -7 .7
KMU-4 1/VPS-7 lJ
According to DMA data, the increase of temperature of the vitreous state of the
binder is the criterion for the post-curing process. This is revealed in a shift towards
temperatures of peak heights of the mechanical losses tangent tg 8 and the regions
of decrease of dynamic shear modulus G', measured in the temperature interval.
Here is the illustration of shielding screens influence on KMU-3/ carbon fibre-
reinforced plastic (time of exposure to the space environment - 1024 days), used in
micrometeoroid protection coverings of the Mir orbital station.
Table 3. The effect of screening KMU-3/ with a layer of the same material on
relative dynamic shear modulus (in %) during 1024 days' exposure to outer space
Number of 293 K 373 K 423 K

KMU-3/ sample
No I -28.5 -26 .9 -33.9
No 2 under No I 17 .4 19 .6 7 .7
No 3 - 17.3 -15. 1 - 10.2
No 4 under No 3 1.6 3.4 6.4
It was ascertained that the operational stability properties of polymeric composite

materials measured at room temperature do not decrease after 102-686 days of
91
exposure to outer space, and rise drastically under increased temperatures .

Maximum increase of relative dynamic shear modulus G'/G 'init. in non-shielded
VPS-7v glass fibre-reinforced plastic amounted to 1.8 times, while maximum
increase of G'/G'init. in non-shielded KMU-3! carbon fibre-reinforced plastic
amounted to 4.1 times.
In unscreened Sample 1, with the carbon fibre-reinforced plastic layer exposed to
outer space, the decrease of dynamic shear modulus G' by 27-34 per cent is
observed, while in Sample 2, the increase of G' is 20 per cent (Table 3). It is evident
that Sample 1 was more affected by destruction and microerosion on the surface of
the binder. The value of thickness of the stripped surface layer constitutes 6.4 per
cent (with thickness of 0.46 mm of the initial sample), which caused reduction of
mechanical strength in the whole temperature interval. Screening of the surface
with glass cloth (Sample 3) reduces degradation of mechanical properties, but also
obstructs the post-curing process in the binder (Sample 4).
e; 3
t5
<II
293K
;:I
~ 373K
~ 2
~ 423K
J!
(J)
u
·S
os
l'l 473K
0- 523K
0
0 250 500 750 1000 1250 1500
Time of Exposure, Days
Figure 3
Effect of the time of exposure on the value G ' of carbon plastic KMU-41 measured at various
temperatures (numbers next to the curves, K)
92
Table 4. Relativechange of dynamicshear modulus(in %) calculated for
the extremetemperatures of thermal cyclesduring exposureto outer space
Extreme 102 days 304 days 382 days 686 days

temperatures, K
190 400 190 400 190 400 190 400
KMU-3! 1.3 -7.4 2.1 -2 2 .4 236.8 -20. 5 231.4
KMU-4! -4 .3 41. 4 -20.3 97 .7 -22.4 105 .0 -26.4 lIlA
VPS-7v -3.0 - 14 .4 -6 .6 10.9 -4 .5 4.6
It is well known that peak temperatures the affect materials of the outer surface
of orbital space stations, when flying in the shade of the Earth and lit by the sun, are
190 K and 363-400 K, respectively. Table 4 shows relative change of the dynamic
shear modulus G'/G'init. calculated for the extreme temperatures of thermal cycles
with respect to the initial sample.
Decrease of dynamic shear modulus of KMU-3/ and KMU-4/ exposed to the
temperature of 190 K is 20-25 per cent as compared to the initial value. When
exposed to the temperature T = 400 K, increase of dynamic shear modulus G I in
KMU-3/ constitutes 230-240 per cent, while in KMU-4/ - 105-110 per cent. A
number of physical criteria and analytical estimates (for example , stress-strain
properties of PCM, forecast for 15 years of operation in work (3]) prove that main
transformations at molecular and supramolecular levels in 5-211B and ENFB
binders are more intensive within the period of 300-400 days of exposure to outer
space; are over within 500-700 days; and thereafter, dynamic shear modulus
practically does not change (Figure 3). VPS-7v material, exposed without shielding,
is characterized by smaller differences of relative strength indexes, while the
increase of its dynamic shear modulus under increased temperatures is distinctly
lower (up to II per cent).
4.0 Conclusion
Mechanisms of changes in measured macroscopic parameters of materials used in

structures designed for micrometeoroid protection during the time of exposure up to
1024-1501 days show the necessity for application of simple shielding screens , that
significantly reduce destruction and micro-erosion effects. On the application of
micrometeor protection panels on the Mir station under a screen-vacuum isolation ,
in our estimation, there is no reduction of mechanical strength in KMU-3/
coverings . As well, post-curing processes predominate over destruction, which
creates additional strength for micrometeoroid protection structures on the Mir
orbital station .
93
5.0 References
I. Startsev, O.V. and Nikishin, E.F.. Ageing of polymeric composite materials in outer space, Mechanics
ofComposite Materials 29, 457-467,1993.
2. Startsev, O.V. and Nikishin, E.F., Properties of adhesive compounds of polymeric composite materials
and thermoplastic polymers during 150 I days of outer space exposure. Proc. ofthe Sixth International
Symposium on Materi als in a Space Environm ent, September 19-23, 1994, ESTEC Noordwijk, The
Nederlands, 1994, pp. 223-233.
3. Startsev,O.V., Isupov. V.V. and Nikishin, E.F., The diagnostics of micrometeor protection for LEO
stations from the results of the in-flight experiments. Proc. ofthe Seventh International Sympo sium on
Mat erials in a Spac e Environment, June 16-20, 1997, Toulouse. France, pp. 367-373, 1997.
4. Startsev,O.V., Isupov, V.V. and Nikishin, E..F.. The gradient of mechanical characteristics across the
thickness of composite laminates after exposure in LEO environment, Polymer Composites, 19, 36-42,
1998.
CLEANLINESS SUPPORT OF VEHICLE FAIRINGS
FORSPACECRAFT
V SITALO, V TIKHIY
Yuzhnoye State Design Office ,
Dnepropetrovsk, Ukraine
Modem features of spacecraft specify special cleanliness requirements not only for the
spacecraft itself but also for the spacecraft fairing during its design, manufacturing,
transportation and launch . Figure 1 illustrates the fairing of a launch vehicle.
Environmental cleanliness of the spacecraft must be maintained in a Class
M6.5 (100000) clean environment per FED-SID-209; the quantity of particles
measuring from 0.5 micron and more should not exceed 3500/1iter, and measuring
from 5 microns and more, 25 particles per liter.
The residual quantity of particulates on internal surfaces forming the fairing
space should not exceed 10 mg/m". The residual quantity of molecular contamination
on internal surfaces forming the fairing space should not exceed 100 mg 1m2 .
Units and assembly units of nonmetallic materials should be outgassed in
vacuum (to pass thermal vacuum treatment) so that during storage and in launch
conditions there should not be more than 1 per cent outgassing, and condensed phase
of no more than 0.1 per cent of the total weight of nonmetallic materials .
To satisfy the above-mentioned cleanliness requirements, a new traditional
riveted construction from aluminum alloy was used. Such a choice guaranteed
minimum weight of the fairing and a short time-frame for project realization.
The principle of reducing to a minimum both molecular and particulate
contamination was chosen in selecting materials for the fairing . For example,
"dusting" soft thermal insulation of basalt mats was replaced by a hard fiberglass one
with aluminum foil on both sides. In addition , to prevent particulate emission its butts
were covered by a special film.
Figure 2 shows the internal surfaces of the fairing . A number of steps were
undertaken to reduce or eliminate completely the sources of contamination. Thus,
germatic, being a source of large particulates, was eliminated . In addition , the varnish
covering the rivets as a source of high out-gassing was also removed. All
technological processes for manufacturing the fairing were selected to rule out the
possibility of delivering contaminated units to assembly.
All units going to assembly are stripped of mechanical and fat contamination
with the help of lint-free napkins, moistened by clean solvent and packed in a
polyethylene cover.
95
J.l. Kleiman and R.C. Tennyson (eds.),
© 2001 Kluwer Academic Publishers .
96
Thermal vacuum treatment of units is carried out to eliminate outgassing of
nonmetallic materials . Specific treatment was determined on material specimens by
experimental measurement of their outgassing level in a vacuum chamber. Units and
assembly units are outgassed at 60°C, pressure 5 + 50 mm Hg for 48 hours.
Assembly of riveted units is carried out according to the following technique :
the structure is assembled, holes are made, metallic shavings are cleaned by a vacuum
cleaner, brushes and napkins, the structure is disassembled, and internal surfaces are
cleaned and then re-assembled.
A characteristic feature of a riveted construction is that not all riveted
connections can be disassembled and cleaned before the finished assembly.
Mechanical contamination in the form of metal particles and shavings accumulate in
these connections and may appear on surfaces during subsequent operations . Hence,
additional cleaning methods such as vibration and high-energy water jets are needed .
A vibration cleaning method consists of using vibration loading to remove
large metallic and nonmetallic particles, using a special vibration bench. Before
vibration loading, the fairing is assembled with the launch vehicle unit, and a
polyethylene screen is placed between them. The assembled construction is placed on
a vibration bench in a vertical position, fastened and loaded along longitudinal and
transversal axes consecutively at a frequency of 20+80 Hz and amplitude 0.3+0.5g, for
8 minutes . The vibration cleaning set-up of the unit is shown in Figure 3. After
vibration loading, the fairing is disassembled and particles which have fallen on the
screen are collected by a vacuum cleaner. Their quantitative and qualitative
composition is analyzed. Experiments have shown the vibration cleaning method is
effective for removing large particles, particularly metal shaving .
Cleaning by high-energy water jet is carried out in special facilities . Distilled
water with 0.01-0.02 per cent potassium bichromat is used. The cleanliness of the
solution is tested by preliminary filtration through a cambric filter for I hour. No
more than 10 mg of dry contamination is allowed. The cleaning is carried out by a
water jet at pressure 5 -6 kg/snr' at speed of movement of the cleaning appliance equal
to 0.8 m/min for 8 hours. The quality of cleaning is tested as described above, with
the dry residual mass not to exceed 10 mg. After cleaning by high-energy water jets,
the fairing is dried by compressed clean air.
A percentage of contamination collected by different consecutive methods of
cleaning has been calculated as follows:
vacuum cleaning - 31%;
vibration cleaning - 16%;
vacuum cleaning - 23%;
cleaning by high-energy water jet - 30%.
Thus the achievement of the required cleanliness level needs the application
of different cleanling methods; cleaning by high energetic water jet is the final and
most delicate operation .
After cleaning the fairing is replaced in a clean room for final assembly.
The assembled fairing is enclosed by a technological cover and vacuumed for
outgassing glue, greases and external thermal insulation which separately, don 't pass
97
thermal vacuum treatment. The outgassing is carried out at pressure 10-2 mm Hg for
48 hours .
Subsequently, the fairing is covered by a special polymeric film, reinforced by
a metal net for removal of static charge .
Another cover of a dense fabric is placed over the fairing and it is transported
to the railway platform, At the cosmodrome, dust covers are cleaned of atmospheric
precipitation and removed, and the fairing passes through additional cleaning
treatments if necessary. With the technological cover removed, the fairing is replaced
in a clean room and after mating with the spacecraft, the launch vehicle is transported
to the launch complex .
During the whole technological cycle of manufacturing the fairing,
cleanliness control is provided according to the following methods .
Particle concentration in the air is checked by a photometric method based on
light scattering by single particle. The light scattered is focused on a cathode of a
photoelectric multiplier that transforms it into electric impulses of specific duration
depending on time of flight and amplitude equal to particle size. Monitoring is carried
out by a PKZV device, with an operational range of 0,5-1,0; 1-2; 2-5; 5-10 and 10-
100 micron . With such techniques, the air of c1eanrooms, the fairing space and
technological gas are under control.
The results of these procedures are presented in Table I. For comparison,
level M6.5 FED-STD-209E requirements are also added. It is evident from the table
that the air cleanliness in the fairing space clean room, and compressed air for drying
the fairing corresponds to M6.5 FED-STD-209E cleanliness level, though particulate
concen-tration in the clean room air is increased below this level.
The residual quantity of molecular contamination on surfaces is checked by
a luminescence method , as fat contamination glows under exposure to ultraviolet
light. Control is carried out by the MIG-IO lamp. The threshold of response is below
100 mg/rrr'.
Particles on the surface are controlled by paper filter "Vladipor." The
method is based on the property of particles that stick to the surface of the filter paper.
The quantity, size and type of particles and fibres are determined with a microscope .
The weight of the particulates is calculated considering the density of the particles
and probability of sticking of a particle to the filter paper .
Control results data are presented in Table 2. For comparison, level 750 of
MIL-STD-1246 requirements are also added . It is evident from the table that the
residual quantity of particulate contamination on internal fairing surfaces doesn't
exceed the established required level.
Qualitative control is accomplished by wiping the surface with a lint-free
napkin . Traces of aluminum alloys on the napkin are allowed.
Measurements of the outgassing levels of nonmetallic materials were carried
out before and after thermal vacuum treatment. At first, specimens were kept in
gennetic chambers for 7 days at +50°C during the day and +20°C at night, that
corresponded to maximum outgassing of the materials while fairing and spacecraft
were together. Chrom atographic tests identified the following substances: acetone ,
98
ethanol , benzol, toluol, utanol , butylacetate, and chlorbenzol. In this situation, mass
loss was 5 -10 times smaller than the permissible level.
Afterwards, the specimens were heated in vacuum (P= 10-2 mm Hg) to a
temperature to which they are heated during launch . The vacuum chamber was
supplied with a plane contact heater and condenser system placed above the specimen .
During investigations the total mass loss and amount of condensed phase
were measured . When the specimens passed thermal vacuum treatments, the
outgassing level was measured once again . The data obtained are presented in Table
3.
The completed investigat ions showed that thermal vacuum treatment of
materials decreased their outgassing level in different ways. The level of outgassing
from fiberglass practically doesn 't change . The level of insolation materials after
thermal vacuum treatment decreases 1.5 +2 times; of rubbers 6+7 times.
Analysis of the data obtained shows that the majority of fairing materials
meets the requirements concerning outgassing , with the exception of some glues,
greases and insulating tapes with sticky layers. However, taking into account their
small contribution to the mass of nonmetallic materials , the total outgassing level will
comply with necessary requirements.
By analyzing the production process described for manufacture and testing of
fairin gs, it is possible to make conclusions about the conformity to modem cleanliness
requirements that provides the necessary quality of launch services for contemporary
spacecraft.
Table I: Summary of Control Results on air cleanliness as obtaine d by PKZV-906 device
Device reading for different size ranges Total quantity of

Conditions particles in air/
one liter
0.5 -1 1-2 2-5 5-10 5-10 0.5 um 5 um
Jlm Jlm Jlm Jlm Jlm and and
more more
The required
levels/M6.5 3500 25
FED-STD 209-E
Empty clean
room after 24 hrs 453 32 23 4 0 512 4
of ventilation
The clean room
after carrying out 3058 3 10 115 14 I 3498 15
the work
The fairing space 1202 61 16 0 0 1279 0
Compressed air
to dry the fairing 60 1 0 0 0 61 0
99
Table 2. Summary of Data on Residual Quantity of Particulate Contamin ation on the Surface Using filter
paper "Vladipor"
Residual Quantity of particulate contamination

Control zone Particles quantity according to level
750 MIL-STD-1246, um mg/m?
50 100 250 500 750
On 0.1 rrr' 105000 9630 23 1 8 1

Specification of area
On the
membrane 1008 92 2 0 0 10
Internal surface of the 14 2 0 0 0 8.9
fairing
9 1 0 0 0 5.7
14 1 0 0 0 7.4
4 1 0 0 0 3.4
15 2 0 0 0 9.2
4 1 0 0 0 3.3
5 1 0 0 0 2.6
Spacecrafts block 10 0 0 0 0 4.4
surface 12 2 0 0 0 8.7
7 0 0 0 0 3.2
6 0 0 0 0 3.0
4 0 0 0 0 1.8
100
Table 3. Outgassing Levels of Nonmetallic Materials
Outgassing level of nonmetallic materials
Materials
Before TVT After TVT
TVL CVCM TML CVCM

1 Fibreglass fabric,
impregnated with 0.59 0.01 0.55 0.002
phenolforrnaldehyde
resin
2 Fibreglass fibres
impregnated with 0.051 0.046 0.032 0.025
phenolforrnaldehyde
resin
3 Rubbers 6.56 0.97 0.57 0.06
4 Insulation materials:
-compounds; 0.16 0.015 0.031 0.015
-cable insulation; 0.03 0.002 0.007 0.0003
- insulation tube. 0.11 0 0.10 0
5 Epoxy resin based glue 1.76 0 0.97 0
101
Figure 1
A unit on vibration bench.
IN SITU XPS STUDIES OF KAPTON EXPOSED TO 5 EV ATOMIC
OXYGEN
E. GROSSMAN and Y. LIFSHITZ

Soreq NRC, Yavne 81800,
Israel
J.T. WOLAN, CK MOUNT and G.B. HOFLUND

Department ofChemical Engineering,
University ofFlorida, Gainesville, FL 32611
Abstract
In this study, a novel hyperthennal oxygen atom source has been used to perform in
situ O-atom erosion of Kapton surfaces at room temperature; this surface has been
examined using X-ray photoelectron spectroscopy before and after exposure to
different fluences of 0 atoms, and then after exposure to air . The XPS data indicate
that the O:C ratio decreases from 0.23 to 0.11 during a 24-hour exposure to a 5 eV
14
O-atom flux of about 1.4 x 10 0 atoms/cm/-s. This decrease in surface oxygen
content is consistent with loss of the carbonyl groups. The nitrogen concentration
increases with O-atom exposure, and a new N chemical state forms. Following the
24-hour, O-atom exposure, the sample was exposed to air for 3 hours and re-
examined using XPS. The O. Nand C concentrations return to the same values
obtained before O-atom exposure due to reaction with O2 in the air. Small amounts
of atmospheric moisture may adsorb as well. Previous data from space and ground
simulations indicate an increase in the 0 1s signal with exposure to atomic oxygen.
The results obtained in this study demonstrate that it is necessary to examine the
chemical effects of O-atom degradation of Kapton without air exposure before
surface characterization.
1.0 Introduction
Oxygen atoms are the most abundant constituents of the Earth's ionosphere at
altitudes ranging from 200-700 km [1-4). Spacecraft orbiting in this zone, referred
to as low Earth orbit (LEO), have velocities of -8 km/sec and collide with residual
• 14 IS 2
oxygen atoms at an effective energy of - 3-7 eV and a flux of 10 -10 O/cm -sec.
Although flight experiments have shown that most organic surfaces are eroded by
exposure to this atomic oxygen [5-8), the mechanism of this degradation is not
103
Protection of Space Materials from the Space Environment, 103-113 .
© 200 I Kluwer Academ ic Publishers.
104
understood. Thus far, O-atom erosion studies have been limited to samples exposed
to atomic oxygen in LEO (STS missions and NASA Long Duration Exposure
Facility [9-11] and in ground simulation facilities [12-14]. In this study, the
samples were exposed to air after the O-atom exposures and before surface analysis.
In this study, a thin polyimide Kapton 300HN film, used in thermal
blankets of spacecraft, has been characterized using X-ray photoelectron spectro-
scopy (XPS) before and after incremental exposures to the flux produced by a novel
electron stimulated desorption (ESD) atomic oxygen source developed by Hoflund
and Weaver [15]. The two primary objectives of the present study are to gain a
better understanding of the chemistry of the erosion of Kapton by atomic oxygen,
and to determine whether air exposure alters oxygen-exposed Kapton surfaces . The
latter objective is very important with regard to interpreting the results of previous
studies , and in the design of future experiments on this topic.
2.0 O-Atom Source Characteristics
The ESD source is ultrahigh-vacuum- (UHV) compatible , operates with the sample
15
at room temperature, and produces a 5eV high purity O-atom flux > 10 atoms em'
8
2S· 1 at a distance of 5 cm from the sample surface, with an 0 +:0 atom ratio < 10. .
These sources are superior to plasma sources in every respect (5 eV 0 atoms
compared to thermal atoms , no UV irradiation and no other oxygen species). Due to
the high reactivity of atoms , this O-atom source and a similar H-atom source have
been used in numerous surface studies [16-21J. The operational concept of the
hyperthermal oxygen atom generator (HOAG) is shown in Figure 1.
conflat flange
reflector
I
D-atom flux !
membrane
targ~t 4 0
4 0
• 0
10'6atomslern'. r
filament /
electrical feed~thru O2 supply tube
Figure 1
Schematic diagram of the atomsource.
105
Ultrahigh purity molecular O2 dissociatively adsorbs on a metallic Ag alloy mem-

brane at the high-pressure side and permeates at elevated temperature (>400°C) to
the UHV side. There, the adsorbed atoms are struck by a directed flux of primary
electrons which results in electron stimulated desorption (ESD) of atoms. Many
processes have to function in series at sufficiently high rates for the system to work.
These include dissociative adsorption of the molecular gas on the metal surface,
permeation of atoms thourough the membrane, and formation of the neutral flux by
ESD. Based on the measured ion energy distribution and mass spectrometric
experiments, the neutral energy distribution has a maximum at about 5 cV with full
width at half-maximum of about 3.6 eV.
3.0 Experimental
An as-received E. 1. du Pont Nemours & Co., Inc. Kapton® 300HN film was wiped
with iso-propanol and then inserted into the ultrahigh-vacuum (UHV) chamber
10
(base pressure <10- Torr) . XPS was performed using a double-pass cylindrical
mirror analyzer (DPCMA) (pHI model 25-270AR). XPS survey spectra were taken
in the retarding mode with a pass energy of 50 eV, and high-resolution XPS spectra
were taken with a pass energy of25 eV using Mg Ko, X-rays (pHI model 04-151 X-
ray source). The sample was tilted 30° off the axis of the DPCMA, and the DPCMA
accepts electrons emitted into a cone 42.6° ± 6° off the DPCMA axis. Data
collection was accomplished using a computer-interfaced, digital pulse-counting
circuit [22] followed by smoothing with digital-filtering techniques[23]. Since XPS
was performed using a defocused source, the XPS data represent an average over
the area analyzed (about 4 mrn in diameter).
XPS spectra were first obtained from an as-received, solvent-cleaned
sample. The sample was then transferred via a magnetically-coupled rotary/linear
manipulator into an adjoining UHV chamber, which houses the O-atom source.
There, the surface was exposed to the hyperthermal 0 flux and re-examined without
air exposure after total exposure times of 20 min. and 24 hours. The approximate
normal distance between the sample face and source was 15 em, at which distance
14
the flux is about 1.4 x 10 atoms/cm-s. The sample remained at room temperature
with a slight increase to - 50°C due to exposure to the X-ray source during XPS data
collection. The substrate temperature was determined using a K-type chomel-alumel
thermocouple.
Next, the sample was exposed to air (room temperature -19°C, relative
humidity -40%) for 3 hours following the 24-hour O-atom exposure. A structural
representation of the polymer with labelled atomic sites is shown in Figure 2 and
reference binding energies (BE)[24] used in this study are given in Table 1. These
reference spectra were used to shift the BE axis to the appropriate position. The
surface morphology of the irradiated samples was analyzed under ambient
conditions by atomic force microscopy (AFM) using a Digital Instruments
106
NanoScope II system with a 140 um scan head. The surface morphology was
detected with a resolution that ranged from 1 nm up to several micrometers .
Figure 2
Molecular structure of poly(elher imide) (Kapton HN) showing labeled atomic sites
Table I. Poly(ether imide) (Kaplon HN) Bes
Cis o Is N Is
1 2 3 4 5 6 1 2 ---
BE (eV) 284.7 285.6 285.7 285.8 286.3 288.6 532.0 533.2 400.6
% 36 10 10 19 10 17 77 23 100
'c Is percentages add 10 102% as staled in the original reference (24).
XPS survey spectra obtained from an as-received, solvent-wiped Kapton surface

before and after the 20-min. and 24-hour, a-atom exposure are shown in Figure 3a,
band c respectively. The spectrum in Figure 3d was taken after the 3-hour air
exposure following the 24-hour a-atom treatment. Peak assignments shown in
Figure 3 pertain to all four spectra. The peaks apparent in these spectra include the
C Is. a Is. N Is and a Auger peaks. Significant changes in relative peak heights
are observed for the C, N. and a features due to the a-atom exposures. An estimate
of the near-surface composition has been obtained from the peak areas in the survey
spectra by assuming that this region is homogeneous and using published atomic
sensitivity factors [25]. The compositions determined in this manner are presented
in Table 2 for the as-entered, 20-min . and 24-hour a-exposed, and air-exposed
surfaces. Unexpectedly, the near-surface a concentration decreases significantly
107
with increasing exposure to the O-atom flux while the concentrations for C and N
increase. After the 3-hour air exposure. the composition returns to that of the
solvent-cleaned sample. The reasons for these compositional changes are described
below.
800 800 400 200 0

Binding Enemv (eV)
Figure 3
XPS survey spectra obtained from a solvent-cl eaned. Kapton film after (a) insertion into the vacuum
system, (b) a 20-min .. and (c) a 24-hour exposure to the hypcrthermal O-atom flux, and (d) a 3-hour air
exposure following the 24-hour O-atom exposure.
High-resolution XPS CIs, 0 Is, and N Is spectra obtained from an as-

received, solvent-wiped Kapton surface before and after a 20-min . and 24-hour, 0-
atom exposure are shown in Figures 4 (spectra a-c) and 5 (spectra a-c) respectively.
Spectra d were obtained after the 3-hour air exposure following the 24-hour. 0-
atom exposure. The chemical species and assigned BEs correspond to the molecular
sites labelled in Figure 2 and values given in Table 1. Significant variations in peak
shapes and positions are observed between the non-exposed, O-atom-exposed and
air-exposed surfaces indicating that the chemical species distribution is altered
significantly by exposure to the O-atom flux and then to air.
The C Is peaks shown in Figure 4 (side A, spectrum a) indicate that the
predominant form of carbon is aromatic at site 1 as shown in Figure 2. In Figures 4
(side A, spectrum bod), the features due to the carbonyl group at site 6 are
eliminated by the O-atom exposure and do not increase with exposure to air as
expected. This site-specific attack is consistent with production of CO2 followed by
desorption [26). Although subtle, the shoulder due to site-4 carbon (and possible
site-3 and site-5 carbon) appears to decrease initially and then increase again with
108
the 24-hour O-atom exposure. This indicates that the relative rate of the loss of the
various carbon species with time is complex.
The irradiation of the Kapton surface with hyperthermal oxygen atoms
causes, besides the chemistry changes, surface modification. Upon successive
irradiation, equilibrium is often reached where the irradiated polymer consisted of
a modified layer that is continuously eroded and advances to the non-affected
material. XPS data reflects mostly the nature of the modified layer (generated by
chemical reactions and/or fragmentation due to scissioning) and less of the non-
affected material .
Table 2. Near-surface composition determined from XPS data obtained from solvent-cleaned, O-atom and
air-exposed Kapton
Surface Composition (at%)

Sample treatment 0 C N
Theoretical 17.2 75.9 6.9
Solvent-cleaned 18.1 77.7 4.2
20-min. O-atom exposure 14.4 78.4 7.2
24-hour O-atom exposure 9.2 83.0 7.8
3-hour air exposure following 17.9 78.2 3.9
24-hour O-atom exposure
29
RFPlasma 28.4 64.9 6.7
9
LEO Environmene 22.2 70.8 7.0
Binding Energy (eV)

Figure 4
XPS (side A) C Is and (side B) 0 Is spectra obtained from a solvent-cleaned Kapton film after (a)
insertion into the vacuum system, (b) after a 20-min, and (c) a 24-hr exposure to the hyperthermal
O-atom flux, and (d) a 3-hr air exposure following the 24-hr O-atom exposure
109
The 0 Is peak shapes and positions obtained from the sample after the
various treatments are shown in Figure 4&. With just a 20-min. exposure to the 0-
atom flux, the contribution from oxygen is significantly reduced from 18.1 to 14.4
at%. This reduction is consistent with the production and subsequent desorption of
CO2 and the loss of site-6 C, and the rate of loss of both site-I and site-2 0 appear
to be similar .
XPS
N 15
Figure 5
XPS N Is spectra obtained from a solvent-cleaned Kapton film after
(a) insertion into the vacuum system, (b) a 20-min ., and (c) a 24-hour exposure
to the hyperthermal O-atom flux. and (d) a 3-hour air exposure following
the 24-hour O-atom exposure .
Water and site-2 0 have BEs which are similar. Any contribution due to
water in these spectra is probably small. After the 24-hour exposure, the near-
surface 0 content is reduced further to 9.2 at%, and new features are apparent at
530.7 and 534.3 eY. The identities of these chemical states are not known. The 0 Is
spectrum shown in Figure 4 (side B, spectrum d) was obtained after exposing the
24-hour, O-exposed surface to air for 3 hours. This spectrum is quite different than
those in Figure 4 (side B, spectrum a-c). The oxygen content is increased from 9.2
to 17.9 at% during the air exposure. The predominant feature in this spectrum lies
at a lower BE than the 0 species in Kapton and may be due to the formation of
aromatic C=O groups, O=C-N groups or C-O-C,groups since their 0 Is BEs lie in
the 530.1 to 532.0 eV range [24J. A distinct feature due to water is not present,
indicating that the increased oxygen concentration is primarily due to reaction with
O2 in the air, and not adsorption of water. In this regard, the O-exposed surface
110
behaves similarly to an ion-irradiated polyethylene surface which also dissociatively

chemisorbs oxygen during air exposure [27]. In both cases, O-exposed Kapton and
ion-irradiated polyethylene, bonds are broken leaving reactive species in the near-
surface region which do not crosslink due to geometrical bonding constraints.
When exposed to O 2 , they cause dissociation of the O2 and bond with it. This view
of the formation of trapped reactive species is consistent with the results of a
molecular dynamic simulation study of ion-irradiated polyethylene by Beardmore
and Smith [28].
The N Is peak obtained from the solvent-cleaned Kapton is shown in
Figure 5 (spectrum a). It has a BE of 400 .7 eV. After just a 20-min. exposure to the
o flux , a distinct shoulder is apparent on the low-BE side of the predominant peak .
After the 24-hour exposure, this shoulder is developed into a well-defined peak with
a BE of 399 .0 eV of equal size to the peak due to the initial N state . After the CO
portions of the Kapton are removed to form CO 2 by the reaction with atomic
oxygen, the N is bonded to the aromatic C as CoN species, which has a N Is BE of
399.0 eV [24]. Furthermore, the N content of the surface increases with increasing
O-atom exposure as shown in Table 2, indicating that the rate of removal of these N
species by atomic oxygen is slower than the removal rates of 0 and C. This feature
is essentially unchanged by the 3-hour air exposure as shown in Figure 5 (spectrum
d) except that it is smaller relative to the 0 Is and C Is features as shown in Figure
2 (spectrum d) . This means that the reaction of the damaged surface with O 2 does
not significantly alter the chemical environments of the two forms of nitrogen
present on the O-exposed surface.
AFM images obtained from an as-received, solvent-wiped Kapton surface
before and after a 20-min. and lO-hour O-atom exposure are shown in Figures 6
(spectra a-c) respectively. After a short exposure for only 20 min ., the surface
morphology has been changed and the surface roughness increased. The roughness
continues to increase with further exposure.
In all of the previous XPS studies on the O-atom erosion of Kapton, the
samples were exposed to air after the O-atom exposure and before analysis by XPS .
Typical examples [29] are given in the bottom two rows of Table 2. The data
presented above the last two rows indicate that increasing O-atom exposure results
in a decreasing 0 concentration in the near-surface region . However, these
examples indicate increased 0 concentrations that result from the air exposure
rather than the O-atom exposure. Thus, the O-atom source and XPS must be
contained in the same UHV system to study the O-atom erosion on Kapton with
XPS .
5.0 Conclusion
A novel hyperthennal oxygen atom source has been used to perform in situ O-atom
erosion of Kapton surfaces at room temperature, and this surface has been examined
using XPS before and after O-atom exposure. The XPS data indicate that the initial
111
attack site on the Kapton is the carbonyl group, to form CO2, which desorbs. Thus,
the XPS 0 concentration decreases from 18.1 to 9.2 per cent during the 24-hour
exposure to the O-atom flux. Previous data from space and ground simulations
indicate an increase in the near-surface 0 content with exposure to atomic oxygen.
This 0 increase is because these surfaces were exposed to air before analysis as
demonstrated in this study. The nitrogen concentration determined using XPS
increases with increasing exposure to the 0 flux as both the 0 and C concentrations
decrease. This indicates that the slow step in the erosion of Kapton is the removal of
N-containing species. Both the C- and N-chemical states are changed significantly
during exposure to the 0 flux. The C bound as C=O is selectively eliminated, and
the N bonded to two C=O groups in Kapton is converted to C-N species. After the
24-hour exposure to the 0 flux, the sample was exposed to air for 3 hours and re-
examined using XPS. The near-surface composition returns nearly to its original
composition primarily through reaction with O2 in the air, but the chemical-state
distributions of the near-surface species are altered. The AFM images indicate that
the pristine Kapton surface is smooth and the O-atom exposure significantly
increases the roughness.
112
(a
(b
(c
Figure 6
AFM images obtained from a solvent-cleaned. Kapton film after (a) insertion into the vacuum system and
(b) 20-min . and (c) a 24·hour exposure to the hyperthermal O-atom flux.
113
6.0 REFERENCES
I. Koontz S. L., Leger L. J., Rickman, S. L., Hakes, e. L., Bui, D. T., Hunton, D. E. and Cross,
J. 8., J. Spacecraft and Rockets, 32, (1995) 473.
2. Cross, J. B., Koontz, S. L. and Hunton, D. E., J. Spacecraft and Rockets, 32, (1995) 496.
3. Jursa A. E. (ed.), Handbook of Geophysics and the Space Environment, Air Force Geophysics
Lab., U. S. Air Force (National Technical Information Service), 1985, Chap 14, pp. 14-38, 14.
4. Hedin A. E., J. Geophys. Res., 88, (1983) 170.
5. Reddy, M. R. , J. Mater. Sci. 30, ( 1992) 281.
6. Packirisamy, S., Schwam D. and Litt, M. H. , J. Mater. Sci. 30 (\995) 308.
7. Koontz, S. L., Albyn , K., Leger, L. J., Hunton, D.E., Cross J.8. and Hakes, CJ., J. Spacecraft
and Rockets 32, (1995) 483.
8. Reddy, M. R., Srinivasamurthy N. and Agrawal, B. L., ESA Journal 16, (1992) 193.
9. Koontz, S. L., Albyn K. and Leger, L. J., J. Spacecraft and Rockets 28, (1991) 315.
10. deGroh K.K. and Banks, BA, Proc. of the lSI LDEF Post Retrieval Symp., Kissimmee, FL,
NASA CP-3134, (1991) 781.
II. Banks, BA, Rutledge, S.K., deGroh, K.K., Mirtich, MJ., Gebauer, L., Olle R. and Hill, C.M.,
Proc. of the 5th Inter. Syrnp. On Materials in a Space Environment, Cannes-Madelieu, France,
(1991) 137.
12. Tennyson, R.e. , Canadian J. of Physics, 69, (199 1) 1190.
13. Cazaubon, 8. , Paillous, A., Siffre J. and Thomas, R., J. Spacecraft and Rockets, 33, (1996)
870.
14. deGroh, K.K,. Terlep JA and Dever, T.M., NASA TM-I05378, Sept. 1990.
15. Hoflund G.B. and Weaver, H ., Meas. Sci. Techno!. 5, (1994) 201.
16. Wolan, J. T., Mount C. K. and Hoflund, G. 8., J. Vac. Sci. Technol. A 15, (1997) 2502.
17. Epling, W.S., Mount ex. and Hoflund, G.8. , Thin Solid Films 304, (1997) 273.
18. Hoflund G.B. and Wolan, J.T., J. Vac. Sci. Technol. B 16, (1998) 1446.
19. Wolan ,J.T., Mount C.K. and Hoflund, G.B., J. Vac. Sci. Technol. A 16, (\998) 2546.
20. Wotan, J.T., Mount C.K. and Hoflund ,G.8. , Appl. Phys. Len. 72, (1998) 1469.
21. Epling, W.S., Mount,C.K., Hoflund,G.B. and Young,V.Y., Appl. Surface Sci. 126, (\998) 235.
22. Gilbert, R.E., Cox D.F. and Hoflund, G.B., Rev. Sci. Instr. 53, (\982) 1281.
23. Savitzky A. and Golay, M.K.E., Anal. Chern. 36, (1984) 1627.
24. Beamson G. and Briggs, D., High Resolution XPS of Organic Polymers: The Scienta
ESCA300 Database, John Wiley & Sons Ltd., Chichester, 1992, p. 214.
25. Wagner, C.D., Riggs, W.M., Davis, L.E., Moulder J.F. and Muilenberg, G.E. (eds.), Handbook
of X-Ray Photoelectron Spectroscopy, Perkin-Elmer Corp., Eden Prairie, MN, 1979.
26. Wydeven,T.W., Johnson,e.e., Golub,MA, Hus, M.S. and Lerner, N.R., Plasma
Polymerization, 108, (1979) 527.
27. Hoflund, G.B., Rex, S.L., Young, V.Y., Ruck, D.M., Eisenreich, J., Trautmann, e., Wagner,
D. and Angert, N., Proc. 9th Int. Conf. Ion Beam Mod. Mater. (IBMM '95) edited by Williams,
J.S., Elliman, R.G. and Ridgway, M.C., Elsevier, Amsterdam (1996) 1059.
28. Beardmore, K. and Smith, R., Nucl. Instrum. Meth. Phys. Res. B 102, (1995) 223.
29. Golub, M. A., Wydeven, T. and Cormia, R.D., Polymer Communication 29, (\ 988) 285.
A LOW COST, LIGHTWEIGHT REUSABLE SENSOR FOR ATOMIC
OXYGEN DETECTION
1.1. OSBORNE, G.T. ROBERTS and S.B. GABRIEL

Department ofAeronautics and Astronautics,
University ofSouthampton, Highfield, Southampton,
Hampshire, UK, S017 lBJ
A.R. CHAMBERS
Department ofEngineering Materials.
University ofSouthampton, Highfield, Southampton,
Hampshire, UK, S017 lBJ
Abstract
Atomic oxygen (AO) erodes many materials used on low Earth orbit (LEO)
spacecraft and so it has become necessary to find ways of measuring the AO flux
encountered by such vehicles. Many techniques have been used for the measurement
of the atomic oxygen environment in LEO; however, their mass and power budgets
often preclude their use on microsatellite missions. Those techniques that are suitable
for such missions frequently suffer from other limitations such as poor accuracy and
the reliance on a coating or a film that, once totally eroded, prevents further
measurements of AD fluxes. Thus, there exists no detector entirely suitable for
accurate determination of atomic oxygen fluxes on microsatellite missions.
Semiconducting detectors offer a promising solution to this problem.
Laboratory investigations have shown that an n-type semiconductor experiences an
increase in resistance when exposed to a flux of AO. The change in resistance is not
permanent; heating the semiconductor to modest temperatures can nearly restore the
original resistance. The ability to regenerate the sensor allows further measurements
to be performed. Preliminary experimental results of such a device are presented and
the development of a lightweight, low cost and low power microsatellite experiment
based on these sensors is described.
1.0 Introduction
A constituent of particular concern in the low Earth orbit (LEO) atmospheric

environment is atomic oxygen (AO). It is well known that AD oxidizes and erodes
many materials such as polymers used in thermal control blankets, and metals such
as the silver interconnects used in solar arrays [1]. Such erosion can impact the
115
J.J. Kleiman and R.C. Tennyson (eds.),
116
operation of the satellite. Atomic oxygen is formed from the photodissociation of

molecular oxygen by solar vacuum ultraviolet light, thus the AO density and hence
the rate of erosion at a given altitude is linked to the solar activity. It has also been
shown that the synergistic effects of combined AO, solar UV and even soft X-ray
exposure can further enhance erosion (2).
Moreover, atomic oxygen is involved with the generation of the glow
observed around many vehicles in LEO. This glow has become known as ' Shuttle
glow' and can interfere with the operation of optical instruments working in the IR,
UV and visible regions of the electromagnetic spectrum (3).
2.0 Atomic Oxygen Sensors
Thus, there has been great impetus to develop devices for measuring AO fluxes not
only for use in orbit but also in ground-based facilities designed to simulate the LEO
AO environment. The requirements of any AO sensor for satellite missions include
low mass, low power consumption, small volume and (ideally) reusability. These
issues become even more stringent for microsatellite missions, which are of
particular interest for the authors.
Several different techniques have been used in space to measure atomic
oxygen fluxes and fluences. The earliest method recorded in the literature is that of
the witness sample. The advantage of this technique is that it is simple, lightweight
and requires no power unless sample heating is provided during AO exposure. The
disadvantages are that it only provides the total fluence, it requires retrieval, and is
subject to errors in the erosion yield values and erroneous mass changes due to post
retrieval contamination.
Mass spectrometers have been widely used not only for the task of AO
measurement but also for the detection of AO/polymer reaction products (4).
However, they appear unattractive for application as devices for small satellite
missions due to the large size, mass and power consumption inherent to the units.
Quartz crystal microbalances (QCMs) have been successfully flown
onboard microsatellite missions. In this application, the piezoelectric crystal is coated
with a material which is reactive with AO; carbon-coated QCMs (C-QCMs) were
used on the BREMSAT mission (5). It is the mass change associated with this
reaction that is recorded by the QCM. thus an erosion yield value is needed to
convert the rate of change of mass into an AO flux. This requirement often leads to
considerable uncertainty in the flux measurement. Once the coating becomes
completely oxidized, no more measurements can be made.
The last technique that has been applied to satellite and sounding rocket
missions is that of thin film actinometry [6,7]. In this method, thin films of materials
which are both reactive with AO and electrically-conductive are deposited onto an
inert and electrically-insulating substrate. Both carbon and silver films have been
used in this manner [8,9]. Whilst the use of carbon would seem beneficial because
the oxides formed in the reactions with AO are volatile (compared to the stable solid
oxide of silver) the erosion yield of carbon is temperature-dependent leading to the
requirement of temperature control (10).
117
The reaction that occurs when the thin films are exposed to AO reduces the
conductor thickness , thereby increasing its resistance. The recorded change in
resistance can be related via the change in thickness to the AO flux by means of an
erosion yield; thus, this method can be susceptible to the same errors as the C-QCM
and witness sample techniques. These sensors become inoperative when the
conductor material has been oxidized to such an extent that the films go open circuit
or when a diffusion-limited regime occurs due to the presence of a growing oxide
layer.
Thus witness samples , C-QCMs and actinometers are all single use devices ,
but benefit from low power consumption (if any) and reasonable mass budgets.
Whilst mass spectrometers can provide continuous time-resolved measurements of
AO flux, they are bulky, heavy and require relatively large amounts of power. So it
can be seen that there is not one atomic oxygen detector that has the ideal
chamcteristics detailed above.
3.0 Semiconductor Sensors
The application of semiconducting materials to gas sensors has received much

attention in recent years [11]. Essentially, these devices work by the interaction of an
adsorbed species on the surface of the semiconductor and the conduction electrons
within the semiconductor. This interaction between chemisorbed species and
electronic structure affects the electrical characteristics of the material .
Metal oxide semiconductors have commonly been used because the ionic
structure of these materials means they have large band gaps and hence , relatively
few intrinsic charge carriers. The majority of the charge carriers are therefore
extrinsic, and so the conducti vity of the sample is highly sensiti ve to any form of
external influence. The action of oxidizing and reducing agents on Il- and p-type
semiconductors . results in opposite responses . Oxidizing agents remove electrons
from the surface regions of the semiconductor and localize them in surface states.
Reducing agents on the other hand introduce electrons into the surface regions .
The adsorption of an oxidizing species such as AO on n-type
semiconductors (where the charge carriers are electrons) reduces the charge carrier
concentration in a relatively thin zone called the depletion layer. This depletion
creates a two layer system between a net positive charge distributed in a space-charge
layer (SCL) and the negatively-charged surface states. In the SCL, the electronic
structure of the semiconductor is modified, forming a potential gradient for the
charge carriers. In this manner , the electrical conductivity of the surface regions
becomes reduced and this can be detected by simple resistance measurements. The
converse is true in the case of adsorption of reducing species on the surface of an n-
type material ; in this case, the conductivity increases in a narrow zone called an
accumulation layer.
In single crystals , the surface regions affected by charge carrier depletion or
accumulation are often thin compared to the bulk thickness , and so the change in
resistance due to chemisorption is small because of a shunting effect. One way to
increase the sensitivity of the detectors is to use thin films of materials. In this case,
118
the region affected by the adsorption of species on the surface is large compared to
the overall film thickness, and so the resistance change becomes a significant fraction
of the pre-exposure resistance. The structure of many thin films (depending on the
method of manufacture) is such that they consist of many small crystallites
interconnected by bridges and asperities. These features often have thicknesses of
similar order to the SCL, further enhancing the resistance changes because, to
conduct a current charge, carriers must cross these structures. The small size of the
bridges means that the whole volume enclosed within them is influenced by the SCL,
rather than just the surface layers.
Whatever the form of semiconductor, the more adsorbed species on the
surface the greater the change in resistance. However, there is a saturation limit
which occurs when the number of atoms, radicals or molecules removed by various
processes such as recombination. desorption, and collisions, balances the number
adsorbed from the surrounding medium. At this point, the concentration of surface
states is stationary and the change in resistance ceases.
Semiconducting detectors have high sensitivities: for example, the
sensitivity of zinc oxide to atomic oxygen has been found to be 109_10 10 atoms em?
4
S-I, which in LEO, would translate to a minimum measurable density of around 10
atoms em? (12). This density is found at altitudes of around 1000 km at average solar
activity.
The main benefit of semiconductor sensors is the ability to recover the
original electrical properties (and therefore the removal of adsorbed species from the
surface) by heating to modest temperatures. This process allows regeneration of the
sensors and hence another measurement of the environment can be made. Potential
disadvantages of this technique are the high thermal coefficients of resistance and the
sensitivity to UV of the semiconductor.
4.0 The Application of Semiconductor Sensors to AO Measurement
4.1 SINGLE CRYSTAL RESULTS
The work conducted at Southampton, England and the European Space Research and
Technology Centre (ESTEC), Netherlands, has been to develop a sensor based on
zinc oxide suitable for operation on a microsatellite in geostationary transfer orbit
(GTO). Initial experimentation was with single crystals because of the homogenous
nature of the surface of such samples. Figure I shows the response of the single
crystal to atomic oxygen exposure in the ESTEC pulsed atomic oxygen source.
Immediately after starting the source at 1Hz, the resistance of the single crystal falls;
the most likely explanation for this is a photoconductivity effect due to the UV
produced by the oxygen plasma. It can be seen that the resistance increases with AO
exposure but the rate of change of the resistance slows down after around 3 minutes,
evidence of the start of sensor saturation.
Switching the laser repetition rate to 7 Hz generates more UV than at 1 Hz
and so the initial resistance response is to decrease just as it does at start-up. The
resistance then increases at a higher rate than during the 1 Hz stage, and reaches a
119
stationary value after -6 minutes elapsed time. The repetition rate was reduced to I
Hz at 9.6 minutes at which time the resistance falls until it reaches a new stationary
value. This behaviour is evidence of flux-dependent rate coefficients. After the laser
was switched off, the resistance begins to climb, an effect often termed the 'after
effect' and is caused by the chemisorption (hence ionization) of previously
physisorbed (unionized) atoms (13).
It was found that heating the single crystals to a temperature of ~3 5 0 K for
~ 10 minutes was sufficient to regenerate the surface properties of the crystal ready
for another exposure. However, difficulties in mounting of the crystals and their
fragile nature has, so far, prevented their use in a space-rated experiment.
1.45E+05 ·, ._ ..__ _ _..•.._ _ ..- _ .._.•........._ .._..•......_._._.-.-.._._ __. : 315
1.40E+05 t:--------MlVrJ~,,---------f
: 314
! 1.35E+05 · .., \J'''''lf\---'y---~' 313 g

j
'ill
, "]""
A '
". . 312 !I
- resistance
=<'" -to- temperature
..................................
1.20E+05 +.,-,.,..,...,.-fT"-~-+.-.".,..,.,.,..,.,...,....,.,.,..,.,...,.,.,.,.,..,.,..,Hr..,.,.,..,.,.,..,f,,,...,....,,..,..r,,.,.,,.,.,,..+.,.,.,......,.,.f
... . . . . . . . . 310
O. 2.00 6.00 8.00 1 .00 12.00 : 14.00 16.00 18.00
Time (min) ;
AO on, 1 Hz AOon,7Hz AO on, 1 Hz AO off
Figure 1
The affect of AO exposure on the resistance of a zinc oxide single crystal
4.2 DESIGN OF TIllN FILM SENSORS
For this reason, it was decided to investigate the properties of thin films for use as the
sensor elements in a satellite experiment. Mounting of thin film sensors is easier than
single crystals because the films can be deposited directly onto a mechanically strong
substrate. The design of the sensor is shown in Figure 2. The substrate used is a pure
(99.6 per cent), smooth «0.076 urn) alumina sheet. The first stage in the
manufacturing process is the deposition of a thick film heater on the back of the
substrate. This heater is fabricated from a ruthenium oxide base in a glass frit, and
electrical contact is provided by the deposition of solderable silver/palladium
contacts. Both the heater and contacts are fired onto the substrate, improving the
thermal contact over what is achievable using standard thick film heaters and organic
glues.
120
Front Back
o = zinc oxide film

E2 = gold contact
-;! = silicon dioxide overlay
o = contact, silver/pa lladium
• = resistor material
IE l 16.94 mm
Figure 2
The layout ofthe thin film sensors and associated heater
To make the sensors described here, a zinc oxide target was sputtered in a
pure argon atmosphere. The sensors are provided with electrical contacts by
evaporating gold tabs that overlap the zinc oxide and substrate . To deconvolute the
effects of temperature and UV exposure from the AO signal, control sensors are
used. This was achieved by overlaying a thin (60 om) silicon dioxide film over two
of the four zinc oxide films per substrate . This completed the sensor fabrication.
4.3 EXPERIMENTAL RESULTS
So far, only preliminary testing has been carried out with the semiconducting thin
films . Figure 3 shows the results of a typical AO exposure for a 0.5 um film,
although 1.0 and 2.0 um films have also been tested. It can be seen that immediately
after switching on the AO source, the resistance of the film starts to increase. The
rate of resistance increase falls with increasing AO fluence until it reaches a stable
value, after which the AO source was switched off. (The pre- and post-exposure
temperature oscillations are caused by changing the sensor position above a
cryogenic pump.)
It has been found that the silicon dioxide overlay intended as an AO barrier
is successful for the 0.5 and 1.0 um zinc oxide films only; it appears to be permeable
for the 2.0 urn zinc oxide film, perhaps due to increased surface roughness of this
film . Also, the films exhibit a hysterisis in that the regeneration stage does not always
bring the resistance of the films back to the pre-exposure value. However, at this time
it is not clear if this is a short-lived effect that will discontinue with further AO
exposure/regeneration cycles or whether it will always be a feature of the sensor
operation.
121
1.06 320
1.05
Y.r : 319
fr
~
~ -
.\
1.04 -----li---------·--· ........ --
I \
..................._- . ....................... .............. . 318
s"
t)
1.03
- ~~ ir""x 1-- resistance 317
g
1 :::~::.~.I1"P.~~.!
~
1;i
....\ . f:!
' iii
~ 1.02
.~
'"
'\
i\
316
~
e
"BII> 8-
___""l ___ d 315 ~
E
<a 1.01 ....... I.! \\
§
0 :11 " ." i
314
1.00
Z
0.99
.'\.../ .l, \
\,..
313
0.98 , ., - ....: .. .. " ,' . . 312

...... . ..
0 2 4 6 8 io 12 14 16 18 20 22 24 26 28 30 32
TIlIlC (mirutes)
Atomic oxygen on Atomic oxygen off
Figure 3
The results of an AO exposure on a 0.5 urn zinc oxide film
5.0 Atomic Oxygen Experiment 2
The opportuni ty arose of flying an atomic oxygen experiment onboard the third
Defence and Evaluation Research Agency (DERA) microsatellite STRV-l c, which
will be launched as a secondary payload with STRV-ld into geostationary transfer
orbit (GTO) by an Ariane-5 launch vehicle. Atomic oxygen experiment 2 (AOE-2)
comprises two units , an electronics unit that contains all of the temperature control ,
resistance measurement, power conversion and data transfer circuitry (14). The
second is a sensor unit (Figure 4) which is a ' patch' designed to be attached to an
exterior surface of the spin-stabilized vehicle. Four silver thin films and eight zinc
oxide films (of the design described previously) are mounted on this patch which is
fabricated from a cerami c material. A harness connects the two units ; this format has
been adopted to improve the flexibility of the experiment. The total mass of the
equipment is less than 350 grams and the maximum power consumption is 3 Watts .
122
Figure 4
The design of the AOE-2 sensor unit
The nature of the GTO is ideal for evaluation of the zinc oxide devices as a
reusable sensor for space-based AO measurements. Atomic oxygen flux will be
received as the satellite passes through the upper atmosphere during perigee; during
this time , the experiment will be in its measurement mode. When STR V-lc
approaches an altitude where the AO flux is much reduced, if not absent (for example
on the approach to apogee), the experiment will be switched into regeneration mode
to heat the semiconductor sensors and thus renew them for the next measurement.
Also, the high radiation doses that will be received (from the radiation belts) during
the 200-day lifetime of AOE-2 will enhance any degradation above the value that
might be expected in LEO during a similar period . This should allow an accelerated
radiation life test equivalent to a much longer duration in LEO .
Construction and preliminary testing of the sensor unit is presently
underway. The experiment AOE-2 is scheduled for delivery in July/August 1998,
with launch planned for the last quarter of 1999.
6.0 Conclusions
It has been shown that a metal oxide semiconductor can be used as a sensor of atomic
oxygen in a ground-based laboratory simulation of the LEO AO environment. In
particular, zinc oxide single crystals and thin films have been tested . It is intended to
utilize these sensors in a space-rated unit. This experiment has been developed
because to date , no other technique for AO measurement is available which satisfies
the microsatellite mission needs of low mass. low power drain, low volume and
123
reusability. It is believed that the zinc oxide sensors described in this paper will
achieve these criteria.
7.0 Acknowledgements
The authors would like to thank M. van Eesbeek and 1. Matcham for their permission
and assistance in using the ATOX source at ESTEC, The Thick Film Unit
(University of Southampton) and K. Lawson (School of Industrial Manufacturing
Science, Cranfield University) for manufacturing the heaters and the zinc oxide
films, and C. Bielby and R. Stansbridge for making the mechanical and electrical
equipment for the laboratory and AOE-2. Lastly we would like to thank DERA for
providing the opportunity of a flight on STRV-lc and C. Maag for his help and
assistance.
8.0 References
I. Raja Reddy, M.R. (1995), Review: Effect of low Earth orbit atomic oxygen on spacecraft materials,
Journal ofMaterials Science 30, 281-307 .
2. Koontz, S., Leger, L., Albyn, K., Cross, 1. (1990), Vacuum ultraviolet radiation/atomic oxygen synergism
in materials reactivity, Journal ofSpacecraft and Rockets 27, no. 3, 346-348.
3. Garrett, H.B., Chutjian, A., Gabriel, S. (1988), Space vehicle glow and its impact on spacecraft systems,
Journal ofSpacecraft and Rockets 25, no. 5,321-340.
4. Visentine, 1.T., Leger, L.J. (1986), Material interactions with the low Earth orbital environment: accurate
reaction rate measurements, Proceedings ofthe NASA Workshop on Atomic Oxygen Effects, Pasadena, CA,
JPL87-14, 11-20, 10"'-II'h November 1986.
5. Drolshagen, G., Van Eesbeek, M. (1991), Atomic oxygen fluence detector for BREMSAT: scientific and
technical description, Proceedings ofthe 5'h International Symposium on Materials in a Space Environment,
Toulouse, 405-409 , September 1991. See also Van Eesbeek, M. (1996), Atomic oxygen fluence detector for
BREMSATscientific and technical result reporting, ESTEC Working Paper 1921.
6. Harris, I.L., Chambers, AR., Roberts, G.T. (1997), A low cost microsatellite instrument for the in situ
measurement of orbital atomic oxygen effects, Review ofScientific Instruments 68, no. 8,3220-3228.
7. Dickinson, P.H.G., Bain, W.e. , Thomas, L., Williams, E.R , Jankins , D.B., Twiddy, N.D. (1980), The
determination of the atomic oxygen concentration and associated parameters in the lower ionosphere,
Proceedings ofthe Royal Society ofLondon A 369, 379-408 .
8. Linton, RC ., Vaughn, 1.A., Finckenor, M.M., Kamenetzky, R.R., DeHaye, R F., Whitaker, A.F. (1993),
Orbital atomic oxygen effects on materials: an overview of MSFC experiments on the STS-46 EOIM-3,
AIAA paper, AIAA-93-4102.
9. Harris, I.L., Chambers, AR., Roberts, GT. (1997), Preliminary results of an atomic oxygen spaceflight
experiment, Materials Letters 31, 321-328 .
10. Caledonia, G.E., Krech, RH., Oakes, D.B. (1994), Laboratory studies of fast atom interactions with
materials, Proceedings of the 6,h International Sympos ium on Materials in a Space Environment, ESTEC,
Noordwijk, The Netherlands, 285-290, September 1994.
II Moseley, PT , Tofield, B.e., (editors) (1987), Solid state gas sensors, lOP publishing Ltd., Bristol.
12. Livshits, AI. , Gutman, E.E., Myasnikov, I.A , Fursa, V.M., Padeev, V.I., Romanovski, Yu.A., Kriman,
E.1. (1981), Atomic oxygen analyser with semiconductor sensors for researching the upper atmosphere of
the Earth, Pribory 1 Tekhnika Eksperimenta, no. 3,177-180.
13. Gutman, E.E. (1984), Effect of the adsorption of free atoms and radicals on the electrophysical properties
of semiconducting metal oxides, Russian Journal ofPhysical Chemistry 58, 801-821 .
14. Gabriel, SR, Osborne, 1.1., Roberts, G.T., Chambers, A.R. (1998), Development of a renewable atomic
oxygen sensor for low Earth orbit, accepted for publication in Journal ofSpacecraft and Rockets .
LARGE-SCALE ECR-CVD PREPARATION OF
INTEGRATED THIN-FILM STRUCTURES
FOR SPACE APPLICATIONS
ROMAN V. KRUZELECKY and ASOKE K. GHOSH

MPB Technologies Inc.,
Space and Special Products Division
151 Hymus Blvd. ,
Pointe Claire, Quebec , H9R 1E9
ETHEL POIRE and DARIUS NIKANPOUR

Canadian Space Agency
Space Technologies
6767 route de l'Aeroport
Saint-Hubert, Quebec , J3Y 8Y9
Abstract
ECR plasma discharges, with their characteristic high activation of reactive

precursors and low ion energies, have facilitated new deposition chemistries, as well
as the low-temperature processing of more delicate substrate materials. MPBT, with
the assistance of the Canadian Space Agency, has extended the many positive
attributes of ECR plasma-processing to the handling of large, flexible substrates such
as those required by the space industry.
ECR processing can be employed to tailor the surface characteristics of existing
space materials such as Kapton and Teflon. facilitating better optimization of the
design of space craft. Since coatings such as dense SiOz do not exhibit reliable
adhesion when directly applied to plastics due to interfacial stress, MPBT has
devoted considerable effort to develop a proprietary process that facilitates the
adhesion of high-quality ECR-CVD SiOz through covalent bonding to various
polymers and resins including Kapton, Polycarbonate, Teflon and even fiberglass in
an integrated thin-film structure. A broad range of thin-film structures can be
prepared by ECR-CVD to meet various space requirements such as AO-resistance,
thermal control and electrostatic protection. An overview is presented of ECR-CVD,
the relevant plasma diagnostics, and the preparation of large-scale ECR-CVD thin-
film structures for space applications. Experimental results are presented on the
wear-resistance, VUV AO-resistance, and thermal characteristics of ECR-CVD SiOz
thin-film structures on Teflon and Kapton.
125
1.1. Kleiman and R.C. Tennyson [eds.),
© 2001 Kluwer Academic Publishers .
126
1.0 Introduction
Many materials that are currently used in space structures, such as Kapton and
Teflon, suffer from degradation and erosion under exposure to atomic oxygen (AO)
and/or vacuum ultraviolet radiation (VUV) [1]. Since atomic oxygen is a
predominant species of the atmosphere at LEO altitudes, the resulting AO/VUV
degradation of these materials makes it difficult to design a space structure with a life
significantly longer than 5 years. Although materials such as Kapton and Teflon
degrade in a AO/VUV environment, they have several other desirable properties for
the thermal control and protection of space structures. Moreover, they have an
extensive history and characterization of use in various space environments.
Plasma processing is widely employed to improve the characteristics of existing
materials either through a modification of the surface structure and/or the deposition
of additional thin-film layers (2). ECR plasma discharges, with their characteristic
high activation of reactive precursors and low ion energies «30 eV), have facilitated
new deposition chemistries, as well as the low-temperature processing of more
delicate substrate materials (3). Rather than finding a new set of materials to match
those already in use with the added quality of AO resistance, MPBT has taken the
approach of modifying the surface of existing materials using ECR-based plasma
processing to improve their AO corrosion resistance. MPBT, with the assistance of
the Canadian Space Agency, has extended the many positive attributes of ECR
plasma-processing to the handling of large substrates such as those required by the
space industry (4).
The large-scale ECR facility at MPBT features an ultra high purity gas-handling
system, a proprietary substrate system that can accommodate subtrates up to five feet
in width, and a proprietary growth zone geometry that minimizes gas-phase
particulate formation. In situ diagnostics include a Langmuir Probe, thickness
monitor and an optical emission spectrometer (OES). Reproducible ECR processing
is accomplished using a recipe-driven process controller.
Figure I. The ECR chamber installed at MPBT

127
In the ECR technique , a perpendicular static magnetic field is employed to

confine the electrons and to induce helical gyrations of the electrons. At a particular
value of the magnetic field (875 gauss for 2.45 GHz microwave excitation) , the
frequency of the electron gyrations is synchronized to the electric field oscillations of
the microwaves, resulting in a resonant energy transfer from the microwaves to the
electrons. This results in maximum coupling of the applied microwave energy to the
electrons. The magnetic field also assists to confine the electrons near the plasma,
resulting in minimal loss through recombination at sidewalls, and correspondingly
high electron densities . This results in the generation of highly reactive plasmas that
are characterized by high ion densities that can exceed 20 ma/cm", yet relatively low
mean ion energies of about 10-20 eV.
The high electron densities that can be achieved by the ECR technique, and the
saddle-field de electric field configuration that can be induced by tailoring the
magnetic field profile, combine to facilitate much stronger excitation and ionization
than could be obtained using the microwave field alone. This facilitates plasma
chemistries that cannot be successfully implemented using MW PE-CVD; such as
ECR-CVD growth of a-SiN :H using SiIL and N2 , due to the strong activation of N/
that is provided by the ECR excitation processes [5]. In contrast, MW-PECVD
results in relatively poor activation of N/ . As a consequence, the MW PE-CVD
technique generally employs NH3 in place of N2 to prepare silicon nitride films,
resulting in a higher hydrogen content and poorer thermal stability of the resulting
MW-PECVD a-SiN :H films.
Electron cyclotron resonance CVD facilitates plasma-assisted processing
(etching or deposition) with high electron and ion densities at total pressures
substantially lower « 5 mTorr) than those afforded by glow discharge techniques (>
10 mTorr). The resulting lower collision rate in the ECR plasma and high ion
densities result in growth kinetics and chemistries for ECR-CVD that are
substantially different than those associated with glow discharge PE-CVD. This is
the basis for ECR-CVD to produce films that are quite different from those that are
produced by MW-PECVD or other glow discharge techniques.
In the preparation of ECR-Si0 2 , an Ar/Oz gas mixture is generally introduced
into the ECR zone at a total pressure between 10,4 and 5·10'3 Torr. The Si precursor,
such as SiIL , is introduced separately near the substrate holder. The ECR
magnetically-induced saddle-field configuration facilitates strong activation of O2+
ions as indicated in the corresponding optical emission spectra. By tailoring the
magnetic field gradient, the ions in the plasma can be extracted and directed onto the
substrate. Due to the low operating pressures, gas-phase interactions in the plasma
stream are minimal. This minimizes gas-phase polymerization and subsequent
particulate formation. facilitating films that are relatively free of any pinholes. In
contrast, appreciable gas-phase reactions can occur in conventional PE-CVD that
result in pinholes and other defects that can lead to sample failure under AO and
VUV exposure.
The chemiabsorbed species interact with the impinging ions in the plasma
stream. Film growth proceeds through heterogeneous reactions at the substrate
surface between the impinging stream of O2+ ions and chemiabsorbed Sill, radicals,
128
as shown in Figure 2. Three reaction paths for ECR-CVD of Si02 have been
identified [6]:
(a) excess O 2: 20 z+SilL=Si02(0H)+nH20
(b) balanced: 02+Si!-L=SiOz+2H2
(c) excess sm, Oz+2Si!-L=Si0 2:H+nH z
Weakly bonded radicals can be sputtered away by the impinging Ar ions, while
strongly adhering radicals are incorporated into the growing film through covalent
bonding.
E:CR-CVD (I OJ T orr)
t ~at t
Figure 2.
Schematic er s .o , growth by conventional PE-CVD [2] and by ECR-CVD
2.0 Experimental Results and Discussion
Plasma processing involves a complex interplay between the formation of reactive

precursors within the plasma, the corresponding plasma/surface interactions and the
resulting desired surface process . The development and control of a plasma process
involves the development of an understanding, or model, of these complex, multi-
variable interrelationships. lnsitu plasma diagnostics have been employed to
optimize the preparation of low-temperature Si02 for MPBT's large-scale ECR-CVD
system. During plasma processing, the presence of luminescent "activated" radicals
in the plasma was monitored using optical emission spectroscopy (OES) . Light
generated within the plasma was collected by a 0.2 mm core UVMS fibre and
coupled to an Oriel 77200 monochromater. The signal was detected . using a
photomultiplier.
Figure 3 shows the OES intensity versus wavelength as measured during a
typical ECR-CVD of SiOz for an ECR discharge ignited in an Ar/Oz/Si!-L gas
mixture at 400 Wand a chamber pressure of 3 mTorr. By a comparison to published
tables of atomic and molecular optical emission spectra [7], one can identify the
main peaks associated with energetic radicals in the plasma. This is somewhat
complicated by the overlap of emission lines due to the various radicals. There is
129
good activation of O2+ ions as indicated by strong emission peaks near 504 and 525
nm [7]. Both O2 (450 nm) and Ar (420. 434 nm) are also strongly activated . There
is a significant OH peak near 306 nm. It suggests that the oxide growth is proceeding
through the excess oxygen reaction path. Residual background water vapor in the
chamber could also contribute to this peak. The presence of the SiH2 peak near 620
nm suggests some dissociation of the SiH4 in the gas phase, with the SiH2 radicals
also contributing to film growth. The relative peak heights of the optical emission
lines depend on the plasma parameters and their effects on the relative formation of
energetic gas-phase radicals.
'i 'i~ 'iN

~ ~~ ~0 ~
O .S
~ lij ~.:o!
~ ~ - 6-r> ~
:x:
..- e» l§;- :. 0""
e-s
~ ~
~ ~- +~
~
0.4 'i t">
-...- ~:.
~
~
+
0 .3
0- ... r>
.$!:P <"">
U':I
U':I
~ 0.2
0
0 .1
~
lij
~
0
200 400 600
Figure 3
Optical emission spectrum of an Ee R discharge in an Ar/0 2/SiIL, gas mixture
at PCH=3 mTorr and PMW =400W
A 4 ft. wide roll of 2 mil Kapton and a 3 ft. wide roll of 5 mil type A Teflon
FEP were obtained from DuPont to examine the growth of Si02 on Kapton and
Teflon by ECR-CVD. Wiping the bare Teflon with a wetted tissue tended to scratch
its surface. As a result, a non-abrasive. wet cleaning procedure was employed for the
Teflon and Kapton prior to introduction into the vacuum system. Prior to film
deposition, the samples underwent a proprietary surface preparation procedure.
Three sets of Si02/InterfacefTeflon samples were prepared at different microwave
powers between 400 and 500 W with a thickness of about 1500 A. A subsequent set
of samples was prepared with a reduced thickness of 600 to 800 A. The substrates
were unheated during the depositions. The maximum substrate temperature due to
plasma induced heating was about 60°C.
The electrical breakdown strength of dielectrics is often employed as a measure
of their density and the presence of defects in their microstructure such as grain
boundaries that could contribute to premature electrical breakdown. The electrical
breakdown field was measured for ECR-Si0 2 deposited onto n-type crystalline silicon
130
using a sandwich configuration of AI dot contacts . Breakdown fields of about 9.5

MY/em were obtained for ECR-Si0 2 films prepared at microwave powers above
400W. These values are comparable to those observed for thermally grown Si02•
This indicates that the ECR-Si0 2 films prepared using the low-temperature ECR-
CVD process at MPBT have a very low density of pinholes and other microdefects
that could contribute to premature electrical breakdown.
The adhesion of stoichiometric ECR-CVD Si02 deposited directly onto various
plastics including Kapton was relatively poor. The films, although continuous,
tended to disengage from the Kapton when rubbed by a blunt tweezer, as shown in
fig. 4a. In subsequent process optimization, a proprietary ECR-based interfacial
plasma treatment was developed which substantially improves the adhesion of Si02
to various plastics including Kapton (see fig. 4b), Teflon and polycarbonate. The
resulting Si02 films , as shown in fig. 4b, exhibit excellent adhesion even after
mechanical wiping and rubbing.
(a)
Figure 4
Surface morphology of (a) SiO,lK apton and
(b) Si02/(Proprietary Interface j/Kapton.
after mechanical wiping .
Several samples of 1.5 urn thick Si02 deposited onto 2" 0 .0. polycarbonate were
subjected to bombardment by 50 um alumina pellets at a velocity of 100 mls. The
corresponding mass loss of the sample was measured as a function of the mass of
alumina consumed. A polycarbonate sample coated with a thick resin protective
coating (about 50 urn) was used as a reference . The resin is employed to protect the
polycarbonate against scratching. The measurements were performed by Liburdi
Engineering. The Si02 coating did not break up upon bambardment by the alumina
pellets but was gradually eroded . As shown in Figure 5, the relatively thin Si02 layer
initially offered significantly better protection to the underlying polycarbonate than
the resin. resulting in a lower material loss. The Si02 layer began to be eroded after
bombardment by about 12 gm of high-velocity alumina. A somewhat thicker layer of
ECR-CVD Si02 could provide good wear resistance for plastics, such as eyewear .
131
20
~15
.s
'"
10 +------------=,......,:::::...----:::::;::>m"'''--------j
~
.&
~ 5+------:,......,:~-__::",....,::::...---------------__I
O~~;:...___.---.--.______,r_-,--__._-.,...____,r_-r-__._-._______,r_---I
o 5 10 15 20 25 30 35
Mass of Sand Consumed [g]
___ Uncoated -mt- SiO Coa1:ed
Figure 5
Erosion of (1.5 um Si0 2) /Polycarbonate (D) and resin-coated Polycarbonat e (.)
in terms of material loss as a function of mass of 50 um alumina particles, at 100 mis, consummed
Figure 6a shows the surface morphology of the bare 5 mil type A Teflon FEP as
received from DuPont. Its surface morphology consisted of smooth islands or
plateaus surrounded by a network of cracks or ridges. The "as deposited" SiOz films
were clear and shiny. Fig. 6b shows the corresponding morphology of a 1500 A thick
SiOz coating on the Teflon (SiOn.310398). The crevices in the underlying Teflon
result in a nodular morphology, consisting of smooth islands over the Teflon plateaus
with continuous bridges over the ridges in the underlying Teflon. The SiOz film was
able to fully bridge the cracks in the Teflon to provide continuous surface coverage
with no significant pinholes . The 1500 to 2000 A thick SiOz films on the soft
underlying Teflon tended to crack when mechanically stressed using a scalpel, as
shown in fig. 7. The microstructure observed in the type A Teflon FEP substrate
may be a contributing factor to the cracking observed for the 1500 A SiOz coatings,
since the SiOz would tend to break in the crevice regions in the underlying Teflon
substrate where there is very high stress on it and no support by the substrate.
A second set of SiOz/lnterfacelTeflon samples was prepared with a thickness of
about 500 to 800 A (SiOn.150498). The incidence of cracking was significantly
reduced for the thinner Si02 coatings. The results indicate that the maximum usable
thickness of the SiOz on Teflon is less than that on Kapton, due to the morphology
and structure of the Teflon. The thinner SiOz films were more flexible and still
provided good coverage of the Teflon.
Since the completion of this work, an additional sample of DuPont's Teflon FEP
was obtained via Sharon Rutledge of NASA's materials group. SEM observation of
this sample indicates a much smoother surface morphology than the Teflon FEP
originally obtained from DuPont. This should encourage a much smoother film
growth, avoiding the nodular film morphology that was observed with the initial
Teflon FEP. Evidently, there is some variation in the surface morphology of different
132
grades of Teflon FEP . This requires further study to determine the best grade for
space applications.
(0) (b)
Figure 6
Morphology of 5 mil thick bare Typ~ A T~f1o n FEP (a) and 1500 A thick
Si011( I ntert:~ce )!Tefl on FEP (b). as deposited at 450 W by ECR-CVD
Figure 7
Magnified view ofa crack in a 1500 A thick Si0 2/(Interface) film
"as deposited" onto 5 mil type A Teflon FEP at 450 W by ECR-C VD
Solar absorptance and thermal emitancc measurements on MPBT's ECR-CVD

Si02 thin films deposited onto 2 mil Kapton were provided by Marc Van Eesbeck
from ESTEC. The results are summarized in Table I. The thermo-optic
characteristics of the coated samples are almost identical to those of the underlying
Kapton.
133
Table I. Thermo-optic characteristics ofECR-CVD Si02/Kapton
Sample Solar Thermal

Absorptanee Emittance
BARE KAPTON 0.319-0.336 0.646
Si01 .241095 0.325 0.639
Si02.251095 0.331 0.627
AO testing was performed at CSA by exposing the ECR-CVD Si02-protected

and bare samples in the afterglow area of a RF plasma in oxygen. The set-up of the
facility allows the prevention of the exposure to the UV/VUV emitted by the plasma.
Samples were tested one at a time for a duration of I hour each . A piece of Kapton R
was used as a reference . Each sample had an area that was masked from exposure to
the plasma using a Kapton mask. The oxygen plasma operating parameters arc given
in the following table.
Table 2. Oxygen plasma operating conditions for AO testing at CSA
Pressure O2 Flow Rate RF Power Duration

(Torr) (seem) (W) (hour)
0.1 10 300 1
A Dektak 3030 ST Surface Texture Analysis System was used to measure the
thickness removed from the samples following the plasma treatment. The technique
consisted in the measurement of any step created at the exposed/unexposed area
intersection after the AO exposure . The results are summarized in Table 3.
Table 3 : Summary of AO tests at CSA
Sample Erosion Depth
Bare KAPTON 10.8um

Si01 .141195/KAPTON <0.01 um
Si02.141195/KAPTON <0.01 urn
Si03.141195/KAPTON <0.01 um
Si02.210695/KAPTON <0.01 um
Bare Aluminized Teflon 241025 urn

Si02.141195/Aluminized Teflon <0.01 urn
Bare 5 mil Teflon FEP 15 urn

Si01310398/Tefion FEP <0.01 urn
Si02.310398/Tefion FEP <0.01IJm
Si03.310398/T eflon FEP <0.01 urn
134
(a) (b)
Figure 8
SEM micrographsof (a) (2000 A Si0 2) /(proprietary interface)/Kaptonfilm
"as deposited" onto Kaptonat 350 W by ECR-CVDafter exposureto CSA's VUVatomic oxygensource,
and (b) bare Kaptonafter a similar AO exposure
Figure 8a shows the SEM micrograph of the morphology of a 2000 A thick

Si02/Interface/Kapton sample prepared by ECR-CVD at 350 W after AO exposure at
CSA. The film remained smooth after the AO exposure with no resolvable change in
morphology or in film thickness . Fig 8b shows the corresponding morphology of bare
Kapton after a similar AO exposure. The bare Kapton was significantly eroded and
pitted by the AO exposure to a depth of about II um.
(a) (b)
; '1 .. •• .• .--:.., · -i : . . .... ~ •

e.. ..••,
y ..
...." •
t ••• " •• ,
• t .' ... . 1 .- .- .•
••••, . / . '
. ... ' ... . . .. . . . J . ' ~ .; . -. II. . ....
f'\ .-:.'" '.•~I ~ ~ . .~. ••••• • . ; . ~. ~

\ .• -~ -. • . - ••!a • • ' ~-7 I •• • lit •
..... . '1._.1< ... " ..... . -• •t
e.'
~.". . ~. ~ ',."1·, •. t .~ . . «...---,
,"' '.'.' •
.
..... l .-""
• 11k . -
.. ..• ~i
\ ,.
. ~" .'
• ,, ' .. ..
.
.....--. ~
t '- .•
;! ..:.t'
' • . . .'..··...
c.: .... .,. . '1'
''~.• ' ...
~~ •....e.!-.J .. ~
• ••
~
:.':J1'
.
'• • ~. lit ,
# - :-
i!' ~ .- . .'.' .. : .
..' ..-...
.~
e. ~
-. .
~..;· ~.·•· - •·-l··~.C ·:·'.~
. '••,.•.. lIPt· .. •• ~ · •• .,
·~·'-~·~ ' .~
~ e ~ •
II~(.
Figure ')
SEM micrographsof (a) (1500 A thick Si0 2)/(proprietaryinterface) film on 5 mil Teflon FEP
as deposited at 450 W by ECR-CVD after exposure to CSA's VUVatomic oxygensource.
and (b) bare Teflon FEP after similar exposure to CSA's AO source.
Figure 9a shows the surface morphology of a (1500 A Si02)/Interface/ Teflon

sample after AO exposure at CSA. The morphology is similar to that of the "as
deposited" films. In some of the tested samples, some additional cracks were formed
after the AO exposure, possibly due to the associated thermal stress. The AO
exposure also encouraged the formation of a moderate concentration of round
135
nodules, possibly due to the local formation of pockets of gas between the coating
and the Teflon. This depended on the SiOz film preparation conditions . No step in
the film thickness could be resolved between masked and exposed sections of the
tested samples. In contrast, as shown in Figure 9b and in Table 3, there was
considerable erosion of the unprotected bare Teflon FEP after a similar AO exposure
at CSA.
4.0 Conclusions
ECR processing can be employed to tailor the surface characteristics of existing

space materials such as Kapton and Teflon, facilitating better optimization of the
design of space craft. Since coatings such as dense SiOz do not exhibit good, reliable
adhesion when directly applied to plastics due to interfacial stress; MPBT has
devoted considerable effort to develop a proprietary process that facilitates the
adhesion of high-quality SiOz to various polymers and resins including Kapton,
Polycarbonate, Teflon and even fiberglass in an integrated thin-film structure. A
broad range of coatings to meet various requirements such as AO-resistance, thermal
control and electrostatic protection, can be prepared by ECR-CVD through alloying,
using appropriate gas mixtures, or by preparing stacked layers of different thin-film
materials .
Teflon FEP, especially aluminized samples, can exhibit significant erosion under
VUV and AO exposure that can surpass the erosion rate of Kapton. The MPBT ECR-
CVD SiOz/Interface coating was able to provide significant improvement in the
Teflon AO resistance. even with the use of relatively thin 600 to 800 A coatings.
with no AO erosion resolvable over protected regions. While some additional testing
and optimization is required, the ECR-CVD SiOz/InterfacefTeflon samples indicate a
good potential for future space applications .
MPBT has developed a high-quality SiOz coating by ECR-CVD at room
temperature using the large-scale ECR facility at .MPBT. Measured specifications for
the coating include:
IThickness Uniformity >95%

Refractive Index(632.8 nm) 1.45
Bonded Hvdroaen Content < 1 at. %
Electrical Breakdown Field 9.5·10"V/cm
~i·O bondstretchina vibrational mode 1076em'
rvlS OpticalTransmittance >92%
200 to 1000 nm) .-
SolarAbsorptance (on Kapton) 0.;33
Jhermal Emittance (on Kapton) 0.635
BHF Etch Rate <30Als
1'.0 and VUV resistance No resolvable etch step using
on Kaoton and Teflon) CSA's AONUV source
136
Acknowledgements:
This workwas partiallyfundedby the Canadian Space Agency.
5.0 References
I. R.C. Tennyson and W.D. Morison, Proceedings TMS Symposium on "Space Environmental effects on Materials",
Anaheim, California, Feb., (1990).
2. D.L. Smith and AS. Alimanda, J. Electrochem Soc. , I, 14%, (1993).
3. K. Shirai and S. Gonda, J. Appl. Phys., 67, 6281, (1990).
4. R.V. Kruzelecky, J. Blezius and AK. Ghosh, Proceedings of the 3-rd International Conference on "Protection of
Materials and Structures from the LEO Space Environment," Toronto, Canada (J.I. Kleiman and R.C. Tennyson, eds.)
Kluwer Academic Publishers, 155-165, (1999).
5. Y. Manabe and T. Mitsuyu, J. Appl. Phys., 66, 2475, (1989).
6. P. Shufflebotham, M. Weise, D. Pirkle and D. Denison, "Biased Electron Cyclotron Resonance Chemical-Vapor
Deposition of Silicon Dioxide Inter-Metal Dielectric Thin Films," in "Plasma Properties, Deposition and Etching," eds.
J.J. Pouch and SA Alterovitz, Materials Science Forum, vol. 140-142, 255-268, (1993).
7. R.W.S. Pearce and AG. Gaydon, "The Identification ofMolecuJar Spectra," Chapman&Hall, London, 241-245, (1985).
NON-SILICONE INORGANIC POLYMER-BASED COATINGS FOR
ATOMIC OXYGEN PROTECTION IN LEO - PART 2
E. POIRE
Canadian SpaceAgency,
Space Technologies, Applied Thermal & Materials Group,
6767 route de l'Aeroport,
Sf-Hubert, Quebec,CanadaJ3Y 8 Y9
B. G. SELLARS
ORTECHCorporation,
2395 Speakman Drive,
Mississauga, Ontario, Canada L5K IB3
D.NGUYEN
Datco Technology Ltd.,
3375 Laird Drive, Unit 3,
Mississauga, Ontario, Canada L5L 5R7
Abstract
Conventional dimethylsiloxane polymers (silicones) have a high erosion rate in

atomic oxygen which makes them unsuitable for long term exposure in the Low
Earth Orbit (LEO) environment. Advanced silicon-oxygen backbone polymers can
exhibit much lower atomic oxygen with vacuum ultraviolet (AO/VUV) erosion rates.
However, there is currently a lack of space-qualified polymer coating systems
available that can withstand the requirements of 30-35 years exposure in LEO .
There is a need for flexible coatings which are stable in the LEO environment.
This paper is a continuation of work presented elsewhere by the authors [1]. It
describes the performance of proprietary non-dimethylsiloxane inorganic polymers
as a coating system resistant to AO/VUV. The protective layer is formed when the
coated surface is exposed to AO. These coatings may be suitable for the protection of
organic materials and composites in LEO. The inorganic polymer is soluble in
common organic solvents and can be applied onto a variety of substrates by
inexpensive spinning, spraying, or dip coating techniques. Kapton'' polyimide is the
primary substrate used in this work, although other materials such as graphite epoxy
composite can also be coated. Performance data of the proprietary coating are
presented including the erosion rate in terrestrial AOIVUV tests and surface
morphology by AFM before and after AOIVUV exposure.
137
J.l. Kleiman and R.C. Tennyson (eds.),
Protection ofSpace Materialsfrom the SpaceEnvironment, 137-143.
© 2001 KluwerAcademicPublishers. .
138
1.0 Introduction
Datco Technology Ltd. and Ortech Corp. developed an advanced atomic oxygen!
vacuum ultra-violet (AO/VUV) resistant coating based on inorganic polymers in a
Canadian Space Agency- (CSA) sponsored STEAR 6 program (1991-2) (Strategic
Technologies for Automation and Robotics) [2]. Unfortunately, the manufacture of
one of the key chemicals was discontinued. The other major component was a
polysilsesquioxane resin and was still manufactured. However, it was felt that the
resin was somewhat 'brittle ' and a more flexible resin system would be required for
applications such as AO protective coatings for wires and wiring harness as well as
having potential as the base resin system for coloured paints (for wire and other
component identification).
Alternate resins were identified and the Datco/Ortech-type coating was refor-
mulated using solutions of polysilsesquioxane and these resins (not mentioned for
reasons of confidentiality). This paper presents performance regarding the resistance
to AO/VUV attack of a set of formulations developed with two alternate resins and a
series of five polysilsesquioxane resins.
2.0 Experimental
2.1 SUBSTRATE AND COATINGS MATERIALS
Flat coupons (5 em x 5 em) of 25 um (I mil) thick Kapton H polyimide were used

for this work. Coating solutions were formulated with selected polysilsesquio-
xane(s) plus a resin "X" and toluene solvent. Twenty wt per cent resins solution
concentrations of 30wt per cent resin "X"170wt per cent silsesquioxane ratios were
prepared for samples 2 through 6. 20wt per cent resin "X"/80wt per cent
silsesquioxane for sample 1. Table I shows the formulations for each coating.
Table 1. Coating formulations for samples I to 6
Resin 'X' Silsesquioxane
A B A B C D E
I * *
2 * *
3 * *
4 • * *
5 * * *
6 * *
139
2.2 COATING METHOD
Coating solution of 20 per cent resin/80 per cent toluene was used since gelation
trials indicated long duration (mixed) solution stability for this ratio. Dissolution and
other preparation steps for these solutions were done in an oxygen-free argon
atmosphere glove box. This precaution was undertaken to avoid hydrolysis of
reagent chemicals after opening of the original container. After the coupons were
cleaned with isopropanol and spin-dried, they were spin-coated with one spray on the
coating solution using a fine mist atomiser, then dried for 30 minutes in a circulating
oven at 100°C (repeated for 3 coats). Coated coupons were cured in a vacuum oven
at a temperature of 200°C for 20 hours. No adhesion promoter was used before
coating. The fine mist atomiser was used to spray coupons at a distance of -3 em
from the spinning coupon (speed = 3500 rpm). This method proved effective at
completely covering the coupon and reducing the ' pinwheel' appearance of coated
samples.
2.3 PLASMA TREATMENT / AO TESTING
Resistance to AO/VUV was tested using facilities at Ecole Polytechnique and at

CSA. Two sets of plasma treatment were performed at Ecole Polytechnique de
Montreal with a reactor described at Ref. 3. TIle set-up and procedure were similar to
the one used in past AOIVUV resistance testing [3-5]. Samples were exposed to a
plasma of oxygen generated by microwave (MW) and a combination of microwave
and radio-frequency (MWIRF) called ' dual-frequency' mode. In these treatments,
samples were in contact with all the species created in the plasma including
UVIVUV. All samples were tested in the same run for a duration of 1 hour. A piece
of Kapton was used as reference and each coated sample had an area not exposed to
the plasma.
The AO testing at CSA was performed by exposing the samples in the afterglow
area of an RF plasma of oxygen. The set-up of the facility allows prevention of
exposure to the UV/VUV emitted by the plasma. Samples were tested one at a time
for a duration of I hour with a piece of Kapton" as reference. Each coated sample
had an area not exposed to the plasma. The operating parameters are shown in Table
2 for the three series of AO testing.
Table 2. Plasma Etch ing Conditions
Plasma Pressure O 2 Flow MW Bias Voltage RF Power Duration

(Torr) (seem ) Power (Vb) (W) (hour)
(W)
MW 0.2 30 200 0 0 I
MW/RF 0.1 20 200 -150 0 I
RF 0.2 10 0 0 300 1
(SSA)
140
For the treatmentperformed at CSA, the AOfluence has been estimated to be F

= 5 X 1020 oxygenatoms/crrr' taking (3):
F = 10-4 hIR [atoms/em'] (1)
where h is the change in thickness in urn (step) and R = 3.0 X 10-24 cm3/o xygen atom,
the etching rate for polyimide Kapton H [6].
2.4 PROFILOMETERY
A Dektak 3030 ST Surface Texture Analysis System was used to measure the
thickness removed from the samples following the plasma treatment. The technique
consisted in measuring the step created at the exposed/un-exposed area intersection
when etching was actually taking place.
2.5 ATOMIC FORCE MICROSCOPY (AFM)
AFM measurements were performedwith a Topometrix TMX 2010 in contact mode

with a 75 urn piezoelectric and a 1 urn by 1 urn pyramidal tip. The RMS roughness
value was the parameter chosen in this work to monitor the change in topography.
RMS roughnessvalue is defined as follow:
RMS = I
/
2
P(j}ij
]X (2)
[
where
(3)
is the Power SpectralDensity in ( 1O-4 llm)2' llm

z(x): line profile in lO-4 llm
L: scan length in urn
dx, x: um
f: urn"
3.1 PROFILOMETERY
The data measured with the Dektak profilometer are presented in Table 3 for the
three series of treatment. Uncoated Kapton was taken as a reference to monitor the
level of etching. The presence of a step is an indication of a loss in mass. which
141
indicates the vulnerability of the material to AO or AOIVUV. '0' values in the table
indicate that no step was detected using this method. The results for the MW plasma
could not be conclusive if taken alone, the etching rate for Kapton being small
(Ium/hour) and close to ' 0' . The dual frequency mode (MWIRF) and the RF
(afterglow) treatment show etching rates of 18 and 15 urn/hour for Kapton
respectively. These treatments are very strong, and as for MW run, show no apparent
loss in mass for the coated samples.
Table 3. Thickness of material rem oved following exposure to plasma of oxygen /atomic
oxygen
Sample Thickness ofthe step (um)
MW Plasma MW/ RF Plasma RF(CSA)
uncoated I 18 15
I 0 - -
2 0 0 0
3 - 0 0
4 0 0 0
5 0 0 0
6 - 0 0
In MW plasma, the active species impinge upon the sample surface at near-
thermal energy. MWIRF plasma, in addition, allows the simulation of particles
impacting on the surface with elevated kinetic energy. Here, the difference in
' nature' between MW and MWIRF could not account for the small etching rate of
uncoated Kapton for MW, since greater etching rates were reported (3). It rather
seems that the parameters of operation were not properly selected. Nevertheless,
results from Table 3 suggest that these coatings may be good candidates for space
applications in LEO regarding AO or AOIVUV resistance.
3.2 AFM AND XPS RESULTS
A great effort was invested in trying to measure a step by AFM for samples showing
no step with the Dektak profilometer. AFM could not be used to fulfil this task.
Even a 1 11m step (Dektak) could not be measured with AFM, the slope of the step
being much wider than the horizontal scan capacity of AFM (maximum of 75 J1Il1).
Table 4 presents the roughness values measured by AFM for the MW plasma
142
treatment. Roughness was computed with an area of 50 um by 50 11m except for

sample 2, for which 40 11m by 40 11m was used. For this reason, sample 2 RMS
rouglmess values can not be compared to the other sample values, since its
computation depend on the area.
Table 4 shows that roughness decreases after the exposure to AO except for
sample I and uncoated Kapton. The behavior for each coating could be associated
with the formation of the SiOx protective Jayer. Results for sample I could be
explained by the fact that it is expected to be a little hard and brittle, and the
"slowest" SiOx layer to form. The other coated samples seemed "smoother." It was
also suspected that the oxygen plasma had the effect of "cleaning" the surface from
contamination. These results suggest that coatings remain smooth once modified by
AO.
Table 4. Roughness measured by AFM
Area RMS (nm)

Sample MW Plasma
un-exposed area exposed area
uncoated 64 81
1 67 1271
2 1,9 1,2
4 87 55
5 130 63
X-ray Photoelectron Spectroscopy (XPS) was also performed on coated samples

before and after AO testings. XPS has shown that coatings were chemically modified
when exposed to AO. An SiOx protective layer tended to form in the presence of AO.
The thickness of the modified layer has not been determined.
4.0 Summary and Conclusion
New non-silicone inorganic polymers-based formulations for AOIVUV-resistant

coatings were tested using facilities at Ecole Polytechnique and the Canadian Space
Agency. Profilometry measurements have revealed no apparent loss of mass for
coated Kapton after exposure to AO/VUV, in conditions of treatment for which
Kapton polyimide loses up to 15 to 18 11m in thickness. XPS has suggested the
formation of the Sif), protective coating after exposure to AO. AFM has shown that
protective coatings were smoother and glassylike compared to coatings not exposed
to AO.
143
The authors wish to thank Mrs. Suzie Poulin for the AFM and XPS measurements
and Prof. Michael R. Wertheimer for the plasma reactor facility (Department of
Engineering Physics, Ecole Polytechniquede Montreal).
6.0 References
1. Poire, E., Sellars, B.G. and Nguyen, D. Non-Silicone Inorganic Polymers Based Coatings for
Atomic Oxygen Protection in LEO, Proceedings of the 43'd International SAMPE Symposium &
Exhibition, Anaheim, CA, pp, 1534-1538, (1998).
2. "Development of an Advanced Coating System for Materials in Space -Phase I. Volume I. Final
Report-Phase I" Contract No. 9F006-1-0024/0I-SW by Datco, ORTECH and ITL, May 29, 1992.
3. Wertheimer, M.R, Czeremuszkin, G., Cerny, J., Klemberg-Sapieha, J.E., Martinu, L. and Kremers,
W., Plasma-Deposited Coatings and Surface Fluorination for Protection of Spacecraft Materials
against Atomic Oxygen Erosion, Proceedings of the 7th International Symposium on "Materials in
Space Environment, Toulouse, France, pp. 393-402, (1997).
4. Lamontagne, B., Wrobel, A.M., Jalbert, G. and Wertheimer, M.R., Large-area Microwave Plasma
Etching of Polyimide, 1. Phys. D: Appl. Phys. 20, pp. 844-850, (1987).
5. Lamontagne, B., Kuttel, O.M. and Wertheimer, M.R., Etching of Polymers in MicrowavelRadio-
frequency O,-CF, Plasma, Can. J. Phys. 69, pp. 202-206, (1990).
6. Banks, BA and Rutledge, S.K., Low Earth Orbital Atomic Oxygen Simulation for Materials
Durability Evaluation. Proc. 4th Int. Symp. on Spacecraft Materials in Space Environment,
Toulouse, France, pp. 371-389, (1988).
RESEARCH ASPECTS OF SCALING-UP IMPLANTOX TECHNOLOGY
FOR PROTECTION OF POLYMERS IN SPACE BY ION IMPLANTATION
Z. A. ISKANDEROVA, J. 1. KLEIMAN, Y. GUDIMENKO

Integrity Testing Laboratory Inc. ,
University ofToronto Institute for Aerospace Studies,
4925 Dufferin Street, North York, Ontario, M3H 5T6, Canada
R. C. TENNYSON
University ofToronto Institute for Aerospace Studies,
4925 DufJerin Street, North York, Ontario, M3H 5T6, Canada
1. G. BROWN, O. R. MONTEIRO
Lawrence Berkeley National Laboratory
Berkeley, California, 94720, USA
B. A. BANKS, M.J. NORRIS,

NA SA Lewis Research Center ,
Cleveland, Ohio, '44135, USA
1.0 Abstract
High-dose Plasma Immersion Ion Implantation, that combines both low-energy ion
and recoil implantation (dynamic ion mixing) and can be performed on flat
materials or intricate three-dimensional parts, was used to enrich the thin surface
layers of high-performance space-related polymers with an appropriate amount of
AI, one of the reactive metal elements. The retained dose and distribution of the
implanted AI and intermixed oxygen was controlled by RBS. Oxygen plasma asher
and fast (E-2-3 eV) atomic oxygen (FAD) beam, that are often used for accelerated
ground-based testing as imitating low Earth orbit (LEO) environmental conditions,
have been used as reactive oxidizing environments for surface conversion and
testing of the implanted polymers. The surface modified materials have been shown
to be resistant to both testing environments, including FAO, which is the main
danger for polymers and other carbon-based materials in space, in LEO. The
content, structure and morphology of the modified surface layers were examined by
XPS and scanning electron microscopy (SEM). It was shown that protective
oxide(s)-based surface structures were formed in the top surface layers of the treated
polymers. Implantation and surface conversion conditions were found, where the
appearance and important thermal-optical properties of treated polymer films, such
145
J.l. Kleimanand R.C. Tennyson (eds.),
146
as solar absorptance and thermal emittance, have been left practically unchanged, or
altered in a specific manner.
2.0 Introduction
Surface modification technology , called Implantox, that is based on high-dose ion

implantation of specially selected metal or semiconductor elements and following
oxidative post-treatment for surface conversion was proposed recently as a new
concept of protection of polymers and carbon-based composite materials in low
Earth orbit [1-3].
Based upon accumulated data, Implantox surface modification, using
conventional ion implantation technique, shows good results for the protection of
polymeric materials for space and some terrestrial applications, where improved
oxidation resistance is required [1-5]. It was continned recently through
experiments that by carefully choosing the elements being implanted and the
implantation conditions, high-current monoenergetic ion beams [1,2,4] and, to
some extent, metal vacuum vapor arc (MEWA) ion sources [3,5] can be used for
these purposes.
However, it takes a significant amount of time, in the hour time scale, to
obtain the required high implantation fluence, that is in the range of (5-7)x10 16cm-Z,
using conventional mid-energy monoenergetic ion beams [1,2,5]. To find an
approach of reducing the implantation time by appropriate choice of advanced
implantation techniques appears very attractive, as well as the idea of being capable
to treat three-dimensional parts, if needed. The second stage of Implantox, surface
conversion to highly protective oxide(s)-based surface structures has been originally
achieved by hours of exposure to unique and low treatment area fast (E-2-3 eV)
atomic oxygen (FAO) beam [1-5]. It also appears very attractive to perform this
treatment in a more conventional scaled-up oxygen plasma facility, and with a
shorter post-treatment time, if possible. Results of the first attempt to use metal
Plasma Immersion Ion Implantation (PIlI) [6-10) to improve oxidation and erosion
resistance of high-performance space-related polymers in highly aggressive
environments, that contain atomic oxygen, are presented here. It is shown, that this
advanced surface modification treatment can be used successfully, with the results
of surface properties modification being sensitive to the treatment conditions and
duration.
3.0 Experimental
3.1 :METAL PLASMA IMMERSION ION IMPLANTATION
Surface modification using metal ion species can be carried out by two closely
related techniques - conventional, line-of-site , high-energy ion implantation using
metal ion beams, or plasma immersion ion implantation using metal plasma, among
147
other approaches . Vacuum-are-produced metal plasma can be used as the ion

feedstock material in an ion source for doing conventional metal ion implantation,
or as the immersing plasma for doing Plasma Immersion Ion Implantation [6,7].
Plasma Immersion Ion Implantation, that can be performed at low energies
and used on flat or intricate 3-dimensional parts, combines both ion implantation
and recoil implantation (dynamic ion mixing) [6-10). In this study, PIlI was used to
enrich the thin surface layers of advanced thermally stable space-related polymer
films with an appropriate amount of a specially selected metal elements, and the
results of PIlI with AI are presented.
Metal plasma immersion surface processing is an emerging tool, whose
characteristics and capabilities are attractive for many applications [6-10]. It is
more suited, but not limited, to higher doses (>I0 17cm-2) and lower energies (E, <
tens of ke V), than the usual ranges of conventional metal ion implantation [6]. Due
to the lower energies, carbonization effects, that are common at ion implantation in
polymers [11-15], can be expected to be significantly reduced. This may reduce the
time for the second stage of Implantox, i.e the post-treatment time for oxidative
surface conversion [2-4]. Carbonization of ion implanted polymers is causing strong
increase in surface conductivity [11-14], and may cause an improvement of surface
mechanical properties [15,16]. However, carbonization is also causing darkening of
implanted polymeric materials, especially at higher energies and doses [1-5,11-13],
that is undesirable in many cases, where unchanged optical or thermal-optical
properties are preferred. It was assumed, that at the proper selection of metals to be
implanted and conditions of PIlI, due to low-energy processing, the appearance of
the treated polymers, as well as solar absorptance a. and thermal emittance s may be
left practically unchanged. This is difficult to achieve at the high doses, that are
required for Implantox treatment, using conventional implantation techniques .
Therefore, higher doses and lower energies are two attractive features for what can
be called PIlI-based Implantox surface modification, minimizing the time and effort
at both implantation and conversion stages. Both those issues are very important for
the potential scale-up of Implantox technology.
Aluminum PIlI was tried first, since AI was expected to interact with
polymers and oxygen at implantation and following oxidative post-treatment, and to
form transparent, colorless oxides. The results are presented for three advanced
space-related high-performance polymer films: Kapton HN (polyimide), Kapton CB
(graphite-filled black Kapton) and Mylar (polyethylene terephtalate) . All of them
are important space-related materials , often used in various space applications ,
mostly in spacecraft thermal control .
The implantation technique used in this work, that consists of using a
metal plasma generated by a vacuum are, and immersing the biased substrate into it
(Figure 1), has been described elsewhere [6-9]. The vacuum arc is a high current
discharge between two electrodes in vacuum in which metal plasma is produced in
abundance. In these studies, small, repetitively-pulsed vacuum arc plasma guns
were used that operated at a pulse duration of about 5 ms, repetition rate about 1 Hz,
and arc current in the range 100-300A. Along with the generated metal plasma, a
flux of macroscopic droplets ranging from 0.1-10 microns in diameter is also
148
produced. It is mostly desirable to remove the solid particulate contamination, and

this was done using a curved "magnetic duct" which stops line-of-sight
transmission of macroparticies while allowing the transport of plasma (Figure 1).
The ions in the plasma exit the filter and are accelerated towards the substrate by
applying a pulsed negative voltage to the substrate (Figure 1).
Substrate Plasma
duct
Controlled
pulse
bias
Plasma
gun
Controllert--~-~ Arc
supply
Figure 1
Schematic drawingof plasmaimmersion ion implantation set-up [7]
149
The bias voltage used in this work for the AI implantation was V=-2 keY, which
corresponds to mean aluminum ion energy of 3.4 keV [8]. At this energy, ions may
be expected to implant into the substrate to a depth of up to (10-20) nm (depending
on the substrate material). Due to the pulsed nature of the applied voltage, the
immersion implantation results from a combination of direct implantation and
recoil implantation (dynamic ion mixing), often yielding a non-Gaussian (flatter)
profile [7].
The arc source used in this implantation treatment was deliberately chosen
to provide a comparatively low implantation time. To prevent overheating of the
polymer films, they were mounted on a water-cooled substrate, and the arc
frequency was maintained at I Hz. Implantation at faster rates can be easily
accomplished by increasing the arc pulse frequency, increasing the pulse duration
(up to an eventual DC operation) or both. The background pressure was 2x10·6 Torr,
and implantation was carried out at ~5xlO·6Torr.
It will be shown later (see Results and Discussion), that depending on the
surface properties to be modified, the surface treatment by PIlI can be performed at
different conditions and/or durations, so that the process can go with or practically
without ion-assisted surface metallization.
For instance, the results of PIlI of the same metal, AI, in the polymers,
mentioned above, at similar conditions are very sensitive to the duration of
treatment (number of pulses). Protection from severe oxidative environments
practically without changing the appearance, i.e. the color and transparency, as well
as other important thermal-optical properties of the polymers (see Results and
Discussion) has been achieved at the mentioned above conditions, with bias voltage
"time on" equal to 2J.ls, "time of' equal to 6f.lS, and with 600 arc pulses. These are
very important features for thermal control polymeric materials for space
applications, and can also be highly preferential for optical and other polymers in
many terrestrial applications in severe oxidizing environments. Thus, we will call
these treatment conditions "preferential." It is worth to mention that in this case
PIlI treatment took only 10 minutes, contrary to implantation with similar fluences,
using conventional ion beam facilities, that takes normally a few hours.
The results of surface analysis, as well as atomic oxygen exposure and
testing are presented here mostly for polymer films, treated at these conditions .
Some results are also presented for the same materials, treated for longer duration .
For instance, increasing the number of arc pulses by ~(20-25) per cent, with all
other conditions being the same, the appearance, as well as some thermal and
optical properties of the polymer films were drastically changed. Kapton HN and
Mylar films became significantly darker, with reduced light transmittance in the
visible, and with increased reflectance (i.e., reduced thermal emittance) in the
infrared for the treated side. Bright lustrous metallic shade was also obvious for the
treated size of both polymers. The treated surface of Kapton CB attained gray matte
appearance (see Results and Discussions). AIl of these changes can be an indication
of both graphitization and, preferably, ion-assisted metallization. which was
150
confirmed by Ratherford Backscattering Spectroscopy (RES) (see also Results and

Discussions).
3.2. SURFACE ANALYSIS AND CHARACTERIZATION, ATOMIC OXYGEN

TESTING
The retained doses of AI and intermixed oxygen, both elements depth distribution,
and surface composition of the polymer films, treated at mentioned above
"preferential" and "high-dose" conditions have been studied by Ratherford
Backscatter Spectroscopy (RBS) [17], using He+ ions at 1.95 MeV. Surface
composition was also analyzed by X-ray photoelectron spectroscopy (XPS),
described in [1,2].
XPS analysis was performed after PIlI as well as after exposure and testing
for different duration in an oxygen plasma asher [18], and mostly in Fast Atomic
Oxygen (FAO) Beam Facility, or Space Simulator at UTIAS, described in [1,2,4).
That is actually a facility for ground-based testing, and can provide a flux of
hyperthennal (fast) (E=2-3 eV) atomic oxygen as high as lxlO 16at • em? • S'I .
Detailed description of the Space Simulator and testing procedure is presented in
[19); for these studies 5 and 8 hours ofFAO exposure was maintained. For oxygen
plasma testing, the samples were placed in SPI Plasma Prep II and in the radio-
frequency (RF) plasma asher. The plasma asher was operated at 13.56 MHz. The
operating gas was air and the pressure was approximately 130 milliTorr.
The samples were exposed for -5-7.5 hours, and the atomic oxygen fluence
was calculated based on a Kapton H effective fluence measurements. Effective
fluence measurements were made using mass loss of vacuum dehydrated Kapton H
polyimide witness samples. The results of testing of the polymer films in an oxygen
plasma asher are often used to imitate long-duration exposure in LEO [20).
Scanning Electron Microscopy (SEM) (see [1,2) for description) was used
to evaluate the possible changes in surface morphology after implantation and after
FAO testing. Thermal emittance, solar absorptance, and transmittance in the
visible was measured for the treated polymer films, and after atomic oxygen
testing.
The thermal emittance (s) of the samples was measured in accordance with
the ASTM specification E-408, using an infrared reflectometer manufactured by
Gier Dunkle Instruments, Model DB 100. The solar absorptance of the samples (a)
was measured using a spectrophotometer of Beckman Instruments, Model DK-2a,
equipped with an integrated sphere. Transmittance of the samples was measured
both before and after oxygen plasma exposure, illuminating the samples with light
from a 670nm laser. Surface sensitivity to the PIlI treatment, exposure to oxidative
environments and post-treatment storage was also explored by water contact angle
measurement, similar to those studies performed after conventional mid-energy ion
implantation, using the procedure, described in [21).
151
3.3 COMPUTER SIMULATION
Since direct implantation of ions in various charged states, as well as absorption of

some metal from plasma and' reactive gas, such as oxygen, from the vacuum
background, as well as recoil implantation are going on together at low-energy
PIlI, the final results might be very sensitive to the treatment conditions and
duration. So, they can not be expected to be predicted easily by conventional
computer simulation. However, one of the important factors, ion charge state
distribution, that can significantly vary for different metals [8] and will influence
the retained dose and depth distribution of implanted elements, was taken into
account.
It was shown in [8], that for PIlI with AI the beam composition, i.e. ion
charge states distribution , can be expressed as
That was taken into account in computer simulation, when Transition of Ions in
Metter (TRIM) [22] computer simulation code has been used, that allows computer
simulation for many polymers, including Kapton and Mylar.
0.011 Combined - 38% AI' + 51% Ai+' + 11% Ai+"

• 0 AI' (2keV)
0.010 At.. (4 keV)
"
a . AI'" (6keV)
0.009 • • Combined
C
•• •
.• .....
.9
E 0.008 •• "
~ ..,
in 0.007
OJ
c: • ••
sen 0.006 " •
E 0.005 .., 0°
o 06
<DO 6
A b-
"6
.".
,,'" ..
0
5 0.004 0
"
:c
iii
c:
Cll
0.003 •~ A. AA 0
" 00
"..-
0 M "00 .......
0.002 ~ 0 <II •
0.001 "" ~"Ettt;

~aaoI' 'Ii eIPa ~.
0.000 000 m 0 ." " " " " B • •
0 1 00 200 300 400 500
Depth (Angstroms) (a)
152
Combined· 38% AI' + 51% AI" + 11% AI-
••
o AI' (2keV)
0.012 •
o
AI" (4keV)
AI'++(6keV)
• Combined
0.010 ••••
0.008 •••
0.006
~
'iii 0.004
c:
Q)
o
0.002
0.000
.
o 100 200 300 400
Depth (Angstroms) (b)
Figure 2
Computer simulation results by TRIM for AI implantat ion in Kapton and Mylar at
V=·2 keY, taking into account the ion charge states distribut ion at PIII :
a) depth distribut ion of AI in Kapton;
b) depth distribut ion of AI in Mylar .
The results of computer simulation of AI implantation at V=-2 keY, taking

into account the charged states distribution, are presented for Kapton and Mylar in
Figure 2. The computational code predicts almost the same penetration depth and
distribution of AI in both polymers, i.e., if without any diffusion-based possible
redistribution, the depth penetration of AI should be close to -20 nm, and with a
similar shape (see Figure 2 a, b). It is clear from these results, that a significant
amount of AI may be expected in a very thin surface layer of the treated polymers.
Some of the real atomic mixing processes, that occur at PIlI of chemically reactive
AI, such as oxygen uptake and intermixing, can not be simulated by TRIM.
Therefore, a significant amount of oxygen, closely repeating Al distribution, can be
also expected in a very thin top surface layer (see Results and Discussions, section
4.2).
153
4.1 RETAINED DOSES AND DEPTH DISTRIBUTION BY RBS
Since Plasma Immersion Ion Implantation of AI was carried out at a chamber

pressure about 5xlO'6 Torr, and in a pulsed mode, significant adsorption of oxygen
can be expected both to the polymer and to the cathode itself between pulses, and
perhaps even a reaction with the highly reactive Al on the cathode surface or on the
polymer surface.
The retained doses and depth distribution of aluminum and oxygen in the
polymers treated at both mentioned above, "preferential" and "high-dose"
conditions, were determined by RBS, and the results are presented in Figure 3 -
Figure 5.
It becomes clear from these results that at "preferential" treatment
conditions (see Figure 3) the average amount of implanted AI was -(4.2-4.7)
· 1016cm' 2, and a significant amount of 0, such as (4.2-5.7) '101 6cm' 2 was intermixed
in the very thin top surface layer. The amount of oxygen in both materials exceeds
the amount of implanted AI. At the "high-dose" implantation (see Figure 4), the AI-
retained dose is -(5 .5-6) .1016cm' 2, that is in good agreement with the higher
implantation time. There is also a clear indication of significant amount of oxygen
uptake in the thin top surface layer at this treatment condition (Figure 4), such as
-(4.5-5) ·IOI6cm·2. From the comparison of the amount of the implanted aluminum
and oxygen uptake, it is clear that in this case the amount of oxygen is always less
than the amount of aluminum. Since no significant carbonization effect has been
found in the treated polymers (see the RBS spectra of pristine and treated polymers
for comparison), it becomes clear that mentioned above change in appearance and
surface properties of "high-dose" treated polymers is caused by the effect of ion-
assisted metallization.
Special attention has been paid to the depth distribution of both aluminum
and oxygen in the treated polymers. It was important to compare the results for
aluminum with the computer simulation data, as well as to confirm the correlation
of oxygen undermixing with aluminum implantation . The results obtained from
RBS data at "high-dose" PIlI (Figure 4), are presented in Figure 5.
154
soo
Kaplon NH
400
Q)
c: AI
c: 300
m
s:
~
§ 200
0
0
100 (a)
~ 600 800 1000 1200

Backscattered Energy (keY)
500
KIiptofI CB
Q)
c: AI
c: 300
ca
s:
~c: Oxygen
::J 200
0
0
(b)
100
400 600 800 1000 1200

Backscattered Energy (keY)
'i)
c: AI
c
II 300
s:
~
!!
c:
::J 200
0
0 (c)
. 100
-400 600 800 1000 1200

Backscatte red Energy (keY)
Figure 3
RBS spectra of a) Kapton HN , b) Kapton CB , and c) Mylar after PIlI with AI at "preferential" treatm ent
conditions (for conditions see text)
155
AI+O implanted Kapton

AI
Implanted
400
unlmplanted
.\
~
C
::::l
o
o 200
(a)
600 800 1000 1200 1400 1600
Energy (keV)
• Implanted
AI
600 unlmplanted
c
~
C
::::l 400 o
o
o
200
(b)
600 800 1000 1200 1400 1600
Energy (keV)
Figure 4
RBS spectra of: a) Kapton HN andb) Kapton CB(Kapton black) after PIlI with AI at "high-dose" treatment
conditions (for conditions see text), in comparison with RBS spectra for pristine polymer films of both
materials
156
AI+O imp. Kaptan
0.40
.• ••••• •.•
0.30 • oCto•
o o·
o
•' 00
':'0 ' 0
• •0
0.20
....
...
. . ..
.)
-:.
0.10 •
(,
.
~
~ '~ -~ .•
o.00 L..-L.......I.--L---L---..l........l.--1-...J..-..L-.L.-L-JI-.l---L---..l..---I---'-~...L-...J
.... '" .. . ..
" '.~ (a)
o 100 200 300 400
depth (e15 atoms/em")
AI+O imp. Kaplon black
0.40
••••
c ••
o ••
15 0.30 • ••
~ • ••
.2 ..:,-:..-:. c-c-,.. . .;.(. ,,0 c•
E 0.20 c. •
...
o
.E ..•. .)
« .:'
'
0.10
.....··.. ., .•.•.. . .
~:.o .:. .:. (. .:'
:. (.
~ '
". " .
0.00 L...-lL.......L--L---L.....L-...L..-..I...-L-J-L--L---L.....L-...I..-..L-L-J-L--L...;J
o 100 200 300 400
depth (e15 atoms/em")
Figure 5
Depth distribution of implanted aluminum and intermixed oxygen in top surface layers of:
a) Kapton HN and b) Kapton CB. treated at "high-dose" PIlI conditions (see Figure 4)
157
When compared with the computer simulation results (Figure 2), it can be seen that
the depth profile of the implanted aluminum is close to the results of computer
simulation for Kapton. The oxygen depth distribution is also in a reasonable
agreement with aluminum depth profile. However, in the experiment the maximum
of AI depth profile is shifted more to the surface, that might happen due to oxygen
intermixing and chemical reactions of aluminum with oxygen.
It is interesting to note, that the surface composition of the main species,
analyzed by XPS, was drastically different, if to compare the ones for the same
polymers, implanted with metal elements, including AI, at middle energy
conventional ion implantation , and with similar doses [1-3]. In the last case, the
implanted species have not been detected at the surface. Oxygen surface
concentration was reduced, and the concentration of carbon was increased due to
carbonization effects [2,3] . At PIlI, using both mentioned above treatment
conditions, a significant surface concentration of AI (in various oxidized states) and
oxygen appeared at the surface of treated materials, also an indication of ion mixing
effects ( see results in Section 4.3).
4.2 ATOMIC OXYGEN EXPOSURE AND TESTING RESULTS
As mentioned above, a severe testing environment, FAO beam, that imitate the
interaction with atomic oxygen in LEO, has been mostly used for testing of the PIlI
treated polymers, with the FAO fluence up to 1.57x102°at·cm-2. For some of the
samples, implantation was followed by hours of atomic oxygen exposure for testing
in the RF plasma asher. Based on the vacuum dehydrated Kapton H witness sample
mass loss measurements, those samples received an effective AO fluence up to
2. 15xl020at·cm-2. Samples, treated at both "preferential" and "high-dose" implan-
tation conditions, have been tested under FAO beam, as well as in oxygen plasma
asher. Testing of samples from both sets has been also performed in the FAO beam
after plasma exposure for a few hours. Mass loss of the implanted samples in the
plasma asher, that can be attributed to the conversion processes, was at least an
order less than for pristine Kapton, and stable, protective surface structures have
been developed on all polymers. For samples, implanted at "preferential"
conditions, both solar reflectance <X. and thermal emittance s have been left
practically unchanged not only by the PIlI treatment, but after oxygen plasma
testing as well. These are very important requirements for polymer films, since they
are often used as thermal control materials in space applications.
Two sets of samples, one as implanted and another one after this plasma
exposure, have been exposed to FAO at the UTIAS Fast Atomic Oxygen Beam for 5
hours and 8 hours, at FAO doses up to - (l-1.5)xI0 2°at . em", For all the implanted
samples the total mass change per surface area was almost two orders of magnitude
less than for the control pristine Kapton HN, i.e, it was at the precision of weight
measurements. The color and appearance of the treated polymer films did not
change, in contrast to heavily eroded pristine Kapton HN witness samples.
The stabilization and protection effects can be explained by examining the
surface content of the implanted materials after FAO testing. According to XPS
158
data, a certain amount of carbon was removed from the surface of every material in
the interaction with FAO, and additional enrichment of AI in oxidized states, as
well as oxygen increase, appeared at the surface (Table 1). These are clear
indications of surface conversion, that was an important stage in Implantox surface
modification, when using middle-energy conventional ion implantation. By
stoichiometry of the surfaces, it can be suggested, that protective, oxide-based
surface layers were finally developed. Oxides-based surface content by XPS was the
same after 5 hrs and 8 hrs ofFAO exposure and testing (see Table 1), that is also an
indication of stable, protective surface layers.
A significant amount of oxygen is present in the PIlI-treated samples. that
was estimated from RBS results as being in the range (4_6)·101 6cm' 2. It is a clear
indication, that oxygen uptake was occurring during the PIlI treatment. Surface
composition of the treated polymers (top surface layer of -3-5 urn) was also
analyzed by XPS, and the results are presented in Table I, together with the data for
pristine and FAO exposed materials. Both aluminum (in different oxidized states)
and oxygen were found at the surface after PIlI in high atomic concentrations (see
Table 1). The surface concentration of carbon was subsequently reduced, if to
compare with pristine materials.
Table 1
Surface compositionof polymerfilms by XPS after PIlI at preferential conditions
and after the followingFAO testing
Surfacecomposition by XPS.at %
Material Treatment C 0 N A]*) Si
Pristine 77.4 15.9 6.35 0.18

PIII,A] 63.7 21.7 0.4 15.2
KaptonHN FAO (5 hrs)**) 14.7 50.4 0.3 32.8 1.7
FAO (8 hrsre) 15 51 0.3 33 1.8
Pristine 75.41 17.0 6.15
PlI, A] 31.8 38.1 0.3 29.7
KaptonCB FAO(5 hrs) 15.7 48.6 0.6 33.7 1.2
FAO (8 hrs) 16 4.8 0.6 34 0.8
Pristine 78 22
PlI, A] 47.2 31.6 0.6 20.5
Mylar FAO (5 hrs) 22 44 0.9 31.5 1.4
FAO (8 hrs) 21.7 45 0.5 32 l.l
159
Protection of the treated polymers from oxidation and accelerated erosion was
clearly illustrated by the difference of surface morphology, analyzed by scanning
electron microscopy (SEM). FAO exposed pristine polymers have been heavily
eroded, with the development of a well-known "carpet-like" surface morphology ; an
example for FAO-exposed bare Kapton HN is presented in Figure 6(a) . On the
contrary, all polymer films that have been treated by PIlI in this study, exhibit
unchanged surface morphology after FAO exposure . A strong contrast was always
evident between the stable, protected surface of treated polymers, and untreated
pristine material, that is illustrated in SEM microphotograph of Kapton in Figure 6
(b).
Surface properties of the modified materials have also been explored by
water contact angle measurements, performed on pristine and implanted polymers,
as well as after different types of oxidative post-treatment and storage at ambient
conditions (Table 2). Water contact angle of all polymer films was significantly
reduced after Al PIlI, in a reasonable agreement with the presence of some oxiders)-
based components in the surface content (see Table 1). It was expected that the
oxide-enriched surface will have an additionally reduced water contact angle after
FAO testing. But, opposite to these expectations, the contact angles for these
samples, measured after -3 weeks of storage, were close to the values for ion
implanted, or even higher (Table 2). Some of the PIlIl treated samples have been
repeatedly washed in acetone or toluene, as in [21], to remove chemically any
possible gaseous or volatile species, adsorbed by the converted surface from the
environment in storage . The contact angles were only slightly reduced by these
washing procedures. However, treatment in a UV/Ozone Cleaning System
(UVOCS), i.e., by exposure to UV radiation in the 254 nm and 185 nm range
(predominantly 254 nm) in the presence of air for 10 minutes reduced the contact
angle practically to full spreading. During the following storage for a few weeks, the
contact angle was again gradually increasing (Table 2).
TABLE 2
Water contact angle of polym er films, treat ed by AI PIlI , following oxidative
post-treatments in various oxidative environments, and after storage
Water contact angle (deg)
Material Pristine PIlI FAO UV+

AI exposure, ozone , Following storage at ambient conditions
storage 10
(3 weeks) min
16 hrs 48 hrs 72hrs 3wks 6wks
Kapton HN 72 44 74 -0 42 58 64 69 70
KaptonCB 66-77 38 73 18 33 53 55 57 57-60

Mylar 77 40 84 16 32 51 53 67 72
Since these samples have already been treated in reactive oxidizing environment,
such as FAO, and proved to be stable in these conditions, the last results can be
viewed as an indication that polymer surfaces treated by PIlI and exposed to atomic
160
Figure 6
Scanning electron microscopy photos of FAO-cxposed Kapton films :
a) Kapton HN untreated (pristine) sample, exposed to FAO for 8 h, x 10,000 ;
b) contrast between PIlI-treated area (left) and untreated area (right), x 5,000
161
oxygen are sensitive to the terrestrial ambient environment. These characteristics,

similar to those in [21]. can be useful for the development of advanced sensors,
based on "marriage" of oxide-based and polymer films sensor applications, with
improved resistance to oxidation.
4.3 THERMAL-OPTICAL PROPERTIES
All polymers, implanted at "preferential" conditions , have not shown any practical
change in both solar absorptance and thermal emittance, as well as transmittance,
and kept the same properties after both FAO and oxygen plasma testing (see Table
3).
Table 3
Thennal-optical properties of polymer films after PlIl at preferential conditions,
and after O-plasma asher testing
Solar Thermal
absorptanc e, emittance, Transmittance
Material Treatmenl al£
a e (:I. =
670 nrn)
PUI (AI») 0.63 0.868 0.72 0.803
0.631 0.8 66 0.72 0.802

Kapton HN POI (AI)+plasma asher")
Pili (AI) 0.945 0865 1.09 n/a
Pili (AI)+plasma asher 0 .941 0.861 1.09 n/a

KaplonCB
Pill (AI) 0.69 0.875 0.78 0.873
PIn (AI) +plasma asher 0.69 0.873 0.79 0.866

Mylar
This is a significant achievement, associated with the AI PIll treatment at

preferential conditions - the polymer films have been made highly resistant to
imitating LEO environmental conditions, and , at the same time, their thermal-
optical properties have been left unchanged and stable. All polymer films of the
second set, implanted with aluminum at "high-dose" conditions, have shown
drastically changed optical and thermal-optical properties. The transmittance of
Kapton HN and Mylar (at A = 670 nm) dropped from 0.80 and 0.87 to 0.22 and
0.15 subsequently. Thermal emittance was significantly reduced, for instance , from
-0.86-0.87 down to 0.58 for Kapton HN and to 0.23 for Mylar, and did not change
after atomic oxygen testing . This is an example , how the major thermal-optical
properties of the treated polymeric films may be altered by Implantox surface
modification, based on metal Plasma Immersion Ion Implantation.
162
5.0 Conclusions
The results of the first attempt to improve oxidation and erosion resistance of a few
high-performance polymers by Plasma Immersion Ion Implantation have shown
significant potential of this surface modification technique for polymer materials
protection from severe oxidative environments . Based on computer simulations and
experimental results, both ion and recoil implantation (dynamic ion mixing) can
play an important role in the modified protective top surface layer formation.
Surface properties modification, as well as the appearance of the treated polymers
were found to be sensitive to PIlI conditions and duration . Exposure and testing in
an D-plasma asher and in FAD beam have shown not only drastically improved
oxidation and erosion resistance after AI PIII at appropriate conditions, but the
capabilities to leave the surface morphology and very important thermal-optical
properties of the materials unchanged , or significantly altered. At the same time, the
change of surface composition after FAD exposure, such as enrichment of the
oxide(s)-based component of the modified surface layer, provides a clear indication
of some conversion processes, undergoing in FAD, at least in the top (3-5 nm)
surface layer.
6.0 Acknowledgments
The authors would like to thank Dr.K.M. Yu for help in RBS measurements, and
Dr. R. Sodhi for assistance in XPS analysis.
7.0 References
I. Iskanderova, Z., Kleiman, I.J., Gudimenko, Yu., Morison, W.O., Tenn yson, R.C., " Improvement of
oxidation and erosion resistance of polymers and composites in spac e environment by ion implantation,"
NIM B 127/128 (1997) 702-709.
2. Iskanderova, Z., Kleiman, 1., Gudimenko, Yu., Cool, G ., and Tenn yson, R.C., " Computer simulation and
experimental studies of ion implantation in polymers for erosion resistance impro vement, " Computer
Modelling of Electronic and Atomic Processes in Solids, Eds. R.C . Tennyson, A E. Kiv, Kluwer
Academic Publishers, 289-299, 1997.
3. Iskanderova, Z.• Kleiman, J., Gudimenko, Yu ., Morison, W.O., Tenn yson, R.C., "Erosion resistance and
durability improvement of polymers and composites in space environment by ion implantation,"
Mat.Chem. Phys. 54,91-97,1998.
4. Iskanderova, Z., et ai, "Surface modification of polymers and carbon-based materials by ion implantation
and oxidative conversion," U.S. Patent 5,683 ,757, Nov. 4, 1997.
5. Iskanderova, Z., Kleiman, J., Gudimenko, Yu., Morison, W.O., Tennyson, R.C., "Comparison of surfa ce
modification of polymeric materials for protection from severe oxidati ve environmen ts using different ion
sources," Surf & Coat. Techno/. 127, 18-23 ,2000.
6. Brown, I.G., Anders, A , Anders, S., Dickinson, M.R., and MacGiIl, RA , "Metal ion implantation:
conventional versus immersi on," J. Vac. Sci. Techno/., BI2 (2 ) 823-827, 1994.
7. I. G. Brown, I.G., Anders, A , Anders, S., Castro , RA , Dickinson, M.R., MacGiIl, RA , Wang, A,
" Synthesis ofunattainable ion implantation profiles - 'Pseudo-implantation ', NIM B 106, 646.650, 1995
8. Brown, I.G., " Vacuum arc ion sources ," Rev. Sci. Instrum. 65 ( 10), 3061-3081 ,1994.
9. Anders, A, Anders, S., Brown, I.G., Yu, K.M., " Increasing the retained dose by plasma immersion ion
implanta tion and deposition," NIM B 102,132-135 , 1995 .
163
10. Brown, I. G., Anders, A , Anders, S., Dickinson, M.R, Ivanov, I.C., MacGill, R. A, Yao, XY ., and Yu,
K.M., "Plasma synthesis ofmetallic and composite thin films with automically mixed substrate bonding,"
NIM B 80/81,1281-1287,1993,
11. Davenas, J., Boiteux, G., and Xu, XL., "Role of the modifications induced by ion beam irradiation in the
optical and conducting properties of polyimide," NIM B 32,136-141, 1988.
12. Davenas , J., Boiteux, G., and Fallavier, M., " Speculations on a mechanism for the ion beam induced
degradation ofpoly imide," NIM B 39,796-799,1989.
13. Fink, D., Milller, M., Chadderton, L. T., Cannington, P. H., Elliman, R. G., and McDonald, D. C.•
"Optically absorbing layers on ion beam modified polymers : a study of their evolution and properties,"
NIM B 32,125-130, 1988.
14. Xu, D., Xu, XL., Du, G.D., Wang, R., Zou, S.C.• Liu, X.H., "Infrared analysis of the irradiation effects
in aromatic polyimide films," NIM B 80/81, 1063-1066, 1993.
15. Lee, E.H., Lewis, M.B., Blau, PJ., Mansur, L.K., "Improved surface properties of polymer materials by
multiple ion beam treatment," J. Mat . Res. 6 (3), 610-628, 1991.
16. Ocher, R, Kluge, A, Zechel-Malonn, S., Gong, L., and Russel, H., "Improvement of surface properties
of polymers by ion implantation," NIM B 80/81, 1050-1054, 1993.
17. Handbook of modem ion beam material analysis, Ed. J. R. Tesmer, M. Nastasi, MRS, 1995.
18. Iskanderova, Z., Kleiman, J., Gudimenko, Yu., Morison, W.D., Tennyson, R.C.• Brown, I.G.• Monteiro ,
O. R., Banks, B. A , Norris, M. J., Proc.Abstracts of 4th Int. Conf " Protection of Materials and
Structures in Low Earth Orbit Space Environment," 23-24 April, UTIAS, Toronto, Canada, 27-29 ,1998.
19. Tennyson, RC., "Atomic oxygen effects on polymer-based materials," Can. J . Phys. 69, 1190-1208,
1991,
20. Koontz, S.L., A1byn, K., Leger, LJ., J. Spacecraft and Rockets, 28, 315, 1991.
21. Iskanderova, Z., Kleiman, 1, Gudimenko, Yu ., Morison, W.D., Tennyson, R.C., "Surface modification of
polymer-based materials by ion implantation - a new approach for protection in LEO," Protection of
Materials and Structures in Low Earth Orbit Space Environment, Ed. J.I.Kleiman and R.C. Tennyson,
Kluwer Academic Publishers , 225-233, 1999.
22 . Ziegler, 1F., TRIM 90 , Transport of Ions in Matter, IBM Corp, 1990.
A COMPARISON OF SPACE- AND GROUND-BASED FACILITY
ENVIRONMENTAL EFFECTS FOR FEP TEFLON
SHARON K. RUTLEDGE AND BRUCE A. BANKS

21000 Brookpark Rd.
Cleveland. Ohio 44135; U.S.A .
MICHAEL KITRAL
Cleveland State University
Euclid Ave. at East 24th Street
Cleveland. Ohio 44115 ; U.S.A.
Abstract
Fluorinated Ethylene Propylene (PEP) Teflon" is widely used as a thermal control

material for spacecraft. However, it is susceptible to erosion, cracking, and
subsequent mechanical failure in low Earth orbit (LEO). One of the difficulties in
determining whether FEP Teflon will survive during a mission is the wide disparity
of erosion rates observed for this material in space and in ground-based facilities.
Each environment contains different levels of atomic oxygen, ions, and vacuum
ultraviolet (VUV) radiation, in addition to parameters such as the energy of the
arriving species and temperature . These variations make it difficult to determine
what is causing the observed differences in erosion rates.
This paper attempts to narrow down which factors affect the erosion rate of
FEP Teflon through attempting to change only one environmental constituent at a
time. This was attempted through the use of a single simulation facility (plasma
asher) environment with a variety of Faraday cages and VUV transparent windows.
Isolating one factor inside of a radio frequency (RF) plasma proved to be very
difficult. Two observations could be made. First, it appears that the erosion yield of
FEP Teflon with respect to that of polyimide Kapton is not greatly affected by the
presence or lack of VUV radiation present in the RF plasma and the relative erosion
yield for the FEP Teflon may decrease with increasing fluence. Second, shielding
from charged particles appears to lower the relative erosion yield of the FEP to
approximately that observed in space, however it is difficult to determine for sure
whether ions, electrons, or some other components are causing the enhanced
erosion.
165
Protection ofSpace Materials from the Space Environment. 165-179.
166
1.0 Introduction
Materials qualification for atomic oxygen durability in low Earth orbit has been
performed in both ground-based facilities and in space. The ground-based facilities
have been developed out of the necessity to study long term atomic oxygen effects to
levels greater than can be achieved with a typical mission on the Space Shuttle .
They also have the advantage of quick turnaround, lower cost per exposure, and
ease of experiment tailorability.
Some questions naturally occur about the applicability of ground-based test
data for use in predicting how a material will perform in low Earth orbit. These
questions arise due to the differences in the species, energy or charged state of the
arriving atoms and the intensity and wavelength of VUV radiation present. Other
parameters also may playa role, such as temperature . There is currently no known
facility, whether thermal plasma or directed beam, which will exactly duplicate
conditions in low Earth orbit [1-7]. Such facilities, however, are needed to screen
materials for use in LEO. So it is important to understand why differences in
erosion rates occur so that the test equipment can be modified to produce results
closer to those in space, or to allow calibration factors to be developed.
Of all the materials, FEP Teflon seems to provide the widest disparity of data
in space- and ground-based systems. Typically, FEP Teflon exposed in RF plasmas
has an erosion yield relative to polyimide Kapton which is an order of magnitude
greater than in LEO. This paper attempts to determine what is present in one type of
simulation facility (RF plasma), that affects the erosion of FEP Teflon, by isolating
selected components in the environment that may have an effect on the erosion rate.
VUV radiation and charged species were the components selected to be the focus of
this study due to the wide belief among the LEO testing community that these are
the important factors.
2.0 Experimental Procedure
2.1. MATERIALS
Sample coupons 2.54 em in diameter, were punched from 0.005 cm thick sheets of
polyimide Kapton HN (DuPont) and fluorinated ethylene propylene (FEP Teflon)
(DuPont). All of the samples were fully dehydrated for 48 hours in a vacuum of 8-13
Pa (60-100 mTorr) prior to weighing and subsequent exposure in the atomic oxygen
plasma. Samples were quickly weighed after dehydration and upon removal from
the vacuum chamber after exposure to atomic oxygen, in order to minimize errors in
mass due to water absorption [8].
2.2 PLASMA ASHER
An SPI Plasma Prep II asher operated on air was used as the environment simulator
for these tests. The asher operates by using RF (13.56 MHz) to create a discharge
167
between two electrodes which surround a glass reaction chamber. A thermal plasma
is produced at an energy of « 0.1 eV. The plasma contains atomic and excited state
species as well as molecules, ions, electrons, and VUV radiation. The amounts and
intensity of the latter five have not been determined. The effective arrival of atomic
oxygen is estimated by determining the reaction rate of polyimide Kapton , whose
space degradation rate is well known. The nitrogen in the plasma has been shown in
earlier tests not to react with the materials that are being tested in this study [8].
Typical vacuum chamber pressure during operation was 16-27 Pa (120-200 mTorr) .
Temperature measured in the past inside the plasma chamber was 65°C [9].
2.3 FARADAY CAGES AND VUV WINDOWS
Aluminum Faraday cages were constructed from single sheets of 0.0127 em

aluminum foil so that boxes with attached lids were formed. Openings under the
lids on the sides produced open re-entrant boxes and prevented line-of-sight
exposure of the samples to the external air plasma. Several modifications were made
to the boxes to allow various plasma components to be excluded or allowed into the
boxes.
Figure 1 shows the various box designs used for testing. Figure lA is a closed
Faraday cage. This design is supposed to block VUV radiation and charged species
from the samples inside while allowing atomic oxygen to scatter into the box
through the openings under the lid on the side of the box. This box design was
modified in Figure IB to include an opening in the lid surrounded by aluminum foil
tabs to hold various windows and filters that were 2.54 em diameter . The window or
filter after insertion was covered with a thin aluminum foil sheet with a circular
hole cut in the centre that was bonded to the surface of the window and box with
acrylic adhesive. This prevented plasma from leaking into the box around the
window.
Al Foil Box with

Attached Lid
Figure IA.
Closed Faraday cage
168
Al Foil Holder
~inAI Foil Pocket

@
Figure lB.
Faraday cage with windowor mesh filter
A magnesium fluoride window was used to allow VUV radiation of

wavelengths greater than approximately 120 urn, and visible light, to impinge onto
one sample placed inside of one end of the box. The transmission versus wavelength
for the window is shown in Figure 2. Another sample could be placed at the other
end of the interior of the box so that it would see all of the same conditions but not
be exposed to the VUV radiation entering through the window. A gold mesh filter
was also used to allow all wavelengths to enter the box but block most of the
charged particles from entering. A single mesh screen spot-welded to an aluminum
foil ring (0.0127 em thick) was found to actually accelerate charged species into the
box, so a double screen was made from two sheets of gold mesh spot-welded to
either side of an aluminum foil ring (0.0127 em thick). The gold mesh was
manufactured by Buckbee Mears to have 80 per cent transparency at roughly 39
lines per centimeter of 0.00267 em diameter gold wire.
In order to keep samples from sliding out from under the window during
exposure due to vibration, a 0.0127 em thick aluminum foil rectangle that just fit
into the interior of the box was cut, as shown in Figure I, with tabs that would fold
around the samples to hold them in place. Each sample was wrapped with
approximately 0.002 cm thick aluminum foil with a 1.905 em diameter hole
punched in the front surface to allow the atomic oxygen to reach the sample of FEP
Teflon or polyimide Kapton HN.
2.4 DATA MEASUREMENT AND ANALYSIS
Changes in mass were recorded for FEP Teflon and polyimide Kapton HN for each
test configuration. The effective fluence (number of atoms of atomic oxygen arriving
per square centimeter over the test duration) was determined by measuring the mass
loss per unit area of polyimide Kapton HN and using Equation 1.
169
100 r-----===:::;::::;:;:;:;:;:;:;;;;;;;;;:;;;;;;;;;:;:;:;:;:;:;:;:;:;:~
90
.~ 80
. '"
~ 70
~ 60
~
a 50
["""' 40
~ 30
~ 20
~ 10
Of-£---+---..,f----+----+---t----i
100 150 200 250 300 350 400
Wavelength (om)
Figure 2.
Transmission of MgF 2 Window
F=MIA (1)
p*E
Where:
F= Fluenee (atoms/em')
M= Mass loss ojpolyimide Kapton (g)
A = Area ofpolyimide Kapton exposed to atomic oxygen (em2)
p= Density ojpolyimide Kapton ( 1.42 g/cm'}
E= Erosionyieldjor polyimide Kapton in LEO (3x10·24 cm'ratom)
The effective erosion yield for FEP Teflon, which is the volume ofFEP Teflon
removed for each effective atom that arrives, was determined by using Equation 2.
The erosion yield was then divided by the known erosion yield for polyirnide Kapton
in low Earth orbit (3xIO· 24 cnr'zatom) to obtain the measurementrelative to a known
standard. Erosion yields relative to polyimide Kapton were then plotted to look for
potential trends in the data.
E _ MIA
(2)
FEP - p*F
Where:
170
EFEP = Erosion yield for FEP Teflon (cnr'zatom)

M = Mass loss ofFEP Teflon (g)
A = Area ofFEP Teflon exposed to atomic oxygen (cnr')
p = Density ofFEP Teflon (2.15 g/cnr')
F = Fluence (atoms/em') (based on polyimide Kapton)
3.1 EXPOSURE DIRECTLY IN THE PLASMA
FEP Teflon exposed in the plasma should see the total environment present
including charged species and VUV radiation . Several tests were conducted with
FEP Teflon and polyimide Kapton samples exposed to the plasma while sitting on
top of glass microscope slides that were approximately 7.62 em x 2.54 cm in area .
The average erosion yield with respect to Kapton measured over approximately the
same fluence exposure was 1.9 ± 0.14. Samples of FEP Teflon and polyimide
Kapton were also placed on top of some of the Faraday cages during testing to
record what the effective atomic oxygen fluence was outside of the Faraday cage. It
was noticed that during these tests, the erosion yield with respect to Kapton in the
plasma but on top of the aluminum surface was lower than that observed on glass
slides (between I and 1.2). This was unexpected, because the plasma environment
was believed to not be affected by the sample holders.
Another test was conducted but this time, the FEP Teflon and polyimide
Kapton were placed in aluminum foil covers and held down to an aluminum sheet
(identical sample holders used in the Faraday cages and described earlier) with
aluminum foil tabs. This placed the samples in good electrical contact with the
aluminum and also allowed them to sit in the same location in the asher as the
samples on glass slides had been during previous testing. The erosion yield for FEP
Teflon with respect to Kapton was even lower for this test (0.374 ± 0.036). A
comparison bar chart with error bars is shown in Figure 3.
It can be seen that the differences in the erosion yields are much greater than
the measurement error. Flux levels varied during each test, but the overall fluence
was approximately 2x10 20 atoms/em/ on average. Table I contains the data for all of
the tests performed including the effective atomic oxygen flux. There does not
appear to be any relationship between flux and relative erosion yield. It is possible
that the surface conductivity may affect the arrival of charged species at the sample
surface or that VUV radiation intensities may be altered near the sample. In order to
try to narrow down the reason for these differences, testing was conducted inside
Faraday cages where some of these environments could potentially be reduced or
isolated.
171
On Glass Slides On Single Mesh Cage On Dou ble Mesh Cage In Aiuminwn lIolder
Figure 3.
Comparison of relative erosionyields for FEPTeflon Exposed In Plasma
Table I. Data From Plasma Asher Exposures
Description Material Time (hrs) Effective Atomic Oxygen Erosion Yield With
Flux (atoms/cmi-sec) Respect to Kapton
IllRIn plasma Kapton 7.98 7.43E15 ± 7.37e14 1.0
on glass
III g In plasma FEP 7.98 7.43E15 ± 7.37el4 2.0065 ± 0.28
on glass
1120 In plasma Kapton 9.18 6.5E15 ± 6.48SEI4 1.0
on glass
1120 In plasma FEP 9.18 6.5E15 ± 6.485EI4 1.968 ± 0.278
on glass
406 In plasma Kapton 4.67 5.11EI5± 1.13E14 1.0
on glass
406 In plasma FEP 4.67 5.l1E15 ± 1.13E14 1.6898 ± 0.050
on glass
409 In plasma Kapton 5 5.14E15 ± 5.13E14 1.0
on glass
409 In plasma FEP 5 5.14E15 ± 5.13E14 1.93±0.271
on glass
402 In plasma Kapton 5.633 1.406E16 ± 5.037E14 1.0
on aluminum
holder
402 In plasma FEP 5.633 1.406EI6± 5.037E14 0.3742 ± 0.03575
on aluminum
holder
172
Description Material Time Ihrs) Effective Atomic Oxygen Erosion Yield With
Flux (atoms/crni-sec) Respect to Kapton
on top of double
mesh cage
415 In plasma FEP 42.1 4.366E15 ± 8.66E13 1.015 ± 0.0284
on top of double
mesh cage

on top of single
mesh cage
410 In plasma FEP 5.05 4.86E15 ± 1.07E14 1.223 ± 0.037
on top of single
mesh cage
410 In single Kapton 5.05 4.72E15 ± 1.856E14 1.0
mesh windowed
cage
410 In single FEP (Under 5.05 4.72EI5 ± 1.856E14 3.53 ± 0.1857
mesh windowed Mesh)
cage
410 In single FEP (Shielded) 5.05 4.72E15 ± 1.856E14 0.0204 ± 0.0167
mesh windowed
cage
415 In double Kapton 42.1 4.666E15 ± 1.64E14 1.0
mesh windowed
cage
415 In double FEP(Under 42.1 4.666E15 ± 1.64E14 0.117 ± 0.00596
mesh windowed Mesh)
cage
415 In double FEP (Shielded) 42.1 4.666E15 ± 1.64EI4 0.0321 ± 0.0027
mesh windowed
cage
130 Closed Kapton 24.3 4.07E15 ± 4.03E14 1.0
Cage with One (shielded)
MgF2 Window
130 Closed FEP (under 24.3 4.07E15 ± 4.03E14· 0.103 ± 0.0143
Cage with One window)
MgF2 Window
130 Closed FEP (shielded) 24.3 4.07E15 ± 4.03E14 0.091 ± 0.0127
Cage with One
MgF2 Window
331 Closed Kapton (under 24.017 8.75E15 ± 3.08E14 0.9432 ± 0.047
Cage with MgF2 window)
Window
331 Closed Kapton 24.017 8.75E15 ± 3.08E14 1.0
Cage with MgF2
Window
401 Closed Kapton 24.8 1.48E16 ±5.19E14 1.0
Cage with MgF2
Window
401 Closed FEP(Under 24.8 1.48E16 ± 5.19E14 0.0325 ± 0.0018
Window
173
Description Material Time (hrs) Effective Atomic Oxygen Erosion Yield With
Flux (atomslcm2·sec) Resnect to Kaoton
401 Closed FEP (Shielded) 24.8 1.48E16 ± 5.19E14 0.0217 ± 0.0011
Cage with MgF2
Window
406 Closed Kapton 48.967 1.283E16 ± 4.5E14 1.0
Cage with MgF2
Window
406 Closed FEP (Under 48.967 1.283E16 ± 4.5E14 0.0327 ± 0.00163
Window
406 Closed FEP (Shielded) 48.967 1.283E16 ± 4.5E14 0.0085 ± 0.00118
Cage with MgF2
Window
522 Closed Kapton 47.88 4.77E15 ± 4.71E14 1.0
Faradav caee
522 Closed FEP 47.88 4.77E15 ± 4.71E14 0.0374 ± 0.0062
Faradav cage
530 Closed Kapton 5.4 6.85E15 ± 6.8E14 1.0
Faradav C32e
530 Closed FEP 5.4 6.85E15 ± 6.8E14 0.195 ± 0.0422
Faradav cage
609 Closed Kapton 5.08 6.64E15±6.86E14 1.0
Faradav caze
609 Closed FEP 5.08 6.64E15 ± 6.86E14 0.15 ± 0.028
Faradav cage
1231A Closed Kapton (under 5 2.55E15 ± 2.56E14 1.0
cage with MgF2 window)
window
1231A Closed Kapton 5 2.55E15 ± 2.56E14 0.982 ± 0.139
cage with MgF2 (shielded)
window
1231B Closed Kapton (under 27.033 3.09E15 ± 3.05E14 1.0
cage with MgF2 window)
window
1231B Closed Kapton 27.033 3.09E15 ± 3.05E14 0.814±0.114
cage with MgF2 (shielded)
window
3.2 EXPOSURE INSIDE A CLOSED FARADAY CAGE
The closed Faraday cage was used in an attempt to prevent VUV radiation from
reaching the surface of the samples by allowing no line-of-sight viewing of the
samples to the plasma. It was also designed to prevent charged species from
entering by allowing the box to float to plasma potential. Several tests at a variety of
fluence levels were conducted with FEP Teflon inside the Faraday cage. Results
varied but the erosion yield for FEP Teflon relative to Kapton was less than 0.2 in
all cases as shown in Figure 4. It was noticed that the relative erosion yield seemed
to decrease with increasing fluence. The data plotted as a function of effective
fluence is shown in Figure 5. The relative erosion yield does appear to take on an
almost exponential decay
174
0.25.-- - - - - - - - - - - -- - - - - - - - - - - -----,
~
~ 0.2
~
ti
~ 0.15
~
~ 0.1
~
$:
z
o
;;; 0.05
s
r.l
o
522 530 609 130 401 406
SAMPLE NUMBER .
Figure 4.
Comparison of relative erosion yields for FEP Teflon inside a closed Faraday cage
with increasing fluence. It is not known why this would occur in an isotropic
plasma. More testing and data are needed to verify this trend .
Another puzzling feature noticed during these tests is that the effective flux
inside and outside of the Faraday cage is approximately the same. One would expect
the flux inside the Faraday cage to be greatly reduced because atoms would need to
scatter in through slots underneath the lid in order to enter the box. However, this
does not appear to be the case. It is possible that the RF field is exciting a plasma
inside the box, but no glow was observed inside during testing. The greatly reduced
relative erosion yield for FEP Teflon also does not support this theory. The relative
erosion yield measured inside the Faraday cage is near that observed in space on the
Long Duration Exposure Facility (LDEF) (0.11 ± 0.002) [10].
3.3 EXPOSURE INSIDE FARADAY CAGE WITH MAGNESIUM FLUORIDE

WINDOW
Because it is difficult to draw conclusions from samples experiencing different fields

and fluxes, a Faraday cage was developed which could expose two samples inside
the same Faraday cage environment, but allow only one of them to have line of sight
to the plasma . This was accomplished by putting a magnesium fluoride window in
the lid towards one end of the Faraday cage box. Initially, polyimide Kapton HN
erosion was compared in order to verify that the window seal did not let
175
0.25 , -- - - - - - - -- -- - - - -- - - -- -- - -- - - - ,
z
~~ 0.2
g
~
~ 0. 15
~
a:
E
~ 0.1
Q
-l
101
>
Z
o 005
~
101
O+--- - - - -+-- - - - --!- - - - - -+-- -- - -+-- - - - ---l

o 5E+20 I E+21 I.5E+21 2E+21 2.5E+21
FL UENCE (atoms/anAl)
Figure 5.
FEP Teflon relative erosion yield as a function of effective atomic oxygen
fluence inside a closed Faraday cage
additional plasma into the box. The results shown in Figure 6 verify that the
polyimide Kapton HN is not affected by VUV radiation and that the seal functions
as expected, because the erosion rates both under the window and shielded from the
plasma are nearly identical and within experimental error.
The relative erosion yield of FEP Teflon was tested in a similar manner ,
but a sample of polyimide Kapton was placed between the two FEP Teflon samples
so that the exposure flux could be measured. Figure 7 contains the data for the FEP
Teflon. The effect of VUV radiation transmitted by magnesium fluoride appears to
be very low. There may be a slight increase in relative erosion yield with VUV
radiation, but not enough to account for the nearly order of magnitude difference
observed when exposed directly in the plasma.
3.4 EXPOSURE INSIDE FARADAY CAGE WITH GOLD MESH SCREEN

Fll.,TER
In order to determine if shorter wavelength VUV radiation plays a role, it was

necessary to block charged species from entering the box, but allow all wavelengths
of radiation to enter. A single gold mesh screen was installed in the Faraday cage
replacing the magnesium fluoride window used in the previous test. FEP Teflon and
176
1.5
Z
0
~
~
..b
1
r.I
~
~
Q
d 0.5
;;:
Z
0
t;;
~
r.I
0
331 123 1A 12318
SAMPLE NUMBER
Figure 6.
Comparison of relative erosion yields for Kapton HNexposed in a Faraday cage
with and without VUV radiation
polyimide Kapton HN were tested in the same manner as previously. During

exposure, a directed plasma beam was observed that was approximately 1 em in
length extending perpendicular to the mesh, centered on it and extending down
through it. The erosion yield of the FEP relative to Kapton in the shielded area of
the box was roughly the same as in the closed Faraday cage test. The relative
erosion yield for the sample under the mesh, however, was 3.53 ± 0.186. This is
much higher than what was observed when the sample was sitting directly in the
plasma. It appears that the mesh screen may have been acting like an electrostatic
accelerator grid increasing the flux of charged species into the box.
Because it was still not possible to determine if other wavelengths of VUV
radiation played a role, a double mesh was developed and installed in place of the
single mesh, making sure that it was electrically connected to the box to provide an
equipotential gap to reduce the flux of charged species entering. The relative erosion
yield was then observed to drop for the sample underneath the double mesh in the
Faraday cage to approximately 0.12 ± 0.006. This value is near that observed inside
the closed Faraday cage and is similar to that observed in space.
177
IIIUnder MgF2Wmdow
5 IIIShielded
§ 0.75
~
~
~ 0.5
1:
~
~
;;:
5 0.25
;
o
130 401 406
SAMPLE NUMBER
Figure 7.
FEP Teflon exposed inside a Faraday cage with and without VUV radiation
3.5 TESTING SUMMARY
Figure 8 contains a comparison chart of samples tested in the plasma asher under
various conditions as compared to the value measured in space. It appears that
placing FEP Teflon inside a Faraday cage will yield results very close to those
observed in low Earth orbit. It does not appear to matter whether VUV radiation can
enter the Faraday cage, only if charged species can enter. The greatl y-reduced
erosion yield for FEP Teflon samples sitting on a conductive plate but directl y inside
the plasma also indicates that charged species may play a role. It is very difficult to
isolate the effects of charged species inside of an RF field. In order to draw more
firm conclusions about what is causing the differences in erosion yields, tests will
need to be conducted in a facility where the sample region is not inside the RF field
such as ina directed beam facility.
178
4 ...-- - - - - - - - - - - - - - - - ---,
3.5
3
2.5
2
1.5
1
0.5
o
Figure 8.
Comparison of FEP Teflon relative erosion yields
4.0 Conclusions
These tests have shown that the accelerated erosion can be reduced by exposing the
samples on a conductive plate or inside a Faraday cage, whether or not the samples
are exposed to the VUV radiation from the plasma. This may provide a means for
testing of samples such as FEP Teflon in plasma environments which more closely
mimics the erosion observed in LEO . More work needs to be done to determine if
electrons, ions , or some effect of the RF field is causing the accelerated erosion of
FEP Teflon in plasma asher facilities.
5.0 References
1. Stidham, C.R. et at. " Low Earth Orbital Atomic Oxygen Environmental Simulation Facility for Space
Materials Evaluation," NASA TM-I06128, 1993.
2. Johnson, c.L. et at. "The Vanderbilt University Neutral O-Beam Facility," 18th International SAMPE
Technical Conference Proceedings, Oct. 7-9, 1986.
3. Arnold, G.S. Peplinski, D.R., "A Facility for Investigating Interactions of Energetic Atomic Oxygen
with Solids," Aerospace Report No. ATR-94 (8540).3,1985.
4. Brinza, D.E. et at. " Production of Pulsed Atomic Oxygen Beams Via Laser Vaporization Methods," 18th
International SAMPE Technical Conference Proceedings, Oct. 7-9, 1986.
5. Rempt, R.D. "Production of a Beam of Ground State Oxygen Atoms of Selectable Energy," 18th
International SAMPE Technical Conference Proceedings, Oct. 7-9,1986.
179
6. Sjolander, G.W. and Bareiss, L.E. "Martin Marietta Atomic Oxygen Beam Facility," 18th International
SAMPE Technical Conference Proceedings, Oct. 7-9, 1986.
7. Caledonia, G.E. et al. "A High Flux Source of Energetic Oxygen Atoms for Material Degradation
Studies," AlAA 85-7015 .
8. Rutledge, S.K. et al. "An Evaluation of Candidate Atomic Oxygen Resistant Materials for Space
Applications in LEO," NASA 1M-I00122, 1986.
9. Rutledge, S.K., Banks, BA and Cales, M., "A Comparison of Atomic Oxygen Erosion Yields of
Carbon and Selected Polymers Exposed in Ground Based Facilities and in Low Earth Orbit," AlAA-94-
2628 .
10. Banks, B.A. "The Use of Fluoropolymers in Space Applications," Modern Fluoropolymers, Scheirs,1.
Ed.; John Wiley & Sons, New York, (1997), 103-111.
EVALUAnON OF LOW EARTH ORBIT ENVIRONMENTAL EFFECTS
ON INTERNATIONAL SPACE STATION
THERMAL CONTROL MATERIALS
JOYCE A. DEVER
SHARON K. RUTLEDGE
Cleveland, OH 44135
MARK M. HASEGAWA
Boeing Space and Defense Systems
Huntington Beach, CA 92647
CHARLES K. REED
Lockheed Martin Vought Systems
Dallas, TX 75265
Abstract
Samples of International Space Station (ISS) thermal control coatings were exposed
to simulated low Earth orbit (LEO) environmental conditions to determine effects on
optical properties. In one test, samples of the white paint coating Z-93P were coated
with outgassed products from Tefzel® (ethylene tetrafluoroethylene copolymer)
power cable insulation as may occur on ISS. These samples were then exposed,
along with an uncontaminated Z-93P witness sample, to vacuum ultraviolet (VUV)
radiation to determine solar absorptance degradation.
The Z-93P samples coated with Tefzel outgassing products experienced
greater increases in solar absorptance than witness samples not coated with Tefzel
outgassing products. In another test, samples of second surface silvered Teflon " FEP
(fluorinated ethylene propylene) , Sit), (where x::::o2)-coated silvered Teflon FEP, and
Z-93P witness samples were exposed to the combined environments of atomic
oxygen and VUV radiation to determine optical properties changes due to these
simulated ISS environmental effects. This test verified the durability of these
materials in the absence of contaminants.
18 1
J.J. Kleiman and R.C. Tennyson (eds.},
Protection ofSpace Materials from the Space Environment , 181-195.
© 2001 Kluwer Academic Publishers. .
182
1.0 Introduction
In low Earth orbit (LEO), solar ultraviolet (UV) radiation and atomic oxygen can
cause changes to the solar absorptance of many thermal control coatings. UV
radiation can cause darkening, indicated by an increase in solar absorptance, of
some polymers and oxides commonly used in thermal control materials. Atomic
oxygen can cause erosion of polymer materials, but may reverse the UV-darkening
effect on oxides. Maintenance of low solar absorptance and high thermal emittance
of thermal control coatings is important to assure the proper operating temperature
of thermal control systems. Thermal control coatings to be used on radiator surfaces
of the International Space Station (ISS) were evaluated for their performance upon
exposure to simulated LEO environments of solar vacuum ultraviolet (VUV)
radiation (wavelength range 115-200 nm) and VUV combined with atomic oxygen
(AO). Solar absorptance measurements were made in vacuum in order to eliminate
the "bleaching effect," or reversal of degradation due to ambient oxygen exposure,
on VUV-induced degradation.
For comparison , solar absorptance measurements were also made in air
before and after the exposure tests. For some samples, thermal emittance was also
measured in air. The objectives of these experiments were to determine changes in
optical properties for various thermal control materials due to exposure to simulated
LEO conditions similar to those expected for ISS. In one test, samples of the white
paint thermal control coating Z-93P (zinc oxide pigment in potassium silicate
binder) were coated with outgassed products from a radiation crosslinked Tefzel
(ethylene tetrafluoroethylene copolymer) to simulate possible on-orbit contamination
from power cable insulation . Samples were then exposed to VUV radiation to
determine the solar absorptance degradation that would be expected due to VUV-
induced darkening of both the Z-93P and the contaminant layer. This test
represented a worst-case scenario for Z-93P solar absorptance degradation because
AO, which may remove some of the contaminants and may reverse the darkening
due to vacuum ultraviolet radiation, was not present during testing. In another set of
tests, samples of silvered Teflon FEP (fluorinated ethylene propylene), SiOx (where
x :::::2)-coated silvered FEP Teflon, and white paint Z-93P coatings were exposed to
AO combined with VUV radiation in order to verify the solar absorptance and
thermal emittance durability of these thermal control materials selected for use on
the ISS. These samples were not exposed to contamination prior to AO-VUV
testing.
2.0 Experimental Procedures
2.1 VUV AND COMBINED AO-VUV EXPOSURE FACILITY
The facility that was used in these experiments for VUV exposure and combined
AO-VUV exposure is described in [IJ and [2J. Four samples of maximum size 2.5
183
em diameter and 0.635 em thickness can be exposed to atomic oxygen and VUV
radiation concurrently or to the individual environmental effects. These four
samples can be measured, in vacuum, for total hemispherical reflectance. A
schematic diagram of the facility is given in Figure 1.
CIRCULATOR/OIRECTlONAL COUPLER/ OUMMY LOAO

MICROWAVE POWER HEAO
THREE STUB TUNER

ELECTROMAGNET
ELECTRON-CYCLOTRON
RESONANCE SOURCE
ELECTROMAGNET
LOW ENERGY
OXYGEN BEAM
DEUTERIUM LAMPS
j~
GLA SS ENCLOSURE
(REMOVABLE FOR
DIRECTED BEAM)
LIGHT SOURCE
- - - - -- -1.71 m _ _----'-,....-_ _-01°1 MONOCHROMATOR
~ 3-AXIS SAMPLE
POSITIONER
Figure I
Atomic oxygen-vacuum ultravioletradiationexposure facility
An ECR (electron cyclotron resonance) source is used to provide thermal

energy atomic oxygen. To avoid direct line-of-sight to the samples from the ECR
source, which also provides intense VUV at the 130 nm oxygen resonance line, a
fixture consisting of glass surfaces and an aluminum foil liner are used to scatter
and redirect the atomic oxygen. Mass change of Teflon FEP, whose in-space erosion
yield has been characterized as 3.64xlO-25 cnr'zatom [3], was used to determine
equivalent atomic oxygen flux and fluence during each AO-VUV exposure test. A
pair of deuterium lamps was used for VUV exposure. These lamps are described in
section 2.2. The in-vacuum reflectance measurement capability is described in
section 2.3.2. Facility pressure was approximately 4xlO-6 to 10-5 torr during VUV
184
exposure of samples without AO, and approximately 4xlO·4 torr during combined
VUV and AO exposure.
2.2 VUVLAMP AND DETECTOR CALIBRAnON PROCEDURES
For each test, a pair of 30 watt deuterium lamps (Hamamatsu model L879-0I) with
magnesium fluoride windows were used to illuminate samples in a four-sample
holder located inside the vacuum chamber as shown in Figure 1. The distance
between the lamp's anode and the centre of the sample holder was approximately
49.5 em where distance was measured along the axis of the lamp nose. The bodies of
the lamps were located outside of the vacuum chamber with their end-tubes pushed
through o-ring compression fittings to provide a vacuum seal. The lamps' end-tube
windows were thus located inside the vacuum chamber. Lamp bodies were cooled
with forced air.
Prior to and following each test, the output of the pair of lamps was
measured in vacuum using a cesium iodide (Csl) phototube (Hamamatsu Model
RII87) with a magnesium fluoride window. The CsI phototube was attached to the
four-sample holder with its axis along the lamp noseaxis with the sensor at the level
of the sample surfaces. The phototube was then placed at each of the four sample
sites to obtain signal at that position. This phototube is sensitive in the wavelength
range between 115and 200 nm, the range of primary outputof the deuterium lamps.
Prior to measuring the output of the exposure lamps, the CsI phototube was
calibrated at 25.4 cm distance with a Hamamatsu L879-01 lamp used as a reference
standard. This lamp had been calibrated at 25.4 cm distance by the National
Institute of Standards and Technology (NIST). The intensity obtained by NIST was
divided by the Csl phototube-measured signal for the reference standard lamp to
determine a calibration factor for the detector for the wavelength range of 115-200
nm. Detectorcalibration data is shown in Table 1.
Table 1. VUV Detector Calib ration Data
Parameter Value
Intensity of NIST-calibrated V1.IV lamp. 25.4 cm
distance. 115-200 nm 3.39xlO -4 W/em 2
Intensity of the sun. air mass zero. 115-200 nm 1.0378x10 ·l W/cm 2
Signal ofNIST-calibratcd V1. IV lamp measured with
CsI detector. 25.4 em distance 1.68 X 10.6 A
Calibr ation factor for Csl detector 201.79 W/(cm 2 • A)
For each pair of lamps measured using the calibrated CsI detector, the
signal obtained in amps was multiplied by the detector's calibration factor to give
the lamp intensity in the wavelength range 115-200 nm. Dividing the lamps'
intensity by the sun's intensity in the same wavelength range provided the number
of suns in the 115-200 nm range. For each VUV exposure test, equivalent sun hours
(ESH) of VUV exposure in the 115-200 nm wavelength range were obtained using
the following equation:
185
(ESH VUV)1 I5-200 nm = (No. of SUnS)II5-200nm x (Exposure duration , hrs.) (1)
Measured lamp intensity data and ESH for each of the exposure tests will be
given in sections 3.1 and 3.2.
2.3 OPTICAL PROPERTIES CHARACTERlZATION PROCEDURES
2.3.1 In-Air Solar Ab sorptance Measurements

Prior to and following VUV or AO-VUV exposure tests, samples were measured for
total hemispherical reflectance in air between 250 and 2500 nm using a Perkin-
Elmer Lambda-9 spectrophotometer equipped with a 15 em integrating sphere. A
Labsphere Spectralon reference standard was used. Solar reflectance, Ps, was
obtained by integrating spectral reflectance over the air mass zero solar spectral
irradiance distribution using standard procedures [4]. Because samples were opaque,
solar absorptance, 0.., was obtained using the following equation:
o.s=l-ps (2)
Error in repeatability of 0. , measurements with the Lambda-9 instrument is

expected to be approximately ± 0.005.
2.3.2 In-Vacuum Reflectance Measurement System : Calibration and Use

The reflectance measurement system, Optronic Laboratories Model 746 (OL 746), is
described in detail in [2]. This system measures total hemispherical spectral
reflectance in the wavelength range from 280 to 2500 nm. Calculations of solar
absorptance were made in this wavelength range using standard procedures [4] and
solar absorptance was calculated using Equation 2.
Because the OL 746 is a single beam instrument , a known reference is
measured in order to calibrate the system prior to measuring test samples.
Calibration factors as a function of wavelength, A, are obtained in the form of
(Cal. factor), = (Reference Reflectance) A. /(measured detector signal) A. • (3)
The reference sample used for these experiments was a sample of YE-71
(white paint thermal control coating comprised of zinc orthotitanate pigment in a
potassium silicate binder) on aluminum, which was located in the integrating sphere
attachment of the OL 746 and protected from VUV and AO exposure. The absolute
reflectance of this sample was previously measured in air using the Perkin-Elmer
Lambda-9 instrument as described in section 2.3.1. These in-air data were used in
Equation 3 as the (Reference Reflectance) A. values.
Samples in the four-sample holder are transported from the AO-VUV
exposure end of the chamber to the measurement end of the chamber using a three-
axis positioner to which the sample holder is mounted, as shown in Figure 1.
186
Because samples slide into place under the integrating sphere for measurement, a
small gap is needed between the integrating sphere port and the sample surface.
This can introduce error in the measurement of reflectance due to loss of light from
this gap. In order to reduce this error, the pre-test, in-air reflectance measurement of
each sample obtained with the OL 746 was compared to an in-air measurement
obtained with the Perkin-Elmer Lambda-9 where samples were in intimate contact
with the integrating sphere. For these in-air data, a set of correction factors was
generated for each sample so that the reflectance at each wavelength measured with
the OL 746 instrument was corrected to the value given by the Lambda-9. These
spectral correction factors were applied to each subsequent reflectance spectrum
obtained with the OL 746 instrument. Error in repeatability of o , measurements
with the OL 746 instrument is expected to be approximately ± 0.0I.
2.3.3 In-Air Thermal Emittance Measurements

A Gier-Dunkle model DB-100 portable emissometer was used to measure room
temperature total emittance, E, normal to sample surfaces. The instrument provides
an integrated value of normal emittance over the wavelength range of 5 to 25 urn.
Error in repeatability of measurement of E is expected to be approximately ± 0.005.
2.4 TEFZEL-CONTAMINATED Z-93P EXPOSED TO VUV WITHOUT AO
2.4.1 Samples
The white paint coating Z-93P was applied to aluminum substrates of 2.38 em
diameter. Coating thickness was approximately 0.10-0.13 mm. Samples of Z-93P
were coated with approximately 100 om of outgassed products from a radiation
crosslinked Tefzel as described fully in [5). Samples of Tefzel were heated to
approximately 149°C under vacuum (10.6 to 10.7 torr) to drive off contaminant
products. Z-93P samples were mounted on a water-cooled sample plate approx-
imately 15.2 cm from the contamination source along with a quartz crystal
microbalance (QCM) used to measure thickness of the deposited contaminants. A
low pressure mercury near ultraviolet (NUV) lamp illuminated the sample plate,
including the QCM, during and after the deposition for a total NUV exposure of 16
hours. It was expected that NUV would fix the contaminants and reduce re-
evaporation of the contaminants from the sample plate. Samples purposely
contaminated with Tefzel outgassing products in this manner were labeled 365-1,
365-2, and 365-3. A sample of Z-93P labeled A-100 was not exposed to Tefzel
outgassing products and was used as a witness sample during VUV exposure testing.
Although this sample was prepared from a different Z-93P paint batch than those
which were coated with Tefzel outgassing products, it was expected to be a typical
Z-93P sample comparable in VUV-durability to the 365-1, -2, and -3 samples in
order to determine the effects of the Tefzel contamination.
187
2.4.2 VUV Exposure and Sample Characterization Procedures

The three Tefzel-contaminated Z-93P samples and one witness Z-93P sample were
exposed in vacuum to four increments of VUV exposure. Total test duration was 234
hours. In addition to in-air reflectance data obtained before and after the test, in-
vacuum reflectance data were obtained prior to VUV exposure and immediately
after each increment. The VUV lamp pair used to illuminate the four sample holder
was calibrated prior to the first exposure increment and after the final increment.
VUV intensity and equivalent sun hours (ESH) for each sample will be provided in
section 3.1.
2.5 AO-VUV EXPOSURE OF SILVERED TEFLON AND SIOx-COATED

SILVERED TEFLON
2.5.1 Samples
Two separate tests were conducted to evaluate uncoated and SiOx (where x "" 2)-
coated silvered Teflon FEP. In AO-VUV Test 1, three samples of SiOx-coated
silvered Teflon FEP labeled SST-I , SST-2 and SST-4 and one Z-93P sample labeled
ZPLM-l were tested. In AO-VUV Test 2, three samples of uncoated silvered Teflon
FEP labeled UST-2, UST-3 and UST-4 and one Z-93P sample labeled ZPLM-5 were
exposed. Table 2 shows the composition of each sample including the thickness of
each layer. All samples were 2.38 em diameter.
Table 2. Samples for AD -VUV Exposure
Test Sample Sample Layer I Layer 2 Layer3 Layer 4 Layer 5

Position Label on AI substrate
AO-VUV I SST-I acrylic adhesive Inconel~ Silver TeflonFEP SiD,
Test I 27.5 nm 100nm 0.140mm 115 nm
2 SST-2 acrylic adhesive Inconel Silver Teflon FEP SiO,
27.5 nm )OOnm 0.140mm 115nm
3 ZPLM-) Z-93P - - - --
0.10 to 0.13 mm
4 SST-4 acrylic adhesive lnconel Silver Teflon FEP SiD,
27.5 nm 100nm 0.140mm 115 nm
AQ-VUV I UST-3 acrylic adhesive Inconel Silver TeflonFEP -
Test 2 27.5 nm 100nm 0.254mm
2 UST-2 acrylic adhesive lnconel Silver Teflon FEP -
27.5 nm 100nm 0.254rnm
3 ZPLM-5 Z-93P - -- - -
0.IOtoO.13 mm
4 UST-4 acrylic adhesive lnconel Silver Teflon FEP -
27.5 nm 100nm 0.254mm
2.5.2 Exposure Procedures

In AO-VUV Test 1, the samples described in Table 2 were exposed to AO and VUV
radiation simultaneously in four increments . Total test duration was 260.2 hours. In
AO-VUV Test 2, the samples described in Table 2 were exposed to AO and VUV
radiation simultaneously in three increments. Total test duration was 250.7 hours.
188
In addition to in-air reflectance data obtained before and after each test, in-vacuum
reflectance data were obtained prior to AO-VUV exposure and immediately after
each increment. The VUV lamp pair used to illuminate the four-sample holder was
calibrated before and after Test 1. Test 2 was conducted immediately following Test
I; therefore, the calibration data following Test I is valid for the pre-test calibration
prior to Test 2. VUV lamp output following Test 2 was not measured due to
equipment problems; however, it was estimated . VUV intensity, VUV ESH, AO flux
and AO fluence for each test will be provided in section 3.2.
3.1 TEFZEL-CONTAMINATED Z-93P EXPOSED TO VUV WITHOUT AO
3. J.J VUV Lamp Calibration

The measured number of VUV suns for each sample position, before and after VUV
exposure testing, are shown in Table 3. It is evident that the VUV lamps
experienced a significant loss of intensity due to the total 234 hours of use. It is
likely that contaminants present in the facility, probably due to perfluorinated
polyether vacuum pump oil, were deposited onto the MgF2 windows of the VUV
lamps and were "fixed" onto the windows due to the intense VUV radiation formed
inside the lamp.
Table 3. Measured Number ofVUV Suns at Each Sample Exposure Position
Exposure Sample No. VUVSuns No. VUV Suns

Position Sampl e Typ e Label Pre-test Post-test, 234 hrs
1 Tefzel f-contaminated Z-93P 365-1 8.8 l.l
2 Tefzel-contarninated Z·93P 365 -2 10.5 1.7
3 Z-93P witness A-IOO 13 1.2
4 Tefzcl-contaminated Z-93 P 365 -3 10.6 2.0
It was assumed that the lamp intensity degraded following an exponential

decay function as shown in Figure 2. The number of suns before and after each of
the four exposure increments was interpolated from the exponential decay curve fit
of the data for each exposure position. Then the average number of suns for each
increment was obtained from these data. Table 4 shows the curve fit equations and
the calculated average number of suns for each exposure increment.
189
14 , . - - - - - - - - - - - - - - - - - - - - ,
12
10 -&- PositionI
-s - Position2
-<lo- Position3
J 8
.. <)- Position4
~ 6
0+------..-----r------,-----r-------1
o 50 100 150 200 250
OperatingTime (h)
Figure 2
VUV lamp intensity as a function of operating time
Table 4. Calculated Number of VUV Suns
Sample Holder Position I 2 3 4

Samnle Label 365-1 365-2 A-100 365-3
Exponential decay curve y= y= y= y=
fit, exp(2 .2- exp(2.3 - exp(2 .6- exp(2.4-
x=operating hours. 0.0088x) 0.0077x) 0.0 lOx) 0.007Ix)
y=no. of VUV suns
Avg. no. VUV suns

Inc. I 7.2 8.9 10.4 9.0
Inc. 2 4.7 6.1 6.4 6.4
Inc. 3 3.1 4.2 3.9 4.6
Inc. 4 1.8 2.6 2.1 2.9
3.1.2 Solar Absorptance Measurements

The solar absorptance values in the wavelength range 280-2500 nrn measured in
vacuum at each VUV exposure increment are shown in Figure 3. Table 5 shows
solar absorptance values before and after VUV exposure and the overall change in
solar absorptance for each sample. Also shown are changes in solar absorptance per
ESH of VUV exposure.
As indicated in Table 5. the change in solar absorptance per ESH for
Tefzel-contaminated Z-93P is 1.9 to 2.7 times that of the uncontaminated Z-93P
witness sample. Therefore. it can be concluded that deposition of the outgassed
products of Tefzel onto Z-93P will cause degradation in solar absorptance of Z-93P
190
to occur at a faster rate than solar absorptance degradation of Z-93P which has not
been exposed to Tefzel outgassing products. Equivalent ISS radiator VUV exposure
years are shown for reference. However, it is not valid to assume that these would be
the actual changes in solar absorptance for the equivalent ISS radiator LEO
exposure duration, because other LEO environmental effects to which ISS radiators
will be exposed were not simulated here. This test provides a worst-case scenario for
solar absorptance degradation of Tefzel-contaminated Z-93P, particularl y because
atomic oxygen, to which many of the thermal control surfaces on ISS will be
exposed, was not present during these tests. Other testing has shown that, following
VUV exposure of Tefzel-contaminated Z-93P samples, atomic oxygen exposure
causes reversal of solar absorptance degradation [5].
Table 5. In-Vacuum Measurements: Effect ofVUVon Solar Absorptan ce of Tefzel-Contarninated Z-93 P and
Z-93 P Witness Sample
Total VUV Equivalent Solar Absorptanc e, a,

Sam ple Exposure. ISS Radiator Before After t1a, t1a,!ES H
ESH VUV Exposure, Exposure Exposure
years
Tefzel-contam ., 1670 4.9 0.154 0.208 0.054 3.2x I0 "
36 5- 1
Tefzel-contam.. 2234 6.5 0.156 0.207 0.05 1 2.3xlO ·5
36 5-2
Tefzel-contam ., 2397 7.0 0.168 0.236 0.068 2.8xlO "
365-3
uncontaminated, 2167 6.3 0.13 3 0.159 0.026 L2 xlO"
A- IOO
0.24
0.22
--
I
Sl
0.2
0.18
~
~ 0.16 ~ -_ .---- - . ---- -£

!il _ --- - A: ~
'0
r:J]
0.14
0.12
--- t - -- 365-2
-D--365-1
A-100
--&-365-3
OJ
0 500 1000 1500 2000 2500
VUVExposure, ESH
Figure 3
Solar absorptance ofTefZel-contaminated Z-93 P samples (365 -1, -2, and -3)
compared to an uncontaminated Z-93 P witness sam ple (A-I 00)
191
Table 6 shows solar absorptance values for the wavelength range 250-2500 nm
measured in air for the Tefzel-contaminated Z-93P samples and the Z-93P witness
sample before and after VUV exposure.
Table 6. In-Air Measurements : Effect ofVUV on Solar Absorptance of Tefzel-Contaminated Z-93P and Z-93P
Witness Sample
TotalVUV Solar Absorptance, a,

Sample Exposure, Before After M, ~a,!ESH
ESH Exposure Exposure
Tefzel-contarn ., 365-1 1670 0.144 0.182 0.038 2.3xl0"
TetZel-contam.,365-2 2234 0.147 0.183 0.036 1.6xl0·5
Tefzel-contam ., 365-3 2397 0.146 0.188 0.042 1.8xlO·5
uncontaminated, A-I00 2167 0.111 0.132 0.021 0.96xlO·5
A comparison of the I!.as values between the in-air data (Table 6) and in-
vacuum data (Table 5) indicate that the in-air values are lower. This difference is likely
due to bleaching , or reversal, of VUV-induced darkening upon exposure of the samples
to ambient conditions. Such bleaching would cause a reduction in the solar absorptance
upon removal of samples from the vacuum chamber. The initial (before exposure) in-
vacuum o, values for the samples are higher than the in-air values, indicating an
increase upon exposure to the vacuum environment. While instrument error may
account for some of the deviation, the consistency of an increase in a s suggests another
effect, possibly facility contamination.
3.2 AO-VUV EXPOSURE OF SILVERED TEFLON AND SIOx-eOATED

SILVERED TEFLON
3.2.1 VUV Intensity and AO Flux Calibration

Table 7 provides VUV lamp calibration data for the AO-VUV exposure tests. As
indicated, the lamp intensity after AO-VUV Test 1 was higher than that prior to the
test by approximately 10-20 per cent. In general, VUV lamp output does not increase
with use as was observed here. Instead, lamp output has been observed to degrade.
Degradation is minimal when lamps are used in the presence of AO, because
contaminants are removed by AO as fast as they can be UV-fixed onto the lamp
windows. The increase in output observed during AO-VUV Test 1 is expected to be
due to AO-removal of contaminants that may have been present on the lamp windows
prior to use.
Because only two data points were obtained, it was not possible to determine
whether the increase in lamp output occurred immediately or gradually. Because the
change in output was fairly small, it was assumed to follow a linear increase with time
for purposes of calculating the average number of suns for each exposure increment as
shown in Table 7. It was further assumed that by the conclusion of Test 1, all
contaminants were removed from the lamp window surfaces and that further use
during Test 2 would cause lamp output to degrade following typical observations of 5
per cent intensity loss during a 250 hour test. Because facility problems prevented
192
VUV intensity calibrations from being conducted following Test 2, the average number
of suns for this test were calculated using this assumption . Calculated values for the
average number of suns during Test 2 are shown in Table 7.
Tabl e 7. Number ofVlIV Suns at Each Sample Position for AO-VllV Tests
Sample No. VUV No. Vl N Percent Avg. No.

Position Suns Suns Increase in of Suns,
Before After No. Suns, AO-VUV Avg, No. of Suns,
AO-VlN AO-VlIV AO-VllV Test 1 AO-VUV Test 2
Test I Test I Test I
I 3.0 3.6 19.4 3.3 3.5
2 3.9 4.5 16.5 4.2 4.4
3 4.9 5.4 9.7 5.1 5.2
4 3.7 4.0 9.6 3.9 3.9
The effective atomic oxygen flux values during AO-VUV Test 1 and Test 2
were obtained by measuring the mass of Teflon FEP witness samples before and after
each test. Test I flux was measured to be 4.73x10 15 atoms/em's, and Test 2 flux was
measured to be 4.62x10 15 atoms/em's ,
3.2.2 Optical Properties Measurements

Results of in-vacuum solar absorptance measurements for samples exposed in AO-
VUV Test I and Test 2 are shown in Table 8 and Table 9, respectively. In-air
measurements for both tests are shown in Table 10. The purpose of these tests was to
determine the effects of combined AO and VUV on SiOx-coated silvered Teflon and
uncoated silvered Teflon as compared to Z-93P. Because each sample received a
somewhat different ratio of AO fluence/VUV ESH, for simplicity in interpreting data,
results are displayed in tabular form and not in graphical form. Based on the ~as
values, there are no consistent increases or decreases in solar absorptance with
increasing AO-VUV exposure. The values of ~as, indicate changes that are within
experimental error of the measurement throughout the tests. For the coated and
uncoated silvered Teflon materials and Z-93P, it was verified that at AO and VUV
fluences representative of 2.5 to 4 years for ISS radiators, there are no significant
changes in solar absorptance. In-air measurements also indicate changes within
experimental error of the measurements.
193
Table 8. In-Vacuum Measurements: Effect of AO and VUV on Solar Absorptance

ofSiOx-coated Silvered Teflon (SST) and Z-93P (ZPLM) Samples
Sample VUV AO Fluence ISS Radiator ISS Radiator o, 8.a.,

ESH (atoms/ern") Equivalent Equivalent from to
VUVyears AO years
SST-I 0 0 0 0 0.092
156 8.12x 10 2O 0.45 0.74 0.081 -0.011
462 2.40 x10 21 1.34 2.18 0.088 -0.004
46 5 2.4lx10 21 1.35 2.19 0.091 -0.001
853 4.43x10 21 2.48 4.03 0.087 -0.005
SST-2 0 0 0 0 0.093
201 8.12x 10 2O 0.58 0.74 0.081 -0.012
593 2.40x10 21 1.72 2.18 0.091 -0.002
597 2.4lx10 21 1.73 2.19 0.083 -0.010
21
1095 4.43x10 3.18 4.03 0.089 -0.004
SST-4 0 0 0 0 0.102
2O
184 8.12x10 0.53 0.74 0.090 -0.0 12
21
544 2.40x10 1.58 2.18 0.094 -0.008
547 2.4lx1021 1.59 2.19 0.096 -0.006
1004 4.43x10 21 2.92 4.03 0.086 -0.016
ZPLM- I 0 0 0 0 0.148
2O
245 8.12x10 0.71 0.74 0.150 0.002
21
724 2.40x10 2.10 2.18 0.160 0.012
728 2.4lxI0 2 \ 2.12 2.19 0.151 0.003
1337 4.43xI0 21 3.89 4.03 0.151 0.003
Table 9. In-Vacuum Measurements: Effect of AO and VUV on Solar Absorptance

of Silvered Teflon (llST) and Z-93P (ZPLM) Samples
Sample VUV AO Fluence ISS Radiator ISS Radiator a., 8.a.,

ESH (atoms/em' ) Equivalent Equivalent AO from to
VUV years years
llST·2 0 0 0 0 0.050
287 1.08x10 2 \ 0.83 0.99 0.064 0.014
785 2.97xI0 21 2.28 2.70 0.052 0.002
1103 4.17xI0 2 \ 3.2 1 3.79 0.056 0.006
llS T-3 0 0 0 0 0.057
228 1.08x10 21 0.66 0.99 0.067 0.010
624 2.97x10 21 1.81 2.70 0.056 -0.001
878 4.17x10 21 2.55 3.79 0.053 -0.004
ll ST-4 0 0 0 0 0.050
254 1.08xl 0 21 0.74 0.99 0.064 0.014
695 2.97x1021 2.02 2.70 0.057 0.007
978 4. 17x1021 2.84 3.79 0.055 0.005
ZPLM-5 0 0 0 0 0.119
339 1.08x10 21 0.98 0.99 0.121 0.002
927 2.97xI0 2 \ 2.69 2.70 0.124 0.005
1304 4.17x10 21 3.79 3.79 0.118 -0.001
194
Table 10. In-Air Measurements : Effect of AO and VUV on Solar Absorptance ofSiOx·coated
Silvered Teflon. uncoated Silvered Teflon and Z-93P
Sample VUV AO Fluence ISS Radiator ISS Radiator a, ~a,

ESH (atoms/em') Equivalent Equivalent from t,
VUVyears AOyears
SST-[ 0 0 0 0 0.072
853 4.43x[O" 2.48 4.03 0.070 -0.002
SST-2 0 0 0 0 0.072
1095 4.43x[O" 3.18 4.03 0.074 0.002
SST-4 0 0 0 0 0.069
1004 4.43 x10" 2.92 4.03 0.062 -0.007
UST-2 0 0 0 0 0.064
1103 4.17xlO" 3.21 3.79 0.060 -0.004
lIST-3 0 0 0 0 0.065
878 4.17xIO'\ 2.55 3.79 0.060 -0.005
UST-4 0 0 0 0 0.064
978 4.17x10 21 2.84 3.79 0.06[ -0.003
ZPLM-l 0 0 0 0 0.132
1337 4.43x10 21 3.89 4.03 0.125 -0.007
ZPLM-5 0 0 0 0 0.117
1304 4.17)(1021 3.79 3.79 0.110 -0.007
Thermal emittance values for samples exposed in AG-VUV Tests I and 2

were measured as shown in Table II. Samples were measured following AG-VUV
exposure and values were compared with emittances measured for pristine samples.
Table II . In-Air Measurements : Effect of AO and VUV on Thermal Emittance

ofSiOx-coated Silvered Teflon. uncoated Silvered Teflon and Z·93P
Sample VUV AO Fluence ISS Radiator ISS Radiator & As from

ESH (atoms/em") Equivalent Equivalent pristine
VVV years AO years sample
SST-5 0 0 0 0 0.810
SST· I 853 4.43xI0" 2.48 4.03 0.81 I 0.001
SST-2 1095 4.43xI0" 318 4.03 0.809 -0.001
SST-4 1004 4.43x10" 292 4.03 0.808 -0.002
lJST -1 0 0 0 0 0.859
lJST-2 1103 4.17)(10" 3.21 3.79 0.853 -0.006
lJST-3 878 4.17xI0" 2.55 3.79 0.857 -0.002
llST-4 978 4.17xI0" 2.84 3.79 0.857 -0.002
ZPLM-2 0 0 0 0 0.911
ZPLM·I 1337 4.43xI0" 3.89 4.03 0.911 0
ZPLM-5 1304 4.17xI0" 3.79 3.79 0.909 -0.002
All changes in emittance were within experimental error, indicating no

significant changes in emittance due to AG-VUV exposure. A reduction in emittance
for silvered Teflon would be expected if enough of the Teflon material was eroded by
195
atomic oxygen . For the exposure fluences provided in AO-VUV Tests 1 and 2, erosion
was not significant enough to cause emittance degradation.
4.0 Conclusions
VUV-induced solar absorptance degradation of the of Z-93P white paint ISS radiator
coating occurred at a faster rate for samples coated with outgassed products from
radiation crosslinked Tefzel than for samples not coated with Tefzel products when
both were exposed to VUV equivalent sun hours representative of 4.5 to 7 years on
ISS. Tefzel power cable insulation is a possible source of contamination for ISS
radiators. Solar absorptance increases of approximately 0.04 for the Tefzel-
contaminated Z-93P coatings were indicative of worst-case testing where AO, which
could reverse solar absorptance degradation, was not present.
SiOx·coated and uncoated silvered Teflon samples, along with Z-93P witness
samples, were exposed to the combined environments of AO and VUV for ISS radiator
equivalent exposures of 2.5 to 4 years . These samples were not previously exposed to
outgassed products from Tefzel. Samples exposed in these tests showed changes in
solar absorptance and thermal emittance that were within instrument error indicating
durability of these materials to the ISS environment for this level of exposure where
contaminants were not present.
5.0 References
I. Bailey, v., Janicki , G. C., and Haulik, T.. eds., Proceeding s of the 38'h International SAMPE Sympo sium
and Exhibition . Volume 38. Book 1, Stidham. C. R., Stueber, T. J., Banks, B. A , Dever, J. A , Rutledge, S.
K., and Bruckner, E. J., "Low Earth Orbital Atomic Oxygen Environmental Simulation Facility for Space
Materials Evaluation. " pp. 649-663 , Society for the Advancement of Material and Process Engineering,
1993. Also: NASA Technical Memorandum 106128. May 1993.
2. Bailey, V., Janicki, G. C., and Haulik, T., eels.. Proceedings of the 3f3'h International SAMPE Sympo sium
and Exhibition. Volume 38. Book 1. Dever, J. A . Rutledge, S. K., Bruckner, E. J.• Stidham. C. R., Stueber,
T. J., Booth., R. E., "The Effects of Simulated Low Earth Orbit Environments on Spacecraft Thermal Control
Coatings," pp. 694-706 , Society for the Advancement of Material and Process Engineering, 1993. Also:
NASA Technical Memorandum 106146 , May 1993.
3. Banks, B. A, Dever. J. A , Gebauer. L., Hill. C., " Atomic Oxygen Interactions with FEP Teflon and Silicones
on LDEF," LDEF - 69 Months in Space, NASA CP 3134 . Part 2, June 1991, pp. 801·815.
4. ASTM E 903 . " Standard Test Method for Solar Absorptance, Reflectance. and Transmittance of Materials
Using Integrating Spheres." E 903-82 (Reapproved 1992).
5. Hasegawa. M. M.. "VUV Darkening of Z-93P contaminated With Outgassed Products from Tefzel lnsulated
Power Cables," McDonn ell Do uglas Aerospace Report No. MDC 96H0305. January 1996.
SOME ASPECTS OF THE DEGRADATION OF FEP TEFLON MATERIAL
UNDER THE ACTION OF VUV, THE SUN AND GROUND TEST
VUV FACILITIES
A. MILINTCHOUK,
Atox ltd
Moscow, Russia
M. VAN EESBEEK,
EuropeanSpace AgencyIESTEC,
The Netherlands
1.0 Introduction
FEP Teflon® is widely used as a thermal blanket material in space applications . There
is growing evidence, however, of the degradation of FEP Teflon during space flights.
From recent publications [1, 2] it can be concluded that soft X-ray radiation from solar
flares is one of the main factors influencing this degradation. There are also many
ground-based experiments where the degradation of FEP under exposure to VUV-
sources has been observed [3-7].
At the same time, in VUV-tests with the Hamamatsu L879 VUV lamp, no
mechanical degradation was observed after 24 hours of ground testing [I]. The sample
was maintained under a constant purge of dry nitrogen to avoid contami-nation of the
VUV-lamp window and to eliminate the presence of any amount of oxygen. The
behaviour of such exposed FEP samples was identical to the virgin sample as shown by
mechanical testing.
The film surface however was eroded. Initially, no attention was paid to this fact
because etching is a common feature during VUV-tests.
Radiation, such as VUV-radiation in the reaction with materials, partially reflects
and is partially absorbed by the material. The absorbed radiation energy is then partly
spent on irreversible chemical reaction and emissivity; the main part of the absorbed
energy is, however, transformed into heat. This heating process during irradiation is not
something newly reported [8], but this paper will deal with the investigation and
considerat ion of the heat output during VUV-radiation tests. Some details of
calculations and concerns about accelerated VUV-tests interpretations are presented .
197
Protection ofSpaceMaterialsfrom the Space Environment, 197-209.
© 2001 KluwerAcademicPublishers.
198
2.0 Calculation of the heating of FEP Teflon material under sun illumination
Figures 1a and 1b present the sun irradiation spectrum. The ranges of VUV and IR are
presented in Table 1 and Table 2.
'"00
W.m..z . ..."..1
'"
~~
'500
'000
j
,JJ -. <,
r-,
)
n5
-------- -
Figure la
Sun Irradiation spectrabetween 0 and 2 urn
.rrr--:
r-.
1000
W .m~.pm-1
I
'00
10
r <,
-,
/ -, -,
/
0 .1
~/ -,
0.001
0.' '0 ""A 100
Figure Ib
Sun Irradiation spectra between O.I and 100um

199
Table I. VUV Solarspectrum
Wavelength Wavelength Solar spectral irradiance Integrated solar

(nm) (11m) (W/m-2'llm-l) irradiance (W / m·2)
115 0.115 0.007 0.0025
120 0.12 0.9 0.0047675
125 0.125 0.007 0.007035
130 0.13 0.007 0.00707
140 0.14 0.03 0.007255
150 0.15 0.07 0.007755
160 0.16 0.23 0.009255
170 0.17 0.63 0.013555
180 0.18 1.25 0.022955
190 0.19 2.71 0.042755
200 0.2 10.7 0.10981
2.1. ENERGY ABSORPTION OF FEP DUE TO THE ACTION OF SOLAR VUV

RADIATION
For VUV radiation, some experimenters take the range between l l S and ISO nm with
an integrated solar irradiance value of 7.5xl 0,7 W/cm 2, while others go up to 200 nm,
resulting in an integrated solar irradiance 1.07-10'5 W/cm2• In both cases the individual
photon energy is higher than the bonding energy of the C-F bonds, which is 486 kJ/mol
or 5.04 eVlbond.
In the first case, during 45 minutes of solar illumination the heat input is 0.002
J/cm2, while in the case of the 115-200 nm approach , this value is 44 times higher or
2
0.089J/cm •
For the calculations, one also needs to know at least the FEP absorption spectrum.
Unfortunately, the absorption spectrum ofFEP is unknown in the VUV region.
The PTFE spectrum shows an intense absorption band at 160 nm with a weak
absorption tail towards the low energy side and an increasing absorption continuum at
wavelengths less than 125 nm. The extinction coefficient at 160 nm is around 104_
105cm-1 and at 125 nm, it is 10 times less [3].
2.2. ENERGY ABSORPTION OF FEP DUE TO THE ACTION OF SOLAR IR

RADIATION
FEP is transparent in the visible region of the spectrum. The infrared spectrum of FEP
is presented in Figure 2. In the IR range, FEP has absorption bands at 4.3 urn, between
7.7 um and 8.8 urn, at 10.2 um and at 13.2 um. The absorption coefficient ofFEP in
the IR-region is calculated using the IR spectra of 25, 50 and 100 urn thick film
material. This coefficient is given in Figure 3.
The absorption coefficient of FEP at 8-9 urn is highest with a = 3x10-3 nm - I,
while the illumination by the sun at 8-9 microns is 0.5 Watt/rrr'. The absorption
200
Table 2. [R SolarSpectrum
Wavelength (urn) Solarspectral irradiance Integrated solar irradiance
(W / rn-z- um-I) (W Im-2 )
4.3 7.[ 1342.7
4.4 6.5 1343.4
4.5 5.92 1344
4.6 5.35 1344.6
4.7 4.86 1345.[
4.8 4.47 1345.6
4.9 4.[1 1346
5 3.79 1346.4
6 1.82 1349.2
7 0.99 135Q.6
8 0.585 1351.4
9 0.367 1351.9
10 0.241 1352.2
II 0.[65 1352.4
12 0.[ 17 1352.5
13 0.0851 1352.6
14 0.0634 1352.7
15 0.0481 1352.7
16 0.0371 1352.8
17 0.0291 1352.8
18 0.0231 1352.8
19 0.0186 1352.9
20 0.0152 1352.9
25 0.0061 7 1352.9
30 0.00297 1353
35 0.0016 1353
40 0.000942 1353
50 0.00391 1353
60 0.0019 1353
80 6.16E-05 1353.1
100 2.57E-05 1353.1
120 1.26E-05 1353.1
150 5.23E-06 1353.1
200 1.69E-06 1353.1
250 7.00E-07 1353.1
300 3.40E-07 1353.1
400 1.IOE-07 1353.1
1000 0 1353.1
201
coefficient for FEP from 9 microns up to 15 microns is around a =10·4_10 -5 nm" and the
corresponding illumination is 0.6 Watt/m 2• The absorption value is 8.37xI0-5 W-cm".
One crrr' of FEP absorbs 0.226 J energy in the IR-region during 45 minutes of direct
sun illumination .
~ .~ ~
~
\i l ~
., • J \ . ' ~,
.
:~
/ ",to'
v ,1 /1
0.8
',,("1
, J\
~- 05 MIL
\/
"
0.6 - -1 MIL I h
.. .. ' 2 MIL
!, i ' '\h~
r
, f .I
, " \I.\1.J1
:;
0.'
.
n
~
,
0.2
I{
I
V.
o
4000 3300 JOOO 2500 2000 1300
Figure 2.
Infrared Spectrum ofFEP for 0.5,1 and 2 mil
Taking into account the absorption spectrum of FEP, Kirchhoff's radiation law and
an emissivity coefficient in the IR-region for FEP of 0.8 (average of absorption and
emission from 8 to 15 urn, where FEP absorbs), one can calculate that lcnr' of FEP
absorbs 0.226J·0.2 = 0.0452 J and emits the rest.
2.3. CALCULATION OF FEP HEATING
From 2. I and 2.2 it can be found that during 45 minutes of direct sun illumination
of an area of I cm 2 receives 4.72xI0-2 1. From the expression
dT = dQ/Cp.m ( 1)
where
dT = increase of temperature;
Cp = specific heat of FEP, 1.17 J/g ;
m = mass of FEP material under 1 cm2 surface
the temperature increase for a 125 micron film can be calculated for spacecraft in LEO,
such as MIR.
202
0.001
a (nlll)
!
I r,
I I
0.0001
/"1_
I
~,~V
/l.) I
10"
I I I
2 4 6 8 10 12 14 16 18
Figure3.
Absorption coefficient of FEP in IR-region

203
This increase in temperature, only related to the radiative heat input from the sun,
is no more than a few K. During the next shadow period the heat absorbed in the
material will be diffused in the bulk of thermal blankets or will be emitted into space.
Heating of FEP blankets is, in such a case, negligible.
3.0 Calculations of the heating of FEP Teflon material during ground VUV
tests
3.1. SOME FEATURES OF VUV FACILITIES
Some characteristics of the sources frequently used in ground VUV simulation

testing are presented in Table 3.
Table3. Summaryof data on the intensities of varioussourcesused in the presentwork(see also 3.3.)
Irradiation Source
Kr- Xe-Iamp LDD- LD-3 Hamamatsu YUY-
lamp 400 lamp simulator
PhTlNT
Maximum Deposited 1.6·E-3 U5'E-3 2.7·E-3 2.7·E-3 3.1-5.2·E-5 3.0·E-4
Energy,W/cm2
Photons, nm 123.6 147 >177 >115 115-400 5-200
Temperature rate in 63 53 10.7 107 1.23-2.06 2.22
100nm layer,Kls
Temperature shift in 11.78 9.9 2.0 20 0.23-0.38 2.22
top IOOnm layerof 125
micronFEP film
1. Kr-lamp light at 123.6 nm; 1= 1014_10 15 photons-cmi-s" [6]

2. Xe-lamp at 147 nm (E= 8 eV). Intensity I = 1014_10 15 photons-cm'i.s" [6]
3.LDD-400 light longer than 177nm (E<7 eV); 1= I j'cm,2 'hou{1 [6]
4. LD-3 light longer than 115nm (E< 10.78 eV). 1= 10 Jocm,2·hou{l. [6]
5. Hamamatsu lamp L879 3.3x10 13 photons or 2.6x10 14eV cmi-s". [9,10]
6. VUV-simulator PhTINT Light 5-200 nm. Intensities up to 1.08 j'cm,2'hou{1 [11,12],
Different facilities use different sources with different energy ranges or photon
energy, and with different fluences .
3.2. CALCULATIONS OF FEP MATERIAL TEMPERATURE CONDUCTIVITY
First, one should know how fast heat and temperature distribution in FEP material
occurs, assuming a heat input by radiation from the front.
The heat conductivity (~) of FEP is 0.209 j·S,I·m,I·K,I . The thermal diffusivity
(temperature conductivity, n) ofa material is given by:
204
ct.= yCp'p , (2)
where
p == density.
From this formula, the temperature conductivity for FEP is found to be 8.31'10-4 cm2/s.
From formula [8]
(3)
one can calculate the period of time (t) for the temperature expansion in a sample with
thickness h :
(4)
For instance, assuming a 125 micron thick film, this period of time is 0.187 sec.
This value can help in estimating the degradation process in FEP.
3.3. ESTIMATION OF TEMPERATURE RATE IN THE SURFACE LAYEROF

FEP RESULTING FROM THE ACTION OF DIFFERENT VUV SOURCES.
The attenuation length in FEP for VUV photons is less than 10 nm. For instance,
photons with an energy Eph == 30eV have an attenuation length of 12.9 nm, and photons
with Eph=50 eV have a penetration length of22.5 nm (see Figure 4.).
0.1
\
0.01
~.. ........
........................ ..
.
.........................
0.001
"
.,., ......, .., ..
00001
50 100 150 200
l(nm)
Figure 4.
Attenuation depth in FEP

205
If one assumes an average depth of penetration for these VUV photons of 100nm,
then one can calculate the volume and mass of FEP under I crrr' surface, taking into
account a density of2.15 g-em" .
v = Icm2.J OOnm = Jcm 2'1 0,5 cm = 10'5 ern"; (5)

m = 10'5 cm 3·2.15g.cm, 3= 2.J5·IO'5 g.
The specific heat for FEP is 1.17 J g'\·K· 1• A layer with a thickness of 100nm and
an area of l crrr' (mass is 2.15·IO'5g) has a specific thermal capacity Q=2 .52·IO'5 J/K.
This means that in order to change the temperature on this layer with a thickness of J00
nm with I K, the material needs to absorb 2.52'10'5 1.
According to Ref. 6, the fluence or intensities of the Xe-lamp vary between F=IO\4
photons-cm'<s" and F=2.4-1O'4 photons-emis" . It means an energy density variation
between 1.35'10,4 J·cm'2·s'\ or 3.24'10'4 J·cm'2· s, l . The corresponding temperature
increases are then dT = 1.35'10,4 / 2.52'10'5= 5.36 Ks'! and dT = 3.24'10 '4/2.52'10'5=
12.86 Ks'\.
Similarly, one can calculate the temperature increases due to the other UV-sources:
- in the case ofLDD-400, dT = IIKs'1
- in the case ofLD-3 , dT=JIO Ks".
Hence, one has a top layer where the heating process is very sharp and the rest of
the volume which will absorb the heat that is lost through conduction out of this top
layer.
From this calculation, we can assume that several seconds after the start of a VUV-
test, the etching process should begin. But the presence of the bulk of the material ,
which absorbs heat as well, is cause for a delay in the start of the FEP Teflon material
etching process. Furthermore, taking into account the period of time needed to have a
homogeneous temperature distribution, as calculated earlier in 3,2, one can roughly
estimate what will be the minimal temperature increase that a top layer of 100 nm in a
125 micron thick film "will see". It means that one has a top layer, where all radiation
has been absorbed and where the temperature is at least higher than the bulk values
presented in Table 3.
If the top layer (i.e. the layer in which the radiated energy is absorbed) is, for
example, smaller by a factor of J0 (i.e, J0 nm), the rise in temperature will be 10 times
higher. Every source has its own specific emission spectrum, and thus its own
absorption profile in the material, which will define the actual temperature rise in the
material.
3.4. ESTIMATION OF THE ETCHING PROCESS INDUCTION TIME
The etching process starts after some induction time . In each particular case, this
induction time is different.
A prediction of the induction time could be made taking into account a melting
point of FEP Teflon material at 275-290 0 C, and assum ing that at this temperature, the
evaporation of low molecular weight products starts. The heating process up to 275-
290 0 C is characterized by the volume of material (or the thickness of the film when the
illuminated surface is equal to the actual surface of the sample) and the initial
temperature of film. For instance, I crrr' of 125 micron FEP film requires 8.49 J of
206
absorbed energy to reach a temperature of290° C. So the etching process due to VUV-
radiation is characterized by an induction period. The induction time can be calculated
from the formula for the heat ing of a material
(6)
where
T = temperature at which the evaporation is starting
TI = initial temperature.
When the temperature of the material arrives at the melting point and etching is
started, what remains unknown is the speed ofthe etching process.
3.5 FEP TEFLON ETCHING SPEED ESTIMATION DUE TO VUV-RADIATION
The etching rate can be estimated from the enthalpy of polymerization for FEP. L1H =
enthalpy of polymerization for FEP from the monomers is ISS kJ/mole or 3.lkJ/g (it is
assumed that CF2 is the monomer of FEP). This is one of the largest values for
polymers.
Assume now that 100 nm of FEP absorbs all VUV radiation . The depolymerization
of 2.52'10'5 grams of FEP needs 3.1 [kJ/g) ·2.52·IO,5 [g] = 0.07 J of energy. Assume
roughly 10,1 J including all losses for evaporation, melting , structure transit ion at 1270
C. So for the etching of I micron of FEP the amount of energy needed is I 1.
4.0 Discussion
4. I. SET UP OF GROUND VUV-TESTS
The above calculations are true for an "adiabatic" process, i.e, a process without
heat losses through heat transfer mechanisms with bulk or environment.
The material set-ups in VUV tests are different in most cases, and there are
uncertainties caused by e.g.:
- fixation of the sample in the sample holder, which includes many unknowns
related to heat exchange (thermal contact) ;
- for non-resonance sources, such as a Hamamatsu lamp, LDD-400 and LD-3 and
others , there is only approximate knowledge of the attenuation length in FEP and the
spectral intensities; consequently, the dose/depth distributions for the photons of
different wavelength is uncertain ;
- the irradiated surface does not always coincide with the real sample surface in a
VUV-test, or the beam is not necessarily homogeneous. This means that the total
amount of sample material must be taken into account for all calculations of heating and
etching.
4.2. ENERGY TRANSFORMATIONS IN FEP TEFLON MATERIAL IN GROUND

VUV-TESTS CONDITIONS
The process of absorption of VUV-radiation is a radiation process and not a photo-

chemistry process. That is why the absorption spectrum in the VUV range is absent
because this would requ ire discrete absorption lines.
207
VUV-radiation, as ionized radiation in reactions with materials, is partially

reflected and partially absorbed by the material. This absorbed energy of VUV-
radiation is then partialy spent on irreversible chemical reactions and emissivity, but the
main part of the absorbed energy is transformed into heat.
For emission, a material needs to have enough energy levels between which excited
particles can "travel ;" during these jumps, these excited particles could emit
electromagnetic radiation . In the case of FEP, we have only a few energy levels and
only in the IR-region. FEP does not emit in the visible and VUV region . So one could
state that under the condition ofP < or equal 10'6 Torr , the process ofVUV-irradiation
of FEP Teflon material is, to a first approximation, adiabatic .
4.3. EXAMPLES OF THE ESTIMATION OF INDUCTION TIME
On the basis of previous experiments [3, 6], one can estimate some value of preheating.
In [3], the changes of the mechanical properties start only after several hours of
irradiation . The last point of measurements without mechanical changes is 10000 sec.
14
The intensity of the VUV-source is 4xlO photon.cmi.s", the photon energy is 10 eV.
The FEP Teflon material had a thickness 105 microns. Using formula (I), preheating
input is 6.41 J.
In [6], the etching process starts after the material is irradiated with a total fluence
of 1.1xlO 18 photons with an energy of8 eV. Unfortunately, no data about the thickness
of the material is available .
4.4 ETCHING PROCESS IN GROUND VUV SIMULATION
In [4], after the fluence 2'10 18 photons ern" with energy of 147 nm, the mass loss is
measured to be 1.5'10'3 g.cm", That means 2.56 J etched 6.9 microns of FEP or 2.69
microns per I Joule of deposited energy.
In [6], FEP Teflon was irradiated 140 minutes by an Xe-resonance lamp with an
intensity of 2.4x10 14 photon.cm'i.s" . Measured mass losses were 0.34-0.36 rug/ern"
This means 0.647 micron per I joule of energy.
The results of a long duration exposure experiments using a VUV Hamamatsu
lamp L879 were presented in Ref. 10. These experiments have shown that FEP exposed
for 1000 hours to this deuterium lamp lost 20711g'cm,2. This corresponds to 0.96
microns. The energy input was between 112 and 187 Jcrn". Etching of I micron ofFEP
needs, in that case, between 116 - 195 J.
These different experiments have shown a large range in energy from 0.37
J'um" up to 195 J'llm,l.
4.5. ENERGY DISTRIBUTION AND DOSE RATE CALCULATION IN THE TOP

LAYER OF FEP DURING GROUND VUV TESTS
The maximum energy output from the different sources is presented in Table 3. Assume
now an average depth for the energy deposition of 10 nm, based on the attenuation
length estimates . We assume that 66 per cent of all VUV energy is absorbed in 10 nm
of material.
208
From this, one can estimate the dose rate for some UV sources . The mass of FEP
for a Icm xlcm xlOnm material is 2.15.10- 6 g. This results in the following maximum
dose rates for the different sources:
Kr-resonance source : 49 Mrad/s
Xe-resonance source : 41 Mrad/s
LD-3 source : 82 Mradls
These values result in extremely high doses, and will result In complete
degradation of this FEP top layer.
4.6 .VUV DEGRADATION IS A CASE OF THERMO DEGRADATION
During VUV ground simulation two stages exist;

I st stage: a preheating induction stage when practically all the energy input is spent
on the heating of the bulk of the material;
2nd stage: the bulk of the material keeps a constant temperature and the energy is
spent on the etching process. There is an equilibrium between energy input, temperature
inside the material and the etching rate.
When FEP was irradiated in the presence of a dry nitrogen flow [I], neither
thermal degradation of FEP nor any changes of mechanical properties were observed.
Only the surface layer of FEP was etched due to the VUV flux of the Hamamatsu
deuterium lamp.
In the case of an "adiabatic" VUV testing [3-7,10] thermal destruction is observed
which depends on:
a) the thickness of the FEP film. The thick film has a higher capacity for absorption
of energy. So thicker films degrade less than thinner ones. This can be seen from
experiments in table 7 [6] where the comparison for different thickness is present.
b) duration of continuity (meaning duration of the test without switching the
radiation source on and oft) ; this means that preheating of the material needs a definite
time .
c) the presence of other factors such as an AO flux.
5.0 Conclusions
The estimation of heating during VUV ground simulation has been presented. The
VUV process is mainly a thermo-degradation process in the bulk and etching of the
surface, and is characterized by the induction time of preheating and a stage of constant
etching rate.
A rough estimate of the heating distribution process during ground simulation has
shown a significant role of thermo-degradation of the material during ground
simulation.
Ground simulation of VUV-radiation should be evaluated taking into account heat
input of the VUV-sources.
209
6.0 References
1. Milintchouk, A , Van Eesbeek, M., Levadou, M., Harper, T., Influence of X-RaySolar FlareRadiation on
Degradation ofTeflon in Space, Journal ofSpacecraft and Rockets, vol.34, No.4, 542-548, (1997).
2. Milintchouk.A, Van Eesbeek, M., Holmes-siedle, A., Levadou, F., Degradation of Materials under the
Action of Soft X-Ray Radiation from Solar Flares, Proceedings of the 7th International Symposium on
"Materials in Space Environment," Toulouse, France, 87-96,(1997).
3. Van Eesbeek, M., Levadou, F., Tupikov, V.I., Chemiavsky, A.\., Khatipov, SA, Milinchuk, V.K.,
Stepanov, V.F., Milintchouk, AV ., Degradation of Teflon PTFE and FEP Exposed to Far Ultraviolet
Radiation, Proceedings of the 6th International Symposium on "Materials in Space Environment,"
ESTEC, Noordwijk, The Netherlands, 149-164, (1994).
4. Van Eesbeek, M., Levadou, F., Skurat, V.E., Dorofeev, Yu.l., Vasiliets, V.N., Barbashev, EA,
Degradation of Teflon FEP Due to VUV and Atom ic Oxygen Exposure, Proceedings of the 6th
International Symposium on "Materials in Space Environment," ESTEC, Noordwijk, The Netherlands,
165-173, (1994).
5. V. \. Toupikov, V.\., Khatipov, SA, Chemiavsky, A.\., Stepanov, V.F., Van Eesbeek, M., Levadou, F.,
Degradation of Mechanical and Electrophysical Properties of Teflon FEP Films under Combined Action
of Far Ultraviolet and Thermal Cycling, Proceedings of the 7th International Symposium on "Materials
in Space Environment," Toulouse, France, 77-85, (1997).
6. Skurat, V.E., Barbashev, EA, Budashov, I.A., Dorofeev, Yu.l., Nikiforov, A.P., Temovoy, AI., Van
Eesbeek, M., Levadou, F., The Separate and Combined Effects of VUV Radiation and Fast Atomic
Oxygen on Teflon FEP and Silicon Carbide, Proceedings of the 7th International Symposium on
"Materials in Space Environment," Toulouse, France, 267-279, (1997).
7. Grossman, E., Noter, Y., Liftshiz, Y., Oxygen and VUV Irradiation of Polymers: Atomic Force
Microscopy (AFM) and Complementary Studies, Proceedings of the 7th International Symposium on
"Materials in Space Environment," Toulouse, France, 2I7-223, (1997).
8. Milinchuk, V.K., Klinshpont, E.R., Tupikov, V.I., The Radiation Stabilityof OrganicMaterials, Energo-
Atomizdat, Moscow, Russia, (1994).
9. Coucke, P., Van Eesbeek, M., Development of a Facility for the Simulation of High Energy Solar
Radiation, Proceedings of the 7th International Symposium on "Materials in Space Environment,"
Toulouse, France, 97-100, (1997).
10. Weihs, B., Van Eesbeek, M., Secondary VUV Erosion Effects on Polymers in ATOX Atomic Oxygen
Exposure Facility, Proceedings of the 6th International Symposium on "Materials in Space
Environment," ESTEC, Noordwijk, The Netherlands, 277-283, (1994).
II. Verkhovtsva, E.'L, Yaremennko, V.\., Telepnev, V.D., Gas-JetSimulator of Solar VUVand Soft X-Ray
Radiation and Irradiation Effecton Some Materials, Proceedings of the 7th International Symposium on
"Materials in Space Environment, "Toulouse, France, 119-124, (1997).
12. Abraimov, V.V., Negoda, AA ., Zavalishin, AP., Kolybaev, L.K., Complex Simulation of SpaceFactors,
Space Science and Technology, 1 n2-6, 76-80(in Russian), (1995).
EVALUATION OF THE COF MDPS CHROMIC ACID ANODIZATION
IN LEO ENVIRONMENT
C.ACKET
Aerospatiale Espace et Defense. Etablissement d'Aquitaine,
Saint Medard-en-Jalles, France
J. MARCO, A. PAll..LOUS, R. REULET, V. VIEL

ONERA , Departement d 'Environnement Spatial,
Centre d'Etudes et de Recherches de Toulouse, France
Abstract
The Columbus Orbital Facility (COF), which constitutes the European space module
of the International Space Station, is expected to be exposed to the low Earth orbit
(LEO) environment for 15 years. Aerospatiale is in charge of developing the
Meteoroid and Debris Protection Shield (MDPS). As part of this program,
Aerospatiale, in cooperation with ONERA, has investigated specific properties of the
Chromic Acid Anodization (CAA) used on the external aluminum panels. The
purpose of the study was to assess experimentally the behaviour of these thermo-
optical surfaces when exposed to vacuum, ultraviolet radiation (UV) and atomic
oxygen (AO) in the laboratory . In addition , plasma breakdown susceptibility was
investigated .
In order to simulate space environment effects, combined tests have been
performed . They included UV radiation up to 7200 equivalent sun hours (ESH), and
Atomic Oxygen (AO) up to 5xl0 20 atoms/em'. Some samples received only UV
irradiation or AO. and others were exposed to UV + AO. The thermo-optical
properties (spectral reflectance, solar absorptance a s, emissivity f:N) were measured
before and after the tests.
For the samples exposed to UV, variations of as were observed. The results
have been compared to ESTEC and LDEF CAA, and they have been extrapolated
using empirical models to estimate end-of-life (EOL) values. Concerning the
samples exposed only to AO, the as variation was not large. After exposure to UV
then to AO, a s came back to its initial value. but f:N seemed to decrease slightly.
The tendency observed in a s variation is in good agreement with LDEF
observations. In addition, spectral modifications were observed at different test steps
(air. vacuum, UV, AO); they are correlated to outgassing and solar absorptance
degradation . Under simulated ionospheric plasma, discharges were found to occur
above a bias threshold of about -80V .
211
J./. Kleiman and R.C. Tennyson (eds.},
Protection of Space Materialsfrom the Space Environment, 211-228.
212
1.0 Introduction
The Columbus Orbital Facility (COF), which constitutes the European space module
of the International Space Station, is expected to be exposed to the low Earth orbit
(LEO) environment for 15 years. Aerospatiale is in charge of developing the
Meteoroid and Debris Protection Shield (MDPS). The MDPS panels are composed
of two parts; a ftrst bumper made of 6061 T6 aluminum treated with a speciftc
Chromic Acid Anodization (CAA); and a second bumper made of a Nextel mattress
and a Kevlar/resin composite.
For this program, the materials constituting the first and the second bumper have
been previously characterized over several aspects; thermal cycling, outgassing,
flammability, humidity. As part of this program, Aerospatiale in co-operation with
ONERA, has investigated the speciftc properties of the CAA used on the external
aluminum panels. The study purpose was to assess experimentally the behaviour of
these thermo-optical surfaces in space, and to verify their ability to meet the
following program speciftcations involving solar absorptance U s and normal
emissivity (EN) :
- Begirming-Of-Life (BOL): U s ~ 0.3 and EN ~ 0.4 at room temperature.
- End-Of-Life (EOL): U s ~ 0.4 and EN ~ 0.4 at room temperature.
UV radiation and atomic oxygen are the main degrading components which
have to be taken into account when evaluating the degradation of thermo-optical
materials placed in the Low Earth Orbit environment. CAA surfaces exposed to the
solar UV radiation are degraded; the alumina layer becomes light-absorbing,
probably owing to the formation of colour centres in the outer dielectric layer,
resulting in a solar absorptance increase. The CAA surfaces are not expected to be
affected to a large extent by atomic oxygen. Results obtained on LDEF indicate that
the effect of O-atoms on CAA probably counterbalances somewhat the solar
absorptance increase which is due to UV radiation.
In order to verify these effects, CAA samples were exposed to a simulated solar
UV radiation; we also exposed to Atomic Oxygen, samples already exposed to UV
and virgin samples.
Concerning UV radiation, Space Station Freedom was expected [1] to
experience during its 30-year life, between 31000 ESH (for earth-facing surfaces)
and 77000 ESH (for space-facing surfaces), with most surfaces receiving around
52000 ESH; ram- and wake-facing surfaces receive about 35 per cent more than
surfaces parallel to the velocity direction. According to these values, the maximum
UV exposure for 15 years is 38500 ESH. A previous long-term test performed on a
CAA in the framework of an ESA-ESTEC program, showed that degradation leveled
off only after 5000 UV ESH [3]. During the present testing, stabilization was not
observed at this level, therefore, the irradiation duration was extended as far as
availability of the facility allowed to; the irradiation was ended at 7200 ESH.
In LEO, the effect of O-atoms is mainly experienced on ram surfaces,
perpendicular to the velocity vector. The fluence of atomic oxygen for 15 years at the
altitude of the International Space Station was estimated to be between
2xl0 22 atoms/em? at a constant density and 8x1O 22 atoms/em" at a constant flight
altitude profile. As such, a fluence could not be achieved in a reasonable test
213
duration and also because low absorptance variations were expected, the screening
test wasfinally reducedto 5x1020 atoms/em".
In the Space Station orbit, the effect of the plasma on CAAs is presumed to be
negligible if the electrical field through the thin dielectric oxide layer remains low.
Previous laboratory experiments have shown that a potential difference between a
conductive structure and a plasma can provide electrical discharges when a
conductive substrate is covered with a dielectric layer. Solar panels working at high
voltages are present onboardSpace Station Alpha; if the ISSA plasma contactordoes
not perform as expected, electrical discharges due to LEO plasma will damage the
CAA materials [2]. The discharges which are observed on anodized surfaces show
electromagnetic transients similar to those obtained with an inverted gradient under
electron irradiation. In the present testing, the electrostatic behaviour of biased
samplesin the presence of plasma was investigated in orderto find the bias threshold
required to trigger discharges and to analyze the discharge transients. Long-term
optical properties due to discharges were not investigated in this study.
2.0 Material elaboration and specimens
2.1. MATERIAL ELABORATION
The analysis of program specifications induced the necessity of having a thin CAA
on a polished aluminum surface in order to obtain low values.of emissivity and of
solar absorption. In a first phase, usual CAA treatments applied to the laminated
surfaces and polished aluminum surfaces were tested. Such CAA treatments were
not compliant with the specifications, so a specific processhad to be developed.
The designed process, applied to aluminum 6061 T6, has been developed and
testedat the Aerospatiale CentralLaboratory. All the parameters (voltage, processing
times, bath temperatures and concentrations) are controlled and performed as
required by the process. It involves three steps; polishing, anodizing and sealing:
a) polish shine:
- Alkaline degreasing
- water rinsing
- polish shine in specific chemical bath
- water rinsing
- neutralization
- hot water rinsing
The first samples were hand polished (50 mm x 50 mm) and a full scale
cylindrical panel (1400 mm x 1200 mm) had been chemically polished in an
industrial plant.
b) Chromic Acid Anodization:
- alkalinedegreasing
- water rinsing
- deoxidizing
- water rinsing
- CAA under 20V during a specific dwell time
214
- waterrinsing
c) sealing:
- sealing
- waterrinsing
- wiping
2.2. SPECIMEN DESCRIPTION
Aluminum samples were composed of plates of three sizes: (20 mm x 20 mm for

microscopy analysis, 50 mm x 50 mm for optical measurements and 300 mm x 350
mm for plasma exposure).
2.3. MICROSTRUCTURAL ANALYSIS
The sealed CAA layer is 1.3 urn thick. Its columnar structure is seen clearly in
Figure 1. Figure 2 shows the "com flake" structure of the sealing.
Figure I. Microstructure of CAA
Figure 2. Microstru cture of sealing.

215
2.4. INITIAL SOLAR ABSORPTANCE AND EMISSIVITY
Optical properties (cx,s solar absorption, 8N emittance) were:

- cx,s = 0.09 ± 0.02 and 8N = 0.05 ± 0.02 measured after the polish shine
treatment,
- cx,s = 0.27 ± 0.Q3 and 8N = 0.30 ± 0.03 measured after the Chromic Acid
Anodization treatment on 50 mm x 50 mm samples.
3.0 Facilities and Specific Test Conditions
Simulation tests were all carried out in specific facilities existing in the DESP
laboratories.
3.1 SOLAR ULTRAVIOLET RADIATION
The irradiations were performed in the SEMIRAMIS facility (described in [4]), in

vacuum, with in situ spectral measurements of the optical reflectance within the solar
wavelength range. With such a system, the outgassing effects during the initial
transition air-to-vacuum, and also the air reentry effects during the final transition
vacuum-to-air, can be investigated. The in situ optical measurements consist of
spectral reflectance measurements made relative to a reference sample. The system is
composed of a spectrophotometer Perkin-Elmer Lambda 9, associated with a lateral-
sample integrating sphere which is mobile in vacuum. Because this in situ system
does not give absolute measurements, additional measurements were also performed
in air (they can be considered as absolute measurements, because, in this case, an
integrating sphere with a central sample is used, associated with a Cary 2300
spectrophotometer). These additional measurements were used to obtain in air,
before irradiation, spectral coefficients used to recalibrate the relative spectra. The
solar reflectance (Rs) is the weighed integral of the spectral reflectance versus the
solar emission spectrum 1(1..) in the range 250-2500 om:
2500 f2500
Rs = R(A) · 1(1..) · dA/ 1(1..) · tIJ... (1)
f 250 250
Relative uncertainty is ~ ~ ± 0.01, including noise and reproducibility. For

opaque samples like CAA, solar absorptance is deduced from the difference:
cx,s=l-R s (2)
The sample-holder was equipped with 6 CAA samples and one aluminum
witness. We measured the optical reflectance in 15 positions, each large sample
measured on four points on its surface.
The vacuum was maintained below 10"4 Pa during the whole test duration (2.5
months). The temperature was maintained at 30°C at the sample-holder. This value
corresponded to a temperature lower than 35°C at the level of the samples during
irradiation.
216
The residual contamination rate was measured during the tests using a
temperature controlled Quartz Crystal Microbalances (QCM) model MK9 from
QCM Research. Their sensitivity is 4.43 x 10-9 g-em -2 ·Hz -I (assuming a density
equal to 1, this corresponds to 0.443 A·Hz -1). A first QCM embedded in the sample-
holder was exposed to the irradiations at the same temperature as the samples. A
second QCM set was placed at the bottom of the irradiationchamber and exposed to
a residual background of UV after reflections on the sample-holder and irradiation
chamber walls. Low mass increases, corresponding to about 20 A on the sample-
holderand 85 Aat the bottom of the chamber, were measured.
UV irradiation was performed in the range 5 to 7 solar UV constants using a
solar UV simulator based on a 6500 W short-arc Xenon source, with complete
rejection of visible and infrared light above 400 nm. The irradiation uniformity was
± 20 per cent. Irradiationwas continuous, except for small switch-offperiods shorter
than one day (for flux measurements, lamp changesand optical measurements).
Normal emittance BN was measuredonly in air, before and after the test, using a
portable reflectometer, Gier & Dunkle DB100.
3.2 ATOMICOXYGEN
AO testing was carried out in the CASOAR facility. This facility uses a pulsed
atomic oxygen source whose principle was established by Caledonia & Krech at PSI
in the USA. It is based on a laser-induced breakdown of molecularoxygen followed
by a rapid expansionof the recombining plasma, resulting in the production of a flux
of nearly mono-energetic oxygen atoms. The source of CASOAR gives oxygen
atoms with velocities adjustable from 5 to 11 km·s -I (nominally: 8 km-s") with an
estimatedproductionof approximately 1015 atoms per cm2 per pulse. During the test,
the sample-holder was placed at 40 em from the nozzle throat; therefore, the
bombardedzone was 16 em in diameter. The velocity was adjusted to 8 km·s -I . Flux
rates were estimated from mass loss measured on Kapton samples placed between
the tested samples, and considering a Kapton reactivity of 3xlO -24 cnr'-atom" . The
sample-holder was equippedwith 6 CAA samplesand 8 Kaptonsamples.
The samples were characterized, in air, before the test and after AO exposure
steps correspondingto the cumulatefluences lxlO-2o• 3xlO-20 and 5x1020 atoms-em -2
by:
a) spectral reflectance and solar absorptance in the range 300-2400 run, using a
Cary 2300 spectrophotometer equipped with an integrating sphere using a central
sample; the large sampleswere measured on four points,
b) normal emissivity &N, by means of a portable Gier & Dunkle DB100
reflectometer,
c) mass (weighing at ± 20 ug),
d) visual appearance (naked eye and optical microscope).
3.3 PLASMA
The tests were performed in the JONAS simulation chamber. It is composed of a

large chamber (working volume ~15m3) made of non- magnetic steel, equipped with
217
a neutralization system for the effects of the earth 's magnetic field . A high power
cryogenic pumping unit ensures a limit pressure of 2xlO -s hPa in about 15 hours.
The plasma source is of the Kauffman synthetic plasma type with the follo wing
characteristics: residual pressure in the presence of plasma 4xlO -6 hPa , plasma
potential close to 0, argon plasma density in the range 104_106 em" , electronic
temperature in the range of 1000-2 000 K drift energy of argon ions about 15-20 eV
(V ~ 8 kmls). Plasma dens ity and electronic temperature are measured and controlled
by means of the current-voltage characteristics of a Langmuir probe . The setup is
speciall y des igned for studying electrostatic charg e phenomena in the presence of an
ionospheric plasma representing the LEO space environment. During the test , the
source setting parameters were an electronic temperature of 1500 K and a plasm a
density close to 5x I0 5/cm3 .
4.0 Results and discussion
4.1 SOLAR ULTRAVlOLETRADIATION
4. J.J Spectral reflectance measurements

On the initial spectrum measured in air (Figure 3), an absorption band centered at
365 nm was observed; its origin has not been identified. A similar band had been
found in a previous study [5] on another CAA (oxide thickness about 4 rum)
performed for ESA-ESTEC (results from this contract were pub lished in [3] but only
solar absorptance curves were given). For wavelengths longer than 500 nm,
oscillations in reflectance can be explained by optical interference provoked by the
1.3 rum transparent anodization layer. During the air-to -vacuum transition and the
consecutive outg assing period , a spectral shift of the oscillations towards the lower
wavelengths is evidenced; it is probably due to a refra ctive index change consecutive
to the outgassing of the oxide layer.
100
80
~
1!....
ell 60
0
C :- - Alr
S0
~ 40 !----- V&cuum (O ESH)
Gi
a:
20
o j I I I I I I I I I I I I I I I I I I I I I I I
0 500 1000 1500 ~ooo 2500
Wavelength (nm)
Figure 3. Air-to-vacuum transition.

218
During the first UV irradiation step (240 ESH), the absorption band at 365 nrn
vanishes completely, and a general reflectance decrease is simultaneously observed
below 800 nrn (Figure 4). The thermal exposure of a blank sample at 210°C for 12
hours provoked the same complete resorption of the 365 nm absorption band. This
effect, also observed in [5], is not explained. It does not seem to be correlated to a
loss of absorbed water . Infrared spectra performed at the end of the test did not show
a sensible difference in absorbed water peaks. Furthermore, it is not possible to
recover this absorption band on the sample exposed to 210°C when putting it in
contact with water.
10O }
80
......
~
L
CIl 60
o
l::
ell
i......:.....- V"" ~um(O ESH) I
(j 1 UV 240 ESH
CIl
:;: 40 L -. I
CIl
a:
20
0 - '- ---'>--I ->--+--+--+- 1 -t-

' -+- , - + -+---+-t---+--t I I I I I I I I
0 500 1000 1500 2000 2500

Wavelength (nm)
Figure 4. Degradation after 240 ESII
During the UV irradiation steps from 240 to 7200 ESH, a steady degradation of
the spectral reflectance mainly below 800 nm is observed (Figure 5) and is probably
due to the formation of color centers.
100 - - UV240ESH
- - - - - UV 575 ESH
.. .. .. . . . UV955 ESH
80
_ . _ . - . UVI790 ESH
, 0
e, _ .. - .. - UV 2660ESII
CIl 60
u - - UV 3270 ESH
l::
flCIl - - - - - UV3995ESH
.. .. ... . . UV4835 ESH
~
a: - .- .- . UV 5400 ESH
- " - " - UV6130ESH
- - UV 6431 ESH
- - - - - UV 7170ESH
I I I I I I I I I I I I I
o 500 1000 1500 2000 2500
.... . .. UV 7200 ESH
Wavelength (nm)
Figure 5. Degradation from 240 up to 7200 ESH

219
During the final operations, 20 hours in vacuum and nitrogen admission,

no variation is seen . During air admi ssion, an opposite effect to that observed during
the air-to-vacuum transition ha ppens : a spectral shift of the spectral oscillations
towards higher wavelength s, but without appre ciable change in amplitude (Figure 6).
100
80
. ~
- - -- ,
~
::!!
L
Gl 60 - - UV 7200 ESH I
0
c:
S0 I
~ 40
/
Qj
a:
20
0
0 500 1000 1500 2000 2500
Wavelength (nm)
Figure 6. Vacuum-to-a ir transition
4.1.2 Solar absorptance coefficie nt

The CAA solar absorptance variation vs. the UV dose is given in Figure 7. This solar
absorptance variation tlo. s = o. Sfinal - o.Sinitial is the mean value taken on the 15
measurements, corrected by subtraction of the aluminum solar absorptance increase.
A small absorptance decrease which is ascribable to outgassing and resorption of the
absorption band centered at 365 nm until 240 ESH is observed at the beginning;
then, up to 7200 ESH , as increases with a low tendency towards leveling off.
'0 .0 5 0
•
0.04 0
0 .0 3 0 -
/
/
. / /-------/'" ••....--:
0 .020 - / -
/ -
0 .010 - •
/
0 .000 1 / - ·
•• 2000 4000 6000 8000
-0 .010 - .
-0 .0 2 0 -
Ultraviolet s (ESH)
Figure 7. Mean e AA solar absorptance variation after correction with aluminum data
220
Scanning Electron Microscopy has shown that the microstructure was not
modified by UV irradiations.
4.1.3 Solar absorptance extrapolations

The objective was to extrapolate the experimental values obtained for 7200 ESH to
the end-of-life (EOL) value of a I5-year mission which corresponds to 38500 ESH.
Data from the present test have been compared to those obtained in a previous
test performed for ESA-ESTEC [2, 4), see Figure 8. In both cases, the values have
been corrected for contamination as described above. The ESA CAA testing was
performed with UV radiation alone up to 10000 ESH, then with particles and UV up
to 14500 ESH. Data below 10000 ESH showed a fast stabilization of the degradation
under UV. The curve was well fitted with an exponential expression :
~as = A + B exp(-t/r) (3)
(t is the time expressed in ESH, 1: a time constant expressed in ESH, A and B are
constants without dimension, A being the EOL value).
0 .1 - - '
"-
III
(5 0 .08 - -- - - - - -- , ESTECCAA
VI
I •
'0
t:
o [:
.- III
CIl
U 0 .06 : .-:
.~
·-w-·.-.
..~ Q. IV- - -::::----",' ,..J:::;.,-= MOPS CAA
"-
III
"-
0 ::,lJ....._ ~=:- :..,;
> VI ../
2-g 0 .02 - -=" - - -- - - - - - - - "'---
:l
(5
VI
.J:J
et -' -
1-'
-0 .02·- '- - ---- - - - - - - - - - ' -- - --
o 5000 10000 15000 20000
Ultravio lets (ESH)
Figure 8. Compari son of Aerospatiale and ESTEC CAA degradation s
Application of such a model to the MDPS CAA, gives A = 0.048, B = - 0.054,

the stabilization of degradation being nearly achieved at 15000 ESH with ~as ::::
0.047. Figure 9 shows the measured as values and the error envelopes considering an
uncertainty on absorptance ~as = ± 0.0 I and an uncertainty on the UV dose of ± 20
per cent. The EOL estimated a s value is in the range 0.31-0.33. The validity of this
empirical model is not demonstrated and is only based on the similarity of the
degradation behaviour with the CAA from ESTEC.
221
0.1 ~
.
III
0 .09
0.08
1
t
"0til
- c: ()
0 .07
o Ql 0 .06
.2 ~ 0 .05
r
+
-!-
1
"..- ~
_ - - - - - - - - - - - - - - - - - - - - 0 .057
/...c----_ - - - - - - - - - - - - - - - - - 00.037
.047
. ..
:§a 0 .04
III 0 0 .03
> III 002
~
I I
/r9.:6-~
rr ,/---
T In /
!.g
:l 0 :01 y~;/
"0til o .~J+I'------'------'------f--------<
.c
< -0.01 "ef 10000 20000 30000 40000
-0 .02 t
-0 .03 .;
Ultraviolets (ESH)
Figure 9. Exponential model
Another empirical model has been applied to LDEF chromic acid anodization by
Wilkes [6]. In this experiment, the samples were placed in the ram direction and
received UV in synergy with atomic oxygen. This model is based on an expression
a s = exp [A+B In(t)] that Wilkes has used on many materials (paints, FEP films,
silicones, CAA, etc). This model, too pessimistic as says the author, fits the
measured values (excepted the first one at time 0) on one sample (sample C63)
exposed for 19.5 months in space. During this period, LDEF was in a high orbit and
for this reason, exposed mainly to UV, the AO flux being negligible. Another sample
(sample C61) exposed during 69.2 months on LDEF, received UV but also a large
quantity of AO due the orbital decrease of LDEF during this long period. For this
last sample the final absorptance value is close to the initial, and the model is not
applicable. The author says that the appearance of the sample is "mottled, washed
out, suggesting a different damage mechanism occurring late in the mission ." The
fact that AO has a contrary effect compared to that of UV is confirmed by numerous
measurements practiced on the LDEF structure [Ref. 5] that shows that on the
leading edge, the absorptance increase is negligible compared to the increase on the
trailing edge not exposed to AO.
For these reasons, the Wilkes's model may be used to obtain an upper limit for
the variations which are expected for the JvlDPS in space, and which are mainly due
to UV and contamination. Using this model on our data gives the curve in Figure 10;
the error envelopes are obtained considering an uncertainty on absorptance ~as = ±
0.01 and an uncertainty on the UV dose of ± 20 per cent; the regression is applied
between 240 and 7200 ESH with the expression:
~as + ~aso = exp [A+B In(t)] (4)

222
with the initial solar absorptance a.so = 0.297 and coefficients A =-1.50, B
4.68xlO·2 . The corresponding estimated EOL a.s value is in the range 0.32-0.35.
0.1
...ra 0 .08 _ - - - - 0.081
'0 _ - - - - - - - -.. .- - - 0.067
-
Ul
0 Ql
0.06
------ __ - - - 0.054
s:: o
0 s::
;; ra
ra Do
.;:
..... 0.04
--- ---
ra 0 0 .02
> Ul
C1/,c
.. ra
~ 0
'0
Ul
,c 10000 20000 30000 40000
<I: -0.02
-0.04
Ultraviolets (ESH)
Figure 10. Wilkes's model
4.1.4 Emissivity variations

Normal emissivity, EN, was measured in two locations for the large samples and in
only one position for the small samples. We obtain a slight decrease measured on all
the samples (except on a reference sample maintained "on the shelf' in air at the
laboratory during the test) corresponding to a mean value ~EN = -0,018 ( EN initial
value 0.30, final value 0.282). As the emissometer reproducibility is stated to be ±
0.002, this variation is imputable to the UV exposure,
Such a decrease was not observed during the previous long test involving the
ESA chromic acid anodization [2, 4), but in this longer test (18000 ESH total), the
conditions were not the same because particle irradiations were also applied. Based
on a comparison with blank quality assurance samples. the stability of the emissivity
was also reported on the anodized LDEF structure [7], with and without atomic
oxygen; nevertheless, it was also reported that an increase of 0.03 had been measured
on the non-exposed samples; as the absorptance of these non-exposed surfaces also
increased, it was supposed that contamination had occurred. LDEF exposed surfaces
could also have been contaminated; in such a case a decrease of emissivity could
have been cancelled by this contamination .
4,2 ATOMIC OXYGEN
4.2.1 Spectral reflectance measurements

Two kind of samples have been exposed to AO: virgin samples and samples already
exposed to UV.
223
On virgin samples (Figure 11), the first 1020 atoms-em? provoked a slight
reflectance decrease and induced the beginning of resorption of the absorption band
at 365 nm. Additional exposure to atomic oxygen. at the cumulated steps 3xl0 20 and
5x1020 atoms.em? slightly intensified these effects.
100
90
80
~ 70
fl 60 - - - - before AD
c 50
n 40
Gl
- -
after 1e20 atlcm2
- after 3e20 atlcm2
;; 30
a: - - - after 5e20 atlcm2
20
10
O+-- - - + - - - -- ! - - - --!--- - -!-----I
o 500 1000 1500 2000
wavelength (nm)
Figure 11. AO effects on a virginsample
On the samples already exposed to UV (Figure 12), the first AO exposure

provoked an important increase of the optical reflectance and a nearly complete
recovery of the spectrum towards the initial state. It can be remarked that the
absorption band at 365 nm which vanished after UV irradiation reappeared slightly
upon AO bombardment.
100 T
90 T
80
~ 70 /r-/"""-..--- -~\ /
fl 60 /
ne 40
Gl
50
,
;; 30 I
- -- before AD
a: 20
- - - - - . after 1e20 atlcm2
10
O+--- - t -- - - t - -- - +-- -- - + - - --j
o 500 1000 1500 2000 25011
wavelength (nm)
Figure 12. First AOexposure on a UVirradiated sample
Additional AO exposure (Figure 13) provoked a general reflectance decrease,

but also an enhancement of the absorption band at 365 nm. These effects are
probably due to several competing mechanisms: destruction of UV-created colour
centres, oxidation of the outermost surface, creation of other color centre species,
etc.
224
100
90
80
~ 70
B 60
n 40
c
III
50
- - - - - . after 1e20 atlcm2

iii
a:
30
- - - after 3e20 atlcm2
20
10
- - - after 5e20 atlcm2
0 - 1 - - - - - - + - - - - + - - - --+-----+-------1
o 500 1000 1500 2000 2500
wavelength (nm)
Figure 13. Additional AO exposure ona UV irradiated sample
4.2.2 Solar absorptance

The mean values of solar absorptance variation corresponding to the measurements,
after AO bombardment of both the virgin samples and the samples already exposed
to UV, is given in Figure 14.
For the samples only exposed to AO, a slight as increase is obtained:
~as::::: +0.017 at the end of the test. For the samples already exposed to UV, a
decrease of solar absorptance is obtained at the first step of AO exposure : ~as
::::: -0.04. This decrease exactly counterbalances the increase previously obtained upon
UV irradiation . A very slight increase of about +0.0I is obtained during further AO
steps.
• UV.AO -------0---- AO
0 .050 ~ -- -
0.045
0.040
A
-...,
0 .035
0 .030
0 .025
/
/
/
'\
, -\
0 .020
/
,
--
0 .015
/ '\
/ ~
0 .010
/ \~
.---
--
0 .005 .~
1/
0 .000
Before After UV o 1. £.. 20 3. £..20 5. £.. 20
UV al/cm2 al/cm2 al/cm2
Figure 14. AO effects on solar absorptance
These results confirm LDEF flight data [I, 7] but also ground test results on
similar materials.
For the LDEF CAA materials , once a threshold atomic oxygen fluence was
reached (around IxlO 20 atoms/em'), the average solar absorptance rapidly returned to
pre-flight values. Below this threshold, a weak effect of UV exposure on solar
225
absorptance was revealed by comparative studies on all LDEF sides, indicating an

absorptance increase of approximately 0.008 per 1000 ESH, for ESH numbers higher
than 6000. It was apparent that atomic oxygen fluences above the threshold value
(lxl0 2o atoms/em') reversed or removed the UV degradation effects on absorptance
[8].
During on-ground tests performed at Boeing Aerospace Co. and NASA Langley
Research Center [9], CAAs were tested under a fluence of 7 x 1020 atoms/ cnr'. The
authors said that "there was no change in either solar absorptance and emittance of
the four samples and no mass loss."
4.2.3 Normal emissivity:

The mean values of normal emissivity show a decrease of about ~eN = - 0.033 after
5xl020 atoms/em' . This variation corresponds to 10 per cent of the initial value (eN
initial value 0.30, final value 0.27). This result is not in agreement with the results
discussed previously .
4.2.4 Ma ss loss and visual appearance:

Mass loss measurements were not accurate enough to be used. Concerning visual
appearance (naked eye and optical microscope), yellowing of the sample was
observed after UV irradiation disappeared after exposure to atomic oxygen.
Scanning Electron Microscopy has shown that the microstructure was not
modified by AO exposure .
4.3 PLASMA
A detection device (Figure 15) developed in the framework of previous contracts

[10], was used during the test. It takes the form of a cylinder with a diameter of
approximately 15 em and makes it possible to simulate two parts of a satellite :
- The first part designated 'the box' houses the discharge detection probes
sheltered from the plasma , a probe for charge transfer from the sample to the box
(pearson probe at the level of the sample, oscilloscope channel #1), a probe for the
charge transfer from the whole satellite to the surrounding plasma, (0.1 n resistor at
the level of the box, oscilloscope channel #2).
- The second part, represented by the sample under study with its dielectric
surface exposed to the plasma, simulates an element external to the satellite which
can have a different potential.
The device also makes it possible to configure the discharge according to the
size of the satellite by varying the value of capacitor Ca and the sample's surface
area. In this configuration , discharge mechanism always results in the expulsion of a
quantity of charges equivalent to that contained in capacitor Ca.
In this test, the value of Ca was 10 nF and the following test procedure was
applied:
- after the samples have been mounted in the chamber, they were outgassed for
about 50 hours, because previous studies had shown that the resistivity of aluminum
oxide layers was only reached after a very long outgassing time,
226
- the dielectric performance of the anodized layer was measured by applying a

stepwise increasing negative bias to the alwninum substrate .
•
Box
Pearson
probe
i(l)2NV
eAA Ionospheric
sample Plasma
Figure 15. Plasma measurement device
Different tests done with small samples (2x2 and 5x5 em) have shown that
samples needed to be larger than the cylinder diameter in order to allow the box
containing the analysis probes to benefit from a wake effect, to avoid a deviation of
plasma's drifting ions upstream of the samples, and to obtain the complete discharge
of the capacitor. The discharges obtained with a large sample 35x28 cm were
satisfactory as the complete discharge of the capacitor Ca was obtained.
The final results were the observation of discharges located on the face of the
CAA exposed to the plasma. The polarization threshold required to trigger
discharging was about -80 V. Less than 5 discharges per hour were observed whose
effects were negligible on the visual appearance of the samples . The existence of the
discharge was shown by the simultaneous presence of two current transients at the
levels of the anodized sample and of the analysis device (box). The two transients
recorded during a given discharge were similar as far as the signal shape, amplitude
and duration are concerned. The peak currents were weak, of the order of 200 rnA,
the duration of the transients was extremely variable, from 10 to 50 us,
5.0 Conclusions
Long term tests have been performed on Chromic Acid Anodizations designed by
Aerospatiale for use in the Columbus Orbital Facilities Micrometeoroids and Debris
Protective Shield. The analysis of the measured thermo-optical properties has
demonstrated that the CAA developed by Aerospatiale is compliant with the initial
.MOPS thermo-optical specifications.
227
Upon 7200 ESH UV irradiation, corresponding approximately to 3 years in

space for most exposed surfaces, the degradation was found to lead to a solar
absorptance increase of about +0.04, and to a slight decrease of emissivity. An
extrapolation, based on an empirical model previously applied to LDEF results,
indicates an upper limit of the absorptance degradation of +0.08 at the end of the
mission (15 years). However, in the real space environment, the synergism of a
stronger UV radiative spectrum with the Space Station-induced contamination could
affect the actual end-of-life value.
Under atomic oxygen alone, laboratory results show a very low increase in solar
absorptance. On the samples already exposed to .UV radiation, atomic oxygen
induces a strong tendency to return to the pre-UV irradiation spectral features, with
quasi-recovery leading back to the solar absorptance initial value. These results
confirm the synergistic effects of oxygen atoms and UV radiation observed in space
on materials exposed during the LDEF mission. Solar absorptance degradation was
indeed observed on LDEF on sides exposed preferentially to UV, whereas the
degradation was less important on the UV-facing sides also exposed to the oxygen
atom flux.
We observed a low decrease in emissivity for samples exposed to UV and/or to
AO which seems not to have been observed on LDEF. We assume that this
difference could be due to LDEF contamination.
Plasma discharges were observed when CAA samples were exposed to
simulated ionospheric plasma. The consequences are negligible when considering
visual appearance in a short period of time, but the long-term effects were not
investigated.
Aerospatiale would like to thank Alenia Aerospazio for financial support of this
study and for permission to publish these results.
7.0 References
1. David, K.E., Babel, H.W., "Optical property degradation of anod ic coatings in the space station low
Earth orbit," A1AA 92-21 62, AL4A Materials Specialists Conf erence. Coating Techn ology For
Aerospace Systems. Dallas, TX, April 16-17, 1992.
2.1. A Vaughn, J.A., Carruth, M.R., Jr., "Electrical Breakdown Currents on Large Spacecraft in Low
Earth Orbit," Journal ofSpacecraft and Ro ckets. Vol. 31, No.1, Jan-Feb , 1994-
3. Marco, 1., Paillous, A, Gourrnelon, G., "Thermal control coatings : Simulated LEO degradation under
UV and particle radiation," Sixth International Sympo sium on Mat erials in a Space Environment, The
Netherlands, September 19-23, 1994,
4. Marco, 1., Paillou s, A , Levadou, F., "Combined radiation effects on optical reflectance ofthermal
control coatings ," Fourth Eu ropean Symp osium on Spacecraft Materials in Space Env ironment,
Toulouse , France, September 6-9, 1988,
5. Marco, 1., Paillous , A, Bourrieau, 1., Chourreau, A., Oscar, H., "Essais de longue duree sur
revetements de regulation thermique," Contrat ESTEC 10023/92!NVNB, Etude CERT 439700 ,
Rapport Final, Tomes I et 2, juin 1994.
6. Wilkes, D. R., Miller, E.R., Zwiener, J.M., Mell, RJ., "The continuing materials analysis of the thermal
control surfaces experiment (S0069)," WEF 2nd Post-Retrieval Sympo sium, part 3, pp. 1061-1075 .
228
7. Sampair, T.R ., and Berrios, W.M., "Effects of Low Earth Orbit Environment on the Long Duration
Exposure Facility," LDEF First Post-Retrieval Symposium, NASA CP-3134 , 1991, pp. 935-944.
8. Golden, J.L., "Anodized aluminum on LDEF," LDEFMaterials Workshop, NASA Langley Research
Center, November 19-22, 1991.
9. Dursch, W., Slemp, W.S., "Eva luation of chromic acid anodized aluminum foil coated composite tubes
for the Space Station truss structure ," 33rd International SAMPE Symposium, March 7·10. 1988.
10. Reulet, R., Roussel, J.F., Levy, L., Sarrail, D., "Structure polarisee dans un plasma ionospherique -
nature, seuil et dimensionnement des decharges," Contrat CNES - Rapport Final RF/4533.11 -Aout
1997.
A DATABASE FOR THE SELECTION OF SURFACES AND MATERIALS
FOR SPACE- AND GROUND-BASED APPLICATIONS
SUSAN H. McCALL, ANTHONY 1. CLARK, ROBYN A. ELLIS,

RORIK A. HENRIKSON, ALICE E. PIOTROWSKI,
LORRAINE A. PIOTROWSKI, MICHAEL A. REID, JOHN W. RODNEY
Stellar Optics Research International Corporation (SORIC)
78 Normark Drive, Thornhill, Ontario.
Canada, L3T 3RJ
ROBERTP. BREAULT
Breault Research Organization. Inc. (B.R.O., Inc.)
6400 East Grant Road. Suite 350, Tucson. Arizona,
USA, 857 J5-3862
MARSHALL 1. McCALL
York University, DepartmentofPhysics and Astronomy
4700 Keele Street, Toronto. Ontario,
Canada, M3P JJ3
1.0 Introduction
Stellar Optics Research International Corporation (SORIC) has compiled the

world's largest databank of spectral reflectance data and Bi-directional Reflectance
Distribution Function (BRDF) data for black, white, reflective and transmissive
surfaces and materials which are commonly employed within or viewed by ground-
based and space-based instrument applications. These surfaces and materials, and in
particular the black and white colored materials, typically are used for: stray light
control, thermal control, calibration, and visual target cues.
From this databank, SORIC has created an interactive software data
engineering tool, SOLEXISTM, which enables the selection of surfaces and materials
for ground and space-based applications, as well as identification of vendors, optical
characteristics, and other properties such as space-environment effects data, thermal
and physical data, and chemical data. Multiple, ranked user-selected criteria are
employed for searches.
229
Protection of Space Materialsfrom the Space Environment. 229-233.
230
2.0 Features and Applications
SOLEXIS is PC-based, and Windows' 95-, Windows 98- and Windows NT-
compatible. It addresses surfaces and materials applications for stray light control,
thermal control, calibration, and visual target cues. It is an interactive data
engineering tool which enables one to search/sort and sort/select surfaces and
materials based on: (I) optical data, (2) physical data, (3) chemical data, (4)
electrical data, (5) mechanical and physical data , (6) space-environment effects
data, and (7) vendor contact information. Production and manufacturing issues are
also encompassed. Digital plots and images can be readily printed. A digital
nomogram has been programmed which calculates the effects of low Earth orbit on
numerous materials for the space environment. The theoretical foundations of the
model were developed by R. C. Tennyson of the University of Toronto's Institute for
Aerospace Studies (UTIAS, Downsview, Ontario, Canada).
The applications for which SOLEXIS data are useful include :
• Apertures • Integrating Spheres • Radiometers
• Calibration Screens • Laser Terminators • Choppers
• Lens Edges • Cold shields • Optical Mounts
• Dewars • Optical References • Detectors
• Solar Collector Absorbers • Planar Mirrors • Diffusion Screens
• Targets and Visual Cues • Radiators • Thermal Blankets
• Projection Screens • Instrument Enclosures
• Room Enclosures & Curtains· Standards of Reflectance
• Telescope Mirrors, Domes , Vanes , Baffles
• Solar Collector Concentrators
The construction types and processes covered by the materials and surfaces
include:
• paints and primers; lacquers and varnishes
• bulk materials (metals, glasses, polymers , ceramics, glass composites, etc.)
• films , foils, tapes, meshes
• fibres , papers, textiles, fabrics
• dyes, inks, marking materials
• surfaces and films created by processes such as:
- anodization, - electrodeposition,
- plasma spray deposition, -chemical vapor deposition,
- vacuum deposition, - proprietary processes, etc.
3.0 Scattering Data in SOLEXIS
A large part of the emphasis of SOLEXIS is scattering data , given its great
importance to stray light analysis in optical design, and since the need for such a
global resource has been recognized for many years [1,2]. The scattering data in
231
SOLEXIS consists of Bi-directional Reflectance Distribution Function (BRDF) and

Bi-directional Transmittance Distribution Function (BTDF) data. Both will be
referred to as Bi-directional Scattering Distribution Function (BSDF) data for the
remainder of this paper. A series of previous papers described various aspects of,
motivation for, and requests for involvement in the SOLEXIS database [7,8,9,10,
11, 12].
To have the BSDF data accessible on a CD in PC-Windows format is a major
milestone for the scattering and optics industry, since the need for this type of data
is urgent, not only for the aerospace and defense industries, but for all fields of
optics and photonics. The BSDF data presentation in SOLEXIS conforms to the
ASTM standard (American Society for Testing and Materials) for BSDF
measurements and data recording [3,4,5] . Some of the problems and solutions in
organizing BSDF data were outlined by Klicker et. al [6]. The data comes from: (1)
formerly proprietary, private sources in U.S.A., Canada, Russia; (2) publicly
accessible international research papers and reports; and (3) manufacturer's
research (published and unpublished). SORIC gathered most of the very large
privately-held BSDF and thermo-optical data resources in North America (i.e.,
thousands and in some instances tens of thousands of data). SORIC continues to
gather data, and in particular is compiling and requesting more information about
surfaces and materials manufactured outside North America.
4.0 Conclusion
The SOLEXIS BSDF and other data are accessible in two ways: as a stand-alone
CD-ROM, and in conjunction with existing commercially-available stray-light and
optical design and analysis software. Strong emphasis is placed on black and white
surfaces and materials in SOLEXIS. SORIC is currently working with Breault
Research Organization, Inc., Lambda Research Corporation, and OptiCAD
Corporation to interface SOLEXIS with the ASAP, TracePRO, and OptiCAD
optical design/analysis packages, respectively. Discussions are ongoing with
manufacturers of other prominent optical design and analysis software packages to
create similar linkages, and SORIC invites more.
SOLEXIS enables users to overcome the following problems formerly
encountered by those who tried to find data and/or select surfaces and materials for
ground- and space-based applications (particularly for black and white materials) :
• some of the old, reliable standby surfaces/materials are no longer available.
• properties data is difficult to find; manufacturer's data is often insufficient;
published data is scarce.
• the names and formulae of the surface preparations, particularly the paints,
have changed over time, causing confusion when one is trying to compare the
published or printed data with the presently commercially-available product.
• it is cumbersome and time consuming to identify the source of existing data;
creating an in-house database can generally take more time than the designer has to
select the surface.
232
• there are inaccuracies in the published data of which the designer may not be
aware.
• there are many selection criteria, and very little information on the selection
process.
SOLEXIS overcomes the above problems and also:
• substantially reduce the costs, time, and scope of a measurement program
and/or pre-selection process.
• substantially improves the ultimate system performance.
• minimizes the risk of system failure and, thus, the cost of damage control.
• substantially increase the comprehensiveness of proposals
SORIC created SOLEXIS with the generous assistance of a large number of people,
companies, and institutions in the optics and space community. The main
participants are listed below, and apologies are given to those who may have been
accidentally left out from the list.
• Stellar Optics Research International Corporation (SORIC)
• Canadian Space Agency (Optics Technology Section)
• The National Research Council of Canada's Industrial Research Assistance
Program
• Photonics Research Ontario , an Ontario Center of Excellence
• The Royal Bank of Canada.
• The SPIE (Society for Photo Optical Instrumentation Engineers) .
Several dozen expert advisors assisted SORIC with SOLEXIS, most of which
were listed in a previous paper[13] . There were two key, chief expert advisors; Dr.
Robert P. Breault, President and CEO, Breault Research Organization, Inc.,
(Tucson, AZ, USA);and Dr. Arthur W. Springsteen, Principal Scientist, Director of
Advance Development, Labsphere, Inc. (North Sutton, NH, USA).
The following institutions, listed in alphabetical order, donated digital data
curves and/or documents, funding, key software, and volunteer labor.
• Breault Research Organization, Inc., Robert Breault, Michael Gauvin,
Raymond Castonguay, Wayne Hall, Wally Davidson
• Canadian Space Agency (Saint-Hubert, QC, Canada) : Fernand Rheault, Lome
Schmidt
• Jet Propulsion Laboratory (pasadena, CA, USA): Paul Willis
• Labsphere, Inc. (North Sutton, NH, USA): Arthur Springsteen
• Oak Ridge National Laboratories (re: the Optical Materials Information
Service) (Oak Ridge, TN, USA): Marty Marchbanks, Brigham Thomas, Raymond
Miller
• Schmitt Measurement Systems in Portland , OR, USA (formerly Toomay
Mathis and Associates (Bozeman, MT, USA): John Stover, Kelly Kirchner, Marvin
Bernt, Marvin Ball
233
• University of Arizona, Optical Sciences Center (Tucson, AZ, USA): William

Wolfe, James Palmer
The following persons sent various reference papers and reports pertinent to the
database project:
• Bruce Banks, NASA Lewis Research Center (Cleveland, OH, USA)
• Leon Begrambekov (LIDASA Ltd. - Moscow, Russia)
• S-K. Chang (Spar Aerospace Ltd.,- Brampton, ON, Canada)
• M. Deshpande, Illinois Institute of Technology Research Institute (IITRI)
(Chicago, IL, USA)
• Robert Harned (Hughes Danbury Optical Systems, Inc., Danbury, CT, USA)
• Zelina Iskanderova (UTIAS, Toronto, ON, Canada)
• Thomas Leonard (Leonard Research Corporation, Dayton, OH, USA)
• Donald Shepard (Martin Marietta, Denver, CO, USA)
• Sheldon Smith (NASA Ames, Moffett Field, CA, USA)
• Timothy Wise (Opto-Mech Interface Organization, Boulder, CO, USA)
• Hundreds of manufacturers of black and white surfaces and materials, who
gave SORIC non-published product data.
Approximately 50 additional advisors, donors, and contributors are listed in
reference 13.
6.0 References
1. Breault, R., "Stray Light Technology Overview of the 1980 Decade (and a Peek Into the Future)." SPIE
Proceedings, pg. 2, Vol. 1331 , 1990 .
2. Breault, R., Current Technology of Stray Light, SPIE Proceedings, pg. 4, Vol. 675, 1986 .
3. ASTM Designation E 1392-90, Standard Practice for Angle Resolved Optical Scatter Measurements no
Specular or Diffuse Surfaces , ASTM Vol. 06 .01, p. 892 (Approved in 1990, Published in 1991)
4. Leonard, T . A , and Pantoliano, M., BRDF Round Robin, SPIE Proceedings, pg. 226, Vol. 967, 1988 .
5. Leonard, T. A, Pantoliano, M., and Reilly, 1., Results ofa C02 BRDF Round Robin, SPIE Proceedings,
pg. 444 , Vol. 1165, 1989 .
6. Klicker, Kyle A , Fuhrman, David M., Bjork, Donald R., A BSDF database , SPIE Proceedings, pg. 270 ,
Vol. 1331, 1990 .
7. McCall, S. Breault, R., Henrikson, R., Piotrowski, A, Rodney,1., Piotrowski, L., "A Black and White
Surfaces and Materials Database," SPIE Proceedings, Vol. 2260, 1994.
8. McCall , S., Begrambekov, L., Pierre, 1., "A Review of Black and Reflective Spectrally Selective Surfaces
Developed in the Former U.S.S.R.," SPIE Proceedin gs, Vol. 2260, 1994.
9. McCall, S., Pierre, 1., Mabee, A , Tennyson, R., Morison, D., Kleiman, J., Iskanderova, Z., Gudimenko, J.,
"The Atomic Oxygen Exposure Effects Module ofthe Database for the Properties of Black, White,
Reflective and Transmissive Spectrally Selective Surfaces," SPIE Proceedings, Vol. 2260, 1994 .
10. McCall , S., Sinclair, R., Pompea, S., Breault, R., "Spectrally Selective Surfaces for Ground and Space-
Based Instrumentation: Support for a Resource Base," SPIE Proceedings, pg. 497 , Vol. 1945, 1993.
11. Pompea, S., and McCall , S., "Outline of Selection Processes for Black Baffle Surfaces in Optical
Systems," SPIE Proceedings, pg. 92, Vol. 1753, 1992.
12. McCall , S., Pompea, S., Breault, R., Rcgens, N., "Reviews of Black Surfaces for Space and Ground-
Based Optical Systems," SPIE Proceedings, pg. 158, Vol. 1753, 1992 .
13.McCall , S. Breault, R., Henrikson, M. Reid, A 1. Clark, R. A Ellis, A Piotrowski, L. Piotrowski , 1.
Rodney, "A Globally Accessible Bi-directional Scattering Distribution Function Software Data Tool,"
SPIE Proceedings, Vol. 3426, 1998.
THERMAL CONTROL MATERIALS AND COATINGS
CURRENTLY BEING DEVELOPED IN FRANCE
FOR USE IN LOW EARTH ORBIT
F. GUERARD
MA P
Zone Industriel/e
Rue Clement Ader
09100 Pamiers
J.e GUILLAUMON
T/TS/AEIMTE/TH
CNES
18 Avenue Edouard Belin
31401 Toulouse Cedex 4
1.0 Introduction
CNES's Thermal department is working closely with the French company MAP to
develop materials and coatings for thermal control in space [1-4] . The materials
being developed are adhesives, varnishes, greases and coating products whose main
characteristic is the low outgassing rate in space. The thermal control coatings are
black, white or aluminum paints with special properties suited to the various orbits
used during space missions .
For the thermal control of low Earth orbiting (LEO) satellites , we have
developed different coatings particularly resistant to the conditions encountered in
this orbit (atomic oxygen, vacuum, temperature, UV, particle radiation, etc.).
Three of the latest coatings that can be used in LEO, namely a high-temperature
black paint, an OSR-type coating known as "RSF," and a varnish to protect Kapton
from atomic oxygen are in different stages of development and will be described
below.
2.0 Coatings Developed for Use in LEO
2.1 IDGH-TEMPERATURE BLACK PAINT
This paint is designed for application on the heat shield of a micro- or nanosatellite.
Heat shields are designed to protect the satellite when it is transferred from a
geostationary orbit to a lower orbit at an altitude of around 600 to 1000 km. The
235
J.1. Kleiman and R.C. Tennyson (eds.),
Protection ofSpace Materials from the Space Environment, 235--242.
236
transfer takes place through elliptic orbits with a perigee of 120 to 125 km. Initially
of 12 hours, the orbits are gradually circularized and are reduced to 90 minutes after
1 to 2 months. At the perigee, temperatures can reach 250°C for a perigee of 125 km,
and 600°C at 120 km. At such altitudes, atomic oxygen may lower emissivity, which
must remain above 0.85.
The high-temperature black paint developed by CNESIMAP consists of a
polyphenylsiloxane (PPS) binder cross-linked at between 300 and 600°C, and a
metal oxide pigment resistant to oxidizing atmosphere at high temperatures. The
paint adheres well to steel or aluminum alloys. It has been subject to the qualification
tests outlined below.
2.1.1 Thermal And Optical Properties

- Thickness of 25f.\.m: a s = 0.93
l::N = 0.79
- Thickness of 40f.\.m: as = 0.92
l::N = 0.88
2.1.2 Thermal Cycling In A Vacuum

- Test conditions: Low temperature: - 170°C
High temperature: + 130°C
- Number of cycles: 200 duration: 90 minutes
- Pre-test: Very good adherence
as = 0.93
l::N = 0.88
- Post-test: Very good adherence
as = 0.93
l::N = 0.88
2.1.3 Bombardment By Atomic Oxygen At High Temperatures

This test was performed in the CASOAR chamber belonging to CERT/ONERA
IDERTS in Toulouse [5].
The total fluence received by the heat shield in a highly elliptic orbit has been
estimated at 1021 atoms/em' if we consider that the shield may be exposed for 15 to
30 minutes to the fluence of oxygen molecules received continuously on one side
perpendicular to the velocity vector at an altitude of 120 km. The amount of fluence
received by exposed samples is estimated by exposing samples of a material with a
known reactivity (Kapton polyimide) over the bombarded surface, then measuring
the resulting erosion. As Kapton's thermal stability is limited to 400°C, the
temperature is restricted to 350°C so as not to bias atomic oxygen fluence
estimations.
The four paint samples are placed on the sample holder 30 em from the nozzle
so as to monitor any degradation in the coating, emissivity is measured in situ at
350°C, ex situ at 20°C, as is the solar absorption factor (as) and the loss in mass of
samples at various stages of bombardment. Table 1 shows the variations in solar
237
absorption factor (mean figures for the four samples) versus received atomic oxygen
dose.
TableI
ATOMIC OXYGEN FLUENCE as

atoms/em'
o 0.87
1.20 x 1020 0.86
4.00 x 1020 0.86
8.00 x 1020 0.86
The solar absorption factor (Us) varies very little between the initial state and
after exposure to atomic oxygen (there is a slight bleaching of around ~as = -0.01),
which shows that the high-temperature paint is very stable. Table 2 and Figure I
show the variations in infrared emissivity at 350°C (650 0K) and 20°C (300 0K)
(mean for the 4 samples) versus received atomic oxygen dose.
Table2
ATOMIC OXYGEN FLUENCE E @350°C

atoms/em'
o 0.85 0.85
1.20 X 1020 0.89 0.85
4.00 X 1020 0.90 0.86
6.50 X 1020 0.90
8.00x 1020 0.90 0.86
0.95 -
0.9 ..
0.95~
c.e +
: :
0.75 +
0 .7 T
0.55 -
- - eoslion 5S0K
0.55 - .,..-II-eg$llon JOOK
0 .5 - - - - - - _
o ss-ec
fluenee Caromeslem21
Figure I
Infra-Red Emissivity Versus AtomicOxygen Fluence
238
Infrared emissivity, especially at 350°C, significantly increases after exposure to

atomic oxygen, right from the first step. At 20°C, the increase in emissivity remains
small (+0.0 I). In both cases, this increase is beneficial as the aim is to obtain a highly
emissive coating.
This increase is explained by the erosion of the paint's surface, which increases
roughness and therefore infrared emissivity. Table 3 shows the variations in mass of
the samples measured in both mg and % (mean for the 4 samples) versus received
atomic oxygen dose.
Table3
ATOMIC OXYGEN FLUENCE MASS LOSS MASS LOSS

atoms/em' mg %
o o o
1.20 X 10'0 2.844 1.18
4.00 X 10'0 3.006 1.26
8.00 X 10'0 3.206 1.34
The loss in mass of the paint samples is around I per cent from the outset of
atomic oxygen bombardment. It is likely that such a loss is mainly due to paint
outgassing at 350°C, as the loss due to erosion by atomic oxygen is no doubt
considerably smaller. This is characterized by the results of the paint's outgassing test
according to standard ASTM E 59577/84, given below:
TML== 0.47 %
RML== 0.45 %
CVCM == 0.10 %
2.2. RSF
OSR or SSM-type solar reflectors are composed of quartz plates or polymeric metal-
plated film (silver or aluminum). These coatings are expensive, difficult to use and
have a relatively high surface mass (around 300 g/rrr'), In order to improve such
reflectors, we have sought inexpensive materials with a low density (around I) which
are both easy to use and able to be applied on metal deposits (such as polished
aluminum). Following is a description of the characteristics and properties of these
new coatings, currently in the industrial development stage.
2.2.1 Physical Characteristics

- Surface mass: RSF alone, thickness of 50~m: 50 g/m2
RSF on metal: 100 g/m2
RSF on metal-plated polymer : 80 g/m"
- Adherence : using adhesive 3M92: very good on aluminum
- Flexibility: good
2.2.2 Thermal And Optical Properties

- Solar absorption factor: aluminum RSF
239
a s= 0.10
- Infrared emissivity factor
EN = 0.80 to 0.85
- Optical transmission between 200 and 1000 nm: above 98%
2.2.3 Thermal Cycling In Vacuum

Test condition s: Low temperature: - 170°C
Number of cycles: 200 duration: 90 minutes
Pre-test: adherence .very good
a s =0.10
EN = 0.83
Post-test: adherence : very good
a s = 0.10
EN = 0.83
2.2.4 Outgassing / Contamination Test (ASTM E 595- 77/84)

TML=0.144 %
RML=0.106 %
CVCM = 0.043 %
2.2.5. Bombardm ent By Atomic Oxygen
These tests are currentl y underway at the CERTIONERAIDERTS in Toulouse. They

simulate bombardment by atomic oxygen with a fluence of 1020 atoms/em',
Intermediate results show that this coating does not suffer any degradation of
thermal or optica l properties, but we are awaiting confirmation on completion of the
tests.
2.2.6. Exposure To Solar Ultraviolet Rays
This test was perform ed at the CERTIONERAIDERTS [Ref6]. It consists of

exposing coating samples using a UV lamp which reproduces as faithfully as
possible the solar spectrum, outside the atmosphere, within the 200 to 300 nm and
300 to 400 nm bands. We simulated exposure to UV rays of around 1200 ehs.
Exposure to ultraviolet rays is continuous within intervals of 0, 50, 200, 500, 700,
1000 and 1200 EHS. Results: as seen in Figure 2 below, degradation of the solar
absorption factor a s remains below 0.01 . This is negligible and comparable to the
degradat ion of an OSR, but below that of an FEP/Ag SSM:~as = 0.01 to 0.02.
2.3 PROTECTING KAPTON FROM ATOMIC OXYGEN
Much work has been carried out to find a means of protecting Kapton from the
effects of atomic oxygen [Ref?, 8, 9]. We used a simple, inexpensive method able to
treat large surface areas of Kapton and easy to industrialize . Polysiloxane coatings
contain few easily-oxidized groups and are therefore little affected by the action of
240
atomic oxygen. The main characteristics and properties of the protective coating for
Kapton are given below .
..=
";j
;;. 0 .2a
.,
u
c
~
Q. 0 . 23
...e
II>
.c
-e o.ra
...
~
"0
CI)
0 . 13
~
=
"0
II>
.c 0 .0 8
<
.s
C.0 3
;
l """!=== I.. 2
0 .0 2
0 20 0 <0 0 800 sao 10 00 ' 2 00
Ultraviolet irradiation duration (ESH)
Figure2
RSF: Solar Absorptance Versus UV Irradiation Duration
2.3.1 THERMAL AND OPTICAL PROPERTIES

For a coating between 5 and 10 urn thick:
a s Kapton alone 0.3 8 a s Kapton + protection 0.38
EN Kapton alone 0.64 EN Kapton + protection 0.80
Transparency between 200 and 1000 run above 98%
2.3.2. MECHANICAL PROPERTIES

Adherence on Kapton: very good (adhesive 3M92)
Flexibility on a 4 mm 0 mandrel: very good.
2.3.3. Thermal Cycling In Vacuum

Test conditions: Low temperature: - 170°C
Number of cycles: 200 duration: 90 minutes
Pre-test: adherence: very good
241
eN = 0.80
Post-test: adherence: vel)' good
a s = 0.38
eN = 0.80
2.3.4 Outgassing / Contamination Test (ASTM E 595-77/84)

TML=0.1l4 %
RML =0.098 %
CVCM = 0.035 %
0 . 33
; 0 .2 3
c.....
o."
.c
<... 0 .18
=
Q
en
.; 0 .1 J
Q."
.c
<
.: 0 .08
c
.~
01
-5 0 .0 3
._ - - - -- ----- -
;>
·0 .0 2
o 2 00 40 0 6 00 800 10 0 0 120 0
Ultraviolet ir rad iation duration (ESH)
Figure 3
Solar Absorptance Of Kapton Protect ed Versus UV Radiation Duration
2.3.5 Bombardment By Atomic Oxygen

As for the RSF coating, tests are currently under way at the CERT/ONERAIDERTS.
The y simulate a fluence of atomic oxygen of 1020 atoms/em' . Initial results show that
the Kapton is protected efficiently. Full results will be availabl e soon.
242
2.3.6 Exposure To Solar Ultraviolet Rays

The relevant test has been described in section 2.2.6. Figure 3 gives the variation in
solar absorption factor (a.s) of the coated Kapton for total exposure of 1200 ESH.
Degradation is minimal (below 0.01) and does not affect the mechanical properties
of the film (adherence, flexibility).
3.0 Conclusion
This paper has presented the latest developments in the field of coatings for use in
LEO carried out by the CNES Thermal department in cooperation with French
company MAP.
The high-temperature black paint performs well in an oxidizing atmosphere
(atomic oxygen) at temperatures around 350°C. It is planned to use this paint on the
heat shields used for braking nano- or microsatellites.
The RSF coating performs well when exposed to ultraviolet rays and is
promising in that its surface mass is below that of SSMs or OSRs, it is easy to use
over large areas and is relatively inexpensive. This product is currently in the
industrialization phase in conjunctionwith potentialusers. The industrializedproduct
is planned to be fully qualified in 1999.
Using our process, the protective coating for Kapton may be used to treat large
areas inexpensively. The qualification, then industrialization of this process is
scheduled for late 1998 - early 1999, for marketingin 1999.
Finally, research and developmentwork to improve current products or produce
new coatings continues in order to improve performance and increase the lifetime of
space vehicles in low Earth orbits.
4.0 References
1) Guillaumon , lC., "New Space Paints," p. 239·244, 3rd Symposium On Spacecraft Materials On
Space Environment, ESTEC Noordwijk, 1-4 October, 1985,
2) Guillaumon, lC., Guerard, F., "Thermal Control Coatings," p. 427-433 5th Sympo sium On Spacecraft
Materials In Space Environment , Toulouse, September, 1991.
3) Guillaumon, J.C., Guerard , F., "Thermal Control Paints For Satellites," Protection Of Material s And
Surface Finishes From The Low Earth Orbit Space Environment, Toronto, February 20-21, 1992.
4) Guerard, F., Guillaumon, J.e "Thermal Control Paints And Various Materials For Space Use," p 457-
458, 7th Symposium On Materials In Space Environment, Toulous e, 16·20 June, 1997,
5) Viel, V., Paillous, A., Siffre, J., "Effet De L'oxygene Atomique Sur Un Bouclier De Protection
Thermique, ' Rapport CRiOXATOM, 17 September. 1997.
6) Marco, 1 , "Irradiat ions Par Ultraviolets Et Particules De Peintures,' Rapport Fevrier,1998.
7) Banks, B. A, "Considerations And Performance Prediction Of Leo Protection Coatings," Protection
Of Materials And Surface Finishes From The Low Earth Orbit Space Environment, Toronto,
February 20-21, 1992.
8) Pippin, G., "Space Station Protective Coating Development," Protection Of Materials And Surface
Finishes From The Low Earth Orbit Space Environment, Toronto , February 20.2, 1992 .
9) Wertheimer, M.R., "Multipurpose Protective Coatings For Spacecraft Applications." Protection Of
Materials And Surface Finishes From The Low Earth Orbit Space Environment, Toronto , February
20-21, 1992.
MODIFICA nON OF THERMAL CONTROL PAINTS BY PHOTOSIL1M
TECHNOLOGY
1.1. KLEIMAN, Y. GUDIMENKO, Z. ISKANDEROVA

Integrity Testing Laboratory Inc. (ITL)
4925 Dufferin Street, Toronto, Ontario, Canada, M3H 5T6
R. C. TENNYSON AND W.O. MORISON

University of Toronto Institut e for Aerospace Studi es (UT/AS)
4925 Dufferin Street, Toront o, Ontario, Canada, M3H 5T6
Abstract
A study was conducted at UTIAS/ITL to use the recently developed PHOTOSIL 1M

technology [I] for fast atomic oxygen (FAO) erosion resistance improvement of
polymer-based thermal control paints. A modified version of PHOTOSIL1M process
was used to treat the painted samples. Aeroglaze A276 white paint, Aeroglaze Z306
black polyurethane coating and epoxy-polyamide primer were used in the study. The
paints were applied to stainless steel coupons, approximately 50 urn thick. White
paints for thermal control application are generally used on the outer surfaces and black
paints on internal spacecraft components .
Atomic oxygen exposure testing was conducted at UTIAS. The samples were
exposed to fast atomic oxygen beam with energy - 3.0 eV at an average flux of 1.2
'\01 6 atoms/cm-s for a period of ten hours and were held at constant temperature for
the duration of the test. The paint density was estimated at 1.3 g/crrr' and the erosion
yield was calculated from mass loss measurements . SEM was used to control the
surface morphology changes. Solar absorptance and thermal emittance of a selected
number of samples, before and after PHOTOSIL 1 M treatment was measured according
to ASTM standards .
The results of surface radiative thermal optical properties are, usually, given as
the ratio of total solar absorptance (a) to thermal emittance (e), i.e. ale. The ale ratio
did not change significantly after the PHOTOSIL1 M treatment. Scanning electron
microscopy did not reveal any noticeable surface morphology changes . The mass loss
measurements failed to notice any changes in the mass in the PHOTOSIL1 M treated
samples after FAO testing.
243
1./. Kleiman and R.C. Tennyson (eds.},
244
1. 0 Introduction
The erosion effects of atomic oxygen (AO) on organic materials is a major limitation
for using resin binders in thermal control coating for space vehicles in LEO. The
degradation of the resin matrix by atomic oxygen (AO) is mainly due to polymer bond
breaking and subsequent molecular fragmentation to the final level of volatile products,
leading to the erosion of the polymer . It causes severe damage to thermal control
components through mass loss and degradation of their thermal-optical properties . It is
known that polyurethane-based paints on exterior surfaces of space structures erode
away under UV and AO exposure . Silicone based paints have been shown to exhibit
resistance to degradation by AO due to in situ generated oxide-based layer that
functions as a barrier to AO erosion. Many studies were performed on the effects of
AO and VUV on the properties of thermal control paints [2-8].
Aeroglaze 2306 is a flat black thermal control paint that consists of fumed
silica and carbon pigments in a polyurethane binder. The 2306 paint was the primary
thermal control coating on all LDEF interior structural members and experiment tray
bottoms. The 2306 on the interior surfaces, which were not subjected to AO or UV,
showed good durability. On exterior surfaces and the leading-edge tray clamps, the
2306 was almost completely eroded away from the composite substrate to which it was
applied . The red colouration characteristic of the primer pigment was visible, and
significant erosion into the composite substrate was observed. Based on the coating
thickness, the erosion rate of 2306 is estimated to be from 5'10-25 g-atom" and up to
2.6'10'24 g-atom", depending on the position of the sample with regard to the ram
direction [7]. In a much shorter exposure on the EOIM-3 mission the 2306 was also
eroded and the surface became more diffuse [5].
Aeroglaze A276 is a white thermal control paint made with titanium dioxide
pigment in a polyurethane binder that has been used on many short term space
missions. It was known to degrade moderately under long term UV exposure and to be
susceptible to AO erosion [2-5]. Approximately 100 A276 disks were measured for
absorptance and emittance making A276 one of the most extensively studied materials
on the LDEF. Samples painted with the A276 paint and protected from AO or facing
the trailing edge underwent darkening , changing from white colour to tan and
eventually dark brown after six years in LEO exposure. This nonrecoverable darkening
is due to a UV degradation of the polyurethane resin portion of the coating.
Unprotected leading edge samples remained very white. Apparently, as the exposed
A276 surfaces degraded they were also eroded by AO, leaving a fresh, undamaged
surface. The AO eroded the polyurethane portion of the paint, leaving behind paint
pigment particles. The total erosion depth was found to be on the order of 10 um, with
the erosion yield estimated to be (1-2.3) '10-24 g-atom" [7]. In a short space exposure on
the EOIM-3 mission, the A276 paint showed typical increased diffuseness due to the
erosion of the binder [3].
The LDEF results had shown that the absorptance of silica-protected materials
increases on orbit [2-6]. The selection process of materials and coatings applied to
spacecraft surfaces must take into account both the solar spectrum and the spacecraft
thermal radiation characteristics. In this regard a knowledge of two important
properties , solar absorptivity (ex) and infrared thermal emissivity (e), is necessary.
245
2. 0 Experimental
Two paints were used in the study, namely, Aeroglaze Z306 Black Polyurethane finish
and Aeroglaze A276 White finish, that have been applied over an Epoxy-Polyamide
Primer (per MIL-P-23377). The stainless steel surface was cleaned with Methyl Ethyl
Ketone. The two components constituting the primer were thoroughly mixed in their
individual containers. Mixing of Component I (pigmented solution) with Component II
(clear solution) of the primer, was in a one volume to one volume. A spray gun
(BADGER 250) was used to apply the primer. The drying time for the primer was more
than 7 hours under atmosphere conditions. The stainless steel support with the primer
dry film with thickness 10!lm was used for the black paint deposition and with the
primer film thickness about 25 um for the white paint deposition. The black and white
paints were deposited by the same procedure as the primer. "Z" series paints are
moisture curing systems. The curing process was conducted for 7 days at room
temperature and a minimum of 50% relative humidity. The dry films thickness was in
the range 25-38 urn for black paint samples and 51-76 urn for white paint samples.
The PHOTOSIL1M is a three-stage process consisting of a photo-activation,
liquid-phase silylation and stabilization stages [I] . The photo-activation experiments
were conducted in a commercially available UV/Ozone cleaning system (UVOCS), that
generates UV emission in the 185 and 254 nm range, in the air atmosphere. The films
were kept at a distance of 25 mm from the UV lamps. UV intensity (irradiance) was
measured using Black-Ray ultraviolet intensity meter that detects light of 220 to 290
nm (peak sensitivity at approximately 254 nm) and was found to be 2.8 mw /cm'. In
addition to existing organic molecules, ozone, atomic oxygen, and singlet oxygen are
generated under these conditions.
Liquid-phase silylation was carried out in a specially designed plastic vessel
by immersing the painted stainless steel substrate in a solution containing 77% decane
as the solvent and 23% bis(dimethyamino)dimethylsilane (B[DMA]DS) as a silylating
agent. After the treatment at the prescribed temperature and time, the film was rinsed
with decane and blown dry under atmosphere conditions.
Stabilization of the functionalised surface is required for converting
organosilicon layerto a more stable silicon-oxide type structure. The same set-up as for
photo-activation was used for stabilization of the functionalised surfaces. At this stage,
the silicone in the material binds with the active oxygen species and forms a protective
oxide-based surface structure. Carbon, hydrogen, and nitrogen atoms can be partially
eliminated, by forming volatile oxidation products [I].
The thermal-optical properties of two coupons with white and black paint
were evaluated before and after the Photosil process. The total solar absorptance, 0:,
was measured by first evaluatingthe reflectance of the surface. The reflectance, in turn,
was measured in accordance with the ASTM specification E-903. A
spectrophotometer manufactured by Beckman Instruments, Model DK-2A was
employed for these measurements. The normal thermal emittance (emissivity) E of the
samples was measured in accordance with the ASTM specification E-408 using an
infrared reflectometer manufactured by Gier Dunkle Instruments, Model DBI00.
Scanning electron microscopy (SEM) coupled with energy dispersive x-ray
microanalysis (EDS) were used to study the morphology and the elemental composition
246
of the investigated samples. The SEM was employed in both, the secondary electron (SE)
and the backscattered electron compositional (BEC) mode. The energy dispersive x-ray
microanalysis system is equipped with a light element detector. This allowed for detection
of all elements from boron and higher.
The attachment of the B[DMA]DS to the paints surface as well as the surface
composition after the stabilization were confirmed by x-ray photoelectron spectroscopy
(XPS) analysis. XPS measurements were obtained on a Leybold MAX 200 x-ray
photoelectron spectrometer with both a monochromatised Al Kn x-rays (1486.6 eV)
source and a non-monochromatised Mg Ko x-rays (1253.6 eV) source. The sources were
operated at 15 kV/30 rnA and 15kV/20 rnA respectively. High resolution spectra of
selected regions of interest were obtained with both sources. Typical full-width-at-half-
maximum for the Ag 3ds/2 peak are 0.67 eV for the AI source and 0.88 eV for the Mg
source. The energy range of the spectrometer was calibrated to place Cu 2P3/2 and Cu 3P3/2
at 932.67 eV and 75.16 eV respectively. Relative atomic % were derived from spectra
obtained in a low-resolution mode with the Mg source which were normalised to unit
transmission of the spectrometer by means of software supplied with the instrument. Peak
distortion due to different charging was not apparent, and high resolution peaks was
shifted to place the C Is maxima to 284.6 eV.
Water contact angle measurements were used at all stages of Photosil
treatment for controlling the incorporation of oxygen functionalities. Advancing contact
angles between the polymer surface and droplets of water were measured using 2-3 III
droplets. The atomic oxygen exposure testing was conducted at the University of
Toronto Institute for Aerospace Studies (UTIAS) . The samples were exposed to fast
16
atomic oxygen (FAO) with energy 3.0 eV at average flux of 1.2xl 0 atoms/crrrs for
20
a period of ten hours , i.e. to a fluence of _3' 10 at/em', and were held at constant
temperature for the duration of the test. More details on the FAO system can be found
in ref [6].
3.0 Results
The results obtained from the UV activation of the 2306 black paint samples showed
incorporation ; of oxygen functionalities (Table 1). The untreated film exhibits
hydrophilic properties with contact angle 78°. The UV treatment leads to reduction in
the advancing contact angle to 12°. Silylation of the UV activated black paint with
B[DMA]DS produces samples with silicon incorporated into the surface . The silylation
leads to formation of a hydrophobic surface with advancing contact angle of about
100°.
Table I. Watercontactangle with 2306 blackpaint surface photoactivatedfor varioustimes
Time, min 0 5 10 15 20
Contact angle, 78 60 49 13 12
deg
247
The stabilization stage belongs to oxidative-degradation processes that

eliminate hydrocarbon moieties and enrich the surface region with SiOx -based
structures (Table 2). Since the nitrogen signal detected by XPS on the surface
decreases, it can be suggested that a thick enough silylation region was formed. Table 2
shows a decrease in both CIO and C/Si ratios.
The results of surface radiative properties are, usually, given as the ratio of
total solar absorptance ( ex) to thermal emittance (e ), i.e. ex/e. As can be seen from
Table 3 the cde ratio did not changed significantly after the PHOTOSIL ™ process.
Table2. Surface composition (at%)of Aeroglaze 2306 blackpaint film by XPS
Sample Treatment XPS chemical content, % at

code
C 0 N Si F
Black paint Control 70.59 27.73 1.56 0.12 0
Photosil 34.39 39.89 0.49 25.23 0
Table 3. Summary of thermal opticalproperties for Aeroglaze 2306 black paintsample beforeand afterthe
PHOTOSIL™ process
Control Sample After PHOTOSIL™ Process
Total Solar Absorptance (ex ) 0.964 0.969
Thermal Emittance, ( e ) 0.887 0.883
ex I e 1.086 1.097
The photo-act ivation of the white A276 paint produced even stronger
reduction in the contact angle. As can be seen from table 4, the contact angles dropped
to 0 degrees after a 25 minute activation process. It should be noted, however, that
after the silylation and stabilization stages the angles returned to the values close to
pristine samples (see table 4).
248
Table 4. Water contact angle with the surface of the Aeroglaze A276 white paint after different stages of
PHOTOSIUM process
Treatment Photoactivation Silylation Stabilization
Time, min 0 5 10 15 20 25
Contact 92 80 59 36 10 0 112 98
angle, dea
As the XPS results indicate, the content of Si and 0 increased dramatically in

the upper layers of PHOTOSIL™-treated samples (see table 5), suggesting formation
of a SiOx-type of structure .
Table 5. Summary ofresults of surface composition (at%) of Aerogalze A276 white paint samples as
obtained by XPS
Sample Treatment XPS chemical content, % at

code
C 0 N Si F
White paint, Control 68.23 18.96 3.51 8.87 0.43

A276
Photosil 35.25 36.54 0.59 27.40 0.22
Table 6 presents the results of thermal optical properties measurements for the
A276 painted samples before and after the PHOTOSIL™ treatment. As can be seen
from the table, the thermal properties of the painted samples practically did not change
after the treatment.
Table 6. Summary of thermal optical properties for Aerglaze A276 white paint sample
before and after the PHOTOSIL™ Process
White Paint (Aerozlaze A276)

Before Photosil After Photosil
Total Solar Absorptance (a) 0.278 0.275
Thermal Emittance (E: ) 0.884 0.881
ole 0.31(4) 0.31(2)
Samples of both paints before and after the PHOSIL™ treatment were
exposed to atomic oxygen testing in the UTIAS facility. The PHOTOSIL™ treated
Z306 coupons changed their color slightly as can be seen from Fig. 1 that shows two
half-discs of Z306 painted samples that were tested in AO and put side-by-side for
249
comparison. As can be seen from Figure 1, the lower left quadrant that represents
an untreated sample masked from AO is slightly different in color from the lower
right quadrant that represents a treated sample masked from AO. However, after the
AO exposure, the color of the treated coupon returned to original (compare the
lower left quadrant with the upper right one in Figure 1). The results from the AO
exposure (Figures 1-3) indicated that the PHOTOSIL-treated samples did not
experience any AO erosion. Similar AO stability was confirmed by SEM studies for
the A276 paint.
Figure I
Aeroglaze Z306 painted couponsafter different treatments. Upperleft quadrant: Control,untreatedsurface,
maskedfrom AO during the exposure; Upperright quadrant: Controluntreatedby PHOTOSILand exposed
to AO (see text for details); Lowerleft: Treated by PHOTOSILand maskedfrom AO: Lowerright: Treated
by PHOTOSILand exposed to AO. Magnification 6x
250
Figure 2
Aeroglaze Z30 6 black paint control sample surfa ce not treated by PHOTOSIL after exposure to AO
(see text for details). Secondary SEM image. Magnification 2,000
Figure 3
Aeroglaze Z306 black paint control sample surfa ce treated by PHOTOSIL after exposure to AO
(see text for details). Secondary SEM image. Magnification 2,000
251
The erosion yields were remarkably low for the Z306 and A276 samples after
the PHOTOSIL™ treatment. Table 7 presents the measured yields for samples treated
by PHOTOSILTM process that underwent the AO exposure and compares them to the
erosion yield of Kapton" HN, a well studied, "standard" material.
Table 7: Erosion yields of Z306 and A176 paints as measured afterAO exposure
Material Erosion Yield Erosion Yield Compared

3/atom)
(cm to Kapton®HN
Z306 control 0.48'10.24 0.160

24
Z306 PHOTOSIL™ treated 0«0.05.10. ) 0 «0.016)
24
A276 control 0.39'10. 0.130
24
A276 PHOTOSIL I M treated 0 «0.002'10- ) 0 «0.0007)
24
Kapton® HN 3.0'10- 1.0
The changes in the thermal optical properties were also found to be very
small. Table 8 shows these changes, as obtained in the present work and compares them
to similar data from other sources [7,8].
Table 8: Summary of changes in thermal optical propert ies of Z306 and A276 paints.
Changes in Thermal
Paint Flown on Optical Properties Source
6.a 6.€
AK-512-w/Funct ional
equivalent ofA276 RCC- I· 0.0 0.0 [8]
A276 LDEF +0.34 -0.01 [7]
AK-243/Functional
equivalent of Z306 RCC-I -0.06 -0.01 [8]
A276 UTIAS FAO -0.005 -0.001 [this work]
Z306 tested 0.0 -0.004 [this work]
• - RCC stands for Removable Cassettes Container program on the MIR space Station [8).
4. 0 CONCLUSIONS
The PHOTOSIL™ treatment of coupons painted with Z306 and A276 thermal control
paints produced a stable, atomic oxygen resistant, Si-O-C rich surface on the
polyurethane binder of the paints. The surface modificat ion process reduced drastically
the erosion of the paints while leaving the thermal optical properties almost unchanged.
252
5.0 AKNOWLEDGEMENTS
The authors wish to acknowledge the help of MMO and CRESTech, two Provincial
Centers of Excellence in Ontario in supporting the development of the PRHOTOSIL™
process.
6. 0 REFERENCES
I. Gudimenko , Y., Iskanderova, Z., Kleiman, J.I., Cool, G., Morison, D.and Tennyson, R.C. ,Proceedings
of the 7th International Symposium on " Materials in Space Environment", Touluse, France, 16-20,
403-410, (1997).
2. Wilkes, D.R., Brown, MJ. , Hummer L.L. and Zwiener ,J.M., in Proceedings of the First LDEF Post-
retrieval Symposium, Ed. AS. Levine, NASA Conference Publication 3134 (2), 899, (1992).
3. Whitaker, AF., Kamenetzky, R.R., Finckenor, M.M. and Norwood, J.K., in Proceedings of the
Second LDEF Post-retrieval Symposium, Ed. AS. Levine, NASA Conference Publication 3194 (3),
1125, (1993).
4. Golden, J.L., in Proceedings of the Second LDEF Post-retrieval Symposium, 1993, Ed. AS. Levine,
NASA Conference Publication 3194 (3), 1099,(1993) .
5. Finckenor, M.M., Linton, R.C., Kamenetzky, R.R. and Vaughn, J.A , "Thermal Control Materials on
EOIM-3", LDEF-69 Months in Space, Proceedings of Third Post-Retrieval Symposium, November
1993, NASA Conference Publication 3275 Part 3,1025-1035.
6. Harada, Y., Mell, RJ., and Wilkes, D. R., "The Effect of the Space Environment on Thermal Control
Coatings", Proceedings Optical Surfaces Resistant to Severe Environments ed. S. Musikant, SPIE vol.
1330 , 90-101, (1990).
7. Reddy, R., Review: Effects of Low Earth Orbit Atomic Oxygen on Spacecraft Materials, J. of Mat.
Sci., 30, 281-307, (1995).
8. Tribble, AC. , Lukins, R., Watts, E., et al. "United States and Russian Thermal Control Coating
Results in Low Earth Orbit", Journal ofSpacecraft and Rockets, Vol. 33, No.1 , pp. 160-166, (1996).
9. Tennyson, R.C., "Atomic Oxygen Effects on Polymer-Based Materials", Canadian Journal of
Physics, 69, 1190-1208, (1991) .
SIMULATED SOLAR FLARE X-RAY AND THERMAL CYCLING
DURABILITY EVALUATION OF HUBBLE SPACE TELESCOPE
THERMAL CONTROL CANDIDATE REPLACEMENT MATERIALS
Kim K. de Groll, Bruce A. Banks

National Aeronautics and Space Administration
Lewis Research Center
Cleveland, Ohio 44135
Edward A. Sechkar, David A. Scheiman

NYMA,lnc.
Brook Park. Ohio 44142
Abstract
During the Hubble Space Telescope (HST) second servicing 1ll1SSlOn (SM2),
astronauts noticed that the multi-layer insulation (MLI) covering the telescope was
damaged. Large pieces of the outer layer of MLI (aluminized Teflon@ fluorinated
ethylene propylene (Al-FEP)) were cracked in several locations around the telescope.
A piece of curled-up Al-FEP was retrieved by the astronauts and was found to be
severely embrittled, as witnessed by ground testing.
The National Aeronautics and Space Administration (NASA) Goddard
Space Flight Center (GSFC) organized an HST MLI Failure Review Board (FRB) to
determine the damage mechanism of the Al-FEP in the HST environment, and to
recommend a replacement thermal control outer layer material to be installed on HST
during the subsequent servicing missions. Candidate thermal control replacement
materials were chosen by the FRB and tested for environmental durability under
various exposures and durations by GSFC and NASA Lewis Research Center
(LeRC). This paper describes durability testing at LeRC of candidate materials which
were exposed to charged particle radiation, simulated solar flare X-ray radiation and
thermal cycling under load. Samples were evaluated for changes in solar absorptance
and tear resistance . Descriptions of environmental exposures and durability
evaluations of these rnaterials are presented.
1.0 Introduction
The Hubble Space Telescope (HST) was launched on April 25, 1990 into low Earth
orbit (LEO) and is the first mission of NASA's Great Observatories program. It is a
telescope capable of performing observations in the near-ultraviolet, visible and near-
253
Ll. Kleiman and R.C. Tennyson ieds.),
254
infrared (0.115 toIOOO urn). The HST was designed to be serviced on-orbit to
upgrade scientific capabilities. The first servicing mission (SMI) occurred in
December 1993, after 3.6 years in space. The second servicing mission (SM2) was in
February 1997, after 6.8 years in space. Servicing missions are planned for mid 2000
and early 2003 [1].
The HST is covered with two types of thermal control materials, radiators and
multi-layer insulation (MLI) blankets, which passively control temperatures on-orbit.
Both of these thermal control materials have metallized FEP as the exterior layer.
During SMI , astronauts retrieved and returned Al-FEP and Ag-FEP MLI blanket
material. The Al-FEP (2 blankets retrieved) covered the two magnetometers, and the
Ag-FEP MLI covered the solar array drive arm (SADA). Analysis of the retrieved
blanket materials revealed that the FEP exterior layer was embrittled on high solar
exposure surfaces [2,3]. Surfaces which received the highest solar exposures (16,670
equivalent sun hours (ESH) for the Al-FEP, and 20,056 ESH for the Ag-FEP) had
some small through-thickness cracks in the 5 mil FEP at stress locations [2,3]. Solar
facing bonded 2 mil Al-FEP on the SADA power harness, which was also retrieved,
had many cracks and had lost total mechanical integrity (pulverization) [4].
During SM2, severe cracking of the MLI outer layer material was observed on
the Light Shield (LS), Forward Shell (FS) and equipment bays of the telescope.
Astronaut observations combined with photographic documentation of HST taken
during SM2 has revealed extensive cracking of the MLI in many locations around the
telescope, with solar facing surfaces being particularly damaged [11. The MLI outer
layer at several of the longest cracks was observed to be curled up and lifted away
from the next MLI layer. Figure I shows severely degraded MLI on HST as
photographed during SM2.
,
HST LS /
samp le, "
curled ...J
Figure I . Extensi ve tearing/curling damage lU MLi exterior layer on HST as photographed during SM2 .
The tightly curled-up piece was cut off by an austron aut and retrieved for analysis (SM2 LS sampl e) prior
to patching the area .
255
The worst of the MLI outer layer cracks were patched during the last EVA
(extravehicular activity) day. Contingency patches were placed over the cracks on the
light shield, Bay 8 and Bay 10. Due to limited material, lanyards were used on Bay 7
to hold the cracked MLI in place. Prior to patching the upper light shield crack, the
tightly-curled outer layer was removed and retrieved for post-mission analysis (see
Figure I). After SM2, GSFC established an HST MLI Failure Review Board (FRB)
to determine the cause of degradation of the FEP MLI on HST, and to identify a
replacement material to be installed during subsequent servicing missions. Extensive
investigation of the optical, chemical and mechanical properties of the retrieved SM2
FEP have been conducted and compared with SMI and pristine FEP results [1,5,6].
Simulated LEO environmental exposure testing of pristine FEP was also conducted to
help determine the cause of degradation ofFEP on HST [7,8].
The replacement thermal control material was required to meet HST's stringent
thermal requirements (end-of-life a../s ~ 0.28) and to maintain its structural integrity
for at least 10 years on-orbit. Candidate replacement materials were chosen and
various sets of materials were exposed to combinations of electron/proton radiation,
atomic oxygen, simulated solar flare X-rays, thermal cycling and near ultraviolet
(NOV) radiation at various facilities in order to evaluate their LEO durability. A
summary of all the sets of candidate materials has been written by Townsend et. al
[9]. The simulated solar flare X-ray and thermal cycling exposure testing which was
conducted at NASA Lewis Research Center (LeRC) is discussed in this paper.
2.0 Materials and Experimental Procedures
2.1 CANDIDATE MATERIALS
The HST MLI FRB members suggested and considered 17 replacement thermal
control materials. Each replacement material was scored using a multiplicative
evaluation formula based on nine performance criteria chosen by the board [9]. Six
candidate materials were chosen from the list of 17 for environmental durability
testing (materials 1-6). Because of the stringent thermal requirements (end-of-life
a../s s 0.28) replacement choices were limited. Several types of metallized FEP
bonded to scrim were chosen because of their excellent optical properties, and
because the radiator surfaces on HST (bonded FEP, as opposed to blanket material)
maintained their structural integrity as documented during SM2 [9]. Four additional
materials were also chosen for testing: the current AI-FEP MLI material (material 8)
to be used to verify environmental damage as witnessed on HST, an Optical Coating
Laboratory, Inc. (OCLl) coated sample (material 7) because it is used on some HST
exterior surfaces and HST management wanted to check its performance, and two
materials (9 and 10) chosen by the FRB Chair which were fundamentally different
than the first six materials [9]. These 10 materials are listed in Table 1. Materials 3, 4
and 6 (vacuum-deposited AI) were purchased from Dunmore, who used their
proprietary non-UV-darkening polyester adhesive. Materials I, 2 and 5 (silveri
Inconel) were purchased from Sheldahl, who used their proprietary non-UV-
darkening polyester adhesive. Material 7 came from GSFC stock, material 8 was sup-
256
Table I. HST Thermal Control Candidate Replacement Materials
Material Sample Id . Candidate Material

I BI.I and M2.1 10 mil FEP/Ag/Incon el/adhesive/Nomex" (polyphenylene isophthalate)
scrim
2 B1.2 and M2.2 5 mil FEP/Agllnconel/adhesive /fiberglass scrimladhesive/2 mil Kapton
3 B1.3 and M2.3 10 mil FEP/AI/adhesivelNomex scrim
4 Bl.4 and M2.4 5 mil FEP/AI/adhesive/fiberglass scrirn/adhesive/2 mil Kapton
5 B1.5 and M2.5 5 mil FEP/Agllncone1/adhesivelNomex scrim
6 B1.6 and M2.6 5 mil FEP/Al/adhesivelNomex scrim
7 B1.7 and M2.7 OCLI multi-layer oxide UV blockerl2 mil while Tedlar
8 B1.8 and M2.8 5 mil FEP/ AI (current HST material)
9 B1.9 and M2.9 Si0 2/ Al 20 3/ AgiAl 20 3/4 mil stainless steel
10 B1.0 and M2.0 Proprietary Teflon@FEP/AZ93 White Paint/Kapton
plied by Lockheed Martin Missiles and Space, material 9 was fabricated at GSFC,
and material 10 was provided by AZTechnology [9].
As previously mentioned, sets of samples were subsequently exposed to different
environmental exposures (electron/proton radiation, atomic oxygen, simulated solar
flare X-rays, thermal cycling and near-ultraviolet radiation) at various facilities. The
samples that were exposed to simulated solar flare X-rays at LeRC were B1.1-B1.8
and B1.0, and M2.1-M2.8. The samples that were exposed to thermal cycling at
LeRC were samples B1.1-B1.8 (samples M2.1-2.8 were thermal-cycled at GSFC).
Prior to environmental exposure, all samples were purposely cut through
approximately one quarter of the width, 5.08 em (2") from the top [91 . This was done
to provide a stress concentration area during environmental testing. Samples in the
'B ' set were exposed to charged particle radiation at the Boeing Company's
Radiation Effects Laboratory (Boeing). These samples were 3.81 em (1.5") wide and
12.7 em (5") long. Samples in the 'M' set were exposed to charged particle radiation
at the NASA Marshall Space Flight Center (MSFC) Space Environmental Effects
(SEE) facility. These 3.81 cm (2") wide samples were cut to 6.7 cm (2.6") in length
to fit in MSFC's facility.
2.2 OPTICAL PROPERTY CHARACTERIZATION
Optical properties were obtained using a Perkin-Elmer /..-9 spectrophotometer

equipped with a 150 mm integrating sphere. Total spectral reflectance and total
spectral transmittance were obtained from 250 to 2,500 nm. The spectral data were
convoluted into the air mass zero solar spectrum over the same wavelength range and
integrated to obtain solar total reflectance (Pt) and solar total transmittance (ft) . Solar
absorptance (a.s ) was calculated by subtracting Pt from 1 for the opaque samples, and
by subtracting Pt + Lt from 1 for the transmitting samples. Instrument repeatability is
±O.005.The thermal emittance (6) of Al-FEP retrieved during SMl did not change
257
with environmental exposure [2], and therefore was not measured in this test
program. Typically, the g for 5 mil Al-FEP is 0.81 [10].
2.3 SURFACE AND MASS CHARACTERIZATION
Mass measurements were obtained using a Sartorius balance. Three readings were
obtained and averaged for each measurement. The balance has an uncertainty of ±50
ug. Overall photographs of the samples were obtained using a Polaroid Land camera.
Optical micrographs and sample examinations were conducted using an Olympus
SZH stereo microscope. Micrographs were obtained at magnifications between 12
and 101 times. A JEOL 6100 scanning electron microscope (SEM) was used for high
magnification imaging. Samples were coated with a thin (r::;75A) palladium conduc-
tive coating prior to SEM examination .
2.4 CHARGED PARTICLE RADIATION
Samples were exposed to charged particle radiation at either MSFC or at Boeing prior
to environmental exposure at LeRC. At MSFC, the M samples were exposed to 10
years ' equivalent HST dose (SM3-201O) of electron and proton radiation [11].
Samples were exposed to 50 keY, 220 keY (for 10 mil FEP) or 500 keY (for 5 mil
FEP) electrons and 700 keY protons simulating the dose versus depth profile which
was calculated for HST. The samples (2 inch wide) were loaded under a 10 lb. tensile
load during exposure . Samples were designed to be stressed during environmental
exposures. A 10 lb. load was applied to the samples because that is the load (for a 2
inch wide sample) which approximates the stress calculated to be necessary to hold 5
mil FEP flat (on HST) when it would otherwise curl-up tight like the retrieved SM2
sample when unconstrained . At Boeing, the B samples were exposed to a 10 year
fluence of electron and proton radiation [12]. Rather than simulating dose versus
depth, Boeing simulated the total fluence using 40 keY electrons (3.1 x 1013 e/cm'')
and 40 keY protons (2.7 x 1010 p/cm-) . Ultraviolet radiation from a xenon arc source
(i.e. a continuum from 200-400 nm) illuminated the samples at approximately 1.2 UV
suns intensity during radiation exposure, resulting in approximately 5 UV solar hours
exposure . These samples (1.5 inch wide) were also loaded under a 10 lb. tensile load
during irradiation.
2.5 SIMULATED SOLAR FLARE X-RAY EXPOSURE
The X-ray fluences predicted for HST from SM3 to the year 2010 are 397.4 J/m2 for
1-8 A X-rays, 28.05 J/m2 for 0.5-4 A X-rays and 0.00967 J/m2 for 0.124-0.5 A x-
rays. These fluence values were calculated at GSFC based on GOES data, which
were extrapolated based on the l l-year solar cycle. The fluence is substantially
higher for the 1-8 Ax-rays (12,396-1,550 eV) than for the more energetic X-rays.
These lower energy solar flare X-rays are often referred to as soft X-rays . Although
FEP is transparent at visible wavelengths, it is highly absorbing between the vacuum
ultraviolet to soft X-ray range of 25 to 3000 eY [8]. If one looks at the X-ray
attenuation length versus photon energy for FEP, as seen in Figure 2, it becomes
258
apparent that energy below 3000 eV is primarily absorbed in the top 20 11m of the
film. Energy between 3000 eV and 6000 eV is primarily absorbed in the bulk of an
150
o '--~_~::::::.l -L ..J
o 2000 4000 6000
Photon energy, eV
Figure 2. X-.dYattenuation length versus photon energy (0-6000 ev j for FEP.
FEP film 127 11m thick. The transmittance versus energy curve for 127 11m thick FEP
shows that 84 percent of photons with energy of 10,000 eV pass through the film [8].
It was therefore concluded that irradiation with continuous X-rays below 10 keY
would provide the X-ray energies that should be necessary for bulk film degradation
as seen on HST.
The samples were exposed to simulated solar flare X-rays in a modified
electron beam evaporator system. A water-cooled molybdenum (Mo) target was
irradiated with a 10 keY electron beam. The target was angled to allow the highest
flux of X-rays to irradiate the candidate materials. The electron beam current was run
low enough to prevent any evaporation of the target material. A model AXUV-20HEI
absolute XUV silicone photodiode, produced by International Radiation Detectors
Inc., was used to measure the X-ray flux during each sample exposure run. The
photodiode has 100 percent quantum efficiency over the range of photon energies
produced by the source. Two sheets of 2 11m AI foil were used as a barrier between
the target and the photodiode and samples during the exposures. The AI foil blocked
energetic electrons from the target, and blocked the detector from visible light from
the electron beam emitter. A photographic cloth was used to block room light to the
detector.
To maximize the X-ray flux, only one candidate sample was exposed at a
time. With each candidate sample, a witness AI-FEP (5 mil FEP) tensile sample was
exposed to evaluate the effect of X-ray radiation on the tensile properties of AI-FEP.
Figure 3 is a drawing of the electron beam system configured to expose the HST
candidate samples to X-rays. Figure 4 shows the X-ray spectrum as a function of
energy for a 10 keY electron beam bombarding a Mo target, and with the X-rays
259
passing through a 4 um AI barrier. This spectnun was obtained using a windowless

energy-dispersive spectrometry (EDS) detector on a scanning electron microscope.
With an electron beam energy of 10 keY, an Mo target will not produce any K-alpha
characteristic radiation (17,476 eV), eliminating the very intense K-alpha charac-
teristic peak. Yet, as can be seen in Figure 4, there is still a fairly substantial Mo L-
alpha line at 2,293 eV (which appears wider in this spectra than the actual peak width
HST Candidate ;- FEP Witness

sample", / tensile sample
\ .......~ -4. _.. _
..:.-'C:. ".,..·.r,\'
t·.' . t~·~-·
,I"c:wr";r; C'"~,.·~-:;.·~':' •.
''r. ~l'''''\,U.. ·I~\.
· "1-y:.• •
•.';~ ·fl· ,I : ,l , l' ;!.,!•.!-'t.:.J• ..,I , ../''''/. a:c.~ r~·.....
·t ,'.
.' .- ,~
,. . ",.,.
" " ' / ' 1I.....'r"'ll
.-'- /;tl~\. i·"'· 1:1
. (.~'t!t.r."
f."T i
f.
It;
".1 '
I
, " rr '., • ~L"'("
• ' ·"' ·i··4 '~ 'f -t:.,... ,L r ......,. :
~1 (,I ~~::;.}:t-t·~I.t:!: 1::-;~tli
. l
.!
,'-~ .iu:i:,..:.. ·r...! '1-~""I'
...' / ,+ ·j·f·,Ll'.:-".j.:t-·r·!}·1
0 ·J-.L.·.. ·~->. ' j ~'~" r.~:7'.,
...., - ! ' .,',. t"-~ t..
1-"\,,,1,,,
. /.;~..
t' {~."
t.
l... ,I
"
J,IfF ~L 11......
~·hL+.J,.I .. ;- AI foil barrier shown
(t. :~ ..1.
~t~ t',I/
~·..!..:>-_ T••~/:.r
-(.11..
1
t''J' ' J .-..:;":!r-t ·n -;~· o-U-f + f..:.t. -!·::::tgr partly removed (actual
I I ~"'~ ·t' " ,'", I."~ v- s: ll'?1....., .... : ...-.1/:)'- I f '1 d I
'
Phatadiode -J " ':"'~~~!I ,, ~~~ U --' ·~;.":'L.t~:~" ..\1~";;
"';;'iI,'tj.J.~ r- .
.,.., i!o~'
:~ \ ~" I' ....'1'. '~.i!'.' ~,
.' • . / .~ ' . '.
01 covere samp es
and phatodiode)
'.
f".
,I
t:
~..... "''\. \.~.
... . ,. '~~""r:;'
""......~ ... ' ..
l ·",,~~t . .
•
I
[','
- - ..
<, ...
.. \
\
4
. t \ \
Ma x-ray
\ 'I
target~" "
I
~
'I
-,
J
i
Electron ....-..;;:........... I"

~..
beam --" '~.... ,.>
-" 'c .f··
.......17"0.
Figure 3. Drawing of the electron beam system configured to expose the HST candidate samples to
M,) x-ruys.
which is a detector artifact). As can be seen in Figure 4, there is Bremsstrahlung, or

continuous radiation between the energy range of 3000 to 6000 eV, and above, where
bulk damage is anticipated to occur in FEP. The actual fluence exposures for each
candidate sample are listed in Table 2. Due to electron beam gun controller
fluctuations, the [mal fluence values varied somewhat from the desired 397.4 J/m2
(457 ± 50 A). After simulated solar flare X-ray exposure characterization, the M
samples were returned to GSFC for thermal cycling.
260
2.6 THERMAL CYCLING
Samples B I.I-B I.8 were thermal-cycled in LeRC's Rapid Thermal Cycling Facility.
The facility consists of two smaller chambers, hot over cold held at fixed
temperatures, and frames that shuttle the samples between the two chambers . The hot
side is heated via resistive heater elements and a low-flow circulating fan to minimize
the gradient. The cold side is cooled using liquid nitrogen which is fed into the
chamber through a vented pipe; the nitrogen sprays away from the samples. Gaseous
1.0
0.8
.~
(/)
c:
-
Q)
.s
Q)
0.6
>
~
Q) 0.4
a:
0.2
0.0
Figur e 4. X -ray spectrum as a functi on of energy for a 10 keV electron beam impacting a Mo target . and
with the x-rays passing through a 4 11m AI barrier.
nitrogen fills the chambers and provides an inert atmosphere. Additionally, there is a
nitrogen line to provide cooling on the hot side and a heater on the cold side to
improve the temperature stability, and to provide some temperature profiling if
needed. Typically, samples are mounted to frames attached to lever arms. The lever
arms are air-driven which raises and lowers the samples between the two chambers .
A flange on the end of the frame seals the frame in each chamber, preventing thermal
leakage between the two chambers . The facility is computer-controlled ; all valves and
relays are operated based on temperature sensors in the chambers and on or near the
samples . This control allows the samples to cycle as rapidly as possible.
The candidate samples were cycled from _100° C to +50°C. The thermal mass of
the samples was extremely low and therefore, a thermocouple was not attached to the
samples but only placed freely next to them. The candidate samples were each
261
tensioned in metal fixtures under 10 lb. loads, (1,778 psi for sample B1.8) and four of
these fixtures were then mounted to each of the two chamber frames . The weight of
these chamber frames was above the capability of the air cylinders, and counter-
weights were added to the lever arms to obtain transition between the two
temperature chambers. The chambers are set up vertically and therefore must work
with or against gravity. The additional dynamic forces caused by the motion of the
frames during transition may have added some non-thermal stresses to the coupons
during cycling . These samples were cycled for 1000 cycles.
Table 2. X-ray Fluences and Tensile Properties for Witness AI-FEP
Sample X-ray Fluence Tensile Properties of Companion

(J/m 2 ) 5 mil Al-FEP Witness Samples
UTS(MPa) % Elongation
81.1 431 15.4 163

81.2 466 15.2 154
81.3 447 19.1 226
81.4 541 14.9 151
81.5 521 16.3 178
81.6 439 17.3 201
81.7 488 15.1 160
81.8 540 14.8 150
81.0 401 16.9 183
M2.1 411 16.6 190
M2.2 399 14.8 150
M2.3 429 18.3 217
M2.4 414 16.0 177
M2.5 463 14.8 154
M2.6 399 17.4 202
M2.8 515 15.4 161
Average 457 ± 50 16.1 ± 1.3 176 ± 25
5 mil FEPIAI 0 19.2 ± 1.8 197± 20
3.1 CHARGED PARTICLE RADIATION
Optical micrographs, overall photographs, mass measurements and optical properties

of the candidate materials after electron/proton radiation exposure by either Boeing
262
(B samples) or MSFC (M samples) were obtained at LeRC prior to simulated solar
flare X-ray exposure. Details of the effect of electron/proton radiation exposure on
candidate materials are reported in references II and 12. In summary, the MSFC
samples experienced small changes in solar absorptance (-0.001 to +0.004) and
generally slight mass losses, attributed to vacuum dehydration [12]. The Boeing
samples experienced small changes in solar absorptance (-0.003 to +0.006) with
radiation exposure [11]. The error of these solar absorptance measurements is 0.02,
indicating that no change in solar absorptance was measured after electron/proton
exposure. Mass measurements were not reported.
3.2 SIMULATED SOLAR FLARE X-RAY EXPOSURE (SAMPLES M AND B)
Therewere generally very small decreases in mass with X-ray exposure (-0.003 to -
0.426 %). The mass measurements are provided in Table 3. The optical properties
remained essentially unchanged with X-ray exposure (a pristine NIST second surface
mirror was optically characterized as a control). The optical properties are listed in
Table 4. Optical microscopy showed no propagation, or change in the purposely-
induced cut mark, or in the general appearance of the samples with X-ray exposure.
Table 3. HST Candidate Sample s Mass Measurements
Sample Post e-/p+ Mass Post % Post T.e. Mass % Change % Change
(g) X-ray Mass Change from Post from Post
X-ray e-/D+
B1.1 2.75763 ±O .OOOIO 2.75639 ± 0.00001 -0.045 2.75607 ± 0.00022 -0.012 -0.057
B1.2 1.95599 ± 0.00016 1.95516 ± 0.00004 -0.042 1.95445 ± 0.00004 -0.079 -0.079
B1.3 2.73842 ± 0.00039 2.7375 6 ± 0.00010 -0.031 2.73648 ± 0.00010 -0.039 -0.07 1
B1.4 1.93944 ± 0.00019 1.93884 ± 0.00008 -0.031 1.93713 ± 0.00003 -0.088 -0.119
B1.5 1.43776 ± 0.00009 1.43781 ± 0.00002 +0.003 1.43727 ± 0.00003 -0.038 -0.034
B1.6 1.43027 ± 0.00015 1.42919 ± 0.00003 -0.076 1.42852 ± 0.00004 -0.047 -0. 122
B1.7 0.48096 ± 0.00006 0.48087 ± 0.00002 -0.019 0.48066 ± 0.0001 -0.044 -0.062
B1.8 1.29512 ±0.00005 1.29487 ± 0.00008 -0.019 1.29569 ± 0.00004 +0.063 +0.044
B1.0 1.45386 ± 0.00005 1.44767 ± 0.00017 -0.426 - - -

M2.1 1.96282 ± 0.00002 1.96165 ± 0.00003 -0.060 - - -
M2.2 1.34966 ± 0.00007 1.34761 ± 0.00002 -0.152 - - -
M2.3 1.96205 ± 0.0000 1 1.96047 ± 0.00012 -0.081 - - -
M2.4 1.32761 ± 0.00001 1.32544 ± 0.00006 -0.163 - - -
M2.5 0.96760 ± 0.00010 0.96634 ± 0.00004 -0.130 - - .
M2.6 1.00858 ± 0 1.00706 ± 0.00016 -0.151 - - -
M2.8 0.88820 ± 0.00005 0.88817 ± 0.00003 -0.00 3 - - -
263
Tab le 4. HST Candidate Samples Optical Properties
Sample Post e"p · Post X-ray Absolute Post Thermal Change from
Change Cycling Post X-ray
p, a, p, a, a, p, a, a,
B1.1 0.923 0.077 0.923 0.077 0 0.860 0.140 +0.063
Bl.2 0.932 0,068 0.933 0.067 -0.001 0.930 0.070 +0.003
Bl.3 0.847 0.153 0.847 0.153 0 0.852 0.148 -0.005
Bl.4 0.814 0.186 0.814 0.186 0 0.812 0.188 +0.002
B1.5 0.927 0.073 0.924 0.076 +0.003 0.883 0.117 +0.041
Bl.6 0.858 0.142 0.858 0.142 0 0.865 0.135 -0.007
Bl.7 0.662 0.338 0.666 0.334 -0.004 0.667 0.333 -0.001

(*0.252) (*0.251)
Bl.8 0.889 0.111 0.888 0.112 +0.001 0.880 0.120 +0.008
NIST2023 0.883 0.117 0.883 0.117 0 0.883 0.117 0
B1.0 0.844 0.156 0.847 0.153 -0.003 - - -

M2.1 0.923 0.077 0.925 0.075 -0.002 - - -
M2.2 0.933 0.067 0.936 0.064 -0.003 - - -
M2.3 0.838 0.162 0.840 0.160 -0.002 - - -
M2.4 0.824 0.176 0.827 0.173 -0.003 - - .
M2.5 0.924 0.076 0.927 0.073 -0.003 - - -
M2.6 0.859 0.141 0.861 0.139 -0.002 - - -
M2.8 0,863 0.137 0.865 0.135 -0.002 - - -
NIST 2023 0.883 0.117 0.886 0.114 -0.003
• Sample has a transmittance of 0.064. which was not measured post e'(p" exposure; NIST 2023: second
surface AI mirror
There were small decreases in tensile properties of the companion AI-FEP

witness samples exposed to X-ray radiation. The ultimate tensile strength (UTS)
decreased from 19.2±1.8 to 16.1±1.3 MPa (2792±263 to 2349±l94 psi). The percent
elongation to failure decreased from 197±20 to 176±25 percent for pristine and X-ray
exposed samples, respectively. The tensile properties of the witness samples are listed
in Table 2. The average X-ray fluence was 457±50 11m2 , It has been suggested that
solar flare X-rays are primarily responsible for the degradation of FEP on HST [13].
The X-ray fluence (1-8 A) for HST at the time of SMI has been calculated by GSFC
to be 131.8 11m2 for the magnetometer surfaces which received 11,339 ESH
exposure. Yet, the UTS for this sample was 13.6 MPa (1974 psi) and the percent
elongation to failure was only 41.0 percent [8]. These results, along with additional
X-ray exposure tests reported in [8], and vacuum ultraviolet radiation and soft X-ray
exposure tests reported in [7], indicate that solar flare X-rays alone can not be the
primary contributor to degradation of FEP on HST. This is consistent with the fact
that MLI degradation occurred on all sides ofHST, not only solar facing sides.
264
3.3 THERMAL CYCLING (SAMPLES Bl.l-B1.8)
Samples were originally loaded under 10 Ibs. of tension. After 252 cycles for samples
Bl.l through B1.4, and 285 cycles for samples B1.5-B1.8, thermal cycling
automatically stopped and the chamber was opened up. Two samples had tom in half
(B1.2 and B1.4, both with fiberglass scrim and Kapton-backed). Sample BI.8 (the
current material AI-FEP with no scrim) had tom about 90 percent through the width
of the sample. The frames were then inverted to keep the samples from falling over
during additional cycling, resulting in a 9 lb. tensile load for the remainder of cycling.
Additional thermal cycling did not appear to cause further crack propagation. Figure
5 shows the candidate samples in the test facility after 1000 thermal cycles. The final
tension was determined for the intact samples, and ranged between 5.05 Ibs.
(sample B1.8) to 8.19 Ibs. (sample B1.7). Figures 6-8 show several samples as
received at LeRC (after electron/proton exposure) and after thermal cycling (and x-
ray exposure).
Generally, there were vel)' small decreases in mass with thermal cycling «0.047
percent). These results are listed in Table 3. The optical properties for the samples
after thermal cycling are listed in Table 4. Thermal cycling caused significant
absorptance increases in sample Bl.l (+0.063) and sample B1.5 (+0.041). These
samples have areas which appear yellowed with thermal cycling. Both of these
samples have Ag and the Nomex (vs. fiberglass) scrim. Sample BI.8 had an
absorptance increase of +0.008 with thermal cycling. Samples BU and B1.6 had
slight decreases in absorptance (-0.005 and -0.007, respectively). Both these samples
have AI and the Nomex scrim Samples Bl.2 and BI.4 (Ag and AI, respectively) with
fiberglass scrims had essentially no absorptance change.
Optical microscopy provided evidence that the yellowed area of samples B1.1
and B 1.5 are associated with the adhesive. Figure 9a shows that the yellowed area has
a distinct pattern. Figure 9b shows the same area, photographed with back-lighting.
Figure 9c shows the backside of the sample. It can be seen in Figure 9c that the
polyester adhesive in the rectangular regions in between the fibers has occasionally
cured with adhesive-voids.
Table 5. Initial Cut Length s and Tear Lengths after Thermal Cycling
Sample Initial Cut Tear Length Candidate Material

Id. Length(mm) (mm)
BI.l 10.2 (0.40") 1.8 (0.07") 10 mil FEP!Ag/Inconel/adhes ive/Nomex scrim
BI.2 9.\ (0.36") 28.7 (1.13")' 5 mil FEP!Ag/lnconel/adhesive/fiberglass scrimladhesive!2 mil
Kapton
BI. 3 9.4 (0.37") 5.3 (0.2\") \0 mil FEP!Al/adhesive/Nomex scrim

BI.4 9.\ (0.36 ") 28.7 (1.13")' 5 mil FEP!Al/adhesive!fiberglass scrimladhesivel2 mil Kapton
B1.5 8.9 (0.35") 2.8 (0.11") 5 mil FEP!Agllncone l/adhesivelNomex scrim
BI. 6 9.\ (0.36") 5.\ (0,20") 5 mil FEP!Al/adhesiveINomex scrim
81.7 9.9 (0.39") 5.3 (0.2\") OCLI multi-layer oxide UV blocker!2 mil white Tedlar
81.8 9.4 (0.37") 24.6 (0.97") 5 mil FEP!AI (current HST material)
, Sample tore in half
265
81.1 81.2 81.3 81 .4
81.5 81 .6 61.7 61.8
Figure 5. Candidate samples (B 1.1 - B 1.8) unde r tension in the Rapid Thermal Cycling Facility after
1000 thermal cycles .
266
(a)
(b)
Figure 6. Candidate sa mple B 1.4 (5 mil FEP/AI/tibe r glass scrirn/adhesive/2 mil KapIOn®),
(a) After e-/p+ radiation exposure . (b) After e-/p+ radiation exposure, x-ray exposure
and thermal cycling.
267
(a)
(b)
Figure 7. Candidate sample 131.6(S mil FEP/Al/adhesive/Nomex® scrim). (a) After e-/p+
radiation exposure. (b) After e-/p+ radiation exposure, x-ray exposure and
thermal cycling.
268
(a)
(b)
Figure ll. Candid ate witne ss sample /) l .lll5 mil FEP/AI). (a) After e- /p+ radiation expo sure.
(b) After e-/p+ radiation exposure , x-ray expo sure and thermal cycling.
269
~.
:~;~
'~
it
.".,.
• "
I ~j
o:. __k-
\.:-
-..
~ ~
..~),.~
r;,.,....
~
..".
~
~ I I '. ;
/
-..... -.....
I~ Yellowed-+I O.8mm (b) O.8mm
area
(c) ....... (d)

0.8mm 0.1 mm
Figure 9. Candidate sample 81,5 after thermal cycling showing discoloration pattern. (a) Yellowed disco lored
area. (b) Same area as 9(a) with back lighting . (c) Backside of sampl e showing adhesive void areas . (d)
Close-up of an adhesive void (cracks in metal layers).
270
A close-up of such a void in the adhesive is shown in Figure 9d. The fine lines
are cracks in the metal layers due to thermal cycling. Looking at the yellow pattern in
9a and the pattern of transmitted light in 9b associated with adhesive-voids, it
becomes apparent that the discoloration is associated with the adhesive . It is not clear
why the discoloration was not across the entire sample. As previously mentioned,
both the discolored samples have Ag and Nomex scrims and were supplied by
Sheldahl with their non-UV darkening adhesive (FEP/Agllnconel/adhesivelNomex
scrim) . The Ag sample supplied by Sheldahl with a fiberglass scrim and Kapton
substrate (B 1.2) did not discolor even though it had the same adhesive
(FEP/ Ag/lnconel/adhesive/fiberglass scrim/adhesivelKapton). The AI samples
provided by Dunmore with AI and Nomex scrim (B1.3 and B1.6) did not yellow with
thermal cycling (FEP/ Al/adhesivelNomex scrim) . It therefore appears that there is a
discoloring problem with the SheldahJ adhesive and Nomex scrim upon heating.
Optical microscopy was used to document the extent of damage which occurred
at the purposely-cut area with thermal cycling under load. The initial cut lengths and
tear lengths propagated during thermal cycling are listed in Table 5. The fiberglass/
Kapton samples (B1.2 and B1.4) tore in half. The tear lengths for the Nomex scrim
samples ranged from 1.8 rom (0.071") to 5.3 rom (0.209"). The Ag/lnconel Nomex
samples (Bl.l and B1.5) had shorter tears than the AI Nomex samples (B1.3 and
B1.6) of the same FEP thickness. The Nomex scrim on the Al samples was not as
orthogonal as the AglInconel samples (as can be seen by comparing B1.1 and B 1.6 in
Figure 5) which may have contributed to the small additional tearing . Sample Bl.7
had a very fine hairline crack which was 5.3 rom long, and the witness AI-FEP
sample (Bl.8) had a tear length of 42.6 rom. Figure 10 shows tearing of the cut tip in
sample B 1.6. The backside of the sample shown in Figure lOb shows the tear
extending to the second fiber bundle, with no damage to the uncut fiber bundles near
the propagated tear.
Several tests were run to understand the exact mechanism responsible for the tear
propagation of the thermal-cycled samples. Because the tensile properties of FEP
decrease with temperature, loaded samples were exposed to the high end temperature
(50°C) of the thermal cycling range to see if tearing which occurred with thermal
cycling was due to decreases in tensile strength with temperature . Four 5 mil AI-FEP
samples with identical sample dimensions and purposely-cut marks as the candidate
samples were loaded under tension (2.5, 5.0, 7.5 and 10 lb. loads) and exposed to
50°C for 90.5 hours (approximately the time the candidate samples were thermal-
cycled) . There was no observed propagation of the cut with exposure under any of the
applied loads. AI-FEP samples which were exposed to 50°C under load were then
brought to liquid N2 temperature while under tension to see if the additional load due
to contraction caused tear propagation . There was no observed propagation of the cut
with exposure under any of the applied loads.
Four AI-FEP (5 mil) samples, which had not been previously exposed to
electron/proton or X-ray exposures , were thermal-cycled under varying loads (2.5,
5.0, 7.5 and 10.0 Ibs.). These samples also had geometry and cut marks identical to
the candidate samples . After 250 cycles, the facility was stopped and opened up. No
signs of tear propagation were observed at 250 cycles for any of the samples .
271
Figure 10. Candidate sample B1.6 after thermal cycling. (a) Close up of purposelycut region
showing fine tear which propagated duringthermal cycling. (b) Backside of
sampleshowing the tear extending to the secondfiber bundle beyond the cut.
272
After 1,000 cycles, small or no tears were observed for the 2.5, 5.0 and 7.5 lb.
loaded samples, while the 10 lb. loaded sample was tom almost as far as candidate
sample B1.8. Because of the thermal cycling test set-up, additional loads are imposed
on the samples as they move from the hot to the cold chamber, and vice-versa.
Therefore, another set of 5 mil AI-FEP samples were prepared and mechanically-
cycled under load (2.5, 5.0, 7.5 and 10 lb. loads) in the thermal cycling facility (at
room temperature). After 250 mechanical cycles, the 2.5 and 5.0 lb. samples had no
tear initiation, while the 7.5 and 10 lb. samples had very small tear initiations. After
1000 cycles, there was no tear propagation in either the 7.5 or the 10.0 lb. samples.
These results indicate that the tearing which occurred in sample B 1.8 was due to
thermal cycling (versus thermal exposure) under the high load (10 lb.). The prior
radiation exposure does not appear to have contributed to tear propagation in samples
B1.8 because radiation exposure was not necessary to cause tear propagation. Also,
mechanical cycling under load was not sufficient without thermal cycling to cause
tear propagation.
The cracks which propagated on HST were found to have very smooth surfaces.
Figure 11 shows examples of these smooth, featureless crack morphologies. Figure
lla is an electron micrograph of an in-space-propagated crack from the LS MLI
sample retrieved during SM2. Similar smooth crack surfaces were found on cracks
located near cable holes in the magnetometer MLI retrieved during SMl, as shown in
Figure llb and llc. These smooth cracks are believed to have occurred due to slow
crack growth under low stress in the presence of a degrading environment [9].
Similar findings were observed for the SADA MLI retrieved during SMl, and have
been described as being similar to stress-cracking of glassy materials [4]. When the
embrittled 5M2 MLI sample is purposely cracked by bending the FEP space-exposed
surface in tension, the crack surface is not smooth and glassy in appearance, but more
fibrous as seen in Figure 12. This also supports the idea that the smooth cracks
occurred slowly in the space environment. The Al-FEP sample which was thermal-
cycled under a 10 lb. load, with no prior radiation exposure , had a very different
surface morphology. Its crack surface has wave-like striations, as seen in Figure 13.
273
Figure II . Smoothcrack morphologies of in-space propagated cracks

in HST FEP. (a) SM2 LS FEP. (b) SMI MSS FEP near a
cable hole. (c) Close-upof crack surface in II (b).
274
,- ... -
Y"'4: :..:::-
~
'- L.J
-10 rnm
Figure 12. 5 M2 FEP after bending the space exposed FEP surface in tension. Th e tension induced
fast-propag ated crack has a fibrous. textured surface morphol ogy.
275
Figure 13. Surface morphology of FEP tear propagated during thermalcycling under load.
This surface has ductile appearing, wave-like striations.
276
This sutface is much more ductile than the glassy-looking sutfaces which
cracked in space. The crack morphology of this thermal-cycled sample is identical to
that of sample Bl.8, which did receive prior radiation exposure [9). The additional
radiation exposures (electron/proton and X-ray) did not affect the crack propagation
morphology in this sample. The smooth crack morphology which occurred on HST
was not observed for the Bl samples which were thermal-eycled under tensile
loading, although one sample with scrim backing (B 1.5) had areas which were
relatively smooth [9). However, relatively smooth crack morphology was observed
with most of the GSFC-tested samples containing scrim which were thermal-eycled
under constraint loading (taped at the edges) and experienced crack propagation (9).
Based on the environmental testing conducted at LeRC, candidate samples Bl.3
and Bl.6 (AI and Nomex scrim, prepared by Dunmore) performed the best. Samples
Bl.2 and Bl.4 (with fiberglass scrim and Kapton substrates) tore in half during
thermal cycling under load. Samples BI.I and B1.5 (with Ag and Nomex scrim,
prepared by Sheldahl) had significant increases in solar absorptance with thermal
cycling which were attributed to an interaction of the adhesive and the Nomex scrim
with heating. Sample Bl.7 (an OCLI-coated sample) has a high initial absorptance,
and was tested because the HST project wanted to check its performance. Because the
solar absorptance ofBl.6 was lower than Bl.3, the HST MLI candidate sample Bl.6
was considered the single best performer (using solar absorptance as the determining
factor between Bl.3 and Bl.6).
4.0 Additional Testing
After sample characterization at LeRC (mass, optical properties and optical

microscopy) the candidate samples were returned to GSFC for further testing and/or
analyses. The M2 sample set was thermal-cycled under constraint loading at GSFC
and then exposed to NUV radiation. Other candidate sample sets were exposed to
environments including electron/proton radiation, atomic oxygen and thermal cycling
(sample set MI), thermal cycling of large samples (samples set Ll) and NUV
radiation (sample set 01) (9).
5.0 Final Ranking
When testing of all candidate samples was completed, the HST MLI FRB met to
review the results and re-vote on the candidate materials using the original
multiplicative evaluation formula based on the original nine performance criteria.
During the intervening months, the HST project decided to use the sample 9 material
(Si021AI20/ AgiAI20/stainless steel) as the outer layer on all bay repairs. The HST
MLI FRB then evaluated the test results for use on the HST LS and FS areas only.
Based on this criteria, two of the original 10 materials (materials 9 and 10) were not
considered in the final evaluation. Material 9 is not practical for the repairs because
of handling (stainless steel sheets), and material 10 had problems with particulate
contamination and UV darkening [9).
277
The final ranking of the candidate materials is listed in Table 6. Material 6 (5 mil
FEP/Ag/adhesivelNomex scrim) was ranked first, and recommended by the FRB as
the replacement thermal control material to be installed on HST during SM3. Because
of concerns of UV darkening of the adhesive, some of the material 6 samples
underwent additional NUV exposure testing at GSFC. Five samples, were exposed
for 2000-3000 ESH. No change in solar absorptance was measured.
Table 6. Final Ranking ofHST Thermal Control Candidate Replacement Materials
Rank" Material Candidate Material

1 6 5 mil FEP/Al /adhesivelNomex scrim
2 3 10 mil FEP/AI/adhesivelNomex scrim
3 8 5 mil FEPIAI (current HST material)
4 I 10 mil FEP/AglInconeIladhesivelNomex scrim
5 5 5 mil FEPIAg/Inconel /adhesive/Nomex scrim
6 2 5 mil FEPIAg/Inconel /adhesive/fiberglass scrimladhesive/2 mil Kapton

7 7 OCLI multi-layer oxide UV blocker /2 mil white Tedlar
8 4 5 mil FEPIAl /adhesive/fiberglass scrimladhesive/2 mil Kapton
• I is the most favorable candidate
6.0 Summary and Conclusions
The outer layer of ML! (5 mil Al-FEP) on HST is degrading in the space
environment. An HST ML! FRB was chartered to determine the cause of degradation
of the FEP, and to recommend a replacement material. The replacement thermal
control material was required to meet HST's stringent thermal requirements (end-of-
life als s 0.28) and to maintain its structural integrity for at least 10 years on-orbit.
Candidate replacement materials were chosen through a multiplicative
evaluation formula based on nine performance criteria for durability testing. Various
sets of candidate replacement materials were exposed to combinations of
electron/proton radiation, atomic oxygen, simulated solar flare X-rays, thermal
cycling and near ultraviolet radiation at various facilities in order to evaluate their
HST on-orbit durability. Two sets of samples (BI and M2) previously exposed to
charged particle radiation were exposed to X-rays at LeRC. One sample set (Bl) was
also thermal-eycled under load at LeRC.
There were very small decreases in mass, and no changes in optical properties,
and no change in the original cut, of the candidate materials with simulated solar flare
X-ray exposure to IO-yearHST fluences (SM3-201O). Small decreases were observed
in the tensile properties of companion 5 mil Al-FEP witness samples with X-ray
exposure. Generally, there were very small changes in mass with thermal cycling.
Thermal cycling caused significant absorptance increases in samples Bl.l (+0.063)
and B1.5 (+0.041) which were attributed to an interaction of the Nomex scrim with
278
the Sheldahl adhesive upon heating. Samples BU and B1.6, with AI and Nomex
scrims had absorptance decreases with thermal cycling (-0.005 and -0.007, respec-
tively). Samples Bl.2 and Bl.4, with fiberglass scrims and Kapton substrates, tore in
half during thermal cycling under load. SampleB1.8 (currentHST MLI material) tore
about 90 percent of the width during thermal cycling. Tear propagation of the BI
samples was attributed to thermal cycling under a high load. The prior radiation
exposures did not appear to have an additional effect on tearing, and no tearing
occurred due to mechanical load cycling. The crack morphology of sample B1.8 (5
mil AI-FEP) exposedto radiation, then thermal-cycled under a high load (10 lb.) was
ductile in appearance, unlike the glassy crack morphology of the brittle FEP from
HST. The crack morphology of radiation-exposed scrim containing samples thermal-
cycled under a low load (constraint loading) at GSFC most closely resembled in-
space-propagated cracks from retrieved HST FEP. Based on X-ray and thermal
cycling testing at LeRC, samplesBU and B1.6 performed the best, with B1.6 having
a lower solar absorptance.
Upon completion of testing all candidate samples, the HST MLI FRB met to
review the durability results and re-vote on the candidate materials using the original
multiplicative evaluation formula. Material 6 (5 mil FEP/AI/adhesivelNomex scrim)
was ranked first, and recommended by the FRB as the replacement thermal control
material to be installed on HST during SM3.
7.0 Acknowledgments
The authors would like to thank the following for their contributions to this program:
Tom Stueber of NYMA Inc. and Alice Jalics of ClevelandState University for tensile
tests; Kevin Malinowski of Cleveland State University for assistance with thermal
cycling; and Dr. Steven Pepper of NASA Lewis Research Center for X-ray spectrum
characterization. We would also like to thank Patti Hansen (the HST MLI FRB
Chair), Jackie Townsend of NASA Goddard Space Flight Center, and Joyce Dever of
NASALewis Research Center for helpfuldiscussions.
8.0 References
1. Hansen, Patric ia A, Town send. Jacqueline A., Yoshikawa Yukio, Castro, David J., Triolo, Jack 1. and
Peters, Wanda C., "Degradation of Hubble Space Telescope Metallized Teflon FEP Thermal Control
Materials," Science of AdvancedMaterials and Process Engineering Series, 43, 1998.
2. Zuby , Thomas M., de Groh, Kim K. & Sm ith, Daniela C., " Degradation of FEP Thermal Control
Material s Returned from the Hubble Space Telescope," Proceedings of the Hubb le Space Telescope
Solar-Array Workshop, Noordwijk, The Netherlands, May 30-31, 1995, ESA WPP-77, pp. 385-402;
also NA&4 TM-104627, Decemb er 1995.
3. De Groh, Kim K. & Smith, Daniela C., "Investigation of Teflon FEP Embrittlement on Spacecraft in
Low Earth Orbit," Proceeding s ofthe 7th International Symposium on Materials in Space Environment.
Toulouse, France , June 16-20, 1997, ESA SP-399 ; also NA&4 TM-l13 153, 1997.
4. Van Eesbeek, Marc, Levadou , Francois & Milintchouk, Andrei, "Investigation on FEP from PDM and
Harness from HST-SA1 ," Proceeding s of the Hubble Space Telescope Solar-Array Workshop,
Noordwijk, The Netherlands, May 30-31 ,1995, ESA WPP-77, pp. 403-4 16.
5. Townsend, Jacqueline A., Hansen, Patricia A., Dever, Joyce A. and Triolo, Jack J., "Analysis of
279
Retrieved Hubble Space Telescope Thermal Control Materials. Science of Advanced Materials and
Process Engineering Series . 43, 1998.
6. Dever, Joyce A, de Groh, Kim K , Townsend, Jacqueline A and Wang, L. Len, " Mechanical Properties
Degradation of Teflon FEP Returned From the Hubble Space Telescope," Presented at the 36th
Aerospace Sciences Meeting & Exhibit, Reno, NV, January 12-15, 1998, AlAA-98 -0895, NASA
TM-1998-206618 .
7. Dever, Joyce A , Townsend, Jacqueline A, Gaier, James R. and Jalics, Alice I., "Synchrotron VUV and
Soft X-ray Radiation Effects on Aluminized Teflon FEP," Science ofAdvanced Materials and Proce ss
Engineering Series, 43. 1998.
8. Banks, Bruce A , de Groh, Kim K, Stueber, Thomas J. and Sechkar, Edward A , "Ground Laboratory
Soft X-ray Durability Evaluation of Aluminized Teflon FEP Thermal Control Insulation," Science of
Advanced Mat erials and Process Engineering Series, 43. 1998.
9. Townsend, Jacqueline A , Hansen, Patricia A, McClendon, Mark W., Dever, Joyce A and Triolo, Jack
J., "Evaluation and Selection of Replacement Thermal Control Materials for the Hubble Space
Telescope," Science ofAdvancedMaterials and Process Engineering Series. 43, 1998.
10. Henninger, John H., " Solar Absorptance and Thermal Emittance of Some Common Spacecraft
Thermal-Control Coatings," NASA RP-1121, 1984.
11. Edwards, David L. and Vaughn, Jason A, "Charged Particle and Atomic Oxygen Exposure of
Candidate Replacement Materials for the Hubble Space Telescope," NASA MSFC Test Report, October
15, 1997.
12. The Boeing Company, " Radiation Exposure Testing - HST Thermal Blanket," Boeing Test Report,
November, 1997.
13. Milintchouk., Andrei, Van Eesbeek., Marc, Levadou, Francois and Harper, Tim, " Influence of X-ray
Solar Flare Radiation on Degradation of Teflon in Space," J. ofSpacecraft and Rock ets, Vol. 34, No.4,
July-August 1997, pp. 542-548.
RADIATION-INDUCED EFFECTS IN SIO z PROTECTIVE COATINGS
ON POLYMERIC SPACECRAFT MATERIALS
C. COlA, AC. FOZZA, AND M.R. WERTHEIMER*

Groupe des Couches Minces.
Departement de Genie Physique et Genie des Materiaux,
Ecole Polytechnique , Montreal. QC H3C 3A 7 • Canada
G. CZEREMUSZKIN
Polyplasma Inc. Montreal. QC H3T IP I. Canada
A HOUDAYER
Groupe des Couches Minces et Laboratoire des Faisceaux loniques,
Universite de Montreal, Montreal, QC H3C 3J7
*Corresponding author ,' e-mail : mwertheimer@mail.polymtl.ca
Abstract
It is now well-recognized that polymeric materials used for the construction of low
Earth orbital (LEO) spacecraft can be protected against atomic oxygen (AO) erosion
by thin protective coatings of Si02, for example. We deposit such coatings by
Plasma-Enhanced Chemical Vapor Deposition (PECVD), a technique which results
in extremely strong covalent chemical bonding at the substrate/coating interface.
However, the thin sub-urn Si02 coating is transparent to both ultraviolet photons (Iv
;:: 180 nm) and to energetic particles such as keV and MeV protons, quite abundant
in LEO. We report the results of experiments in which SiOrcoated polymers
(Kapton®, Polyimide and Mylar® PET) are exposed to two types of high-energy
radiation: on one hand, VUV-VIS photons (Iv ;:: 112 nm) from a hydrogen plasma
lamp, which have been shown to give rise to photochemically-generated gaseous
reaction products from the polymer and, on the other hand, 100 keV protons from the
Tandetron accelerator at the Universite de Montreal, corresponding to multi-year
exposure in LEO. Possible modification or damage is assessed on the basis of
mechanical testing (using a CSEM Microscratch Tester), and by microscopy,
compared with untreated witness samples.
1.0 Introduction
In low Earth orbit (LEO), all spacecraft are subjected to various and very destructive
281
1./. Kleiman and R.C. Tennyson (eds.),
Protection ofSpace Materials from the Space Environment , 281-290.
282
components of the LEO environment, such as ionizing radiation (keV electrons,
MeV protons), vacuum ultraviolet (VUV) photons, hyperthermal atomic oxygen
(AO), and other factors such as extreme temperature variations. For these reasons,
polymer-based materials suffer relatively rapid erosion (mass loss) and surface
roughening, leading to irreversible degradation of the physical characteristics
(optical, thermal, electrical, and mechanical) for which these materials are designed
[1-3]. To overcome these problems, the polymer surfaces can be protected by a thin
inorganic coating [4,5]. Althoughprotective materialshave been the object of several
studies, and have been refined over the years, their ability to protect the underlying
polymers does not necessarily remain unaltered by the harsh LEO environment,
especially during long-term exposure: VUV photons and low energy protons (-100
keV-l MeV) can deposit substantial amounts of energy near the interface of the
coating with the polymer. In this report, we present preliminary investigations of the
possible effects of low energy proton irradiation (100 keV) and VUV radiation on
the mechanical properties of the protective coating-polymer interface. We present
here results for Kapton-H® and Mylar® coated with PECVD Si02, on which we
simulated LEO exposure conditionsup to 50 years in duration.
2.0 Experimental Methodology
2.1 SAMPLE PREPARATION
Two types of samples (thin foils) were investigated in this study, namely (A)
Kapton-H®, and (B) Mylar®, both 50flIIl thick. The thin protective Si02 coating was
deposited using a large-scale, dual-frequency microwave/radiofrequency (MW/ RF)
PECVD reactor [5-7].
The coating was about 0.3 11m thick, and was proven to satisfy the following
criteria: a) chemical resistance to attack by AO, and very low density of defects such
as "pinholes"; b) hardness and high resistance to abrasion; c) flexibility, that is,
satisfactory resistance to damage under flexural deformation; d) insignificant effect
on the thermal radiative (infrared absorptivity/ emissivity) characteristics of the
substrate material ; and e) high adherence [5]. Stoichiometric Si02 coatings were
obtained by PECVD reaction of an inexpensive, non-toxic precursor (hexame-
thyldisiloxane, HMDSO), with O2 as the oxidizer. In this work, we used a continuous
reactor described elsewhere [5,7], at constant MW power (PMW = 100 W), constant
Rf'-induced de bias voltage (Vb = -150V), and constant pressure (p = 80 mTorr).
2.2 PROTON IRRADIAnON PROCEDURE
For this study, the new 1.7 MY Tandetron accelerator of the Laboratoire de physique
nucleaire (Universitede Montreal)was used for all the irradiations. The proton beam
was raster-scanned over a region slightly larger than a tantalum collimator (2.0 x 1.5
em), which was located just in front of the targets; it served to define the irradiated
area, and thus the fluence.
The various coated film samples were mounted on a target holder located
283
directly in the centre of the irradiation chamber. With this configuration, it is not
possible to use a Faraday cup to measure the beam current, so that beam
measurements on the targets are uncertain due to secondary emission. We therefore
calibrated a BPM (Beam Profile Monitor) , located just upstream of the irradiation
chamber, versus the Faraday cup response, for various current values in the absence
of specimens. These calibration measurements were repeated prior to and after each
sample irradiation. Taking into account the variations in this calibration and other
sources of uncertainty. the total fluence is estimated to be known to within 7 per
cent.
o
Coa ling I - Dcp \h - )
Polyme r (Kuptonl Penetrarlon
depth ( 1.I Uln )
Figure I : TRIM simulation ofi on distribution for 100 keV protons

on Kapton® protect ed with 0.3 urn 0[Si0 2•
The samples were irradiated with 100 keY protons to total doses of
approximately 3.3 kGy (3.2 x 1010 p/cnr'), 6.5, 32.5, 130 and 325 kGy. These doses
correspond to exposures in LEO of roughly 0.5, I, 5, 20 and 50 years, respectively
[8]. Monte Carlo simulations using TRIM 98.01 code calculations [9] indicate that
the range of 100 keY protons in Kapton and Mylar is approximately l.l um; thus,
the protons are fully stopped just below the coating-substrate interface . The proton
energy of 100 keV was chosen for its abundance in LEO, to maximize the damage at
the interface and, finally, because it was the minimum energy available from the
Tandetron accelerator. Figure I represents a TRIM simulation of the ion distribution,
with recoils, projected on the Y-plane (the direction of the incoming protons), on 50
um thick Kapton covered with a 265 IUn SiOz protective layer.
284
2.3 VUV lRRADIATION PROCEDURE
The experimental system used, shown schematically in Figure 2, comprises two

major parts, namely the plasma source and the sample chamber. We used a
microwave (MW) resonant cavity with a 2.45 GHz power supply as the plasma
source, the plasma being excited in a quartz tube with an internal diameter of 13.5
mm and a length of 400 mm. A calibrated MgFz window (cut-off wavelength Ac
=112 nm) is placed between the sample chamber and the plasma reactor. The
samples were placed at the exact position of the centre of the sample chamber, co-
axially with the MgFz window and facing it. The radiation fluxes at the sample
position from a hydrogen plasma (100 seem, 1 Torr Hz) at various powers have been
precisely determined by careful spectroscopic calibration experiments using, in line-
of-sight arrangement with the plasma source, an absorption cell and a VUV-
spectrophotometer (0.2 m, Acton Research Corp., VM 502).
The radiation flux used in this experiment yields, for the selected time of
exposure, a fluenee of about 1.8 J/cm2, which is roughly equivalent to 2 months in
LEO. The plasma reactor and the VUV monochromator have been described in more
detail elsewhere [10,11]. After irradiation, the samples were stored under vacuum
before being studied further.
2.4 MICROSCRATCH TESTING
After irradiations, samples were cut into equal portions for scratch tests. Non-
irradiated and irradiated samples were scratched, using a CSEM MicroScratch tester.
The scratch is achieved by applying an increasing load through a Rockwell "C"
diamond stylus, while it is passed over the substrate surface at constant speed (see
Figure 3).
III.
Vacuum
Figure 2. Schematic of the VUV irradiation apparatus,

based on a H2 plasma "lamp"
285
The procedure was repeated four times for every value of received dose,
and the mean value was calculated.
Figure 3. Schematic of the Microscratch Test
The purpose of this test is to simulate damage, for example, LEO debris-
impact generated defects (e.g. cracks) that act as sites for undercutting of the
polymer by AO (12), but mainly to evaluate possible changes in the coating adhesion
and resistance as a function of the radiation dose. In other words, the test is intended
to evaluate the ability of the coating to protect the underlying polymer during lengthy
simulated sojourns in space. The stylus loads at which various types of defects
appear in the Si02 coating are observed and recorded, and their variation is
considered as an index of radiation-induced changes to the film/substrate interfacial
region [13). The samples were about 2 cm2 in area and were each subjected to four
scratch experiments. Two samples (taken from different locations on the material)
were studied for each value of radiation dose, in order to provide statistically
meaningful results.
The CSEM MicroScratch tester is computer controlled, and allows the user
to bring the sample under a microscope after a scratch experiment, for optical
characterization of defects. Since the substrate holder is indexed, every observed
defect on the sample can be linked to a particular value of the applied normal force
and the corresponding friction force exerted on the sample by the travelling stylus.
Table l. Parameters used to perform microscratch experiments.
Stylus Radius 50 urn

Ranae of aoolied forces 0-1200 mN
Loading rate 150 mN/min
Substrate speed 0.83 mm/min
Scratch lenath 3.3-6.3 mm
286
3.1 PROTON IRRADIATION
For any given sample, we can distinguish at least three types of defects, which are
created at three different, reproducible values of the applied force on a same scratch,
as shown on Figure 4, and we find that they vary with the radiation dose. These
defects can be associated with two failure modes of the coating/substrate pair,
namely cohesive and adhesive failure .
The first type of defect, observed at the lowest load, is a cohesive failure of
the Si02 coating. It is characterized by the appearance of thin cracks within the
coating, that follow the patterns of tensile stress release lines. At this point, no
damage of the interface has yet occurred.
OmN Load increase

I
u---~-[jl.-
t t t · , '-
First defect
HF -deleds
First det2chment
Critical Load - Lc
Figure 4. Schematic repres entation of the three types of defects
observed in microscratch testing.
The second type of defect, at significantly higher load., is adhesive failure at

the SiOipolymer interface. The cracks in the coating now occur under a compressive
stress, and they are accompanied by a regular, periodic delamination. Since adhesion
is the characteristic of interest in this study, the load to create this type of defect is
referred to as the critical load, L, This failure mode was observable both with the
microscope and on a plot of friction force versus applied normal force, produced by
the scratch tester software, where an abrupt change in slope occurs.
The third type of defect, at still higher load, is also an adhesive failure . It
consists of an abrupt increase in linear density of the type 2 compressive cracks , and
is also accompanied by a change in slope of the above-described plot. This kind of
phenomenon was also observed by Rats et al. [14].
Figures 5 (a) and (b) show plots of the applied load values for types 2 and 3
287
defects, versus equivalent dose (in years), for Si02IPET and Si02IPI samples,
respectively. In the former case, the values of loads are roughly one-half of the latter,
and the trends of the two curves are opposite: proton irradiation appears to improve
adhesion in the Si02IPI case. but a degradation is noted for the case of Si02IPET.
While the interpretation of these observations must be considered somewhat
speculative at this time, it would appear that cross-linking in the irradiated region
(i.e.e-Ium below the SiOipolymer interface) tends to dominate in the case of PI, and
a dominance of chain scission reactions may account for weaker adhesion in the case
of PET. This latter view is supported by spectroscopic ellipsometry, Figure 6, which
clearly shows the formation of a new low-index layer in the irradiated Si02IPET
specimen. Without this new interfacial layer in the ellipsometric model, no
satisfactory fit to experimental data could be achieved.
Since the Si02 coatings were thin, (- 0.3Ilm), type I defects appeared at
low values of load, about 120 mN and 50 mN for PI and PET, respectively.
Therefore, these do not appear on tile scales of Figures 5 (a) and (b). After a small
initial rise for PI. tile loads for inception of the type I defects decrease to about 80
mN and 30 mN. for PI and PET respectively. for the highest (50 years) equivalents
doses.
3.2 VUV IRRADIATION
Figure 7 shows tile measured threshold values of applied force for the appearance of
types I, 2 and 3 defects for virgin and VUV-irradiated PET and PI samples. As
mentioned earlier, the fluence used for tile latter corresponds to roughly two months
in LEO.
In both cases we note some degradation (decrease in load) after irradiation;
this is contrary to proton irradiation. where only PET showed this behaviour. Since
the thin SiO~ coating is transparent to the photons with A. > 180 om, and both PET
and PI absorb UV radiation quite strongly in this spectral region, these observations
suggest a dominance of chain scission reactions, accompanied by the formation of
low molecular weight (gaseous) species. This is compatible with the observations of
Hollander et al. (15) and Fozza et al. [II) . It is surprising that the relatively low
VUV dose should have such an important effect, and this question is presentIy being
studied in greater depth.
4.0 Conclusions
PECYD Si02 coatings on polymers offer excellent protection against attack of the
polymer surface by atomic oxygen (AO) [4,5]. Furthermore, these coatings adhere
strongly to the substrate by virtue of an "interphase" region of covalent bonding [16].
288
3OOr------------------,
PET
• •
- •
0. 150
<'
•
1I...L.....o-.....L.....o...-'--.L..............L...........L..--...J'-----...~
o 0.2 0.5 1 2 5 10 20 50 100
a Equivalent proton irradiation dose [years]
g)O , - - - - - - - - - - - - - - - - ,
..... 700 · • PI
~~f r · .-;
~:"II~
~~~:
O. 0.2 0.5 1 2
.5 10 20
•
50 100
b Equivalent proton irradiation dose [years]
Figure 5. Applied force values for the appearance of types 2 ("L, ") and 3 defect s
versus proton irradiation dose.
In this work we have used a recently-modified microscratch test, which

permits a ten-fold reduction in applied load compared with our earlier work [13], to
probe the interfacial region of proton- or VUVIUV-irradiated SiOzlPET and SiOzlPI
samples. This refined technique allows us to distinguish three types of failure modes
in the coating/substrate pairs, at loads ranging from 50 to about 500 mN, for a 50
IJ.m diamond stylus. After proton irradiation (at doses equivalent to up to 50 years in
space), SiOzlPI shows apparent enhancement of adhesion, while that in SiOzlPET
degrades. Both sample materials show apparent degradation after VUVIUV
irradiation equivalent to a few months in LEO, but these data are preliminary and are
being verified.
289
If I IRRADITED (50 YItS )
40\)r---"'---""'--~-_--, 400r---------~--__, 25
Joo 20
200 15
<J -E
100 . 10
o
.1 00 --"- _.__t . __. .. .... .. ... __ --' ...........__ ~ 0
200 400 600 800 1000
Wavelength (nm)
1 .80 1.80
1.70 1.70
1.60 PET 5i02 film 1.60 PET Si02film
n 1.50 n 1.'0
1.40 1.40 ·
1.30 1.30
1.2~50 0 50 100 ISO 1.2?,O 0 SO 100 ISO 200

Distance from Subsl..ate (nm) Distance from Substrate (NIl)
Figure 6. Spectroscopic ellipsometry results

for control - and Proton- irradiated samples
J 7~r-------------------.
Silica coatedPET
JUU
IA: PET
..."""e
UnexJl. UV Unexp. try UnexJl. UV
Silica coated Kapton
..,
o"
o-l
Unex p, uv Unex p, uv Unex p. uv
Figure 7. Applied force values for the appearance of types 1,2 ("Lc"), and 3 defects
in control samples (Unexp .) and after a total dose equivalent to 2 months in LEO .
290
We are also preparing to expose coated polymers simultaneously to the
combined (synergistic) effects of protons and VUV irradiation as encountered in
space, along with thermal cycling. TIle possibility of in-situ ellipsometry and
mechanical testing are also being considered.
5.0 Acknowledgments
This work is being funded by the Natural Sciences and Engineering

Research Council of Canada (NSERC). within the Industrial Research Chair (IRC)
on Low-Pressure Plasma Processing of Materials. It is a continuation of contractual
work for the Canadian Space Agency (CSA) by Polyplasma Inc. under STEAR RFP
6 (protection of materials in space environment). The authors are grateful to A.
Bergeron for ellipsometric measurements.
6.0 References
1. Leger, L.1., NASA TM-58246, May, 1982.

2. R.C. Tennyson, Can. J . Phys. 69.1190., 1990 .
3. Tennyson, R.C., The Behavior of Systems in the Space Environment, eds. R.N. deWitt, D. Duston and
A K. Hyder Kluwer, Dordrecht, p.233. 1993.
4. Zimick, D.G., Wertheimer, M.R., Balmain, K.G.. and Tennyson, R.C., AL4A J . Spacecraft Rockets 28,
652. 1991.
5. Wertheimer, M.R.• Czerernuszkin, G.• Cerny, 1.. Klemberg-Sapieha, J.E., Martinu, L., and Kremers,
W.. Proc. 111 Int. Symp. On "Materials in space environment, " ESA SP.399, p. 393,1997.
6. Martinu, L., Klemberg-Sapieha, 1.E., Kuttel, a .M., Raveh, A . and Wertheimer, M.R., J. Vac. Sci.
Technol., A 12, 1360, 1994,
7. Wertheimer. M.R., Thomas. H.R.. Perri. MJ.. Klemberg-Sapieha, J.E., and Martinu, L.. Pure and
App!. Chem . 68,1047,1996.
8. Cross, E.L.,MASc. Thesis. U. of Toronto. ISBN: 0612192482,1996.
9. Ziegler. 1.F., Biersack. 1.P.. and Littmark. U.. TRIM. version 98.01. in: Stopping and Range ofIons in
Solids, I,Pergamon Press, N.Y.. 1985.
10. Fozza, AC . Kruse, A., Hollander. A , Ricard, A. and Wertheimer, M.R., J. Vac. Sci. Techno!. A 16
72,1998 .
11. Fozza, AC ., Roch, 1., Klemberg-Sapieha. J.E., Kruse. A ,Hollander, A , and Wertheimer, M.R.,
Nuc!. Intrum Meth. Phys. Res. B. 131,20,1997.
12. Banks, B.A, and Rutledge, S.K.. Proc . 4'11 Int. Symp. On Spacecraft Materials in Space Environment,
Toulouse, France, p. 371, 1988.
13. Houdayer, A , Cerny, J.. Klemberg-Sapieha, J.E.. Czeremuszkin, G.. and Wertheimer. M.R., Nucl .
Intrum . Meth. In Phys. Res . B 131,335. 1997.
14. Rats, D., Hajek, V.. Martinu, L., Thin Solid Films 340, 33,1999.
15. Hollander, A, Klemberg-Sapieha, 1.E., and Wertheimer, M.R., J . Polym. Sci-A: Polym. Chem . 34 ,
1511,1996.
16. Da Silva-Sobrinho, AS., Schiihler, N., Klemberg-Sapieha, J.E., Wertheimer, M.R., Andrews, M.,
and Gujrathi, S.C., J . Vac. Sci. Techno!' A 16,2021 ,1998.
INFLUENCE OF ELECTRON ANDPROTON FLUXES ON OPTICAL
PROPERTffiS OF PROTECTIVE ENAMEL
S. N. SKOVOROD'KO, A. V. KURll...OVICH, V. YA. MENDELEYEV

AND V. N. VASIL'YEV
Institute for High Temperatures,
The Russian Academy ofSciences.
Izhorskaya 13/19. Moscow 127412,
Russian Federation
1.0 Introduction
The efficiency of the protection of materials from low Earth orbit (LEO) space
environment is determined by the stability of their physical properties during
operation. An important component is associated with optical properties. In this
paper, optical polarization characteristics of protective enamel irradiated by electron
and proton fluxes are investigated experimentally.
The protective enamel is based on a composite ceramic. Modelling the fluxes of
charged particles irradiating the enamel was performed using electron and proton
accelerators. The optical properties of the samples were determined from the
following polarization characteristics of scattered laser light including : degree of
polarization, azimuth and ellipticity.
The following influences that imitate the Low Earth Orbit were available for the
experiments :
- weak influence : using a 20 kV electron beam;
- middle influence: using a 150 kV electron beam;
- strong influence : using a 150 kV proton beam.
The polarization characteristics were measured in the specular direction into the
angular range from 15 degrees to 85 degrees. All measurements were carried out by
a polarometer - ellipsometer with following characteristics :
Incident radiation parameters :
- wavelength is 632.8 nm ( He-Ne laser)
- polarization rate is 100 per cent
- azimuth is 45 degrees from incidence plane
- ellipse is 0.0.
Measurement accuracy:
- for polarization rate - 1 per cent
- for azimuth - 0.5 degree
- for ellipse - 0.01
291
292
2.0 Results
The resuls of irradiation experiments for weak influence, middle influence and
strong influence are presented in numeric form in Table 1; measurement results
for conditions without any irradiation source are also given in Table 1 for
comparison.
Table 1 Numeric results

Incidence Ellipsometry parameters
angle Polarization rate Azimuth Elliptic ity
without influence
15 6,87 130,7 -0,16
25 6,38 132,9 -0,19
35 4,15 122,8 -0,23
45 4,9 95,7 -0,47
55 8,28 66,3 -0,21
65 23,89 56,9 -0,05
75 68,47 50,9 -0,01
85 96,67 46 Om
weak influence
15 5.87 133.3 -0.14
25 3.87 135.0 -0.16
35 3.97 127.1 -0.35
45 4.13 95.8 -0.49
55 5.81 67.3 -0.22
65 17.30 54.4 -0.12
75 51.12 49.2 ·0 .08
85 97.0 8 45.7 -0.05
middle influence
15 6.31 135.0 ·0. 17
25 5.02 132.6 ·0 .17
35 4.26 131.0 -0.36
45 2.96 99.3 -0.51
55 6.26 61.7 -0.32
65 18.75 52.2 -0.10
75 50.86 51.8 -0.06
85 91.71 46.5 -0.05
strong influence
15 7.70 133.5 ·0.09
25 5.37 129.9 -0.19
35 4.36 124.9 -0.20
45 3.31 82.9 -0.36
55 10.43 60.8 -0.17
65 40.73 52.5 -0.08
75 84.24 48.2 -0.05
85 97.94 45.8 -0.05
293
The three major characteristics, namely, the polarization, the azimuth and the
ellipticity vere also plotted versus the angle of specular reflection for all four cases,
as shown in Table 1. Figures 1 through 3 present these results.
00
,-
i
80 - - no influence i
/
- - - - -- weak influence ,
i .:
middle influence r
60
- - -- - strong influence , .- "
,-i i
Q.
, i ,.,-
40 ,- i
,- ;'
,- _.,
t
i
,-i ,;7
20 ,- -'
,-
.... -.: ~., ."
_ . . . . ... ~~ .. . _~i
O l...-....L._ -'----'-_ ........---'_ ........._ I . . . -........ _.l----'-_-'-----'_ ---'------'

20 30 40 50 60 70 80
Angle, degrees
Figure 1
Dependen ce of the polarization rate on influence volume
- - no influence
- - - -- - weak influence
.... .. _ . 9 ... .__ - _ . : . : _ , ~ _" _ ~~.
-,-. ---- ----- middle influen.::e

, ~. -
.",". --- --- stronginfluence
"
-,
-,
" -
-.
30
o l-.----L_~----L_-'--_.L _ _'___.L_-'--_____l._~_____l._ ........_ __L----'

20 40 50 80 70 80
Angle, degrees
Figure 2.
Dependenc e ofazimuth on influence volum e
294
0,8 - - no influence
· · · · · · ~k inluence
......... middleinfluence
0,4 _._._.•. strong influehce
>-
'i l 01)
:§. ....:-
'-~~ ..
'ii ~ .. . .....
~.
-0,4
-0,8
50 70
Angle, degrees
Figure 3
Dependenc e of ellipse on influence volum e
3.0 Summary
It can be seen that the 'strong influence' sample surface has optical properties
different from the others, due to the presence of proton beams as heavy particles.
Several ranges of the light incidence angles at which the influence of the
beams on the optical properties was detectable were found. The influences of the
electron beams on the polarization characteristics were identical. However, the
electron and proton beams changed the polarization characteristics differently. The
proton beam caused an increasing degree of polarization, ellipticity and reducing
azimuth of the scattered light. Maximum deviations of the degree of polarization,
azimuth and ellipticity parameters were found to be 10 per cent, 10 degrees and 0.15,
respectively. Under the action of the electron beams, the degree of polarization and
ellipticity was found to decrease. Maximum variations in the degree of polarization
and ellipticity were 5 per cent and 0.1, respectively. The influence on the azimuth
was weak.
METEOROID HAZARD IN THE LEO SPACE ENVIRONMENT
Ya. S. YATSKIV
Main Astronomical Observatory
National Academy ofSciences of Ukraine
252650 Kyiv-22 , Ukraine
V.G.KRUC~~O
Kiev University Observatory
254053 Kiev, Ukraine
Yu. 1. VOLOSHCHUK
Kharkiv Technical Universityfor Radioelectronics
3 J0726 Kharkiv , Ukraine
Abstract
Two different approaches to the problem of meteoroid hazard in the LEO space
environment are considered. The traditional approach is based on use of the
Integrated Influx Density (lID) of space particles and bodies derived for the wide
range of masses (IO -12 s M s 1022 g) for calculation of the probability of collision
between spacecraft and these bodies. Another approach relies upon an indirect
method for detection of numerous parent bodies by means of observations of meteor
showers. Using the Kharkiv Meteor Data Base (KMDB), the probabilities of collision
between parent bodies and the Earth were calculated.
1.0 Traditional Approach to the problem of Meteoroid Hazard in the LEO

Space Environment
The traditional approach to the problem of meteoroid hazard in the LEO space
environment relies upon the use of the Integrated Influx Density (lID) of space
particles or bodies of different masses, for derivation of the probability of collision
between spacecraft and these objects. lID is the number of objects per square em and
per sec (N) of which masses are not less than M (N and M are expressed in log-scale):
Log N = A + B Log M (I)
To determine the coefficients A and B of (I), different types of measurements

have been used (see, Lebedinets, 1980).
295
Protection of SpaceMaterialsfrom the Space Environment, 295-303.
296
Figure 1 illustrates the different types of the lID (Table 1) and the general
equation of (1) derived by LSM:
Log N = - 18.23 - 0.938 Log M, (r = 0.9813) (2)
Table 1. UO derived for different intervals of masses of space object s
Interval of masses Designation of Coefficient in (1)

in g. intervals in Fig.1 A B References
10 ·12_ 10 .• 01 - 15.1 - 0.6 Mazets , 1973
10-11 _ 10-7 O2 ·14.1 · 0.55 Lebedinets, 1980
10 .6 _ 10 2 M - 18.8 - 1.25 Lebedinets, 1980
10 2.10 7 B -19.1 - 0.62 McCrosky,1968
1.9.10 6 -1.9.10 11 MO(a») - 20.13 -0.73 ReVelle, 1997
1.9.10 6 -1.9.10 10 MO(a)2 - 19.42 - 0.83 Nemtchinov et al., 1997
2.10 8 _10 15 L - 25.14 -0.49 Whipple, 1967.
1.3.10 13 _1.3.10 22 AlP) -22.56 - 0.65 Whipple ,I967; K1acka,1992;

....................................................................................................................................... ~~~~~~~!~.~..:!..~.I::..~??~: .
(a) : To determine coefficients A and B within this interval of masses, we have used fireball
observations carried out by the Sandia Lab in 1994 - 1996 (Reynolds, 1992; Tagliafeni, 1996) and
measurements of atmospheric acoustic waves by the U.S. Air Force in 1960-1980 (ReVelle, 1997) . Taking
the value of atmospheric entry velocity V = 21 krn/sec, derived from the analysi s of 'Prairie network
fireball data' (McCrosky et al., 1976; 1977), we have made the tran sition from energy of impact to mass of
object .
(b) : To determine the coefficients A and B within this interval of masses , we have used results by
Whipple (1967) based on analysis ofthe Apollo astero ids with masse s i > 2.8 ,10 15 g and results of studying
the mass distribution ofasteroids (Rabinovitz et. al., 1994; Klacka, 1995).
Defining the lID in terms of number of space objects N" per year over the entire
Earth, of which masses are not less than M, the equation (2) becomes
Log N" = 7.99 - 0.938 Log M (3)
Equations (2) and (3) could be used for different estimations of meteoroid impacts
with the Earth and spacecraft. For the masses of particles M ~ 10 -6 g and an effective
surface of spacecraft of 100 square meters, the probabilities of collision between
meteoroidsand the LEO spacecraft during a year were computed (Table 2).
297
Table 2. Probability of a collision between a spa cecraft and a meteor part icle of given mass, and the
thickn ess of the shield required for protection of spacecraft.
Impacl energy for Min. thickness

Mass of particl e Probability V = 30 km -sec · 1 of shi eld
in g. P x 103 in 10'Oergs. in em.
1.9.10 ·2 450 1.50
1.2 4.50 0.30
9.7 0.45 1.15
600 0.00(4) 0.03
Log N (cm'f-sec")
·10
-25
-40 -t-- - , - - - r -- - - , , . -----.- - - - , - - - - - , -- - - , - -- --,
-1 5 -10 -5 o 5 10 15 20 25
Log M (g)
Figure I. Th e lID of spac e bodies of whi ch mass es ar e not less than M.

N is the number of spac e bodies per squar e em, and per sec.
On the basis of the equation (2) and the data obtained by McCrosky et al. (1976,
1977) one could reach the following conclusions :
a) Each day, about 400 meteoroids with masses not less than 1 kg enter the
Earth 's atmosphere .
b) Some of these meteoroids (about 30 per cent) could approach the Earth as
close as 30 kIn or even less. The motion of such meteoroids in the atmosphere is
accompanied by shock waves and thermal explosions which could be real hazards for
nearby aircraft.
c) During one year, about 870 meteorites fall to the Earth.
298
2.0 Kharkiv Meteor Database (KMDB)
For decades, B. Kashcheev's team at the Kharkiv Technical University for Radio-
electronics has carried out radar observations of meteors. Thus, the Kharkiv meteor
Database (KMDB) was developed which is one of the most complete at the present
time . It consists of:
• Total number of meteor orbits 233543
• Number of selected meteor orbits according to
Voloshchuk et aI. (1995) 159787
• Number of meteor showers selected according to the
condition that each shower is formed by not less than 5 orbits 5160
• Number of meteor showers selected according
to aphelion distance Q < 4.8 a.u. 4810
1.0
45 90 135 180
1.0
0.0
0.0 0.5 1.0
1.0
N/Ntl
0.0 q
0.0 1.0 0.0 0.5 1.0
1.0
N/N..
0.0 ro
o 90 180 270 360 o 90 180 270 360
Figure2.
The normalized distributions of the meteorshowers depending on the orbital parameters i, e, q, and co
(see text for definitions).
Figure 2 illustrates the normalized distributions of meteor showers

depending on the orbital parameters (i, e, q, and UJ, see next two pages for
definitions). Columns marked as I and II in figure 2 demote an apparent distribution
and distribution corrected for selection effect, respectively. Some differences
between cases I and II are noticeable. The distribution II differs significantly from
the distribution of AAA-asteroids (Apollo, Amor and Aton asteroids), which have no
orbits with 1>90 °.
299
Figure 3 illustrates the two-dimensional distributions of meteor showers

depending on (q, i) and (e, i); (q is the perihelion distance, i is orbit inclination , e is
the eccentricity of orbit). The area marked by solid lines is covered by AAA-
asteroids . The differences between distributions of meteor showers and AAA-
asteroids is clearly visible in the function of orbital parameters proposed by Lidov,
1961 (see Figure 4).
OB
0 6
02
O '------'....:....:'---'""-'--.:.....;---'--~~-=-----"-------'~-'---.:.:..:----" i
o 30 6U 12B ' 50 18 0
e
O .B
0 .6
0 .'
0 .2
Figure 3
Two-dimensional distributions of meteor showers depending on orbital
parameters. The area marked by solid lines is covered by AAA-asteroids
300
J1 AAA
14
SHW
0.7
.r
'.
'.
- 0.7
- 14 L.-_~ __ ~ __ ~ ---' V
-D.3 -0.2 - 0.1 o 0.1 0.2 0.3 04
Figure 4
Two-dimensional distributions of meteor showers,
depending on parameters It and v
We find that:
v = e2 (0.4 - sirr'i-sirrro), J.! = [a (1 - e2) ] 1/2 cosi (4)
where (0 is the argument of the perihelion.

KlvIDB was used for studying the variation of the lID of meteor particles in time
and in orbital parameters of spacecraft (Figures 5, 6 and 7).
We have found that the lID has no diurnal variation, and has very small (up to 10
per cent) inter-annual variation. At the same time, there is significant intra-annual
variation.
3.0 New Approach to the Problem of Meteoroid Hazard
A new approach to the problem of meteoroid hazard relies upon the idea of using the
KMDB for search of numerous parent bodies of meteor showers . In principle, this idea
is not new, and is employed here with some reservations concerning the number of
meteor orbits which could form a meteor shower. We have presumed in this paper that
a minimum of 5 orbits forms one shower. Based on this new approach we can detect
such NEO 's which are not observed by traditional techniques. Then we can compute
the probability of collision between parent bodies and the Earth. The non-zero
probability of such collision is defined by
(R/2)2 - D ~ 0, (5)
301
where
R = 0.6/a - 112 - 2, D = 0.4 + (2 - lIa)'1l
2
- v,
a is semi-major axis of an orbitof a meteor body.
N·2·1012 cn; 2s·!

1.3 , . .
1.1
09 i\ / -,
-,
/
\- / '---V -, !
/
07
'y/'
0.5
. i\ /
0.3
0 .1
1
i-+- 6 9
\/
10 11
Montll
12
Figure 5
Intra-annual variation oflID of particles of which masses
are not less than 5.10 .6 g (according to Voloshchuk et al., 1997).
N.2.10 9 cm 2s- t
110
90 ..' . ..,
1
70 .
....
50
i-
-. 1
6,
2
-, ..- J ~ -. v
/
30
10 ~ ""
............~ ~'
,.,'
/'
-: I
'/
'+.
~
o
•· ... i·+··,t· ---s:::,
30 60 90
j
120 150 180 210 240 270 300 330 360

/ -.+ --
.
Figure 6
The lID of particles depending on the true anomaly of a satellite of the MIR-type
(H.... = 365 krn, Hmin = 335 km, e = 0.002, i = 52°). I • frontal pad of platform
2 • back pad of platform ;3 - spherical surface.
302
': :~T\;r~:~T:~~Ft~r::;: rt:~

!
60
.~ j
\1
1~ .......~
i
_ _+ ········t---- ·····f ..... .....!
40 ! !
; !
l
20 .) ... ..... i .
.......= '=d
06..:......_.J..:::::o.*:::l=:::I..._.L..:.:~=~---'i...-.;=-=..£:::
o 30 60 90 120 150 180 210 240 270 300 330 360
Figure 7
The lID of particles depending on the true anomaly of a satellite of the
MOLNIY A-type (H m", = 40000 km, HnUn = 550 km, e = 0.74, i = 15°) for the
sake of argument of perihelion ro = 90° and for longitude of ascending node Q = 0°.
I - frontal pad of platform ; 2 - back pad of platform; 3 - spherica l surface
Table 3. Probability of collision between Parent Bodies (PB) of meteor showers and th e Earth .
NumberofPB
Probability intervals colliding th e Earth Percenta ge of total
6
x 10 with given number of PB
probability
0 -0.1 3762 84.8
0.1 - 0.2 415 9.36
0.2 - 0.3 113 2.55
0.3 - 0.4 45 1.01
0.4 - 0.5 29 0.65
0.5 - 0.6 19 0.43
0.6 - 0.7 14 0.32
0.7 - 0.8 7 0.16
0.8 - 0.9 3 0.07
0.9 - 1.0 2 0.04
1.0 -1.1 6 0.14
1.1 - 1.2 2 0.04
0-3.3 4435 100.
4.0 Conclusion
1. The meteor data base complied at the Kharkiv Technical University for
Radioelectronics allows us to propose a new approach to the problem of meteoroid
hazard in the LEO environment. It relies upon the search for numerous parent bodies
303
of meteor showers, and calculation of probabilities of collision between these parent

bodies and spacecraft.
2. In spite of some deficiencies of an indirect search method for parent bodies of
meteor showers, the preliminary results of our study could provide additional
information for detection of NEG.
5.0 References
1. Klacka, J., Mass distribution and structure of the asteroid belt, Acta Astronomica et Geophysica Univ.
Comenanie XVII, 62-76,1995,
2. Lebedinets, V.N. Dust in the upper atmosphere and cosmic space. Meteors, Hidrometeoizdat,
Leningrad, (in Russian), 1980.
3. Lidov, M.L, Evolution of orbits of artificial satellites of planets under an operation of gravitational
perturbations of bodies, Artificial satellites ofthe Earth 8, 5-45, (in Russian) , 1961.
4. Mascts, E.P. Micrometcoroids in space, in N.B.Divary (ed.} Dust in atmosphere and in near-earth
space , Nauka, Moskow" pp. 13-23 . (in Russian), 1973.
5. McCrosky, R.E. Distribution of large meteoritic bodies , Smithson . Astrophys. Obs. Spec . Rep ., (1968)1
280.
6. McCrosky, R.E., Shao, C.- Y., and Posen, A , Prairie network fireball data . 1. Summary and orbits,
Center for Astrophysics. Prep. Ser. I 665 , 1976.
7. McCrosky, R.E., Shao, C.- Y., and Posen, A Prairie network fireball data . II. Trajectories and light
curves, Center for Astrophysics. Prep. Ser. I 721, 1977.
8. Nemtchinov, LV., Svetsov, V.V., Kosarev, I.B., Golub', A.P., Popova, O.P., Shuvalov, V.V., Spalding,
R.E., Jacobs, C, and Tagliaferri, E. Assessment of Kinetic Energy of Meteoroids Detected by Satellite-
Based Light Sensors, learus 130,259-274, 1997,
9. Rabinovitz, D.L., Bowell, E., Shoemaker, E., and Muinonen, K., The population of Earth-crossing
asteroids, in T.Gehrals (ed .), Hasard due to comets and asteroids. Univ, Arizona Press , Tucson, pp.
285-312, 1994.
10. ReVelle, D.O., Historical Detection of Atmospheric Impacts by Large Bolides Using Acoustic-Gravity
Waves, Repr. from Near-Earth Objects. V. 822 ofAnnals ofthe New York Academy ofSciences . May
30, 284-302, 1997.
11. Reynolds, D.A, Fireball observation via satellite. in G.H . Canaven, lC. Solem and J.D.G. Rather
(eds.), Draft Proceed. Of the Near-Earth Objects Interception Workshop, Los Alamos National Lab.,
pp. 221-226, 1992.
12. Tagliaferri, E. Satellite observations of large meteoroid impact, Meteoroid Impact Workshop, Sandia
National Laboratories,A1buquerque, NM, 1996.
13. Voloshchuk, Yu.l., Kashcheev, B.L., and Podolyaka, B.A A Meteor Complex near Earth's Orbit :
Sporadic Background, Showers Associations. I. The Technique for Selection of Showers and
Associations from a Large Sample oflndividual Meteors , Astron . Vestn. 29, 439-449, 1995,
14. Voloshchuk, Yu.l. , and Kashcheev, B.L., The Meteor Complex near the Earth's Orbit: Sporadic
Background, Streams, and Associations. 11. A Comparative Analysis of Some Statistical Characteristics
of Stream and Sporadic Meteoric Bodies, Astron. Vestn.30, 541-560,1996.
15. Voloshchuk, Yu.l., Vorgul, AB., and Kashcheev, B.L. The Meteor Complex near the Earth's Orbit:
Sporadic Background, Streams, and Associations. Ill. Sources of Stream and Sporadic Meteoric Bodies,
Astron . Vestn. 31, 345-369,1997.
16. Whipple, F..L., On maintaining the meteoritic complex, Smithson. Astrophys. Obs. Spec. Rep . I 239,3-
45,1967.
CHARACTERIZING THE ANNUAL METEOROID STREAMS
MARTIN BEECH
Camp ion College, and Department ofPhysics
The University ofRegina
Regina. Saskatchewan. Canada S4S OA2
Abstract
We review the properties of the annual meteoroid streams. Special emphasis is

placed upon the understanding of the spatial and temporal properties of the principal
meteor showers. Impact probabilities of stream meteoroids on space platforms are
also discussed.
1.0 Introduction
The Solar System is literally awash with particulate debris derived from
cometary out-gassing and the triturization of asteroids. Riding the swell of this
particulate ocean. spacecraft in Earth orbit must inevitably suffer an occasional
impact and consequently, the distinguishing features of the sea are worthy of our
attention. Indeed , the meteoric complex, as we should more correctly call the sea of
ejected fragments, is a permanent and dynamic feature of the inner Solar System and
is therefore a fixed and non-negligible component of any spacecraft's environment.
There is much popular confusion surrounding meteoric nomenclature. For
example, the words ' meteor," meteoroid ' and 'meteorite' are often and incorrectly
used interchangeably. The point is, one can 'see' a meteor, but not necessarily a
meteoroid. More precisely, the occurrence of a meteor can be discerned with the eye,
recorded photographically, or be detected by an appropriate radar system . In
contrast, a meteoroid is understood to be any interplanetary body that can potentially
produce a meteor if it chances to enter the Earth 's atmosphere. A meteor is
associated, therefore, with the light and ionization phenomenon produced by the
ablation of a meteoroid within the Earth's atmosphere. Also, on the basis of these
distinctions, the Solar System contains ' meteoroid streams ,' but not ' meteor
streams.' Further, a ' meteor shower' occurs whenever the Earth passes through a
meteoroid stream . Meteorites are identified as those fragments of a meteoroid that
have survived passage through the Earth 's atmosphere and have landed on its surface
(Beech and Youngblood, 1994).
The size limit beyond which an object is identified as an asteroid rather than a
meteoroid is determined partly by semantics and partly by instrumental techniques
305
Protection ofSpace Materialsfrom the Space Environment, 305-329.
306
(Beech and Steel. 1995). Objects smaller than about 10 meters cannot be detected
with modem-day search techniques in cis-lunar space. Consequently, meteoroids are
taken to be those solid objects that are smaller than 10m in size. A lower limit can
also be determined by ablation conditions. For an object to produce an ionization
trail when it encounters the Earth's atmosphere. it must be larger than about 100 urn
(e.g., have a mass greater than about lO-6g). Particles smaller than 100~m are simply
termed dust particles.
The number of meteoroids within a given mass range increases with decreasing
mass. Formally, the cumulative flux of meteoroids with masses greater than or equal
to rno is F(m ~ rno) a rno(1 . s), where s is the mass index defined so that the number of
meteoroids dN in the mass range m to m + dm is dN = m,sdm. Typically 1.3<s <2.3
for meteoroid streams and s e 2.2 for the sporadic background.
The meteoroid environment, at the Earth's orbit, is composed of two distinct
populations; the meteoroid streams and the sporadic background. Meteoroids and
dust particles from the sporadic background intercept the Earth's orbit throughout the
year and are derived from several broad radiants distributed across the celestial
sphere. Meteoroid streams, which we note also contain dust particles, are comprised
of a large number of particles that have similar heliocentric orbits. I will not dwell
upon the physical properties and characteristics of the sporadic background in this
article. Indeed, aspects of the average meteoroid flux, radiant distribution and
temporal behaviour of the sporadic background have been discussed in numerous
recent reviews (see e.g., Griin et. 01., 1985, and Brown and Jones, 1995).
Historically, the space-user community has largely ignored the meteoroid stream
component of the meteoric complex. There are sound reasons, discussed below, for
this general practice, but recently it has been realized that there are occasions when
extremely high meteoroid fluxes can be realized from normally 'quiet' meteoroid
streams. The conditions leading to the appearance of outbursts is part-and-parcel a
consequence of the manner in which meteoroid streams are formed and ' maintained.'
We shall briefly review the formation process in the next section.
2.0 The Formation of Meteoroid Streams
Meteoroid streams typically form in the wake of aging comets. Giovanni Schiaparelli
was the first person to demonstrate this observationally in 1867, when he was able to
show that the orbital elements of the Perseid meteoroid stream were almost identical
to those of comet 109P/Swift-Tuttle.
Our modem-day understanding of the cometary decay process is based upon the
pioneering work of Whipple (1950). In Whipple's model, cometary nuclei are
' pictured' as kilometer-sized ice-structures with embedded dust grains. As the
nucleus approaches perihelion the Sun's radiation flux warms the surface ice
resulting in sublimation and the onset of an outflow of gas. The resulting molecular
outflow carries with it the solid particles that were embedded in the icy matrix. It is
the ejected particles that become the stream meteoroids. This is true, provided the
meteoroids are (a) not so large that the gas drag cannot overcome the gravitational
attraction of the nucleus. and (b) they are not so small that they are ejected from the
307
Solar System on hyperbolic orbits through the Sun's radiation pressure. That this
model must be essentially correct was clearly demonstrated by the GIOTO fly-by of
the nucleus of IPlHalley in 1986 (Keller, Kramm and Thomas, 1988).
The basic planar-geometry of the meteoroid ejection process is illustrated in the
upper panel of Figure 2 (to be discussed later) . We imagine that the comet has a
velocity Vc, which will vary according to heliocentric distance and the semi-major
axis of its orbit , and that tile meteoroid has an ejection velocity V00 ' If the meteoroid
is ejected at an angle a. to tile Sun-comet line, then the heliocentric velocity VM of
the meteoroid will be
. (1)
The meteoroid ejection velocity. as given by Whipple 's formula, is typically

between 10 to 100 mis, depending upon the size of the cometary nucleus, the
meteoroid density and the heliocentric distance at the time of ejection. Such
velocities are very small compared to that of the parent comet, which might typically
be of the order of 30 kmls . In other words , the meteoroids only slowly move away
from the comet. Conservation of energy and angular momentum arguments dictate
that the orbit of a newly ejected meteoroid will have a semi-major axis , a' , and an
eccentricity, e', given by
at = 886 a (1- P) / { 886 (I - 2 a P/ r) - (VM2 - Vc2) } (2)
e t 2 = I - (a / at) (1- e2)/(1 - P) t (3)
where a and e are the semi-major axis and eccentricity of the comet's orbit, r is the
heliocentric distance at the time of ejection and P is the ratio of the solar radiation
pressure to solar gravity . Gustafson (1994) shows that ,
P= F,ad. / Fgrav= C, Q (Nm), (4)
where Q is an efficiency factor for radiation pressure weighted to the Sun 's
spectrum.. C, is a constant equal to 7.6 xlO-5 g/cnr', and (A'm) is the area to mass
ratio of the meteoroid. Q is not entirely particle size independent, but for meteoroid
masses greater than about 1O -12g, it is essentially a constant and consequently P
varies as lI8, where 8 is the radius of the meteoroid. In principal P is of less
importance for the larger meteoro ids. but tins need not be true if meteoroids have a
"fluffy" or "birds nest" -like structure.
Equations (2) and (3) dictate that tile orbital characteristics of stream meteoroids
are not , in general, the same as those of the parent comet. Some meteoroids, for
example, will have smaller semi-major axes , others larger ones than that of the
parent comet. The meteoroids and comet will have essentially the same perihelion
distances, and arguments of perihelion, however. The meteoroid stream has a ' shape'
akin, therefore, to that of a pinched elliptical torus, with the 'pinch ' being at the
perihelion point. Figure I illustrates this characteristic, and shows the orbits of
several Leonid meteoroids. One has to be a little careful , however, in interpreting
308
Figure I, in the sense that the orbits shown are based upon observations that have
inherent errors, and that some of the orbital variation is no doubt a result of planetary
perturbations.
Equation (2) can be combined with Kepler's 3rd law to calculate the lead and lag
time with which meteoroids will follow the parent comet to perihelion. Figure 2
shows the variation in semi-major axis and return time for O.Olg meteoroids ejected
from comet 55prrempel-TuttIe at tIle time of perihelion passage . The spread in
return time amounts to about 3.4 years . At Ig the variation in return times is smaller,
amounting to 1.1 years . At IO·sg the spread has grown to 7.5 years.
As the return times of the meteoroids and parent comet become more and more
' mixed, ' so the Earth will begin to sample meteoroids each time that it passes near to
the stream node . Newly-forming meteoroid streams, in contrast, will charac-
teristically only produce meteors at those times when the parent comet is near
perihelion. The n-Puppid meteor shower , for example, represents a stream in genesis
and appreciable numbers are only seen every five years when the parent comet,
26P/Grigg-Skjellerup, rounds the Sun. The August Perseid meteor shower, on the
other hand, is a 'well mixed ' stream in that near equal numbers of meteors are seen
each year irrespective of the position of 109P/Swift-Tuttle in its orbit. This being
said , however. we note tI13t significantly enhanced Perseid meteor rates were
observed for a number of years before and after the perihelion passage of Comet
Swift-Tuttle in December 1992. The enhanced ' activity-spike' in observed Pcrseid
meteors was the result of jovian perturbations acting upon those meteoroids ejected
into the stream at perihelion passages in 1862 and 1737 (Wu and Williams, 1993).
If the orbital period of a meteoroid is P' and that of the parent comet is P, then
the loop formation time will be of the order uP, where n is the number of revolutions
required to yield the equality n(P' - P) = P/2. That is, after n revolutions the ejected
meteoroid will have accumulated a half-period difference to the parent comet. Since
the period can be related to the orbital semi-major axis through Kepler's 3rd law, we
have
n = I I 2(P'fP - I) and P'fP = (a'/a)3/2 (5)
and (a'/a) is given by equation (2). If we assume a typical meteoroid ejection

velocity of 10 m/s, then the maximum value of (V M2 - VC2) will be of the order
0.02Vc. With this approximation we have calculated the minimum loop formation
times shown in Table I.
The simple planar-model discussed above can be used to predict the general,
short-term behaviour of stream meteoroids, but a detailed picture can only be derived
through a full three-dimensional numerical simulation including the effects of
radiation pressure and planetary perturbations. As a case in point, Figure 3a shows
the nodal crossing time and heliocentric distance for Leonid meteoroids encountered
one orbit after ejection, with ejection taking place at the perihelion point of comet
55prrempel-Tuttle (I have used the 1965 perihelion conditions in the calculation).
In this case , the only variable is the ejection angle «, measured in the plane of the
comet's orbit. The meteoroids clearly fall along a straight line, and a seemingly
309
Table I. Loop formation times for several meteor showers. The Draconids and z-Puppids are
examples of streams that have not completed their loop formation phase.
Shower Revolution Loop formation

Number, n time (yrs)
Leonids 35 llOO
Orionids 15 llOO
Perseids 13 1700
Draconid s 110 700
n-Puppids 130 650
simple distribution of meteoroids emerges. Figure 3b, however, illustrates the results
of a full three-dimensional calculation including planetary perturbations . Meteoroids
are ejected into the stream from the entire sun-pointing hemisphere, and ejection is
allowed to take place over the orbital arc corresponding to q s r(au) :5: 2.5.
The upper limit to the radial arc is set by the observation that the sublimation of
cometary water ice becomes inefficient for heliocentric distances in excess of about
2 au. Clearly in tins second case the nodal distribution has more 'structure.'
Certainly, the numerical model suggests that none of the O.lg Leonid meteoroids
ejected in 1965 will cross tile Earth 's orbit in 1998 (one orbital period of the comet
later). Importantly as well, only continued numerical integration, including planetary
perturbations, will be able to determine if they are likely to intersect the orbit, say,
three orbital periods hence. Indeed, in general it will be the effects of planetary
perturbations that "shape" the geometrical properties of a meteoroid stream rather
than the initial ejection conditions.
3.0 Observational Characteristics
Meteoroid streams have historically been identified through naked-eye observations

of annual meteor showers. A number of meteor showers have been catalogued by
photographic surveys and several day-time showers have been discovered via
forward -scatter radio and/or radar techniques. The classic "working list" of
meteoroid streams published by Cook (1973) identifies 54 night-time and four day-
time meteor showers. The more refined, and indeed, more realistic, working list of
annual meteor showers published by the International Meteor Organization (IMO)
identifies 38 night-time and 12 day-time meteor showers. The most comprehensive
study of annual meteor showers published to date, however, is that by Jenneskins
(1997), and he presents an overview of 49 night-time showers. The inherent problem
in establishing the existence of an annual meteor shower through visual observations
is that in most cases only a few meteors are seen per hour at the time of maximum .
For the space-user community this, however, is not a major problem since the
meteoroid flux from minor annual showers is entirely negligible.
A meteor shower may be characterized according to its duration, the date on
which it reaches a maximum display. the typical yearly-averaged number of visual
meteors per hour at the time of shower maximum. and the position of the stream
radiant at tile time of maximum activity. While about 50 annual meteor showers are
310
recognized, there are 12 principal night-time and six principle day-time annual
meteor showers. These showers are distinguished by having zenithal hourly rates
(ZHR) in excess of ten meteors per hour at shower maximum. Table 2 is a summary
of the principal meteor showers.
Table 2: Th e prin cipal meteor sho wers. The solar longitud e at th e time of showe r maximum, the
rad iant rig ht ascension and declination are given in degrees. V. is th e geocentric velocity with which the
meteoroids encounter th e Earth. Th e population index r is giv en in column 6 (see text for detail s). Th e
number density gives th e numb er of meteoroids of mass greater than 10-4 gm per 109 km) at shower
maximum. Th e last six entries in the table correspond to the principal day-time meteor showers. Th e ZHR
at maximum and meteoroid numb er density n9 are related through equatio n (6). Number densities given in
parenth esis ass ume a population index of 2.5. Data taken from Rendtel, Arlt and McBeath (1995),
Jennis kens (I 994} and Yrjo la and Jenn iskens (1998).
Shower A at max RA Dec. Vg (kmls) r Peak Number

ZHR density, n9
Bootids 283.3 230 +49 41 2.5 133 480
Lyrids 32.1 271 +34 49 2.7 13 25
n -Aquarid s 45.5 338 - 01 66 2.7 37 15
S. 0- Aqua rids 125 339 -16 41 3 2 180

.3 0
a· 127 307 -10 23 2 10 140
Caoricomi ds .0
Perseids 140.0 46 +58 59 2.5 84 50
u -Aurigids 158.6 84 +42 66 2.7 10 4
Orioni ds 208 95 +16 66 3.1 25 115
S. Ta urids 223 52 +13 27 2.3 10 130
N. Ta urids 230 58 +22 29 2.3 10 130
Leonids 235.3 153 +22 71 3.4 23 7
Geminid s 262.0 112 +33 35 2.6 88 810
Ursids 270.7 217 +76 33 3.4 12 440
o Cetids 59 28 -4 35 * 10 (80)
Arietids 77 44 +24 38 2.7 20 140
c; Perse ids 78 62 +23 27 * 20 (560)
~ Tau rids 97 86 +19 30 * 20 (330)
Sextantids 184 152 0 32 * 10 (120)
Cap/Sa gittari ds 312 299 -15 * * (10 ?) (100)

311
The ZHR is a standardized number derived from the visual observational data. It
is defined as being the number of meteors that would be seen per hour if the shower
radiant was at the observer 's zenith and the limiting visual magnitude was +6.5. The
ZHR may be directly related to the spatial number density of meteoroids per 109
knr'. The procedure by which the observed meteor rates are converted to a ZHR and
the method of conversion to a spatial number density are outlined in Koschack and
Rendtel (1990) and Brown and Rendtel (1996). The spatial number density, n9, of
meteoroids per 109 knr' is given by
n, = 2.78xlO·13 ZHR C(r) / A(r) Vg , (6)
where, C(r) = (13.10 r - 16.45), is a correction factor to account for the total number
of meteors of visual magnitude +6.5 and brighter in the observers field of view, A(r)
= 37,200 (r - 1.3)'0 748, is the effective collecting area in km 2 for a ' standard'
observer , Vg is the velocity of the meteoroid as it enters the Earth's atmosphere in
km/s, and r is the population index defined so that r = NCMv + 1)/NrMv), where
NCMv) is the total number of meteors of visual magnitude Mv or brighter observed in
a prescribed interval of time.
The population index is a measure of the gradient of the cumulative number
versus meteor magnitude diagram for a given shower. Visual observations indicate
that the population index is not necessarily constant with time, and consequently
average values for the principal showers are given in Table 2. In general, however,
the greater the population index. the greater the relative number of fainter visual
meteors. Hughes (1978) has shown that the population index can be related to the
meteoroid mass index through the relationship :
s = 1 + 2.5 loglO(r) (7)
The ZHR may also be used to derive a flux, F(m 2 roo), of meteoroids of mass
greater than or equal to mo per m2 per second. We find
F(m 2 roo) = 2.78xlO·IO ZHR C(r) / A(r) , (8)
where roo is the mass of the stream meteoroid that produces a meteor of peak visual
magnitude +6.5 at the observers' zenith. The peak visual magnitude of a meteor is in
general a function of the initial meteoroid velocity, the initial meteoroid mass and the
zenith angle at which is observed. Verniani (1973) gives a useful empirical
relationship that may be used to evaluate mofor stream meteoroids. When the zenith
angle is zero and the visual magnitude is +6.5 we have,
26 1
mo(g) = 900.5/ V/ , (9)
where the velocity is expressed in km/s. The meteoroid flux at masses different to m,
may be evaluated with the aid of the differential mass index s, such that
(10)
312
The annual variation in the meteoroid flux (meteoroids per second per m2, to a
limiting meteoroid mass of m, = 10-5 g) for the principal meteor showers is shown in
Figure 4.
It is clear from Figure 4 that under typical conditions, the sporadic meteoroid
flux dominates that of the stream meteoroids, at a limiting mass of 1O-5g, by at least
an order of magnitude . TIle greatest stream meteoroid fluxes are realized at solar
longitudes between 70 <'A. < 140 and 260<'A. < 285. Importantly, it should be noted
that meteor showers can undergo periodic outbursts of activity resulting in a ZHR
some 10 to several thousand times higher than the yearly mean. Under these
circumstances, the stream meteoroid flux will dominate that of the sporadic
background. Predicting the times of stream outbursts, however, is not an easy task -
but see below .
Figure 4 highlights the existence of several important differences between the
fluxes of sporadic and shower meteoroids . The overall sporadic meteoroid flux is
essentially constant over the course of a year , and the meteoroids encounter the Earth
from essentially all directions. The flux from shower meteoroids is highly time-
dependent and the direction of origin varies from one stream to the next. While the
sporadic meteoroid flux dominates that of the stream meteoroids at low masses (e.g.,
m S 19), the flux of shower meteoroids will tend to dominate in higher mass ranges
(e.g., m ~ 19). This is partly a result of the smaller mass indices associated with
meteoroid streams, and partly a result of the young - that is collisionally non-evolved
- nature of stream meteoroids . Indeed, meteor showers tend to be 'obvious' to a
visual observer simply because they are rich in ' large' meteoroids, and because the
shower meteors can be traced to a distinct radiant. It is also the case that the
brightness of a meteor is related to the meteoroid Earth encounter velocity to the
fourth power, and typically shower meteoroids have higher encounter velocities than
sporadic meteoroids (see the section on impact probabilities below).
A limiting mass condition must be satisfied for a meteoroid to remain on a
bound , elliptical orbit after ejection from a parent body. This limiting condition is
determined by the stream eccentricity, e, and the solar radiation field (Harwit, 1963).
The condition is given as ~ < (l - e)/2, where ~ is the ratio of the solar radiation
force to that of gravity . It was shown earlier that ~ may be expressed in terms of the
meteoroid density and mass, yielding
mlim (g) = 4 X 10-11 / p (I - e)3 (II)
Meteoroids with masses less than mlirn will rapidly leave the stream on unbound
orbits, and consequently . mlim sets the lower mass limit for meteoroids within a
stream. Stream eccentricities (see Table 4 below) range from 0.65 < e < 0.99, and
meteoroid densities are of order 0.8 g/cnr' in the mass range between 10-3 to 10-5 g
(Vimiani, 1973). Typically , therefore , we expect 1O.9<mlim (g)<5xlO-5 . Babadzhanov
(1993) has recently reviewed the literature on meteoroid densities and argues that
stream meteoroids have much higher average densities than that indicated by
Vimiani (1973) . In particular, he argues (as others have before him) that meteoroid
densities can be correlated with properties of the mean stream orbit. Specifically, the
313
smaller the orbital semi-major axis and the smaller the perihelion distance, the
greater the average meteoroid density. So, for example, meteoroids in the Leonid and
Perseid streams are more friable than meteoroids within the Geminid or 8-Aquarid
streams because meteoroids in the latter two streams move along orbits with small
perihelia (0.14 and 0.1 au respectively - see Table 4 below).
The duration, S ; of a meteor shower can vary from a few days, as in the case of
the Bootids, to in excess of a month or more, as in the case of the Perseids. Data
from the IMO working list of meteor showers indicate that < Ii(deg. solar long.) > =
27.6 ± 17.4 for the main night-time showers and that < Ii(deg. solar long.) > = 39.8 ±
19.4 for the day-time meteor showers. The reason the average duration of the day-
time meteor showers is greater than that of the night-time showers is biased by the
fact that they mostly have radiants near to the ecliptic plane (see below). Figure 5
indicates the times during which the Earth samples meteoroids from the principle
meteor showers.
Meteor showers are typically named after the constellation in which the radiant
is placed at the time of shower maximum. That meteor showers actually have a
radiant is a consequence of the stream meteoroids encountering the Earth along
parallel paths - recall that by definition stream meteoroids have similar heliocentric
orbits. The distribution of meteor shower radiants on the celestial sphere is shown in
Figure 6.
Typically, the meteors from a given stream emanate from a region interior to a
circle of 10 degrees radius about the radiant center. This is partly a consequence of
small variations in the meteoroid encounter velocities and partly due to stream
evolution effects. Likewise, the position of the radiant center advances as the Earth
moves through the stream. The diurnal drift typically amounts to about 0.8 degrees in
RA and some 0.3 degrees in declination. Since spacecraft are only likely to
encounter stream meteoroids at the time of shower maxima, however, stream specific
diurnal drifts are not discussed here.
Averaged over the year and weighted by the ZHR at maximum, the majority of
stream meteoroids will be sampled from the region of the celestial sphere bounded
by 0 s Dec. ~ +60 and 0 ~ RA ~ 180. This result can be presented in an alternate
fashion as illustrated in Figure 7. In this figure, histograms of the distribution of
shower radiants with respect to the ecliptic and the apex of the Earth 's way are
shown. The majority of stream meteoroids are encountered at elongations of between
60 to 80 degrees to the apex of the Earth's way and from latitudes within ±10
degrees of the ecliptic.
As the Earth cuts through a meteoroid stream, the number of visual meteors
observed per hour will vary in accordance with the intercept geometry and the spatial
distribution of meteoroids within the stream. Typically, the observed activity profiles
(ZHR versus time diagrams) are symmetric about the time of shower maximum -
although there are some exceptions to this rule. Cour-Palais (1969), for example,
presented a number of shower activity profiles that were either 'top hat' or 'triangle'
-like functions of time. Naively, one might expect an activity profile to follow a
Gaussian function, but it has become common in recent years to measure shower
activity in terms of a simple power-law function. Jenniskens (1994), for example,
uses a function of the form
314
ZHR(A.) = ZHRmax 1O'BII. ,I.maxl (12)
where A. is the solar longitude, Amax is the solar longitude at the time of shower
maximum, and where B is a constantprofile parameterthat varies from one stream to
the next. Table 3 is a summary of the profile parameters for the principal meteoroid
streams. The larger the value of the profile parameter, the more sharply-peaked is the
activity profile. Equation (12) may be used in conjunction with the data in Tables 2
and 3 to produce approximate activity profiles for each of the major annual meteor
showers.
Table 3. Profile parameters for the principal meteor showers. Data taken from
Jenniskens (1994) and Yrjola and Jenniskens (1998).
Shower B (de!!:·) Shower B(de!!:')

Bootids 1.84 N. 0.03
Taurids
Lvrids 0.22 Leonids 0.39
n-Aquarids 0.08 Geminids 0.39
S. Il-AQuarids 0.09 Ursids 0.61
Ct.- 0.04 o-Cetids 0.08
Capricomids
Perseids 0.20 Arietids 0.10
o-Aurigids 0.19 i;-Perseids 0.05
Orionids 0.12 B-Taurids 0.08
S. Taurids 0.03 Sextantids 0.08
4.0 ImpactProbabilities
To estimate impact probabilities on space platforms some measure of a stream's

meteoroid fluence is required. The fluence F may be evaluated from the integral of
the flux as a function of time. vis
F = f F(A.) ax , (13)
where the flux has been explicitly written as a time-dependant function and the
integration period covers the time of shower maximum. The fluence may be
evaluated in units of meteoroids per square meter by substitution of equations (8) and
(12) into (13). We find,
F = 2 (ZHRmax / B In1O) [ I - 10-B5 I (14)
where Amax - 8 to Amax + 8 is the time window (measured in degrees of solar

longitude) over which the fluence is evaluated. Typically, one can safely assume a
time interval of 8 equal to one solar day to evaluate the shower fluence - the
contributionfor times in excess of one day either side of the maximum will be small.
315
Impact probabilities are evaluated from the product of the fluence and the
spacecraft area exposed to the stream radiant. Clearly, the larger the area exposed to
the radiant, the greater the impact probability . This result indicates a simple impact
mitigation scenario by feathering solar panels, the largest components of a
spacecraft, to the radiant (shower radiant coordinates are given in Table 1). Also, the
instantaneous meteoroid flux is clearly at its greatest value at the time of shower
maximum (see Figure 4), the greatest instantaneous impact risk must also, therefore,
be realized at the time of shower maximum. In other words, the impact risk to a
particular spacecraft may be assessed in terms of the time of shower maximum, and
the direction of the radiant at the time of shower maximum .
While impact probabilities may be evaluated as a function of time and radiant
position, the impact risk must also take into account the potential impact velocity.
The velocity with which a typical sporadic meteoroid encounters the Earth's
atmosphere is 20 kmls . Greater impact velocities are possible, but are less common.
The distribution of sporadic meteoroid impact velocities, in terms of a probability
function , is shown in Figure 8.
Superimposed upon the sporadic velocity probability function in Figure 8 is a
histogram of stream meteoroid impact velocities. For the histogram, the y-axis
corresponds to the number of streams rather than a probability . It is clear from the
diagram that high encounter velocit ies are a characteristic of meteoroid streams. The
median stream velocity is twice that of the median sporadic velocity, and there are a
significant number of meteoroid streams that have encounter velocities in excess of
50 kmls.
The encounter velocity enters the risk assessment through the velocity moment
V". The exponent a will be either one or two according to whether one is interested
in momentum or kinetic energy transfer. If the generation of impact plasma is to be
considered, then a will be or order four . Mass for mass, the collisional effects of
stream meteoro ids, compared to that from sporadic meteoroids, will scale as the
order of 2"'.
5.0 Orbital characteristics
The mean-stream orbital elements of the principal meteor showers are given in Table
4. The elements are primarily taken from Lindblad (1987) and correspond to data
derived from a number of photographic surveys . All of the streams listed in Table 4,
with the exception of the o-Capricornids, have been associated with parent objects .
Since, however, the time over which a typical comet is active (between 103 to 104
years) is much shorter than the estimated coherence time of a meteoroid stream
4 5
(about 10 to 10 years) , the existence of some meteoroid streams, even prominent
ones, without a parent is not entirely unexpected.
Four of the streams listed in Table 4 are associated with comets having
retrograde orbits : the n-Aquarids and the Orionids are associated with comet
IPlHalley, the Perseids are associated with comet 109P/Swift-Tuttle, and the
Leonids with comet 55PlTempel-Tuttle. The Geminid stream has the smallest semi-
316
major axis of all of the known meteoroid streams and it is associated with the near-
Earth asteroid (3200) Phaethon . The Lyrids are the only stream within the principal
group to be associated with a long-period comet, comet 1861 Gl Thatcher, which
has an orbital period of 415 years. The Bootids, the day-time Arietids and the
Southern 8-Aquarids are associated with comet 96PIMachholz 1, while the Southern
and Northern Taurids, as well as the day-time p-Taurids and the s-Perseids are
associated with comet 3PlEncke and its extensive meteoroid complex (Steel et. al.,
1994).
Table 4. Orbital elements for the principal annual meteoroid streams : a is the orbital semi-major axis
in astronomical units, e is the orbital eccentricity, i is the inclination to the ecliptic, OJ is the argument of
perihelion and Q is the longitude of the ascending node . Data taken from Lindblad (1987), except (a)
which is from Lindblad and Porubcan (1994), and (b) which are taken from Cook (1973).
Shower a (AU) e i(") 00(") Q(")
Bootids 3.04 0.677 71.9 171.1 282.4

Lvrids 28.0 0.968 79.0 214 .3 31.7
n-Aquarids 5.0 0.882 165.7 95.9 45 .5
S. Ii-Aquarids 3.09 0.967 26.2 149.5 309 .6
a- Capricornids 2.42 0.758 7.3 270.2 126.9
Perseids'" 24.0 0.961 113.7 151.7 139.0
Orionids 11.5 0.951 164.3 82.7 28.2
S. Taurids 2.2 0.82 5.8 113.2 33 .4
N. Taurids 2.3 0.83 3.3 293.4 228 .1
Leonids 15.0 0.931 162.3 173.1 234.5
Geminids 1.41 0.899 23.5 324.3 260 .2
Ursids 5.8 0.836 52.6 207 .5 270 .7
Arietids'" 1.6 0.94 21 29 77
C:-Persids(b) 1.6 0.79 0 59 77
I3-Taurids(b) 2.2 0.85 6 246 277
A number of the showers listed in Table 4 have been active for many centuries .
Hasegawa (1993) has recently collected together historic records and observations of
meteor showers and outbursts. The oldest known account of a meteor shower is that
of the Lyrids seen in 687 Be when "stars fell like rain." The Perseids are known to
have been active as long ago as 36 AD, and the Orionids may be traced as far back as
585 AD. The oldest account of the Leonids dates to 902 AD, while the earliest
observation of the Bootids dates to 1748. The ancient heritage of the principal
meteor showers is not surprising given that they typically deliver a near constant
number of meteors per hour at maximum each year at the present epoch. The
principal streams essentially present a uniform distribution of meteoroids per unit
volume of space at the Earth's orbit each year (see Table 1).
6.0 Outburst Streams
Jenniskens (1995) has recently reviewed the literature pertaining to meteor showers
that undergo bouts of outburst activity. The outburst streams tend to, but not always,
317
produce appreciable numbers of meteors when the Earth samples the stream shortly
before or shortly after the parent comet has passed through its closest nodal point to
the Earth's orbit. Table 5 is a summary ofthe principal outburst streams .
Table 5: The principal outburst streams . See text for details .
Shower A. Parent comet Last Next

at major perihelion return
max. outburst
Lyrids 3 1861 G I Thatcher 1803 2276
2
x-Puppids 3 26P/Grigg-Skjellerup 1982 2002
3
Draconids 1 21P/Giacobini- 1985 1998
96 Zinner
Orionids 2 IPlHalley 1993 2062
08
(J.- 2 unknown 1995 ?
Monocerotids 37
Leonids 2 55P/Tempel - Tuttle 1966 1998
35
Andromedids 2 3D/Biela 1885 2120 (?)
45
Ursids 2 8P/ Tuttle 1986 2008
71
There is no strict definition for the attainment of outburst status. Essentially, the
label is justified if the observed meteor rates are at least several times higher than the
annual average. In other words, the term ' outburst' must be applied in a stream-
specific manner. The term 'meteor storm' is also used in the literature, and this is
generally taken to mean a 'strong' outburst. Beech et. al., (1995) suggest that the
storm label be restricted to those occasions when the observed ZHR exceeds 1000
meteors per hour - but there is no generally accepted convention on this point.
The Draconid meteor shower, associated with comet 21P/Giacobini-Zinner, was
first observed in 1926, and has subsequently undergone outbursts at storm levels in
1933, 1946 and 1985. Beech et. al., (1996) estimate that the impact probabilities
associated with the 1933 and 1946 storms were 8 x1O-4 and 1 x1O-3 percent per m2
respectively. It is not clear when. or if, outburst activity will be seen from this stream
again, although it is reasonably clear from the visual observations that enhanced
meteor rates are seen at the times of the perihelion passage of 21P/Giacobini-Zinner.
The comet was at perihelion on November 3rd• 1998.
The Lyrids are an example of what Jenniskens (1995) has called a far from
comet outburst stream. The last major outburst of the Lyrids occurred in 1803, some
58 years before Comet Thatcher arrived at perihelion. Similarly, the Lyrid storm of
1122 took place about 91 years before the comet returned to perihelion. There are
intriguing hints within the observational data base that the Lyrids undergo a lower
level outburst cycle with a period of about 12 years. It has been suggested that the
stronger Lyrid outbursts and storms are related to a secondary fragment that broke
away from the main nucleus, but Arter and Williams (1997) have argued that the 12-
318
year cycle is in fact due to gravitational perturbations by Jupiter.

Meteors from the n-Puppid stream were first observed in 1972, and
subsequently notable numbers of meteors have only been seen in those years when
the parent comet, 26P/Grigg-Skjellerup has been at perihelion. Being a 'new' and
still-forming meteoroid stream, the n-Puppids don't so much under go outbursts, but
periodic returns. The most recent Orionid outburst occurred in 1993 some seven
years after the perihelion passage of comet IP/Halley. There is some evidence to
suggest that the Orionids undergo a 12 year outburst cycle (Jenniskens, 1995). Ifthis
is indeed the case, the cause is probably similar to that suggested for the Lyrids and
is the result of Jovian perturbations. Very little is known about the long-term
behaviour of the Ursids other than the fact that they are a far from comet outburst
stream.
The Leonids are by far the most interesting and important of the outburst
streams listed in Table 5. The historic record indicates that of the 25 meteor storms
recorded during the last millennium, II have been due to the Leonid stream. High
Leonid meteor rates can be associated with the return of the parent comet to
perihelion, and the outbursts are modulated. therefore, by the 33.3 year orbital period
of the parent. Storms do not occur every time Comet Tempel-Tuttle rounds the Sun,
but it is often the case that enhanced activity is seen for several years either side of
the time that the comet passes through the near-Earth nodal point.
The Meteor Physics Group at the University of Western Ontario has over the
past several years developed a numerical model to describe the orbital evolution of
Leonid stream meteoroids. The findings of Brown and Jones (1996) indicate that
outburst activity. potentially to storm levels, was highly likely in 1998 and 1999, and
possibly the year 2000. The IMO observations of the Leonid meteor shower
collected during the past several years indicate that Leonid stream activity is on the
increase, and it is anticipated that the up-coming outbursts will occur at a solar
longitude of 235.16 ± 0.04 (Brown and Arlt, 1998). This solar longitude is in fact the
same as that of the peak of the 1966 Leonid storm, and in essence, the stream has
suffered no significant planetary perturbations since the last perihelion passage of
55P/Tempel-Tuttle in 1965.
The penultimate stream listed in Table 5, the Andromedids, is associated with
comet 3D/Biela. the nucleus of which split sometime between 1830 and 1845. In
1872 and 1885, the Earth cut through the stream at a position close to where the
comet would have been had it not split, and intense meteor storms were recorded
(Babadzhanov ct. al., 1991). Since the early 1900s, planetary perturbations have
carried the Andromedid stream away from an Earth-crossing orbit and the shower is
now no longer detectable.
The Earth will be able to sample the Bielid stream again beginning circa 2120
(Carusi ct. al., 1985), however, and enhanced rates may once more be seen. Figure 9
shows the time variation of the nodal distance of the 3DIBiela as calculated by
Carusi ct. al., (1985). The radiant of the new Bielid shower, should it appear, will be
in Pegasus and it will reach maximum in early March.
At the present time, the meteoroid fluences resulting from outbursts may only be
studied in a parametric manner. That is, the ZHR at outburst maximum and the time
during which the activity is enhanced are not predictable, for any stream, with any
319
degree of certainty. Beech et. al., (1997), however, have outlined a procedure for
estimating Leonid meteoroid fluxes under storm conditions, and this methodology
may be generalized to any other stream.
7.0 Acknowledgments
The authour gratefully acknowledges the travel and research funds provided by the
Natural Sciences and Engineering Research Council of Canada.
8.0 References:
l .Arter, T., and Williams, I. P.. Mon . No t. Roy. astron. Soc., 286, 163 - 172, 1996.
2. Babadzhanov, P. 8. , Wu , Z., William s, I. P., and Hughes , D. W., Mon . Not. Roy . astron. Soc., 253 , 69-
74,1991.
3. Babadzhanov, P. 8., , Meteoro ids and their Parent Bodies, Stohl, 1., Williams, I..P. (Eds.) A stronomical
Inst. Slovak Acad. Sci., Bratislava. pp. 295 - 301, 1993.
4. Beech, M., Brown. P., Jones, 1., and Webster , A R., and Worsfold, R., Proce edings of the Second
European Conf erence on Spac e De bris. ESOC, Darmstadt , Germany, ESA SP-393 , 387 - 389 , 1997.
5. Beech, M., Brown . P., Jones, J.. and Webster. A R.. Adv. Space Res. 20,1509 -1512,1996.
6. Beech, M., and Steel. D., Q. 1. R. astro. Soc. 36. 281 - 284 , 1995.
7. Beech, M., and Youngblood, R.• The Observatory. 114,312,1994.
8. Brown, P., and Arlt, R., WGN, Journal ofIMO, 26:1, II - 12, 1998.
9. Brown, P., and Jone s, 1., Earth , Moon and Planets, 68. 223 - 245, 1995.
10. Brown, P., and Jone s. 1.. Physic s, Chemistry and Dynam ics of Interplanetary Dust , Gustafson, 8. A
S., and Hanner . M. S. (Eds .), ASP Conference Series. 104. 113 - 116. 1996.
II. Brown , P., and Rendtel , 1.. Icarus, 124.414 - 428, 1996.
12. Carusi, A, Kresak , L., Perozzi, E., and Valsecchi, B. G., Long-term Evolution ofShort -period Comets.
A Hilger, Bristol , 1985.
13. Cook, A F., Evolutionary and Physical Propert ies of Meteoroids, eds., Hemenway, C. L., Millman, P.
M., and Cook , A F., NASA SP-319 , NASA, Washington. pp. 183 -191,1973.
14. Cour-Palais, 8. G.. Meteoroid Envirorunent model-I 969. NASA SP·80J3. 1969.
15. Grun, E., Zook, H., Fechtig, H., and Giese , R. H., Icarus, 62, 244 - 272, 1985.
16. Gustafson, B. A S., Ann. Rev. Ea rth Planet. Sci., 22. 553 - 595, 1994.
17. Harwit , M., 1. Geophys. Res., 68, 2171 - 2180.1963.
18. Hasegawa, I.. Meteoroids and their Parent Bodie s. Stohl. J., and Williams, I.P, (Eds.) Astronomical
Inst. Slovak Acad. Sci., Bratislava. pp. 209 - 223 ,1993.
19. Hughes, D. W., Cosmic Dust, ed. McDonnell, 1. A M. Wiley, New York, pp. 123 - 185, 1978 .
20. Jenniskens, P., A stronomy and A strophy sics. 287 , 990 - 1013, 1994 .
21. Jenniskens, P., A stronomy and A strophysics, 295, 206 - 235 , 1995.
22. Keller, H. U., Kramm, R., and Thomas, N., Natur e, 331 , 227 - 231, 1988.
23. Koschack, R., and Rendtel, 1., WGN J. IMO. 18, 119 - 141, 1990.
24. Lindblad , D. A. The Evolution of the Small Bodies of the Solar System. Soc. ltaliana di Fisica,
Bologna, 1987.
25. Lindblad, B. A, and Proubcan , V., Planet. Space Sci.. 42, 117 - 122, 1994.
26. Rendtel, J., Arlt, R., and Mcbeath, A, Handbookfor Visual Meteor Observers , IMO Monograph #2,
1995.
27. Steel, D., Asher. D. J., Napier, W. M., and Clube , S. V. M., Hazards due to Comets and Asteroids,
Gehrels , T. (Ed .). University of Arizona Press. Tu cson . pp.463 - 477, 1993.
28. Vimina, F., J. geophys.Res. 78.8429 - 8462.1973.
29. Whippl e, F.• Ap. 1. 111.375.1950.
30. Wu, Z., and Williams, I. P.. Mon. Not. R. As tron. Soc . 264,980 - 990, 1993 .
31. Ytjola, I., and Jenniskens. P., Astronomy and Astrophysics, 330, 739 -752. 1998.
320
Figure 1. Orbits of Leonid meteoroids projected ontothe ecliptic plane. The axes givethe scale in astronomical
units. The orbitsillustrate the idea that meteoroid streams have a pinched elliptical toms -like structure, with the
' pinch' beingat the perihelion point.
321
v..,.
t?-=-4--~--... ToSoo
••' v..
.............
v.
"5'
~
.~ 10.4
ell 55PfTempel-Tuttle
....
o
'(ij'
E
~ 10.2
E
Q)
C/)
10.0 + -r-e- ..,.L
-80 -60 -40 -20 o 20 40 60 80

Ejection angle (deg.)
400
300 material ahead
200 of comet
lil
>- 100
ell
~
0
t:i
-100
material behind
-200 comet
-300
-80 -60 -40 -20 0 20 40 60 80
Ejection angle (deg.)
Figure 2. Orbital characteristics of meteoroids following ejection. Thetop schematic illustrates the geometry of the
modeland defines the anglea. The upperpanelshows thevariation of thesemi-major axis with a - as desaibed by
equation (2). The horizontal line indicates the semi-major axis of CometTempel-Tuttle. The lower panel shows
the lead and lag timesby whichmeteoroids returnto perihelion compared withthat of the comet,
322
0.990 . . . , . - - - - - - - - - - - - - - - - - - - - - - - ,
0.989
Earth'sheliocentric distance: A. =235.2
Q)
0.988 .,E
01
c
'iii
III
0.987 §
Iii
'0
o
C
0.986
nodaldistance (1965)
S-
III
........
Q)
o
c 0.985
-
III
I II
'5
0.984
materialahead material behind
of comet comet
0.983
0.982
0.981 nodaldistance(1998)
0.980
-600 -400 -200 o 200 400 600
~T (days)
Figure 3a. Nodaldistribution of Leonidmeteoroid; - planar geometry model The nodal crossingtime (oompared
with that of CometTempel-Tuttle)and nodal distanc:c is plottedfor meteoroid; ejectedat pcrihc\ion and with-9O
Sa (dcg.) S +90. The calculations arc basedupon the 1965 orbital parameters for cometS5PfTempel-Tuttlc. The
nodal distances of Comet Tempel-Tuttle in 1965 and 1998 are indicated, along with the Earth's heliocentric
distanceat the timeof the annualLeonidshower, Sectextfor details.
323
0.990
Earth's heliocentric distance: A. = 235.2

0.989
IQ)
0.988
I:§
Ig>
, '~
0.987 I§
,~
0.986
,2
S' I nodal distance (1965)
~
Q)
c: 0.985
0
co
Ui
'5
0.984
0.983 material ahead material behind

of comet comet
0.982
0.981 nodal distance (1998)
0.980 -t--,.--,------.---,.--..,--+--,...---,---.---,--,--t-
-600 -400 -200 o 200 400 600
ilT (days)
Figure 3b. Nodal distribution of Leonid meteoroids • full numerical simulation including radiation pressure,
hemispherical ejection and planetary perturbations. The orbital evolutionof 1000 meteoroids were followed to
produdcthe diagram. See text for details.
324
I I I I
10-8
Sporadic
10-9 - G)
en ro
.rt c
3-' ..,
I
"0
"'CP
I 0'> , 2. ~.CD
-l
..,
ro
-
N~ 0) » 0. 0. 0
.0 (J) (J) 0
-~
E (J)
ro
0: 0:
c
.., c0)
~.
0:
(J)
-'e
(J)
(J) 0. j
( /)
(J)
en
ro
'0 10-10 - ~ -l f-
o, "0 0)
-
o 0)
ro ::s c..,
2CD oro @
0:
g ro 0:
X
~=
,0. 0: (J) (J)
:;]
q:: r »(J)
(J)
0..,
-e .0
<:,- r
~.c
0.0) ::s ro
(J) ~. ~ a: o
::J
0. (J)
(J) 0:
10-11 - Q, (J)
f-
»
c
..,
cO·
0:
(J)
10.12 I I I I I I
o 40 80 120 160 200 240 280 320 360

Solar longitude (deg.)
Figure 4. Annual stream meteoroid flux (at shower- maximum) to a limiting mass of 1O-5g. The line labeled
'sporadic' corresponds to the yearly avenge non-shower meteoroid flux (see, e.g., GriIn et: al., 1985).
325
AUR
:f:
PRINCIPAL NIGHT-TIME SHOWERS
y
Kqtol)'mbols
T m Vc:mal equinox
LYR • Lyrids(36)
EAQ • 1\ Aquarids (45)
DAQ • Ii Aquarids (125)
ACA • a. Caprioomids (127)
-
PER • Pmcids(I40)
GEM _BOO AUR • a. Aurigids (159)
OR! Oriooids(208)
TAU • S. Taurids(223)
N. Taurids (230)
LEO • Lconids (235)
GEM m Geminids (262)
BT~ ~~ - - - - - - . : : ZPE OCE

URS • Ursids(271)
BOO • Bootids(2S3)
ARI
OCE • 0 CdicIs(59)
ARJ • Arictids (17)
ZPE • t; Perseids (78)
SEX BTA • II Taurids(97)
SEX • SextanIids (184)
CIS = CaplSagjuarids (312)
PRINCIPAL ~Y-TIME SHOWERS
( y
TH'S ORBIT
-C,{/
Figure 5. The Earth encountertimes for the principal meteorshowers, The arcs indicate the times, measured in
degrees of solar longitude, over which the Earth samplesmeteoroids from the stream. The radial positions of the
arcs are not intended to rcpreseot spatial distances - they are separatedfor clarity of recognition. T is the Spring
Equinox and this point approximately indicates the day of March 21st Solar longitudes of 90, 180 and 270
degrees correspondto the approximate datesof June 21st, September 23rd and December 2200 respectively. The
numbers in bracketsfollowing the showeridentification codesindicatethe solar longitude of maximum activity.
326
90
80 + Urs
70
60
50 And Aur
...
Ora
o
o
o 0800
T
40
Gem o
. 0 ... Lyr
30 Ari zPe 0 Leo
~~ bTa
T
-
Cl
20 OTau , o r i o
Q)
-c
'-'"
10 o
Sex n Aqu
c=
.Q 0 ········. ······O····························0 ...JA:.....························_·_·················;··S·A~~·····8·····
coc:
Q)
-10
oCe 0 d) 0
(3
Q)
Cl
-20
o 0 0 00Sc . C/S qg
o 0
-30 o
-40 o uPh o
Phe+
ppu~ o
-50 o o
o o OaCe o
-60
o OdPa
-70
Typical 10 deg .
-80 radiant radius
-90
o 20 40 60 80 100120140160180200220240260280300320340360
Figure 6. Annualmeteor shower radiant distribution. Open circles correspond to the annual showen discussed by
lcnniskcos (1994) . The • symbols indicate the primary outburst showen (see table 5 for details). The A symbol
indicates a principal day-timc shower. while the T symbol indicateS a shower that has historically undergooc
outburstsat storm levels. The typical radiant area corresponding to a radius of 10 degrees is illustrated in the lower
left-handcomer.
327
.sg
1 - ..-- f-
C
til
~
a... .---
Ql
.D
E
::J
C
"0 ~
Ql
.~
iii - ~ ~
...o
E
'---- ~
Z
o I
-,
II I
I I
-90 -80 -70 -60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60 70 80 90

Ediptic latitudein degrees
.sg
- ~ f-
C
til
15
~ ~
00-
o
...
Q)
.D I--
E
::J ~ ~
C
"0
Ql
.~ I--- I--
(ij
E
o
I--
z .----
o h I
n
I I
o 10 20 30 40 50 60 70 80 90 100 110 120 130 140

Elongation fromthe Apex of the Earth's way
Figure 7. Radiantdistribution of the principal meteorshowers with respect to eclipticand withrespectto the apex
of the Earth's way.
328
I I I
10
9 - r---
(i)
"C
...
'0
8 - ,..--
-0
Q)
Q)
E
0
'6 7 l-
...
«l
0
c.
<I)
~ 6
.......
0 e--
~
:0
«l
.0
...
0 5 - -
0-
<i'
E
«l
... - -
-Q) 4 ,..-- I--
<I)
"C
'e
0
2Q) 3 - - I-- -
-...
E
1\
Q)
.0
E
2 - I-- f-- -
:::l
Z
1 -
1\<,
0
r--
I
o 10 20 30 40 50 60 70 80
Velocity (kmls)
Figure 8, Velocity distribution of sporadic and streammeteoroids. This is a composite diagram • the histogram
corresponds to the binnedstream data, whilethe cwve indicates the probability distribution for sporadic meteors.
Note the y-axisis to be read as a number in the case oftbe annualstreamhistogram and as a probability for the
sporadic meteor curve,
329
___ Ascending node

- Descending node
........... Earth's orbit (1 AU)
5
1600 1700 1800 1900 2000 2100 2200 2300 2400

Year
Figure 9. Time evolution of the nodalpointsfor comet 3D/Biela. Circa 2120 the Earth will potentially sample
meteoroids fromthe streamagain. Orbital data takenfromCarosi et. al., (1985).
OXIDATION KINETICS OF VB GROUP METALS BY ATOMIC AND
MOLECULAR OXYGEN
AG. GUSAKOV, S.A RASPOPOV, AG. VOROPAYEV,

M.L. ZHELUDKEVICH, A.A. VECHER
Department ojChemistry, Belarus State University
Skoriny Ave. 4, Minsk 220080, Belarus
1.0 Introduction
Atomic oxygen is the main constituent of Earth 's atmosphere at altitudes of 200-300
km, and is known to be the main cause of the degradation of parts of space apparatus
exposed to outer space. Of the VB group metals, only tantalum has been tested
previously in atomic oxygen [1,2], but at much higher temperatures (2400-2800 K)
when reaction products (faO and TaO 2 ) evaporate immediately on formation .
Dissolution of oxygen in tantalum occurs simultaneously with oxide formation at
such high temperatures. It was found in [1, 2] that the probability of oxygen
chemisorption on a metal surface is close to unity. Under conditions in the work
presented, solid oxide scales are formed on metal surfaces. Processes of oxygen
dissolution and oxide growth are separated in time.
2.0 Experimental methods
Studies of the kinetics of the interaction of metals with atomic and molecular oxygen
at low pressures have been conducted using a vacuum flow system pumped by a 600
I . s" diffusion pump with polyphenyl ether fluid and a liquid nitrogen trap. The
experimental setup is described in refs. [3,4]. Accelerated material tests were
conducted at oxygen pressures in the range of 0.01-0.1 Pa which corresponds to an
incident flux of 1016 - 1017 atoms' em" . S·I . Atomic oxygen was produced by
passing diluent-O, gas mixtures through a microwave discharge before entering the
specimen chamber. The dissociation degree of oxygen was determined by N0 2
cherniluminescent titration. It could be varied in the range 4.5 to 45 per cent by
varying the diluent-O, mixture composition which allowed additional information
about the reaction mechanisms to be obtained.
Specimens were metal strips 100-150 microns in thickness, heated resistively to
the temperature of the experiment (773-1573 K) by alternating current. Temperature
was maintained constant during the experiment. Kinetic data were obtained by
monitoring the specimen's electrical resistance with a four-probe method. This
technique allowed the investigation of both the dissolution of oxygen in metal, and
331
332
the growth of oxide film on the surface, inasmuch as both processes result in an
increase of specimen resistance from which kinetic data can be computed . In the
latter case, the resistance increase is due to the reduction of strip thickness with the
formation of non-conducting oxide surface film. In the former case, a more
complicated model, taking into account uneven spatial distribution of dissolved
oxygen within the specimen (4), has been used to describe experimental
resistance/time curves and, consequently, to obtain the kinetic data.
Other methods such as gravimetry, metallography and X-ray diffraction studies
(determination of oxygen concentration via lattice parameters) have been utilized to
supplement the electrical resistance measurements, and have confirmed the kinetic
data obtained in most cases.
3.0 Results and discussion
3.1 TANTALUM
Figure I shows an example of resistance/time dependencies for tantalum interactions

with molecular and dissociated oxygen, other factors being equal. Two sections with
different slopes can be distinguished in the dependencies: the initial section
corresponds to dissolution of oxygen in tantalum; the final section is due to growth of
oxide film on a metal surface. One can also see from Figure I that the action of
atomic oxygen greatly increases the oxidation rate for both stages.
1.5
1.4
~1.3
1.2
1.1
1.0 f"!!--'--'----'--'--J....--'---'--'---'_L-.1--.J.......J
o 40 80 120 160 200 240
Time, min
Figure I
Electrical resistance vs. time plot for oxidation of tantalum in molecular and dissociated oxygen
at 1073 K and incident flux of 1017 atoms' cm·2 • s·\
Experiments at different values of oxygen pressure and dissociation degree have

been carried out in order to clarify the kinetic mechanisms. These showed that first-
order kinetics are observed at the stage of oxygen dissolution in metal.. Reaction rate
is proportional to oxygen incident flux in the case of molecular oxygen, which allows
the introduction of the magnitude of the probability of molecular oxygen absorption
333
by tantalum surface (absorbed fraction of incident molecules). In the case of
dissociated oxygen, atoms and molecules are absorbed independently with
corresponding probabilities, i.e.
(1)
where Z abs is the oxygen flux absorbed by metal; Z m and Z a are the incident
fluxes of molecular and atomic oxygen respectively; 8 m and 8 a are the absorption
probabilities of molecular and atomic oxygen respectively by a metal surface.
Such kinetics indicate that chemisorption of atoms or molecules on a metal
surface is the rate-limiting step both for molecular and atomic oxygen. Equation (1)
was utilized for calculation of the atomic oxygen absorption probability from the data
obtained by oxidation of tantalum in dissociated oxygen.
Figure 2 presents the dependence of the absorption probabilities of molecular
and atomic oxygen on temperature . The absorption probability for atomic oxygen is
well above that for molecular oxygen (by two orders of magnitude), in particular at
relatively low temperatures . In addition, the activation energy for atomic oxygen (29
kJ/mol) is much less than for molecular oxygen (74 kJ/mol).
~ 1E + 0 .,.-,r---r-,.---r---r--r-...---.-...--..---,....--,-.,--,----r.....
:i
~ 1E-1
ro
o
§t 1E-2
z
o
f= 1E-3 MOLECULAR OXYGEN
0.- E=74 +/- 2 kJ/mol
<=>::
o 1E-4
CI)
ro 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4
«: 3
10 /r, K
1
Figure 2
Temperature dependence of the oxygen absorption probability
for the stage of dissolut ion (Arrhenius plot).
As shown by tile measurements of the tantalum lattice parameter and

microhardness, strong oversaturation of the metal surface with dissolved oxygen
takes place during tantalum oxidation in dissociated oxygen at relatively low
temperatures (below 873 K).
Linear oxidation is also observed at the final stage of the oxidation process
(growth of Ta205scale). Consequently, tile rate-determining step is not diffusion via
oxide film but chemical reaction at tile metal/oxide or oxide/gas interface. This
agrees with the fact that Ta205possesses high oxygen ionic conductivity. The rate of
334
tantalum decrease for oxide formation (Z Ta) during oxidation in molecular and
atomic oxygen is given in Table 1 as a function of temperature.
It follows from the data presented that atomic oxygen offers substantial
acceleration of oxidation at tins stage as well. Decrease in the oxidation rate is
observed above 1373 K, which correlates with the beginning of the formation of
high-temperature cc-mcdification of Ta20s as found by X-ray diffraction studies .
Dependencies of oxidation rate on oxygen incident flux and dissociation
degree have been investigated for the stage of scale growth as well . A model has
been suggested to explain tile obtained results based on the assumption that at this
stage, the oxidation rate is determined by the reaction of oxygen chemisorbed on the
Ta20 Ssurface in accord with tile Langmuir-Hinshelwood mechanism [3), namely, the
rate is proportional to tile surface coverage with chemisorbed oxygen. The final
formula is:
(2)
where k m and k a are the complex rate constants for molecular and atomic oxygen
respectively . As follows from Equation (2). Zfa / Zm should be a linear function of
Z a / Z m in the case of dissociated oxygen. whereas Z Ta should be proportional to
~Z m in the case of molecular oxygen . This does agree with the experiment.
Enhancement of oxidation in atomic oxygen is due to a higher chemisorption
probability of oxygen atoms at tile surface of oxide film as compared with molecules.
Table I. Reaction rate of tantalum with molecular and dissociated oxygen at the stage of oxide
scale growth at incident flux of 1017 atoms' cm·2 • s·' .
Temperature / K
in molecular oxygen at the dissociation degree of

.., _.- ~_ _ -_ _ _ - - - _ 28.7% _ _
~ _ _.
973 0.79 2.8
1073 1.8 3.6
1173 2.7
1373 4.4 6.2
1473 2.8 3.9
1573 3.5 4.7
3.2. NIOBIUM
Similarly to tantalum, two stages of the oxidation process are observed for niobium :
(1) dissolution of oxygen in a metal lattice; (2) parabolic growth of oxide scale [5].
At the first stage, linear oxidation rate law and linear dependence of the reaction
rate on incident flux are found for niobium as well. Equation (1) is also correct.
Figure 3 shows the temperature dependence of the absorption probabilities of
molecular and atomic oxygen by niobium as a function of temperature. A very
335
similar plot has been obtained for PWC-ll alloy (Nb-Iat.% Zr - 0.7 at.% C). The
absorption probability of atomic oxygen depends only slightly on temperature in all
temperature ranges studied. As for the molecular oxygen absorption probability, it is
temperature-independent above 1073 K as well, whereas strong temperature
dependence is observed at lower temperatures (activation energy is about 115
kl/mol). Difference in the reaction probabilities of atoms and molecules amounts to 2
orders of magnitude around 773 K.
ATOMIC OXYGEN, E=12 kJ/mal

•
•
MOLECULAR
OXYGEN
E=115 kJ/mol
0.8 0 .9 1.0 1.1 1.2 1.3

3 -1
10 rt, K
Figure 3
Dependence ofthe probability of oxygen absorption by niobium
at the stage of dissolution as a function oftemperature.
Parabolic oxidation was found at the stage of oxide scale (consisting mostly of
NbO and NbOz) growth, i.e., the rate of reaction is determined by diffusion of ions
through the scale. Nonetheless, the action of atomic oxygen results in a substantial
increase in oxidation rate for this stage, also. The values of the parabolic rate
constant for oxide film growth are given in Table 2.
Table 2. Parabolic rate constant for oxidation of niobium in molecular and dissociated oxygen
at the stage of scale growth at incident oxygen flux of 1017 atoms' cm·2 • s"
Temperature / Parabolic rate constant / mkm2 . min- 1

K
in molecular oxygen at the dissociation
..............................................................................................................~~~E~~ .~f..~.~:7.~ ..
973 0.23
1073 0.17 0.30
1173 0.20 0.45
1273 0.50 0.93
From the viewpoint of Wagner's theory of oxidation, the rise of the oxidation
rate on the action of oxygen in an atomic state in conditions of diffusion control can
be explained by that surface coverage of oxygen and, consequently, chemical
336
potential of oxygen at the surface of oxide film increases strongly due to high
chemisorption probability of oxygen atoms on the oxide surface. TIlls leads to an
increase of the rate of ionic diffusion via oxide film.
3.3 VANADIUM
Only the first stage of the oxidation process has been studied for vanadium -
dissolution of oxygen. Similar to other VB group metals, linear rate law is observed
with absorption probabilities of atoms and molecules independent of corresponding
incident fluxes. Dependencies of the probabilities on temperature are presented in
Figure 4. Difference in the absorption probabilities of molecular and atomic oxygen
is not as large for vanadium (approx. threefold). The corresponding activation
energies are close as well.
MOLECULAR
OXYGEN, E=54 +1- 8 kllmal
0.8 1.1 1.2
Figure 4
Arrhenius plot of oxygen absorption probabilit y by vanadium at the stage of dissolution .
4.0 Conclusions
Two stages are observed for VB group metals oxidation: (1) dissolution of oxygen in
metal and (2) growth of oxide film. No change of the kinetic mechanism was found
at both stages when molecular oxygen is substituted with atomic oxygen. but higher
oxidation rates are observed in dissociated oxygen due to higher chemisorption
probabilities of oxygen atoms on metal or oxide surfaces. TIlls difference is
particularly large at relatively low temperatures.
At stage (I), reaction probability with oxygen decreases when going from V to
Ta in the VB group in the temperature range investigated (773-1573 K). Oxygen
chemisorption on a metal surface is the rate-controlling step both for molecular and
atomic oxygen. The difference between the reaction probabilities for atomic and
molecular oxygen is about 1-2 orders of magnitude.
At stage (2), linear oxidation is observed for tantalum, the rate-controlling step
being the reaction of oxygen chemisorbed on oxide scale surface according to the
Langmuir-Hinshelwood mechanism. In the case of niobium, parabolic rate law is
337
observed at this stage, i.e., diffusion control takes place, but atomic oxygen still
offers enhancement of oxidation.
5.0 References:
1. Rosner, D.E., Chung, H. and Feng, H., Me rall. Trans., 5,2303, 1974,
2. Rosner, D.E., Chung, H. and Feng, H., Faraday Trans. 1,72,842,1976.
3. Gusakov, AG., Raspopov, S.A., Vecher.. A A and Voropayev, AG.,J. Alloys Comp., 202, 67,1993.
4. Raspopov S.A, Gusakov, AG., Voropayev, AG., Vecher, A.A, Savitski, AA. and V.K. Grishin J.
Chern. Soc., Faraday Trans., 92, 2775, 1996.
5. Raspopov S.A, Gusakov, AG., Voropayev, AG., Vecher, A.A, J. Alloys Comp., 205, 191, 1994.
A NEW APPROACH IN DEVELOPING METHODS
FOR BETTER PROTECTION OF MATERIALS AND STRUCTURES
FROM LEO SPACE ENVIRONMENT
A.K. GALYTSKY, A.I. STEGNYY, V.P. GUSYNIN, 8.M. SHEVCHUK

Federation OfCosmonautics a/Ukraine
National Space Agency 0/ Ukraine
/2 , Trostyanetskaya SIr., Kiev 253/75, Ukraine
1.0 Introduction
Rapid scientific and technological progress, accompanied by a high accident rate of

human-machine systems, created the necessity of developing strategies and methods
for increased reliability of complex technological systems. In accordance with this
philosophy, a new way to increase protection of materials and structures from
hazardous influence of space is suggested.
The key operational tool to achieve this proposed approach was developed
by one of the authors (A.K. Galytsky) and is based on information technology
forecasting space weather disturbances. Allowances were made for special
technological conditions during production of materials and electronic units,
choosing the most optimal time (during periods of "quiet space" when the
probability of accidents is minimal) to launch space vehicles, and for testing both
aircraft and complex systems. The protection strategy proposed could lead to an
increase in material strength and in reliability of the equipment, as well as lowering
the rate of human errors.
Numerous research studies confirm that our biosphere is affected by solar
and geophysical factors. Convincing proof exists that solar activity impacts on the
rate of accidents in transportation [1 ,2], influences the rate of technology-caused
catastrophes [3], affects the change in physical and chemical properties of solutions
and materials [4,5,6] and physiological processes in the human body [5]. In a study
dealing with the influence of cosmic factors on the biosphere it was stated [5] that
there are no objects or processes on the Earth where the influence of space would not
be felt to some degree.
One of the most pressing problems at the present time is the transition from
theoretical studies to the development and implementation into practice of
technologies for increasing the reliability of complex systems and prevention of
technology- and/or ecology-related catastrophes. As a result of interdisciplinary
research, a new approach based on information technology forecasting the periods of
increasing probability of accidents in different spheres of human activity was
developed in the last few years. Extensive testing of the proposed concept conducted
in the last eight years confirmed its wide functional applicability. Information
339
J.t. Kleiman and R.C. Tennyson (eds.) ,
Protection ofSpace Materials from the Space Environment . 339-349.
© 200t Kluwer Academic Publishers .
340
technology can become an important tool for preventing ecological and

technologically-generated catastrophes, and lowering accidents in LEO space
environment, in industry and in transportation.
2.0 Theoretical Background
The influence of space factors on material objects is expressed through the changes
of their physical and chemical properties. Strong disturbances in the characteristics
of space factors can cause changes in materials structure, the concentration of defects
and dislocations, etc. Such changes , in tum, may induce a deterioration of physical
and chemical properties in the materials that may affect their sealing properties,
induce electrolytic corrosion , parasitic currents, and lead to power cuts. One should
also take into account the impact of cosmic disturbances on material objects . Very
often, disturbances with seemingly insignificant energy levels can produce
substantial effects on the structure of materials.
The reliability of human-machine system functioning depends, to a great
extent, upon human factors. Cosmic disturbances, therefore, in addition to
influencing the physical and chemical properties of materials, may cause sharp
changes in physiological processes of a human-operator. In such cases, the
probability of erroneous actions leading to accidents and catastrophes increases
greatly .
Thus, the complex impact of cosmic disturbances upon technological
systems and the human operator is the most important factor limiting the reliability
of human-machine systems. There are periods of time, during powerful cosmic
disturbances, when the probability of accidents is very high, especially in complex
technological systems. If such periods can be predicted, recommendations can be
made to abstain from terrestrial activities like spacecraft launches, air flight tests,
conduct of air shows, and testing of vital complex control systems to avoid accidents
and catastrophes.
For instance, after a number of accidents that inflicted financial and human
losses, the USA Department of Defense had to cancel all Air Force flights due to a
high rate of accidents in September 1997 and October of 1989 (both periods were
characterized by strong solar-geophysical disturbances). A study of conditions in the
LEO space environment in these periods could have thrown more light upon the
cases of the rise in the accident rate.
3.0 Methodology
In the process of developing information technology of forecasting, a number of

tools including standard statistics, time-series analysis and such processing methods
as correlation analysis, spectral analysis, period-gram analysis and smoothing
techniques were used.
Long term interdisciplinary research of a wide range of processes in nature
and society has shown that, due to the nonlinear and non-stationary character of real
341
physical processes, standard statistical methods are not applicable for their
evaluation. They wouldn't allow adequate resolution of analysis and forecasting of
time-series dynamics of natural-process problems and, in many cases, may even lead
to the loss of information. As a result, strong disturbances in space factors that may
have a large energetic and informational impact upon a wide range of processes on
the Earth, and which trigger mechanisms that cause accidents and catastrophes are
lost or are dismissed as artifacts.
To resolve this problem, special approaches and methods were used to
obtain and to model the integral characteristics of energy processes in space. The
modeled curve shows the dynamics of space energy processes. The new and primary
idea of the information technology of forecasting developed is based on the fact that
the biosphere is considered as a complex highly-organized system with a multitude
of interrelated links. The key moment is the assumption about synchronization of
macro and micro-levels in the biosphere (confirmed by our own and our
predecessors' experimental studies).
In the mathematical model, the main links between the processes in
interplanetary space are integrated. The computer modeling, based on the system
analysis of functional links between different solar-terrestrial processes on macro
and micro-levels, allows forecasting of cosmic disturbances and periods (days) with
increased probability of accidents in different spheres of human activity for one year.
The information technology allows the prediction for one year of such delicate
information-energy changes in interplanetary space that is impossible to measure
with modem instruments and which precede solar flares, ionosphere disturbances
and magnetic storms that often are the triggering mechanism of accidents
One of the important features of the information technology concept is the bio-
monitoring data analysis block, which increases the reliability and sensitivity of the
forecasting system. With the help of analysis algorithms developed, the information
component of bio-systerns response to those cosmic disturbances that precede solar
flares, ionosphere disturbances and magnetic storms, and that cannot be registered by
the contemporary measuring devices is selected. Evaluation and confirmation of
information forecasting technology concept has been conducted since 1989 when
one of the authors (A.K.G.) began work on its development. At the end of each year
the changes in integral characteristics of energy processes in interplanetary space
were calculated for the next year to come (Figures 1 and 2), using computer
modeling. Thus, the calculations were done at the end of 1989 for 1990, at the end of
1990 for 1991 etc. During the next eight years, monitoring of the major parameters
was conducted on a daily basis. Among the parameters monitored, are the dynamics
of solar-geophysical factors, the occurrence of technologically-generated accidents
and ecological catastrophes. The data on air accidents for 1984-95 were also used
[7].
The analysis of data on accidents and catastrophes in the aerospace
industry, transportation, and on technology-generated and ecological catastrophes
has shown that, in most cases, space disturbances may be assumed as the triggering
mechanism. As a rule, the periods of highest accident rates were tied to the
characteristic points (maxima, minima, points of sign change, changes in the
342
monotonous behavior of the curve) on the curve of changes of the integral

characteristic of energetic processes in interplanetary space.
The relation of accident rates in 1996-97 versus the most characteristic
points/intervals on the modeled curve of energy processes' changes in interplanetary
space are shown in Figures I and 2. According to the analysis conducted, the most
dangerous periods are those of activation of energetic processes in space (the rising
part of the curves in Figures I and 2). Among them, the highest accident probability
corresponds to the characteristic points in the vicinity of the beat nodes and to the
maximum deviation of the long period component of the modeled curve (dotted line
in Figures I and 2). These observat ions are true at periods, when the restructuring of
periodic structure of energy processes in space is observed that is accompanied by
disturbances of gravitational and/or electromagnetic fields and solar plasma fluxes.
Within such periods, changes in physical and chemical properties of materials and in
physiological processes in a human organism may occur, caused by the influence of
cosmic factors.
Figures 3 and 4 demonstrate the dependence of accident rates in aviation on
solar activity. Analyzing the biggest air-crashes of Boeing planes within the last 20
years (Figure 4) one could observe that they are grouped around minimums (1974,
77,85,86,96) and maximums of solar activities' l l-year cycle (1979, 80, 88, 89).
There are intervals where the periodic structure of wave processes in interplanetary
space changes, having an impact on physical-chemical properties of materials and a
human-operator through the solar wind or changes of gravitational and
electromagnetic fields. Hence, during such periods, the probability of malfunctions
in complex technological systems and the number of errors of a human-operator will
increase.
4.0 Validity of the Results
The validity of the proposed concept of information technology of forecasting was

confirmed by analyzing a large pool of experimental data, accumulated in the last 22
years of research and eight years of testing of the models developed. To confirm the
distribution series of accidents and catastrophes within a year, a calculation was
conducted where it was compared (using Student and Fisher criteria) to a model of
random numbers series. The obtained data t = 2.349 > tSI =2.0; F=5.561 > FSI = 3.9
confirmed the validity of the difference between the two series compared, as
mentioned above (p<O.05). The results obtained serve as an additional
acknowledgement of the proposed concept that allows the prediction of the time
distribution of accident rates during the year.
5.0 Discussion
Despite the existence of persuasive proof of the influence of certain cosmic factors
on accidents, in many cases, the parameters of the solar-geophysical environment
that can be evaluated by modem measuring instruments cannot be identified as the
o liM
0.00
~=-----=.::..==; - -f!
I02.1~~I.
0.04 ,--
t ', -/ 1
Z
:J
:>:: 0.0]
L;.;
00- illmil
> (\J, if-At I 1\ ~I • p;une, crash tl Iin--I
o
e<: II
I~
Z
w
02.07. - Boe ing- 7H crush
S: - 0.1l2
/ : « ,
~
:Il /
/
,,
0
U
· 0.04 1--+ "
- ' 06.12. - plant crush (Japan)
" ' .. U
- 0.00 _ »: »: _
--
O g );.-e-: 0) 22. . 1051 nr St.ling or hydra ulic
SjI,'tm in Allanl is
-~'~--I./---~-~,-- ' -
- O.OR
- 02.24. - Columbia 10Sllh::~
1996
I 2 I J ~4 _1 5 -1..'_ " , -_6 7 8 9 10 ~I I 12
Figure I.
Var iations of the integral characte ristic of cosmic energy processes in interplanetary space (computer
modeli ng for 1996) and their conne ctions with maj or accidents and malfunctions. Dotted line - long term VJ
~
component of the integral chara cter istic. VJ
344
04.02. .• US Airforce plane crash (Colorado) 05.18. . malfunct ion in Zaporozhskaya NPS
04.05. - power cut in spaceship Columbia 05.19. . malfunction in Chernobyl NPS
CO 04.08. - emergency landing of Columbia 05.20. . unsuccessful launch ofZcnilh
~ 1 -.;:---'--~
<D
o
o
CO
o
oI
07.26. •• plane crash in Belgium
CO
o
o
1
2 8
1997
Figure 2a.
Variations of the integral characteristic of cosmic energy processes in interplanetary space (computer
modeling for the first halfof 1997) and their connections with majoraccidents and malfunctions. Dotted
line- long tenn component of the integral characteristic.
345
07.3 I. - plane crash (USA) main compuler crash (M ir)

08.01.·· oxygen generator malfunct ion (Min
08.01. - computer crash (Tok yo Stock
Exchange) . US and German A irf or cc planes
.>:
coll ision (Angola)
09.14. F·II? crash (Balt imore)
09.14. . ra ilway accident (Ind ia)
I
. power CUI in space stat ion Mir
09.15. .. Mir hardly avoided
co llisio n with satell ite
plane crash in Indone sia - unexpla ined vibra tion s o f
An Museum in Paris
0.08
11.02. - unsucccssfullaunch of
~ 0 .06 Brasilian satellite
27
::>
ffi 0 .04
c.
>- 0.02 3D
0
0: 11
uJ
Z 0 .00
22
uJ
~ -0.02 11.20. - unsucccssfullaunch of

~ satellite from Columbia
en
8 -0.04
08.25. - oxygen generator 3
malfunction (Mir)
-O~ to I.... . ,~, . . I

09.22. - main computer crash (Mir)
9 10
Iill ..... "
11 t2
..,...,..1,..........""
09.22. - railway accident (M ilan)
997
09.22. - tra ins collision ( Philippine)
Figure 2b.
Variations of the integral characteristic of cosmic energy processes in interplanetaryspace (computer
modeling for the second half of 1997)and their connections with major accidents and malfunctions.
Dotted line - long term component of the integral characteristic.
346
a ,so
'40
~
eli
a:: 'zo
w
a:l
--
Z
0
E-
1(
>
36
::E
::J
Z
I-
'00
'0
0
1( Q.. SO
-o
CIl
0 Z8
U til
S '"'
CIl
Z
1( ~O
a::
CIl Z4
CIl
::J
-
Z %0
CIl
~ 'S
W
-
0
U
U
'Z
1(
1986 1988 1990 1992 1994 1996
Figure3.
Correlation between solar activity (data fromZurich Royal Observatory) and aviation accidentsin USSR
and CIS (data from the newspaper Kievskiye Vedomosry), a - sunspot numbers, w; b- accidents
347
October 1989 - very high

rate of accidents in US
Air Force
08.19.80 12.21.88
05.25.79 I
I
200
~
vi
~ 160 09.0 1.83
~
lQ 03.03.74
03.27.77
~
;>
Z 120 11. 12.96
bll.< i
til
Z 80
;>
til
40
1970 1974 1978 1982 1986 1990 1994
Figure4.
Correlation between solar activity(data fromZurich Royal Observatory) and major Boeing accidents
(data from [2])
348
reasons that have caused them. In a substantial number of recorded accident cases,
neither on the eve, nor during them, were there magnetic storms or solar flares, or
other complex meteorological conditions that are considered the most powerful
influencing processes in the biosphere.
A number of theories were put forward to explain the interrelation between
cosmic and terrestrial events. One of the most popular opinions is the idea about the
electromagnetic interrelation. Data exist, however, that can not be explained from
this point of view [8-11]. It was shown that various effects exist like fluctuation in
radioactivity measurements and the study of variations of a vacuumed screened
mechanical quartz resonator that cannot be explained by fluctuations of external
electromagnetic fields.
The results of studies of fluctuation phenomena in physico-chemical
systems showed interrelations that cannot be explained by electromagnetic influence.
These studies confirm that, in reality, other cosmic physical influences exist having
non-electromagnetic and probably gravitational natures, the origin of which must be
associated with cosmic objects and the Sun above all. Our own studies demonstrated
that different bio-systems (humans and animals) react to some delicate disturbances
originating in interplanetary space that are not detectable by ordinary measuring
instruments, some days in advance. Based on the arguments mentioned above, other
approaches in forecasting of cosmic disturbances and their influence of terrestrial
processes should be developed and used.
The important advantage of the proposed concept lies in the fact that the
information technology developed allows the forecasting of energy-informational
changes in interplanetary space that are inaccessible by contemporary measuring
instruments and that precede solar flares and magnetic storms, and can be considered
as triggering mechanisms for accidents. The integral characteristic modeled of
energy process changes in interplanetary space reflects the set of changes of high-
energy particles of solar plasma and gravitational and electromagnetic fields. Using
the proposed concept, there appears to be a chance to substantially increase the
reliability of the complex human-machine system interaction by using timely
forecasting of the periods of increased accident rates. Furthermore, a prospective
strategy of preventing accidents in different spheres of human activity can be
suggested based on the following four rules:
I. Using the information technology developed of accident forecasting that covers:

• Monitoring of solar-geophysical factors,
• Monitoring of the state of complex technological and electronic systems
and their units (permanent controlling of physico-chemical and mechanical
properties of materials),
• Bio-monitoring,
• Mathematical modeling and forecasting of cosmic energy-informational
processes changes.
2. Developing of new materials with increased reliability, optimizing technological
processes by choosing special sites for the manufacturing facilities and
determining the best time for production during special states of solar-
geophysical activity.
349
3. Choosing the most optimal time (during the period of "quiet space" when the
probability of accidents is minimal) to launch space vehicles, and to test both
aircraft and complexsystems.
4. Creating a service network to warn about probable extreme situations in
different industries.
6.0 Summary
The research conducted proved that with the help of information technology one
could forecast periods and days with the strongest probability of accidents caused by
unfavorable influences of cosmic disturbances on technological systems and human-
operator up to one year in advance. In addition to forecasting disturbances on a
global scale, this information technology allows the forecast oflocal disturbances.
Consequently, the information technology developed of forecasting periods
and days with increased probability of accidents could be an effective instrument in
determining the optimum time for launching spacecraft, testing aircraft, developing
materials with new physico-chemical properties, optimizingand increasing reliability
of telecommunication network functioning, preventing technologically-generated
and ecological catastrophes in differentspheres of humanactivity.
7.0 References
I. Massamura, c., "Strong influence of solar activity on frequency of traffic accidents," Influence of
solar activity on biosphere, Nauka, Moscow , m 50-54, 1971 (in Russian) .
2, Vasylyk, P.V., Galytsky A.K., "The influence of solar-geophysical factors on human organism,
statistics of traffic accidents and problems of forecasting," Cybernetics and Computer Technology.
Medical Cybernetics , Allerton Press, N.Y., 90, 8-11,1992.
3. Epov, A.B.,"Time and space laws of technologically-generated catastrophes and their connection to
natural phenomena," Mathematical methods ofanalyzing cycle in geology 7. 7, 55-60, 1996.
4. Avdonina, E.N., Lukyanov, Y.V., "Solar-geophysical effects in measuring resultswith the methods
of liquid scintillation counting and statistics of nucleardisintegration," Biophysics 40, 2, 967-881,
1995 (in Russian) •
5. International Symposium on Cosmic Correlation in Biological and Physico-chemical processes:
Pushchino, September 26-0ctober I, 1993, Biophysics 40,4,1995 (in Russian).
6. Shnol, S.E., Palamarchuk, L.E., Mikonova, M.V., "Preliminary results of research of impactof space
irradiation of non-electromagnetic natureon physical and biological systems," Biophysics 40, 4, 965-
975, 1995 (in Russian).
7. Enders, J.H., Dodd, R., Tarrel, R., Khatwa, R., Roelen, A., Karwal, A., "Accident sample listing,"
Flight Safety Digest, March, 26-28, 1996-
8. Garib, F. Yu. Et aI., "The influence of tide changes of Gravityon periodicity of expressreceptors T-
lymphocytes in vivo," Biophysics 40, 4, 1995 (in Russian).
9. Rabshtein, et al. "The influence of gravity changes in regulation of system response of blood
circulation and breathing on orthostas," Biophysics 40, 4, 1995 (in Russian).
10. Klotchek, N.V., Palamarchyk, L.E., Nikonova, M.V.. "Preliminary results of research of impact of
space irradiation of non-electromagnetic natureon physical and biological systems," Biophysics 40,
4, 1995 (in Russian).
11. Piccardi, D., Chemicalfundamentals ofmedical climatology, Hydrometeoizdat, Leningrad, 1971 (in
Russian).
Author Index
A I
Acker, C. 211 Ikeda, J. 181
Iskanderova, l .A. 33,145,243
B Isupov, V.V. 85
Banks, Bruce A. 1,145,165,253
Beech, Martin 305 K
Breault,R. P. 229 Kinoshita, H. 73
Brown, Ian G. 145 Kitral, M. 165
Bungay, Corey L. 51 Kleiman, Jacob I. 33, 145,243
Kruchinenko, V.G. 295
C Kruzelecky, R.V. 125
Chambers, A.R. 115 Kurilovich, A.V. 291
Clark, A. J. 229
Coia, C. 281 L
Czeremuszkin, G. 281 Lifshitz, Y. 103
D M
de Groh, Kim K. 51 ,253 McCall,M.L. 229
Dever,J. A. 181 McCall, S.H. 229
Devries, M. J. 51 Marco, J. 211
Mendeleyev, V.Ya. 291
F Milintchouk, A. 197
Fozza, A. C. 281 Minton, T. K. 15
Frandsen, AX 15 Monteiro, O. R. 145
Morison, W.O. 243
G Mount, C.K. 103
Gabriel, S.B. 115
Galytsky, A. K. 339 N
Garton, DJ . 15 Nikanpour, D. 125
Ghosh, A.K. 125 Nikishin, E.F. 85
Grossman, E. 103 Nguyen, D. 137
Gudimenko, Yu. 33, 145,243 Norris, Mary Jo 1, 145
Guerard, F. 235
Guillaumon, J.C. 235 0
Gusakov, A.G. 331 Ohmae, N. 73
Gusynin, V.P. 339 Osborne, J.J. 115
H P
Hasegawa, M.M. 181 Paillous, A.. 211
Henderson, R.A. 229 Piotrowski,A. E. 229
Hoflund, G.B. 103 Piotrowski, L. A. 229
Houdayer, A. 281 Poire, Ethel 125, 137
351
352
R
Raspopov, S. A. 331
Reed,Ch.K. 181
Reid, M.A. 229
Reulet, R. 211
Roberts, G.T. 65
Rodney, J.W. 229
Rutledge, Sharon 165,181
S
Schmidt, L.R. 131
Scheiman, D.A. 253
Seale, J.W. IS
Sechkar, Edward 253
Sellars, B.G. 137
Shevchuk,B.M. 339
Sitalo, V. 95
Skovorodko, S.N. 291
Startsev, a.v. 85
Stegnyy, A.. I. 339
Stueber, T. J. 1
T
Tagawa, M. 73
Tennyson, R.C. 33, 145,243
Tikhiy, V. G. 95
Tiwald, T.E. 51
U
Umeno, M. 73
V
Van Eesbeek, Marc 197
Vasil'yev, V. N. 291
Vecher, A. A. 331
Viel, V. 211
Voloshchuk, Yu. I. 295
Voropayev, A.G. 331
W
Wertheimer, M.R. 281
Wolan,J .T. 103
Woollam, J. A. 51
Y
Yatskiv, Va. S. 295
Z
Zheludkevich, M. L. 331
Subject Index
A BRDF 229
Aclar 41
adherence 238, 240,282 C
aluminum 85, 145, 170, 186, 211, 238 Carbon fiber reinforced plastic
aluminum foil 85, 167,258 (CFRP) composites 34,85
aluminum alloy(s) 85,95,236 charged particle
atomic force microscopy (ir)radiation 256,257,291
(AFM) 105,137 cleanliness 95
atomic oxygen (AO) clean room 97
-attack 1,15, 73 coatings 1,15,181 ,235
-average flux 115 (also see protecti ve coatings)
-angular distribution 15 collis ion (probability) 295
-bombardment 236,238 Columbus Orbital Facility
-durability evaluation 33 (COF) 211
-degradation of materials 15,33 composite materials 42,85
-effective fluence 2, 61, 150, 68 (see also CFRP composites)
-erosion 2, 15, 33,251 , 81 combined AO-VUV exposure
-predictive model 33 facility 182
-erosion yield 33, 165, 170,244 conversion 65
-exposure to 73, 81, 103, 166 contaminants 181
-laboratory simulation by contamination 95
-CASOAR facility 216,236 cosmic factors 339
-afterglow RF plasma 133,139 crack (s) 2,87,254,269,273,286
-ECR plasma source 183 chromic acid anodization 21 1
-ESD source 104
-plasma asher 60, 145, 166,171 D
-pulsed laser source 17,74,216 damage 253,281
-oxygen plasma 35,60,145 darkening 182,244
-UTIAS facility 41,150,246 database 229
-oxidative degradation by 1 dehydration 166
-predict ive model of eros ion 33 defect (s) 2
-react ion probability 2 degradation 15,181
-reaction products 20,116 density 169
-recombination 4 diffusion coefficient 88
-resistance to 33, 126 dimethylsiloxane 137
-scattering 3, 23 durability
-sensor(s) 115 -evaluation 33,253
-VUV resistance 51,126,142,182 -of polymers 33,253
-undercutting 1
accidents 339 E
Auger spectroscopy 74 ECR-CVD 125
ECR plasma processing 126
B ECR (plasma) source 182
Beginning-Of-Life (BOL) 212 electron
binding energies 105 -beam 259,291
"bleaching effect" 182 -flux 291
boron nitride 29 -radiation 257
353
354
electron/proton radiation 256 hydrocarbon surface 15
- exposure 261,262 high temperature 236,239
erosion effects 33 high-temperature black paint 235
erosion rate 244 hyperthermal
erosion yield 33, 165, 170,244 -atomic oxygen (AO) 74,282
Equivalent Sun Hours(ESH)184,212,240 -AObeam 18
End-Of-Life (EOL) 211,212,220 -beam 20
-OA generator 104
F -molecular oxygen(MO) 74
fast atomicoxygen
(FAO) 33,145,243,246 I-J
fabric 100 impactprobability 305
fairings 95 Implantox technology 145
Faraday cage(s) 165,167,168 indexof refraction 51
FEP/Tetlon 51,165,197,253 information technology 329
(also see Teflon) inorganic protective coatings 137
-aluminized (AI-FEP) 68, 253 International Space Station 52, 181,211
-silvered(AglFEP) 191,193,256 ion(s) 146, 165
fiber(s) 43,97 ion implantation 145, 146
fiberglass 100,246 ion-irradiated 110
tlexibility 233,282,240 ionization 127
tluorinated polymers 1,33 ionizing radiation 282
fluorine 36,37
friction coefficient 73 K
friction force trace(s) 79,80,81 Kapton
-AFM images 112
G -bindingenergies 108,109
glass -degradation 103
-fiberreinforced plastics 85,86 -effective tluence 2,61,150,168
-netting 85 -erosion yield in LEO 169
-fiber cloth 85 -polyimide 36,236,281
glue 100 -relative erosion yield 176
graphite 48 -structure 106
ground-based facility 2 -surface modification 108,147
ground laboratory facility 2 -surface morphology 106, 112, 160
(see also AO laboratory simulation) -thermal blankets 104
-thermal-optical properties 240
H -thermal stability 236
Halar 41 -XPS spectra 107
heat shield 235 - high resolution 108,109
honeycomb system(s) 86 KaptonCB 147
Hubble SpaceTelescope(HSP) 52,253 Kapton HN 147,166
-thermal control replacement KaptonH 7,58,282
materials 253 Kharkiv MeteorData Base 295
hydrocarbon polymers 1,33 kinetic of oxidation 332
hydrophobic 247
hydrophilic 247
355
L -characteristics 243,291
lacquer 85,87 -constants 53,61,62
large scale ECR facility 126 -density 52, 68
LEO environment 52,108,137 -microscopy 264
Leonid meteoroid 307 -properties 53, 181,263
Long Duration Exposure Facility on-orbit 254
(LDEF) I, 104, 211 outgassing 95,181,211,217
linear oxidation 333 outgassing/contamination tests 239
low Earth orbit oxide(s)
(LEO) 1,15,I03, 115,137, 181 ,235,281 -Iayer(s) 51,65,217
low temperature 240 -thickness 65,66,217
lubricant(s) 73 -forming polymers 51, 71
Lyrids 317 oxygen atom(s) 15, 108
oxygenplasma(s) 67, 145, 165
M
mass loss 2,5,41,168,216,238,243 p
mechanical properties 91, 240 paints
metals (VB group) 331 -aluminum 235
meteor -thermal control 235,243
- orbit 298 -black 235,243
- showers 298,299,300 -white 235,243
meteoroid and debris protection shield -Z-93P 181
(MDPS) 211 particle radiation 235
meteoroid hazard 295 PECVD 281
meteoroid stream(s) 305 PEEK 47
microhardness 85,89 Photosil technology 243
(micro)meteoroid protection 86,211 Pigment(s) 182,236, 244
-panels 92 plasma asher 60, 145, 157,165,182
MIR Orbital Station 85, 86, 90, 92 plasma chamber 167
microsatellite 115, 121,235 plasma immersion ion
MicroScratch testing 284 implantation (PIlI) 145
modelling I polymer-based composites 33
molecular oxygen 18, 74, 331,333 polymer composite materials (PCM) 85
molibdenum disulfide (MoS2) polymer materials
-single crystals 73 -Aclar 41
-sputtered films 73 -CV-1144-0 silicone 51
Monte-Carlo computational model I -epoxy resin 5208 47
multi-layer insulational -Kapton 15,36, 51, 103, 150, 236
(MLI) blankets (HST) 254 -Kynar 38
MLI Failure Review Board (FRB) 255 -Mylar 151,281
Mylar 151,281 -PAEBI 51
-PMR-15 47
N-O -polycarbonate 125
niobium 334,335 -polyethylene (PE) 38
a-atom 20,37,103,107 -polyphenylsiloxane (binder) 236
optical -Tedlar 38
-emission spectrometer 126 -Tefzel 38,181
356
-Victrex (polyethersulfone) 47 Residual Gas Analysis (RGA) 74,78

-Udel (polysulfone) 47 Resistance (to AO and VUV) 125,182
polymer materials rivetedconstruction(s) 95
-durability in space 1,33 resins
predictive model 33 -pheno1fonna1dehyde 100
porosity 52 -epoxy 100
products roughness 238
-reactive 15 rubber 100
-non-reactive 15
-outgassing 181 S
-volatile 35 "Salut" Orbital Station 85
protective coatings -Design Bureau 86
-enamel 291 Scanning Electron Microscopy
-defects I (SEM) 42, 145,243,250,257
-durability 2 surface morphology 42,45, 160,243
-non-silicone, inorganic 137 scratch testes) 284
-polysiloxane 239 scrim
-sro, 142,181 -fiberglass 256
-Si0 2 215,281 -Nomex 256
-Si0 2lInterfacelKapton 134 Second-Surface Mirrors (SSM) 238
-Si0 2lInterfacelTeflon 129,132 sensors 115
protective coatings by semiconductor oxide(s) sensor(s) 117
-PECVD 281 shielding 85
-Si0 2 by ECR-CVD 125 silica (Si0 2) 125,281
profilometry 140 silicone 51
protonfluxes 291 silver 115
proton (ir)radiation 257,286 silylation technology 245
silveractinometer(s) 116
Q simulated LEOenvironment 182
Quartzcrystal microbalance soft X-ray 197
(QCM) 75, 116, 180,216 solar
QCMsensors 116 -absorptance 133, 150, 182,211 , 236, 243
quartzplates 238 -array 115
-flares 253
R -x-rayts) 258
radiator(s) 254 SOLEXISTM database 229
reaction products 20 space (in space) 145,165,254
reliability 339 -applications 73,197,229
rotatable mass-spectrometer (detector) 15 -durability 2
reactive products 16 -exposure 86
non(reactive) products 16 -factors 85
reflectance 185,215 -mass loss 2
-spectral 217 -mission 2,73
radio-frequency (RF) -testing 2
plasma 1,108,138,166 spacecraft materials 1,125,165
reinforced plastics 85 spectral reflectance 229
relative erosion yield 165, 176
357
surfaces UV 33,51,21 1,235

-black 229 -absorption band 58
-reflective 229 -irradiation 16,58,216,211
-transmissive 229 -lamp 54
-white 229 -exposure 58
surface conversion 51,145 -induced modification 59
surface modification of -radiation 82,211
-Kapton CB (black) 147 vanadium 336
-Kapton HN 147 varnish 235
-Mylar 147 VASE (variable angle spectroscopic
-CV-1144-0 51 ellipsometry) 54
-PAEBI 51 vibration cleaning 96
surface modification by void(s) 65,269
-ion implantation 145 volatileproducts (species) 1,16,77,245
-Photosil technology 243 VUV
-AOand UV 51 -action 197
-attenuation length 204
T -degradation 184
tantalum 331 -durability 186
telescope 253 -etching process 205
tear resistance 253 -exposure 184,187
Teflon 35,51,165,I81 ,197 -Iamp(s) 184,187,188,203
tensilestrength 263 -groundsimulation testing 203
Tefzel 38 -radiation 205, 206
Tefzeloutgassing products 181 -sources 203, 208
tensileproperties 263 -test(s) 214
tensile strength 263 -combined with AO 51,I82,I26,142
thermal controlmaterials 181,235,254
-blankets 115,254 W
-coatings 181,235 water absorption 166
-paint(s) 236,243 water contactangle 159,246
thermal cycling 253,259,264,268 wavelength 51,129,168,217
thermal emittance 133,150,211 ,236,245 witness sample 116,181,258,263,281
threads 100
time-of-flight (TOF) X-y-Z
-distributions 19 X-rayphotoelectron spectroscopy
-spectra 74, 75 (XPS) 73,103,142,150,158,246
total mass loss (TML) 98,238 -data 108
transmission 239 -spectrum 77, 81
transparency 240 -spectra 77,108,109
transmittance 161 -survey spectra 107
tribological X-ray radiation 258
-properties 78 -flux 258
-testing 76 -fluence 258,261
TRIM(computer) simulation 151,283 -spectrum 258
zinc oxide sensors 123
v-v
ultimate tensilestrength 263
UHV-friction tester 76
Space Technology Proceedings
1. J.C. van der Ha (ed.): Mission Design & Implementation ofSatellite Constellations.
1998 ISBN: 0-7923-5210-6
2. J.1. Kleiman and R.C. Tennyson (eds.): Protection ofMaterials and Structuresfrom
the Low Earth Orbit Space Environment. 1999 ISBN: 0-7923-5540-7
3. J-J. Miau and R. Holdaway (eds.): Reducing the Cost ofSpacecraftGround.2000
ISBN: 0-7923-6174-1
4. 1.1. Kleiman and R.C. Tennyson (eds.): Protection ofSpaceMaterialsfrom the Space
Environment. Proceedings of ICPMSE-4, Fourth International Space Conference,
held in Toronto, Canada, April 23-24, 1998.2001 ISBN 0-7923-6981-5
Forfurther information on this series see www.wkap.nllseries.htrnlSPTP
KLUWER ACADEMIC PUBLISHERS - DORDRECHT / BOSTON / LONDON

Protection of Space Materials From The Space Environment

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Protection of Space Materials From The Space Environment

Uploaded by

Copyright:

Available Formats

PROTECTION OF SPACE MATERIALS

FROM THE SPACE ENVIRONMENT

ISBN 978-94-010-3838-6 ISBN 978-94-010-0714-6 (eBook)

Printed on acid-free paper

Section A: AONUV Effects, Evaluation and Modeling

1. Monte Carlo Computational Modeling of the Energy

2. Dynamics of Atomic-Oxygen-Induced Degradation of Materials 15

3. Generalization of the Predictive Model of Erosion of Carbon-Based 33

4. Spectroscopic Ellipsometry Characterization of Polymers Modified 51

5. Effect of Atomic Oxygen Exposures on the Tribological Properties 73

6. Influence of Shielding Upon the Structure and Properties 85

7. Cleanliness Support of Vehicle Fairings for Spacecraft 95

8. In situ XPS Studies of Kapton Exposed to 5 eV Atomic Oxygen 103

Section B: Instrumentation, Sensors, New Technologies

10. Large-Scale ECR-CVD Preparation ofIntegrated Thin-Film Structures 125

11. Non-Silicone Inorganic Polymer-Based Coatings for Atomic Oxygen 137

12. Research Aspects of Scaling Up the Implantox Technology 145

Section C: Thermal Control in LEO

13. A Comparison of Space and Ground-Based Facility 165

14. Evaluation of Low Earth Orbit Environmental Effects on International 181

15. Some Aspects of the Degradation ofFEP Teflon Material 197

16. Evaluation ofCOF MDPS Chromic Acid Anodization 211

Section D: Charged Particles/Soft X-Rays/Meteoroidsffhermal Cycling

20. Simulated Solar FlareX-Rayand Thermal Cycling Durability 253

21. Radiation-Induced Effects in Si0 2 Protective Coatings on 281

22. Influence of Electron and ProtonFluxes on Optical Properties 291

23. Meteoroid Hazard in the LEO SpaceEnvironment 295

24. Characterizing the Annual Meteoroid Streams 305

25. Oxidation Kinetics of VB Group Metals by Atomic and 331

Author Index 351

• The Canadian Space Agency; the Space Station Program Office

Jacob Kleiman, Integrity Testing Laboratory Inc.

Th is publication presents the proceedings of ICPMSE-4, the fourth

These proceedings are organized into four sections:

a) AOIVUV and Radiation Effects, Evaluation and Modelling;

Jacob Kleiman, Chairman/Organizing Committee/ICPMSE-4

Session A: AO\VUV Effects, Evaluation and Modelling

Session B: Instrumentation, Sensors, New Technologies

Session C: Thermal Control in LEO

Session D: Charged Particles/Soft X-rayslMeteoroids/

Ll. Kleiman and R.C. Tennyson (eds.),

Table I. Computational modelparametersand reference valuesfor LEO atomicoxygeninteraction with Kapton.

Initial impact reaction probability 0.11

Where: N = Number of Monte Carlo atoms needed to simulatean effective

M = Number of atoms for wide defect (atoms)

This calibration technique is shown schematically in Figure 1. Because the

3.0 Results And Discussion

Probability of Reaction Upon Initial Impact = k exp -(EJE» [2]

The consequence of this dependence of probability of reaction on initial

Figure I. Monte Carlo computational model calibration.

Figure 2. Monte Carlo computation model predictedundercut cavity for a fluence of

TIMOTHY K. MINTON, JAMES W. SEALE, DONNA J. GARTON,

We have investigated the mechanisms that lead to degradation of hydrocarbon-based

3.1 5.3 6.5 "10.8 eV

Figure 2. Translational energy distributions of atomicand molecularoxygen components in the

3. Initial Reactions with a Liquid Hydrocarbon Surface

Theinitialreactionsbetweenthe beamand a hydrocarbon surfacewere probedwiththe use

mlz =17 (OH+)

o 400 800 1200 0 400 800 1200

Figure3. Representative time-of-flight distributions ofOH products following reactionofoxygen

<Ef > = 1.8 eV

o 400 800 1200 0 400 800 1200