Professional Documents
Culture Documents
Protection of Space Materials From The Space Environment
Protection of Space Materials From The Space Environment
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.
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
9. A Low Cost, Lightweight Reusable Sensor for Atomic Oxygen Detection 115
1.1. Osborne, G. T. Roberts, S.B. Gabrieland A.R. Chambers
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
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
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;
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.
ix
Introduction
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 .
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
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
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
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]
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:
Where k = 0.1165
Eo= activation energy = O.26eV
E = energy of atomic oxygen impact, eV
6
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_ . .-
111
H
500j.lm
4.5 eV
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
Abstract
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.
© 2001 Kluwer Academic Publishers.
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
O~I,
'\
I, t
\
\
II
t \
II I \
,I
0
0 5 10 15 20 25
Energy, eV
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.
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
Flight time, us
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
0.6
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
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.
@\ 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
Kapton Squalane
<Ei > = 6 eV =
<E? 6 eV
700
~
Olo.6w----------l
1400
c m/z =28 (CO· D m/z= 2
700
01W'~----------.;-l
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
...e
~
U)
400
::::s 300
0
--....
z
CJ
200
100
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
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
2. Z. A. Iskanderova, J. I. Kleiman, Yu. Gudimenko, and R. C. Tennyson, J. Spacecraft and Rockets 32,5
(1995).
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).
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
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
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.
(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
"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
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.
s;LEO -M-'y'(l-o)
I
p (2)
(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 .
4.0 Conclusions
(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
~
~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
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*
KIM K. DE GROH
Nation Aeronautics and Space Administration
Lewis Research Center
Cleveland, Ohio '
Abstract
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
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
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
2.3.1. 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
.......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
Figure 5.
Index of refraction depth profile showing the depth and amount ofUV induced modification into PAEBI film
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
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
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].
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:
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) .
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
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 (%)
§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
-'Q) 2.0
- - Oxide After 15 min. AO Exposure, n
I-
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
only the phosphorous oxygen double bond (P=O). Thus, there is likely a large amount
of void surrounding the phosphorous oxygen molecule that remains.
4.1 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.
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
-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
-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 .
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
References
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
J./. Kleiman and R.C. Tennyson (eds.),
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.
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
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.
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.
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.
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 • • , ,
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
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
2.0
.- 1.5
-z~
o
u,
0.5
Friction
0,0
-0.5 L...-L-........................L...-...........--'---'-....L-""""----'--...........--L--'--'---'--.........,
2.5
2.0
-
Z
~ 1.5
o
LL 0.0
-0.1
-0.2
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' •• • • • ••
*
IJ..
-. 3
~
...-:..:..~~. : ~~~~ .....:..-:-.:.;~.:- ::: .... ~...: . ~. :-;i ····..~. -; .~ ~-7:-:··- ·
o • • • • • • • • ••• • • • •• • • • • • • • 0' • " • •• •
2
...-.
Z
.......
8 Load
~
0
IJ.. 0 Friction
-0.25 L-.I...-'--'..............L....I..--'--"--'-....L-L-.l...-'--'---I--'--'--'--.......L......I...-.............~
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
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
J./. Kleiman and R.C. Tennyson (eds.),
Protection ofSpace Materials from the Space Environment, 85--93.
© 2001 Kluwer Academic Publishers. .
86
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.
2.5
C 2.0
.su
....
<..::
<I
0 1.5
U
..
0
' ;;
1.0
~ 0.5
Q
0.0
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
. •
..-
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
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
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
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
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.),
Protection ofSpace Materials from the Space Environment, 95-101.
© 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 2. Summary of Data on Residual Quantity of Particulate Contamin ation on the Surface Using filter
paper "Vladipor"
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
Materials
Before TVT After TVT
Figure 1
A unit on vibration bench.
IN SITU XPS STUDIES OF KAPTON EXPOSED TO 5 EV ATOMIC
OXYGEN
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
J./. Kleiman and R.C. Tennyson (eds.),
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.
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
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
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
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.
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.
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
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
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
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
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.
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.
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.
1.40E+05 t:--------MlVrJ~,,---------f
: 314
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
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
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.
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
Figure 3
The results of an AO exposure on a 0.5 urn zinc oxide film
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
Abstract
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.
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
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.
~
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
(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]
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
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.
(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
(a) (b)
.. ..• ~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.
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
Acknowledgements:
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
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
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.
MW 0.2 30 200 0 0 I
RF 0.2 10 0 0 300 1
(SSA)
140
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.
RMS = I
/
2
P(j}ij
]X (2)
[
where
(3)
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
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.
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
uncoated 64 81
1 67 1271
2 1,9 1,2
4 87 55
5 130 63
5.0 Acknowledgements
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
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
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.),
Protection ofSpace Materials from the Space Environment, 145-163.
© 2001 Kluwer Academic Publishers.
146
as solar absorptance and thermal emittance, have been left practically unchanged, or
altered in a specific manner.
2.0 Introduction
3.0 Experimental
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
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
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
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.
"6
.".
,,'" ..
0
5 0.004 0
"
:c
iii
c:
Cll
0.003 •~ A. AA 0
" 00
"..-
0 M "00 .......
0.002 ~ 0 <II •
••
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 .
soo
Kaplon NH
400
Q)
c: AI
c: 300
m
s:
~
§ 200
0
0
100 (a)
Q)
c: AI
c: 300
ca
s:
~c: Oxygen
::J 200
0
0
(b)
100
'i)
c: AI
c
II 300
s:
~
!!
c:
::J 200
0
0 (c)
. 100
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
~
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
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")
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).
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
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
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
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
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
MICHAEL KITRAL
Cleveland State University
Euclid Ave. at East 24th Street
Cleveland. Ohio 44115 ; U.S.A.
Abstract
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.
2.1. 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].
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].
Figure IA.
Closed Faraday cage
168
Al Foil Holder
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
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
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
415 In plasma Kapton 42.1 4.366E15 ± 8.66E13 1.0
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
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
Cage with MgF2 window)
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
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].
z
~~ 0.2
g
~
~ 0. 15
~
a:
E
~ 0.1
Q
-l
101
>
Z
o 005
~
101
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.
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
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
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
NASA Lewis Research Center
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.
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.
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
exposure of samples without AO, and approximately 4xlO·4 torr during combined
VUV and AO exposure.
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.
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
Measured lamp intensity data and ESH for each of the exposure tests will be
given in sections 3.1 and 3.2.
o.s=l-ps (2)
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.
2.4.1 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
2.5.1 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.
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.
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
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
0.24
0.22
--
I
Sl
0.2
0.18
~
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
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.
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
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 ,
Table 10. In-Air Measurements : Effect of AO and VUV on Solar Absorptance ofSiOx·coated
Silvered Teflon. uncoated Silvered Teflon and Z-93P
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
J./. Kleiman and R.C. Tennyson (eds.),
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
.rrr--:
r-.
1000
W .m~.pm-1
I
'00
10
r <,
-,
/ -, -,
/
0 .1
~/ -,
0.001
0.' '0 ""A 100
Figure Ib
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].
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.
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.
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.
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
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
Different facilities use different sources with different energy ranges or photon
energy, and with different fluences .
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.
(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.
0.1
\
0.01
~.. ........
........................ ..
.
.........................
0.001
"
.,., ......, .., ..
00001
50 100 150 200
l(nm)
Figure 4.
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" .
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.
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.
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.
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 .
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.
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.
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
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.
© 2001 KluwerAcademic Publishers.
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.
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
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.
Simulation tests were all carried out in specific facilities existing in the DESP
laboratories.
2500 f2500
Rs = R(A) · 1(1..) · dA/ 1(1..) · tIJ... (1)
f 250 250
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
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)
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
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
- - 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)
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)
'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.
(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)
Ultraviolets (ESH)
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:
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)
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)
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)
100
90
80
~ 70
B 60
n 40
c
III
50
• 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
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
4.3 PLASMA
•
Box
Pearson
probe
i(l)2NV
eAA Ionospheric
sample Plasma
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
6.0 Acknowledgements
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
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
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.
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
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
5.0 Acknowledgements
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
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.
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.
© 2001 Kluwer Academic Publishers.
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.
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
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
Table3
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.
a s= 0.10
- Infrared emissivity factor
EN = 0.80 to 0.85
- Optical transmission between 200 and 1000 nm: above 98%
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
Figure2
RSF: Solar Absorptance Versus UV Irradiation Duration
eN = 0.80
Post-test: adherence: vel)' good
a s = 0.38
eN = 0.80
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
Figure 3
Solar Absorptance Of Kapton Protect ed Versus UV Radiation Duration
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
Abstract
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°.
Time, min 0 5 10 15 20
Contact angle, 78 60 49 13 12
deg
247
Table 3. Summary of thermal opticalproperties for Aeroglaze 2306 black paintsample beforeand afterthe
PHOTOSIL™ process
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
Time, min 0 5 10 15 20 25
Contact 92 80 59 36 10 0 112 98
angle, dea
Table 5. Summary ofresults of surface composition (at%) of Aerogalze A276 white paint samples as
obtained by XPS
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
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
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
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.),
Protection ofSpace Materialsfrom the Space Environment. 253-279.
© 2001 KluwerAcademicPublishers.
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.
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
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.
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].
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 .
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.
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
·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
Figure 3. Drawing of the electron beam system configured to expose the HST candidate samples to
M,) x-ruys.
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.
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.
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
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
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
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
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
,- ... -
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).
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.
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
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
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.
© 2001 Kluwer Academic Publishers.
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.
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).
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 )
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
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
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.
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.
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.
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.
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
J 7~r-------------------.
Silica coatedPET
JUU
IA: PET
..."""e
..,
o"
o-l
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
6.0 References
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
J./. Kleiman and R.C. Tennyson (eds.),
Protection ofSpace Materialsfrom the Space Environment. 291-294.
© 2001 Kluwer Academic Publishers.
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.
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
Figure 1
Dependen ce of the polarization rate on influence volume
- - no influence
- - - -- - weak influence
.... .. _ . 9 ... .__ - _ . : . : _ , ~ _" _ ~~.
-.
30
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.
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):
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
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.
Log N (cm'f-sec")
·10
-25
-1 5 -10 -5 o 5 10 15 20 25
Log M (g)
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
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).
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)
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.
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
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
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
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
J./. Kleiman and R.C. Tennyson (eds.),
Protection ofSpace Materialsfrom the Space Environment, 305-329.
© 2001 Kluwer Academic Publishers.
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.
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)
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 ,
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
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.
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.
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).
o Cetids 59 28 -4 35 * 10 (80)
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
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 :
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
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
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
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).
4.0 ImpactProbabilities
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,
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"'.
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).
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).
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 .
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
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/)
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
0.986
,2
S' I nodal distance (1965)
~
Q)
c: 0.985
0
co
Ui
'5
0.984
0.982
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
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)
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
.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
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
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.
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
J./. Kleiman and R.C. Tennyson (eds.),
Protection ofSpace Materials from the Space Environment. 331-337.
© 2001 Kluwer Academic Publishers.
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.
3.1 TANTALUM
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·\
(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).
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
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.
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"
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
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
1.0 Introduction
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
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
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
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
.>:
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
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(
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
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
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:
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
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