Professional Documents
Culture Documents
The Use of Tribocharging in The Electrostatic Beneficiation of Lunar Simulant
The Use of Tribocharging in The Electrostatic Beneficiation of Lunar Simulant
The Use of Tribocharging in The Electrostatic Beneficiation of Lunar Simulant
3, MAY/JUNE 2009
Abstract—The use of tribocharging as a potential method to and it may significantly reduce the process’ complexity. In
provide sufficient charge to several different lunar simulants for addition, shipping oxygen from Earth is weight prohibitive,
electrostatic beneficiation was investigated. The objective was to and therefore, investigations into the potential production of
determine whether specific minerals of interest (e.g., ilmenite) that
are present in lunar regolith could be enriched in concentration by breathable oxygen from oxidized mineral components are a
beneficiation that would therefore allow for more efficient extrac- major research initiative by the National Aeronautics and Space
tion for in situ resource utilization use. The production of oxygen, Administration (NASA). The feasibility of extracting oxygen
water, and other resources on the Moon from raw materials is from the FeO component by hydrogen reduction in terrestrial
vital for future missions to the Moon. Successful separation of soils has been demonstrated [2], and a number of oxygen
ilmenite was achieved for a prepared simulant (KSC-1), which is a
mixture of pure commercially supplied pyroxene, olivine, feldspar, production technologies are being tested [3].
and ilmenite, in a 4 : 4 : 1 : 1 ratio, showing proof of concept when Ilmenite is a mineral of interest on the Moon as a source
tribocharged against three different charging materials, namely, of iron, titanium, and oxygen (FeTiO3 ) [2], and therefore,
Al, Cu, and PTFE. Separation by chemical composition was also any method to increase its natural abundance in excavated ore
observed for existing lunar simulants JSC-1 and JSC-1A; however, before processing would be a considerable advantage.
the interpretation of the separation was difficult due to the com-
plex mineralogy of the simulants compared to the simple prepared The lunar regolith grains and the lunar environment are ideal
mixture. for triboelectrification and electrostatic separation, with the
lack of moisture preventing the grains from sticking together,
Index Terms—Beneficiation, lunar simulant, tribocharging,
work function, X-ray photoelectron spectroscopy (XPS). the lack of atmospheric gaseous breakdown allowing higher
voltages to be used on the separation plates, and the gravity only
I. I NTRODUCTION being one-sixth that of Earth that will allow the particles’ longer
charge contact times and time between the plates to enhance the
Two production lunar simulants from the NASA Johnson sizes. The XPS data were obtained on a Kratos XSAM 800
Space Center were used in this paper, namely, JSC-1 and the spectrometer at a background pressure of 1 × 10−9 torr, using
newer JSC-1A. These simulants were developed to replicate a Mg Kα (hυ = 1253.6 eV) X-ray source. The X-ray source
as close as possible the chemistry of lunar soil brought back used was run at 150 W power with a beam diameter of 4 mm.
from the Apollo missions. They are principally composed of The collected spectra peak positions were referenced to the C1s
basalts, containing phases of plagioclase, pyroxene, olivine, and peak at 285 ± 0.5 eV. Wide survey scans were collected from
ilmenite. In addition, a simulant of exact known composition 0–1100 eV at a pass energy of 80 eV in 1-eV steps with a 50-ms
was developed in house at KSC, named KSC-1, consisting dwell time to determine the overall elemental composition. The
of a mixture of pure commercially supplied minerals, namely, relative atomic concentrations of the detected elements were
pyroxene (LiAlSi2 O6 ), olivine ((Mg, Fe)2 SiO4 ), feldspar calculated and normalized to 100% using sensitivity factors
((Na,K,Ca)AlSiO3 O8 : SiO2 ), and ilmenite (FeTiO3 ), in a supplied by the instrument manufacturer from known certified
4 : 4 : 1 : 1 ratio to reproduce the approximate mean composition standards. Detection limits for XPS are approximately 0.1–
of lunar dust. This simulant was used to test the efficiency of the 1.0 at%, depending upon the sensitivity of the elements. The
beneficiation apparatus. escape depth of the ejected photoelectrons is on the order of
The actual lunar regolith is a powdery dust with a grain 50–100 Å, and hence, XPS is a very surface-sensitive tech-
size ranging from 45 to 800 μm, with most falling between nique. It must be emphasized that the area of analysis is
45 and 100 μm [9], [10]. It has low electrical conductivity and ∼16 mm2 , and hence, the information obtained is an average
dielectric losses, permitting the accumulation of electrostatic over this area. SEM images were taken on a Hitachi S3500
charge [9]. Previous studies performed on beneficiation of microscope using a beam energy of 20 kV at a magnification
pulverized coal by Trigwell [7] had shown that the efficiency of of ×100. Samples of each grain size fraction were mounted
beneficiation using tribocharging was a factor of particle size, using conductive carbon paint and then gold coated to prevent
with low efficiency for the smaller particle sizes (< 50 μm) due charging up under the electron beam.
to the larger charge-to-mass (Q/M ) ratio obtained by smaller Due to the known strong adhesion of fine powder to surfaces,
particles. This caused problems with the dust strongly adhering preliminary tests were performed to determine the size fraction
to all surfaces, interfering with the contact charging process. to best use in the beneficiation experiments. An aliquot of each
As part of this paper, the simulants were sieved into a range of size fraction was placed on a flat sheet of each of Al, Cu,
particle sizes to determine the most efficient size range for this and PTFE, and the sheet was inclined until the majority of
method. the dust aliquot slid off. The particle size range and incline
Three different materials were investigated for the triboelec- angle at which the sliding occurred for each sheet material
trification process, i.e., aluminum, copper, and polytetrafluo- were recorded. It was observed that particles > 50 μm at an
roethylene (PTFE). These materials were selected because they angle > 50◦ provided the best results for all materials, showing
offer a wide variation in work functions (aluminum—4.28 eV, a reproducible > 95% clearing. The worse case of sticking
copper—4.65 eV, and PTFE—5.75 eV) [7]. The difference was observed for the < 25-μm-size range on PTFE where the
between the work function of each material and the simulant in- entire aliquot of simulant stayed adhered to the sheet, even
fluences the charge obtained by the grains. The resulting Q/M when it was raised up to 90◦ . Based on these data, for all
ratio was expected to lead to variations in mineral separation subsequent experiments, the 50–75-μm-size range was chosen
and allow for optimization of the process. for use. Feldspar, spodumene, olivine, and ilmenite were sieved,
The first objective of this paper was to determine the vari- and the 50–75-μm-size ranges were taken and mixed in the
ation of surface chemical composition as a factor of particle 4 : 4 : 1 : 1 ratio to make KSC-1. XPS was then performed on the
size because this would affect the beneficiation efficiency. The mixed KSC-1 to determine surface composition to compare it to
second objective was to determine the amount and polarity JSC-1 and JSC-1A.
of the transferred charge imparted by the different materials To provide tribocharging under vacuum, cups were machined
to the simulant by measuring the Q/M first in air to show from PTFE, Cu, and Al. Each cup had a hole in the bottom
proof of concept and test the system, and then in vacuum. The and a fine screen (120 mesh) to support the simulant made
third part of this paper was to perform the beneficiation on of the same material. The Cu and Al cups were resistively
KSC-1, JSC-1, and JSC-1A using the three different material heated using a nichrome wire wrapped around the cup capable
chargers. of heating the cup and contents to 130 ◦ C to remove moisture
from the simulant. Unfortunately, it was not possible to heat the
simulant in the PTFE cup. The heater wire was isolated from
II. E XPERIMENTAL S ETUP
the cup by Kapton tape. A small motor attached to the cup,
The JSC-1 and JSC-1 A simulants were sieved into different isolated by PTFE and a ceramic spacer to prevent the motor
grain size fractions, namely, > 100, 75–100, 50–75, 50–25, and from heating up, vibrated the holding cup, causing the simulant
< 25 μm (only for JSC-1), and each fraction was analyzed us- to fall through the mesh. The cup setup is shown schematically
ing X-ray photoelectron spectroscopy (XPS) to determine min- in Fig. 1. The cup was mounted on top of a solid block of
eral surface composition. Scanning electron microscopy (SEM) material (either PTFE, copper, or aluminum), the composition
was performed to determine any variation in size distribution of which is the same as the cup, in which a channel that is
within each size fraction. The objective was to determine if composed of a “zigzag” series of inclines greater than 50◦ has
there was a compositional correlation between different particle been cut out (Fig. 2). The voltage to the vibrating motor can
Authorized licensed use limited to: University of Newcastle. Downloaded on November 06,2023 at 03:05:03 UTC from IEEE Xplore. Restrictions apply.
1062 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 45, NO. 3, MAY/JUNE 2009
wall of the Faraday pail with the outer sheath grounded. It was
found that the measured charge on the electrometer steadied
after about 2 s. The inside container of the Faraday pail could
be removed, and the accumulated mass of the collected pow-
der was weighed relative to the preweighed empty container.
The Q/M measurements were first taken in air to test the
equipment and provide a comparison to results on dry dust in
vacuum. Five runs were taken for each sample, and the mean
value was plotted.
The charge separator was constructed of two parallel copper
plates of dimensions 6 × 12 separated by 3 in. The plates
were run at +15 and −15 kV supplied by positive and negative
30-kV Glassman MJ series 15-W regulated dc modules. The
theory and design of this separator have been described else-
Fig. 1. Schematic of the shaker cup.
where [6], [7]. The dust accumulated on each plate and on the
filter paper in the collection box at the bottom of the plates was
brushed off and collected in separate vials and then analyzed
by XPS. Again, five runes were taken for each material, and the
mean value was recorded.
Fig. 3. SEM images of the sieved fractions of the JSC-1 lunar simulant. (a) < 25 μm. (b) 25–50 μm. (c) 50–75 μm. (d) 75–100 μm. (e) > 100 μm.
(×100 magnification).
TABLE I
JSC-1, JSC-1A, AND KSC-1 LUNAR SIMULANT SURFACE COMPOSITION MEASURED BY XPS (RELATIVE ATOMIC PERCENT)
AS A F UNCTION OF S IEVED S IZE F RACTION
TABLE II
BENEFICIATION DATA FOR KSC-1 SIMULANT SHOWING RELATIVE COMPOSITION CHANGES COMPARED TO THE ORIGINAL
COMPOSITION OF THE SEPARATED FRACTIONS (−ve PLATE, +ve PLATE, AND BOTTOM TRAY)
FOR THE T HREE D IFFERENT C HARGING M ATERIALS , N AMELY , Al, Cu, AND PTFE
V. C ONCLUSION
JSC-1 simulant is complicated, and contains other Fe- and A direct correlation between the work function of the JSC-1
Ti-bearing minerals, which makes the interpretation of the sep- simulant and that of the charging material was established.
aration data based on work-function differences alone complex. From this, a work function of ∼5.4 eV was determined for
Along with the high concentrations of C detected in JSC-1, at the simulant. However, this must be emphasized that this was
this point, in the project and time constraints, it was decided determined for the bulk material composition and may vary
to not continue with this simulant and rather focus on the between the minerals.
newer JSC-1A. Very little surface compositional difference was observed, as
3) JSC-1A: Table IV shows the XPS beneficiation data for measured by XPS, for the various size fractions for both JSC-1
the JSC-1A simulant as a function of charger material. The and JSC-1A. This allowed for the 50–75-μm size fraction to be
first observation was that the separation concentrations were used and represent chemically all size fractions.
not as dramatic as that for the KSC-1 simulant. This was to The twin-plate separator proved successful in separating
be expected because the presence and amount of ilmenite in back out a large percentage of ilmenite from the KSC-1 mixed
JSC-1A are not as defined as the individual components used simulant by charge with full recovery of the material. Some
to make up the KSC-1 simulant. However, there were some significant variations in compositions were observed in the sep-
distinct differences in the Fe and Ti concentrations between arated fractions of the JSC-1 and JSC-1A simulants; however,
the plates for all three chargers above the 5% error that indi- there seems no correlation or consistency in terms of the known
cated a degree of separation of ilmenite. The most significant relative work functions of the charging materials and that of
changes are for the Na concentrations for the Cu and PTFE the bulk simulant. However, the mineralogical compositions
chargers, which may show the presence of different plagioclase of the simulants were significantly more complicated than that
(albite) phases. of KSC-1, as shown by previous Raman data, with numerous
Authorized licensed use limited to: University of Newcastle. Downloaded on November 06,2023 at 03:05:03 UTC from IEEE Xplore. Restrictions apply.
1066 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 45, NO. 3, MAY/JUNE 2009
TABLE III
BENEFICIATION DATA FOR JSC-1 SIMULANT SHOWING RELATIVE COMPOSITION CHANGES COMPARED TO THE ORIGINAL COMPOSITION
OF THE S EPARATED F RACTIONS (−ve P LATE , +ve P LATE , AND B OTTOM T RAY ) FOR J UST THE Al C HARGER
TABLE IV
BENEFICIATION DATA FOR JSC-1A SIMULANT SHOWING RELATIVE COMPOSITION CHANGES COMPARED TO THE ORIGINAL
COMPOSITION OF THE SEPARATED FRACTIONS (−ve PLATE, +ve PLATE, AND BOTTOM TRAY)
FOR THE T HREE D IFFERENT C HARGING M ATERIALS , N AMELY , Al, Cu, AND PTFE
other iron phases present. This made the interpretation of the [11] The Lunar Regolith Simulant Materials Workshop, Huntsville, AL:
XPS data extremely difficult. Marshall Space Flight Center, Jan. 24–26, 2005. [Online]. Available: http://
est.msfc.nasa.gov/workshops/LRSM_docs/LRSM_Abstract_Book.pdf
The preliminary data presented herein indicate that tri- [12] D. S. McKay, J. L. Carter, W. W. Boles, C. C. Allen, and J. H. Allton,
bocharging lunar simulant, followed by a twin-plate charge “JSC-1: A new lunar soil simulant,” in Engineering, Construction, and
separator, is potentially a viable method for mineral separation Operations in Space IV. Reston, VA: Amer. Soc. Civil Eng., 1994,
pp. 857–866.
and enrichment of lunar soils. [13] Z. Sternovsky, S. Robertson, A. Sickafoose, J. Colwell, and M. Horanyi,
“Contact charging of lunar and Martian dust simulants,” J. Geophys. Res.,
vol. 107, no. E11, pp. 15-1–15-8, 2002.
R EFERENCES [14] J. G. Captain, S. Trigwell, E. E. Arens, J. E. Captain, J. W. Quinn, and
C. I. Calle, “Tribocharging lunar simulant in vacuum for electrostatic
[1] C. C. Allen, J. C. Graf, and D. S. McKay, “Sintering bricks on the moon,” beneficiation,” in Proc. AIP Conf. STAIF, Albuquerque, NM, Feb. 11–15,
in Engineering, Construction, and Operations in Space IV. Reston, VA: 2007, vol. 880, pp. 951–956.
Amer. Soc. Civil Eng., 1994, pp. 1220–1229.
[2] R. J. Williams and O. Mullins, “Enhanced production of water from
ilmenite,” in Proc. Lunar Planetary Sci. XIV, 1983, pp. 551–558.
[3] Eagle Engineering, “Conceptual design of a lunar oxygen pilot plant
lunar base systems study (LBSS) task 4.2,” NASA Johnson Space Center,
Houston, TX, EEI Rep. No. 88-182, 1988. Steve Trigwell received the B.Sc. degree in physics
[4] W. M. Agosta, “Electrostatic separation and sizing of ilmenite in Lunar from Leicester University, Leicester, U.K., the M.S.
soil simulants and samples,” in Proc. Lunar Planetary Sci. XV, 1984, degree in materials engineering from San Jose State
pp. 1–2. University, San Jose, CA, and the Ph.D. degree in
[5] W. M. Agosta, “Electrostatic concentration of lunar soil minerals,” in engineering science from the University of Arkansas,
Lunar Bases and Space Activities of the 21st Century, W. W. Mendall, Ed. Little Rock.
Houston, TX: Lunar Planetary Inst., 1985, pp. 453–464. He is currently the Manager of the Applied Sci-
[6] M. K. Mazumder, Wiley Encyclopedia of Electrical and Electronics ence and Technology Department in the Applied
Engineering, J. G. Webster, Ed. New York: Wiley, 1999. Technology Directorate of ASRC Aerospace Cor-
[7] S. Trigwell, “Correlation between surface structure and tribocharging of poration at Kennedy Space Center in Florida. The
powders,” Ph.D. dissertation, Dept. Appl. Sci., Univ. Arkansas, Little Department consists of three emphasis areas; the
Rock, 2002. Applied Chemistry Laboratory, the Corrosion Laboratory, and the Electrostatics
[8] P. K. Srirama, R. A. Sims, A. S. Biris, J. D. Wilson, and M. K. Mazumder, and Surface Physics Laboratory. His group is currently involved in all areas
“Development of a portable electrostatic charge separator,” in Proc. ESA- of lunar exploration for NASA’s commitment to returning to the Moon. His
IEEE Joint Meeting Electrostatics, 2003, pp. 241–247. current research interests include studying the surface electronic and chemical
[9] G. H. Heiken, D. T. Vaniman, B. M. French, Eds., Lunar Source Book: A properties of Apollo lunar regolith. He is also currently a Graduate Faculty
User’s Guide to the Moon. Cambridge, U.K.: Cambridge Univ. Press, Member in the Applied Science Department and an Affiliated Scientist with
1993. the Nanotechnology Center, University of Arkansas.
[10] I. I. Inculet and D. R. Criswell, “Electrostatic beneficiation of ores on the Dr. Trigwell received the NASA Space Flight Awareness Award for recogni-
moon surface,” in Inst. Phys. Conf. Ser. No. 48, 1979, pp. 45–53. tion in support of the vision for space exploration in 2007.
Authorized licensed use limited to: University of Newcastle. Downloaded on November 06,2023 at 03:05:03 UTC from IEEE Xplore. Restrictions apply.
TRIGWELL et al.: USE OF TRIBOCHARGING IN THE ELECTROSTATIC BENEFICIATION OF LUNAR SIMULANT 1067
James G. Captain received the B.S. degree in chem- Jacqueline W. Quinn received the B.S. degree in
istry from the University of Florida, Gainesville, in civil engineering from the Georgia Institute of Tech-
2000, and the M.S. degree in chemistry from the nology, Atlanta, and the M.S. and Ph.D. degrees in
Georgia Institute of Technology, Atlanta, in 2001. environmental engineering from the University of
From 2002 to 2005, he was an Analytical Chemist Central Florida,Orlando.
doing radiochemical analysis in conjunction with the Since 1989, she has been with the Kennedy Space
Georgia Department of Natural Resources for their Center (KT-D-3), NASA, FL, working first on the
environmental monitoring program in proximity to environmental controls and life support subsystem of
the Savannah River Site. Since 2005, he has been a the shuttle fleet prior to transitioning into research
Research Scientist with the Kennedy Space Center and development. Her current research ranges from
(ASRC-24), FL, working on a variety of environ- environmental cleanup to lunar soil enrichment.
mental and space-related programs. Dr. Quinn received NASA’s Government and Commercial Invention of the
Year, was inducted into the Space Technology Hall of Fame, and is a Tech
Museum Laureate.
Authorized licensed use limited to: University of Newcastle. Downloaded on November 06,2023 at 03:05:03 UTC from IEEE Xplore. Restrictions apply.