BASIC PHYSICAL AND CHEMICAL PROCESSESIN SPACE RADIATION EFFECTS ON POLYMERS

E. Kamaratos

Department of ChemistryChristopher Newport College

Newport News, VA 23606

J. W. WilsonNASA Langley Research Center

Hampton, VA 23665

C. K. ChangDepartment of Chemistry

Christopher Newport CollegeNewport News, VA 23606

Y. J. XuDepartment of Physics

Old Dominion UniversityNorfolk, VA 23508

Large Space Systems Technology - 1981Third Annual Technical Review

November 16-19, 1981

37

https://ntrs.nasa.gov/search.jsp?R=19820010403 2020-05-17T09:12:36+00:00Z

SPACE RADIATION PHYSICS AND CHEMISTRY

Space ionizing radiation-induced changes in a structural material areinitiated by energetic charged particle impact which transfers kineticenergy of the particle via its coulomb field into electronic excitationenergy of the material. In complex molecules, the electronic energy isusually shared among various vibrational modes through potential surfacecrossing phenomena which depend intimately on the molecular structure(ref. I). Radicals produced through bond rupture are usually placed on ahighly repulsive potential surface which may subsequently lead to reactionsnot accessible at thermal energies in the early phases. The time progres-sion proceeds under conditions of near thermal equilibrium after <1 nsecand reactions are akin to the more usual chemical processes (ref. 2). Inthe solid state, long lived radicals result from their lack of mobility andchemical activity ends only after long time periods which depend on thetemperature. It is natural for laboratory studies to begin with finalchemical products to infer events leading to their production, It is morenatural from a physicist's point of view to consider the initial impact andenergy handling processes. In the future we should meet in the middle ofthe diagram with some understanding of the radiation degradation of poly-mers. In the present report we are interested mainly in the energy absorp-tion and the immediately following events.

ENERGY

ABSORPTION

RELAXATION

&

HOT RADICAL

REACTIONS

DIFFUSION

&

CHEMICAL

REACTION

LONG-LIVED

RADICAL

REACTION

FINAL

PRODUCTS

10 -_6 SEC 10_ 1,_ 11-,--10-

SEC

T_oo

PHYSICAL

INTERACTION

CHEMICAL

PRODUCTS

38

ENERGY ABSORPTION

There are three major questions concerning the energy absorptionevent. First we should consider how the energy is imparted to the moleculefollowed by the second question of how the energy was handled by the systemleading to the final collision products. These final collision productsare the chemically active species which lead to the ultimate change in thematerial. Of nearly equal importance to what is produced is the thirdquestion relating to the spatial distribution of various product types inrelation to the charged particle trajectory. We shall consider these

questions in more detail.

ENERGY ABSORPTION

A) INTO WHAT MOLECULAR MODES OF THE POLYMER IS THE ENERGY DEPOSITED?

B) WHAT ARE THE CORRESPONDING COLLISION PRODUCTS?

C) WHAT IS THE SPATIAL DISTRIBUTION OF THESE COLLISION PRODUCTS?

39

COLLISIONAL ENERGY LOSS IN POLYMERS

The energy transferred per unit distance traveled by a passing chargedparticle to molecules as derived by Bethe (ref. 3) is shown. That thepassing coulomb field appears as a field of virtual photons is evidenced bythe strong dependence of the stopping power on the optical dipole oscilla-tor strengths, fn, and the corresponding excitation enery levels, En.As chemical bonds are altered, there is a shift in the excitation energies,En, and a shift in oscillator strengths as well (ref. 4), consequentlythe energy deposit depends on the specific molecular structure. The energydeposit, dE, over a distance, dx, of the trajectory is spread laterallyfrom the particle path according to cylindrical symmetry. Ehrenfest'stheorem requires the interaction time to be short compared to the oscilla-tor's period and empirically the photon transfer probability to a levelEn is given (ref. 5) by Pn. Clearly the most energetic molecularstates are excited near the particle path and the lateral extent of excita-tion expands with increasing particle velocity. Note that the high energyionization processes (i.e. large En) lie nearest the particle path butthe energetic secondary electrons migrate far from the initial particletrajectory (ref. 6). A fundamental quantity is the mean excitation energy,I, which is related to dipole oscillator strengths, fn, and energylevels, En, as shown. This quantity will be studied for effects due tochemical bonding.

STOPPING POWER

4BNZ_E 4D___EE: [ ] Z FN iN [2MV-_-_2]

DX MV2 N EN

LATERAL EXTENT

ENPN " EXP [-(--",j,-H-_--)R]

FUNDAMENTAL QUANTITY

Z IN(1) = Z FN [N(E N)m

4o

LOCAL PLASMA MODEL FOR POLYMERS

The calculation of excitation energies and oscillator strengths ofpolymers is beyond present day computational ability. Considerable simpli-fication can be made through replacement of the molecule by an equivalentspatial dependent plasma (ref. 7). The stopping power is now expressed interms of the local plasma density, p(r), and the corresponding plasma fre-

quency, Up. The lateral extent of the energy deposit, Pp, and the meanexcitation energy, I, are now also given in terms of the local plasmamodel. To effectively use this model for molecular calculations, means ofsimple approximation to the electron density of the molecule must befound. For this purpose we used the Gordon-Kim electron gas model (ref. 8)of molecular bonds for the estimation of chemical bond effects (refs. 9-I0). The main value of the local plasma model is it's utter simplicity.The accuracy of the local plasma model is tested here by applying the rela-tion between the mean excitation energy, I, and the local plasma density asa fundamental quantity to ionizing radiation interaction.

STOPPING POWER

4nNZ_E 4DE=[ ,, ]DX MV2

=p = Vq,E2p(R)/M

LATERAL EXTENT

J P(R) CN [2MV] D3R

THWp

XHUpPp - EXP [-(--"_rH-_-) R]

FUNDAMENTAL QUANTITY

Z IN(I} = S p(R) iN (YHmP) D3R

41

CHEMICAL BOND EFFECTS ON MEAN EXCITATION ENERGIES

The mean excitation energies of several molecules related to the poly-mer problem have been calculated according to the local plasma approxima-tion. The values labeled "atomic" are those obtained using the Braggrule (ref. ii), in which chemical bond effects are neglected, with theaccurate atomic values of Inokuti and coworkers (ref. 12). The presenttheoretical results are shown for comparison. In addition, experimentalresults are also shown for which the merit of the present theory is clearlydisplayed. We conclude from these comparisons that the l.ocal plasma modelgives an adequate representation of particle impact with the polymer,although the coupling between the electronic mode excited and the remainingvibrational-rotational modes of the polymer are yet undetermined, Insearch of empirical evidence of the polymer's reaction to the impact wehave examined specific fragmentation data from mass spectroscopy.

I : EXP [Z-I I P(R) _N(YH_p)D3R]

ATOMIC PRESENT THEORY EXPERIMENTAt

CEl4 35. I 44.7 42.8

(CH2} x 43.5 55.0 53.4

C6H6 50.6 60.6 61.4±1.9

H2 15.0 18.9 18.5±0.5

GRAPHITE 62.0 76.1 78.5±1.5

42

STRUCTURE OF AN EPOXY POLYMER

A repeating unit is shown in the structure of an epoxy resin producedby curing tetraglycidyl methylenedianiline with diamino diphenyl sulfone(ref. 13). This resin is one of the resins that are reported to be underinvestigation for possible use in composite materials for applicationsranging from Army materiel to aircraft and space structures. An interest-ing characteristic of this polymer is the presence of several tertiaryamino functional groups in the repeating unit that impart certain chemicalproperties to the polymer like, e.g., those of a weak base. These chemi-cal properties may lead to degradation of the mechanical properties, inaddition to increase in weight, on chemical reaction (e.g., with C02 andatmospheric pollutants). The association and subsequent reaction of theamino groups with molecules of water may have an effect on the NMR lineobserved with the soaked polymer that was recently ascribed to a state ofthe water molecules "perhaps hydrogen bonded to the polymer" (ref. 14). Inaddition to the amino functional group there are two additional functionalgroups in the repeating unit: the -OH (hydroxyl) and the -S02 (sulfonyl)functional groups. The presence of functional groups may also be importantin the determination of transients formed after the transfer of energy fromionizing radiation to a polymer. Since electrons penetrate into matterdeepest and are the primary constituents of the space ionizing radiation,and since all ionizing radiation in space leads to the formation of second-ary electrons on interaction with matter, it is of general interest toinvestigate the effect of electron impact on molecules first. From thisviewpoint it is of interest to study the role of functional groups in elec-tron impact fragmentation of molecules. Because of the predominant produc-tion of positive ions (usually, but not always, by orders of magnitude)over that of negative ions on electron-impact and the larger availabilityof mass spectra for positive ions, this investigation started with positiveions first. However, the possibility of an important role for negativeions should not be discounted.

#C / OH OH _ /N N I I \'_ S N \ _./.."I"/ O_I ___/ H2_CH _CH2_ /CH 2 _CH _CH2"_ IIX_--/

N

Ctt 2

N o "1

\ .-. -. o. ,, .

0

'N 0\

\\

\N

\\

\\

\N

N\

\\

o I

I

43

ELECTRON-IMPACT MASS SPECTRA OF MOLECULES WITH AN AMINO GROUP

Mass spectra of some molecules that contain amino functional groupsare shown (ref. 15). It has been observed experimentally that among sever-al chemical functional groups, including the hydroxyl and amino groups,that are present as substituents in aliphatic chemical compounds the aminofunctional group influences the chemical bond scission and ionization onelectron impact the most strongly (refs. 16-17). The extent of substitu-tion on the amino group itself appears to increase the influence of theamino group on the dissociative ionization. The dominant" bond scission isthat between m and _ carbon atoms in hydrocarbon chains, leading to theformation of imonium ions. In aromatic amines with no _ and # carbonatoms in aliphatic groups the parent ion or the alkyl substituted aminogroup ion may be formed predominantly.

93 C_HTN 62-53-3Benzenamine

PhNH2

., .,i......... d. ,L 1

149 CsoHlsN

Benzenamine, N,N-diethyl-

Et2NPh

91-66-7

10 20 30 40 SO eO 70 I10 I1_ 100 110 120 130 140 150

177 C12HIaN

Benzenamine, N-(2,2-dimethylpropyl)-N-methyl-

PhNMe CH2 CMe3

53927-61-0

0 _

'il0 _

d.. 1...... ill I ......

] i" • LT • • v , r

44

ELECTRON-IMPACTMASSSPECTRAOFLARGEMOLECULESWITHAMINOGROUPS

Structural formulae and electron-impact mass spectra are shown foradditional molecules that have amino functional groups as their substitu-ents and, in addition, appear to be constituents to some extent of therepeating unit in the structure of the polymer. The fragmentation patternon electron-impact of these and other compoundsto be investigated may leadto the identification of someof the sites of bond scission and of ions andradicals that may be formed under the action of ionizing radiation on thepolymer.

198 C:3H.N2

Benzenamine, 4,4'-methylenebis-

101-77-9

0:

'il:I0 a,

;o _ ;o _o _o _o 7o _ .o loo ,;o Go ,so .o ,6o

II., ,. , , , , .... , ,_oo 17o leo _oo 2oo 2_o 22o 23o 24o 25o 2eo 2To 2eo no $oo

268 C:THzoN20

Methanone, bis[4-(dimethylamino)phenyl]-

90-94 -8

; • • w w • v • w • •

.... i .... Li ....

• • w • ! w

,.o ,,o ,fo ,.o _ 2,o 2_o _o 2.o 25o _o 2;o 2_o _ _o

45

ELECTRON-IMPACT MASS SPECTRA OF MOLECULES WITH A HYDROXYL GROUP

The electron-impact mass spectra of several melecules that contain thehydroxyl functional group are shown, The hydroxyl group in aliphatic com-pounds appears to favor ionization and scission of the bond between _ andcarbon atoms with respect to the hydroxyl group, although much less domi-nantly compared to the amino group.

198 C14H140

Benzeneethanol, a-phenyl-

PhCH2 CH( OH)Ph

614-29-9

100,

100

eo

40

20

o .... ;',;o ,_o ,80 ,_o'_o 2;0 2_o 2_o 2;0 _ _o 2;0 2;0 _ _o

192 ClaH2oO

Phenethyl alcohol, a--butyl-/_-methyl-

7661-43-0

Me ( CH2 ) S CH(OH) CHM_Ph

100

o- ;o _ _o _o 50 _o 7'o ;o ;o ,;o Go ,_o ,;o ,_ ,;o

46

ELECTRON-IMPACT MASS SPECTRA OF MOLECULESWITH BOTH A HYDROXL AND AN AMINO GROUP

When both an amino group and a hydroxyl group are present in a mole-

cule, the amino group appears to form a more abundant, if not the mostabundant, fragment ion on electron-impact, namely the =C=NH + (imonium

ion) compared to the =C=0H + (oxonium ion), the ion from the hydroxyl

group.

75 CaHaNO 78-91-11-Propanol, 2-amino-

HOCH2 CH( NH2 )Me

;o 2o 3o ,o go /o 70 ao ,o 1_o 1,o ,20 Go 1_o ,50

75

2-Propanol, 1-amino-

C3H9NO 78-96-6

H2 NCH2 CH ( OH ) Me

10 20 30 40 so 60 70 80 90 100 110 120 130 140 150

75

1-Propanol, 3-amino-C3HgNO 156-87-6

H2 N ( CH2 ) 3

JJ .._

...... _ :., •_0 20 30 40 60 70 80 90 100 110 120 1 0 1 1._0

47

ELECTRON-IMPACT MASS SPECTRA OF MOLECULES WITH A SULFONYL GROUP

The electron-impact mass spectra of certain molecules that contain the-SO 2- functional group are also shown, lons formed by scission of C-S andS-O bonds appear to form in addition to the parent ion. The nature of thesubstituents influences the dissociative ionization involved.

248 C12H12N202S

Benzenamine, 4,4'-sulfonylbis-

80-08 -0

H2N SO2-- NH

218

Benzene, 1, l'-sulfonylbis-C,2HzoO2S 127-63-9

PhSO2Ph

0

10 20 30 40 50 60 70 60 90 100 110 120 130 140 150

48

SITES OF MAJOR DISSOCIATIVE IONIZATION ON THE POLYMER ACCORDING TOELECTRON-IMPACT MASS SPECTRA

Sites of possible major ionization and bond scission are shown on arepeating unit of the polymer formed by curing tetraglycidyl methylene-dianiline with diamino diphenyl sulfone. The determination of possiblesites is based on electron-impact mass spectrometric evidence.

SITES OF MAJOR DISSOCIATIVE IONIZATION

ACCORDING TO ELECTON-IMPACT MASS SPECTRA

./

I CH 2

I

O

CH 2 _ II\ o

\\\

\\

\

i \\

i \\

49

LONGITUDINALENERGYABSORPTIONFORPROTONSPENETRATINGINTOTHEPOLYMERANDA COMPOSITEMATERIALOFTHEPOLYMER

The density of chemically active species is of utmost importance indetermining the chemical products within an activated region since theyaffect the rates of various competing chemical reactions. We shall nowconsider the spatial inhomogeneity of the initial energy deposit. Theenergy absorption as a function of penetration depth is shown for monoener-getic protons of three energy levels. It has been calculated according tothe stopping power of Anderson and Ziegler (ref. 18). The energy depositper unit volume is quite high due to the proton's low speed and limits tilepenetration to short distances (note that the penetration depth isexpressed in units of areal density which is the distance of penetrationtimes the material density). Penetration depths and energy deposit in thisgraph for low energy protons should be compared with electron data of thefo]lowing graph.

1000250 keY

I tJ J

-" 800 ''I /

600/ i"I

° Y1eN 2_22_N 20O ,

I

lII)

I0 ,0

PROTON

500key/

II/I

!/

//

graphite/epoxy 1 000 keY70/30 ,,

.... i lI lO/lOO ,

f II I

j/I II

1.11./ // "//// /

| |

lxlO -3 2x10 -3

Penetration depth (g/cm 2)

i-3

3xlO

5O

LONGITUDINAL ENERGY ABSORPTION FOR PASSAGE OF ELECTRONS THROUGHTHE POLYMER AND A COMPOSITE MATERIAL OF THE POLYMER

The energy absorption for electrons is more dispersed in the materialas a result of the electron's smaller mass and higher velocity. Shown hereis the distribution of energy deposit as a function of penetration depthcalculated according to the electron transmission coefficient and averagetransmitted energy (ref. 19). The declining peak values as a function ofdepth results from the increased multiple scattering of the electrons ran-domizing their direction of travel at larger depths. Such effects are ofprincipal importance in defining dose profiles within the material.

7

'E 6

4

e-_ 1

0

2.50keY

/A/! :

// Ij

'7 i

ELECTRON

graphite/epoxy70/30o/1oo

500keV1 000key

o.I 0.2 0.3 0.4

Penetration depth (g! cm2)

0.5

51

52

EXCITATION AND IONIZATION DENSITY NEAR A IO0-KeV-PARTICLE PATH

The lateral dispersion of the energy deposit extends over two energyabsorption regions, the core and the penumbra. In the core, the energytransferred from a passing charged particle to a material shows radialdependence that depends on the mechanism of the coulombic interaction con-sidered. One process involves the coulombic interaction between thecharged particle and the electron of one individual atom or molecule, inwhich case the radial dependence of the energy transferred is givenapproximately (ref. 5a) by equation (i). Another process involves thecollective longitudinal excitation of valence electrons of the atoms ormolecules in the condensed phase, in which case the radial dependence isgiven approximately (ref. 5b) by equation (2) (for singly charged projec-tiles)

Pn = exp(-Enr/3hv) (i)

p(r) = exp(-2mpr/v)/r 2

Clearly, the highest energy transfers are restricted to small radialdistances. Even so the highest energy transfers result in energetic secon-dary electrons which generally travel far from their sites of productionand deposit their energy far from the trajectory of the original passing par-ticle forming a penumbra of energy deposited outside the core region. Thecore and penumbra of i00 KeV protons and electrons are shown clearly in thegraph. The core region of the proton is limited to a few atomic distancesby the proton's low velocity while the electron core to about 6 nm accord-ing to equation (I). The proton-produced penumbra is limited by the secon-dary electron mass to proton mass ratio, which limits the energy trans-ferred to secondary electrons. The electron produced penumbra extends faroff the graph to the right. There are great differences between the exci-tation densities achieved by the passage of the two particles. Withoutdoubt, rapid recombination will characterize the chemistry in the proton'score region. A more detailed analysis is required to better define thechemistry of these various regions of exposure.

er_!

E

.m

(--O

1021

1018

10150

100 keY

electron

Radial distance (nm)

EXCITATION AND IONIZATION DENSITY NEAR A IO00-'/eV-PARTICLE PATH

At higher particle energy the same features are apparent in the trackstructure but the radial distances are expanded. The proton core regionstill appears on the graph but the electron core extends to the right ofthe present graph. Qualitatively, we anticipate the proton core chemistryto be radically different from the remainder of the exposed regions. Thechemical change in any solid material will be limited to a small cylindri-cal region around the individual particle trajectory. Early in the expo-sure these cylindrical regions will not overlap and their evolution willdevelop independently of one another. Late in the exposure of highlyirradiated material the chemistry of later tracks will generally be alteredby the prior change of the material caused by a previous track.

A

I

E

em

0Q_

eI

xL,I.I

1024

1021

1018

10,150

1 000 keY

electron

i I I , I ! ! !

2 4 6I I

8 10

Radial distance (nm)

53

GEOSTATIONARY DOSE IN ORGANIC COMPOSITES SHOWING GEOMETRIC AND MODELUNCERTAINTIES

The accumulated 30 year dose for geostationary orbit is shown in the

graph. It is represented by exposure of : 10 11 rad at : 0.5 p depths dueto low energy protons and _ 5xlO tad at _ 50 _ due to electrons. At thesevalues, each portion of material receives energy by the near passage ofseveral particles. Two effects could result from such multiple track expo-sure: (a) if sufficiently close in time, then chemically active species ofthe two trajectories may alter the chemical end product; (b) even if longtimes elapsed between the passage of the two particles then the secondpasses through previously altered material. Of interest is the number oftraversals which overlap within a given time interval for an acceleratedtest of the 30-year geostationary exposure.

1012

1011

Upper limit

10IC

Lower limit

IO8

.1 1

Thickness, mils.

10 100

54

PROBABILITY OF OVERLAP OF TWO-PARTICLE TRACKS WITHIN TIME At(30-YEAR GEOSTATIONARY EXPOSURE ACCELERATED BY 104 )

The probability of overlap of two particle tracks within a time inter-val At is shown for an accelerated test at 30 year geostationary exposurelevels. The test is assumed completed in one day (24 hours). It is seenthat the more active chemical species (At _ i usec) will rarely be affectedby such events. A small effect is expected for the slow chemical processes(At _ I msec) in the penumbra region of the electron exposure. Effects onlong lived radicals (At = I sec) are likely in nearly all regions. Most ofthe exposure is seen to take place in previously disturbed material (AtI day). What is clear from these figures is that material changes resultfrom the combined effects of spatially compact individual particle tracks.

PROTON ELECTRON

AT CORE PENUMBRA CORE PENUMBRA

I _SEC 3X10 -li 3X10 -e 5X10 -_ 2XlO -5

] MSEC 3x10 -e 3x10 -s 5xlO -_ 2x10 -2

1SEC 3X10 -s 3XlO -2 5XlO -] "1

I DAY " I " I " I " I

55

SUMMARYAND CONCLUSIONS

A useful model of charged particle impact with the polymer has beenpresented.

Principal paths of molecular relaxation have been identified.

The radiation-induced changes will be confined to local regions aboutindividual particle tracks.

The main additional effects of accelerated testing may result fromeffects because of possible higher temperatures and long-lived freeradicals.

Chemically active source terms are now reasonably defined. Attentionmust now be given to subsequent processes.

56

References

. a. Rosenstock, H. M.; Wallenstein, M. B.; Wahrhaftig, A. L.; andEyring, H.: Absolute Rate Theory for Isolated Systems and the MassSpectra of Polyatomic Molecules. Proc. Natl. Acad. Sci. U.S., vol.38, 1952, pp. 667-678. b. Rosenstock, H. M.: Theory of MassSpectra. A General Review. Adv. Mass Spectrom, vol. 4, 1968,pp. 523-545.

. Kupperman, A.: Diffusion Model of the Radiation Chemistry of AqueousSolutions. Radiation Research, North-Holland Pub., 1967, pp. 212-234.

• Bethe, Hans A.: Zur Theorie der Durchgangs schneller Korpurskular-strahlen durch Materie. Ann. Physik, vol. 5, 1930, pp. 325-400.

. Platzman, Robert J.: Influences of Details of Electronic Binding onPenetration Phenomena, and the Penetration of Energetic Charged Par-ticles through Liquid Water. Symposium on Radiobiology; The BasicAspects of Radiation Effects on Living Systems (Oberlin College, June14-18, 1950), 1952, pp. 139-173.

, a. Wilson, John W.: Solar-Pumped Gas Laser Development. NASA TM81894, December 1980. b. Ritchie, R. H.; and Brandt, Werner:Primary Processes and Track Effects in Irradiated Media. RadiationResearch; Biomedical, Chemical and Physical Perspectives. AcademicPress, Inc., 1975, pp. 315-324.

. Katz, R.; Ackerson, B.; Homayoonfar, M.; and Sharma, S. C.: Inactiva-tion of Cells by Heavy lon Bombardment. Radiation Res., vol. 47,1971, pp. 402-424.

. Lindhard, J.; and Scharff, M.: Recent Developments in the Theory ofStopping Power. Principles of the Statistical Method. Penetration ofCharged Particles in Matter (Conference at Gatlinburg, Tennessee,September 15-18, 1958.). NAS-NRC Publ. 752, 1960, pp. 49-55.

, Gordon, R. G.; and Kim, Y. S.: Theory for the Forces Between Closed-Shell Atoms and Molecules. J. Chem. Phys., vol. 56, 1972, pp. 3122-3133.

, Wilson, John W.; and Kamaratos, Efstathios: Mean Excitation Energyfor Molecules of Hydrogen and Carbon. Phys. Lett., vol. 85A, 1981,pp. 27-29.

10. Wilson, J. W.; Chang, C. K.; Xu, Y. J.; and Kamaratos, E.: lonic BondEffects on Mean Excitation Energy for Stopping Power. To be publishedin J. Appl. Phys., Feb. 1982.

11. Bragg, W. H.; and Kleeman, R.: On the _ Particle of Radium and theirLoss of Range in passing through various atoms and molecules.Philos. Mag., vol. i0, 1095, pp. 318-340.

57

12. Dehmer, J. L.; Inokuti, M.; and Saxon, R. P.: Systematics of momentsof dipole oscillator-strength distributions for atoms of the first andsecond row. Phys. Rev., vol. 12A, 1975, pp. 102-121.

13. Long, E. R., Jr.: Electron and Proton Absorption Calculations for aGraphite/Epoxy CompositeModel. NASATechnical Paper 1568. November1979, p. 16.

14. Lawing, D.; Fornes, R. E.; Gilbert, R. D.; and Memory, J. D.: Tem-perature dependenceof broadline NMRspectra of water-soaked, epoxygraphite composites. J. Appl. Phys., vol. 52, 1981, pp. 5906-5907;and references therein.

15. Heller, S. R.; and Milne, G. W. A. (eds.): EPA/NIHMassSpectral DataBase; 1980 Cumulative Indexes. NSRDS-NBS63, Suppl. i.

16. Remberg,G.; Remberg,E.; Spiteller-Friedmann, M.; and Spiteller, G.:Massenspektren SchwachAngeregter Molek_le. 4. Mitteilung. Org. MassSpectrom., vol. I, 1968, pp. 87-113.

17. Remberg, G.; and Spiteller, Go: Uber den Einfluss von Substituentenauf die prim_ren Abbaureaktionen aliphatischer Verbindungen im Massen-spectrometer. Chem.Ber., vol. 103, 1970, pp. 3640-3660.

18. Anderson, H. H.; and Ziegler, J. F.: The Stopping and Rangesof lonsin Matter. Vol. 3. Hydrogen Stopping Powers and Ranges in All Ele-ments. PergamonPress, 1977.

19. Mar, B. W.: An Electron Shielding Analysis for Space Vehicles.Nucl. Sci. and Eng., vol. 24, 1966, pp. 193-199.

58