Ind. Eng. Chem. Prod. Res. Dev. 1980, 19, 103-106 (13) Kaelble, D. H. J . Adhesion, 1970, 2, 66. (14) Kulkarni, R. D.; Goddard, E. D.; Kanner, B. Ind. Eng. Chem. Fundam. 1977, 16, 472. (15) Lee, W. A,; Rutherford, R. A. "Polymer Handbook", 2nd ed.; Brandrup, J., Immergut, E. H., Ed.; Wiley: New York, 1975; p 111-139. (16) Noli, W.; Steinbach, H.; Sucker, C. J . Polym. Sci. C . 1971, 34, 123. (17) Owen, M. J.; Denis. C. J . Cell. Plast. 1977, 13, 264. (18) Owen, M. J.; Evans, J. I.B r . Polym. J . 1975, 7 , 235. (19) Owen, M. J.; Thompson, J. Br. Polym. J . 1972, 4 , 297. (20) Owen, M. J.; unpublished Dow Corning value. (21) Owens, D. K.; Wendt, R. C. J . Appl. Polym. Sci. 1989; 13, 1741. (22) Pittman, A. G. "High Polymers", Vol. 25, "Fluoropolymers"; Wall, L. A,, Ed.; Wiley-Interscience: New York, 1972; p 419. (23) Plumb, J. B.; Atherton, J. H. "Block Copolymers"; Allport, D.C.; Janes, W. H., Ed.; Applied Science Publ. Ltd.: London, 1973; p 305. (24) Quaal, G. J.; unpublished Dow Corning value. (25) Riedo, F.; Czencz, M.; Liardon, 0.; Kovats, E. S.M l v . Chim. Acta 1978, 61, 1912. (26) Roe, R. J. J . Phys. Chem., 1988, 72, 2013.
103
(27) Ross, S. Rensselaer Polytechnic Institute Bulletin No. 63; New York, 1950. (28) Shafrin, E. G. "Polymer Handbook", 2nd ed.; Brandrup, J.; Immergut, E. H.. Ed.; Wiley: New York, 1975, p 111-221. (29) "Tables of Interatomic Distances and Configurations in Moiecules and Ions", Spec. Publ. No. 11: The Chemical Society: London, 1958. (30) Tobolsky, A. V. "Properties and Structure of Polymers", Wiley: New York, 1960; p 67. (31) Voronkov, M. G.; Mileshkevich. V. P.; Yuzhelevskil, Yu A. "The Siloxane Bond"; Livak, J.. Transl.; Consultants Bureau: New York. 1978; p 12. (32) Wu, S. "Polymer Blends"; Paul, D. R.; Newman, S., Ed.; Academic Press: New York, 1978; Vol. 1, p 243. (33) Zisman, W. A. Adv. Chem. Ser. 1964, No. 43, 1. (34) Zisman, W. A. "Symposium on Adhesion and Cohesion", Weiss, P., Ed.; Elsevier: New York, 1962; p 176.
Received for review June 18, 1979 Accepted November 12, 1979
The Relationship of Chemistry to Electronic Properties of Graphite Intercalation Compounds: A Short Review Lawrence B. Ebert' and Joseph C. Scanlon Corporate Research-Science Laboratories, Exxon Research & Engineering Company, Linden, New Jersey 07036
Recent reports have suggested the electronic conductivity of some intercalation compounds of graphite to be comparable to that of copper. The origin of this conductivity increase may be found in an analogy between intercalation compounds of graphite and simple molecular basis radical anions and cations. The validity of this analogy is demonstrated by magnetic resonance experiments which demonstrate both the presence of redox chemis?ry in intercalation and the existence of delocalized electrons on the graphite planes following intercalation.
Much excitement has been generated by recent reports of highly conducting intercalation compounds of graphite (Fischer and Thompson, 1978). While the ultimate technical value of these compounds in unclear, the properties of the intercalates can be understood in relatively simple chemical and physical terms. T o present the relationship between the physical property, electronic conductivity, and the chemical identity of the compounds, this paper is divided into three parts. Initially, the basic terminology of intercalation is defined. Secondly, data are presented which show the existence of high conductivity in certain graphite intercalation compounds, and, finally, the key chemical cause of the high conductivity, charge transfer between host and guest, is analyzed.
Terminology As is seen in Figure 1, graphite is composed of sheets of sp2hybridized benzenoid carbon. This layered structure accounts not only for the lubricating properties of graphite but also allows insertion, within the interlamellar voids, of various chemical species. This insertion, referred to as intercalation, increases the distance between carbon planes without severely disrupting the planar benzenoid carbon array. The chemistry of graphite intercalation is amphoteric, with either chemical oxidants or reductants inserting with facility. Along the c axis, the direction perpendicular to the planes, intercalation may proceed with a certain periodicity. As is seen in Figure 2, the concept of staging describes this c axis order and may be considered to be the ratio of carbon planes to guest planes. Thus a first stage compound is most concentrated. Ordering within the 0196-4321/80/1219-0103$01 .OO/O
I. Conductivities of Some Common Metals at Room TemDerature'
Table
element silver copper gold aluminum iron graphite
u
(ac m ) - '
6.3 x 10' 5.9 x 10' 4.3 x l o 5 3.8 x 10' 1.0 x l o 5 7.3 x l o 4
After Ubbelohde ( 1 9 7 6 )
plane is frequently observed, with a common array involving sixfold coordination of intercalant at a distance of 4.9 A, as illustrated in Figure 3. Ordering along the a and c axes is sometimes combined; curiously, in the case of C8K this ordering yields an orthorhombic, rather than a hexagonal, Bravais lattice. For more detailed discussion of intercalation compound terminology and structure, the reader may consult recent reviews (Ebert, 1976; Herold, 1977). Intercalation Compounds of Graphite as Synthetic Metals Let us turn now to the question of conductivity in graphite compounds. As is seen in Table I, graphite itself, along the a axis, is a fair conductor, a factor of nine lower in conductivity than silver a t room temperature. The question naturally arises, can chemical modification enhance the conductivity of graphite? In point of fact, prior to World War 11, A. R. Ubbelohde constructed a double hypothesis which predicted that graphite intercalation compounds should be good elec0 1980 American
Chemical Society
104
Ind. Eng. Chem. Prod. Res. Dev., Vol. 19, No. 1, 1980 Table 11. The A Axis Conductivities of Intercalated Highly Oriented Pyrolytic Graphite Are Largea r.f. method,
bridge method ua,
2.7 x 2.7 x 2.2 x 1.0x 9.5 x 5.9 x
HOPG C,,HNO,
c,,A~F,
I I
C,Rb C8K cu
I I
I
a,,+
I
a
I+
Figure 1. The hexagonal modification of graphite. Typically, a. = 2.45 A and co = 6.7 A.
lo4 lo5
11.4 1.96 105 0.23 105 .__ 104 4.8 x 104 105 ___
copper silver lithium
I I
(2" C,AsF5 (1") graphite C16AsF5
a
SECOhC
ElRST
THIRD
2.463
1.4155
a=4.926
Figure 3. The in-plane C8X network characteristic of a number of compounds.
tronic conductors (Ubbelohde, 1976). The two points of the hypothesis were, number one, that graphite intercalation compounds were analogous to molecular aromatic charge transfer compounds as potassium naphthalenide, and, number two, that such charge transfer, in the case of graphite, would create an ensemble of delocalized conducting electrons. Intercalation of graphite, even under the primitive conditions of a laboratory demonstration, does increase the conductivity of graphite. Precise measurements on highly oriented pyrolytic graphite, given in Table 11, suggest that the conductivity may in fact rival that of copper. This increase in the conductivity along the a axis is a common feature of intercalation with either chemical oxidants or chemical reductants (Ubbelohde, 1976). Conversely, the conductivity along the c axis usually is diminished by intercalation.
(ac m ) - '
a = aau;'
2.6 X l o 4 3.3 x 105 6.3 x los 9.1 x 104
___
-_-
2.3 X l o 2 1.4x 105 2.7 x l o 6
._-
21
___
After Zeller et al. (1977).
metal
Figure 2. The concept of staging reflects periodicity along the c axis. The stage of a compound may be defined as the ratio of graphite layers to intercalant layers.
( a cm)-'
Table 111. Conductivity and Density of States for Metals
iTAG1l.G
I Ir
anisotropy,
ua,
UC,
compound ( 5 2 cm).'
States/(eV cm3).
density of states" 1.8 x 1.6 X 1.5 x 3.7 x 2.8 x 6.5 x
1022 10'' 102' lo2' 1020
conductivityb 5.9 x 6.2 x 1.2 x 5.8 x 4.7 x 7.3 x
105
10' 105
lo5 lo5 104
(n ern).'.
What is the origin of the conductivity increase? According to the Ubbelohde viewpoint, it is an increased carrier density arising from transfer of charge between the host graphite and the guest molecules. Nevertheless, conductivity is actually a product of terms, involving not only the carrier concentration but also the mobility of the carriers. Recently completed magnetic resonance experiments shed some light on this problem (Weinberger et al., 1978). Performing the Shumacher-Slichter experiment on graphite/AsF,, the authors determined the graphite intercalation compounds to have a density of states lower than that of normal metals, as illustrated in Table 111. In fact, the density of states of the graphite intercalation compounds is about the value for graphite, even though the conductivity of the intercalates is on a par with the good conductors. As conductivity is proportional to the product of carrier density (density of states X Fermi level) and mobility, Weinberger inferred from this that the high conductivity of graphite intercalates arose from an increase in mobility rather than an increase in carrier concentration. At this point, one might conclude that the basic questions in the conductivity of graphite compounds have been answered. Indeed, intercalating almost anything into graphite does increase the a axis conductivity. The reason for this increase is probably a balance between an increase in carriers and an increase in mobility. Chemistry and conductivity are intrinsically linked, because the electrons added to or removed from the aromatic planes are, in part, the carriers of the electrical current. Furthermore, the process of intercalation seems to create carriers which scatter less than the carriers of intrinsic graphite. Unfortunately, the next steps to take in conductivity optimization are not clear. If one tries to increase the carrier density by using stronger oxidants, one observes the formation of sp3, rather than sp2,carbon hybridization forms (Selig et al., 1977). Enhancement of carrier mobility cannot be systematically made until an understanding of the mobility increase on intercalation is attained. Although notable advances in band structure calculations have been made (Holzwarth et al., 1978), studies to date have been restricted to graphite/alkali metal compounds. While possessing simple, well-defined structures with atomic basis intercalant species, these compounds
Ind. Eng. Chem. Prod. Res. Dev., Vol. 19, No. 1, 1980 105 _-METAL
H A L I D E ) SERVE T
M
H O X I D I Z E AN3 INTERCALATE GRAPHITE
"GRAPHITE/AsF5"
I S A CHARGE TRANSFER COMPOUND
G E N E R A L I Z E D SCHEME F O R O X I D A T I V E I N T E R C A L A T I O h : G
t
XI yClx
+.
G/Ix t yRe
I
=
I N T E R C A L A T I N G METAL H A L I D E
Ox
=
O X I D I Z I N G S P L C I E S , AS METAL H A L I D E OR HALOGEN
Re
=
REDUCED SPECYES
*The r a t i o of AsF;
t o AsF5 i n t h e f i n a l "C10AsF5" compound i s
THE REDUCED S P E C I E S MAY F O W A FCNDAMEhTAL P O R T I O N OF THE I V T E R C A L A T I O N
a s s i g n e d by t h e amount o f AsF3 r e l e a s e d d u r i n g vacuum t r e a t m e n t .
COMPOUND ( X e F 4 , X e F 6 , IF,)
X-ray a b s o r p t i o n edge measurements, however, suggest t h e f i n a l
NOR I T MAY NOT ( A s F 5 ) .
( A F T E R EBERT AN3 S E L I G ( 1 9 7 ' ; )
Figure 4. The charge-transfer hypothesis as applied to oxidative intercalation by metal halides. For compounds as FeC1, and AsF5, the metal halide serves both to oxidize and to intercalate the graphite. Both neutrals and anions are intercalated.
suffer the complication of "alloy-like'' identity, arising from the intercalation of a metal into graphite. Unlike in the acceptor compounds (as C8AsF,), the carriers in the graphite/alkali metal compounds exhibit a great deal of intercalant character, as evidenced by an increased c axis conductivity over graphite and wide EPR lines arising from metal atom spin orbit interactions. To more fully utilize intercalation compounds, one requires a more detailed understanding of the distribution of electron charge in the compounds. For instance, is all the charge transferred to the graphite planes delocalized and available for applications in conductivity and catalysis? Are the intercalated species heterogeneous with respect to oxidation state and ionicity?
The Chemistry of Charge Transfer in Graphite Intercalation Compounds The final portion of this paper thus dwells on subjects related to the chemistry of intercalation compounds, with emphasis on determination of the charge transfer aspects of the materials. In seeking evidence for such charge transfer, we are attempting to bridge the gap between intercalation compounds of graphite and molecular basis complexes as potassium naphthalenide. One of the major tools in this characterization will be magnetic resonance, involving both electron paramagnetic resonance and nuclear magnetic resonance performed on solid samples. If the charge transfer hypothesis, depicted in Figure 4, is correct, we might expect to see evidence for intercalant molecules existing in an oxidation state other than the original. Whether or not these reduced species are visible in the solid phase product of the intercalation depends on the situation. With strong oxidants, only the reduced species may intercalate, while with weak oxidants, intercalation may involve primarily the initial oxidation state. Both nuclear magnetic resonance and electron paramagnetic resonance have successfully been used to observe changes in oxidation states on intercalation (Ebert and Selig, 1977). With xenon-fluorine compounds, the chemical shift of the fluorine NMR resonance is a direct measure of the oxidation state of the xenon (Frame, 1969);with the compound of nominal stoichiometry "CI9XeF,", the fluorine chemical shift is that corresponding to Xe(1V) and not to Xe(V1). With certain transition metals, determination of the g shift of the EPR resonance can effectively monitor oxidation state. Starting with Cr(V1) compounds as CrOBor Cr02C12,one sees an ESR resonance at g = 1.97, which may be associated with a Cr(V) species in the graphite. Combining magnetic resonance with other tools as microanalysis and mass spectroscopy can give a more com-
p r o d u c t t o be C;:'67(AsF;)2,3 although
( A S F ~ ) ~( ,B~a r t l e t t e t a l . ,
1978)
NMR measurements show t h e presence o f e i t h e r AsF5 or
AsF6' b u t n o t AsF3 i n " C 1 0 A ? F 5 " .
Figure 5. Analysis of charge transfer in C1&sF6 based on fluorine nuclear magnetic resonance of the solid and mass spectroscopic analysis of thermally evolved volatiles. Table IV. Previous Physical Experiments Predict a Low Hydrogen Yield. Estimations of Ionic Character in C,Cs expt
% ionic Cs
reference
optical Knight shift Mossbauer
67 55 50
Hennig ( 1 9 6 5 ) Carver ( 1 9 7 0 ) Campbell et al. ( 1 9 7 7 )
plete picture of product identity. Figure 5 illustrates an analysis of charge transfer in graphite/AsF, ("C1&F5'') based on these tools. Charge transfer is determined from the amount of AsF3 trapped, while final product identity is determined by NMR. To the extent that residual AsF, may still be present in the solid phase, such an analysis would underestimate the degree of charge transfer. Significantly, optical reflectance experiments (Fischer, 1978) support this low degree of charge transfer, although such experiments in turn might be blind to oxidative degradation to sp3 forms. As is seen in Table IV, physicists have applied not only nuclear magnetic resonance but also other techniques to determine the ionicity of alkali metals in graphite/alkali metal intercalation compounds. With respect to the resonance experiment, one is measuring the Knight shift, which is the "chemical shift" between alkali metal atom and alkali metal ion; the magnitude of this shift is proportional to the square of the ionicity of the species. For instance, the 0.29% shift reported by Carver for CBCs corresponds to a 55% fraction of ionic cesium (Carver, 1970). One chemical way to evaluate the numbers of Table IV may be to interact the graphite/alkali metal compounds with protic acids and measure hydrogen evolution (Bergbreiter and Killough, 1978). The two extremum cases are 100% ionic: C8-M+ + H 2 0 100% atomic: C8M + H 2 0
-
-
8C
C8-H+ + MOH (1)
+ MOH + 0.5H2 (2)
While Bergbreiter reported rather low yields of hydrogen (16% of that expected for reaction 2) suggesting a large ionic character, our measurements, performed in the solution state, suggest as low as a 60% ionic character, which is in harmony with results of previous physical techniques. The product of the reaction of C8M compounds with water reveals much about the intrinsic charge transfer nature of intercalation compounds. If we wrote the above eq 1 as if C8-M+were a large radical anion (analogous to naphthalene-), we might expect covalent, hydroaromatic
106
Ind. Eng. Chem. Prod. Res. Dev., Vol. 19, No. 1, 1980 I
I
I
400 G SCAN
tt
I
,
/
I
I
I
L
A ' I\
400 G SCAN
, 1 a
t
t I
Figure 7. Electron spin resonance of polycrystalline solid phase product of reaction of CsCs with water. Full scale is 400 G, frequency is 9.5 GHz, and peak is referenced to solid diphenylpicrylhydrazyl.
1
I 1
I
I
I
b
Figure 6. Cesium nuclear magnetic resonance of C8Cs (a) and solid phase C8Cs/H20 product (b). While the starting C8Cs has atomic character in the cesium ensemble, the product material does not. Dispersion mode plots are referenced to Cs,CO,(aq).
from or added to the aromatic planes. The high conductivity along the aromatic planes in intercalation compounds arises both from the change in carrier density resulting from this chemistry and from changes in t h e mobility of t h e carriers. Chemical and physical experiments are in harmony in estimating the amount of charge transfer in some compounds. Further experiments are required both to elucidate the balance between carrier concentration and mobility in conductivity enhancement and to determine the degree of competition between charge transfer and generation of sp3 hybridized carbon. Acknowledgment We thank E. Frey for manuscript preparation and K. E. Shenton for literature analysis. Literature Cited
products C8H (analogous to 1,4-dihydronaphthalene). Without "benzylic" carbons, however, the graphite anion is actually inert to water, yielding instead products of the form C164.6M+o.6(H20)2-3 (Ebert and Matty, 1978). To demonstrate this assignment, we have analyzed both C8Cs and the C8Cs/H20 product by magnetic resonance (Ebert, 1978). As shown in Figure 6, the 133Csresonance of C8Cs is shifted 0.24% downfield from Cs+, while the product of the C&S/H20 reaction is unshifted from Cs'. Thus, while the intercalated starting material is similar in ionicity to that of Carver, the product material does contain cesium species which are totally ionic. The nature of the anion may be perceived from electron spin resonance. Although starting C8Cs has no visible ESR absorption, the C8Cs/H20product does have an easily observed resonance. As seen in Figure 7 , the absorption is wide, of approximately 120 G derivative extremum separation, with this width presumably arising from spin orbit interactions. In essence, however, the observed signal is due to electrons delocalized on graphitic planes and serves to support the analogy between aromatic radical anions and intercalation compounds.
Bartlett, N., Biagoni, R. N.,McQuillan. B. W., Robertson, A. S., Thompson, A. C., J . Chem. Soc., Chem. Commun., 200 (1978). Bergbreiter, D. E., Killough, J. M., J . Am. Chem. SOC.. 100, 2126 (1978). Campbell, L. E., Montet, G. L., Perlow, G. J., Phys. Rev. B , 15, 3318 (1977). Carver, G. P., Phys. Rev. 8 ,2, 2284 (1970). Ebert. L. B., BUN. Am. Phys. Soc.. 23, 185 (1978). Ebert, L. B., Matty, L., INOR 18, 176th National Meetlng of the American Chemical Society, Miami Beach, Fla., Sept 11-17, 1978. Ebert, L. B., Ann. Rev. Mt. Sci., 6, 181 (1976). Ebert, L. B., Selig. H.. Mat. Sci. Eng.,31, 177 (1977). Flscher, J. E., Thompson, T. E., Phys. Today, 31, 36 (1978). Fischer. J. E., J . Chem. Soc., Chem. Commun., 544 (1978). Frame, H. D., Chem. Phys. Lett., 3, 182 (1969). Hennig, G. R., J. Chem. Phys., 43, 1201 (1965). Herold, A., Mat. Sci. Eng., 31, 1 (1977). Holzwarth, N. A. W.. Rabbi, S.,Girifalco, L. A,, Phys. Rev. 8,18, 5190 (1978). Selig. H., Gallagher, P. K., Ebert. L. B., Inorg. Nucl. Chem. Lett., 13, 427 ( 1977). Ubbelohde, A. R., Carbon, 14, 1 (1976). Weinberger, B. R., Kaufer, J., Heeger, A. J., Fischer, J. E., Moran, M., Holzwarth, N. A. W., Phys. Rev. Lett., 41, 1417 (1978). Zeller, C., Foley, G. M. T., Falardeau, E. R., Vogel, F. L., Mat. Sci. Eng., 31, 255 (1977).
Summary Graphite will react with oxidants and reductants to form compounds in which charge density has been removed
Presented a t the 13th Middle Atlantic Regional ACS Meeting, Monmouth College, West Long Branch, N.J., March 1979, Division of Inorganic Chemistry.
Received for review June 22, 1979 Accepted October 9, 1979