1877
Anal. Chem. 1980, 52, 1877-1881
A significant advantage peculiar to this industry is that practically all of the radiologically "hot" sample components are separated from the solution that is actually injected into the ion chromatograph. This means that the instrument does not become contaminated with radionuclides to the extent that it would with direct sample injection. This significantly reduces personnel exposure to radiation and in practice saves the trouble and expense of trying to clean up or replace (and correctly dispose of) the "hot" columns.
LITERATURE CITED (1) Small, H.; Stevens, T. S.;Bauman, W. C. Anal. Chem. 1975, 4 7 , 1801-1809. ( 2 ) Mulik, J. D.; Sawicki, E. Environ. Sci Techno/. 1979, 13, 804-809. (3) Nordman, Joseph "Qualitative Testing and Inorganic Chemistry"; Wiley: New York, 1957; p 376. (4) Drucker, K Z.Elektrochem. 1912, 78,236.
RECEIVED
for review May 15, 1980. Accepted July 11, 1980.
Determination of Molecular Weight Distributions of Polymerized Petroleum Pitch by Gel Permeation Chromatography with Quinoline Eluent R. A. Greinke" and L. H. O'Connor Union Carbide Corporation, Carbon Products Division, Parma Technical Center, P.O. Box 6 1 16, Cleveland, Ohio 44 10 1
The technique of gel permeation chromatography (GPC) using quinoline eluent has been applied to measure molecular weight distributions in thermally polymerized petroleum pitches. The molecules present In the toluene-soluble fraction of polymeric pitch eluted ideally in quinoline. A study of low molecular weight model pericondensed polynuclear aromatic hydrocarbons, which elute nonideally in most GPC eluents, revealed that the elutlon volume in quinoline was closely related to the maximum length of the molecule. The molecules present in the toluene-insoluble-quinoline-soluble fraction of polymerized pltch eluted nonideally, and the elution volume was a function of the concentration of these molecules. Ideal elution behavior was obtained for these molecules after reductive hydrogenatlon. A linear GPC polymerized pitch calibration curve was obtained for the 450-2000 molecular weight range. The technique is applicable for obtaining molecular weight distributions of petroleum pitch mesophases and semicokes which have quinoline-insoluble contents approaching 100 %
.
Pitch, defined as the solid fusible residue obtained from the pyrolysis of organic materials, generally has been produced from coal, petroleum, and pure compounds. Pitches are complex in constitution and are usually composed of mixtures of polynuclear aromatic hydrocarbons (PAHs) and heterocyclic compounds. Pyrolysis or carbonization of pitch leads to polymerization of these compounds resulting in polymeric pitches a n d finally infusible coke (1-3). T h e aromatic and heterocyclic components in pitch are quite similar in chemical structure but differ in molecular size and shape. The molecular weight of these components reflects the extent of polymerization and is, therefore, an important property of pitch materials. Quantitative characterization of pitches and polymeric pitches for molecular weight distribution by gel permeation chromatography has generally been plagued by nonideal elution behavior of the smaller molecular weight constituents (4-10) and by insufficient solubility of the higher molecular weight constituents (11). The nonideal elution behavior, particularly noticeable for the pericondensed polynuclear aromatics, has been observed in tetrahydrofuran (8),toluene 0003-2700/80/0352-187750 1.OO/O
( 9 ) , benzene (9, IO), and methylene chloride ( 4 , 5 ) . This behavior, sometimes attributed to adsorption ( I O ) , generally results in a nonlinear calibration curve a t low molecular weights for petroleum residuals (12,13)and polymeric pitches (14). Bergmann et al. (9) have reported that the anomalous elution behavior of pitch molecules disappears when 1,2,4trichlorobenzene is used as the gel permeation chromatographic (GPC) solvent. However, interfering negative peaks were observed in 1,2,4-trichlorobenzene for the GPC elution of pitch samples (14). Lewis and Petro (14) reported the first use of GPC to obtain molecular weight distributions of polymerized pitches. In their procedures the polymeric pitch was separated into pyridinesoluble (PS) and pyridine-insoluble (PI) fractions. After lithium reduction of the PI fraction, separate GPC curves were obtained on both fractions by using toluene eluent. The use of toluene, however, limited the GPC evaluation to no more than 70% by weight of a polymeric pitch containing 50% PI. T h e lithium-reduced P I fractions contained significant amounts of toluene insolubles (TI) and the T I portion of the PS fraction was also insoluble in toluene. In search for a better solvent, Tillmanns et al. (15) employed quinoline as a GPC eluent for pitches. In our present work, the application of quinoline as a GPC solvent for polymeric petroleum pitches is investigated. We have found that the low molecular weight pitch molecules, including the pericondensed molecules, elute ideally in quinoline, that a linear calibration curve of the logarithm of the molecular weight vs. elution count is obtained for pitch molecules with a molecular weight range from 450 to 2000, that polymeric pitches or semicokes with initial quinolineinsoluble (QI) contents near 100% now can be completely evaluated by GPC after chemical reduction of the samples, and that basic kinetic pitch polymerization data can be obtained directly from GPC curves since one chromatogram is now obtained on the entire sample rather than the two chromatograms previously obtained on the P I and PS fractions (14).
EXPERIMENTAL SECTION Instrumentation. Gel P e r m e a t i o n C h r o m a t o g r a p h y . The GPC experiments were performed with B Waters Associates Model 200 GPC instrument, equipped with a differential refractometer. 1980 American Chemical Society
1878
ANALYTICAL CHEMISTRY, VOL. 52, NO. 12, OCTOBER 1980
Quinoline produced by Koppers Corp. was distilled in glass. Five GPC columns were packed with Styragel and were of the following pore sizes: 1000 A, 500 A, 250 A, 60 A, and 60 A. Quinoline was pumped at 1.0 mL/min and syphoned every 2.5 min (one elution count equals 2.5 mL of quinoline). The following set of temperatures was used: injector, 87 "C; oven, 87 "C; degasser, 90 "C; refractometer base plate, 86 "C; refractometer heat exchanger, 93 "C; and syphon, 68 "C. Oven temperatures above 100 O C resulted in some column degradation. Vapor Pressure Osmometry. The Corona Wescan Model 232A molecular weight apparatus was used for the measurement of the number-average molecular weight of GPC fractions. The solvent employed was pyridine. Reductive Hydrogenation of Polymeric Pitch. The TI fraction from approximately 2 g of polymerized pitch was isolated by Soxhlet extraction. The TI fraction was solubilized for GPC evaluation by reductive hydrogenation with lithium in ethylenediamine. The ethylenediamine was purified by refluxing with sodium for 48 h followed by slow distillation in the presence of sodium. Aproximately 0.8-1.0 g of the TI fraction of the pitch was lithium reduced according to standard procedures (16, 17). To prepare the sample for GPC, we mixed the reduced TI fraction and the unreduced toluene-soluble (TS) fraction in the weight ratio based on the TS and TI extraction percentages and on the refraction index response differences between the T S and reduced T I fractions. (The Development of the Analytical Method section contains additional discussion of this procedure.) Aproximately 0.14.5 g of this mixture was added to quinoline and heated a t 80 "C for 30 min. The hot solution was filtered through a Krueger filter under nitrogen pressure just prior to injection into the GPC. Reductive Ethylation of Polymeric Pitches. Reductive ethylation has been used for solubilization of coal (18). The results indicate that higher coal solubility can be obtained by this method than by lithium reduction (18). Similarly improved solubility was also obtained for polymerized pitch in this study. This additional treatment was applied to those polymeric pitch samples that could not be completely solubilized in quinoline with lithium reduction. GPC Calibration. The GPC calibration between 450 and 2000 molecular weight units was achieved by fractionating a polymerized petroleum pitch on the GPC and subjecting the fractions to vapor-phase osmometry to determine number average molecular weight. The polymeric pitch sample was a mixture of the TS fraction and the hydrogenated TI fraction. Reductive ethylation was used for calibration of the highest molecular weight fractions in the polymeric pitch. The quinoline was removed from each collected fraction prior to osmometry by placing them in a tared glass culture dish and by heating for 30 min at 86 "C in a vacuum oven until dry. The evaporation procedure was repeated twice after redissolving the sample in pyridine. The repeated dissolution in pyridine was necessary to remove the last traces of quinoline because the presence of trace amounts of quinoline in the sample lowers the osmometer number average molecular weight. The GPC calibration for molecular weights below 450 was achieved by using a series of PAH model compounds. (The calibration procedure could have been simplified by using quinoline as the osmometer solvent. However, the high boiling point temperature of quinoline was found to be beyond the capability of the osmometer.) Heat Treatment of Pitches. The polymerized petroleum pitches were prepared by heating the starting pitches in a nitrogen atmosphere at 400 "C. RESULTS AND DISCUSSION Development of t h e Analytical Method. P r e p a r a t i o n of P o l y m e r i c Pitch f o r G P C Evaluation. Nonideal GPC elution behavior was observed for the higher molecular weight pitch molecules in quinoline. GPC curves of a polymeric petroleum pitch containing no QI but significant amounts of TI are shown in Figure 1. The weight of pitch injected into the GPC for curve B was three times larger than that for curve A. (Both curves were normalized to produce equal areas.) Even though the same sample was injected, the shapes of the curves were considerably different. For the larger sample size, the resulting GPC curve showed a higher concentration of
W
z l
(i:
PI 2i LL
a
A
u
I
I
& I
1
75
70
65
60
55
L\
53
G P C ELUTIOU C 3 U N T S (INCREASING MOLECULAR
45 hElGb.'a)
Nonideal elution behavior of an unreduced polymeric peboleum pitch containing toluene insolubles: curve A, 30 mg of sample; curve B, 90 mg of sample (GPC areas normalized). Figure 1.
larger molecules. Altgelt (19) and Dawkins (20) also reported this same phenomenon for other GPC systems and attributed the faster-than-expected elution of molecules to a gel-solute repulsion. They (19,20)found that the degree of repulsion increased with increasing molecular weight and that the GPC curves also were influenced by concentration. This nonideal elution behavior of the molecules present in the TI-QS fraction of pitches was eliminated after reductive hydrogenation. Reductive hydrogenation or reductive ethylation also resulted in solubilizing the QI portion of polymeric pitches. T h e TI fraction of polymeric pitches had to be separated before reduction, because the TS fraction, if hydrogenated, was soluble in the wash water used to remove the lithium salts from the product. For example, lithium reduction of the TS fraction of a polymeric petroleum pitch resulted in only a 38% product yield. The loss of pitch molecules present in the TS fraction resulted in a n inaccurate GPC curve. Therefore, polymeric pitch containing both TS and T I fractions was first extracted, the T I fraction was hydrogenated, and the hydrogenated T I fraction was remixed with the TS fraction for GPC evaluation. The response of the refractive index detector to the hydrogenated T I fraction was found to be considerably different than to the nonhydrogenated TS fraction. For example, the refractive index responses of TS fractions from four polymeric petroleum pitches were compared to the hydrogenated T I fraction from these four polymeric pitches. T h e GPC areas of the four TS fractions compared to the areas of four hydrogenated T I fractions were 3.02, 2.95, 2.85, and 2.91 times larger when equal weights of the fractions were injected into the GPC. An average detector response correction factor of 2.9 was used when remixing the hydrogenated T I fraction with the TS fraction of a polymeric petroleum pitch. G P C Retention Times of Model P A H s . Many authors have observed that model PAHs exhibit anomalous GPC elution behavior on gels (4-20). A particular problem has been the nonideal elution of the pericondensed PAHs, such as coronene (9). The elution behavior of a number in PAHs in quinoline was studied, and t h e results are shown in Table I. T h e elution times of t h e PAH is quinoline were calculated relative to benzene. The relative retention times of the same model compounds eluting in tetrahydrofuran (8) are also shown in Table I. In quinoline, the pericondensed PAHs, such as benzcoronene, coronene, and pyrene, elute more nearly according to their size. In tetrahydrofuran, these compounds are held back and elute at approximately the same retention time as benzene. T h e holdup of the pericondensed PAH is worse in toluene, where coronene elutes at a retention time of 1.13 relative to benzene (9). The near ideal elution behavior of the pericondensed compounds in quinoline may be related to the polarity of this solvent. Adsorption of solutes on gels
ANALYTICAL CHEMISTRY, VOL. 52, NO. 12, OCTOBER 1980
1879
Table I. Relative GPC Retention Times of Model Aromatic Hydrocarbons in Quinoline and Tetrahydrofuran E l u t i o n Time Quinoline
(b) (c) Relative E l u t i o n Time Tetrahydrofuran
Relative (b)
(a) C omp ound
Structure
Size,
'8,
m-quinquephenyl
21.8
0.79
0.73
p-quaterphenyl
20.2
0.82
0.81
benzcoronene
14.3
0.83
0.99
p-terphenyl
16.0
0.84
0.86
tet racene
14.3
0.85
0.93
chrysene
14.3
0.86
0.93
car on e n e
12.3
0.88
1.01
phenanthrene
11.9
0.89
0.95
anthracene
12.3
0.90
0.95
biphenyl
11.9
0.91
0.92
pyrene
11.6
0.93
0.98
naphthalene
10.1
0.97
0.97
toluene
8.0
1.0
0.98
benzene
7.0
1.0
1.0
(a) (b) (c)
Maximum L e n g t h o f ~ o l e c u l e ( 8 ) R e t e n t i o n t i m e i n t h e two s o l v e n t s r e l a t i v e t o b e n z e n e Data f r o m r e f e r e n c e ( 8 )
is strongly influenced by solvent polarity (20, 21). The data of Table I indicate that the smaller molecules in polymerized pitch or the molecules present in the TS fraction should elute ideally in quinoline. GPC Calibration Curve for Polymeric Pitch. As described in the GPC calibration section, the calibration was performed with GPC fractions collected from unreduced (TS fraction) and reduced (hydrogenated TI fraction) polymeric petroleum pitch. A logarithmic plot of the osmometric number average molecular weight of the GPC fractions vs. elution
counts is shown in Figure 2. For molecular weights less than 450, the elution times of the model PAHs in Table I are plotted. Least-squares fit of the osmometer results of the GPC pitch fractions indicate that the data between molecular weights 450 and 2000 could be described by log (mol wt) = 4.314 - 0.0224(counts) (1) The exclusion limit of the columns occurs a t elution count 45 (approximate molecular weight of ZOOO), and the permeation limit occurs at count 75 (approximate molecular weight of 450).
1880
ANALYTICAL CHEMISTRY, VOL. 52, NO. 12, OCTOBER 1980 I
IO30
-
c
l
70
60
5C
G P C ELUTION
80
a 9C
CCUNTS
Figure 2. GPC calibration curve for polymeric petroleum pitch (molecuhr weights above 450 are osmometric values, molecular weights below 450 are model PAH; eluent quinoline).
A linear plot between molecular weights 450 and 2000 and a clearly visible permeation limit below a molecular weight of 450 indicates ideal elution behavior. The molecular weight values of the calibration curve of Figure 2 are considerably lower than those reported for the GPC separation of petroleum residuals (12, 13). T h e differences may be related to the solvents employed for obtaining osmometric Mn values. Speight and Moschopedis (22) have shown that osmometric A& values obtained for asphaltenes were considerably lower in more polar solvents than in nonpolar solvents, such as benzene, toluene, or tetrahydrofuran. The use of pyridine as an osmometer solvent appears to have eliminated molecular aggregation resulting in accurate Mn values. T o verify the accuracy of the calibration curve, we hydrogenated the T I fraction of a polymeric pitch and subjected the fraction to evaluation by both the osmometer and the GPC. A,, values of 1010 and 1037 were measured, respectively. Column P e r f o r m a n c e of Polymeric P i t c h Separation. T h e performance of GPC columns for the separation of a specific material can be quantitatively evaluated in terms of the product of two column parameters (23, 24): u , the standard deviation of the GPC peak of a monodispersed polymer, and D2, which is related to the slope of the linear portion of the chromatographic calibration curve. The uDz column performance criterion also can be applied to polymeric pitch separation as follows. Equation 1 can be rearranged to
;=c
LE-ICIU
ccurrs
Figure 3. GPC curve of polymeric petroleum pitch, 1 h heat treatment at 400 OC, pyridine insolubles = 10%. w m
z
0, m c
$1
n
XI
an
counts = 192.6
-
44.6 log mol wt
1
85
8C
75
I
73 65 60 G P C ELUTION C 3 J U T S
1
55
5C
15
Figure 4. GPC curve of polymeric petroleum pitch, 63 h heat treatment 400 O C , quinoline insolubles = 73%.
at
The calculated values of u and D 2 were used in eq 6 and 7 to calculate the relative errors, Mn* and Mw*(errors only caused by column band broadening), between molecular weights calculated from the experimental calibration curve, and molecular weights calculated from the infinitely resolved curve, (23).
(mexptl,
(2)
Balke and Hamielec (24) state that the GPC calibration curve can be represented as counts = (1/D2) In D,
-
1/D2 In mol wt
(3)
By using eq 2 and 3 t o solve for D z , one obtains
D 2 = O.O516/count (or 0.0206/mL)
(4)
Two PAHs, p-quaterphenyl and benzcoronene, were employed as "monodispersed" polymer standards t o obtain the n (25) value u
= A/h2n
(5)
where A = area of the PAH standard peak and h = height of the PAH standard peak. The elution volumes of the two standards were approximately 76 counts, a value near the permeation limit of the columns. Ideally, higher molecular weight standards should be employed since u could be underestimated (26). Unfortunately, higher molecular weight PAHs are not available. An average u value of 0.723 counts (or 1.80 mL) was obtained for the two standards. This value does not appear to be smaller than t h a t normally expected for the Styragel columns (27).
For the polymeric pitch system Mn* = -7 X and MW* = f 7 X T h e unusually small error is related to the very small value of D 2 or to the slope of the calibration curve. Compared to other polymer systems separated by GPC (231, the molecular weight of pitchlike molecules does not change rapidly as a function of elution volume. The pitch molecular weight separation range of the Styragel columns is 450-2000 or approximately 0.4 decade (Figure 2), which is considerably less than t h a t observed for other polymer systems such as polystyrene. Applications. By using the recommended procedure, after solubilization by ethylation, one can obtain molecular weight distributions of entire polymeric pitches which initially contain quinoline insolubles approaching 100 % . The applications, therefore, include mesophase pitches (28) through semicokes. Figures 3-5 illustrate GPC molecular weight distributions of petroleum pitch that had been heated for 1, 63, and 263 h a t 400 "C. The GPC curves visually show for the first time the
ANALYTICAL CHEMISTRY, VOL. 52, NO. 12, OCTOBER 1980
1881
that the slow rate-determining step is the formation of a free radical (33). The use of GPC, with quinoline as an eluent, coupled with the proper chemical sample preparation, has allowed the accurate evaluation of molecular weight distributions of polymeric petroleum pitches and has also provided a unique tool for directly monitoring pitch polymerization.
ACKNOWLEDGMENT The authors thank M. J. Kovelan and H. W. Leggon for assistance in developing the GPC calibration curve. The helpful discussions with I. C. Lewis and R. Didchenko, Union Carbide Corp., and A. E. Hamielec, McMasters University, are also acknowledged.
LITERATURE CITED
75
70
65
60
SPC E L - I S N
55
5C
45
C3UhTS
Figwe 5. GPC curve of polymeric petroleum pitch, 263 h heat treatment OC, quinoline insolubles = 100% (sample was reductively
at 400
ethylated).
progression of pitch polymerization to a semicoke. The GPC curve shown in Figure 5 indicates that many molecules in the semicoke are large and are excluded from the pores of the gel. The polymeric petroleum pitches of Figures 4 and 5 contain 100% mesophase. T h e reproducibility of the procedure was evaluated by subjecting a polymerized petroleum pitch to repeated analysis. The coefficient of variation of five determinations was 1.9% for M,, and 0.3% for the dispersion, M w / M n . The kinetics of the polymerization process occurring during carbonization have been monitored by solubility measurements (29, 30). The formation of pyridine insolubles a t the mesophase stage was shown to follow first-order kinetics (29, 30). Since GPC curves can be obtained for the entire polymerized pitch, the kinetics of pitch polymerization can now be measured by GPC. The disappearance of the 400-700 molecular weight molecules in a petroleum pitch a t 400 "C was followed by GPC between 3 and 18 h reaction time. The reaction obeyed first-order kinetics and the experimentally measured rate constant was 0.7 X s-'. The kinetics of the disappearance of the pyridine-soluble fraction in the same petroleum pitch was also measured a t 400 OC between 3 and 18 h reaction time. A first-order rate constant of 1.4 X 10-j s-l was calculated. T h e larger rate constant observed for pyridine-insoluble formation suggests t h a t factors other than polymerization, such as the cosolvent effect (31),and changes in the carbonhydrogen framework, such as aromatization and loss of side chains (31, 32), also accelerate the formation of pyridine insolubles. The first-order reaction for petroleum pitch polymerization has been interpreted mechanistically by assuming
Lewis, 1. C. Polym. Prepr., Am. Chem. Soc.,Div. Polym. Chem. 1973. 14, 380. Blayden, H. E. Transitions Non Radiat. Mol., R e m . SOC.Chim. Phys., ZOih, 1969 1969, 15. Singer, L. S. Transitions Non Radiat Mol., Reun. SOC. Chim. Phys., 20th. 1969 1969. 21. Oelei, HI H.; Latham, D. R . ; Haines, W. E. Sep. Sci. 1970, 5 , 657. Oelert, H. H.; Weber. J. H. ErdoelKohle, Erdgas, Petrochem. 1970, 2 3 , 484. Hirsh, D. E.; Dooley, J. E.; Coleman, H. J. Rep. Invest.-U.S., Bur. Mines I.Q ?.A.., 7. 0 7. 5 ...
Thompson, R. E.; Sweeney, E. G.; Ford, D. C. J . Polym. Sci.. Polym. Chem. Ed. 1970. 8 . 1165. Edstrom, T.; Petro, 6 . A. J . Polym. Sci., Part C 1968, 21, 171. Bergmann. I. G.;Duffy, L. J.; Stevenson, R. B. Anal. Chem. 1971, 4 3 , 131. Popi, M.; Fahnrich, J.; Stejskal, M. J . Chromatogr. Sci. 1976, 14, 537. Lewis, I. C.; Ddchenko, R. Carbon 7 6 , Int. Carbon Conf., Prepr., Znd, 1976 1976, 385. Albaugh, E. W.; Query, R. C. Anal. Chem. 1976, 48. 1579. Reerink, H.; Lijzenga, J. Anal. Chem. 1975, 4 7 , 2160. Lewis, I. C.; Petro, B. A. J . Polym. Sci., Polym. Chem. Ed. 1976, 14, 1975. Tillmanns, F. H.; Ulsomer, W.; Pietzka. G. Carbon 76. I n f . Carbon Conf., Prepr., Znd, 1976 1976, 385. Reggel, L.; Friedel. R. A.; Wender, I. J . Org. Chem. 1957, 2 2 , 891. Brooks, J. D.; Silberman, H. Fuel 1962, 4 1 , 67. Sternberg. H. W.; Delle Donne, C. L.; Dantages, P.; Moroni, E. C.; Markby, R. E. Fuel 1971, 50, 732. Altgelt. K. H. Sep. Sci. 1970, 5 , 777. Dawkins, J. V. J . Polym. Sci., Polym Phys. Ed. 1976, 14, 569. Oelert, H. H. Fresenius' 2. Anal. Chem. 1969, 2 4 4 , 91. Speight, J. G.;Moschopedis, S. E. Fuel 1977, 5 6 , 344. Yau, W. W.; Kirkland, J. J.; Bly, D. D.; Stocklosa, H. J. J. Chromatogr. 1976, 125, 219. Balke, S . T.; Hamielec, A. E. J. Appl Polym. Sci. 1969, 15, 1381. James, A. T.; Martin, A. J. P. Analyst(L0ndon) 1952, 77, 915. Yau, W. W.; Kirkland, J. J.; Bly, D. D. "Modern Size-Exclusion Liquid Chromatography": Wiley: New York, 1979; p 109. Yau, W. W.; Kirkland, H. J.; Bly, D. D. "Modern Size-Exclusion Liquid Chromatography": Wiley: New York, 1979; p 108-1 13. Brooks, J. D.; Taylor, G. H. I n "Chemistry and Physics of Carbon"; Walker, P. C., Ed.; Marcel Dekker, New York, 1968; Voi 4, p 263. Honda, H.; Kimura, H.; Sanda. Y.; Sugawara. S.;Faruta, T. Carbon 1970, 8. 141. Singer, L. S.;Lewis, I. C. Carbon 1978, 16, 417. Farcasiu, M.; Mitchell, T. 0.; Whitehurst, D. D. "Abstracts of Papers", 172nd National Meeting of the American Chemical Society, San Francisco, CA, Aug 1976; American Chemical Society: Washington DC, 1976; FUEL 21. Schlosberg, R. H.; Gorbaty, M. C.; Aczel. T. J . Am. Chem. SOC.1978, 100, 4188. Yamada, Y.; Oi, S.; Tsutsui, H.; Kitajima, E.; Tsuchitoni, M.; Kakiyarna, H.; Honda, H. Abstracts, 12th Conference on Carbon; American Carbon Society: Pittsburgh, PA, 1975; p 271.
RECEIVED for review April 22, 1980. Accepted July 8, 1980.