J. Phys. Chem. 1993,97, 12924-12927
12924
Ca Fullerene as Carbon Source for Diamond Synthesis G. Bocquillon' and C. Bogicevic CNRS, Laboratoire de Physico-Chimie des MatCiaux, 1 Place Aristide Briand, 921 95 Meudon Cedex, France
C. Fabre and A. Rassat ENS, Laboratoire d' Activation Molhxlaire, 24 rue Lhomond, 75231 Paris, France Received: July 30, 1993; In Final Form: September 10, 19936
The Cm fullerene was used as the carbon source to carry out the catalyzed synthesis of diamond, at 6.7 GPa and temperatures between 1200 and 1850 'C; the catalysts used were Ni, Co, and a Co alloy. Under these conditions, the graphite used as the carbon source gave a diamond yield of about 90%; the yield with Cm depended on the catalyst. With Co, about the same quantity of diamond was obtained from Cm as from graphite, whereas almost no diamond was synthesized when pure Cm was used as the carbon source with N i and the Co alloy as catalysts. The diamond yield recovered to a high level with these two catalysts when a small quantity (about 10%) of graphite was added to the Cm fullerene. This effect of the addition of graphite to a non-graphite carbon in the synthesis of diamond is different from that observed with amorphous carbons. In all these experiments under high pressure and high temperature with Ni, Co, or Co alloy, the Ca fullerene was destroyed and transformed into diamond, graphite, or poorly crystallized carbon.
2000
Introduction Many previous experiments show clearly that the structure of the carbon starting materials is a determining factor in direct and catalyzed diamond synthesis.'" Since the discovery that bulk quantities of fullerenes, the new family of carbon allotropes, can be obtained in workable quantities,' it is of interest to synthesize diamond from the C a fullerene, the most abundant of the family. The C bounding angle is 120° in graphite and 109O in diamond, and in the Cw molecule this angle has the two following values 120° and 108O, which can perhaps facilate the synthesis of diamond. Up to now few studies under pressures were performed using Cm as starting material and they seem somewhat inconsistent. Yoo et ai. shocked isentropically C a films to 69 GPa and about 2200 K and obtained diamondlike metastable carbon phases.* Regueiro et al."' claimed that C a transformed to diamond under a rapid, nonhydrostatic compression at pressures of 20 f 5GPa and room temperature. However, all other compression experiments at ambient temperature up to 35 GPa never succeeded in producing diamond.12-'4 On the other hand Webb and QadriIs tried to synthesize diamond from a mixed Ca/C70 fullereneusing a Ni catalyst at 6GPa between 1100 and 1500 ' C and observed either graphitization or a new X-ray diffraction powder pattern (for the higher temperatures), but no diamond. It is these last experimentswhich were again attempted using a pure Cm fullerene as carbon material and transition metals or alloys as catalysts. Experimental Section
The catalyzed synthesis of diamond were performed under high pressure and high temperature, from a carbon source (generallygraphite) in the presence of transition metals or alloys, in a P-Tregion located between the eutectic melting temperature and the graphitdiamond equilibrium line16J7(see Figure 1). Our experiments were carried out in a belt type apparatus; the high-pressure cells with indirect heating are illustrated in Figure 2. The indirect heating allows to use smaller samples and provides a better thermal homogeneity and a better temperature reproducibility than the direct heating in which the heating electrical Abstract published in Advance ACS Absrracrs, November 1, 1993.
0022-3654/93/2091- 12924$04.00/0
1 aoo
/
GRAPHITE GRAPHITE
____----Y
1600
I I-
1400
eutectic N i - C
DIAMOND
1200
5.0
5.5
6.0
P
6.5
7.0
- GPa
Figure 1. Pressure-temperature region of catalyzed diamond synthesis (TfNi: melting curve of Ni).
current goes through the carbon4atalyst mixture. The sample was constituted by a carbon pellet between two metal catalyst disks; it was surrounded by boron nitride so that the graphite of the heater could not react with it. The experiments were performed at 6.7 GPa and temperatures between 1200 and 1850 'C (see Figure 1). The pressure at high temperature was determined based on the melting curve of silver.'* The silver melting temperature was measured in a separate run, using a metal-sheathed Ni-Cr/NiA1 thermocouple of 0.5-mm outer diameter placed in the center of a small silver sample in the synthesis cell. In this run the same loading and heating cycle as those of synthesis experiments was used. The temperature read was corrected of the effect of pressure on the thermocoupleemf using the equation given by Getting and Kennedy19 and the pressure was determined by the following (0
1993 American Chemical Society
Cm as Source for Diamond Synthesis
The Journal of Physical Chemistry, Vol. 97, No. 49, 1993 12925
TABLE I: Most Significant Experiments of Diamond Synthesis Carried out at 6.7 CPa’ exDt 209
211
213
215
celltypeA TIOC) 1410 caia1yst I11 carbon C60
A 1410 I 207
A 1410 I11 207
B
diamond e yield
91
90
216
219
217
A B A B B 1410 1410 1410 1410 1410 1410 I I I I11 I11 I C60 207 C60 C60 207 C 6 0 + 10% 207 91 92 c 92 91 91
221
A 1410 I11 C60+ 10% 207 91
222
326
328
A A A 1380 1410 1410
330
332
333
A 1410 1 1 C60
A 1520 1 C60
A 1780 I C60
A 1820 I C60
A A 1820 1410 I I1 207 207
e
e
e
89
I
1
207
amC a m C + 10% 207 0 e 82
91
1
329
334
335
92
($1 a Catalyst
I, 11, I11 are Ni, Co and Co alloy respectively. Carbons 207, C60, am C are graphite, Cw fullerene and amorphous carbon respectively.
-- 5
I _ _
and the yield was calculated by the ratio of diamond mass over initial carbon mass. When the quantity of diamond was very small, the presence of diamond was detected by the optical microscopic observation, confirmed by X-ray diffraction and indicated by c in the Table I. Results
-
1 ) graphite heater 2 ) boron nitride 4 ) carbon source - 5 ) graphite
-
3) catalyst
Figure 2. High-pressure cell for diamond synthesis (detail of the sample environment).
equation18 based on the results of Mirwald et al.:20
+
T = a b P + cp + d@ where T i s the silver melting temperature in “C at the pressure P i n GPa and a = 959.3, b = 64, c = -0.56, and d = -0.12. The temperatures in the synthesis cell were measured in a separate run using a Ni-Cr/Ni-AI thermocouple corrected for the effect of pressure on its emf and a calibration curve T =f(W), temperature versus heating power, was established and extrapolated for the highest temperatures. Experiments were performed using different carbons and catalysts. The catalysts used were nickel, cobalt, and a cobalt alloy, the carbons were a spectographic graphite Gr207, purchased at “Le Carbone Lorraine”, which contained less than 50 ppm ash, a Cm fullerene, and an amorphous carbon termed amC. To obtain pure Cm, a mixture of approximately 80% of c 6 0 and 20% of C70 was extracted from a soot with boiling toluene; Cm was then separated from the mixture by column chromatography on neutral alumina followed by solvent drying2’ and was submitted to a further heat treatment in vacuum a t 170 OC for 6 h. The amorphous carbon amC was prepared by pyrolysis of trindan at 900 “C and it showed no Debye-Scherrer X-ray diffraction pattern. Most of the diamond synthesis runs were performed in the high pressure A type cell (Figure 2); the runs with C ~ were O always done with the same pressure P, temperature T, and time t cycle as a run with graphite, which was used as reference for the diamond yield. In some other experiments, the B type cell was used to be sure that the P and T conditions were exactly the same for graphite and Cm, and unexpected results were obtained (see next section). The most significant runs are given in Table I. The recovered samples were chemically treated to dissolve the metal catalyst and to eliminate the unreacted carbon. After the elimination of the metal by nitric acid or aqua regia the diamonds were generally visible under the optical microscope. The carbon was oxidized either by a mixture of HNO3 and H2SO4 or by a mixture of CrO2 and H2S04and in some cases the two treatments were used successively. The recovered diamonds were weighed,
First it must be underlined that almost all the experiments were done well inside the diamond stability region, far from the graphitediamond equilibrium line and with a P-T-t cycle which gives a very high yield of diamond when graphite is used as carbon material. Therefore small variations of pressure or temperature from one experiment to another should not affect the yield. In these runs the recovered diamonds were blackandslightly sintered. The results obtained with the Cm were very different according to the catalyst used, Co, on the one hand, or pure Ni and Co alloy, on the other hand. Diamond Synthesis from Pure Cm and Ni or a Co Alloy. The experiments carried out with pure Ni or Co alloy gave no or few diamonds (see Table I, experiments 209, 216, and 330). The results obtained with Ni are similar to those obtained by Webb and QadriI5 with a mixed C,50/C7,3. N o experimental details are known concerning their work, but one can assume that they are classical experiments of solvent-catalytic diamond synthesis and therefore of the same type as ours. The fact that they obtained no diamond and that we found a very small amount can be due to the pressure which was a slightly higher in our experiments or froma great ability todetect thediamond produced. Therefore, it is possible to conclude that the pure c 6 0 fullerene reacts in the same way as a mixed Cm/C,o fullerene for the Ni-catalyzed synthesis of diamond under high pressure and high temperature. Diamond Synthesis from Cmand Ni or a Co Alloy in the F’resence of a Small Amount of Graphite. To verify that the lack of diamond in these runs did not come from the greater compressibility of Cm fullerene than that of graphite22 creating a great pressure loss in the cell, experiments with the B type cell were undertaken (experiments 215 and 217 in Table I). In this B type cell, the synthesis was carried out at the same time with graphite and Cm to be sure that the P and T conditions were exactly the same for the two carbons, and then if the pressure was too low to synthesize diamond, graphite would not transform either. The unexpected result of these experiments was that a 90% diamond yield was obtained from c 6 0 fullerene as from graphite. This suggests that graphite has an initiator effect in the diamond synthesis from Cm and this effect was confirmed by the experiments 219 and 221 in A cell where the carbon sample was composed of a 5 mg of c 6 0 fullerene pellet between two fine layers of 0.25 mg of graphite (in all experiments the carbon sample weighed 5.5 mg), the graphite was in direct contact with the catalytic metal. Diamond Synthesis from Pure Cm and Co. In contrast with these experiments, experiment 329 with pure Co as catalyst and Ca alone gave a diamond yield of 80% (close to that obtained from graphite) showing that the behavior of pure Co is very different from that of pure Ni or Co alloy. This difference is not
Bocquillon et al.
12926 The Journal of Physical Chemistry, Vol. 97, No. 49, 1993
understood at present: it could come from a greater reactivity of Co with Cm, a greater ability of Co to form compounds with Cm, or a difference in the degree to which the various metals wet the Cm. All these factors could play a role in the catalyticdiamond ~ynthesis23.~~ and further experimentswith Co and C ~are O needed.
Discussion The initiator effect of graphite observed above in the experiments with pure Ni and Co alloy could perhaps by explained in the followingway. In the solvent+atalyst method, the diamonds are obtained in a P-T region located between TE,the melting temperature of the carbon-metal eutectic and TH, the transformation temperature from diamond to graphite. These two temperatures depend on the nature of the carbon material and of the metal. When graphite is used, it diffuses into the metal in the solid state during the heating period and the melting occurs at the eutectic temperature Te, the dissolution continues until the solution is supersaturated and then diamonds crystallize.23 When Cm is used as carbon starting material with pure Ni or Co alloy, it can be assumed that the metal and the C ~ did O not react in the solid state and the metal could not melt because the experiments were done at temperatures lower than the melting temperature of the pure metal. The Cm is a large molecule and it could not diffuse into the metal unit it was broken down. In the experiments 215 and 217 where B cell were used or 219 and 221 where a small amount of graphite was inserted between the catalyst and the Cm one could conclude that the graphite allowed the solid solution of carbon in the metal to form and the melting could occur at the eutectic temperature. Then the Cm could decompose in the melted metal, dissolve like the graphite and recrystallize as diamond. To verify the preceding hypothesis the experiments 332 and 333 were done in a P-T region located above the melting point of pure Ni25 and below the diamond-graphite transition temperature. It is to be noted that in this region the error in temperature can be large owing to the double extrapolation: of the pressure effect on the thermocouple emf and of the powertemperaturecalibration curve. However, experiment 334 proves that the conditions of pressure and temperature are inside the diamond stability area, and the aspect of the sample after the experiment shows that the Ni melted. In this experiment, no more diamond was obtained than in those performed with c60 alone. Thus, it must be concluded that graphite not only lowers the melting temperature of Ni but also has an initiator effect. This initiator effect of graphite has already been emphasized by Vereshchagin et al.3 and Shipkov et a1.26 to explain the dependence of diamond synthesis on the nature of the starting carbon. Studies of carbon solutions in molten iron and nickel have shown that at carbon concentrations above 1.5 wt %, microclusters of graphite packets floating in the liquid metal can be observed together with the true atomic solution. These microclusters of graphite in the stability area of diamond can transform into diamond seeds by a solid state transformation. When the diamond seeds reach a size greater than the critical size, the diamond phase can grow. If the seed size is lower than the critical size, diamond cannot form even in its thermodynamical stability field and thus nongraphitic carbons must be graphitized before diamond formation can occur. An attempt was made to determine if, under the conditions of our synthesis, this initiator effect of graphite can be observed with other carbons, such as amorphous carbons, known to give no diamond when they are used as starting carbons under similar conditions. This is what motivated the experiments 326 and 328 which were performed in the same manner and conditions as the experiments216 and 2 19, respectively,and gave a totally different result. As with the other amorphous carbons, the amC carbon alone gives no diamond. When a small amount of graphite is added (experiment number 328) very few diamonds (impossible
to weigh) were formed, certainly from the graphite only. Previous experimentsof Tsuzuki et al.5made with other amorphous carbons show no initiator effect of graphite and even some amorphous carbons have a poisoning effect on the graphitdiamond transformation. This and our results with amC carbon show that the effect of an addition of graphite to Cm or to other non-graphite carbon for thecatalytic synthesisof diamond was entirelydifferent and certainly linked to the properties of Cm. If the mechanism described previously by Vereshchagin and Shipkov (mentioned above) is valid for our conditionsof synthesis, it is possible that a small amount of graphite accelerated the graphitization of CMat high pressure and temperature. Recent work on the rate of decomposition of Cm and C70 in shows that pure Cm is more stable than a mixture of Cm and (270,and it was thought that possibly the ruptured C70 cages trigger the breakup of Cm cages. Perhaps, in the same way, the presence of some graphite triggered the graphitization of Cm, in the diamond synthesis experiments. A more detailed study of the behavior of Cm alone and with graphite at high pressure and high temperature and of the product of decomposition are necessary to elucidate this point.
Conclusions The synthesis of diamond using Cm as the carbon source has been carried out at 6.7 GPa and temperatures lying between 1200and 1850 OC withCo,Ni,andaCoalloyassolvent-catalysts. Under these conditions the synthesis performed with graphite gave a very high yield of diamond. The synthesis using Co as catalyst gave almost the same yield of diamond with either Cm or graphite as the carbon source. When pure Ni or Co alloy was used as the catalyst nearly no diamond was formed when pure Cm was used as the carbon source, whereas the diamond yield recovered to the level of that obtained with pure graphite as soon as a small amount (less than 10%)of graphite was added to Cm. This indicates that graphite has an initiator effect in the transformation of Cm into diamond with some catalysts. In all these experiments with Co, Ni, and Co alloy under high pressure and high temperature, the structure of Cm was destroyed and it was transformed either to diamond, graphite, or a poorly crystallized carbon which has a diffuse X-ray diffraction pattern.
Acknowledgment. Thanks are due to Dr. C. Loriers-Susse for her interest in this work and many fruitful discussions and to F. Clerc for the chemical treatment of the samples. References and Notes (1) Wentorf. R. H. J . Phys. Chem. 1965,69,3063. (2) Kasatochkin, V. I.; Shterenberg, L. E.; Slesarev, V. N.; Nedoshivin, Yu. N. Sw. Phys. Dokl. 1971,IS, 930. (3) Vereshchagin, L. F.; Kalashinikov, Ya. A.; Shalimov, M. D. High Temp.-High Press. 1975, 7 , 41. (4) Tsuzuki,A.; Hirano, S.I.; Naka, S. J . Mater. Sci. 1984,19,1153. (5) Tsuzuki,A.; Hirano, S.I.; Naka, S. J. Mater. Sci. 1985,20, 2260. (6) Higashi, K.; Onodera, A. Physica 1986, 139,IIOB,813. (7) KrHtschmer, W.; Lamb, L. D.; Fostiropoulos, K.; Huffman, D. R. Nature 1990. 347. 354. (8) Yo0,’C. S.; Nellis, W. J.; Sattler, M. L.; Musket, R. G. Appl. Phys. Lett. 1992,61,3, 273. (9) Regueiro, M. N.; Monceau, P.; Hodeau, J.-L. Nature 1992, 355, 237. (10) Regueiro, M. N.;Abello,L.; Lucazeau,G.; Hodeau, J.-L. Phys. Rev. B 1992,46, 15, 9903. (1 1) Regueiro, M. N. Adv. Mater. 1992,4,6, 438. (12) Snoke, D. W.; Raptis, Y. S.;Syassen. K. Phys. Rev. B 1992,45,24, 14419. (13) Moshary, F.; Chen, N. H.; Silvera, I. F. Phys. Rev. Lett. 1992,69, 3, 466. (14) Szwarc, H.. private communication. No diamond was obtained by the compression of C ~ up Q to 35 GPa with shear stresses superimposed by an anvil rotation. (15) Webb, A. W.; Qadri, S.B. Bull. Am. Phys. Soc. 1992,37, 1, 406.
Ca as Source for Diamond Synthesis (16) B0venkerk.H. P.;Bundy,F.P.;Hall, H.T.;Strong,H.M.; Wentorf, R. H. Nature 1959, 184, 1094. (17) Bundy, F. P.; Strong, H. M.; Wentorf, R. H. Chem. Phys. Carbon 1973, IO, 213. (18) Bocquillon, G.; Clerc, F.; Loriers-Susse, C. High Press. Res. 1993, 11, 177. (19) Getting, I. C.; Kennedy, G. C. J . Appl. Phys. 1970, 41, 11, 4552. (20) Mirwald, P. W.; Getting, I. C.; Kennedy, G. C. J . Geophys. Res. 1975, 80, 151. (21) Kriza, G.; Ameline, J.-C.; Jerome, D.; Dworkin, A.; Szwarc, H.; Fabre, C.; Schiitz, D.; Rassat, A.; Bernier, P.; Zahab, A. J . Phys. I (France) 1991, I , 1361.
The Journal of Physical Chemistry, Vol. 97, No. 49, 1993 12927 (22) Fischer, J. E.; Heiney, P. A,; McGhie, A. R.; Romanow, W. J.; Denenstein, A. M.; McCauley, J. P., Jr.; Smith, A. B.,I11 Science 1991,252, 1288. (23) Wedlake,R. J. TheProperriesofDiamod,Field, J. E.,Ed.;Academic Press: London, 1979; pp 501-535. (24) Burns, R. C.; Davies, G. J. The Properties of NaturalandSynthetic Diamond; Field, J. E., Ed.; Academic Press: London, 1992; pp 395-422. (25) Cannon, J. F. J. Phys. Chem. ReJ Data 1974, 3, 3, 781. (26) Shipkov, N. N.; Kalashnikov, Ya. A., et al. Tsoetnye Mertally 19%0, 10, 68. (27) Chibante, L. P. F.;Pan, C.; F’ierson, M. L.;Haufleer, R. E.; Heyman, D.Carbon 1993, 31, 185.
