TNew preparation of superconducting alkali-metal fullerides utilizing monomethylamine as solvent S. Cooke, S. Glenis, X. Chen, C. L. Lin and M. M. Labes* Departments of Chemistry and Physics, Temple University, Philadelphia, PA 19122, USA Monomethylamine is employed as a solvent for the preparation of alkali-metal (K, Rb, Cs and mixtures) doped C60lfullerene (c60). Not only is the initial solution homogeneous, but also the extended temperature range of this solvent, as compared to ammonia, provides favourable conditions for the preparation of single phase, well defined materials with high superconducting volume fractions. The materials were characterized by dc magnetization and X-ray diffraction measurements. Since the first report in 1991 of superconductivity in alkali- metal fullerides,' substantial effort has been devoted to developing appropriate procedures for the intercalation of alkali-metal, alkaline-earth-metal and lanthanide ions into [60lfullerene (c60).It is known that superconductivity occurs in M3C60 (M =K and Rb), which crystallizes in a face-centred cubic lattice. In this structure there are three sites associated with each c60 molecule: one is a relatively large, octahedrally coordinated position, while two smaller sites are tetrahedrally coordinated. There are a number of techniques that have been investigated in an attempt to prepare this structure with high phase purity and a high volume fraction of superconductivity. In this work, the solvent monomethylamine is used in the preparation, and considerable success is achieved in obtaining well defined materials.The synthesis of superconducting alkali-metal-doped c60 can be achieved by a number of solid-state and solution methods. One of the most recent is the rapid synthesis of superconducting alkali-metal fullerides using a microwave-induced argon plasma method.2 Among the solid-state tech- niques, perhaps the most efficient are: (a) the direct reaction of stoichiometric amounts of c60 with metal vapo~r,~.~ and (b)a method which uses M6C60 as a starting material and then mixes an appropriate amount of c60 with it to obtain the desired stoichiometry.' Although the solid-state techniques provide a useful method for producing single-phase supercon- ducting compounds, there are some difficulties with the hetero- geneous nature of the starting materials leading to limited accuracy of stoichiometric control in method (a), and possible disruption of the lattice during the mixing process in method (b).Thus, there is a wide variation in the volume fraction of superconductivity observed in products of these solid-state reactions. Alternative solution techniques using toluene6,' and ammonia*>' have also been developed to comple- ment the solid-state techniques. The samples prepared using toluene as a solvent for c60 typically show quite small diamag- netic fractions (1-5%). When ammonia is employed as a solvent, the samples frequently show some depression in the superconducting transition temperature, T,.In some cases, dynamic pumping for a few days is required in order to remove any residual ammonia, presumably resident at both octahedral and tetrahedral sites, and to obtain similar T, values to those achieved for samples prepared by solid-state reaction." One of the limitations of the use of ammonia as a solvent is that the reaction is restricted to being conducted in its liquid temperature range of -78 to -33 "C. Alkylamines share the property of being solvents for both alkali-metals and c60. It is for this reason that we undertook, for the first time, the preparation of Rb3C60, Rb2CsC, and K3C60 from monomethyl- amine solutions. Monomethylamine extends the operating temperature range for conducting the reaction from -93 to -6 "C.The monomethylamine technique is found to furnish an efficient method for producing a variety of intercalation compounds which typically have large superconducting volume fractions as well as the same T, and lattice parameters achiev- able by other solid-state and solution techniques. Dissolving alkali-metals in liquid amines creates a reductive environment, the well known Birch reduction." There is, therefore, concern that preparation of the alkali-metal fullerides in amine solvents may be accompanied by protonation of c60 if any adventitious source of protons, such as a trace of water, is available. The rates of such reactions are undoubtedly very slow at the temperatures involved, and there has been no analytical evidence of substantial protonation occurring. In this work, the criteria for 'purity' are the values of the super- conducting parameters.Experimental The c60 used was Hoechst's 'Gold Grade' and was used without further purification. All alkali-metals were obtained from Ae_s_ar with purities of >99% and were used without further purification. The monomethylamine (98%) was obtained from Matheson & Co. and required further purifi- Cation as described below. Nitrogen (99.999% Grade 5) was obtained from Airco. c60 (75mg) was weighed into a glass reactor containing a quartz-encapsulated stirring bar. The reactor was then evacuated and brought into a nitrogen atmosphere glove box, where appropriate masses of alkali- metals were added. Potassium (25 mg) was added to a second reactor containing only a quartz-encapsulated stirring bar.Both reactors were then closed, removed from the glove box, and connected to a manifold system to which methylamine and nitrogen could be added. The system was purged of oxygen and moisture. The amine was then first condensed into the reactor containing the potassium. The potassium was allowed to react with the liquid monomethylamine until a permanent dark blue colour appeared, characteristic of solvated electrons. The amine was then distilled over into the second reactor containing both c60 and the alkali metal. This reactor was kept at -10 to -6 "C for 1-3 h. In the case of the ternary Rb,CsC, sample, rubidium and caesium will react completely with 75 mg of c60 in methylamine in 1-3 h, rates that are much slower than those in ammonia.Sodium and potassium require much longer times to react completely in methylamine. Additional amine was distilled over when necessary. Note that the reaction process is accompanied by a change of colour from blue to brown for Na, K and Rb and to red-brown in the case of Cs. After the J. Muter. Chem., 1996, 6(1), 1-3 reaction was complete, the amine was allowed to boil off, leaving the solid intercalation compound on the reactor walls. The reactor was sealed and vacuum-annealed in a sand bath for 1 h at 100°C. The reactor was then returned to the glove box where the product was scraped out and packed into glass or quartz ampoules for further vacuum-annealing in a Thermolyne Type 21 100 tube furnace.Rb& and Rb2CSC6, were annealed at 375 "C for 1-2 days. X-Ray diffraction (XRD) measurements were performed using an INEL powder diffractometer. The radiation used was generated from a fixed target Cu anode operated at 40 kV and 20 mA. A flat HOPG (004) monochromator selects the charac- teristic Cu-Ka,,, radiation (1.5424 A). Estimates for grain (particle) size were made using the Scherrer formula. The temperature dependence of the magnetization was measured using a Quantum Design SQUID magnetometer. Results and Discussion Fig. 1 shows the room-temperature XRD profile of Rb& prepared from monomethylamine solution. The pattern con- sists of a number of reflections, all of which can be indexed on the basis of a face-tentred cubic (fcc) unit cell with a lattic? parameter of 14.43 A, slightly larger than the value of 14.39 A obtained in a solid-state preparation.' Unreacted c60 appears to be minimal.If one examines the XRD pattern carefully, one can find shoulders on the 200, 220 and 311 reflections, which may arise from the presence of RblC60. Further annealing of the sample results in reduction of the size of these shoulders, but Goes not remove them completely. A particle size of ca. 220 A can be deduced from the Scherrer formula. Particle sizes vary with sample preparation and annealing conditions, and constitute another important variable in defining the super- conducting properties. A preparation with a Rb2CsC6, stoichiometry p!oduced a single phase with a lattice constant of ca.14.45A (Fig. 2), .......1 .........J . . .... J 10 20 30 40 50 2Bldegrees Fig. 1 Powder XRD pattern of the Rb3C6,, sample - cn which is comparable with the latticeo parameter obtained from the solid-state preparatjon (14.49 A)." The particle size is estimated to be ca. 205 A for this sample. It is known that in Rb2CSC60, the Rb ions mainly occupy the tetrahedral sites while the Cs ion occupies the octahedral interstitial site.13 Fig. 3 shows the temperature dependence of the dc magnetiz- ation of the Rb3C60 sample which was first cooled to 6 K in zero magnetic field. The zero-field-cooled (ZFC) curve was obtained by applying a magnetic field of 10G to the sample and warming it slowly to 35 K.The field-cooled (FC) curve was obtained by cooling the sample again to 6 K in the same magnetic field. The superconducting transition temperature of both ZFC and FC curves of 28-29 K is comparable to that obtained for samples prepared by direct solid-state reaction. The shielding (ZFC) and Meissner (FC) diamagnetic fractions, estimated by comparison to the ideal value of -1/4n: for a long cylinder, are 90 and 8%, respectively. Annealing the sample for 2 days at 375°C is required to produce the high diamagnetic fraction. The reaction temperature does not appear to have a significant effect on T,. In this work, Rb& was reacted at temperatures varying from -78 to -6 "C, and T,remains at 28-29 K over the entire temperature range.Fig. 4 shows the temperature dependence of the magnetiz- ation for a sample of Rb2CSC60. Superconductivity occurs with an onset temperature of 30-31 K, similar to the T, value obtained from preparation via solid-state rea~ti0n.l~The shielding and Meissner diamagnetic fractions were found to 0 0 0 H=lOG 0 Fig. 3 Temperature dependence of the zero-field-cooled (ZFC) and field-cooled (FC) magnetic susceptibility of Rb&, measured at an applied field of 10 G ma -0.006t 0 0 00 O FC DOe oDo.*o.~o*O 0 0 D 0 H=lOG 0 -0.018 e ZFC ooDI e DO* ** I.).. . . . . . . . . . I I ......... 10 20 30 40 50 2Bldegrees Fig. 2 Powder XRD pattern of the RbzCSC6, sample Fig.4 Temperature dependence of the zero-field-cooled (ZFC) and field-cooled (FC) magnetic susceptibility of RbzCSC6, measured at an applied field of 10 G 2 J. Muter. Chem., 1996, 6(1), 1-3 be 28 and 14%, respectively. There is an extended tail below T, in the magnetization curve, instead of the expected sharp drop. The same phenomenon is also observed in the shape of the curve of R~,C~C,,samples prepared by sohid-sfate reac-tions. This extended-tail behaviour may be associated with the small particle size of the powder sample and the existence of intergranular voids. Conclusions Alkali-metal fullerides can be easily obtained by precipitation from liquid monomethylamine which, because of its extended temperature range, appears to be a very useful medium for such syntheses.The technique works particularly well with the more reactive alkali metals rubidium and caesium, whereas reactions with potassium and sodium will require extended reaction times. The high volume fraction of superconductivity observed in Rb3C60 reflects the accuracy of the end-product stoichiometry.T,and lattice constan+ If the prepared materials agree quite closely with publisLd values using solid-state techniques. Helpful discussions with Professor J. E. Fischer are gratefully acknowledged. Research sponsored by the Air Force Office of Scientific Research, Air Force Systems Command, USAF, under Grant AFOSR F49620-93-1-0018. References 1 A. F. Hebard, M. J. Rosseinsky, R.C. Haddon, D. W. Murphy, S. H. Glarum, T. T. M. Palstra, A. P. Ramirez and A. R. Kortan, Nature (London), 1991,350, 600. 2 R. E. Douthwaite, M. L. H. Green and M. J. Rosseinsky, .I. Chem. Soc., Chem. Commun., 1994, 2027. 3 M. J. Rosseinsky, A. P. Ramirez, S. H. Glarum, D. W. Murphy, R. c. Haddon, A. F. Herard, T. T. M. Palstra, A. R. Kortan, S. M. Zahurak and A. V. Makhija, Phys. Rev. Lett., 1991,662830. 4 K. Holczer, 0. Klein, S-M. Huang, R. B. Kaner, K-J. Fu, R. L. Whetten and F. Diederich, Science, 1991,252, 11 54. 5 J. P. McCauley, Q. Zhu, N. Coustel, 0. Zhou, G. Vaughan, S. H. J. Idziak, J. E. Fischer, S. W. Tozer, D. M. Groski, N. Bykovetz, C. L. Lin, A. R. McGhie, B. H. Allen, W. J. Romanow, A. M. Denenstein and A. B. Smith, J. Am.Chem. Soc., 1991, 113, 8537. 6 H. H. Wang, A. M. Kini, B. M. Savall, K. D. Carlson, J. M. Williams, K. R. Lykke, P. Wurz, D. H. Parker, M. J. Pellin, D. M. Gruen, U. Welp, W-K. Kwok, S. Fleshler and G. W. Crabtree, Inorg. Chem., 1991,30,2838. 7 H. H. Wang, A. M. Kini, B. M. Savall, K. D. Carlson, J. M. Williams, M. W. Lathrop, K. R. Lykke, D. H. Parker, P. Wurz, M. J. Pellin, D. M. Gruen, U. Welp, W-K. Kwok, S. Fleshler, G. W. Crabtree, J. E. Schirber and D. L. Overmyer, Inorg. Chem., 199 1,30,2962. 8 D. R. Buffinger, R. P. Ziebarth, V. A. Stenger, C. Recchia and C. H. Pennington, J. Am. Chem. Sac., 1993,115,9267. 9 W. K. Fullagar, I. R. Gentle, G. A. Heath and J. W. White, J. Chem. Soc., Chem. Commun., 1993, 525. 10 R. D. Boss, J. S. Briggs, E. W. Jacobs, T. E. Jones and P. A. Mosier-Boss, Physica C, 1991, 243, 29. 11 A. J. Birch and G. Subba Rao, Advances in Organic Chemistry, ed. E. C. Taylor, Wiley-Interscience, New York, 1972, vol. 8, p. 1. 12 R. M. Fleming, A. P. Ramirez, M. J. Rosseinsky, D. W. Murphy, R. C. Haddon, S. M. Zahurak and A. V. Makhija, Nature (London), 1991,352,787. 13 I. Hirosawa, K. Tanigaki, J. Mizuki, T. W. Ebbesen, Y. Shimakawa, Y. Kubo and S. Kuroshima, Solid State Commun., 1992,82,979. 14 K. Tanigaki, T. W. Ebbesen, S. Saito, J. Mizuki, J. S. Tsai, Y. Kubo and S. Kuroshima, Nature (London), 1991,352,222. Paper 5105304H; Received 8th August, 1995 J. Muter. Chem., 1996, 6( l), 1--3