Search

Since the initial discovery of fullerenes nearly a decade ago [1], material scientists have focused attention on the possibility of encapsulating one or more metal atoms inside these spheroidal carbon frames. The experimental realization of macroscopic quantities of endohedral metallofullerenes (Am@C2n, n=30-55) in the early 1990's has heightened interest in developing this new class of tunable materials with possible electronic and/or optical applications [2,3]. They have been characterized by a number of spectroscopic techniques, for example, scanning tunneling microscope [4,5], EXAFS [6,7] and x-ray diffraction and electron microscopy [8]. However, low production yields and purification difficulties have hampered the development of this new class of materials. The soluble product distribution usually consists of high levels of the empty-caged fullerenes C60, C70, C84 and decreasing levels of the higher fullerenes, while the endohedral metallofullerene fraction usually constitutes less than 1% of the total soluble yield. Furthermore, the endohedral metallofullerene fraction consists of molecules with different numbers of metal atoms encapsulated (m=1-3), cage sizes (C2n) and isomers of the same mass (e.g., Er2@C82). The purification process is further complicated by the chemical reactivity of several endohedral metallofullerenes [9] in aerobic environments. For several years, we have been involved in a collaborative effort to develop methodology for detection, isolation, and characterization of endohedral metallofullerenes. The focus of the present study is on fullerenes encapsulating metals from Group II1b, (Sc@C2n, Y@C2n, and La@C2n) and the lanthanide series metal (Er@C2n).

Amorphous carbon films are (a-C:H) of interest because of their useful physical properties. They are extremely hard and chemically inert, resisting degradation by both acids and alkalis. They are insoluble and can be conformably coated onto virtually any substrate. These properties make the films ideal protective coatings on magnetic disks and tools. We have studied several thin (one to two micron) films prepared by plasma enhanced chemical vapor deposition with varying radiofrequency fields strengths to determine structural differences at the atomic level. Several properties of the films, such as hardness and wear rate, are dependent on deposition power. We have found that the sp2/sp3 ratio increases with increasing deposition power. Thus, films that are harder are more “graphitic” and less “diamondlike”. The films studied here contain 11–16 atomic percent hydrogen, most of which is associated with sp3 carbon sites. At least two distinct phases of hydrogens exist. Variable temperature studies reveal that, in contrast to amorphous hydrogenated silicon, proton linewidths in carbon films are temperature dependent, suggesting some molecular motion is present at room temperature.

We report here the arc-production and spectroscopic characterization of fullerene-encapsulated metal atoms and metal-atom clusters. In particular, both solution and solid-stateelectron paramagnetic resonance (EPR) spectra of LaC82', YC82', SCC82' and Sc3 C82 have been obtained. Additional species containing rareearth atoms and clusters have been produced. The results suggest, for example, that the three scandium atoms in Sc3C82 form a molecule In the shape of an equilateral triangle — as was previously suggested for Sc3 molecules Isolated In a cryogenic rare-gas matrix. The spectraof the MC82 species (M =La, Y, Sc) exhibit small hyperfine couplings and g-values close to 2, suggesting that they can be described as a +3 metal cations within – 3 fullerene radical anion cages. Sc2C2n species — the most abundant metailofuilerenes In the scandlum-fullerene mass spectrum — are EPR-silent, even though Sc2 is EPR-actlve In a rare-gas matrix at 4.2K. A broader Implication of this work Is that production of macroscopic quantities of metal-containing fullerenes may make possible the fabrication of exotic new structures with regular arrays of metal atoms or clusters Isolated in fullerene molecules, resulting in new typesof host/guest nanostructured materials.

We recently demonstrated that C60 and C70, as well as other fullerenes, can be deposited and accumulated on surfaces using laser ablation of graphite In an Inert gas atmosphere. After learning of the work of Kratschmer et al. Indicating the presence of C60 In carbon soot, we showed that samples consisting almost exclusively of C60 and C70 can be sublimed from such soot. Vibrational Raman spectra of C60 and C70 were obtained from these samples. The C60 spectrum Is consistent with the calculated spectrum of Buckmlnsterfullerene, and ?he strongest three lines can be assigned on the basis of frequency and polarization. The NMR spectrum of dissolved C60 was then obtained, and found to consist of a single resonance, establishing the Icosahedral symmetry of this molecule. STM images of the C60 molecules on a Au(111) crystal face show that these clusters form hexagonal arrays with an Intercluster spacing of 11.0 A and are mobile at ambient temperature. Distinctly taller species evident In the arrays are believed to be C70 clusters. Vibrational Raman and infrared spectra have also been obtained for separated C60 and C7 0. Five years have passed since Kroto, Heath, O'Brien, Curl and Smalley, inspired by the work of R. Buckminster Fuller, realized that there was a most elegant solution to the problem of explaining the extraordinary inertness and stability observed for 60 atom pure-carbon clusters 1 . The solution proposed was the beautifully symmetric arrangement familiar as the soccerball pattern - a truncated icosahedron consisting of 12 pentagons and 20 hexagons. As Kroto would write: "This structure necessitated the throwing of all caution to the wind ... and it was proposed immediately; after all, it was surely too perfect a solution to be wrong." 2 The experimental evidence required to substantiate this intuitive leap proved difficult to obtain, however. While many ingenious experiments indirectly supported the picture that not only C60 but all of the observed carbon clusters with an even number of atoms (n > 24) can exist in the form of 2-d closed nets of 12 pentagons and ( n/2 - 10 ) hexagons 3 , doubts persisted, and some investigators suggested alternative structures for these clusters 4 , 5. Over the last few months the gap between intuition and experiment has been closing rapidly. With the discovery that macroscopic samples of fullerenes can be produced and accumulated 6-8 it has become possible to rapidly obtain a wealth of new information about them 6, 7, 9-13. In this paper we describe the course of our work on carbon cluster deposition, vibrational spectroscopy, NMR spectroscopy, and scanning tunnelling microscopy during this period of time. Early this year, we began an effort to develop a new approach to studying carbon clusters by trying to deposit them on surfaces using laser ablation of graphite under a static inert gas atmosphere 8. A highly sensitive surface analytical mass spectrometer 14 was used to analyze the deposited material. The observed mass spectrum, shown in Figure 1, showed a broad range of fullerenes, with C60 and C70 particularly prominent.

We have investigated the structure and dynamics of C60 and C70 with 13C NMR spectroscopy. In solution, high-resolution spectra reveal that C60 has a single resonance at 143 ppm, indicating a strained, aromatic system with high symmetry. This is strong evidence for a C60 “soccer ball” geometry. A 2D NMR INADEQUATE experiment on 13C-enriched C70 reveals the bonding connectivity to be a linear string, in firm support of the proposed “rugby ball” structure with D5h symmetry, and furnishes resonance assignments. Solid state NMR spectra of C60 at ambient temperatures yield a narrow resonance, indicative of rapid molecular reorientation. Variable temperature T1 measurements show that the rotational correlation time is ∼ 10−9s at 230 K. At 77 K, this time increases to more than 1 ms, and the 13C NMR spectrum of C60 is a powder pattern due to chemical shift anisotropy (tensor components 220, 186, 40 ppm). At intermediate temperatures a narrow peak is superimposed on the powder pattern, suggesting a distribution of barriers to molecular motion in the sample, or the presence of an additional phase in the solid state. A Carr-Purcell dipolar experiment on C60 in the solid state allows the first precise determination of the C60 bond lengths: 1.45 and 1.40Å.