Alternative isomeric structures of C70 have been investigated using semiempirical and ab initio quantum chemical techniques. As in the case of C60, these isomers are characterized by the presence of pentagonal rings adjacent to each other. The lowest energy alternative isomer of C70 has only one pair of edge-sharing pentagons and lies ≈ 1.4 eV higher in energy than the ground state. This energy difference is smaller than that for the lowest energy alternative isomer of C60 which contains two pairs of adjacent pentagons and lies ≈ 2.0 eV higher in energy than its ground state.

We report on the preparation of C60/C70 by the arc method. Evidence of monofunctional C60O, and C60-CH2 has been observed by mass spectrometry. Dependent on their preparation conditions, high resolution transmission electron microscopy of crystals consisting of mixtures of C60/C70 show different structure. Needle-like crystals grown from toluene show superlattice formation while single crystals obtained by sublimation have fcc structure. UV-VIS-absorption spectra in toluene are compared to those of films of various thickness and crystallinity.

The optical absorption spectra for thin films of sublimed C60 and C70 are presented. Both transmission spectroscopy and photothermal deflection spectroscopy are used to obtain the spectrum over a wide energy range (0.4 to 6.2 eV) for the same film. The optical gap varies for the different fullerenes. The value for C60 is ∼ 1.6 eV and the corresponding transition appears to be optically forbidden given the weakness of the absorption. Structure is observed on the “gap edge” which can be attributed to vibronic transitions. The gap region can be described in terms used for amorphous semiconductors, having features such as an Urbach edge, and sub-gap defect absorption. The similarity suggests that the nature of fullerene thin films may be analogous to amorphous silicon and carbon films. These films have both a molecular and an amorphous semiconductor nature.

The photoluminescence (PL), X-band ODMR. and LESR of C60 films and C60 isolated in a toluene/polystyrene glass matrix (C60:T/PS) is described. The delocalized triplet LESR and ODMR of C60:T/PS are similar to previously reported LESR. In films, however, the ODMR indicates that the delocalized triplet is larger and distorted by neighboring molecules. In addition, another, localized triplet, and a narrow PL-enhancing line at g = 2.0017 ± 0.0005 are observed. These features are similar to those observed in several π-conjugated polymers. The former is attributed to a localized triplet of size similar to a 5- or 6-membered ring, tentatively on a face adjacent to another C60 molecule; the latter is believed to result from intermolecular polaron recombination.

The present paper reports on positron lifetime measurements on Ceo/C 70 fullerite powder as a function of temperature and quasi-hydrostatic pressure in order to give an estimate of the positron annihilation site in the fullerite lattice. A single-component positron lifetime of 402 ps is observed which significantly decreases under quasi-hydrostatic pressure. From this and the soft intermolecular properties of the fullerites one can conclude that the positron is annihilated rather in the intermolecular space than inside the fullerene molecules. However, positron trapping at lattice defects, which are observed by high-resolution electron microscopy, cannot be ruled out.

The early stages of the microwave assisted chemical vapor deposition of diamond on fused silica and silicon substrates were examined with Raman spectroscopy and scanning electron microscopy. Grain size as a function of time was determined for both substrates. Grains formed on fused silica were larger, with smoother growth surfaces than those formed on silicon substrates under the same conditions. For deposition on silica, the particle morphology changes from cuboid to cubo-octahedral for deposition times between 5 and 15 minutes. Also, the glass surfaces were etched during the pretreatment and deposition stages. These results are discussed in terms of mass transport limited growth, and chemical interactions between the gas-phase and the substrate surface.

The early stages of a gas phase pretreatment for diamond nucleation in an oxy-acetylene flame were investigated. The pretreatment involved a low oxygen-to-acetylene ratio (Rf = 0.93) performed at 15 mm from the torch on a scratched Si substrate. The nature of the carbon species deposited was analyzed using X-ray photoelectron spectroscopy, Auger electron spectroscopy, and scanning electron microscopy. The pretreatment process produced diamond crystals after 30 seconds and a complete film in the center of the deposition area after 180 seconds. Furthermore, the pretreatment process kept the initial Siob2 layer less than the thickness generated by the control conditions, (Rf = 0.97, d = 10mm). The success of the pretreatment was attributed to changes in the flame structure and chemistry from the control conditions.

Diamond films were grown using a computer controlled oxy-acetylene torch that allowed precise programming of O2 and C2H2 gas profiles. This enabled more accurate control of the diamond growth chemistry. With this torch, systematic and complex variations in gas chemistry were applied during growth and the resulting film morphology and film quality was investigated. Plan view scanning electron microscopy (SEM) revealed predominately (111) faceting as a result of a sawtooth variation in the O2:C2H2 gas ratio from 0.97 to 1.00. Raman spectroscopy indicated mostly sp3 bonding, with little non-diamond component present. Furthermore, crosssectional SEM showed small grains during normal growth (O2:C2H2 = 0.97), whereas grain size increased significantly after onset of a sawtooth variation in the gas ratio.

Diamond thin films were grown on a scratched silicon crystal surface by a novel CVD technique. The substrate was exposed to a bombardment of sputtered carbon atoms from a graphite target in a helium DC glow discharge, and subsequently exposed to atomic hydrogen generated by a hot tungsten filament. The resulting diamond films were characterized by Raman spectroscopy and SEM. Deposited film quality, and growth rate are presented as a function of carbon flux, and atomic hydrogen flux. The observed increase in growth rate with atomic hydrogen indicates that a surface reaction mechanism may be responsible for growth. The saturation of the utilization of carbon confirms that the diamond growth is probably a surface reaction. Based on this work we propose that the growth of diamond films in the sequential CVD reactor is most likely governed by surface reactions, and that the necessity of gas phase precursors can be precluded.

Growth rates of homoepitaxial (110), (111), and (100) diamond films were experimentally determined, for the first time, in a hot filament reactor using methane and carbon tetrachloride as the carbon source. Methane concentrations from 0.07 % to 1.03 % in H2 were studied at a substrate temperature of 970°C. Growth rates were found to be crystal-face dependent with respect to methane concentration, being linear or first order for the (100)-orientation, sublinear for (110), and sigmoidal for (111). The observed growth kinetics of (111) suggest the viability of an acetylene mechanism for (111), along with the methyl radical mechanism at methane concentrations above 0.73%. CC14 concentrations from 0.06% to 0.69% in H2 were also investigated at a substrate temperature of 970°C. Growth rate behavior was similar to that of methane for all three crystal faces.

The temperature dependence of the growth rates was also crystal-orientation dependent. At substrate temperatures above 730°C, growth rates are thought to be mainly transport limited, yielding effective activation energies of 8±3, 18±2, and 12±4 kcal/mole for (100), (110), and (111) orientations, respectively. At substrate temperatures below 730°C, growth rates are thought to be surface reaction rate-limited, with an overall effective activation energy of 50±19 kcal/mole for the three crystal-orientations studied.

A study was performed on polycrystalline copper versus that of Si(100) utilizing a negative substrate bias to enhance diamond nucleation. The biasing pretreatment and subsequent growth of the diamond were performed via microwave plasma chemical vapor deposition and the initial stages of nucleation were characterized by in-vacuo surface analysis. The biasing pretreatment step proved to have a tremendous influence on the nucleation density pertaining to Si(100) substrates, however, the nucleation density on polycrystalline copper was only increased slightly. A highly graphitic surface coverage of roughly 10A was evident on the copper substrates prior to the detection of diamond and was quite stable in thickness in as early as 15 minutes of biasing. The Si(100) substrates, however, were characterized by the formation of a carbide with some form of nondiamond carbon present on the surface throughout the biasing pretreatment.

Diamond films prepared by microwave plasma chemical vapor deposition, using a gas mixture of methane and hydrogen with ethanol, were formed on silicon substrates. Highresolution transmission electron microscopy was employed to characterize the microstructure at interface regions. It was found that the diamond crystals were grown on an amorphous carbon layer. Twins and stacking faults were observed at the regions interfaced with the amorphous carbon layer, suggesting that the defects may already exist in the nucleation stage and at the very first stage of growth. Also, some diamond nuclei embedded in the amorphous layer were observed.

The early stages of diamond thin film growth under various growth conditions have been investigated using Scanning Atomic Force Microscopy and Scanning Electron Microscopy. A series of experiment with various source gas combinations, substrate surface preparation methods and different substrates have been carried out to determine the effect on the diamond nucleation. The highest early nucleation density was obtained with CH4(1%)/H2 process among other gas combinations used. It is suggested that the atomic oxygen content should be kept minimum during the initial nucleation period of CVD diamond. It has been observed that diamond nucleation does not necessarily require diamond seeding dust. The silicon substrate polished with alumina powder showed diamond nucleation along the alumina scratched lines. In the process of diamond crystal growth, screw dislocations and stacking faults have been observed.

Diamond films grown in the microwave plasmas of CO(7–8%)-O2(0–2.2%)-H2 systems in the range of 130–750°C were characterized by scanning electron microscopy, Raman spectroscopy, and cathodoluminescence (CL) studies. The films grown in the CO-O2-H2 system had much better crystallinity than those grown in the CO-H2 system. This was because oxygen extremely purified diamond films by suppressing polyacetylene inclusion, and prohibited the vacancy formation in the crystallites. These oxygen functions have indicated the possibility that high quality diamond films (FWI-tM of the diamond Raman peak =4.0–4. lcm−1) close to natural diamond (FWHM=3.0cm−1) were obtained in the CO(8%)-O2(2.2%)-H2 system between 400 and 750°C. Though crystallinity deterioration occurred at 130°C, the obtained film (FWHM=10.2cm−1) in the CO(8%)-O2(2.2%)-H 2 system was of good crystallinity comparable to those (FWHM=7–21cm−1) grown by conventional CVD processes and gas systems between 590 and 1327°C. The CO-O2-H2 microwave plasma was concluded to be one of the best environment for the low temperature growth of highly purified diamond films of good crystallinity.

A mass spectrometric analysis of heavy hydrocarbons (HHCs) during hot-filament CVD diamond growth was performed together with in situ monitoring of the growth rate. Many HHCs were detected and tentatively identified. Of all HHCs studied, only diacetylene shows good correlation with the diamond growth rate under various deposition conditions. Its possible role is discussed.

Resonance-enhanced multiphoton ionization (REMPI) has been used to detect the methyl radical CH3 within 0.5 mm of the substrate during CVD diamond growth. A strong dependence of the CH3 REMPI signal near the surface on substrate temperature is observed, which is not seen further from the surface. Below 1000 K, the observed temperature dependence may be characterized by an activation energy of approximately 4 ± 1 kcal/mole. The cause of the methyl depletion at low temperatures is not yet clear, but may be due to either gas-phase recombination near the surface or surface chemistry. The same qualitative behavior is observed for different substrate-filament distances and for gas compositions from 0.5% to 5% CH4 in H2.

Adsorption and subsequent surface processes of hydrogen and hydrocarbon species on the diamond C(111)-(2×1) bare surface have been investigated by AES, XPS, LEED, TDS and laser non-linear spectroscopies (SHG and SFG) under ultrahigh vacuum condition. Extremely low coverage of atomic II causes the transition of the hare (2×1) surface to the (1×1) structure. Exposure of the. bare (2×1) surface to CIIx species from hot filament activated methane leads to the formation mainly of CH3 species. The conversion of the surface structure from (2×1) to (1×1) with exposure does take place but not as easily as for hydrogen. These studies are now being extended to adsorption of F by exposures to XeF2 There arc two stages of saturation for 17 adsorption on the C(111)-(2×1) surface. This is interpreted to indicate that there are at least two adsorption sites for the F atoms. Whereas adsorption of II atoms at coverages less than 5% of a monolayer causes the surface lattice transition from (2×1) to (1×1) structure, the (2×1) structure is retained at all levels of fluorine adsorption. However the C(111) (1×1) surface with over half a monolayer of II adsorption behaves quite differently with respect to F adsorption.

To investigate some of the fundamental differences between halogen and hydrogen assisted diamond film growth we have performed several calculations related to the <100> diamond surface. The models used in these investigations include ten-layer periodic slabs of free standing fluorinated diamond films as well as isolated clusters [C21F6H20]. For purposes of comparison, we have also performed calculations on models of the hydrogenated <100> surface. The calculations are performed within the density-functional framework using LCAO and LAPW computational methods. We have considered two geometries of a monofluoride surface. The first surface, best described as an ideal l×l surface with a monolayer of ionically bonded fluorines, exhibits a metallic density of states in contrast to a 2×l reconstructed surface with chemically bonded fluorines that is found to be insulating. We compare theoretical carbon core level shifts with experimental values and discuss growth models based on these surface calculations.

Surface properties - stability and reconstruction - of clean and hydrogenated diamond (100) have been studied by real temperature molecular dynarnic (MD) simulations using an approximate density functional (DF) theory expanding the total electronic wave function in a minimal basis of localized atomic valence electron orbitals (LCAO - ansatz). The clean surface is highly unstable against a spontaneous dimerization resulting in a 2×1 reconstruction. Atomic hydrogen in the gas phase above the top surface at all temperatures and H2 molecules approaching the center of the dimer bond at room temperature are reactive in breaking the dimer π-bonds forming a monohydrogenated surface which maintains a stable 2×1 structure but with elongated surface C-C dimer bonds remaining stable against continuing hydrogen supply. The dihydrogenated surface taking a 1×1 structure, because of steric overcrowding dynamically becomes unstable against forming a 1×1 (alternating) di-, monohydrogenated surface. As first elementary reaction processes which may be discussed in relation to diamond growth we studied the thermal adsorption of CH3 and C2H2 onto a clean 2×l reconstructed (100) diamond surface.

The desorption kinetics of H2, HD, and D2 from a CVD-grown diamond sample have been measured using temperature programmed desorption. The sample was exposed to hydrogen and deuterium at ∼ 100 K. Molecular hydrogen does not react with the diamond surface but atomic hydrogen chemisorbs. The sample is then heated to 1500 K at a rate of 6 K/s. The hydrogen desorption spectrum from these polycrystalline specimens is surprisingly simple, consisting of a single peak near 1270 K for all coverages studied. The width of the desorption peak is ∼ 100 K. Neither the width nor the peak maximum shift with increasing hydrogen coverage. There is no isotope effect to the desorption. No other desorption products (water, small hydrocarbons) are detected. Kinetic parameters can be extracted from these spectra, and these can be compared to hydrogen desorption from natural diamond single crystal surfaces. These parameters can be used to model diamond growth processes.