Diamond films were deposited by chemical vapor deposition using a radio- frequency induction plasma operating at 130 torr. Linear growth rates of polycrystalline diamond films ranged from 18 to 37 μm h-1. For a fixed substrate temperature of 1000°C the input methane-hydrogen ratio was varied from 2% to 10%. Over this range the resulting film morpologies changed from faceted ball-like structures to well-faceted diamond, then to non-faceted balls, and for the well- faceted films increases in methane-hydrogen ratio caused the film texture to shift toward the <100> direction. During these experiments gas sampled through an orifice in the center of the substrate was delivered to a gas chromatograph for measurement of stable hydrocarbon species. As the input methane-hydrogen ratio was increased the measured methane-acetylene ratio decreased. The gas chromatograph measurements showed marked differences from measurements made for an RF reactor with somewhat different flow geometry operating at atmospheric pressure.

Growth rates for homoepitaxy of diamond (100) and (111) by hot-filament chemical vapor deposition were measured via in situ Fizeau interferometry and the surface morphologies were subsequently characterized by atomic force microscopy (AFM). (100)-oriented growth from 0.5% CH4 in H2 exhibited pure Arrhenius behavior, with an activation energy of 17±1 kcal/mol, up to a substrate temperature of 1100°C. Addition of oxygen to the feed gas resulted in an increased growth rate below 900°C, a maximum growth rate between 900 and 1000°C, and etching (of diamond) above 1050 - 1100°C. However, the presence of oxygen apparently had less effect on the surface morphology than did the (100)-to-(111) growth rate parameter α, determined directly from the relative growth rates of (100) and (111) substrates mounted side by side. During homoepitaxial growth from 0.5% CH4 in H2 at 875°C of ca. 1-micron-thick films,α = was 2.2 without oxygen and 1.3 for growth with 0.14% O2. The (100) film grown with α = 2.2 was quite smooth, while that with α = 1.3 was covered by numerous hillocks and penetration twins. AFM analysis revealed surprisingly little difference between the (111) films despite the considerable difference in α. Implications of these results for the growth mechanism are discussed.

A simple model of the competition between the growth of diamond and amorphous, graphitic carbon has been extended to include oxygen within the mixture of gases in a chemical vapour deposition (CVD) process. The model is based on linear rate equations which describe the growth and removal of layers of carbonaceous material on the surface of a substrate placed within the CVD reactor. The main effect of oxygen is to tie up a proportion of the available gas-phase carbon in the form of CO, which reduces the flux of reactive carbonaceous material onto the surface. Hydrogen atoms can remove carbonaceous material from the surface, and in addition can perform the key task of cementing such groups into place on the surface to make diamond. Starting with these simple assumptions, it is possible to account for the shape of the diamond domain in the C-H-O phase diagram, namely the boundaries with the domains of graphitic growth and no-growth.

In order to study the effects of gas phase transport on the growth of hot-filament chemical vapor deposited (HFCVD) diamond, crystallites were selectively grown on a pre-nucleated, oxygen plasma patterned silicon wafer. Growth rate differences across the substrate were observed from scanning electron micrographs. The deposition system was then modeled with a three dimensional finite difference scheme that employed gas phase diffusion of a single growth limiting species from the hot filament to the surface coupled with a first order surface reaction. The variations in the predicted gas phase concentration directly above the surface were adjusted to match the observed growth rate differences through the Dahmk6hler number which was then used to calculate a first order surface reaction coefficient. This value was compared to published reaction coefficients for the abstraction of a surface H-atom by a gas-phase H-atom.

A method has been developed for producing large area single-crystal diamond plates, suitable for optical and electronic applications. It starts with orienting and closely packing a set of diamond seed crystals with (001) top faces. This assembly, or mosaic, is then joined by a single-crystal overgrowth using a CVD process. A number of assembling techniques have been tested for compatibility with homoepitaxial diamond growth by hot filament assisted CVD and/or growth and etching by the acetylene-oxygen combustion flame. Furthermore a two-step process is described. First an initial layer (20-50/μm) is deposited by hot filament assisted CVD at a low growth rate in order to bridge the gap between the seeds. Subsequently the fast growth rate of the acetylene-oxygen combustion flame is employed to increase the layer thickness (>250,μm). It was found that both the basic mosaic process as well as the two step process can produce a single-crystal diamond layer on top of mosaics consisting of seed crystals with well aligned crystallographic directions. The width of the gaps between the seed crystals (up to 25 μm) was found to be less critical, while the orientation of the side faces and the direction of the misorientation (i.e. the step flow direction) seem not to effect the successful overgrowth. Apart from the alignment of the seed crystals the most severe problem, which has to be overcome in order to obtain one single-crystal overgrowth, is the occurrence of penetration twins in the joint regions. The largest mosaic structure -up to now- overgrown by CVD consists of seven seed crystals and has a surface area slightly in excess of 1 cm2

Development of a diamond homoepitaxial deposition process that utilizes water and-ethanol at a growth temperature of ∼600°C is described. Topographies are excellent, and etch-pit densities (EPD) are in the 106 cm-2 range when growth is done on type Ia C(100) substrates.-This process has been used to epitaxially join diamond single crystals that were bonded in close-proximity to each other. This process of “tiling” single crystal diamonds in close proximity in-order to manufacture a large-area diamond single crystal template is also described. Specially-prepared diamonds that have had their faces and edges oriented to { 100} were coated with-heteroepitaxial Ni, then pressed onto a Si wafer while being heated in an inert gas atmosphere.-The resulting bond is excellent; thereby permitting our 600°C diamond deposition process to-epitaxially join the diamonds. A diamond wafer cutting technology has been addressed using a-specific sequence consisting of: ion implantation, homoepitaxial diamond growth, annealing, and-contactless electrochemical etching. This “lift-off” method of cutting has thus far resulted in a 2mm×O.5mm×17.5μm transparent, synthetic, free-standing, single crystal diamond plate being-fabricated. Raman spectroscopy and EPD show the plate to be comparable to our best-homoepitaxial diamond.

Growth of large area, single or almost single crystal diamond is of great importance to the electronics industry. In this work, single crystal diamonds were implanted with C+ ions, inducing a subsurface damage layer in the diamond lattice. Homoepitiaxial diamond films were then grown on the implanted crystals using a microwave plasma CVD reactor. Films grown on on-axis substrates were dominated by large numbers of hillocks, renucleation and penetration twins, while miscut substrates exhibited stepflow growth. The homoepitaxial layers were separated from the substrate by a water-based etch which selectively attacks the subsurface damage layer of the diamond lattice. The films were analyzed by Raman scattering, scanning electron microscopy (SEM), optical microscopy, photo- and cathodoluminescence, and x-ray diffraction. CVD growth on adjacent, oriented substrates formed a single, continuous diamond layer. The resulting homoepitaxial film quality, orientation, defect density and it's relationship to the underlying substrates were compared at various points on the surface, particularly the region which overgrew the gap between different substrates.

Diamond films were grown on (100) and (110) oriented natural diamond substrates by hot filament assisted chemical vapor deposition (HFCVD) to thicknesses of 7 to 100μm. Raman spectroscopy was used to measure tensile stresses of up to ∼2GPa for some of the (110) films. The development of stress was attributed to the incorporation of impurities (Re, Mo, and H). Impurity concentrations were greater at the interface than through the film thickness. Up to∼11% H and 50ppm Re were measured in the films with secondary ion mass spectrometry (SIMS). Homoepitaxial diamond films were further characterized using a seven crystal high resolution x-ray diffraction system. This new characterization tool allowed the separation of the effects of mosaicity from those of variation in lattice parameter.

We have carried out diamond deposition by microwave plasma-assisted chemical vapor deposition (MPCVD) on planar and nonplanar diamond substrates at temperatures between 850° and 2000° C. The nonplanar and planar substrates were the polished (100) surfaces of type Ia natural diamond anvils and isotopically pure (99.95% 12C) type Ila synthetic crystals, respectively. The extremely low level of lattice and optical defects in the isopure substrates relative to natural crystals makes them ideal for optical characterization and defect studies, as well as high performance optical and electronic applications [1's (1) geometry, (2) type (defect density), and (3) temperature. The phase purity and defect structure of the films were analyzed by micro-Raman and low temperature photoluminescence (PL) spectroscopy. Optical and atomic force microscopy (AFM) were used to examine the surface morphology of the films. High temperature growth on the nonplanar anvil tip was anisotropic and extremely rapid (50-400 μm/hr), with significant incorporation of optically active defects. A possible model for the observed rapid growth is discussed. The films grown at lower temperatures on planar isopure substrates showed a dense array of oriented, stepped, square growth hillocks with penetration twins at their tops, while those grown at higher temperatures exhibited reasonably smooth, <110> oriented macrosteps. The higher temperature planar films showed a larger variety of incorporated optical centers than the lower temperature films.

The heteroepitaxial nucleation and eventual growth of large area single crystal diamond films has long eluded researchers interested in tapping it's many enabling properties, specifically in the field of active electronics. The uncertainty surrounding the diamond nucleation mechanism(s) and corresponding inability to carefully control this process are often blamed for the difficulty in achieving true heteroepitaxial growth. Biasenhanced nucleation (BEN) has been shown to provide in-situ control of the nucleation process. Subsequent advancements in both nucleation and deposition stages has resulted in highly oriented diamond films, approaching single crystal quality yet still plagued by arrays of medium to low angle grain boundaries that can degrade the electronic transport properties. To further improve upon these results and achieve large area, single crystal films it is believed that development must focus on the more fundamental problems of diamond nucleation. This paper presents a review of recent progress pertaining to the bias-enhanced process and focuses on data specific to the epitaxial nucleation dilemma.

Homoepitaxial diamond films were deposited on {100} and {111} oriented substrates using microwave plasma assisted CVD. The growth rate was measured in situ using laser interferometry. Various amounts of 15N2 were admixed to the process gas (0-50 ppm). The growth rate on {100} faces was found to increase significantly (by a factor 1.8) with increasing 15N2 content. In contrast, on { 111 } faces only a minor increase of the growth rate upon nitrogen admixture was observed. These findings are in perfect agreement with the observed influence of nitrogen contaminations on the α-parameter, as derived by the X-ray texture analysis of polycrystalline diamond films [1]

Near-single-crystal diamond films have been obtained in a number of laboratories recently. The optimization of nucleation density by using a bias-enhanced nucleation (BEN) method is believed to be a critical step. However, the condition of optimized nucleation has never been clearly delineated. In the present report, a novel quantitative technique was established to monitor the nucleation of diamond in-situ. Specifically, the induced current was measured as a function of nucleation time during BEN. The timedependence of induced current was studied under various methane concentrations as well as substrate temperatures. The optimized nucleation condition can be unambiguously determined from the current-time plot. Besides the in-situ current probe, ex-situ x-ray photoelectron spectroscopy (XPS) and atomic force microscopy (AFM) were also used to investigate the chemical and morphological evolution. Characteristic XPS and AFM features of optimized nucleation is discussed.

Many applications of thin films, especially for electronics devices, require that these materials, which are often anisotropic, have a very high thermal conductivity as well as uniform areal properties to ensure that reproducible performance be attained. These factors necessitate that measurements of thermal transport properties are required both to provide absolute application values for different heat flow directions as well as to evaluate uniformity and homogeneity of a wafer. For diamond, the combination of a very high thermal conductivity with limited size and form of available specimen presents unique challenges to the experimentalist. As a result, a modification of the ac calorimeter method has been developed to evaluate the thermal diffusivity of thin films.

Details of the technique will be provided together with examples of its use to evaluate thermal diffusivity and thermal conductivity of different CVD diamond film composites having thicknesses from 10µm to 600µm and free-standing films. In addition, results using this method will be compared. with those obtained by other techniques involved in a recent international round-robin measurements program designed to evaluate a potential standard method(s).

performance that can be achieved by utilizing a diamond heat - sink design which minimizes junction - to - case thermal resistance. Effects of the thermal conductivity of the substrate material, the thermal conductivity of the die attach material, the substrate thickness, and the die attach thickness onl Ihe thermal resistance are addressed. The results indicate that the temperature increase could be 3 to 4 times less with diamond heat-sinks when compared to other materials.

Many applications for lasers require them to run at high power densities, and subsequent thermal effects can degrade their operation or limit their performance. The incorporation of a thin film of diamond in-between a laser diode and its heatsink can bring about improved laser operation through a reduction in operating temperature and temperature gradients. This paper describes the use of finite element techniques to determine the temperature distribution within a double-heterostructure laser diode and its heatsink. In the model, the diode is mounted p-side down on a copper heatsink with various thicknesses of diamond film being incorporated between the diode and the heatsink. The importance of using the correct dimensions for the diamond film are demonstrated. Various heatsink-diamond film configurations were experimentally constructed and the temperature of the active laser region measured by the nulling technique (monitoring the lasing wavelength as temperature changes). This temperature was compared to that predicted for the active region from the numerical model in order to verify the accuracy and limitations of the numerical model.

The main advantages of stationary luminescence as a characterization technique of diamond are its relatively simple technical realization and a wide information gained on the impuritydefect structure of crystal. The physical reasons of capability of luminescence come from a very high Debye temperature of the diamond crystal, a strong interatomic bonding, and a relatively weak electron-phonon coupling in optical centers. As a result, a majority of optical centers in diamond possess the luminescence spectra with well resolved zero-phonon lines assisted by vibration-related replica, many of them retaining this spectral structure to room temperature and even above. As zero-phonon line spectrum is normally very sensitive to any external forces applied to the crystal (temperature, mechanical stress, magnetic and electrical fields), the optical centers can serve as sensitive markers, providing much information on the state of the didmond lattice surrounding them.

The value of the luminescence increases when diamond is characterized as a material for electronic and, particularly, optoelectronic applications. In this area luminescence really proves to be a unique characterization tool giving a direct information on optical and electronic properties (for instance, impurity and defect content, lifetime of charge carriers, efficiency of radiative recombination etc.), to be used in the devices.

In this review the basic parameters of the luminescence centers of diamond are considered and discussed from the view-point of the information they provide. Cathodoluminescence of diamond films and ion implanted layers is given as an example of luminescent characterization of diamond. The semiconducting structures with insulating active zones are discussed as a promising application of diamond for light emitting devices operating at high temperatures.

A cathodoluminescence (CL) study of free- and bound-exciton recombination in high purity high pressure high temperature (HPHT) synthetic diamond is presented, including temperature dependence measurements of the free-exciton intensity and luminescence decaytime. The results are compared with those from device quality diamond produced using the chemical vapour deposition (CVD) process.

Ultraviolet and X-ray photoelectron spectroscopy techniques have been employed in a preliminary study of the electronic structure of polycrystalline diamond films that have been grown on Si substrates by if-plasma enhanced chemical vapor deposition using water/ethanol growth chemistries. In particular, polycrystalline diamond films with distinctly different surface morphologies and Raman scattering characteristics have been investigated. Corresponding ultraviolet photoemission spectra from air-exposed samples have shown the presence of a prominent low-energy secondary electron emission peak indicative of a negative electron affinity (NEA) surface. Chemical stability of the polycrystalline diamond NEA surface has been demonstrated following conventional acid cleans and hydrogen plasma processing. In contrast, an oxygen (20%)/Ar plasma exposure has been shown to extinguish the photoemission of low-energy secondary electrons and remove the NEA. However, by employing a high-temperature anneal at 750 °C for 15 min in ultra-high vacuum the NEA surface can be restored. Compared to NEA single crystal diamond surfaces the photoexcited low-energy electron emission from chemical vapor deposited polycrystalline diamond films is more robust.

An evaluation of the use of intermediate layers for promoting adhesion between diamond and Fe and Ni base alloys is presented. The lack of adhesion between diamond and such alloys has prevented its use as a protective coating in applications at intermediate temperatures. In this study we use a combination of plasma assisted deposition techniques together with controlled bias of the substrate in order to deposit intermediate layers, and simultaneously achieve a thorough intermixing of the elements at the interface, in order to increase the adhesion between the layers. We have compared the performance of different carbide forming metals, as well as other materials. Properties considered in the selection of the appropriate intermediate layers are: chemical affinity with carbon, thermal expansion coefficient, mechanical properties, and adhesion to diamond and to the base metal as well. In the case of multiple layers, adhesion strength was measured after every new layer was deposited, in order to identify any weak links of the composite structure. Transmission electron microscopy was used to determine the microstructure and phases. An analysis is provided of the resulting performance on the basis of the microstructure.

Various optical techniques, including Photothermal Deflection Spectroscopy (PDS), ellipsometry, and transmission measurements, have been used to examine the optical properties of diamond like C films deposited at various laser power densities on Si and fused quartz. The absorption coefficients of these films are typically above 5 × 104 cm−1 throughout the visible part of the spectrum, decreasing quasi-exponentially in the infra-red. Larger sp2/sp3 bonding ratios caused by deposition at lower laser power settings noticeably increase film absorption coefficients. While Tauc plots of our absorption data yield bandgaps of 1.5-1.0 eV, the quality of these fits is unimpressive and the slopes are unphysically low. Rather than modelling these materials as narrow gap semiconductors, we have pursued another avenue: treating the films as a mixture of small graphitic particles embedded in a diamond, sp3-like matrix. We will discuss the results of this approach as well as the evolution of the optical properties after thermal anneals.