Diacuond thin filnis can be grown epitaxially on cubic born nitride (c-BN) surfaces using DC plasilia cheniical vapor desposition. As a substrate of this experinient, high pressure synthesized c-BN particles (2OO---5OOuni in diatueter) are used. At the early growth stage of the dianiond filuis on c—BN lll} , {lOO} and i22l} surfces, it is found that the growth wanner of the films is nucleation—growth type showing the island density of about 1011 cm 2• The coalescence of the growing island is also found by SEM observation and the continuous filn is formed at the film thick of 2000A and l,uni for the substrate surfaces of {lll} and {lOO} respectively. The epitaxial re1ations on typical c-BN surfaces such as till] and lOO} are {ll1} , (110) diaiuond 1/ Cll1} , (110) c—BN {lOO} , (110) diauiond /1 1OO} , (110) c—BN. The (100) epitaxial growth of diamond on C221} of c-BN is also observed. This is interpreted by twinning introduced in the diaciond filtns during crystal growth.

Color centers in diamond have a potential for opto-electronic applications. Laser action of H3 centers in natural diamond has been already reported. Our interest is focused on the possibility of laser action using synthetic diamond. A first attempt was made to produce H3 centers with high density in synthetic diamond. Optical properties of the H3 centers were investigated. Sample-dependent luminescence decay time and long lived transient absorption were observed. Laser action using the H3 centers in synthetic diamond occurred though the efficiency was as low as 0.1%.

The first order Raman line and the prominent features in the second order Raman spectrum of diamond crystals interpreted on the basis of polarization selection rules and symmetry provide energies of the critical points of the phonon dispersion curves with considerable precision. Similarly elastic constants are deduced from Brillouin shifts with high precision. Absolute Ra cross-sections are accessible from a simultaneous observation of the Raman spectrum and Brillouin components. Diamond films can be characterized on the basis of this body of information. 1.

The carbon-hydrogen chemical vapor transport system was examined in accordance with a four-stage transport model. A result of this examination is that graphite co-deposition could be avoided when diamond is deposited from gas solutions under-saturated with respect to diamond. Actual deposition experiments showed that this unusual requirement can be fulfilled but only for the condition that the transport distance between the carbon source and the substrate surface is short. In such a case diamond can be deposited equally from super-saturated as well as from under-saturated gas solutions. On the basis of thermodynamic considerations a possible explanation of this unusual phenomenon is given. It is shown that there is a possibility of deposition of diamond from both super-saturated as well as under-saturated gas solutions but only on the condition that they are in a non-equilibrium state generally called the activated state. A model of the diamond deposition process consisting of two steps is proposed. In the first step diamond and graphite are deposited simultaneously. The most important carbon deposition reaction is C2H2(g) + 2 H(g) C(diamond graphite) + CH(g). The amount of co-deposited graphite is not a direct function of the saturation state of the gas phase. In the second step graphite is etched according to the most probable reaction C(graphite) + 4 H(g) CH4(g). Atomic hydrogen in a super-equilibrium concentration is necessary not only to etch graphite but also to precipitate and graphite. 1.

The synthesis of well-crystallized diamond has been achieved by a wide variety of CVD techniques including plasma assisted methods (DC, RF and microwave), hot wire methods and by oxyacetylene combustion. These can be differentiated from other carbon coatings (graphite, vitreous carbons, and DLC) by Raman spectroscopy, X-ray diffraction and scanning electron microscopy. The high linear growth rates reported for the small area flame deposition of diamond and the recent demonstration of optically transparent continuous diamond films using 'torch' growth methods has led to speculation that by enlarging the area of uniform combustion it might be possible to deposit tens of square centimeters of transparent diamond at reasonable rates using relatively inexpensive equipment. We have previously reported that samples prepared by oxyacetylene combustion at low pressure were highly non-uniform and that higher oxygen/acetylene ratios were required than those found successful at atmospheric pressure using common brazing or welding torches. This paper reports low pressure flat flame combustion growth at still higher 02/C2H2 gas ratios and a new Raman feature at —1280 cm1 which was observed in some experiments in addition to the expected diamond peak at 1332 cm. Using data from previously published work on a similar flame composition, we speculate that the locations in the flame yielding diamond (as opposed to non-diamond forms of solid carbon) correspond to regions of the highest atomic hydrogen mole fraction (partial pressure) which in turn is closely related to the highest partial pressure of numerous reactive hydrocarbon radicals including the methyl radical. The increased oxygen/acetylene ratio needed for diamond deposition in these experiments may well be related to the much lower than adiabatic flame temperature expected and observed in previous work on oxyacetylene combustion at low pressure using a water cooled flat flame burner.

A dc arc discharge plasma torch has been developed for chemical vapor deposition (CVD) diamond growth. The apparatus and process parameters are described. Free-standing polycrystalline diamond samples of 50 mm by 50 mm by a few mm have been grown at high rates. The Raman spectra of the samples show little nondiamond structure. Transmission electron microscopy indicates that the diamond is highly twinned and has a high defect concentration. The infrared spectra indicate the presence of hydrogen contamination in the diamond via absorption bands associated with carbon-hydrogen motion. 2.

We have applied magneto-microwave plasma to be successful for the first time in deposition of diamond at far lower pressure than the conventional high pressure plasma CVD. The important point of the low pressure plasma deposition system is to set the electron cyclotron resonance (ECR) condition (875 G in the case of a 2. 45 GHz microwave) at the deposition area. The high density plasma ( above '' '' /cm3 ) necessary for high quality diamond formation has been obtained by the effective microwave absorption near the magnetic field satisfying the ECR condition. The plasma is quite uniform at the discharge area (160 mm in diameter) and uniform diamond films with high quality have been obtained. From the investigation of diamond formation from 50 Torr to lO2 Torr in the same deposition system it is definite that the lower pressure lowers the formation temperature of diamond to 500C and the effective species for diamond formation are low energy radicals. 1.

We have done a series of experiments on diamond microcrystal formation in flowtubes. The system is designed to separate the discharge used to create atomic hydrogen from the organic molecules used as a carbon source. This creates a simplified chemical environment in which the species concentrations are kinetically rather than thermodynamically controlled. The flowtube enables us to examine kinetics of diamond formation under a variety of conditions and gives us some information about the rate of nucleation independently of the growth rate. 1.

We have investigated the process of UV laser evaporation of diamond-like carbon films on metallic and semiconducting alloys. High purity targets in vacuum were irradiated with a pulsed KrF excimer laser beam (248 nm). Adhesion of diamond-like carbon films to several materials was improved by applying in situ germanium interlayers. Various process parameters were varied to determine the optimum conditions for deposition. The diamond-like films and germanium interlayers were structurally characterized by inelastic light scattering and scanning electron microscopy. Mechanical properties such as indenter hardness and friction coefficient were also measured. Preliminary results indicate that films deposited at low substrate temperature and high laser fluence have superior mechanical properties.

A microwave plasma enhanced chemical vapor deposition system is characterized using optical emission spectroscopy and mass spectrometry. CH4 CH2 CH4 and CO were used as carbon source gases. The effects of 02 addition to the feed gas is examined. Emission from CH in the plasma is observed and CH4 is a stable reaction product for all carbon source gases used. 02 is fully consumed and converted to H20 and CO. Emission from C is observed for all hydrocarbon gases when 02 is added but is absent when CO is the carbon source gas. Addition of 02 also dramatically affects the relative amount of reaction products as the carbon in the system is converted to CO. 1.

The biaxial modulus and residual (post deposition) stress of polycrystalline diamond (PCD) films deposited by microwave plasma CVD is determined using the bulge test technique. This method involves measuring the deflection of a circular membrane under an applied differential pressure. A calibration parameter for the bulge test is determined by evaluating the biaxial modulus of a silicon specimen standard. The film is characterized using X-Ray diffraction. Preliminary results yield a biaxial modulus value of 960 GPa for the PCD film. 1. INTRODUCFION Diamond inherently exhibits several unique physical and optical properties enabling its use in various applicalions. Its large bandgap (5. 45 evi) and covalently bonded aliphatic sp hybridized carbon atoms arranged t. eirahedrally result in an optical transparency in wavelength from 220 to 2500 nm and wavelengths above 6000 nm (beyond mid-infrared). Diamond is extremely hard highly thermally conductive and has both low coefficient of thermal expansion as well as high elastic modulus. This combination of properties results in diamond being extremely resistant to thermal shock as is characterized by the thermal shock resistance figure of merit R: (1v)nc R Ea where v equals the Poisson ratio a is the fracture strength ic is the thermal conductivity E the elastic (Young''s) modulus and a is the coefficient of thermal expansion2. Similarly the damage velocity threshold v1 for brittle materials impacted by water drops has been has been related to the

Whirling arm rain erosion tests and scanning electron microscope (SEM) tribonietry were performed in parallel with germanium (Ge) flats coated with sputtered DLC films ranging in thickness from 0. 25 zin to ''1 in. Our aim was to establish some correlation between delamination resistance of the DLC under rain drop impact and under environmentallycontrolled tangential shear conditions. The DLC tested at rain erosion velocities of 112 156 and 201 in. s1 (250 350 and 450 mph) showed a correlation between DLC thickness and mnicrofracturecaused delamination where the pit density decreased and pit size increased with coating thickness. Sliding a CVD polycrystalline diamond covered a-SiC pin tip against DLC-coated e flats undr controlled loads speeds and in 1. 33 x 10 Pa (1 x 10 torr) vacuum of the SEM triboineter abrasively removed the thin films inunediately. The thick films lasted for 3 to 10 cycles followed by catastrophic delamination from the wear path in large flakes. The data indicate that delamination of the DLC under any of the test conditions is controlled mainly by the increasing internal stresses developing with progressively higher film thicknesses. These tests served as a prelixainary study using high temperature SEM tribometry measuring abrasion resistance of DLC and polycrystalline diamond films as an approximation for rain erosion resistance under ultrahigh velocity and thermal shock conditions. SPIE Vol. 1325 Diamond Optics /11 (1990) / 99

Diamond films grown under a variety of deposition conditions in hot-filament or microwave-plasma CVD reactors were characterized by Raman and cathodoluminescence (CL) spectroscopies. The magnitudes of the following four Raman spectral features were observed to vary from specimen to specimen in a correlated manner: (1) the linewidth of the diamond Raman line (2) the intensity of the tails of the diamond Raman line at several half-widths from the peak (3) the intensity ratio of the sp2-bonded carbon Raman band to the diamond Raman line and (4) the intensity ratio of the broad photoluminescence (PL) background that underlies the Raman spectrum to the diamond Raman peak. We suggest that each of these features varies with the abundance of sp2-bonded carbon. As a function of deposition conditions the sp2-bonded carbon content increases with methane fraction and with substrate temperature and decreases with oxygen fraction. A cyclic variation with deposition time is observed for one set of hot-filament depositions. The CL spectra of these specimens consist of several distinct components attributed to impurities point defects and dislocations. The CL intensity is found to be vary from specimen to specimen approximately inversely with the intensity of the visible-excited PL background in the Raman region. This correlation is strongest for the dislocation-related 2. 85 eV CL band. A model of competing recombination between two sets of defects is proposed to explain the inverse correlation. 1 .

A variety of cathodoluminescence depending on crystal perfection impurities and defects such as vacancies and interstitials have been investigated in undoped boron-doped and nitrogen-doped CVD diamonds. The free exciton recombination radiation is sensitive to the perfection and purity of diamonds and is located in C100sectors. The bound exciton recombination radiation depends on the amount of boron in diamonds. The 5RL center is strong in the undoped diamonds after neutron irradiation and supposed to be related to intrinsic defects such as seif-interstitials. The 2 . 16 eV and the 3 . 19 eV centers are observed prominently in the nitrogen-doped diamonds after electron irradiation and subsequent annealing. The origins of the 2. 16 center (nitrogen-vacancy complex) are located in sectors. 1.

The growth surfaces of CVD diamond films are usually rough with polycrystalline crystallographic habits which presents a severe problem if CVD diamond films are to be used in infrared optics. Several methods are described in this paper in an effort to solve this problem. A polishing process was used to reduce the surface roughness by polishing the rough growth surface with a heated cast iron scaife. For polished films, near 70% transmittance was obtained over the whole range of 600-4000 cm-1, while the transmittance for non-polished films were much lower and varied strongly with the wavenumber. Absorptions believed due to carbon-hydrogen stretching bands and a silicon carbide phase were observed in the transmission spectra of polished diamond films.

Room temperature Fourier transform spectroscopic transmissivity data in the far-infrared (10 - 360 cm1) were used to derive a model for the complex refractive index (n - ik) of polycrystalline diamond films grown by microwave plasma enhanced chemical vapor deposition. Transmission measurements were also made at a frequency of 19. 5 cm1 utilizing a CO2 optically pumped submillimeter laser. Due to their polycrystalline nature the diamond films exhibited a surface roughness with crystallite sizes ranging from 2 -8 p. m. This roughness introduced a diffuse scattering loss at the shorter wavelengths and is accounted for in the determination of the optical properties of the diamond. Residual absorption proportional to 2 is observed at longer wavelengths (1A cm1). 1.

Infrared transmission reflection and scatter of freestanding polycrystallin diamond films have been measured. Raman spectra have also been obtained. Films with microcrystalline as well as (1 1 1 ) and (1 00) faces exposed are compared using scanning electron microg raphs . Altered gas mixtures anc temperature of substrate cause changes in surface microstructure and will bE discussed in regard to the effect on optical properties. 1 .

A parametric influence on the morphology of CVD diamond has been investigated. While the gas pressure has a minor effect on the morphological development of diamond in the optimum temperature range, the influences by the substrate temperature, the methane concentration in hydrogen, and the gas flow rate are found to be substantial. With an increase of substrate temperature or gas flow rate over the range in this study, the surface of diamond films changes from the f 100) face dominating to the { 1 1 1 } face prevailing. With an increase of methane concentrations from 0.5% to 2%, the morphology of diamond films changes from the triangular 1 1 1 ) faces to the square f 100) faces. Further increasing methane concentrations to 5% results in a film with sub-.tm crystallite aggregates. These experimental variables change the diamond morphology possibly by influencing the free energies of crystallographic faces and the atom mobility on the surface.

The microstructure of polycrystalline diamond films grown by microwave plasma assisted chemical vapor deposition (PACVD) have been observed as a function of growth temperature substrate identity and surface condition. Our highest microwave PACVD growth rates have been achieved in (1 10) axis normal oriented polycrystalline diamond films. Results indicate thai at growth temperatures below 650C kinetically dominated processes induce the formation of a preferential (1 10) axis normal orientation in diamond films with micron scale microstructure1 This orientation can be sustained on silicon boron nitride and silicon nitride substrates to filrr thicknesses in excess of 60 microns through the occurrence of (I I I) twin defects. Such fiIm have the high density and generally uniform microstructure required for optical applications. 1.

The diamond film polishing apparatus with hot metals has been presented . Diamond films deposited by both microwave plasma CVD and arc discharge plasma jet CVD are used as workpieces . As a diamond film sways on a polishing plate made of an iron or a nickel heated to 75O''95O C it is finished to flat and glass-like surfaces without any exfoliation. Not only swaying speed but also polishing pressure in this apparatus are extremely low in comparison to a conventional mechanical polishing method. When a cast iron and a molybdenum are used as a polishing plate polishing is not advanced . The polishing rate is highest in a vacuum atmosphere . Among gas atmospheres the rate is highes t in a hydrogen atmosphere . These result show that a diamond film is polished by the diffusion of carbon into a polishing plate. When a diamond film surface is too rough to polish by this apparatus the surface planing with YAG laser is applied then polishing is conducted on the planed surface. 1 .

Optically smooth surfaces can be produced on initially rough polycrystalline diamond film through the combined use of reactive ion etching and high temperature lapping on Fe metai Protective thin film barriers are first applied to the diamond surface to restrict the reactiv oxygen or hydrogen ion etching process to regions of greatest roughness. When the overaJ surface roughness has been reduced sufficiently by etching mechanical lapping of the surfac on an Fe plate at temperatures of 730C-900C in the presence of hydrogen can be used t produce surface roughnesses of less than 10 nm as measured by profilimetry. The tw techniques are complementary for flat surfaces while the reactive etching process alone can b used with shaped substrates to produce a surface finish suitable for LWIR optical applications. 1.

Polycrystalline diamond films have been synthesized in an oxygen-acetylene flame (15-20 tm thick and in a filament-assisted CYD reactor (10-60 p. m thick). The quality of the diamond was measured with a Raman microprobe optical and electron microscopes an FTIR and a UV-Vis-NIR spectrophotometer. The hemispherical transmittance and reflectance of several films was measured in the UV visible and infrared with integrating spheres. The transmittance and reflectance was found to depend on which surface of the film faces the spectrophotometer''s beam this result can be explained by total internal reflection effects and the different roughness scales of the film''s two surfaces. Preliminary results on the polishing of several FACVD films using a heated steel disk a rhenium filament and a hydrogen environment are discussed. An Auger depth profile of the steel disk shows graphite at the surface of the disk iron carbide in the bulk and a depletion of carbon in the bulk near the surface. The films'' surface roughness before and after polishing was measured with a profiometer and local roughness averages of A have been achieved. Results of optical scatter measurements made on films before and after polishing are presented. 1.

The results of room-temperature photoconductivity measurements on free-standing diamond films are reported. The films were grown on Si(100) substrates by hot filament-assisted chemical vapor deposition (CVD) from a methane/hydrogen mixture and ranged in thickness from 40 to 100 pm. The observed photocurrents in unintentionally doped films increased monotonically with increasing excitation energy. The films are found to exhibit photocurrent excitation similar to that observed for bulk diamond. In films doped with either N or Li the photocurrent exhibited broad structure superposed on the monotonic background. The photocurrent was found to depend on the chopping frequency of the excitation light decreasing with increasing chopper frequency indicative of trapping center dominated recombination dynamics. Schottky barrier heights were determined from the photoresponse for Au on CVD diamond film and on (100) oriented single crystalline type ila natural diamond. The measured barrier heights were 2. 02 and 2. 24 eV respectively in good agreement with previously measured values. A second barrier height was obtained from a threshold for internal photoemission at lower energies P4. 35 eV. We were able to observe for the first time an optical enhancement of 20X in the photocurrent using an optical biasing technique. 1.

Diamondlike carbon (DLC) is deposited by electron cyclotron resonance radio-frequency (ECR-RF) hybrid plasma method. The films are hard with optical bandgap ranges between 1. 1 and 1. 8 eV depending on the deposition parameters. Waveguide structures have been fabricated in DLC using liftoff technique. The films are also characterized by Electron energy loss spectroscopy and Raman spectroscopy. 2.

Protective and anti-reflective optical coatings of diamond-like carbon have been deposited on germanium and silicon substrates from glow discharges of cyclohexane. Films of excellent quality were obtained with high deposition rates using this non-hazardous feed gas. A parametric study was carried out to determine the dependence of the process characteristics and the optical properties of the films on the electrode bias gas pressure and flow. The study determined the contours for film refractive index and deposition rates over the parameter space and also revealed information on the dependence of infrared absorption and film adhesion. Important information on process scaling was determined. A simple dependence of negative self-bias on electrode design was obtained by using electrodes of different sizes. Changes in bias as a result of the coating of grounded reactor surfaces were measured. The effect of feed gas flow on the film and process properties provided insight into the impact of reactant depletion on the deposition process. 258 / SPIE Vol. 1325 Diamond Optics 111(1990)

Experimental investigations on diamond films prepared in oxyacetylene combustion flames at atmospheric pressures have been carried out by a detailed analysis of laser Raman spectroscopy scanning electron microscopy high resolution electron diffraction x-ray photoelectron spectroscopy and Auger electron spectroscopy. The Raman spectra of the films1deposited on Si (1OO substrates clearly indicate a well defined 1334 cm signature of the sp bonded carbon (diamond). A systematic Raman analysis has been carried out to establish the rnxed nature of the carbon bonding1in these films by 1monitoring the 1334 cm line as wll as the 1350 cm D line and 1580 cm G line both reminescent of the sp bonded graphitic Raman signatures of carbon. The optimum deposition conditions necessary for obtaining a predominant diamond film growth have thus been established. The scanning electron microscopic investigations indicate that the growth morphologies of the films are dominated by and 3ioo3 surfaces. The morphological features have a spatial variation and also exhibit ball''like features along with the well faceted crystallites. The electron diffraction studies establish cubic diamond crystal structure with a lattice parameter of 3. 55 A. The high resolution studies also indicate the presence of other impurity phases the origin of which are attributed to the thermodynamically non-uniform situation at the substrate-gas interface. Core level photo-electron spectroscopic studies carried by monitoring the Si (2p) and C (is) core levels at 98. 1 and 285 eV

Using highpressure synthetic diamonds of type Tb NV or 112 center is introduced dominantly by controlling the dose and the annealing temperature. Some optical properties are studied to explore a possibility for application to optoelectronics. For NV center the holeburning effect of the zerophonon line has been examined on the following aspects: the distribution for the magnitude of internal strain the burning and erasing processes of a hole and the temperature dependence of the hole width. For 112 it has been found that photochromism occurs by illumination of light shorter than 600nm wavelength. A tentative model is proposed on the basis of the experimental results. _i .

The use of photothermal radiometry to obtain the thermal diffusivity of CVD diamond is analyzed. The finite sizes of the heating beam and the radiation detector are taken into account. Heating a finite circularly symmetric region of the specimen surface is compared with uniform heating of the specimen surface. Thermal wave propagation in thick and thin diamond films in vacuum in air and on several substrate materials is modelled. 1 .

Most green diamonds owe their coloration to radiation exposure that produces the GR]. absorption band. Depending upon the conditions of radiation either a surface color or a bodycolor can result. Natural green diamonds of known geographic origin especially those with a conspicuous bodycolor are extremely rare. A group of seven natural greenish bluetogreen bodycolor type-Ia diamonds from alluvial deposits in Guyana display a distinct but sometimes subtle internal blue-to-green color zoning and a visiblerange absorption spectrum consisting of in addition to the N3 N2 and GR1 bands weak-to-moderate 503 (H3) 496 (H4) and 595 nm bands. We suggest that this combination of features is characteristic of at least one category of natural bodycolor green diamonds. 1.

Diamond membranes are being developed for x-ray lithography masks. Typically these 1-3 im thick membrane films are deposited on Si substrates using microwave driven plasmas. To obtain smooth films with uniform tensile stress spray and electrophoretic deposition of 0. 1 zm diamond seeds were used to control the initial nucleation and growth of the diamond films. The films have a room ternperature tensile stress of 25-125 MPa. The temperature dependence of the stress is due to the tensile growth stress of the diamond film and the thermal stress of the diamond-Si layer structure. The films have a biaxial modulus of 800 GPa. X-ray lithography masks have been made and used to print patterns with x-rays from a synchrotron source. 1 .

Vie have developed an analytical color fluorescence electron microscope (ACFEM) which is now being studied as a possible tool for evaluating diamonds. The ACFEM permits observation using colors corresponding to cathodoluminescence (CL) wavelengths in the visible region (400700nm) to distinguish the type size and distribution of emission centers and emission bands of diamonds. The ACFEM is an effective tool for determing the conditions used for synthesizing chemical vapor deposition (CVD) diamonds and is far superior to xpS and SIMS as a tool for analyzing dopants. 1.

In our program for ion beam polishing of microwave-assisted chemical vapor deposited diamond films using a planarizing layer we have sought to develop the technique by polishing AT-cut quartz disks used in rate monitors as an intermediate objective. Using an argon ion beam and spin-coated photoresist as the planarizing layer we successfully reduced the peak-to-valley (p-v) surface roughness from 5 pm to 0. 2 pm. In the case of diamond films whose etch rate is minute under argon bombardment reactive oxygen ions constituted the beam. The planarizing overcoat designed so that its etch rate would match that of the diamond film was a mixture of photoresist and titanium-silica. With this combination we reduced the diamond films'' surface roughness from 6 jim and 1 pm to 217 nm and 35nm p-v and mis respectively. This successful diamond film polishing opens the field of infrared optics to these materials. 1.

Polycrystalline diamond films with high growth-rate have been synthesized by dc arc discharge plasma CVD in a mixture gas of CH4 (1) and 112 (99). The diamond films are deposited on water-cooled silicon and molybdenum substrates at gaseous pressure of about 200 Torr. The typical arc discharge is performed at 200V and 4A while the hydrogen flow rate is about 3000 3500 sccm. The crystallinity of diamond films prepared are characterized by Xray differaction (XRD) Raman scattering spectroscopy and scanning electron microscopy (SEM). It is verified by XRD and Raman measurements that the synthesized diamond films are identified as natural cubic diamond structure and contain substantially no graphite or amorphous carbon. SEM photographs show that the crystal grain size reachs 60 80 im with good crystal habit and the average growth rate of diamond films deposited during 4 hours is about 40 - 60 pm/h. As shown by SEM photographs the diamond grain size obviously depends on the local nucleation density. 1.

Hard carbon coating such as diamond and diamond-like c :bon (also referred to as amorphous carbon) have received considerable attention for tribological applications due to their high hardness high modulus and desirable surface properties. Unfortunately most of the deposition techniques induce high substrate temperatures that temper traditional bearing steels and reduce the substrate load-carrying capability. Therefore to effectively use these desirable coatings a lower temperature deposition technique is required. Ion beam deposition offers essentially ambient temperature conditions accurate control ofprocess parameters and good coating-substrate adhesion. To use these attributes a test program was initiated to deposit mass analyzed high purity carbon and methane ions onto molybdenum and 440C bearing steel for subsequent characterization by Raman spectroscopy and friction-wear tests. Preliminary results for a coating deposited from a CO source showed an amorphous carbon/microcrystalline graphite structure which exhibited very high microhardness and a 3-fold reduction in coefficient of friction for unlubricated tests compared to untreated 440C steel. In addition incrementally increasing the applied load up to a factor of 5 resulted in progressively lower coefficients of friction only a minor increase (about 11) in the wear scar depth and no dramatic coating delamination or damage. Therefore an amorphous carbon/graphite coating applied to 440C steel at ambient temperature exhibits solid lubricating film characteristics with extremely high load-carrying capability. *Work performed under Martin Marietta Independent Research and Development Project D-8 1R Materials Technology. 116