Diamond Growth Rate Research Articles

Growth and characterisation of CVD diamond wires for X-ray detection

Abstract In order to gain insight into the mechanism of diamond growth on metallic wires (50–500 μm diameter), we have investigated the deposition process in a hot filament chemical vapour deposition (HFCVD) reactor. A two-dimensional model of our HFCVD reactor was developed in order to interpret the experimental data which indicate an evident influence of the substrate wire diameter on the gas species concentration and on the diamond growth rate. The theoretical model indicates a reduction of the growth rate as the substrate diameter increases due to the reduction of hydrogen and the active radicals concentrations at the substrate as its growth area increases. In view of a possible application of such samples as X-ray detectors, we investigated their electrical properties by analysing the current/voltage characteristic in a metal/diamond/metal geometry at different temperatures. The activation energy of the conductivity is found to be dependent on the applied electric field. The experimental data support a Poole-Frenkel conduction with overlapping centres. Finally the response of our diamond wires to X-ray radiation flux was investigated. A sensitivity of about 5 × 10 −6 A s Gy −1 was measured under a 100 keV Bremsstrahlung X-ray flux and the photocurrent is almost linear with respect to X-ray tube current.

Diamond growth by injecting thermally decomposed chlorine atoms into methane/hydrogen mixture

Crystalline diamond has been successfully deposited by injecting thermally decomposed Cl atoms into CH4/H2 in a hot-tube system at an extremely high flow velocity (in the convection-dominant mass transport region). Diamond growth rate increased with increasing the total flow rate, suggesting the increase of [Cl]/[H] ratio near the growth surface. Film quality also improved with increasing the total flow rate as well as reducing the reactor pressure. Both the quality and film growth rate were enhanced as the inlet [Cl2] increased, due to the increase of total radical concentration. Two distinct growth activation energies were measured ranging from 3.6 kcal/mol in the substrate temperature range of 600–750 °C to 7.9 kcal/mol in the temperature range of 400–600 °C. Owing to the extremely short residence time and low gas temperature, carbon species near the growth surface remained almost the same as the input carbon source. By employing almost pure CH4 or C2H2 near the substrate surface, the CH3 radical was shown to be a more efficient diamond growth precursor than C2H2. With almost pure C2H2 near the surface, diamond deposition was negligible in a wide range of conditions on either silicon or diamond surfaces.

Effect of chlorine addition on diamond growth using methane/hydrogen reactants

The effect of Cl2 addition on diamond growth using CH4/H2 was studied in a hot-filament chemical vapor deposition reactor. Cl2 was mostly converted into HCl before arriving at the hot filament. Cl2 addition enhanced the diamond growth rate. The growth rate enhancement was more pronounced at a lower growth temperature and lower reactor pressure, implying that the CH3 and CH4 interconversion reactions were not in partial equilibrium below 5 Torr. Cl2 addition also reduced the activation energy by the same value for both pressures at 1 and 5 Torr, suggesting that the H and Cl exchange reaction was in partial equilibrium. However, Cl2 addition degraded the film quality, and the deterioration became more pronounced at a higher Cl2 concentration or lower substrate temperature. The effects of chlorine are suggested to include: (i) the fast reaction of Cl with surface bonded C–H; (ii) the fast reaction of Cl with CH4; (iii) the weak ability of Cl to adsorb on C=C; and (iv) the weak ability of Cl to break C–C to remove the improperly adsorbed CH3.

Kinetics of the initial stages of CVD diamond growth on non-diamond substrates: surface catalytic effects and homoepitaxy

Abstract Recently, we have reported that the density of chemical vapour deposition (CVD) diamond particles obtained on non-diamond substrates abraded with a diamond slurry can be enhanced by a few orders of magnitude by abrasion with a mixed metal/diamond slurry. In this work, we present growth kinetics studies of isolated diamond crystallites (prior to coalescence into a continuous film), which are formed at the initial stages of deposition on metal/diamond and diamond-pretreated silicon. The metals used were Ti, Fe and Cu. It has been found that the temperature dependence of the diamond growth rates on differently pretreated substrates can be described by an Arrhenius plot. From these plots, it has been concluded that at the initial stages of deposition the activation energy for CVD diamond formation, as observed for diamond-abraded substrates, can be substantially decreased as a result of substrate surface pretreatment with a mixed metal/diamond slurry. However, after coalescence of the crystallites into a continuous film, the activation energies for metal/diamond and diamond-pretreated substrates were found to be similar. Based on micro-Raman measurements it has been found that at the initial stages of deposition, the quality of diamond that was grown on the metal/diamond abraded substrates is higher than that deposited on the diamond-abraded ones. In contrast, after a long deposition time, the quality of diamond grown on the differently pretreated substrates was found to be similar. It is suggested that the observed effects are associated with the metal's catalytic activity towards the hydrogenation of unsaturated hydrocarbons, namely enhancing the concentration of sp 3 -bonded hydrocarbon fragments. At the initial stage of deposition, before a stable substrate is formed, this enhancement leads to an increase in the diamond growth rate, lowering of the activation energy for its homoepitaxial growth on diamond residues left by pretreatment, and improvement of its quality.

Low temperature plasma process based on CO-rich CO/H2 mixtures for high rate diamond film deposition

A low temperature process (350 °C<T<500 °C) for nanocrystalline diamond film growth by microwave plasma-enhanced chemical vapor deposition has been developed that yields high rates (up to 2.5 μm/h) at relatively low plasma powers (0.5 kW). In contrast to the widely used H2-rich mixtures of CH4 or CO and H2 that exhibit monotonic decreases in the diamond growth rate as T is reduced from 800 to 400 °C, CO-rich mixtures of CO and H2 exhibit an increase and sharp maximum as T is reduced. The results suggest a different diamond growth mechanism from the CO-rich mixtures.

Optimization of MW-PACVD diamond deposition parameters for high nucleation density and growth rate on Si3N4 substrate

Abstract We use orthogonal designs of experiments to optimize the process parameters for high diamond nucleation density and growth rate. The diamond films were deposited on Si3N4 substrate in a MW-PACVD reactor with a gas mixture of methane and hydrogen. The process pressure, the substrate temperature, the total gas flow and the methane concentration were varied in selected ranges. The optimal conditions for the nucleation and growth stages are different. This investigation proves that the nucleation is favored by low process pressures and methane concentrations between 1.2 and 1.6% in hydrogen, while the growth rate is enhanced mainly by increasing the reactor pressure. By using statistical analysis the variance was calculated in order to evaluate the degree of influence of the different parameters and their interaction effects. Trend analysis was done to describe the effects of each parameter on the results.

Lithium addition during CVD diamond deposition using lithium tert.-butanolat as precursor

Abstract Since doping of diamond is of great interest for electronic applications, the possibilities for the introduction of lithium into CVD diamond during the deposition process were examined. Lithium tert.-butanolat (C4H9OLi) is volatile enough to be transported by a small hydrogen flow to the reactor. Diamond films were grown by the hot-filament method under various deposition conditions (Li-precursor concentrations, CH4 concentrations, filament temperature, substrate temperature). The influences of diamond growth rates and the changes in diamond morphology show that Li influences the diamond deposition process. The diamond crystals of the coatings are well faceted, mainly showing a (111) morphology. SIMS depth profiles prove that lithium is present in the diamond films. The amount of lithium is constant within the film. It is significantly increased at the diamond/substrate interface.

Mineral inclusions in diamonds from the Sputnik kimberlite pipe, Yakutia

The Sputnik kimberlite pipe is a small “satellite” of the larger Mir pipe in central Yakutia (Sakha), Russia. Study of 38 large diamonds (0.7-4.9 carats) showed that nine contain inclusions of the eclogitic paragenesis, while the remainder contain inclusions of the peridotitic paragenesis, or of uncertain paragenesis. The peridotitic inclusion suite comprises olivine, enstatite, Cr-diopside, chromite, Cr-pyrope garnet (both lherzolitic and harzburgitic), ilmenite, Ni-rich sulfide and a Ti-Cr-Fe-Mg-Sr-K phase of the lindsleyite-mathiasite (LIMA) series. The eclogitic inclusion suite comprises omphacite, garnet, Ni-poor sulfide, phlogopite and rutile. Peridotitic ilmenite inclusions have high Mg, Cr and Ni contents and high Nb Zr ratios; they may be related to metasomatic ilmenites known from peridotite xenoliths in kimberlite. Eclogitic phlogopite is intergrown with omphacite, coexists with garnet, and has an unusually high TiO 2 content. Comparison with inclusions in diamonds from Mir shows general similarities, but differences in details of trace-element patterns. Large compositional variations among inclusions of one phase (olivine, garnet, chromite) within single diamonds indicate that the chemical environment of diamond crystallisation changed rapidly relative to diamond growth rates in many cases. P- T conditions of formation were calculated from multiphase inclusions and from trace element geothermobarometry of single inclusions. The geotherm at the time of diamond formation was near a 35 mW/m 2 conductive model; that is indistinguishable from the Paleozoic geotherm derived by studies of xenoliths and concentrate minerals from Mir. A range of Ni temperatures between garnet inclusions in single diamonds from both Mir and Sputnik suggests that many of the diamonds grew during thermal events affecting a relatively narrow depth range of the lithosphere, within the diamond stability field. The minor differences between inclusions in Mir and Sputnik may reflect lateral heterogeneity in the upper mantle.

Growth of Diamond Anvils for High-Pressure Research by Chemical Vapor Deposition

ABSTRACTGem quality diamond crystals are employed as anvils in high-pressure diamond cell research. Homoepitaxial growth experiments by microwave plasma-assisted chemical vapor deposition (MPCVD) have produced 1.76 mm (diameter) by 0.65 mm (thickness) sized diamonds. We report fundamental studies on diamond growth rate and quality as a function of reactor pressure and methane concentration, in a hydrogen plasma. By varying the growth conditions, large, single crystal diamond can be produced, which is ideal for manufacturing high pressure anvils.Traditional high pressure, high temperature (HPHT) techniques for production of synthetic diamond anvils are extremely expensive and chemical vapor deposition (CVD) provides an economically viable alternative. We report diamond growth rates up to 0.32 mg/hr, which are comparable to HPHT growth rates, and crystal quality approaching that of gem diamond. When perfected, diamond anvils produced from chemical vapor deposition methods could replace those manufactured by high pressure, high temperature synthesis.

Chemical vapour deposition of diamond at low substrate temperatures using fluorinated precursors

Diamond deposition from the fluorinated hydrocarbons CF 4 and CHF 3 was investigated in a hot filament chemical vapour deposition (HFCVD) system. The systems H 2CF 4, H 2CHF 3 and H 2CH 4 were compared under equivalent conditions. Our results demonstrate that below 650 °C the diamond growth rate for H 2CHF 3 gas mixtures is higher than that for the H 2CH 4 system. Even at 400 °C it was possible using H 2CHF 3 to measure considerable growth rates with in situ laser interferometry. Mass spectrometry proved the formation of HF at a significant concentration for the system H 2CHF 3. This is due to the abstraction of F from CHF 3 by atomic hydrogen. CF x radical species are postulated from CHEMKIN calculations. They are supposed to play an important role in the diamond growth process using CHF 3. With H 2CF 4 mixtures neither diamond nor amorphous graphitic carbon could be deposited. Mass spectrometry showed the inertness of CF 4 under the conditions of our investigations with the absence of the F abstraction reaction.

Modelling the substrate-temperature-dependence of diamond film growth

A model has been developed which describes the process of substrate-temperature-dependent diamond film growth. Detailed calculations were performed for the dependence of the diamond growth rate on the substrate temperature and the flux ration of hydrogen- and carbon-containing species above the surface. The theoretical results were fitted well with the experimental observation and data for diamond deposition by the hot-filament CVD method, illustrating that the model gives a satisfactory description of the diamond formation process.

Low temperature diamond growth using fluorinated hydrocarbons

Diamond thin film deposition using gas mixtures of CH 4, CHF 3 and CF 4, respectively, was studied using a hot-filament CVD technique. The diamond growth rate was measured in-situ with laser-interferometry and the film quality was determined using Raman spectroscopy. Compared with the conventional CH 4/H 2 gas mixtures, with CHF 3/H 2 the substrate temperature could be considerably decreased to 400 °C with still significant growth rates. Contrary to this positive behaviour of CHF 3/H 2, no deposition at all had been observed for the CF 4/H 2 gas mixture.

Growth of diamond in an electrodeless capacitive radio frequency discharge: W. Putz et al. (Technical University, Vienna, Austria). Int. J. Refractory Metals Hard Mater., vol 14, no 1/3, 1996, 173–178

Using the developed two-dimensional model of a d.c. discharge CVD reactor we have examined two possibilities of diamond growth rate enhancement. The first deals with the noble gas addition to inlet hydrogen-methane reactive mixture of a d.c. discharge. The second deals with application of a pulsed d.c. discharge. In the model, heat and mass transfer, electrodynamics phenomena, plasma-chemical and surface kinetics have been taken into account in a self-consistent manner. The electron-molecule reaction rate coefficients in methane/hydrogen/noble gas mixtures are calculated by solving the Boltzmann equation. Our calculations show that the dilution of the CH4H2 mixture with argon allows the input power to reduce substantially and to increase the tolerable CH4H2 inlet ratio. Diamond growth rate in a pulsed discharge may be several times higher than in a continuous d.c. discharge with the same power input. The overall enhancement depends on the relationship between the pulsed width and pause time.

Nucleation and bulk film growth kinetics of nanocrystalline diamond prepared by microwave plasma-enhanced chemical vapor deposition on silicon substrates

Real time spectroscopic ellipsometry has been applied to characterize the substrate temperature (T) dependence of the deposition rates for nanocrystalline diamond thin films prepared by microwave plasma-enhanced chemical vapor deposition on seeded Si substrates. With the real time capability, it is possible to determine the rates at which the diamond mass thickness (i.e., volume per area) increases during the early nucleation and bulk film growth regimes. The increases in the nucleating and bulk diamond growth rates with T for 400<T<800 °C are consistent with activation energies of ∼17 and 8 kcal/mol, respectively. The results reported here provide insights into the nature of the low-T growth rate limitations for diamond films on nondiamond substrates.

Boundaries of the diamond domain in the C - H - O diagram of carbon film deposition

Chemical vapour deposition experiments using gas mixtures of hydrocarbons with carbon dioxide can produce films of either diamond or a graphitic, amorphous form of carbon. The boundaries of the diamond growth domain in the triangular C - H - O gas composition diagram appear to be straight lines. It is shown that this is consistent with a simple model of the competition between the deposition of diamond and amorphous carbon phases. Gas phase carbonaceous fragments can attach to the surface of a growing film. Hydrogen in the gas phase etches poorly bound carbonaceous material from the surface, and also assists the incorporation of carbon into the diamond lattice. Oxygen in the gas phase forms carbon monoxide, and oxygen-bearing species also etch carbonaceous material. With these simple ingredients a rate equation approach is developed which also predicts the rate of diamond growth within the diamond domain, and indicates that oxygen can improve the efficiency of the process.

Diamond growth enhancement in d.c. discharge CVD reactors. Effects of noble gas addition and pulsed mode application

Using the developed two-dimensional model of a d.c. discharge CVD reactor we have examined two possibilities of diamond growth rate enhancement. The first deals with the noble gas addition to inlet hydrogen-methane reactive mixture of a d.c. discharge. The second deals with application of a pulsed d.c. discharge. In the model, heat and mass transfer, electrodynamics phenomena, plasma-chemical and surface kinetics have been taken into account in a self-consistent manner. The electron-molecule reaction rate coefficients in methane/hydrogen/noble gas mixtures are calculated by solving the Boltzmann equation. Our calculations show that the dilution of the CH 4 H 2 mixture with argon allows the input power to reduce substantially and to increase the tolerable CH 4 H 2 inlet ratio. Diamond growth rate in a pulsed discharge may be several times higher than in a continuous d.c. discharge with the same power input. The overall enhancement depends on the relationship between the pulsed width and pause time.

Two-dimensional simulation of a hot-filament chemical vapor deposition reactor

A two-dimensional model of a hot-filament (HF) CVD reactor has been developed to study the gas-phase and surface processes of diamond growth. Full transport equations were solved numerically to calculate the gas temperature, fluid flow, and species concentration fields. Catalytic chemistry at the filament surface was considered. The distribution of the hydrogen atom concentration and the gas temperature in an HFCVD reactor were obtained analytically. The expressions for diamond growth rate and the hydrogen atom destruction coefficient at the substrate were derived from the surface kinetics. The gas-phase reaction mechanism was obtained using a brute force sensitivity analysis. The calculated results were compared with existing experimental data.

Fabrication of diamond films at low temperature by pulse-modulated magneto-active microwave plasma CVD

A magneto-active microwave plasma chemical vapour deposition technique was developed by pulse modulation of the discharge to reduce the time-averaged microwave power for diamond film synthesis at low temperature. Due to a threshold power being required to start growth, the practical growth rate obtained by using the pulse-modulated plasma became three times larger than that obtained by using a continuous plasma of time-averaged power near the threshold. The methyl radical density was measured in continuous or pulse-modulated plasma by infrared laser absorption spectroscopy and compared to the growth rate. The time-averaged radical density was also enhanced by pulse modulation; it was up to 1.3 times larger than that in the continuous plasma. Though the correlation between density and diamond growth rate was not clear, the number of carbon atoms supplied as radicals was larger than the actual growth rate by almost two orders of magnitude.

Probable optimum hydrogen coverage of chemical vapor deposited diamond (111) surface

The isolated single dangling bonds on the diamond growth surfaces are believed to be the active bond sites for the surface growth of diamond in chemical vapor deposition (CVD) systems. If more than one dangling bonds connect with each other at some areas of the surface, they may reconstruct and form π-chain or dimer bonds because of the higher partial surface energies at such areas. These π-chain or dimer bonds are the templates for graphite growth and can be restrained by the etching action of the active hydrogen atoms in the deposition atmosphere. Assuming that the occurrence probability of each dangling bond at any site of the surface is random, the probability functions of a bond site occupied by an isolated dangling bond or one of the dangling bonds in a dangling bond cluster have been worked out respectively. Then, the typical experimental curve of the CVD diamond growth rate can be reasonably explained. When the probability of a surface bond site occupied by an isolated single dangling bond is maximum, an optimum hydrogen coverage of 0.86 monolayer (ML) of the CVD (111) diamond surface can be achieved and this corresponds to the highest surface growth rate of diamond. This mechanism is proposed by us as a reduced diamond surface growth model.

Determination of diamond growth rate in a flow tube geometry as a function of measured atomic hydrogen density

Present diamond deposition reactors have been very successful in furthering our understanding of how diamond grows at low pressures. However, their primary disadvantage is that changes in reactor conditions (e.g., filament temperature or microwave power, substrate temperature, and CH4:H2 concentration ratio in methane-based reactors) change all chemical species and their relationships. Changes in diamond growth created by independent changes to a single species are not directly accessible. For example, it is known that making atomic hydrogen is required for the growth of diamond films. However, changes to film growth rate have not been tied quantitatively to changes in the amount of atomic hydrogen available. A high-yield atomic hydrogen source was fabricated and pretested in a diamond growth setup. A previously developed atomic hydrogen sensor was used to measure atomic hydrogen density output as a function of both power and pressure. Diamond was then grown as a function of atomic hydrogen density, and growth rate was shown to be linear with density with a lower bound relationship of 0.27 μm/h/1016 cm−3. Since the magnitude and trend of the data appear to be inconsistent with other reported results, they are analyzed with respect to a simplified growth model. Based on reasonable assumptions, we conclude that the data are consistent with growth rate being independent of surface atomic hydrogen concentration and linear with methyl radical concentration. Because sufficient methyl radical density is produced through the interaction of atomic hydrogen with methane with correspondingly negligible reduction in total atomic hydrogen density, it is inferred that growth rate linearity with atomic hydrogen enters only through the production of methyl radical.