Keith N. Rosser

Low temperature diamond growth using CO2/CH4 plasmas: Molecular beam mass spectrometry and computer simulation investigations

Microwave plasma enhanced chemical vapor deposition has been used to grow diamond films at substrate temperatures down to 435 °C using CO2/CH4 gas mixtures. An Arrhenius plot of growth rate as a function of substrate temperature yields a value for the activation energy for the growth step of 28 kJ mol−1. This is lower than that measured previously for CH4/H2 systems and hints at a different gas-surface chemistry when using CH4/CO2 plasmas. Molecular beam mass spectrometry has been used to measure simultaneously the concentrations of the dominant gas phase species present during growth, for a wide range of plasma gas mixtures (0%–80% CH4, balance CO2). The CHEMKIN computer package has also been used to simulate the experimental results in order to gain insight into the major reactions occurring within the microwave plasma. The calculated trends for all species agree well with the experimental observations. Using these data, the model for the gas phase chemistry can be reduced to only four overall reactions...

Dissociation dynamics of HCN(DCN) following photoexcitation at 121.6 nm

The technique of H(D) atom photofragment translational spectroscopy has been applied to the photodissociation of HCN(DCN) at 121.6 nm. Analyses of the H(D) atom time‐of‐flight spectra reveal the partner CN fragment to be formed predominantly in its A 2Π excited electronic state. Branching into the H/D+CN(B 2∑+) product channel accounts for a few percent of the total fragment yield, but we discern no evidence for any contribution from the product channel yielding H/D atoms in conjunction with ground state CN(X 2∑+) fragments. The majority of the CN (A) fragments are formed in their v’=0 level but with a markedly bimodal rotational state population distribution. This bimodality has been rationalized in the light of the available information regarding the form of the potential energy surface of the excited 1Π state of HCN(DCN) populated following photoexcitation at 121.6 nm.

Investigations of the plume accompanying pulsed ultraviolet laser ablation of graphite in vacuum

The plume accompanying 193 nm pulsed laser ablation of graphite in vacuum has been studied using wavelength, time and spatially resolved optical emission spectroscopy and by complementary Faraday cup measurements of the positively charged ions. The temporal and spatial extent of the optical emissions are taken as evidence that the emitting species result from electron–ion recombination processes, and subsequent radiative cascade from the high n,l Rydberg states that result. The distribution of C neutral emission is symmetric about the surface normal, while the observed C+ emission appears localized in the solid angle between the laser propagation axis and the surface normal. However, Faraday cup measurements of the ion yield and velocity distributions, taken as a function of scattering angle and incident pulse energy, indicate that the total ion flux distribution is peaked along the surface normal. The derived ion velocity distributions are used as input for a two-dimensional model which explains the obse...

Unravelling aspects of the gas phase chemistry involved in diamond chemical vapour deposition

We describe laser and mass spectroscopic methods, and related modelling studies, that have been used to unravel details of the gas phase chemistry involved in diamond chemical vapour deposition (CVD) using both H/C (i.e. hydrocarbon/H2) and H/C/O (e.g. CO2/CH4) gas mixtures, and comment on the relative advantages and limitations of the various approaches. In the case of the more extensively studied hydrocarbon/H2 systems we pay particular emphasis to investigations (both experimental, and 2- and 3-dimensional modelling) of transient species like H atoms and CH3 radicals, their spatial distributions within the reactor and the ways in which these distributions vary with process conditions, and the insight provided by such investigations into the chemistry underpinning the diamond CVD process. These analyses serve to highlight the rapid thermochemical cycling amongst the various hydrocarbon species in the reactor, such that the gas phase composition in the vicinity of the growing diamond surface is essentially independent of the particular hydrocarbon source gas used. Such applies even to the case of hot filament activated C2H2/H2 gas mixtures, for which we show that CH3 radical formation (hitherto often presumed to involve heterogeneous hydrogenation steps) can be fully explained in terms of gas phase chemistry. Diamond growth using H/C/O-containing gas mixtures has traditionally been discussed in terms of an empirically derived H–C–O atomic phase composition diagram (P. K. Bachmann, D. Leers, H. Lydtin and D. U. Wiechert, Diamond Relat. Mater., 1991, 1, 1). Detailed studies of microwave activated CO2/CH4 gas mixtures, accompanied by simpler zero-dimensional thermochemical modelling of this and numerous other H/C/O-containing input gas mixtures, provide a consistent rationale for the ‘no growth ’, ‘diamond growth’ and ‘non-diamond growth’ regions within the H–C–O atomic phase composition diagram.

CVD diamond wires and tubes

Abstract Diamond has been uniformly deposited onto the surface of thin metal wires using hot filament CVD. The diamond-coated wires are stronger and stiffer than the uncoated wires. Subsequent etching of the metal core in a suitable chemical reagent allows free-standing diamond tubes to be made, the typical dimensions being 1 cm long with an internal diameter of 10–150 μm. The formation of a thick, chemical-resistant carbide layer at the metal-diamond interface when using Ti and W wires is investigated.

Resonance enhanced multiphoton ionisation probing of H atoms in a hot filament chemical vapour deposition reactor

We demonstrate some of the merits and limitations of using multiphoton ionisation (MPI) spectroscopy, resonance enhanced at the two photon energy by the 2S1/2 state, to detect H atoms within a hot filament reactor used for diamond chemical vapour deposition (CVD). Subsequent analysis of the Doppler broadened lineshapes obtained in this way allows determination of spatially resolved, relative H atom number densities and gas temperatures. The effects of H2 pressure and flow rate, filament temperature and radial distance from the filament on the relative H atom number densities and the gas temperature profiles have each been investigated for the case of pure H2 and, in some cases, in the presence of added CH4. The present findings complement and extend previous measurements obtained using alternative insitu detection methods, and are generally consistent with current models of the gas phase chemistry prevailing in low power diamond CVD reactors. They also serve to refine the role of the hot filament in the H atom production process, and to illustrate continuing ambiguities regarding the way in which the local relative H atom number densities vary with filament temperature, with the presence of trace quantities of added hydrocarbon, and the interpretation of these dependencies.

Ion imaging studies of Cl(2P3/2) fragments arising in the visible photolysis of BrCl: Measurement of orientation, alignment, and alignment-free anisotropy parameters

The photodissociation dynamics of jet-cooled BrCl molecules have been investigated at four different wavelengths in the range 425–485 nm by high-resolution velocity map ion imaging. Four images of the Cl(2P3/2) atomic fragments are recorded at each photolysis wavelength with the probe laser polarization, respectively, linearly aligned and vertical (i.e., perpendicular to the detection axis), right circularly polarized, horizontally linearly polarized (i.e., parallel to the detection axis) and left circularly polarized on successive laser shots, thereby ensuring automatic mutual self-normalization. Appropriate linear combinations of these images allow quantification of the angular momentum alignment of the Cl(2P3/2o) fragments [i.e., the correlation between their recoil velocity (v) and their electronic angular momentum (J)] in terms of the alignment anisotropy parameters s2, α2, η2, and γ2, and determination of the “alignment-free” recoil anisotropy parameter, β0, as a function of parent excitation wavele...

Sulfur doping of diamond films: Spectroscopic, electronic, and gas-phase studies

Chemical vapor deposition (CVD) has been used to grow sulfur doped diamond films on undoped Si and single crystal HPHT diamond as substrates, using a 1% CH4/H2 gas mixture with various levels of H2S addition (100–5000 ppm), using both microwave (MW) plasma enhanced CVD and hot filament (HF) CVD. The two deposition techniques yield very different results. HFCVD produces diamond films containing only trace amounts of S (as analyzed by x-ray photoelectron spectroscopy), the film crystallinity is virtually unaffected by gas phase H2S concentration, and the films remain highly resistive. In contrast, MWCVD produces diamond films with S incorporated at levels of up to 0.2%, and the amount of S incorporation is directly proportional to the H2S concentration in the gas phase. Secondary electron microscopy observations show that the crystal quality of these films reduces with increasing S incorporation. Four point probe measurements gave the room temperature resistivities of these S-doped and MW grown films as ∼20...

A reinvestigation of the near-ultraviolet photodissociation dynamics of the methyl radical

Abstract The translational energy spectrum of H Rydberg atoms resulting from the photolysis of the methyl radical (CH 3 ) at 216.3 nm has been recorded. The use of pulsed supersonic jet flash pyrolysis of azomethane (CH 3 N 2 CH 3 ) as a clean source of methyl radicals has allowed a clear analysis of the data, eliminating possible interference from other molecular species. The measured time-of-flight of the resulting H atoms indicate that the partner CH 2 fragments resulting from predissociation of CH 3 (B) radicals are formed in their low-lying a 1 A 1 excited state, as predicted by theory. An earlier contradictory report of this dissociation process, using methanethiol (CH 3 SH) as a photolytic precursor to generate methyl radicals, has been reinterpreted in terms of secondary dissociation of ground-state thomethoxy radicals (CH 3 S) yielding electronically excited thioformaldehyde CH 2 S(A 1 A 2 )+H( 2 S).

Diamond deposition in a hot-filament reactor using different hydrocarbon precursor gases

A hot-filament reactor was used to deposit polycrystalline diamond films upon single-crystal Si substrates using hydrocarbon/H a gas mixtures. We studied the effect upon the deposition process and resulting film properties by varying the hydrocarbon gas from CIH x to C4H x alkanes. This was done maintaining a constant carbon-to-hydrogen ratio, but using a substantially lower-than-normal filament temperature (1500°C) in order to highlight differences in activation barriers and in the chemistry of the diamond-forming step. It was found that with increasing hydrocarbon chain length the deposition rate decreased, from a value of about 0.4/~m h -l for methane/H 2 mixtures to less than 0.07/xm h 1 for butane/H 2. This was accompanied by an increase in the relative proportion of amorphous carbon to diamond present in the films. After one hour deposition the diamond grain size remained constant at about 20 nm, irrespective of the precursor gas. The measured Knoop hardness of the films also decreased when using process gases other than methane. We also studied the effect of changing the bond order in C2H~, precursor gases (ethane, ethene, ethyne) but found that this had no effect on either the deposition rate or the film quality.