Characterization of crystallographic defects in homoepitaxial diamond films by synchrotron X-ray topography and cathodoluminescence

Crystallographic defects in a p-type homoepitaxial diamond film grown by microwave plasma-assisted chemical vapor deposition on a synthetic high-pressure high-temperature type-Ib (001) substrate were characterized by synchrotron radiation X-ray diffraction topography (XRT) and cathodoluminescence (CL). CL mapping indicated typical luminescent spots corresponding to the band-A emission around 420 nm. The band-A spots correspond to the spots observed by XRT for both diffraction vectors g = [044] and [404], and are considered to be mixed dislocations with a dislocation direction t = [001]. Typical dislocations in the film, such as edge and perfect 60° dislocations were determined by utilizing the relationship between the diffraction vector g, Burgers vector b, and the dislocation line vector t.     

A cathodoluminescence study of boron doped {111}-homoepitaxial diamond films

In this work we use cathodoluminescence (CL) at liquid helium temperature to investigate the boron incorporation in {111}-homoepitaxial diamond films, grown outside the visible plasma ball by the Microwave plasma-assisted chemical vapor deposition (MPCVD) technique. The boron concentration of this set of films covers the whole possible doping range divided into four parts: Low doping (5 × 10¹⁶ < [B] < 1.5 × 10¹⁹ cm^? 3), high doping (1.5 × 10¹⁹ < [B] < 3 × 10²⁰ cm^? 3), heavy doping (3 × 10²⁰ < [B] < 2 × 10²¹ cm^? 3), and phase separation range ([B] > 2 × 10²¹ cm^? 3). The phase separation occurs for very high boron concentrations, between the diamond phase (sp³ carbon) and the other components of the layer, namely sp² carbon and boron. A part of them is accumulated outside the diamond lattice.This detailed cathodoluminescence investigation of {111}-homoepitaxial diamond films has led to determining the doping range of the films and following the evolution of their crystalline quality when the boron concentration increases. In addition, a comparison between {111} and {100} films in the same doping ranges has been undertaken.     

785 nm Raman spectroscopy of CVD diamond films

Raman spectroscopy is a powerful technique often used to study CVD diamond films, however, very little work has been reported for the Raman study of CVD diamond films using near-infrared (785 nm) excitation. Here, we report that when using 785 nm excitation with 1 µm spot size, the Raman spectra from thin polycrystalline diamond films exhibit a multitude of peaks (over 30) ranging from 400–3000 cm^− 1. These features are too sharp to be photoluminescence, and are a function of film thickness. For films > 30 µm thick, freestanding films, and for films grown in diamond substrates the Raman peaks disappear. This suggests that the laser is probing the vibrations of molecular units at the grain boundaries of the disordered crystallites present at the interface between the diamond and substrate.     

Morphology dependence of cathodoluminescence spectra of CVD diamond film

Translucent polycrystalline diamond film made by a DC plasma CVD method was investigated with cathodoluminescence spectroscopy and topography. Shift and/or splitting were found in peaks of several luminescence bands, 575, 532, 480, 460, 389, 270 (5RL) and 235 nm (FE), suggesting considerable magnitude of residual stress is present in the film. The behavior of the splitting was different for the different peak, indicating that the symmetries of the associated defects are different. The film after heat treatment at 6 GPa and 1800 °C was found to exhibit new luminescence bands approximately 300 and 500 nm. Splitting of the 575 nm peak was recognized to decrease with the heat treatment, suggesting that the residual stress was reduced.     

Photoelectron spectroscopy of CVD diamond

We report a study of the electronic structure of thin diamond films by core state (X-ray photoelectron spectroscopy (XPS)) and valence band (UV photoelectron spectroscopy (UPS)) photoelectron spectroscopy. These techniques were used to investigate the different phases in the initial growth of polycrystalline diamond films on Si(100) substrates. The films were deposited by a standard microwave technique as well as by bias-enhanced microwave plasma chemical vapour deposition in a dilute mixture of methane in hydrogen. The influence of sample preparation (such as prebiasing or prescratching of the silicon surface with different sized diamond powder prior to deposition) on nucleation density and electronic structure was also investigated by scanning electron microscopy and photoelectron spectroscopy (PES) respectively. The surface composition was probed as a function of deposition time. The XPS data reveal the formation of an SiC phase at the early stage on nucleation, preceding the gradual growth of diamond. At intermediate stages a combination of different carbon phases was observed. The atomic structure of the interface phases is discussed and a growth model is proposed. Valence band spectra of the different samples show the extreme sensitivity of PES to impurities and to surface properties such as reconstruction. The obtained data were compared to valence band measurements of natural diamond and other forms of carbon and to some extent to data obtained with Raman spectroscopy.     

Thermionic electron emission from low work-function phosphorus doped diamond films

Thermionic electron emitters are a key component in applications ranging from travelling wave tubes for communications, space propulsion and direct energy conversion. As the conventional approach based on metallic emitters requires high operating temperatures the negative electron affinity (NEA) characteristic of diamond surfaces in conjunction with suitable donors would allow an electronic structure corresponding to a low effective work function. We have thus prepared phosphorus-doped polycrystalline diamond films on metallic substrates by plasma assisted chemical vapor deposition where an NEA surface characteristics was induced by exposure of the film surface to a hydrogen plasma. Thermionic electron emission measurements in an UHV environment were conducted with respect to the Richardson–Dushman relation observing an emission current at temperatures < 375 °C. Measurements were terminated at 765 °C without significant reduction in the electron emission current indicating a stable hydrogen passivation of the diamond surface. A fit of the emission data to the Richardson equation allowed for the extraction of emission parameters where the value of the materials work function was evaluated to 0.9 eV. This value could well be the lowest measured work function of any known material.     

Thermionic emission from phosphorus (P) doped diamond nanocrystals supported by conical carbon nanotubes and ultraviolet photoelectron spectroscopy study of P-doped diamond films

Materials with low work function values (< 2 eV) are highly in demand for low temperature thermionic electron emission, which is a key phenomenon for waste heat recovery applications. Here we present the work function reduction of phosphorus (P) doped (i) diamond nanocrystals grown on conical carbon nanotubes (CCNTs) and (ii) diamond films grown on silicon substrates. Thermionic emission measurements from phosphorus doped diamond crystals on CCNTs resulted in a work function value of 2.23 eV. The CCNTs provide the conducting backbone for the P-doped diamond nanocrystals and the reduced work-function is interpreted as due to the presence of midband-gap state and no evidence for negative electron affinity was seen. However, ultraviolet photoelectron spectroscopy studies on phosphorus doped diamond films yielded a work function value of ~ 1.8 eV with a negative electron affinity (NEA) value of 1.2 eV. Detailed band diagrams are presented to support the observed values for both cases.     

Cathodoluminescence of phosphorus doped (111) homoepitaxial diamond thin films

A homoepitaxial diamond film with a thickness of approximately 1 μm was grown on the (111) surface of a type Ib diamond substrate with dopant concentration (PH₃/CH₄) of 500 ppm. Bound exciton recombination radiation due to the neutral donors has been observed in the peaks at 5.18 and 5.32 eV. The dominant peak at 5.18 eV is associated with one TO phonon. The peak at 5.32 eV is the radiation without phonon emission due to the localization of donors. This appearance of non-phonon emission is one of the important evidences to prove the two peaks at 5.18 and 5.32 eV to be the recombination radiation of bound exciton due to the donor. The binding energy of the free exciton to the donors, which is deduced from the energy difference between free and bound exciton, is 90 meV.     

Characterization of substrate off-angle effects for high-quality homoepitaxial CVD diamond films

We have studied the substrate off-angle effects for the crystalline quality of the homoepitaxial diamond films mainly by using steady-state cathodoluminescence (CL) and time-resolved photoluminescence (PL) measurements. By means of the microwave plasma chemical vapor deposition method under high-power microwave power with high methane concentrations, the homoepitaxial diamond films were grown on the high-pressure/high-temperature-synthesized (HPHT) Ib (001) substrates inclined along either <110> or <100> direction by different off-angles ranging from 2° to 5°. In spite of high growth rates, we have succeeded in improving crystalline quality by employing the HPHT substrates with considerably large off-angles. Both steady-state CL and time-resolved PL measurements clearly indicate that larger off-angles lead to better crystalline quality of the homoepitaxial film, suggesting that further improvements in crystalline quality can be expected when using substrates having even larger off-angles.     

NEXAFS spectroscopy of CVD diamond films exposed to fusion relevant hydrogen plasma

A series of CVD diamond films have been exposed to hydrogen plasma in the linear magnetized plasma device, MAGPIE, with various applied sample stage biases between 0 V (no applied bias) to − 500 V. The plasma-induced damage to the surface structure of the diamond films has been investigated by Near Edge X-ray Absorption Fine Structure (NEXAFS) spectroscopy in both the Auger electron yield (AEY) and total fluorescence yield (TFY) modes. The key diamond NEXAFS spectral features (diamond core exciton and second absolute band gap) are found to be diminished following plasma exposure as measured in the surface sensitive, AEY spectra, whilst these features remain unchanged relative to an unexposed diamond reference film as measured using the bulk sensitive, TFY spectra. These results, in conjunction with SRIM simulations, show definitively that the damage to the surface of the diamond films is restricted to the scale of the penetration depth of the H ions and no damage is induced at greater depths. The power and sensitivity of NEXAFS spectroscopy in assessing damage to the surface of diamond from fusion-relevant plasma-surface interactions are demonstrated.     

机械法抛光加工金刚石膜研究

在平面研磨机上使用金刚石抛光盘对金刚石膜进行了抛光实验。通过观察金刚石膜机械抛光过程中表面形貌的变化，并结合拉曼光谱分析，提出了金刚石抛光盘抛光金刚石膜的抛光机理主要是微切削和压力破碎作用，同时还可能有少量金刚石转变为非晶碳的作用。通过单因素实验研究，发现金刚石盘的粒度对抛光金刚石膜材料去除率的影响最为显著。转速越高，表面粗糙度越小，压力和粒度存在最优值，超过该值后，表面粗糙度并不会随之降低。     

Characterization of d.c.jet CVD diamond films on molybdenum

Diamond films grown using a thermal plasma technique are characterized using a variety of techniques. The relationships between the chemistry, morphology, and mechanical properties are explored using microscopy, Raman spectroscopy, Auger electron spectroscopy, and X-ray photoelectron spectroscopy. The characteristics of films grown using two different nucleation enhancement techniques are shown. Films grown using high methane concentrations at the beginning of growth produce large grained columnar films, whereas films grown on substrates which have been treated with a diamond polishing step show nanocrystalline structures. Variations in sp³ and sp² bonding and peak shifts are tracked through the thickness of the film, corresponding to variations in the methane concentration during growth. Stresses are measured using peak shifts and beam bending techniques. Adhesion is tested using indentations, and is shown to increase both as growth temperatures and surface roughness increase.     

High pressure annealing of CVD diamond films

CVD diamond films were annealed from 600 to 1900 °C at 7.7 GPa in a toroidal high pressure (HP) apparatus, always inside the diamond-phase stability region. The annealed films were analyzed by Raman and infrared (IR) spectroscopy and the results showed that the diamond grains remained stable while the non-diamond carbon phases and impurities, responsible for the intricate film structure, changed after processing. For the HP annealing from 600 to 1300 °C, there were no major changes in the Raman spectra of the film, however, the film became easily broken and the IR spectra indicated a high reactivity of carbon with chemical elements from the environment. After annealing at 1500 °C and 7.7 GPa, the formation of diamond-like (DLC) and graphitic structures in-between the diamond grains were observed, while the reaction with the environment elements decreased. For higher temperatures, the DLC and graphitic structures persisted up to 1700 °C and the film incorporated OH in large amounts. The results showed that the non-diamond carbon species are susceptible to the HP annealing, and structural modifications in between the diamond grains are significant for temperatures above 1300 °C at 7.7 GPa.     

Electronic properties of CVD polycrystalline diamond films

Electronic properties of polycrystalline diamond films have been analysed using steady-state photoconductivity under UV (λ=190 nm) and red (λ=633 nm) illumination. We look at the effect of several post-deposition treatments: annealing in vacuum at 400°C, successive annealing under air and methane or the reverse, and X-ray exposure. These are shown to affect the photoresponses in different manners. Globally, the mobility–lifetime products of samples are found in the range 10⁻⁷–10⁻⁶ cm² V⁻¹ under UV illumination, and ratios of photoresponses in UV and red are in the range 10³–10⁵. First results obtained using the modulated photocurrent technique are also presented.     

砷化镓上生长金刚石薄膜的研究

以CH4和H2为气源,用微波辅助等离子体装置,在10.0 mm×7.0 mm的砷化镓基底上沉积了CVD金刚石薄膜,用扫描电子显微镜观察沉积效果,拉曼光谱表征沉积质量,分析薄膜附着力与砷化镓材料性能的关系。结果表明,当基体温度为600℃,气压为5 kPa,甲烷浓度为2.0%时,在砷化镓片表面上沉积出了CVD金刚石薄膜,晶粒尺寸均匀,晶形完整、规则,晶界非常清晰。     

Photoluminescence studies of type IIa and nitrogen doped CVD diamond

Photoluminescence (PL) studies were carried out on CVD, type IIa and high purity HPHT diamond samples irradiated with electrons of energies between 150 and 300 KeV; near threshold energies for carbon displacement. The majority of PL spectra are obtained using a 488-nm lasing line, with samples cooled to approximately 7 K. Of particular interest is the behaviour of the self-interstitial related centre, 3H, at 503.5 nm. The centre is particularly sensitive; its formation varies significantly with dose and dose rate and is severely quenched with incident laser power in excess of 10 mW. 3H is the dominant centre in highly doped (50–100 ppm) nitrogen samples, for doses between (10¹⁹–10²⁰) el/cm², but reduces with higher doses. In lower nitrogen (few ppm) samples, the centre is considerably weaker after equivalent doses, comparable to the Raman line. In type IIa crystals, creation of 3H varies considerably from sample to sample. Upon annealing, 3H is at an optimum between 310 and 330 °C for type IIa diamonds and vanishes by 400 °C. Indications show these temperatures increase slightly as nitrogen content is increased. Migration of the centre well outside the irradiated area is frequent, tens of microns after irradiation and hundreds of microns post annealing. Other centres of interest include GR1, the neutral diamond vacancy, which is found to be created linearly with dose and be rate independent. Using 325 and 457.9 nm lines the TR12 centre was studied. It has a strong dose rate dependence, growing as dose rate raised to a power of approximately 2 and is unaffected by annealing up to 700 °C. A 244-nm line was used to study the 5RL centre and contrary to some reports was observed in samples containing approximately 0.1 ppm of nitrogen. PL provides an extremely sensitive way of measuring the nitrogen concentration in diamond, to levels of less than 0.1 ppm. The problem remains how to obtain an accurate measurement.     

X-ray topography studies of dislocations in single crystal CVD diamond

X-ray topography has been used to study single crystal diamond samples homoepitaxially grown by microwave plasma-assisted chemical vapour deposition (CVD) on high pressure high temperature (HPHT) and CVD synthetic diamond substrates. Clusters of dislocations in the CVD diamond layers emanated from points at or near the interface with the substrate. The Burgers vectors of observed dislocations have been determined from sets of {111} projection topographs. Dislocations have line directions close to the [001] growth direction and are either edge or 45° mixed dislocations. Where groups of dislocations originated at isolated points they tended to be of the edge variety. Where the substrate surface was deliberately damaged before growth, two sets of dislocations were observed to have propagated from each line of damage and there was a tendency for dislocations to be of the 45° mixed variety with a component of their Burgers vector parallel to the polishing direction. It is demonstrated that X-ray topography can be used to deduce the growth history of CVD synthetic diamond samples produced in multiple growth stages.     

Photoluminescence and positron annihilation measurements of nitrogen doped CVD diamond

A range of techniques have been used, to determine the nitrogen levels in a series of polycrystalline diamonds. The crystals were grown by the chemical vapour deposition (CVD) technique and EPR measurements indicate that they have a single substitutional nitrogen (N_s) concentration between 10 ppb and 50 ppm. Photoluminescence (PL) spectroscopy provides an extremely sensitive way of detecting the nitrogen vacancy complex, in both the neutral (N–V)⁰ and negative charge states (N–V)⁻, down to the 10 ppb range and below. It has been observed that for diamonds with a single substitutional nitrogen content of approximately (0.5–2.0) ppm, the nitrogen vacancy complex exists in both charge states, with an almost equal abundance. Below this level, the complex exists with greater regularity in the neutral charge state, with the (N–V)⁻ centre dominating at (N_s) levels greater than 2 ppm. Electron irradiation of so-called ‘high nitrogen’ films, with an (N_s) value in excess of 15 ppm, show an incredibly high abundance of the interstitial related 3H centre. Such samples show little sign of the neutral vacancy, GR1, although some absorption is evident in the tail of the (N–V)⁻ centre. No evidence of the negative vacancy, the ND1 centre, was observed in PL spectra. For equivalent electron doses on ‘low N’ samples, typically with an (N_s) level of below 1 ppm, the GR1 centre completely dominates the PL spectra. For all diamonds the intensity of the nitrogen vacancy complexes are seen to reduce within the irradiated region. Annealing studies of high N samples reveal that by 900 °C all the vacancies are annealed with many being trapped at the nitrogen to form a large concentration of the (N–V)⁰ centre. Positron annihilation measurements have been performed on the samples. As positron annihilation is sensitive to vacancy concentrations, it is expected to yield lifetime measurements that are strongly dependent on the bulk (N–V) concentration. Doppler broadening spectra have been recorded as the samples were illuminated with light varying from 325 to 785 nm. The shape ‘S’ parameter was found to increase upon illumination and relax back to its ground state, post-illumination. The relaxation time was found to vary between 30 and 200 hours and be strongly dependent on both the nitrogen concentration and the wavelength of light used. Work is currently on-going, and includes angular correlation of annihilation experiments. Lifetime measurements show clear evidence of the formation of positronium in the voids that are present in the CVD material.     

Secondary electron emission from CVD diamond films

Secondary electron emission from boron doped diamond polycrystalline membranes (hole concentration 5×10¹⁸ cm⁻³), prepared by microwave plasma assisted CVD, was investigated in both the reflection and transmission configurations. The model of secondary electrons behavior taking into account the distribution and diffusion mechanism of secondary electrons is proposed to explain the yield dependencies on primary electron energy in both configurations. The model predicts the SEE yield K=19 at the primary electron energy E₀ close to 1 keV for reflection configuration and K=3–7 at E₀=15–30 keV for transmission configuration for polycrystalline films used in the study. Experimental measurements of the SEE yield vs. primary electron energy (18 at E₀=950 eV for the reflection scheme and 3.5–4 at E₀=25 keV for the transmission one) are found to accord well with the theoretical results. Estimations, which were made using the model, show that SEE yield in transmission configuration can be increased up to 60 for the primary electron energy of about 10 keV. Since such high yields in transmission scheme may be obtained in monocrystalline membrane, another approach using porous polycrystalline diamond membranes is considered. Porous diamond membranes having SEE yield in transmission scheme of more than 10 at the primary electron energy E₀=1 keV were fabricated.     

Highly and heavily boron doped diamond films

From the high ionization energy E_i = 0.368 eV and high solid solubility ≥ 1.4 × 10²² cm^− 3 of boron in diamond, metallic conductivity is expected on the boron impurity band within the band gap (Mott model). On the contrary, the numerical models used to describe the superconductivity of metallic diamond mainly use the Bardeen model with a Fermi level within the valence band. Taking into account the decrease to zero of E_i through the high [B] range and the band gap narrowing through the high and heavy [B] ranges, both specific of the Bardeen model, we discuss the validity of the Mott and Bardeen models from the literature and cathodoluminescence and Raman experiments. They agree with the Mott rather than the Bardeen model. Several experiments independently show the coupling of boron related levels with the zone centre optical phonons which soften for heavy [B]. The Mott model might explain the similar range of the superconductivity temperature of homoepitaxial and polycrystalline films from the similarity of their boron impurity band.