Search

The p-to-n-type conversion of particular B-doped homoepitaxially grown diamond layers upon deuterium plasma treatment was discovered three years ago. However, many questions regarding the reproducibility of the effect for samples of different origins remain unanswered up to now, in particular the role of the electrical contacts and the possibility of a surface inversion layer being responsible for the n-type conductivity, the thermal stability and origin of the donor. Here we address the above questions. We show that the p-to-n conversion is closely related to the presence of defects, which can vary, on a macroscopic scale, in different regions of the same sample. The p-type regions containing a small concentration of defects are found to be converted to n-type upon deuteration. In contrast, the regions counting a high defect density exhibit very low p-type conductivity related to passivation of the boron acceptors by deuterium accompanied by a large uptake of D on defects. We show that the n-type conversion can be equally observed with three different kinds of contacts (silver paint, implanted and evaporated). We prove that the n-type conversion is a bulk effect and not a surface effect. We find that the thermal stability of the n-type complex is limited to temperatures lower than 200 °C. The temperature dependence of the carrier concentration measured by the Hall effect at different stages of the conversion process is fitted following the formalism describing the conduction mechanisms in a partially compensated semiconductor. The fact that the donor concentration in the as-deuterated state is very close to the boron concentration strongly suggests that the new-formed donor complexes contain both D and B.

From the high ionization energy Ei = 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 Ei 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.

A general view of the key issues which have to be surmounted for building elementary devices such as Schottky diodes on homoepitaxial p-type diamond layers is presented. The growth process of double homoepitaxial layers, comprising a stack of both high and low boron doped films, is described and boron concentrations are analysed by means of cathodoluminescence. The methods used for masking, performing lithography on 3 × 3 mm 2 samples and etching are delineated. Several ohmic contacts are compared. Surface treatments necessary for the Schottky diode technology are analysed and discussed. Two methods for achieving adherent Schottky contacts are compared and the electrical characteristics of diodes are studied as a function of the morphology, defect content and doping of the diamond layer, metal stack which is used, temperature history of the sample and passivation post-processes such as dry etching.

A set of MPCVD films was prepared from 0.5% CH 4 /H 2 and B 2 H 6 with B/C in the gas phase from 0 to 8000 ppm. When the excitation wavelength of their Raman spectra is switched from 632.8 to 325 nm: The Fano induced dip switches from high to low wave number side of the “1332 cm − 1 ” signal. The deformation and the downward shift of this signal decrease. The G band, diamond second order, CH x bands appear. The evolution of these bands with increasing B / C is reported and discussed.

Homoepitaxial films containing from 2 × 10 16 to 1.75 × 10 21 B−cm − 3 are deposited by Microwave Plasma Chemical Vapor Deposition on Ib substrates. From their full continuity for all [B] with those found for the lower [B], the high energy structures in their cathodoluminescence spectra originate from recombination of free and boron bound excitons, with no signal from band to band transitions. Therefore, the metallic conduction takes place on a boron impurity band with a Fermi level outside the valence band in metallic diamond, in agreement with the original Mott model. The energy for the recombination of the boron bound exciton assisted by the TO phonon decreases from 5.211 to 5.036 eV when [B] increases from ≈ 1.5 × 10 19 to 6 × 10 20 cm − 3 , then remains constant. An interpretation of the results somewhat similar to that used for GaP:N might account for the excitons bound to the boron atoms of the impurity band.

The conversion of the conductivity of B-doped homoepitaxially grown diamond layers from p- to n-type upon deuteration and its reconversion to p-type following annealing is extensively studied. Several B doped samples have been converted to n-type when exposed to a deuterium plasma at approximately 500 °C. The n-type features are related to D uptake of the samples. The donors, thus formed, have an ionization energy of approximately 0.34 eV below the conduction band edge and a high RT (room temperature) mobility (up to 430 cm2/V s), as determined by Hall effect measurements as a function of temperature. In the B-doped layers that do convert to n-type upon deuteration, the D to B concentration ratio is found to be in the range of one to two. Higher D uptake (D to B ratio ≫2) must be connected to B-doped layers containing a large amount of growth defects, which trap the D. The exact nature of the donor is yet unknown. However, the reversibility of the effect and its relationship with the D concentrations suggest that deuterium is involved in the formation of some complex with B or with some other defects.

We study the Raman spectra of heavily boron ([B]>3×1020 cm−3) polycrystalline and homoepitaxial diamond films. A detailed comparison of the structures of these spectra can be done only after normalizations. As [B] increases in polycrystalline and homoepitaxial films, (i) the ‘1225 cm−1’ appear identical (ii) the maximum of the ‘500 cm−1’ band shifts to lower wavenumbers while its intensity increases (ii) the overall shapes of the ‘500 cm−1’ bands remain similar while their full widths at half maximum are higher in the polycrystalline films. From these characteristics and previous works, the ‘500 cm−1’ band is proposed to originate from local vibrational modes of boron pairs.

Diamond deposition is performed at 900 °C on Ib substrates in the middle of the plasma ball from gas mixtures containing 1% CH 4 and 0, 0.125, 0.25, 0.375 and 0.5% O 2 at a pressure of 50 Torr. Smooth homogeneous homoepitaxial films of good quality whose characteristic depends systematically on the percentage of O 2 are obtained on the backside of the substrates. Weak structures (in the 10 −5 –10 −3 range compared to the 1332 cm −1 peak) appear in their Raman spectra excited by 632.8 or 514.5 nm lasers. They correspond to (i) N-vacancy centers (2.156 eV luminescence), but no Si-vacancy center (no 1.681 eV luminescence) (ii) Ultra nanocrystalline (diameter 4–6 nm) diamond and micro crystalline graphite residual phases which are detected from their specific signals. Lower concentrations of both residual parasitic phases and N complexes are obtained in the 0.125–0.375% O 2 film.

Schottky contacts on the (1 0 0) surface of homoepitaxial boron-doped diamond films are investigated. The potential barrier height of a metallic interface build in a ultra-high vacuum (UHV) chamber by the deposition of one monolayer of titanium, subsequently covered with a thick film of silver, is studied. From X-ray photoelectron spectroscopy spectra, a barrier height of 1.7 eV is derived. However, after annealing in air, the current–voltage characteristics analysed at several temperatures allow the plot of a Richardson's diagram, which yields a 3.4 eV barrier height, higher than half the diamond bandgap. Annealing of the Ti layer at 400 °C in UHV induces carbide formation, which ensures a good adhesion of the metallic coating. Comparison of the series resistance of diodes implemented either only on a moderately doped diamond layer (a few 1017 cm−3) or on p/p+ double layer is done. In the former case, the 0.38 eV activation energy for boron acceptors ionisation is clearly identified while it decreases to 0.06 eV in the latter case, with an absolute value of the series resistance typically five orders of magnitude lower at ambient temperature. Consequences and applications of these findings are presented.

BOLD (Blind to the Optical Light Detectors) est un projet regroupant plusieurs laboratoires, instituts, et industriels europeens ; il est dedie aux developpements et aux tests de nouveaux detecteurs pour l'observation du Soleil dans l'ultraviolet. Notre investigation repose sur les proprietes des solides a large bande interdite, tels que le diamant ou les nitrures d'aluminium et de gallium (AlGaN). L'initiative est proposee en vue des instruments UV du futur Orbiteur Solaire de l'Agence Spatiale Europeenne (ESA), mais plusieurs autres applications scientifiques ou industrielles en beneficieront. L'insensibilite au visible (solar-blindness) des nouveaux senseurs sera tres avantageuse. Elle permettra une plus grande sensibilite des instruments, et donc plus de resolution spatiale, et une plus haute cadence. Une technologie fondee sur le diamant ou sur les AlGaN plutot que sur le silicium autorisera aussi une resistance accrue aux rayonnements ionisants (rad-hardness), et donc la stabilite indispensable aux conditions extremes que connaitra l'Orbiteur Solaire. Nous presentons ensuite quelques resultats de campagnes de mesures entreprises dans le cadre du projet BOLD, et decrivons le programme envisage pour obtenir en quelques annees des cameras UV tres innovantes.

Hydrogen is a key impurity in diamond since it is unintentionally incorporated in all chemical vapor deposition (CVD) grown diamond layers.Its presence in the material can grossly affect its electrical and optical properties.Theoretically, hydrogen has been predicted to be present in diamond in one of the three charge states, H , H and H .Moreover it may form complexes q 0 y with impurities, native defects or with other hydrogen atoms.This paper is comprised of two parts: (a) a review of previous results of studies investigating different aspects of the diffusion of hydrogen (deuterium) in various kinds of diamonds.The diamonds studied are: undoped type IIa diamonds, undoped CVD diamond layers containing growth defects only, p-type B-doped homoepitaxially CVD grown diamond layers or B ion implanted type IIa diamonds and n-type P doped homoepitaxially CVD grown diamond or N-doped type Ib natural diamonds.Hydrogen is introduced in diamond by exposing the diamond surface to hydrogen plasma or by using hydrogen ion implantation.The following issues are discussed: (1) the influence of the interaction between H and the dopants and defects on the hydrogen diffusion. (2) The kinetic of (B, H), (P, H ) and (N, H) pair formation and dissociation. (3) The modification of the optical and electrical properties as a result of hydrogen incorporation and annealing. It is found that, under certain conditions, H diffuses into the B containing layer and it passivates B acceptors.In contrast, no H diffusion could be observed in n-type diamonds, up to 1000 8C. (b) Recent results of our group regarding other aspects related to the diffusion of H in diamond are presented.These include results on: (i) the influence of ion implantation related defects on the diffusion of deuterium.For this study type IIa samples implanted with B or non-dopant ions are used. (ii) The determination of the charge state of H or Hydefects complex as a function of diamond type.For that, annealing under bias is applied to deuterated diamond layers.We show that the presence of implantation defects retards the deuterium diffusion in a B-implantation doped diamond, demonstrating that D strongly interacts with defects, thus inhibiting diffusion.The new-formed complexes deteriorate the electrical properties of the diamonds and are very stable up to high temperatures.We confirm that, as expected, in highly B-doped CVD diamond layers, H diffuses as a positive ion.In lightly B-doped homoepitaxial layers, however, D is incorporated in complexes which seem to be negatively charged. � 2003 Elsevier Science B.V. All rights reserved.

Hydrogen diffusion is investigated in boron-doped diamond from the redistribution profiles under thermal annealing of an initial step-like profile of (B,H) pairs located in a shallow region below the surface. For boron concentrations in the range 2×1019–1×1020 cm−3, the profiles are Gaussian-like and effective diffusion coefficients can be calculated. From the diffusion activation energy, we deduce a (B,H) pair dissociation energy of 2.5 eV. For lower boron concentrations, the redistribution of hydrogen is inhibited by the presence of deep traps for hydrogen which might be associated with structural defects.

c ´´ Abstract The incorporation of hydrogen in n-type diamond has been investigated through its non-intentional introduction during the growth and through post-growth diffusion experiments. In the first case, (111) n-type diamond homoepitaxial films doped with phosphorus have been grown by microwave plasma chemical vapor deposition from a CH qD mixture. The phosphorus 42 concentrations were in the range 1.3=10 -3=10 cm . The concentration of deuterium incorporated is usually 100 times 18 19 y3

Simultaneous etching of NID, lightly boron-doped, phosphorus-doped homoepitaxial CVD films and Ib HPHT crystal was achieved at RT in a home-made reactor producing O2 ECR plasma which was homogeneous on approximately 200 cm−2. Etching rates varying from 60 (at −30 V) to 40 or 120 nm/min (at −140 V) are obtained with n- or p-type-doped samples, respectively. These etching rates are higher than those with O2 RF plasma, and those with previous inhomogeneous O2 ECR plasma with samples at 100 °C.

b ¨¨ Abstract The n-type doping of diamond by sulfur and phosphorus has been investigated. Diamond thin films were grown epitaxially on (111) oriented synthetic type Ib crystals with phosphine added in the reactant gas (wPxywCxs100-1500 ppm). Each grown sample was characterized by FTIR absorption and conductivity vs. temperature measurements. The infrared signature of neutral substitutional phosphorus appeared as two peaks located at 523 and 562 meV. Activation energy of the conductivity was approximately 490-550 meV above 400 K. All characterizations performed so far suggest a successful n-type doping by phosphorus. Sulfur-doped diamond thin films were grown on (100) oriented synthetic Ib crystals with H S (wSxywCxs500-5000 2 ppm) added in the gas phase. Secondary electron microscopy, AFM and Raman scattering indicate a high quality homoepitaxy of the diamond film. The addition of H S induces a light green-gray or gray color of the doped film. The films are highly resistive 2 after growth and oxidation of the surface, as usual undoped diamond. However, after light-illumination the electrical and photoelectrical properties change drastically. The results are interpreted in term of compensated semiconductor, taking into account contamination by boron atoms. 2002 Elsevier Science B.V. All rights reserved.

MicroRaman signals excited at 457.8 nm have been obtained from some areas of heavily boron doped polycrystalline diamond films which have been used for electrochemical reactions. We could identify sp3 CH, CH2, CH3, all with narrow full width at half maximum. From comparisons to IR absorption from the bulk of undoped films and from the (111) and (100) surfaces of IIa crystal, and to Raman signals from the bulk of undoped polycrystalline films, we ascribed our micro Raman signals to CHx from the film surface. The sp3 CH2 and CH3 components appear more stable than the sp3 CH under CW laser exposure. This implies chemical bonding of H with C atoms of the film surface for all the sp3 CHx sites. This difference with the CHx at the surface of IIa crystals might originate from the heavy boron-doping of our films.

A long-term decay of the photocurrent under red light has been observed and analysed in polycrystalline CVD diamond films. The kinetics of this photocurrent decay is non-exponential and can rather be described by a combination of several exponential functions or by a stretched-exponential function. This relaxation is only observed if the samples are preliminary exposed to band-to-band UV light excitation during a few minutes and it is perfectly reproducible. This long-term relaxation of the below-gap photocurrent following UV exposure is not observed on a homo-epitaxial crystalline CVD layer, where the photocurrent remains almost constant. Therefore, it is related to the disorder in the CVD polycrystalline diamond films and to the broad distribution of localised gap states, which are mainly created by the grain boundaries. The results are discussed in terms of defect influence in a range of diamonds in relation with their nature and synthesis process.

BOLD (Blind to the Optical Light Detectors) is an international initiative dedicated to the development of novel imaging detectors for UV solar observations. It relies on the diamond and nitride materials that have lately undergone key advances. The investigation is proposed in view of Solar Orbiter UV instruments, for which the expected properties of the new sensors—visible blindness and radiation hardness—will be highly beneficial. Solar Orbiter is a selected Flexi mission of the European Space Agency (ESA). Despite various improvements over the last few decades, the present UV detectors exhibit limitations inherent to their actual technology. Yet the utmost spatial resolution, temporal cadence, sensitivity, and photometric accuracy will be decisive for the forthcoming space solar missions. The advent of imagers made of a large bandgap semiconductor would surmount many weaknesses, thus opening up new prospects and making the instruments cheaper. As for the ESA Solar Orbiter, the aspiration for wide bandgap semiconductor-based UV detectors is still more sensible, for the spacecraft will approach the Sun where the heat and the radiation fluxes are high. We depict motivations and present activities and programme to achieve revolutionary flight cameras within the Solar Orbiter schedule.

Point contact methods using metallic needles are widely applied for electrical measurements on the diamond semiconductor surfaces because of the difficulties of alloyed ohmic contact preparation. Contradictions and discussions in the literature concerning the resistance values and the shape of I–V characteristics made necessary a detailed study of the point contact measurements. In this work two- and four-point methods were applied to HPHT bulk crystals and MPCVD monocrystalline epitaxial layers in order to evaluate the applicability of these methods, to estimate the current conductivity character of such metal tip–diamond junctions, and to measure the electrical resistance of these crystals. Some samples underwent various treatments: annealing in vacuum; hydrogenation in molecular hydrogen; deuteration in RF and MW plasmas. Conductivity character of the contacts was determined and the applicability of point contacts was demonstrated. In contrast to some earlier results, only linear I–V characteristics were found. Conducting superficial layer connected with the presence of hydrogen was not observed. After the microwave deuterium plasma treatment, just the contrary, a highly resistive superficial layer posed obstacles before the measurements.

A set of type IIa diamond crystals was processed by cold implantation and rapid annealing at 1050°C ex situ annealing at 1400°C and investigated by ESR measurement. The ESR spectra of CIRA and post-annealed samples show an isotropic line at g =2.003, a set of anisotropic hyperfine lines, and a pair of isotropic hyperfine lines centred at g =2.003 with Δ H =28 G. The intensities of each signal increase with the implantation dose. After the second annealing step the intensities of all signal was reduced, excepted for the centre with Δ H =28 G which increases. This signal was ascribed to a new centre related to phosphorus, probably neutral substitutional phosphorus.

Hydrogen-acceptor interactions are investigated in boron-doped diamond through deuterium diffusion experiments followed by SIMS measurements and through infrared absorption spectroscopy. From deuterium diffusion, we show that BD interactions can be properly demonstrated in low compensation B-doped homoepitaxial layers. However, the presence of defects in such layers strongly affects this interaction. The degree of passivation of boron acceptors by deuterium depends on the diffusion temperature. At 550°C or below, the B and D concentrations exactly match giving rise to a complete disappearance of the absorption bands related to the electronic transitions of neutral boron acceptors. Under thermal annealing above 500°C, (B,D) pairs dissociate and neutral boron acceptors recover. At deuterium diffusion temperatures of 700°C, the B passivation is absent.

Drastic differences in diffusion of H (deuterium) in diamond, B-doped by ion implantation and during homo-epitaxial film growth, and its influence on electrical properties are found by SIMS depth profiling, and by electrical (Hall effect) measurements. Type IIa natural diamond, B-doped by ion implantation and high quality homo-epitaxial B-doped diamond films were subjected to D plasma treatment under similar conditions. The results of SIMS measurements clearly show a huge difference in D diffusion profile for these two samples. While the sample doped during growth was totally deuterated, the implanted one showed only minor D penetration. Electrical measurements indicated that while the homo-epitaxial samples became insulating or showed strong decrease in their free hole concentration following deuteration, the identical treatment to the B-ion implanted sample caused only slight changes in electrical properties. The electrical properties and their dependence on annealing are correlated with the deuterium diffusion into diamond. A possible mechanism of (B, H) and (defect, H) pair formation is suggested as a possible explanation of the observed differences.

As the boron incorporation level and the wavelength of the exciting laser were varied, we observed systematic modifications of the Raman spectra of homoepitaxial diamond films. A pronounced change in the lineshape of the zone-center optical Raman peak as well as a wide and structured signal at lower wavenumbers appeared simultaneously when the boron incorporation was increased above an abrupt threshold around 3×10 20 cm −1 . This threshold was found to depend on the excitation laser wavelength. Possible origins for the wide peaks at 500 and 1225 cm −1 are also discussed.

Investigation of deep levels in the depletion zone of boron-doped homoepitaxial diamond films is achieved from analysis of the transient responses of Schottky junctions. The active layer consists of a homoepitaxial film grown by microwave plasma decomposition of a very small percentage of B2H6 mixed with CH4 in hydrogen on synthetic Ib diamond substrates. Rectifying contacts are implemented with either aluminium, gold or erbium after suitable pre- and post-treatments. Reverse-current transients due to the thermal emission of holes from deep-level traps into the valence band are recorded and analysed with a Fourier transform deep-level transient spectroscopy (Fourier DLTS) procedure as a function of temperature up to 520°C. At each temperature, a thermal emission time constant is extracted. Arrhenius diagrams are deduced. A deep level at 1.40±0.07 eV above the valence band is detected in samples that have been submitted to a 800°C annealing in ultrahigh vacuum. Another deep level at 1.16±0.03 eV above the valence band exists only if an oxygen radio-frequency plasma treatment is applied to the sample before the Schottky contact deposition. Thermal capture cross-sections are also deduced.

The diffusion and out-diffusion properties of hydrogen have been investigated in homoepitaxial layers of CVD boron-doped diamond which have been exposed to a deuterium microwave plasma. The diffusion experiments show that the onset of (B, H) pair dissociation occurs at 550°C. The hydrogen diffusion activation energy is 1.4 eV. Hydrogen out-diffuses at 880°C and the diamond surface barrier limits the out-diffusion process. In r.f. plasma exposed diamond epilayers, the hydrogen out-diffusion spectra are composed of a peak at 720°C related to the subsurface hydrogen accumulation layer and another one at 840°C attributed to the hydrogen adsorbed at the surface.

The Raman spectra and voltammograms in 0.1 M KCl, 0.1 M KNO 3 and 0.1 M KCl+0.1 M KNO 3 solutions of polycrystalline diamond electrodes vary systematically versus their boron content in the gas phase (200 to 14 000 ppm). Both exhibit the same threshold (2000 ppm, about 1.2×10 20 B cm −3 ) corresponding to the appearance of a low-wavenumber continuum in Raman spectra and to a factor of 10 3 jump in the electrochemical current. In 0.1 M KNO 3 solution, the reduction of nitrates from −0.6 to about −2 V cathodic potential occurs with low hydrogen reduction. The correlation between the electronic and electrochemical properties of these electrodes is discussed.

Undoped polycrystalline diamond presents two thermoluminescent (TL) peaks at 480 and 560 K which may potentially be used for ionising radiation dosimetry. In order to increase the TL response sensitivity, CVD diamond samples doped with various boron concentrations have been elaborated. The aim of this work is to investigate the TL properties of these samples in relation with the boron concentration. For concentrations lower than 15 ppm (3×10 18 cm −3 ), the intensity of the 480 and 560 K TL peaks are effectively the most important, but for concentration values higher than 15 ppm, the TL signal decreases. Simultaneously, a new peak appears at 340 K. TL mechanisms related to these peaks are discussed.

In this paper, the evolution of the infra-red absorption of high-quality epitaxial diamond films versus boron concentration is reported. Homoepitaxial diamond films were grown by microwave-plasma-assisted chemical vapour deposition on type Ib crystals. From 2×10 17 to 8×10 20 cm −3 of boron were incorporated in the diamond from the gas phase during growth. The main absorption features are (1) a single phonon absorption around 160 meV, (2) a series of three lines at 304, 347 and 363 meV due to excited states of the bound hole, (3) a photoionisation continuum, and (4) a phonon replica of these electronic transitions. The evolution of these absorption with the boron concentration is presented and discussed. A strong increase in the photoionisation cross-section for boron concentrations above 5×10 18 cm −3 is observed. New linear relationships between the boron concentration and the integrated absorption at 347 meV and the photoionisation are proposed. From the evolution of the lines due to transitions to boron-excited states, we deduced the width of the impurity band versus the boron concentration.

High-quality homoepitaxial diamond films have been grown on 100-type Ib synthetic diamond, and boron doped by diborane from 5×1016 to 8×1020 cm−3. The current–temperature characteristics from 300 to 1000 K have been measured and fitted by the general expression of the conductivity in a compensated semiconductor in the ionisation regime. The saturation of the conductivity is observed between 680 and 1000 K for the samples doped below 2×1017 cm−3. In the range 1018–1019 cm−3, an activation energy of 185 meV is observed between 500 and 1000 K and around 368 meV between 300 and 500 K. The activation energy of 185 meV corresponds to half the boron ionisation energy, and is related to the conduction regime in an uncompensated semiconductor. The evolution of these activation energies and of the room-temperature resistivity with the boron doping are presented.

The various phases obtained after annealing of molybdenum films on diamond free and encapsulated by Si3N4 are determined from Rutherford backscattering spectrometry and X-ray diffraction. Virgin molybdenum films always contain some oxygen, which hinders carbon diffusion and prevents the formation of Mo2C. Oxygen is removed by annealing of the free molybdenum films above 700°C, and full formation of Mo2C is obtained at 1000°C. However, annealing at 1100°C induces the formation of a superficial carbon layer, with carbon inclusions in the carbide. A two-step process, with annealing of the molybdenum virgin film at 900°C, then from 100 to 1100°C after encapsulation with Si3N4, allows the formation of the suitable full Mo2C film, with less stringent conditions.

The X-ray diffraction peaks of undoped and boron-doped polycrystalline diamond films have been recorded up to 8 × 1020 B-cm−3. The lattice parameter varies slightly according to the crystallographic direction [111], [220] and [311] of the crystalline perpendicularly to the substrate. For undoped films, it is lower in the [220] direction. In all directions, we measure a large expansion coefficient of the diamond lattice versus the boron incorporation [B]. Microstrains appear independent of [B], while coherent domains have a maximum size around 3 × 1019 B cm−3. We suggest contribution of both the different of size of boron and carbon atoms (Vegard's law) and of holes in the impurity band of boron (but with a positive “deformation potential”) to describe the lattice expansion from boron incorporation.

Mo films are deposited at 500 K and 300 K by simple (sample I) and magnetron (sample II) sputtering respectively on diamond films formed by microwave plasma chemical vapour deposition. The samples are then annealed from 500 to 1100 °C. Although as-deposited samples incorporate O, significantly for sample I and weakly for sample II, they form respectively ill-crystallized β-Mo 2 C and well-crystallized α-Mo 2 C. Well-crystallized α-Mo 2 C appears after annealing of sample I at 900 °C which induces a drastic decrease in the O concentration. The high resistivity, 420 μΩ cm, of as-deposited sample I decreases to 40 μΩ cm after annealing at 900 °C. Similarly, the as-deposited resistivity of 120 μΩ cm of sample II decreases after annealing above 600 °C, which reduced its O content. The α-Mo 2 C resistivity is determined to be 80 μΩ cm, lower resistivities of the metallic films originating from Mo in Mo-α-Mo 2 C mixtures. Annealing at 1100 °C induced the formation of a superficial carbon film.

A boron-implanted marker and secondary ion mass spectroscopy analysis were used for diamond etching assessment. Trace amounts of oxygen in argon etch polycrystalline diamond films during annealing at 800 °C. The deposition of a protective PECVD hydrogenated amorphous silicon nitride layer prevents such etching during annealing at least up to 1300 °C. The hydrogen losses of Si3N4 during heating, from 10% after growth to 0.5% at 1000 °C, do not reduce the diffusion barrier efficiency. This is attributed to Si-N bond reconstruction during H effusion. The influence of high temperature annealing on the diamond film was studied by IR transmission spectroscopy and X-ray microanalysis. No carbon compound was found on the diamond surface after annealing and Si3N4 removal. However, a large amount of silicon carbide was observed after annealing above 1100 °C at the interface between the diamond film and the silicon substrate.

Optical absorption coefficients, and especially those below 103cm−1, cannot be determined in diamond thin films (thickness < 10 μm), because of their low values and the very large optical diffusion induced by the surface roughness. Both problems are overcome when the absorption coefficient is derived from the measurement of the heat absorbed in the film by mirage effect (or PDS for photothermal deflection spectroscopy). This technique appears especially well suited to the study of the defects and doping in polycrystalline diamond. For the first time, the absorptance induced by the localized states introduced by implantation of polycrystalline diamond thin films and its recovery after annealing have been shown. In agreement with the results of other techniques, the number of localized states increases with the implantation dose, up to amorphization between 1015 and 1016 B cm−2. For amorphized layers, a graphitic layer appears after annealing.

Annealing at 1573 K of boron-implanted polycrystalline diamond films on silicon substrates introduces spurious effects originating from the shunt of the implanted layer by the substrate. The shunt resistance is dominated by the resistance to reach the substrate from the implanted region. It is ascribed to the grain boundaries of polycrystalline diamond. After the removal of the substrate, the electrical resistance of the implanted layer exhibits the expected behaviour for donor compensation. However, the higher activation energy for the high temperature resistance of the implanted layer requires the introduction of the partial ionization of the donor in this range of temperature and a high degeneracy factor gd = 1000 for the donor level.

Photoconductivity and photovoltaic signals of diamond films were measured at 300 K between 0.5 eV and 2 eV. We obtained bands around 0.5, 1.08, and 1.38 eV, and a plateau around 1.9 eV, which are ascribed to transitions from the valence band to localized states in the gap. The concentration of defects involved in the 1.08 eV band decreases after annealing at 600 °C. Different polarization offsets are needed to cancel these bands. This difference is ascribed to the occurrence of different defects according to the orientation of the crystallites under the contact. Space-charge zones are assigned to diamond in contact with an intermediate layer under the contact.

International audience; The electrical properties of boron doped diamond layers after deuterium diffusion have been investigated by Hall effect and capacitance voltage measurements. It is found that (i) the deuterated boron doped diamond layers are inhomogeneous after the deuterium diffusion, resulting in conducting and insulating areas; (ii) negative and positive Hall voltages are measured on the same boron doped sample after deuterium diffusion, depending on the contact geometry (negative with van der Pauw and positive with mesa etched Hall bar); and (iii) in the conducting area, the majority of the boron-deuterium complexes are not ionized after the deuterium diffusion. The detailed electrical measurements using mesa-insulating Hall bar structures revealed that the existence of inhomogeneous regions with conducting and insulating areas is the most probable source of wrong n-type conductivity in deuterated boron doped diamond layers of this work. In the light of this result, the possibility of an incorrect assignment of a shallow donor previously reported in deuterated boron doped diamond is discussed. Finally, confirmation is given that the boron deuterium complexes created after the deuterium diffusion are neutral in p-type diamond.

Recently, a new method using carbon implantation at 300 K in bulk copper followed by excimer laser annealing has been proposed to synthesize diamond, with contradictory results. We use this technique on copper thin films with carbon implantation at 77 and 300 K. From Rutherford backscattering spectrometry, the C, Cu and Si profiles have been deduced. They suggest solid state diffusion during annealing. The carbon depth profiles are found to be mainly dependent on the implantation temperature. The Raman spectra are well fitted by the same three and sometimes four gaussian components. The fourth component centred between 1305 and 1340 cm−1 is tentatively assigned to diamond.

CVD diamond detectors are of interest for charged particle detection and tracking due to their high radiation tolerance. In this article, we present, for the first time, beam test results from recently manufactured CVD diamond strip detectors and their behavior under low doses of electrons from a β-source and the performance before and after intense (>10 15 /cm 2 ) proton- and pion-irradiations. We find that low dose irradiation increase the signal-to-noise ratio (pumping of the signal) and slightly deteriorate the spatial resolution. Intense irradiation with protons 2.2×10 15 p / cm 2 lowers the signal-to-noise ratio slightly. Intense irradiation with pions 2.9×10 15 π / cm 2 lowers the signal-to-noise ratio more. The spatial resolution of the diamond sensors improves after irradiations.

(Received 10 August 2011; accepted 14 November 2011; published online 15 December 2011)[doi:10.1063/1.3665435]This article was originally published online on 9 August 2011 without all of the author’s corrections. AIP apologizes forthese errors.The article was correct as it appeared in the printed version of the journal. All online versions of the article were correctedon 15 August 2011.

CVD diamond shows promising properties for use as a position sensitive detector for experiments in the highest radiation areas at the Large Hadron Collider. In order to study the radiation hardn ess of diamond we exposed CVD diamond detector samples to 24~GeV/$c$ and 500~MeV protons up to a fluence of $5\times 10^{15}~p/{\rm cm^2}$. We measured the charge collection distance, the ave rage distance electron hole pairs move apart in an external electric field, and leakage currents before, during, and after irradiation. The charge collection distance remains unchanged up to $1\ times 10^{15}~p/{\rm cm^2}$ and decreases by $\approx$40~\% at $5\times 10^{15}~p/{\rm cm^2}$. Leakage currents of diamond samples were below 1~pA before and after irradiation. The particle indu ced currents during irradiation correlate well with the proton flux. In contrast to diamond, a silicon diode, which was irradiated for comparison, shows the known large increase in leakage curren t. We conclude that CVD diamond detectors are radiation hard to 24~GeV/$c$ and 500~MeV protons up to at least $1\times 10^{15}~p/{\rm cm^2}$ without signal loss.