I.M. Buckley-Golder

Piezoresistive effect of boron-doped diamond thin films

Abstract Double-layer structures consisting of a nominally undoped diamond film and an overlying boron-doped diamond film on Si substrates were used for investigations into the piezoresistive effect of polycrystalline diamond films. Various resistor structures were fabricated by conventional photolithography and by reactive ion etching in an oxygen plasma. The resistor structures were pasted onto a defined deflectable bending beam to permit measurement of the longitudinal and transverse piezoresistive effects. The longitudinal piezoresistive effect is greater than the transverse effect. An increase in the gauge factor was observed with decreasing dopant concentration. The maximum longitudinal gauge factor at room temperature is 5.4. An increase in the gauge factor from 5.4 to 13.7 was measured in the course of a temperature increase from 27 to 60°C.

Post-processing of diamond and diamond films: a review of some Harwell work☆

Abstract Diamond is an attractive material for electronic applications: however natural stones are often too small and inhomogeneous to be useful, so several processes have been developed to produce synthetic thin film diamond. Work on an ion beam process for epitaxial growth of thin film diamond is presented. Using C + ion beams of up to 100 keV at temperatures from 350–1050°C, diamond thin films of up to 15 μm can be deposited at rates of up to 5 μm h −1 . The ion beam method has also been modelled to take growth mechanisms, defect incorporation and film failures into account. The electrical behaviour of diamond has been studied by introducing electrically active dopants via ion implantation and annealing. Boron has been demonstrated to be a p-type dopant while lithium has been identified as an n-type dopant. A p–n junction has been produced by ion implanting a type 2B diamond stone with lithium. This junction displayed good I–V characteristics at room temperature and retained excellent properties at temperatures up to 350°C.

High temperature Young's modulus of polycrystalline diamond

Abstract The Youngs modulus of polycrystalline diamond grown by microwave assisted chemical vapour deposition was determined by a dynamic three-point bending measurement between room temperature and 750°C. The room temperature Youngs modulus was approximately one-half of the theoretical value of 1143 GPa. The lower Youngs modulus was traced back to voids and microcracks and therefore a smaller effective sample cross-section. Also, the density of the the polycrystalline samples of 3.387 g cm −3 was smaller compared with single crystal diamond. Up to temperatures of 600°C the Youngs modulus remains approximately constant. At higher temperatures the onset of diamond etching in air leads to a strong reduction of the Youngs modulus. At 750°C the Youngs modulus drops linearly with −10.4 GPa/min to one-thired of its initial value before sample fracture occurs. The onset of diamond etching at these temperatures was proved by thermogravimetric measurements.

Etching of polycrystalline diamond and amorphous carbon films by RIE

Abstract Amorphous carbon films and doped and undoped polycrystalline diamond films were patterned with RIE in an oxygen plasma. Insitu measurement of the etch rate gave different, depth-dependent etch rate values. For the amorphous films, an etch rate of 150 nm/min to 600 nm/min was measured; for the polycrystalline films the etch rate measured was 10 nm/min to 60 nm/min. The etched polycrystalline films exhibit a columnar structure, which can be traced back to the anisotropic etching behaviour. Suitable masking layers include PECVD oxide and nitride with excellent selectivity towards diamond. The etch rate rises with increasing RF power and decreases with increasing oxygen partial pressure.

Space-charge-limited current flow and trap density in undoped diamond films

Abstract Nominally undoped, polycrystalline diamond films of various qualities were deposited on p-type silicon substrates by microwave chemical vapour deposition. The diamond films were characterized by Raman spectroscopy, scanning electron microscopy and electrical measurements. Current-voltage measurements on one wafer show that the current flow is space charge limited and influenced by traps. The energetic distribution of the trap densities was determined by analysis of the current-voltage characteristic. The trap densities vary over an energy range of 0.2 eV from 10 14 to 10 17 cm −3 eV −1 . The energetic trap distribution and trap densities are not locally constant.

Very low resistivity AlSi ohmic contacts to boron-doped polycrystalline diamond films

Abstract The effects on the contact resistivity of annealing Al Si (99:1) contacts on diamond with different B concentrations have been investigated. The change of the current-voltage characteristic from rectifying to ohmic on lightly doped samples, and the drop of the contact resistivity by orders of magnitude for more heavily B-doped samples after annealing at 450 °C in N2, are attributed to the formation of SiC at the metal-diamond interface. The existence of the SiC interface has been verified by X-ray-induced photoelectron spectroscopy. In addition, the contact resistivity is a very sensitive function of doping at high doping concentrations. Contact resistivities as low as about 10−7 θ cm2 have been achieved.

Electrical characterization of Al/Si ohmic contacts to heavily boron doped polycrystalline diamond films

Mesa etched transmission line model (TLM) test structures with different contact lengths have been fabricated on heavily boron doped polycrystalline diamond films. The behavior of the contact and contact end resistance measurements can be fully explained using the TLM. No influence of the grain size on the contact resistivity has been observed. High surface boron doping concentrations led to low contact resistivities, in agreement with numerical calculations. Annealing of Al/Si–diamond contacts at 450 °C in N2 leads to lower contact resistivities due the formation of SiC at the metal–diamond interface. The temperature dependence of the specific contact resistivity can be described well with a tunneling model before annealing. After annealing no useful fit is possible, indicative of the fact that the SiC interface layer acts as defect layer.

High temperature Raman studies of diamond thin films

Raman spectroscopy has become the definitive technique for assessing the quality of diamond thin films. Not only can the diamond/graphite contents be determined, but more detailed information, for example, about the domain size and the stress associated with a coating can be gleaned. The results of a Raman microprobe study of synthetic diamond coatings on silicon and alumina substrates, at elevated temperatures (up to 750°C) in controlled atmospheres of hydrogen, argon and oxygen are presented. The position of the Raman band associated with crystalline diamond (1332 cm−1) was monitored as a function of temperature. From the shift of the Raman band from its natural position, an associated stress value can be obtained.

How to fabricate low-resistance metal-diamond contacts

Abstract Three types of mesa etched contact resistance test structures with Al Si , TiAu, TiWN-Au contacts are compared. The contact resistivity was calculated using transmission-line model (TLM) theory which is fully applicable to diamond. For Al Si contacts the lowest contact resistivity was obtained after annealing at 500 °C in dry nitrogen. The lowest contact resistivity for as-deposited contacts was found for TiAu. In general, the contact resistivity drops with increasing levels of B dopant and depends on the metallization scheme. The dependence of the contact resistivity on the surface concentration of B for as-deposited Al Si contacts can be fitted by a power-law indicative of spatial inhomogenous areas in the contact region leading to an additional spreading resistance component. The contact resistivity depends exponentionally on the operating temperature for as-deposited Al Si contacts but the thermal coefficient is independent of the doping level for heavily B-doped diamond films, which is typical for tunneling.

Degradation mechanisms of passivated and unpassivated diamond thermistors

Abstract The purpose of this investigation was to examine the failure mechanisms of prototype thermistors as a function of temperature in air. A hot stage Raman microprobe was used to examined device degradation mechanisms at elevated temperatures by monitoring the diamond Raman signal and simultaneously measuring the resistance of the thermistor. An unpassivated device became non-conducting (resistance greater than 20 MΩ) at approximately 610°C in laboratory air. Although the diamond film remained intact under these conditions, the contacts were found to degrade via a combination of oxidation and carburization. Formation of an insulating titania phase caused the device to fail. A silica device passivation scheme was investigated for the protection of the contacts and diamond film from oxidation at elevated temperatures. Oxidation tests made on silica-coated diamond films showed a beneficial corrosion protection effect at 650°C in laboratory air. The passivated thermistor operated up to marginally higher temperatures of 670°C. At this temperature the passivation cracked and some spalling occurred. Post-failure analysis indicated that the contacts had oxidised and also that the insulating diamond had been etched. In contrast, the boron-doped resistive element of the thermistor was relatively unaffected, suggesting different susceptibilities to oxidation between the doped and insulating diamond.