An investigation of thin amorphous carbon-based sputtered coatings for MEMS and micro-engineering applications

The work presented in this thesis describes an investigation into the production and properties of thin amorphous C films, with and without Cr doping, as a low wear / friction coating applicable to MEMS and other micro- and nano-engineering applications. Firstly, an assessment was made of the available testing techniques. Secondly, the optimised test methods were applied to a series of sputtered films of thickness 10 - 2000 nm in order to: (i) investigate the effect of thickness on the properties of coatingslcoating process (ii) investigate fundamental tribology at the nano-scale and (iii) provide a starting point for nanotribological coating optimisation at ultra low thickness. The use of XPS was investigated for the determination of Sp3/Sp2 carbon bonding. Under C 1s peak analysis, significant errors were identified and this was attributed to the absence of sufficient instrument resolution to guide the component peak structure (even with a high resolution instrument). A simple peak width analysis and correlation work with C KLL D value confirmed the errors. The use of XPS for Sp3/Sp2 was therefore limited to initial tentative estimations. Nanoindentation was shown to provide consistent hardness and reduced modulus results with depth (to < 7nm) when replicate data was suitably statistically processed. No significant pile-up or cracking of the films was identified under nanoindentation. Nanowear experimentation by multiple nanoscratching provided some useful information, however the conditions of test were very different to those expect for MEMS and micro- / nano-engineering systems. A novel 'sample oscillated nanoindentation' system was developed for testing nanowear under more relevant conditions. The films were produced in an industrial production coating line. In order to maximise the available information and to take account of uncontrolled process variation a statistical design of experiment procedure was used to investigate the effect of four key process control parameters. Cr doping was the most significant control parameter at all thicknesses tested and produced a softening effect and thus increased nanowear. Substrate bias voltage was also a significant parameter and produced hardening and a wear reducing effect at all thicknesses tested. The use of a Cr adhesion layer produced beneficial results at 150 nm thickness, but was ineffective at 50 nm. Argon flow to the coating chamber produced a complex effect. All effects reduced significantly with reducing film thickness. Classic fretting wear was produced at low amplitude under nanowear testing. Reciprocating sliding was produced at higher amplitude which generated three body abrasive wear and this was generally consistent with the Archard model. Specific wear rates were very low (typically 10-16 - 10-18 m3N-1m-1). Wear rates reduced exponentially with reduced film thickness and below (approx.) 20 nm, thickness was identified as the most important control of wear.

A study of the electrical and optical properties of unhydrogenated neon-sputtered amorphous silicon

Films of amorphous silicon (a-Si) were prepared by r.f. sputtering in a Ne plasma without the addition of hydrogen or a halogen. The d.c. dark electrical conductivity, he optical gap and the photoconductivity of the films were investigated for a range of preparation conditions, the sputtering gas pressure, P, the target-substrate spacing, d, the self-bias voltage, Vsb, on the target and the substrate temperature, Ts. The dependence of the electrical and optical properties on these conditions showed that various combinations of P, d and Vsb, at a constant Ts, giving the same product (Pd/V sb) result in films with similar properties, provided that P, d and Vsb remain vithin a certain range. Variation of Pd/Vsb between about 0.2 and 0.8 rrTorr.cm!V varied the dark conductivity over about 4 orders of magnitude, the optical gap by 0.5 eV and the photoconductivity over 4-5 orders of magnitude. This is attributed to controlling the density-of-states distribution in the mobility gap. The temperature-dependence of photoconductivity and the photoresponse of undoped films are in support of this conclusion. Films prepared at relatively high (Pd/Vsb) values and Ts=300 ºc: exhibited low dark-conductivity and high thermal activation energy, optical gap and photoresponse, characteristic properties of a 'low density-of-states material. P-type doping with group-Ill elements (Al, B and Ga) by sputtering from a composite target or from a predoped target (B-.doped) was investigated. The systematic variation of room-temperature conductivity over many orders of magnitude and a Fermi-level shift of about 0.7 eV towards the valence-band edge suggest that substitutional doping had taken place. The effects of preparation conditions on doping efficiency were also investigated. The post-deposition annealing of undoped and doped films were studied for a temperature range from 250 ºC to 470 ºC. It was shown that annealing enhanced the doping efficiency considerably, although it had little effect on the basic material (a-Si) prepared at the optimum conditions (Pd/Vsb=0.8 mTorr.cm/V and Ts=300 $ºC). Preliminary experiments on devices imply potential applications of the present material, such as p-n and MS junctions.