JOURNAL OF MATERIALS SCIENCE LETTERS 12 (1993) 119-121 Electrical and strain-sensitive behaviour of sputtered gold films M. M. NAYAK, S. SRINIVASULU*, K. RAJANNA*, S. MOHAN*, A. E. MUTHUNAYAGAM Liquid Propulsion Systems Centre, Indian Space Research Organisation, Bangalore 560 008, India *lnstrumentation and Services Unit, Indian Institute of Science, Bangalore 560 012, India In recent years there has been much interest in the study of the strain sensitivity of vacuum-deposited thin films because of their possible use as strain gauges. In general, all electrically conducting mater- ials exhibit strain sensitivity. The strain sensitivity is an electromechanical property of the material, most commonly referred to as the "gauge factor". Ideally a practical strain gauge taust possess fairly good strain sensitivity and stability of resistance to chang- ing temperature. Therefore, it is important to study the effects of both strain and temperature if a material is to be evaluated as a strain gauge. Several investigators have studied the strain-sensitive pro- perty of gold films [1-5]. However, no systematic study of sputtered gold films has been reported with respect to strain gauge applications. In this letter we report a systematic study of the strain-sensitive behaviour (electrical resistance- strain characteristics) and the temperature co- efficient of resistance (TCR) of sputtered gold films in the thickness range 30-130 nm. Gold films were prepared by a d.c. sputtering technique (using a Univax 300 vacuum system with a turbo-rotary pumping combination, from Messrs Leybolds, FRG), at a constant pressure of 8 Pa, on microscope glass slides. A gold target of diameter 50 mm (obtained from Messrs Leybold, FRG; purity 99.999%) was used. Pure argon (Iolar grade, purity 99.999%) was employed as the sputtering gas. The target-substrate distance of 20 mm was maintained. The cathode potential applied was 1200 V. Gold films of different thicknesses were sputtered by varying the deposition time. The thickness of the gold film was measured using a multiple-beam interferometer. Soon after deposition, the samples were loaded in a four-point bending apparatus for gauge factor measurement. The strain (e) value was calculated using [6] e = 4tI61/d z where t is the thickness of the glass substrate, 6 is the deflection in/xm and d is the distance between the rollers. Simultaneously the relative change in resistance (AR/R) was measured using a 51-digit Keithley (model 195 A) digital multimeter. The gauge factor was computed using [7, 8] F = (AR/R)/e After measuring the gauge factor, the samples were loaded in a jig [9] to measure the TCR. The films were annealed thoroughly in the present work because residual defects are expected to give incon- 0261-8028 ©1993 Chapman & Hall sistent characteristics when the films are used for device applications such as strain gauges. After complete annealing it is believed that films exhibit stable and repeatable characteristics in their beha- viour. The gauge factor of the film was again measured after the TCR measurement. The variation in resistance with temperature of the sputtered gold film is shown in Fig. 1 for a typical film thickness of t ~-130 nm. The arrows indicate the variation in film resistance as the temperature is decreased and increased. It can be seen that initially the resistance increased slightly (portion AB) and then began to decrease gradually (portion BC), finally reaching a minimum (C). Heating was stopped after observing this minimum and the film was allowed to cool to room temperature (portion CD). The linear portion obtained during cooling was used to calculate the TCR value. The tem- perature-resistance behaviour is similar to that found in observations by other investigators [10-12]. However, it is important to note that the reduction in resistance due to annealing is relatively low in the present study compared with that in the case of evaporated metal films [12]. This observation sup- plements the fact that the sputtered films contained fewer defects than evaporated films. The typical temperature-resistance plot for a thinner film (t ~ 30 nm) shown in Fig. 2 indicates that the tem- perature corresponding to minimum resistance has to be critically controlled, because even a small rise in temperature above the critical value results in a large increase in resistance. However, upon cooling the variation in resistance is linear, as in the case of thicker films. It has been observed that the variation in the 28 B 26 - ~ D 16 14 I t I t I I 25 75 125 175 225 275 300 Tempecatuce (°C) Figure I Variation in resistance with temperature for gold film of thickness t ~ 130 nm. 22 C~ 2O 119