JOURNAL OF MATERIALS SCIENCE LETTERS 14 (1995) 1496-1498 Lead-zirconate-titanate micro-tubes G. R. FOX Laboratoire de Ceramique, Ecole Polytechnique FOd6rale de Lausanne, Lausanne CH-1015, Switzerland Piezoelectric tubes have been used to produce composite sensors and actuators with enhanced response and an improved range of tuneability. For thin-walled tubes, the contributions from the d33 and d31 piezoelectric coefficients are additive, giving large effective piezoelectric responses that improve with decreasing tube diameter and wall thickness [1]. Ceramic tubes made by extrusion techniques are limited to minimum diameters in the 100 ~m size range and wall thickness on the order of 10/xm [2]. An alternative technique for tube fabrication based on fugitive phase and vapour deposition techniques extends the size range to include diameters in the 10 ~m range and wall thicknesses on the 1 ~m scale. This technique has been demonstrated by forming micro-tubes of ZnO and multi-layer micro-tubes of Pt/ZnO/Pt [3, 4]. Not only do piezoelectric micro- tubes show promise for making highly responsive composite sensors and actuators, but their cylindrical geometry may uniquely suit certain micro-electro- mechanical applications. Materials in the PbZrxTil_xO 3 (PZT) system have been studied in great detail. They exhibit some of the largest known piezoelectric responses and are used commercially on a large scale [5]. For these reasons, it was of interest to attempt to extend the micro-tube fabrication process to include PZT. This letter reports on the fabrication of PZT micro-tubes by combining multi-magnetron reactive sputter deposition and fugitive phase techniques. Mono-filament polyester fibres with a diameter of 23/xm were mounted on a fibre holder using a polymer cement. Sputter deposition was used to coat the fibres while they were rotated at a rate of 5.6 rpm. The basic sputtering system and the holder used for fibre rotation are similar to those described in a previous publication concerning the fabrication of ZnO micro-tubes [3], but additional magnetron sputtering sources were in61uded in the system to allow for the deposition of PZT coatings by multi- magnetron reactive sputtering. Three magnetron sources were mounted, with angular separations of 45 °, on the perimeter of a vertical cylindrical vacuum chamber. The rotating fibre holder was placed at the centre of the chamber where the spatial distributions of the material sputtered from the Pb, Zr and Ti metal targets overlapped. The distances between the targets and the fibres were 18 cm for the Pb target and 20cm for the Zr and Ti targets. An Ar/O 2 gas mixture with an oxygen partial pressure, Po2, of 0.350 _+ 0.010Pa and total pressure, P, of 0.500 _+ 0.005 Pa was used for the sputtering pro- 1496 cess. The targets were sputtered using applied d.c. powers of PWpb= 31-34 W, PWzr = 250 W and PWyi = 445 W. A quartz crystal microbalance (QCM) was used to monitor the deposition rate of the sputtered material during the deposition process and to establish the relative deposition rates of the oxidized Pb, Zr and Ti species. The ratios of the cation deposition rates were chosen (by adjusting the power applied to each target) to give a nominal composition of PbZr0.52Ti0.4803. After pre-sputtering the targets, obtaining nearly steady-state deposition conditions, and adjusting the deposition rates to the proper ratios, a shutter used to mask the fibres was removed and deposition onto the fibres was begun. A constant deposition rate of approximately 3 nm min-1 was maintained during the entire deposition process. After coating, 2-3 cm sections of fibre were wrapped in Pt foil and annealed using a two-step process. Firstly, the fibres were heated to 550 °C at a rate of 5 °C min -1 and held at this temperature for 20 rain. This annealing step burns away the polymer fibre. The second annealing step was a 5 °C min -1 heat from 550 to 650 °C and a hold at 650 °C for 20 rain. This crystallizes the sputter deposited PZT, which is amorphous in its as-deposited state. The annealed samples were cooled at approximately 5 °C min- 1. Scanning electron microscopy (SEM; Jeol 6300F) was used to observe fracture cross-sections and the surfaces of annealed samples. X-ray diffraction (XRD) patterns were obtained using a 114.6 mm diameter Debye-Scherrer camera with a 0.5 mm diameter aperture and Ni-filtered CuK~ (40 kV and 35 mA) radiation. Films were exposed for 5 h and a densitometer (Carl Zeiss, Germany) was used to obtain plots of intensity against 20. Coated polyester fibres were flexible and could be handled easily for subsequent processing and analy- sis. After annealing, fibre sections between 1 and 2 cm remained on the Pt foil, but the samples were fragile. SEM observations revealed that the annealing process resulted in the formation of micro-tubes with a 19/xm inside diameter and uniform 0.8/xm wall thickness, as shown in Fig. 1. The diameter of the tubes remained uniform over the entire length, as illustrated by Fig. 2. A shrinkage of 17% is indicated by comparing the diameter of the polyester fibres and the inside diameter of the annealed tubes. The reduction in diameter could result from a combina- tion of polyester fibres shrinkage due to heating during the coating process and shrinkage due to crystallization and densification of the PZT during 0261-8028 © 1995 Chapman & Hall