Mixing of phosphorus and antimony ions in silicon by recoil implantation

Abstract

The mixing of two different ion species by recoil implantation has received considerable interest recently because the technique appears to be an attractive way of modifying the surface properties of an ion-implanted region. By recoil implantation, one could introduce almost any two ion species simultaneously into a substrate and study their respective properties or any mutual interactions. Possible applications are in the formation of low Schottky barrier contacts [1, 2] and in the formation of high-low homojunctions. This work examined the effects of mixing phosphorus and antimony in silicon by recoil implantation. The main purpose was to investigate the electrical properties after ion mixing and to compare the results with those achieved using other techniques. By investigating the annealing behaviour, we also studied the different degrees of activation. (1 00>-oriented p-type silicon was used in this work. It was supplied by Wacher Chemitronic. The resistivity was between 10 and 20f~cm. The wafers were first cleaned using standard IC cleaning procedures and rinsed dry prior to metal deposition. An Edwards E-beam evaporator was used and the thickness of the deposited layer was monitored using a quartz crystal. The film thickness should be uniform within + 1 0 % . A Varian 200-1000 ion implanter was used and the beam energy was set at 120 keV. Phosphorus was used as the ion source and the dosage was varied from 1 x 1014 to 2 × 1016 cm 2. After recoil implantation, the metal film, antimony in this case, was removed using a mixed solution of HNO3 : HC1 : H20 (1 : 1 : 1). Isochronal annealing was carried out in a nitrogen atmosphere for 30 rain durations. The annealing temperature was between 600 and 1100 ° C. Sheet resistance was measured using the standard four-point probe technique. Fig. 1 shows a typical plot of how the sheet resistance of the recoil implanted silicon samples varied with annealing temperature. The results were for samples with four different antimony thicknesses. The phosphorus dosage was 1 x 1015 cm -2. The data for samples with an antimony thickness of less than 20 nm showed no appreciable difference and the magnitudes are in basic agreement with the results reported by Dearnaley et al. [3]. From the figure, it is apparent that more than one effect influences the annealing characteristics. Fig. 2 shows the sheet resistance of the samples as a function of the implant dosage when the samples are annealed at different temperatures. The annealing time was set at 30 min. Only the results from two different types of samples were shown. The solid curves are for samples without a surface antimony layer and the dotted curves are for samples coated with 80nm antimony. The observed linear relationships for the samples with no antimony indicated that the sheet resistance has a similar power dependence on the implant dosage and that the annealing temperature (from 700°C upwards) only affected the degree of carrier activation. At very high dosages, saturation effects were observed at the lower annealing temperatures. The antimony recoil implanted samples exhibited similar behaviour when the annealing temperature was above 1000°C indicating most probably the same type of activation behaviour as the directly implanted samples. The sheet resistances were somewhat higher because of the scattering of the phosphorus ions by the antimony layer. Note the difference in the slopes between the samples annealed at 800°C and those annealed above 1000 ° C. Figs 1 and 2 clearly demonstrate the effects of antimony recoil implantation in silicon. The initial decrease of the sheet resistance with increasing anneal-