Towards improving the quality of semiconducting diamond layers doped with large atoms

Abstract

Abstract Previous studies on phosphorus- and aluminium-dopant activation in ion-implanted diamonds indicated that vacancies can either act to compensate the dopants or, when annealed at a temperature where they diffuse (600°C), they can interact to passivate these same dopant atoms. In order to improve the quality of diamond layers doped with large atoms, ways must be found to counteract such passivation. In this study, multiple cold implantation–rapid annealing (CIRA) steps were used to study the activation of phosphorus atoms which had been implanted into diamonds. To prevent large-scale passivation by the vacancies, the annealing temperature was chosen to be only 500°C. The results prove that enough phosphorus-dopant atoms can be activated to exceed the compensating vacancy density, and that the density of the activated donor atoms can be steadily increased by merely increasing the number of CIRA steps. In the case where the phosphorus-donor density N PD just exceeded the density of the compensating vacancy acceptors N V , conduction occurred with an activation energy of ≈0.62 eV. With increasing R = N PD / N V , the measured activation energies decreased. A model is proposed that relates this decrease to the negative electron affinity (NEA) of diamond. For the number of CIRA steps done to date, the resulting resistances of the layers are, unfortunately, still too high to allow any Hall-effect measurements. These high resistances can also be a result of the large density of unannealed vacancies, as well as the fact that the implanted layers only have an average depth of 80 nm, which is less than the surface roughness of the diamonds. Thermal EMF (Electro-Motive Force) measurements, whilst maintaining a stationary temperature gradient over the length of the diamond, confirmed n-type conduction.