Persistent n-type photoconductivity in p-type ZnO

Abstract

Research activity on ZnO has increased over the past few years, with particular interest in potential electronic and optical device applications such as transparent field-effect transistors (FETs) and light-emitting diodes (LEDs). High-quality bulk and epitaxial samples have been prepared using a variety of growth techniques; however, progress on ZnO-based devices has been limited by the lack of reliable and reproducible p-type doping. Unintentionally doped ZnO films usually exhibit n-type conduction, generally attributed to interstitial H or Zn, oxygen vacancies, or substitutional impurities such as Al Zn serving as shallow donors. Recent efforts have demonstrated p-type conduction using N, P, As and Sb as acceptor dopants, with hole concentrations as high as 10 19 cm −3. In this work, the electrical properties of N- and P-doped p-type ZnO are characterized by temperature-dependent Hall-effects and photo-Hall-effects. An MBE-grown ZnO:N homoepitaxial layer exhibits weak p-type conduction with an acceptor energy E A≈90 meV in the dark and n-type photoconduction with a peak electron mobility at low temperature μ n>850 cm 2/V s under blue/UV light. This n-type photoconductivity persists for days when the sample is maintained in the dark, under vacuum, at room temperature. A sputtered ZnO:P film shows degenerate p-type conduction with p≈4×10 18 cm −3 and a hole mobility μ p≈3 cm 2/V s in the dark at room temperature. Under blue/UV light exposure, this P-doped sample undergoes a classic-mixed conduction transition from p-type to n-type where the carrier concentration exhibits a singularity, the Hall mobility ( μ H) goes to zero and both change sign as the temperature is increased. However, the n-type photoconductivity persists and no transition from n- back to p-type is observed upon subsequent cooling. Sequential, 400 K anneals with the sample in the dark and under vacuum cause the mixed conduction transition to reappear and shift to progressively higher temperatures, ultimately returning the sample to its original p-type state. A surface-layer model provides qualitative agreement with the observed behavior.