Published online by Cambridge University Press: 10 January 2018
The unique electronic structure of diamond and its excellent thermal properties allow a broad range of possible applications; from electron sources to RF electronics. However, knowledge of the surface energy levels is essential to produce efficient, high-quality devices. We investigate the valence band position and resulting negative electron affinity for hydrogen terminated diamond under ambient, low vacuum and ultra-high vacuum (UHV) conditions. There was a -0.5 eV change in valence band position causing a negative electron affinity shift from -1.1 eV under UHV to -0.6 eV in ambient pressure. We compare the photoemission current under each environment to predict the ability of the sample to be used as an electron source. The maximum emission was observed when the sample displayed the largest negative affinity. A scanning photoemission measurement is demonstrated to highlight the superior photoemission yield from the hydrogen terminated diamond surface compared to the stainless steel contact. A scanning Kelvin probe measurement is shown to illustrate a method of analyzing the contact potential difference across the diamond surface. Within high-power RF electronics, devices are likely to be operating at increased temperatures so knowledge of the impact of temperature on the energy levels is important. We study the valence band and Fermi level positions for hydrogen terminated diamond from room temperature (300K) to 425K under low and UHV conditions. The Fermi level moved below the valence band edge at increased temperature, illustrating the effect of the 2D hole gas at the surface. We also analyzed the photoemission characteristics and found an increase in yield with increasing temperature. The measurement techniques used to evaluate the energy levels of diamond: photoemission spectroscopy and Kelvin probe measurements, in ambient and vacuum, allow analysis to be completed in minutes. This offers an initial analysis alternative to elucidate more information and predict performance prior to the more time-consuming full device manufacture and characterization.