Electric fields occur in all semiconductor devices and the intensity of the field is a fundamental parameter that varies depending on the device geometry, material quality, and impurities. The fields are more fundamental than the voltages, but cannot be directly measured using a voltmeter. Thus, semiconductor devices rely on ideal electrostatic models to estimate the field distributions and predict how devices might fail at high field operation, called dielectric breakdown. This phenomena is paramount for developing the next generation of high power devices using ultra wide band gap (UWBG) materials, such as Ga2O3, AlN, and BN. In order to independently measure the electric field, rather than voltage, we use a spectroscopic fingerprint for the electric field. An E-field sensitive spectral fingerprint was discovered in the 1950’s by Franz and Keldysh, in which there is a red shift and broadening of the absorption spectrum with electric field in semiconductors. In several materials, the absorption becomes modified by a secondary effect, which is the mutual electron (-) and hole (+) attraction, which also changes the spectra to exhibit a peak structure. Much like an atom, an exciton (electron-hole bound pair), creates its own energy levels also show atom-like energy level shifting with electric field, called the Stark effect, discovered in the atomic emissions lines in 1913. This eXciton-Franz-Keldysh (XFK) effect was predicted back in the 1960’s and 1970’s, but not experimentally quantified. Using spatially-resolved photocurrent spectroscopy, we are now using the XFK model to fit the spectra and determine the local electric fields in GaN and in Ga2O3. We are developing this method to be able to map out E-fields in semiconductor devices.
See: Verma and Adnan et al. APL (2020) and Adnan and Verma et al. Phys. Rev. Appl. (2021)