Model for doping-induced contrast in photoelectron emission microscopy

• Published in: Journal of applied physics 2002-01, Vol.91 (1), p.469-475
• Main Authors: Ballarotto, V. W., Siegrist, K., Phaneuf, R. J., Williams, E. D.
• Format: Article
• Language: English
• ISSN: 0021-8979; 1089-7550
• DOI: 10.1063/1.1423399

We present a model that describes doping-induced contrast in photoelectron emission microscopy by including the effect of surface state distributions and doping-induced band gap reduction. To quantify the contrast, the photoyield from the valence band for near-threshold photoemission is calculated as a function of p-type doping concentration in Si(001). Various surface state distributions appropriate for a native-oxide covered Si device are investigated in order to determine the effect on doping-induced contrast. The lower limit on the number of surface states necessary for doping-induced contrast to occur is approximately 5×1013 cm−3. An interesting result is that neither the position nor the energy distribution of the surface donor states affects the contrast, which corresponds to approximately a factor of 2 change in intensity for each decade change in doping density. However, the overall intensity increases with any one of: increased surface state density, narrowing of surface state distribution, or increased energy of surface states with respect to the valence band. The band bending profile generated by the model predicts that doping-induced contrast will be affected by varying the incident photon energy. Experimentally, we verify this prediction by imaging with photon energies between 4.5 and 5.2 eV.