AVS 53rd International Symposium
    Surface Science Monday Sessions
       Session SS2-MoA

Paper SS2-MoA7
Achieving a Low Surface Recombination Rate Despite a High Surface Defect Density: the Role of Charge Carrier Concentration

Monday, November 13, 2006, 4:00 pm, Room 2004

Session: Functionalization of Semiconductor Surfaces
Presenter: D.J. Michalak, California Institute of Technology
Authors: D.J. Michalak, California Institute of Technology
N.S. Lewis, California Institute of Technology
Correspondent: Click to Email

Low interfacial electron-hole recombination rates are essential for low-noise electronic devices and high-efficiency solar energy converters. This interfacial recombination rate is dependent on both the electrical trap state density, N@sub T@, and the concentrations of electrons, n@sub s@, and holes, p@sub s@.@footnote 1@ Reports of low recombination rates have often been attributed to a low N@sub T@,@footnote 2@ without a measurement of n@sub s@ and p@sub s@, partly because the importance of n@sub s@ and p@sub s@ has not been fully recognized and partly because an accurate evaluation of n@sub s@ and p@sub s@ can be difficult. In this work, surface recombination rates of stable silicon/liquid junctions, observed using a contactless rf photoconductivity technique, were compared with n@sub s@ and p@sub s@ values obtained from Mott Schottky analysis. Our results demonstrate that recombination rates can only be correlated with N@sub T@ when n@sub s@ = p@sub s@. Otherwise, the recombination rate was always low due to a large n@sub s@ or p@sub s@ even in the presence of a large N@sub T@. The full impact of this work was further realized through a study of the recombination rates of H-Si immersed in solutions of 48% HF, 40% NH@sub 4@F, or buffered HF (BHF) because such measurements are often performed for in situ monitoring of the surface quality during wafer processing steps. Our results demonstrate that only HF contacts can be used for in situ monitoring because n@sub s@ = p@sub s@. For NH@sub 4@F or BHF contacts, low recombination rates were observed only because n@sub s@ is much larger than p@sub s@, and no information about N@sub T@ can be inferred from these measurements. @FootnoteText@ @footnote 1@ Hall, R.N., Physical Review 1952, 87, (2), 387, and Shockley, W., and Read, W.T., Physical Review 1952, 87, (5), 835-842.@footnote 2@ Msaad, H., et al., J. Electron. Mater. 1994, 23, (5), 487-491, and Yablonovitch, E. et al., Phys. Rev. Lett. 1986, 57, (2), 249-252.