Surface photovoltage phenomena: theory, experiment, and applications

TLDR: The theoretical concepts, experimental tools, and applications of surface photovoltage (SPV) techniques are reviewed in detail in detail as discussed by the authors, where the theoretical discussion is divided into two sections: electrical properties of semiconductor surfaces and the second discusses SPV phenomena.

About: This article is published in Surface Science Reports.The article was published on 1999-12-01 and is currently open access. It has received 1499 citations till now. The article focuses on the topics: Surface photovoltage.

Figures

Fig. 30. Current-voltage characterics of an electron gun ± CdS sample vacuum diode: (I) in the dark, (II) under super-bandgap illumination, (III) difference between the former two curves, in arbitrary units (after Steinrisser and Hetrick [239]).

Fig. 57. SPV spectra of device quality a-Si:H, deposited (a) without and (b) with H2 dilution on tin oxide coated glass (after Fefer et al. [305]).

Fig. 44. SPV spectra of a GaP(1 0 0) surface: (a) Extrapolation of the indirect bandgap. (b) Extrapolation of the direct bandgap following the subtraction of the indirect bandgap contribution (after Adamowicz and Szuber [290]).

Fig. 83. SPV spectra of ZnO:Al/u-ZnO/CdS/CIGS structures for various annealing times (after Gal et al. [402]).

Fig. 84. (a) Spectral planar photoconductivity of GaAs layers: Filled circles ± -doped layer. Empty circles ± homogeneous layer. Dashed line ± absorption spectrum. (b) SPV spectra of the -doped layer: Filled circles ± back-SCR-related SPV. Empty circles ± front-SCR-related SPV (after Bednyi et al. [342]).

Fig. 90. (a) Numerical simulation of SPV spectra of the nÿ-InP/n-In0:53Ga0:47As structure for various InP overlayer thicknesses (curve for 200 nm is shifted up by 20 mV for clarity). Inset: magnitude of the spectral `knee' at 1.35 eV vs. InP overlayer thickness for Ec 275 meV and Ec 225 meV. (b) Experimental SPV spectra of the same structure for various overlayer thicknesses (after Leibovitch et al. [21]).

Frequently asked questions

Q1. What contributions have the authors mentioned in the paper "Surface photovoltage phenomena: theory, experiment, and applications" ?

An in-depth review of quantitative analyses, which permit the extraction of various important surface and bulk parameters, follows. 

Q2. How did they modify EFs at the clean Si(111) surface?

By studying both p- and n-type samples over the entire doping range available to them, they were able to modify EFs at the clean Si(111) surface over a 0.2 eV range slightly below midgap. 

Q3. Why do the authors use the asymptotic-slope-intersection approach for qualitative?

Since by using the extrapolation technique or the derivative approach one must assume a priori the dominance of optical crosssection or energy-distribution effects, respectively, the authors usually use the simple, if less accurate, asymptotic-slope-intersection approach for performing qualitative analyses. 

Q4. Why is L satis®ed even at low doping levels?

The reason is that for other materials, L is at least a fraction of a micron, so that L w is satis®ed even at fairly low doping levels. 

Q5. Why does the p-type semiconductor have a negative SPV?

Due to the different signs of the equilibrium surface potential, this would result in a positive SPV in n-type semiconductors and a negative SPV in p-type semiconductors. 

Q6. What is the dc component of the electrostatic force used for topography measurements?

Yasutake [262] noted that the dc component of the electrostatic force (see Eq. (3.13)) is inherently less local than, e.g., van der Waals forces used for topography measurements. 

Q7. How can one ascertain that the sample has relaxed?

When working with a Kelvin probe, it is possible to ascertain that the sample has relaxed by monitoring the CPD in the dark as a function of time. 

Q8. Why are the bulk spectra insensitive to surface transitions?

The latter spectra are inherently insensitive to surface transitions because the photocurrent is collected from the entire bulk of the sample, so that the contribution of the SCR is typically negligible. 

Q9. What is the maximum time for which a transient can be adequately followed?

From Eq. (3.9b), the maximal time for which a transient signal can be adequately followed is bounded by RiCins, which is typically of the order of msec. 

Q10. What is the reason why the authors claimed that the obtained L cannot be used as a qualitative?

these authors claimed that the obtained L cannot even be used as a qualitative criterion for cell performance because in some of the simulations samples which yielded a smaller Lapp were calculated to exhibit a better, rather than worse, solar cell performance. 

Q11. How many times have they examined the relation between the spot and electrode size in scanning systems?

More recently, Faifer et al. [538] have examined the relation between the spot and electrode size in scanning systems and concluded that the latter should be bigger than the former by at least one diffusion length on either side in order to reduce errors due to lateral diffusion effects. 

Q12. What is the overall increase rate of free holes and trapped electrons at the semiconductor surface?

The overall increase rate of free holes (ps) and trapped electrons (nt) at the semiconductor surface is then given by:dps dt 1 e Jph ÿ Rsp; dnt dt Rsn ÿ Rsp; (2:74where Rsp (Rsn) is the surface recombination rate of holes (electrons). 

Q13. What was used for the identi®cation of many symmetry-allowed?

This was used for the identi®cation of many `symmetry-allowed' and `symmetryforbidden' quantum transitions in ZnSe/ZnMgSSe MQW structures.