Two-electron oxidations at illuminated N-type semiconducting silicon electrodes: Use of chemically derivatized photoelectrodes

Abstract The two-electron reductants N,N,N ′ ,N ′ -tetramethyl- p -phenylenediamine (TMPD), biferrocene (BF), bis(fulvene) diiron (BFD), and differrocenylmethane(VI) can all be oxidized contrathermodynamically to the dications at illuminated n-type Si photoanodes. Even for TMPD 2+ /TMPD + and BF 2+ /BF + , which have formal potentials near the top of the valence band of Si, the oxidation process to form the dication is up to ≈0.6 V contrathermodynamic at sufficiently high light intensity. But, exept for VI, each of these two-electron reductants exhibits two, one-electron, waves in cyclic voltammetry rather than one, two-electron wave as expected for an ideal n-type semiconductor/liquid junction where the top of the valence band at the interface is more positive on the potential scale than the formal potential of the solution couples. VI is unique, since the two electrons are known to be removable at nearly the same potential. Surface states provide a rationale for the observation of multiple waves at illuminated n-type Si. TMPD + oxidation is only reproducibly accomplished using n-type Si derivatized with (1,1 ′ -ferrocenediyl)dichlorosilane, I, which protects the surface from deleterious photoanodic SiO x growth on the surface at the more anodic excursions generally needed to observe TMPD + →TMPD 2+ . Dichlorodiferrocenylsilane, III, a two-electron reductant serves as a derivatizing reagent for Pt and n-type Si and gives surfaces which are similar to those previously generated by derivatization with I or trichlorosilylferrocene, II. The oxidizing power of derivatized photoanodes (or those having a high density of surface states) depends on light intensity in that the ratio surface-attached oxidant/surface-attached reductant is a measure of the oxidizing power; the ratio of attached ferricenium/ferrocene depends on light intensity and hence the oxidizing power of the surface is light intensity dependent. Drawing an analogy between redox active derivatives and surface states, the oxidizing power of the photoanode with surface states is measured by the ratio of unfilled to filled surface states, which at a given potential appears to be light intensity dependent as for the derivatized electrode.