Photoelectrochemistry of Redox-Active Self-Assembled Monolayers Formed on n-Si/Au Nanoparticle Photoelectrodes

Controlling the chemistry of the electrode–solution interface is critically important for applications in sensing, energy storage, corrosion prevention, molecular electronics, and surface patterning. While numerous methods of chemically modifying electrodes exist, self-assembled monolayers (SAMs) containing redox-active moieties are particularly important because they are easy to prepare, have well-defined interfaces, and can exhibit textbook photoelectrochemistry. Here, we investigate the photoelectrochemistry of redox-active SAMs on semiconductor/metal interfaces, where the SAM is attached to the metal site instead of the semiconductor. n-Si/Au photoelectrodes were fabricated using a benchtop electrodeposition procedure and subsequently modified by immersion in aqueous solutions of (ferrocenyl)hexanethiol and mercaptohexanol. We explored the relevant preparation conditions, finding that after optimization, we were able to obtain canonical cyclic voltammetry for a surface-bound redox molecule that could be turned on and off using light. We then characterized the optimized electrodes under varying illumination intensities, finding that the heterogeneous electron transfer kinetics improved under higher illumination intensities. These results lay the foundation for future studies of semiconductor/metal/molecule interfaces relevant to sensing and electrocatalysis.


Table of contents
We estimated the thickness of the SAM to be ≈30 Å using Bain and Whitesides's method: 1 where I SAM and I Au are the intensities of the Au 4f 5/2 peak, d SAM is the thickness of the SAM (in Å), and λ SAM is the attenuation wavelength (42 Å). 1,2 This is considerably thicker than we expected for a monolayer with a sixcarbon chain.However, Bain and Whitesides's method assumes that the surface is flat. 1 In our samples, the electrodeposited Au nanoparticles introduce significant roughness to the surface (Figure 2), leading to an underestimation of the attenuation wavelength and an overestimation of the SAM thickness. 1

Section S2. Electrochemical impedance spectroscopy (EIS) characterization
Figure S2 shows reciprocal square capacitance versus voltage plots for n-Si/Au/Fc-SAM recorded in a 0.1 M HClO 4 electrolyte over a -1 to 0 V range with an impedance measurement made every 10 mV.The electrochemical cell was housed inside a Thorlabs dark box.Equation S1 relates the capacitance of the space charge region (C SC ) to the potential of an electrode versus a reference (E): 3 where C sc is the space charge capacitance (in F), q is the fundamental charge on an electron (=1.6•10 -19 C), ɛ is the dielectric constant of the semiconductor (11.7 for Si), 4 ɛ 0 is the permittivity of free space, N d is the bulk dopant carrier concentration (in cm -3 ), A is the electrode area (=0.071 cm 2 ), E is the applied potential (in V), E fb is the flat band potential (in V), k B is Boltzmann's constant, and T is the absolute temperature (in K).We calculated the space charge capacitance using the imaginary component of the impedance and equation S3: where v is the frequency (in Hz), and Z" is the imaginary component of the impedance (in Ω).We fit the data to the Mott-Schottky equation to find the flat-band potential and the bulk dopant concentration.The x-intercept of the linear portion of the the C -2 vs. E plot corresponds to E fb + k B Tq -1 , while the slope of the line is related to bulk dopant concentration, N d : The conduction band edge (E cb ) can be estimated using equation S5: where N c is the effective density of states of the conduction band for Si (=2.8•10 19 ). 4 Finally, the valence band edge (E vb ) is calculated by adding the band gap energy (E g ; =1.1 eV for Si) to the conduction band edge.

Figure S3 .
Figure S3.Reciprocal square capacitance plots of n-Si/Au and n-Si/Au/FcHT-SAM electrodes in the dark in 0.1 M HClO 4 .The solid lines represent the linear regression of the linear portion of the plot -in both cases R 2 > 0.99; frequency = 50 kHz, amplitude = 5 mV, E step = 10 mV; reference = SCE, counter = glassy carbon rod.

Figure S9 .Figure S10 .
Figure S9.CVs of n-Si/Au/FcHT-SAM electrodes in 0.1 M HClO 4 at scan rates between 0.1 and 1.0 V s -1 (a) before and (b) after 1 minute of vortex mixing.Arrow indicates increasing scan rates.Reference = SCE, counter = glassy carbon rod.

Figure S13 .
Figure S13.Examples of non-unique fits of the same experimental data.

Figure S14 .
Figure S14.Deconvolution of the anodic portion of the CVs that are presented in Figure 6 of the main text.N-Si/Au/FcHT samples were prepared with 0.2 mM HAuCl 4 in the electrodeposition bath and incubated in 1 mM FcHT solutions for (a) 15, (b) 30, or (c) 60 minutes.Symbols are the raw data, the red, blue, and purple traces are individual components and the black trace is the sum of the fits.

Table S1 .
Summary of peak fitting parameters from FigureS12.

Table S2 .
Summary of fit fitting parameters from FigureS13.