Designing semiconductor-electrochemical junctions for bioinspired energy transduction

TL;DR

This work proposes bioinspired nonenzymatic electron bifurcation by designing semiconductor-electrochemical junctions that spontaneously split two-electron donors into separate downhill and uphill electron streams. The approach uses energy-landscape engineering at an n-p-electrolyte interface, with a three-way junction enabling selective charges and suppressing short circuits; a quantitative model demonstrates how $I_{ ext{inj}}$ and $I_{ ext{rec}}$ scale with $V_{ ext{bias}}$ and how $ ext{η}_{ ext{eff}}$ can approach unity at low bias. The key contributions include a detailed nonadiabatic ET framework for interface kinetics, a model demonstration of exponential current growth and efficiency tradeoffs, and concrete design principles (layered heterojunctions, bulk heterojunctions, surface area, and proton-coupled kinetics) for higher performance. The results point to practical bioinspired voltage-conversion, high-open-circuit-voltage generation, and catalysis-driven energy transduction, with implications for both devices and understanding biological machinery.

Abstract

Long ago, life discovered how to efficiently push electrons thermodynamically uphill to lower potential by harnessing energy released by an equal number of electrons moving downhill. Known as electron bifurcation, this form of energy transduction has never been observed in the absence of natural enzymes. To successfully bifurcate electrons, a system must block short-circuit electron transfers that allow all electrons to flow downhill, while maintaining productive reactions. It is difficult to design systems that catalyze these highly-selective electron flows while minimizing free energy dissipation. Using theories of electron transfer and charge transport, I introduce semiconductor-electrolyte junctions that spontaneously bifurcate electrons analogously to natural enzymes (bifurcating junctions). I simulate a simple but illustrative bifurcating junction with typical material properties, and discuss how more complicated designs could achieve higher performance.

Figures (3)

• Figure 1: Overview of biological electron bifurcation and energy landscape analogy with semiconductor np junctions. (A) A simple electron bifurcating enzyme oxidizes a two electron donor molecule D, shuttling each electron down a different "branch" of cofactors to reduce two final acceptors, one at high-potential ($\ce{A_H}$) and the other at low-potential ($\ce{A_L}$). (B) When operating in the bifurcating (forward) direction, the electron donor $\ce{D^=}$ undergoes a series of redox reactions when bound to the active site (protons are not shown for simplicity). (1) First, the first cofactor in the high-potential branch is transiently oxidized by a thermal (i.e. a "hole" moves from the $\ce{A_H}$ pool to $\ce{H_1}$. This allows $\ce{H_1}$ to oxidize $\ce{D^=}$ forming the highly reactive (often radical) species $\ce{D^-}$ that can (2) inject a high-energy electron into the low-potential branch. The second electron does not follow the first electron to the high-potential branch because $\ce{H_1}$ is no longer oxidized ($\ce{H_1}$ is maintained in a reduced $\ce{H^-_1}$ state by the $\ce{A_H}$ pool). Lastly, the high-energy electron is pulled out of the low-potential branch to reduce $\ce{A_L}$. (C) An energy landscape with steep (hundreds of meV) energy landscapes maintains the cofactors in redox states that prevent short-circuit electron transfers. The low-potential branch is maintained in an oxidized state (few "holes" to accept short circuit electrons) and the low-potential branch is maintained in an oxidized state (few electrons available to short circuit). The $\ce{A_L}$ and $\ce{A_H}$ redox pools are out of equilibrium (positive $\Delta G_{\text{bias}}$), so electrons from the $\ce{D^=}$ pool flow thermodynamically uphill to the $\ce{A_L}$ pool. (D) There is a strong analogy between the electron energy landscape in (C) and the band bending in the depletion region of np semiconductor junctions, with the same suppression of charge carriers. An applied bias $|e^-|V_{\text{bias}}$ between electron and hole quasi-fermi levels $\mu_{e^-}$ and $\mu_{h^+}$ is analogous to $\Delta G_{bias}$ in the enzyme energy landscape.
• Figure 2: Proposed electron-bifurcating semiconductor-electrolyte junction (A) An electron-bifurcating junction uses a tailored semiconducting electrode to spontaneously bifurcate electrons from a two-electron species ($\ce{DH^-}$) toward p- and n-doped regions with a voltage $V_{\text{bias}}$ applied across them (if doping is undesired. (B) Interfacial electron transfer reactions that result in spontaneous electron bifurcation across the n-p junction. (a) First, a thermal fluctuation brings a hole ($\ce{h+}$) from the p-doped region to the surface of the depletion region, and performs a one-electron oxidation of $\ce{DH^-}$ to the radial $\ce{D^{\cdot -}}$ species. (2) This radical species has sufficient reducing power to inject a charge into the conduction band. (3) Finally, the electron flows to the n-doped region while oxidized $\ce{D}$ diffuses back to the counter electrode to be refilled with electrons. (C) On the left is shown the energy of the valence and conduction band edges at the interface and on the right the energies are shown corresponding to the first ($\ce{DH^-/D^{\cdot-}}$) and second ($\ce{D^{\cdot -}/D}$) oxidation potentials of the two-electron redox species. Reversible electron transfers from $\ce{DH^-}$ to the valence band ($v$) and $\ce{D^{\cdot -}}$ to the conduction band $c$ will contribute to successful bifurcating current, but energy wasting short-circuit reactions are also possible (D) Productive reactions conserve the reducing power of the reactants and contribute to bifurcating current. Short circuit reactions (which can be prevented, vida infra) reduce a high-potential acceptor (blue) from a low-potential donor (red) and waste free energy.
• Figure 3: Model bifurcating junction and its performance in simulation (A) Model electron bifurcating junction (to scale). The dimensions of the model were chosen so that the depletion region could be fully accommodated within the device. (B) Simulated performance of the model bifurcating junction. The injected $I_{\text{inj}}/A$ (blue) and recombination $I_{\text{rec}}A$ (orange) currents both increase exponentially with $V_{\text{bias}}$, but with the predicted different exponential factors (Equations \ref{['Eq:Iinj']} and \ref{['Eq:Irec']}). Both currents are divided by the surface area $A$ of the interface for meaningful comparison. Thus, at low currents the bifurcation efficiency $\eta_{\text{eff}}$ (red, in percent) is very high. Short-circuits dramatically reduce efficiency when $V_{\text{bias}} \approx 0.2-0.25$ V or greater, where the bifurcated current is $\sim 1 - 10$ nA/cm$^2$.