One of the most important chemical processes is interfacial charge transfer. It encompasses the entire field of chemistry and is involved in most heterogeneous chemistry and surface catalysis.
Our understanding of this problem prior to our work was based primarily on electrochemical measurements in which the picture for electron transfer at electrode surfaces was thought to be universally in the nonadiabatic or weak coupling limit. One has to think back to undergraduate chemistry courses in which the Helmholtz double layer is described and details of electronic coupling between discrete molecular/ionic acceptors at the interface were described as a weak interaction that limited current (reaction rates). This picture needed to be challenged as there are clearly surface adsorbed molecules, strong interacting with surfaces, to the point the electronic interaction is sufficient to cause bonding. There is a whole range of spatial locations, spatial minima, in the structure of the Helmholtz double layer with distances short enough to support strong coupling between molecular and solid state electronic levels.
The conundrum was that there is a huge mismatch in the electronic density of states across the discrete discontinuity defined by the electrode surface and the molecular acceptor or redox couple. On the electrode side of the half space, one is dealing with an electronic continuum of states, with electronic density of states defined by the band structure of the electrode material. In contrast, the molecular acceptor/donor state is truly a discrete state with the degree of electronic mixing determined by the wavefunction overlap with the electronic band states. It is one of the few examples where there is such a large mismatch in electronic density of states, where continuum states meet singularly discrete quantum states. It is what makes this problem so fascinating. How should we think about this problem – arguably one of the most important forms of chemistry dating back to Faraday’s day? At the time, we didn’t even know how to treat the fundamental electronic coupling leading to electron exchange, i.e., the fundamental process leading to electrochemistry.
The above problem statement has even more importance in that most solar energy conversion concepts rely on a semiconductor to tune a threshold absorption band to match the maximum power spectrum of the sun and store the absorbed photon energy through charge separation – interfacial charge transfer. The overall conversion efficiency is reduced by any nonradiative relaxation processes competing with interfacial electron transfer – the energy storage step. At the time, there was a great deal of interest in the prospect of hot electron injection to increase the overall solar energy conversion from a theoretical maximum of 33% for a threshold absorber (at solar power max) to potentially 66% if the charge transfer process could occur faster than electron relaxation via electron phonon relaxation processes (Art Nozik’s concept, NREL). This prospect is only possible if the interfacial electron transfer step occurred on the 100 fs time scale to compete with thermalization losses though electron-phonon scattering. There were no direct measurements prior to the group’s work and conventional wisdom had pinned electron transfer at electrode surfaces, metal or semiconductor, as occurring the nonadiabatic or weak coupling limit. In other words, it was thought that interfacial electron transfer would be slow, much slower than nonradiative relaxation, thermalization of hot electrons with lattice phonons.
To go after this problem, the group developed 5 different spectroscopic methods to selectively study different aspect of the interfacial charge transfer process. These include Surface Restricted Grating Spectroscopy, Surface Acoustic Wave Generation, In Situ Space Charge ElectroOptic Sampling, Fs Photoemission Spectroscopy, and Time Correlated Fluorescence Quenching Spectroscopy at Atomically Flat Interfaces. The Surface Restricted Grating method exploited space charge fields to focus reactive minority carriers to the surface reaction plane to directly observe them undergo interfacial charge transfer. This holographic method is based on modulated changes in the index of refraction due to the written grating pattern stored as carrier density modulation. The extremely high sensitivity of this method due to the effectively zero background (relative to standard pump probe) made it possible to detect changes amounting less than 10-5 of a monolayer of reactive flux or charge carriers. This sensitivity was essential to avoid charge screening of the surface field and loss of spatial restriction of the charge to the reaction plane. This work showed 100 ps dynamics for hole carrier transfer for TiO2 systems (effective hole mass was found to be >> 1 rather than much smaller predicted value much less than 1) or fully thermalized charge transfer despite predictions for hot hole carrier injection for this system. GaAs studies found ps dynamics that still implicated thermalized charge transfer but taking into account the mismatch in density of state across the interface, this result illustrated thatthe electronic coupling was in the adiabatic or strong coupling regime – NOT weak coupling. The Space Charge ElectroOptic Sampling studies used the change in index of refraction accompanying charge separation, and resulting changes in surface fields, in noncentrosymmtric materials – directly in the surface space charge region – to give the fastest possible time resolution or highest effective bandwidth possible for detecting electric field transients. We held the record for 10 fs time scales to changes in electric field (it was all optical with no leads or RC factors to limit bandwidth. The optical detection was directly in the junction – could not be faster time resolution). This work was the first to follow the field acceleration of charge carriers to the reaction plane, to correctly describe the carrier arrival time for connecting to charge transfer and the field accelerated carriers’ energy. With Fs Photoemission we were able to full resolve photoexcited electron relaxation at the surface region (which could be different/faster than bulk). Surface Acoustic Wave studies were initially thought to be related to the electrostatic structure of the double layer but were later shown to be due to surface acoustic waves confined in the liquid and solid half spaces interfering with one another (so called Stonely modes). This information was still valuable in thinking about the double layer structure and affect on the solid half space. The work on time resolved fluorescence quenching using the “poor man’s ultraclean surface” (quote from Bruce Parkinson) SnS2 really broke this problem for us. We showed that by time resolving the fluorescence and ground state bleach (to prove quenching due to interfacial charge transfer) with sensitivity to changes in transmission to 10-8 – absolute shot noise limited detection using high speed time scans - we demonstrated interfacial charge transfer could occur within 40 fs, i.e., on the 10 fs time scale. This primary reaction step is actually faster than the nuclear fluctuation timescales normally associated with chemical processes and electron transfer in particular. What was going on? In order for a chemical process to be irreversible through a barrier (to not recross), the reaction coordinate must undergo dephasing to collapse the quantum aspects of this process and localize the reactive crossing on the product state surface. In all chemistry, till this work, this dephasing involves coupling to a nuclear continuum of states for which the sampling time is dictated by the time scale for nuclear fluctuations to sample different configurations within the continuum of possible nuclear configurations. What we realized is that the same dephasing requirement will occur if the reaction coordinate couples to an electronic continuum as well, in this case the electron going from a discrete molecular state into the continuum as defined by the semiconductor/electrode electronic band structure. You can think of the electron from the molecular acceptor as being a bullet fired out of a proverbial chemical reaction/gun into vacuum. There is no chance the bullet will go back into the gun as the mismatch in available space makes this impossible. The bullet’s motion in this analogy leads to a charge separated state. This work redefined the upper limit speed limit to chemical reaction dynamics. The time scale of chemistry is no longer defined by nuclear fluctuations or sampling times (as for the A prefactor in the Arrhenius expression) but by the electronic coupling to an electronic continuum. The lifetime of the excited state due to electron injection into the electronic continuum of the solid state conduction band gives a direct measurement of the electronic coupling. These studies were the first to provide unambiguously this fundamental information defining interfacial electron transfer.
The above body of worked showed that electron transfer at interfaces is more generally in the strong coupling or adiabatic coupling not the nonadiabatic coupling limit as assumed from electrochemical measurements. It is now clear, electrochemical current measurements (even rotating electrode approaches) are limited by the slowest rate, which involves translational diffusion from outside the Helmholtz double layer to be within wavefunction overlap. These measurements do not measure the fastest steps which the above methods do as well as provide different information. This work illustrated that for dye sensitizer concepts for solar energy conversion/solar cell development, the molecular acceptor excited state only has to lie above the conduction band to ensure strong, adiabatic, electronic coupling with electron transfer times on the femtosecond timescale – to give a direct measurement of the electronic coupling – the most fundamental parameter in understanding interfacial electron transfer. This work solved the conundrum in the mismatch in the electronic density of states between conduction bands and discrete molecular acceptor AND showed how simple it is to ensure strong electronic coupling for solar cell development.
There has been an enormous world-wide effort to develop low cost dye sensitized solar cells to replace Si solar cells for renewable energy technology platforms. There have been very promising results with dye sensitized TiO2 solar cells and very impressive conversion efficiencies with perovskite materials. The basic concepts of electronic coupling discussed above seem to be rediscovered ever so often in perusing the literature in this field. The hope is the above work helped sort out the most fundamental issue with respect to electronic coupling – and how to have the desired charge transfer process (energy storage step) occur faster than nonradiation/thermalization loss channels for a given material.
The above body of work essentially mapped out all the relevant photophysical pathways for interfacial charge transfer. It is akin to mapping out all the photophysical processes for molecular photophysics as embodied in Jablonski diagrams – but here for surface electron transfer. All the relevant photophysics in the above figure have been mapped out using the above different spectroscopies. This body of work was reviewed in the J. Phys. Chem. Feature Article Lanzafame et al 1994. This work also was part of a monograph, “Surface Electron Transfer Processes” by R. J. Dwayne Miller, George McLendon, Arthur J. Nozik, Wolfgang Schmickler, and Frank Willig, Wiley, 1995. closing this chapter in the Miller group research efforts.