Artificial Photosynthesis: Is Computation Ready for the Challenge Ahead?

A tremendous effort is currently devoted to the generation of novel hybrid materials with enhanced electronic properties for the creation of artificial photosynthetic systems. This compelling and challenging problem is well-defined from an experimental point of view, as the design of such materials relies on combining organic materials or metals with biological systems like light harvesting and redox-active proteins. Such hybrid systems can be used, e.g., as bio-sensors, bio-fuel cells, biohybrid photoelectrochemical cells, and nanostructured photoelectronic devices. Despite these efforts, the main bottleneck is the formation of efficient interfaces between the biological and the organic/metal counterparts for efficient electron transfer (ET). It is within this aspect that computation can make the difference and improve the current understanding of the mechanisms underneath the interface formation and the charge transfer efficiency. Yet, the systems considered (i.e., light harvesting protein, self-assembly monolayer and surface assembly) are more and more complex, reaching (and often passing) the limit of current computation power. In this review, recent developments in computational methods for studying complex interfaces for artificial photosynthesis will be provided and selected cases discussed, to assess the inherent ability of computation to leave a mark in this field of research.

Monte Carlo simulation and thermodynamic integration applied to protein charge transfer

We introduce a combination of Monte Carlo simulation and thermodynamic integration methods to address a model problem in free energy computations, electron transfer in proteins. The feasibility of this approach is tested using the ferredoxin protein from Clostridium acidurici. The results are compared to numerical solutions of the Poisson-Boltzmann equation and data from recent molecular dynamics simulations on charge transfer in a protein complex, the NrfHA nitrite reductase of Desulfovibrio vulgaris. Despite the conceptual and computational simplicity of the Monte Carlo approach, the data agree well with those obtained by other methods. A link to experiments is established via the cytochrome subunit of the bacterial photosynthetic reaction center of Rhodopseudomonas viridis.

Long-range electron-electron interaction and charge transfer in protein complexes: a numerical approach

With application to the nitrite reductase hexameric protein complex of Desulfovibrio vulgaris, NrfH2A4, we suggest a strategy to compute the energy landscape of electron transfer in large systems of biochemical interest. For small complexes, the energy of all electronic configurations can be scanned completely on the level of a numerical solution of the Poisson-Boltzmann equation. In contrast, larger systems have to be treated using a pair approximation, which is verified here. Effective Coulomb interactions between neighbouring sites of excess electron localization may become as large as 200 meV, and they depend in a nontrivial manner on the intersite distance. We discuss the implications of strong Coulomb interactions on the thermodynamics and kinetics of charging and decharging a protein complex. Finally, we turn to the effect of embedding the system into a biomembrane.

Charge transfer through a fragment of the respiratory complex I and its regulation: an atomistic simulation approach

We simulate electron transfer within a fragment of the NADH:ubiquinone oxidoreductase (respiratory complex I) of the hyperthermophilic bacterium Aquifex aeolicus. We apply molecular dynamics simulations, thermodynamic integration, and a thermodynamic network least squares analysis to compute two key parameters of Marcus' theory of charge transfer, the thermodynamic driving force and the reorganization energy. Intramolecular contributions to the Gibbs free energy differences of electron and hydrogen transfer processes, ΔG, are accessed by calibrating against experimental redox titration data. This approach permits the computation of the interactions between the species NAD+, FMNH2, N1a-, and N3-, and the construction of a free energy surface for the flow of electrons within the fragment. We find NAD+ to be a strong candidate for the regulation of charge transfer.

Simulating biological charge transfer: Continuum dielectric theory or molecular dynamics?

We study the thermodynamic parameters of Marcus's theory of charge transfer, the driving forces and the reorganization energies, using two widely applied approaches to bioenergetic problems that seem to be radically different: continuum dielectric theory via a numerical solution of Poisson's equation, and the thermodynamic integration approach based upon classical Newtonian molecular dynamics, as perfomed by Na et al., PCCP 19, 18,938 (2017). With application to a nitrite reductase NrfHA protein heterodimer, we obtain an excellent agreement between the respective driving forces with an r.m.s. deviation of 1.7 kcal/mol, and a lower limit to the reorganization energies. The computational methods turn out to be mutually supportive: molecular dynamics can be used to determine the parameters of a dielectric theory computation, which on the other hand can be used to properly rescale the reorganization energies and partition them into aqueous and protein contributions. In addition, we use the electrostatic approach to study the influence of Ca²⁺ ions on the free energy landscape of charge transfer.

Thermodynamic integration network study of electron transfer: from proteins to aggregates

We describe electron transfer through the NrfHA nitrite reductase using a thermodynamic integration scheme. Driving forces are hardly affected by dimerization, but the transport mechanism only emerges simulating the dimer.

Decaheme Cytochrome MtrF Adsorption and Electron Transfer on Gold Surface

Emergent electrical properties of multiheme cytochromes have promising applications. We performed hybrid simulations (molecular dynamics, free energy computation, and kinetic Monte Carlo) to study decaheme cytochrome, MtrF adsorption on an Au (111) surface in water and the electron transfer (ET) efficiency. Our results reveal that the gold surface's dehydration serves as a crucial driving force for protein adsorption due to large surface tension. The most possible adsorption orientation is with the ET terminal (heme5) approaching the gold surface, which yields a pathway for ET between the substrate and the aqueous environment. Upon adsorption, protein's secondary structures and central domains (II and IV) bonded with heme-residues remain relatively stable. MtrF surface mobility is dictated by thiol-gold interaction and strong binding between Au(111) and peptide aromatic groups. ET transfer rate across protein heme-network along the solvent-to-surface direction is slightly larger than that of the reverse direction, but lower than that of the solvation structure.