When electrons and protons get excited

The coupling of electron and proton transfer reactions is central to a broad range of biological and chemical processes, including photosynthesis, respiration, and various solar energy devices. Elucidation of the fundamental physical principles underlying proton-coupled electron transfer (PCET) reactions (1–5), in which both electrons and protons are transferred, is critical for a complete understanding of these energy conversion processes. This level of understanding is particularly important for the development of alternative renewable energy sources. The detailed mechanistic study of PCET reactions is challenging because of the complexity of the biological and chemical systems associated with such reactions. As illustrated by a study in PNAS (6), however, the application of modern spectroscopic techniques to relatively simple model systems provides an unprecedented level of mechanistic detail about PCET reactions. This pioneering work lays the foundation for future experimental and theoretical investigations to further unravel the intricacies of coupled electron and proton motions.

The report by Westlake et al. (6) in PNAS centers on the PCET reactions induced by optical excitation of hydrogen-bonded dyes. One of the model systems studied is a phenol molecule hydrogen-bonded to an amine base, as depicted in Fig. 1A. Optical excitation of this system induces intramolecular charge transfer (ICT), which can be viewed as an electron transfer from the oxygen to the nitro group within the phenol molecule. This shifting of the electron density away from the oxygen decreases the strength of the O-H bond and leads to proton transfer from the oxygen to the nitrogen in the base. The experiments implicate two different mechanisms for this PCET process. The first mechanism is sequential, where the intramolecular electron transfer is followed by proton transfer. The second mechanism is a concerted electron–proton transfer (EPT) mechanism in which the electron and proton transfer reactions occur simultaneously during the optical excitation.

Fig. 1.

Schematic illustration of the optically excited ICT and ICT-EPT processes. (A) Nitrophenyl-phenol hydrogen bonded to t-butylamine. Electron and proton transfer are depicted by arrows. These charge transfer reactions may be accompanied by a change in the torsion angle θ. (B) Schematic depiction of the diabatic states, where the green ellipse indicates the covalent bond involving the transferring hydrogen, the ground state (GS) corresponds to the hydrogen covalently bonded to the oxygen of the phenol, the ICT state corresponds to the electron transferred from the oxygen to the nitro group in the phenol and the hydrogen covalently bonded to the oxygen of the phenol, and the ICT-EPT state corresponds to the electron transferred from the oxygen to the nitro group in the phenol and the hydrogen covalently bonded to the nitrogen of the amine in an elongated bond. (C) Schematic depiction of the potential energy surfaces corresponding to the ICT and ICT-EPT states as functions of the position of the hydrogen nucleus, r_p, and another solute coordinate, Q, which could be the torsion angle θ or a combination of solute modes. Optical excitations from the ground state to the ICT (red arrow) and ICT-EPT (blue arrow) states as well as the subsequent relaxation pathways are shown schematically. Although proton transfer occurs during optical excitation to the ICT-EPT state (green ellipse in B), the hydrogen nucleus relaxes on a slower timescale (blue path). Image courtesy of Alexander Soudackov and Brian Solis (Department of Chemistry, Pennsylvania State University).

These two mechanisms are differentiated by the nature of the excited state populated directly upon optical excitation. In the ICT state, the hydrogen is still covalently bonded to the oxygen, but in the ICT-EPT state, the hydrogen becomes covalently bonded to the nitrogen of the base. The main difference between these two excited states, which are depicted schematically in Fig. 1B, is the shift of the electronic charge associated with a covalent bond from the O-H to the H-N at the hydrogen-bonding interface. Because the optical excitation occurs on a much faster timescale than nuclear rearrangements, the nuclei, including the transferring hydrogen nucleus, are assumed to remain stationary during the optical excitation. Thus, in the ICT-EPT excited state, the covalent H-N bond is elongated because the hydrogen nucleus is still in its initial position, where it was covalently bonded to the oxygen. Experimental evidence for the population of both the ICT and ICT-EPT excited states is provided by femtosecond transient absorption measurements. In these experiments, two distinct spectroscopic signatures that are consistent with the ICT and ICT-EPT states are observed. Further confirmation is provided by coherent Raman measurements.

The presence of two spectroscopically accessible states corresponding to the ICT and ICT-EPT states defined in Fig. 1B implicates the two different PCET mechanisms. Direct population of the ICT state (red arrow in Fig. 1C) corresponds to the sequential mechanism with electron transfer followed by proton transfer. In this case, proton transfer corresponds to a transition from the ICT state to the ICT-EPT state after or in conjunction with vibrational relaxation processes (red path in Fig. 1C). This type of photoinduced excited state proton transfer mechanism has been studied by other groups (7–9). In contrast, direct population of the ICT-EPT state (blue arrow in Fig. 1C) corresponds to the concerted EPT mechanism. This mechanism is associated with the virtually instantaneous change in the electronic charge distribution from the ground state (GS) to the ICT-EPT bonding arrangement (Fig. 1B) during optical excitation, but vibrational relaxation of the hydrogen nucleus within the ICT-EPT state occurs on a slower timescale after this optical excitation (blue path in Fig. 1C). To our knowledge, this type of concerted EPT mechanism has not been reported previously.

This work calls into question the traditional definition of proton transfer as the movement of a hydrogen nucleus between a donor and acceptor atom. According to this perspective, a proton transfer reaction is defined in terms of the change in electronic configuration rather than the movement of the hydrogen nucleus. In this case, a proton transfer reaction is defined as the movement of electronic charge density from the donor–hydrogen covalent bond to the acceptor–hydrogen covalent bond. Such a change in electronic configuration without the motion of the hydrogen nucleus is special to optically induced processes, where the optical excitation is much faster than nuclear motion. The proton is in a highly excited vibrational state immediately after optical excitation and subsequently relaxes to its equilibrium state. The investigation of the electronic and nuclear dynamics under these nonequilibrium conditions is particularly interesting because of the quantum mechanical behavior of hydrogen nuclei associated with their light mass and the complexity of electron–proton correlation arising from their attractive electrostatic interaction (10). The traditional Born–Oppenheimer separation between electronic and nuclear motions based on mass and timescale differences may break down in these cases (11, 12). Thus, these types of experiments force us to reevaluate previous conceptions and allow us to explore unusual yet fundamentally interesting behavior of electrons and nuclei under previously unattainable conditions.

Furthermore, these experiments lead to several provocative questions and provide a challenge for theoreticians who aim to model these types of processes. What do the potential energy surfaces of the ground state and excited ICT and ICT-EPT states look like? Using electronic structure methods, these excited-state potential energy surfaces can be calculated as functions of specified nuclear coordinates. For the system shown in Fig. 1A, these surfaces could be generated as functions of the position of the transferring hydrogen nucleus and the torsion angle in the phenol molecule, as depicted in Fig. 1C. The equilibrium position of the hydrogen nucleus corresponds to that of the covalent

The report by Westlake et al. in PNAS centers on the PCET reactions induced by optical excitation of hydrogen-bonded dyes.

O-H bond in the ICT state and the covalent H-N bond in the ICT-EPT state. The equilibrium torsion angle may be significantly different for the ground state than for the ICT and ICT-EPT states. Furthermore, these types of excited state surfaces may exhibit conical intersections, as depicted in Fig. 1C.

After the potential energy surfaces are generated and characterized, another question arises. How do the nuclei in the system move on these potential energy surfaces? Quantum dynamical and mixed quantum/classical dynamical methods can be used to simulate the motion of the nuclei. After photoexcitation, the nuclei are in a nonequilibrium configuration on the excited ICT and ICT-EPT states. The relative probabilities of the various nuclear pathways will be different for the ICT and ICT-EPT states. In addition, the nuclei may not behave classically, particularly the hydrogen nucleus, which is light enough to exhibit quantum mechanical behavior such as tunneling (13). For this reason, concerted PCET reactions may be described in terms of mixed electron–proton vibronic states rather than purely electronic states (14). The nuclei may also move on more than one potential energy surface, requiring the inclusion of nonadiabatic transitions among the surfaces (15, 16).

In addition to the nuclei in the solute molecules, the motion of the solvent nuclei may also be coupled to the PCET reaction. After the optically induced ICT, the solvent will be in a nonequilibrium configuration because the solvent polarization will correspond to the equilibrium configuration for the ground state rather than the excited state. For a polar solvent, the dipoles of the solvent molecules will be oriented to stabilize the charge distribution of the ground state. PCET reactions may be described in terms of one or two collective solvent coordinates that describe the change in the equilibrium solvent polarization associated with electron and proton transfer (17). Depending on the relative timescales of charge transfer and solvent relaxation, the dynamics of these collective solvent coordinates may be coupled to the PCET process.

Clarifying the fundamental principles of coupled electron and proton motions in relatively simple model systems is a critical step for understanding more complex biological and chemical systems. The synthesis and experimental characterization of such model systems provides the foundation for theoretical studies that can assist in elucidating the roles of the various components of these processes. In turn, these insights may assist in the design of more efficient energy conversion devices that take advantage of the strong coupling between electrons and protons.