Perspective pubs.acs.org/JPCL
Probing Photosynthetic Energy and Charge Transfer with Two-Dimensional Electronic Spectroscopy Kristin L. M. Lewis and Jennifer P. Ogilvie* Department of Physics and Biophysics, University of Michigan, Ann Arbor, Michigan 48109-1040, United States ABSTRACT: Two-dimensional electronic spectroscopy (2DES) has emerged as a powerful method for elucidating the structure−function relationship in photosynthetic systems. In this Perspective, we discuss features of two-dimensional spectroscopy that make it highly suited to address questions about the underlying electronic structure that guides energy- and chargetransfer processes in light-harvesting materials. We briefly describe a pulse-shaping-based implementation of two-dimensional spectroscopy that is making the method widely accessible to problems spanning frequency regimes from the ultraviolet to the mid-infrared. We illustrate the utility of 2DES in the context of our recent studies of the primary energy-transfer and charge separation events in the photosystem II reaction center, discussing remaining challenges and speculating about exciting future directions for the field of multidimensional spectroscopy.
P
would yield useful information.14 The basic pulse sequence for a 2D Fourier transform spectroscopy measurement is illustrated in Figure 1. The two axes of a 2D spectrum are obtained by
hotosynthetic organisms have developed extraordinarily efficient mechanisms for harvesting solar energy; elaborate antenna arrays collect solar energy and transfer it to photochemical reaction centers with near unit quantum efficiency on a 10−100 ps time scale.1 Once inside of the reaction centers, this energy is converted to a stable charge separation that drives slower biochemical reactions, ultimately producing chemical energy. The architecture of the light-harvesting and reaction center complexes contains important secrets about how to efficiently capture, transfer, convert, and store solar energy, our most abundant energy source.2 A wealth of spectroscopic studies has aimed at understanding the design principles of photosynthetic systems. Such systems present unique challenges to the spectroscopist; the choice of pigments, their particular protein environments, relative distances, and orientations produce broad inhomogeneous transitions that reflect varying degrees of electronic coupling, disorder, and dynamical fluctuations of the protein matrix. The ultrafast nature of the energy-transfer and charge separation processes further complicates an already demanding spectroscopic problem. Recently, two-dimensional electronic spectroscopy (2DES) has emerged as a valuable tool for probing the structure−function relationship in natural photosynthetic complexes.3−8 The aim of this Perspective is to discuss aspects of 2DES that make it a useful probe of electronic structure and energy and charge transfer in multichromophoric systems. We illustrate these points in the context of our 2DES studies of the photosystem II reaction center.6 These studies employed a particularly simple implementation of 2DES that is readily adapted from a standard pump−probe experiment.9−13 We briefly describe this implementation, which should facilitate the wider adoption and further development of multidimensional methods, bringing them to bear on a broader range of problems. Two Dimensional Electronic Spectroscopy: What Is It and Why Use It? It was realized as early as 1976 by Ernst that twodimensional (2D) spectroscopy in other frequency regimes © 2012 American Chemical Society
Figure 1. Pulse sequence for 2D Fourier transform spectroscopy.
Fourier transforming with respect to the t1 and t3 time variables, producing excitation frequency ω1 and detection frequency ω3, respectively, for a given waiting time t2. To date, there have been considerably more 2D experiments in the infrared (2DIR) than at visible frequencies. This can in part be attributed to the relatively more difficult experimental implementation of 2D spectroscopy at higher frequencies; obtaining reliable frequency axes requires precise knowledge of the time delays to be Fourier transformed, and interferometric measurements require high phase stability.15 Early work by the Jonas group,16,17 along with technical developments using diffractive optics18,19 and pulse shaping,9−13,20,21 have paved the way for a wide range of applications of 2DES. A number of recent review papers20,27,28 and two monographs22,23 summarize work in the 2DIR and 2DES fields.15,24,25 Figure 2 depicts several types of information provided by 2DES that is of particular importance for studying photosynthetic systems. Figure 2A illustrates how inhomogeneous Received: December 3, 2011 Accepted: January 27, 2012 Published: January 27, 2012 503
dx.doi.org/10.1021/jz201592v | J. Phys. Chem. Lett. 2012, 3, 503−510
The Journal of Physical Chemistry Letters
Perspective
Figure 2 also illustrates how 2D spectra provide information about electronic coupling in a simple two-pigment model. Photosynthetic systems contain pigments with a range of relative separations and orientations, leading to Coulombic coupling that varies from weak, where excitations are respectively localized on individual pigments, to strong, where excitations are shared among pigments and an excitonic picture is appropriate.1 In Figure 2B, we depict the case of weak Coulombic coupling, where the relative orientation and separation are favorable for energy transfer between pigment b (donor) and pigment a (acceptor) via the Förster mechanism. In this case, the t2 = 0 2D spectrum shows peaks along the diagonal corresponding to the individual transitions of the pigments, revealing the inhomogeneity of the transitions. No cross-peaks are observed because excitations on each pigment are localized. For t2 > 0, a cross-peak appears below the diagonal, indicating that energy absorbed by the donor has been transferred to the acceptor. In Figure 2C, we imagine the case of strong Coulombic coupling between the two pigments. In this case, a common ground state gives rise to cross-peaks in the t2 = 0 2D spectrum. Such cross-peaks provide a direct signature of excitonic coupling. Polarization-dependent measurements can provide information about the relative orientation of the coupled transitions.27 The Fourier transform implementation of 2DES provides a key advantage over methods such as pump−probe spectroscopy for unraveling ultrafast kinetic processes. Consider a system as illustrated in Figure 3, with a series of closely spaced excited Figure 2. (A) Cartoon 2D spectra for an inhomogeneously broadened system. The t2 = 0 spectrum shows the degree of inhomogeneous broadening. For t2 ≫ 0, spectral diffusion occurs, and the remaining asymmetry in the diagonal/antidiagonal width is related to static disorder. (B) Energy level structure and cartoon 2D spectra for a twopigment system in the case of weak coupling. At t2 = 0, no cross-peaks are present in the 2D spectrum, while at later waiting times, the appearance of a cross-peak below the diagonal indicates energy transfer. (C) Energy level structure and cartoon 2D spectra for a twopigment system in the case of strong coupling. At t2 = 0, cross-peaks indicate excitonic coupling via a common ground state. For t2 > 0, the cross-peaks carry the information about energy-transfer kinetics. We note that here, we have portrayed simplified absorptive 2D spectra, omitting excited-state absorption features.
broadening appears in a 2D spectrum and how its physical origin can be understood. In photosynthetic systems, inhomogeneous broadening occurs when pigments experience different local protein environments. Both static and dynamic disorder contribute to these differences. In general, a 2D spectrum recorded at a given t2 allows correlations to be made between excitation and detection frequencies. At t2 = 0, any particular transition that is excited will not yet have undergone dynamical processes that could change its resonant frequency; thus, its resonant frequency will be detected at its excitation frequency. This produces a diagonally elongated 2D spectrum at t2 = 0, where the widths along the diagonal and antidiagonal directions indicate the respective inhomogeneous and homogeneous broadening. As the system evolves (t2 > 0), the correlations between excitation and detection frequency will be lost if dynamical processes cause fluctuations in the transition frequencies (a process termed spectral diffusion).15 If the inhomogeneity arises from static disorder, the diagonal elongation will persist. Thus, the time evolution of the asymmetry of the 2D peak shape is an indicator of static versus dynamic disorder.26
Figure 3. Illustration of the benefits of Fourier transform 2DES for distinguishing between different kinetic models. To resolve rapid energy transfer, a short pulse excites states a, b, and c and, subsequent energy transfer to the lowest-energy state occurs via a sequential (A) or parallel (B) pathway where kcb ≈ kca ≫ kba. Shown to the right of the relaxation scheme are the 2D DAS derived from a fit of two exponentials to the 2DES data. Here, the solid circles refer to amplitude decay, while the dashed circles indicate growth of amplitude. When the time scales of energy transfer from state c and state b are not wellseparated, 2D DAS with overlapping features are obtained.
states. In this system, we are interested in distinguishing between two different kinetic schemes that describe energy 504
dx.doi.org/10.1021/jz201592v | J. Phys. Chem. Lett. 2012, 3, 503−510
The Journal of Physical Chemistry Letters
Perspective
inserting a pulse shaper into one arm of a standard transient absorption setup.9 This pump−probe method of 2DES uses a partially noncollinear beam geometry in which a pulse-shaper transforms a single pump pulse into a pair of pump pulses with a programmable time delay and relative phase. Demonstrated first at IR frequencies,10 it has been extended to the visible11,13 and ultraviolet30 regimes; our particular experimental implementation of 2DES is detailed in Figure 4. With collinear pump
transfer from state c to a; a sequential pathway where energy flows from c to b, then b to a, and a parallel pathway where direct transfer of c to a also occurs. If the states are wellseparated and energy-transfer time scales are slow, pump− probe spectroscopy can readily meet this challenge; for example, selective excitation of b or c would allow observation of the growth in population of state a. However, when the separation between states is small and/or the kinetics are too fast, selective excitation becomes impossible; to achieve the requisite time resolution, the bandwidth of the exciting pulse may become broader than the separation between exciton levels. Thus, selective excitation comes at the expense of time resolution. By separating the excitation into two steps, Fourier transform 2DES decouples the frequency selectivity of excitation with the time resolution of the experiment, enabling excitation frequency resolution determined by the sample itself.15 The rich kinetic information in 2DES is sometimes reflected in relatively small changes in peak shapes as a function of t2. To highlight these small changes, we have developed a decomposition analysis that yields two-dimensional decay-associated spectra (2D DAS) that reveal the 2D spectral signatures of different kinetic intermediates. This procedure is in the same spirit as the use of decay-associated spectra (DAS) and lifetime density maps derived from pump−probe data to identify kinetic intermediates and test kinetic models.28,29 To extract the 2D DAS from 2DES data, we independently fit a series of exponential decays to each frequency−frequency point and then cluster points with similar lifetimes into 2D DAS.6 The additional frequency axis provided by 2DES reveals the excitation frequency dependence of particular kinetic processes. Figure 3 illustrates how decomposition into 2D DAS readily distinguishes between the sequential and parallel kinetic schemes discussed above. For illustrative purposes, we make the simplifying assumption that the rates of transfer among the exciton levels obey kcb ≈ kca ≫ kba. This produces easily separated 2D DAS components as indicated, showing that the sequential and parallel schemes are readily distinguished. Excitation-frequencydependent resolution of kinetics is also helpful when the time scales are not well-separated. In 2D DAS, the decay and growth components are spread out along the excitation axis, facilitating their separation. This is analogous to incorporating the information from a multitude of pump−probe experiments performed with narrow-band selective excitation but without the sacrifice in time resolution required for excitation selectivity.
Figure 4. Experimental setup for 2DES with a pulse-shaper in the pump−probe geometry. NOPA: noncollinear optical parametric amplifier; WLG: white light generation. The Dazzler is a commercially available acousto-optic pulse-shaping device.
pulses, this technique simplifies many of the challenges of 2DES.16 In particular, “absorptive” spectra, which are free from broadening refractive contributions, are automatically obtained in this method. The use of a pulse-shaper to generate the pump pair provides phase-cycling opportunities similar to those used in NMR to extract and enhance signals of interest and improve data quality in scattering samples.9,31 While absorptive 2D spectra are often desired, rich information is also present in the constituent “rephasing” and “nonrephasing” spectra, particularly for studying electronic and vibrational coherences.32,33 Should the individual rephasing and nonrephasing signals be desired, we have shown that they can be readily obtained via phasecycling.13 2DES of the Photosystem II Reaction Center. Photosystem II is a multisubunit dimeric membrane-spanning complex that catalyzes the most thermodynamically demanding reaction in living systems, the splitting of water into O2 and reducing equivalents necessary to fix carbon dioxide and produce organic molecules.34,35 Understanding the link between structure and function in this unique system remains a grand challenge in biology. The primary processes in photosystem II are the absorption of light energy by the light-harvesting antennae and the transfer of energy to the photosystem II reaction center (PSII RC). The PSII RC is composed of the D1 and D2 proteins that bind the redox-active cofactors necessary for the process of solar energy conversion. Simplified preparations containing the D1 and D2 proteins as well as cytochrome b559 but lacking the antennae proteins and reducing side plastoquinone acceptors are frequently used to study the processes of energy and charge transfer within the PSII RC. These preparations, which we will refer to as simply PSII RC (also called D1D2-Cyt b559 photosystem II reaction center), stop short of the final charge separation that precedes water splitting, creating a
Despite intensive efforts, the mechanisms and time scales of the energyand charge-transfer processes in the PSII RC are not yet fully understood. 2DES Made Simple. Many different experimental implementations now exist for 2DES, varying in both the beam geometries used and the methods employed for the generation of the time delays. One particularly simple approach amounts to 505
dx.doi.org/10.1021/jz201592v | J. Phys. Chem. Lett. 2012, 3, 503−510
The Journal of Physical Chemistry Letters
Perspective
short-lived radical pair PD1+PheoD1−. The primary charge separation processes are thought to be representative of intact photosystem II,29,36 and numerous spectroscopy studies have aimed at understanding the structure−function relationship of this system.37−39 Despite intensive efforts, the mechanisms and time scales of the energy and charge-transfer processes in the PSII RC are not yet fully understood. One significant difficulty is that the constituent pigments have severely overlapping absorptions, even at 77 K (shown in Figure 5B). The PSII RC
Figure 6. (A) 2DES of the PSII RC at 77 K at waiting times of t2 = 28 and 215 fs. (B) The first (
