Solar-to-Chemical Conversion

Efficient unidirectional singlet energy transfer among chromophores relies on adequate spectral overlap between the emission of the donor pigment and the absorption of the acceptor pigment and on successful avoidance of alternative excited‐state decay pathways. Suitable energy gradients that involve sacrificing some of the excitation energy ensure rapid transfer along the productive direction. The principal theory of excitation energy transfer is that of Förster resonance energy transfer (FRET) [27]. This relates the rates of excitation energy transfer to the dipole–dipole interactions between the donor and acceptor chromophores. The transfer rate according to FRET is determined by the relative orientations of the local electronic transition dipoles and the distance between them, and the theory formally applies to pigments that are spatially well separated, or equivalently when the electronic coupling between pigments is negligible. These conditions are rarely fully met in actual biological antenna complexes; therefore FRET most often provides a rough approximation at best. In this case a more appropriate approach is based on extensions of the Redfield theory [28] that treat the strong excitonic coupling non‐perturbatively and take into account interactions with the environment. The close proximity of chromophores within the protein matrices implies strong electronic interaction between donor and acceptor pigments that results in shared excitonic energy levels that lie lower than those of the independent pigments. It is currently accepted that quantum coherence plays a key role in the remarkable efficiency of light harvesting in natural photosynthesis [29–31].

Considering biomimetic approaches to artificial versions of antenna complexes, certain challenges become immediately apparent. The pigments used in biological light harvesting are rather small molecules that are elaborately positioned and electronically fine‐tuned by a “smart” protein matrix, resorting only to limited extent to covalent bonding. In this respect, it is tempting to consider the possible role of template‐guided assembly of pigments as opposed to the conventional synthetic approach of covalently linking arrays of chromophores. It is remarkable that even when using several molecules of the same pigment, the protein matrix can modulate their absorption profile to produce a range of site energies with well‐defined spatial distribution. This is important for expanding the spectral range and light‐harvesting ability of the antenna beyond the intrinsic features of a given pigment, but its directed nature makes it also the basis of a crucial functionality: the creation of energy gradients within the antennae that enable efficient transfer of excitation energy to the reaction centers. It is likely that similar functionality could be built synthetically by utilizing ordering of distinct chromophores rather than manipulating the properties of a given pigment in a site‐dependent manner. The high effective concentration of chromophores achieved in antenna complexes also appears hard to achieve in artificial analogs while avoiding concentration quenching (self‐absorption) and seems to be possible in nature only because of the pigment organization imposed by the protein scaffold. Finally, the adaptability to changing light conditions, for example, by rerouting excitation energy transfer pathways, and the intrinsic photoprotection mechanisms of photosynthetic enzymes and antenna complexes are features that would be difficult, though not impossible [32], to replicate outside biology.

Artificial light‐harvesting complexes [33, 34] can be conceived in the context of supramolecular chemistry as arrays of chromophores coupled through covalent linkages based on dendrimer architectures or assembled on scaffolds. Dendrimers [35, 36] lend themselves to V‐shaped or circular arrangements of covalently bonded chromophores and can exhibit inherent directionality in excitation energy transfer at the molecular level. An important design concept in this area is the linkage of different but complementary chromophores to enable wide spectral coverage and to create energy gradients for efficient excitation energy transfer cascades. An additional requirement for a synthetic antenna complex is the successful interfacing with the synthetic reaction center. The coupling of the two modules should ensure efficient energy transfer from the light‐harvesting system to the charge separation site while avoiding uncontrolled perturbation of the latter by the former.

Dendrimeric synthetic antenna complexes have been constructed using transition metal complexes of Ru or Os, bridged by oligopyridine‐type ligands [37]. The choice of connecting ligand is important because it does not merely bring together mononuclear complexes into a closely packed ensemble, but it determines the overall nanoscale architecture of the dendrimeric structure and controls both the local electronic properties of the metal‐based units and the electronic coupling between them [34]. Earth‐abundant first‐row transition metal complexes feature much less prominently in the field of light harvesting compared to 4d and 5d elements, but research efforts are currently directed toward changing this paradigm [38]. Multiporphyrin arrays, particularly utilizing zinc porphyrins, represent another common pattern in the context of bioinspired light‐harvesting complexes [33, 34]. On the other hand, light‐harvesting dendrimers constructed using solely organic subunits have also been explored [42]. Figure 3.5depicts two examples representative of ideas mentioned above. Another molecular approach worth mentioning is based on host–guest constructions, where molecules such as organic dyes or transition metal complexes are accommodated within internal cavities of dendrimers [45, 46]. Research and development of artificial light‐harvesting systems is an active field with immense potential and diversity that goes far beyond what can be covered in the present chapter, so the interested reader is referred to the primary literature for more details.

Figure 3.5Examples of synthetic approaches to multi‐chromophore arrays: (a) a nine‐porphyrin array unit comprising a central free‐base porphyrin core that acts as final acceptor and is surrounded by eight energy‐donating zinc porphyrins [43].

Source: Choi et al. [41]. (b) Dendrimer consisting of a terrylenediimide (TDI) core with four attached perylenemonoimides (PMI) and eight peripheral naphthalenemonoimides (NMI) [44].

Source: Balzani et al. [34].

Referring back to Figure 3.2that depicts the main components involved in oxygenic photosynthesis, it is useful to translate that scheme into a corresponding energy/electron flow diagram, the so‐called Z‐scheme of photosynthesis shown in Figure 3.6. The essential features are the utilization of two charge separation events (at P680 of PS‐II and P700 of PS‐I) with distinct potentials and the closely spaced electron transfer cascades that contribute to stabilization of charge separation and unidirectionality of electron transfer. It is noted that purple bacteria utilize only a type II reaction center that lacks the water oxidizing ability of PS‐II, whereas green sulfur bacteria utilize only type I centers, related to PS‐I. A one‐step process is in principle harder to apply to water splitting because of constraints placed on the reduction potential of the excited reaction center chromophore: it should be more positive than the water oxidation potential yet more negative than the hydrogen evolution potential. This creates limitations regarding the minimum excitation energy required to drive a single‐step water splitting process. By contrast, the two‐step process embodied in the Z‐scheme of oxygenic photosynthesis relaxes these constraints by utilizing two photons per electron transferred from water to the final electron acceptor and hence being able to use sunlight of lower energy that what would have been necessary otherwise. In the context of artificial photosynthesis, an implementation of the Z‐scheme (for instance, in multi‐junction photovoltaic devices) would similarly offer higher flexibility in the choice of materials and redox linkers/mediators, requiring only that the excited‐state potential of the reaction center at the oxidative side be lower than that of the reaction center at the reducing side.