Synthetic Organic Design for Solar Fuel Systems

Abstract From the understanding of biological processes and metalloenzymes to the development of inorganic catalysts, electro‐ and photocatalytic systems for fuel generation have evolved considerably during the last decades. Recently, organic and hybrid organic systems have emerged to challenge the classical inorganic structures through their enormous chemical diversity and modularity that led earlier to their success in organic (opto)electronics. This Minireview describes recent advances in the design of synthetic organic architectures and promising strategies toward (solar) fuel synthesis, highlighting progress on materials from organic ligands and chromophores to conjugated polymers and covalent organic frameworks.


Introduction
Since the demonstration of photoelectrochemical water splitting using the semiconductor (SC) TiO 2 and the H 2 evolution catalyst (HEC) platinum, inorganic materials have dominated the field in both number and efficiency. [1] They still undergo fast development, but often lack understanding at an atomic level, which limits the flexibility and fine-tuning capabilities needed to rationally improve (photo)catalytic performance.
Biology,t hrough photosynthesis and fuel-making enzymes even in non-photosynthetic organisms,p rovides blueprints for the design of dyes and catalysts with outstanding performance such as the light-harvesting complexes,H 2evolving hydrogenase (H 2 ase) and the CO 2 reductases. [2] The activity of these biocatalysts relies on the choreography of evolutionarily-developed design principles,i ncluding:a ctive sites with an optimized primary and outer coordination sphere to stabilize reaction intermediates,e fficient energy and electron transfers (ETs), well-aligned electroactive ligands and electron relays,a sw ell as substrate and product channels.
Although some of these concepts are being implemented toward artificial photosynthesis and (photo)catalysis,e specially in the design of molecular electrocatalysts,most reports focus on inorganic systems.I na ddition to coordination complex catalysts and natural archetypes,( semi-)organic (hybrid) materials have emerged in the field of catalysis. [3] Themodularity and amenability displayed by these materials offer af ertile ground for integration in catalytic schemes, which have led to rapid developments in organic photo-and electrochemistry. [4] Here,w es ummarize the progress toward developing electro-and photo-catalysis systems for fuel synthesis enabled by organic design, organized according to molecular and polymeric concepts.A lthough the synthetic strategies described in this mini-review focus on the fuel forming, reductive half-reaction, analogous design can also be employed in water oxidation catalysis with oxidatively robust organic architectures.

Ligand Design for Catalysts
The[NiFe]-or [FeFe]-active site in H 2 ase and the O 2 -evolving [CaMn 4 ]cluster in Photosystem II (PSII) inspired the development of early structural biomimetics as molecular catalysts. [5] Catalysts were initially designed to mimic the first coordination sphere of the enzyme active site,which led to Fe 2 S 2 -type HECs delivering modest performances, [5f, 6] and having the propensity to decompose into active particles. [7] Tu ning the ligands substituents to affect the electronic density and electrochemical properties was also extensively investigated. [5e,f,6c] Amore recent approach is to innovate modulation of the second and outer coordination spheres of the metal center with organic residues in an attempt to replicate the multifunctionality found in enzymes ( Figure 1). [6c, 8] Ac lass of Ni-containing HECs containing aP 2 N 2 ligand (Dubois catalyst) displays high activities due to pendant basic tertiary amines that promote proton transfer. [11] HEC 1 with further arginine (Arg) residues displayed reversible H 2 production/oxidation in acidic aqueous solutions (Figure 2). [8c] Theh igh turnover frequency (TOF) of 300 s À1 was attributed to Arg-Arg interactions that aid positioning of the pendant-amine groups in close proximity to the Ni center.
From the understanding of biological processes and metalloenzymes to the development of inorganic catalysts,e lectro-and photocatalytic systems for fuel generation have evolved considerably during the last decades.Recently,organic and hybrid organic systems have emerged to challenge the classical inorganic structures through their enormous chemical diversity and modularity that led earlier to their success in organic (opto)electronics.T his Minireview describes recent advances in the design of synthetic organic architectures and promising strategies toward (solar) fuel synthesis,highlighting progress on materials from organic ligands and chromophores to conjugated polymers and covalent organic frameworks.  [9] and (b) synthetic Fe-porphyrin [10] with bound CO 2 (in red) stabilized by outer coordination sphere interactions (in blue).
Tu ning porphyrin ligands to improve the performance of aC O 2 reduction catalyst (CRC) was demonstrated with Fetetraphenylporphyrins (TPPs). [8d, 10,12] Fe-TPPs display acatalytic onset potential (E cat )o fÀ1.40 Vv s. standard hydrogen electrode (SHE) in N,N-dimethylformamide (DMF). [12] The catalytic performance was optimized by stabilization of the initial Fe 0 -CO 2 adduct upon addition of positively charged N,N,N-trimethylanilinium groups in the ortho position of the four phenyl groups in 2 ( Figure 2). This modification resulted in amore anodic E cat of À0.95 Vvs. SHE with a3-fold greater catalytic current than the corresponding "para"c atalyst, while also delivering high Faradaic efficiencies (FEs) toward CO with limited degradation over 84 hofelectrolysis in DMF containing phenol and H 2 O. [8d] Thecatalysis-enhancing effect was attributed to Coulombic interactions of the positively charged moieties with the carboxylate borne from the Fe 0 -CO 2 adduct.
Other concepts to improve performance have been reviewed elsewhere and include (non-exhaustively): modulating steric hindrance around the metal core, [13] isolating catalysts on surfaces via anchoring groups, [14] providing Brønsted acid groups to deliver proton relays, [8b] H-bonded and multimetallic systems for electrostatic stabilization of CO 2 -bonded intermediates, [15] and tailored CO 2 -fitting clefts. [10] Importantly,aswater oxidation represents ascalable and readily available source of electrons for (solar) fuels production, many molecular catalysts for the oxygen evolution reaction have been developed with ligand design toward modulating the catalystso uter coordination sphere also representing an active area of research. [5a, 16] In particular,Ru coordination complexes currently display benchmark perfor-mance and more in-depth reviews on this topic can be found elsewhere. [17]

Dyes
Photosensitizers (PSs) can harvest light to drive asuitable electrocatalyst for solar fuel synthesis.C ommon homogeneous photocatalytic systems employ commercial dyes,e .g., Ir and Ru complexes,t hat are regenerated by as acrificial electron donor (SED). Purely organic PSs,now ubiquitous in dye-sensitized solar cells (DSCs), have been much less investigated for the demanding requirements of artificial photosynthesis such as aqueous conditions and endergonic multi-electron processes. [18] Common limitations include their lack of solubility in aqueous media, modest stability and short-lived excited states that impede diffusional ET.Building on chromophoric units,d rawbacks can be overcome by molecular engineering to yield suitable PSs (Figure 3).
Fori nstance,t riazatriangulenium 3 ( Figure 4) displays intense visible-light absorption (l max = 530 nm; e = 8800 m À1 cm À1 )w ith ar elatively long excited-state lifetime of its singlet state (14 ns at pH 4.5). [20] This allows in solution for ad iffusion-controlled reductive quenching by ascorbic acid (AA) to generate the organic radical 3C À with excellent stability due to the planar scaffold incorporating three electron-donating nitrogen atoms for delocalization of the radical. As aresult, efficient ET to amolecular Co HEC was   An alternative PS-design strategy to improve intermolecular ET is the promotion of intersystem crossing toward along-lived triplet state.Halogenation of aborondipyrromethene (bodipy) PS yields diiodide-bearing 4 ( Figure 4) that has ac onsiderably shorter fluorescence lifetime than the corresponding iodine-free PS (0.13 vs.6 .0 ns,r espectively), indicating fast intersystem crossing. [21] PS 4 reached aTON PS of 600 when employed at low concentrations with amolecular Co HEC and triethanolamine (TEOA) as aS ED in acetonitrile under Xe/Hg lamp irradiation.
Organic dyes have also been involved in colloidal dyesensitized SC photocatalysis (DSP) systems toward H 2 evolution and CO 2 reduction (Figure 5a). [22] Building upon DSC principles,D SP systems are assembled through attachment of an anchor-bearing PS to the surface of aSCparticle, together with ac o-attached catalyst. [18a,22a] TheS Cd elivers dual functionality as it provides ascaffold for immobilization to enable fast charge separation and allows accumulation of multiple long-lived charges in the SC to drive catalysis.T he regeneration of the photoionized PS generally relies on aSED.
As eries of diketopyrrolopyrrole (DPP) dyes with modulated energetic and non-energetic parameters (e.g., steric hindrance,position and nature of the solubilizing side chains) were studied in DSP systems. [23] When attached to platinized TiO 2 and placed under simulated sunlight, PS 5 ( Figure 4) delivered ap erformance (TON PS in 5 j TiO 2 j Pt % 2700) superior even to the corresponding phosphonic acid-bearing Ru trisbipyridine-based assembly.The performance of the PS in DSP systems was shown to ultimately depend on the orthogonal adequacyo fP Sd esign and external parameters (pH, SED,c hemical catalyst and mechanistic details).
Aseries of hydrophobic perylene monoimide (PMI) dyes were functionalized with five different anchoring groups: carboxylic acid, phosphonic acid, acetylacetone,pyridine-2,6dicarboxylic acid and hydroxyquinoline.T hese PSs were investigated in DSP using platinized TiO 2 nanoparticles for H 2 evolution in water. [24] TheC O 2 H-bearing PS 6 ( Figure 4) delivered the best performance in acidic and neutral pH with as tability beyond 3days and aT ON PS of % 11 000. The activity decreased at higher pH due to desorption of the PS from the SC surface.I nc ontrast, albeit yielding modest activity,t he phosphonic acid-bearing PS 7 ( Figure 4) enables good stability due to the anchoring groupsb etter resistance to hydrolysis.T hese results highlight an ongoing challenge in DSP,w here electron injection-promoting properties are not yet readily compatible with robust anchoring.
Thet unable electronic properties of organic chromophores allow their use for the absorption and conversion of low-energy photons.P S8 (Figure 4) shows the extended conjugation of ab odipy to ap henothiazine,w ith its donor-p conjugated linker-acceptor (d-p-A) organization leading to as trong panchromatic absorption up to 700 nm with l max = 638 nm (e = 123,000 m À1 cm À1 ). Attachment of 8 on platinized hierarchical porous TiO 2 resulted in aDSP system toward H 2 evolution, with aTON PS of 11 100 after 10 hofirradiation (l > 400 nm, 100 mW cm À2 )i nt he presence of AA as SED.A n apparent quantum yield (QY) of 1.0 %a t7 50 nm was recorded. [25] Overcoming the parasitic,f ast electron recombination in PS-catalyst systems as well as suppressing the need for aSED are current challenges in colloidal and homogeneous schemes. Anchoring ad ye and catalyst onto ap -type SC (p-SC) electrode to produce ad ye-sensitized photocathode (DSPC, Figure 5b)isapromising strategy inspired by p-type DSCs to address these limitations. [3c, 26] PS 9 (Figure 4) was co-anchored to an anostructured NiO substrate together with ac obaloxime HEC. [27] Thep ush-pull design in 9 locates the dyesh ighest occupied molecular orbital close to the NiO surface,thereby promoting hole injection and intermolecular ET to the HEC upon photoexcitation. Af ully assembled tandem photoelectrochemical cell with ad ye-sensitized photoanode demonstrated overall water splitting. [27] Incident photon to current conversion efficiency analysis identified the photocathode as the overall systemsb ottleneck, ascribed to the limitations of NiO (e.g.,s hort hole diffusion length). Alternative p-SCs such as CuCrO 2 (combined with PSs 5 and 7)a nd LaFeO 3 have recently been proposed as ap otential replacement for NiO. [28]

Molecular Electro-and Photo-catalysis
Photocatalysts require structures with high molar absorption and ar eactive catalytic center. [29] Thed ifficulty of combining these two properties in one discrete molecule led

Angewandte
Chemie to the development of architectures that covalently or supramolecularly bind ad ye to an electrocatalyst, [30] where the photoexcited PS triggers intramolecular ET to the catalyst site.Such dyads often rely on aprecious metal-based PS,but organic chromophores are emerging.
Ap hotocatalyst composed of ac obalt diimine-dioxime HEC linked to the carboxylate anchor-bearing PS 11 (Figure 6) was used as adyad in aDSPC (PS-cat.,F igure 5b). [30d] TheP S sp ush-pull design locates the lowest unoccupied molecular orbital close to the HEC and thereby promotes its intramolecular reduction upon irradiation. As aresult, aNiO j 11 DSPC displayed an early photocurrent onset potential of + 0.61 Vv s. SHE, and chronoamperometry at À0.18 Vv s. SHE in pH 5.5 electrolyte solution under simulated solar irradiation resulted in aF E H 2 of 8-10 %.
Another P2H (13, Figure 7) is capable of photo-catalyzing the production of H 2 O 2 from O 2 when immobilized on NiO. [33] H 2 O 2 has potential as al iquid fuel with an energy density comparable to that of compressed H 2 .B ulk electrolysis of NiO j 13 at %+ 0.20 Vv s. SHE (pH 6) under 623 nm LED irradiation for 24 hr esulted in near-unity FE and aT ON > 12,000. Theproduction of H 2 O 2 was attributed to the lightdriven reduction of O 2 by 13 into the superoxide radical anion O 2 C À ,which further disproportionates into H 2 O 2 .
Flavins are organic electro-and photo-catalysts studied toward chemical oxidation chemistry that commonly use O 2 as the final electron acceptor in order to produce H 2 O 2 . [34] Similar to quinone,f lavin-derivatives exist in three redox states:n eutral (flavin-quinone FMN,F igure 7), 1e À -reduced, and 2e À -reduced (flavin-hydroquinone, FMNH 2 ). Upon irradiation with blue light, the FMN chromophores form asinglet excited state,apotent oxidant with E(FMN*/FMNC À ) =+ 1.53 Vv s. SHE. Rapid intersystem crossing (7.8 ns in water) also produces atriplet state that triggers ET. FMNH 2 is often ultimately oxidized to FMN in presence of O 2 .A lternative mechanisms could also be occurring such as FMN* converting 3 O 2 into ar eactive 1 O 2 species. [34c]

Polymeric Systems
Whilst Nafion remains the most commonly used polymer, alternatives are quickly being developed for solar fuel applications.E lectropolymerization was employed early to surface anchor molecular catalysts via pyrrole,v inyl and Figure 7. Metal-free compounds for electro-and photo-driven fuel production. Figure 6. Photocatalysts for H 2 production based on transition metalfree PSs. Figure 8. Proposed H 2 evolution mechanism for 12. [32] Angewandte Chemie methacrylate groups for applications in water oxidation, proton and CO 2 reduction. [35] Recent reports have considerably expanded the scope of bespoke polymers toward solar fuel production with applications ranging from scaffolds and catalysts to PSs.

Scaffolds for Electrocatalysts
Thea ctive sites of enzymes are embedded into polypeptide scaffolds.S ynthetic polymeric matrices can similarly integrate HECs or CRCs to provide better stability,a nd functionalities to stabilize catalytic intermediates or allow for surface anchoring. [36] An amine-containing metallopolymer 14 (Figure 9) derived from 2-(dimethylamino)ethylmethacrylate and aF e 2 S 2type HEC operates in pH neutral aqueous solution with high current densities,aTONo f2 .6 10 4 ,a no perational lifetime of 6days and even retained activity under aerobic conditions. Theh igh performance is possibly due to the protonation of amine side-chains that facilitates proton transport to the Fecatalyst while shielding it from O 2 reduction products. [36d] Ac obaloxime HEC was integrated into ac ross-linked copolymer via ap yridine ligand that also contained pyrene and ethylene glycol groups (15,F igure 9). Interfacing this structure with carbon nanotubes resulted in as tandalone electrode achieving higher TONsa nd stability compared to an electrode with the immobilized monomeric cobaloxime HEC.T he improved performance was attributed to entrapment of the otherwise labile Co HEC in the polymeric matrix and improved proton transport from the ethylene glycol moieties.T his work highlights the potential benefits by considering the choice of co-monomers,i ndependent of the HEC unit itself. [36a] Asimilar copolymeric approach was applied toward COselective CO 2 reduction using aC ob is(terpyridine) CRC. Tw oc oordination copolymers were prepared comprising the CRC,phosphonic acid anchoring groups,and either amethyl or decyl moiety to tune hydrophobicity surrounding the catalyst core in 16 and 17,r espectively (Figure 9a nd 10 a). After integration of the copolymers into bespoke inverse opal TiO 2 electrodes (pore Ø = 750 nm, Figure 10 b), electrolysis demonstrated higher selectivity toward CO vs.H 2 for the more hydrophobic,d ecyl-based copolymer,h ighlighting the possibilities for improved product selectivity offered by tuning the catalystso uter sphere environment. [36e] Poly-4-vinylpyridine coordinated to aC op hthalocyanine (CoPc) in sub-stoichiometric conditions resulted in polymer 18 ( Figure 9) that was interfaced with graphite electrodes.A better electrolysis performance for CO 2 reduction was achieved compared to the corresponding pyridine-coordinated, molecular CoPc,w hich was attributed to outer sphere effects of uncoordinated pyridines. [36b,g] Aseries of polymeric frameworks with positively charged ammonium salts,p henyl, or negatively charged trifluoroborate groups was designed to alter the catalytic activity of ac ovalently bound Re CRC. [36c] Electrochemical studies in organic solvent showed that the quaternary ammoniumcontaining polymers 19 ( Figure 9) have as ignificantly lower E cat toward CO evolution ( % 300 mV) compared to the free, molecular catalyst. In contrast, the trifluoroborate polymers displayed an egative shift in potential and catalytic activity was not observed. This illustrates how ac harged polymeric framework can influence the catalytically active species without changing the primary coordination environment around the reactive center.
Using as emi-biological approach, [37] am ultifunctional polymer was employed as ahydrogel to provide stabilization and entrapment to an O 2 -sensitive [NiFe]-H 2 ase toward biofuel cell applications. [37a,d] Thep olyamine-based polymer 20 (Figure 9) was synthesized from ab ranched poly(iminoethylene) functionalized with electroactive methyl viologen (MV) units. [37a] TheMVunits in 20 act as an electron relay for enzymatic catalysis and reduce O 2 to protect the H 2 ase.Afew micrometer-thick polymer film is sufficient to protect the

Polymeric Dyes and Photocatalysts
Light harvesting and charge conducting SCs are ubiquitous in lightweight optoelectronics applications and typically produced from conjugated polymers (CPs). CPs can be produced under mild conditions with molecularly tunable optoelectronic and physicochemical properties. [38] Recently, they have also emerged as promising materials for photocatalytic fuel production delivering high catalytic activities, often when combined with added or residual Pd/Pt nanoparticles. [3d] In some cases,t horough metal removal and purification has been shown to eliminate activity,d emonstrating that residual Pd-even at ppm-level (e.g.from crosscoupling reactions)-plays as ignificant role in the H 2 evolution abilities of some polymers. [39] CPs have evolved from linear polymers to non-crystalline microporous polymer networks,c arbon nitrides and carbon dots,p olymer dots (Pdots) and covalent organic frameworks (COFs) (Figure 11). [40] Carbon nitride derivatives (and other triazine-based systems) and carbon dots (along with other carbon nanoparticles) are attracting much attention, [41] and they can be fabricated from purely organic precursors (e.g., melamine,u rea, cyanamide,c itric acid and aspartic acid) via pyrolysis and solvothermal procedures at relatively high temperatures. [3b,41a,c, 42] Modulation of their photocatalytic activity generally involves solid-state approaches to introduce morphological alteration, doping and composite construction. [3b,41b] Given that the synthetic procedures,t uning approaches and properties of these materials differ significantly from the molecularly-defined polymeric materials described in this section, we will not examine them further. Reviews on carbon nitrides and carbon dots can be found elsewhere. [3b,42a] Hereafter, we focus on polymer engineering strategies to enhance activities,i ncluding improved light-harvesting properties,porosity and crystallinity.
CPs offer the possibility of fine-tuning SC properties such as the energy levels and resulting optical band gap (E g ), through the selection of monomeric building blocks and via modular polymerization strategies.A ne arly report showed that poly(p-phenylene) (PPP,F igure 12 a) acts as ap hotocatalyst for H 2 evolution, despite showing low activity and requiring UV irradiation. TheE g could be reduced from 2.9 to 2.7 eV (for PPP and 21,respectively) with the introduction of ad ibenzo[b,d]thiophene sulfone moiety and then further to 2.1 eV upon introduction of planarizing ethynyl groups in 22 (Figure 12 a). [40a, 44] Thesmaller E g combined with accelerated charge separation allowed for an increase in the photocatalytic rate from 1492 to 6023 mmol H 2 h À1 g À1 in the presence of aSED.
This strategy was also employed in 3D microporous CPs, where high surface area polymers were prepared by adjusting the ratio of four monomers (Figure 12 b). [45] Thecorresponding E g thereby gradually decreased from 2.95 to 1.94 eV following an increase in pyrene content. [46] In the absence of an externally added metal catalyst, the polymers showed ag radually enhanced photoactivity for H 2 evolution with an optimal E g of 2.33 eV for 23.Polymer 23,obtained with aratio of 2:1f or 1,4-benzene diboronic acid:1,3,6,8-tetrabromopyrene,e xhibited the highest activity of 174 mmol H 2 h À1 g À1 (and am id-range surface area compared to polymers of other ratios). Further reduction in E g of the polymers led to alower rate,w hich was ascribed to increased nonradiative deactivation in the pyrene-rich polymers.
Thehigh hydrophobicity of CPs commonly results in their aggregation in aqueous solution into ab ulk material composed of micrometer-size particles with low surface area (i.e., water-polymer interface) and extended travelling distance for charge carriers.T oo vercome these issues,p orous systems as well as optimized precipitation and gel-promoting methods have been developed.
Thelength of the p-extended linker affects the porosity of microporous networks synthesized via condensation, as shown in polymers 24-27 ( Figure 13). [47] Polymer 24 with

Angewandte
Chemie as hort para-phenylene spacer exhibits micropores,w hereas 25 and 27 contain longer spacers and subsequently integrate micro-and mesopores.P olymer 26 shows am uch broader pore size distribution due to the longest polyphenylene spacer in the polymer network. As ar esult, the Brunauer-Emmett-Te ller (BET) surface areas were 669, 750, 564 and 834 m 2 g À1 for 24, 25, 26 and 27,respectively.Despite their similar E g ,the corresponding photoactivities were 134, 598, 908 and 620 mmol H 2 h À1 g À1 ,r espectively,w ith the highest rate for 26 and ascribed to its nanoparticular morphology,b etter wettability and large surface area. This highlights the possibility to influence the activity of polymers via fine-tuning of morphological variations.
An alternative approach was developed with the charged amphiphilic PMI dye 28 (Figure 14), which can self-assemble into ribbon-type supramolecular polymers via hydrophobic collapse. [43a] Moreover,atsufficiently high concentrations,the charged supramolecular polymers produce highly hydrated 3D network hydrogels,w hich display ah igh degree of crystallinity.T his leads to the PMI losing its individual excitonic character and behaving as an ensemble with photoinduced excitons spreading out over multiple PMI units within the crystalline ribbons.The 28-based hydrogels formed in presence of poly(diallyldimethylammonium) chloride can host aw ater-soluble Dubois Ni HEC,u ltimately producing aTON H 2 of % 340 under irradiation in the presence of AA as SED.
Pdots represent another family of organic SC assemblies used in photocatalysis with diameters from 1to100 nm. [43b,48] Thesmaller size of Pdots compared to bulk materials reduces the distance for photogenerated charges to migrate to the surface,w hich decreases the recombination probability.A Pdot suspension can be generated by the nano-precipitation method using CPs and awater-soluble polymer.F or example, the synthetic polymer 29 and the matrix PS-PEG-CO 2 H ( Figure 15) were solubilized in THF and injected into pure water under sonication to produce asuspension after solvent evaporation. [49] Absorption up to 660 nm and aD -A architecture in 29 allowed for an excellent H 2 evolution rate of up to 50 mmol H 2 h À1 g À1 under LED irradiation in presence of AA as SED.A lthough 29 contains traces of Pd (0.1 wt %), theoretical calculations suggested that nitrogen atoms in the benzothiadiazole units may provide the reactive sites for the formation of H 2 .
Thel inear homopolymer of dibenzo[b,d]thiophene sulfone 30 produced emulsion particles when synthesized from mini-emulsions of toluene droplets in water with waterstabilizing sodium n-dodecyl sulfate (SDS,F igure 15). The latter exhibited ah igh surface area of 16 m 2 g À1 ,c ontained 0.4 wt %P d, and produced H 2 with an excellent rate of 61 mmol H 2 h À1 g À1 under visible light irradiation in the presence of TEA. [50] Inspired by bulk heterojunction-based solar cells and emerging examples of photoelectrodes, [51] as imilar approach was extended to the preparation of heterojunction nanoparticles using ab lend of ad onor polymer 31 and an onfullerene acceptor 32,i nt he presence of sodium 2-(3thienyl)ethyloxybutylsulfonate (TEBS) as as tabilizing agent ( Figure 15). [52] TheT EBS affinity of its exposed aromatic units of 32 is believed to control the nanomorphology of these particles into an intermixed D/A blend. SDS,o nt he other hand, promotes an inefficient core-shell morphology.A30:70 blend content of 31:32 is optimal for efficient exciton dissociation and formation of % 82 nm particles.F ollowing platinization, the photocatalyst displayed as ignificant H 2 evolution rate of % 64 mmol H 2 h À1 g À1 under broadband visible light illumination, and an external QY exceeding 5% from 660 to 700 nm.
Crystallinity can improve the efficiency in conjugated systems as it favors charge transport and separation. COFs are an emerging class of 2D/3D polymers and an example of highly crystalline organic building units combined into extended covalent structures. [53] Thew ell-defined pores, excellent stability and fine-tuned physicochemical properties of COFs make them appealing candidates as PSs and catalysts for fuel production. [54] Ad ibenzo[b,d]thiophene sulfone moiety can be integrated into aC OF (33,F igure 16), which leads to ah igh photocatalytic performance of 10 mmol H 2 h À1 g À1 when used in presence of Pt and AA as SED. [43c] TheCOF allows broad visible light absorption (E g =    [47] Angewandte Chemie 1.85 eV) and relatively long excited state lifetimes (t avg = 5.56 ns) in aqueous suspensions.T he high efficiencyw as ascribed to its good wettability and al arge BET area of 1288 m 2 g À1 from its 28 pore size diameter.
Ab ipyridine-containing COF was recently post-synthetically modified with aRecomplex to afford the photocatalyst 34 ( Figure 17). [55] Thelatter delivers aCOproduction rate of 1040 mmol CO g À1 h À1 with 81 %s electivity over H 2 ,a cross 17.5 hofillumination (TON CO % 19) in acetonitrile containing TEOA. Computational results support that ET occurs from the light-absorbing COF backbone to the Re CRC upon photoexcitation. Crystallinity and porosity were key factors in the activity of such materials,a sa na morphous,l ow porosity analogue showed almost no catalytic activity.

Conclusions
Thed evelopment of synthetic organic architectures for fuel catalysis has experienced rapid progress during the past decade and we have summarized the wealth of approaches that originated from integrating organic designs.T ailor-made organic structures have followed several approaches,ranging from small molecules to polymers acting as light harvesters, (photo)catalysts and environment modifiers.O verall, integrations of such organic architectures with inorganic and biological components resulted in innovative hybrids such as DSP,D SPC and COF systems.H igh performances have already been reached, especially toward photocatalytic H 2 evolution (> 60 mmol h À1 g À1 ), making these systems competitive with inorganic counterparts.
Thepossibility of molecular engineering has played avital role in achieving these recent developments,w ith rational design enabling the integration of anchoring abilities,i mproved reactivity,i ntense and wide light absorption, high surface area, efficient charge separation and transport, and so forth. Synergistic combinations of such properties can in principle result in further enhanced catalytic activity.N evertheless,t he structure optimization of one parameter often collaterally impacts other properties,a nd deconvolution of individual effects remains challenging.
Despite these significant advances,m any opportunities for further exploration persist. Ab etter understanding of photocatalytic fuel mechanisms for polymers,COFs and other carbon nanoparticle-based materials is desirable to reach better designs toward higher performance.T here is scope for metal-free electrocatalysts and the assembly of CO 2 -reducing and full water-splitting systems.Aparticular opportunity lies ahead in the exploration of redox transformations beyond classical solar fuels applications such as organic electro-and photoredox catalysis.T he relatively unexplored possibilities offered by tuning the environment around the catalytic center and PS bears many promises and organic chemistry also allows for the development of nanoreactors to enable controlled catalysis in ac onfined environment. [56] More robust and red-light absorbing PSs [57] are also in demand as well as ab etter understanding of aqueous media-organic system interfaces and the development of oxidative chemistry. [58] Finally,t he integration of organic materials with biological systems, [59] and the prospect of their high throughput analysis by robotics [60] represent further exciting avenues of future research. We therefore envision many possibilities to employ organic chemistry in the future development of electro-and photocatalytic systems.

Conflict of interest
Theauthors declare no conflict of interest.