Synthetic Photoelectrochemistry

Abstract Photoredox catalysis (PRC) and synthetic organic electrochemistry (SOE) are often considered competing technologies in organic synthesis. Their fusion has been largely overlooked. We review state‐of‐the‐art synthetic organic photoelectrochemistry, grouping examples into three categories: 1) electrochemically mediated photoredox catalysis (e‐PRC), 2) decoupled photoelectrochemistry (dPEC), and 3) interfacial photoelectrochemistry (iPEC). Such synergies prove beneficial not only for synthetic “greenness” and chemical selectivity, but also in the accumulation of energy for accessing super‐oxidizing or ‐reducing single electron transfer (SET) agents. Opportunities and challenges in this emerging and exciting field are discussed.


Introduction
Chemical synthesis by visible light is the fundamental process for biological photosynthesis on Earth. However,CO 2 and H 2 O, and most organic molecules,d on ot absorb visible but ultraviolet light. Naturessolution is chlorophyll, acolored pigment that absorbs visible-light energy to drive the process. Researchers have made efforts towards artificial photosynthesis with visible light ever since Giacomo Ciamiciansvision in the turn of the 20th century (1912). [1] By mimicking the concept of nature,b ut stripping down the complexity of interlinked photosystems into defined single-molecule photocatalysts,r esearchers found that transition-metal complexes such as bipyridyl complexes of Ru II and Ir III can harvest visible-light photons to become powerful excited-state single electron transfer (SET) agents for redox processes,and enjoy sufficiently long lifetimes (700-1100 ns) [2] to undergo diffusion-controlled redox events.I nitial reports came as early as the 1980s, [3] and the field of "visible light photoredox catalysis (PRC)" erupted in the turn of the 21st century. [2,4] In this context, seminal papers demonstrated the synthetic applications of Ru II and Ir III bipyridyl complexes. [5][6][7] With sustainability and cost at the forefront of minds in academia and chemical industry, [8] researchers were quick to challenge the presence of rare mid-row transition metals with examples of organophotocatalysts such as eosin Y, Rose Bengal, and acridinium salts,a sn oted in seminal papers and reviews. [4d-g, 9-11] Recently,t he use of more sustainable transition-metal-based coordination compounds such as those of iron, nickel, and copper, whose excited-state lifetimes are much shorter (rendering their application more challenging), are starting to receive attention. [12] PRC is attractive for avariety of reasons reviewed elsewhere, [4,13] but arguably the biggest advantage is that use of visible light precludes direct excitation of substrates (leading to difficult-to-control highenergy pathways and decomposition), selectively transferring energy to the photocatalyst chromophore.
Another vehicle for SET chemistry,w hich has been undergoing arenaissance in recent years,issynthetic organic electrochemistry (SOE). Theapplication of electrical current in organic synthesis dates back as far as the Faraday and Kolbe electrolysis reactions from the 1830s to 1840s; [14] far earlier than Ciamiciansvision for artificial photosynthesis.A number of efforts [15][16][17][18][19][20] in the last two decades have brought SOE to the fore in organic chemistry. [21] SOE is advantageous for several reasons that are well-documented, [21] but arguably the biggest advantage of SOE is the ability to dial in any potential, and the redox window is in theory only limited by the tolerance of the reaction solvent.

Visible-Light Photoredox Catalysis:The Limits
Af undamental problem in visible-light PRC is that the energy of processes is constrained by the energy of visiblelight photons (400-700 nm;ca. 1.8-3.1 eV). Inevitably,not all of this energy is accessible to the photocatalyst;l osses occur due to intersystem crossing/non-radiative pathways,w hich can account for up to approximately 0.6 eV in the case of Ru II complexes. [2a] Ultimately,t he energy available to ap hotocatalyst from excitation by as ingle visible-light photon is typically insufficient for challenging chemical transformations such as the conversion of CO 2 and H 2 Oi nto glucose and water, [22] or the direct SET activation of many moieties of interest to organic chemists.F or example,S ET oxidations of hydrocarbon CÀHb onds, [23] electron-neutral/poora romatic p-systems, [24] carbonyl groups, [25] and ethers [25] require potentials of + 2.4-3.5 Vv s. SCE, while reductions of aromatic p-systems, [26] aryl chlorides, [27] and silyl halides [28] require potentials of À2.6-3.4 Vv s. SCE. In order to engage challenging moieties,visible-light PRC has thus far relied on tricks that circumvent direct SET activation. ForC ÀHo r carbonyl activations,t hese can include 1) in situ generated radical or radical ions that undergo hydrogen atom transfer Photoredox catalysis (PRC) and synthetic organic electrochemistry (SOE) are often considered competing technologies in organic synthesis.Their fusion has been largely overlooked. We review stateof-the-art synthetic organic photoelectrochemistry,grouping examples into three categories:1)electrochemically mediated photoredox catalysis (e-PRC), 2) decoupled photoelectrochemistry (dPEC), and 3) interfacial photoelectrochemistry (iPEC). Suchsynergiesp rove beneficial not only for synthetic "greenness" and chemical selectivity, but also in the accumulation of energy for accessing super-oxidizing or -reducing single electron transfer (SET) agents.O pportunities and challenges in this emerging and exciting field are discussed. (HAT) chemistry, [29] 2) excited states that directly engage in HATc hemistry, [30] or 3) proton-coupled electron transfer (PCET). [31] Naturess olution to the "energy problem" is to accumulate the energies of multiple photons. [22] Mimicry of such at echnique has proven elusive to researchers until recent years.T he concept of consecutive photoelectron transfer (conPET) was disclosed by Kçnig and co-workers using ap erylene diimide or RhB as the organophotocatalyst, to cleave C À Xb onds that could not be cleaved by as ingle quantum of visible-light energy. [11d,e] Following absorption of one quantum of visible-light energy and then reduction by asacrificial SET donor (e.g., Et 3 N), the formed radical anion absorbs the second quantum of visible-light energy.U ltimately,asuper-electron donor is formed in situ by accumulation of visible photons.Although the subsequent chemistry may be redox-neutral, the requirement for as acrificial electron donor to ensure asufficient concentration of photoexcitable radical anion is undesirable.T his strategy may not be so general because it requires design of photocatalyst architectures that absorb visible light both in their ground state and in their radical ion state.

Synthetic Organic Electrochemistry:The Limits
Afundamental problem in SOE is that the conductivity of organic solvents is typically low (compared to aqueous systems). Ah igh "ohmic drop" exists between the two separated electrodes,n ecessitating high cell potentials for useful reaction conversions.S uch potentials may be high enough to encourage unselective,deleterious redox processes when applied to the organic substrate of interest. Thec ell potential is the sum of electrode potential and ohmic drop.By employing ah igh concentration of supporting electrolyte (such as n-Bu 4 NPF 6 ), the solution conductivity can be increased and the ohmic drop decreased; [32] however, the amphiphilic electrolyte is generally (not always [32b] )d ifficult to separate from the desired product(s) after the reaction. A different strategy that allows reactions to proceed at milder electrode potentials is "mediated" electrolysis or "redox catalysis". [21,33] Here,am ediator transports holes [17] or electrons [27,34] to/from the electrode surface from/to the substrate. However,t he redox power of mediators is limited to the redox potential of their radical ion or their ion forms.

Photoelectrochemical Organic Synthesis
Visible-light PRC and SOE have enjoyed adramatic rise in popularity in the last decade,p artly because of the drive towards green chemistry and sustainability but fundamentally because of their use as SET methods for straightforward access to organic free radicals that can be used in synthesis.In terms of their ability to perform redox chemistry,P RC and SOE are often thought of as competing technologies,a nd their fusion has thus far been largely overlooked ( Figure 1). This Review explores synthetic photoelectrochemistry as the next evolutionary stage of PRC and SOE. State-of-the-art examples are presented. Forthe purposes of this Review,we separate the examples into i) electrochemically mediated photoredox catalysis (e-PRC), where the electrochemical and photochemical components have interdependent roles providing an explicit benefit within the chemical process; ii)decoupled photoelectrochemistry (dPEC), where electrochemical and photochemical components have separate, discrete roles;a nd iii)interfacial photoelectrochemistry (iPEC), where reactions occur at photoelectrode surfaces. In this Review,wefocus only on the use of organic substrates and exclude the photoelectrolytic splitting of water and solar fuel production. Forour justification of nomenclature and for recommendations to users of this technology,see Section 3.4.

Photoexcitation of Electrochemically Generated Ions
One fundamental, exciting branch of e-PRC is the photoexcitation of electrochemically generated ions. [35][36][37][38][39] Here, ab ase redox energy level is provided by electrochemistry (e.g.,aradical anion). Then, redox energy is provided from photoexcitation to generate super-redox agents in atransient fashion ( Figure 2). As the mediator is regenerated and accumulates both electrons and photons to overcome the activation energy barrier,t he term "electromediated photoredox catalyst" (e-PRC) can be coined. Considering the molecular orbital transitions of 9,10-dicyanoanthracene (DCA) as an example of ar ecently reported [38] reducing e-PRC,t he LUMO (y 2 )o fD CA is first populated with an electron by cathodic current, thus becoming SOMO-2 (y 2 )of DCAC À .P hotoexcitation promotes an electron from the MO-1 (y 1 )t o the SOMO-2 (y 2 ), thus effecting SOMO-HOMO inversion. [38] This also occurs in the complementary scenario with PTZ as an oxidizing e-PRC; [35a] the removal of an electron by anodic current turns HOMO-4 into SOMO-4. An electron is then promoted from MO-1t oS OMO-4 by 514 nm light. [40] In both cases,the e-PRC becomes adoublet excited state.
Thec ombination of photochemistry with electrochemistry within the context of organic synthesis was first disclosed by Moutet and Reverdy, [35] who photoexcited electrochemically generated radical ions.V isible-light photoexcitation (> 400 nm) of the phenothiazine (PTZ) radical cation, generated electrochemically at controlled potential (E 1/2 (PTZ) =+0.79 Vv s. SCE), [35c] in the presence of 1,1diphenylethylene (DPE, E p/2 ox =+1.57 Vvs. SCE), [20j] leads to oxidation of DPE and regeneration of phenothiazine ( Afew years later, Moutet and Reverdy reported that the electrogeneration of N,N,N',N'-tetraphenyl-p-phenylenediamine (TPPD) radical cations and their photoexcitation with UV light (366 nm) enabled oxidation of benzyl alcohol (3)to benzaldehyde (4;F igure 3B). [35b] Interestingly,t he oxidation of substituted benzyl alcohols,1 -phenylethanol, or benzhydrol did not lead to the corresponding ketones,b ut rather to the symmetrical ethers.H ere,e -PRC is tentatively written because not enough details (yields,conversion) were reported to determine whether PTZ is catalytic in the first example. Whilst the second example could be considered to be the first report of e-PRC,the process took place with only about three turnovers and the yield was not reported.
Following these [35] and other early reports (generally investigated in an analytical/fundamental context), [36,37] photoelectrochemistry in organic synthesis did not receive attention until very recently.T his naturally follows on from the resurgence of SET chemistry in organic synthesis thanks to PRC and SOE, which have been popularized in the last decade.  In terms of SET oxidation, among the most powerful photoredox catalysts are the acridinium salts (Mes-Acr + ) developed by Fukuzumi and co-workers. [41] Seminal papers by the group of Nicewicz employed these organophotocatalysts in the oxidation of alkenes to radical cations,which could be intercepted by nucleophiles in an anti-Markovnikov-type reaction. [10a,b] Moreover,d irect oxidation of arenes was achieved, and their nucleophilic trapping with heterocyclic nucleophiles gave rise to ap alladium-free Buchwald-Hartwig-type reaction. [10d] However,t he former reaction was limited to styrenes or highly electron-rich (trisubstituted) alkenes with atethered nucleophile.The latter was limited to electron-rich arenes (anisoles) because the redox potentials of mono-/disubstituted alkenes (E p ox =+2.37 Vv s. SCE) [25] and of benzene (+ 2.48 Vv s. SCE) [39] lie beyond the redox potential of the acridinium excited state (+ 2.06 Vvs. SCE). [4f] One way that researchers overcame this limitation was by employing DDQ,which forms avery powerful excited triplet state (+ 3.18 Vv s. SCE) that can engage unactivated or electron-deficient arenes. [42,43] Photocatalytically generated arene radical cations can be intercepted by nucleophiles such as 5 to give aminated arenes such as 6 as demonstrated by the groups of Kçnig ( Figure 4) [43,44] and others. [45] However,DDQ is moderately expensive and is prone to promiscuous reactivity with arene substrates or amine nucleophiles (Figure 3D), as well as other functional groups, [46] through ground-state oxidation chemistry.
Photoexcitation of electrochemically generated cations allows for redox potentials that are notably more positive than those achieved even with *Mes-Acr + and avoids the complications associated with PRC using DDQ.Inanelegant and seminal e-PRC example,L ambert and co-workers reported the oxidation of unactivated arenes and their coupling with heterocyclic amines ( Figure 5). [39] Under anodic oxidation at af ixed potential (+ 1.50 Vv s. SCE), colourless trisaminocyclopropenium cation (TAC + )w as oxidized to its dication radical (TACC 2+ ; E 1/2 = 1.26 Vv s. SCE), which is strongly coloured. Excitation of TACC 2+ with visible light (ca. 600 nm) provided the superoxidant *TACC 2+ (E 1/2 = + 3.33 Vvs. SCE), which oxidized unactivated arenes to their radical cations.T he remarkable potential of *TACC 2+ was rationalized by time-dependent density functional theory (TD-DFT) calculations,w hich revealed aS OMO-HOMO level inversion leaving alow-lying hole in the HOMO.E thyl 1H-pyrazole-4-carboxylate (13)undergoes nucleophilic addition to the benzene radical cation, generating (upon loss of aproton) an aryl radical. Oxidation of the aryl radical, either by TACC 2+ or by the carbon (felt) anode,f ollowed by loss of ap roton, furnishes product 14.P roton reduction was proposed as the corresponding cathodic half-reaction, as gas bubbles were observed. Control reactions confirmed that no reaction occurred without light, current, or TAC. For comparison, direct electrolysis was performed at fixed  potential (+ 3.0 Vv s. SCE) and gave polymeric material, exemplifying the advantage of the mild conditions of e-PRC. Thereaction tolerated benzene and even chloroarenes to give products 15 and 16,a lbeit in modest yield. Substituted triazoles,b enzotriazoles,a nd purines were successful partners,a ffording products such as 17 and 18.N oo xidation of aldehyde-, ketone-, or ester-bearing pyrazoles was observed. Thee xpansion of scope to unactivated or electron-deficient arenes represents ak ey advantage over Nicewiczso riginal report. [10d] In acomplementary fashion, cathodic current can be used to generate radical anions photoexcited to generate superreductants.L ambert, Lin, and co-workers reported the reduction of chloro-and bromoarenes such as 19 using photoexcited 9,10-dicyanoanthracene radical anion (*DCAC À ), [38] itself generated by cathodic reduction of DCA by aporous carbon anode ( Figure 6). Theextraordinarily high reduction potential of À3.2 Vv s. SCE was proposed to arise from aSOMO-HOMO level inversion and ahighly unstable filled antibonding orbital, as confirmed by TD-DFT calculations. [38] Theg enerated aryl halide radical anions fragment to afford halide anions and aryl radicals,t he latter of which were successfully trapped with B 2 pin 2 ,S n 2 Me 6 ,o rh eteroarenes to give products such as 21-24.O xidation of sacrificial Zn anode was proposed as the corresponding half-reaction. Them ethod provides ak ey advantage over palladiumcatalysed functionalizations used to generate similar products, which suffer when coupling partners contain Lewis basic groups (such as the precursor to 23)asthey alter the course of catalysis by coordination.

Replacing Sacrificial Redox Agents with Current
Although the former subsection likely presented am ore fundamental and potentially ground-breaking advantage of e-PRC in organic synthesis,r eplacement of sacrificial redox agents is another very important aspect offered by e-PRC that appeals to asustainability and industry perspective (Figure 7). Xu and co-workers reported on the C À Ha lkylation of heteroarenes with trifluoroborates under e-PRC ( Figure 8). [47] Photoexcited 9-mesityl-10-methylacridinium (*Mes-Acr + )i sapotent oxidant (E p red =+2.06 Vv s. SCE) capable of SET oxidation of isopropyl trifluoroborate (26; E p ox %+ 1.50 Vvs. SCE) [48] to its secondary alkyl radical. The alkyl radical adds to the protonated quinoline 25-H + in aM inisci-type manner,w hich, followed by loss of ap roton and SET oxidation (either by ground-state Mes-Acr + (E p/2 red = À0.57 Vv s. SCE) or by the anode) affords product 27.
Mes-Acr + is regenerated by anodic oxidation of Mes-AcrC at ar eticulated vitreous carbon (RVC) anode.Awide substrate scope of heteroarenes were employed, including isoquinolines,p henanthridines,p hthalazines,b enzothiazoles, acridines,and purines,affording products such as 28-31.T he reaction conditions tolerated secondary and tertiary amines as well as secondary alcohols and alkynes,which would all be prone to oxidation under direct electrolysis at high potentials.
Lambert and Huang reported S N Ar reactions of unactivated aryl fluorides under e-PRC (Figure 9). [49] Here,p hotoexcited 2,3-dichloro-5,6-dicyanoquinone (DDQ) was sufficiently oxidizing (E p red =+3.18 Vvs. SCE) to engage chlorofluoroarenes such as 32 in SET oxidation. In terms of the heteroarene partner,the substrate scope was similar to that of  the previous report involving the photoexcited dication *TACC 2+ . [39] Heteroarenes bearing aldehydes and esters were tolerated, affording products such as 35 and 36.Alcohols such as ethanol and acetal-protected galactose,aswell as tert-butyl carbamate,were also well-tolerated as nucleophiles (products 37 and 38). Redox potentials for the oxidation of polyhalogenated benzenes are unavailable in the literature,l ikely because they exceed the redox potential window of the solvent. It is interesting that although *TACC 2+ (E p red = + 3.33 Vv s. SCE) is am ore potent oxidant than *DDQ,i t afforded al ower yield of 34.T his suggests that matching of redox potentials is not always areliable predictor of successful SET chemistry and that other factors such as precomplexation of mediator and substrate (Section 3.3), might be important. Elsewhere,o xidation of unactivated alcohols was recently achieved under e-PRC with riboflavin tetraacatate as the photocatalyst and thiourea as aHAT co-catalyst. [50] Here, the role of anodic current was to regenerate riboflavin from its dihydroquinone form.

Decoupled Photoelectrochemistry (dPEC)
Sheffold and Orlinski reported ap hotoelectrochemical 1,4-addition of acyl groups to a,b-unsaturated carbonyl compounds ( Figure 10). [51] Cathodic current reduced vitamin B 12a (Co III )o raC o II macrocyclic complex 42 to give Co I complex 43,which reacted with anhydride 39.Photochemical cleavage of the Co III ÀCb ond of 44 presumably afforded an acyl radical 45,p rimed for 1,4-addition to 40 to give 46.T he authors claimed that HATf rom the solvent to 46 yielded product 41.SETreductions of 45 (to give an acyl anion primed for 1,4-addition) or 46,f ollowed by proton transfer from the solvent, could not be ruled out. Thea uthors did not specify the anodic half-reaction or the anode materials.Here,photochemistry and electrochemistry were handled as discrete processes,representing the first example of decoupled photoelectrochemistry (dPEC).
In addition to the HLF reactions of N-alkyl sulfonamides to afford pyrrolidines,2,2,2-trichloroacetimidates (and benzimidates) were employed to afford oxazolines (products 51 and 52). Va rious heterocycle-bearing substrates were tolerated despite the anodic potential and in situ generated molecular iodine.A cid hydrolysis of the oxazolines gave rise to pharmaceutically valuable (protected) 1,2-amino alcohols (product 53). This work follows on from electrochemical HLF reactions reported by MuÇiz, [53] yet exhibits ak ey advantage in its use of low anodic potentials for the oxidation of iodide to molecular iodine.S uch potentials are less positive than the redox potentials of electron-rich arenes and other functional groups,a nd thus the mild conditions allow for excellent redox chemoselectivity.S tahl demonstrated [52] that previously reported electron transfer/proton transfer/electron transfer (ET-PT-ET), [53] proton-coupled electron transfer (PCET), [54] and bromide-mediated electrochemical HLF reactions [55] all failed to convert 47 into product 48,instead yielding ac omplex mixture of products.

Interfacial Photoelectrochemistry (iPEC)
In interfacial photoelectrochemistry (iPEC), also known as "photoelectrocatalysis", ap hotoelectrode is coated in ap hotoresponsive (typically,s emiconductor) material whose band gap corresponds to the energy of avisible-light photon. An applied or "bias" potential (E AP )i su sed to improve charge carrier separation upon irradiation (preventing recombination and generation of heat). Forp hotoanodes,a pplied potential followed by irradiation promotes an electron from the valence band to the conductive band, generating ah ole that is used for oxidation chemistry ( Figure 12). [56] Hu, Grätzel, and co-workers recently reported the use of aphotoelectrochemical cell in organic synthesis as an example of iPEC. [57] After setting the photoelectrochemical cell at afixed applied potential (+ 1.13 Vv s. SCE), ah ematite (a-Fe 2 O 3 ) photoanode was irradiated with blue LEDs and was rendered highly oxidizing (valence band =+2.30 Vv s. SCE). Anisole was oxidized to its radical cation, primed to nucleophilic attack by arange of aromatic heterocycles (such as 54)inan overall CÀHa mination of electron-rich arenes to furnish products such as 55-59 ( Figure 13). In the absence of light, higher applied potentials (+ 1.93 Vvs. SCE) were required to access the desired chemistry and in decreased yield. Direct electrolysis (in the dark) with ac onductive glassy carbon electrode (+ 1.73 Vv s. SCE) gave poorer yields and side products that were absent when the photoelectrochemical cell and light were used.
Thefact that the arene scope was limited to electron-rich arenes,m irroring the original amination report of Nicewicz and co-workers, [10d] is unsurprising considering that the hematite band gap (2.3 Vv s. SCE) is similar to the redox potential of *Mes-Acr + (E p red =+2.06 Vv s. SCE). The markedly different ortho/para (o/p)s electivity between the   two reports [10d, 57] was attributed by the authors to the hexafluoroisopropanol solvent. Thea uthors proposed that HFIP creates ah ydrogen-bonding network that favours substitution at the ortho position, and that the fundamental photoelectrochemical process proceeds through the same intermediates as the photochemical example. [10d] One possibility not yet considered is that precomplexation of acridinium photocatalyst and anisole (Section 3.3), or precomplexation of the anisole with the photoanode,m ay encourage stereoelectronic effects that bias the selectivity.
Several reports of oxidation of simple organic molecules by iPEC exist, for example,a lcohol oxidations. [58] However, such reactions generally occur in aqueous solvent systems, and certain photoanode materials are known to undergo photocorrosion in aqueous solvent systems. [59] Sammis,B erlinguette and co-workers reported oxidations of tetralin (60), benzyl alcohol (3), and cyclohexene (62)i nM eCN under iPEC using aBiVO 4 photoanode and a100 WXelamp fitted with an AM1.5G filter as simulated sunlight, to give products 61, 4,a nd 63,r espectively ( Figure 14). [60] N-Hydroxysuccinimide (NHS) was employed as as oluble,t ransparent holetransfer mediator [15,61] between the photoanode and the substrates.F or oxidations of 60 and 62,i tw as necessary to employ t-BuOOH as the external oxygen source.T he same oxidations could be achieved under electrochemical potential only (E cell =+1.8 Vvs. Ag/AgCl) with aglassy carbon anode/ cathode,l eading the authors to assume that this potential matched the pseudo-standard potential of NHS.However,the authors noted that their iPEC method, which operates at a1.0 Vlower applied potential than the electrochemical cell, expects energy savings of 60 %. Although product yields were modest, the authors noted that the ability to perform organic synthesis at as olar-to-electricity efficiency (h = 1.3 %) close to that of traditional photoelectrochemical water oxidation (h = 1.7 %) is important because of the higher value of the organic products.
Related to this report is the iPEC CÀHo xidation of cyclohexane by aWO 3 photoanode [62] and the iPEC oxidation of benzylic alcohols by aBiVO 4 /WO 3 photoanode [63] reported by Sayama and co-workers,w hich both showed ad rastic decrease in the applied potentials required for oxidation in the presence of light. Sayama further employed the BiVO 4 / WO 3 photoanode in the iPEC oxidative dimethoxylation of furan 64 mediated by bromide ions (Figure 15). [64] In the first step,oxidation of bromide anions by the photoanode afforded apool of bromine cations.After 5C of charge had passed and furan in MeOH had been added, the dimethoxylated product 65 was obtained in very good yield.   All of these reports exemplify the advantage of iPEC in leveraging the energy of visible light to offset the high applied electrode potentials otherwise needed, thus affording better selectivity and energy efficiency in chemical redox processes. [57,60,[62][63][64][65] Thei nitial modification of the photoelectrode with applied potential, followed by energy top-up via the selective delivery of light energy,isconceptually similar to e-PRC involving photoexcitations of electrochemically generated ions.The advantage of iPEC is that it does not rely on the generation of ac hromophore in solution, and can directly engage substrates that do not absorb visible light. The disadvantage is that iPEC cannot reap the energy benefits of e-PRC,w hich can access very high redox potentials ( Figure 13 vs.F igure 5). While most examples of iPEC to date have dealt with simple chemical transformations,i PEC will undoubtedly occupy an important role in redox transformations of more complex organic substrates in the future.

Practical Execution and Experimental Rigor
Thus far, the synthetic photoelectrochemistry examples reported herein have been conducted in custom-built (transparent) electrochemical reaction vessels.T hese generally fall into two categories ( Figure 16): a) an undivided glass "pot cell"/"beaker cell"/undivided glass voltammetry setup, [35,47,53,58,61,63] or b) ad ivided glass "H-type" cell with ag lass or membrane frit. [39,49,52,63] These are all standard academic reactors used for SOE, [66] which can be easily irradiated with visible light. It is widely accepted that one of the drivers behind the renaissances of PRC and SOE in the last decade is the availability of reactor equipment. Indeed, visible-light photoredox and synthetic organic electrochemical batch reactors have now been standardized and some are commercially available, [67][68][69] addressing the long-standing plague of practical irreproducibility in both fields.T he design of suitable and standardized synthetic photoelectrochemical equipment will carry its own set of challenges,b ut fortunately,photoelectrochemical cells that have been developed for hydrogen production, such as the Cappicino PEC cell (EPFL Switzerland), [70] the PortoCell (UPorto), [71] and designs by Redoxme AB [72] could be readily adapted for synthetic applications in organic solvents.Another challenge is the need for more rigorous control experiments (in absence of either light, applied potential, or e-PRCs) to ensure that both the photochemical and electrochemical components are necessary and beneficial to the reaction.

Flow Photoelectrochemistry
Both PRC and SOE suffer upon scaling up in batch mode due to the physical constraints governing transfer of photons to the reaction or electrons to/from the reaction. The relationship between absorbance A,e xtinction coefficient e, path length l,and molar concentration c is given by the Beer-Lambert law [Eqn. (1)].R earrangement to Equation (2) shows the exponential relationship between the transmitted intensity I and the absorbance A,w hich highlights the fundamental challenge faced in the scale-up of photochemical processes.G eneral theory predicts that for at ypical 50.0 mm reaction with aphotocatalyst loading of 1mol %(0.5 mm)and e = 11280 m À1 cm À1 (452 nm absorption band of Ru(bpy) 3 Cl 2 ), 90 %o ft he light is absorbed at l = 0.2 cm from the reactor surface. [73] This tiny path length highlights the importance of the surface area to volume (SAVR) ratio in photochemical processes.
A ¼ log 10 I 0 I ¼ elc ð1Þ While SOE in macrobatch reactors has been achieved on an industrial scale,p henomena such as interelectrode ohmic drop,m ass transfer, reaction selectivity,o re nvironmental factors have presented barriers to various synthetic processes. [74] In SOE, the rate-limiting step of electrochemical reactions is generally how quickly the reagents can reach the proximity of the electrode surface (within which electron transfer can occur) by mass transfer,r ather than the kinetics of the chemical reaction. [66] Thecell current I cell (and, in turn, the cellsproductivity) is given by aderivative of the Butler-Volmer equation [Eqn.
(3)] and is related to the number of moles of reagent to be converted n,F aradaysc onstant F, electrode surface area A (cm 2 ), mass transfer coefficient k m , and reagent concentration c. Hence,the cell productivity can be increased by increasing the electrode surface area and by mixing (increasing k m by decreasing the size of the diffusion layer). Thetime-dependent fractional conversion X of amasstransfer-limited reaction of volume V is given by Equation (4). [66] This demonstrates the key importance of efficient mixing (increasing k m )a nd largest possible electrode SAVR.
Continuous flow (CF) is aglobally recognized technology within chemical industries and academia [75] that is especially useful in photochemistry [76] and electrochemistry, [77] because the flow of ar eaction mixture through small-diameter (mmmm) channels 1) allows shorter path lengths for light transmission, 2) minimizes separation of electrodes ("ohmic drop"), allowing wasteful electrolytes to be eliminated or used in smaller amounts,3 )enhances mixing or user control over mixing by laminar or turbulent flow regimes,a nd 4) increases SAVR. Indeed, CF has even enabled multigram-

Angewandte Chemie
Reviews to kilogram-scale photochemical [78] and SOE operations. [79] Just as CF has enabled PRC and SOE separately,i ti s expected to be an enabling platform for synthetic photoelectrochemistry.
Reports of flow photoelectrochemistry have thus far focussed on simple chemical transformations.F or example, Behm and co-workers reported the oxidation of formic acid to CO 2 by aphotoanode in CF. [80] Athin film of reaction mixture was pumped over af luorine-doped tin oxide(FTO)/TiO 2 photoanode under irradiation from aH g(Xe) (200 W) lamp ( Figure 17). Such ac onfiguration is suitable for certain chemical transformations,but may not be suitable for organic synthesis in general. This is due to shielding of the photoelectrode via absorption of UV light by reactants flowing atop it, facilitating potential side reactions/slow kinetics derived therefrom. Visible-light e-PRC or iPEC in CF has the advantage of selective delivery of light to the coloured e-PRC mediator in the flow path or through the flow path to the photoelectrode,r espectively.Aconceptual CF photoelectrochemical reactor for synthesis is shown in Figure 18. Here, groove channels are etched into the working electrode,which is covered with aborosilicate glass window.Anion-exchange membrane is sandwiched between the working electrode and counter electrode (here,asacrificial counter electrode is assumed). Thew orking electrode could be replaced with aphotoelectrode for iPEC.Additional groove channels could be incorporated into/above the counter electrode if as olution-phase half-reaction is necessary. [81]

Precomplexation and Redox Processes beyond the Electrochemical Solvent Window
Electrogenerated and photoexcited e-PRCs (*PTZC + , *DCAC À ,* TACC 2+ )d iscussed herein (Section 2.1.1) are rare, doublet excited states.T he ultrashort lifetime of doublet excited states (fs to ps) [40,82] is shorter than the timeframe for diffusion control and should prohibit outer-sphere SET events.While the mechanisms of such excited-state processes are still unclear, precomplexation is likely responsible for ultrafast quenching (inner-sphere SET) of e-PRCs and successful reactions.F or example, p-p stacking to generate ap recomplex that is photoexcited has been proposed to explain reactions involving excited perylene diimide radical anions and arenes. [11d,83] Such ap henomenon may likewise rationalize Lambertse -PRC oxidation of unactivated arenes [39] by *TACC 2+ as an e-PRC ( Figure 19).
Thee lucidation of such precomplexation mechanisms presents ac hallenge and demands the use of advanced spectroscopic,s pectroelectrochemical, and theoretical (computational) tools. [84,85] Thea bility to generate super-oxidants and super-reductants in situ and within close proximity to the substrate of interest (by an e-PRC-substrate precomplexation) may allow redox processes to take place at potentials beyond those available from PRC and beyond those normally tolerable by the organic solvent in which the reaction takes place ( Figure 20). Thereby,e -PRC may create a" realm" for extremely challenging SET processes such as direct oxidations of carbonyl compounds,sulfones,fluorinated aromatics, and hydrocarbons.D irect reductions of amides,e thers,S i À X bonds (X = Cl, F, OSiR 3 ,OR), sulfoxides,and sulfides may be possible.T he potentials that would be required in such scenarios by SoE would no doubt lead to decomposition/poor chemoselectivity.F inally,anotable challenge is the inability to measure redox potentials of substrates that lie beyond the redox window of the solvent. [86] Here,c omputational methods [87] to estimate redox potentials may prove useful.

Nomenclature
One key challenge for the scientific community is the adoption of appropriate and consistent nomenclature in this rapidly developing field of synthetic photoelectrochemistry. This is important not only to avoid misunderstanding between the different strategies and concepts herein, but also to avoid confusion with other distinct research fields and phenomena such as photoelectrolysis and the photoelectric effect. In fact, reports discussed herein (that do not involve photoelectrodes) are divided in their use of the terms "photoelectrochemistry", [47] "photoelectrocatalysis", [50] and "electrophotocatalysis", [38,39,49] while others carefully avoided using such terms. [52]

Reviews
TheI UPAC definition [88] of "photoelectrochemistry" is a "term applied to ah ybrid field of chemistry employing techniques whichcombine photochemical and electrochemical methods for the study of the oxidation-reduction chemistry of the ground or excited states of molecules or ions.Ingeneral, it is the chemistry resulting from the interaction of light with electrochemical systems."T herefore,w er ecommend "synthetic photoelectrochemistry" as ab lanket term that can encompass all of the examples presented herein.
In electrochemistry," electrocatalysis" refers to catalysis of electrochemical reactions by surface states of electrodes (e.g.,P ti sa ne lectrocatalyst for proton reduction to H 2 ), [89] where the role of catalysis is to lower the activation energy barrier for heterogeneous electron transfer. Accordingly, "photoelectrocatalysis" and "electrophotocatalysis" have historically referred to the catalysis of electrochemical reactions on the surface states of photoelectrodes.T os pecify the interfacial nature of this chemistry,a si nS ection 2.4, we recommend the terms "interfacial photoelectrochemistry" [60a] or "interfacial photoelectrocatalysis". [57] "Redox catalysts" and "mediators" are well-established terms in SOE for the shuttling of electrons between electrode surfaces and substrates, [21,89] while "photoredox catalysis" is au biquitous term in its field. [2,4] All examples in Section 2.1 can be considered as "photoredox catalysis", either at an elevated redox energy level or with electrochemical turnover. Thec atalysis of most interest to the synthetic chemist (that which is depicted in every example) [38,39,47,49,50] is indeed the photocatalytic cycle of the redox catalyst, not catalysis of the interfacial reaction involved in its generation/regeneration. In light of all of the above,w er ecommend the terms "electrochemically mediated photoredox catalysis (e-PRC)" and "electromediated photoredox catalysts" (e-PRCs).

Summary and Outlook
Synthetic photoelectrochemistry is as wiftly emerging research field following renaissances in its respective parent technologies,p hotoredox catalysis (PRC) and synthetic organic electrochemistry (SOE) that have taken place over the last decade. [90,91] To simplify the technology for users,this Review sets precedent for grouping historic and recent reports into three categories of photoelectrochemistry:e lectrochemically mediated photoredox catalysis (e-PRC), decoupled photoelectrochemistry (dPEC), and interfacial photoelectrochemistry (iPEC). Thef undamental advantages that derive from the fusion of PRC and SOE are expected to 1) broaden the accessible redox window of SET chemistry, [35,38,39] 2) enable milder conditions that allow greater functional group tolerance and chemoselectivity, [49,52,57] and 3) increase energy savings and atom economy. [60,64] Practical challenges in the execution of synthetic photoelectrochemistry could be addressed by an equipment and expertise interface with research fields of photoelectrochemical cells for water splitting and photovoltaic cells,while flow chemistry is expected to offer significant benefits to the transmission of light/electrons, [76,77] kinetics,a nd scalability of photoelectro- Figure 20. Redox potential scale showing current limitations of PRC and direct electrolysist echnologies and opportunities for e-PRC. [86] Angewandte Chemie Reviews chemical reactions. [76,77] We are particularly excited by the concept of an e-PRC-substrate precomplexation. [11d, 83] Further understanding of this concept is of critical importance, with potential to leverage it to improve the kinetics of SET processes as well as to control redox chemoselectivity.