Electrode Materials in Modern Organic Electrochemistry

Abstract The choice of electrode material is critical for achieving optimal yields and selectivity in synthetic organic electrochemistry. The material imparts significant influence on the kinetics and thermodynamics of electron transfer, and frequently defines the success or failure of a transformation. Electrode processes are complex and so the choice of a material is often empirical and the underlying mechanisms and rationale for success are unknown. In this review, we aim to highlight recent instances of electrode choice where rationale is offered, which should aid future reaction development.


Introduction
With high control of the rate,location and driving force of electron-transfer processes,electrochemistry is uniquely positioned to provide selectivity and sustainability benefits for the preparation of organic compounds.Asthese opportunities are being realised by academic and industrial research groups worldwide,the field of synthetic organic electrochemistry has received renewed interest over the last 5-10 years. [1][2][3][4][5] New synthetic methodology and reactivity has been developed, including processes that are more inexpensive,s afer and produce less waste than "classical" approaches. [6][7][8][9][10][11][12] In addition, the relative ease with which the technique can be scaled is demonstrated by the fact that several industrial organic electrochemical processes have been developed. [13][14][15][16][17][18][19] As the electron transfer between electrode and solutionphase electrolyte is heterogenous,t he development of synthetic organic electrochemical reactions requires close attention to parameters that are not traditionally encountered by organic chemists.A sw ell as optimising the applied current density or potential difference across ac ell, electrochemical reactions can be performed in either batch or flow cells,a nd divided or undivided cells.H owever,i tis the electrodes that constitute the most important difference,a st he success or selectivity of aparticular transformation is highly dependent on the material. Not only does the electrode material itself determine the mechanism of electron transfer, but the electrode separation distance,s hape and size determine the submerged surface area, the field homogeneity and the resulting current density;a ll of which can affect the reaction outcome.W hile the electrode material is an additional parameter that requires optimisation, it can be exploited to control and change the selectivity of areaction, and provides opportunities to vary reactivity through electrode-catalysis, (electrocatalysis), mediator-modified or chemically-modified electrocatalysis.
Thea bility of specific materials to give unique outcomes and determine the selectivity for synthetic electrochemical reactions has long been recognised. [20,21] Classical examples include the anodic oxidation of acetic acid in aqueous solutions ( Figure 1A), in which the identity of the products and their distribution varies with different anode materials, [22][23][24][25] and the reduction of acrylonitrile in which the reaction products strongly depend on the cathode material ( Figure 1B). [26][27][28] In this reaction, the formation of adiponitrile (1)w ith cadmium and steel electrodes is am ega-tonne per annum industrial process [15] that is used in the production of nylon-6,6, thus exemplifying the importance of the control that the electrode material imparts,and the possible ramifications of its variation. Thechoice of electrode material can impart am ore binary outcome by switching reactivity on or off.Classic examples of this include the cathodic hydrodimerization of formaldehyde to ethylene glycol, wherein product is only obtained with the use of mercury or carbon cathodes and no product is observed with lead or cadmium ( Figure 1C). [29] In addition, memory of chirality (enantiomeric excess) was only observed with the use of graphite anodes in adecarboxylative etherification reaction ( Figure 1D). [30] In am ore recent example,Xureports adrastic change in yield when exploring the electrode material in an aromatic CÀHf unctionalisation reaction with electrochemically-generateda midinyl radicals ( Figure 1E). [31] Va rying the electrolyte or the applied current had arelatively minor effect on the yield of the reaction, but replacing the reticulated vitreous carbon (RVC) anode with Pt completely shut down reactivity,whereas graphite gave an intermediate outcome.
While the differences in the outcomes of these reactions with different electrode materials are stark, the high complexity of electrode processes commonly renders the generation of conclusive explanations very difficult. Indeed, it has been noted elsewhere that it is "impossible to select the optimum electrode for agiven process on atheoretical basis.Instead, an empirical approach must be used". [21] While it is true that an empirical approach is currently the most efficient strategy to optimise ar eaction, an appreciation of the influence of electrode materials and ag reater understanding of electrode processes should lead to amore informed approach and new opportunities in the field. In addition to this,p oor reproducibility is am ajor challenge that accompanies the use of electrochemistry in organic synthesis,a nd differences in the electrode material, grade and source all contribute to this problem. Thus,a na ppreciation of the important factors associated with the electrode should facilitate an improved rationalisation of the differences between reported and achieved yields or selectivity.
In this review,weinitially summarise the most important practical and reactivity considerations for electrode materials in organic electrochemistry.T hen, our goal is to highlight The choice of electrode material is critical for achieving optimal yields and selectivity in synthetic organic electrochemistry.T he material imparts significant influence on the kinetics and thermodynamics of electron transfer,and frequently defines the success or failure of atransformation. Electrode processes are complex and so the choice of amaterial is often empirical and the underlying mechanisms and rationale for success are unknown. In this review,weaim to highlight recent instances of electrode choice where rationale is offered, which should aid future reaction development. examples in which the performance of ap articular electrode material is found to be unique and important. We place an emphasis on contributions to the literature from the last decade,w hile focussing on synthetic organic transformations and practical considerations on regular laboratory scales.I t should be noted that although an explanation on electrode choice is given in an increasing number of cases,o ft he protocols that we have surveyed from the last decade,o nly as mall percentage (ca. < 5%)p rovide some supporting insight. Reaction selectivity and yields can also very much depend on other reaction parameters,a nd so it is not always clear if the electrode material itself exclusively defines the observed difference.A st his point adds to the ambiguity,t he examples have been selected as carefully as possible,i n preference to providing an exhaustive coverage.T hus,t he reader is referred to an umber of earlier review articles that are also relevant to these themes. [21,[32][33][34][35][36][37][38][39][40] Beyond the scope of this review are photoelectrodes, [41][42][43][44][45][46][47] and other practical aspects of electrochemistry,w hich have been addressed in recent tutorial reviews. [48][49][50][51][52][53][54] 2. Electrode Selection

Practical Aspects
Thep rimary judgement of candidate materials will be based on their performance in the reaction, i.e., yields and selectivity,but current efficiencies,obvious signs of corrosion, cost, availability and machinability are other critical factors, the relative importance of which will vary according to the specific process.I no ther applications of electrochemistrysuch as those focussed on energy or bulk scale commodity production-small, single digit differences of efficiencyg ains can be extremely critical, for example from the use of ap recise grade of graphite.H owever,i no rganic synthesis where the scales are comparatively smaller, larger gains in yield or complete switches in selectivity become more important. This is because the cost of the electrode material and the man-hours that are required to optimise ap rocess must be balanced against the costs of the reagents and the value of the product. Forexample,asthe price of electricity is typically low compared to the contents of areaction mixture, achieving small gains in current efficiencyi sn ot the highest

Angewandte Chemie
Minireviews 18868 www.angewandte.org priority for reaction optimisation. This will only become ac onsideration if the scale is increased and the value of product is lowered.
Whilst electrodes can in principal be made from any conductive material, in order to make an appropriate choice there are an umber of mechanical and electrochemical properties to consider.Anidealised electrode material should be inexpensive,n on-toxic,s table to aw ide range of temperatures,pressures and solvents,yet able to be manipulated into forms for electrode construction, such as rods,w ires,p lates, foams and meshes.M ost electrodes consist of as ingle material, but as upport material combined with an electroactive coating, such as Pt, can also be used. [20] In organic solvents,w hich are more resistive than aqueous systems,t he use of 3-dimensional, high surface area electrodes is advantageous,a st hey impart decreased current density and cell potential. Thus,t he use of RVCo rc arbon felt can increase productivity as higher currents can be applied. [37,55] Between electrode materials,s urface area can vary dramatically,f or example,the surface area of a"smooth electrode" can be up to 3orders of magnitude lower than aporous surface,such as platinised platinum. [56] An electrode should be stable and resist corrosion. An exception to this is when the electrode is sacrificial, for example when metal ionisation is intended as ac ounter electrode process,orwhen the metal ions are used to stabilise aproduct, such as in acarboxylation reaction. [57] Degradation of electrodes by mechanical action can occur as aconsequence of convection forces in the reaction vessel, such as the release of graphite particulates,w hich requires separation via filtration. In addition, fragile materials,s uch as low pore density RVC, can lead to difficulties in physical handling and manipulation. Swelling of the electrode can also be problematic with certain electrode material/electrolyte combinations.
Theu se of electrodes with high resistivity leads to an ohmic (IR) drop,which creates the requirement for ah igher cell potential. This excess energy input is likely lost as heat into the reaction medium, which is inefficient and may be deleterious to the reaction outcome. [58] On an industrial scale, this can limit the choice of materials to only those that are highly conductive,o rr equire special electrode architectures. [21] Once am aterial is formed into an electrode,alow ohmic resistance connection should also be made.

Reactivity Aspects
Them echanism for electron transfer at an electrode occurs between two limiting scenarios.I nt he first limiting case (Figure 2A), the electrode surface is intimately involved in the mechanism of electron transfer and acts as acatalyst in the reaction;i.e., electrocatalysis. [52] Theproducts,mechanism and kinetics of electrode reaction in this case are highly dependent on the composition of the electrode material, meaning that small differences may be extremely significant in determining the outcome of the reaction. Conversely,inthe second limiting case ( Figure 2B), the electrode is completely inert and provides as ource or sink of electrons that are transferred in an outer-sphere manner between the substrate and electrode.T he identity of the products formed, and the mechanism and kinetics of their formation should be independent of the material.
Thep otential required beyond that necessitated by thermodynamics to drive ar eaction at ap ractical rate is referred to as the overpotential (h). [59] Theo bserved overpotential in ap articular system is as um of the individual overpotentials for each step in the process,such as adsorption, charge-transfer,d esorption and mass-transport (diffusion, convection and migration) overpotentials.A st he electrode material dictates the mechanism of electron transfer, it is the biggest contributing factor to the overall overpotential of ap rocess.T his important factor will be responsible for outcome variations observed during reaction optimisation.
Form any reactions,s uch as the Hydrogen Evolution Reaction (HER) or Oxygen Evolution Reaction (OER), the decrease in overpotential through new electrode materials is the subject of intense investigation. [60][61][62][63][64][65][66][67][68][69] Small efficiency gains will translate into large cost savings when these processes are conducted on scale.H owever,o fg reater importance to synthetic organic electrochemistry is the selectivity changes or suppression of side reactions that are enabled by the different overpotentials for each process on different electrode materials.A ne xample of this control in as ubstratereduction reaction is to suppress competing proton reduction (HER) by choosing ac athode material that has ah igh overpotential for this process.I ndeed, the overpotentials on common electrode materials varies considerably for the HER and OER, Table 1a nd Figure 3. Al ow overpotential for the desired redox reaction will not only ensure the reaction can be driven more efficiently but will improve selectivity over competing processes.T he overpotential for solvent oxidation or reduction can also vary significantly on different electrode materials. [70] This variation has implications for the width of the potential solvent window that is available to ar eaction and therefore to the extent of redox chemistry that can be performed, Figure 4.

Angewandte Chemie
Minireviews 18870 www.angewandte.org mass balance and yield of product ( Figure 5A). Strategies for grafting organic compounds onto electrode surfaces for intentional surface modification have also been reported. [71][72][73][74] However,u nintentional grafting can vary in degree depending on the specific redox event, electrolyte and electrode material. Ther esult is ap assivated electrode with decreased electrode activity due to the formation of an electrically insulating layer. Electrode passivation can be detected by cycling acyclic voltammetry (CV) experiment and observing the current decrease with each cycle, [75][76][77] with the current not being restored to its original value until the electrode is cleaned. Examples of electrode passivation are tight oxide films on metals that are formed at high anodic potentials, [78,79] insoluble oxidation products,p olymer deposits generated by anodic oxidation of olefinic,a romatic or phenolic compounds, [80][81][82] or solutions of HF or ionic liquids. [83,84] Optimisation of the electrode material is akey task in remedying this effect ( Figure 5B). Other methods that have been shown to be effective include pulse electrolysis, [85] sonication, [86] alternating the polarity of the electrodes (which can also effect the reaction selectivity or yield), [87][88][89] and the use of mediators to shuttle redox equivalents from the electrode to the substrate in the bulk phase. [75,90] Alternatively,the addition of additives can increase the solubility of the insulating polymer in the electrolyte or protect the electrode surface,w hich has been shown to be highly effective in arecent electrochemical Birch reduction. [91]

Trends
Thef actors that contribute to the choice of electrode material vary and can be very specific. Then umber of electrode materials available has increased over time and trends of use have changed and evolved. Fore xample,l ead and mercury were previously preferred due to their high hydrogen overpotential (h H )a nd stability to acidic media. With mercury being in the liquid state,t he surface is constantly renewed and can remain clean and free of impurities.H owever,c oncern over the high toxicity of these metals has limited more recent wide-spread use and hence other materials have attracted greater attention. Modern organic electrochemical methodology relies more heavily on platinum, which is robust, easy to clean and redox stable,a s well as carbon-based electrodes that are more inexpensive and thus appropriate when the scale of areaction renders the cost of platinum prohibitive ( Figure 6). [130,131] Glassy carbon is the most commonly used carbon material, which is the fullerene allotrope of carbon, [132] and includes the highsurface area foam form, RVC. [34] Graphite is also acommonly used form of carbon electrode,w hich is less chemically inert than glassy carbon but more conductive [133] and is less expensive.T he diamond allotrope of carbon can also be used, Boron Doped Diamond (BDD) has emerged as aunique material and is becoming increasingly popular. [134][135][136][137] There has also been evidence for the emergence of new materials, metals or alloys used as electrodes in organic synthesis,s uch as leaded bronze,t antalum, niobium or molybdenum. [138][139][140][141] No doubt this trend will continue as the electrode processes with each material become better understood, wider range of materials are adopted, and the further development of idealised materials.

Electrocatalysis:S pecific Adsorption and Surface Interactions
At the extremity of the first limiting case (Figure 2), the electrode surface is explicitly involved in the reaction mechanism through specific adsorption and surface interactions.Aswell as providing the required redox equivalents,the electrode surface serves to catalyse the reaction, and is thus known as electrocatalysis. SavØant defined "Electrocatalysis" as the term to "name catalysis of electrochemical reactions by surface states of the electrode", [52] which is distinct from mediated electrolysis that employs molecularly defined catalysts.T he precise nature of these interactions varies, depends on each specific reaction and can often be the subject of much debate.Nevertheless,theoretical models are improving and can now describe catalytic reactions in great detail. [142] Thes trength of interaction (adsorption vs.d esorption), the timing and order of electron transfers and the concertedness of steps are all relevant when considering the mechanism. Adsorption describes avariety of more specific interactions of as ubstrate with the electrode,s uch as strong electrostatic interactions, p-interactions and chemical bonds.Aswell as the  smooth material itself,t he sites of binding and catalysis may be impurities,e dge or end atoms,d eposited nanoparticles, thin films or single atoms of asecondary or different material to the bulk material. [143][144][145] Thes trength of interaction between substrate and electrode should be strong enough to trigger ar eaction, but not too strong that the product fails to dissociate and desorb.This balance is known as the Sabatier principle [146] and has been shown to contribute to the bell curves observed for rates of electrocatalytic HER. [147,148] In organic electrochemistry,i ti s common for products to avoid dissociation from the electrode ( Figure 5A), which can lead to decomposition and alow mass balance at the end of reaction.

Working Electrode Material
In aclassical example,the extent of electrocatalysis in the oxidative decarboxylative Kolbe and Hofer-Moest reactions has been the subject of much debate in the literature over the years. [149][150][151][152] In one study,t he product distributions from the use of ap latinum and carbon anodes were compared. It was found that the ratio of 1-electron vs.2 -electron oxidation products (i.e., ratio of products-from-radicals over productsfrom-cations) was much greater with platinum anodes than with graphite anodes (Figure 7). [153] Carbon anodes are more efficient than platinum anodes at removing asecond electron, to form ac ation (with ap roton loss). [145] Thed ifference was proposed to be due to agreater tendency of radicals to adsorb onto carbon because of the presence of paramagnetic centres in the material. Thus,t he adsorbed radicals on carbon undergo further oxidation to form ac arbocation that is then electrostatically repelled and primed to react with nucleophiles.However,the radicals produced on aplatinum surface are largely desorbed and so participate in radical reactions. This effect has also been recorded in other transformations, such as in the electrochemical cyclisation of dienes,i nw hich Moeller observes ad ifference in the efficiencyo f1 -v s. 2electron pathways when using platinum and carbon anodes. [154] More recently,e lectrocatalysis has been especially noted for cathodic processes;i th as even been remarked that "it seems uncertain that totally inert electrodes exist […] within the cathodic range". [155] In particular, the dehalogenation of aryl and alkyl-halides with different cathodic materials has been the subject of significant investigation. [21,[155][156][157] Theo verpotential, concertedness and degree of interaction varies with different electrode materials and can lead to the formation of different product distributions.F or example,the use of silver cathodes significantly decreases the over-potential necessary to cleave aC À Xb ond, [157] compared to mercury or glassy carbon cathodes.I nt he reduction of linear alkyliodides on smooth silver cathodes,t here is evidence for the transient formation of [Ag + ÀR] I À species on the interface. [155] The formation of such species will stabilise the radical and ensure al ower reduction overpotential. Compared with glassy carbon electrodes,c opper has also been found to show exceptional electrocatalytic properties,e ither as as mooth metal or when deposited onto aconducting surface. [158] Thee lectrocatalytic dehalogenation of aryl-halides can occur via astepwise electron transfer-cleavage mechanism, or aconcerted process.Recent analysis of an extensive range of cathode materials revealed as trong dependencyo ft he mechanism of debromination on the electrode material ( Figure 8). [159] Electron transfer coefficients (a)g ive an indication of how reactant or product-like the transition state is in terms of its electrical behaviour.T hese were extracted from CVs by analysis of the difference between the peak potential and half-wave potential, and used as an indication for the mechanism. Avalue of a significantly lower than 0.5 indicates ac oncerted mechanism, whereas as tepwise mechanism will either have an a significantly higher than 0.5 if cleavage is the RDS or only slightly lower than 0.5 if ET is the RDS.T hus,r eduction potentials and electron transfer coefficients were measured by cyclic voltammetry for the reduction of different aromatic bromides on different electrode materials.Only 4ofthe 11 materials showed reduction features in the CV.Agand glassy carbon were found to both follow ac oncerted mechanism:A ge xhibited ar emarkable electrocatalytic activity with a0.9 Vlower overpotential than

Angewandte Chemie
Minireviews 18872 www.angewandte.org glassy carbon. This overpotential difference is aconsiderable thermodynamic improvement and demonstrates the significant effect that materials can have on the overpotential of an electrochemical redox event. Cu and Zn electrodes were found to give step-wise mechanisms,w ith rate determining ET and cleavage steps,r espectively.
Thec athode material can determine the distribution of products from areaction. Thereduction products of alkylhalides have been reported by Peters to vary according to the cathode material. [160,161] On vitreous carbon cathodes, ndecane and 1-decene are yielded from the reduction of iododecane,whilst on asilver cathode,adimeric product was formed. When testing secondary alkyl halides,itwas interesting to note that the product distribution switched, such that dimers were predominately formed at carbon-based cathodes.
Avoiding the adsorption of reagents and the subsequent electrocatalysis of competing side-reactions can be critical to the success of ad esired reaction. This can be achieved using an electrode material with ah igh overpotential for the competing processes.F or example,i nareduction reaction, competing proton reduction can be avoided through the use of acathode with ahigh overpotential for that process,such as glassy carbon, mercury or lead. Lead has found use for this reason in the deoxygenation of amides and sulfoxides (Figure 9A). [162] Amides are thermodynamically difficult to reduce and the presence of acid is necessary to provide the equivalents of protons.T herefore,i tw as important to use am aterial that preferentially reduces amides over protons, and lead was found to be superior for that.
In the reduction of menthone oximes to menthylamines ( Figure 9B), Waldvogel screened cathode materials to avoid any competing proton reduction. [119,163] Those with only am oderate hydrogen overpotential, such as titanium, iron, copper, zinc, indium, tin and bismuth, all failed to produce the desired product, and so it was necessary to use ah igh overpotential material, such as lead or mercury.The electrode material also influenced the selectivity of the reaction: whereas,m ercury and cadmium cathodes led to pronounced diastereoselectivity,l ead or copper/lead gave either no or little selectivity.T he authors proposed that good diastereoselectivity was due to stabilisation of the reactive intermediate by stronger binding to the electrode surface and slowing down conformational switching.
Thee fficiencyo ft he electrochemical reductive carboxylation of imines to yield N-bromoamino acids also depends on the cathode material (Figure 10 A). [164] In this case,t he yield was proposed to correlate with the strength of substrate adsorption onto the electrode surface.S ilver was noted to exhibit pronounced specific adsorption of imines,which leads to ah igher concentration of imines on the electrode surface, and therefore more facile decomposition and accelerated imine dimerisation. Highest yields were reported using nickel cathodes.W hen adapting the reaction into flow,A tobe considered cathode materials with ah igh overpotential for carbon dioxide reduction (Figure 10 B). [165] In this example, the overpotentials (determined by linear sweep voltammetry) correlated strongly with yield. Glassy carbon gave the highest efficiency,f ollowed by graphite,t hen platinum and lastly silver.R eduction of the imine to aradical anion is necessary for aproductive reaction to take place.Any competing direct CO 2 reduction decreases charge efficiency and reaction yields.

Angewandte
Chemie cathodes,t he reaction was completely unsuccessful (Figure 11 A). Only the use of silver produced any desired 3,5dichloropicolinic acid (3)p roduct, ac ompound with significant pharmaceutical relevance.I tw as proposed that the overpotential for proton reduction was lower with the use of Cu and Ni, making substrate reduction more difficult. The productivity of silver cathodes compared to glassy carbon (GC) cathodes was ascribed to an electrocatalytic effect of silver that was not possible on carbon (Figure 11 B). Interestingly,t he selectivity of hydrodechlorination was found to be dependent on the pH as well. Electrostatic forces engendered high selectivity for the 3,5-substituted isomer (3)atpH3,and the 3,6-substituted isomer (4)a tp H13( 97 %) (Figure 11 B).

Counter Electrode Material:Electrogenerated Base
As well as tuning the overpotential on the working electrode,the overpotential on the counter electrode reaction is also important to consider and has often been shown to be key to the success of ar eaction. Of particularly frequent consequence is the reduction of protons to evolve hydrogen (HER) at the cathode to form abase.The concept of forming electrogenerated bases (EGBs, Figure 12 A) [167,168] in situ from ap ro-base for utilisation in as ynthetic transformation was first reported in 1967. [169] Using electrochemistry allows the concentration and basicity of the reaction to be carefully controlled, [170] and changing the counterion influences both the stability and reactivity of the EGB. [171] To generate bases electrochemically (EGB), the reduction potential of the pro-base must be less negative than any other species in the reaction (including the product), which renders the overpotential for proton reduction vital for success to be achieved. By promoting hydrogen evolution over other potential reductive processes,t he choice of cathode material influences the outcome of anodic transformations. [138,172,173] An example of this is in the synthesis of (E)-vinylsulfones from cinnamic acids,i nw hich Wang found as ignificant dependence of the reaction on the counter electrode material (Figure 12 B). [174] Whilst the reaction is oxidative with respect to the substrates,t he cathodic generation of base is required for the deprotonation of the carboxylic acid. Platinum was the best performing cathode material, which has the lowest overpotential for proton reduction, whilst glassy carbon, which has ah igher overpotential, resulted in ad ecreased yield. Materials with am edium h H performed in-between these two cases.Thus,alower potential difference is necessary with Pt under constant current electrolysis conditions,w hich limits competing reduction processes.S imilarly,Z hang switched from carbon to platinum counter electrodes in an electrochemical Hofmann rearrangement, to more readily form methoxide on the cathode and found that yields improved (Figure 12 C). [175] Thee lectrochemical oxidative formation of N-centred radicals and their intramolecular cyclisation onto alkenes has been developed by Moeller and Xu. [176][177][178][179][180][181][182] Electrode materials were thoroughly analysed by Wirth when the process was transferred to aflow cell set up. [183] Interestingly,itwas found that the choice of counter electrode material had as tronger

Angewandte Chemie
Minireviews 18874 www.angewandte.org influence on yield than the choice of the working electrode material. Ayield of over 90 %w as achieved with ap latinum cathode,a round 60 %w ith nickel, and no reaction occurred with ag raphite cathode (Figure 12 D). These results directly correlate with the over-potentials for proton reduction (Pt = À0.09 V; Ni = À0.32 V; graphite = À0.47 V, Table 1). Yields were consistent (ca. < 5%)b etween the anodic materials tested with aP tc athode.C orrect choice of cathode material therefore meant the anode material could be chosen by cost, rather than due to its influence on the reaction.
In an elegant example of the importance of counter electrode material, Xu recently reported acomplete selectivity switch when the cathode material was switched from platinum to lead. [184] Ther eaction is an oxidative TEMPO catalyzed intramolecular arene amination from oxime 5 ( Figure 13). When platinum was used as the counter electrode,t he low overpotential for proton reduction resulted in the N-oxide product 6 remaining intact. However,when lead was employed, the deoxygenated heterocycle 7 was returned. This is because lead has am uch higher hydrogen overpotential for proton reduction, meaning proton reduction is more difficult and so N-oxide 6 now preferentially reduced on the counter electrode.

Modified Electrode Surfaces
Them odification of electrode surfaces to aid catalysis of areaction and decrease the overpotential for electron transfer is at echnique that is well established, especially for energyrelated applications,s uch as the HER, OER and CO 2 reduction. [60][61][62][63][64][65][66][67][68] Forsynthetic applications,the immobilisation or tethering of electron transfer mediators onto electrode surfaces,e ither covalently [185] or non-covalently [186] has been shown to improve the efficiency of reactions.Carbon electrodes are especially effective supports for catalysts as they can be readily functionalised. [74] Forexample,oxidation produces ah igh density of surface carboxyl groups with which amide bonds can be formed, or the single electron reduction of diazonium cations reveals arene radicals that readily combine with graphitic electrodes. [187,188] Further details of the mediator-immobilisation approach are,however, beyond the scope of this review,a nd the interested reader is directed to other relevant reviews. [74,189] Theb ulk surface modification of electrodes through, for example,p olymer coating or nanoparticle deposition is comparatively less well exploited for organic synthesis compared to energy applications. [190] Ar ecent example demonstrated that the in situ generation of an active Mo V layer on the surface of aMoanode was responsible for greatly enhanced yields in the dehydrogenative coupling of arenes ( Figure 14). [138] Whilst this arene coupling reaction could be performed with BDD,P t, Au,V ,C ro rWelectrodes,t he efficiency was less optimal than with the use of Mo.Only very low levels of molybdenum were detected by mass spectrometry in the electrolyte solution, which is evidence that the active Mo V species is only present on the surface and not released into solution.

Double Layer Control
Upon application of ap otential to an electrode in solution, an ordering process occurs to form as tructure of oppositely charged ions and solvent molecules at the surface, commonly known as the Helmholtz double layer. There have been several theoretical models proposed for this interfacial region (Helmholtz, Gouy-Chapman, Stern) but the precise behaviour depends on the nature of electrode (material and surface properties), as well as the composition of the electrolyte (supporting electrolyte,s olvent). [21] Unlike under aqueous conditions,the structure and thickness of the double layer in organic solvents is not well-understood. Nevertheless,t he structure determines the potential distribution close to the surface and the uniformity of current. Thed ouble layer thus influences the local driving force for electron transfer,which determines the kinetics of electron transfer.
Waldvogel manipulated the interfacial region in the reduction of menthone oximes through the addition of quaternary ammonium salts ( Figure 9B). [119] These small, hard cationic species form acompact and robust layer on the cathode surface (Figure 15 A). It was found that di-or polyammonium salts separated by an alkyl chain also gave superior reaction outcomes,p ossibly due to an entropic effect. Theh ard, lipophilic layer was able to exclude both solvent and protons from the surface,a nd decrease sidereactions.T he ammonium salts serve to attenuate corrosion of the lead cathode,suppressing the formation of lead sulfate deposits to keep as hiny intact surface.T he adsorption of ammonium cations also serves to increase the hydrogen overpotential of the cathode by reducing the rate of the HER. This effect was further studied in an amide deoxygenation reaction. [118] It was proposed that the cationic layer still allows the tunnelling of electrons to reduce substrate,b ut protons are repelled due to coulombic repulsion. By avoiding competing proton reduction, the double layer protects the electrode from corrosion, and leads to an improved performance.Another example of the use of quaternary ammonium  salts to suppress hydrogen evolution was demonstrated by Bhanage in the reduction of N-alkoxyamides to esters. [191] Reactive species,s uch as cations or radicals,c an readily react with solvent molecules to form undesired products. However,w hen formed at an electrode within the double layer, solvent can be excluded, which can aid reactivity and enhance selectivity.A ne legant example of this is from Xu, who reported an alkynyl-hydroxylationr eaction of alkenes, which is highly regio-and chemo-selective and proposed to only be successful because the selectivity-determining alkyne addition step occurs within the polarised double layer (Figure 15 B). [192] Then egatively charged alkynyltrifluoroborate nucleophile is attracted to the region and creates al ocalised high concentration, while competing neutral nucleophiles, such as water, are excluded. Oxidation at an electrode was found to be essential for these reasons,a sp hotochemical or chemical oxidation conditions facilitated direct reaction with water in preference to alkyne.
Moeller also demonstrated that an ordered double layer can improve selectivity in an intramolecular cyclisation reaction by promoting cyclisation over solvent trapping (Figure 15 C). [193][194][195] Thea nodic oxidation and the ensuing cyclisation occur within the ordered environment of the double layer at the anode surface,w hich slows diffusion and excludes the methanol from the electrode surface that could otherwise interfere with cyclisation. Silyl protection of the internal alcohol moiety of the substrate was still necessary to prevent the intramolecular trapping by this competing nucleophile.

Inert Electrodes
At the other end of the scale to the significant involvement of the electrode and high levels of electrocatalysis is electron transfer from an inert electrode that does not participate in the mechanism and has little substrate or intermediate adsorption. An outer-sphere-type electron transfer mechanism is more dominant, which results in high overpotentials for specific reactions.T he best-known inert material is boron doped diamond (BDD), the use of which in organic electrosynthesis has primarily been driven by Waldvogel. [36,[196][197][198][199][200] Although, the level of interaction of an electrode in ar eaction is very difficult to determine,B DD has the highest known overpotential for the oxygen and hydrogen evolution reactions,which indicates very low levels of interaction. Because of this,B DD also offers av ery high potential window and is highly chemically inert. However,i t has been shown that the level of boron doping can actually affect selectivity, [201] and sp 2 non-diamond carbon impurities alters the potential window, [144] implying that the material is not perfectly inert. BDD has been reported to yield differences in selectivity to other materials in various reactions, such as CO 2 reduction. [202][203][204] However,h erein, we describe several synthetic organic examples that have required the use of am ore inert electrode,w hich BDD has fulfilled. More specific features of BDD and its general use has been well reviewed elsewhere. [137,[205][206][207] Theelectrochemical CÀHamination of arenes via Zincke intermediates (8)was reported by Yoshida using acarbon felt anode and platinum plate cathode,F igure 16. [208] However, the scope was limited to electron rich rings containing methoxy groups.I na ne ffort to widen the scope toward arenes with less electron density,W aldvogel explored the use of different anode materials in the reaction. [200] While retaining the Pt counter cathode in adivided cell set up,the use of carbon felt or fleece anodes were confirmed to give moderate or poor results for the amination of alkylated arenes.T hese porous carbon materials have high surface area, which causes diffusion of the radical cation away from the electrode to be more difficult, as it is liable to adsorb and oxidise further. Platinum, glassy carbon and isostatic graphite anode materi-

Angewandte Chemie
Minireviews als all returned very poor yields,w ith electrode fouling observed with the former two and corrosion with the latter. However,switching to aBDD anode resulted in asignificant boost in yield, up to 60 %. CV studies were conducted to gain greater insight into the enhanced performance of BDD compared to platinum and glassy carbon. CV traces of xylene (red) and xylene with pyridine (blue) were recorded and compared ( Figure 16). With the use of Pt and glassy carbon anodes,t he oxidation feature of xylene disappeared upon addition of pyridine.H owever,w ith aB BD anode the CV trace was unaffected, and the oxidation of xylene was retained. This trend correlates with the outcome of the amination reaction at each anode material.
An electrochemical dimethoxylation was ak ey step in Nishiyama and Einagass ynthesis of (AE)-parasitenone. [209,210] Theuse of BDD and Pt for the oxidation of 9 gave excellent yields of the desired product 10.H owever,g lassy carbon or the use of chemical oxidants returned ad ifferent, aldehyde product 11 (Figure 17 A). Anodic oxidation of 9 leads to the radical cation 12,f rom which 11 is formed from methoxide deprotonating the benzylic position (via route c). Although methoxide attack of the ring leads to the product 10 (via route a), the anode material dependency on the reaction selectivity suggests an alternative electrode material-dependent mechanism. ESR studies revealed the formation of methoxyl radicals,most efficiencywith aBDD anode,toalesser extent with Pt, but not at all with GC (Figure 17 B). Thea uthors proposed that these data signal that methoxyl radical attack onto the radical cation is the leading pathway to product 10 (via route b)a nd were used to explain the selectivity differences observed with each anode material. As methoxyl radicals are highly reactive,a ni nert electrode proved to be essential for this transformation.
Theu se of aB DD anode led to the highest yields in the challenging oxidation of isoeugenol (13)to(AE)-Licarin A(14) (Figure 17 C). [210] This was similarly proposed to be due to the more efficient formation of methoxyl radicals on BDD. Lower yields of the desired product 14 and overall mass balance were observed with Pt and lower still with glassy carbon. Theo ther side-products that also required the formation of highly reactive radicals were also formed in greater quantities with BDD.I nterestingly,t he oxidation of isoeugenol on BDD in hexafluoro isopropanol (HFIP) give the homo-coupled product, diisoeugenol. [211] Waldvogel tested the influence of anode materials in the cross-coupling of phenols and arenes.P reliminary studies revealed that the use of carbon electrodes gave only homocoupled adducts. [196] Platinum plates improved the yield and selectivity of cross to homo-coupled ratio to 1:1, but aswitch to BDD gave afurther enhancement in the selectivity (1.5:1). Further optimisation led to aset of improved conditions that contained methanol or water as an additive ( Figure 18). [212] When the electrode material was varied again, BDD,P ta nd Figure 17. A) Twopossible mechanisms for the oxidative methoxylation of 9 (a) and (b) to give 10.A ldehyde 11 is formed from (c);B)ESR spectra reveal methoxyl radicals, leading to mechanism (b);C )formation of methoxyl radicals are more efficient on BDD anode in oxidation of isoeugenol.

Angewandte
Chemie GC all gave excellent selectivity,b ut BDD proved to be the superior anode material with respect to yield. These observations led the authors to propose the formation of alkoxyl radicals,s tabilised by the HFIP environment and which mediated the generation of the phenoxyl radical intermediate. Further studies revealed the alcohol additive beneficially altered the oxidation potentials of the substrates. [213] Nonetheless,t he use of BDD as an inert electrode had ap ositive effect on the outcome of the reaction.

Sacrificial Electrode
Theu se of am etal anode with av ery low oxidation potential will leech metal ions into the solution upon its oxidation. In this case,itistermed asacrificial electrode,asit is being consumed stoichiometrically.T his approach is practical, facile and so frequently applied as ac ounter electrode process in electrochemical reduction reactions. Common choices for asacrificial anode include zinc, magnesium or aluminium. In many cases,t he choice is dictated by price or toxicity concerns,w ith little effect on the reaction observed. However,commonly,the liberated ions play arole in the reaction by coordinating reactants or products,a nd maintaining high conductivity.Care should be taken to avoid reduction of the liberated ions on the cathode to avoid short circuiting the system, hence the use of highly reducing metals that thermodynamically disfavour this process.
Ther eduction of organohalides is ar eaction in which as acrificial anode is frequently used, [214,215] in particular for carboxylation reactions by coupling with CO 2 as an electrophile. [57,139,[216][217][218][219][220] Them etal ions liberated from the anode stabilise the carboxylate product, which also helps to prevent anodic Kolbe-type reactions of the carboxylate. [57,165,220] A more recent example is the use of either Mg or Al anodes by Baran in an electrochemical Birch reduction (Figure 19 B). Thereaction was shown to be highly scalable and remarkably tolerant to av ery wide range of substrates. [91] In an earlier study of the same reaction (Figure 19 A), [86] Mg anodes were also found to give the highest yields,which was proposed to be due to Mg 2+ ions acting as electron transfer catalysts or stabilising anionic intermediates and promoting their reduction. Mechanistic studies were undertaken by Baran to elucidate if liberated Mg 2+ ions played ab eneficial role in the reaction. Theaddition of MgBr 2 as an additive only served Figure 18. BDD gives best yields for phenol/arenec oupling reaction, which produces reactive radical intermediates.    www.angewandte.org Table 3: As ummary of the key properties and example uses of electrode materials for organic electrosynthesis.

Angewandte Chemie
Minireviews 18880 www.angewandte.org to decrease the product yield, although it seems likely that its reduction may compete with substrate reduction. In the absence of stirring, diminished yields could also be correlated to smaller electrode separation distances.T hese data suggest the diffusion of metal ions to the anode is deleterious to the reaction and so the liberated ions are not mechanistically relevant.
Rather than as acrificial anode producing only waste products,i tc an be used to liberate reagents into solution in ac ontrolled manner that matches reaction progress,o ften yielding results that are not possible by other means.I na n elegant recent example of this,S evov reported the use of an aluminum anode in the deoxygenation of phosphine oxides ( Figure 20). [221] Thes acrificial electrode oxidises to liberate aluminum ions into solution that combine with chloride ions to form an amine-stabilised AlCl 3 complex. This in situ generated Lewis acid activates the phosphine oxide,p roducing aless negative potential for its reduction and subsequent deoxygenation. As 2electrons are required for the phosphine oxide reduction, and 3are removed from Al 0 to give Al 3+ ,an additional quantity of added AlCl 3 is required to balance the stoichiometry and ensure high reaction efficiency,w ithout which, alower performance was observed.
Metal anodes have been employed as sacrificial electrodes by Willans to provide metal ions into solution, at acontrolled rate and with control of the oxidation state,i no rder to generate organometallic complexes that are not otherwise obtainable (Figure 21 A). Examples of this approach include the use of copper,iron and manganese anodes in the presence of NHC and salen ligands,t of orm the corresponding complexes. [222][223][224] Ther eaction is ap aired process (Figure 21 B), that is,t he ligand precursor is reduced on the cathode and the metal ions are produced on the anode.T he procedures are remarkably straightforward and yield high purity complexes without the necessity for column chromatography.S imilarly,M ellah demonstrated Sm II complexes could be efficiently prepared through the use of as acrificial samarium anode,a nd are important catalysts. [225][226][227]

Summary and Outlook
While ar ational choice of electrode material for use in organic electrochemical transformations cannot yet be made readily and reliably,herein, we have highlighted where efforts have moved beyond screening and empirical investigations.In many cases,e fforts to understand the influence of electrode material through analytical electrochemistry and ap hysical organic chemistry approach have led to the elucidation of trends.Such trends and insight may be applied more broadly, which will lead to enhanced efficiencies and new opportunities.D ue to the complexities and variation of electrodesubstrate interactions in organic transformations,i ti sl ikely that experimentation will remain necessary,e ven when the choice is guided by principles.T oa id exploration of the breadth of materials available,t wo tables are appended summarising key materials,their properties,and applications in electrosynthesis (Tables 2a nd 3).
Thec riteria for an ideal electrode material is that it is inexpensive,non-toxic,stable,manipulatable,resist corrosion and, most importantly,p rovide high yields and exquisite selectivity.W hile an umber of materials perform extremely well and fit many of these criteria, it is clear that there is currently no material that meets all of them. These criteria are also reaction-specific,ascost, selectivity and yield have to be balanced against the cost of product and ease of access to it by other means.T he development of new materials and the design of robust electrocatalysts for organic synthesis also still lags many other applications of electrochemistry,y et may provide new opportunities for the field. While the electrode material remains akey optimisation parameter,itholds great opportunity to impart new reactivity and greater reaction efficiency.