Recent Progresses in the Preparation of Chlorinated Molecules: Electrocatalysis and Photoredox Catalysis in the Spotlight

Abstract

Among halogenated molecules, those containing chlorine atoms are fundamental in many areas such as pharmaceuticals, polymers, agrochemicals and natural metabolites. Despite the fact that many reactions have been developed to install chlorine on organic molecules, most of them rely on toxic and hazardous chlorinating reagents as well as harsh conditions. In an attempt to move towards more sustainable approaches, photoredox catalysis and electrocatalysis have emerged as powerful alternatives to traditional methods. In this review, we collect the most recent and significant examples of visible-light- or current-mediated chlorination published in the last five years.

1. Introduction

The incorporation of halogen atoms into molecules is of great importance in organic synthesis as well as in materials science and medicinal chemistry. Halogenated, and more specifically chlorinated, compounds are not only widespread in nature (Figure 1a) [1,2,3], but also essential intermediates in the preparation of complex molecules and active pharmaceutical ingredients.

Chloroalkanes are all-round building blocks in organic chemistry. They are fundamental precursors to alcohols, ethers, amines and thioethers, which are prepared through displacement by appropriate nucleophiles. They are also used in the Finkelstein reaction to make more reactive iodoalkanes [4]. Moreover, organochlorine compounds can be converted into synthetically important organometallic species, such as Grignard and organolithium reagents, and can produce reactive radical intermediates through halogen atom transfer (XAT) [5]. In the area of pharmaceuticals, both aromatic and aliphatic chloro compounds are key intermediates in the preparation of a large variety of drugs. Some of them are active ingredients themselves and used as antibiotics (e.g., floxacillin and dicloxacillin), anesthetics (e.g., halothane and isoflurane) and anti-inflammatory drugs (e.g., alclometasone) (Figure 1b) [6,7,8,9]. In addition, vinyl chloride, used in the plastic manufacture for the preparation of polyvinyl chloride (PVC), is among the largest petrochemicals in world production [10]. Some of them also find application as paint additives [11], and in food packaging [12]. Finally, some chlorinated hydrocarbons (e.g., dichloromethane, chlorobenzene and tetrachloroethylene) are used as solvents in organic chemistry and in cleaning applications, such as degreasing and dry cleaning.

Therefore, the continuous interest towards the development of new methods to incorporate chlorine atoms is not surprising. The formation of C-Cl bonds can be achieved via three general approaches: electrophilic addition to C-C multiple bonds, nucleophilic displacement of leaving groups or radical chlorination of C-H bonds. Alkenes are readily converted in the corresponding dichloro and monochloro derivatives upon reaction with Cl₂ or HCl, respectively. Chlorine can also be used for the chlorination of aromatic hydrocarbons. In such cases, a Lewis acid (FeCl₃, AlCl₃) is necessary to generate a strong chlorinating species able to overcome the low reactivity of the substrate. Chloroalkanes can be prepared starting from available alcohols upon conversion of the hydroxyl group into a better leaving group. This can be carried out through sulfonylation, followed by replacement with a chloride salt. Alternatively, alcohols can be activated in situ with thionyl chloride [13], or with PPh₃/CCl₄ in the Appel reaction [14]. Finally, the radical C-H chlorination is a powerful tool for the synthesis of this class of compounds, since the direct substitution of hydrogen atoms in largely available alkanes is highly desirable. This last strategy relies on the formation of a carbon-centered radical, through hydrogen atom transfer (HAT), that is quenched, for instance, by one of the several available N-chloroimides 1a–e (Figure 2).

Despite the many applications of electrophilic and radical chlorination reactions, most of them require hazardous or toxic reagents, an inert atmosphere and harsh reaction conditions such as a high temperature or an excess of the chlorinating agent and initiator. These features are difficult to reconcile not only with the high chemo-, regio-, site- and stereoselectivity requirements in modern organic synthesis, but also with the present need for alternative and more sustainable processes.

In such a scenario, photoredox catalysis and electrocatalysis offer alternative ways to efficiently prepare a large variety of chlorine-containing molecules. Both approaches provide access to high-energy intermediates (e.g., radicals) under very mild reaction conditions, by promoting the movement of single electrons in redox processes. In photoredox catalysis, they are triggered by single-electron transfer (SET) events produced by a photocatalyst upon photoexcitation, while in electrocatalysis, they occur on the surface of metal electrodes connected to an external circuit. Recent reviews on the halogenation of organic molecules have shown how powerful the modern tools based on visible light [15,16] and electrochemistry are [17]. Even though photoredox catalysis and electrolysis present several differences, as recently shown by a comprehensive review by Rovis [18], they are conceptually very close. For such a reason, in this review, we collect the most recent progress in the preparation of chloro compounds by means of electrocatalysis and photoredox catalysis, to highlight similarities and differences between the two approaches.

2. Chlorination Enabled by Visible Light Photoredox Catalysis

Photoredox catalysis is based on the ability of photocatalysts, transition-metal complexes or organic dyes to undergo an electronic transition from the ground state (PC) to the excited state (PC*) upon irradiation with visible light. Once produced, the excited state can be quenched by reagents through single-electron transfer in either the oxidative or the reductive quenching cycle. In a reductive quenching cycle (Scheme 1a), SET from the donor (D) to PC* results in the ground state PC•− and the oxidized donor (D+). A subsequent SET from PC•− to an acceptor (A) results in the restoration of the photocatalyst and the reduced acceptor A−. Alternatively, in an oxidative quenching cycle (Scheme 1b), SET from PC* to an acceptor (A) results in the ground state PC•+ and the reduced acceptor (A−). A subsequent SET from a donor (D) to PC•+ regenerates the photocatalyst (PC), affording the oxidized donor (D+).

2.1. Photoredox Electrophilic Amplification of NCS and Related Reagents

In 2016, König demonstrated that photocatalysis is a valuable alternative way to activate NCS for the chlorination of arenes [19]. N-Chlorosuccinimide 1a is a common chlorination reagent but requires activation. Nevertheless, it is more reactive than N-chloroamines, as a consequence of the two carbonyl groups on the succinimide, which reduce the electron density on the nitrogen and increase the electrophilicity of the chlorine. NCS oxidation generates an even more reactive intermediate 2, because of the formation of a N-radical cation, possibly capable of giving S_EAr even on electron-poor arenes. The oxidation potential of 1a (E⁰(C/C+) = +1.10 V) does not match that of the photoexcited state of [Ru(bpy)₃]Cl₂ 3 (E⁰(Ru^II*/Ru^I) = +0.77 V). This means that its activation via the reductive quenching cycle is not possible. However, in an oxidative quenching cycle, the so-formed Ru(III) complex is a stronger oxidizer (E⁰(Ru^III/Ru^II) = +1.29 V). Indeed, the authors reported that, in the presence of (NH₄)₂S₂O₈ as the sacrificial oxidant, anisole 4 was selectively chlorinated at the para position, affording 4-chloroanisole 5 in 92% yield (Scheme 2). Without photocatalytic activation, they did not observe any conversion after three hours, and product 5 was produced in less than 5% yield after sixteen hours, despite the complete conversion of the substrate. Nevertheless, the electrophilic amplification was not sufficient to chlorinate electron-poor arenes such as acetanilide and acetophenone.

To unlock the direct electrophilic amplification of NCS without the use of sacrificial oxidizers, photocatalysts possessing a higher oxidizing photoexcited state are necessary. In 2019, Lamar et al. reported the use of organic dyes for the introduction of chlorine into aromatic substrates [20]. Choosing naphthalene 7 as the model reactant (E⁰(C/C+) = +1.64 V), they extensively screened organic photocatalysts, using white LED irradiation under aerobic conditions. Methylene green 6 ended up being the best one, producing 1-chloronaphthalene 8 in 72% yield. Oxygen was responsible for the regeneration of the ground-state photocatalyst, as demonstrated by the formation of H₂O₂. Nevertheless, in order to enhance the efficiency of the catalytic process, 10 mol% of (NH₄)₂S₂O₈ was also added, and the molarity of the reaction was increased from 0.1 M to 0.2 M in acetonitrile. Under the optimized conditions, a library of (hetero)aromatic molecules bearing electron-donating and withdrawing groups was chlorinated in good to excellent yields (selected products 9–12) (Scheme 3).

In an attempt to further evolve towards sustainable processes, Lamar and LeBlanc described two new catalytic systems based on FDA-approved food dyes (Fast Green FCF 13 and Brilliant Blue FCF 14) capable of functionalizing aromatics and heteroaromatics under mild conditions [21]. The methodology was exploited for the chlorination of known pharmaceuticals, such as lidocaine 15 and phenazone 16, as well as for the preparation of agrochemicals (nitrofungin 17 and chloroxylin 18) (Scheme 4). Additionally, the two organocatalysts were employed with two distinct N-chloro reagents, NCS 1a and DMDCH (1,3-dichloro-5,5-dimethylhydantoin 1e). Mechanistic investigations of the separate systems indicated two different modes of activation. Fast Green FCF 13 indeed acted as a light-promoted photoredox catalyst oxidizing NCS 1a, while Brilliant Blue FCF 14 served as a chlorine-transfer catalyst with DCDMH 1e, not necessarily involving visible light.

Similarly, other chlorinating reagents can undergo electrophilic amplification by exploiting visible light photoredox catalysis. Trichloroisocyanuric acid (TCCA, 1d), for example, which usually requires activation under acidic conditions, was used to chlorinate a large group of arenes 20a and heteroarenes 20b under air and irradiation with white light using the inexpensive brilliant green 19 [22]. Chlorinated derivatives 21a and 21b were isolated with yields between 35% and 96% (Scheme 5). Noteworthy, the reported methodology was highly efficient even on electron-poor rings, presented short reaction times, with some substrates undergoing chlorination in 5–10 min, and was scalable up to the gram scale with no erosion in the yield. In comparison to other chlorinating agents, such as Palau’chlor [23], the TCCA/BG system was proven as a valuable complementary method and, in many examples, provided a superior alternative for the chlorination of arenes and heteroarenes.

N-Chlorosuccinimide 1a has not only been used in S_EAr reactions for the production of chlorinated aromatic compounds, but also to quench radical (cation) intermediates. In 2020, Wu described a visible light chlorination of benzylic C-H bonds, employing NCS 1a and Fukuzumi’s dye (Acr⁺-Mes, 24) [24]. The reaction, conducted in dichloromethane under argon and blue LED irradiation and described in Scheme 6, converted a library of substituted toluenes 22, affording the corresponding benzyl chlorides 23 in moderate to good yields after 4 h. The authors conducted an extensive study on the reaction mechanism, based on quenching analysis of the photocatalyst. They proposed a mechanism involving the initial oxidation of NCS 1a by the photoexcited catalyst (Acr^•-Mes^·+), generating Acr^•-Mes, which can reduce a second equivalent of NCS to afford NCS• and regenerate the ground-state Acr⁺-Mes. Subsequently, NCS•− releases further Cl⁻ with the generation of succinimidyl radical 25, which abstracts a hydrogen atom from substrate 22 to give benzylic radical 26. NCS may react with radical 26, affording benzylic chloride 23 and succinimidyl radical 25, which results in chain propagation.

Under similar conditions, Mal and co-workers were able to develop a chlorinative cyclization of aryl alkynoates 27, to synthetize coumarins 29. Additionally, in this report, the mechanistic hypothesis was based on data coming from several techniques (Stern–Volmer, electrochemical and EPR analysis), together with on–off and radical trapping experiments. The reaction involved the oxidation of 27 by SET from the excited photocatalyst and dearomatization through cyclization to give spiro compound 28. Subsequent aryl migration and final HAT from the succinimidyl radical generated chloroisocoumarins 29 (Scheme 7).

2.2. Photoredox Chlorination Using Alternative Chlorinating Reagents

Despite the excellent results obtained with NCS 1a and related reagents, several groups have devoted great efforts towards the development of methodologies that rely on the use of alternative, more sustainable reagents capable of incorporating chlorine atoms into molecules. In 2017, Hu demonstrated that the catalytic system reported the year before by König and depicted in Scheme 2 was also effective in the activation of NaCl, a more sustainable chlorine source [25]. The oxidative quenching cycle of [Ru(bpy)₃]Cl₂ was triggered by using Na₂S₂O₈ and used to oxidize Cl⁻ to Cl⁺ or its equivalent (HClO), probably via a chlorine radical intermediate. The in situ generated electrophilic chlorine was then incorporated into several aromatic substrates, showing good yields and functional group tolerance. The authors demonstrated the potential utility of the reaction by preparing clofibrate 32a, 32b (precursor of glibencamide), and 32c (precursor of metoclopramide), (Scheme 8).

Halogenidric acids are valuable alternatives as halogen atom sources. This concept was demonstrated by König, who used the organophotocatalyst 4CzIPN 33 to chlorinate electron-rich arenes 30 under blue light and an oxygen atmosphere, with a mixture of HBr and HCl (Scheme 9) [26]. The catalyst has a dual role, as it is able to oxidize both bromide and chloride anions to their respective radical species. Bromine radicals recombine to bromine as the active brominating agent. The in situ formed bromo arenes are subsequently attacked by chlorine radicals, yielding the desired chlorinated compounds.

The chloride ion (introduced in the form of a salt or HCl) was also employed for electrophilic aromatic oxidative chlorination using biomimetic or bioinspired approaches. In particular, in 2016, Hering, inspired by flavin adenine dinucleotide (FAD)-dependent halogenases, replaced the biomolecules FAD and NADH₂ with the inexpensive organic dye riboflavin tetraacetate and anisyl alcohol as the sacrificial reducing agents [27]. By irradiating at 455 nm under air, an equimolar mixture of acetic and hydrochloric acids was effectively used to add a chlorine atom on aryl ethers and anilines. Similar conditions provided α-monochlorination on aceto- and propiophenone. More recently, Gulder and co-workers managed to combine photocatalysis and biocatalysis by using flavin mononucleotide (FMN) together with a vanadium-dependent haloperoxidase from Acaryochloris marina (AmVHPO) [28]. To avoid the use of sacrificial reductants, they used a redox-active buffer or even water as the electron source. In the reaction vessel, the H₂O₂ produced in situ by FMN was directly used by the peroxidase to halogenate electron-rich (hetero)aromatics using KCl.

Reaction solvents such as dichloromethane can act as chlorinating reagents as well. In 2019, Wu and Cui published a visible-light-mediated chlorination of silanes by accessing silyl radicals using photocatalytic Si-H activation promoted by neutral Eosin-Y 35. Under simple operative conditions, more than twenty silanes 36 were quantitatively converted in the corresponding chlorosilanes 37, widely used reagents in organic synthesis and in materials science (Scheme 10). Moreover, the strategy was also transferred from stirred reactor vessels to continuous-flow micro-tubing reactors. Under flow conditions, the chlorination was amenable to gram-scale production, and, even more importantly, the stepwise chlorination of di- and trichlorosilanes was successfully achieved in a highly selective manner. Regarding the mechanism, silyl radical 38a was generated by the HAT between the light-activated photocatalyst 35 (EY^*) and hydrosilane 36. The derived silyl radical subsequently abstracted the chlorine from CH₂Cl₂ to deliver the silyl chloride product 36 and the chloromethyl radical species (38b). A reverse hydrogen atom transfer (RHAT) between radical 37 and the catalyst (EY^·-H) regenerated the ground-state eosin Y 35 (EY) and produced chloromethane.

Finally, acyl chlorides can be used as well as chlorinating reagents. The chloroacylation of styrenes 39 was reported by Oh in 2020 employing fac-Ir(ppy)₃ 40 as the photocatalyst to cleave the C-Cl bond in the acyl chloride and produce Cl⁻ and a nucleophilic acyl radical 41a [29]. Upon acylation, benzyl radical 41b is oxidized to the corresponding cation 41c, thus regenerating the ground-state photocatalyst. The interception of chloride from 41c finally produces β-chloroketones 42 (Scheme 11). The authors pointed out the essential role of KHCO₃, which prevented the dehydrochlorination of the product by keeping a neutral pH during the course of the reaction.

2.3. Photoredox Chlorination of Nitrogen-Containing Aliphatic Molecules

Alkane chlorination is probably one of the first reactions that undergraduate students meet during their studies, and they also learn that, because of the similar BDEs, more than one C-H bond is usually chlorinated, producing mixtures of regioisomers difficult to separate. In addition to this, the standard radical halogenation requires high temperatures, radical initiators and hazardous reagents (Cl₂). Photoredox catalysis can provide for milder reaction conditions, the choice of safer and alternative chlorinating reagents and also site selectivity, for example, when the substrate contains nitrogen atoms. One classic example of how this high regioselectivity is achieved is represented by the Hofmann–Löffler–Freytag reaction [30], producing cyclic amines, typically pyrrolidines, by the thermal or photochemical decomposition of N-chloroamines in the presence of a strong acid (H₂SO₄ or CF₃CO₂H) (Scheme 12).

When N-chloroamine 43a is protonated, the homolytic cleavage of the N-Cl bond under the influence of heat, light or other initiators affords radical cation 43b and the chlorine radical. The ammonium radical undergoes a favored intramolecular 1,5-HAT to afford alkyl radical 43c which, capturing a chlorine radical, forms alkyl chloride 43d. The alkyl chloride is later cyclized under the influence of a base, producing pyrrolidine 43e. The key to the regioselectivity relies on the site-selective hydrogen abstraction occurring on the intermediate containing the N-centered radical, accessible through photoredox catalysis from N-Cl- or N-OX (X = H or R)-containing substrates.

In 2018, Leonori proved that the remote functionalization of cyclic and acyclic oximes 44 was indeed possible via a radical-promoted C-C and C-H bond cleavage cascade, by using the aforementioned Fukuzumi acridinium salt 24 [31]. The methodology provided access to a series of γ-chlorinated nitriles and imines, depending on the structure of the starting material. When cyclic oximes 44 were used, the formation of the corresponding imidyl radical 45a by the excited photocatalyst triggered C-C bond cleavage to open the ring, producing a terminal cyano group and a C-centered radical 45b. The radical was then quenched by NCS 1a to afford chlorinated nitrile 46. When linear oximes 47 were used instead, the imidyl radical 48a underwent 1,5-HAT, affording radical imine 48b. NCS 1a was again responsible for the chlorination, this time of both the radical and the imino group, producing stable γ-Cl,N-Cl imines 49 (Scheme 13).

Additionally, aliphatic amines can undergo a similar γ-chlorination. In 2020, Yu published a methodology to functionalize sulfonamides using sodium hypochlorite [32]. More than twenty sulfonamides 50 were successfully chlorinated with good to excellent yields, even at the gram scale (Scheme 14). Chlorination at the allylic or benzylic position, as well as across a cyclohexyl moiety, was not feasible. According to the mechanistic hypothesis, the reactive N-chloroamine 51a was produced directly in the reaction mixture and underwent single-electron reduction to N-centered radical 51b from photoexcited [Ru(bpy)₃]Cl₂. 1,5-HAT generated radical 51c reacting with a second equivalent of 51a, initiating a chain reaction (as suggested by the quantum yield measurement of 3.23), and producing product 52.

To conclude this section, one example of cyclic amine chlorination is presented. Aryl piperidines 53 were functionalized under aerobic conditions, in the presence of NCS 1a, affording 3,3-dichloro-2-hydroxy-piperidines 54 via two subsequent chlorinations of an intermediate enamine, followed by the capture of O₂ by an α-amino radical species (Scheme 15) [33].

To gain insights into the mechanism, the authors ran control experiments. When 2,2,6,6-tetramethylpiperidinooxy (TEMPO) and 2,6-di-tert-butyl-4-methylphenol (BHT) were added to the reaction mixture as radical scavengers, the formation of the product was completely suppressed, indicating that a free radical process was involved in the transformation. Additionally, when the model reaction was performed under a molecular oxygen (¹⁸O₂) atmosphere, the ¹⁸O-labeled 54 was isolated, confirming that the oxygen in the product was coming from the molecular oxygen in the air.

3. Chlorination Enabled by Electrocatalysis

Electrocatalysis pivots on the use of electrical energy to apply a potential across a pair of electrodes immersed into a solution of the components to be electrolyzed. In an electrolysis experiment, four key features are present: (1) an anodic oxidation, (2) a cathodic reduction, (3) conservation of the charge in the solution and (4) the presence of a soluble supporting electrolyte to ensure the low electrical resistance of the solution. Common to many organic electrochemical transformations, only one of the two electrodes (working electrode) generates a useful product, while on the counter electrode, a non-productive reaction takes place. For net-oxidative processes, proton reduction to H₂ at the counter electrode is the most common non-productive reaction. For net-reductive reactions, sacrificial oxidation of an amine or the anode itself (e.g., Zn, Mg, Al, Cu) is commonly encountered. In this section, there are many examples in which both electrodes produce useful intermediates. Some common mechanisms in electrolysis are depicted in Scheme 16 and are those herein presented.

One major difference between the two approaches is that while in photoredox catalysis, the redox events take place on the same site (i.e., the photocatalyst), in electrocatalysis, they occur on two distinct places (cathode and anode). This feature makes it possible to physically separate the two events, when necessary, via the use of a divided cell, provided that a method to accommodate the movement of ions from one half-cell to the other is employed. This is usually achieved by interfacing the two half-cells with a separator, such as a sintered-glass frit, a porous ceramic, a porous polymer sheet or a semipermeable ion-selective membrane (e.g., Nafion). The other fundamental difference is that, in direct electrolysis, redox events occur at the interface between the electrode and solution, which can produce a localized, high concentration of radicals. In contrast, redox chemistry in photoredox catalysis occurs in the solution and provides a low radical concentration. As a result, the coupling of two transient radicals is feasible in electrochemistry if they are generated at the same electrode, whereas this is more challenging in photoredox catalysis.

3.1. Electrocatalyzed Dichlorination of Alkenes

Just like photoredox catalysis is employed (in combination with NCS 1a or chloride salts) to replace Cl₂ in the electrophilic chlorination of arenes, electrocatalysis is a valuable alternative to achieve alkene chlorination under mild and more sustainable reaction conditions. In 2017, Lin described a manganese-catalyzed alkene dichlorination, with MgCl₂ as the chlorine source, producing H₂ and Mg(OAc)₂ as the sole by-products (Scheme 17) [34]. Optimizing the dichlorination of indene 55, the authors observed that the direct electrolysis of LiCl on the Pt electrode was not satisfying, giving poor yields of 56 and no syn:anti selectivity. Therefore, a redox-active metal was added to impart kinetic control over the difunctionalization, and they obtained good results, in terms of diastereoselectivity, with Mn(OTf)₂. The high diastereoselectivity came from a combination between the steric and electronic properties of the chlorine transfer agent Mn(III)−Cl 57. Further improvements were observed when LiCl was replaced with MgCl₂. The reaction was extended to several styrenes and alkenes, producing the corresponding dichloro derivatives in good yields and diastereoselectivity. Moreover, oxidatively sensitive functional groups (alcohols, aldehydes, amines) were tolerated.

An interesting alternative for the preparation of vicinal dichlorides 62 was recently presented by Morandi. The methodology involved an electronically assisted shuttle paradigm, with inexpensive 1,2-dichloroethane (DCE) 59 used as the chlorine source [35]. In the cell, two consecutive cathodic reductions of DCE are leveraged to produce chloride anions in the reaction medium, releasing ethene 63. Following this, the oxidation of Cl⁻ to Cl⁺ affords an intermediate chloronium ion 61, which is converted in the final product 62 via ring opening from Cl⁻ in an overall redox-neutral process (Scheme 18). Optimal reaction conditions featured graphite electrodes in an Et₄NBF₄ solution (0.1 M in DCE), at 50 °C. MnCl₂ was used as the mediator. In many cases, a solution of 1,1,1,2-tetrachloroethane 60 in MeCN was used instead of neat DCE 59 to obtain better yields. Under optimized conditions, thirty alkenes were efficiently dichlorinated. Noteworthy, soil contaminated with lindane (gamma-hexachlorocyclohexane) 63 could also be used for the synthesis of 1,2-dichlorododecane 64.

In 2021, Hilt published the first electrochemical cis-dichlorination of alkenes [36]. The key reagent in such a transformation was PhSeCl 67, which has a double role (Scheme 19). On one side, it can be oxidized at the anode, in the presence of tetrabutylammonium chloride (TBAC, 66), to form PhSeCl₃ 67a, but on the other side, it can undergo a non-catalyzed anti-chloroselenation of cyclohexene 68, affording intermediate 67b. PhSeCl₃, being a strong electrophile, may induce a nucleophilic chlorination on intermediate 67b, with inversion of the configuration, producing cis dichlorocyclohexane 69. The reaction could be applied to cyclic as well as linear olefins, also bearing hydroxyl and acethoxyl moieties. Lower yields were observed on styrenes, and the authors attributed the result to the redox lability of the substrates.

3.2. Electrocatalyzed Heterofunctionalization of Alkenes

After having developed the electrocatalyzed dichlorination of olefins 70, Lin applied an anodically coupled electrolysis to merge two distinct oxidative events and therefore make the heterofunctionalization of alkenes possible. In particular, he developed a protocol to incorporate the trifluoromethyl moiety using Langoi’s reagent 71 (CF₃SO₂Na) [37]. This salt is directly oxidized on the graphitic carbon to the electrophilic trifluoromethyl radical 71a (anodic event A), capable of reacting with electron-rich alkenes, affording the alkyl radical 71b. Concomitant oxidation of Cl⁻, assisted by Mn(OAc)₂ (anodic event B), generates the radical chlorinated reagent Mn(III)-Cl 72, able to quench intermediate 71b, giving the final product 73 (Scheme 20). The two events could be merged thanks to the different natures of the two radical species. The alkene is indeed more likely to react with a transient free radical, such as 71a, than with the persistent metal-based radical 71c. The authors investigated the scope of the reaction, and several different styrenes and alkenes were efficiently trifluoromethylchlorinated (selected products 74–76).

The same group later applied a similar concept to achieve alkene chloro-alkylation by anodic generation of a carbon-centered radical from malonitrile through proton-coupled electron transfer [38].

When anodic and cathodic events are combined, the chloro-chalcogenation of alkenes is possible (Scheme 21). This was demonstrated by Chen, who reported a cobalt-catalyzed electrochemical oxychlorination of styrenes 77 to α-chloroacetophenones 78 [39]. To obtain reproducibility, reticulated vitreous carbon (RVC) electrodes were used. Despite prolonging the reaction times, due to the lower current density, they provided better results. Like manganese in the previous examples, cobalt was fundamental in mediating the anodic oxidation of MgCl₂ to a Co(III)-Cl radical species 79a. Oxygen was directly reduced on the cathode to a persistent radical superoxide ion, capable of quenching the transient benzyl radical 79b, affording alcohol 79c. β-Chloroketone 80 was formed by the oxidation of 79c, but the author did not investigate whether the process was electrochemical or not.

With excellent atom economy, Lei used sulfonyl chlorides 81 as bifunctional reagents to produce β-chloro(vinyl)sulfones 83 via a redox-neutral chlorosulfonylation of alkenes and alkynes 82, Scheme 22 [40]. In the designed reaction, the cathodic reduction of sulfonyl chloride 81, followed by the cleavage of the S-Cl bonds, produces Cl⁻ and the sulfur-centered radical 84a, which reacts with the substrate to give intermediate 84b. Chloride anions are not wasted, but oxidized at the anode, with the assistance of MnCl₂, to give Mn(III)-Cl 72. Final radical coupling produces the chlorinated product 83, regenerating the manganese catalyst. The reaction was extended to several styrenes and aryl acetylenes. Moreover, a radical ene–yne cyclization was attempted on propargyl amine 85, and pyrrolidine 86 was isolated in 50% yield.

3.3. Electrocatalyzed Arene Chorination

As highlighted in the Introduction, photocatalysis and electrochemistry are closely related, despite some differences: one among all seems to be the choice of the substrates. Indeed, photoredox catalysis has mostly been used as an alternative mode of activation for aromatic electrophilic substitution, thus affording chloroarenes, with few methodologies applied to alkenes. The opposite trend appears when looking into electrochemistry. The methodology reported thus far in this review is the addition of one of two chlorine atoms to olefins. Nevertheless, there are some recent examples of aromatic chlorination performed under electrochemical conditions.

For example, in 2021, Fang and Guo reported a C-5 selective chlorination of 8-aminoquinoline 87, catalyzed by copper acetate [41]. Dichloromethane was used as the source for chlorine radicals, being formed at the anode by the oxidation of small amounts of Cl⁻ released in the electrolytic cell upon DCM heterolysis, according to the authors’ hypothesis. Aminoquinoline 87 is able to coordinate copper by the two nitrogen atoms, and intermediate 88a can evolve towards product 89 through two different catalytic cycles. One possibility involves the direct addition of the chlorine radical to produce chlorinated intermediate 88b, with the concomitant reduction of the metal center to Cu(I). Anodic oxidation produces arenium ion 88d (Scheme 23, path B). Alternatively, the same species can be formed from complex 88a after anodic oxidation to radical cation 88c and the subsequent interception of the chlorine radical. In such a catalytic cycle, there is no change in the oxidation state of copper. Wheland intermediate 88c evolves to the final product 89, regenerating the metal catalyst. Protons are reduced at the cathode to equilibrate the overall anodic process. A fairly high current of 100 mA was essential for obtaining satisfying results.

The substrate scope of the reaction was explored, and moderate to excellent yields were obtained when substituted benzamides bearing electron-donating or electron-withdrawing groups were employed. Electron-rich substrates exhibited higher reactivity compared with electron-deficient ones. The different positions of the substituents on the benzene ring did not remarkably affect the reaction efficiency. In addition, alkyl amides were smoothly chlorinated in good yields. The preparation of 89 was also scaled up to the gram scale using a flow electrolytic cell, allowing the synthesis of 1.19 g of product 89 in 24 h (91.2% yield).

8-Aminoquinolines are peculiar heteroaromatic rings, and due to the distance between the two nitrogen atoms, they have been extensively used as the directing group for the ortho-functionalization of benzoic acids. The process takes advantage of the proximity between the dicoordinated metal center and the C-H bond, which can undergo metalation, producing a reactive metallacycle. In such a context, palladium is among the most used metals, and an electrochemical version of the process has been reported for the synthesis of ortho-chlorobenzamides, by Kakiuchi and co-workers [42]. After the initial screening of the reaction conditions, they observed that unsubstituted 8-aminoquinoline 87 was not the ideal bidentate ligand, since it could undergo chlorination itself, probably by a mechanism similar to the one later exploited by Fang. To avoid such a side reaction, a 5,7-dichloro-8-quinolinyl group was chosen as the directing unit. After testing several conditions, they optimized the process which occurred at 90 °C, with a 100 mA current, in the presence of 10 mol% PdCl₂. Crucial for the success of the reaction was the setup, consisting of a divided cell, which prevented the reduction of palladium at the cathode. Cl⁺ or synthetically equivalent species were generated in the anodic semi-cell, with chloride coming from the cathodic reduction of HCl. The reaction was applied to several substituted benzamides 90, and the corresponding chlorinated products 91 were recovered in good to excellent yields (Scheme 24). In all cases, dichlorination occurred to some extent, producing around 10% of the undesired product 92, with the only exception of meta-bromo and -CF₃ substituents, which strongly deactivated the aromatic ring toward a second substitution. Finally, palladium-catalyzed C−H chlorination under anodic oxidation conditions was applied to the convergent synthesis of vismodegib 95 from pyridine 93 and benzamide 94.

Prior to the alkene dichlorination using dichloroethane as the source of Cl⁺ species reported by Morandi and previously described in this review, Jiao and co-workers explored bifunctional electrocatalysis as a possible way to perform aromatic chlorination, decomposing DCE 59 to vinyl chloride 96 and hydrochloric acid [43]. They initially investigated the viability of the electrolytic dehydrochlorination of dichloroethane 59, and they observed that a 100 mA current in a solution of n-Bu₄NOH in DCE 59 led to the formation of vinyl chloride 96 (48%), ethylene 63 (46%) and HCl. Since chloride oxidation to Cl⁺ could be coupled at the anode, they tested and demonstrated the possibility of a one-pot aromatic chlorination. Under optimized conditions, they were indeed able to efficiently chlorinate various anilides, bioactive compounds and heterocycles (selected examples 97a–c) (Scheme 25).

In the Introduction of this review, we highlighted how the chlorination reaction is among the most straightforward approaches to introduce chloride into molecules, but it comes with several drawbacks when excess chlorine gas is used, such as handling Cl₂ itself and the highly corrosive HCl that is generated. Electrocatalysis could be helpful in this scenario, assuming that chlorine is generated in situ. First, the rate of Cl₂ generation could be regulated by electrochemical parameters, leading to minimum escape from the reaction and facilitating reaction handling under ambient conditions. Second, the presence of an excess of the chloride source could be feasible, avoiding the generation of HCl. In 2021, Cheng et al. published the first example of an electrochemical aromatic chlorination with in situ generated chlorine. The authors identified trichloroacetonitrile 98 as the optimal Cl₂ source, being decomposed on a graphite felt cathode, in the presence of tetraethylammonium chloride as the supporting electrolyte. In optimizing the chlorination of para-chloroacetamide 99, other solvents were tested (CCl₄, CHCl₃ and DCE), but only traces of product 100 were observed. Moreover, without CCl₃CN, no product was formed, thus excluding a mechanism involving the direct oxidation of the electrolyte. The existence of chlorine was confirmed by the tetrachlorination of N-tosyl diallylamine, as previously reported by Fu [34], which excluded a radical pathway. In addition to this, they demonstrated that CCl₃CN 98 could donate all three chlorine atoms, producing acetonitrile. Product 100 was in fact produced in 91% yield after three hours when 0.67 equivalents of CCl₃CN 98 were used. In addition to this, the reaction could also be scaled up, and 153 g (1 mole) of para-fluoroacetamide 101 gave product 102 in 90% NMR yield and 73% isolated yield (Scheme 26).

3.4. Electrocatalyzed Miscellaneous Reactions

In the section dedicated to the photochemical functionalization of nitrogen-containing substrates, the C-C bond cleavage on cyclic oxime esters was presented. A similar strategy can be used in electrolysis to access chlorinated products from alcohols, again via ring opening. The concept was presented for the first time by Browne and Morrill, who reported in 2019 the electrochemical deconstructive chlorination of cycloalkanols 103, to synthesize γ- and β-chloroketones 105 [44]. Additionally, they demonstrated that by employing micro-reactor technology and a recirculating flow, the method could be performed at the gram scale, with incorporated purification. The reaction pivots on the formation of alkoxy radicals directly by O-H homolysis, not a trivial process to achieve under mild reaction conditions, due to the high BDE of RO-H bonds (~105 kcal/mol). The formation of radical 104a does not occur on the electrode surface, but by means of the Mn(III)-Cl species 72, which is also used as an active chlorinating species. The formation of the alkoxy radical also produces HCl, with protons being reduced to hydrogen on the cathode surface. In a similar process to that described by Leonori with oxime esters, alkyl radical 104b is formed upon β-scission from alkoxy radical 104a and, similar to other electrochemical processes, undergoes chlorination by a second equivalent of Mn(III)-Cl 72 (Scheme 27). The full scope of the electrochemical process was explored starting with the deconstructive chlorination of cyclobutanols. It has been observed that 1-arylcyclobutan-1-ols containing aromatic systems with electron-donating groups at the ortho or para positions underwent decomposition. This instability was attributed to the ionization of the C–OH bond in the presence of Brønsted and/or Lewis acids, forming unproductive stabilized carbocations. In such cases, the issue was overcome by employing a syringe pump addition of the substrate over two hours and using tetrabutylammonium acetate as the supporting electrolyte. With a choice of two suitable reaction conditions in hand, they converted a variety of 1-arylcyclobutan-1-ols 103 into the corresponding γ-chlorinated ketones 105 in good to excellent isolated yields. In addition, a representative selection of 1-aryl- and 1-alkylcyclopropan-1-ols 103 could be readily transformed to β-chlorinated ketones 105. Furthermore, to demonstrate scalability, the batch process was translated to a recirculating flow electrochemical setup. Not only did switching from batch to flow allow for a gram-scale synthesis, but also, due to the decreased distance between the electrodes in the flow, the supporting electrolyte was not required, thus making the overall process even more efficient from a sustainability perspective.

Electrochemistry can be exploited not only to achieve synthetically relevant ring-opening processes, but also for the opposite purpose. Indeed, if a nucleophilic group is tethered to the substrate being chlorinated, it is not surprising that a cyclization can be promoted by a careful design of the reaction. One example, from 2020, is the electrochemical dearomative chlorocyclization on tryptamine and tryptophols 106, leading to biologically and pharmaceutically relevant hexahydropyrroloindolines 108 [45]. Without the help of any metal catalyst, LiCl is oxidized at the anode to an electrophilic Cl⁺ species, capable of giving an electrophilic addition to the electron-rich indolic substrate 106, leading to the dearomatized chloronium intermediate 107. The tethered hydroxyl or acetylated amino group is then able, by nucleophilic ring opening, to afford the final tricyclic product 108. Hydrogen evolution at the cathode, from the supporting acetic acid, ensures redox neutrality (Scheme 28). The method was developed more specifically to obtain brominated derivatives; for such a reason, the scope was restricted to seven chlorinated products. The authors hypothesized the direct anodic oxidation of chloride ions by comparing the yield of chloro (spanning from 50 to 86%) and bromo derivatives (higher than 90%). According to Lei and co-workers, the result was consistent with the higher oxidation potential for Cl⁻.

To conclude this section on electrolysis, three examples of non-aromatic C(sp²)-H chlorination, all from 2021, are reported. The first one, by Liu, is a chlorination of electron-deficient C-H bonds in quinones 109, coumarins 110 1,3-diketones 111 and quinoxalines 112 using six equivalents of HCl [46]. The reaction occurred under simple experimental conditions, in an undivided cell, with a graphite felt anode and a platinum cathode, using a solution of Et₃NBF₄ in acetonitrile, affording chlorinated products 113–116 with moderate to good yields (Scheme 29). Despite the non-aromatic nature of both the starting material and products, the reaction should actually be considered an aromatic chlorination, at least according to the mechanism hypothesized by the authors to explain the chlorination of benzoquinone 109. Cathodic reduction of benzoquinone 109 may easily generate the corresponding aromatic hydroquinone 117, which undergoes electrophilic chlorination with chlorine from the anodic oxidation of Cl⁻. Anodic oxidation of the chlorinated hydroquinone 118 gives the desired product. This was not indicated by the authors, but H₂ should be generated at the cathode to ensure redox neutrality.

Liu et al. prepared a series of 3-chlorochromones 120 by the electrochemical chlorination of enaminones 119 [47]. The reaction was based on a cascade process, initiated by the formation of chloronium intermediate 121a upon the addition of Cl₂ to the electron-rich enaminone group. Chlorine is electrochemically generated on the anode from the oxidation of NaCl. The tethered phenolic moiety opens the chloronium ion by intramolecular formation of a new C-O bond, affording α-chloro emiaminal ether 121b. Elimination of Me₂NH₂Cl produces the final chromone. Additionally, in such a process, protons were reduced on the cathode (Scheme 30). The scope was restricted to eight examples, but in the same paper, they also applied the reaction to the synthesis of bromo- and iodochromones. The presented method was also scaled up, although a dramatic drop in the yield was already observed on a two-millimole scale.

As the last example, Morrill et al. published a method for the oxidative Z-selective C(sp²)–H chlorination of tertiary acrylamides 122, providing access to a broad range of synthetically useful Z-β-chloroacrylamides 123 in good yields, Scheme 31 [48]. They optimized a reaction based on MgCl₂ as both the chloride source and electrolyte in MeCN:AcOH (7:1) using galvanostatic conditions (i = 10 mA), a graphite anode and a platinum cathode at 25 °C for 2 h under N₂. More than 20 products were isolated in good to excellent yields (average yield 73%). They observed that several substituents could be incorporated into the C(2)-aromatic unit, including electron-donating, electron-withdrawing and halogen substituents. Additionally, the introduction of sterically demanding groups into the acrylamide starting materials 122 did not negatively impact upon product 123 formation (Scheme 23). The reaction should initiate with anodic chloride oxidation to form a chlorine radical, which regioselectively adds to the electron-deficient acrylamide 122 to furnish the tertiary C-centered radical 124a. One possibility is then to have a parallel paired electrolysis, with 124a being intercepted by a chlorine radical (path A) to form the dichlorinated intermediate 124b. The final loss of HCl should give acrylamide 122. Alternatively, in a sequential paired electrolysis, radical 124a could be further oxidized to form carbocation 124b, with subsequent deprotonation providing access to the observed product 123, or it could be intercepted by Cl⁻ to form dichlorinated compound 124c, which could also generate 123 in such a hypothesis via hydrochloric acid loss. In all cases, hydrogen gas should be generated at the cathode.

To demonstrate product utility, the authors succeeded in a palladium-catalyzed Suzuki cross-coupling and in amide hydrolysis. In addition to this, to demonstrate scalability, the batch process was translated to a flow electrochemical setup, employing a syringe pump in combination with a commercially available Ammonite8 flow electro-reactor.

4. Conclusions

At this point, we would like to stress, one more time, how profoundly the recent developments in photo- and electrochemistry are changing the way we synthetize molecules. Every day, they are disclosing new intriguing retrosynthetic disconnections that can be helpful for the late-stage, problematic functionalization of complex molecular structures. Additionally, and maybe more importantly, they are providing us with more sustainable and safer alternatives to perform “easier” transformations with respect to traditional methods, at least at a laboratory scale, generating radicals or cations under mild and controlled reaction conditions. The formation of carbon–chlorine bonds, via direct chlorination, is a perfect example of this improvement. At the same time, we should point out some differences between these two techniques, such as the preference for aromatic chlorination in photocatalysis versus the difunctionalization of alkenes in electrocatalysis. Photoredox catalysis might present some limitations in the design of the reaction. First, there are plenty of photocatalysts that can be tested before finding the best one. Even though it can be helpful to look at redox potentials, these sometimes fail, due to the reaction conditions being far from the ideal ones used in cyclic voltammetry. Second, most of the visible-light-promoted aromatic substitutions presented here need NCS or a related reagent, much safer than Cl₂, but less efficient in an atom economy perspective, producing a stoichiometric amount of side products that must be separated from the main product. Electrochemical chlorination usually takes advantage of chlorides or hydrochloric acid, which results in low- to zero-waste processes. Moreover, many of them are successfully scaled up, usually with the help of flow reactors. A limitation of applying electrocatalysis is the experimental setup because it requires more specialized devices to control the intensity of the current or the voltage, and expensive electrodes. This, at the moment, makes it easier for many research groups to start working on photoredox catalysis, for example, with household lights or LED strips and irradiating whatever the reaction vessel elected. In conclusion, photoredox catalysis and electrocatalysis are surely taking a leading role in the preparation of organochlorine compounds, and it is expected that even more efficient methodologies will be developed in the near future, potentially capable of overcoming the present issues with selectivity or scalability, thus making it possible to introduce such techniques to the chemical industry.