Organic Photoredox Reactions in Two‐Molecule Photoredox System

Using our recent relevant results, this account shows the featured reactivities of two‐molecule photoredox systems compared to one‐molecule photoredox systems. The low efficiency of electron transfer processes, such as photoinduced and back‐electron transfer, in the two‐molecule photoredox system, furnishes unique products through different pathways. The facile replacement of photoredox catalysts with appropriate oxidation/reduction potentials in this system provides valuable insights into photoredox reactions.


Introduction
Organic photoredox reactions are versatile, powerful, and environmentally friendly tools for transforming organic compounds since the 1980s, owing to the mild generation of unique highly reactive intermediates such as radical ions or radicals, which were previously developed by Pac, [1] Mizuno, [2] Fukuzumi, [3] Okada, [4] Arnold, [5] Mariano, [6] Albini, [7] Griesbeck, [8] et.al.However, the substrate, reagent, and product scopes were considerably limited, and the control of reactive intermediates was not efficient at this time.Thus, the synthetic application of photoredox reactions to organic reactions was unsuccessful.This decreased the number of published reports in this field in the latter half of the 1990s.[11][12] However, the overall utility of radical reactions is limited by the toxicity of these substrates and reagents, difficulties associated with the removal of tin-containing byproducts (Bu 3 SnX), and harsh conditions (high temperatures).Organic photoredox reactions provide a cleaner and more environmentally friendly process for generating radicals under mild conditions.The knowledge of radicals, desire for environmentally friendly organic reactions, and new technologies such as LED and flow reactors were refocused on organic photoredox reactions in the 2010s by MacMillan et al., [13][14][15] and the number of studies on organic photoredox reactions increased rapidly (renaissance of organic photoredox reactions [16] ).
The organic photoredox reaction systems are broadly classified into "one-molecule photoredox system" (Scheme 1a) and "two-molecule photoredox system" (Scheme 1b).The former efficiently oxidizes and reduces substrates under light irradiation using a photoredox catalyst (PC) and exhibits only one PC redox cycle.One-molecule photoredox catalysts, such as transition metals (e. g., Ru and Ir), [13] 1,2,3,5tetrakis(carbazol-9-yl)-4,6-dicyanobenzene (4CzIPN), [14] and Fukuzumi catalysts, [15] are used in this system, leading to efficient photoredox reactions even at low PC concentrations (< 1 mol%), owing to one redox cycle and the long lifetimes of both the excited and reduced states of PC.However, these PCs are expensive and exhibit slightly low stability in water, acids, and bases.In addition, dramatic changes in the oxidation and reduction abilities of these PCs are difficult to achieve.By contrast, the latter, which was previously called "redox-photosensitized reaction" by Pac, [1] has two redox cycles of PCs.The radical cation of the electron-donor (ED) and radical anion of the electron-acceptor (EA) function as the oxidant and reductant, respectively, through photoinduced electron transfer (PET). [1,17]This two-molecule photoredox catalyst, a combination of inexpensive phenanthrene (Phen) and 1,4-dicyanobenzene (1,4-DCB), leads to relatively inefficient electron transfer processes, such as PET and back electron transfer (BET) because of the two redox cycles and short lifetimes of the excited state of PC, radical cation from ED, and radical anion from EA.This requires high concentrations of PCs (2.5-50 mol%) and a long irradiation time.However, the low efficiency of two-molecule photoredox systems results in unique products via different pathways.In addition, the replacement of ED and EA with other PCs having oxidation and reduction potentials was straightforward.This account focuses on some features and advantages of the two-molecule photoredox system based on our recent results.

Photoinduced Decarboxylative Radical Reactions of Carboxylic Acids in Two-Molecule Photoredox System
Recently, the photoinduced decarboxylation of aliphatic carboxylic acids using a photoredox catalyst was established as an effective and more environmentally friendly method for generating alkyl radicals under milder conditions such as room temperature than typical method using alkyl halides, AIBN, and Bu 3 SnH with heating. [18]This is because carboxylic acids are ubiquitous in nature and easy to handle; decarboxylation releases the nonflammable and nontoxic product CO 2 under mild conditions; and light is a traceless reagent.Although the photoinduced decarboxylative radical reactions of carboxylic acids in the presence of various electron acceptors, such as iminium salts, [19] acridine, [20] benzopyridines, [21] phthalimide, [22] 1-cyanonaphthalene, [23] 1,4dicyanonaphthalene, [24] and tetracyanobenzene, [25] were well described in the 1980s and 1990s, the used carboxylic acids were limited to aryl, vinyl, and aryloxy acetic acids except one case. [21]Furthermore, addition reactions between the generated alkyl radicals and radical anions of the electron acceptors determine the product.In 2007, the first report of direct decarboxylative radical generation from aliphatic carboxylic acid 1 using two-molecule photoredox catalysts, Phen and 1,4-DCB, under ultraviolet (UV) irradiation (mainly 313 nm, 100 W high-pressure mercury lamp through a Pyrex filter (λ > 280 nm)) was published by us (Scheme 2). [26]The process is initiated by PET from Phen (ED) to 1,4-DCB (EA), generating the radical cation Phen (ED), which oxidizes carboxylate ions 2 to afford carboxy radicals 3, which readily decarboxylate generating the corresponding alkyl radicals.The generated alkyl radicals then react with various radical acceptors to form the respective products in high yields without the abovementioned side reactions (in addition to the radical anions of the electron acceptors).For example, photoinduced decarboxylative radical reactions conducted using two-molecule photoredox catalysts are useful in reduction, [26] deuteration, [27] addition to C=C, [28] C=N, [29] and C=O [30] double bonds, addition to aromatic rings, [31] peptide modification, [32] cyclization, [28a,33] ring expansion, [30] dicyanoarene substitution, [34] polymerization, [35] and carbanion generation [36] under mild conditions.The ED replacement of Phen with dibenzo[g,p]chrysene (DBC) [37] or EA replacement of 1,4-DCB with 9,10-dicyanoanthracene (DCA) [38] or 9cyano-10-methoxycarbonylanthracene (CMA) [39] enables the usage of visible light (18 W blue LED, 405 nm) as the excitation source.Similar visible light-induced decarboxylative radical reactions in a one-molecule photoredox system using Ir, Fukuzumi, or 4CzIPN catalysts were reported in 2014 (Ir [40] and Fukuzumi [41] catalysts) and 2016 (4CzIPN [42] ).In the following section, the featured results of photoinduced decarboxylation in a two-molecule photoredox system are described.Intermolecular radical addition to alkenes is considered a powerful and facile approach for CÀ C bond formation and alkyl radicals generated from carboxylic acids through photoinduced decarboxylation can attack alkenes to furnish adducts.28a] In the case of high alkene concentrations (> 1 M), photoinduced decarboxylative radical polymerization with acrylic esters and acrylamides proceeded, and the mass spectrum of the obtained polymer in Scheme 4 shows smooth radical polymerization by this method. [35]This is because the low efficiency of PET in the two-molecule photoredox system led to a low concentration of alkyl radicals.When a onemolecule photoredox catalyst such as Ir or Fukuzumi was used, a short oligomer was obtained because the high efficiency of BET prevented radical polymerization and promoted termination.By contrast, using a two-molecule photoredox catalyst decreased the efficiency of BET (termination) in forming a high-molecular-weight polymer.Thus, the low efficiency of PET and BET in the two-molecule photoredox system resulted in successful radical polymerization via the photoinduced decarboxylation of carboxylic acids.

Yasuharu Yoshimi received his
Additionally, two-molecule photoredox system with Phen and 1,3-dicyanobenzene (1,3-DCB) produced the unique ringconstrained γ-amino acids 5 in moderate yields through the sequential intermolecular radical addition and reductive radical cyclization of phenylalanine and tyrosine derivatives 4 with alkenes such as acrylamides and acrylic esters via photoinduced decarboxylation as reported by our group (Scheme 5a). [43]The radical generated by the photoinduced decarboxylation of 4 was intermolecularly added to the alkene, and the radical 6 intramolecularly attacked the benzene ring to furnish 5 via BET and protonation because the low efficiency of BET to radical 6 in the two-molecule photoredox system promoted this radical cyclization (Scheme 5b).However, using a onemolecule photoredox catalyst, such as Ir and Fukuzumi, produced only alkene adducts without forming 5 because of the efficient BET to radical 6.This sequential reaction can also be applied to the direct tethering of dipeptides to yield unique ring-constrained tetrapeptides (Scheme 5c).
A typical example of a two-molecule photoredox system is the photoinduced decarboxylation of benzoic acids 7. [39,44] In contrast to the photoinduced decarboxylation of aliphatic

P e r s o n a l A c c o u n t T H E C H E M I C A L R E C O R D
carboxylic acids, attempts for the direct photoinduced decarboxylation of benzoic acids 7 to generate aryl radicals using photoredox catalysts failed because the rate (k = 2.0×10 6 s À 1 ) of decarboxylation of aryl carboxy radical 9 (Scheme 6) is lower than that (primary: k = 2.0×10 9 s À 1 , secondary: k = 6.5×10 9 s À 1 , tertiary: k = 11.0×10 9 s À 1 ) of alkyl carboxy radical 3.However, only our two-molecule photoredox system functioned efficiently for the direct photoinduced decarboxylation of 7 using biphenyl (BP)/1,4-dicyanonaphthalene (DCN) or BP/DCA or BP/CMA under UV-or visible-light irradiation with mild heating (30 °C), respectively (entries 1-3, Table 1).The photoinduced decarboxylation of 7 using the Ir and Fukuzumi catalysts was unsuccessful, and the starting material was recovered entirely (entries 4-6) because of the low rate of decarboxylation of the aryl carboxy radical 9 and high rate of BET between 9 and the reductant part of the Ir and Fukuzumi catalysts (Scheme 6).By contrast, the low BET efficiency of the two-molecule photoredox system led to successful decarboxylation of benzoic acids bearing various substituents with slight heating.The resulting aryl radicals reacted with electron-deficient alkenes, diboron, and acetonitrile to furnish adducts 8, aryl boronate esters, and the reduction product, respectively.A similar trend was observed for the UV-or visible-lightinduced decarboxylation of primary carboxylic acids, such as L-glutamic acid methyl esters 10 a at the side chain with ethyl acrylate 11 A (Table 2). [45]In addition, our two-molecule photoredox system is the only one that allows the photoinduced decarboxylation of 10 a because of the low rate of decarboxylation of the primary carboxy radical, similar to that of benzoic acids.The high oxidation potential of the primary carboxylate ion of 10 a and weak donating ability of the generated primary alkyl radical, which lowers the rate of radical addition to 11 A, decreased the yield of 12aA; however, sidechain modification of 10 a without racemization was achieved in the two-molecule photoredox system.We connected Lglutamic/aspartic acid methyl esters 10 a,b and alkenes 11B-E at the side chain using this system to yield side-chain functionalized aspartic/glutamic acid derivatives 12 under mild conditions (Table 3).
One of the advantages of a two-molecule photoredox system is the facile replacement of the photocatalysts with other photocatalysts to modulate the oxidation/reduction potential of the substrate.This feature demonstrates the selective photo-induced decarboxylation of N-Boc L-glutamic acid 13, which has two different carboxyl groups (Scheme 7). [37]The first photo-induced decarboxylation of 13 using DBC and 1,4-DCB with acrylonitrile in the presence of 2 equiv. of KOH occurred only at the α-position carboxy group to furnish adduct 14 in a high yield (81 %).This is Scheme 6. Proposed mechanism of the photoinduced decarboxylative radical reactions of 7.

Table 1. Photoinduced decarboxylative radical addition of benzoic acid 7
Table 2. Photoinduced decarboxylative radical addition of primary carboxylic acid such as 10 a. SCE in acetonitrile, oxidation potential of carboxylate ion at the side-chain: + 1.25 V vs. SCE in acetonitrile) (Scheme 7a).

P e r s o n a l A c c o u n t T H E C H E M I C
The second photoinduced decarboxylation of the obtained product 14 with methyl methacrylate using Phen and CMA afforded adduct 15 in moderate yield (Scheme 7b).Thus, the selective photo-induced decarboxylation of carboxylic acids with different carboxyl groups enabled the use of an appropriate combination of ED and EA in the developed system.

Photoinduced Deboronation of Arylboronic Acids in a Two-molecule Photoredox System
Aryl radicals are useful reactive intermediates in reactions that introduce aryl groups but are difficult to generate and control because of their high reactivity.Recently, mild aryl radical generation from arylboronic acids 16 and arylboronic acid esters 17, which are stable, commercially available substances, via photoinduced deboronation in a two-molecule photoredox system under UV or visible light irradiation has been reported (Scheme 8a). [37,39,46]Direct evidence of the generation of aryl radicals in this system was obtained by trapping TEMPO in the presence or absence of alkenes (Scheme 8b). [46]In addition, deuterium-labeling experiments using d 3 -acetonitrile showed the existence of competitive radical reactions between the radical addition to the alkene and hydrogen abstraction from acetonitrile (Scheme 8b).As shown in the proposed mechanism (Scheme 8c), the radical cation of ED oxidizes the arylborate ions to generate aryl radicals via deboronation, and the sequential BET and protonation yielded the adducts similar to decarboxylation of carboxylic acids.In contrast to alkyl radicals, the reaction of aryl radicals required five equivalents of the alkene because of the competitive hydrogen abstraction and addition to alkene processes (Scheme 8c).The milder conditions and higher efficiency of the deboronation of arylboronic acids compared to those observed in the decarboxylation of benzoic acids improved the yield of alkene adducts.
In particular, the generation of aryl radicals bearing electronwithdrawing groups (e. g., R = Cl, CN) proceeded smoothly to yield the corresponding adducts in moderate yields.In addition, the mild generation of naphthyl and phenanthryl radicals from the corresponding arylboronic acids successfully improved the yield of alkene adducts.Recently, we investigated the effects of ED and countercations on the photoinduced deboronation and decarboxylation of aryl radicals. [47]Determining the precise oxidation potentials of the arylborate and benzoate ions obtained through CV poses challenges owing to several factors, including the equilibrium between the substrate and corresponding ate ions in the presence of a base, sequential deboronation and decarboxylation reactions, and the inter-action between the electrode and ate ion of the substrate.However, replacing ED in the two-molecule photoredox system yielded valuable insights that suggested the possibility of roughly estimating the oxidation potentials of the arylborate and benzoate ions.
Table 4 shows the effects of ED and the counter-cation on the photoinduced deboronative radical addition of arylboronic acid derivatives 17 a-c to acrylonitrile, serving as the radical trapping reagent, in the two-molecule photoredox system using CMA as the EA.To assess the effects of ED and the countercation, we compared the yield of the adduct in the photoreaction after 2 h of irradiation, before complete consumption of 17, by replacing the ED and base.NaOH or KOH was used as the base because LiOH is insoluble, and CsOH is unsuitable for this photoreaction because of the heavy-atom effect.In the absence of ED, the low efficiency of direct PET between the excited state of CMA and arylborate ion of  (the substituent attached to the boron atom and electrondonating or electron-withdrawing substituents on the benzene ring).By contrast, the counter-cation only slightly influenced the oxidation potential.In addition, the oxidation potentials of the arylborate and benzoate ions could be approximately estimated using the tendency of the yield of adduct replacing the ED.For example, the estimated oxidation potential of arylborate ion from 17 a is ~+ 1.1 V vs. SCE in CH 3 CN because the yield of adduct was significantly changed between the use of DBC (+ 1.17 V) and Anth (+ 1.09 V) as ED.
Similarly, the oxidation potentials of the arylborate and benzoate ions estimated using this method are shown in Scheme 9. [47]

Summary and Outlook
From the aforementioned results, our two-molecule photoredox system shows several advantages: (1) it employs inexpensive, stable, and neutral organic photoredox catalysts, (2) exhibits distinct reactivity compared to one-molecule photoredox systems using Ir and Fukuzumi catalysts, because of the lower efficiency of BET from the radical anion of EA, (3) allows a facile replacement of these photocatalysts with other photocatalysts for modulating the oxidation/reduction potential of the substrate, (4) enables the replacement of only an ED while maintaining the same EA, facilitating a change in the oxidation potential while keeping the reduction potential of the photoredox catalyst.However, this modification is impossible for the one-molecule photoredox catalyst because of the simultaneously changing oxidation and reduction potentials in the modification process of the one-molecule photoredox catalyst.Further development of useful and unique organic photoredox reactions in two-molecule photoredox systems is underway.
Ph.D. in 2002 from Osaka Prefecture University under the supervision of Professor Kazuhiko Mizuno.He became an assistant professor at University of Fukui in 2002 and was promoted to full professor in 2022.He was a visiting researcher at Tulane University in 2001 (Professor V. Ramamurthy, Organic Photochemistry) and University of Wisconsin, Madison, in 2011 (Professor S. H. Gellman, Peptide Chemistry).His research focuses on the development of synthetically useful organic photoredox reactions and modification and preparation of amino acids and peptides using organic photoredox reactions.
A L R E C O R D Chem.Rec.2024, 24, e202300326 (5 of 9) © 2023 The Authors.The Chemical Record published by The Chemical Society of Japan and Wiley-VCH GmbH because the primary carboxylate ion of 13 at the side-chain cannot be oxidized by the radical cation of DBC (oxidation potential of carboxylate ion at the α-position: + 0.95 V vs.
Scheme 8. Generation of aryl radicals via photoinduced deboronation of arylboronic acids 16 or arylboronic acid esters 17 in a two-molecule photoredox system.

Table 4 .
Visible-light-induced deboronation of 17 a-c replacing ED in a two-molecule photoredox system.P e r s o n a l A c c o u n t T H E C H E M I C A L R E C O R D Chem.Rec.2024, 24, e202300326 (7 of 9) © 2023 The Authors.The Chemical Record published by The Chemical Society of Japan and Wiley-VCH GmbH