Visible-light-driven indation of aryl halides: arylindium(iii) reagents bridging single- and two-electron reactivity Authors: Anton A. Gladkov, Liubov I. Panferova, Vitalij V. Levin, Alexander D. Dilman Published: 2026 ABSTRACT A photoinduced indation of aryl and alkyl halides enabling efficient access to organoindium reagents is described. The method is operationally simple, exhibits broad functional group tolerance, is insensitive to air and moisture, and proceeds in high yields. The transformation can be achieved with or without a photocatalyst. It was proposed that starting from indium(i) bromide and bromide anion, a highly active low-valent indium species is formed under irradiation with 400, 455, and even 520 nm LEDs, which is capable of performing halogen atom transfer (XAT) to generate aryl radicals, finally leading to arylindium reagents. This reactivity mode is exceptionally rare for metals in main-group chemistry and typically associated with transition metals. The arylindium reagents can, in turn, serve as aryl radical precursors under photoredox conditions, establishing a reversible interplay between the formation and cleavage of the carbon–indium bond. This dual reactivity unlocks an unprecedented reactivity mode for arylindium reagents as radical sources in photocatalysis. Introduction Activation of aryl halides leading to the formation of organometallic reagents is a fundamental synthetic process, crucial both for the construction of C–C bonds1–5 and for the installation of functional groups.6–11 Historically, lithium and magnesium metals were employed as activators to obtain the corresponding organometallics,12,13 which served as nucleophilic reagents, while less polar metals are notably less efficient for the direct insertion into the C(sp2)-X bond. On the other hand, arylzinc, arylboron, and aryltin reagents have found widespread applications in aryl–aryl cross-coupling reactions, with each metal having its own reactivity profile14 (Scheme 1A). An apparent advantage of less polar organometallics is expanded functional group tolerance and enhanced stability towards air or moisture. In this regard, organoindium reagents offer an excellent option in terms of air/moisture stability,15,16 with an additional benefit of having low toxicity.17 However, methods to access arylindium reagents are quite limited. Thus, in earlier literature, triarylindium reagents were generated from the corresponding aryllithium derivatives,18–23 which negates these advantages with respect to tolerance/sensitivity (Scheme 1B). Arylmagnesium reagents can also be generated in situ to be directly converted to triarylindiums.24 Other methods include reaction of aryl mercury with Rieke indium25 and indation of aryl diazonium salts.26 An attractive general approach would be direct insertion of elemental indium into the carbon–halogen bond. However, due to low reactivity of elemental indium, which is further exacerbated by batch-to-batch variation of indium surface,27 these reactions work only with aryl iodides.28 Another issue associated with the use of elemental indium is inevitable, due to trivalency of indium, formation of mixtures of organometallic reagents such as mono- and diorganoindiums. Recently, we reported that indium(i) bromide is very efficient for converting alkyl halides into alkylindiums formed as a single In(iii) species.29 However, application of the latter method to aryl halides is problematic, apparently, owing to stronger C(sp2)-halide bonds and lower reactivity of aryl halides in SET-driven reductive processes. Arylindium reagents. A modern approach to the synthesis of organic derivatives of less polar metals relies on the generation of carbon-centered radicals followed by their interception by a metal center30,31 (Scheme 1C). Despite the recent proliferation of photocatalytic methods for radical generation based on single-electron transfer (SET) processes32 and halogen atom transfer (XAT),33,34 their application to the generation of sp2-hybridized radicals35–37 is less developed compared to their sp3 counterparts. The formation of an organometallic reagent as the final product is particularly rare, as most metals corresponding to organometallic reagents employed in cross-coupling chemistry do not readily interact with radicals. Consequently, successful metalation under photocatalytic conditions has thus far been achieved primarily for the preparation of aryl derivatives of boron38,39 and tin.40–42 Herein, we report an unprecedented property of anionic indium(i) halide complexes to behave as effective XAT/SET agents under visible-light irradiation (Scheme 1D). An equimolar mixture of indium(i) bromide and lithium bromide in solution absorbs visible light and performs activation of difficult-to-activate aryl halides. This observation has enabled the development of a mild, photoinduced synthesis polyfunctional arylindium reagents via the generation and subsequent interception of aryl radicals to form the carbon–indium bond. Moreover, the resulting arylindium compounds can, in turn, serve as precursors to aryl radicals under photoredox conditions, establishing their role as a reservoir for these reactive intermediates. Arylindium reagents have previously been employed in palladium-catalyzed cross-coupling.43,44 However, the reactions involving monoarylindium(iii) dihalides have not received widespread use due to the limited number of suitable electrophilic partners, which are mainly restricted to aryl iodides. To demonstrate the practical value of organoindium reagents obtained via light-promoted insertion of indium(i) bromide, they were successfully employed in an improved palladium-catalyzed cross-coupling with aryl iodides, bromides, and chlorides. Collectively, the presented results not only overcome the key limitations that have hitherto constrained the development of organoindium chemistry but also lay the conceptual groundwork for the utilization of indium(i) complexes in photochemistry. Results and discussion Photoinduced indation of aryl- and alkyl halides 4-Fluorobromobenzene 1a was selected as a model substrate, and its reaction with indium(i) bromide was monitored by 19F NMR (Table 1). In the presence of lithium bromide and cyanoarene-type photocatalyst 3DPA2FBN45 (0.25 mol%) in acetonitrile under 400 nm LED irradiation, the complete conversion of 1a was achieved in 2 h, resulting in a 97% yield of the arylindium reagent 2a. It should be noted that the reaction proceeds in a heterogeneous manner, as mixing of all reaction components results in disproportionation of indium(i) bromide, leading to the formation of indium(iii) bromide and a finely divided powder of indium(0).46 Conducting the reaction in deuterated acetonitrile (d3-MeCN) enabled full characterization of 2a by 1H, 13C, and 19F NMR. The spectra revealed the formation of two organoindium species, which were assigned to monoaryl (ArInBr2) 2a-M and diaryl (Ar2InBr) 2a-D based on HRMS (ESI-TOF) data. The yield of 2a-D varied from 1% to 17% depending on the reaction conditions (see SI for a detailed analysis). For clarity, all yields of reagent 2a reported here represent the combined yield of species 2a-M and 2a-D. Reagent 2a was obtained in 93% yield even in the absence of 3DPA2FBN under 400, 455, or 520 nm LED irradiation (entries 1–3). However, with green light a longer reaction time (48 h) was required to achieve complete conversion of 1a. The reaction also demonstrated high tolerance to air atmosphere and moisture, affording 2a in 97% and 95% yield, respectively (entries 4 and 5). Given the heterogeneous nature of the reaction, we investigated the effect of temperature fluctuations on the yield of 2a. The efficiency remained virtually unchanged, both when the temperature was lowered to 5 °C and when it was raised to 30 °C (entries 7 and 8). Reducing the loading of the InBr/LiBr reagent system to 1.1 equivalents led to a slight decrease in the yield of 2a to 89% (entry 9). Although heterogeneous reactions are generally considered to exhibit lower reproducibility than their homogeneous counterparts, our experiments demonstrated that this protocol is robust toward trace amounts of moisture and air, temperature variations, and reduced loadings of the InBr/LiBr reagents. This robustness underscores the practical utility of the method. Table 1Optimization of photoinduced indation of aryl bromides aYields determined by 19F NMR with p-F-toluene as an internal standard.bBased on 1a, taking into account two aryl groups in 2a-D. Optimization of photoinduced indation of aryl bromides Yields determined by 19F NMR with p-F-toluene as an internal standard. Based on 1a, taking into account two aryl groups in 2a-D. The reaction proceeded with decreased yields in the presence of the oxidative photocatalysts 8Cl4CZIPN47 or 4CzIPN47 (50% and 71% yield, respectively) (entries 10 and 11). In contrast, the use of more reductive photocatalysts – 3DPAFIPN,45 [Ir(dttbpy)(ppy)2][PF6], Ir(ppy)3, and N-phenylphenothiazine – afforded reagent 2a in high yields ranging from 92% to 96% (entries 12–15). Notably, within this series, the formation of the diarylindium species 2a-D increased (see SI for a detailed analysis). With thioxanthone as a triplet sensitizer,48 the photoinduced indation was slower, affording 2a in 64% yield and leaving 31% of unreacted 1a according to 19F NMR analysis (entry 16). Solvent screening revealed that reagent 2a was obtained in high yield in acetone and DMSO (92% and 83%, respectively) (entries 17 and 18). In contrast, lower yields were observed in THF, EtOH, and dimethylacetamide (DMAc), along with an increased formation of the protodehalogenation byproduct 3a (entries 19–21). This result is consistent with competitive hydrogen atom transfer (HAT) from these solvents to the aryl radical intermediate. While most photoinduced transformations require low concentrations, our reaction proceeded efficiently at concentrations up to 1.5 M, delivering reagent 2a in 96% yield (entry 22). The dilution of the reaction mixture to 0.5 M had no adverse effect, but further dilution to 0.25 M reduced the yield to 84% (entries 23–24). Replacing lithium bromide with lithium chloride did not reduce the yield, which offers a potential advantage in cost savings when scaling up (entry 25). Tetrabutylammonium bromide (TBABr) was also effective, although a slight increase in byproduct 3a was observed (entry 27). The use of lithium iodide led to a significantly lower yield of 52% (entry 26). The reaction did not proceed under thermal conditions at 100 °C in the dark (entry 29) or when indium(i) bromide was replaced with metallic indium (entry 32). The yield of reagent 2a dropped sharply to 19% in the absence of lithium bromide (entry 31), which we attribute to the critical role of the halide anion in facilitating the formation of the light-absorbing and redox-active anionic indium(i) species in solution. After successful optimization, we decided to evaluate whether manipulations with arylindium reagents could be performed under a non-inert atmosphere. Although reports on the stability of triarylindium complexes with DMAP (Ar3In·DMAP) are known,19 such tests had not been conducted on monoarylindium reagents. For this purpose, reagent 2a was prepared under standard conditions, and then the solution was transferred into an NMR tube under air. The yield of all reaction products was monitored over two months by means of 19F NMR spectroscopy using an internal standard; after each 19F NMR spectrum was recorded, the tube cap was removed, and the solution was left open to air (Scheme 2). As can be seen in the plot, the combined yield of reagent 2a (solid grey line) does not decrease during the 30 days and remains at 96%. On the 45th day, the yield of 2a slightly decreases to 93%, and by the end of 60 days to 92%. At the same time, a slight increase in the protodemetalation product 3a (solid red line) to 4% is observed. It is noteworthy that initially, reagent 2a consists of a mixture of reagents 2a-M (dotted blue line) and 2a-D (dotted green line) with yields of 82% and 7%, respectively, and by the end of the seventh day, the Schlenk equilibrium49 completely shifts toward reagent 2a-M, while the total yield remains the same 96%. Air stability test of reagent 2a. aYields were determined by 19F NMR with 1-fluoronaphthalene as an internal standard. Given the well-known affinity of indium(iii) species toward water, we sought to probe whether water coordinates to arylindium reagent 2a. For this purpose, the 1H NMR spectrum of water in CD3CN was first recorded, followed by that of reagent 2a. In the latter case, the characteristic singlet of free H2O at δ 2.44 ppm was absent, whereas a broad singlet at δ 3.57 ppm was observed instead. Upon addition of one equivalent of water to the NMR sample containing reagent 2a in CD3CN, the integral intensity of the resonance at δ 3.57 ppm increased, while the signal of free water at δ 2.44 ppm did not appear. These observations are consistent with coordination of water to indium(iii) species present in solution, giving rise to hydrated arylindium and/or indium(iii) bromide complexes (ArInBr2·xH2O and/or InBr3·xH2O) (see the SI for details). Taken together, these experiments indicate that although arylindium reagents are capable of coordinating water, such coordination does not result in significant protodemetallation upon preparation, handling, or under reaction conditions employed in this work. Since disproportionation of indium(i) bromide provides elemental indium, which is present in the final reagent, the mixture was centrifugated, and the resulting homogeneous solutions of the reagents 2 were analyzed. Thus, it contained low-valent indium complexes, specifically dibromoindate(I) anion [InBr2]− as identified by HRMS (ESI-TOF) (see SI for details). Prior to examining the scope of the photoinduced indation, the requirement for pre-treatment of reagents 2 had to be addressed (Scheme 3). Since these indium(i) species were found to impair cross-selectivity in subsequent palladium-catalyzed cross-coupling reactions (see below), a protocol was developed to introduce a mild oxidant into the reaction mixture. This additive was required selectively to consume the indate [InBr2]−, with any excess being readily removable by evaporation. Methyl iodide (Method A) and 1,2-dibromotetrafluoroethane (Method B) were employed as quenching agents for this purpose. Thus, treatment of the reaction mixtures with one equivalent of either reagent at room temperature for two hours resulted in the complete degradation of Li+[InBr2]−, yielding a methylindium species (Method A) or indium(iii) bromide (Method B), respectively. The choice of quenching method was guided by the structure of the starting substrate 1. Generally, Method B offers broader utility owing to its superior functional group tolerance and because it avoids the formation of potentially interfering methylindium byproducts, which is inherent to Method A. For the model arylindium reagent 2a, stirring the reaction solution in air for 24 h also resulted in oxidation of the low-valent indium(i) species. Although this alternative protocol was demonstrated only for reagent 2a, it provides a convenient method for removing indium(i) species in this case and is discussed below in the context of the Pd-catalyzed cross-coupling reactions (see the SI for details). Scope of arylindium(iii) reagents. aRatio between Ar–InBr2 (2) and Ar–H (3) was determined by titrating an aliquot of the reaction solution with iodine and subsequent analysis by GC-MS. bConcentration of the reagent 2 was determined by iodometric titration. cYield determined by 1H NMR with 1,3,5-trioxane as an internal standard. dReaction was performed on a 0.5 mmol scale in MeCN or CD3CN using modified irradiation times (see SI for details). eInBr/LiBr (1.1 equiv.) were used. fReaction was performed at a concentration of 0.75 M. gInBr/LiBr (2.0 equiv.) were used. hYield of 2 based on the isolated corresponding aryl iodide obtained after iodination. iConcentration of the reagent 2 was determined by 1H NMR with 1,3,5-trioxane as an internal standard. j400 nm 10 W irradiation was used. kConcentration of the reagent 2ax was determined by 19F NMR with PhCF3 as an internal standard. lYield of 2ax was determined by 19F NMR with PhCF3 as an internal standard. Since arylindium reagents 2 were not amenable to isolation in pure form by column chromatography, the efficiency of their formation was assessed by three complementary analytical methods. The results are summarized in Scheme 3 (see SI for details). First, the ratio of Ar–InBr22 to Ar–H 3 was determined by GC-MS analysis after quenching an aliquot of the reaction mixture with iodine, which resulted in the formation of the corresponding aryl iodide (see SI for details). Second, the concentration of arylindium reagent 2 after pre-treatment by Method A or Method B was determined by iodometric titration, and the resulting standardized solution was used in all subsequent transformations. Third, the yield of an independently prepared solution of arylindium reagent 2, obtained under identical reaction conditions but without pre-treatment, was determined by 1H or 19F NMR spectroscopy using an internal standard. The corresponding crude reaction solutions of organoindium reagents 2 in MeCN or CD3CN were characterized by 1H, 13C, and 19F NMR, as well as HRMS (ESI-TOF). With analytical methods for characterizing the organoindium reagents 2 established, we next explored the scope of the photoinduced indation reaction. To demonstrate the scalability of our photochemical indation protocol, we initially prepared four stock-solutions at 8 mmol scale – two solutions of reagent 2a and two of reagent 2b. While full conversion to reagent 2a was achieved within two hours on a 0.5 mmol scale of aryl bromide 1a, a 16-fold scale-up required only a modest extension of the irradiation time to 6 hours using a LED matrix of comparable power. The photoinduced indation to form reagent 2b proceeded with similar efficiency. Having obtained duplicate 8 mmol-scale solutions of 2a and 2b, one solution of each pair was processed according to Method A and the other according to Method B. This allowed us to evaluate potential overestimation in iodometric titration due to the presence of the methyl indium byproduct formed in Method A, relative to the cleaner Method B treatment. The titration results for all four samples are summarized in Scheme 3. Two principal conclusions emerge: (1) reagents 2a and 2b are formed in high yield, as indicated by titers approaching the theoretical maximum of approximately 0.85–0.9 M for quantitative yields; (2) the methylindium species generated during Method A contributes negligibly to the titration endpoint, which is rationalized by the significantly faster reaction of the arylindium species 2 with iodine compared to that of methylindium(iii) dihalide. To demonstrate synthetic utility, all subsequent photoreactions for preparing reagents 2 were performed on a 0.5–3 mmol scale. Notably, the reaction conditions exhibited exceptional functional group tolerance, considerably broadening the scope of accessible organoindium reagents relative to classical methods. A diverse set of functional groups proved compatible, including fluoride 2ab, chloride 2d, trifluoromethyl 2q, acetyl 2f, ester 2ai, nitrile 2ao, amino 2u, methoxy 2b, tosylate 2o, amide 2p, sulfonamide 2n, and diethyl acetal 2ax, affording the corresponding organoindium reagents in good to excellent yields (62–97%). For comparison, under photocatalyst-free conditions, reagent 2c was obtained in 91% yield using 455 nm LED irradiation, whereas reagent 2f was obtained in 84% yield with 520 nm LED irradiation, albeit requiring a longer reaction time. Substrates bearing amide functionalities required prolonged irradiation to reach full conversion, furnishing the corresponding arylindium reagents 2p and 2x in 77% and 91% yields, respectively. An extended reaction time of 12 hours was also necessary for the synthesis of the reagent from 4-bromoquinoline 2ap (70%). Even the strongly electron-donating 5-bromo-1,2,3-trimethoxybenzene was successfully transformed, affording the corresponding reagent 2c with a good titer (0.87 M) and minimal formation of the protodebromination side product 3c. The use of 1-bromo-4-iodobenzene as a substrate was investigated. Pleasingly, upon reducing the loading of InBr/LiBr to 1.1 equivalents, indation proceeded with complete selectivity for the C–I bond, leaving the C–Br bond intact and affording reagent 2e in 71% yield after only 10 min of irradiation. We next examined aryl bromides containing metal atoms in the para position. Under the standard conditions, arylindium reagents bearing boryl (Bpin, 2g) and stannyl (SnBu3, 2h) functionalities were successfully obtained in 85% and 63% yields, respectively. To the best of our knowledge, such reagents are unprecedented and provide a platform for orthogonal downstream transformations. Functionalities bearing C–H bonds sensitive to radical HAT abstraction were also tolerated: reagents containing free benzylic hydroxymethyl 2i (93%) and aminomethyl 2j (60%) groups were prepared. Furthermore, substrates with protected phenolic groups were compatible. Notably, triflate and mesylate groups remained intact during photoinduced indation, affording reagents 2k and 2l in 95% and 97% yields, respectively, with excellent ArInBr2 (2)/Ar–H (3) ratios. Terminal alkenyl 2s and protected alkynyl 2t groups were also incorporated successfully in good yields (77% and 75%, respectively). A series of para-aminophenyl indium reagents 2u–x was also synthesized in good to excellent yields (72–91%). Only in case of substrate 1v the indation required an increased loading of InBr/LiBr (2.0 equiv.) to achieve full conversion, presumably owing to strongly donating and/or coordinating character of the dimethylamino group. Surprisingly, the reagent containing an acidic trifluoroacetamide group 2w proved stable, with no significant accumulation of the protodemetalation product 3w over time. We then turned our attention to ortho-substituted aryl bromides. Reagents bearing ortho-methyl, acetyl, amino, ester, phenyl, and mesityl groups (2ac–ah) were obtained in moderate to good yields (41–83%). As shown in Scheme 3, these ortho-substituted reagents were typically formed with lower yields and increased amounts of the protodebromination products 3, consistent with steric hindrance impeding the capture of the ortho-substituted aryl radical by indium bromide. This trend was corroborated by the increasing proportion of 3 with greater steric bulk in the substrate. Iodination of aliquots of 2ac, 2af, 2ag, and 2ah revealed trace amounts of 2-iodoacetonitrile (ICH2CN). This suggests that the sterically hindered aryl radicals can competitively abstract a hydrogen atom from acetonitrile, yielding 3ac, 3af, 3ag, 3ah, and a cyanomethyl radical, which is subsequently captured by indium(i) bromide to form cyanomethylindium(iii) bromide (see SI for details). In addition, reagents 2ag and 2ah were subjected to iodination, and the isolated yields of the corresponding aryl iodide/Ar–H mixtures were consistent with the formation of arylindium reagents 2ag and 2ah in 48% and 25% yields, respectively. For reagent 2ag, this value closely matched the yield determined by 1H NMR spectroscopy using an internal standard. In contrast, a significant discrepancy was observed for reagent 2ah. We attribute this difference to the sterically hindered iodination of mesitylindium reagent 2ah. The scope was also extended to heteroaryl bromides. 3-Bromopyridines with electron-withdrawing groups at the 5-position provided the corresponding heteroarylindium reagents 2ai and 2ao in good yields. However, prolonged irradiation after complete conversion led to increased formation of byproducts 3ai and 3ao. In addition, solutions of these pyridylindium reagents exhibited a brownish hue, suggesting potential oligomerization. Accordingly, their use in subsequent Pd-catalyzed cross-coupling (see below) resulted in lower cross-selectivity, possibly, due to catalyst deactivation by these oligomeric species. Reagents 2al and 2am, derived from 5-bromopyrimidine and 3-bromopyridine, were obtained in good yields. However, this required reducing the LED power to 10 W and limiting the reaction time to 2 hours. Under standard, more vigorous conditions, initial high conversion was followed by near-complete protodemetalation of the initially formed heteroarylindium species upon extended irradiation. This intriguing observation hints at a potential SET between the products 2al/2am and the excited photocatalyst or a photoexcited anionic indium(i) complex (see SI for details). Bromopyridines with donor substituents (e.g., amino, methoxy) were smoothly converted to the corresponding indium reagents 2ak and 2an. To highlight the versatility of the method, the synthesis of 2aj was successfully performed on a 2 mmol scale in the absence of 3DPA2FBN, affording the product in 87% yield (1H NMR), with full characterization by 1H, 19F NMR, and HRMS (ESI-TOF). Condensed heterocycles such as a TIPS-protected 5-bromoindole and 3-bromobenzo[b]thiophene were also viable substrates, affording reagents 2aq and 2ar in 60% and 61% yields, respectively. Finally, five-membered heterocyclic bromides such as derivatives of furan and thiophene were successfully employed to furnish the corresponding heteroarylindium reagents 2as–au in moderate yields (35–79%). Notwithstanding the broad scope demonstrated, our methodology encounters certain limitations. Photoinduced indation proved incompatible with aryl bromides bearing aldehyde and carboxylic acid functionalities. For example, for p-bromobenzaldehyde, GC-MS analysis revealed complete consumption of the starting material, affording only minor quantities of a product along with formation of a pinacol-type homocoupling dimer. In the case of p-bromobenzoic acid, extensive protodehalogenation was observed by GC-MS, whereas only trace amounts of the corresponding arylindium reagent were formed. Furthermore, p-bromonitrobenzene was not tolerated – a finding congruent with the established role of nitroarenes as inhibitors of SET processes.41 In this case, GC-MS indicated full substrate conversion with only trace amounts of p-bromoaniline detectable. Additionally, some five-membered nitrogen heterocycles – 3-bromopyrazole and 4-bromothiazole derivative were unsuccessful substrates, undergoing complete conversion to the corresponding protodebrominated heterocycles under the standard reaction conditions (see SI for details). To underscore the generality of our method, we sought to demonstrate that the photoinduced indation protocol extends beyond aryl halides to encompass traditionally challenging vinyl and alkyl halides. This investigation aimed to establish that photochemical conditions provide unique access to organoindium reagents that are unobtainable via thermal pathways, and that analogous organozinc reagents also fail to form under these conditions. The feasibility of conducting these transformations in the absence of a photocatalyst was also a key point of evaluation. Thus, isopropenyl bromide 1av, cyclopropyl bromide 1aw and 1-bromo-3,3-diethoxy-1,1-difluoropropane 1ax were successfully converted into the corresponding indium reagents 2av, 2aw, 2ax under irradiation without photocatalyst in good yields. Of special note, these organic bromides did not give the organoindium reagents under thermal conditions, and were ineffective or less effective in reaction with elemental zinc in attempts to prepare organozinc reagents (for detailed studies, see SI). In addition to the challenging organoindium reagents 2av, 2aw and 2ax, a stock solution of reagent 2ay was obtained on a 3 mmol scale in 90% yield after only 3 h of irradiation. For comparison, the same alkylindium reagent required 4 h under thermal conditions (80 °C) and was obtained in a significantly lower yield (41%) (see SI for details). To conclude this section, we sought to prepare the phenylindium(iii) dihalide 2az from phenyl chloride (Scheme 4). Disappointingly, under standard conditions in the absence of a photocatalyst, even prolonged irradiation (72 h) afforded only trace amounts of 2az. However, replacing acetonitrile with DMF and increasing the concentration of all reagents enabled the formation of the desired reagent 2az in satisfactory yield. To the best of our knowledge, this represented the first example of the synthesis of a monoarylindium(iii) reagent from an unactivated aryl chloride. Photoinduced indation of PhCl. aDetermined by GC-FID after iodination of an aliquot of the reaction mixture. Synthetic applications of arylindium(iii) reagents in Pd-catalyzed reactions With the scope of organoindium reagents established, we first targeted the optimization of their palladium-catalyzed cross-coupling with aryl halides. Since literature indicated that monoarylindium(iii) dihalides had participated in palladium catalysis only with aryl iodides,28,50,51 we sought conditions that would be equally effective with aryl bromides and chlorides. Our investigation began with 4-iodoanisole 1b-I as the model electrophile for coupling with reagent 2a (pre-treated via Method A), monitored by 19F NMR spectroscopy (Table 2). A system comprising Pd(OAc)2 (1.5 mol%), SPhos (3.0 mol%), [Cu(MeCN)4]BF4 (3.0 mol%), and LiCl in DMAc at 80 °C achieved complete conversion within 1 hour, delivering the biaryl product 4 in 96% isolated yield. Importantly, no cross-coupling product from the methyl indium byproduct (generated in Method A) was detected. Catalyst loadings could be reduced six-fold [Pd(OAc)2 (0.25 mol%), SPhos (0.5 mol%), and [Cu(MeCN)4]BF4 (1.0 mol%)], affording full conversion within 2 hours and product 4 in 93% yield (TON = 372), highlighting the efficiency of the system. The yield was unaffected by the addition of stoichiometric amounts of water. Performing the reaction at 20 °C significantly reduced the rate, yielding 72% of 4 after 48 hours, consistent with transmetalation being rate-limiting. The more economical Pd(PPh3)2Cl2 instead of Pd(OAc)2/SPhos provided a slightly diminished yield of 89%, and switching the pre-treatment of 2a to Method B had no adverse effect. Interestingly, preparing arylindium reagent 2a under the standard photochemical conditions in the presence of air, followed by stirring under air for 24 h, also resulted in oxidation of the soluble indium(i) species. The resulting air-pretreated reagent 2a was directly compatible with the standard Pd-catalyzed cross-coupling conditions, providing biaryl 4 in 96% yield (Table 2, bottom). Table 2Optimization of Pd-catalyzed cross-coupling reaction aYields determined by 19F NMR with p-F-toluene as an internal standard.bIsolated yield.cLiCl (4.0 equiv.), Pd(OAc)2 (5 mol%), SPhos (10 mol%), [Cu(MeCN)4]BF4 (10 mol%) were used. Optimization of Pd-catalyzed cross-coupling reaction Yields determined by 19F NMR with p-F-toluene as an internal standard. Isolated yield. LiCl (4.0 equiv.), Pd(OAc)2 (5 mol%), SPhos (10 mol%), [Cu(MeCN)4]BF4 (10 mol%) were used. Pleasingly, the optimized system proved effective with p-bromoanisole 1b, albeit requiring a proportional increase in catalyst loading to 5 mol% to furnish 4 in 97% yield. In stark contrast, employing p-chloroanisole 1b–Cl afforded merely 6% yield of 4, with rapid formation of palladium black halting the reaction. This suggests a shift in the rate-determining step to oxidative addition for electron-rich aryl chlorides, leading to Pd(0) accumulation and aggregation – a known deactivation pathway in related couplings.52 Acceptor-type aryl chlorides participated successfully under the standard conditions (see below), though coupling donor-type aryl chlorides will require further catalyst development. Notably, full conversion in the model coupling between 1b-I and 2a could not be achieved in THF, DME, or MeCN, even with 5 mol% catalyst loading. We hypothesized that indium(iii) bromide, generated as a stoichiometric byproduct, might act as a catalyst poison, potentially via formation of heterobimetallic Pd–In complexes analogous to those proposed in Negishi coupling.52 Supporting this, adding one equivalent of InBr3 to the standard reaction suppressed the yield of 4 to 73%. Crucially, the concurrent addition of InBr3 and four equivalents of HMPA restored the yield to 96%. This confirms that polar aprotic solvents like DMAc or HMPA are essential, likely by sequestering InBr3 and preventing catalyst deactivation. We further compared efficiency of our protocol with the one-step cobalt/palladium system reported by Yoshikai and co-workers.51 While the authors noted no interference from the cobalt catalyst in the subsequent cross-coupling, under our conditions, adding 5 mol% of the cobalt catalyst to the coupling mixture reduced the yield of 4 from 98% to 40% at 30 min reaction run, requiring 4 hours to reach 90% yield. This demonstrates that the cobalt catalyst significantly retards the palladium-catalyzed step. A series of control experiments validated the optimized system. Omitting the copper co-catalyst did not markedly affect the yield of 4, though increased homocoupling of 2a was observed by 19F NMR (see SI for details). Excluding LiCl, or both LiCl and the copper salt, reduced yields to 89% and 93%, respectively. The absence of the SPhos ligand diminished the yield to 60%. Conducting the reaction under air lowered the yield to 78%. Most tellingly, skipping the pre-treatment Method A of 2a resulted in a 67% yield of 4, with GC-MS analysis revealing significant homocoupling of 1b-I. This corroborates the presence of low-valent indium species (for example, Li[InBr2]) in the untreated solutions, which compromise selectivity. With optimized cross-coupling conditions in hand, we next explored the scope of Pd-catalyzed cross-coupling (Scheme 5). Conditions A are analogous to optimized conditions B, with the sole difference being an increased loading of catalysts and additives. Initially, a brief screening of the electrophilic component was conducted by reacting various aryl electrophiles with reagent 2b (processed using Method A). p-Iodonitrobenzene reacted with reagent 2b within just 30 minutes, affording product 5 in an excellent 96% yield. Aryl halides bearing p-acetyl, methyl, chloro, fluoro, and ethenyl groups reacted within 2 hours, delivering products 6–10 in good to excellent yields (63–95%). Biaryl 11, containing both an amide and a free hydroxyl group in the starting aryl bromide, was also obtained in good yield. We next examined sterically hindered aryl halides. The reaction with o-iodoaniline proceeded within the same 2-hours timeframe, providing the desired biaryl 12 in 70% yield. To evaluate the maximum steric demand tolerated by the electrophilic component, we performed a cross-coupling between 2-bromo-1,3,5-triisopropylbenzene and reagent 2b. The initial attempt under standard conditions B was unsuccessful, stalling at 30% conversion (GC-MS). Switching the solvent to HMPA enabled the cross-coupling to reach 80% conversion of the starting aryl bromide, affording product 13 in 64% isolated yield. It was observed that certain nitrogen-containing heterocycles decelerated the cross-coupling, presumably by coordinating to the palladium catalyst and thereby impeding efficient transmetalation. Based on these observations, transmetalation proceeds more slowly in DMAc and HMPA than in DME or EtOH. For instance, 2-iodopyrazine reacted sluggishly under standard conditions A and required a solvent switch to DME, which allowed product 14 to be obtained in a high 83% yield. Bromopyridines bearing formyl and nitrile groups reacted smoothly with reagent 2b, furnishing biaryls 15 and 16 in 82% and 88% yield, respectively. Among five-membered heteroaryl halides, 2-iodothiophene, a bromothiazole bearing a donor dimethylamino substituent, and 2-bromothiazoles functionalized with a methyl or ester group at the 5-position all underwent cross-coupling with reagent 2b, affording the corresponding products 17–20 in good to excellent yields. The synthesis of product 20 required a solvent change to ethanol for the reason noted above. Unfortunately, product 21, containing a thiadiazole moiety, could not be obtained under our conditions. Finally, we performed two sequential cross-couplings on dichloropyrazine. The double cross-coupling product 22 was isolated in a good yield of 63%, although this required switching to conditions B, along with an increase in the loading of 2b (2.6 equiv.), the reaction temperature to 100 °C and an extension of the reaction time to 12 hours. Following this brief screening of the electrophilic component, the majority of the previously synthesized arylindium reagents were subjected to cross-coupling with various aryl halides. To demonstrate the robustness of the developed cross-coupling protocol, we generated mixed examples without specifying a particular electrophilic partner. Arylindium reagents containing hydroxymethyl 2i and trifluoroacetamide 2w groups underwent smooth cross-coupling under conditions A with acceptor-substituted aryl iodides, delivering the corresponding biaryls 23 and 24 in high yields. Product 25, featuring a free phenolic group, was obtained in a good 74% yield, although this required an increase in the palladium catalyst loading to 2 mol%. Given that our palladium-catalyzed cross-coupling proceeds in the absence of base, we anticipated that arylindium reagent 2g would exhibit orthogonal reactivity with respect to the Bpin group. Pleasingly, product 26 was selectively obtained in a good 67% yield, illustrating the synthetic potential of arylindium reagents in transformations orthogonal to Suzuki–Miyaura cross-coupling. Free and Boc-protected 5-bromoindoles also underwent smooth cross-coupling with reagents 2a and 2l, respectively, affording biaryls 27 and 28 in excellent yields. Notably, replacing the Pd(OAc)2/SPhos catalytic system with Pd(PPh3)2Cl2 enabled selective cross-coupling with aryl iodides while leaving the aryl bromide moiety untouched, both when present in the arylindium reagent (biaryl 29) and in the electrophilic component (biaryl 39). Scope of Pd-catalyzed cross-couplings. Isolated yields are shown. a2b (2.6 equiv.) was used. bLiCl (1.5 equiv.), Pd(OAc)2 (2 mol%), SPhos (4 mol%), [Cu(MeCN)4]BF4 (4 mol%) were used. cPd(PPh3)2Cl2 (5 mol%) instead of Pd(OAc)2/SPhos was used. Indium reagents 2v, 2u, and 2x participated in cross-coupling with methyl 3-iodobenzoate, 5-bromopyrimidine, and o-iodobenzonitrile, furnishing products 30, 31, and 32 in moderate yields (44–66%). Arylindium reagent 2y, bearing donor substituents, reacted with an aryl bromide containing an acceptor ester group at the ortho position under conditions B, affording biaryl 33 in a high 84% yield. The reaction between electron-accepting arylindium reagent 2n, functionalized with an acceptor sulfamoyl group at the para position, proceeded under conditions B with the extremely donor bromarene – 4-bromo-N,N-dimethylaniline over four hours, delivering product 35 in a moderate 47% yield. Arylindium reagents 2p, 2m, and 2q underwent cross-coupling with aryl halides bearing mesylate, sulfamoyl, and Bpin groups at the para position, providing the corresponding cross-coupling products 36–38 in moderate yields. A complex example, product 40, was also obtained, wherein the electrophilic component incorporated multiple sensitive functional groups simultaneously: an unprotected indole, an ester, and a protected hydroxybenzyl moiety. We next evaluated ortho-substituted arylindium reagents 2ad, 2ag, 2ae, and 2af in the cross-coupling reactions. Notably, no cross-coupling product 41 was observed when reagent 2ad was combined with 4-iodoanisole at 80 °C. However, increasing the reaction temperature to 100 °C enabled the isolation of product 41 in 71% yield. Increasing the steric demand at the ortho position of the arylindium reagent necessitated even higher reaction temperatures. Reagent 2ag failed to react with 3-iodobenzaldehyde even at 100 °C; raising the temperature to 120 °C was required to achieve full conversion of the starting aryl iodide and deliver the desired product 42 in moderate yield. The reaction between arylindium reagent 2ae and 3-bromopyridine also proceeded at 120 °C, affording cross-coupling product 43 in high yield. We also performed the coupling of ortho-acetyl-substituted arylindium reagent 2af with ortho-iodochlorobenzene at 120 °C, furnishing product 44 in a good 72% yield. Subsequently, a range of heteroarylindium reagents were introduced into cross-coupling reactions with various aryl and heteroaryl halides. Heteroarylindium reagents 2am, 2as, 2at, and 2au reacted smoothly with 4-bromoacetophenone, 5-bromo-2-acetylthiophene, and protected para-halogenated anilines, delivering the corresponding cross-coupling products 45, 46, and 52–54 in good to excellent yields (49–88%). Pyridyl indium reagents 2ai and 2ao were cross-coupled with 4-iodoanisole and 5-bromo-2-fluoropyridine, respectively, affording biaryl 48 in a good 64% yield and bipyridine 49 in a diminished 42% yield. Indium reagent 2ap, derived from 4-bromoquinoline, reacted smoothly under conditions B with the highly donor-substituted 5-bromo-1,2,3-trimethoxybenzene, furnishing product 50 in 60% yield. The reaction between 2-chlorobenzo[d]oxazole and reagent 2aq proceeded under conditions B within 24 hours, providing the corresponding bis(heterocyclic) product 51 in a moderate 56% yield. To demonstrate the value of the developed method, we prepared several synthetically valuable biaryls (products 55–58). A Cytisine precursor5355 was obtained in 58% yield via the reaction of reagent 2an with methyl 5-bromonicotinate under conditions B. A Boscalide precursor5456, conventionally synthesized via Suzuki–Miyaura coupling, was successfully accessed using our methodology with an excellent 84% yield through the reaction of reagent 2d with 2-chloronitrobenzene. Ligand 57, required for the preparation of an iridium photocatalyst,55 was obtained by cross-coupling reagent 2ab with 2-bromo-5-(trifluoromethyl)pyridine in ethanol as solvent, with the reaction time extended to 12 hours. Finally, biaryl 58, known to exhibit antitumor activity,56 was isolated in a good 72% yield after performing the cross-coupling under conditions B for 24 hours. It is worth noting that isopropenyl indium reagent 2av participated in cross-coupling with 1-(benzyloxy)-4-iodobenzene under standard conditions A with an extended reaction time of 12 hours, affording the cross-coupling product 59 in a high 90% yield. Likewise, cyclopropyl indium reagent 2aw reacted with a protected p-bromoaniline within one hour, delivering cross-coupling product 60 in a high 84% yield, albeit requiring an increase in the reaction temperature to 100 °C. In a previous study, our group demonstrated the cross-coupling of aryl thioesters with arylindium reagents (Liebeskind–Srogl-type cross-coupling).26 To showcase that the arylindium reagents 2 prepared under photocatalytic conditions are also competent substrates in Liebeskind–Srogl cross-coupling, we conducted a reaction between reagent 2b and thioester 1c (Scheme 6). Product 61 was obtained in 67% yield in the presence of stoichiometric amounts of LiCl (4 equiv.) and copper(i) thiophene-2-carboxylate (CuTC) (1.5 equiv.), along with catalytic quantities of Pd(OAc)2 (5 mol%) and P(OEt)3 (40 mol%), upon heating the reaction mixture at 80 °C for two hours. Liebeskind–Srogl cross-coupling. Isolated yield is shown. Photoredox-activation of arylindium(iii) reagents Because photoinduced indation of aryl halides proceeds via aryl radical intermediates, we were intrigued by the possibility of generating an aryl radical back from arylindium reagents under photoredox conditions. To explore this, we investigated a model photoredox-catalyzed reaction of reagent 2b (prepared according to Method B) with a suitable radical trap such as α-trifluoromethylstyrenes, which are widely used in photocatalysis,57 including for intercepting aryl radicals.58 Thus, α-trifluoromethylstyrene was selected as a model substrate, and its reaction with reagent 2b was monitored by 19F NMR (Table 3). In the presence of lithium bromide and the cyanoarene-type photocatalyst 8Cl4CzIPN (1.0 mol%) in MeCN (0.5 M) under 400 nm LED irradiation, complete conversion of the styrene was achieved within 16 h, affording product 62 in 69% yield (64% isolated yield). Increasing the loading of reagent 2b to 1.5 equiv. and lithium bromide to 2.5 equiv. did not significantly improve the reaction yield (Entry 1). Performing the reaction under 455 nm LED irradiation or at a higher concentration (1 M) resulted in a slight decrease in the yield of product 62 (Entries 2–3). Interestingly, when reagent 2b was not pretreated with dibromotetrafluoroethane, the reaction still proceeded, albeit with a slightly diminished yield of 62% (Entry 4). Replacing the solvent with DMSO led to a drop in yield to 44%, accompanied by incomplete conversion of α-trifluoromethylstyrene as determined by GC-MS (Entry 5). Finally, a series of control experiments were conducted. The reaction efficiency dropped markedly in the absence of lithium bromide additive (Entry 6), and no reaction occurred in the dark (Entry 7). Most notably, when the reaction mixture was irradiated in the absence of 8Cl4CzIPN, a modest yield of product 62 was still observed (Entry 8). This control experiment supports the notion that during the photoinduced indation of aryl bromides, even after complete conversion of the starting material to the arylindium reagent, continued irradiation of the reaction mixture still generates aryl radicals. Since the yield of the arylindium reagent does not decrease upon prolonged irradiation, this suggests that the resulting aryl radical is efficiently trapped by low-valent indium bromides, reforming the arylindium reagent. In support of this claim, irradiation of the reaction with reagent 2b prepared under photocatalyst-free conditions did not afford product 62 (Entry 9). Table 3Optimization of photoredox activation of arylindium reagents aYields determined by 19F NMR with p-F-toluene as an internal standard.bIsolated yield. Optimization of photoredox activation of arylindium reagents Yields determined by 19F NMR with p-F-toluene as an internal standard. Isolated yield. With optimized conditions in hand, we subjected several arylindium reagents 2, prepared according to our previously described method, to photoredox-catalyzed addition reactions to multiple bonds (Scheme 7A). Reagent 2k, bearing a triflate group at the para position, reacted smoothly with α-trifluoromethylstyrene to afford the gem-difluorinated alkene 63 in a good 68% yield. Heteroarylindium reagent 2aj reacted with α-trifluoromethylstyrene in a satisfactory 49% yield (product 64), although it required the addition of sodium acetate to achieve complete conversion of the starting α-trifluoromethylstyrene. Reagent 2s, containing a terminal alkene moiety, also participated in the reaction with α-trifluoromethylstyrene, albeit with a diminished yield of product 65. Product 66, derived from the photocatalytic reaction between N-methyl-N-phenylacrylamide and reagent 2b, was obtained in a moderate 46% yield and required the use of water as a co-solvent. The reaction between reagent 2b and acrylonitrile delivered product 67 in reduced yield and necessitated the addition of a nickel co-catalyst to suppress oligomerization. In addition, we performed a Giese-type photoredox-catalyzed reaction between the alkylindium reagent 2ay and 2-benzylidenemalononitrile. By switching the solvent mixture to MeCN/MeOH (1/1), the product 68 was obtained in a satisfactory 44% yield. The reaction between acceptor olefins and monoalkylindium reagents cannot proceed via a polar mechanism, as demonstrated in our previous work.29 Finally, we were interested in exploring whether arylindium reagents 2 could serve as XAT reagents (Scheme 7B). To this end, we conducted a reaction between N-methyl-N-phenylacrylamide and cyclohexyl iodide in the presence of reagent 2b under conditions analogous to those recently developed by the Leonori group59 for aryl boronic acids as XAT reagents. Product 69 was isolated in 54% yield in the presence of lithium bromide additive and the photocatalyst 8Cl4CzIPN (1 mol%) under 400 nm LED irradiation for three hours. Supporting the XAT mechanism for this transformation, 4-iodoanisole 1b-I was detected by GC-MS (see SI for details). Photoredox activation of arylindium reagents. Isolated yields are shown. aMeCN/H2O (9/1) (0.5 M) as solvent system was used. bAcONa (2.0 equiv.) was used. cNiCl2 diglyme (2.5 mol%), 4,4′-dimethoxy-2,2′-bipyridine (5 mol%) were used. dMeCN/MeOH (1/1) (0.25 M) as solvent system was used. eIrradiation time 8 h. Mechanistic studies To gain deeper insight into the reaction mechanism, several experiments were carried out. Firstly, the reaction performed under standard conditions but in the absence of a photocatalyst afforded 2a in 93% yield (Scheme 8A). The addition of TEMPO under otherwise identical conditions resulted in inhibition of the photoinduced indation. The reaction was also suppressed in the presence of excited state quenchers, such as triplet molecular oxygen, Ni(acac)2, and Cu(acac)2.60 The reaction of phenylacetylene with the InBr/LiBr system in CD3CN using methanol as a proton donor under irradiation with a 400 nm LED led to the formation of (E)-styrylindium(iii) bromide 73 in 18% yield by 1H NMR, while full conversion of the starting phenylacetylene was observed (presumably due to predominant polymerization) (Scheme 8D) (see SI for details). A similar experiment was demonstrated by the Uchiyama group,41 in which the authors postulated the formation of a photoexcited stannyl anion (T1-stannyl diradical). After mixing InBr and LiBr in acetonitrile at room temperature, the rapid disproportionation with the formation of elemental indium took place (Scheme 8I); then, the supernatant (obtained after centrifugation) was transferred to a vial containing aryl bromide 1a. Irradiation of this homogeneous reaction mixture resulted in the formation of product 2a, as detected by 19F NMR spectroscopy. This observation supports the notion that lithium bromide not only facilitates the disproportionation of InBr but also converts a part of InBr into a soluble form, such as Li+[InBr2]− and other anionic complex species. Irradiation of indium(i) bromide in acetonitrile for 24 hours visually demonstrated its disproportionation, although it is important to note that InBr itself does not disproportionate in acetonitrile in the absence of donor additives (Scheme 8E). To confirm the absorption of visible light by indium(i) complexes, a UV-vis experiment was also conducted (Scheme 8F). Substrate 1a, InBr (which is poorly soluble in acetonitrile; the spectrum obtained essentially corresponds to pure acetonitrile; details of the procedure are described in the SI), and LiBr exhibited no absorption in the visible region. Surprisingly, the InBr/LiBr mixture at low concentrations showed absorption only in the UV region (λmax = 276 nm). The mixture InBr/LiBr/1a displayed absorption in the same region (λmax = 274 nm). Upon irradiation of the cuvette containing the solution of InBr/LiBr/1a, the disappearance of the maximum at 274 nm was observed, indicating the occurrence of the reaction. Mechanistic experiments. aIsolated yields are shown. b19F NMR with p-F-toluene as an internal standard. cYield determined by 1H NMR with 1,3,5-trioxane as an internal standard. dYield determined by GC-FID with tetralin as an internal standard. In contrast, the UV-vis spectrum of the supernatant C obtained from the reaction mixture of 1a/InBr/LiBr displayed a pronounced bathochromic shift, giving rise to a broad absorption band extending into the visible region up to approximately 440 nm. Consistent with this long-wavelength absorption, the supernatant itself exhibited a distinct yellow colour. To verify that this behaviour originated from soluble indium(i) species, supernatant B, prepared from an InBr/LiBr mixture containing 0.25 M soluble indium(i) complexes, was examined. Its UV-vis spectrum exhibited a qualitatively similar absorption profile to that of the reaction supernatant. In contrast, the spectrum of supernatant A, containing half the concentration of soluble indium(i) complexes relative to supernatant B, showed the absorption band shifted toward shorter wavelengths, extending only to approximately 420 nm. The concentrations of soluble indium(i) complexes in supernatants A and B were determined by iodometric titration of aliquots prior to the UV-vis measurements (see the SI for details). This experiment, together with the fact that the yield of reagent 2a decreased when the reaction was carried out under more dilute conditions, suggests that it is not a well-defined In(i) complex responsible for light absorption. It is reasonable to assume that the system contains a cocktail of In(i) complexes46,61 in dynamic equilibrium: upon concentration, cluster complexes (absorbing in the visible region) may form, whereas upon dilution, a monomeric form, which is incapable of absorbing visible light, predominates. Collectively, this set of experiments supports the hypothesis that anionic indium(i) complexes absorb visible light and lead to the formation of excited T1-anionic indium(i) complexes. To substantiate the radical nature of the reaction mechanism, a radical clock experiment was conducted (Scheme 8B). Substrate 1d was subjected to the standard reaction conditions, leading predominantly to the formation of the cyclized product 2ay (resulting from a 5-exo-trig cyclization). Subsequent iodination of the reaction mixture afforded alkyl iodide 70 in 77% isolated yield. We have also shown that when the reaction was carried out in the presence of 10 equivalents of triethyl phosphite or 25 equivalents of N-methylpyrrole, the photoinduced indation was inhibited, and products 71 or 72 were obtained in 61% or 42% isolated yield, respectively (Scheme 8C). This indicates that the reaction proceeds through the intermediate formation of aryl radicals, which are known to be readily trapped by phosphites62,63 or pyrroles.64 For further corroboration of the radical mechanism, the standard reaction with 1e was performed in the presence of C-phenyl-N-tert-butyl nitrone (1e-N) as a radical trap (Scheme 8G), and the corresponding nitroxyl radical 1e-T was observed by EPR spectroscopy, which is consistent with literature data65 (for details of the phenyl radical trapping procedure, see SI). To assess the possibility of a chain mechanism, a light/dark experiment was carried out (Scheme 8H). During the dark phases, a slight increase in the yield of aryl iodide 1a-I was observed in certain segments of the plot, although for the most part the yield of 1a-I did not increase in the dark. This experiment indicates that short chain sequences may be operative.66 For the model photoinduced synthesis of arylindium reagent 2a, the apparent quantum yields (AQY)67,68 were determined to be 6.4% and 7.9% in the absence and presence of the photocatalyst, respectively. Although these values provide a useful measure of the overall efficiency of the process, they neither confirm nor exclude the operation of short radical chain sequences, since apparent quantum yields cannot be directly correlated with the extent of chain propagation in heterogeneous photochemical systems. To gain insight into the reaction pathway in the presence of a photocatalyst, Stern–Volmer fluorescence quenching experiments were performed (Scheme 8J). It was found that there was virtually no quenching of the photocatalyst excited state (PC*, PC = 3DPA2FBN) by lithium bromide or aryl bromide 1a. On the other hand, the supernatant (containing Li+[InBr2]−), obtained after centrifugation of the InBr/LiBr mixture in acetonitrile, served as a fluorescence quencher (see SI for details). This experimental observation points to a single-electron oxidation of Li+[InBr2]− by the photocatalyst excited state to generate an InBr2˙ radical intermediate.69 Concerning the photocatalytic reaction of arylindium reagent 2b with α-trifluoromethylstyrene, a Stern–Volmer fluorescence quenching experiment was also performed (see SI for the analysis of the lines). It was found that lithium bromide, as well as arylindium reagent 2b, served as fluorescence quenchers. As follows from the slopes of the Stern–Volmer plots, quenching was most efficient for lithium bromide. This indicates that lithium bromide undergoes single-electron oxidation by the 8Cl4CzIPN*, which is in accordance with the redox potentials [E(Br−/Br˙) = +0.80 V vs. SCE;70 for 8Cl4CzIPN, E(PC*/PC˙−) = +1.58 V vs. SCE47]. Based on the obtained experimental data, the following mechanism can be proposed (Scheme 9). In the absence of a photocatalyst (Scheme 9A), indium(i) bromide, which is insoluble in acetonitrile, undergoes disproportionation in the presence of lithium bromide. Concurrently, lithium bromide facilitates the dissolution of indium(i) bromide by converting it into an anionic complex A. It is known, for example, that the addition of lithium chloride ensures the solubility of insoluble lanthanide salts in organic solvents.71 It is important to note that A does not represent a specific, well-defined complex but rather exemplifies a broader class of anionic indium(i) species. This anionic indium complex A is promoted to its first singlet excited state 1A upon irradiation with 400 nm LEDs. Due to the heavy atom effect,72,73 rapid intersystem crossing (ISC) occurs, populating the triplet indium intermediate 3A*. Subsequent SET from 3A* to aryl bromide 1 generates an aryl radical, a bromide anion, and an indium dibromide radical (InBr2˙). Radical recombination between the aryl radical and the InBr2˙ radical then yields the arylindium reagent 2. The photoinduced formation of phenylindium reagent 2az from phenyl chloride shown in Scheme 4 is proposed to follow the same mechanistic pathway. Proposed mechanisms of photoinduced indation (A–C) and photoredox activation of arylindium reagents (D). By analogy with nucleophilic main-group radicals that promote XAT elementary step34 and with precedents for HInCl2-derived radical dehalogenation of organic halides,74 InBr2˙ is expected to display sufficient nucleophilic character to engage in an alternative XAT pathway with aryl bromide 1, generating an additional aryl radical together with InBr3. The formation of InBr3 would shift the disproportionation equilibrium, promoting the formation of indium(i) bromide and its subsequent conversion into the soluble anionic complex A. The aryl radical generated in this step could then be intercepted by indium(i) bromide to form an indium radical species B (Ar-InBr˙). The latter radical can then engage in a chain XAT process with aryl bromide 1, leading to the formation of reagent 2 and the aryl radical. This new aryl radical can again be trapped by indium(i) bromide to maintain the chain. In the presence of a photocatalyst (Scheme 9B), the anionic indium complex A is oxidized by the excited photocatalyst with the formation of an indium dibromide radical (InBr2˙). This radical can activate aryl bromide 1via a XAT process, yielding an aryl radical and InBr3. Concurrently, the reduction potential of the 3DPA2FBN radical anion (E1/2 = −1.92 V) is sufficiently negative to directly activate aryl bromides through a SET mechanism. The resulting aryl radical is trapped by indium(i) bromide to form the indium radical B. This species can then initiate a XAT chain analogous to the photocatalyst-free pathway, ultimately leading to the formation of reagent 2. Chain termination can occur through several routes (Scheme 9C). First, direct radical recombination between an aryl radical and an InBr2˙ radical yields reagent 2. Second, combination of the indium radical B with an aryl radical produces a diarylindium reagent 2-D. As previously discussed, the formation of such diaryl species was almost invariably observed by 1H, 19F NMR, and HRMS (ESI-TOF). Third, the photoexcited photocatalyst (PC*) might oxidize a complex formed between the indium radical B and a bromide anion, leading to the formation of reagent 2. The proposed mechanism for the addition of arylindium reagents to unsaturated substrates under photoredox conditions is shown in Scheme 9D. Two types of activation of arylindium reagent may be operative. In the major pathway, a bromide anion is oxidized by the photoexcited photocatalyst (PC*) to generate a bromine radical. This radical then attacks the indium center of arylindium reagent 2, liberating an aryl radical. Alternatively, as a minor pathway, the arylindium reagent 2 itself could be directly oxidized by the photoexcited catalyst to generate the aryl radical. Following the addition of the aryl radical to the unsaturated substrate, the resulting radical intermediate C can accept a single electron from the reduced photocatalyst (PC˙−) to afford a stabilized carbanion D. This carbanion is then either protonated by H2O or MeOH to yield products 66–67 or, in the case of gem-difluorinated substrates, undergoes fluoride elimination to give products 63–65. Conclusions We discovered an unprecedented property of low-valent halide complexes of Group 13 elements to generate highly active species upon visible-light irradiation. An equimolar mixture of InBr/LiBr in solution can undergo photoexcitation by visible light, leading to the formation of the S1 state of the anionic indium(i) complex. After rapid ISC due to heavy atom effect, the T1 anionic indium(i) complex is generated, which participates in SET processes with challenging substrates such as aryl bromides and chlorides. This observation enabled the development of a new approach for the generation of aryl or alkyl radicals from the corresponding precursors, followed by their interception with low-valent indium complexes to afford the desired organoindium reagents. The reaction can also proceed in a photoredox-catalyzed variant, in which the anionic indium(i) complex undergoes single-electron oxidation to give a dibromoindium radical (InBr2˙), which then engages in XAT processes with organic halides. This photoinduced indation protocol is remarkable from both fundamental and practical standpoints. Fundamentally, the visible-light photoexcitation of indium(i) halide complexes and its application to radical generation represent an entirely new reactivity paradigm for Group 13 elements. Practically, this method provides rapid and mild access to a wide array of functionalized organoindium reagents from previously inaccessible aryl and alkyl halides. Furthermore, we have achieved a long-standing goal in indium chemistry: the monoarylindium(iii) dihalides prepared via our protocol participate efficiently in Pd-catalyzed cross-couplings not only with aryl iodides but also with aryl bromides and acceptor-substituted aryl chlorides. Importantly, we demonstrated that the synthetic utility of these reagents extends far beyond traditional two-electron processes. Arylindium reagents are also capable of single-electron oxidation under photoredox conditions, effectively functioning as a reservoir for aryl radicals. This dual reactivity profile establishes arylindium compounds as a bridge between classical transition metal catalysis and modern radical chemistry. Author contributions A. A. Gladkov: conceptualization, methodology, investigation, data curation, writing – original draft. L. I. Panferova: investigation, data curation. V. V. Levin: data curation, writing – review and editing. A. D. Dilman: writing – review and editing. Conflicts of interest There are no conflicts to declare. Supplementary Material