Borylation of Aryl Halides Induced by Visible Light Under Open‐To‐Air Environment Authors: Jorge C. Herrera‐Luna, José L. López‐Paz, David Díaz Díaz, M. Consuelo Jiménez, Diego Sampedro, Raúl Pérez‐Ruiz Published: 2026 ABSTRACT Aryl boronate esters play a key role in organic chemistry, making them suitable reagents for subsequent derivatizations, such as preparation of bioactive molecules. The search for novel and sustainable procedures for their synthesis still endeavors the organic community. In this investigation, we developed a new methodology to obtain aryl boronate esters that involved the employment of visible light as energy source, the absence of photocatalyst, and the utilization of purely organic confined media, allowing the reaction to proceed under an open‐to‐air environment. The gel network provided an adequate stabilizing microenvironment to support wide substrate scope for the borylation of commercially available aryl halides containing different functional groups, such as carbonyl, alcohol, esters, amines, and trifluoromethyl, giving in some cases quantitative product yields. Mechanistic evidence from both absorptivity measurements and NMR experiments revealed not only the formation of a ground–state complex but also the nature of this complex via hydrogen bonding. These findings were further explained by DFT calculations to assess the complex formation and the specific photoreactivity induced. Introduction Currently, organoboron compounds are very important building blocks due to their wide applicability, ranging from contemporary organic synthesis to medicinal chemistry or materials science. In particular, aryl boronate esters play a central role in cross‐coupling reactions and late‐stage functionalization strategies, where the development of efficient and sustainable methodologies for their preparation has attracted considerable interest [1, 2, 3]. In line with the principles of green chemistry [4], approaches that operate under mild conditions, avoid transition metals, and minimize energy input are especially attractive. Among catalytic C—H borylation strategies [5, 6, 7, 8], the direct borylation of aryl halides via carbon–halogen (C—X, X=Cl, Br, I) bond activation remains one of the most practical strategies due to the abundance, low cost, and structural diversity of these substrates [9]. Transition‐metal‐catalyzed borylation of organo(pseudo)halides, first described by Miyaura [10], is a powerful method for the synthesis of boronic esters. Catalysts based on Earth‐abundant transition metals such as palladium [11, 12], copper [13, 14], nickel [15, 16], cobalt [17, 18], zinc [19, 20], and iron [21] have been shown to be effective. Limitations of these procedures include the employment of high temperatures together with their high toxicity that require additional removal procedures of trace metal impurities. Photochemical activation has emerged as an attractive alternative, providing a green and sustainable method for efficient synthesis of fine chemicals. Thus, light‐driven borylations of aryl halides have expanded the scope of accessible boron‐containing arenes under mild conditions [22, 23, 24]. Photoinduced reductive activation of aryl halides allows the generation of aryl radicals through single‐electron transfer pathways, which can subsequently engage in synthetically useful bond‐forming reactions. In this vein, diboron reagents such as bis(pinacolato)diboron (B 2 pin 2 ) have been found to be highly effective radical acceptors in these transformations, enabling the direct conversion of aryl radicals into aryl boronate esters [25]. However, to perform the borylation of aryl halides under milder conditions, including photocatalyst‐free, visible‐light irradiation at room temperature under an aerobic atmosphere, seems to be a challenging task. Recently, visible‐light‐induced electron donor−acceptor (EDA)‐based borylation protocols for the synthesis of various aryl boronates under mild reaction conditions were reported [26, 27] although their performance in homogeneous solution was severely compromised under aerobic conditions. We reported a visible‐light‐driven, photocatalyst‐free borylation of thiophenes using B 2 pin 2 with the Hünig's base (N, N‐diisopropylethylamine, DIPEA). The study showed that non‐covalent interactions, especially the hydrogen bond formed between the thiophene carbonyl group and the diboron–base adduct, played a key role in promoting the reaction (Scheme 1A) [28]. Again, this approach was limited by oxygen sensitivity, demanding anaerobic conditions. This drawback was successfully overcome by introducing a supramolecular gel nanoreactor to facilitate aerobic borylation (Scheme 1A) [29]. The gel confined the reaction microenvironment, protecting radical intermediates from oxygen quenching while the ground–state complex absorbed visible light to initiate the reaction. The network of the supramolecular viscoelastic gel avoided undesirable pathways, such as singlet oxygen formation, affording efficient borylation of diverse heteroarenes in air. The substrate scope was, however, restricted to five‐membered, electron‐rich heteroarenes, such as furan, thiophene, selenophene, and pyrrole derivatives. Visible‐light‐driven, photocatalyst‐free borylations of diverse heteroarenes and aryls. Inspired by these findings, the question then arises whether the potential of this photocatalyst‐free borylation not only depends on specific substrate interactions but also could be applied to more common motifs such as six‐membered aryl halides. From a mechanistic point of view, we hypothesize that substrates bearing hydrogen bond acceptors with lone pairs and sufficient electronegativity, such as oxygen, nitrogen, or fluorine could efficiently form the ground state complex via hydrogen bonding (Scheme 1B). In the present work, we have therefore exploited this possibility where the combination of photocatalyst‐free visible‐light activation with supramolecular gel confinement to achieve the borylation of aryl halides under aerobic conditions has been achieved. The in situ generation of the ground state complexes that absorb in the visible‐light region has been demonstrated by spectroscopic measurements whereas a broad variety of products have been obtained, making this strategy an operationally simple, metal‐free, and sustainable route to aryl boronate esters under ambient conditions. To gain further insight into the formation of the ground state complex by hydrogen bonding interaction, a detailed computational investigation using density functional theory (DFT) methods at the M06/aug‐cc‐pVDZ level of theory has been carried out. Results and Discussion Optimizing the Reaction Conditions At the beginning of our investigation, we selected as a model reaction the 4‐bromoacetophenone (1a) as target aryl halide and B 2 pin 2 as the radical trapping agent in the presence of DIPEA as base in a confined medium based on the physical gel formed by G1 (N,N’‐bis(octadecyl)‐L‐boc‐glutamic diamide, molecular structure in Table 1). Hence, irradiation with cold white LEDs under an open‐to‐air environment afforded the desired product 2a with quantitative yield (Table 1, entry 1) after 24 h of irradiation. Control experiments determined the essential role of the supramolecular gel since no formation of the borylated product was detected in acetonitrile (MeCN)/H2O (9/1 v/v) solution (Table 1); furthermore, the absence of light or DIPEA confirmed that these elements played a key role in the chemical transformation (Table 1). TABLE 1Optimization of photocatalyst‐free borylation under visible light irradiation. EntryDeviations from the conditions shown2a Yield/%1— 99 20.1 mmol of 1a instead of 0.02 mmol of 1a 9534 eq of B 2 pin 2 instead of 10 eq7241 eq of DIPEA instead of 2 eq61515 mg/mL of G1 instead of 10 mg/mL of G1 786 G2 (10 mg/mL) instead of G1 587 G3 (40 mg/mL) instead of G1 638Gram scalea 569Natural sunlight instead of LEDsb 7710With TEMPO (10 eq)0 Control experiments Yield of 2a 0% 0% 0% a Conditions: 1a (1 g, 5 mmol), B 2 pin 2 (50 mmol, 10 eq.), DIPEA (10 mmol, 2 eq.), G1 (10 mg/mL), 20 mL of MeCN:H2O (9:1/v:v), 3 days of irradiation under open‐to‐air environment.b 8 h of irradiation.© 2026 WILEY‐VCH GmbH Optimization of photocatalyst‐free borylation under visible light irradiation. Control experiments Yield of 2a 0% 0% 0% Conditions: 1a (1 g, 5 mmol), B 2 pin 2 (50 mmol, 10 eq.), DIPEA (10 mmol, 2 eq.), G1 (10 mg/mL), 20 mL of MeCN:H2O (9:1/v:v), 3 days of irradiation under open‐to‐air environment. 8 h of irradiation. Encouraged by this magnificent result, variations of the above conditions were realized in order to improve the sustainability of the reaction in terms of loading reagents. Thus, a decrease in the amount of the starting material 1a presented practically the same yield for 2a (Table 1, entry 2) whereas reduction either of B 2 pin 2 equivalents or of DIPEA equivalents did not result in better production of 2a (Table 1, entries 3 and 4). Variation of the gel loading revealed a slight dependence of reaction efficiency on the amount of G1 employed (Table 1, entry 5) where the yield of 2a was found to be lower, presumably due to reduction of the diffusion rate of the reactants inside the solvent pools above the optimal G1 concentration. Accordingly, we considered whether the versatility of this photoinduced methodology was exclusively attributable to the self‐assembled architecture of G1, or if gels exhibiting alternative assembly patterns might also function as effective photonanoreactors. As a matter of fact, the model reaction was also achieved in gels G2 (N, N‐((1S, 2S)‐cyclohexane‐1,2‐diyl)‐didodecanamide; see the molecular structure in Table 1) and G3 (1,3:2,4‐dibenzylidene‐D‐sorbitol; see the molecular structure in Table 1) as confined media. Product 2a was successfully obtained in 58% and 63% yield, respectively (Table 1, entries 6 and 7), indicating that the supramolecular viscoelastic gel matrix offers a favorable microenvironment for such photoactivated processes. It is worth mentioning that G1 could be easily separated by filtration (85%–95% of the original weight, see Supporting Information for further details) after the photoreaction. A 1H‐NMR spectroscopic measurement of the recovered G1 confirmed its purity (see Figure S8), and it could be reused in subsequent procedures without detriment to its gelation properties. This finding, therefore, supported the role of viscoelastic gel networks as suitable nanoreactors, which differed from the loss of catalysts activity and/or selectivity with time on stream [30]. The practicability and scalability of this protocol were successfully demonstrated by performing the model reaction at 1 g scale under the standard conditions for 3 days (see Supporting Information for more details), resulting in a 2a yield of 56% (Table 1, entry 8). To quantitatively assess the sustainability of this methodology, key green chemistry metrics were estimated for the gram‐scale model reaction. The transformation exhibits a moderate atom economy (54.7%), which is consistent with the intrinsic limitations of B 2 pin 2 ‐mediated borylation processes. Approximate reaction‐stage PMI and E‐factor values were also estimated for current conditions. The obtained data (see Supporting Information for further details) reveal that the material efficiency is strongly affected by the excess of B 2 pin 2 and by the dilute reaction conditions (Table S4). These limitations are partially mitigated on scale, where the solvent contribution to PMI decreases substantially, and by the possibility of recovering the gelator after the reaction. Therefore, the sustainability advantages of the present protocol arise mainly from its metal‐free and photocatalyst‐free character, visible‐light activation, operation under open‐to‐air conditions, compatibility with sunlight irradiation in the model case, and gelator recovery (see Supporting Information for more details). In addition, the reaction could also be carried out under natural sunlight irradiation, that gave rise to the desired product 2a in 77% yield after 8 h of a whole sunny day (Table 1, entry 9; see Supporting Information for further details). Finally, to provide evidence of the involvement of radical intermediate, the model reaction was carried out in the presence of TEMPO (radical scavenger). The result clearly indicated that the reaction was completely inhibited (Table 1, entry 10). Then, we were interested in confirming the generality of our procedure. For this purpose, the investigation was expanded to the borylation of six‐membered aryl halides containing nitrogen or fluorine (Scheme 2). As representative substrates, 4‐bromo‐N, N‐dimethylaniline (1b) and 4‐bromobenzotrifluoride (1c) were consciously selected as they contain functional groups capable of participating in hydrogen‐bonding interactions, which is presumably crucial for the formation of photoactive ground‐state complexes. To our delight, both substrates reacted under slightly modified conditions compared to the standard ones (vide supra), leading to the corresponding aryl boronate ester in 53% and 51% yield, respectively. This gratifying result supported the mechanistic paradigm that offered an unprecedented approach to the synthesis of valuable aryl boronate compounds. Borylation of 4‐bromo‐N,N‐dimethylaniline and 4‐bromobenzotrifluoride. Scope The scope of this photocatalyst‐free borylation protocol under aerobic conditions was investigated (Scheme 3). Aryl bromides containing carbonyl‐based functional groups showed generally good reactivity. In the case of bromoacetophenone derivatives, the borylation was tolerant in terms of regioselectivity (ortho‐, meta‐ and para‐position) and the corresponding products (2a–c) were successfully obtained. Yields at ortho‐ and meta‐position dropped down in comparison with the para‐position substrate, presumably due to hindrance effects. Other substrates bearing carbonyl groups such as aldehyde or ester functional groups also afforded the desired products (2d–f) in very good yields, whereas amide‐containing substrates were found to be unreactive under the standard conditions (data not shown). Then, we were interested in whether the reaction could occur with not only carbonyl‐based functional groups but also with other oxygen‐based functional groups such as alcohols or ethers. This was the case when using 4‐bromophenol as starting material, where its corresponding boronate ester (2g) was obtained in excellent yield (96%). A similar situation was observed for 4‐iodoanisole as reagent, affording the product (2h) in nearly quantitative yield. The scenario changed when 4‐bromoanisole and 4‐chloroanisole were employed where a pronounced decrease in reactivity was detected, obtaining only trace amounts of the product for the case of the chloride derivative. Scope of the reaction (X = Br, unless otherwise indicated). In the case of low yields, the balance of conversion was the unreactive starting material. Isolated yields in brackets. Next, borylation of halogenated aniline derivatives was studied. The reaction system was applied to bromo‐N, N‐dimethylanilines, to give the boronic ester anilines 3a–c in moderate‐to‐good yields. To note that better results for product 3c in comparison with its analogous 2c could be explained in terms of electronic effects, since substituted anilines are richer in electrons than the acetophenone scaffolds. A non‐conjugated amine such as 4‐bromo‐N, N‐dimethylbenzylamine was successfully borylated (3d) in 72% yield, demonstrating that conjugation between the aromatic ring and the heteroatom was not a strict requirement for reactivity. In addition, the borylation of non‐methylated anilines proceeded to give 3e and 3f in 38% and 10% yield, respectively. The methodology was further extended to heteroaromatic systems. Thus, 2,5‐dibromopyridine and 3‐bromoquinoline provided the corresponding borylation products 3g and 3h in 22% and 49% yield, respectively. To mention that employment of 2‐bromoquinoline as starting reagent led exclusively to the dehalogenated product (data not shown), indicating that subtle changes in halogen position can profoundly influence the reaction outcome. Finally, commercially available bromobenzotrifluoride substrates showed good reactivity for para‐ and meta‐substituted reagents (4a and 4b), whereas ortho‐substituted analog failed to undergo borylation (4c) due to presumably steric hindrance and less favorable hydrogen‐bonding interactions. Mechanism Insights A series of measurements were conducted to elucidate the reaction mechanism of borylation. Control experiments (vide supra) confirmed the need of the three components (1 + B 2 pin 2 + DIPEA). Hence, we first investigated the interaction between these reagents in the ground state by means of UV–Vis absorption spectroscopy. A mixture solution of 1a, B 2 pin 2 and DIPEA under reaction conditions generated a new band in the visible region which indicated the formation of a ground–state complex (Figure S2). The same scenario was observed for both cases 1b + B 2 pin 2 + DIPEA and 1c + B 2 pin 2 + DIPEA (see Figures S3–S4), permitting the absorbance in the visible region and, therefore, the beginning of the photoreaction. To gain further insight into the nature of the ground‐state complexes, NMR experiments were carried out (Figure 1). Based on our previous study [28], a sp3–sp2 diboron species (I) was formed when B 2 pin 2 was treated with DIPEA in water. Thus, the 13C NMR spectrum of 1a in the presence of B 2 pin 2 +DIPEA revealed that the carbonyl signal (δ 199) shifted gradually to high field upon increasing amounts of B 2 pin 2 together with the appearance of a new weak peak, suggesting a hydrogen‐bonding equilibrium between I and the carbonyl group of 1a (Figure 1A). In a similar measurement involving 1b in the absence/presence of B 2 pin 2 +DIPEA (Figure 1B), a low‐field shifting of the aromatic carbons at δ = 108 (C‐Br) and at δ = 150 (C—N) was observed with increasing amounts of B 2 pin 2 ; furthermore, the carbons at the methyl groups attached to the N atom presented the same trend (Figure 1C), confirming again the possibility of a hydrogen‐bonding leading to the formation of 1b···I ground state complex. Finally, the 19F NMR spectroscopic analysis of the starting material 1c in the absence/presence of B 2 pin 2 +DIPEA (Figure 1D) was performed. Interestingly, the fluorine signal (δ − 63.2) shifted weakly in the presence of increasing amounts of B 2 pin 2 , pointing out an interaction between 1c and I. Therefore, these experimental results indicated the formation of a three–component complex in the ground state, where hydrogen bond acceptors with lone pairs and sufficient electronegativity such as oxygen, nitrogen or fluorine played a pivotal role in this type of processes. Hydrogen bonding ground state complex evidence detected by NMR spectroscopy. (A). (a) 13C NMR spectrum of 1a (20 mg, 0.1 mmol) in 0.6 mL CD3CN/D2O 9/1. (b) 13C NMR spectrum of 1a (0.1 mmol) + B 2 pin 2 (0.2 mmol) + DIPEA (0.2 mmol) in 0.6 mL CD3CN/D2O 9/1. (c) 13C NMR spectrum of 1a (0.1 mmol) + B 2 pin 2 (0.5 mmol) + DIPEA (0.2 mmol) in 0.6 mL CD3CN/D2O 9/1. (d) 13C NMR spectrum of 1a (0.1 mmol) + B 2 pin 2 (1 mmol) + DIPEA (0.2 mmol) in 0.6 mL CD3CN/D2O 9/1. (B and C). (a) 13C NMR spectrum of 1b (20 mg, 0.1 mmol) in 0.6 mL CD3CN/D2O 9/1. (b) 13C NMR spectrum of 1b (0.1 mmol) + B 2 pin 2 (0.2 mmol) + DIPEA (0.2 mmol) in 0.6 mL CD3CN/D2O 9/1. (c) 13C NMR spectrum of 1b (0.1 mmol) + B 2 pin 2 (0.5 mmol) + DIPEA (0.2 mmol) in 0.6 mL CD3CN/D2O 9/1. (d) 13C NMR spectrum of 1b (0.1 mmol) + B 2 pin 2 (1 mmol) + DIPEA (0.2 mmol) in 0.6 mL CD3CN/D2O 9/1. (D). (a) 19F NMR spectrum of 1c (15 µL, 0.1 mmol) in 0.6 mL CD3CN/D2O 9/1. b) 19F NMR spectrum of 1c (0.1 mmol) + B 2 pin 2 (0.2 mmol) + DIPEA (0.2 mmol) in 0.6 mL CD3CN/D2O 9/1. (c) 19F NMR spectrum of 1c (0.1 mmol) + B 2 pin 2 (0.5 mmol) + DIPEA (0.2 mmol) in 0.6 mL CD3CN/D2O 9/1. d) 19F NMR spectrum of 1c (0.1 mmol) + B 2 pin 2 (1 mmol) + DIPEA (0.2 mmol) in 0.6 mL CD3CN/D2O 9/1. Computational Studies In order to gather deeper insight into the reaction mechanism, we performed a series of theoretical calculations on the critical steps along the reaction mechanism. As supported by experimental evidence and previous reports [28, 29], the formation of an EDA complex in the ground state modifies the optical properties of the precursor aryl halide, allowing the photochemical reaction. Thus, the formation of the active boron species I, the formation of the complex, and its optical properties were subject to computational studies within the framework of DFT and its time‐dependent (TD‐DFT) version (see Supporting Information for Computational Details). As the first step in the mechanism, the formation of the diboron active species experimentally requires both water and DIPEA to take place. This was experimentally tested under the required reaction conditions and further confirmed by NMR analysis. We computationally further explored the thermodynamic viability of this transformation to assess the potential generalization of the process (see Figure S9). As can be seen, two consecutive steps implying first the interaction between B 2 pin 2 and a water molecule, followed by a hydrogen atom capture by the base, confirm a thermodynamically viable complex at the M06/aug‐cc‐pVDZ level. In turn, this may set the ground for alternative processes with related compounds; in fact, previous studies employing other nitrogen‐like bases or diboron derivatives allowed the reaction [28, 29]. This intermediate I was experimentally found to interact with hydrogen bond acceptor molecules 1a‐c (Figure 1) forming an EDA ground state complex. The energetics and chemical structures of these complexes were further analyzed computationally. In all cases (1a, 1b and 1c) formation of the complexes proved to be exergonic suggesting a favorable formation under the reaction conditions (see Table 2). This agrees with NMR and UV data, as the combination of 1, DIPEA and B 2 pin 2 in ACN/water leads to the formation of new chemical species. Figure 2 shows the computed structure of the EDA complex formed for 1a, as a representative example. Computed EDA complex geometry formed upon addition of 1a, DIPEA, and B 2 pin 2 . Hydrogen bonding experimentally found by NMR is shown, distance in Å. TABLE 2Relative energies of the EDA complexes for 1a‐1c.CompoundEDA complexa 1a 8.39 1b 2.28 1c 7.38a Energy in kcal/mol related to the corresponding separated species.© 2026 WILEY‐VCH GmbH Relative energies of the EDA complexes for 1a‐1c. Energy in kcal/mol related to the corresponding separated species. Due to the quantitative ground state complex formation, the optical properties of the reaction mixture are modified. To further explore the changes in the optical properties (i.e., new, red‐shifted band), TD‐DFT calculations were performed to unveil the changes caused by the complex formation. Figure 3 shows the computed UV‐spectra of 1a and the EDA complex 1a‐I at the M06/aug‐cc‐pVDZ level as an example. No shifting of the computed spectra was performed, as the shape and relative intensities match the experimental results (see Figure 3, inset). Additional functional benchmarking was also performed, with similar results (see Table S5). Corresponding figures for 1b and 1c are included in the Supporting Information (Figures S10 and S11). (a) Optical properties for 1a (blue line) and the EDA boron complex (black line). Both spectra were normalized against the same global maximum to compare relative intensities across the two spectra; no shifting was applied to fit the experimental data. (b) Molecular orbitals involved in the relevant transition. (c) Experimental absorption of 1a (red line) and EDA complex (purple line). A new red‐shifted band appears after the formation of the complex. On one hand, this allows for light absorption under the reaction conditions. This new band is associated with a transition between an orbital mostly centered in I and an orbital centered in 1 (see Figure 3). On the other hand, this excited state is responsible for the photochemical reactivity. The population of the π* orbital eventually leads to halogen atom extrusion and product formation as described previously [29]. Conclusion Overall, this investigation demonstrated that the combination of photocatalyst‐free visible‐light activation with supramolecular gel confinement provides an efficient and sustainable platform for the borylation of aryl halides under ambient and aerobic conditions. The in situ formation of ground‐state complexes enabled by non‐covalent interactions between the aryl halide and the B 2 pin 2 ‐base adduct allowed absorption in the visible region and triggered radical pathways leading to the corresponding aryl boronate esters. The gel matrices acted as effective nanoreactors, likely protecting reactive intermediates from oxygen quenching while supplying a structured microenvironment that enhanced reactivity and broadened the substrate scope to various aryl halides bearing hydrogen bond acceptor functionalities. Experimental evidence combined with spectroscopic studies strongly supports this mechanistic scenario, whereas complementary computational analysis provides additional insight into the nature of the supramolecular assemblies involved. Altogether, these findings established a robust, scalable, metal‐free, and operationally simple strategy for the preparation of aryl boronates, contributing to the advancement of green and practical photochemical methodologies. Funding This work was supported by the Ministerio de Ciencia, Innovación y Universidades (PID2023‐148367OBI00 and PID2024‐161126NB‐I00). Conflicts of Interest The authors declare no conflicts of interest. Supporting information The authors have cited additional references within the Supporting Information [31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47].