Scalable reductive deuteration of (Hetero)Aryl chlorides with D₂O

Abstract

Deuterated compounds serve as powerful tools for investigating reaction mechanisms, tracing molecular pathways, as well as enhancing properties in medicinal and materials science. Herein, we report a nickel-catalyzed deutero-dehalogenation of abundant yet inert aryl chlorides, enabling direct access to deuterated (hetero)arenes using D₂O as the exclusive, economical deuterium source. This reductive cross-coupling strategy overcomes traditional limitations of aryl chlorides and operates under mild conditions. This protocol delivers products with a high degree of deuterium incorporation across a broad range of (hetero)aryl substrates. It also exhibits excellent functional group tolerance and tolerates various sensitive functional groups including anilines, phenols, and organoboron derivatives. A variety of deuterated products have been efficiently prepared via site-selective chlorination intermediates. Moreover, the method is readily scalable to the kilogram level. Extensive mechanistic studies have been carried out to provide insights into the non-radical Ni^I/N^III catalytic cycle. The simplicity, cost-effectiveness, and scalability of this approach make it highly attractive for applications in drug discovery, mechanistic studies, and metabolic research.

Introduction

Deuterium is a cost-effective and nonradioactive isotope of hydrogen that shares similar chemical properties with hydrogen in specific reactions¹. Deuterated compounds are widely used in mechanistic research for isotope labeling, quantitative analysis, pharmacokinetic analysis, and other research domains^{2,3,4,5,6,7,8,9,10}. In 2017, the U.S. Food and Drug Administration (FDA) reached a significant milestone by approving Austedo (deutetrabenazine) as the first deuterated drug. Since then, over 20 therapeutic compounds incorporating deuterium have progressed through various stages of clinical trials^11,12,13,14. In recent years, deuteration has become a popular method for the discovery of new drugs. Notably, in 2022, the FDA approved deucravacitinib as an innovative de novo deuterated medication^15,16,17. By harnessing the unique biological properties of deuterium, its application in medicine offers significant potential to enhance drug absorption, distribution, metabolism, and elimination. It also minimizes the adverse effects often seen with non-deuterated pharmaceuticals¹⁸. Impressively, following the COVID-19 pandemic, the number of scientific publications mentioning “deuteration” has steadily increased. Additionally, more than 20 deuterated drugs and biomarkers have been utilized in efforts to combat the pandemic. Therefore, developing deuteration methods that are green, efficient, economical, and broadly applicable is essential (Fig. 1A).

Fig. 1: Background of deuteration reactions.

A Deuterated pharmaceutical compounds. B Proportion of aryl chlorides and prices of deuterium sources. C Nucleophilic deuteration with a strong base. D Hydrogen isotope exchange (HIE) reaction. E Deuteration by photo- electro- or mechano-chemistry. F Traditional transition-metal catalyzed deuteration. G Catalytic reductive deuteration of aryl chlorides with D₂O.

The choice of aryl halides as starting materials for deuteration is influenced by their reactivity hierarchy. Aryl bromides and iodides are generally preferred owing to their high reactivity in coupling reactions. In contrast, aryl chlorides, although inexpensive and abundant, have historically been underutilized owing to their poor reactivity in early synthetic steps and their limited potential for structural diversification¹⁹. Deuterated water (D₂O) further enhances practicality as a deuterium source, offering the lowest cost and superior storage stability, making it ideal for industrial or high-throughput deuteration (Fig. 1B). Unlocking the deuteration potential of aryl chlorides remains a key challenge, as traditional metal-halide exchange approaches, often employing strong bases like n-butyllithium to generate organolithium intermediates, face significant limitations. These include harsh reaction conditions (low temperatures), the use of hazardous organolithium reagents, and inherently low efficiency with unactivated aryl chlorides^{20,21,22,23,24,25} (Fig. 1C). To address these issues, C–H activation-based deuteration and hydrogen isotope exchange (HIE) strategies have emerged as alternatives for synthesizing deuterated aromatic/alkyl compounds, wherein functional groups are used for the activation of specific C–H bonds^{26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52}. The activated C–H bonds are then exchanged with deuterium sources, yielding deuterated products. However, these methods require predefined functional groups for C–H bond activation, often resulting in incomplete or inaccurate deuteration, constraints that hinder their adoption in biochemistry (Fig. 1D). In parallel, visible light photoredox catalysis^{53,54,55,56,57,58,59,60,61}, organic electrochemistry^{62,63,64,65,66,67,68,69} and mechanical-force-induced synthesis^70,71 have gained traction as sustainable synthetic tools, enabling mild reductive dehalogenation protocols with D₂O (Fig. 1E). These techniques involve the use of electrophilic reagents to induce the formation of the corresponding radicals or organometal reagents that react with deuterium sources to produce deuteration products. However, these approaches require specific photocatalytic, electrocatalytic reaction modules or mechanical force equipment, which restrict their applicability and feasibility. Furthermore, achieving high efficiency in reactions at the kilogram scale is a challenge for those strategies. Therefore, the development of a scalable and economical deuteration approach is crucial to facilitate the extensive use of this approach.

Given these challenges, transition metal^72,73, such as Pd^{74,75,76,77,78}, Ni⁷⁹, Cu⁸⁰ and Co^81,82, catalyzed deuteration using electrophilic aryl halides has attracted significant attention due to their inherent cost-effectiveness and easy availability. However, many methodologies typically rely on nucleophilic deuterated reagents (e.g., deuteroformates, α-deuterated alcohols, or NaBD₄), proceeded through mechanistic pathways involving β-D elimination and reductive elimination steps (Fig. 1F). Recently, Yao and Zhang reported an efficient palladium-catalyzed deuteration reaction of aryl (pseudo)halides with D₂O as deuterium source⁸³. This efficient method for incorporating deuterium into drug molecules using cheap D₂O offers a practical strategy for late-stage deuteration. Despite the great achievements, the reliance on expensive palladium catalysts and limited reactivity with unactivated aryl chlorides warranted further investigation. To overcome these limitations, this study introduced a methodology employing earth abundant nickel as a catalyst and cheap deuterated water as the primary deuterium source to achieve the deuteration of inactivated aryl chlorides, within a reductive cross-coupling framework leading to higher yields during deuteration (Fig. 1G). This protocol offers several noteworthy advantages: a) utilization of D₂O as a cost-effective deuterium source; b) efficient synthesis of structurally diverse deuterated compounds featuring hydroxyl, amine, or boron functional groups, achieving excellent yields, broad substrate compatibility and exceptional chemoselectivity; c) easy scaling to the kilogram level, demonstrating its potential for application in industrial synthesis; and d) simple and mild conditions, which can be easily achieved in common chemical laboratories.

Results

We initiated our investigation by selecting 5-chloro-1,3-dimethoxybenzene (1) as the electrophile and D₂O as the deuterium source to evaluate the nickel-catalyzed reductive deuteration of aryl chlorides. Following a series of systematic investigations, the optimized conditions were established (Table 1). Under standard reaction conditions, using NiCl₂·DME (5 mol%) as the catalyst, 2,9-dimethylphenanthroline (L1, 6 mol%) as the ligand, zinc (3.0 equiv) as the reductant, and DMA as the solvent afforded the desired deuterated product 2 in 99% yield with more than 99% deuterium incorporation at room temperature over 12 h (entry 1). We found that P-ligands failed to afford the deuterated product (entries 2‒3). Furthermore, reactions using alternative bidentate or tridentate N-ligands including 1,10-phen, bipyridine, and terpyridine afforded the product in lower yields with lower deuterium incorporation, indicating the ligand’s critical role in this transformation (entries 4‒10). Subsequent investigations revealed that other nickel(II) catalysts furnished product 2 in marginally reduced yield and with diminished deuterium incorporation (entries 11‒13). Experiments with varying amounts of D₂O indicated that 5.0 equivalents of D₂O afforded the target product in high yield with excellent deuterium incorporation (entries 14–15). Control experiments underscored the crucial roles of the catalyst, ligand, and reductant. Omitting either the catalyst or reductant completely suppressed product formation, while omission of the ligand afforded the product in reduced yield (30%) and D-incorporation (27%) (entries 16–17).

Table 1 Investigation of reaction conditions^a

To assess the generality of the deuteration method, a broad substrate scope was examined under the optimized conditions. A series of structurally diverse aryl chlorides representing electronically and sterically differentiated systems were evaluated to probe functional group compatibility (Fig. 2). Deuteration of polycyclic arenes yielded the deuterated isotopologues 3–4, 7–9 in good to excellent yields (75–96%) with excellent D-incorporation (95–99%). Moreover, the bromo-, iodo-, and pseudo-halogenated (tosylate) arenes similarly served as competent electrophilic partners under optimized conditions, providing corresponding isotopologue 4 in good yields. Aryl chlorides bearing acetyl, benzoyl, sensitive formyl, and ester groups were efficiently deuterated, providing the corresponding isotopologues 12‒28 in 65–94% yields with 84–99% D-incorporations. It is noteworthy that this reaction is highly chemoselective, cleaving the Csp²‒Cl bond while tolerating the Csp³‒Cl bond (28). Selective deuteration of 3,5-dichloroethyl benzoate resulted in complete substitution of all chlorine atoms, affording the deuterated product 29 in 80% yield and 99% D-incorporation. Electronically diverse substituents including both electron-donating (e.g., 2, 32) and electron-withdrawing groups (e.g., 5–6, 31, 33) were well tolerated by the transformation. Generally, electron-poor substrates afforded consistently lower deuterium incorporation than their electron-rich counterparts, even with a larger amount of D₂O. Additionally, several protected anilines and phenols were successfully deuterated to isotopologues 34‒37 in good yields and D-incorporation. This methodology extended to chlorinated heterocycles, delivering N-, O-, and S-containing isotopologues 38‒47 in 67–90% yields with 85–99% D-incorporation. However, the nitro and aliphatic amine could not be compatible under the standard conditions (products 48, 49).

Fig. 2: Scope of deuteration of aryl chlorides with deuterium oxide.

Reaction conditions: aryl chlorides (0.2 mmol), D₂O (1.0 mmol), NiCl₂·DME (5 mol %), 2,9-di-Me-1,10-phen (6 mol %), Zn (3.0 equiv.), and DMA (1.0 mL) at RT for 12 h under N₂. ^a–I instead of –Cl. ^b–Br instead of –Cl. ^c–OTs instead of –Cl. ^dZn (6.0 equiv.), D₂O (2.0 mmol). Isolated yields.

We next applied the methodology to assess its robustness for substrates bearing strong electron-donating groups (phenols and anilines) as well as organoboron compounds and pharmaceutical agents. (Fig. 3). Deuteration of iodinated phenol substrates typically afforded isotopologues 50‒53 in excellent yields (70–93%) with high D-incorporation (89–99%), though the para-substituted analog 50 exhibited lower D-labeling efficiency. Moreover, dihalogenated phenols bearing mixed halogens (Cl, Br, I) exhibited selective deuteration at the more reactive iodo sites, while chloro and bromo substituents remained intact in products 54 and 55. We also found that para-, meta-, and ortho-chloroanilines 57‒59 afforded the deuterated products in good yields. Anilines 61‒64 exhibited higher yields and deuteration efficiency than 60 and 61, potentially due to electronic effects. Remarkably, the ortho-position displayed a higher degree of D-incorporation than the para- and meta-positions. Subsequently, mixed-halogen substrates 62 and 63 displayed selective iodine deuteration while retaining chloro or bromo substituents. Notably, the reaction was successfully scaled up to 10 grams for ortho-deuterated phenol 51 and aniline 59, providing the products in isolated yields of 89% and 80%, respectively, thus demonstrating its potential for industrial-scale deuterium labeling applications. For these electron-rich phenols and anilines, a reaction temperature of 80 °C was required to achieve good yields and deuterium incorporation. Notably, the developed methodology expresses tolerance to boron-containing groups, which are less frequently reported in traditional metal-catalyzed deuteration. Several aryl chlorides containing –Bpin and –Bcat afforded isotopologues 68–74 in excellent yields (61–92%) and D-incorporation (89–99%). Notably, complete deutero-dehalogenation of multi-halogen organoborons 75‒77 proceeded efficiently, delivering consistently high yields (71–87%) and D-incorporation (98–99%). To demonstrate methodological robustness, late-stage deuteration was successfully applied to pharmaceutically relevant compounds. For instance, chlorpropham underwent efficient deuteration to afford 78 with high D-incorporation (80%). Clofibrate was similarly transformed to its deuterated analogue 79 with 99% D-incorporation. This protocol was further extended to fenofibrate and chloromezanone yielding isotopologues 80 and 81 with deuteration efficiency of 99% respectively. Furthermore, dehalogenative deuteration of esters derived from 2-chlorophenol and pharmacologically relevant carboxylic acids (gemfibrozil, naproxen, and probenecid) afforded deuterated drug derivatives 82–84 with 99% deuterium incorporation.

Fig. 3: Deuteration of aryl chlorides with active functional groups.

Reaction conditions: aryl chlorides (0.2 mmol), D₂O (2.0 mmol), NiCl₂·DME (5 mol %), 2,9-di-Me-1,10-phen (6 mol %), Zn (3.0 equiv.), and DMA (1.0 mL) at RT for 12 h under N₂. Isolated yields. ^a–I instead of –Cl. ^b80 °C instead of RT. ^c10 gram-scale reaction yields. ^dZn (6.0 equiv.), D₂O (4.0 mmol). ^eZn (9.0 equiv.), D₂O (6.0 mmol).

Furthermore, sequential reactions were conducted to synthesize deuterated substrates (Fig. 4). The process began with various (hetero)arenes as starting materials, which underwent highly efficient Jiao chlorination under straightforward conditions^84,85. The resulting chlorinated intermediates were then subjected to dehalogenative deuteration under our standard reaction conditions, yielding the corresponding deuterated products in isolated yields ranging from 62% to 96%, with deuterium incorporation between 86% and 99% (85–95). This strategy demonstrates that a range of (hetero)arenes can be site-selectively deuterated through the combination of the useful Jiao chlorination and our dehalogenative deuteration protocol.

Fig. 4: Stepwise hydrogen isotope exchange via Jiao chlorination.

^aZn (6.0 equiv.), D₂O (2.0 mmol). ^b80 °C instead of RT. Isolated yields.

Encouraged by the successful milligram-scale reactions in dehalogenative deuteration, kilogram-scale reactions were then designed to demonstrate the potential of this approach for practical synthesis (Fig. 5). Two deuterated products containing active substituents such as organoboron and amine derivatives were selected for the synthesis. In previous studies, costly reaction conditions were required for the synthesis of benzen-2-d-amine⁸⁶. This study utilized cheaper conditions to synthesize deuterated anilines on a kilogram scale within 16 hours, achieving a 72% yield (1.02 kg) with over 99% D-incorporation (Fig. 5A). Furthermore, previous studies indicated that expensive palladium catalysts and deuterium sources are necessary for the synthesis of 2-deuterated organoboron⁷⁸. Scaled deuteration afforded 1.05 kg of the boron-functionalized isotopologue in 85% isolated yield with > 99% D-incorporation via simple filtration, extraction and distillation, demonstrating its industrial viability (Fig. 5B). To our delight, this kilogram-scale synthesis was successfully executed with a lower catalyst loading (2.5 mol% Ni, 3.0 mol% ligand), achieving excellent deuterium incorporation and a slight reduction in yield within 36 hours, thus establishing a viable process for industrial-scale deuterium labeling.

Fig. 5: Kilogram-scale synthesis.

A Preparation of benzen-2-d-amine in kilogram-scale. B Preparation of 4,4,5,5-tetramethyl-2-(phenyl-2-d)-1,3,2-dioxaborolane in kilogram-scale.

To gain further insight into the reaction mechanism, several relevant experiments were conducted under the standard reaction conditions (Fig. 6). Radical trapping studies employing (1-cyclopropylvinyl)benzene as a radical clock yielded exclusively the non-rearranged deuterated product in 77% isolated yield (Fig. 6A, up). The observed cyclopropane ring integrity indicates the absence of free radical intermediates under standard reaction conditions. In addition, 1-chloro-2-(4-methylpent-3-en-1-yl)benzene was tested as the aryl halide radical trap. The absence of cyclization products further ruled out the possibility of an aryl radical intermediate (Fig. 6A, down)⁸⁷. Then, we used a series of reductants, including heterogeneous powder reductants, such as Mn and Fe, as alternatives to Zn. The reaction using Mn as the reductant yielded a significant amount of the desired product, albeit lower than with Zn, suggesting that the formation of an organozinc intermediate is not essential for the reaction (Fig. 6B, entries 1–3)⁸⁸. Reactions with TDAE [tetrakis(dimethylamino)ethylene] or B₂pin₂ as organic reductant also produced the deuterated product, though in low yields, further supporting that an organozinc intermediate is not required for the reaction (Fig. 6B, entries 4 and 5). To exclude deeper organozinc intermediate involved, sequential mechanistic studies were performed (Figs. 6C and 6D). Firstly, D₂O quenching experiments were conducted. If organozinc was formed as intermediate, detectable deuteration product would be immediately generated when D₂O was added. However, GC–MS analysis revealed trace amount of deuteration product, again suggesting that an organozinc reagent is not involved in this reaction (Fig. 6C). Secondly, a Pd-catalyzed Negishi-type coupling reaction was employed to probe organozinc intermediates⁸⁹. No coupling products were observed, further confirming that organozinc reagents do not participate in this process (Fig. 6D)⁹⁰. These findings collectively ruled out organozinc involvement in the reaction mechanism. An alternative mechanistic hypothesis was that zinc reacts with D₂O to produce D₂, which could act as a reductant for the nickel catalyst⁹¹. To validate this proposed pathway, a hydrogen reduction experiment was performed. The absence of protonated products confirmed that hydrogen generation via this mechanism is negligible (Fig. 6E). Further mechanistic insights were obtained through kinetic isotope experiments. Competitive kinetic isotope (KIE = 6.14) studies demonstrated that neither aryl radicals nor deuterium radicals participate in the reaction. Parallel experiments (KIE = 1.03) revealed that C–D bond formation is not the rate-determining step (Fig. 6F). Additionally, to investigate the influence of different substituent groups on reaction activity, competitive experiments were conducted using substrates bearing N,N-dimethyl, phenyl, and methylsulfonyl functionalities. The reaction results demonstrated that the methylsulfonyl-substituted substrate yielded a significantly higher deuterated product compared to its counterparts. This outcome strongly suggests that electron-withdrawing groups may facilitate faster reaction rates than electron-donating groups (Fig. 6G). To further clarify the specific relationship between the solvent and deuterium incorporation, a series of deuterated solvents were tested as the deuteration source in the absence of D₂O (Fig. 6H). Experimental results revealed that protons inherent in the reaction solvent can partially influence the deuterium incorporation rate of the product, although the protons in solvent are much less reactive than water (entries 7‒8, Fig. 6H). The incomplete deuteration may be due to the presence of trace amounts of water in the deuterated solvents. These findings serve as direct evidences that reduced deuterium incorporation for some products (15, 18, 31, 45) may arise from protonation with the solvent.

Fig. 6: Mechanistic experiments.

A Radical experiments. B Reductant scope experiments. C Organozinc intermediate experiment. D Sequential Negishi experiment. E Hydrogen reduction experiment. F Kinetic isotope experiments. G Competitive experiment. H Experiments with deuterated solvents.

Kinetic studies were performed to elucidate the reaction mechanism (Fig. 7A). Firstly, a time-course study was conducted. The graph shows a gradual acceleration of the rate as the reaction progressed, followed by a gradual decline after 150 min. Detailed kinetic analysis through GC–MS yields highlighted a first-order dependence with respect to the concentration of NiCl₂·DME. In addition, a zeroth-order dependence on the concentration of aryl chloride and D₂O was observed, indicating that the conversion of an intermediate assembled from both the reagents and the Ni catalyst might be involved in the rate-determining step. To further elucidate the mechanism intermediate, we conducted the following experiments. First, we performed the reaction between 1 and stoichiometric amount of Ni⁰ catalyst, specifically Ni(cod)₂, in the absence of Zn powder. Unfortunately, only trace amount of the desired deuterated product was observed, suggesting that Ni⁰ might not the active catalyst species (Fig. 7B). To obtain further information, complex Ar–Ni^II(2,9-di-Me-1,10-phen)Cl was synthesized. This experimental design was guided by the hypothesis that zinc plays a critical role in reducing Ni^II to a nucleophilic ArNi^I species, thereby facilitating the catalytic cycle. As shown in Fig. 7C, the addition of zinc powder significantly enhanced the product yield, directly supporting the hypothesis that zinc-mediated reduction of Ni^II to ArNi^I is an essential step in the reaction pathway^{92,93,94,95,96}. To further probe the involvement of nickel species, the Ni⁰ and Ni^I complexes were synthesized using established methodologies⁹⁷ and subsequently reacted with D₂O. However, no deuterated products were detected when Ni⁰ was employed as the initial nickel catalyst (Fig. 7D, up). In contrast, the stoichiometric reaction of Ni^I complex with substrate 1 and D₂O furnished the desired deuteration product, although in 11% yield, indicating the Ni^I complex might be involved as a key intermediate (Fig. 7D, down). These results provided unambiguous evidence that Ni⁰ does not function as a key intermediate in this transformation. Consequently, the proposed mechanism involving Ni⁰ I underwent oxidative addition to form Ni^II II, followed by reduction to Ni^I III, which would undergo oxidative addition with deuterium oxide to give Ni^III intermediate IV⁹⁸, and finally reductive elimination to yield deuteration products and Ni^I V might be excluded (Fig. 7E, cycle A). Based on the aforementioned experimental results and literature reports^{92,93,94,95,96,97,98}, we propose a plausible reaction mechanism (Fig. 7E, cycle B). Initially, the Ni^I complex VI could be formed under reductive conditions, followed by oxidative addition of aryl chloride to yield Ni^III complex VII. The resulting intermediate is then protonated by D₂O via a Ni(III)-D₂O species VIII to form the deuterated products. The competitive protonation with solvent would decrease the deuterium incorporation efficiency for some highly reactive substrates. Lastly, the active Ni^I complex VI is regenerated for further catalytic cycles via reduction of Ni^III complex IX.

Fig. 7: Intermediates studies and proposed catalytic cycle.

A Kinetic studies. B Transformation of Ni⁰ species. C Catalytic reactions with Ni^II complex. D Conversions of Ni⁰ or Ni^I complex. E Proposed reaction mechanism.

Discussion

In summary, a versatile dehalogenation method was successfully established for readily accessible aryl chlorides using D₂O as the deuterium source under reductive conditions. This protocol entails reductive deuteration via oxidative addition and protonation through a Ni^I process. Our strategy offers a mild and efficient dehalogenative deuteration pathway for a range of inactivated (hetero)aryl chlorides with excellent functional group tolerance including sensitive phenol, anilines, formyl and boronated groups. Further, carbonyls, esters, amides, and diverse N/O/S-heterocycles remained intact under reaction conditions. More importantly, its applicability has been strongly demonstrated by reactions conducted on the kilogram scale. Notably, this robust strategy necessitates 5.0–10.0 equivalents of D₂O and can be performed at room temperature, leading to a high degree of deuterium incorporation. This rapid and cost-effective methodology for the synthesis of highly deuterium-labeled (hetero)aryl compounds holds significant potential for applications in drug development and metabolic studies. Overall, this approach employs a practical, safe, and cost-efficient deuteration strategy that delivers high-value deuterated compounds. Further investigation on reductive deuteration will be continuously explored in our group.

Methods

General procedure for the nickel-catalyzed deuteration

To a 10 mL Schlenk tube was added sequentially NiCl₂^.DME (2.20 mg, 0.010 mmol), 2,9-dimethyl-1,10-phenanthroline (2.25 mg, 0.012 mmol), Zn powder (40.0 mg, 0.6 mmol). The vessel was evacuated and filled with argon (three times), DMA (0.50 mL) was added via syringe and the mixture was stirred at room temperature for 10 min. The deuterated water (20.0–40.0 mg, 1.0–2.0 mmol) was added, followed by the addition of aryl chloride (0.2 mmol) in one portion. DMA (0.50 mL) was subsequently added via syringe. The resulting solution was stirred for 12 h at room temperature. After this time, the crude reaction mixture was diluted with ethyl acetate (10 mL) and washed with water (2.0 mL × 3). The organic layer was dried over Na₂SO₄, filtered, and concentrated. The residue was purified by flash chromatography.