Deuteration of Formyl Groups via a Catalytic Radical H/D Exchange Approach

Abstract

H/D exchange at formyl groups represents the straightforward approach to C-1 deuterated aldehydes. This transformation has been recently realized by transition metal and NHC carbene catalysis. Mechanistically, all these processes involve an ionic pathway. Herein we report a distinct photoredox catalytic, visible light mediated neutral radical approach. Selective control of highly reactive acyl radical in the energy barrier surmountable, reversible reaction enables driving the formation of deuterated products when an excess of D₂O is employed. The power of the H/D exchange process has been demonstrated for not only aromatic aldehydes, but also aliphatic substrates, which have been difficult in transitional metal catalyzed H/D exchange reactions, and for selective late-stage deuterium incorporation into complex structures with uniformly high deuteration level (>90%).

Graphical Abstract

Among isotopes, deuterium perhaps has the broadest impact on almost every sub-discipline in the life, and chemical, material and nuclear sciences and beyond.¹ The recent surge in applications of deuterated pharmaceutical agents has been witnessed by the FDA approval of the first deuterated drug, austedo (deutetrabenazine), in 2017² and the large number of emerging deuterated drug candidates.^{1c, 1d, 2, 3} This has spurred considerable interest in developing synthetic methods that enable to efficiently generate deuterated building blocks.^4,5

Given the broad availability and synthetic versatility of aldehydes, C-1 deuterated aldehydes can provide the quick access to a wide range of highly valued and structurally diverse deuterated building blocks and to pharmaceutically relevant structures.⁶ Several methodologies have been reported for their synthesis (Scheme 1). Conventional methods rely on the reduction and/or then oxidation sequence but using expensive deuterated reducing agents, such as LiAlD₄⁷ and Cp₂ZrDCl.⁸ Recently, more cost-effective approaches using cheap, safe and readily handled D₂O as the deuterium source have been elegantly realized from aryl iodides,⁹ carboxylic acids,¹⁰ or benzyl halides¹¹ by Denmark, Xie, and Sheng, respectively (Scheme 1a). Direct hydrogen-deuterium exchange (HDE) processes with aldehydes without requiring additional functional group transformation represent an even more synthetically efficient strategy. Tuttle and Newman¹² independently reported Ir and Ru catalyzed HDE reactions (Scheme 1b). Despite these impressive studies, it is recognized that significant synthetic challenges are associated with the difficulty of controlling chemoselectivity of non-selective deuteration of aromatic rings and the unsatisfactory deuteration level (14–84%). Furthermore, the reported protocols (Scheme 1a and 1b) mainly work for aromatic aldehydes. More recently, we¹³ and Bertrand and Yan¹⁴ have developed more efficient NHC carbenes promoted HDE processes (Scheme 1c). Mechanistically, all these HDE transformations rely on the ionic reaction pathway.

Scheme 1.

Methods for the synthesis of C-1 deuterated aldehydes.

A radical process for direct HDE offers an alternative route. Implementing the distinct neutral radical process for direct HDE requires the in situ generation of an aldehyde derived acyl radical. A process using acyl radical as a key intermediate has been nicely realized by Xie and co-workers for the synthesis of C-1 deuterated aldehydes (Scheme 1a).¹⁰ In their studies, acyl radicals are produced from carboxylic acid precursors using visible light mediated photoredox deoxygenation with phosphines. Furthermore, direct abstraction of formyl hydrogen with photoredox strategy has been successfully used by Glorious and us for new bond formation recently.^15,16 These processes, involving new bond connection, are generally kinetically controlled and irreversible. It is recognized that the realization of the reversible acyl radical¹⁷ involved chemistry for HDE process must overcome the challenge that the control of intrinsic reversibility enables to manipulate the equilibrium to achieve synthetically useful yields of the desired product. Moreover, achieving the thermodynamic control process could be complicated by an undesired decarbonylation process (Scheme 2).^18,19 Studies indicate that the decarbonylation process is unfavored.²⁰ With these in mind, it is possible to realize the strategy of combining a thermodynamic approach with the use of one of the reaction components in a large excess. Furthermore, the stronger C-D bond (bond dissociation energy (BDE) of C–H as 338 kJ/mol and C-D as 341.4 kJ/mol) also could enhance the C-D bond formation. The investigation from MacMillan and coworkers in an elegant photoredox mediated HIE (hydrogen isotope exchange) for selective deuteration and tritiation of α-C-H of amines make possible for our new HDE for formyl group.^5b

Scheme 2.

A plausible mechanism.

Herein, we wish to report the results of the study, which lead to a visible light mediated photoredox catalytic H/D exchange at formyl groups, (Scheme 1d). In the preparation of this manuscript, a similar work was reported by Wang and coworkers.²¹

Our investigation began with a model deuteration reaction of p-tolualdehyde with D₂O in the presence of a photocatalyst and a HAT (hydrogen atom transfer) catalyst. 9,10-Phenanthrenequinone 2a used as a HATphotocatalyst in our previous study¹⁶ gave 23% deuterium incorporated product 3a (Table 1, entry 2). Switching to the photoredox catalytic system of 4CzIPN as photosensitizer (PS) and methyl 2-mercaptoacetate 2b as HAT gave slightly improved deuterium incorporation (40%, Table S1), but the results were not reproducible. In our extensive reaction optimization effort (Table S1), we accidently found impurity of sodium dodecyl sulfate (SDS) from the detergent contaminated reaction flask could generate reproducible results (40% D incorporation). We surmised that sodium dodecyl sulfate (SDS) might serve as more effective HAT. Glorius et al. found that benzoyloxy radical PhCOO•, produced from PhCO₂Na in the presence of a PS, could selectively and efficiently abstract the hydrogen atom from the formyl group.¹⁵ Therefore, we probed PhCO₂Na. To our delight, a level of 95% D-incorporation and 87% yield were obtained (entry 3). Without PhCO₂Na, almost no deuteration product was detected (entry 7). Screening of thiol HAT catalysts revealed triisopropylsilanethiol 2d to be optimal one (entry 1 vs 3 and 4) and variation of reaction media led to the choice of ethyl acetate (entry 1 vs 5 and 6). In control studies, HAT 2d (entry 8) and light and photocatalyst (entry 9) are necessary for the deuteration process. When the reaction is open to air, p-tolualdehyde was oxidized to benzoic acid (entry 10). The efforts on the optimization of the reaction led to the optimal reaction conditions as follows: 4CzIPN (5 mol%) as photoredox promoter, PhCO₂Na (30 mol%) and triisopropylsilanethiol (2d, 30 mol%) as HAT and D₂O (40 equiv) in anhydrous EtOAc (0.2 M) by 5 W blue LED strip under N₂ at rt. To reduce the introduction of extra protium in reaction as much as possible, isotope exchanging of 2d was performed by stirring 2d solution of anhydrous EtOAc with 10 equiv D₂O for 1 h.

Table 1.

Optimization of the reaction conditions.

Entry	Variation of standard conditionse	Yield [%]b	D ratio [%]d
1a	None	95 (91[c])	97
2	2a instead of 4CzIPN and PhCO2Na	89	23
3	2b instead of 2d	87	95
4	2c instead of 2d	91	82
5	MeCN instead of EtOAc	93	16
6	DMF instead of EtOAc	94	<5
7	without PhCO2Na	93	<5
8	without 2d	93	<5
9	no light or no photocatalyst	95	<5
10	open to air	<5	N.D.

^[a]Standard reaction conditions: 1a (0.2 mmol), 4CzIPN (5 mol%), Sodium benzoate (30 mol%), 2d (30 mol%), D₂O (40 equiv), anhydrous EtOAc (0.2 M), 5 W blue LED strip, N₂, 36 h. Isotope exchanged 2d of 1 M anhydrous EtOAc solution was prepared by following procedure: dissolve 2d (1.0 mmol) in 1.0 mL of anhydrous EtOAc followed by adding D₂O (10 mmol, 1/10 molar ratio of 2d/D₂O); after 1 h stirring, the top EtOAc layer was separated and used in reaction.

^[b]1H NMR yield.

^[c]Isolated yield.

^[d]D (deuteration) % determined by ¹H NMR.

^[e]All solvents are anhydrous. N.D.: not detected.

Based on these results, a possible reaction mechanism is proposed (Scheme 2). Irradiation of photocatalyst 4CzIPN by visible light generates the excited state 4CzIPN*, which oxidizes sodium benzoate 5 to give benzoyloxy radical 7 for a subsequent HAT. The key acyl radical 4 is formed by abstraction of a hydrogen from aldehyde precursor 1. Then the second HAT process between resulting acyl radical 4 and deuterated thiol 10, delivers deuterated aldehyde product 3 and concurrent generation of thiyl radical 8. Finally, the thiyl radical is reduced by the 4CzIPN•‾ to give the thiolate 9 and 4CzIPN to complete the redox cycle. The more basic thiolate anion 9 serves as an internal base to deprotonate the benzoic acid (pKa = 10) to regenerate HAT catalysts 5 and 2d. The latter HAT catalyst can be transformed to deuterated thiol 10 by H-D exchanging (HDE) with D₂O. It is also possible that 9 can be directly protonated to give 10.

With the optimized protocol in hand, the scope of deuteration process was probed. The methodology serves as a mild, general approach for the synthesis of a wide array of deuterated aromatic aldehydes in good yields (up to 99%) and with uniformly high levels (90–98%) of deuterium incorporation (Scheme 3). In general, the substrates of aromatic ring containing electron donating groups (MeO, AcO, AcHN, morpholine) gave slightly better results. It is believed that their enhancement in the nucleophilicity of acyl radical provides higher reactivity in the electrophilic HAT process. Notably, various functional groups, such as free hydroxy (3j and 3k), halogen substituents (3g, 3h, 3i), ester (3l, 3s), and allylic (3m) can be tolerated by the deuteration protocol. Moreover, heteroaromatic aldehydes underwent deuteration smoothly to afford desired products (3p, 3q, 3r).

Scheme 3.

Reaction scope for aromatic aldehydes^a

^aSee the Supporting Information for detailed reaction conditions. Isolated yields (Y); D% determined by ¹H NMR spectroscopy.

Having demonstrated the capacity of the aromatic aldehydes to participate in this radical engaged deuteration process, we turned our attention to aliphatic aldehyde substrates, which have remained an unsolved synthetic challenge for their C-1 deuteration (Scheme 4). Gratifyingly, aliphatic aldehydes are competent substrates, affording the corresponding C-1 deuterated products (3t-3aa) in 52−99% yields and with uniformly high level of D-incorporation (93–98%) under the optimized mild reaction condition (Scheme 3). It is noteworthy that under the mild reaction conditions, the formation of decarbonylated side products is minimized. Nonetheless, in many cases, D-incorporation into other acidic C-H positions, in particular enolizable α-position is observed. Again, a wide array of functional groups such as radical sensitive C=C (3aa, 3am), amide (3ab), ketone (3ac), ester (3ad), and cyanide (3ae and 3af) are tolerated. Moreover, the reaction can be applied to generate aliphatic deuterated aldehydes, which contain synthetically and biologically relevant heterocycles, including pyridine (3ag), indole (3ah), furan (3ai), thiophene (3aj) and pyrimidine (3ak). Finally, aliphatic aldehydes bearing long chain also work smoothly (3al and 3am). It is found that isomerization of the cis C=C double bond in 3am is observed.

Scheme 4. Reaction scope for aliphatic aldehydes^a.

^aSee the Supporting Information for detailed reaction conditions. Isolated yields (Y); D% determined by ¹H NMR spectroscopy.

The capacity of the mild deuteration methodology is also demonstrated in selective C-1 deuteration of aldehyde functionality in complex pharmaceutically relevant structures (Scheme 5a). Native ursodeoxycholic aldehyde, mycophenolic aldehyde and marketed drug indomethacin derived aldehydes are selectively deuterated in good yield and with high deuteration (3an-3ap). Furthermore, ribose- amino acid tryptophan and dipeptide Phe-Gly aldehyde derivatives can be labelled by deuterium with high level deuterium decoration (91–96%, 3aq-3as).

Scheme 5.

Deuteration of aldehyde moiety in structurally complex structures, gram scale synthesis, recycling and reuse of D₂O and intermediate trapping study^[a]

^aSee the Supporting Information for detailed reaction conditions. Isolated yields (Y); D% determined by ¹H NMR spectroscopy.

The method can be used in a gram-scale synthesis of C-1deuterated aldehydes. In comparison to the small-scale process, as shown using p-tolualdehyde 1a as an example, the gram-scale counterpart 3a is formed in a similar yield and with a comparable level of D-incorporation although reduced amounts of reagents and catalysts (only 20 equiv of D₂O, 1.5 mol% 4CzIPN, 20 mol% PhCO₂Na and 8 mol% triisopropylsilanethiol) are used (Scheme 5b). Furthermore, the recovered D₂O containing solvent can be used in a second and third reaction without causing a significant decrease in yield and D-incorporation level (Scheme 5c). These studies demonstrate that the cost for the synthesis of deuterated aldehyde products can be further reduced. Finally, to verify the proposed mechanism, the formed acyl radical was trapped by TEMPO. The product was confirmed by ESI-MS analysis of the crude reaction mixture (Scheme 5d) (SI, Section 6).

In summary, we have developed a visible light mediated, organocatalyzed hydrogen-deuterium exchange (HDE) process for directly converting readily accessible aldehydes to their 1-deutero counterparts using D₂O as the deuterium pool. Distinct from the established transition metal catalyzed ionic HDE processes, this organophotoredox catalytic radical strategy is successfully realized. Notably, this approach not only enables direct HDE of aromatic aldehydes without deuteration on aromatic ring, but also for aliphatic aldehydes and selective late-stage deuterium incorporation into complex structures with uniformly high level (>90%) of deuterium incorporation. We anticipate that the approach will enable facile access to a wide range of deuterated structures, which are highly valuable in the fields of biological, medicinal and organic chemistry.