Cu-catalyzed N-3-Arylation of Hydantoins Using Diaryliodonium Salts.

A general Cu-catalyzed, regioselective method for the N-3-arylation of hydantoins is described. The protocol utilizes aryl(trimethoxyphenyl)iodonium tosylate as the arylating agent in the presence of triethylamine and a catalytic amount of a simple Cu-salt. The method is compatible with structurally diverse hydantoins and operates well with neutral aryl groups or aryl groups bearing weakly donating/withdrawing elements. It is also applicable for the rapid diversification of pharmaceutically relevant hydantoins.

Hydantoin is a heterocyclic scaffold with numerous application areas.[1] The structure has an impressive range of biological activities,[2] likely due to its high density of intermolecular interaction points.[3] It is encountered in several drugs, such as nilutamide, sorbinil, and nitrofurantoin, and constitutes a key molecular component of various agrochemicals.[4] The nucleus also has synthetic versatility,[5] for example, in the preparation of amino acids.[6] Of particular interest are N-arylated hydantoins due to their important biological applications (Figure ).[7]

Figure 1

Some N-3-arylhydantoins with biological properties.

The primary access point to N-arylhydantoins (and other substituted hydantoins) is through de novo cyclization reactions from linear precursors (Scheme ).[8] A plethora of such transformations is known,[9] and each method has advantages and limitations. The direct functionalization of the hydantoin nucleus offers a complementary approach where the substituent(s) of interest are chemo- and regioselectively forged onto the ring itself. This scenario allows divergent modifications of hydantoins as well as potentially quicker and cheaper access to such structures. Direct functionalization remains underdeveloped, perhaps due to its challenging nature.[7c,10]

Scheme 1

Prior Work

For example, the direct N-3-phenylation of unsubstituted hydantoin has, to the best of our knowledge, not been described in the literature.

Recently, a breakthrough was achieved by Petit, Evano, and coworkers (Scheme a), who divulged a Cu-mediated N-arylation protocol.[11] The method exhibited a good reaction scope and mostly good yields. However, it suffered from two disadvantages: (i) the use of stoichiometric Cu and (ii) clear structural limitations of the hydantoin starting material. Whereas 5,5-disubstituted hydantoins were regioselectively arylated at the N-3 position, removing one or both of the C-5 substituents led to significant competing arylation at N-1, limiting the substrate scope. The Cu-catalyzed N-arylation of hydantoins is largely unexplored, and only sparse and highly specialized examples are reported in the literature (Scheme b).[7c,12] Herein we describe a general Cu-catalyzed process for N-3-arylation of hydantoins, a complementary method to the current Cu-mediated protocol (Scheme ).

Diaryliodonioum salts[13] have received increasing attention over the past decade,[14] partially due to their ability to transfer aryl groups to nucleophiles.[15] Our reaction discovery process led to the revelation that unsymmetrical iodonium salts regioselectively arylated the N-3-position of hydantoin in the presence of a simple copper salt and tertiary amine under mild conditions. In contrast, phenyl iodide was an inefficient arylating agent under Cu-catalyzed conditions and did not successfully couple to the hydantoin nucleus.

An overview of the optimization studies is provided in Scheme . The best conditions (entry 1) involved the use of excess aryl(trimethoxyphenyl)iodonium tosylate in the presence of copper(II)nitrate sesquihydrate and triethylamine in toluene at 70 °C for 24 h. Under these conditions, trace amounts (<5%) of the N-1-regioisomer were typically observed as well as small amounts of the corresponding N,N′-bisarylated product (3–10%). The process was also efficient and regioselective when the excesses of phenyl(trimethoxyphenyl)iodonium tosylate was reduced to 1.5 or 1.1 equiv (entries 2 and 3), although the desired target was produced in slightly reduced yield. If a small reduction in yield is tolerable, then the process can therefore be performed with improved atom efficiency. The scope and limitations of the process are shown in Scheme .

Scheme 2

Key Controls and Optimization Data for the Reaction

Yield measured using mesitylene as an internal standard in the 1H NMR analysis of the crude reaction mixture.

Scheme 3

Scope and Limitations of the Catalytic N-3-Arylation Process

5 mol % of catalyst employed.

Reaction performed using 1 mmol of diaryl iodonium salt and extended reaction time (31 h).

Conditions: Hydantoin 1a–g (0.2 mmol, 1.0 equiv), [ArI(TMP)]Tos 2a–s (0.6 mmol, 3.0 equiv), Cu(NO3)2·3/2 H2O (0.02 mmol, 0.1 equiv), triethylamine (TEA) (0.3 mmol, 1.5 equiv), and toluene (2 mL).

Pleasingly, the N-3 arylation proceeded smoothly with structurally diverse hydantoins, including structures lacking substituents on C-5 or N-1.[11] Aryl rings bearing neutral or weakly electron-donating/withdrawing groups were efficiently transferred to the N-3-position on the hydantoin ring. This trend was also observed for disubstituted aryl rings. The method was sensitive to iodonium salts bearing aryl groups congested at the o-position, likely due to steric crowding around the Cu-center in the catalytic process. The relatively small fluoride atom was tolerated, providing the corresponding hydantoins 3j and 3k in moderate yields. The more demanding o-methyl group was not tolerated under our conditions, and the coupled product 3q was not observed in the post reaction mixture. The mild conditions did not cause epimerization at the C-5-position of the hydantoin ring. The coupled chiral hydantoin products ()-3m and ()-3o were obtained in good yield with stereoretention, as indicated by polarimetric analysis. Strongly electron-poor and electron-rich aryl groups were less efficient substrates (3c,d, e), and N-3-arylated hydantoins 3y, 3z, and 3aa were isolated in modest yield.

To our delight, the method was also applicable to a small selection of cyclic imide-type structures[16] (Scheme ) such as succinimide 4a and phtalamide 4b, exemplifying the usefulness of the method in an expanded structural space. The linear imide 4d did not produce the desired product 5d, indicating that the method is not applicable to linear imides, likely due to their less constrained confirmation compared with cyclic imides. Thiohydantoin 4e did not react under our conditions.

Scheme 4

N-Arylation of Other N–H Bonds

Conditions: Imide 4a–d (0.2 mmol, 1.0 equiv), [ArI(TMP)]Tos 2a (0.6 mmol, 3.0 equiv), Cu(NO3)2·3/2 H2O (0.02 mmol, 0.1 equiv), triethylamine (TEA) (0.3 mmol, 1.5 equiv), and toluene (2 mL).

To illustrate the practicality of the method, the well-known antibiotic drug nitrofurantoin 6(17) was N-3-arylated in good yield, providing a small library of nitrofurantoin derivatives 7a–c (Scheme ). The racemate of the aldose reductase inhibitor sorbinil 8(18) was also N-3-arylated in good yield. These examples illustrate the potential methods for the rapid diversification of pharmaceutically relevant agents and highlight the advantage of direct functionalization.

Scheme 5

Regioselective N-Arylation of Pharmaceutically Relevant Agents

Conditions: hydantoin 6 or 8 (0.2 mmol, 1.0 equiv), [ArI(TMP)]Tos (0.6 mmol, 3.0 equiv), Cu(NO3)2·3/2H2O (0.02 mmol, 0.1 equiv), triethylamine (TEA) (0.3 mmol, 1.5 equiv), and toluene (2 mL).

Two different mechanisms involving the Cu(I)/Cu(III) species could be envisaged for this process (Scheme ). Pathway 1 corresponds to the mechanism proposed for Ullman-type C–N bond coupling reactions using aryl iodides.[19] In this process, a Cu(I) species would deprotonate the hydantoin prior to the oxidative addition of the iodonium tosylate. In Pathway 2, the first step would consist of an oxidative addition of the iodonium salt followed by the deprotonation of hydantoin. The latter pathway has been proposed in Cu-catalyzed arylation reactions involving diaryliodonium salts.[20] In both pathways, the final product is formed by reductive elimination of the Cu(III) aryl amido intermediate. The use of Cu(II) precatalyst for Cu(I)-mediated reactions is precedented.[21] Both one-electron reduction and disproportionation mechanisms have been proposed for the reduction of Cu.[22] We believe that in the two mechanisms TEA acts as both a Cu ligand and base, driving the deprotonation of hydantoin. This double function was corroborated by using a catalytic amount of TEA and NaH as a base (see entry 4, Scheme 2). The tosylate anion may also play an important role when coordinated to the Cu center, assisting the intramolecular deprotonation of hydantoin at the N-3 position, which is the most acidic.[22] This could account for the lower yield observed when using bromide or the less basic OTf anion (see entries 5 and 14, Scheme ).

Scheme 6

Two Mechanistic Proposals for the Cu-Catalyzed N-Arylation Process

In summary, we have reported the first general Cu-catalyzed method for the regioselective N-3-arylation of hydantoins. The method utilizes unsymmetrical iodonium salts, a simple Cu salt, and triethylamine under mild conditions. The method is robust and flexible, operates well, with weakly electron-donating and -withdrawing groups, proceeds without epimerization at C-5 of hydantoin, and smoothly arylates structurally varied hydantoins, including pharmaceutically relevant structures such as nitrofurantoin and rac-sorbinil. In-depth mechanistic investigations are currently underway.