KOBu Promoted Selective Ring-Opening N-alkylation of 2-Oxazolines to Access 2-Aminoethyl Acetates and N-Substituted Thiazolidinones

An efficient and simple KOBu promoted selective ring-opening N-alkylation of 2-methyl-2oxazoline or 2-(methylthio)-4,5-dihydrothiazole with benzyl halides under basic conditions has been described for the first time, which provides a convenient and practical pathway for the synthesis of versatile 2-aminoethyl acetates and N-Substituted thiazolidinones with good functional group tolerance and selectivity. KOBu not only plays an important role to promote this ring-opening N-alkylation, but also acts as an oxygen donor.

Introduction 2-Oxazolines are important structural units in pharmaceutical applications and efficient ligands in coordination chemistry, and also valuable protecting or directing groups in catalysis [1][2][3]. 2-Oxazolines are a readily stable class of heterocycles resistant to a range of nucleophiles, bases, or radicals [4,5], which can be easily generated from amino alcohols and carboxylic acids, and from alkenes or epoxides as substrates via alternative synthetic procedures [6]. However, under acidic conditions, oxazolines transform into β-substituted carboxamides through nucleophilic ring opening with S N 2 attack at the C5 position of the ring [7,8]. Recently, Guo's group developed an efficient method for the synthesis of β-nitrate ester carboxamides using tert-butyl nitrite as the nitro source and oxygen as the oxidant through the ring opening of 2-oxazolines [9] (Scheme 1a). Kappe reported a two-step continuous-flow synthesis of N- (2-aminoethyl)acylamides through ring opening/hydrogenation of oxazolines with TMSN 3 as the azide source [10] (Scheme 1a). Coates described a Co 2 (CO) 8 -catalyzed ring-opening hydroformylation of oxazolines for the synthesis of β-amidoaldehydes [11] (Scheme 1a). However, the ring-opening N-alkylation of 2-oxazolines to produce 2-aminoethyl acetate derivatives under basic conditions has not been reported. Thiazolidinone derivatives are important moieties in functional materials and natural products [12][13][14], such as latrunculin that has been obtained from the sponge Cacospongia mycofijiensis [15]. Convenient syntheses of thiazolidinone derivatives are highly attractive to synthetic chemists. However, only a few examples are reported for the synthesis of N-substituted thiazolidinones although the synthetic method of thiazolidinone was described in 1993 [16]. Frost and co-workers explored an efficient ruthenium-catalyzed O-to-S-alkyl migration of N-alkyloxazolidine-2-thiones to synthesize thiazolidinone derivatives through a Barton-McCombie pathway in 2015 [17] (Scheme 1b).

Results and Discussion
To test this ring-opening N-alkylation of 2-oxazoline, benzyl bromide (1a) and 2-methyl-2-oxazoline (2) were chosen as the model substrates for the reaction in the presence of 20 mol % of CuBr 2 , 2 equiv of KOt-Bu in CH 3 CN at 100 °C for 16 h, and a full conversion to the 2-aminoethyl acetate product 3a was obtained (Table 1, entry 1). By changing the copper salt to CuBr or CuI, similar results were detected under the same conditions ( Table 1, entries 2 and 3). Surprisingly, when this reaction was performed without copper salts and decreasing the temperature to 50 °C in CH 3 CN, still a 99% GC yield of the desired product Next, the scope and limitations of this KOt-Bu-promoted ringopening N-alkylation from alkyl bromides with 2-methyl-2-oxazolines were explored using the optimized conditions. As shown in Scheme 2, various benzyl bromides bearing -Me, -t-Bu, -F, and -Cl groups were applied in the synthesis and afforded the tertiary amines 3a-g in 70-85% yield, respectively. Notably, the steric and inductive effects of the substituents did not hamper this ring-opening N-alkylation. Allyl bromide (1h) successfully reacted with 2-methyl-2-oxazoline and produced the corresponding product 3h in good yield (69%). More importantly, bromide-containing enoate derivatives 1i and 1j were easily transferred to the corresponding ring-opened N-alkylated triesters 3i and 3j containing two C=C bonds which could be important as precursors for double Michael additions, and were isolated in 71% and 58% yields.
The above described synthetic system has been evaluated for the ring-opening N-alkylation with benzyl chloride derivatives 4 under similar conditions but at 80 °C, as the chlorides are expected to be less reactive than the corresponding benzyl bromides (Scheme 3). Only 26% yield of 2-aminoethyl acetate compound 3a were observed, however, the addition of 1.0 equiv of I 2 allowed to increase the conversion up to 95% and product 3a was isolated in 78% yield. Other benzyl chlorides bearing -Me, -t-Bu, and -F groups in the para-position were applied to generate corresponding products 3b, 3d, and 3e in moderate to good yields. Furthermore, the reaction proceeded well with heterocycle-containing chlorides such as 2-chloro-5-(chloromethyl)thiophene (4e), leading to 2-aminoethyl acetate product 3k in 66% yield which has potential as bifunctional monomer for polymerization.
Next, the reaction of benzyl bromide (1a) with 2-(methylthio)-4,5-dihydrothiazole was performed under similar conditions but with 2 equiv of KOt-Bu and 2 equiv of I 2 at 80 °C (Scheme 5). However, in this case the N-substituted thiazolidone compound 5a was observed as the only product instead of the above 2-aminoethyl acetate compound. Analogously, the N-substituted thiazolidone derivatives 5a-h were obtained in 63-90% yields from the corresponding benzyl bromides. The electrondonating and the electron-withdrawing groups did not affect these N-alkylation reactions. Interestingly, the reaction tolerates a cyano functional group on the aryl ring of benzyl chloride, and the corresponding thiazolidone 5i (80%), was directly obtained without reaction of C≡N bond. Substrates with pyridyl and thiophene groups were also applied to the synthesis of the corresponding thiazolidine derivatives 5j and 5k in 77% and 50% yields, respectively.
On the other hand, further transformation of the 2-aminoethyl acetate product 3a was investigated, and 88% yield of 2-(dibenzylamino)ethanol (6) was successfully produced in the presence of 2.0 equiv of K 2 CO 3 in MeOH at room temperature for 24 h (Scheme 6). This result indicated that these types of 2-aminoethyl acetate products are useful building blocks for functionalized alcohols.
Next, to gather more information about this reaction, some control experiments were performed under the established ringopening N-alkylation conditions. First, no conversion to the desired product 3a was observed without the addition of KOt-Bu. However, the addition of 0.2 equiv of KOt-Bu gave the 2-aminoethyl acetate product 3a in 39% yield under air and 36% yield under a N 2 atmosphere, whereas 86% yield of 3a were obtained in the presence of 1.0 equiv of KOt-Bu under N 2 conditions (Scheme 7a). These results indicate that KOt-Bu plays an important role to improve the yield of 2-aminoethyl acetate product from this ring-opening N-alkylation. Then, when this reaction is performed in a mixed solvent system (DMC/H 2 O 8:2), only 40% yield of the desired product 3a was produced, while no labeled compound was detected by GC-MS in the 18 O-labeled experiment (Scheme 7b and c). These important results revealed that the oxygen of product 3a does not come from water or air, and it may be transferred from KOt-Bu as supported by Dash and co-worker who demonstrated KOt-Bu can serve as an oxygen source [26]. Therefore, in the present KOt-Bu-promoted ring-opening N-alkylation, KOt-Bu not only plays an important role to promote this type of reaction, but also acts as a nucleophilic oxygen donor during the C=N bond cleavage process to lead the corresponding 2-aminoethyl acetates.
To gain insight into the reaction mechanism, the reaction was repeated in the presence of radical scavengers to evaluate if a radical process is involved in the reaction. Excellent yields of products 3a or 5a were obtained in the presence of the radical scavengers (2,2,6,6-tetramethylpiperidin-1-yl)oxyl (TEMPO), stilbene, or butylated hydroxytoluene (BHT) in the reaction of benzyl bromide with 2-methyl-2-oxazoline or 2-(methylthio)-4,5-dihydrothiazole (Scheme 8). These experimental results suggest that the reaction may proceed through a nucleophilic substitution rather than a radical pathway.
On the basis of the results collected from the control experiments and previous works in the literature, a plausible mechanism for this KOt-Bu-promoted ring-opening N-alkylation is illustrated in Scheme 9. The reaction is proposed to begin with the generation of iminium ether A [27,28], gener-

Conclusion
In summary, we have developed a new and simple transitionmetal-free selective ring-opening N-alkylation of 2-methyl-2oxazoline or 2-(methylthio)-4,5-dihydrothiazole with benzyl halides and allyl halides under mild conditions. Various 2-aminoethyl acetates and N-substituted thiazolidinone derivatives were successfully isolated in moderate to excellent yields. Moreover, in this reaction system, KOt-Bu not only plays an important role to promote this ring-opening N-alkylation, but also acts as an oxygen donor.

Experimental
General procedure for the KOt-Bu-catalyzed ring-opening N-alkylation of 2-oxazolines with benzyl bromides KOt-Bu (0.5 mmol, 56 mg), 2-oxazoline (0.5 mmol), benzyl bromide (1.0 mmol) and DMC (2 mL) were introduced in a tube, equipped with magnetic stirring bar and the mixture was stirred at 50 °C. After 16 h, the progress of the reaction was analyzed by gas chromatography. The solvent was then evaporated under vacuum and the desired product was purified by silica gel chromatography and a mixture of petroleum ether/ ethyl acetate as eluent.
General procedure for the KOt-Bu-catalyzed ring-opening N-alkylation of 2-oxazolines with benzyl chlorides KOt-Bu (0.5 mmol, 56 mg), I 2 (0.5 mmol, 127 mg), 2-oxazoline (0.5 mmol), benzyl chloride (1.0 mmol) and DMC (2 mL) were introduced in a tube, equipped with magnetic stirring bar and the mixture was stirred at 80 °C. After 16 h, the progress of the reaction was analyzed by gas chromatography. The solvent was then evaporated under vacuum and the desired product was purified by silica gel chromatography and a mixture of petroleum ether/ethyl acetate as eluent.
General procedure for KOt-Bu/I 2 -promoted N-alkylation of thiazolidin-2-one derivatives KOt-Bu (1 mmol, 112 mg), I 2 (1 mmol, 254 mg), 2-(methylthio)-4,5-dihydrothiazole (0.5 mmol), benzyl halide (1.0 mmol) and DMC (2 mL) were introduced in a tube, equipped with magnetic stirring bar and the mixture was stirred at 80 °C. After 16 h, the progress of the reaction was analyzed by gas chromatography. The solvent was then evaporated under vacuum and the desired product was purified by silica gel chromatography and a mixture of petroleum ether/ethyl acetate as eluent.

Supporting Information
Supporting Information File 1 Characterization data and copies of NMR spectra.

Funding
We acknowledge the Support of the National Natural Science