Step economy strategy for the synthesis of amphoteric aminoaldehydes, key intermediates for reduced hydantoins

Synthesis of 2-formyl aziridines 4

To a vial equipped with a magnetic stir bar were added the organocatalyst 3 (20 mol%, 26.0 mg), the corresponding α,β-unsaturated aldehydes (0.48 mmol, 1.2 equiv.) dissolved in AcOEt (2.0 mL), NaOAc (1.2 mmol, 3.0 equiv., 98 mg) and TsONHBoc (0.40 mmol, 1.0 equiv., 114 mg). The reaction was stirred for 40 min at room temperature or at −20°C, depending on the substrate. After this period, the reaction was extracted with EtOAc (3×15 mL) and the combined organic layers were dried with anhydrous Na₂SO₄ and concentrated under reduced pressure. The crude residue was purified by column chromatography using silica flash and a mixture of hexane: ethyl acetate as eluent.

The enantiomeric excess of compounds 4a–b were obtained after reduction and benzoylation, while the enantiomeric excess of compounds 4c–e were determined by reduction to the corresponding alcohols. Thus, the crude 2-formyl aziridine 4 was dissolved in methanol (2.0 mL) and NaBH₄ (0.8 mmol, 2.0 equiv., 30 mg) was added to the solution. The reaction was allowed to react over 40 min, at 0°C. After this period, the reaction mixture was extracted with EtOAc (2×15 mL) and the product 5 was isolated by flash chromatography on silica gel, using a mixture of hexane: ethyl acetate (20–30%) as solvent. Aliphatic alcohols 5a and 5b (0.25 mmol) were further dissolved in dichloromethane (18 mL) and submitted to react with benzoyl chloride (1.3 equiv.), DMAP (0.5 equiv.) and Et₃N (4.75 equiv.) over 2 h, at room temperature. The reaction mixture was extracted with dichloromethane (3×15 mL) and washed with saturated solution of NH₄Cl (2×15 mL). The combined organic layers were dried using Na₂SO₄, filtered and concentrated. The crude residue was purified by flash chromatography using a solution of hexane: ethyl acetate (5%) affording benzoate 6.

Synthesis of the amino aldehyde dimers 8

A solution of TBAF (0.25 mmol, 1.0 M in THF, 0.25 mL) was added to a solution of the 2-formyl aziridines 4a or 4b (0.25 mmol, 1.0 equiv.) in 2-MeTHF (1.25 mL). The reaction was heated to reflux and stirred over 2 h. After cooling, the crude mixture was extracted with ethyl acetate (3×15 mL) and washed with saturated aqueous NaHCO₃. The products were isolated by flash chromatography, using a solution of hexane: ethyl acetate 40–50%, affording the desired products as colorless oils.

Sequential synthesis of the reduced hydantoins 9

The first step of the sequential synthesis of reduced hydantoins was performed as mentioned above (synthesis of the amino aldehyde dimers). The reduced hydantoin synthesis was carried out following the procedure previously reported by Yudin et al. [21], with slight modifications. In a vial equipped with a magnetic stir bar, isocyanate (1.5 equiv.) was added to the crude solution of amino aldehyde dimers in HFIP:H₂O (8:2). The reaction was stirred over 1 hour, at room temperature. After total consumption of the starting material, the reaction was extracted with ethyl acetate (3×15 mL) and the organic layers were combined, dried over Na₂SO₄, and concentrated under reduced pressure. The residue was purified by flash chromatography using a solution of hexane: ethyl acetate 40–60%, affording the desired compounds in good to moderate yields, after two steps.

Organocatalyzed synthesis of 2-formyl aziridines

Several methods have been described for the asymmetric synthesis of aziridines [25], including the use of organometallic catalysts [26]. Vesely et al. [27] were pioneers to develop an organocatalyzed synthesis of 2-formyl aziridines from α,β-unsaturated aldehydes and protected amines, using dichloromethane as solvent. Thus, based on this work, we started our studies evaluating alternative solvents [28], [29] for this organocatalyzed protocol. Initially, we have examined MeCN, which furnished the desired product in 70% yield (Table 1, entry 1). Polar protic solvents as EtOH delivered the desired product in lower yield (Table 2, entry 2). The bio-based 2-MeTHF [30] also provided a lower yield, in spite of the excellent enantio- and diastereoselectivity (Table 1, entry 3). At 40°C, we observed a higher yield, but a decrease in the diastereoisomeric ratio (dr) (Table 1, entry 4).

Table 2:

Synthesis of 2-formyl aziridines from aliphatic α,β-unsaturated aldehydes.

Entry	Aldehyde	Structure	Yield (%)	ee (%)	drd
1	4a		85a	97b	92:8
2	4b		66a	99b	92:8
3	4c		74	93c	88:12
4	4d		55	99c	92:8

1. Reactions were carried out using 1a (1.2 equiv), 2 (0.4 mmol), NaOAc (3.0 equiv), catalyst 3 (20 mol%) and EtOAc (2.0 mL). ^aIsolated yield after reduction. ^bDetermined by chiral HPLC after reduction and esterification of the aziridine aldehyde. ^cDetermined by chiral HPLC after reduction of the aziridine aldehyde. ^dDetermined by ¹H NMR of the crude product.

Other solvents as toluene, ethyl acetate, trifluoroethanol (TFE) and γ-valerolactone (GVL) were also evaluated (Table 1, entries 5–8). In the case of toluene, low yield and dr were observed. For our delight, EtOAc furnished an excellent yield, ee and dr. Using TFE, however, no product was observed, and the starting materials were recovered after 2 h. When GVL was used, the desired product was observed, but was not isolable due to the high boiling point of the solvent. Therefore, EtOAc has been selected to continue our study.

In order to evaluate the enantiomeric excess (ee) using chiral HPLC, the aziridine aldehyde 4 was reduced with NaBH₄ to the corresponding alcohol 5 followed by esterification with benzoyl chloride, giving the benzoate derivative 6.

With the optimal conditions in hand (Table 1, entry 6), the scope and limitations of this reaction were explored with different α,β-unsaturated aldehydes. Firstly, trans-2-octenal (1b) was employed, furnishing 4b in 66% yield over two steps with excellent ee and dr (Table 2, entry 2). Aldehydes 1c and 1d were prepared from cis-2-butenediol and 3-phenyl-1-propanol, respectively [31], [32], and then submitted to the aziridination reaction affording the corresponding aziridine aldehydes 4c and 4d in good yields and high ee and dr (Table 2, entries 3 and 4).

In order to further explore this reaction, we then evaluated aromatic α,β-unsaturated aldehydes. trans-Cinnamaldehyde (1e) was the first to be tested, but only the α-aminoaldehyde 7e, which is formed via deprotonation and opening of the aziridine ring, was observed [33]. Aiming to obtain the aforementioned desired aziridine aldehyde, we have carried out the reaction at −20°C and also decreasing the base amount (1.5 equiv.), but once more, only the 7e was observed as product (Table 3, entries 1 and 2). The same result was observed when trans-p-methoxycinnamaldehyde (1f) and trans-p-fluorocinnamaldehyde (1g) were employed, affording the corresponding aminoaldehydes 7f and 7g (Table 3, entries 6–8).

Table 3:

Synthesis of 2-formyl aziridines from aromatic α,β-unsaturated aldehydes.

Entry	R	Base (equiv.)	Temp. (°C)	Product (yield %)	ee (%)	dr
1	H	3.0	−20	7a	nd	nd
2a	H	1.5	−20	7a	nd	nd
3	o-MeO	3.0	25	7b	nd	nd
4	o-MeO	3.0	−20	7b	nd	nd
5a	o-MeO	1.5	−20	7b	nd	nd
6	p-F	3.0	25	7c	nd	nd
7	p-F	3.0	−20	7c	nd	nd
8a	p-F	1.5	−20	7c	nd	nd
9b	o-NO2	1.5	−20	4e (57%)	99%	72:22
10b	o-NO2	3.0	−20	4e (72%)	99%	82:18

1. Reactions were carried out using 1a (1.2 equiv), 2 (0.4 mmol), NaOAc (3.0 equiv), catalyst 3 (20 mol%) and EtOAc (2.0 mL). ^aReaction time=2 h. ^bIsolated yield, ee determined by chiral HPLC after reduction of the aziridine aldehyde.

Nevertheless, the reaction with trans-o-nitrocinnamaldehyde (1h) at −20°C with 1.5 equiv. of base furnished the desired product 4h in 57% yield, 99% ee and 78:22 dr (Table 3, entry 9), while using 3.0 equiv. of the base, a better yield and selectivity were attained (Table 3, entry 10). The observed moderate diastereoseletivity can be a consequence of the corresponding dr of the starting material 1h, which was 88:12. These results suggested that only when strong electron-withdrawing groups are present in the aromatic ring of the α,β-unsaturated aldehydes are able to deliver the aziridine aldehyde products in good yields and stereoselectivities.

Having the aziridine aldehydes in hands, we turned our attention to the one-pot Boc deprotection followed by dimerization to generate the amphoteric aminoaldehydes. We initiated this study by evaluating conventional conditions, as shown in Table 4 (entries 1–3). However, in all cases we observed low yields and the formation of several byproducts – which probably can be related to the excess of the aldehyde, which may react via an aza-Michael aldol reaction [34]. Fusco et al. [35] reported the asymmetric synthesis of aziridines via Boc deprotection using TBAF under reflux of THF. Thus, using this condition, we were able to deprotect the Boc group of 4a and isolated the corresponding dimer 8a in 68% yield (Table 4, entry 4). As an alternative, this reaction also occurred using 2-MeTHF as solvent with full conversion, however in 42% isolated yield.

Table 4:

Optimization of the reaction conditions of the tandem Boc deprotection – dimerization.

Entry	Conditions	Yield (%)c
1a	DCM, (C2H5)3SiH, DIPEA, 0°C, 40 min	–
2a	4M HCl, Dioxane, NaHCO3, rt, 2 h	–
3a	Anisole, TFA, H2SO4, rt, 40 min	–
4b	TBAF, THF, 60°C, 2 h	68
5b	TBAF, 2-MeTHF, 60°C, 2 h	42

1. ^aReaction carried out one-pot. ^bReaction carried out with the isolated aziridine aldehyde 4a (0.25mmol), TBAF (1.0 equiv.), solvent (1.25 mL). ^cIsolated yield after purification by column chromatography.

The use of fluoride in Boc deprotection has been reported for other N-heterocycles [36]. According to the literature, this anion can act either by a nucleophilic attack leading to Boc-F or as a base via hydrogen abstraction, furnishing the desired product, isobutene, CO₂ and HF [34]. Due to the low acidity of the hydrogen of compound 4a, we believe that the fluoride could act as a nucleophile.

Synthesis of reduced hydantoins

Since the aziridine aldehyde do not present a high stability and its alkyl-dimer do not show high tolerance to column chromatography purification, which could explain the moderate yields obtained in this deprotection, we decided to submit this crude product to the [3+2] cycloaddition reaction with isocyanates without any further purification [21]. Fortunately, the formyl aziridines readily reacted with isocyanates in hexafluoroisopropanol (HFIP) as solvent [34]. Thus, using the conditions described by Yudin and co-workers, we carried out the reaction of the crude aziridine aldehyde dimers 8 with different isocyanates using a 8:2 mixture of HFIP:H₂O [21]. The reaction of phenylisocyanate with the crude aliphatic dimers 8a–d afforded the corresponding hydantoins 9a–d in good yields over two steps (Scheme 4). Unfortunately, in our hands the trans-o-nitrocinnamaldehyde dimer derivative did not produce the desired reduced hydantoins as previously reported [21].

Scheme 4:

Scope of the synthesis of reduced hydantoins.^a

^aReactions were carried out using 4 (0.25–0.35 mmol, 0.2M), TBAF (1.0 equiv), 2-MeTHF then, after extraction with EtOAc and concentration, HFIP:H₂O (8:2) and isocyanate (1.5 equiv.) were added. Isolated yield over two steps after column chromatography purification. ^bThe separation of compounds 9a and 9c could not be accomplished using the tested chiral columns.

To further explore the scope of this protocol, aromatic and aliphatic isocyanates were employed. Functionalized aliphatic isocyanates containing the benzyl or allyl group afforded the corresponding hydantoins 9e and 9f in good overall yields. Moreover, electron-donating groups in the aromatic ring of the isocyanate gave the corresponding hydantoin 9g in 49% yield (Scheme 4).

In conclusion, in this work we have developed a two-step organocatalyzed-synthetic route to amphoteric aziridine aldehydes. Then, these compounds were employed in the synthesis of new reduced hydantoins in good overall yields, scaffolds with numerous potential applications as buildings blocks in organic synthesis. In both processes, bio-based and/or greener solvents were used. Although there is still room for improvement in this protocol, especially in what regards the use of safer solvents in the purification steps (which already are fewer when compared to the previously reported approaches), we believe that this work is an important step towards the development of more sustainable routes for the synthesis of important heterocyclic compounds.