Organocatalytic Enantioselective α-Nitrogenation of α,α-Disubstituted Aldehydes in the Absence of a Solvent

A highly efficient enantioselective α-nitrogenation method of α,α-disubstituted aldehydes with azodicarboxylates promoted by a chiral carbamate-monoprotected cyclohexa-1,2-diamine as organocatalyst has been developed. The process was carried out without any solvent, and the corresponding α,α-disubstituted α-nitrogenated aldehydes were obtained with excellent yields and enantioselectivities up to 99% ee. The sustainability of the procedure was established through the calculation of green metrics, such as EcoScale and E-factor. In addition, theoretical calculations have been used to justify the obtained enantioselectivity sense.


■ INTRODUCTION
Enantioenriched α-nitrogenated aldehydes are important building blocks in chemical synthesis, with many applications in medicinal chemistry and the pharmaceutical industry. 1−5 The aldehyde functionality can be transformed into various functional groups, leading to substituted chiral amines in natural products and bioactive substances. 6 Particularly interesting is the synthesis of quaternary α-amino aldehydes as they can be transformed into quaternary α-amino acids, which are building blocks for constructing peptidomimetics or valuable as pharmaceuticals. 7−11 An example of the latter is αaryl−α-alkyl α-amino acid derivatives, which have shown strong inhibitory effects on aldose reductases, a potential target for treating various diabetes-related diseases. 12 Although not too extensively explored, organocatalysis has been found as a direct and convenient method for the asymmetric synthesis of quaternary α-amino aldehydes through the conjugate addition reaction of α,α-disubstituted aldehydes with azodicarboxylates as the electrophilic nitrogen source. Chiral amine-containing species, suitable to generate an enamine nucleophile and, at the same time, coordinate with the electrophile to get a close transition state, have been used as organocatalysts ( Figure 1).

■ RESULTS AND DISCUSSION
The conjugate addition reaction between 2-phenylpropanal (10a) and diisopropyI azodicarboxylate (DIAD, 11a), in the presence of the monocarbamate derived from (1R,2R)cyclohexa-1,2-diamine 9 35 as an organocatalyst (20 mol%), was chosen as a model reaction to optimize the reaction conditions (see Table 1). Initially, we were interested in the behavior of organocatalyst 9 in conventional, solvent-present reaction conditions. Thus, the reaction of 10a (2 equiv) with 11a organocatalyzed by 9 (20 mol%) in several organic solvents at room temperature for 48 h afforded the αnitrogenated α,α-disubstituted aldehyde 12aa in up to 71% ee ( Table 1, entries 1−6). The absolute stereochemistry of 12aa was determined according to the order of elution of the corresponding enantiomers in chiral HPLC reported in the literature (see the Supporting Information).
However, when the same reaction was carried out under solvent-free conditions, the reaction was completed in 24 h, observing similar enantioselectivity for 12aa to the one obtained when the best solvent was used ( Table 1, compare entry 7 with entries 5 and 6). In addition, when the molar ratio of the reagents was modified and 1.2 equiv of 11a were used under solvent-free conditions, the conversion remained unaltered in 24 h, and the enantioselectivity for 12aa increased slightly to 75% ( Table 1, entry 8). This change in the molar ratio is important as the usual excess of the most expensive or not commercially available aldehyde is avoided. Therefore, as solvent-free conditions resulted in a superior methodology, we further optimize this environmentally friendlier procedure.
We explored if the presence of additives was beneficial for the reaction. The addition of an organic base such as 1,4diazabicyclo[2.2.2]octane (DABCO, 10 mol%) gave full conversion in only 8 h, although it was disastrous for the enantioselectivity, obtaining 12aa as a racemic mixture ( Table  1, entry 9). However, the use of acids as additives (10 mol%) also gave full conversions in 8 h while keeping the enantioselectivity ( Table 1, entries 10−14), with the best ee (83%) being achieved using the economic acetic acid as an additive (Table 1, entry 14). Lowering the loading of the organocatalyst 9 down to 10 mol% diminished the enantioselectivity slightly for 12aa (Table 1, entry 15), while keeping the loading of 9 in 20 mol% and increasing the loading of the acid additive up to 20 mol% raised the enantioselectivity of 12aa (87%) ( Table 1, entry 16). A further increase in the amount of acetic acid up to 30 mol% was detrimental to the enantioselectivity (Table 1, entry 17).
Using these conditions, we employed di-tert-butyl azodicarboxylate (11b) as a bulkier electrophilic nitrogen source, although the corresponding nitrogenated aldehyde 12ab was obtained in lower enantioselectivity (Table 1, entry 18). The use of diethyl azodicarboxylate (DEAD) was discarded because it is commercially available in solution due to its explosive potential.
Finally, we lowered the reaction temperature. Thus, when the reaction was carried out at −10°C, the ee observed for 12aa raised slightly ( This change in enantioselectivity with catalyst loading has been previously observed when using this organocatalyst in other enantioselective reactions. 25 This could suggest that when the catalyst concentration reaches a certain value, other species (even combined with the cocatalyst), such as aggregates, could also be present, modifying the enantioselectivity of the process upward. Moreover, in the absence of organocatalyst 9 and in the presence of 20 mol% of AcOH as a cocatalyst, almost no reaction was observed (Table 1, entry 23). In the absence of a background reaction, enantioselectivity should not vary with catalyst loading unless some other processes, such as described above, are occurring.
When using an α-benzylated-α-methyl aldehyde, such as cyclamen aldehyde (10o), the reaction gave the corresponding nitrogenated adduct 12oa with an enantioselectivity of 89% ( Table 2, entry 15). Interestingly, when several α,α-dialkylated aldehydes 10p−s were used as starting materials, the final adducts 12pa-sa were obtained with excellent yields and enantioselectivities ranging from 82 to 95% (Table 2, entries  16−19). These high enantioselections are remarkable as low or We also carried out the synthesis of compound 12qa under the optimal reaction conditions, but using 20 mol% of proline (1) as a commonly used organocatalyst, achieving an enantioselectivity of 59% ee. In addition, the use of 20 mol% of chiral naphthylethanamine 6 gave 12aa in 63% ee and quinine-derived amine 8 afforded 71% ee. These enantioselections are noticeably higher than those reported for similar adducts from 10q or other dialkylated aldehydes when using these catalysts under conventional solvent-including reaction conditions. 15,23,25 To evaluate the "greenness" of our protocol, we calculate the EcoScale and E-factor of the reactions, leading to adducts 12 (see Table S1 in the Supporting Information). Thus, the obtained EcoScale 36 values ranged from 63 to 72, which means, according to the established definitions, that the reaction conditions cannot be ranked as "excellent" (EcoScale > 75) but can be situated in the upper part of "acceptable" (EcoScale > 50). No "inadequate" reaction conditions were observed (EcoScale < 50). Expectedly, the EcoScale values from this solvent-free methodology are higher than the calculated for previously reported solvent-including reaction conditions (see Table S1 in the Supporting Information). In addition, the E-factor 37 for the preparation of all compounds 12 was initially calculated excluding purification materials (no workup is employed), giving values in the range 0.27−0.47 (see Table S1 in the Supporting Information). However, more realistic values (244−346) were obtained when considering the purification by column chromatography, a method present in all previously reported procedures, which also include workup.
We also proved the easy scalability of this solvent-free process by carrying out the conjugate addition of aldehyde 10a with 11a under the conditions of Table 2 but on a 3 mmol scale of 10a. This scaled-up reaction allowed us to isolate 0.96 g (95% yield) of the adduct 12aa in a 93% enantioselectivity, which are almost identical results to those obtained when working on a 0.2 mmol scale ( Table 2, entry 1).
The synthetic usefulness of these enantioenriched nitrogenated adducts 12 was exemplified by preparing the hydrazide-containing oxazolidin-2-one 13 in a 92% yield after reduction of aldehyde 12aa, obtained in the scaled-up reaction (93% ee), and further cyclization (Scheme 1). The subsequent reaction of 13 with methyl bromoacetate in the presence of cesium carbonate, and additional treatment with this base, gave the (R)-oxazolidinone 14 in a 90% yield and the expected 93% ee.
To get further insight into the behavior of organocatalyst 9 when leading to the observed sense of enantioselectivity, we performed DFT theoretical calculations (see the Supporting Information) on substrates 10a and 11a. Thus, the initial enamine formed between aldehyde 10a and catalyst 9 can adopt two different configurations, and the most stable one corresponds to the phenyl group trans to the NH moiety (A-E isomer, Figure 2). The A-Z enamine was found to be 1.4 kcal/ mol higher in energy, affording a predicted 10:1 E/Z enamine ratio. This difference is significant but not enough to discard the participation of both enamines in the reaction (see below).
From this initial point, the nucleophilic attack on the azodicarboxylate looks quite straightforward to compute. However, as previously described by our groups in related reactions, 35 the calculations showed a complex mixture of possible conformations and reacting faces of the two substrates. For example, just to indicate some of the varying factors, the enamine can adopt different conformations, that is, A (E and Z) or B (E and Z) (Figure 2). The approach between the enamine and azodicarboxylate can proceed through diastereomeric endo (C) or exo (D) transition states, and both the nucleophile and electrophile present two reacting Si and Re faces. In the case of the azodicarboxylate, these faces do not generate a new stereogenic center but still strongly affect the approach's selectivity to the enamine. Meanwhile, the two faces of the enamine lead to the R and S enantiomers of product 12aa after forming the corresponding diastereoselective transition states.
Following extensive calculations of all possibilities, structure A-E was found to be the most reactive enamine, where two main factors affect the electrophile's approach. First, the Re face (upper face in A-E representation), which leads to the minor S enantiomer of the product, is less sterically congested, as computed by the difference between TS2 and TS4 ( Figure  3). Meanwhile, an intramolecular hydrogen bond can be formed between the carbamate NH of the catalyst and one of the carbonyl groups in the azodicarboxylate. This H-bond is easier to create and stronger in the lower Si face of the enamine (TS1 vs TS3). Thus, two opposite factors are competing, with the steric one favoring the Re approach and the H-bond leading to the Si face activation. The computed energies of the most preferred structures (TS1−TS4) show the stronger stabilizing effect of the H-bond, which outcompetes the steric destabilization, making TS1-E the lowest transition state in energy and explaining the formation of the experimental R major enantiomer. Related transition states were found for the minor Z enamines, but their activation energies are at least 2.0 kcal/mol higher, indicating that they do not participate in the reaction. It is worth noting that, in the absence of the intermolecular H-bond between the electrophile and enamine, a weaker intramolecular H-bond forms between the NH of the enamine and the carbonyl group of the carbamate, both in the catalyst. Finally, the calculations were repeated with aldehyde 10p, containing methyl and ethyl substituents, and the results were almost identical to the previous ones, reinforcing the agreement between calculations and experiments (see Figure  S106 in the Supporting Information). ■ CONCLUSIONS α,α-Disubstituted aldehydes have been enantioselectively αnitrogenated employing diisopropyl azodicarboxylate as the electrophilic nitrogen source and a simple chiral mono-N-Bocprotected cyclohexa-1,2-diamine as the organocatalyst under    The stabilizing effect of the H-bond is also responsible for the enantioselective formation of the major R isomer. This procedure is efficient, easily scalable, and environmentally convenient for the enantioselective preparation of synthetically useful α,α-disubstituted α-nitrogenated aldehydes.
Materials, methods, experimental procedures, characterization data, 1 H NMR and 13 C NMR spectra, HPLC The Journal of Organic Chemistry pubs.acs.org/joc Article chromatograms, calculation of sustainability metrics, and computational data (PDF)