Asymmetric Conjugate Addition of α,α-Disubstituted Aldehydes to Nitroalkenes Organocatalyzed by Chiral Monosalicylamides from trans-Cyclohexane-1,2-Diamines

Primary amine-salicylamides derived from chiral trans-cyclohexane-1,2-diamines are used as organocatalysts for the enantioselective conjugate addition of α,α-disubstituted aldehydes to arylated and heteroarylated nitroalkenes. The reaction is performed in the presence of 4-dimethylaminopyridine as an additive in dichloromethane as a solvent at room temperature. The corresponding enantioenriched γ-nitroaldehydes are obtained with enantioselectivities up to 95%. Theoretical calculations are used to justify the reasons of the stereoinduction.


Introduction
The asymmetric synthesis of γ-nitroaldehydes has gained great importance during recent years.They are precursors of γ-aminobutyric acid analogues (GABAs), which exhibit a range of pharmacological activities including antidepressant, anticonvulsant, anxiolytic and others [1,2].In addition, GABA derivatives can be potent drugs in the treatment of neurodegenerative disorders [3].Moreover, the presence of the versatile nitro group allows for further transformations to valuable compounds.This is largely due to the utility of the nitro group as a masked functionality to be transformed to a variety of other useful functional groups [4,5], which was well described by Seebach with the term 'synthetic chameleon' [6].
Molecules 2018, 23, 141 2 of 13 enantioselectivity is induced by addition of a transient enamine to the nitroolefin, which is H-bondcoordinated by the NO2 group to the NH's of the catalyst functionality.In the last years, our group has been involved in the use of small-size trans-cyclohexane-1,2diamine-derived monocarbamates, such as the mono-Boc-protected diamine 10 as organocatalysts in the enantioselective conjugate addition reactions of carbonyl compounds to electron-deficient olefins (Figure 2) [25][26][27][28].This paper shows now that a simple monoamidation of a chiral trans-cyclohexane-1,2-diamine with salicylic acid affords a primary amine-bearing salicylamide 11 (Figure 2), suitable to organocatalyze the asymmetric conjugate addition reaction of the 'difficult' α,α-disubstituted aldehydes to nitroalkenes, leading to enantioenriched γ-nitroaldehydes.Theoretical calculations can explain the observed enantioselectivity of the process.

Results and Discussion
The primary amine-salicylamide 11 employed as organocatalyst in this study was prepared by monoamidation of (1S,2S)-cyclohexane-1,2-diamine with phenyl salicylate in refluxing propan-2-ol [29].The search for the most appropriate reaction conditions (Table 1) was carried out using the model conjugate addition reaction of isobutyraldehyde (12a) to trans-β-nitrostyrene (13a).Thus, this reaction organocatalyzed by 11 (20 mol %) in toluene as a solvent at room temperature afforded the corresponding γ-nitroaldehyde (S)-14aa in 79% ee but in a very low yield after 2 d reaction time (Table 1, entry 1).The (S) absolute configuration of the final adduct was determined by comparison of the elution order of the corresponding enantiomers in chiral HPLC with those in the literature [24].The use of dimethylformamide (DMF) as solvent lowered down dramatically the stereoselectivity of the process (Table 1, entry 2), whereas the use of dichloromethane raised the enantioselectivity again up to 84%, but with almost negligible yield (Table 1, entry 3).In the last years, our group has been involved in the use of small-size trans-cyclohexane-1,2-diamine-derived monocarbamates, such as the mono-Boc-protected diamine 10 as organocatalysts in the enantioselective conjugate addition reactions of carbonyl compounds to electron-deficient olefins (Figure 2) [25][26][27][28].This paper shows now that a simple monoamidation of a chiral trans-cyclohexane-1,2-diamine with salicylic acid affords a primary amine-bearing salicylamide 11 (Figure 2), suitable to organocatalyze the asymmetric conjugate addition reaction of the 'difficult' α,α-disubstituted aldehydes to nitroalkenes, leading to enantioenriched γ-nitroaldehydes.Theoretical calculations can explain the observed enantioselectivity of the process.In the last years, our group has been involved in the use of small-size trans-cyclohexane-1,2diamine-derived monocarbamates, such as the mono-Boc-protected diamine 10 as organocatalysts in the enantioselective conjugate addition reactions of carbonyl compounds to electron-deficient olefins (Figure 2) [25][26][27][28].This paper shows now that a simple monoamidation of a chiral trans-cyclohexane-1,2-diamine with salicylic acid affords a primary amine-bearing salicylamide 11 (Figure 2), suitable to organocatalyze the asymmetric conjugate addition reaction of the 'difficult' α,α-disubstituted aldehydes to nitroalkenes, leading to enantioenriched γ-nitroaldehydes.Theoretical calculations can explain the observed enantioselectivity of the process.

Results and Discussion
The primary amine-salicylamide 11 employed as organocatalyst in this study was prepared by monoamidation of (1S,2S)-cyclohexane-1,2-diamine with phenyl salicylate in refluxing propan-2-ol [29].The search for the most appropriate reaction conditions (Table 1) was carried out using the model conjugate addition reaction of isobutyraldehyde (12a) to trans-β-nitrostyrene (13a).Thus, this reaction organocatalyzed by 11 (20 mol %) in toluene as a solvent at room temperature afforded the corresponding γ-nitroaldehyde (S)-14aa in 79% ee but in a very low yield after 2 d reaction time (Table 1, entry 1).The (S) absolute configuration of the final adduct was determined by comparison of the elution order of the corresponding enantiomers in chiral HPLC with those in the literature [24].The use of dimethylformamide (DMF) as solvent lowered down dramatically the stereoselectivity of the process (Table 1, entry 2), whereas the use of dichloromethane raised the enantioselectivity again up to 84%, but with almost negligible yield (Table 1, entry 3).

Results and Discussion
The primary amine-salicylamide 11 employed as organocatalyst in this study was prepared by monoamidation of (1S,2S)-cyclohexane-1,2-diamine with phenyl salicylate in refluxing propan-2-ol [29].The search for the most appropriate reaction conditions (Table 1) was carried out using the model conjugate addition reaction of isobutyraldehyde (12a) to trans-β-nitrostyrene (13a).Thus, this reaction organocatalyzed by 11 (20 mol %) in toluene as a solvent at room temperature afforded the corresponding γ-nitroaldehyde (S)-14aa in 79% ee but in a very low yield after 2 d reaction time (Table 1, entry 1).The (S) absolute configuration of the final adduct was determined by comparison of the elution order of the corresponding enantiomers in chiral HPLC with those in the literature [24].The use of dimethylformamide (DMF) as solvent lowered down dramatically the stereoselectivity of the process (Table 1, entry 2), whereas the use of dichloromethane raised the enantioselectivity again up to 84%, but with almost negligible yield (Table 1, entry 3).The addition of acid or basic additives frequently has proven beneficial in the organocatalyzed conjugate addition of carbonyl compounds to nitroalkenes, not only accelerating the formation of the transient enamine, but supposedly avoiding the formation of stable catalyst-derived byproducts [12].Therefore, we assayed the model conjugate addition reaction in the presence of 4dimethylaminopyridine (DMAP, 20 mol %) as a basic additive in dichloromethane as a solvent, now obtaining adduct (S)-14aa in an 81% isolated yield and with an enantioselectivity of 92% (Table 1, entry 4).However, the use of other basic additives resulted in being not so effective.Thus, the presence of imidazole or pyridine as additives gave lower enantioselections and very poor yields (Table 1, entries 5 and 6), whereas other basic species such as TMG, DBU or DABCO showed less efficiency than DMAP, considering the isolated yield of (S)-14aa (Table 1, entries 7-9).In addition, we also attempted the use of aromatic carboxylic acids as additives, but, in all cases, the achieved enantioselectivities were just moderate and yields were very low (Table 1, entries [10][11][12]. Considering the couple 11/DMAP as the most efficient catalytic combination, we explore the influence of the ratio between both components.Thus, lowering the organocatalyst loading down to 10 mol % and keeping the additive loading to 20 mol % gave rise to (S)-14aa in only 17% yield but in 94% ee (Table 1, entry 13), whereas keeping the catalyst loading in 20 mol % and diminishing the amount of DMAP to 10 mol % maintained the ee unaltered, increasing the final yield (Table 1, entry 14).However, loadings of 20 mol % of 11 and 30 mol % of DMAP allowed to obtain (S)-14aa in 72% yield and 95% ee (Table 1, entry 15).
Expecting to achieve an opposite enantioselection, we also performed the reaction using as organocatalyst ent-11, which was prepared similarly to its enantiomeric counterpart, but using (1R,2R)-cyclohexane-1,2-diamine as chirality source (Figure 3).Using this primary amine as organocatalyst (20 mol %) under the most effective reaction conditions (DMAP (30 mol %), CH2Cl2, rt), adduct (R)-14aa was isolated in the same 95% enantioselectivity than its opposite (S)-enantiomer (Table 1 The addition of acid or basic additives frequently has proven beneficial in the organocatalyzed conjugate addition of carbonyl compounds to nitroalkenes, not only accelerating the formation of the transient enamine, but supposedly avoiding the formation of stable catalyst-derived byproducts [12].Therefore, we assayed the model conjugate addition reaction in the presence of 4-dimethylaminopyridine (DMAP, 20 mol %) as a basic additive in dichloromethane as a solvent, now obtaining adduct (S)-14aa in an 81% isolated yield and with an enantioselectivity of 92% (Table 1, entry 4).However, the use of other basic additives resulted in being not so effective.Thus, the presence of imidazole or pyridine as additives gave lower enantioselections and very poor yields (Table 1, entries 5 and 6), whereas other basic species such as TMG, DBU or DABCO showed less efficiency than DMAP, considering the isolated yield of (S)-14aa (Table 1, entries 7-9).In addition, we also attempted the use of aromatic carboxylic acids as additives, but, in all cases, the achieved enantioselectivities were just moderate and yields were very low (Table 1, entries [10][11][12]. Considering the couple 11/DMAP as the most efficient catalytic combination, we explore the influence of the ratio between both components.Thus, lowering the organocatalyst loading down to 10 mol % and keeping the additive loading to 20 mol % gave rise to (S)-14aa in only 17% yield but in 94% ee (Table 1, entry 13), whereas keeping the catalyst loading in 20 mol % and diminishing the amount of DMAP to 10 mol % maintained the ee unaltered, increasing the final yield (Table 1, entry 14).However, loadings of 20 mol % of 11 and 30 mol % of DMAP allowed to obtain (S)-14aa in 72% yield and 95% ee (Table 1, entry 15).
We were intrigued to determine if the presence of the phenolic OH on the organocatalyst was a determinant for the high enantioselectivity obtained.Thus, we prepared the primary amine-containing benzamide 15 by reaction of (1S,2S)-cyclohexane-1,2-diamine with phenyl benzoate under similar conditions as 11 (Figure 3).However, under the above optimal reaction conditions, this organocatalyst 15 gave rise to adduct (S)-14aa in only a 79% ee (Table 1, entry 17).Therefore, the presence of the phenolic OH in organocatalyst 10 results as being important for achieving a good enantioinduction.It is interesting to note that the use of the monocarbamate 10 as an organocatalyst was not particularly successful, giving rise to the corresponding adduct in low yield and only 65% ee (Table 1, entry 18), whereas it has been able to reach up to 96% ee when employed in a related enantioselective Michael addition of aromatic ketones to nitroalkenes [27].
Molecules 2018, 23, 141 4 of 13 We were intrigued to determine if the presence of the phenolic OH on the organocatalyst was a determinant for the high enantioselectivity obtained.Thus, we prepared the primary aminecontaining benzamide 15 by reaction of (1S,2S)-cyclohexane-1,2-diamine with phenyl benzoate under similar conditions as 11 (Figure 3).However, under the above optimal reaction conditions, this organocatalyst 15 gave rise to adduct (S)-14aa in only a 79% ee (Table 1, entry 17).Therefore, the presence of the phenolic OH in organocatalyst 10 results as being important for achieving a good enantioinduction.It is interesting to note that the use of the monocarbamate 10 as an organocatalyst was not particularly successful, giving rise to the corresponding adduct in low yield and only 65% ee (Table 1, entry 18), whereas it has been able to reach up to 96% ee when employed in a related enantioselective Michael addition of aromatic ketones to nitroalkenes [27].Next, we extended the addition reaction of isobutyraldehyde to other trans-β-nitroalkenes 13 under the most favorable reaction conditions (11 (20 mol %), DMAP (30 mol %), CH2Cl2, rt), the results being summarized in Table 2.The absolute configuration of the known γ-nitroaldehydes 14 was assigned in accordance with the elution order of the enantiomers in chiral HPLC when compared to the literature (see Experimental Section).
When nitroalkene 13k bearing a 2-naphthyl group was employed as Michael acceptor, the corresponding adduct (S)-14ak was obtained in 91% ee (Table 2, entry 11).In addition, the influence of the presence of heteroarylated rings in the nitroalkene was also explored with the use as Michael acceptors of the 3-pyridinyl-and 2-furanyl-containing nitroalkenes 13l and 13m, which gave rise to adducts (S)-14al and (S)-14am in 91 and 92% ee, respectively (Table 2, entries 12 and 13).Next, we extended the addition reaction of isobutyraldehyde to other trans-β-nitroalkenes 13 under the most favorable reaction conditions (11 (20 mol %), DMAP (30 mol %), CH 2 Cl 2 , rt), the results being summarized in Table 2.The absolute configuration of the known γ-nitroaldehydes 14 was assigned in accordance with the elution order of the enantiomers in chiral HPLC when compared to the literature (see Experimental Section).
When nitroalkene 13k bearing a 2-naphthyl group was employed as Michael acceptor, the corresponding adduct (S)-14ak was obtained in 91% ee (Table 2, entry 11).In addition, the influence of the presence of heteroarylated rings in the nitroalkene was also explored with the use as Michael acceptors of the 3-pyridinyl-and 2-furanyl-containing nitroalkenes 13l and 13m, which gave rise to adducts (S)-14al and (S)-14am in 91 and 92% ee, respectively (Table 2, entries 12 and 13).When nitroalkene 13k bearing a 2-naphthyl group was employed as Michael acceptor, the corresponding adduct (S)-14ak was obtained in 91% ee (Table 2, entry 11).In addition, the influence of the presence of heteroarylated rings in the nitroalkene was also explored with the use as Michael acceptors of the 3-pyridinyl-and 2-furanyl-containing nitroalkenes 13l and 13m, which gave rise to adducts (S)-14al and (S)-14am in 91 and 92% ee, respectively (Table 2, entries 12 and 13).We also explored the conjugate addition reaction of other α,α-disubstituted aldehydes with nitroalkene 13a.Thus, when cyclopentanecarbaldehyde (12b) was used, the corresponding Michael adduct (S)-14ba was isolated in an excellent 94% ee (Table 2, entry 14).However, when diphenylacetaldehyde or 2-phenylpropionaldehyde were used as pro-nucleophiles, almost negligible amounts (5%) of the corresponding adducts were detected as racemates.
To get further insight into the origin of the enantioselectivity, we carried out theoretical calculations on the reaction between isobutyraldehyde 12a and nitroalkene 13a in the presence of catalyst 11.A salient feature of this catalyst is the presence of a phenolic OH in the ortho position, which seems to play a key role to enhance the activity and/or selectivity of the process.To understand this intriguing behavior, the results obtained with 11 were compared with those of the less active catalyst 15, and also with the model system 16 (Figure 4), which bears the OH group para to the carbonyl substituent.This species was not checked experimentally but would give us a better understanding of the relevance of the ortho substitution for the reactivity.We also explored the conjugate addition reaction of other α,α-disubstituted aldehydes with nitroalkene 13a.Thus, when cyclopentanecarbaldehyde (12b) was used, the corresponding Michael adduct (S)-14ba was isolated in an excellent 94% ee (Table 2, entry 14).However, when diphenylacetaldehyde or 2-phenylpropionaldehyde were used as pro-nucleophiles, almost negligible amounts (<5%) of the corresponding adducts were detected as racemates.
To get further insight into the origin of the enantioselectivity, we carried out theoretical calculations on the reaction between isobutyraldehyde 12a and nitroalkene 13a in the presence of catalyst 11.A salient feature of this catalyst is the presence of a phenolic OH in the ortho position, which seems to play a key role to enhance the activity and/or selectivity of the process.To understand this intriguing behavior, the results obtained with 11 were compared with those of the less active catalyst 15, and also with the model system 16 (Figure 4), which bears the OH group para to the carbonyl substituent.This species was not checked experimentally but would give us a better understanding of the relevance of the ortho substitution for the reactivity.As expected, the initial formation of an enamine between the catalyst free amine and the aldehyde is followed by attack to the nitrostyrene according to Seebach's synclinal model (endo attack) [30,31], which was confirmed by the much higher energy of other possible approaches, like exo depicted in Figure 5.The synclinal model secures a diastereoselective approach of the reacting faces of enamine and alkene, meaning that the lower face (in our view) of the enamine reacts with the Re face of alkene and the opposite, the upper face of enamine with the Si face of alkene.As a direct consequence, the approach of the alkene from each side of the enamine produces a single diastereoisomer.As expected, the initial formation of an enamine between the catalyst free amine and the aldehyde is followed by attack to the nitrostyrene according to Seebach's synclinal model (endo attack) [30,31], which was confirmed by the much higher energy of other possible approaches, like exo depicted in Figure 5.The synclinal model secures a diastereoselective approach of the reacting faces of enamine and alkene, meaning that the lower face (in our view) of the enamine reacts with the Re face of alkene and the opposite, the upper face of enamine with the Si face of alkene.As a direct consequence, the approach of the alkene from each side of the enamine produces a single diastereoisomer.
attack) [30,31], which was confirmed by the much higher energy of other possible approaches, like exo depicted in Figure 5.The synclinal model secures a diastereoselective approach of the reacting faces of enamine and alkene, meaning that the lower face (in view) of the enamine reacts with the Re face of alkene and the opposite, the upper face of enamine with the Si face of alkene.As a direct consequence, the approach of the alkene from each side of the enamine produces a single diastereoisomer.The different conformations of the both cyclohexyl-amine substituents in the catalysts were taken into consideration.The amine side arms of cyclohexane possess several free rotating bonds, and, among all the possibilities checked, we selected the most stable ones, outlined in Figure 6.In them, the enamine and amide groups occupy the equatorial positions of the cyclohexane, pointing up and down, respectively.The fragment NH-C-C-NH is in staggered conformation.We then located the transition states for 11, corresponding to the different approaches, taking into account all possible H-bonding activation networks of the NH and OH groups (Figure 6).As expected, the amide NH group in 11 is H-binding the nitro group of the electrophile, activating it for the nucleophilic attack, and inducing a good differentiation of both faces of the enamine.The lower face approach in TS1 is much lower in energy (9.8 kcal/mol) than TS2 (15.8 kcal/mol), justifying the experimental selectivity.The origin of the selectivity is clearly linked to the high strain developing in TS2 to accommodate the H-bond between the nitro and amide groups.We could demonstrate that OH is not actively participating in the activation of the nitro group, since the barrier in TS3 is clearly higher (18.4 kcal/mol) than in TS1, even if the amide-NH in TS3 is slightly contributing to increase the acidity of the OH through Brønsted assistance (δ O-H = 1.98 Å).Interestingly, this H-bonding network alternative is more flexible, and the energy difference between TS3 and TS4 is fairly reduced to only 2.0 kcal/mol.The different conformations of the both cyclohexyl-amine substituents in the catalysts were taken into consideration.The amine side arms of cyclohexane possess several free rotating bonds, and, among all the possibilities checked, we selected the most stable ones, outlined in Figure 6.In them, the enamine and amide groups occupy the equatorial positions of the cyclohexane, pointing up and down, respectively.The fragment NH-C-C-NH is in staggered conformation.We then located the transition states for 11, corresponding to the different approaches, taking into account all possible H-bonding activation networks of the NH and OH groups (Figure 6) [32].As expected, the amide NH group in 11 is H-binding the nitro group of the electrophile, activating it for the nucleophilic attack, and inducing a good differentiation of both faces of the enamine.The lower face approach in TS1 is much lower in energy (9.8 kcal/mol) than TS2 (15.8 kcal/mol), justifying the experimental selectivity.The origin of the selectivity is clearly linked to the high strain developing in TS2 to accommodate the H-bond between the nitro and amide groups.We could demonstrate that OH is not actively participating in the activation of the nitro group, since the barrier in TS3 is clearly higher (18.4 kcal/mol) than in TS1, even if the amide-NH in TS3 is slightly contributing to increase the acidity of the OH through Brønsted assistance (δO-H = 1.98 Å).Interestingly, this H-bonding network alternative is more flexible, and the energy difference between TS3 and TS4 is fairly reduced to only 2.0 kcal/mol.Initially, the tentative role of OH in promoting the reaction was assigned to its suitability to enhance the H-bond donor ability of the amide group through an extra activation by Brønsted acidassistance (δO-H = 1.62 Å in TS1).However, to our surprise, the calculations predicted a similar reactivity and selectivity for species 15 and 16 (Figure 4), where the OH group is lacking or cannot interact with the amide because it is positioned at the para position.In fact, the activation energies of transition states TS5 and TS7 are undistinguishable from TS1 (Figure 7), and a similar situation is found comparing TS2, TS6 and TS8.Thus, the presence or absence of the ortho OH group in 11 is not predicted computationally to have a significant effect on both reactivity and selectivity.
In this regard, there is an increasing number of studies indicating that C-C bond formation between enamine and nitroalkene is facile and might not be the rate-determining step in catalyzed Initially, the tentative role of OH in promoting the reaction was assigned to its suitability to enhance the H-bond donor ability of the amide group through an extra activation by Brønsted acid-assistance (δ O-H = 1.62 Å in TS1).However, to our surprise, the calculations predicted a similar reactivity and selectivity for species 15 and 16 (Figure 4), where the OH group is lacking or cannot interact with the amide because it is positioned at the para position.In fact, the activation energies of transition states TS5 and TS7 are undistinguishable from TS1 (Figure 7), and a similar situation is found comparing TS2, TS6 and TS8.Thus, the presence or absence of the ortho OH group in 11 is not predicted computationally to have a significant effect on both reactivity and selectivity.
In this regard, there is an increasing number of studies indicating that C-C bond formation between enamine and nitroalkene is facile and might not be the rate-determining step in catalyzed processes [32][33][34].Instead, the highest energy along the reaction coordinate would correspond to the protonation of the highly stable cyclic intermediates that arise after the C-C bond formation.Thus, the rate-and stereodetermining steps might not be identical.We checked this alternative in our case, and could locate two cyclic intermediates, 17 and 18 (Figure 8), showing energies of −8.6 and −6.1 kcal/mol with respect to the starting materials.Furthermore, the lowest transition state for their protonation (TS9) would be the rate limiting step of our cycle, presenting a Free energy barrier of 28.0 (from 17).A similar protonation cannot obviously be envisioned for catalytic species lacking ortho-phenol, like 15 and 16, and, in those cases, adventitious water must be responsible for the cleavage and protonation of the proposed intermediates, although at a higher energetic cost.As a confirmation, we were able to locate some transition states involving a molecule of water, which are in every case at least 1.5-2.0kcal/mol higher in energy than TS9.Thus, we can confirm that also in our case the stereoselectivity is controlled during the C-C bond formation event, but the reaction rate is governed by the protonation and cleavage of the stable cyclic intermediates.It is at this point where the phenol moiety plays a crucial role, nicely explaining the experimental results.A similar protonation cannot obviously be envisioned for catalytic species lacking ortho-phenol, like 15 and 16, and, in those cases, adventitious water must be responsible for the cleavage and protonation of the proposed intermediates, although at a higher energetic cost.As a confirmation, we were able to locate some transition states involving a molecule of water, which are in every case at least 1.5-2.0kcal/mol higher in energy than TS9.Thus, we can confirm that also in our case the stereoselectivity is controlled during the C-C bond formation event, but the reaction rate is governed by the protonation and cleavage of the stable cyclic intermediates.It is at this point where the phenol moiety plays a crucial role, nicely explaining the experimental results.A similar protonation cannot obviously be envisioned for catalytic species lacking ortho-phenol, like 15 and 16, and, in those cases, adventitious water must be responsible for the cleavage and protonation of the proposed intermediates, although at a higher energetic cost.As a confirmation, we were able to locate some transition states involving a molecule of water, which are in every case at least 1.5-2.0kcal/mol higher in energy than TS9.Thus, we can confirm that also in our case the stereoselectivity is controlled during the C-C bond formation event, but the reaction rate is governed by the protonation and cleavage of the stable cyclic intermediates.It is at this point where the phenol moiety plays a crucial role, nicely explaining the experimental results.

General Information
All the reagents and solvents employed were of the best grade available and were used without further purification.The 1 H-and 13 C-NMR spectra were recorded at room temperature on a Bruker (Bruker, Billerica, MA, USA) AC-400 at 400 MHz and 101 MHz, respectively, using TMS as internal standard.IR spectra were measured on a Nicolet Impact 400D-FT instrument (Thermo Fisher Scientific, Waltham, MA, USA).Electron Ionized Mass Spectrometry (EIMS) spectra were obtained on an Agilent Technologies GC/MS-5973N equipment (Agilent Technologies, Santa Clara, CA, USA) at 70 eV.HR-MS spectra were obtained on an Agilent Technologies 7200 Accurate-Mass Q-TOF GC/MS equipment at using EI at 70 eV.Compounds 11 and ent-11 were obtained as described [29].Nitroalkenes 13 were purchased or prepared following a reported procedure [35], except 13l, which was obtained differently [36].Absolute configuration for adducts 14 was determined according to the order of elution of their enantiomers in chiral HPLC.The absolute configuration of the not described compounds 14ad and 14ae was assigned by analogy.Reference racemic samples of adducts 14 were obtained by performing the conjugate addition reaction using 4-methylbenzylamine (20 mol %) as organocatalyst in toluene as a solvent at room temperature.

General Procedure for the Asymmetric Conjugate Addition Reaction
To a solution of 10, 11, ent-11 or 15 (0.04 mmol), the nitroalkene 13 (0.2 mmol) and DMAP (7.3 mg, 0.06 mmol) in CH 2 Cl 2 (0.3 mL) was added the aldehyde 12 (0.4 mmol) and the mixture was stirred at rt until completion (TLC).The reaction was quenched with HCl 2N (10 mL) and the mixture was extracted with AcOEt (3 × 10 mL).The organic phase was washed with sat.NaHCO 3 (2 × 10 mL), dried over MgSO 4 , and the solvent was evaporated (15 Torr) to get the crude product, which was purified by silica gel chromatography (n-hexane/AcOEt gradients).Known adducts 14 were identified by comparison of their NMR data with those of the literature (Supplementary Materials NMR spectra).Their enantiomeric excesses were determined by chiral HPLC using the conditions described in each case (Supplementary Materials HPLC chromatograms).Not described compounds 14ad and 14ae have been fully characterized.

Molecules 2018 ,
23, 141 2 of 13 enantioselectivity is induced by addition of a transient enamine to the nitroolefin, which is H-bondcoordinated by the NO2 group to the NH's of the catalyst functionality.

Figure 4 .
Figure 4. Model catalysts employed in the computational study.

Figure 4 .
Figure 4. Model catalysts employed in the computational study.

Figure 5 .
Figure 5. Seebach's synclinal model for the approach of enamine and nitroalkene faces.Figure 5. Seebach's synclinal model for the approach of enamine and nitroalkene faces.

Figure 5 .
Figure 5. Seebach's synclinal model for the approach of enamine and nitroalkene faces.Figure 5. Seebach's synclinal model for the approach of enamine and nitroalkene faces.

Figure 6 .
Figure 6.Computed Free energies and structures of the diastereoselective transition states for the C-C bond formation in the presence of catalyst 11.

Figure 6 .
Figure 6.Computed Free energies and structures of the diastereoselective transition states for the C-C bond formation in the presence of catalyst 11.

Figure 7 .
Figure 7. Computed Free energies and structures of the diastereoselective transition states for the C-C bond formation in the presence of the catalysts 15 and 16.

Figure 8 .
Figure 8. Computed Free energies and structures for the formation of the cyclic intermediates and their protonation to the final products.

Figure 7 . 13 Figure 7 .
Figure 7. Computed Free energies and structures of the diastereoselective transition states for the C-C bond formation in the presence of the catalysts 15 and 16.

Figure 8 .
Figure 8. Computed Free energies and structures for the formation of the cyclic intermediates and their protonation to the final products.

Figure 8 .
Figure 8. Computed Free energies and structures for the formation of the cyclic intermediates and their protonation to the final products.

Table 1 .
Screening and optimization of the reaction conditions for the model enantioselective conjugate addition.

Table 1 .
Screening and optimization of the reaction conditions for the model enantioselective conjugate addition.
a Isolated yield after flash chromatography.b Enantioselectivities determined by chiral HPLC.Absolute configuration assigned by the order of elution of the enantiomers in chiral HPLC (See Experimental Section).c Apparent change in the sense of the enantioselectivity because of the application of the Cahn-Ingold-Prelog rules.