Tetrahydropyridines’ Stereoselective Formation, How Lockdown Assisted in the Identification of the Features of Its Mechanism

The multicomponent reaction of aldehydes, cyano-containing C-H acids, esters of 3-oxocarboxylic acid and ammonium acetate led to unexpected results. The boiling of starting materials in methanol for one to two hours resulted in the formation of polysubstituted 1,4,5,6-tetrahydropyridines with two or three stereogenic centers. During the 2020 lockdown, we obtained key intermediates of this six-step domino reaction. A number of fast and slow reactions occurred during the prolonged stirring of the reaction mass at rt. Sequence: 1. Knoevenagel condensation; 2. Michael addition; 3. Mannich reaction; 4. cyclization—fast reactions and cyclization of the product polysubstituted 2-hydroxypiperidine—was isolated after 40 min stirring at rt. Further monitoring proved the slow dehydration of 2-hydroxypiperidine to obtain 3,4,5,6-tetrahydropyridine after 7 days. Then, four-month isomerization occurred with 1,4,5,6-tetrahydropyridine formation. All reactions were stereoselective. Key intermediates and products structures were verified by X-ray diffraction analysis. Additionally, we specified conditions for the selective intermediates’ preparation.

To validate the proposed mechanism (Scheme 1), we monitored the reaction between aldehydes 1, malononitrile 2a, aryl containing esters of 3-oxocarboxylic acids 3 and ammonium acetate in methanol at room temperature (Scheme 3, Table 2). In all cases, 40-45 min stirring of the reaction mass created a dense white precipitate. After filtration and drying, single compounds (by TLC and NMR) were obtained. The 1 H and 13 C NMR spectra of compounds 6 showed one set of signals, indicating the formation of a single diastereomer. The 6d structure is shown in Figure 3. X-ray crystal diffraction data indicated that the structure 6d with four stereogenic centers should be  To validate the proposed mechanism (Scheme 1), we monitored the reaction between aldehydes 1, malononitrile 2а, aryl containing esters of 3-oxocarboxylic acids 3 and ammonium acetate in methanol at room temperature (Scheme 3, Table 2). In all cases, 40-45 min stirring of the reaction mass created a dense white precipitate. After filtration and drying, single compounds (by TLC and NMR) were obtained. The 1 Н and 13 С NMR spectra of compounds 6 showed one set of signals, indicating the formation of a single diastereomer. The 6d structure is shown in Figure 3. X-ray crystal diffraction data indicated that the structure 6d with four stereogenic centers should be defined as methyl (2SR,3RS,4SR,6RS)-5,5-dicyano-2-(4-bromo)phenyl-2-hydroxy-4,6-bis(4-methyl)phenylpiperidine-3-carboxylate.   To validate the proposed mechanism (Scheme 1), we monitored the reaction between aldehydes 1, malononitrile 2а, aryl containing esters of 3-oxocarboxylic acids 3 and ammonium acetate in methanol at room temperature (Scheme 3, Table 2). In all cases, 40-45 min stirring of the reaction mass created a dense white precipitate. After filtration and drying, single compounds (by TLC and NMR) were obtained. The 1 Н and 13 С NMR spectra of compounds 6 showed one set of signals, indicating the formation of a single diastereomer. The 6d structure is shown in Figure 3. X-ray crystal diffraction data indicated that the structure 6d with four stereogenic centers should be defined as methyl (2SR,3RS,4SR,6RS)-5,5-dicyano-2-(4-bromo)phenyl-2-hydroxy-4,6-bis(4-methyl)phenylpiperidine-3-carboxylate.   Table 2. Multicomponent synthesis of (2SR,3RS,4SR,6RS)-2-hydroxypiperidines 6. a.

Entry
Aldehyde  It should be noted that the introduction of alkyl-substituted esters of 3-oxocarboxylic acid 3 (R 1 = Alk) into this reaction did not result in the formation of 2-hydroxypiperidine 6. Apparently, the aryl substituent in position 2 is a "stabilizer" of the molecule as a whole. Thus, we found that 2-hydroxypiperidines 6 are formed as a result of a "fast" domino sequence: Knoevenagel condensation, Michael addition, Mannich reaction and intramolecular cyclization. This sequence of reactions takes only 40 min at rt. Unordinary results were found when one of the reaction mixtures was left for a long time without stirring due to isolation measures in spring 2020. The TLS monitoring of the reaction mixture containing 4-methylbenzaldehyde 1d, malononitrile 2a, methyl 3-(4-bromophenyl)-3-oxopropanoate 3f and ammonium acetate in methanol after one and a half months of standing at rt showed the presence of a new substance, different (according to TLS) from 2-hydroxypiperidine 6d and the final 1,4,5,6-tetrahydropyridine 4t. We monitored this reaction for 4.5 months. Every week, we took samples of the precipitate from the reaction mixture and analyzed it with 1 H NMR spectroscopy ( Figure 4). It should be noted that the introduction of alkyl-substituted esters of 3-oxocarboxylic acid 3 (R 1 = Alk) into this reaction did not result in the formation of 2-hydroxypiperidine 6. Apparently, the aryl substituent in position 2 is a "stabilizer" of the molecule as a whole. Thus, we found that 2-hydroxypiperidines 6 are formed as a result of a "fast" domino sequence: Knoevenagel condensation, Michael addition, Mannich reaction and intramolecular cyclization. This sequence of reactions takes only 40 min at rt. Unordinary results were found when one of the reaction mixtures was left for a long time without stirring due to isolation measures in spring 2020. The TLS monitoring of the reaction mixture containing 4-methylbenzaldehyde 1d, malononitrile 2a, methyl 3-(4-bromophenyl)-3-oxopropanoate 3f and ammonium acetate in methanol after one and a half months of standing at rt showed the presence of a new substance, different (according to TLS) from 2-hydroxypiperidine 6d and the final 1,4,5,6-tetrahydropyridine 4t. We monitored this reaction for 4.5 months. Every week, we took samples of the precipitate from the reaction mixture and analyzed it with 1 H NMR spectroscopy (Figure 4).
Crystal data and the structure refinement of 4r, 5a, 6d, 7 and 8 are shown in Table 4.

Scheme 5.
Stereoselective multicomponent synthesis of (3SR,4RS,5SR,6SR)-6-(4-bromophenyl)-3cyano-2,4-bis(4-fluorophenyl)-3,4,5,6-tetrahydropyridine-3,5-dicarboxylate 8.  Crystal data and the structure refinement of 4r, 5a, 6d, 7 and 8 are shown in Table 4.   Thus, the multicomponent reaction between aldehyde 1, cyano C-H acids 2 (malononitrile or ethyl cyanoacetate), esters of 3-oxocarboxylic acids 3 and ammonium acetate is a six-step domino process (Scheme 6). At the first stage, the Knoevenagel condensation between aldehydes and cyano C-H acid occurs. Ammonium acetate is a catalyst for this reaction. The formation of cyano olefins A under ammonium salts catalysis is already known [58]. The second step of the process is the Michael addition of C-H acid 3 to the electron-poor styrene A to form the Michael adduct B. The formation of close analogues of intermediate B from benzylidenemalononitriles and malononitrile or ethyl cyanoacetate was studied previously by Verboom et al. [48]. The subsequent Mannich reaction of B, aldehyde 1 (second equivalent) and ammonia, which is formed from ammonium acetate, leads to intermediate C. The latter undergoes intra-molecular cyclization with the formation of a substituted 2-hydroxypiperidine 6, which was identified and characterized in this work for the first time. A similar sequence of Knoevenagel condensation -Michael addition-Mannich reaction-intramolecular cyclization was described by Latypova et al. when studying the multicomponent reaction between 1,3-dicarbonyl compounds (two equiv.), formaldehyde and diamines with the formation of substituted bis-1,2,3,4-tetrahydropyridines [59]. None of the intermediates were isolated. Moreover, we tried to isolate C in the course of the work, but failed because in the reaction mass, after 10-30 min from the reaction start, there were many compounds (by TLC) that were almost impossible to isolate due to the rapid reaction rate. Polysubstituted 2-hydroxypiperidines 6 were isolated up to 87% even after stirring at rt for 40 min (see Table 2). The fifth step of the domino process is C dehydration. We established formation of 3,4,5,6-tetrahydropyridines 7, 8. A final isomerization produces 1,4,5,6-tetrahydropyridines 4, 5. Thus, the multicomponent reaction between aldehyde 1, cyano C-H acids 2 (malononitrile or ethyl cyanoacetate), esters of 3-oxocarboxylic acids 3 and ammonium acetate is a six-step domino process (Scheme 6). At the first stage, the Knoevenagel condensation between aldehydes and cyano C-H acid occurs. Ammonium acetate is a catalyst for this reaction. The formation of cyano olefins A under ammonium salts catalysis is already known [58]. The second step of the process is the Michael addition of C-H acid 3 to the electron-poor styrene A to form the Michael adduct B. The formation of close analogues of intermediate B from benzylidenemalononitriles and malononitrile or ethyl cyanoacetate was studied previously by Verboom et al. [48]. The subsequent Mannich reaction of B, aldehyde 1 (second equivalent) and ammonia, which is formed from ammonium acetate, leads to intermediate C. The latter undergoes intra-molecular cyclization with the formation of a substituted 2-hydroxypiperidine 6, which was identified and characterized in this work for the first time. A similar sequence of Knoevenagel condensation-Michael addition-Mannich reaction-intramolecular cyclization was described by Latypova et al. when studying the multicomponent reaction between 1,3-dicarbonyl compounds (two equiv.), formaldehyde and diamines with the formation of substituted bis-1,2,3,4-tetrahydropyridines [59]. None of the intermediates were isolated. Moreover, we tried to isolate C in the course of the work, but failed because in the reaction mass, after 10-30 min from the reaction start, there were many compounds (by TLC) that were almost impossible to isolate due to the rapid reaction rate. Polysubstituted 2-hydroxypiperidines 6 were isolated up to 87% even after stirring at rt for 40 min (see Table 2). The fifth step of the domino process is C dehydration. We established formation of 3,4,5,6-tetrahydropyridines 7, 8. A final isomerization produces 1,4,5,6-tetrahydropyridines 4, 5. Scheme 6. Verified mechanism of the substituted 1,4,5,6-tetrahydropyridines' formation.

General Information
All melting points were measured with a Stuart SMP30 melting point apparatus (Bibby Sterling Ltd., Granton, UK). 1 H and 13 C NMR spectra were recorded with a Bruker AM300 (Bruker, Bremen, Germany) and Bruker DRX 500 (Bruker BioSpin GmbH, Bremen, Germany) at ambient temperature in DMSO-d 6 or CDCl 3 solutions. Chemical shifts values are given in δ scale relative to Me 4 Si. The J values are given in hertz. Only discrete or characteristic signals for the 1H NMR are reported. IR spectra were recorded with a Bruker ALPHA-T FT-IR spectrometer (Bruker Corporation, Bremen, Germany) in KBr pellets. HR-ESI-MS were measured on a Bruker microTOF II instrument (Bruker Daltonik GmbH, Bremen, Germany); external or internal calibration was performed with electrospray calibrant solution (Fluka). All starting materials were obtained from commercial sources and used without purification. All reactions were monitored with thin-layer chromatography (TLC) and carried out with Merck precoated plates DC-AlufolienKieselgel60 F254 (Merck KGaA, Darmstadt, Germany). X-ray crystallographic analyses were performed with Bruker Quest D8 diffractometer (Bruker AXS GmbH, Bremen, Germany).

DFT Calculations
DFT calculations were performed with Gaussian 16 Rev C.01. B3LYP DFT (Gaussian Inc., Wallingford CT, USA, 2016) functional with GD3BJ empirical dispersion correction, and a Def2SVP basis set was used for geometry optimization and calculations of thermodynamics. Data from X-ray diffraction experiment for 7 were used as starting points for geometry optimizations. Cartesian coordinates are given in angstroms; absolute energies for all substances are given in hartrees. The analysis of vibrational frequencies was performed for all optimized structures. All compounds were characterized by only real vibrational frequencies. Wavefunction stability, using stable keyword, was also checked for each molecule. For more information see Supplementary Materials.
For the calculations of the optimized geometries, frequencies and thermodynamics with the following keywords were used: # opt freq b3lyp nosymm def2svp empiricaldispersion = gd3bj test

X-ray Crystallographic Data and Refinement Details
X-ray diffraction data for all compounds were collected at 100 K on a Bruker Quest D8 diffractometer equipped with a Photon-III area detector, using graphite-monochromatized Mo Kα-radiation (0.71073 Å) and the shutterless ϕand ω-scan technique. Relying on the analysis of preliminary collected reflections with the Cell_Now program [60], all crystals of 8 from various batches contained over seven major domains with apparently chaotic orientations. This, along with a chiral space group, seriously impeded data analysis, resulting in six attempts to collect reflection data and to solve the structure. The intensity data were integrated by the SAINT program [61] and were semi-empirically corrected for absorption and decay, using SADABS [62] for 4r, 6d, 5a and 7 or using TWINABS [61] for 8. The structures were solved by direct methods using SHELXT [63] and refined by the full-matrix least-squares method on F 2 using SHELXL-2018 [64]. The crystals of 6d and 7 were refined as inversion twins, for which the absolute structure parameter (Flack) was determined by classical fit [65]. The selected specimen of 8 was refined as a non-merohedral 2-component twin.
All non-hydrogen atoms were refined with individual anisotropic displacement parameters. The locations of atoms H1 (in 4r, 5a) and H1A, H1B (in 6d) were found from the electron density difference map; these H atoms were refined with individual isotropic displacement parameters. All other hydrogen atoms were placed in geometrically calculated positions and refined as riding atoms with relative isotropic displacement parameters. A rotating group model was applied for methyl groups. Mercury program [66] was used for molecular graphics. Crystal data, data collection and structure refinement details are summarized in Table 4.