Simple Synthesis of Complex Amines from the Diels–Alder Adducts of (–)-Cytisine

Abstract

The Diels–Alder reaction of N-benzylcytisine with N-methyl- and N-benzylmaleimides is 100% endo-selective and gives the corresponding syn- and anti-diastereomers in 11–42% isolated yields. The studies of the reaction progress with LCMS and NMR along with detailed quantum chemical calculations revealed that some Diels–Alder adducts are kinetically and their isomers are thermodynamically controlled products. The Pd/C-catalyzed hydrogenation of benzyl-protected cytisine amine derivatives resulted in the removal of the benzyl group and the addition of hydrogen to the C=C double bond to give the corresponding secondary amines in 45–84% yield. The complete reduction of carbonyl groups in a cytisine derivative with LiAlH₄ in THF under reflux afforded the respective tricyclic triamine. Quantum mechanical calculations for the mechanism of the Diels–Alder reaction between the simplest model compounds are presented.

Naturally occurring alkaloid (–)-cytisine (1) is an efficient acetylcholine agonist possessing a strong affinity to nicotinic acetylcholine receptors.[1] Due to its biological activity (–)-cytisine has been used as an active ingredient of antismoking drugs.[2] Various chemical modifications of (–)-cytisine were applied to obtain novel bicyclic molecular modules for rational drug discovery.[3] For example, N-alkylations,[4] reductive aminations,[5] nitrations, halogenations,[6] alkylations, and arylations[8] afforded various biologically active (–)-cytisine analogues.[8] In search of novel antiviral drugs, the Diels–Alder reaction of N-alkyl and N-acyl derivatives of cytisine 2a–e with a series of maleimides 3a–g afforded isomers 4a–h and 5a–h (Scheme [1]), which were separated and completely characterized.[9] Compounds 4 and 5, however, have not been used for the synthesis of molecular scaffolds suitable for simple systematic variations of functional groups.

In the present paper, we report the synthesis, crystallographic, computational, and spectroscopic studies of conformationally restricted tricyclic di- and triamines based on the Diels–Alder adducts of N-benzyl-(–)-cytisine with N-methyl- and N-benzylmaleimides.

We considered N-benzylcytisine (2c) and maleimides 3a and 3e as promising starting materials for the synthesis of the Diels–Alder adducts 4i,j and 5i,j (Scheme [2]), which could be further converted into the corresponding secondary amines via Pd/C-catalyzed hydrogenation. The following LiAlH₄ reduction of the carbonyl groups in the amines was considered as a feasible route to the secondary diamines. As shown in Scheme [2], the Diels–Alder reaction of N-benzylcytisine (2c) with 3a or 3e in refluxing toluene gave mixtures of syn-4i, syn-4j, and anti-5i, anti-5j diastereomers, which were separated by preparative column chromatography (silica gel, acetone/hexane).

The LCMS analysis of the reaction mixtures of the preparative procedures revealed [5i]/[4i] = 2 (40 h reaction, 78% conversion of 2c) and [5j]/[4j] = 5 (16 h reaction, 82% conversion of 2c). The chromatographic separation on silica gel afforded individual isomers 5i (32%), 4i (31%), 5j (42%), and 4j (11%) due to the partial overlap of the peaks to be separated.

In an additional experiment, the progress of the Diels–Alder reaction between 2c and 3a was carefully monitored via LCMS and NMR analysis of the aliquots taken from the reaction mixtures. As shown in Table [1], the amount of anti-isomer 5i increases faster at the beginning of the reaction however, after ca. 24 hours, syn-adduct 4i becomes the major product. This observation indicates that isomers 5i and 4i are products of kinetic and thermodynamic control, respectively.

At 50 °C the reaction between 2c and 3a occurred slowly with 5% conversion of the starting materials and [5i]/[4i] = 10 after 6 hours of heating. The stirring of the solution of 5i in toluene at 110 °C for 6 hours resulted in the retro Diels–Alder reaction with 20% conversion. In this case, only 2c and 3a formed and no trace of isomer 4i was detected.

Due to the sheer complexity of spatial structures of molecules 4i, 4j and 5i, 5j single-crystal X-ray analysis was used for the unambiguous determination of their 3D structures. To obtain appropriate single crystals multiple crystallization attempts have been undertaken from various solvents and solvent mixtures. Compounds 4i, 4j and 5i, 5j are off-white solids readily soluble in CDCl₃, CH₂Cl₂, and benzene but barely soluble in MeCN and DMSO. After many trial­-and-error attempts, it was found that slow evaporation of solutions of 5i in DMSO/CH₂Cl₂, 4i in MeCN, 5j in CH₂Cl₂, and 4j in CH₂Cl₂/MeCN afforded diffraction-quality crystals, which were stable without mother liquors. Single-crystal X-ray diffraction studies revealed that in the crystalline state molecules 4i, 4j have syn (Figure [1a, b]) and 5i, 5j anti (Figure [1c, d]) orientation of the imide residue with respect to the methylene bridge of the 3,7-diazabicyclo[3.3.1]nonane fragment. In all the structures the piperidine ring adopts the most favorable chair conformation.

The ¹H NMR spectra of compounds 4i, 4j and 5i, 5j contain well-resolved signals for two protons of the vinylidene bridges. The integration of the ¹H NMR spectra revealed the numbers of protons in 4i, 4j and 5i, 5j, which are completely consistent with their molecular structures in the crystalline state. The numbers of signals in the ¹³C NMR spectra of 4i, 4j and 5i, 5j correspond to their structures, however, no assignments of those signals have been carried out. Since the structures of compounds 4i, 4j and 5i, 5j had been unambiguously confirmed by single-crystal X-ray analysis, their 2D NMR stereochemical studies in solutions were not carried out. More detailed spectral investigations of compounds 4i, 4j and 5i, 5j and their analogues will be published in due course.

The reaction of compounds 4i, 5i, and 5j with H₂ (10 bar, Pd/C, MeOH) resulted in the removal of benzyl groups from the amine fragment and hydrogenation of the double C=C bond to give amines 6a–c (Scheme [3]). The benzyl group attached to the imide fragment of 5j remained intact under these reaction conditions. Crystallization of crude compounds 6a (MeCN) and 6b (i-PrOH) afforded their analytically pure samples in 55 and 45% yield, respectively. Compound 6c·HBr was obtained in 84% yield via the addition of one equivalent of aqueous HBr to a solution of the crude amine in MeCN.

Single-crystal X-ray analysis unambiguously determined 3D structures of 6a and 6b (Figure [2] a,b) which are similar to the structures of the corresponding starting materials 4i and 5i (Figure [1a,c]). The ¹H and ¹³C NMR spectra, mass spectrometry, and elemental analysis data are consistent with the single-crystal structures of 6a and 6b.

The reduction of carbonyl groups of 6c by LiAlH₄ in THF (reflux 14 h) afforded crude diamine 7 containing about 10% of unidentified impurity. The addition of 2 equivalents of aqueous HBr to a solution of the crude 7 in MeCN afforded a fine precipitate of pure 7·2HBr and well-shaped diffraction-quality crystals, which were mechanically separated. Single-crystal X-ray analysis revealed that the crystals contained 80% of 7·3HBr·MeCN and 20% of 8·3HBr·MeCN. Compound 8 seems to be a product of rearrangement of the hypothetical incomplete reduction intermediate 9. The molecular structure of 7 (Figure [2c]) is similar to the scaffold structures of 6b, 5i, and 5j and confirms the 3D structure of its starting material 6c, which unfortunately was not determined by direct single-crystal X-ray analysis. The tertiary ammonium groups N1⁺H and N2⁺H of 7 form single hydrogen bonds to bromide anions Br2^– and Br3a^–. The secondary ammonium group N3⁺H is hydrogen-bonded simultaneously to Br1a^– and Br3a^– and bromide anion Br3a^– forms two hydrogen bonds to N⁺3H and N2⁺H. The positions of C, H, N, and Br atoms in 8 are the same as in 7. The OH group of 8 forms hydrogen bonds to Br2 as shown in Figure [2d]. The integration of ¹H NMR signals of the co-crystal revealed the presence of 7 and 8 in a 4:1 ratio. These components could not be separated and compound 8 was not isolated as an individual compound. The LCMS study showed that 7 and 8 have the same retention time but the MS spectrum of the single peak contained the masses of both components.

We have undertaken an in-depth quantum mechanical (QM) investigation of the mechanism for the Diels–Alder reaction between the simplest model reagents 2a and 3a (Scheme [1]), which leads to isomers 4k and 5k (R¹ = X = Me).

Starting material 2a, products 4k and 5k, and the corresponding transition states have two conformational degrees of freedom: the interconversion between the chair and boat conformers of the piperidine fragment and the flipping of the NMe group between the axial and equatorial positions. This results in four possible conformers, which are shown in Figure [3] for compound 2a. Conformations 2a(A,B), 4k(A,B), 5k(A,B) possess the piperidine ring in its chair conformation, and the NMe group occupying equatorial (A) and axial position (B). Likewise in 2a(C,D), 4k(C,D), and 5k(C,D) the piperidine ring adopts the boat conformation (Figure [3]) and the NMe groups are either equatorial (C) or axial (D). The transition states for the formation of 4k(A–D) and 5k(A–D) from the corresponding isomers of 2a and 3a will be denoted as TS 4(X) and TS 5(X) where X is A–D.

According to gas-phase QM computations (Table [2]), the most stable conformer is 2a(A), which agrees with a solid-state structure determined by single-crystal X-ray diffraction.[10] The conformer 2a(B) has 4.9–5.4 kcal/mol higher energy and electronic activation energy for 2a(A) → 2a(B) is estimated at 8.7–8.8 kcal/mol. The energy of the third possible conformer 2a(C) is 5.8–6.2 kcal/mol higher compared to 2a(A) and the activation energy for 2a(A) → 2a(C) (13.3–14.0 kcal/mol) is nearly two times higher than for 2a(A) → 2a(B). The least stable conformer 2a(D) has its energy 10.3–11.0 kcal/mol higher than 2a(A) and the activation barrier for 2a(A) → 2a(D) is 14.0–14.7 kcal/mol.

The reaction of 2a(A) with 3a gives 4k(A) and 5k(A) with the electronic reaction energies between –28 to –25 kcal/mol depending on the applied method. Notably, the conventional MP2 method tends to overestimate these values and to lower the respective energetic barriers on account of exaggerated long-range correlation effects. Nevertheless, after appropriate scaling of spin components, the RI-SCS-MP2 [and next DLPNO-CCSD(T)] reaction energies lie very close to each other (typically, within 1 kcal/mol). An inclusion of zero-point energies to the electronic reaction energies increases the latter by ca. 4 kcal/mol. We note that the RI-SCS-MP2 reaction energies calculated with the rather small 6-31G* basis set do not significantly vary (i.e., by 0.4 kcal/mol or better) from their aug-cc-pVTZ counterparts indicating the presence of almost identical equilibrium geometries. This observation is also confirmed through the respective single-point DLPNO-CCSD(T)/aug-cc-pVTZ calculations utilizing both RI-SCS-MP2 equilibrium geometries with these two basis sets. All calculated reaction energies indicate the exothermic character of the reaction (see Table [2]) with the small thermodynamic preference of 4k(A) over 5k(A) (ca. 2 kcal/mol).[11] Hence, these calculations predict that under thermodynamic control conditions compounds 4i,j have to be the major product. As shown in Table [2], RI-SCS-MP2/6-31G*(aug-cc-pVTZ) calculations, as well as DLPNO-CCSD(T) ones, predict very close activation energies for the formation of 4k(A) (8.0–11.7 kcal/mol) and 5k(A) (7.5–11.9 kcal/mol).

Consequently, no preferential formation of any product is predicted in this case based on kinetic reasons. Theoretical modeling of the Diels–Alder reaction between conformer 2a(B) and 3a revealed almost isoenergetic products 4k(B) and 5k(B) with the reaction energies of ca. –20 kcal/mol. However, TS 4k(B) could not be located since the axial NMe group of 2a(B) sterically block the reaction center precluding the 3a molecule to come closer and form the transition state. Hence, the direct formation of product 4k(B) from 2a(B) and 3a is predicted to be highly improbable. In contrast, the transition state of the reaction 2a(B) + 3a → 5k(B) was located and the corresponding activation energy was found to be about 15 kcal/mol [DLPNO-CCSD(T)]. The reaction 2a(C) + 3a is also predicted to be exothermic for both products, however in this case 5k(C) has by 2.3–2.5 kcal/mol lower energy than 4k(C). The transition state structures TS 4k(C) and TS 5k(C) differ considerably and the difference in activation barriers for the formation of 5k(B or C) from 3a and 2a(C) is 5.5–6.2 kcal/mol.

With the most accurate DLPNO-CCSD(T) method, the electronic activation energy for the formation of 5k(C) is 15.1 kcal/mol and 20.6 kcal/mol for 4k(C). Such a striking effect prompted us to analyze the energetic and geometric characteristics of these transition states more in detail. Table [3] reports the deformation energies E _def on account of a geometry change associated with the reorganization of the reactants into the respective TS structures. It can be seen that 3a does not exhibit any significant difference in E _def between in the transition states leading to 4k(C) and 5k(C) whereas 2a(C) requires 4.2 kcal/mol more energy to form TS 4k(C) (24.5 kcal/mol) than TS 5k(C) (20.3 kcal/mol). As the electronic activation energy difference for these cases at the RI-SCS-MP2/6-31G* level of theory is 6 kcal/mol, the remaining 1.8 kcal/mol can be attributed to the interaction energy E _int difference in TS, disfavoring the formation of 4k(C). Thus, both bonded and nonbonded contributions to the TS discriminate against the formation of TS 4k(C).

As collected in Table [4], TS 4k(C) is more asynchronous with its incipient bond length difference of 0.208 or 0.181 Å at the RI-SCS-MP2/6-31G* and RI-SCS-MP2/aug-cc-pVTZ level, respectively, whereas for TS 5k(C) these values are 0.129 and 0.127 Å with the same methodologies applied. This asynchronicity of TS 4k(C) reflects, in turn, the steric hindrance since 3a cannot come close to 2a because of the repulsion to the hydrogen atom of the CH₂ group in 2a. Both products show, however, negligible asynchronicities which are also conspicuously reflected in a small reaction energy difference (cf. Table [2]).

For the least stable conformation 2a(D), the Diels–Alder reaction with 3a is associated with ca. 4.5–5.0 kcal/mol higher values of reaction and activation energies than those for 2a(C) (see Table [2] and Figure [1]) and are less likely to contribute significantly to the whole reaction mechanism. However, even in this case, the calculations predict the preferable (faster) formation of 5(D) over 4(D).

The endothermic character of the conformational transitions from 2a(A) to 2a(B–D), predicted at 0 K by QM methods, suggests that at the experimental temperature of the Diels–Alder reaction (383 K) the conformational equilibria are likely to be more shifted towards 2a(B–D). This would increase the probability of their reaction with 3a preferably giving compound 5k over 4k.

It should be noted that the calculated reaction energies for 2a + 3a = 4k + 5k are comparable to the experimentally determined enthalpy values for the Diels–Alder reaction of N-hydroxy- and N-hydroxyethylmaleimide with furfuryl alcohol.[12] Similarly, our calculated activation energy values are comparable to the experimentally found ones for the Diels–Alder reaction between furan with N-phenylmaleimide[13] as well as DFT calculated values for eight different Diels–Alder reactions.[14]

In conclusion, N-benzylcytisine (2c) reacts with N-methyl- and N-benzylmaleimides (3a and 3e) in refluxing toluene to give two endo-isomeric Diels–Alder adducts 4 with syn- and 5 with anti-orientation of the imide group with respect to the methylene group of the 3,7-diazabicyclo[3.3.1]nonane fragment. Preparative yields and simple workup procedures for compounds 4i, 4j and 5i, 5j make them promising starting materials for the simple synthesis of structurally complex and conformationally restrained amines. The corresponding diamines 6a–c can be prepared in high yield via Pd/C catalyzed complete hydrogenations of 4i and 5i, 5j, which in the removal of the benzyl protection from the amino group and are the addition of hydrogen to the C=C bond. Further complete reduction of the carbonyl groups in diamine 6c with LiAlH₄ readily affords triamine 7 containing exclusively sp³-hybridized carbon atoms.

Quantum electron-correlation computation methods suggest that possible conformers of N-methylcytisine (2a) may react with N-methylmaleimide (3a) to give both 4k and 5k isomers. The most stable isomer of 2a, having the chair conformation of the piperidine fragment and equatorial orientation of the N-Me group is predicted to lead to the slightly preferential formation of 4k over 5k under thermodynamic control conditions and to a nearly 1:1 mixture of 4k and 5k under kinetically controlled conditions. However, the three more energetic conformers of 2a are predicted to lead to the preferential formation of isomer 5k due to either lower activation barrier or steric reasons in accordance with the experimental results for 4i, 4j and 5i, 5j reported here.

Reagents and solvents were purchased from commercial suppliers (Sigma-Aldrich, Apollo Scientific, Fisher, Acros Organic, and Alfa Aesar) and used without further purification. All reactions were performed under argon or N₂ atmosphere. Chromatography solvents were HPLC grade and used without further purification. All solvent mixtures are quoted as volumes before mixing (v/v). Analytical TLC was performed using ALUGRAM Xtra SIL G/UV254 plates. Column chromatography was performed on Combiflash RF 200 using Kieselgel Merck 60 (230–400 mesh) as the stationary phase. NMR spectra were recorded on a Varian Gemini 2000 using an internal deuterium lock. Chemical shifts (δ) of ¹H and ¹³C nuclei are measured in parts per million (ppm) relative to TMS as an internal standard. LC-MS and GC-MS spectra were recorded on an Agilent 1100 LCMSD SL instrument (electrospray ionization (ESI)) or Agilent GC7820A/MSD5977B system [electron impact ionization (EI), ionization energy 70 eV]. The internal normalization method allowing the quantitative determination of a sample component by relating its peak area to that of the total peak was used. Conversion measurements were performed using the same GC-MS instrument (column: HP-5ms UI, 30–0.25 mm, 0.25 μm; carrier gas: He at 1 mL/min; temperatures: injector 250 °C, oven program 50 °С initial temperature for 1 min, then ramp to 300 °С at 20 °С/min, then hold the final temperature for 5 min; MSD transfer line 280 °C, MSD source 230 °C, MSD quad 150 °C; injection parameters: split ratio 200:1, 0.5 μL injected; MS parameters: mass scan range 35–550, ionization energy 70 eV). Elemental analyses were carried out on a LECO CHN-900 analyzer. The values of elemental analyses reported herein are averages of three independent measurements. Optical rotation angles were measured on an MCP 300 polarimeter.

(1R,5S)-3-Benzyl-1,2,3,4,5,6-hexahydro-8H-1,5-methanopyrido[1,2-a][1,5]diazocin-8-one (2с)

A suspension of K₂CO₃ (25 g, 0.178 mol) in a solution of (–)-cytisine (1; 17.1 g, 0.1 mol) and benzyl bromide (20.4 g, 0.12 mol) in MeCN (150 mL) was stirred at rt (20 °C) for 16 h. The reaction mixture was filtered off, the precipitate was washed with MeCN (2 × 20 mL), the combined filtrates were evaporated under reduced pressure, and the residue was partitioned between DCM (200 mL) and H₂O (100 mL). The organic layer was separated and dried (MgSO₄) overnight, filtered, and diluted with EtOAc (200 mL). The solution was concentrated to a volume of 100 mL. The crystallized product was filtered, washed with EtOAc, and dried in vacuum to afford 2c; yield: 25 g ((90%).

Spectroscopic data were identical to that reported in the literature.[15]

Synthesis of Compounds 4i, 4j and 5i, 5j

A solution of N-benzyl-(–)-cytisine (2c; 2.3 g, 8.2 mmol) and 3a (3.3 g, 29 mmol) or 3e (5.5 g, 29 mmol) in toluene (7 mL) was stirred under reflux for 40 h (for 4i, 5i) or 16 h (for 4j, 5j).

In the case of 3a: The excess of 3a precipitated from the reaction mixture at rt was filtered off. The filtrate was evaporated in vacuo and the residue was purified by column chromatography on silica gel (acetone/hexane) to give 5i (1.08 g) and 4i (1.0 g); total yield: 63%.

In the case of 3e: The solvent was removed under reduced pressure and the residue was subjected to column chromatography on silica gel (acetone/hexane) to give 1.6 g of 5j and 0.4 g of 4j; total yield: 52%.

3aR,4S,8S,12R,12aS,12bR)-10-Benzyl-2-methyl-3a,7,8,9,10,11,12,12b-octahydro-1H-4,12a-etheno-8,12-methanopyrrolo[3′,4′:3,4]pyrido[1,2-a][1,5]diazocine-1,3,5(2H,4H)-trione (4i)

Yield: 1 g (30%); mp 173–174 °C; [α]_D ²⁰ +6.34 (c 0.177, MeOH).

¹H NMR (CDCl₃, 400 MHz): δ = 7.32–7.16 (m, 5 H), 6.26 (dd, J = 7.5, 5.9 Hz, 1 H), 6.17 (dd, J = 7.5, 1.6 Hz, 1 H), 3.89 (ddd, J = 6.0, 3.2, 1.5 Hz, 1 H), 3.70 (d, J = 12.4 Hz, 1 H), 3.58 (d, J = 8.0 Hz, 1 H), 3.50 (d, J = 13.0 Hz, 1 H), 3.43 (d, J = 13.0 Hz, 1 H), 3.42–3.29 (m, 2 H), 3.24 (dd, J = 8.0, 3.2 Hz, 1 H), 2.82 (s, 4 H), 2.46 (d, J = 6.6 Hz, 1 H), 2.46 (s, 1 H), 2.26 (ddd, J = 18.7, 11.5, 2.2 Hz, 2 H), 1.74 (t, J = 3.5 Hz, 2 H).

¹³C NMR (CDCl₃, 101 MHz): δ = 176.43, 175.78, 173.23, 140.46, 138.69, 130.52, 129.65, 128.93, 127.91, 64.53, 63.74, 61.00, 56.71, 48.25, 47.93, 46.06, 42.59, 34.72, 27.80, 27.01, 25.75, 17.34.

MS (ESI): m/z calcd for C₂₃H₂₆N₃O₃ [M + H]⁺: 392.19; found: 392.2.

Anal. Calcd for C₂₃H₂₅N₃O₃: C, 70.57; H, 6.44; N, 10.73. Found: C, 70.48; H, 6.50; N, 10.65.

(3aS,4R,8S,12R,12aR,12bS)-2,10-Dibenzyl-3a,7,8,9,10,11,12,12b-octahydro-1H-4,12a-etheno-8,12-methanopyrrolo[3′,4′:3,4]pyrido[1,2-a][1,5]diazocine-1,3,5(2H,4H)-trione (4j)

Yield: 0.4 g (11%); mp 145–148 °C; [α]_D ²⁰ +13.33 (c 0.123, MeOH).

¹H NMR (CDCl₃, 400 MHz): δ = 7.33–7.22 (m, 3 H), 7.26–7.16 (m, 6 H), 7.18–7.08 (m, 1 H), 6.11 (dd, J = 7.5, 5.9 Hz, 1 H), 6.03 (dd, J = 7.5, 1.6 Hz, 1 H), 4.45 (s, 2 H), 3.86 (ddd, J = 6.0, 3.2, 1.6 Hz, 1 H), 3.72–3.63 (m, 1 H), 3.52–3.37 (m, 3 H), 3.40–3.27 (m, 2 H), 3.18 (dd, J = 8.0, 3.2 Hz, 1 H), 2.82 (dd, J = 10.4, 2.7 Hz, 1 H), 2.42 (t, J = 3.2 Hz, 1 H), 2.24 (ddd, J = 12.3, 4.6, 2.2 Hz, 2 H), 2.20–2.13 (m, 1 H), 1.70 (q, J = 3.1, 2.4 Hz, 2 H).

¹³C NMR (CDCl₃, 101 MHz): δ = 175.55, 174.91, 172.70, 139.90, 138.15, 135.45, 129.81, 129.24, 128.66, 128.61, 128.29, 128.04, 127.32, 63.84, 62.92, 60.42, 55.71, 47.51, 46.97, 45.35, 42.48, 41.74, 33.86, 27.01, 26.18.

MS (ESI): m/z calcd for C₂₉H₃₀N₃O₃ [M + H]⁺: 468.22; found: 468.2.

Anal. Calcd for C₂₉H₂₉N₃O₃: C, 74.50; H, 6.25; N, 8.99. Found: C, 74.38; H, 6.34; N, 8.91.

(3aS,4R,8S,12R,12aR,12bS)-10-Benzyl-2-methyl-3a,7,8,9,10,11,12,12b-octahydro-1H-4,12a-etheno-8,12-methanopyrrolo[3′,4′:3,4]pyrido[1,2-a][1,5]diazocine-1,3,5(2H,4H)-trione (5i)

Yield 1.08 g (33%); mp 196–197 °C; [α]_D ²⁰ –55.0 (c 0.17, MeOH).

¹H NMR (400 MHz, CDCl₃): δ = 7.28–7.11 (m, 5 H), 6.24 (dd, J = 7.8, 6.0 Hz, 1 H), 6.09 (dd, J = 8.0, 1.6 Hz, 1 H), 3.92 (ddd, J = 5.6, 3.3, 1.5 Hz, 1 H), 3.54 (d, J = 13.3 Hz, 1 H), 3.47 (d, J = 13.1 Hz, 1 H), 3.34 (dd, J = 13.4, 6.8 Hz, 1 H), 3.32–3.22 (m, 2 H), 3.19 (dd, J = 7.8, 3.3 Hz, 1 H), 3.17–3.05 (m, 2 H), 2.87–2.78 (m, 1 H), 2.82 (s, 3 H), 2.32–2.23 (m, 1 H), 2.20–2.11 (m, 2 H), 1.98–1.89 (m, 1 H), 1.79 (dq, J = 13.7, 3.3 Hz, 1 H).

¹³C NMR (CDCl₃, 101 MHz): δ = 175.67, 175.34, 171.71, 138.31, 135.04, 129.06, 128.74, 128.16, 127.20, 63.11, 63.02, 60.46, 55.87, 48.66, 47.77, 44.97, 42.89, 29.76, 26.67, 26.04, 24.97.

MS (ESI): m/z calcd for C₂₃H₂₆N₃O₃ [M + H]⁺: 392.19; found: 392.2.

Anal. Calcd for C₂₃H₂₅N₃O₃: C, 70.57; H, 6.44; N, 10.73. Found: C, 70.51; H, 6.46; N, 10.70.

(3aR,4S,8S,12R,12aS,12bR)-2,10-Dibenzyl-3a,7,8,9,10,11,12,12b-octahydro-1H-4,12a-etheno-8,12-methanopyrrolo[3′,4′:3,4]pyrido[1,2-a][1,5]diazocine-1,3,5(2H,4H)-trione (5j)

Yield 1.6 g (42%); mp 170–172 °C; [α]_D ²⁰ –38.0 (c 0.16, MeOH).

¹H NMR (CDCl₃, 400 MHz): δ = 7.31–7.10 (m, 10 H), 6.13 (dd, J = 7.9, 6.0 Hz, 1 H), 6.02 (dd, J = 7.9, 1.6 Hz, 1 H), 4.48 (d, J = 3.0 Hz, 2 H), 3.92 (ddd, J = 5.9, 3.2, 1.6 Hz, 1 H), 3.52 (d, J = 13.4 Hz, 1 H), 3.44 (d, J = 13.1 Hz, 1 H), 3.37–3.26 (m, 3 H), 3.22–3.05 (m, 3 H), 2.84–2.77 (m, 1 H), 2.30–2.10 (m, 3 H), 1.97–1.87 (m, 1 H), 1.83–1.72 (m, 1 H).

¹³C NMR (CDCl₃, 101 MHz): δ = 175.29, 174.99, 171.71, 138.31, 135.36, 135.03, 128.97, 128.74, 128.65, 128.19, 128.12, 127.18, 63.20, 62.99, 60.27, 56.06, 48.55, 47.76, 45.06, 42.82, 42.63, 29.81, 26.65, 26.00.

MS (ESI): m/z calcd for C₂₉H₃₀N₃O₃ [M + H]⁺: 468.22; found: 468.2.

Anal. Calcd for C₂₉H₂₉N₃O₃: C, 74.50; H, 6.25; N, 8.99. Found: C, 74.44; H, 6.31; N, 8.91.

Synthesis of Compounds 6; General Procedure

A suspension of Pd/C in a solution of the corresponding N-benzyl protected derivative 4 or 5 (10 mmol) in MeOH (100 mL) was vigorously stirred under H₂ atmosphere (10 atm) for 96 h at 30 °C. The catalyst was filtered off, washed with MeOH, and the filtrate was evaporated in vacuo to give the crude product as a viscous oil; which was purified by crystallization.

(3aS,4R,8S,12R,12aR,12bS)-2-Methyloctahydro-1H-4,12a-ethano-8,12-methanopyrrolo[3′,4′:3,4]pyrido[1,2-a][1,5]diazocine-1,3,5(2H,4H)-trione (6a)

This compound was prepared from 4i (0.95 g, 2.4 mmol) according to the general procedure. The oily crude product was crystallized from MeCN; yield 0.41 g (55%); mp 188–190 °C; [α]_D ²⁰ +51.35 (c 0.197, MeOH).

¹H NMR (CDCl₃, 400 MHz): δ = 3.87–3.73 (m, 2 H), 3.59 (dd, J = 13.6, 7.5 Hz, 1 H), 3.49 (d, J = 13.6 Hz, 1 H), 3.24 (ddd, J = 9.8, 3.8, 1.7 Hz, 1 H), 3.04 (td, J = 3.5, 2.4 Hz, 1 H), 2.99 (s, 2 H), 2.99–2.91 (m, 1 H), 2.90 (dd, J = 10.9, 2.4 Hz, 1 H), 2.76 (dd, J = 12.5, 2.2 Hz, 1 H), 2.09 (dt, J = 7.1, 3.3 Hz, 1 H), 2.01–1.68 (m, 4 H), 1.72–1.53 (m, 1 H), 1.39 (ddd, J = 13.8, 10.4, 4.6 Hz, 1 H).

¹³C NMR (CDCl₃, 101 MHz): δ = 177.13, 176.46, 173.22, 60.81, 53.63, 49.71, 47.19, 46.17, 42.89, 38.77, 33.18, 30.14, 26.58, 26.37, 25.11, 19.45.

MS (ESI): m/z calcd for C₁₆H₂₂N₃O₃ [M + H]⁺: 304.16; found: 304.2.

Anal. Calcd for C₁₆H₂₁N₃O₃: C, 63.35; H, 6.98; N, 13.85. Found: C, 63.25; H, 7.04; N, 13.77.

(3aR,4S,8S,12R,12aS,12bR)-2-Methyloctahydro-1H-4,12a-ethano-8,12-methanopyrrolo[3′,4′:3,4]pyrido[1,2-a][1,5]diazocine-1,3,5(2H,4H)-trione (6b)

This compound was prepared from 5i (0.95 g, 2.4 mmol) according to the general procedure. The crude oily product was crystallized from i-PrOH; yield 0.33 g (45%); mp 211–217 °C (dec.); [α]_D ²⁰ –35.3 (c 0.187, MeOH).

¹H NMR (400 MHz, DMSO-d ₆): δ = 3.40 (dd, J = 13.6, 7.1 Hz, 1 H), 3.37–3.26 (m, 2 H), 3.15–3.07 (m, 1 H), 3.00 (d, J = 12.2 Hz, 1 H), 2.90–2.67 (m, 5 H), 2.68 (dt, J = 3.7, 2.0 Hz, 1 H), 2.65–2.51 (m, 4 H), 2.17 (dd, J = 14.6, 10.3 Hz, 1 H), 2.02–1.89 (m, 2 H), 1.72–1.60 (m, 2 H), 1.55–1.36 (m, 2 H).

¹³C NMR (101 MHz, DMSO-d ₆): δ = 176.87, 176.38, 172.00, 60.19, 52.73, 48.04, 47.22, 46.25, 42.99, 40.40, 38.61, 29.53, 25.92, 25.67, 25.45, 24.62, 18.19.

MS (ESI): m/z calcd for C₁₆H₂₂N₃O₃ [M + H]⁺: 304.16; found: 304.2.

Anal. Calcd for C₁₆H₂N₃O₃: C, 63.35; H, 6.98; N, 13.85. Found: C, 63.24; H, 7.05; N, 13.73.

(3aR,4S,8S,12R,12aS,12bR)-2-Benzyloctahydro-1H-4,12a-ethano-8,12-methanopyrrolo[3′,4′:3,4]pyrido[1,2-a][1,5]diazocine-1,3,5(2H,4H)-trione (6c)

This compound was prepared from 5j (1 g, 2.1 mmol) according to the general procedure. The addition of 47% aq HBr (1.8 mL, 0.01 mol) to the solution of the crude oily product in MeCN resulted in the precipitation of the hydrobromide, which was filtered off, washed with MeCN, and dried in a vacuum to give 6c·HBr; yield: 0.82 g (84%); mp 367–369 °C; [α]_D ²⁰ –32.75 (c 0.16, MeOH).

¹H NMR (400 MHz, DMSO-d ₆): δ = 8.45 (s, 2 H), 7.39–7.25 (m, 5 H), 4.58 (d, J = 2.7 Hz, 2 H), 3.64–3.51 (m, 2 H), 3.42 (d, J = 13.7 Hz, 1 H), 3.30 (dd, J = 13.8, 6.8 Hz, 1 H), 3.20 (qd, J = 6.5, 5.9, 3.6 Hz, 2 H), 3.08 (dd, J = 14.0, 3.2 Hz, 2 H), 2.85 (s, 1 H), 2.76–2.70 (m, 1 H), 2.30 (s, 1 H), 2.02 (d, J = 14.2 Hz, 2 H), 1.93 (t, J = 8.7 Hz, 2 H), 1.64 (t, J = 11.9 Hz, 1 H), 1.34–1.18 (m, 2 H).

¹³C NMR (101 MHz, DMSO-d ₆): δ = 176.16, 175.74, 173.86, 135.74, 128.50, 127.76, 127.67, 59.06, 47.89, 45.91, 44.91, 43.72, 42.76, 41.95, 38.63, 27.92, 26.47, 23.37, 22.48, 17.12.

MS (ESI): m/z calcd for C₂₂H₂₆N₃O₃ [M + H]⁺: 380.19; found: 380.2.

Anal. Calcd for C₂₂H₂₆BrN₃O₃: C, 57.40; H, 5.69; N, 9.13. Found: C, 57.40; H, 5.69; N, 9.13.

(3aR,4S,8S,12R,12aS,12bS)-2-Methyldodecahydro-1H-4,12a-ethano-8,12-methanopyrrolo[3′,4′:3,4]pyrido[1,2-a][1,5]diazocine (7)

A solution of oily crude base 6c (0.95 g, 2.5 mmol) in toluene (2 mL) was added in 3 portions to a stirred solution of LiAlH₄ (0.8 g, 21 mmol) in anhyd THF (15 mL) under argon atmosphere. The reaction mixture was slowly warmed up and stirred under reflux for 14 h. After cooling down to rt, the mixture was diluted with benzene (10 mL) and carefully neutralized with 40% aq NaOH (1 mL). The mixture was stirred at rt for 7 h, filtered, and the filtrate was evaporated in vacuum. The residue was dissolved in benzene (15 mL) and the solution was stirred with 40% aq NaOH (1 mL) for 7 h. The benzene layer was separated and the solvent was removed under reduced pressure to give a viscous oil. The oil was dissolved in MeCN and neutralized with 47% aq HBr (0.8 mL, 4.5 mmol). After 3 days, the sample consisted of fine white powder and a few well-shaped single crystals. The crystals were removed from the sample and kept under mother liquor and the powder was filtered off to give 7·2HBr as a white crystalline powder; yield: 0.74 g (58%); mp 274–277 °C; [α]_D ²⁰ +2.67 (c 0.15, MeOH).

¹H NMR (400 MHz, D₂O): δ = 7.49 (s, 5 H), 4.46 (d, J = 13.0 Hz, 1 H), 4.38 (d, J = 12.8 Hz, 1 H), 3.70 (s, 1 H), 3.59 (d, J = 13.3 Hz, 1 H), 3.52–3.38 (m, 2 H), 3.33 (dd, J = 11.3, 2.3 Hz, 1 H), 3.20 (d, J = 12.4 Hz, 2 H), 3.11–2.96 (m, 3 H), 2.87 (d, J = 12.1 Hz, 1 H), 2.24 (d, J = 11.3 Hz, 1 H), 2.13–2.00 (m, 2 H), 1.89 (d, J = 15.9 Hz, 1 H), 1.86–1.74 (m, 1 H), 1.74–1.60 (m, 4 H), 1.55 (s, 1 H).

¹³C NMR (CDCl₃, 101 MHz): δ = 131.09, 130.79, 130.77, 130.04, 58.11, 56.92, 56.30, 55.64, 54.67, 50.85, 47.02, 37.21, 32.37, 27.36, 26.44, 26.41, 25.90, 17.98.

MS (ESI): m/z calcd for C₂₂H₃₂N₃O₃ [M + H]⁺: 338.25; found: 338.2.

Anal. Calcd for C₂₂H₃₃Br₂N₃: C, 52.92; H, 6.66; N, 8.42. Found: C, 52.83; H, 6.74; N, 8.37.

Single-Crystal X-ray Analysis

All crystallographic measurements were performed on a Bruker Smart Apex II diffractometer operating in the ω scan mode Mo-Kα radiation (λ = 0.71078 Å). The structures were solved by direct methods and refined by the full-matrix least-squares technique in the anisotropic approximation for nonhydrogen atoms using the Bruker SHELXTL program package.[16] All CH hydrogen atoms were placed at calculated positions and defined as the ‘riding’ model.

4i: C₂₃H₂₅N₃O₃ crystal size 0.51 × 0.37 × 0.22 mm³, monoclinic, P2₁, T = 173K, a = 11.864(3) Å, b = 6.699(1) Å, c = 12.991(3) Å, β = 110.403(6)°, V = 967.8(3) ų, ρ _c =1.343 g cm^–3, 2θ max = 54.9, Z = 2, μ = 0.09 mm^–1, F(000) = 416, 263 parameters, R1 = 0.0351, wR₂ = 0.0826 (for 4162 reflections [I > 2σ(I)]), R = 0.0396, wR(F2) = 0.0854 (for 4527 unique reflections), R(int) = 0.0352, S = 1.04, ρ (min/max) = –0.22/ 0.32 e Å^–3.

4j: C₂₉H₂₉N₃O₃ crystal size 0.49 × 0.40 × 0.30 mm³, orthorhombic, P2₁2₁2, T = 173K, a = 6.4896(12) Å, b = 10.907(2) Å, c = 33.620(6) Å, V = 2379.7(8) ų, ρ _c = 1.305 g cm^–3, 2θ max = 51.6, Z = 4, μ = 0.085 mm^–1, F(000) = 992, 316 parameters, R1 = 0.0575, wR2 = 0.1255 (for 5186 reflections [I > 2σ(I)]), R = 0.0683, wR(F2) = 0.1305 (for 5990 unique reflections), R(int) = 0.0456, S = 1.034, ρ (min/max) = –0.241/0.245 e Å^–3.

5i: C₂₃H₂₅N₃O₃ crystal size 0.44 × 0.38 × 0.14 mm³, orthorhombic, P2₁2₁2₁, T =173 K, a = 7.119(1) Å, b = 11.407(2) Å, c = 23.857(5) Å, V= 1937.4(6) ų, ρ _c =1.342 g cm^–3, 2θ max = 54.0, Z = 4, μ = 0.090 mm^–1, F(000) = 832, 263 parameters, R1 = 0.0407, wR2 = 0.0842 (for 3537 reflections [I > 2σ(I)]), R = 0.0533, wR(F2) = 0.0892 (for 4227 unique reflections), R(int) = 0.0576, S = 1.010, ρ (min/max) = –0.219/0.186 e Å^–3.

5j: C₂₉H₂₉N₃O₃ crystal size 0.38 × 0.20 × 0.18 mm³, orthorhombic, P2₁2₁2₁, T = 296(2) K, a = 8.635(1) Å, b = 9.332(2) Å, c = 29.644(5) Å, V = 2388.8(7) ų, ρ _c = 1.300 g cm^–3, 2θ max = 50.48, Z = 4, μ =0.085 mm^–1, F(000) = 992, 316 parameters, R1 = 0.0560, wR2 = 0.1132 (for 3923 reflections [I > 2σ(I)]), R = 0.1039, wR(F2) = 0.1318 (for 6210 unique reflections), R(int) = 0.0521, S = 1.035, ρ (min/max) = –0.295/ 0.167 e ^–3.

6a: C₁₆H₂₁N₃O₃ crystal size 0.48 × 0.31 × 0.20 mm³, orthorhombic, P2₁2₁2₁, T = 123(2) K, a = 7.590(1) Å, b = 10.103(2) Å, c = 18.636(4) Å, V = 1429.1(4) ų, ρ _c = 1.410 g cm^–3, 2θ max = 53.00, Z = 4, μ = 0.099 mm^–1, F(000) = 548, 204 parameters, R1 = 0.0349, wR2 = 0.0707 (for 2636 reflections [I > 2σ(I)]), R = 0.0427, wR(F2) = 0.0739 (for 2962 unique reflections), R(int) = 0.0441, S = 1.035, ρ (min/max) = –0.195/ 0.183 e Å^–3.

6b: C₁₆H₂₁N₃O₃ crystal size 0.50 × 0.45 × 0.13 mm³, orthorhombic, P2₁2₁2₁, T = 123(2) K, a = 6.365(1) Å, b = 8.3462(2) Å, c = 26.512(5) Å, V = 1408.5(4) ų, ρ _c = 1.431 g cm^–3, 2θ max = 53.00, Z = 4, μ = 0.10 mm^–1, F(000) = 648, 204 parameters, R1 = 0.0393, wR2 = 0.0945 (for 3200 reflections [I > 2σ(I)]), R = 0.0453, wR(F2) = 0.0970 (for 3546 unique reflections), R(int) = 0.0371, S = 1.045, ρ (min/max) = –0.232/ 0.284 e Å^–3.

0.8 (7·3HBr·MeCN) 0.2 (8·3HBr·MeCN): C₂₄H₃₇Br₃N₄O_0.2 crystal size 0.46 × 0.38 × 0.05 mm³, orthorhombic, P2₁2₁2₁, T =173(2) K, a = 8.688(1) Å, b = 16.388(3) Å, c = 18.390(4) Å, V = 2618.5(8) ų, ρ _c =1.576 g cm^–3, 2θ max = 53.02, Z = 4, μ = 4.641 mm^–1, F(000) = 1256, 310 parameters, R1 = 0.0354, wR2 = 0.0693 (for 4401 reflections [I > 2σ(I)]), R = 0.0557, wR(F2) = 0.0754 (for 5419 unique reflections), R(int) = 0.0954, S = 1.035, ρ (min/max) = –0.44/0.54 e Å^–3.

Computational Details

Since the choice of a proper computational methodology is crucial to obtain reasonable reaction and activation energies for the typical Diels–Alder reactions, a set of electron-correlation methods based on the second-order Møller–Plesset (MP2)[17] and coupled-cluster (CC)[18] theory has been employed here. We deliberately omitted density-functional theory because of its inability to treat London dispersion interactions leading to significant underestimation of reaction energies (even with an empirical dispersion term included), although the respective activation energies may still be acceptable. The conventional MP2 method, on the contrary, results in quite accurate reaction energies but much lower activation energies on account of an overestimation of the dispersion energy contribution stemming from an unbalanced electron correlation.[19] A well-established remedy involves different scaling of the parallel and antiparallel spin components of the correlation energy leading to the spin-component-scaled MP2 (SCS-MP2) approach,[20] which has been used in this work. To gain confidence in the SCS-MP2 findings, the domain-based local pair natural orbital (DLPNO) scheme of the CC method involving single-, double-, and perturbative triple excitations [i.e., DLPNO-CCSD(T)][21] has also been employed here as a high-accuracy benchmark method. The single-point calculations based on the latter method relied on the TightPNO truncation thresholds by Liakos et al.[22] suitable for very accurate modeling of nonbonded and conformationally flexible systems.

All these calculations have been performed with the Orca program package, version 4.2.[23]

Two orbital basis sets were employed in this work: Pople’s 6-31G*[24] and Dunning’s aug-cc-pVTZ.[25] The first one ensured fast evaluation of the MP2 energy and gradients essential for the geometry optimizations as well as for the numerical Hessian matrix generation needed for the next transition state optimization or the frequency calculation. The second one was used to refine the obtained 6-31G* energetics either through the respective geometry optimizations (when affordable) or single-point calculations. In the case of correlation fitting basis sets needed for the resolution-of-the-identity (RI) approximation, we relied either on the automatic auxiliary basis set generation (AutoAux feature in Orca)[26] in conjunction with the 6-31G* orbital basis set or on the aug-cc-pVTZ/C basis set of Weigend et al.[27] for the aug-cc-pVTZ orbital basis set.

Geometry optimizations were performed with the MP2/6-31G* and MP2/aug-cc-pVTZ model chemistries followed by single-point DLPNO­-CCSD(T)/aug-cc-pVTZ computations. Tight values of the SCF convergence criteria (TightSCF keyword in Orca, with the energy change of 10^–8 Hartree between the two consecutive SCF cycles) were used throughout this work whereas the default geometry convergence criteria (energy change = 5·10-6 Hartree, RMS and a maximum gradient of 1·10^–4 a.u. and 3·10^–4 a.u., respectively; RMS and maximum displacement of 2·10^–3 a.u. and 4·10^–3 a.u., respectively) were utilized for all systems. The optimized geometries were characterized as minima or first-order saddle points on the potential energy surface through the eigenvalues of the (numerically) calculated Hessian matrix leading to zero or one negative eigenvalue for the minimum and transition-state structure, respectively. To start the transition state optimization, the MP2/6-31G* Hessian matrix for the appropriately chosen initial structure was calculated first, whereas the subsequent MP2/aug-cc-pVTZ transition state optimizations had the previously calculated MP2/6-31G* Hessian matrix as an initial starting basis for the optimizer. Zero-point energies (ZPE) were calculated from the obtained harmonic vibrational frequencies without any scaling.

The reaction energy ΔE was calculated as the energy difference between the products and reactants (in their lowest-energy conformations) while the activation energy ΔE ^≠ corresponds to the energy needed for the reactant to reach a given transition state. Additionally, transition state structures were characterized by deformation energies, E _def, which represent the energy required for a reactant to distort to its transition-state geometry. Overall, the activation energy of a typical Diels–Alder reaction can be partitioned into the following contributions:

ΔE ^≠ = E _def (diene) + E _def (dienophile) + E _int

where E _def (diene) and E _def (dienophile) are respective deformation energies of the diene and the dienophile whereas E _int reflects the remaining interaction energy between these reactants at the transition state.

Conflict of Interest

The authors declare no conflict of interest.

• References

• 1a Blom AE. M, Rego Campello H, Lester HA, Gallagher T, Dougherty DA. J. Am. Chem. Soc. 2019; 141: 15840
  CrossrefPubMedSearch in Google Scholar
• 1b Butler MS, Robertson AA. B, Cooper MA. Nat. Prod. Rep. 2014; 31: 1612
  CrossrefPubMedSearch in Google Scholar
• 1c Philipova I, Stavrakov G, Vassilev N, Nikolova R, Shivachev B, Dimitrov V. J. Organomet. Chem. 2015; 778: 10
  CrossrefSearch in Google Scholar
• 2a Etter J.-F. Arc. Intern. Med. 2006; 166: 1553 
  CrossrefPubMedSearch in Google Scholar
• 2b West R, Zatonski W, Cedzynska M, Lewandowska D, Pazik J, Aveyard P, Stapleton J. N. Engl. J. Med. 2011; 365: 1193
  CrossrefPubMedSearch in Google Scholar
• 3 Rouden J, Lasne M.-C, Blanchet J, Baudoux J. Chem. Rev. 2013; 113: 712
  CrossrefPubMedSearch in Google Scholar
• 4a Sakhautdinov IM, Mukhamet’yanova AF, Dosniyazova AG, Vinogradova VI, Lobov AN, Yunusov MS. Chem. Nat. Comp. 2019; 55: 398
  CrossrefSearch in Google Scholar
• 4b Liu C, Watt DS, Frasinyuk MS, Sviripa VM, Zhang W, Bondarenko SP. US 20180344862 A1, 2018
  PubMed
• 4c Przybyl AK, Maj E, Wietrzyk J, Kubicki M. J. Mol. Struct. 2019; 1176: 871
  CrossrefSearch in Google Scholar
• 5 Lator A, Gaillard QG, Merel DS, Lohier J.-F, Gaillard S, Poater A, Renaud J.-L. J. Org. Chem. 2019; 84: 6813
  CrossrefPubMedSearch in Google Scholar
• 6 Tsypysheva I, Petrova P, Koval’skaya A, Lobov A, Sapozhnikova T, Makara N, Gabdrakhmanova S, Zarudii F. Nat. Prod. Res. 2021; 35: 207
  CrossrefPubMedSearch in Google Scholar
• 7 Marrière E, Rouden J, Tadino V, Lasne M.-C. Org. Lett. 2000; 2: 1121
  CrossrefPubMedSearch in Google Scholar
• 8 Gallagher TC, Rego Campello H. PCT Int. Appl WO 2018033742 A2, 2018
  PubMed
• 9a Tsypysheva IP, Lobov AN, Kovalskaya AV, Vinogradova VI, Suponitsky KY, Khursan SL, Yunusov MS. Tetrahedron: Asymmetry 2013; 24: 1318
  CrossrefSearch in Google Scholar
• 9b Tsypysheva IP, Lobov AN, Kovalskaya AV, Petrova PR, Ivanov SP, Rameev SA, Yunusov MS. Nat. Prod. Res. 2014; 29: 141
  CrossrefPubMedSearch in Google Scholar
• 9c Tsypysheva IP, Borisevich SS, Lobov AN, Kovalskaya AV, Shamukaev VV, Safiullin RL, Khursan SL. Tetrahedron: Asymmetry 2015; 26: 732 
  CrossrefSearch in Google Scholar
• 9d Tsypysheva I, Koval’skaya A, Petrova P, Lobov A, Borisevich S, Tsypyshev D, Fedorova V, Gorbunova E, Galochkina A, Zarubaev V. Tetrahedron 2019; 75: 2933
  CrossrefSearch in Google Scholar
• 10 Freer AA, Robins DJ, Sheldrick GM. Acta Crystallogr., Ser. C 1987; 43: 1119
  CrossrefSearch in Google Scholar
• 11 This result is consistent with the stability trend reported earlier for the Diels–Alder reaction of N-phenylmaleimide with 2a: Borisevich SS, Kovalskaya AV, Tsypysheva IP, Khursan SL. J. Theor. Comput. Chem. 2014; 13: 1450048
  CrossrefSearch in Google Scholar
• 12 de Oliveira JC, Laborie M.-P, Roucoules V. Molecules 2020; 25: 243
  CrossrefPubMedSearch in Google Scholar
• 13 Kornilov DA, Kiselev VD, Anikin OV, Kolesnikova AO, Shulyat’ev AA. Russ. J. Org. Chem. 2019; 55: 7
  CrossrefSearch in Google Scholar
• 14 Black K, Liu P, Xu L, Doubleday C, Houk KN. Proc. Natl. Acad. Sci. U. S. A. 2012; 109: 12860
  CrossrefPubMedSearch in Google Scholar
• 15 Honda Е, Takahashi R, Namiki H. J. Org. Chem. 2005; 70: 499
  CrossrefPubMedSearch in Google Scholar
• 16a Sheldrick GM. Acta Crystallogr., Sect. A. 2008; 64: 112
  CrossrefPubMedSearch in Google Scholar
• 16b CCDC 1967962–1967968 contain the supplementary crystallographic data for this paper. The data can be obtained free of charge from The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/getstructures.
• 17 Møller C, Plesset MS. Phys. Rev. 1934; 46: 618
  CrossrefSearch in Google Scholar
• 18a Pople JA, Head-Gordon M, Raghavachari K. J. Chem. Phys. 1987; 87: 5968
  CrossrefSearch in Google Scholar
• 18b Raghavachari K, Trucks GW, Pople JA, Head-Gordon M. Chem. Phys. Lett. 1989; 157: 479
  CrossrefSearch in Google Scholar
• 19 Goumans TP. M, Ehlers AW, Lammertsma K, Würthwein E.-U, Grimme S. Chem. Eur. J. 2004; 10: 6468
  CrossrefPubMedSearch in Google Scholar
• 20 Grimme S. J. Chem. Phys. 2003; 118: 9095
  CrossrefSearch in Google Scholar
• 21a Riplinger C, Neese F. J. Chem. Phys. 2013; 138: 034106
  CrossrefPubMedSearch in Google Scholar
• 21b Riplinger C, Sandhoefer B, Hansen A, Neese F. J. Chem. Phys. 2013; 139: 134101
  CrossrefPubMedSearch in Google Scholar
• 21c Riplinger C, Pinski P, Becker U, Valeev EF, Neese F. J. Chem. Phys. 2016; 144: 024109
  CrossrefPubMedSearch in Google Scholar
• 22 Liakos DG, Sparta M, Kesharwani MK, Martin JM, Neese F. J. Chem. Theory Comput. 2015; 11: 1525
  CrossrefPubMedSearch in Google Scholar
• 23a Neese F. Wiley Interdiscip. Rev. Comput. Mol. Sci. 2012; 2: 73
  CrossrefSearch in Google Scholar
• 23b Neese F. Wiley Interdiscip. Rev. Comput. Mol. Sci. 2018; 8: e1327
  CrossrefSearch in Google Scholar
• 24a Hariharan PC, Pople JA. Theor. Chim. Acta 1973; 28: 213
  CrossrefSearch in Google Scholar
• 24b Hehre WJ, Ditchfield R, Pople JA. J. Chem. Phys. 1972; 56: 2257
  Search in Google Scholar
• 25a Dunning TH. Jr. J. Chem. Phys. 1989; 90: 1007
  CrossrefSearch in Google Scholar
• 25b Kendall RA, Dunning TH. Jr, Harrison RJ. J. Chem. Phys. 1992; 96: 6796
  CrossrefSearch in Google Scholar
• 26 Stoychev GL, Auer AA, Neese F. J. Chem. Theory Comput. 2017; 13: 554
  CrossrefPubMedSearch in Google Scholar
• 27 Weigend F, Köhn A, Hättig C. J. Chem. Phys. 2002; 116: 3175
  CrossrefSearch in Google Scholar

Corresponding Author

Publication History

Received: 13 April 2021

Accepted after revision: 02 July 2021

Article published online:
16 August 2021

© 2021. Thieme. All rights reserved

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany

• References

• 1a Blom AE. M, Rego Campello H, Lester HA, Gallagher T, Dougherty DA. J. Am. Chem. Soc. 2019; 141: 15840
  CrossrefPubMedSearch in Google Scholar
• 1b Butler MS, Robertson AA. B, Cooper MA. Nat. Prod. Rep. 2014; 31: 1612
  CrossrefPubMedSearch in Google Scholar
• 1c Philipova I, Stavrakov G, Vassilev N, Nikolova R, Shivachev B, Dimitrov V. J. Organomet. Chem. 2015; 778: 10
  CrossrefSearch in Google Scholar
• 2a Etter J.-F. Arc. Intern. Med. 2006; 166: 1553 
  CrossrefPubMedSearch in Google Scholar
• 2b West R, Zatonski W, Cedzynska M, Lewandowska D, Pazik J, Aveyard P, Stapleton J. N. Engl. J. Med. 2011; 365: 1193
  CrossrefPubMedSearch in Google Scholar
• 3 Rouden J, Lasne M.-C, Blanchet J, Baudoux J. Chem. Rev. 2013; 113: 712
  CrossrefPubMedSearch in Google Scholar
• 4a Sakhautdinov IM, Mukhamet’yanova AF, Dosniyazova AG, Vinogradova VI, Lobov AN, Yunusov MS. Chem. Nat. Comp. 2019; 55: 398
  CrossrefSearch in Google Scholar
• 4b Liu C, Watt DS, Frasinyuk MS, Sviripa VM, Zhang W, Bondarenko SP. US 20180344862 A1, 2018
  PubMed
• 4c Przybyl AK, Maj E, Wietrzyk J, Kubicki M. J. Mol. Struct. 2019; 1176: 871
  CrossrefSearch in Google Scholar
• 5 Lator A, Gaillard QG, Merel DS, Lohier J.-F, Gaillard S, Poater A, Renaud J.-L. J. Org. Chem. 2019; 84: 6813
  CrossrefPubMedSearch in Google Scholar
• 6 Tsypysheva I, Petrova P, Koval’skaya A, Lobov A, Sapozhnikova T, Makara N, Gabdrakhmanova S, Zarudii F. Nat. Prod. Res. 2021; 35: 207
  CrossrefPubMedSearch in Google Scholar
• 7 Marrière E, Rouden J, Tadino V, Lasne M.-C. Org. Lett. 2000; 2: 1121
  CrossrefPubMedSearch in Google Scholar
• 8 Gallagher TC, Rego Campello H. PCT Int. Appl WO 2018033742 A2, 2018
  PubMed
• 9a Tsypysheva IP, Lobov AN, Kovalskaya AV, Vinogradova VI, Suponitsky KY, Khursan SL, Yunusov MS. Tetrahedron: Asymmetry 2013; 24: 1318
  CrossrefSearch in Google Scholar
• 9b Tsypysheva IP, Lobov AN, Kovalskaya AV, Petrova PR, Ivanov SP, Rameev SA, Yunusov MS. Nat. Prod. Res. 2014; 29: 141
  CrossrefPubMedSearch in Google Scholar
• 9c Tsypysheva IP, Borisevich SS, Lobov AN, Kovalskaya AV, Shamukaev VV, Safiullin RL, Khursan SL. Tetrahedron: Asymmetry 2015; 26: 732 
  CrossrefSearch in Google Scholar
• 9d Tsypysheva I, Koval’skaya A, Petrova P, Lobov A, Borisevich S, Tsypyshev D, Fedorova V, Gorbunova E, Galochkina A, Zarubaev V. Tetrahedron 2019; 75: 2933
  CrossrefSearch in Google Scholar
• 10 Freer AA, Robins DJ, Sheldrick GM. Acta Crystallogr., Ser. C 1987; 43: 1119
  CrossrefSearch in Google Scholar
• 11 This result is consistent with the stability trend reported earlier for the Diels–Alder reaction of N-phenylmaleimide with 2a: Borisevich SS, Kovalskaya AV, Tsypysheva IP, Khursan SL. J. Theor. Comput. Chem. 2014; 13: 1450048
  CrossrefSearch in Google Scholar
• 12 de Oliveira JC, Laborie M.-P, Roucoules V. Molecules 2020; 25: 243
  CrossrefPubMedSearch in Google Scholar
• 13 Kornilov DA, Kiselev VD, Anikin OV, Kolesnikova AO, Shulyat’ev AA. Russ. J. Org. Chem. 2019; 55: 7
  CrossrefSearch in Google Scholar
• 14 Black K, Liu P, Xu L, Doubleday C, Houk KN. Proc. Natl. Acad. Sci. U. S. A. 2012; 109: 12860
  CrossrefPubMedSearch in Google Scholar
• 15 Honda Е, Takahashi R, Namiki H. J. Org. Chem. 2005; 70: 499
  CrossrefPubMedSearch in Google Scholar
• 16a Sheldrick GM. Acta Crystallogr., Sect. A. 2008; 64: 112
  CrossrefPubMedSearch in Google Scholar
• 16b CCDC 1967962–1967968 contain the supplementary crystallographic data for this paper. The data can be obtained free of charge from The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/getstructures.
• 17 Møller C, Plesset MS. Phys. Rev. 1934; 46: 618
  CrossrefSearch in Google Scholar
• 18a Pople JA, Head-Gordon M, Raghavachari K. J. Chem. Phys. 1987; 87: 5968
  CrossrefSearch in Google Scholar
• 18b Raghavachari K, Trucks GW, Pople JA, Head-Gordon M. Chem. Phys. Lett. 1989; 157: 479
  CrossrefSearch in Google Scholar
• 19 Goumans TP. M, Ehlers AW, Lammertsma K, Würthwein E.-U, Grimme S. Chem. Eur. J. 2004; 10: 6468
  CrossrefPubMedSearch in Google Scholar
• 20 Grimme S. J. Chem. Phys. 2003; 118: 9095
  CrossrefSearch in Google Scholar
• 21a Riplinger C, Neese F. J. Chem. Phys. 2013; 138: 034106
  CrossrefPubMedSearch in Google Scholar
• 21b Riplinger C, Sandhoefer B, Hansen A, Neese F. J. Chem. Phys. 2013; 139: 134101
  CrossrefPubMedSearch in Google Scholar
• 21c Riplinger C, Pinski P, Becker U, Valeev EF, Neese F. J. Chem. Phys. 2016; 144: 024109
  CrossrefPubMedSearch in Google Scholar
• 22 Liakos DG, Sparta M, Kesharwani MK, Martin JM, Neese F. J. Chem. Theory Comput. 2015; 11: 1525
  CrossrefPubMedSearch in Google Scholar
• 23a Neese F. Wiley Interdiscip. Rev. Comput. Mol. Sci. 2012; 2: 73
  CrossrefSearch in Google Scholar
• 23b Neese F. Wiley Interdiscip. Rev. Comput. Mol. Sci. 2018; 8: e1327
  CrossrefSearch in Google Scholar
• 24a Hariharan PC, Pople JA. Theor. Chim. Acta 1973; 28: 213
  CrossrefSearch in Google Scholar
• 24b Hehre WJ, Ditchfield R, Pople JA. J. Chem. Phys. 1972; 56: 2257
  Search in Google Scholar
• 25a Dunning TH. Jr. J. Chem. Phys. 1989; 90: 1007
  CrossrefSearch in Google Scholar
• 25b Kendall RA, Dunning TH. Jr, Harrison RJ. J. Chem. Phys. 1992; 96: 6796
  CrossrefSearch in Google Scholar
• 26 Stoychev GL, Auer AA, Neese F. J. Chem. Theory Comput. 2017; 13: 554
  CrossrefPubMedSearch in Google Scholar
• 27 Weigend F, Köhn A, Hättig C. J. Chem. Phys. 2002; 116: 3175
  CrossrefSearch in Google Scholar

• Supplementary Material