Synthesis of bis-spirocyclic derivatives of 3-azabicyclo[3.1.0]hexane via cyclopropene cycloadditions to the stable azomethine ylide derived from Ruhemann's purple

  1. Alexander S. Filatov1ORCID Logo,
  2. Olesya V. Khoroshilova1ORCID Logo,
  3. Anna G. Larina1ORCID Logo,
  4. Vitali M. Boitsov2ORCID Logo and
  5. Alexander V. Stepakov1,3ORCID Logo

1Saint-Petersburg State University, Universitetskaya nab. 7/9, 199034, St. Petersburg, Russian Federation
2Saint-Petersburg National Research Academic University of the Russian Academy of Sciences, ul. Khlopina 8/3, 194021, St. Petersburg, Russian Federation
3Saint-Petersburg State Institute of Technology, Moskovskii pr. 26, 190013, St. Petersburg, Russian Federation

  1. Corresponding author email

Associate Editor: S. Bräse
Beilstein J. Org. Chem. 2022, 18, 769–780.
Received 13 Feb 2022, Accepted 17 Jun 2022, Published 29 Jun 2022

A non-peer-reviewed version of this article has been posted as a preprint


A reliable method for the synthesis of bis-spirocyclic derivatives of 3-azabicyclo[3.1.0]hexanes through the 1,3-dipolar cycloaddition (1,3-DC) reactions of cyclopropenes to the stable azomethine ylide – protonated form of Ruhemann's purple (PRP) has been developed. Both 3-substituted and 3,3-disubstituted cyclopropenes reacted with PRP, affording the corresponding bis-spirocyclic 3-azabicyclo[3.1.0]hexane cycloadducts in moderate to good yields with high diastereofacial selectivity. Moreover, several unstable 1,2-disubstituted cyclopropenes were successfully trapped by the stable 1,3-dipole under mild conditions. The mechanism of the cycloaddition reactions of cyclopropenes with PRP has been thoroughly studied using density functional theory (DFT) methods at the M11/cc-pVDZ level of theory. The cycloaddition reactions have been found to be HOMOcyclopropene–LUMOylide controlled while the transition-state energies for the reaction of 3-methyl-3-phenylcyclopropene with PRP are fully consistent with the experimentally observed stereoselectivity.

Keywords: azomethine ylides; cycloaddition; cyclopropenes; DFT calculations; spiro heterocycles


Spiro compounds (molecules containing at least two rings with only one common atom) are an important class of both synthetic and naturally occurring substances. Many biologically active natural products have a spirocyclic skeleton in their structure [1,2]. In this regard, there is interest in studying heterocyclic spiro compounds for drug discovery. These compounds were found to exhibit a broad range of biological activities, including antioxidant [3], anticancer [4], antidiabetic [5], and antibacterial [6] properties. It is also worth noting that spiro compounds have found application in agriculture as fungicides [7], as well as in materials science as organic semiconductors [8]. The 3-azabicyclo[3.1.0]hexane framework is a valuable structural fragment found in natural compounds [9-11]. It is used in pharmaceuticals [12-15] and key intermediates [16,17]. Compounds containing a 3-azabicyclo[3.1.0]hexane moiety are antagonists of morphine-induced antinociception [18], histone deacetylase inhibitors [13], and opioid receptor antagonists [15]. In our recent studies, great attention was paid to developing methods for the synthesis of spiro[3-azabicyclo[3.1.0]hexanes] based on 1,3-dipolar cycloaddition reactions involving azomethine ylides and cyclopropene dipolarophiles, and also the in vitro activity of some synthesized compounds has been explored [19-24]. To generate azomethine ylides, we used a classical method based on the reaction of cyclic carbonyl compounds with α-amino acids. Μono-, bi-, and tetracyclic carbonyl compounds were utilized in these studies. 1,3-Dipolar cycloaddition of cyclopropenes to azomethine ylides generated from primary or secondary α-amino acids and carbonyl compounds such as alloxan, isatin, tryptanthrin, and 11H-indeno[1,2-b]quinoxalin-11-one were performed in a multicomponent fashion [19-21,24]. In these reactions, the azomethine ylides generated in situ are highly reactive and cannot be isolated as individual compounds. At the same time, the reaction of ninhydrin and proline results in the formation of the stable azomethine ylide. This 1,3-dipole demonstrated high reactivity towards diverse cyclopropenes, including parent cyclopropene [22]. Note that the reactions of cyclopropenes with azomethine ylides from ninhydrin were also successfully carried out in a multicomponent fashion [23]. Mention should also be made of the recent advances in developing enantioselective approaches to the synthesis of 3-azabicyclo[3.1.0]hexane derivatives. Deng and co-workers reported the first asymmetric 1,3-dipolar cycloaddition of azomethine ylides and cyclopropenes catalyzed by a chiral Cu-(CH3CN)4BF4/Ph-Phosferrox complex for the construction of 3-azabicyclo[3.1.0]hexane derivatives [25]. Another concise enantioselective approach towards 3-azabicyclo[3.1.0]hexanes is based on a Cp*Ir-catalyzed reductive amination/cyclization of enantiopure cis-cyclopropane dicarbonyls [26]. The strategy based on azomethine ylide cycloadditions to cyclopropenes enables ready access to a wide range of spiro-fused 3-azabicyclo[3.1.0]hexanes (Scheme 1a). Inspired by our recent achievements, we have focused on developing an approach to the synthesis of bis-spiro[3-azabicyclo[3.1.0]hexanes] (compounds containing spiro units at the 2,4-positions of the 3-azabicyclo[3.1.0]hexane moiety) that are a hitherto unknown class of spirocyclic compounds. When considering a synthetic approach to obtaining such compounds, we have turned our attention to a tetrasubstituted azomethine ylide – the N-protonated Ruhemann's purple (PRP). This stable ylide was first utilized as a 1,3-dipole in cycloaddition reactions by Grigg and co-workers [27]. This research group demonstrated that the reactions of PRP with dipolarophiles such as N-phenylmaleimide, maleic anhydride, and methyl propiolate resulted in the formation of the corresponding bis-spirocyclic cycloadducts (Scheme 1b). It is clear that PRP has been studied in reactions with a small number of dipolarophiles. Accordingly, it is impossible to fully evaluate the synthetic potential of PRP only on the basis of this study. Positioning the current study as a continuation of a series of our works, in which cyclopropenes are utilized as dipolarophiles, we have described a synthetic route to bis-spirocyclic derivatives of 3-azabicyclo[3.1.0]hexane through cyclopropene cycloadditions to stable azomethine ylide PRP (Scheme 1c).


Scheme 1: Early studies concerning cyclopropene cycloadditions to azomethine ylides and cycloaddition reactions involving protonated Ruhemann's purple (PRP).

Results and Discussion

The study commenced with testing the feasibility of the cycloaddition reaction between protonated Ruhemann's purple (PRP, 1) [28,29] and cyclopropene dipolarophiles 2. 1,2,3-Triphenylcycloprop-1-ene (TPC, 2a) [30] was chosen as a model substrate since this alkene had worked well in cycloaddition reactions with various stabilized azomethine ylides [19-24]. Initially, we considered the reaction conditions suggested by Grigg and co-workers in the study [27]. An equimolar mixture of 1 and 2a was dissolved in tetrahydrofuran (THF), and the resulting mixture was maintained at reflux. After heating for 2 h, completion of the reaction was indicated by both the colour change of the reaction mixture and TLC (thin-layer chromatography) analysis. The proposed cycloadduct was isolated after recrystallization from methanol (MeOH) in 75% yield. Eventually, the constitution of 3a was unambiguously corroborated by NMR spectra. Given the results of previous studies [22,23] concerning cycloadditions of ninhydrin-derived azomethine ylides to cyclopropenes, it was suggested that the stable azomethine ylide 1 appears to react diastereoselectively with TPC (2a), resulting in the formation of bis-spirocyclic meso compound 3a in which the three phenyl substituents attached to the cyclopropane ring are oriented in the same direction (Scheme 2). This suggests that the diastereomer 3a resulting from the approach of the 1,3-dipole 1 from the less-hindered face of TPC (2a) is more favorable than the opposite diastereoisomer 3a' (Scheme 2). Subsequently, our hypothesis about the relative configuration of 3a was confirmed by X-ray analysis that was carried out for the related compound 3e (vide infra).


Scheme 2: The pilot experiment aimed at studying the cycloaddition reaction between the protonated form of Ruhemann's purple (1) and 1,2,3-triphenylcyclopropene (2a).

Subsequently, our efforts focused on the optimization of the reaction conditions for improving the yield of the cycloadduct 3a. A broad range of solvents was screened at different temperatures. As presented in Table 1, aprotic solvents such as 1,4-dioxane, acetonitrile, and dimethylformamide (DMF) at 65 °C favored the formation of the desired cycloadduct 3a. 3-Azabicyclo[3.1.0]hexane derivative 3a was obtained after recrystallization from MeOH in yields 67, 70, and 61%, respectively (Table 1, entries 2–4). In contrast to aprotic solvents, alcohols such as methanol (MeOH) and ethanol (EtOH) were absolutely unsuitable for carrying out this reaction owing to incompatibility of 1,3-dipole 1 with this medium (Table 1, entries 5 and 6). PRP (1) was found to immediately undergo a proton transfer reaction with alcohols, converting it into the conjugate base (Ruhemann's purple). When using dichloromethane as a solvent, the reaction took place at reflux temperature. However, in this case, the yield of 3a significantly decreased to 42% due to incomplete conversion of 1 and 2a (Table 1, entry 7). Also, we attempted to carry out this transformation in aprotic solvents at room temperature. Unfortunately, even after 12 h, we failed to completely convert reactants 1 and 2a into product 3a in both cases (Table 1, entries 8 and 9). Accordingly, it was concluded that tetrahydrofuran is the most appropriate solvent for carrying out the cycloaddition reaction between 1 and 2a. To reach full conversion of the reactants, it is necessary to conduct the reaction at reflux for 2 h (Table 1, entry 1).

Table 1: Optimization of the reaction conditionsa,b.

[Graphic 1]
Entry Solvent Temperature (°C) Time (h) Yield of 3a (%)c
1 THF reflux 2 75
2 1,4-dioxane 65 2 67
3 CH3CN 65 2 70
4 DMF 65 2 61
5 MeOH reflux 2 NRd
6 EtOH 65 2 NRd
7 CH2Cl2 reflux 12 42
8 THF rt 12 44
9 CH3CN rt 12 47

aAbbreviations: NR – no reaction, rt – room temperature. bReaction conditions: 1 (1 equiv), 2a (1 equiv), solvent. cIsolated yield. dAzomethine ylide 1 exclusively underwent deprotonation in protic solvents.

With the optimized conditions in hand, 1,2-diphenylcyclopropenes 2bi differently substituted at the C3 position were tested as dipolarophiles to evaluate the effect of C3-substituents on the 1,3-DC reaction (Scheme 3). 1,2-Diphenylcyclopropene (2b) [31] smoothly underwent the cycloaddition reaction to azomethine ylide 1 to form bis-spiro 3-azabicyclo[3.1.0]hexane 3b in 78% yield (Scheme 3). Remarkably, the structure of cycloadduct 3b was additionally verified by X-ray analysis (see Supporting Information File 1, Figure S26 and Table S1). The reaction of 3-ethyl-substituted cyclopropene 2c [32] with PRP (1) also proceeded with full diastereofacial selectivity, giving the corresponding cycloadduct 3c in acceptable yield (72%). On treatment of cyclopropenes 2d [32], 2e [33] both containing multiple bonds with stable azomethine ylide 1, the 1,3-DC reaction occurred in a highly chemo- and diastereoselective manner and brought about the formation of 3-azabicyclo[3.1.0]hexane cycloadducts 3d and 3e in 69% and 91% yields, correspondingly (Scheme 3). As shown by these experiments, the cyclopropene double bond demonstrates higher reactivity compared to exocyclic double or triple bonds due to ring strain. It is noteworthy that we managed to determine the relative configuration of cycloadduct 3e by carrying out the corresponding X-ray structural analysis (Figure 1 and Table S2 in Supporting Information File 1). As anticipated, the azomethine ylide 1 cycloaddition to cyclopropene 2e led to the formation of the diastereomer with all three substituents (at the cyclopropane ring) oriented in the same direction.


Scheme 3: Synthesis of meso-3'-azadispiro[indene-2,2'-bicyclo[3.1.0]hexane-4',2''-indene] derivatives 3bg via 1,3-DC reactions of N-protonated Ruhemann's purple (1) with 3-substituted 1,2-diphenylcyclopropenes 2bg.


Figure 1: ORTEP representation of the molecular structure of 3e.

A notable substituent effect on the reactivity of cyclopropene dipolarophiles 2 was observed for the reactions between PRP (1) and derivatives of 2,3-diphenylcycloprop-2-ene-1-carboxylic acid 2fi. The amide 2f and nitrile 2g [34] were found to be less active dipolarophiles towards the azomethine ylide 1 than hydrocarbons 2ae. As a result, the corresponding cycloadducts 3f and 3g were obtained in moderate yields (58% and 55%, respectively) as single diastereomers (Scheme 3). Notably, it took longer reaction times (6 h) for achieving full consumption of 1 while significant amounts of alkenes 2f and 2g remained unreacted. In turn, both the ester 2h and acid 2i [35] proved to be totally unreactive towards PRP (1) under the optimized reaction conditions (THF, reflux). Based on these observations, we concluded that electronic properties of the substituent at the C3 position of a cyclopropene ring have a major impact on the reactivity. A more detailed discussion of this issue is presented in the section devoted to the DFT (density functional theory) computational study. The constitution of compounds 3bg was established by analyzing 1H and 13C NMR spectra. In line with the structure of meso compound 3e, the relative configuration, that is shown in Scheme 3, was assigned to cycloadducts 3a, 3c, 3d, 3f, and 3g. Thus, the cycloadditions of 3-substituted-1,2-diphenylcyclopropenes 2a, 2cg to azomethine ylide 1 were found to exclusively proceed by a pathway in which PRP (1) approaches the cyclopropenes 2 from the more sterically available side of 2a, 2cg (from the side containing a hydrogen substituent).

Next, unsymmetrically 3,3-disubstituted cyclopropenes 2jl were studied in the reaction with 1 to determine whether these 1,3-DC reactions would similarly proceed with high diastereofacial selectivity. The reaction of 3-methyl-3-phenylcyclopropene (2j) [36] with ylide 1 resulted in the formation of a 1:1 adduct and the cycloaddition with cyclopropene 2j occurred at slower rates than with 3-monosubstiuted 1,2-diphenylcyclopropenes. Approximately 6 h at reflux were needed to complete the reaction. The analysis of the 1H NMR spectrum of the crude mixture indicated that the formation of cycloadduct 4 occurred with a high level of diastereocontrol (Scheme 4). The other observed signals were at the limit of detection and whether they actually correspond to a minor diastereomer could not be ascertained unambiguously. The major product 4 was purified by recrystallization of the crude mixture from MeOH and obtained in 62% yield. The structure and relative configuration of cycloadduct 4 were unequivocally established on the basis of its two-dimensional (2D) NMR spectrum (1H,1H nuclear Overhauser effect spectroscopy (NOESY), see Supporting Information File 1, Figures S17–S19). This experiment showed that azomethine ylide 1 also predominantly undergoes cycloaddition to the prochiral cyclopropene 2j from the less sterically hindered side. This is consistent with previous experiments in which 3-substituted-1,2-diphenylcyclopropenes were utilized as dipolarophiles.


Scheme 4: The reaction of protonated Ruhemann's purple (1) with 3-methyl-3-phenylcyclopropene (2j).

Regretfully, neither methyl 1-methylcycloprop-2-enecarboxylate (2k) [37] nor 3-methyl-1,2,3-triphenylcycloprop-1-ene (2l) [30] reacted with ylide 1 in THF at reflux (Scheme 5). The inactivity of substrate 2k appears to be caused by electronic effects while the tetrasubstituted cyclopropene 2l seems to be too bulky to undergo a cycloaddition reaction with 1.


Scheme 5: Attempts to carry out the cycloaddition reactions between 3,3-disubstituted cyclopropenes 2k,l and azomethine ylide 1.

To the best of our knowledge, PRP (1) is one of the few stable azomethine ylides. This fact prompted us to test this azomethine ylide 1 as an effective trap for capturing unstable cyclopropenes (Scheme 6). At first, 1-chloro-2-phenylcyclopropene (2m) [38], that is stable only in a solution, was studied as a dipolarophile in reaction with 1. Thus, a freshly prepared carbon tetrachloride solution of 1-chloro-2-phenylcyclopropene (2m) was treated with PRP (1) in THF medium. To avoid cyclopropene decomposition, the reaction was carried out at room temperature with stirring. Twelve hours later, we were pleased to note that azomethine ylide 1 was fully consumed as a result of the cycloaddition reaction with cyclopropene 2m. The corresponding bis-spiro compound 5a was readily isolated in 82% yield after recrystallization from ethanol. Similarly, we succeeded in carrying out reactions between unsymmetrically 1,2-disubstituted cyclopropenes 2n, 2o and PRP (1). 1-Methyl-2-phenylcyclopropene (2n) and 1-phenyl-2-(trimethylsilyl)cyclopropene (2o) synthesized from reliable precursor 1,1,2-tribromo-2-phenylcyclopropane [39,40] were immediately added to a THF solution of nitrogen ylide 1 at room temperature, resulting in the formation of cycloadducts 5b and 5c in satisfactory yields (74% and 79%, respectively). Finally, we attempted to trap parent cyclopropene (2p) as a cycloadduct with PRP (1). In one of our previous studies [22], we have demonstrated that stable ninhydrin-derived azomethine ylide – DHPO (2-(3,4-dihydro-2H-pyrrolium-1-yl)-1-oxo-1H-inden-3-olate) is an excellent chemical trap for detecting parent cyclopropene. Having carried out the experiment with ylide 1 as a trap in a similar fashion, we found that PRP (1) is unreactive towards parent cyclopropene (2p) under these conditions. Cyclopropene 2p, generated in situ from allyl chloride [41] and driven into a chilled tube containing PRP (1), appeared to undergo free-radical polymerization when increasing the temperature to 25 °C. Despite this failed experiment, in general, PRP (1) has established itself as a highly reactive 1,3-dipole towards cyclopropene dipolarophiles 2.


Scheme 6: The reactions of protonated Ruhemann's purple (1) with unstable cyclopropenes 2mp.

In this study, we did not confine ourselves exclusively to carrying out laboratory experiments. We turned to DFT calculations (M11 density functional theory) [42-46] to interpret the experimental results. At the beginning of the computational study, we evaluated the relative stability of prototropic tautomers which are formed during protonation of Ruhemann's purple. Although there was conclusive evidence of the structure of PRP (1) in the Grigg's study (proven by X-ray analysis) [27], we aimed to establish a stability order of tautomers on the basis of calculated relative values of the Gibbs free energy. Upon treatment of Ruhemann's purple with hydrochloric acid, three tautomers 1, 1', and 1'' may be theoretically formed, i.e., protonation could occur at the nitrogen atom, the oxygen atom or the carbon atom, respectively (Scheme 7). According to literature data [27], the nitrogen atom of Ruhemann's purple is considered to be the most basic site in the molecule. We carried out full geometry optimization of all possible tautomers 1, 1', 1'', and aza-allylic anion Ruhemann's purple to calculate the Gibbs free energy change for the corresponding acid–base reactions (Scheme 7). As expected, the calculation data showed that the betaine form 1 is the most thermodynamically stable of all tautomers (ΔG = −4.9 kcal/mol). It is also not surprising that the O-protonated form 1', which is both a ketone and an enol, is found to be the most unfavorable (ΔG = 10.8 kcal/mol). In contrast to the O-protonated tautomer 1', both the N-protonated 1 and C-protonated 1'' tautomers do not contain an enol functional group. It is therefore likely that the lowest stability of the O-protonated form 1' compared to tautomers 1 and 1'' is due to the favorability of the C=O bond over the C=C bond.


Scheme 7: The acid–base reaction of Ruhemann's purple with hydrochloric acid and relative Gibbs free energy change (ΔG, kcal/mol) for acid–base reactions resulting in the formation of three protonated forms of Ruhemann's purple.

Next, using the equations recommended by Parr [47] and Domingo [48], we calculated global electrophilicity indexes (GEI, ω) for both PRP (1) and cyclopropenes 2 to determine which type of electron demand takes place during cycloaddition reactions. The global value of the electrophilicity index for azomethine ylide 1 (1.29 eV) revealed that this compound displays a moderate electrophilicity (Table 2, entry 1). By comparing the global electrophilicity indexes of reactants 1, 2ap (Table 2, entries 1–17), we concluded that the cyclopropene cycloadditions to PRP (1) appear to be inverse electron demand (IED) reactions, i.e., they can be considered as the interaction of the lowest unoccupied molecular orbital (LUMO) of the azomethine ylide 1 with the highest occupied molecular orbital (HOMO) of cyclopropenes 2. It follows that the greater the nucleophilicity of the cyclopropene substrate 2 is, the higher is its reactivity towards the 1,3-dipole 1. In fact, during the experiments, 1,2-diphenylcyclopropenes 2ae containing either an alkyl or an alkenyl substituent at the C3 position were found to demonstrate a high reactivity in the 1,3-DC reactions with PRP (1). In contrast to substrates 2ae, cyclopropene dipolarophiles 2fi bearing electron-withdrawing groups are less nucleophilic. This is evidenced both by fairly large ω values of these cyclopropenes (Table 2, entries 7–10) and by the reaction outcomes. Introducing electron-withdrawing groups at the C3 position of the 1,2-diphenylcyclopropene framework instead of alkyl or alkenyl substituents results in a decrease in the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energies of such cyclopropenes, thereby increasing the HOMOcyclopropene–LUMOylide energy gap (Table 2, entries 7–10). In practice, this has an impact on the cyclopropene reactivity towards azomethine ylide 1. For example, the reactions of 2,3-diphenylcyclopropene-1-carboxylic acid derivatives 2fi with PRP (1) either bring about the formation of the corresponding cycloadducts 3 in poor yields or result in the formation of complex mixtures. Thus, having carried out an analysis of global reactivity descriptors, we have found a correlation between the structure of cyclopropene substrates 2 and their reactivity towards PRP (1).

Table 2: FMO energies (a.u.), electronic chemical potential (μ, eV), chemical hardness (η, eV), and global electrophilicity index (ω, eV) for PRP (1) and cyclopropenes 2.a,b,c.

Entry Compd HOMO LUMO μ η ω NEDc IEDc ED
1 1 −0.3110 −0.0207 −4.51 7.90 1.29
2 2a −0.2867 0.0614 −3.06 9.47 0.50 10.13 7.24 IED
3 2b −0.2815 0.0637 −2.96 9.39 0.47 10.19 7.10 IED
4 2c −0.2802 0.0650 −2.93 9.39 0.46 10.23 7.06 IED
5 2d −0.2860 0.0612 −3.06 9.45 0.50 10.13 7.22 IED
6 2e −0.2910 0.0591 −3.16 9.53 0.52 10.07 7.35 IED
7 2f −0.2925 0.0557 −3.22 9.47 0.55 9.98 7.39 IED
8 2g −0.2996 0.0526 −3.36 9.58 0.59 9.89 7.59 IED
9 2h −0.2935 0.0581 −3.20 9.56 0.54 10.04 7.42 IED
10 2i −0.2947 0.0571 −3.23 9.57 0.55 10.02 7.45 IED
11 2j −0.3179 0.1212 −2.68 11.95 0.30 11.76 8.09 IED
12 2k −0.3805 0.1350 −3.34 14.03 0.40 12.14 9.79 IED
13 2l −0.2853 0.0618 −3.04 9.44 0.49 10.14 7.20 IED
14 2m −0.3107 0.0820 −3.11 10.68 0.45 10.69 7.89 IED
15 2n −0.2972 0.0944 −2.76 10.66 0.36 11.03 7.52 IED
16 2o −0.3019 0.0824 −2.99 10.46 0.43 10.70 7.65 IED
17 2p −0.3574 0.1556 −2.74 13.96 0.27 12.70 9.16 IED

aAbbreviations: HOMO – highest occupied molecular orbital, LUMO – lowest unoccupied molecular orbital, NED – normal electron demand, IED – inverse electron demand, ED – electron demand. bFMO energy values (eV) were computed by using HF/6-311g single point calculation on the M11/cc-pVDZ optimized geometries. cEnergy gaps for both possible HOMO–LUMO interactions between PRP (1) and cyclopropenes 2 are given in eV.

In the next step, we concentrated on studying the mechanism of the 1,3-DC reactions between PRP (1) and cyclopropenes 2. At first, it was planned to explore the mechanism of 1,3-DC reactions involving cyclopropenes substituted at the C3 position. It is precisely for these cyclopropenes that two diastereomeric cycloadducts can be theoretically obtained. In fact, only one of the two epimers is formed, i.e., the cycloaddition reaction proceeds with complete diastereofacial stereoselectivity. Our goal was to identify the nature of transition states associated with two theoretically possible diastereomers and to find out if the experimentally observed diastereomer is more kinetically favorable than the opposite one.

The reaction between 3-methyl-3-phenylcyclopropene (2j) and PRP (1) that led to the stereoselective formation of cycloadduct 4 was studied in detail for this purpose (Scheme 8). In analyzing the potential energy surface (PES) of this cycloaddition reaction using the M11 functional together with the cc-pVDZ basis set, we could locate the four transition state structures ΤS-4-endo, TS-4-exo, TS-4'-endo, and TS-4'-exo that were associated with a concerted mechanism of the 1,3-dipolar cycloaddition. Two of them, namely TS-4-endo and TS-4-exo correspond to the two nitrogen invertomers 4-endo and 4-exo of cycloadduct 4 while the other two TS-4'-endo and TS-4'-exo associate with nitrogen invertomers 4'-endo and 4'-exo of cycloadduct 4', respectively. Next, we determined the Gibbs energy of activation for nitrogen inversion in cycloadducts 4 and 4' to figure out if the invertomers undergo rapid interconversion. Having carried out full geometry optimization of both transition states TS-4-NI and TS-4'-NI corresponding to pyramidal inversion in cycloadducts 4 and 4' and two pairs of invertomers, we have found that each of the diastereomers 4 and 4' is a mixture of two easily interconverting invertomers (ΔG4-exo->4-endo = 1.2 kcal/mol and ΔG4-endo->4-exo = 2.1 kcal/mol, ΔG4'-exo->4'-endo = 0.2 kcal/mol, and ΔG4'-endo->4'-exo = 3.8 kcal/mol). Since rapid interconversion takes place, it does not matter via which transition state (ΤS-4-endo or ΤS-4-exo) cycloadduct 4 is formed. The same pattern is consistent with cycloadduct 4'. It should be mentioned that the endo approaches (ΔG1+2j->4-endo = 19.3 kcal/mol and ΔG1+2j->4'-endo = 20.3 kcal/mol) are much more profitable than the exo ones (ΔG1+2j->4-exo = 23.4 kcal/mol and ΔG1+2j->4'-exo = 24.8 kcal/mol). It follows that cycloadducts 4 and 4' appear to be formed via ΤS-4-endo and TS-4'-endo, respectively. In turn, the Gibbs energies of activation corresponding to the endo approaches indicate that cycloadduct 4G1+2j->4-endo = 19.3 kcal/mol) is more kinetically favorable than its epimer 4'G1+2j->4'-endo = 20.3 kcal/mol). The small free energy difference (1.0 kcal/mol) between the two competing pathways is in good agreement with the experimental data and explains the reason why the cycloaddition reaction does not result in the exclusive formation of 4.


Scheme 8: Plausible mechanism of the 1,3-DC reaction of protonated Ruhemann's purple (1) with 3-methyl-3-phenylcyclopropene (2j) and corresponding DFT calculations (relative Gibbs free energy change between reagents, transition states and possible products are given in kcal/mol).

Additionally, the reaction of PRP (1) with 1-chloro-2-phenylcyclopropene (2m) was investigated to calculate the Gibbs energy of activation for this cycloaddition reaction and to compare this value with the Gibbs free energy barrier calculated for the reaction between 1 and 2j (Scheme 9). The azomethine ylide 1 cycloaddition to the chloro-substituted cyclopropene 2m was also found to proceed by a one-step mechanism via two transition states TS-5a-endo and TS-5a-exo that bring about invertomers 5a-endo and 5a-exo of cycloadduct 5a, respectively. According to the values of the Gibbs energy of activation, the endo cycloaddition (ΔG = 12.2 kcal/mol) significantly prevails over the exo one (ΔG = 14.5 kcal/mol). When comparing the values of the Gibbs energies of activation calculated for the above-mentioned reactions, it was established that the reaction involving 1 and 2mG = 12.2 kcal/mol) should proceed significantly faster than the cycloaddition between 1 and 2jG = 19.3 kcal/mol). Thus, the calculation data are in full accordance with the experimental results, taking into consideration the fact that the reaction between 1 and 2m smoothly occurs at room temperature while the reaction involving 3-methyl-3-phenylcyclopropene (2j) requires harsher conditions.


Scheme 9: Plausible mechanism of the 1,3-DC reaction of protonated Ruhemann's purple (1) with 1-chloro-2-phenylcyclopropene (2m) and corresponding DFT calculations (relative Gibbs free energy change between reagents, transition states and possible invertomers are given in kcal/mol).


In conclusion, we have developed a convenient and diastereoselective approach for the synthesis of bis-spirocyclic derivatives of 3-azabicyclo[3.1.0]hexane through cycloaddition reactions of a stable azomethine ylide – protonated Ruhemann's purple to cyclopropenes. The cycloaddition reaction is compatible with a broad scope of cyclopropenes. DFT calculations revealed that the cycloaddition reactions are under kinetic control and belong to the class of inverse electron demand 1,3-DC reactions. We believe that the outcome of this work will serve as a basis for developing synthetic approaches to other bis-spirocyclic derivatives of 3-azabicyclo[3.1.0]hexane via cycloadditions of tetrasubstituted azomethine ylides with cyclopropenes.

Supporting Information

Supporting Information File 1: Experimental details for the synthesis and characterization of all compounds, copies of 1H NMR and 13C NMR spectra, X-ray data and details of calculations.
Format: PDF Size: 4.4 MB Download


This research made use of resources from the X-ray Diffraction Center, Computer Center, Center for Magnetic Resonance, Educational Resource Center of Chemistry, Center for Chemical Analysis and Materials of Saint-Petersburg State University.


We gratefully acknowledge the financial support from the Russian Foundation for Basic Research (Project no. 20-33-90325). This study was supported by the Ministry of Science and Higher Education of the Russian Federation (0785.00.Х6019). V.M.B. is grateful to the Ministry of Education and Science of the Russian Federation (0791-2020-0006) for the financial support.


  1. Saraswat, P.; Jeyabalan, G.; Hassan, M. Z.; Rahman, M. U.; Nyola, N. K. Synth. Commun. 2016, 46, 1643–1664. doi:10.1080/00397911.2016.1211704
    Return to citation in text: [1]
  2. Khanna, P.; Panda, S. S.; Khanna, L.; Jain, S. C. Mini-Rev. Org. Chem. 2014, 11, 73–86. doi:10.2174/1570193x1101140402101831
    Return to citation in text: [1]
  3. Acosta-Quiroga, K.; Rojas-Peña, C.; Nerio, L. S.; Gutiérrez, M.; Polo-Cuadrado, E. RSC Adv. 2021, 11, 21926–21954. doi:10.1039/d1ra01170g
    Return to citation in text: [1]
  4. Ramdani, L. H.; Talhi, O.; Taibi, N.; Delort, L.; Decombat, C.; Silva, A.; Bachari, K.; Vasson, M. P.; Caldefie-Chezet, F. Anticancer Res. 2016, 36, 6399–6408. doi:10.21873/anticanres.11237
    Return to citation in text: [1]
  5. Goyard, D.; Kónya, B.; Chajistamatiou, A. S.; Chrysina, E. D.; Leroy, J.; Balzarin, S.; Tournier, M.; Tousch, D.; Petit, P.; Duret, C.; Maurel, P.; Somsák, L.; Docsa, T.; Gergely, P.; Praly, J.-P.; Azay-Milhau, J.; Vidal, S. Eur. J. Med. Chem. 2016, 108, 444–454. doi:10.1016/j.ejmech.2015.12.004
    Return to citation in text: [1]
  6. Chande, M. S.; Verma, R. S.; Barve, P. A.; Khanwelkar, R. R.; Vaidya, R. B.; Ajaikumar, K. B. Eur. J. Med. Chem. 2005, 40, 1143–1148. doi:10.1016/j.ejmech.2005.06.004
    Return to citation in text: [1]
  7. Li, J.-J.; Liang, X.-M.; Jin, S.-H.; Zhang, J.-J.; Yuan, H.-Z.; Qi, S.-H.; Chen, F.-H.; Wang, D.-Q. J. Agric. Food Chem. 2010, 58, 2659–2663. doi:10.1021/jf903665f
    Return to citation in text: [1]
  8. Saragi, T. P. I.; Spehr, T.; Siebert, A.; Fuhrmann-Lieker, T.; Salbeck, J. Chem. Rev. 2007, 107, 1011–1065. doi:10.1021/cr0501341
    Return to citation in text: [1]
  9. Liu, H.-W.; Walsh, C. T. Biochemistry of the cyclopropyl group. In The Chemistry of the Cyclopropyl Group; Rappoport, Z., Ed.; John Wiley & Sons: Chichester, UK, 1987; pp 959–1025. doi:10.1002/0470023449.ch16
    Return to citation in text: [1]
  10. Netz, N.; Opatz, T. J. Org. Chem. 2016, 81, 1723–1730. doi:10.1021/acs.joc.5b02815
    Return to citation in text: [1]
  11. Krow, G. R.; Cannon, K. C. Org. Prep. Proced. Int. 2000, 32, 103–122. doi:10.1080/00304940009356278
    Return to citation in text: [1]
  12. Micheli, F.; Cavanni, P.; Arban, R.; Benedetti, R.; Bertani, B.; Bettati, M.; Bettelini, L.; Bonanomi, G.; Braggio, S.; Checchia, A.; Davalli, S.; Di Fabio, R.; Fazzolari, E.; Fontana, S.; Marchioro, C.; Minick, D.; Negri, M.; Oliosi, B.; Read, K. D.; Sartori, I.; Tedesco, G.; Tarsi, L.; Terreni, S.; Visentini, F.; Zocchi, A.; Zonzini, L. J. Med. Chem. 2010, 53, 2534–2551. doi:10.1021/jm901818u
    Return to citation in text: [1]
  13. Moffat, D.; Patel, S.; Day, F.; Belfield, A.; Donald, A.; Rowlands, M.; Wibawa, J.; Brotherton, D.; Stimson, L.; Clark, V.; Owen, J.; Bawden, L.; Box, G.; Bone, E.; Mortenson, P.; Hardcastle, A.; van Meurs, S.; Eccles, S.; Raynaud, F.; Aherne, W. J. Med. Chem. 2010, 53, 8663–8678. doi:10.1021/jm101177s
    Return to citation in text: [1] [2]
  14. Lunn, G.; Banks, B. J.; Crook, R.; Feeder, N.; Pettman, A.; Sabnis, Y. Bioorg. Med. Chem. Lett. 2011, 21, 4608–4611. doi:10.1016/j.bmcl.2011.05.132
    Return to citation in text: [1]
  15. McHardy, S. F.; Heck, S. D.; Guediche, S.; Kalman, M.; Allen, M. P.; Tu, M.; Bryce, D. K.; Schmidt, A. W.; Vanase-Frawley, M.; Callegari, E.; Doran, S.; Grahame, N. J.; McLean, S.; Liras, S. Med. Chem. Commun. 2011, 2, 1001–1005. doi:10.1039/c1md00164g
    Return to citation in text: [1] [2]
  16. Li, T.; Liang, J.; Ambrogelly, A.; Brennan, T.; Gloor, G.; Huisman, G.; Lalonde, J.; Lekhal, A.; Mijts, B.; Muley, S.; Newman, L.; Tobin, M.; Wong, G.; Zaks, A.; Zhang, X. J. Am. Chem. Soc. 2012, 134, 6467–6472. doi:10.1021/ja3010495
    Return to citation in text: [1]
  17. Topczewski, J. J.; Cabrera, P. J.; Saper, N. I.; Sanford, M. S. Nature 2016, 531, 220–224. doi:10.1038/nature16957
    Return to citation in text: [1]
  18. Komine, T.; Kojima, A.; Asahina, Y.; Saito, T.; Takano, H.; Shibue, T.; Fukuda, Y. J. Med. Chem. 2008, 51, 6558–6562. doi:10.1021/jm800800c
    Return to citation in text: [1]
  19. Filatov, A. S.; Knyazev, N. A.; Molchanov, A. P.; Panikorovsky, T. L.; Kostikov, R. R.; Larina, A. G.; Boitsov, V. M.; Stepakov, A. V. J. Org. Chem. 2017, 82, 959–975. doi:10.1021/acs.joc.6b02505
    Return to citation in text: [1] [2] [3]
  20. Filatov, A. S.; Knyazev, N. A.; Ryazantsev, M. N.; Suslonov, V. V.; Larina, A. G.; Molchanov, A. P.; Kostikov, R. R.; Boitsov, V. M.; Stepakov, A. V. Org. Chem. Front. 2018, 5, 595–605. doi:10.1039/c7qo00888k
    Return to citation in text: [1] [2] [3]
  21. Filatov, A. S.; Knyazev, N. A.; Shmakov, S. V.; Bogdanov, A. A.; Ryazantsev, M. N.; Shtyrov, A. A.; Starova, G. L.; Molchanov, A. P.; Larina, A. G.; Boitsov, V. M.; Stepakov, A. V. Synthesis 2019, 51, 713–729. doi:10.1055/s-0037-1611059
    Return to citation in text: [1] [2] [3]
  22. Filatov, A. S.; Wang, S.; Khoroshilova, O. V.; Lozovskiy, S. V.; Larina, A. G.; Boitsov, V. M.; Stepakov, A. V. J. Org. Chem. 2019, 84, 7017–7036. doi:10.1021/acs.joc.9b00753
    Return to citation in text: [1] [2] [3] [4] [5]
  23. Wang, S.; Filatov, A. S.; Lozovskiy, S. V.; Shmakov, S. V.; Khoroshilova, O. V.; Larina, A. G.; Selivanov, S. I.; Boitsov, V. M.; Stepakov, A. V. Synthesis 2021, 53, 2114–2132. doi:10.1055/a-1360-9716
    Return to citation in text: [1] [2] [3] [4]
  24. Filatov, A. S.; Selivanov, S. I.; Shmakov, S. V.; Larina, A. G.; Boitsov, V. M.; Stepakov, A. V. Synthesis 2022, 54, 1803–1816. doi:10.1055/a-1700-3115
    Return to citation in text: [1] [2] [3]
  25. Deng, H.; Yang, W.-L.; Tian, F.; Tang, W.; Deng, W.-P. Org. Lett. 2018, 20, 4121–4125. doi:10.1021/acs.orglett.8b01686
    Return to citation in text: [1]
  26. Lahtigui, O.; Forster, D.; Duchemin, C.; Cramer, N. ACS Catal. 2022, 12, 6209–6215. doi:10.1021/acscatal.2c01827
    Return to citation in text: [1]
  27. Grigg, R.; Malone, J. F.; Mongkolaussavaratana, T.; Thianpatanagul, S. J. Chem. Soc., Chem. Commun. 1986, 421–422. doi:10.1039/c39860000421
    Return to citation in text: [1] [2] [3] [4]
  28. Moore, S.; Stein, W. H. J. Biol. Chem. 1948, 176, 367–388. doi:10.1016/s0021-9258(18)51034-6
    Return to citation in text: [1]
  29. Wigfield, D. C.; Buchanan, G. W.; Croteau, S. M. Can. J. Chem. 1980, 58, 201–205. doi:10.1139/v80-032
    Return to citation in text: [1]
  30. Breslow, R.; Dowd, P. J. Am. Chem. Soc. 1963, 85, 2729–2735. doi:10.1021/ja00901a012
    Return to citation in text: [1] [2]
  31. Longone, D. T.; Stehouwer, D. M. Tetrahedron Lett. 1970, 11, 1017–1019. doi:10.1016/s0040-4039(01)97895-5
    Return to citation in text: [1]
  32. Padwa, A.; Blacklock, T. J.; Getman, D.; Hatanaka, N.; Loza, R. J. Org. Chem. 1978, 43, 1481–1492. doi:10.1021/jo00402a002
    Return to citation in text: [1] [2]
  33. Gilbertson, R. D.; Weakley, T. J. R.; Haley, M. M. J. Org. Chem. 2000, 65, 1422–1430. doi:10.1021/jo9915372
    Return to citation in text: [1]
  34. White, E. H.; Winter, R. E. K.; Graeve, R.; Zirngibl, U.; Friend, E. W.; Maskill, H.; Mende, U.; Kreiling, G.; Reisenauer, H. P.; Maier, G. Chem. Ber. 1981, 114, 3906–3915. doi:10.1002/cber.19811141213
    Return to citation in text: [1]
  35. Jones, W. M.; Stowe, M. E.; Wells, E. E., Jr.; Lester, E. W. J. Am. Chem. Soc. 1968, 90, 1849–1859. doi:10.1021/ja01009a030
    Return to citation in text: [1]
  36. Rubin, M.; Gevorgyan, V. Synthesis 2004, 796–800. doi:10.1055/s-2003-44368
    Return to citation in text: [1]
  37. Alnasleh, B. K.; Sherrill, W. M.; Rubin, M. Org. Lett. 2008, 10, 3231–3234. doi:10.1021/ol8011138
    Return to citation in text: [1]
  38. Henseling, K.-O.; Weyerstahl, P. Chem. Ber. 1975, 108, 2803–2808. doi:10.1002/cber.19751080839
    Return to citation in text: [1]
  39. Lin, H.-C.; Tsai, R.-T.; Wu, H.-P.; Lee, H.-Y.; Lee, G.-A. Tetrahedron 2016, 72, 184–191. doi:10.1016/j.tet.2015.11.024
    Return to citation in text: [1]
  40. Lee, G.-A.; Chang, C.-Y. Tetrahedron Lett. 1998, 39, 3013–3016. doi:10.1016/s0040-4039(98)00447-x
    Return to citation in text: [1]
  41. Binger, P.; Wedemann, P.; Goddard, R.; Brinker, U. H. J. Org. Chem. 1996, 61, 6462–6464. doi:10.1021/jo960728r
    Return to citation in text: [1]
  42. Peverati, R.; Truhlar, D. G. J. Phys. Chem. Lett. 2011, 2, 2810–2817. doi:10.1021/jz201170d
    Return to citation in text: [1]
  43. Dunning, T. H., Jr. J. Chem. Phys. 1989, 90, 1007–1023. doi:10.1063/1.456153
    Return to citation in text: [1]
  44. Cossi, M.; Rega, N.; Scalmani, G.; Barone, V. J. Comput. Chem. 2003, 24, 669–681. doi:10.1002/jcc.10189
    Return to citation in text: [1]
  45. Schlegel, H. B. J. Comput. Chem. 1982, 3, 214–218. doi:10.1002/jcc.540030212
    Return to citation in text: [1]
  46. Fukui, K. J. Phys. Chem. 1970, 74, 4161–4163. doi:10.1021/j100717a029
    Return to citation in text: [1]
  47. Parr, R. G.; Szentpály, L. v.; Liu, S. J. Am. Chem. Soc. 1999, 121, 1922–1924. doi:10.1021/ja983494x
    Return to citation in text: [1]
  48. Pérez, P.; Domingo, L. R.; Aurell, M. J.; Contreras, R. Tetrahedron 2003, 59, 3117–3125. doi:10.1016/s0040-4020(03)00374-0
    Return to citation in text: [1]

Interesting articles

Alexei N. Izmest’ev, Galina A. Gazieva, Natalya V. Sigay, Sergei A. Serkov, Valentina A. Karnoukhova, Vadim V. Kachala, Alexander S. Shashkov, Igor E. Zanin, Angelina N. Kravchenko and Nina N. Makhova

M. Fernanda N. N. Carvalho, M. João Ferreira, Ana S. O. Knittel, Maria da Conceição Oliveira, João Costa Pessoa, Rudolf Herrmann and Gabriele Wagner

© 2022 Filatov et al.; licensee Beilstein-Institut.
This is an open access article licensed under the terms of the Beilstein-Institut Open Access License Agreement (, which is identical to the Creative Commons Attribution 4.0 International License ( The reuse of material under this license requires that the author(s), source and license are credited. Third-party material in this article could be subject to other licenses (typically indicated in the credit line), and in this case, users are required to obtain permission from the license holder to reuse the material.

Back to Article List

Other Beilstein-Institut Open Science Activities

Keep Informed

RSS Feed

Subscribe to our Latest Articles RSS Feed.


Follow the Beilstein-Institut


Twitter: @BeilsteinInst