Activating harmful small molecules under mild conditions: theoretical insights into cinchonine-based valorization of CO₂, CS₂, and COS

Abstract

DFT calculations have been employed to deeply investigate the mechanism of CO₂ cycloaddition to aziridines catalyzed by cinchonine hydrochloride salt, forming oxazolidin-2-ones under ambient conditions (room temperature, 0.1 MPa CO₂). Computed energy barriers align with experimental observations and support a dual activation mechanism involving hydrogen bonding and nucleophilic attack at the aziridine carbon atom. The theoretical study also accounts for the observed regioselectivity, rationalizing the preference for nucleophilic attack at the more substituted aziridine carbon atom. Consistent with experimental findings, the calculations reveal that the reaction efficiency is influenced by the nature of the substituent at the aziridine nitrogen atom, explaining the lack of reactivity observed with N-aryl aziridines due to steric and electronic factors that hamper the reaction. Furthermore, the DFT study suggests that COS and CS₂ can be activated for analogous cycloaddition reactions. Although these transformations involve higher energy barriers compared to CO₂ cycloaddition, the formation of oxazolidin-2-thiones and thiazolidin-2-thiones is predicted to be feasible under slightly elevated temperatures (for CS₂) or near-ambient conditions (for COS). These findings highlight the potential of cinchonine hydrochloride salt as an efficient, biocompatible and cost-effective catalyst for the sustainable valorization of small harmful molecules under mild conditions.

---

Introduction

The activation of small molecules has long been a central theme in chemical research, driven by the challenge of breaking strong bonds in simple, thermodynamically stable species. Historically, much of this work has focused on inert molecules such as N₂, O₂ and CH₄, which are starting materials in processes like nitrogen fixation, oxidation reactions, and hydrocarbon functionalization.^1,2 During the last decades, scientific interest has progressively moved toward carbon dioxide. Although CO₂ is similarly unreactive, it presents a distinctive combination of scientific and environmental urgency. Its massive atmospheric abundance, mainly due to anthropogenic emissions, and its well-established role in global warming have made its activation a priority in the context of sustainable chemistry. As a result, CO₂ is now viewed not only as a waste product but also as a potential carbon feedstock, whose efficient transformation could contribute to both climate mitigation and resource circularity.

Advances in catalytic^3–5 and electrochemical methods^6–8 have enabled the efficient conversion of CO₂ into a variety of useful chemicals⁹ and fuels¹⁰ such as carbon monoxide, formic acid, methanol, methane, and a range of value-added compounds. In light of the development of a circular and sustainable economy, the establishment of new and efficient processes for CO₂ valorization has become interesting for the synthesis of fine chemicals. One particularly intriguing reaction is the 100% atom-efficient cycloaddition of CO₂ into organic substrates, such as epoxides^11,12 and aziridines,^13–20 to produce high-value chemicals like cyclic carbonates and oxazolidin-2-ones, respectively, which have several applications including as pharmaceutical agents.^21–23 Classic CO₂ activation typically requires the co-presence of an electrophilic and a nucleophilic promoter. In the synthesis of oxazolidin-2-ones, the electrophile, generally a metal center, interacts with the nitrogen aziridine atom to render one of the two carbon atoms more susceptible to nucleophilic attack. The consequent cleavage of the C–N bond transfers electron density to the nitrogen atom, which in turn becomes capable of activating the CO₂ molecule (Scheme 1). The need for both electrophilic and nucleophilic functionalities led to the development of various binary and bifunctional metal-based catalysts for the CO₂ cycloaddition to aziridines.^4–10,13 More recently, metal-free systems,^{14,15,18,19,24} which integrate both functional components enabling efficient aziridine activation and ring-opening without the use of external metal catalysts, have also been extensively developed.

Scheme 1 General reaction mechanism of CO₂ cycloaddition to aziridines (left); CO₂, COS and CS₂ cycloaddition products to aziridines (right).

Although CO₂ has received widespread attention as a platform molecule for sustainable synthesis, its thio-analogues, carbon disulfide (CS₂) and carbonyl sulfide (COS), have remained largely overlooked.

Since the early 1970s, when the reactivity of CS₂ toward aziridines was first investigated,^25,26 research on COS and CS₂ has primarily focused on their interactions with metal complexes.^27,28 Only in recent years, the interest in these triatomic molecules has expanded, with growing attention to their cycloaddition reactions with epoxides and aziridines, particularly via metal–organic framework (MOF)-based catalytic systems in combination with nucleophilic co-catalysts.^29–32 The limited experimental examples reported for the CS₂ cycloaddition to aziridines rely on expensive binary systems involving amidato divalent lanthanide complexes, among which the most active is represented by the europium derivative used in combination with DBU (1,8-diazabicyclo[5.4.0]undec-7-ene).²⁹ Other examples include a cerium-based framework,³⁰ a porous 3D cobalt–organic framework assembled by [Co₁₅] and [Co₁₈] nanocages³¹ and a 3D material assembled by twisted cylindrical [Dy₂₄] cages.³² In all cases, TBAB (tetrabutylammonium bromide) was the co-catalyst; either the metal centre or the positive channel of the MOF-based material acts as a Lewis acid to activate the aziridine, while the external nucleophile promotes the ring-opening reaction enabling CS₂ insertion and product formation. To date, no examples of active bifunctional organocatalytic systems have been reported. The above-described behavior aligns with the structural and electronic similarities between COS, CS₂, and CO₂, highlighting their potential for analogous activation and transformation pathways.^27,29,33 It is important to note that CS₂ and COS not only cause environmental effects but can also contribute to various health disorders, particularly affecting the human reproductive and nervous systems.^34,35 Therefore, rather than eliminating them through traditional combustion methods, treating and reusing these compounds as carbon and sulfur resources offers an attractive alternative to reduce the presence of toxic sulfur-containing pollutants in the atmosphere. In this view, CS₂ and COS could serve as valuable C1-building blocks for the synthesis of sulfur-containing heterocycles,^36–38 such as thiazolidin-2-thiones and oxazolidin-2-thiones (Scheme 1), and other functionalized materials,³⁹ offering new opportunities in the field of small-molecule activation and sustainable sulfur chemistry.

It is important to underline that the activation of small molecules, such as CO₂, CS₂, and COS, should occur under mild reaction conditions to ensure a favorable balance between the energy required for the transformation and the amount of CO₂ effectively utilized. If harsh conditions are employed, such as high temperatures, elevated pressures, or energy-intensive inputs, the environmental benefit of using these molecules as feedstocks may be offset by additional CO₂ emissions generated during the process. For this reason, the development of catalytic systems capable of operating efficiently under ambient or near-ambient conditions is essential to maximize the net carbon benefit and make these transformations truly viable from both an energetic and environmental standpoint.

Given the growing demand for low-toxicity synthetic procedures, the use of eco-friendly bifunctional organocatalysts has become increasingly attractive. Recently, we reported a combined experimental and computational investigation of the catalytic properties of the metal-free bis-protonated porphyrin TPPH₄Cl₂ (TPP = dianion of tetraphenylporphyrin) in promoting the cycloaddition of CO₂ into N-alkyl aziridines to produce N-alkyl oxazolidin-2-ones at 100 °C and 1.2 MPa of CO₂ pressure.¹⁶ Computational analysis revealed that the reaction occurred, thanks to porphyrin/aziridine synergic CO₂ activation in which the protonated core of porphyrin acts, by establishing hydrogen bonding with the CO₂ oxygen atom, as an electrophilic center to facilitate the nucleophilic attack of the aziridine nitrogen atom to the CO₂ carbon atom. The study paved the way for developing other bifunctional organocatalytic systems, bearing the same nucleophilic moiety (chloride anion) but different electrophilic NH⁺-containing species, to promote the synthesis of N-alkyl oxazolidin-2-ones under milder experimental conditions.⁴⁰ In this context, some of us recently published a study on the catalytic activity of hydrochloride salts of DBU, quinine, and cinchonine,⁴⁰ which were active at room temperature and 0.1 MPa of CO₂ pressure. In view of the very good results achieved, we decided to investigate the electronic/energetic features of the mechanism of these reactions by using DFT calculations. In addition, the potential catalytic activity of cinchonine hydrochloride salt to promote the cycloaddition of other triatomic harmful molecules such as CS₂ and COS has been investigated in silico. The obtained computational results pave the way for developing future efficient and eco-compatible catalytic processes for the synthesis of fine chemicals by recycling waste.

Results and discussion

Cycloaddition of CO₂ to 1-butyl-2-phenylaziridine (1_butyl) promoted by cinchonine hydrochloride (3)

In light of remarkable data published for the synthesis N-alkyl oxazolidin-2-ones under mild experimental conditions,⁴⁰ detailed computational analysis was carried out on the model CO₂ cycloaddition to 1-butyl-2-phenylaziridine (1_butyl) yielding 3-butyl-5-phenyloxazolidin-2-ones (2A_butyl + 2B_butyl) (Scheme 2), catalyzed by the naturally derived cinchonine hydrochloride salt (3). Although particular emphasis was placed on elucidating the reaction mechanism catalyzed by 3, the mechanisms involving hydrochloride salts of DBU and quinine, organocatalysts 4 and 5 respectively, were also investigated and the collected data are reported in the SI. The structures of 3, 4, and 5 organocatalysts were optimized at the B97D-DFT level of theory,⁴¹ and are shown in Fig. 1. The solvent effects have been taken into account by using the CPCM model^42,43 for acetonitrile that is the solvent employed in experimental studies already published on the CO₂ cycloaddition to aziridines.⁴⁰ Additional methodological details are provided in the SI.

Scheme 2 Synthesis of 3-butyl-5-phenyloxazolidin-2-ones 2A_butyl and 2B_butyl by CO₂ cycloaddition to 1-butyl-2-phenylaziridine (1_butyl).

Fig. 1 Optimized structure of hydrochloride salts 3, 4 and 5. The hydrogen atoms were hidden for the sake of clarity, except for those linked to heteroatoms. Selected distances are given in Å.

According to the published experimental results,⁴⁰ 1-butyl-2-phenylaziridine (1_butyl) was chosen as the model substrate to investigate the catalytic mechanism of the reaction promoted by 3, and CH₃CN was the modeled reaction solvent.

Drawing from prior data on the TPPH₄Cl₂-mediated reaction,¹⁶ the first step of the computational analysis focused on the possible formation of adduct 6_butyl in which CO₂ is positioned between hydrochloride salt 3 and aziridine 1_butyl. As shown in Fig. 2, 6_butyl was computationally identified with a free energy cost of +15.0 kcal mol⁻¹ resulting from a balance between the favorable enthalpic contribution of −6.5 kcal mol⁻¹ and a severe unfavorable entropic contribution. The local electrophilicity index (ω_K⁺) of the two aziridine carbon atoms, namely C2 and C3, was estimated by using the method developed by Domingo et al.⁴⁴

Fig. 2 Optimized structure of adduct 6_butyl. The hydrogen atoms were hidden for the sake of clarity, except for those linked to a heteroatom. Selected distances are given in Å and O–C1–O angle in degrees (°).

The calculations revealed that the ω_K⁺ value for C2 is twofold higher than that of C3, suggesting a more favored nucleophilic attack at C2 rather than at the C3 carbon atom. This is in line with the experimentally observed reaction regioselectivity,⁴⁰ as 3-butyl-5-phenyloxazolidin-2-one (compound 2A_butyl in Scheme 2) was always detected as the major isomer. Consequently, the mechanism yielding isomer 2B_butyl, deriving from the nucleophilic attack on C3, was not further studied by DFT calculations.

In adduct 6_butyl, aziridine nitrogen atom N2 acts as a nucleophile toward CO₂, which loses its linearity with the O–C1–O angle reduced to 135°, as confirmed by the appearance of an IR-active vibration at 1759 cm⁻¹, associated with the asymmetric stretching of the C–O bonds. The chloride anion remains distant from the C2 center, with a Cl⋯C2 distance of 3.76 Å whose reduction to 2.80 Å revealed the presence of a transition state, designated TS_6-7butyl (Fig. 3).

Fig. 3 Optimized structure of transition state TS_6-7butyl and intermediate 7_butyl. The hydrogen atoms were hidden for the sake of clarity, except for those linked to a heteroatom. Selected distances are given in Å and O–C1–O angle in degrees (°).

In TS_6-7butyl, the approaching chloride anion weakens the N2–C2 bond, which is elongated by 0.15 Å compared to adduct 6_butyl. This facilitates the initial shift of electron density toward the aziridine nitrogen N2, enhancing its nucleophilicity toward CO₂. This effect is further supported by a 0.07 Å shortening of the N2⋯C1 distance and a more pronounced bending of the CO₂ moiety, with the O–C–O angle reduced by 4° with respect to adduct 6_butyl. From an energetic perspective, the transition from 6_butyl to TS_6-7butyl involves an estimated free energy barrier of +4.3 kcal mol⁻¹. The transition-state nature of TS_6-7butyl was confirmed by the presence of a single imaginary frequency at −170 cm⁻¹, corresponding to the approach of the chloride ion to the C2 atom and the complete cleavage of the C2–N2 bond. Following TS_6-7butyl, the system evolves toward the minimum-energy structure 7_butyl (Fig. 3) that displays the complete formation of the C2–Cl bond (1.90 Å) and the corresponding cleavage of the N2⋯C2 bond (2.48 Å). The C1–N2 bond is also fully formed, with a length of 1.42 Å. The formation of intermediate 7_butyl was estimated to be exergonic, with a free energy change of −12.6 kcal mol⁻¹. At this point, the electron density originally localized on the N2 center has shifted toward the oxygen atoms of the original CO₂ moiety, leading to a strengthening of the O⋯H hydrogen bond, as indicated by a 0.23 Å shortening compared to the corresponding distance in TS_6-7butyl.

As confirmed by this O⋯H shortening, the oxygen atom of intermediate 7_butyl becomes electron-rich and capable of performing a nucleophilic attack to the C2 center yielding oxazolidin-2-one 2A_butyl by a ring-closure step and the regeneration of catalyst 3. Intermediate 8_butyl (Fig. 4) was achieved through the formation of the transition state TS_7-8butyl, whose structure was computationally identified (Fig. 4).

Fig. 4 Optimized structure of transition state TS_7-8butyl, intermediate 8_butyl and product 2A_butyl. The hydrogen atoms were hidden for the sake of clarity, except for those linked to a heteroatom. Selected distances are given in Å and O–C1–O angle in degrees (°).

In TS_7-8butyl, the O–C2–Cl moiety adopts a quasi-linear arrangement with an angle of 165°, and the C2 atom approaches a quasi-planar geometry. The nucleophilic approach of oxygen atom to C2 initiates the displacement of the chloride ion with an associated free energy barrier of +7.4 kcal mol⁻¹. The transition state nature of TS_7-8butyl is supported by the presence of a single imaginary frequency at −216 cm⁻¹, corresponding to the formation of the O–C2 bond and the cleavage of the C2–Cl bond. Subsequently, intermediate 8_butyl is obtained with a free energy gain of −20.1 kcal mol⁻¹ (Fig. 4).

The complete release of 3-butyl-5-phenyloxazolidin-2-one (2A_butyl) from 8_butyl is exergonic by −5.1 kcal mol⁻¹ accompanied by the restoration of the salt 3, able to perform the activation of a new aziridine moiety. The overall free energy gain, associated with the complete catalytic cycloaddition of CO₂ to 1-butyl-2-phenyl aziridine 1_butyl yielding 2A_butyl, was estimated to be exergonic by −11.2 kcal mol⁻¹ (Fig. 5).

Fig. 5 Free energy (kcal mol⁻¹) pathway for the cycloaddition reaction of CO₂ to 1-butyl-2-phenyl aziridine 1_butyl yielding 3-butyl-5-phenyloxazolidin-2-one 2A_butyl.

The energy profile of the reaction (Fig. 5) reveals that the largest cost for the synthesis of 2A_butyl is +19.3 kcal mol⁻¹, with the main disfavoring contribution associated with the formation of the initial adduct 6_butyl. For comparison purposes, the cycloaddition of CO₂ to 1_butyl was also investigated in the presence of catalysts 4 and 5. The free energy pathways for the three catalytic processes are reported in the SI (Fig. S1). No substantial changes in free energy were identified, except for the free energy stabilization of compounds analogous to intermediate 7_butyl, namely 7’_butyl (involving catalyst 4) and 7”_butyl (involving catalyst 5) within 6 kcal mol⁻¹. The maximum free energy costs are within +20.8 kcal mol⁻¹, corresponding to the formation of compounds 6’_butyl and 6”_butyl, analogous to the initial adduct 6_butyl, and to the subsequent transition states TS_{6’-7’butyl} and TS_{6”-7”butyl}. The energy barriers for TS_{6’-7’butyl} and TS_{6”-7”butyl} fall within the range of +6.2 to +9.2 kcal mol⁻¹. All the free energy values associated with the single step of the processes are listed in Fig. S2.

Effect of the N-aziridine substituent on the reaction efficiency

Already published experimental data⁴⁰ highlighted that the steric and electronic nature of the substituent on the aziridine nitrogen center may drastically influence the reaction efficiency. No conversion was observed when 1-(3,5-bis(trifluoromethyl)phenyl)-2-phenyl aziridine (1_aryl) was used in the CO₂ cycloaddition catalyzed by 3. In contrast, when a cyclohexyl substituent was present at the nitrogen atom, the reaction proceeded with a low efficiency, affording the corresponding oxazolidin-2-one in 21% yield.⁴⁰ To clarify the influence of the substituent at the aziridine nitrogen atom on the reaction efficiency, DFT computational analysis was carried out. The study was run as already described for the reaction involving 1-butyl-2-phenylaziridine 1_butyl. First, two adducts with structures similar to that of 6_butyl were optimized by reacting CO₂ and 3 either with 1-(3,5-bis(trifluoromethyl)phenyl)-2-phenylaziridine (1_aryl) or 1-cyclohexyl-2-phenylaziridine (1_cyhexyl), obtaining 6_aryl and 6_cyhexyl respectively. As shown in Fig. 6, the optimized structure of 6_aryl displays an N2⋯C1 distance as large as 3.02 Å, as well as a very weak hydrogen bond between H(NH⁺) and O(CO₂), with an H⋯O distance of 2.64 Å.

Fig. 6 Optimized structure of adducts 6_aryl and 6_cyhexyl, presenting 1-(3,5-bis(trifluoromethyl)phenyl)-2-phenylaziridine and 1-cyclohexyl-2-phenylaziridine, respectively. The hydrogen atoms were hidden for the sake of clarity, except for those linked to a heteroatom. Selected distances are given in Å and O–C1–O angle in degrees (°).

These features, together with the nearly unperturbed linear structure of the CO₂ moiety, suggest that the occurrence of the process is unlikely when an aryl substituent is present at the N2 aziridine center, in line with the experimental results. For this reason, the CO₂ cycloaddition to 1-(3,5-bis(trifluoromethyl)phenyl)-2-phenylaziridine was not investigated further. Conversely, when 1-cyclohexyl-2-phenylaziridine is the involved substrate, the adduct 6_cyhexyl exhibits an activated CO₂ characterized by a bent O–C–O structure with an angle of 135° and a computed IR-active stretching at 1746 cm⁻¹. From the energy viewpoint, the formation of adduct 6_cyhexyl requires a free energy cost of +19.2 kcal mol⁻¹, higher than that needed for achieving adduct 6_butyl from 1-butyl-2-phenylaziridine (+15 kcal mol⁻¹). Even in this case, the obtained results are in line with experimental data, which highlighted a less efficient process when 1-cyclohexyl-2-phenylaziridine was employed as the starting material instead of 1-butyl-2-phenylaziridine.

The free energy barrier for conversion of 6_cyhexyl to the transition state TS_6-7cyhexyl, whose optimized structure is reported in Fig. S3, is estimated to be +6.5 kcal mol⁻¹ while the overall energy barrier for transforming the starting reagents into TS_6-7cyhexyl is +25.7 kcal mol⁻¹, a quite high value for a reaction performed at room temperature. The whole free energy associated with the formation of oxazolidin-2-one 2A_cyhexyl from 1-cyclohexyl-2-phenylaziridine (1_cyhexyl) and CO₂ is less exergonic than that of the same process involving 1-butyl-2-phenyl aziridine 1_butyl (-7.5 kcal mol⁻¹ versus −11.2 kcal mol⁻¹). The complete energy profile of the CO₂ cycloaddition to 1-cyclohexyl-2-phenyl aziridine is reported in Fig. S4.

Cycloaddition of CS₂ to 1-butyl-2-phenylaziridine (1_butyl) promoted by cinchonine hydrochloride (3)

A precedent computational analysis has predicted the feasibility, at least in silico, of the CS₂ cycloaddition to N-alkyl aziridines to provide thiazolidin-2-thiones in the presence of bifunctional TPPH₄Cl₂.⁴⁵ The study revealed a free energy contribution of −22.1 kcal mol⁻¹ associated with the cycloaddition of CS₂ to 1-butyl-2-phenylaziridine 1_butyl to provide the corresponding thiazolidin-2-thione 9A_butyl.⁴⁵ Based on these results, we studied the reaction between CS₂ and 1-butyl-2-phenylaziridine 1_butyl also in the presence of catalyst 3. In this regard, the adduct 10_butyl (Fig. 7) was optimized.

Fig. 7 Optimized structure of adduct 10_butyl. The hydrogen atoms were hidden for the sake of clarity, except for those linked to a heteroatom. Selected distances are given in Å and O–C1–O angle in degrees (°).

Despite the favorable enthalpic contribution of −2.1 kcal mol⁻¹, adduct 10_butyl was optimized with a free energy cost of +21.5 kcal mol⁻¹. It should be noted that the energy cost is 6.5 kcal mol⁻¹ larger than that of the analogous process involving CO₂ (formation of adduct 6_butyl), possibly due to a weaker hydrogen bonding between the proton of 3 and the sulfur atom of CS₂, as confirmed by the long S–H distance of 2.28 Å.

After the formation of adduct 10_butyl, the activated aziridine substrate can be attacked by a chloride nucleophile through the transition state TS_10-11butyl (Fig. 8) with a free energy barrier of +7.7 kcal mol⁻¹. The overall barrier of +29.2 kcal mol⁻¹ for the conversion of the reactants into TS_10-11butyl was larger than that calculated for CO₂ activation and it can be overcome by performing the reaction at higher experimental temperatures. Transition state TS_10-11butyl features a quasi-linear Cl–C2–N2 arrangement with an angle of 151° and a weakened C2–N2 bond, whose length is stretched by 0.15 Å compared to that in 10_butyl. In TS_10-11butyl, the chloride approaches the C2 atom with a consistent shortening of Cl⋯C2 distance by 1.0 Å.

Fig. 8 Optimized structure of transition state TS_10-11butyl and of intermediate 11_butyl. The hydrogen atoms were hidden for the sake of clarity, except for those linked to a heteroatom. Selected distances are given in Å and O–C1–O angle in degrees (°).

The complete formation of the C2–Cl bond in intermediate 11_butyl (Fig. 8) was estimated to be exergonic by −25.9 kcal mol⁻¹.

Similar to the CO₂ activation, the sulfur atom in 11_butyl may perform a nucleophilic attack on the C2 center to form the 5-membered ring and regenerate the catalyst 3. The transition state TS_11-12butyl, shown in Fig. S5, was obtained with a free energy barrier of +13.6 kcal mol⁻¹, while the intermediate 12_butyl (Fig. 9) was obtained with the large free energy gain of −29.9 kcal mol⁻¹.

Fig. 9 Optimized structure of the intermediate 12_butyl and thiazolidin-2-thione, 9A_butyl. The hydrogens were hidden for the sake of clarity, except for those linked to a heteroatom. Selected distances are given in Å and O–C1–O angle in degrees (°).

The complete release of thiazolidin-2-thione 9A_butyl, shown in Fig. 9, was achieved with a further free energy gain of −9.1 kcal mol⁻¹. The overall formation of compound 9A_butyl by CS₂ cycloaddition to aziridine 1_butyl is depicted in Fig. 10.

Fig. 10 Free energy (kcal mol⁻¹) pathway for the cycloaddition reaction of CS₂ to 1-butyl-2-phenyl aziridine 1_butyl forming thiazolidin-2-thione 9A_butyl.

A comparison between the energy profiles of CS₂ (Fig. 10) and CO₂ (Fig. 5) cycloaddition to 1_butyl reveals key differences. While a higher barrier is required for the formation of the initial adduct 10_butyl and for reaching the transition state TS_10-11butyl in the case of CS₂ activation, more energy is released during CS₂ activation rather than during CO₂ activation. In summary, in silico analysis predicts that the energy barriers for the catalytic cycloaddition of CS₂ to aziridine rings are not prohibitively high, although temperatures above room temperature might be necessary to promote the reaction.

Cycloaddition of COS to 1-butyl-2-phenylaziridine (1_butyl) promoted by cinchonine hydrochloride (3)

Until now, only the activation of symmetric triatomic molecules has been investigated. This raises the question if the reaction can also proceed when a non-symmetric substrate, such as carbonyl sulfide COS, is used in the cycloaddition to aziridines.

Unlike previous cases, different structural isomers can be obtained starting from 1-butyl-2-phenylaziridine 1_butyl and COS. Depending on which heteroatom is involved in the cyclization step, 3-butyl-5-phenylthiazolidin-2-one (13A_butyl) or 3-butyl-5-phenyloxazolidin-2-thione (14A_butyl) can be formed (Fig. 11).

Fig. 11 Potential products of the cycloaddition of COS to 1-butyl-2-phenylaziridine (1_butyl).

To begin the computational analysis of the reaction between COS and 1_butyl, both compounds shown in Fig. 11 were optimized. Isomers 13B_butyl and 14B_butyl, derived from the nucleophilic attack to the less electrophilic carbon atom C3 of the aziridine ring, were not theoretically modelled in view of the unfavorable energy costs related to their formation (see below).

Preliminary DFT calculations revealed that product 13A_butyl is more stable than 14A_butyl by 10.4 kcal mol⁻¹ in free energy. Although the energy difference between the two potential products is significant, the processes yielding both isomers were investigated. Given the non-symmetrical nature of COS, both adducts 15_butyl (evolving in 13A_butyl) and 16_butyl (evolving in 14A_butyl) (Fig. 12) were optimized, featuring the alternative involvement of either oxygen or sulfur in hydrogen bonding, respectively. Adduct 15_butyl was estimated to be 1.3 kcal mol⁻¹ more stable than 16_butyl, mainly due to more efficient hydrogen bonding when oxygen, rather than sulfur, is involved.

Fig. 12 Free energy (kcal mol⁻¹) pathway for the cycloaddition reaction of CS₂ to 1-butyl-2-phenyl aziridine 1_butyl forming thiazolidin-2-thione 9A_butyl. Selected distances are given in Å and O–C1–O angle in degrees (°).

Starting from the reactants, adduct 15_butyl is obtained with a free energy cost of +15.9 kcal mol⁻¹, while the formation of 16_butyl requires a slightly higher cost of +17.2 kcal mol⁻¹. For adduct 15_butyl, the N2–C2 bond cleavage via nucleophilic attack of chloride proceeds through transition state TS_15-17butyl (Fig. S6) with an associated free energy barrier of +5.8 kcal mol⁻¹. The system then evolves toward intermediate 17_butyl (Fig. 13), with a free energy gain of −16.7 kcal mol⁻¹. Accordingly, an overall energy barrier of +21.7 kcal mol⁻¹ must be overcome to reach TS_15-17butyl from the separate reactants.

Fig. 13 Optimized structure of the intermediate 17_butyl. The hydrogens were hidden for the sake of clarity, except for those linked to a heteroatom center. Selected distances are given in Å and O–C1–O angle in degrees (°).

Accordingly, an overall energy barrier of +21.7 kcal mol⁻¹ must be overcome to reach TS_15-17butyl from the separate reactants. The potential nucleophilic attack of chloride on the C3 center has been also investigated highlighting a free energy barrier of +17.8 kcal mol⁻¹ to obtain TS_15-17butylC3, shown in Fig. S6, from adduct 15_butyl. Since the overall estimated free energy barrier for reaching TS_15-17butylC3 is as high as +33.7 kcal mol⁻¹, this mechanism was discarded.

In the case of adduct 16_butyl, where the sulfur atom is involved in hydrogen bonding, the estimated free energy barrier for reaching TS_16-18butyl is +11.5 kcal mol⁻¹, resulting in an overall barrier of +28.7 kcal mol⁻¹ from the isolated reactants to TS_16-18butyl. The system then proceeds to intermediate 18_butyl, with a free energy gain of −25.4 kcal mol⁻¹. Thus, in view of the quite high calculated barrier compared to that of TS_15-17butyl, the energy pathway for the formation of 14A_butyl was discarded.

Starting from 17_butyl, the transition state TS_17-19butyl was computed with a free energy barrier of +7.9 kcal mol⁻¹, in which the sulfur center of COS can perform a nucleophilic attack to the C2 center yielding the intermediate 19_butyl that evolves into the final product 3-butyl-5-phenylthiazolidin-2-one 13A_butyl. The complete release of 13A_butyl and restoration of catalyst 3 is accompanied by a free energy gain of −6.9 kcal mol⁻¹. A complete free energy pathway for the production of 3-butyl-5-phenylthiazolidin-2-one 13A_butyl, starting from COS and aziridine 1_butyl promoted by metal-free 3, is depicted in Fig. 14 and shows an overall free energy gain of −20.8 kcal mol⁻¹.

Fig. 14 Free energy (kcal mol⁻¹) pathway for the cycloaddition reaction of COS to 1-butyl-2-phenyl aziridine 1_butyl forming 13A_butyl.

To better summarize and compare the reactivity of the investigated triatomic molecules, the energy profiles of their cycloaddition to 1-butyl-2-phenyl aziridine 1_butyl catalyzed by 3 were superimposed, as shown in Fig. 15.

Fig. 15 Superimposed free energy (kcal mol⁻¹) profiles of the cycloaddition reaction of CO₂, COS and CS₂ to 1-butyl-2-phenyl aziridine 1_butyl catalyzed by 3.

A clear difference in overall energy gains can be observed between the formation of oxazolidine-2-one 2A_butyl and its sulfur-containing analogues (9A_butyl and 13A_butyl). The overall energy gains calculated for CS₂ and COS cycloaddition to 1_butyl were nearly twice as high as that for CO₂ (−20.8 kcal mol⁻¹ and −22.1 kcal mol⁻¹ versus −11.2 kcal mol⁻¹), suggesting a thermodynamic preference for the formation of oxazolidin-2-thiones and thiazolidin-2-thiones over oxazolidin-2-ones.

However, the activation barrier computed for the formation of the key transition state was higher for CS₂ (+29.2 kcal mol⁻¹) compared to CO₂ (+19.3 kcal mol⁻¹), indicating that more forcing reaction conditions may be required to achieve the product formation. In contrast, the energy barrier to reach the first transition state from COS is only slightly larger than that calculated for CO₂ (+21.7 kcal mol⁻¹ versus +19.3 kcal mol⁻¹), suggesting that its cycloaddition could proceed efficiently under the similar mild conditions successfully employed for the 3-catalyzed oxazolidin-2-one synthesis. These findings support the feasibility of COS cycloaddition to aziridines promoted by catalyst 3 under experimental conditions only slightly more drastic than those validated for CO₂ transformations.

Conclusions

In conclusion, the energetic/structural DFT analyses suggest a mechanism for the CO₂ cycloaddition to the aziridine ring for the synthesis of oxazolidin-2-ones, efficiently promoted by cinchonine hydrochloride 3 under very mild conditions (RT and 0.1 MPa of CO₂ pressure). The calculated energy barriers are compatible to the employed experimental conditions and the DFT study confirmed the double activation of CO₂ through hydrogen bonding interactions and nucleophile attack of the aziridine nitrogen atom to the CO₂ carbon atom. The theoretical analysis explains the dependence of the catalytic efficiency on the steric hindrance and/or electronic effects of the N-aziridine substituents, providing a rationale for the lack of reactivity of N-aryl aziridines, observed experimentally.

To extend the activation of triatomic molecules mediated by 3 from CO₂ to CS₂ and COS, the DFT study reported here will be fundamental for managing in near future experimental reactions involving CS₂ and COS, which until now have only been theoretically predicted. The computational analysis revealed that catalyst 3 could be a potential candidate for efficiently mediating both CS₂ and COS activation. Even if theoretical calculations underlined that both processes involve higher energy barriers than those optimized for the CO₂ activation, acquired data support a future experimental study on the CS₂ and COS cycloaddition to aziridine in the presence of 3. By comparing the calculated energy barriers, the cycloaddition of COS is predicted to be feasible at temperatures close to the ambient one, under conditions therefore similar to those observed for the 3-catalyzed valorization of CO₂. In contrast, the use of CS₂ would require slightly higher temperatures. In both cases, the synthesis of thiazolidin-2-thiones and oxazolidin-2-thiones is predicted to be feasible using an inexpensive and biocompatible catalyst, such as cinchonine hydrochloride salt (3), under mild reaction conditions.

Conflicts of interest

There are no conflicts to declare.

Data availability

The datasets supporting this article are included in the SI. See DOI: https://doi.org/10.1039/d5nj03189c.

Additional data are available from the corresponding author upon reasonable request.

Acknowledgements

CD and EG thank Università degli Studi di Milano for the PSR 2023 grants. GM acknowledges the CINECA award under the ISCRA initiative, for the availability of high-performance computing resources and support.

References

1. X. B. Li, Z. K. Xin, S. G. Xia, X. Y. Gao, C. H. Tung and L. Z. Wu, Chem. Soc. Rev., 2020, 49, 9028–9056.
2. L. Wang, W. Chen, D. Zhang, Y. Du, R. Amal, S. Qiao, J. Wu and Z. Yin, Chem. Soc. Rev., 2019, 48, 5310–5349.
3. D. Intrieri, C. Damiano, P. Sonzini and E. Gallo, J. Porphyrins Phthalocyanines, 2019, 23, 305–328.
4. C. Damiano, M. Cavalleri, L. Invernizzi and E. Gallo, Eur. J. Org. Chem., 2024, e202400616.
5. T. Yan, H. Liu, Z. X. Zeng and W. G. Pan, J. CO2 Util., 2023, 68, 102355.
6. K. Wiranarongkorn, K. Eamsiri, Y. S. Chen and A. Arpornwichanop, J. CO2 Util., 2023, 71, 102477.
7. L. Li, X. Li, Y. Sun and Y. Xie, Chem. Soc. Rev., 2022, 51, 1234–1252.
8. L. Rotundo, R. Gobetto and C. Nervi, Curr. Opin. Green Sustainable Chem., 2021, 31, 100509.
9. Q. Zhang and X. Jin, Chem. – Eur. J., 2025, 31, e202500933.
10. M. Aresta, A. Dibenedetto and E. Quaranta, J. Catal., 2016, 343, 2–45.
11. S. Kaewsai and V. D’ Elia, J. Organomet. Chem., 2025, 1039, 123799.
12. W. Natongchai, D. Crespy and V. D’ Elia, Chem. Commun., 2025, 61, 419.
13. S. Arayachukiat, P. Yingcharoen, S. V. C. Vummaleti, L. Cavallo, A. Poater and V. D’Elia, Molecular Catalysis, 2017, 443, 280–285.
14. P. Yingcharoen, W. Natongchai, A. Poater and V. D’ Elia, Catal. Sci. Technol., 2020, 10, 5544–5558.
15. C. Damiano, P. Sonzini, G. Manca and E. Gallo, Eur. J. Org. Chem., 2021, 2807–2814.
16. M. Cavalleri, C. Damiano, G. Manca and E. Gallo, Chem. – Eur. J., 2023, 29, e202202729.
17. C. Damiano, P. Sonzini, M. Cavalleri, G. Manca and E. Gallo, Inorg. Chim. Acta, 2022, 540, 121065.
18. P. Sonzini, C. Damiano, D. Intrieri, G. Manca and E. Gallo, Adv. Synth. Catal., 2020, 362, 2961–2969.
19. P. Sonzini, N. Berthet, C. Damiano, V. Dufaud and E. Gallo, J. Catal., 2022, 414, 143–154.
20. C. Damiano, A. Fata, M. Cavalleri, G. Manca and E. Gallo, Catal. Sci. Technol., 2024, 14, 3996–4006.
21. A. Z. Bialvaei, M. Rahbar, M. Yousefi, M. Asgharzadeh and H. S. Kafil, J. Antimicrob. Chemother., 2017, 72, 354–364.
22. D. McBride, T. Krekel, K. Hsueh and M. J. Durkin, Expert Opin. Drug Metab. Toxicol., 2017, 4, 491.
23. F. Moureau, J. Wouters, D. Vercauteren, S. Collin, G. Evrard, F. Durant, F. Ducrey, J. Koenig and F. Jarreau, Eur. J. Med. Chem., 1992, 27, 939–948.
24. G. Bresciani, M. Bortoluzzi, G. Pampaloni and F. Marchetti, Org. Biomol. Chem., 2021, 19, 4152–4161.
25. C. S. Dewey and R. A. Bafford, J. Org. Chem., 1965, 30, 491–495.
26. T. A. Foglia, L. M. Gregory, G. Maerker and S. F. Osman, J. Org. Chem., 1971, 36, 1068–1072.
27. K. K. Pandey, Coord. Chem. Rev., 1995, 140, 37–114.
28. M. Guo, B. Dong, Y. Qu, Z. Sun, L. Yang, Y. Wang, I. L. Fedushkin and X. J. Yang, Chem. – Eur. J., 2025, 31, e202403652.
29. Y. Xie, C. Lu, B. Zhao, Q. Wang and Y. Yao, J. Org. Chem., 2019, 84, 1951–1958.
30. Y. Shi, D. Wen and S. Q. Zhao, Inorg. Chem., 2025, 64, 4387–4392.
31. W. Ding, X. Tang, S. Jin, Z. Li, D. Xu, X. Kang and Z. Liu, Green Chem., 2024, 27, 218–226.
32. Y. Shi, B. Tang, X.-L. Jiang, Y.-E. Jiao, H. Xu and B. Zhao, J. Mater. Chem. A, 2022, 10, 4889–4894.
33. A. J. Plajer and C. K. Williams, Angew. Chem., Int. Ed., 2022, 61, e2021044952022.
34. K. Sieja, J. von Mach-Szczypiński and J. von Mach-Szczypiński, Med. Pr., 2018, 69, 317–323.
35. A. W. Demartino, D. F. Zigler, J. M. Fukuto and P. C. Ford, Chem. Soc. Rev., 2017, 46, 21–39.
36. R. Morales-Nava, M. Fernández-Zertuche and M. Ordóñez, Molecules, 2011, 16, 8803–8814.
37. A. Khalaj and M. Khalaj, J. Chem. Res., 2016, 40, 445–448.
38. A. Biswas and S. Hajra, Adv. Synth. Catal., 2022, 364, 3035–3042.
39. M. Sengoden, G. A. Bhat and D. J. Darensbourg, Green Chem., 2022, 24, 2535–2541.
40. L. Invernizzi, C. Damiano and E. Gallo, Chem. – Eur. J., 2025, 31, e202500473.
41. S. Grimme, J. Comp. Chem., 2006, 27, 1787–17991.
42. V. Barone and M. Cossi, J. Phys. Chem. A, 1998, 102, 1995–20013.
43. M. Cossi, N. Rega, G. Scalmani and V. Barone, J. Comp. Chem., 2003, 24, 669–681.
44. L. R. Domingo, M. Ríos-Gutiérrez and P. Pérez, Molecules, 2016, 21, 7482016.
45. C. Damiano, N. Panza, J. Nagy, E. Gallo and G. Manca, New J. Chem., 2023, 47, 4306–4312.

---