Mapping the catalytic landscape of triphenylborane (BPh₃)-catalyzed CO₂-epoxide coupling to carbonates: an in silico approach to solve substrate-dependent selectivity†

Nikunj Kumar^a and Puneet Gupta*^ab
aComputational Catalysis Center, Department of Chemistry, Indian Institute of Technology, Roorkee 247667, Uttarakhand, India
^bCenter for Sustainable Energy, Indian Institute of Technology Roorkee, Roorkee 247667, Uttarakhand, India. E-mail: puneet.gupta@cy.iitr.ac.in

First published on 9th June 2025

---

---

1. Introduction

Carbon dioxide (CO₂) is a prominent greenhouse gas naturally regulated through the global carbon cycle, involving terrestrial plants, oceans, and microorganisms.¹ The primary drivers of CO₂ emissions are the extensive combustion of fossil fuels and various anthropogenic activities.^2–4 Since the industrial revolution, CO₂ emissions have risen significantly, surpassing the natural absorption capacity due to simultaneous deforestation and ocean acidification, thereby contributing to global warming.⁵ However, CO₂ is not merely an environmental concern; it also represents a promising C1 feedstock that is abundant, non-toxic, non-flammable, and readily available as a sustainable carbon source.^6,7 In the context of environmental sustainability and resource valorization, developing innovative technologies for the catalytic conversion of CO₂ into high-value products has become a focal point in academia and industry.⁸ Among the diverse CO₂ utilization strategies, its coupling with epoxides to synthesize cyclic-carbonates (CCs) or poly-carbonates (PCs) has garnered significant interest due to its 100% atom economy and high carbon utilization efficiency, aligning with the principles of green chemistry.^9,10 CCs serve as green polar aprotic solvents,¹¹ Li battery electrolytes,¹² and precursors for pharmaceuticals.¹³ Meanwhile, PCs are valuable materials for synthesizing polyols used in various applications¹⁴ and offer a sustainable alternative to conventional synthetic polymers derived from non-renewable petroleum feedstocks.¹⁵

In the absence of external catalysts, the coupling reactions between CO₂ and epoxides require high temperatures and pressures to overcome the energy barriers associated with CO₂ activation and its conversion into CCs or PCs.^16,17 To circumvent these obstacles, multifarious catalytic systems have been investigated, including metal porphyrins,¹⁸ salen metal complexes,¹⁹ metal oxides,²⁰ alkali metals,²¹ and metal–organic frameworks (MOFs),²² all of which have demonstrated effectiveness in facilitating the CO₂ to CC conversions (Fig. 1a). Similarly, for the transformation of CO₂ into PCs, the pioneering work of Inoue and coworkers, who first reported the use of diethylzinc for the copolymerization of CO₂ and epoxides, laid the foundation for significant advancements in the design of highly efficient catalytic systems.^23,24 Over the past decade, a diverse array of metal-based catalysts featuring transition metals and earth-abundant main group metals have been extensively explored for PC production.^25–30 Despite their high catalytic activity, metal-based systems often face challenges such as potential metal leaching, recycling difficulties, and complex synthetic procedures, which limit their industrial applicability.³¹ Consequently, research efforts have increasingly focused on the development of metal-free catalytic systems, driven by their cost-effectiveness, environmental sustainability, and enhanced chemical and thermal stability.³²

In recent years, boron-based catalysts have emerged as promising candidates in the field of metal-free catalysis of CO₂ conversion, primarily due to the intrinsic Lewis acidity of the boron center (Fig. 1b).^33–39 Zhang and coworkers demonstrated that boronic acids, in combination with tetrabutylammonium iodide (TBAI), can effectively catalyze the synthesis of CCs from CO₂ and various epoxides.³⁴ Liu and coworkers reported a binary catalytic system comprising tetrahydroxydiboron and TBAI, which facilitated the cycloaddition of CO₂ with epoxides.³⁵ Independently, Lu and coworkers showed that boron oxide (B₂O₃), in conjunction with tetrabutylammonium bromide (TBAB), can serve as an efficient catalyst for CO₂-epoxide cycloaddition.³⁶ More recently, our research group conducted a computational study that provided valuable mechanistic insights into the B₂O₃-catalyzed cycloaddition reaction, shedding light on key reaction pathways and activation modes.³⁷ Beyond cyclic-carbonate synthesis, boron-based catalysts have also been explored to copolymerize CO₂ with epoxides to produce PCs. Feng, Gnanou, and coworkers reported the first example of a boron-based catalytic system for this transformation, utilizing a binary system comprising triethylborane (BEt₃) and onium salts such as bis(triphenylphosphine)iminium chloride (PPNCl) to facilitate the copolymerization of CO₂ with both propylene oxide (PO) and cyclohexene oxide (CHO).³⁸ Building on this approach, Zhang and coworkers demonstrated that BEt₃, in combination with a tertiary amine (TEA), could also effectively catalyze the copolymerization of CO₂ and PO.³⁹

Inspired by the work of Gnanou and coworkers,³⁸ Kerton and colleagues explored the use of triphenylborane (BPh₃) as an alternative to BEt₃ for catalyzing the coupling of CO₂ with epoxides.⁴⁰ Interestingly, their study revealed substrate-dependent selectivity, leading to the formation of either CC or PC products. Using the same binary catalytic system comprising BPh₃ and PPNCl, they observed that PO preferentially formed CC, whereas CHO favoured PC formation (Fig. 1c). Experimental kinetic studies further unveiled a distinctive feature of BPh₃ catalysis. Unlike many other metal-free systems, where the reaction order with respect to CO₂ is typically one or zero, this system exhibited an inverse dependence, resulting in an overall reaction order of −1 for CO₂. However, the fundamental factors governing this unique catalytic behaviour remain unclear, necessitating further investigation. Additionally, the effect of substituting ethyl groups with phenyl groups at the boron center and its role in dictating substrate-dependent selectivity remains an open question. To the best of our knowledge, no comprehensive computational study has sketched the catalytic landscape of BPh₃-mediated CO₂-epoxide coupling. Herein, we employ density functional theory (DFT) calculations to elucidate the reaction mechanism of BPh₃-catalyzed CO₂ and epoxide coupling. Through distortion–interaction analysis (DIA) and non-covalent interaction (NCI) analysis, we aim to provide fundamental insights into the origins of substrate-dependent selectivity in this system.

2. Computational details

Computational calculations were performed using the ORCA quantum chemical software package (version 5.0.3).^41,42 Geometry optimizations for all minima and transition states were conducted in the gas phase using the hybrid M06-2X^43,44 density functional, in combination with the def2-SVP⁴⁵ basis set, with the defgrid3 tight integration grid. The selection of the M06-2X functional was based on its proven reliability in modelling boron-catalyzed coupling reactions, ensuring accurate predictions of reaction mechanisms.^46–48 To enhance computational efficiency, the RIJCOSX⁴⁹ approximation was employed. Vibrational frequency analysis was performed at the same level of theory to confirm the nature of the stationary points and to obtain zero-point energy (ZPE) and thermal corrections at the corresponding experimental temperatures. Intrinsic reaction coordinate (IRC) calculations⁵⁰ were carried out to ensure that transition state correctly connects the reactant and product minima on the potential energy surface. Since the experimental studies were conducted under solvent-free conditions,⁴⁰ no solvation model is applied in our calculations. To further refine the energies, single-point energy calculations were performed on the optimized geometries using the M06-2X functional with the larger def2-TZVP⁴⁵ basis set. All reported relative Gibbs energies in the mechanistic study are obtained at the M06-2X/def2-TZVP//M06-2X/def2-SVP level of theory. Non-covalent interaction (NCI) analyses were conducted using the Multiwfn program,⁵¹ and NCI isosurfaces were visualized with VMD 1.9.3.⁵² The three-dimensional molecular structures were rendered using the CYLview program.⁵³ The computational model employed in our mechanistic study is illustrated in Fig. 2.

3. Results and discussion

The result and discussion section is divided into distinct subsections to systematically analyze the catalytic behaviour of borane systems in CO₂-epoxide coupling reactions. We first investigated the BPh₃-catalyzed CO₂ with propylene oxide (PO) and CO₂ with cyclohexene oxide (CHO) coupling reactions, highlighting their mechanistic pathways and product selectivity. To further understand the substrate-dependent selectivity observed in BPh₃ catalysis, we performed a detailed computational analysis to elucidate the underlying factors influencing reaction outcomes. Next, we extend our study to BEt₃-catalyzed CO₂-epoxide coupling reactions, examining the BEt₃-catalyzed reactions of CO₂ with PO and CHO, respectively. A comparative evaluation of BPh₃ and BEt₃ systems provides deeper insights into how boron substitution influences product selectivity.

3.1 BPh₃ catalyzed coupling reaction of CO₂ and propylene oxide (PO)

The Gibbs free energy profile for the BPh₃-catalyzed coupling reaction of CO₂ and PO is depicted in Fig. 3. To account for the role of the bulky PPNCl co-catalyst, its catalytic influence was modelled using a chloride ion (Cl⁻), as its primary contribution stems from the Cl⁻ active center. This approach aligns with previous computational studies investigating the role of quaternary ammonium halide salts in catalysis.^37,54–56 The reaction begins with the formation of adduct Int1_a, in which one BPh₃ molecule coordinates to the PO group, activating it, while a second BPh₃ forms a complex with a chloride ion (BPh₃_Cl⁻); these two species then associate to form the Int1_a adduct. Subsequently, Int1_a undergoes ring opening via the transition state TS1_a, facilitated by the direct nucleophilic attack of Cl⁻ on the epoxide, leading to concurrent C_PO–O_PO bond cleavage and C_PO–Cl bond formation. This step proceeds with an energy barrier of 27.2 kcal mol⁻¹ (relative to BPh₃ + PO). The subsequent introduction of a CO₂ molecule and desorption of one BPh₃ results in the formation of a stable intermediate Int2_a, characterized by weak van der Waals interactions between CO₂ and the alkoxide (O⁻) species.

Introducing an additional BPh₃ molecule in the vicinity of CO₂ leads to the formation of a new species, Int3_a, which undergoes CO₂ insertion via a transition state, TS2_a, with a reduced activation barrier of 20.9 kcal mol⁻¹ (relative to Int2_a). In TS2_a, two BPh₃ molecules are involved–one coordinating the epoxide and the other interacting with the CO₂ molecule. Following the nucleophilic attack of the alkoxide (O⁻) on the CO₂ carbon center via TS2_a, the reaction proceeds to form the intermediate Int4_a.

From Int4_a, the reaction can proceed via two different pathways: one leading to the formation of cyclic-carbonate (CC) and the other resulting in poly-carbonate (PC) formation. The departure of the weakly coordinated BPh₃ molecule from the epoxide oxygen generates the intermediate Int5_a, which undergoes intramolecular cyclization through the transition state TS3_a. This ring-closing transition state involves the formation of a C_PO–O_CO2 bond, accompanied by the simultaneous cleavage of the C_PO–Cl bond. The activation barrier for this step is 23.9 kcal mol⁻¹ (relative to Int5_a), ultimately yielding Int6_a (the CC product coordinated to BPh₃ species and regenerated Cl⁻ ion). Alternatively, in the poly-carbonate formation pathway, the dissociation of BPh₃ from Int4_a, followed by the introduction of a BPh₃-activated PO molecule, leads to the formation of Int7_a. This intermediate subsequently undergoes epoxide addition through the transition state TS4_a, which requires an activation barrier of 28.6 kcal mol⁻¹ (relative to Int2_a). In TS4_a, nucleophilic attack by the alkoxide (O⁻) on the newly introduced PO molecule results in an epoxide ring opening, yielding Int8_a. This species serves as the key intermediate for poly-carbonate propagation, where further reaction can proceed through alternating CO₂ insertion and epoxide addition steps, facilitating polymer chain growth.

A comparison of the activation barriers for the ring-closing step (TS3_a, ΔG^‡ = 23.9 kcal mol⁻¹) and the epoxide addition step (TS4_a, ΔG^‡ = 28.6 kcal mol⁻¹) indicates that the formation of the CC product is more favorable than PC formation. This aligns well with experimental observations, where the BPh₃-catalyzed coupling of CO₂ and PO yields the CC product. Among all the mechanistic steps, the epoxide ring-opening transition state TS1_a exhibits the highest activation barrier (ΔG^‡ = 27.2 kcal mol⁻¹), identifying it as the rate-determining step. This computational finding is consistent with the experimental proposed rate-determining step and kinetic studies, showing a first-order dependence on PO, PPNCl, and BPh₃. Furthermore, experimental investigations by Kerton and coworkers⁴⁰ revealed that lower CO₂ concentrations accelerated reaction rates and enhanced overall substrate conversion. As illustrated in Fig. 4, introducing CO₂ to the BPh₃ coordinated PO species BPh₃_PO leads to the de-coordination of the PO molecule, resulting in the formation of a stable inactive species BPh₃_CO₂.

This inactive species inhibits the coordination and activation of the epoxide, which is essential for the coupling reaction, thereby suppressing catalytic activity. The formation of this inactive species increases the overall activation barrier to 31.3 kcal mol⁻¹, which is not reasonable for catalysis at 100 °C. This computational insight provides a mechanistic explanation for the experimentally observed inverse rate dependence on CO₂ concentration (reaction order with respect to CO₂ = −1). It suggests that an excess of CO₂ slows down the epoxide ring-opening step by forming an inactive species, thus decelerating the overall reaction.

3.2 BPh₃ catalyzed coupling reaction of CO₂ and cyclohexene oxide (CHO)

The Gibbs free energy profile for the BPh₃-catalyzed coupling reaction of CO₂ and CHO is presented in Fig. 5. The reaction begins with formation of adduct Int1_b, followed by ring opening via TS1_b, where nucleophilic attack by the chloride ion induces cleavage of the C_CHO–O_CHO bond while simultaneously forming the C_CHO–Cl bond. This step proceeds with an activation barrier of 22.3 kcal mol⁻¹ (relative to BPh₃ + CHO). The subsequent introduction of CO₂ and desorption of one BPh₃ results in the formation of the oxyanion intermediate Int2_b, characterized by weak van der Waals interactions between C_CO2 and the alkoxide O⁻. In Int2_b, when an additional BPh₃ molecule interacts with CO₂, leading to the formation of Int3_b. This additional interaction facilitates CO₂ activation via the transition state TS2_b, which requires an activation barrier of 20.6 kcal mol⁻¹ (relative to Int2_b). In TS2_b, the alkoxide (O⁻) attacks the carbon center of CO₂, yielding the intermediate Int4_b.

This boron-carbonate species, Int4_b, serves as a branching point for both CC and PC formation pathways. In the CC formation route, the weakly coordinated BPh₃ molecule dissociates from O_CHO, generating the intermediate Int5_b, which can subsequently undergo intramolecular cyclization through the five-membered transition state TS3_b. This step features the formation of a C_CHO–O_CO2 bond, accompanied by the simultaneous breaking of the C_CHO–Cl bond, eventually yielding Int6_b (where the CC product remains coordinated to BPh₃, and Cl⁻ is regenerated). However, unlike the BPh₃-catalyzed reaction of CO₂ and PO, the ring-closing step in the CO₂–CHO system requires a significantly higher energy barrier (ΔG^‡ = 59.1 kcal mol⁻¹, relative to Int2_b), making CC formation kinetically unfavourable. Alternatively, in the PC formation pathway, the dissociation of BPh₃ from Int4_b, followed by addition of activated CHO molecule by BPh₃, leads to the formation of Int7_b. The polymerization process proceeds via epoxide addition through the transition state TS4_b, with an activation barrier of 18.4 kcal mol⁻¹ (relative to Int2_b). In this step, nucleophilic attack by the alkoxide (O⁻) on the second CHO molecule induces epoxide ring opening, yielding Int8_b. Once Int8_b is formed, polymer chain growth can continue through alternating CO₂ insertion and epoxide addition steps. In summary, the energy barrier required for CC product formation (ΔG^‡ = 59.1 kcal mol⁻¹) is significantly higher than the activation barrier of PC formation (ΔG^‡ = 18.4 kcal mol⁻¹). This substantial energy difference renders the PC formation pathway kinetically favourable providing a mechanistic rationale for the experimental observation of PC product in BPh₃-catalyzed CO₂ and CHO coupling reactions.

3.3 Origins of substrate-dependent selectivity in BPh₃ catalyzed coupling reactions

To gain deeper insights into the substrate-dependent selectivity observed in BPh₃-catalyzed CO₂ and epoxide coupling reactions, we performed distortion/interaction analysis (DIA)^57–59 on the two key competing transition states: ring-closing (leading to the CC product) and epoxide addition (leading to the PC product) for both CO₂ with PO and CO₂ with CHO coupling reactions (Fig. 6). In this approach, each transition state was fragmented into two components: the catalyst (Cat) and the substrate (Sub). Single-point energy calculations were then conducted to determine the distortion energy (ΔE_dist) and interaction energy (ΔE_int) between the fragments (see ESI† section S3 for further details). For the CO₂ and PO system, the total distortion energy in the epoxide addition transition state TS4_a (ΔE_dist = 30.7 kcal mol⁻¹) is 5.2 kcal mol⁻¹ lower than that in the ring-closing transition state TS3_a (ΔE_dist = 35.9 kcal mol⁻¹). However, the interaction energy between the distorted fragments is significantly higher in TS3_a (ΔE_int = −12.6 kcal mol⁻¹) compared to TS4_a (ΔE_int = −1.8 kcal mol⁻¹). These results suggest that the strong interaction energy between the fragments in TS3_a plays a dominant role in favoring cyclic-carbonate formation in BPh₃-catalyzed CO₂ and PO coupling.

In contrast, for the CO₂ and CHO system, both the total distortion energy and interaction energy are higher in the ring-closing transition state TS3_b (ΔE_dist = 100.1 kcal mol⁻¹, ΔE_int = −43.0 kcal mol⁻¹) compared to the epoxide addition transition state TS4_b (ΔE_dist = 27.2 kcal mol⁻¹, ΔE_int = −11.1 kcal mol⁻¹). The significantly higher distortion energy in TS3_b makes the ring-closing step less favorable than epoxide addition, leading to a preference for poly-carbonate formation. In summary, the differences in distortion and interaction energies between the ring-closing and epoxide addition transition states play a crucial role in determining the product selectivity. In the CO₂ and PO system, strong interaction energy in the ring-closing step promotes cyclic-carbonate formation, whereas in the CO₂ and CHO system, the high distortion energy in the ring-closing step disfavours cyclic-carbonate formation, making poly-carbonate the preferred product.

In addition, non-covalent interaction (NCI) analysis⁶⁰ was conducted to further investigate the differences between the ring-closing transition states in the BPh₃-catalyzed coupling reactions of CO₂ and PO (TS3_a), and CO₂ and CHO (TS3_b), respectively. The NCI plots (Fig. 7) for TS3_a and TS3_b reveal a significant increase in steric repulsion around the five-membered ring formation when the substrate changes from PO to the bulkier CHO. The enhanced red isosurfaces in TS3_b indicate pronounced steric hindrance during the ring-closing step, making this transition state energetically less favorable. These findings strongly suggest that the high activation barrier observed for the ring-closing step in the CO₂ and CHO system is a direct consequence of increased steric repulsion. Furthermore, this conclusion is consistent with our DIA results, which showed significantly higher distortion energy for TS3_b compared to TS3_a, reinforcing the mechanistic rationale for substrate dependent selectivity in BPh₃-catalyzed CO₂ and epoxides coupling reaction.

3.4 BEt₃ catalyzed coupling reaction of CO₂ and epoxides

In their experimental study, Gnanou and coworkers utilized BEt₃ as a catalyst for the coupling of CO₂ and epoxides.³⁸ Unlike the BPh₃ system reported by Kerton and coworkers,⁴⁰ where substrate-dependent selectivity was observed, the BEt₃ system exclusively yielded poly-carbonate (PC) products from both propylene oxide (PO) and cyclohexene oxide (CHO). To investigate how substituting ethyl groups with phenyl groups at the boron center influences product selectivity, we extended our mechanistic study to BEt₃-catalyzed CO₂-epoxide coupling reactions. We first explored the mechanistic pathways of BEt₃-catalyzed CO₂ and PO coupling, and the corresponding Gibbs free energy profile is provided in Fig. 8. The reaction begins with the formation of the adduct Int1_c, followed by epoxide ring opening via the transition state TS1_c, where chloride ion attacks PO. This step proceeds with an activation barrier of 22.9 kcal mol⁻¹ (relative to BEt₃ + PO). The subsequent addition of CO₂ and desorption of one BEt₃ results in the formation of a stable intermediate Int2_c, where the alkoxide anion interacts with CO₂ through weak van der Waals forces. From Int2_c, coordination of an additional BEt₃ molecule to CO₂ in Int2_c generates Int3_c, which undergoes CO₂ insertion via a transition state TS2_c with an activation barrier of 12.4 kcal mol⁻¹ (relative to Int2_c) to yield intermediate Int4_c. The additional BEt₃ molecule stabilizes the negative charge on CO₂ in TS2_c through interaction with its Lewis acidic boron center, thereby lowering the energy barrier for CO₂ insertion step.

Following the formation of Int4_c, the reaction can proceed via two competing pathways. In the first pathway, dissociation of the weakly coordinated BEt₃ generates Int5_c, which undergoes intramolecular cyclization through a five-membered transition state TS3_c (ΔG^‡ = 26.0 kcal mol⁻¹, relative to Int5_c) to yield Int6_c (the CC product coordinated to BEt₃ with a regenerated Cl⁻ ion). In the alternative pathway, displacement of BEt₃ in Int4_c by an incoming BEt₃-activated PO molecule leads to the formation of Int7_c. This intermediate undergoes epoxide addition via transition state TS4_c, requiring an activation barrier of 19.0 kcal mol⁻¹ (relative to Int7_c). Here, nucleophilic attack by the alkoxide ion induces ring opening of the second epoxide, yielding Int8_c, the key intermediate for poly-carbonate propagation. Subsequent alternating CO₂ insertion and epoxide addition steps drive poly-carbonate formation. The BEt₃-catalyzed coupling of CO₂ and CHO follows a similar mechanism, and the corresponding Gibbs free energy profile is presented in Fig. S5 (see ESI† section S5 for more details).

3.5 Comparative evaluation of BPh₃ vs. BEt₃ catalytic systems

A comparison of the energy profiles for BPh₃- and BEt₃-catalyzed CO₂ and PO coupling reactions (Fig. 4 and 8) reveals a key difference in their reactivity. In the BPh₃ system, the pathway leading to CC formation (ΔG^‡ = 23.9 kcal mol⁻¹) is more favorable than the PC formation pathway (ΔG^‡ = 28.5 kcal mol⁻¹). However, the BEt₃ system exhibits the opposite trend, where the transition state TS3_c (ΔG^‡ = 26.0 kcal mol⁻¹) associated with CC formation has a 7 kcal mol⁻¹ higher energy barrier than TS4_c (ΔG^‡ = 19.0 kcal mol⁻¹), which leads to PC formation. To understand this reactivity difference, we performed DIA on the competing transition states TS3_c and TS4_c (Fig. 9). Although the interaction energy (ΔE_int) in TS3_c (−13.6 kcal mol⁻¹) is slightly higher than that in TS4_c (−12.0 kcal mol⁻¹), the total distortion energy (ΔE_dist) in TS3_c (38.9 kcal mol⁻¹) is significantly greater than in TS4_c (30.5 kcal mol⁻¹). This indicates that distortion energy plays a dominant role in BEt₃-catalyzed CO₂ and PO coupling, where the lower distortion energy of the epoxide addition transition state (TS4_c) makes it relatively more stable, favoring PC formation over CC formation. A further comparison of DIA results for the ring-opening and epoxide addition transition states in BPh₃ and BEt₃ systems (Fig. 6a and 9) reveals additional insights. The overall effect of ΔE_dist and ΔE_int on the activation barrier for the ring-opening transition states (TS3_a and TS3_c) is similar in both catalytic systems. However, in the epoxide addition transition states (TS4_a and TS4_c), ΔE_dist values are nearly identical (30.7 vs. 30.5 kcal mol⁻¹), whereas ΔE_int in the BEt₃ system (−12.0 kcal mol⁻¹) is significantly higher than in the BPh₃ system (−1.8 kcal mol⁻¹). This difference arises from steric factors. In TS4_a, the bulky phenyl groups in the BPh₃ catalysts lead to intermolecular steric repulsion, reducing effective attractive interaction between the ‘Cat’ and ‘Sub’ fragments. Such steric hindrance is absent in TS4_c due to small ethyl groups in BEt₃, allowing stronger interactions. These results suggest that substituting ethyl groups with bulky phenyl groups at the boron center reduces intermolecular interactions during the epoxide addition step. Consequently, this effect makes PC formation less favorable as compared to CC formation in the BPh₃ system, whereas in the BEt₃ system, stronger intermolecular interactions facilitate PC formation over CC formation.

As for the CO₂ and CHO coupling reactions, both BEt₃ and BPh₃ catalytic systems favour PC formation. A comparative DIA analysis of the ring-closing and epoxide addition transition states in both systems (Fig. 6b and S7†) reveals that in both cases, the high distortion energy during the ring-closing step disfavours CC formation, making PC formation the preferred pathway (see ESI† section S5 for more details). In summary, our detailed mechanistic study provides a clear explanation for the substrate-dependent selectivity observed in BPh₃-catalyzed CO₂-epoxide coupling reactions, while also demonstrating how substituting bulky phenyl groups with ethyl groups at the boron center eliminates this selectivity in BEt₃-catalyzed reactions.

Conclusions

In this study, we conducted a detailed mechanistic investigation into the triphenylborane (BPh₃)-catalyzed coupling of CO₂ with epoxides, aiming to uncover the origins of its substrate-dependent selectivity. While BPh₃ catalysis yielded cyclic-carbonate (CC) from propylene oxide (PO), it exclusively favoured poly-carbonate (PC) formation from cyclohexene oxide (CHO), highlighting a distinct selectivity pattern. To further understand this behaviour, we extended our study to triethylborane (BEt₃)-catalyzed CO₂-epoxide coupling, where both PO and CHO yielded PC products, demonstrating the absence of selectivity. This comparative evaluation provided critical insights into how boron substitution influences catalytic performance and product distribution.

Our results indicate that epoxide ring opening is the rate-determining step, aligning with experimental mechanistic studies. Additionally, we found that high CO₂ concentrations in the initial stage can lead to the formation of an inactive species that inhibits epoxide coordination and activation, thereby suppressing catalytic activity. This finding provides a mechanistic explanation for the experimentally observed inverse rate dependence on CO₂ concentration.

Through distortion/interaction analysis (DIA) and non-covalent interaction (NCI) analysis, we established that in the BPh₃-catalyzed CO₂ and PO system, the weaker intermolecular interactions in the epoxide addition transition state render PC formation less favorable, leading to a preference for CC formation. However, in the BEt₃-catalyzed CO₂ and PO system, stronger intermolecular interactions in the corresponding transition state stabilize PC formation, overriding the selectivity observed in BPh₃ catalysis. Moreover, for CO₂ and CHO coupling reactions, both BPh₃ and BEt₃ catalytic system exhibit high distortions in the ring-closing transition state, making CC formation kinetically unfavourable, thereby explaining the exclusive formation of PC in both cases.

This study demonstrates that substituting ethyl groups with phenyl groups at the boron center fundamentally alters catalytic behaviour, affecting intermolecular interactions, transition state stability, and overall reaction energetics. Our findings provide a mechanistic rationale for the substrate-dependent selectivity in BPh₃ catalysis while also offering valuable insights for the development of tailored boron-based catalysts for selective CO₂ utilization in both cyclic-carbonate and poly-carbonate synthesis.

Data availability

The data supporting this article have been included as part of the ESI.†

Author contributions

Nikunj Kumar: writing – original draft, visualization, methodology, investigation, formal analysis, data curation, conceptualization. Puneet Gupta: writing – review & editing, supervision, resources, project administration, funding acquisition, formal analysis.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

N. K. thanks CSIR for the SRF fellowship (09/143(1026)/2020-EMRI). P. G. acknowledges SERB (CRG/2021/000759) for financial support and National Supercomputing Mission (NSM) for providing computing resources of ‘PARAM Ganga’ at IIT Roorkee, which is implemented by C-DAC and supported by the Ministry of Electronics and Information Technology (MeitY) and Department of Science and Technology (DST), Government of India.

References

1. M. Alves, B. Grignard, R. Mereau, C. Jerome, T. Tassaing and C. Detrembleur, Catal. Sci. Technol., 2017, 7, 2651–2684 RSC .
2. S. Chu, Science, 2009, 325, 1599 CrossRef CAS PubMed .
3. F. Sher, S. Z. Iqbal, S. Albazzaz, U. Ali, D. A. Mortari and T. Rashid, Fuel, 2020, 282, 118506 CrossRef CAS .
4. K. H. S. M. Sampath, P. G. Ranjith and M. S. A. Perera, Energy Fuels, 2020, 34, 13369–13383 CrossRef CAS .
5. P. M. Cox, R. A. Betts, C. D. Jones, S. A. Spall and I. J. Totterdell, Nature, 2000, 408, 184–187 CrossRef CAS PubMed .
6. C. Maeda, Y. Miyazaki and T. Ema, Catal. Sci. Technol., 2014, 4, 1482–1497 RSC .
7. F. D. Meylan, V. Moreau and S. Erkman, J. CO2 Util., 2015, 12, 101–108 CrossRef CAS .
8. W. D. Jones, J. Am. Chem. Soc., 2020, 142, 4955–4957 CrossRef CAS PubMed .
9. J. Chen, G. Chiarioni, G. J. W. Euverink and P. P. Pescarmona, Green Chem., 2023, 25, 9744–9759 RSC .
10. W.-W. Yu, X.-G. Meng, Z.-Y. Gan, W. Li, Y.-L. Zhang and J. Zhou, Catal. Sci. Technol., 2024, 14, 6215–6223 RSC .
11. B. Schäffner, F. Schäffner, S. P. Verevkin and A. Börner, Chem. Rev., 2010, 110, 4554–4581 CrossRef PubMed .
12. K. Xu, Chem. Rev., 2004, 104, 4303–4417 CrossRef CAS PubMed .
13. V. Besse, F. Camara, C. Voirin, R. Auvergne, S. Caillol and B. Boutevin, Polym. Chem., 2013, 4, 4545–4561 RSC .
14. J. Wang, H. Zhang, Y. Miao, L. Qiao, X. Wang and F. Wang, Green Chem., 2016, 18, 524–530 RSC .
15. P. P. Pescarmona and M. Taherimehr, Catal. Sci. Technol., 2012, 2, 2169–2187 RSC .
16. I. Omae, Coord. Chem. Rev., 2012, 256, 1384–1405 CrossRef CAS .
17. W. Zhou, Q. W. Deng, G. Q. Ren, L. Sun, L. Yang, Y. M. Li, D. Zhai, Y. H. Zhou and W. Q. Deng, Nat. Commun., 2020, 11, 1–9 CrossRef PubMed .
18. Z. Li, Z. Su, W. Xu, Q. Shi, J. Yi, C. Bai, N. Wang and J. Li, ChemCatChem, 2020, 12, 4839–4844 CrossRef CAS .
19. R. R. Shaikh, S. Pornpraprom and V. D'Elia, ACS Catal., 2018, 8, 419–450 CrossRef CAS .
20. A. I. Adeleye, D. Patel, D. Niyogi and B. Saha, Ind. Eng. Chem. Res., 2014, 53, 18647–18657 CrossRef CAS .
21. W. Natongchai, S. Posada-Pérez, C. Phungpanya, J. A. Luque-Urrutia, M. Solà, V. D'Elia and A. Poater, J. Org. Chem., 2022, 87, 2873–2886 CrossRef CAS PubMed .
22. F. N. Al-Rowaili, U. Zahid, S. Onaizi, M. Khaled, A. Jamal and E. M. AL-Mutairi, J. CO2 Util., 2021, 53, 101715 CrossRef CAS .
23. S. Inoue, H. Koinuma and T. Tsuruta, Makromol. Chem., 1969, 130, 210–220 CrossRef CAS .
24. S. Inoue, H. Koinuma and T. Tsuruta, J. Polym. Sci., Part B: Polym. Lett., 1969, 7, 287–292 CrossRef CAS .
25. S. D. Allen, D. R. Moore, E. B. Lobkovsky and G. W. Coates, J. Am. Chem. Soc., 2002, 124, 14284–14285 CrossRef CAS PubMed .
26. D. J. Darensbourg, J. C. Yarbrough, C. Ortiz and C. C. Fang, J. Am. Chem. Soc., 2003, 125, 7586–7591 CrossRef CAS PubMed .
27. Z. Qin, C. M. Thomas, S. Lee and G. W. Coates, Angew. Chem., Int. Ed., 2003, 42, 5484–5487 CrossRef CAS PubMed .
28. A. C. Deacy, E. Moreby, A. Phanopoulos and C. K. Williams, J. Am. Chem. Soc., 2020, 142, 19150–19160 CrossRef CAS PubMed .
29. C. Chatterjee and M. H. Chisholm, Inorg. Chem., 2011, 50, 4481–4492 CrossRef CAS PubMed .
30. M. R. Kember and C. K. Williams, J. Am. Chem. Soc., 2012, 134, 15676–15679 CrossRef CAS PubMed .
31. R. M. Izatt, S. R. Izatt, R. L. Bruening, N. E. Izatt and B. A. Moyer, Chem. Soc. Rev., 2014, 43, 2451–2475 RSC .
32. M. Cokoja, M. E. Wilhelm, M. H. Anthofer, W. A. Herrmann and F. E. Kühn, ChemSusChem, 2015, 8, 2436–2454 CrossRef CAS PubMed .
33. Y. Ge, G. Cheng and H. Ke, J. CO2 Util., 2022, 57, 101873 CrossRef CAS .
34. J. Wang and Y. Zhang, ACS Catal., 2016, 6, 4871–4876 CrossRef CAS .
35. Y. Luo, F. Chen, H. Zhang, J. Liu and N. Liu, J. Org. Chem., 2023, 88, 15717–15725 CrossRef CAS PubMed .
36. L. Y. Zhao, J. Y. Chen, W. C. Li and A. H. Lu, J. CO2 Util., 2019, 29, 172–178 CrossRef CAS .
37. N. Kumar and P. Gupta, J. Catal., 2024, 439, 115787 CrossRef CAS .
38. D. Zhang, S. K. Boopathi, N. Hadjichristidis, Y. Gnanou and X. Feng, J. Am. Chem. Soc., 2016, 138, 11117–11120 CrossRef CAS PubMed .
39. Y. Wang, J. Y. Zhang, J. L. Yang, H. K. Zhang, J. Kiriratnikom, C. J. Zhang, K. L. Chen, X. H. Cao, L. F. Hu, X. H. Zhang and B. Z. Tang, Macromolecules, 2021, 54, 2178–2186 CrossRef CAS .
40. K. A. Andrea and F. M. Kerton, ACS Catal., 2019, 9, 1799–1809 CrossRef CAS .
41. F. Neese, F. Wennmohs, U. Becker and C. Riplinger, J. Chem. Phys., 2020, 152, 224108 CrossRef CAS PubMed .
42. F. Neese, Wiley Interdiscip. Rev.: Comput. Mol. Sci., 2022, 12, 1–15 Search PubMed .
43. R. Valero, R. Costa, I. D. P. R. Moreira, D. G. Truhlar and F. Illas, J. Chem. Phys., 2008, 128, 114103 CrossRef PubMed .
44. Y. Zhao and D. G. Truhlar, Acc. Chem. Res., 2008, 41, 157–167 CrossRef CAS PubMed .
45. F. Weigend and R. Ahlrichs, Phys. Chem. Chem. Phys., 2005, 7, 3297–3305 RSC .
46. G. W. Yang, Y. Y. Zhang, R. Xie and G. P. Wu, J. Am. Chem. Soc., 2020, 142, 12245–12255 CrossRef CAS PubMed .
47. Y. Y. Zhang, G. W. Yang, R. Xie, X. F. Zhu and G. P. Wu, J. Am. Chem. Soc., 2022, 144, 19896–19909 CrossRef CAS PubMed .
48. J. Wang, Y. Zhu, M. Li, Y. Wang, X. Wang and Y. Tao, Angew. Chem., Int. Ed., 2022, 61, e202208525 CrossRef CAS PubMed .
49. F. Neese, F. Wennmohs, A. Hansen and U. Becker, Chem. Phys., 2009, 356, 98–109 CrossRef CAS .
50. H. P. Hratchian and H. B. Schlegel, J. Chem. Theory Comput., 2005, 1, 61–69 CrossRef CAS PubMed .
51. T. Lu and F. Chen, J. Comput. Chem., 2012, 33, 580–592 CrossRef CAS PubMed .
52. W. Humphrey, A. Dalke and K. Schulten, J. Mol. Graphics, 1996, 14, 33–38 CrossRef CAS PubMed .
53. C. Y. Legault, CYLview, ver. 1.0b, Université de Sherbrooke, QC, 2009, http://www.cylview.org Search PubMed .
54. R. Babu, R. Roshan, Y. Gim, Y. H. Jang, J. F. Kurisingal, D. W. Kim and D. W. Park, J. Mater. Chem. A, 2017, 5, 15961–15969 RSC .
55. C. J. Zhang, S. Q. Wu, S. Boopathi, X. H. Zhang, X. Hong, Y. Gnanou and X. S. Feng, ACS Sustainable Chem. Eng., 2020, 8, 13056–13063 CrossRef CAS .
56. J. Tang, M. Li, X. Wang and Y. Tao, Angew. Chem., Int. Ed., 2022, 61, 1–8 Search PubMed .
57. X. Hong, Y. Liang and K. N. Houk, J. Am. Chem. Soc., 2014, 136, 2017–2025 CrossRef CAS PubMed .
58. F. M. Bickelhaupt and K. N. Houk, Angew. Chem., Int. Ed., 2017, 56, 10070–10086 CrossRef CAS PubMed .
59. N. Goswami, N. Kumar, S. Bag, P. Gupta and D. Maiti, ACS Catal., 2023, 13, 11091–11103 CrossRef CAS .
60. E. R. Johnson, S. Keinan, P. Mori-Sánchez, J. Contreras-García, A. J. Cohen and W. Yang, J. Am. Chem. Soc., 2010, 132, 6498–6506 CrossRef CAS PubMed .

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5cy00281h

This journal is © The Royal Society of Chemistry 2025