Tetra- and penta-coordinated aluminum complexes with bis-amidine ligands and their reactivity for efficient CO₂ transformation to cyclic carbonates

Abstract

The development of efficient catalysts for CO₂ transformation into value-added products is crucial for sustainable chemistry. In this work, we report the synthesis and characterization of a new series of tetra- and penta-coordinated aluminum complexes supported by bis-amidine ligands featuring a rigid naphthalene bridge and their application as catalysts in the reaction of carbon dioxide with epoxides to form cyclic carbonates. Interestingly, while tetra- and penta-coordinated aluminum complexes bearing methyl groups are efficient when used with TBAI as a co-catalyst, their iodo analogs proved to be inefficient in mono-component catalysis, requiring the use of additional TBAI. Computational results of the proposed systems reveal the necessity of achieving a delicate balance between the nucleophilic properties of iodide atoms and the steric configuration of the complexes. Experimental and computational studies offer valuable insights into the role of ligand structure in aluminum-based catalysts, highlighting the potential of main-group metals in CO₂ fixation.

---

Introduction

Research in the field of coordination chemistry has contributed enormously to the current development of transition metal chemistry, which plays a crucial role in homogeneous catalysis. Indeed, the properties of transition metal catalysts can be carefully tuned through ligand design and metal selection, making them applicable in a wide variety of fields. Thus, transition metal complexes can act as catalysts to accelerate reactions, activate substrates, and/or reduce the energy costs of many industrial processes (olefin metathesis, hydroformylation, polymerization processes, etc.) and as catalysts to directly convert CO₂ into high value-added chemicals.^1–3 Utilizing CO₂ as a raw material for organic compound production adds value and could help offset the costs of its capture and storage. Additionally, it offers an alternative to reduce dependence on fossil fuels for chemical production.^4,5 In this context, the formation of cyclic carbonates from epoxides and carbon dioxide has emerged as one of the most extensively studied methods for carbon dioxide fixation in recent years. On the other hand, cyclic carbonates are valuable as solvents for electrolytes in lithium-ion batteries, as polar aprotic solvents, and as intermediates in the synthesis of different small molecules and polymers.^6,7 These cyclic carbonates can be synthesized through reactions between epoxides and CO₂, using several organometallic complexes as catalysts (Scheme 1). In these catalytic processes, systems combining Lewis acids and nucleophiles are required. The metal center serves as the Lewis acid, activating the epoxide, while a cocatalyst provides a halide that acts as a nucleophile, inducing ring-opening.⁸

Scheme 1 Synthesis of cyclic carbonates from epoxides and CO₂.

As an alternative to transition metals, nowadays the chemistry of main-group element complexes, capable of mimicking transition metal behavior, is experiencing significant development. In particular, recent research has demonstrated that aluminum derivatives, the third most abundant element in the Earth's crust with low toxicity and low production costs, behave quite similarly to transition metal complexes and are able to activate strong bonds in small molecules (H₂, N₂, CO, and CO₂).^9,10

Since the choice of ligands is essential in each organometallic process, amidinate ligands, which are endlessly versatile in their potential structure and substitution patterns, have attracted attention for the stabilization of aluminum species.¹¹ Indeed, they present high basicity with a pK_a value of approximately 12 and therefore exhibit good affinity for protons and Lewis acids.^12,13 Amidinate aluminum complexes have shown promising catalytic activity in various processes, such as ring-opening polymerization, alkyne hydroboration and the conversion of CO₂ to cyclic carbonates.^11k,l,n

In recent years, our research group has reported a series of aluminum amidinate complexes as catalysts for synthesizing cyclic carbonates from CO₂ and epoxides.^14–16 The first symmetric aluminum bimetallic complex I (a double aluminum amidinate, see Fig. 1) proved to be an excellent catalyst (conversions between 62 and 100% in 24 h) in the presence of tetrabutylammonium iodide (TBAI) as a co-catalyst.¹⁴ Bi- and tri-metallic aluminum-amidinate complexes were also reported (Fig. 1, II and III, respectively) as efficient catalysts for the formation of cyclic carbonates from terminal epoxides and CO₂ at 50 °C under 1 bar of CO₂ pressure without solvent.¹⁵ In 2021, a pentacoordinate bisamidinate iodide aluminum complex IV (Fig. 1) was reported, which is one of the first examples of a neutral non-zwitterionic aluminum complex used as a single-component catalyst for the synthesis of carbonates (catalytic activity at 80 °C and 1 bar of CO₂, without a co-catalyst).¹⁶

Fig. 1 Double aluminum amidinate, bi- and tri-aluminum amidinate and pentacoordinate bis-amidinate complexes described by our group.

As part of our ongoing interest in designing new aluminum complexes for catalytic applications, this work reports the synthesis and structural characterization of a new series of tetra- and penta-coordinated aluminum complexes with amidinate ligands featuring a rigid naphthalene bridge (1–7, Fig. 1). The performance and reactivity of these complexes were evaluated as catalysts in the reaction of carbon dioxide with epoxides to form cyclic carbonates. The influence of the substitution pattern on the amidinate ligand was explored to understand how these variations affect the catalytic activity and selectivity of the corresponding aluminum complexes.

Results and discussion

Synthesis of aluminum complexes

Aluminum complexes 1–4 were synthesized through protonolysis reactions between the corresponding bis(amidine) ligands (L₁H₂ or L₂H₂) and one or two equivalents of Al(CH₃)₃ (Scheme 2). These reactions were easily monitored by gas bubble evolution (CH₄). Complex 1 was obtained through the reaction of one equivalent of L₁H₂ and one equivalent of Al(CH₃)₃ in dry benzene at 60 °C for 2 hours because of the high steric hindrance. Complex 1 was isolated as a yellow solid in 93% yield. The formation of a pentacoordinate aluminum complex was established by NMR spectroscopy. The ¹H NMR spectrum showed the disappearance of the resonance corresponding to the N–H groups of ligand L₁H₂. Additionally, the methyl group coordinated to the aluminum atom is easily identified in ¹H and ¹³C NMR spectra, each exhibiting a characteristic signal in the negative region, respectively, at −0.34 ppm and −4.28 ppm, in good agreement with previously reported pentacoordinate aluminum complexes.¹⁶ The ¹H and ¹³C NMR resonance patterns support the existence of a plane of symmetry in this complex. On the other hand, L₂H₂ reacts much faster and smoothly with one equivalent of Al(CH₃)₃ in CH₂Cl₂ at room temperature, affording the pentacoordinate aluminum complex 2 as a yellow solid in 90% yield within only one hour. Similarly, the methyl group coordinated to the central aluminum atom appears at −0.89 ppm and −3.99 ppm in the respective ¹H and ¹³C NMR spectra.

Scheme 2 Synthesis of penta-coordinated aluminum (1 and 2) and tetra-coordinated bis-aluminum (3 and 4) complexes.

The bis-aluminum complex 3 was synthesized using 1 equivalent of L₁H₂ and two equivalents of Al(CH₃)₃ in dry toluene at room temperature for 1 hour. Complex 3 was obtained as orange crystals in 83% yield. The formation of the corresponding bis-aluminum complexes was established by NMR spectroscopy and corroborated by XRD analysis (Fig. 2). The ¹H NMR spectrum exhibited two signals in the negative region at −0.02 and −0.33 ppm, each integrating for 6 H, indicating that each aluminum atom is tetracoordinated with one amidinate ligand and two CH₃ groups in its structure. In the ¹³C NMR spectrum, the signals of the methyl groups coordinated to the aluminum atoms appeared at −5.41 and −9.16 ppm.

Fig. 2 Molecular structure of 3 (hydrogens are omitted for clarity). Thermal ellipsoids are shown at 50% probability. Selected bond lengths (Å) and angles [°]: C16–Al1 1.948(4), C17–Al1 1.956(4), Al1–N2 1.942(3), Al1–N1 1.948(3), Al2–N3 1.942(3), Al2–N4 1.954(3), C11–N2 1.333(4), C11–N1 1.375(4), C26–N4 1.335(4), C26–N3 1.369(4), C31–Al2 1.954(4), C32–Al2 1.949(4); N2–C11–N1 108.5(3), N4–C26–N3 108.7(3).

Finally, the analogous aluminum complex 4 was obtained when L₂H₂ is mixed with two equivalents of Al(CH₃)₃. The NMR data are very similar to those of complex 3 with signals at −0.33 and −0.75 ppm that integrate for 6 H in the ¹H NMR spectrum.

To gain more insight into these electronic structures, quantum chemical calculations were performed. Each molecular structure was optimized by employing density functional theory (DFT) calculations at the ωB97XD/def2-TZVP level of theory. Geometric indexes, such as τ₅ and τ₄, were computed at the DFT level. Molecular electrostatic potentials (MEPs) were calculated for all systems and are shown in Fig. 3. At first glance, this analysis underscores notable variations in electronic distribution that can be attributed to the effects of substitution either at the aluminum center or at the ligand.

Fig. 3 Molecular electrostatic potential (MEP) for all complexes (isosurface 0.01 au). NPA charges (q_i) for selected atoms are displayed.

The DFT-optimized molecular structures of complexes 1 and 2 confirmed a pentacoordinate Al center with τ₅ indexes¹⁷ of 0.35 and 0.28, suggesting a distorted square pyramidal geometry at the Al centers. The computed equilibrium Al–C distances (bond orders) were 1.967 Å (0.56) and 1.965 Å (0.58), for complexes 1 and 2, respectively. Fig. 3 reveals that the Al atom in complex 1 exhibits a depletion of the electron density shown as a more positive atomic charge, while the C atom exhibits a more pronounced negative charge compared to complex 2. This observation is further supported by the comparative analysis of bond distances, bond orders, and atomic charges. Specifically, the longer bond distance and lower bond order, coupled with the greater charge separation of the Al–C bond in complex 1, unequivocally indicate a more polar Al–C covalent bond in this system compared to complex 2. Besides the small differences observed, these indexes still highlight the effect of the different R groups in both ligands. The τ₄ index¹⁸ computed from the XRD and DFT structures was 0.95 and 0.94, which supports the use of the DFT structures for structural comparison. However, for a fair comparison, only the DFT-derived indexes are discussed. The τ₄ indexes computed from the DFT structures were found to be 0.95 and 0.96 for complexes 3 and 4, respectively, confirming the distorted tetrahedral structure at Al. Interestingly, the DFT structures of the aluminum complexes reported slightly different Al–C bond distances for each Al center in complex 3 of 1.963 Å and 1.957 Å, respectively. However, the respective bond order remains similar for both distances around 0.67. The indistinguishable bond distances of N–C in the NCN unit and almost equal N–Al bond distances proved the formation of a delocalized system.

In a previously reported article, we have established that THE incorporation of a nucleophile (iodide) at the aluminum center could be an interesting strategy to obtain a single-component catalyst (avoiding the use of ammonium iodide as a cocatalyst). Indeed, the reaction mechanism has been shown to require an epoxide–iodide exchange at the aluminum atom, with an activation free energy of ∼20 kcal mol⁻¹ for styrene oxide.¹⁶

For the pentacoordinate aluminum complexes, the formation of the respective iodide derivatives was achieved by reacting the corresponding complex (1 or 2) with one equivalent of I₂ using a method previously described (Scheme 3).

Scheme 3 Synthesis of complexes 5 and 6.

Compounds 5 and 6 were obtained as brown solids in 95% and 96% yields, respectively. The products were fully characterized by NMR spectroscopy. These reactions were easily monitored by ¹H NMR, with the disappearance of the signal corresponding to the methyl group at the aluminum center.

The NMR patterns of both iodo-complexes 5 and 6 again indicate molecular symmetry.

The optimized DFT structures of iodine-containing complexes (complexes 5 and 6) show Al–I bond distances of 2.565 Å (a bond order of 0.82) for complex 5 and a slightly shorter bond length of 2.547 Å (a bond order of 0.85) for complex 6. Notably, the NBO charges at the iodine centers remain consistent, with only minor variations in the electron population at the aluminum center. This significant electron density difference between Al and I suggests a polarized covalent bond, an essential characteristic for their function as single-component catalysts (vide infra). The MEP of complexes 5 and 6 (Fig. 3) indicates charge accumulation on the iodine atom, with delocalization across the ligand backbone. This delocalization is more pronounced in complex 6, extending between the naphthalene and phenyl groups due to the complex's relatively planar conformation.

In the case of the bis-aluminum complexes 3 and 4, attempts to replace the methyl groups with iodide using I₂ with the present procedure were unsuccessful in our hands, resulting in mixtures of products that were difficult to identify. As an alternative synthetic route, BI₃ was used as a source of iodides to carry out the substitution. The reaction between 3 and 2 equivalents of BI₃ was carried out in benzene at room temperature for 1 hour (Scheme 4), resulting in the isolation of compound 7 as yellow crystals (91% yield). As before, the evident disappearance of signals in the negative region of the ¹H NMR spectrum was a good indicator of reaction completion. Complex 7 presents an extremely low solubility in common solvents, preventing us from acquiring the corresponding ¹³C NMR spectrum. However, the formation of the bis-(iodo)complex was established by DRX analysis (Fig. 4), which revealed that the aluminum atoms exhibit a distorted tetrahedral geometry.

Scheme 4 Synthesis of 7.

Fig. 4 Molecular structure of 7 (hydrogens are omitted for clarity). Thermal ellipsoids are shown at 50% probability. Selected bond lengths (Å) and angles [°]: Al1–N1 1.894(2), Al1–N2 1.906(2), Al1–I2 2.4847(9), Al1–I1 2.4949(9), C11–N2 1.331(3), C11–N1 1.371(3), Al2–N3 1.901(2), Al2–N4 1.905(2), Al2–I3 2.4987(8), Al2–I4 2.4996(8), C24–N4 1.330(3), C24–N3 1.385(3); N2–C11–N1 108.8(2), N4–C24–N3 108.8(2).

For the bimetallic complex 7, the MEP surface reveals a distinctive electronic distribution influenced by the presence of two Al–I bonds. By comparing the atomic charges at the Al and I centers, we gain insights into the polarity of these bonds.¹⁹ The results show that the Al–I bonds in complex 7 are approximately 20% less polar than those in complexes 5 and 6. Consequently, this reduced polarity of the two covalent bonds in complex 7 may suggest lower reactivity, limiting its efficacy as a single-component catalyst for cyclic carbonate formation.

Surprisingly, the replacement of methyl groups in complex 4 with iodides to form the halogenated complex led to a thermally unstable species. Indeed, at low temperature (−78 °C), the ¹H spectrum indicates that the reaction proceeds similarly to the previously reported complexes 1–3. Above this temperature, decomposition occurred, making it impossible to isolate the iodocomplex.

Catalytic study for the synthesis of cyclic carbonates

With complexes 1–7 in hand, we explored their use as catalysts for the preparation of a range of cyclic carbonates (9a–h) from their corresponding epoxides (8a–h) and CO₂ under solvent-free conditions. The results of the catalytic runs are shown in Fig. 5. The reactions were carried out at 80 °C under 1 bar of CO₂ pressure for 24 h employing a combination of 1.5 mol% of complexes 1 or 2 and 1.5 mol% of TBAI as a cocatalyst and 0.75 mol% of complexes 3 or 4 with 1.5 mol% of TBAI for dual component systems. For one-component systems, only 1.5 or 0.75 mol% of catalysts 5, 6 or 7 were used. As expected, under these reaction conditions, no polycarbonates were obtained. Catalysts 1–4 associated with TBAI achieved moderate to excellent yields (50–97%) for the preparation of cyclic carbonates 9a–9h (Fig. 5).^8,11,14–16 No clear trend can be identified regarding the effectiveness of these systems based on the electronic or steric effects triggered by the R substituent in the starting epoxide, demonstrating that these systems exhibit good efficiency and broad scope.

Fig. 5 Conversion of cyclic carbonates 9a–h employing catalysts 1–7. ^a Determined by ¹H NMR spectroscopy of the crude reaction mixture.

Generally, complexes 3 and 4 exhibited slightly better efficiency than complexes 1 and 2. For the mono-component catalysts 5, 6, and 7, the results were disappointing. Indeed, no catalytic activity could be detected for catalysts 5 and 7, and extremely low conversions were observed with catalyst 6 for three substrates (8, 3 and 9% conversion for substrates 9a, 9e, 9g, respectively). Specifically, the absence or the poor catalytic activity of compounds 5, 6 and 7 can be attributed to the poor polarity of the Al–I bonds as shown by the MEP surface (Fig. 3) that probably prevents (or limits) the liberation of the iodide atom. The X-ray data obtained in the case of complex 7 corroborate this hypothesis, with Al–I bond lengths rather short (between 2.4847 and 2.4996 Å) when compared to the one observed in a similar active one-component system (2.553 Å) with a similar ligand reported by our research group.¹⁶

To further understand the absence of reactivity of 5 and the limited reactivity of 6, we examined the geometries of these complexes by DFT when functioning as single-component catalysts. A critical comparison with the reactive system IV (Fig. 1) suggests that the pivotal step in these new complexes is the initial stage of the mechanism, involving the epoxide–iodide exchange at the aluminum center. The transition-state structure (TS_IV+a) of complex IV during this exchange with styrene oxide (a) is depicted in Fig. 6. Concurrently, steric maps illustrating the computed free volume in the frontal regions of systems 5 and 6 are also presented in Fig. 6.

Fig. 6 Illustration of the transition state structure for the reaction between complex IV and styrene oxide (a). Key methyl groups in the ligand are highlighted in red, while the epoxide is rendered transparent for enhanced clarity (inset box). Steric maps of complexes 5 and 6 are presented, showing the computed free volume percentage in the frontal region of the complexes. Arrows indicate the entry pathway of the epoxide and the exit route of iodide atoms.

It is important to note that complexes IV and 5 differ only in the R substituent linked to the amidine moiety, with methyl in IV and ^tbutyl in complex 6. Although this substitution might seem minor, the bulkier ^tbutyl substituent in complex 5 significantly hampers its reactivity as a single-component catalyst. The steric maps in Fig. 7 further detail the approximate entry channel for the epoxide and the exit channel for the iodide atom, consistent with the transition state structure shown on the left. As illustrated, the ^tbutyl group on the amidine moiety, along with the methyl groups on the phenyl rings, obstructs both the entry channel for the epoxide and the exit channel for the iodide, thereby inhibiting the reaction mechanism.

Fig. 7 Conversion of cyclic carbonates 9a–h employing catalysts 5, 6, and 7. ^a Determined by ¹H NMR spectroscopy of the crude reaction mixture.

Conversely, the low reactivity of complex 6 towards the selected epoxide can be attributed to more open entry and exit channels, coupled with a larger free volume. This is achieved by retaining the bulky ^tbutyl groups while positioning the methyl group on the phenyl rings at the para position. Computational analysis of the systems suggests that achieving an optimal balance between the nucleophilic nature of the iodide atoms and the steric demands of the complexes, which facilitates favorable entry and exit channels for the exchange of epoxide and iodide atoms, is crucial for the future design of single-component catalysts.

To investigate whether the lack of catalytic activity of complexes 5, 6, and 7 as one-component systems is due to the non-dissociation of the iodide ligand, we employed these complexes in the presence of TBAI as a cocatalyst, keeping all the other experimental conditions unchanged.

The conversions are depicted in Fig. 7, demonstrating that the simple addition of TBAI as a cocatalyst leads to an active system for every studied substrate. The efficiency of the mono- and bimetallic catalysts is very similar for every substrate. It is worth noting that the catalytic activity of iodoaluminum complexes is slightly lower than that of their methylated analogs.

These new catalytic systems show interesting reactivities, comparable to those previously described in the literature.^{11k–n,14–16} The recyclability of the catalyst, due to the type of catalysis (homogeneous catalysis) and the experimental conditions (without solvent, mol%…), is not possible.

Conclusions

In this study, we synthesized and characterized a series of tetra- and penta-coordinated aluminum complexes with bis-amidine ligands and evaluated their catalytic performance in the conversion of CO₂ and epoxides into cyclic carbonates. The results highlight the significant influence of ligand structure on the catalytic efficiency and selectivity of these aluminum complexes. While tetra- and penta-coordinated aluminum complexes 1–4 bearing methyl groups are efficient when used with TBAI as a co-catalyst, their iodo analogs 5–7 proved to be inefficient in mono-component catalysis, requiring the use of additional TBAI. Computational results of the proposed systems reveal the necessity of achieving a delicate balance between the nucleophilic properties of iodide atoms and the steric configuration of the complexes, which is pivotal for facilitating effective entry and exit channels in the design of future single-component catalysts. This study underscores the potential of aluminum-based catalysts as sustainable alternatives for CO₂ fixation. Future research should focus on further ligand modifications (electronic and steric) to enhance the catalyst reactivity and obtain efficient single-component catalysts for cyclic carbonate synthesis.

Experimental

General information

All manipulations were performed under an inert atmosphere of argon or nitrogen using standard Schlenk line and glovebox techniques. Dry, oxygen-free solvents were employed. Reagents were obtained from commercial suppliers unless otherwise stated. N,N′-(naphthalene-1,8-diyl)bis(2,2-dimethylpropanamide),²⁰N,N′(naphthalene-1,8-diyl)bis(2,2-dimethylpropanimidoyl chloride),²⁰ L₁H₂,²¹ and L₂H₂ ²¹ were synthesized following reported procedures. 1D and 2D NMR spectra were recorded with the following spectrometers for ¹H and ¹³C: Bruker Avance II 300 MHz, Avance III HD 400 MHz, and Avance I and II 500 MHz spectrometers. The chemical shifts were counted positively relative to the low field and expressed in parts per million (ppm). Mass spectrometric analyses were performed using a Maldi micro MX micro-mass spectrometer with a pyrene matrix (ratio of product/matrix : 1/100). Melting points were measured with a capillary electrothermal Stuart SMP40 apparatus, and samples were prepared in the glovebox before the analysis. Single-crystal X-ray data were collected at low temperature (193(2) K) on a Bruker D8 VENTURE diffractometer equipped with a PHOTON III detector and using MoK_α radiation (λ = 0.71037 Å). Phi and omega scans were performed for data collection, and an empirical absorption correction was applied.²² The structures were solved by the intrinsic phasing method (ShelXT)²³ and refined by the full-matrix least-squares method on F².²⁴ All non-hydrogen atoms were refined with anisotropic displacement parameters. All hydrogen atoms were refined isotropically at calculated positions using a riding model, except for those of the water molecules in the structure of compound 3, which were not located in the difference Fourier maps. For 3, some electron density was difficult to model; therefore, the SQUEEZE tool²⁵ of PLATON was used to remove the corresponding electron density from the intensity data.

Deposition numbers 2477633 (3) and 2477634 (7) contain the supplementary crystallographic data for this paper.

Complexes 1–7 were fully optimized at the ωB97XD/def2-TZVP level of theory, with the resulting structures confirmed by frequency calculations at the same level. Population and bond order analyses were conducted using the NBO3.1 program, interfaced with Gaussian16 Rev B.01. Bond orders were specifically calculated using the Mayer definition, recommended as a suitable method for characterizing bonds in metal complexes due to its incorporation of the overlap matrix in the bond order determination.²⁶ Molecular electrostatic potentials were computed for all systems at the same level of theory.

Synthesis of complex 1. Al(CH₃)₃ (8.12 mg, 0.11 mmol) was added to a solution of L₁H₂ (60 mg, 0.11 mmol) in dry benzene. The mixture was stirred for 2 h at 60 °C. After the solvent was removed, the solid residue was washed with pentane. A yellow solid was obtained (59.5 mg, 0.10 mmol, 93%). Melting point: 206 °C. ¹H NMR (C₆D₆, 400 MHz): δ 7.42 (t, J_HH = 8.2 Hz, 4H, C₁₀H₆), 7.22 (t, J_HH = 7.8 Hz, 2H, C₁₀H₆), 6.86 (s, 6H, C₆H₃), 2.08 (s, 12H, CH₃), 1.06 (s, 18H, C(CH₃)₃), −0.34 (s, 3H, CH₃–Al). ¹³C{¹H} NMR (C₆D₆, 100 MHz): δ 181.6 (NCN), 144.8 (C₆H_3ipso), 142.2 (C₁₀H_6ipso), 137.4 (C₁₀H_6ipso), 133.8 (C₆H_3ipso), 128.3 (C₆H₃), 126.5 (C₁₀H_6ipso), 125.2 (C₁₀H₆), 124.6 (C₆H₃), 123.4 (C₁₀H₆), 121.0 (C₁₀H₆), 39.9 (C̲(CH₃)₃), 29.5 (C(C̲H₃)₃), 19.7 (CH₃), −4.3 (CH₃–Al).

Synthesis of complex 2. Al(CH₃)₃ (28.6 mg, 0.39 mmol) was added to a solution of L₂H₂ (200 mg, 0.39 mmol) in dry CH₂Cl₂. The mixture was stirred for 1 h at room temperature. After the solvent was removed, the solid residue was washed with pentane. A yellow solid was obtained (192 mg, 0.35 mmol, 90%). Melting point: 238 °C. ¹H NMR (CDCl₃, 400 MHz): δ 7.42 (d, J_HH = 8.2 Hz, 2H, C₁₀H₆), 7.30–7.24 (m, 2H, C₁₀H₆), 7.18 (d, J_HH = 7.4 Hz, 2H, C₁₀H₆), 7.00 (d, J_HH = 7.9 Hz, 4H, C₆H₄), 6.81 (d, J_HH = 7.9 Hz, 4H, C₆H₄), 2.31 (s, 6H, CH₃), 1.34 (s, 18H, C(CH₃)₃), −0.89 (s, 3H, CH₃–Al). ¹³C{¹H} NMR (CDCl₃, 100 MHz): δ 181.0 (NCN), 143.7 (C₆H_4ipso), 142.7 (C₁₀H_6ipso), 136.8 (C₁₀H_6ipso), 132.3 (C₆H_4ipso), 129.3 (C₆H₄), 125.2 (C₁₀H₆), 124.6 (C₆H₄), 123.0 (C₁₀H₆), 122.5 (C₁₀H_6ipso), 120.3 (C₁₀H₆), 40.5 (C̲(CH₃)₃), 31.1 (C(C̲H₃)₃), 21.0 (CH₃), −4.0 (CH₃–Al).

Synthesis of complex 3. A 2 M solution of Al(CH₃)₃ in heptane (1.9 mL, 3.8 mmol) was added to a solution of L₁H₂ (1.0 g, 1.9 mmol) in toluene (20 mL) in a Schlenk flask. The mixture was stirred for 2 h at room temperature. After the solvent was removed, the solid residue was washed with pentane (5 × 10 mL). An orange solid was recrystallized from pentane at −30 °C. Orange crystals were obtained (1.07 g, 1.7 mmol, 83%). Melting point: 200 °C. ¹H NMR (C₆D₆, 500 MHz): δ 7.44 (dd, J_HH = 8.1, 1.0 Hz, 2H, C₁₀H₆), 7.18–7.16 (overlapped, 2H, C₁₀H₆), 7.11–7.07 (m, 2H, C₁₀H₆), 7.00–6.91 (m, 6H, C₆H₃), 2.55 (s, 6H, CH₃), 2.52 (s, 6H, CH₃), 0.95 (s, 9H, C(CH₃)₃), −0.02 (s, 6H, CH₃–Al), −0.33 (s, 6H, CH₃–Al). ¹³C{¹H} NMR (C₆D₆, 125 MHz): δ 186.9 (NCN), 144.6 (C₁₀H_6ipso), 142.4 (C₆H_3ipso), 137.9 (C₁₀H_6ipso), 133.9 (C₆H_3ipso), 132.8 (C₆H_3ipso), 129.0 (C₆H₃), 128.7 (C₆H₃), 127.1 (C₁₀H_6ipso), 126.6 (C₁₀H₆), 126.6 (C₁₀H₆), 125.7 (C₁₀H₆), 125.1 (C₁₀H₆), 41.3 (C̲(CH₃)₃), 29.8 (C(C̲H₃)₃), 20.7 (CH₃), 20.5 (CH₃), −5.4 (CH₃–Al), −9.2 (CH₃–Al). MS (Maldi-TOF) m/z: 643.21 ([M]⁺), 629.17 ([M − CH₃]⁺), 615.16 ([M − 2(CH₃)]⁺).

Synthesis of complex 4. A 2 M solution of Al(CH₃)₃ in heptane (0.6 mL, 1.2 mmol) was added to a solution of L₃H (300 mg, 0.60 mmol) in toluene (6 mL) in a Schlenk flask. The mixture was stirred for 2 h at room temperature. After the solvent was removed, the solid residue was washed with pentane (5 × 10 mL). A yellow solid was obtained (295.0 mg, 0.48 mmol, 79%). Melting point: 221 °C. ¹H NMR (CDCl₃, 300 MHz): δ 7.63 (d, J_HH = 8.2 Hz, 2H, C₁₀H₆), 7.32 (t, J_HH = 7.7 Hz, 2H, C₁₀H₆), 7.16 (s, 8H, C₆H₄), 6.99 (d, J_HH = 7.3 Hz, 2H, C₁₀H₆), 2.36 (s, 6H, CH₃), 0.89 (s, 18H, C(CH₃)₃), −0.33 (s, 6H, CH₃–Al), −0.75 (s, 6H, CH₃–Al). ¹³C{¹H} NMR (C₆D₆, 125 MHz): δ 183.9 (NCN), 143.7 (C₁₀H_6ipso), 142.7 (C₆H_4ipso), 137.1 (C₁₀H_6ipso), 134.7 (C₆H_4ipso), 129.9 (C₆H₄), 126.4 (C₁₀H₆), 126.1 (C₁₀H₆), 125.2 (C₁₀H₆), 41.2 (C̲(CH₃)₃), 31.0 (C(C̲H₃)₃), 20.9 (CH₃), −7.9 (CH₃–Al), −11.0 (CH₃–Al).

Synthesis of complex 5. Benzene was added to 1 (200.0 mg, 0.35 mmol) and I₂ (88.7 mg, 0.35 mmol) in a Schlenk flask. The mixture was stirred for 1 h at room temperature. After the solvent was removed, the solid residue was washed with pentane (3 × 5 mL). A brown solid was obtained (228.0 mg, 0.33 mmol, 95%). Melting point: 196 °C (decomposition). ¹H NMR (C₆D₆, 400 MHz): δ 7.54 (dd, J_HH = 21.5, 7.8 Hz, 4H, C₁₀H₆), 7.33–7.26 (m, 2H, C₁₀H₆), 7.04–6.94 (m, 6H, C₆H₃), 2.30 (s, 12H, CH₃), 1.09 (s, 18H, C(CH₃)₃). ¹³C{¹H} NMR (C₆D₆, 100 MHz): δ 183.3 (NCN), 142.9 (C₆H_3ipso), 140.4 (C₁₀H_6ipso), 137.0 (C₁₀H_6ipso), 134.3 (C₆H_3ipso), 128.4 (C₆H₃), 126.5 (C₁₀H_6ipso), 125.5 (C₆H₃), 125.3 (C₁₀H₆), 124.3 (C₁₀H₆), 121.8 (C₁₀H₆), 39.8 (C̲(CH₃)₃), 29.2 (C(C̲H₃)₃), 20.4 (CH₃).

Synthesis of complex 6. CH₂Cl₂ was added to 2 (200.0 mg, 0.37 mmol) and I₂ (93.3 mg, 0.37 mmol) in a Schlenk flask. The mixture was stirred for 1 h at room temperature. After the solvent was removed, the solid residue was washed with pentane (3 × 5 mL). A brown solid was obtained (234.0 mg, 0.36 mmol, 96%). Melting point: 207 °C. ¹H NMR (CDCl₃, 400 MHz): δ 7.47 (dd, J_HH = 7.2, 1.7 Hz, 2H, C₁₀H₆), 7.33–7.24 (m, 4H, C₁₀H₆), 7.01 (d, J_HH = 8.0 Hz, 4H, C₆H₄), 6.83 (d, J_HH = 8.1 Hz, 4H, C₆H₄), 2.30 (s, 6H, CH₃), 1.35 (s, 18H, C(CH₃)₃). ¹³C{¹H} NMR (CDCl₃, 100 MHz): δ 182.3 (NCN), 141.7 (C₆H_4ipso), 140.6 (C₁₀H_6ipso), 136.8 (C₁₀H_6ipso), 133.7 (C₆H_4ipso), 129.5 (C₆H₄), 125.2 (C₆H₄), 125.1 (C₁₀H₆), 124.0 (C₁₀H₆), 122.2 (C₁₀H_6ipso), 121.0 (C₁₀H₆), 40.6 (C̲(CH₃)₃), 30.7 (C(C̲H₃)₃), 21.1 (CH₃).

Synthesis of complex 7. Benzene (10 mL) was added to 3 (200.0 mg, 0.31 mmol) and BI₃ (243.0 mg, 0.62 mmol) in a Schlenk flask. The mixture was stirred for 1 h at room temperature, and a white precipitate was observed. After the solvent was removed, the solid residue was washed with pentane (3 × 5 mL). A white solid was recrystallized from benzene at room temperature. Yellow crystals were obtained (307.0 mg, 0.281 mmol, 91%). Melting point: 203 °C (decomposition). ¹H NMR (C₆D₆, 300 MHz): δ 7.48–7.39 (m, 4H, C₁₀H₆), 7.07–7.02 (m, 2H, C₁₀H₆), 6.93–6.83 (m, 6H, C₆H₃), 2.83 (s, 6H, CH₃), 2.70 (s, 6H, CH₃), 0.95 (s, 18H, C(CH₃)₃). MS (Maldi-TOF) m/z: 583.3 ([M − 4I]⁺).

General catalysis protocol. Epoxide (0.2 mL) and the catalysts (1–7) were placed in glass vials with a magnetic stirring bar in an autoclave reactor. The reaction mixture was stirred at 80 °C under 1 bar of CO₂ pressure for 24 h. For complexes 1 and 2, the conditions were 1.5 mol% of catalyst and 1.5 mol% of TBAI as a cocatalyst; for complexes 3 and 4, 0.75 mol% of catalyst and 1.5 mol% of TBAI; for complexes 5 and 6, 1.5 mol% of catalyst without a cocatalyst; and for complex 7, 0.75 mol% of catalyst without TBAI. Complexes 5, 6, and 7 were again tested under the same conditions mentioned before with the addition of 0.75 mol% of TBAI. All the reactions were carried out under solvent-free conditions.

Author contributions

Conceptualization, D. M. and R. S. R.; investigation, A. A., F. G. Z. and N. K.; X-ray structural studies: S. M-L.; DFT calculations, N. V.-E.; writing – original draft preparation, A. A., F. G. Z., R. S. R., E. M. and D. M.; writing – review and editing, A. A., F. G. Z., R. S. R. E. M. and D. M.; supervision, E. M., R. S. R and D. M.; project administration, R. S. R. and D. M.; and funding acquisition, D. M. and R. S. R. All authors have read and agreed to the published version of the manuscript.

Conflicts of interest

There are no conflicts to declare.

Data availability

Additional data supporting this article have been included as part of the supplementary information (SI). Supplementary information: additional experimental information, NMR spectra, crystallographic data and DFT calculations. See DOI: https://doi.org/10.1039/d5dt02221e.

CCDC 2477633 and 2477634 contain the supplementary crystallographic data for this paper.^27a,b

Acknowledgements

This work was supported by the Centre National de la Recherche Scientifique (CNRS), the Université de Toulouse, ECOS–Sud Chili (No. C19E04), and FONDECYT (project no. 1200748; 1230537). A. A. acknowledges funding from Ph.D. ANID 2019 fellowship no. 21190209. F. G. Z. acknowledges funding from Ph.D. ANID 2021 fellowship no. 21210371. N. V. E. acknowledges VRID Iniciación Grant No. 2024001183INI at the Universidad de Concepción and Fondecyt Regular Projects 1220355 and 1221803. This research was supported by the high-performance computing system of PIDi-UTEM (SCC-PIDi-UTEM CONICYT-FONDEQUIP-EQM180180).

References

1. J. Pospech, I. Fleischer, R. Franke, S. Buchholz and M. Beller, Angew. Chem., Int. Ed., 2013, 52, 2852–2872,  DOI:10.1002/anie.201208330.
2. F. Godoy, C. Segarra, M. Poyatos and E. Peris, Organometallics, 2011, 30, 684–688,  DOI:10.1021/om100960t.
3. J. Takagi, K. Takahashi, T. Ishiyama and N. Miyaura, J. Am. Chem. Soc., 2002, 124, 8001–8006,  DOI:10.1021/ja0202255.
4. N. A. Tappe, R. M. Reich, V. D'Elia and F. E. Kühn, Dalton Trans., 2018, 47, 13281–13313,  10.1039/C8DT02346H.
5. Q. Liu, L. Wu, R. Jackstell and M. Beller, Nat. Commun., 2015, 6, 5933,  DOI:10.1038/ncomms6933.
6. J. Yun, L. Zhang, Q. Qu, H. Liu, X. Zhang, M. Shen and H. Zheng, Electrochim. Acta, 2015, 167, 151–159,  DOI:10.1016/j.electacta.2015.03.159.
7. J. H. Clements, Ind. Eng. Chem. Res., 2003, 42, 663–674,  DOI:10.1021/ie020678i.
8. P. P. Pescarmona, Curr. Opin. Green Sustainable Chem., 2021, 29, 100457,  DOI:10.1016/j.cogsc.2021.100457.
9. Handbook of Aluminum: Volume 2: Alloy Production and Materials Manufacturing, ed. G. E. Totten and D. S. MacKenzie, New York, Basel, 2003.
10. K. Hobson, C. J. Carmalt and C. Bakewell, Chem. Sci., 2020, 11, 6942–6956,  10.1039/D0SC02686G.
11. (a) M. P. Coles, D. C. Swenson and R. F. Jordan, Organometallics, 1997, 16, 5183–5194,  DOI:10.1021/om9706323; (b) M. P. Coles and R. F. Jordan, J. Am. Chem. Soc., 1997, 119, 8125–8126,  DOI:10.1021/ja971815j; (c) M. P. Coles, D. C. Swenson and R. F. Jordan, Organometallics, 1998, 17, 4042–4048,  DOI:10.1021/om9802358; (d) S. Dagorne, I. A. Guzei, M. P. Coles and R. F. Jordan, J. Am. Chem. Soc., 2000, 122, 274–289,  DOI:10.1021/ja992104j; (e) D. Doyle, Y. K. Gun'ko, P. B. Hitchcock and M. F. Lappert, J. Chem. Soc., Dalton Trans., 2000, 4093–4097,  10.1039/B005824F; (f) D. Abeysekera, K. N. Robertson, T. S. Cameron and J. A. C. Clyburne, Organometallics, 2001, 20, 5532–5536,  DOI:10.1021/om010752h; (g) H. A. Jenkins, D. Abeysekera, D. A. Dickie and J. A. C. Clyburne, J. Chem. Soc., Dalton Trans., 2002, 3919–3922,  10.1039/B205875H; (h) B. Clare, N. Sarker, R. Shoemaker and J. R. Hagadorn, Inorg. Chem., 2004, 43, 1159–1166,  DOI:10.1021/ic035035p; (i) A. L. Brazeau and S. T. Barry, Chem. Mater., 2008, 20, 7287–7291,  DOI:10.1021/cm802195b; (j) F. Qian, K. Liu and H. Ma, Dalton Trans., 2010, 39, 8071–8083,  10.1039/C0DT00272K; (k) Y. R. Yepes, J. Martínez, H. R. Sánchez, C. Quintero, M. C. Ortega-Alfaro, J. G. López-Cortés, C. G. Daniliuc, A. Antiñolo, A. Ramos and R. S. Rojas, Dalton Trans., 2020, 49, 1124–1134,  10.1039/C9DT03808F; (l) K. Hobson, C. J. Carmalt and C. Bakewell, Inorg. Chem., 2021, 60, 10958–10969; (m) S. Banerjee, C. A. Kumar, S. Bose, S. K. Sarkar, S. K. Gupta, N. Graw, C. Köhler, R. Herbst-Irmer, D. Stalke, S. Dutta, D. Koley and H. W. Roesky, Z. Anorg. Allg. Chem., 2021, 647, 1735–1743; (n) C. Ni, X. Ma, Z. Yang and H. W. Roesky, Eur. J. Inorg. Chem., 2022, e202100929,  DOI:10.1002/ejic.202100929; (o) F. Gómez Zamorano, M. J. Rojas, S. Mallet-Ladeira, A. R. Cabrera, J. Garo, J.-M. Sotiropoulos, E. Maerten, D. Madec and R. S. Rojas, Molecules, 2025, 30, 3842–3859,  DOI:10.3390/molecules30193842.
12. T. Ishikawa and T. Kumamoto, in Superbases for Organic Synthesis, John Wiley & Sons, Ltd, 2009, pp. 49–91,  DOI:10.1002/9780470740859.ch3.
13. A. A. Aly and A. M. Nour-El-Din, ARKIVOC, 2008, 1, 153–194,  DOI:10.3998/ark.5550190.0009.106.
14. D. O. Meléndez, A. Lara-Sánchez, J. Martínez, X. Wu, A. Otero, J. A. Castro-Osma, M. North and R. S. Rojas, ChemCatChem, 2018, 10, 2271–2277,  DOI:10.1002/cctc.201702014.
15. Y. R. Yepes, C. Quintero, D. O. Meléndez, C. G. Daniliuc, J. Martínez and R. S. Rojas, Organometallics, 2019, 38, 469–478,  DOI:10.1021/acs.organomet.8b00795.
16. S. Saltarini, N. Villegas-Escobar, J. Martínez, C. G. Daniliuc, R. A. Matute, L. H. Gade and R. S. Rojas, Inorg. Chem., 2021, 60, 1172–1182,  DOI:10.1021/acs.inorgchem.0c03290.
17. A. W. Addison, T. N. Rao, J. Reedijk, J. van Rijn and G. C. Verschoor, J. Chem. Soc., Dalton Trans., 1984, 1349–1356,  10.1039/DT9840001349.
18. L. Yang, D. R. Powell and R. P. Houser, Dalton Trans., 2007, 955–964,  10.1039/B617136B.
19. N. Villegas-Escobar, D. E. Ortega, D. Cortés-Arriagada, R. Durán, D. Yepes, S. Gutiérrez-Oliva and A. Toro-Labbé, J. Phys. Chem. C, 2017, 121, 12127–12135,  DOI:10.1021/acs.jpcc.7b00278.
20. M. V. Yakovenko, A. V. Cherkasov, G. K. Fukin, D. Cui and A. A. Trifonov, Eur. J. Inorg. Chem., 2010, 3290–3298,  DOI:10.1002/ejic.201000330.
21. A. Acuña, S. Mallet-Ladeira, J.-M. Sotiropoulos, E. Maerten, A. R. Cabrera, A. Baceiredo, T. Kato, R. S. Rojas and D. Madec, Molecules, 2024, 29, 325,  DOI:10.3390/molecules29020325.
22. Bruker, SADABS, Bruker AXS Inc., Madison, Wisconsin, USA, 2008.
23. G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Adv., 2015, 71, 3–8,  DOI:10.1107/S2053273314026370.
24. G. M. Sheldrick, Acta Crystallogr., Sect. C: Struct. Chem., 2015, 71, 3–8,  DOI:10.1107/S2053229614024218.
25. A. L. Spek, Acta Crystallogr., Sect. C: Struct. Chem., 2015, 71, 9,  DOI:10.1107/S2053229614024929.
26. A. J. Bridgeman, G. Cavigliasso, L. R. Ireland and J. Rothery, Dalton Trans., 2001, 2095–2108,  10.1039/b102094n.
27. (a) CCDC 2477633: Experimental Crystal Structure Determination, 2025,  DOI:10.5517/ccdc.csd.cc2p55nq; (b) CCDC 2477634: Experimental Crystal Structure Determination, 2025,  DOI:10.5517/ccdc.csd.cc2p55pr.

---