Homocoupling of C₁ building blocks by a heterometallic Mg–Al complex

Abstract

A heterometallic magnesium–aluminium hydride complex is reported for the selective homologation of C₁ building blocks (CO and CNXyl, Xyl = 2,6-Me₂C₆H₃) to form products containing C₂ and C₃ chains. CO reacts to form known deltate or ethynediolate motifs with different outcomes depending on the steric demands of the ligand on magnesium. In contrast, CNXyl reacts through a combination of carbon–carbon bond formation and hydride migration to form a novel metalated ethene diamidolate ligand [RN¹C¹=C²(H)N²R]³⁻.

---

Carbon monoxide (CO), carbon dioxide (CO₂), and isocyanides (CNR, R = alkyl, aryl) have been extensively used as C₁-building blocks for the construction of more complex organic molecules.¹ For example, the Fischer–Tropsch (F–T) process combines CO and H₂ and can be viewed as a controlled polymerisation, hydrogenation, and deoxygenation of carbon monoxide.^2–4 Despite its impact, the F–T process suffers from low product selectivity forming both a distribution of carbon chain lengths and mixtures of alkanes, alkenes or oxygenates depending on the catalyst and precise conditions. In recent years, there has been growing interest in using C₁ building blocks to form atomically precise carbon chains.^5–8 The use of well-defined, homogenous, metal complexes has allowed identification of new selective chemical transformations. In particular, systems that contain two (or more) metal sites capable of synergistic binding and activation of C₁ building blocks have led to some remarkable examples of forming complex products with C₂–C₆ carbon chains and cycles.

For example, we previously showed that the aluminium(I) compound [{(Ar¹NCMe)₂CH}Al] (1, Ar¹ = 2,6-iPr₂C₆H₃) reacts with CO in the presence of transition metal carbonyl complexes to produce C₂, C₃ and C₄-chains (Scheme 1).^9–11 Similar reactions occur between 1 and CN(1-Ad) (1-Ad = 1-adamantyl), with addition of [Mo(CO)₆] impacting the selectively of product formation.¹² The Jones group combined their low-oxidation state magnesium(I) complex [{(Ar²NCMe)₂CH}Mg]₂ (Ar² = 2,4,6-Me₃C₆H₂ or 2,6-Me₂C₆H₃) with sub-stoichiometric amounts of [Mo(CO)₆] to facilitate the cyclohexamerisation of CO to a benzene-hexolate moiety.¹³ In the absence of [Mo(CO)₆], related magnesium(I) species form either deltate or ethenediolate complexes, with the selectivity controlled by the sterics of the ligand on magnesium.^14–16 The magnesium hydride complexes [{(Ar³NCMe)₂CH}MgH]₂ (2a, Ar³ = 2,6-iPr₂C₆H₃; 2b, Ar³ = 2,6-EtC₆H₃) have also been shown to react with CO to generate cis-ethenediolate and cyclopropanetriolate complexes respectively.^17–19 Related magnesium alkyl complexes can promote the dimerisation and trimerisation of CO either through thermal or catalytic routes.^20–22 We have shown that tetrameric magnesium hydride clusters effect the coupling and deoxygenation of CO,²³ and the cross-coupling of CO and CNXyl.²⁴ In related studies, both anionic aluminium and magnesium complexes have been demonstrated to promote the coupling of either CO or CNR, achieving the formation of C_2–6 chains.^25–30

Scheme 1 Reactions of 3a and 3b CO to form deltate and ethynediolate complexes.

Despite the significant advances described above, there has been limited investigations into coupling reactions of CO (or isocyanides) mediated by complexes with heteronuclear metal–metal bonds. Herein we report the homologation of C₁ building blocks to C₂ and C₃ chains by reaction with an in situ generated heterometallic Mg–Al hydride complex.³¹ We show that this heterometallic can participate in known reactivity to form deltate and ethynediolate motifs. But more importantly demonstrate a new mode of reactivity of an isocyanide that involves both carbon–carbon bond formation and hydride transfer to the carbon chain, generating a previously unknown trianionic ethene diamidolate fragment [RNC=C(H)NR]³⁻ that bridges magnesium and aluminium centres. DFT calculations are used to probe the electronic structure of the metalated ethene diamidolate and rationalise the most likely mechanism for its formation.

The reaction of 1 with the magnesium hydride complex 2a was originally reported by Harder and co-workers,³¹ and leads to the formation of the corresponding heterometallic Mg–Al hydride species 3a. Herein we generated 3a, and an analogue 3b, in situ and reacted them directly with either CO or CNAr (Ar = 2,6-dimethylphenyl). Addition of excess CO to 3a over 24 h at 25 °C in benzene-d₆ led to the formation of the deltate complex 4a (Scheme 1). 4a results from a threefold reaction of CO and incorporates both Mg and Al sites from the heterometallic precursor. In the solid state, 4a contains a dimeric structure with two deltate ligands bridging magnesium sites. The ¹H and ¹³C NMR spectroscopic data, are suggestive of a dynamic process which renders the three binding sites chemically and magnetically equivalent in solution. The hydride ligand remains unreacted and bound to aluminium appearing as a broad resonance at δ = 3.97 ppm in the ¹H NMR spectrum. 4a crystallises alongside a minor byproduct 4b derived from deprotonation and metalation of one the methyl groups of the β-diketiminate backbone (see ESI† for details). While the mechanism to form 4b remains unclear, it can be readily separated from 4a through fractional crystallisation in hexamethyldisiloxane.

The reaction outcome proved sensitive to the steric demands of the ligand on magnesium. Addition of CO to an in situ generated sample of 3b over 24 h at 25 °C in benzene-d₆ led to the formation of the ethynediolate complex 5 (Scheme 1). 5 was characterised by diagnostic resonances at δ = 47.9 and 48.2 ppm in the ¹³C NMR spectrum, consistent with an unsymmetrical ethynediolate moiety. In the solid-state, the C–C and C–O bond lengths of the deltate and ethynediolate fragments in 5a, 5b, and 6 are consistent with those reported previously (Fig. 1).^14,15

Fig. 1 Crystal Structures of 4a, 5, and 6. Aryl groups on the β-diketiminate ligands are drawn as wireframe. Hydrogen atoms with the exception of hydride ligands have been omitted for clarity.

Reaction of 3a with 2,6-dimethylphenylisocyanide for 1 h at 25 °C in benzene-d₆ led to formation of 6 (Scheme 2). 6 contains a metalated ethene diamidolate ligand with a trans-geometry obtained by coupling of two isocyanides and hydride transfer from aluminium to carbon. This new C₂ fragment in 6 can be viewed as originating from a 2 : 1 reaction of CNXyl : hydride. This is the first time a simple metalated ethene diamidolate ligand, [RN¹C¹=C²(H)N²R]³⁻, has been observed to form from reaction of an isocyanide and metal hydride complex. In C₆D₆ solution, 6 was characterised by a diagnostic proton of the ethene moiety H¹ which resonates at δ = 6.03 (s) ppm. This correlates with ¹³C environments C¹ and C² at δ = 148.3 and 139.2 ppm respectively.

Scheme 2 Reaction of 3a with CNXyl to form complex 6 which contains a metalated ethene diamidolate ligand.

In the solid-state, 6 retains a heterometallic structure with a newly formed ethene diamidolate ligand which coordinates the Al atom through a κ²-C¹,N² binding mode and the Mg atom through a κ¹-N¹ binding mode The [RN¹C¹=C²(H)N²R]³⁻ moiety is formally a trianionic ligand. The Al–C¹ and Al–N¹ bonds lengths are 1.999(2) and 1.8514(19) Å respectively. The C¹–N¹, C¹=C², and C²–N² bond lengths take values of 1.416(3), 1.354(3) and 1.453(3) Å respectively (Fig. 1).

DFT calculations were conducted to gain further insight into the electronic structure of the new [RN¹C¹=C²(H)N²R]³⁻ motif in 6, alongside its mechanism of formation from CNXyl and 3a. Geometries were optimised with the B3PW91-D3 functional using a hybrid basis-set comprised of 6-311+G* (C,H,N,O) and SDDAll (Mg, Al). Solvation was considered using the polarizable continuum model (benzene). Single point corrections to energies were conducted with the 6-311+G** triple zeta basis-set.

NBO calculations on 6 suggest that the [RN¹C¹=C²(H)N²R]³⁻ motif adopts a conjugated structure. The Wiberg Bond Indices (WBIs) take values of N¹–C¹ (1.16), C¹=C² (1.68) and C²–N² (1.03). These data can be interpreted in terms of an alternating array of single and double bonds, however the N–C WBIs are slightly greater than expected for a single bond suggesting a degree of delocalisation of the N lone-pair into the conjugated system. Binding to metals is primarily ionic with NPA charges on Al (+2.10) and Mg (+1.82) being highly positive and those on N¹ (−0.94), C¹ (−0.44), C² (−0.23) and N² (−0.99) uniformly negative. To further probe the electronic structure the parent non-coordinated trianion [RN¹C¹=C²(H)N²R]³⁻ was calculated, this species shows a similar electronic structure to the coordinated fragment found in 6, consistent with a large ionic contribution to the bonding to the metal sites.

Further calculations were undertaken to shed light on the mechanism that leads to the formation of 6 (Fig. 2). These calculations were initiated from 3a. The lowest energy pathway identified is initiated from binding of CNXyl to the magnesium site of 3a to form Int-1. This step is followed by the insertion of CNXyl into the Mg–Al bond via TS-1 (ΔG^‡_{298 K} = 18.9 kcal mol⁻¹) and directly generates a dimetallated imine functional group in intermediate Int-2. The structure of Int-2 is similar to the established intermediates reported in CO homologation.¹⁵ Int-2 can undergo a reorganisation into the thermodynamically more stable isomer Int-3 by TS-2 (ΔG^‡_{298 K} = 11.7 kcal mol⁻¹).

Fig. 2 Calculated pathway for the formation of 6 from 3a and 2 equiv. of CNXyl. B3PW91-D3/6-311+G**/PCM (benzene)//B3PW91-D3/6-311+G* (C,H,N,O)/SDDAll (Mg)/PCM (benzene). Gibbs energies, values in kcal mol⁻¹.

Subsequent coordination of another molecule of CNXyl to Int-3 forms the intermediate Int-4. Int-4 contains both adjacent imine and isocyanide ligands, which are suitably arranged for carbon–carbon bond formation. Nucleophilic attack of the metalated imine unit on the isocyanide group occurs via TS-3 (ΔG^‡_{298 K} = 24.6 kcal mol⁻¹) to generate Int-5.^32,33 Along the potential energy surface from Int-4 → TS-3 → Int-5 the Wiberg Bond Indices (WBIs) for the forming C¹–C² bond increase (0.13 → 0.45 → 1.07) while those for C²–N² bond decrease (2.02 → 2.10 → 1.63). NPA charges are consistent with the imine carbon acting as the nucleophilic site with the charge on C¹ (−0.80 → −0.45 → 0.12) becoming more positive as the reaction progresses. Hydride transfer occurs through first a conformational change from Int-5 to Int-6, followed by hydride migration to the C¹ position by TS-4 (ΔG^‡_{298 K} = 8.9 kcal mol⁻¹) to form Int-7. Species related to Int-5 and Int-6 have been reported during the coupling of isonitriles with low-oxidation state magnesium complexes containing homonuclear Mg–Mg bonds.³⁴ A subsequent 1,2-migration from C¹ to C² via TS-5 (ΔG^‡_{298 K} = 13.9 kcal mol⁻¹) ultimately affords the experimentally identified product 6.

In summary, a heterometallic magnesium–aluminium hydride complex reacts with CO to form known C₂ and C₃ fragments namely ethynediolate and deltate ligands. Reaction with the isocyanide CNXyl (Xyl = 2,6-dimethylphenyl) leads instead to both C–C coupling and hydride transfer to form a new [RN¹C¹=C²(H)N²R]³⁻ ligand. To our knowledge, this is the first example of homocoupling and reduction of an isocyanide to form such a fragment. DFT calculations support a mechanism for C–C bond formation through nucleophilic attack of a dimetallated imine functional group on a coordinated isocyanide. Our findings demonstrate that heterometallic complexes containing metal–metal bonds are suitable reagents to form carbon chains from CO and isocyanides. More importantly we demonstrate that these species give rise to new reactivity leading to selective formation of new products not yet observed with other main group or transition metal complexes.

Conflicts of interest

There are no conflicts to declare.

Data availability

Details of experimental procedures, characterisation data and computational modelling is provided in the ESI.†

Acknowledgements

The EPSRC is thanked for funding (EP/S036628/1).

References

1. A. Eschenmoser, Tetrahedron, 2007, 63, 12821–12844.
2. A. Y. Khodakov, W. Chu and P. Fongarland, Chem. Rev., 2007, 107, 1692–1744.
3. M. E. Dry, Catal. Today, 2002, 71, 227–241.
4. H. Schulz, Appl. Catal., A, 1999, 186, 3–12.
5. S. Fujimori and S. Inoue, J. Am. Chem. Soc., 2022, 144, 2034–2050.
6. R. Y. Kong and M. R. Crimmin, Dalton Trans., 2020, 49, 16587–16597.
7. J. M. Parr and M. R. Crimmin, Angew. Chem., Int. Ed., 2023, 62, e202219203.
8. M. J. Evans and C. Jones, Chem. Soc. Rev., 2024, 53, 5054–5082.
9. R. Y. Kong, M. Batuecas and M. R. Crimmin, Chem. Sci., 2021, 12, 14845–14854.
10. R. Y. Kong and M. R. Crimmin, J. Am. Chem. Soc., 2018, 140, 13614–13617.
11. M. Batuecas, R. Y. Kong, A. J. P. White and M. R. Crimmin, Angew. Chem., Int. Ed., 2022, 61, e202202241.
12. C. Zhan, F. Dankert, Z. Jiang, B. Wang, D. Munz and J. Chu, Angew. Chem., Int. Ed., 2023, 62, e202307352.
13. A. Paparo, K. Yuvaraj, A. J. R. Matthews, I. Douair, L. Maron and C. Jones, Angew. Chem., Int. Ed., 2021, 60, 630–634.
14. K. Yuvaraj, I. Douair, D. D. L. Jones and C. Jones, Chem. Sci., 2020, 11, 3516–3522.
15. K. Yuvaraj, I. Douair, A. Paparo, L. Maron and C. Jones, J. Am. Chem. Soc., 2019, 141, 8764–8768.
16. K. Yuvaraj, J. C. Mullins, T. Rajeshkumar, I. Douair, L. Maron and C. Jones, Chem. Sci., 2023, 14, 5188–5195.
17. R. Lalrempuia, C. E. Kefalidis, S. J. Bonyhady, B. Schwarze, L. Maron, A. Stasch and C. Jones, J. Am. Chem. Soc., 2015, 137, 8944–8947.
18. M. D. Anker, M. S. Hill, J. P. Lowe and M. F. Mahon, Angew. Chem., Int. Ed., 2015, 54, 10009–10011.
19. M. D. Anker, C. E. Kefalidis, Y. Yang, J. Fang, M. S. Hill, M. F. Mahon and L. Maron, J. Am. Chem. Soc., 2017, 139, 10036–10054.
20. M. Wang, Y. Cai, T. Rajeshkumar, L. Maron and X. Xu, Chem. – Eur. J., 2025, 31, e202404594.
21. D. T. Nguyen, R. Mondal, M. J. Evans, J. M. Parr and C. Jones, Angew. Chem., Int. Ed., 2025, e202500264.
22. K. Yuvaraj and C. Jones, Chem. Commun., 2021, 57, 9224–9227.
23. W. Yang, A. J. P. White and M. R. Crimmin, Angew. Chem., Int. Ed., 2024, 63, e202319626.
24. W. Yang, A. J. P. White and M. R. Crimmin, Chem. Sci., 2024, 15, 11807–11813.
25. M. J. Evans, M. D. Anker, C. L. McMullin and M. P. Coles, Chem. Sci., 2023, 14, 6278–6288.
26. M. J. Evans, M. G. Gardiner, M. D. Anker and M. P. Coles, Chem. Commun., 2022, 58, 5833–5836.
27. A. Heilmann, M. M. D. Roy, A. E. Crumpton, L. P. Griffin, J. Hicks, J. M. Goicoechea and S. Aldridge, J. Am. Chem. Soc., 2022, 144, 12942–12953.
28. H.-Y. Liu, R. J. Schwamm, S. E. Neale, M. S. Hill, C. L. McMullin and M. F. Mahon, J. Am. Chem. Soc., 2021, 143, 17851–17856.
29. A. O'Reilly, M. D. Haynes, C. L. McMullin, S. Harder, Z. R. Turner, D. O'Hare, J. R. Fulton and M. P. Coles, Chem. Commun., 2024, 60, 7204–7207.
30. M. J. Evans and C. Jones, Inorg. Chem., 2023, 2023, 14393–14401.
31. S. Brand, H. Elsen, J. Langer, S. Grams and S. Harder, Angew. Chem., Int. Ed., 2019, 58, 15496–15503.
32. W. Chen, Y. Zhao, W. Xu, J.-H. Su, L. Shen, L. Liu, B. Wu and X.-J. Yang, Chem. Commun., 2019, 55, 9452–9455.
33. W. Uhl, U. Schütz, W. Hiller and M. Heckel, Chem. Ber., 1994, 127, 1587–1592.
34. M. Ma, A. Stasch and C. Jones, Chem. – Eur. J., 2012, 18, 10669–10676.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental procedures and details of calculations (PDF), crystallographic data for 4a, 4b, 5 and 6 (CIF) and computational coordinates (XYZ). CCDC 2363835–2363837 and 2425919. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d5dt01351h

---