Reactivity of a triamidoamine terminal uranium(VI)-nitride with 3d-transition metal metallocenes

Abstract

Reactions between [(Tren^TIPS)U^VI≡N] (1, Tren^TIPS = {N(CH₂CH₂NSiPrⁱ₃)₃}³⁻) and [M^II(η⁵-C₅R₅)₂] (M/R = Cr/H, Mn/H, Fe/H, Ni/H) were intractable, but M/R = Co/H or Co/Me afforded [(Tren^TIPS)U^V=N-(η¹:η⁴-C₅H₅)Co^I(η⁵-C₅H₅)] (2) and [(Tren^TIPS)U^IV–NH₂] (3), respectively. For M/R = V/H [(Tren^TIPS)U^IV–N=V^IV(η⁵-C₅H₅)₂] (4), was isolated. Complexes 2–4 evidence one-/two-electron uranium reductions, nucleophilic nitrides, and partial N-atom transfer.

---

In recent years molecular uranium-nitrides have attracted burgeoning attention due to their importance as actinide electronic structure benchmarks and in small molecule activations.^1–4 The search for isolable terminal uranium-nitrides was accomplished by some of us just over a decade ago, first with [Na(12C₄)₂][(Tren^TIPS)U^V≡N] (Tren^TIPS = {N(CH₂CH₂NSiPrⁱ₃)₃}³⁻; 12C₄ = 12-crown-4 ether)⁵ in 2012 and then [(Tren^TIPS)U^VI≡N] (1) in 2013.⁶ The Tren^TIPS ligand has proven to be a ‘privileged’ ancillary ligand for terminal uranium-nitrides,^7–10 and indeed the only other ligand class to have supported an isolable terminal uranium-nitride linkage is the siloxide ligand (Bu^tO)₃SiO¹⁻ used by Mazzanti.¹¹ In addition to terminal uranium-nitrides, a variety of low- (two-) coordinate bridging uranium-nitrides are now known, including U≡NAM (AM = Li, Na, K, Rb, Cs),^6,8,12,13 U=N=An (An = U, Th),^13–28 and U≡N–M complexes (M = Mo, Rh, Ir, Mo).^29,30 The latter remain few in number, likely largely reflecting the limited synthetic methodologies available for constructing such linkages: M = Mo was accessed by partial nitride transfer from Mo to U,²⁹ and M = Rh and Ir compounds were made by photolysis of azido precursors.³⁰ We decided to examine the potential of 1 to construct heterobimetallic nitride-bridged complexes since it already has a terminal,^5–10 nucleophilic nitride installed at uranium which could in principle simplify its use in synthesis.

Here we report on our findings, where we have examined the reactivity of 1 towards 3d transition metal metallocenes [M^II(η⁵-C₅R₅)₂] (M/R = V/H, Cr/H, Mn/H, Fe/H, Co/H, Co/Me, Ni/H). The reactions with M = Cr, Mn, Fe, and Ni appeared to proceed but proved intractable. However, reactions with M/R = Co/H, Co/Me, and V/H produced isolable derivatives that evidence one- and two-electron reductions of uranium, nucleophilic nitrides, and partial N-atom transfer.

In separate reactions, Scheme 1, mixing [(Tren^TIPS)U^VI≡N] (1) with [M^II(η⁵-C₅H₅)₂] (M = Cr, Mn, Fe, Ni) in cold (−78 °C) toluene afforded, after solvent was removed, crude brown solids. However, in all cases no products could be isolated cleanly. ¹H NMR spectroscopy revealed numerous paramagnetically shifted resonances (up to 66 ppm wide range of resonances, Fig. S1–S4, ESI†) and hence the product identities and/or extent of decomposition is unclear.

Scheme 1 Synthesis of 2–4 from 1 and intractable reaction outcomes. The by-products are either not known or are not shown for clarity.

In contrast to the reactions between 1 and M = Cr, Mn, Fe, and Ni, with M = Co an identifiable product could be obtained, Scheme 1. Specifically, treating 1 with nineteen valence electron [Co^II(η⁵-C₅H₅)₂] afforded the uranium(V)-imido complex [(Tren^TIPS)U^V=N-(η¹:η⁴-C₅H₅)Co^I(η⁵-C₅H₅)] (2) as red crystals. However, 2 co-crystallises with variable quantities of 1 and [Co^II(η⁵-C₅H₅)₂] (Fig. S5 and S6, ESI†). Indeed, a variable-temperature ¹H NMR study (Fig. S7, ESI†) revealed the dominance of 2 at low temperature (−60 °C) and a greater proportion of 1/[Co^II(η⁵-C₅H₅)₂] at higher temperature (25 °C), and hence 2 is in equilibrium with 1 and [Co^II(η⁵-C₅H₅)₂]. Whilst the optimal practical ratio for the reaction was found to be two equiv. of [Co^II(η⁵-C₅H₅)₂] to 1 we could only ever isolate 2 as a mixture (A) co-crystallised with 1 and [Co^II(η⁵-C₅H₅)₂]. Although the [Co^II(η⁵-C₅H₅)₂] can be sublimed out of A, when redissolved in addition to 2 resonances for 1 and [Co^II(η⁵-C₅H₅)₂] are still observed in the resulting ¹H NMR spectrum demonstrating an immutable equilibrium.

Nucleophilic attack of eighteen valence electron [Co^III(η⁵-C₅H₅)₂]⁺ is known,³¹ and whilst a radical reaction cannot be discounted the radical chemistry of 1 is quite slow in the absence of strong light,⁶ so we propose that [Co^II(η⁵-C₅H₅)₂] (E^0′ = ∼−1.32 V vs. Fc)³² initially reduces 1 to give “[Co^III(η⁵-C₅H₅)₂]⁺[(Tren^TIPS)U^V≡N]⁻”, and then nucleophilic attack of a cyclopentadienyl ring by the nitride occurs. The nucleophilic attack rehybridises one of the cyclopentadienyl carbon atoms from sp² to sp³, formally forming a Co^I-cyclopentadiene unit, hence retaining an eighteen valence electron cobalt moiety.

Given the issue in isolating 2, its characterisation was probed using A as far as was reasonably practicable. The ¹H NMR spectrum of A exhibits resonances for 2 over the range 23.5 to −4.2 ppm (Fig. S5 and S6, ESI†). Of most salience, in addition to one cyclopentadienyl ring resonance of 5H (9.7 ppm) two pairs of 2H each for the η⁴-diene portion of the cyclopentadiene ring are located at 17.9 and 10.6 ppm, and the H-atom residing on the ring sp³ C-atom resonates at −1.5 ppm. We recorded the UV/Vis/NIR spectra of 1 and [Co^II(η⁵-C₅H₅)₂] and then subtracted them from the corresponding spectrum of A to unambiguously identify absorptions that correspond to 2 (Fig. S12–S14, ESI†). Of most interest is the near infrared region, where four absorptions (ε = ∼10–30 M⁻¹ cm⁻¹) are found at ∼6000, ∼7100, ∼9000, and ∼10 600 cm⁻¹ which represent ²Γ₄ to ²Γ₄, ²Γ₄, ²Γ₄, and ¹Γ₅ + ¹Γ₆ absorptions, respectively, that are characteristic of uranium(V) in C_3v symmetry.³³

The solid-state structure of 2 was determined, Fig. 1, confirming its formulation and also exo-attack by the nitride. The U1–N5 distance of 1.925(3) Å is longer than the terminal U^VI≡N distance of 1.799(7) Å in 1 and group 1 capped and terminal (Tren^TIPS)U^V≡N distances (1.801(7)–1.840(3) Å),^5,6,8,12 slightly shorter than (Tren^TIPS)U^V=NR distances (∼1.95 Å),⁶ though similar to [(Tren^TIPS)U^V≡NM]₂ (AM = Li, Na, K, Rb, Cs) U^V≡N distances (1.833(4)–1.929(6) Å).^5,8 The N5–C34, C34–C35, and C34–C38 distances of 1.475(5), 1.516(5), and 1.525(15) Å are consistent with N–C and C–C single bonds, and the presence of the Co-bound diene is reflected by C35–C36, C36–C37, and C37–C38 distances of 1.414(5), 1.422(5), and 1.414(6) Å. All other distances in 2 are as anticipated. Overall, the metrical data are consistent with 2 being a uranium(V)-imido complex consistent with the UV/Vis/NIR data.

Fig. 1 Molecular structure of 2 with selective labelling at 120 K and displacement ellipsoids at 50%. Hydrogen atoms except for H34 are omitted for clarity.

Density functional theory (DFT) calculations on 2 (Fig. S26, S27 and Tables S1–S3, S6, ESI†) reveal a somewhat delocalised picture, however the principal UN- and Co-related bonding combinations could be identified and natural bond orbital (NBO) and natural localised molecular orbital (NLMO) analyses identify the σ²π⁴ bonding motif of the imido (Fig. S27, ESI†). The computed charges and spin densities are consistent with U^V/Co^I. The U–N_imido Nalewajski–Mrozek bond order is 2.73, and quantum theory of atoms-in-molecules (QTAIM) analysis reveals a UN 3,−1-bond critical point with a ρ value of 0.18 that is typical of a uranium(v)-imido complex.⁶

Noting the reaction between 1 and [Co^II(η⁵-C₅H₅)₂], we examined the analogous reaction with [Co^II(η⁵-C₅Me₅)₂], Scheme 1: [Co^II(η⁵-C₅Me₅)₂] is a stronger reducing agent (∼−1.93 V vs. Fc)³² compared to [Co^II(η⁵-C₅H₅)₂], meaning an excess of Co-reagent would be less likely to be needed possibly simplifying purification, and the former is sterically more congested which may impede exo-addition. Thus, we treated 1 with one equiv. of [Co^II(η⁵-C₅Me₅)₂], and after work-up and recrystallisation isolated the previously reported emerald green amido complex [(Tren^TIPS)U^IV–NH₂] (3).^6,9,34

The formation of 3 seems at first surprising, but can be rationalised. Assuming that the reaction proceeds by U-reduction to form “[Co^III(η⁵-C₅Me₅)₂]⁺[(Tren^TIPS)U^V≡N]⁻”, protonation to give [(Tren^TIPS)U^V=NH] could occur, and it is known that oxidation of [(Tren^TIPS)U^IV=NH]⁻ results in the formation of 3 and 1via disproportionation of [(Tren^TIPS)U^V=NH].³⁴ Alternatively, given the reducing nature of [Co^II(η⁵-C₅Me₅)₂], it could be that double reduction of 1 occurs to give “[(Tren^TIPS)U^IV≡N]²⁻”, which would be very reactive. Indeed, the closely related complex [(Tren^TIPS)U^IV≡NLi₂]₂ contains bridging nitrides and of all the group 1 cations is only stable with Li because of the highly polarised nature of the U^IV≡N linkage with substantial destabilising charge accumulation at the nitride.³⁵ We note that (C₅Me₅)¹⁻ can provide protons via tuck-in/tuck-over complexes,³⁶ and that [Co^II(η⁵-C₅Me₅)₂] can act as a H-atom shuttle,³⁷ and either process could potentially expedite the formation of 3 from 1.

Since the exact nature of divalent group 4 metallocenes can be ambiguous, we lastly examined the reaction of 1 with [V^II(η⁵-C₅H₅)₂], Scheme 1. Accordingly, a 1 : 1 mixture was stirred in toluene, and after work-up the red complex formulated as [(Tren^TIPS)U^IV–N=V^IV(η⁵-C₅H₅)₂] (4) was isolated in 82% yield.

The solid-state structure of 4 confirms its gross formulation, Fig. 2. The U1–N5 distance of 2.261(9) Å is much longer than the U–N distances in 1⁶ and 2 but similar to the U–N amido distance in 3 (2.228(4) Å).^6,34 Whilst the U1–N5 distance in 4 would be incompatible with U/V oxidation state combinations of VI/II and V/III, IV/IV and III/V were possible. However, the U–N_amide and –N_amine distances of 2.261(9)–2.285(8) and 2.674(10) Å are suggestive of U^IV over U^III.³⁸ The V1–N5 distance was found to be 1.680(9) Å, which compares to a V–N distance of 1.665 Å in [Me₃SiN=V^IV(η⁵-C₅H₅)₂].³⁹ Hence, 4 can be considered to result from partial N-atom transfer from U to V.

Fig. 2 Molecular structure of 4 with selective labelling at 120 K and displacement ellipsoids at 50%. Hydrogen atoms are omitted for clarity.

The ¹H NMR spectrum of 4 (Fig. S8, ESI†) covers the range 72.5 to −6.6 ppm; the resonance at 72.5 ppm corresponds to the vanadocene moiety, with the remaining resonances spanning 32.5 to −6.6 ppm which is qualitatively consistent with U^IV. The ²⁹Si{¹H} NMR spectrum of 4 (Fig. S9, ESI†) exhibits a resonance at −76.2 ppm which falls in the range of U^IV complexes.⁴⁰

The UV/Vis/NIR spectrum of 4 (Fig. S15–S17, ESI†) above 22 500 cm⁻¹ is dominated by charge-transfer bands and a prominent absorption is found at ∼18 000 cm⁻¹ (ε = ∼1500 M⁻¹ cm⁻¹). Below 15 000 cm⁻¹ the spectrum evidences weak (ε = ∼30–70 M⁻¹ cm⁻¹) f–f absorptions. The NIR region has the appearance of U^IV,³ but we could not completely rule out the broad feature at 18 000 cm⁻¹ being f–d transitions of U^III rather than d–d transitions of V.³

Given the potential ambiguity of the U/V oxidation states in 4 we turned to quantum chemical calculations. However, DFT geometry optimisation always led to the U–N and V–N distances being too short and long, respectively (both ∼1.95 Å). Therefore, to first resolve the oxidation state question we turned to state-averaged complete active space self-consistent field (SA-CASSCF) calculations using the unoptimised crystal structure of 4 with an active space of 3 electrons in 12 orbitals (3d and 5f) examining low spin (S_tot = 1/2, 4′) and high spin (S_tot = 3/2, 4′′) multiplicities (see ESI† for details). The ground state is found to be dominated by U^IV (5f²) and V^IV (3d¹) configurations, consistent with the foregoing characterisation data overall. Interestingly, the ground Kramers doublet after spin–orbit coupling is dominated by S_tot = 1/2 states, suggesting that there is an antiferromagnetic interaction between the V^IV and U^IV ions. Furthermore, the calculations show that there significant covalency and crystal field splitting of the 3d- and 5f-orbitals quenching the orbital angular momentum of 4 (see ESI†).

To confirm the SA-CASSCF findings, we collected variable-temperature SQUID magnetometry data on powdered 4 in an external 0.1 T field (Fig. S18, ESI†). The effective magnetic moment of 4 is 2.71μ_B at 300 K and this decreases steadily until at 8 K (1.21μ_B) when it drops more rapidly reaching 0.42μ_B at 1.8 K. The magnetic moment for 4 falls far more quickly with decreasing temperature than for isolated U^IV in [{(Me₃Si)₂N}₃U^IV=E]⁻ (E = O, NSiMe₃),⁴¹ suggesting antiferromagnetic coupling which is also implied by a maximum in the χ_Mvs. T plot of 4 at 4.8 K. The magnetisation at 1.8 K and 7 T (Fig. S21, ESI†) of 0.24μ_B mol⁻¹ is also far smaller than the sum of an isolated [{(Me₃Si)₂N}₃U^IV=E]⁻ and a free S = 1/2,⁴¹ again reflecting the presence of U–V magnetic exchange.

The X-band EPR spectrum of powdered 4 (Fig. S22, ESI†) exhibits an eight line spectrum (⁵¹V, I = 7/2) with g = 1.971 (A_x,y(⁵¹V) = 35 MHz and A_z(⁵¹V) = 220 MHz). However, this is incompatible with the low-temperature SQUID magnetometry data which indicates a S_eff = 1/2 state with g ∼ 0.7. Indeed, the SA-CASSCF results suggest a strongly axial ground doublet state (Table S5, ESI†), and previous work has continuously highlighted the effective high-symmetry behaviour of pseudo-C₃ U^IV fragments.^6,8,41 Taking the data together, we suggest that 4 is actually EPR silent, and that due to the high sensitivity of EPR a trace impurity has been observed instead. We suggest that this is most likely [HN=V^IV(η⁵-C₅H₅)₂] given the similarity of our EPR data to related vanadium(IV)-imido EPR data which are also isotropic.⁴²

To probe the nature of the U–N–V linkage in 4 we performed DFT single point energy calculations on the 4′ and 4′′ spin-state formulations (Fig. S28–S31, Tables S1–S5, S7, ESI†). DFT computes 4′ to be 0.96 kJ mol⁻¹ more stable than 4′′ which again suggests antiferromagnetic coupling. For 4′ the α-spin HOMO (268a) and HOMO−1 (267a) are 5f character, and the β-spin 267b orbital is the 2a_1g orbital of a bent metallocene (sd-hybrid). HOMOs−12, −20, and −21 reveal principally V–N π, π, and σ-bond interactions with weaker U–N π- and σ-components, respectively, and these interactions are also found in the NBO and NLMO analysis confirming that the V–N and U–N bonds are largely of imido and amido character, respectively. Inspection of 4′′ reveals a very similar bonding picture, except that after HOMO (269a) and HOMO−1 (268a) which are 5f-character the 2a_1g orbital is now found as HOMO−2 (267a) in the α- rather than β-spin manifold, and then the analogous V–N π, π, and σ-bond combinations are now HOMOs−13, −21, and −22. The computed bond order, charge, spin density, NBO, NLMO, and QTAIM data (Tables S1–S3 and S7, ESI†) are consistent with 4 being described as a U^IV/V^IV complex where partial N-atom transfer from U to V has occurred.

To conclude, we have examined the reactivity of 1 towards a range of 3d-transition metal metallocenes. Although several metals (Cr, Mn, Fe, Ni) did not give tractable products, cobaltocene generated a uranium(V)-imido that results from one-electron reduction of uranium and nucleophilic attack of a cyclopentadienyl ligand by the nitride. In contrast, using decamethylcobaltocene resulted in two electron reduction of uranium and formation of a uranium(IV)-amido complex. The reaction of 1 with vanadocene resulted in a two electron redox couple resulting in U^IV and V^IV centres; since the nitride in 4 can be described as being formally of amido- and imido-type bonding character towards U and V, respectively, then 4 can be regarded as representing partial N-atom transfer from U to V. Nevertheless, there is clearly some electronic communication across the U–N–V linkage resulting in antiferromagnetic U–V exchange coupling. These complexes expand the still limited range of transition metal capped uranium-nitrides, and whilst demonstrating that constructing heterobimetallic actinide–nitride–metal linkages certainly benefits from starting with the nitride pre-installed at the actinide ion the resulting chemistry can still be complex and dictated by the nature of the transition metal fragment.

We thank the EPSRC (EP/K024000/1, EP/K024000/2, EP/M027015/1, EP/T011289/1), Royal Society (UF071260, UF110005, RG080285, RG110238, URF191320), ERC (239621, 612724, 851504), University of Nottingham, University of Manchester, Australian National University, and the University of Manchester Computational Shared Facility and the EPSRC UK EPR National Research Facility (EP/W014521/1 and EP/V035231/1) for supporting this work.

Data availability

Data are available in the ESI,† from the CCDC, or from the authors on request. CCDC codes 2373740 (2) and 2373741 (4).

Conflicts of interest

There are no conflicts to declare.

Notes and references

1. S. T. Liddle, Inorg. Chem., 2024, 63, 9366–9384.
2. M. Keener, L. Maria and M. Mazzanti, Chem. Sci., 2023, 14, 6493–6521.
3. S. T. Liddle, Angew. Chem., Int. Ed., 2015, 54, 8604–8641.
4. D. M. King and S. T. Liddle, Coord. Chem. Rev., 2014, 266–267, 2–15.
5. D. M. King, F. Tuna, E. J. L. McInnes, J. McMaster, W. Lewis, A. J. Blake and S. T. Liddle, Science, 2012, 337, 717–720.
6. D. M. King, F. Tuna, E. J. L. McInnes, J. McMaster, W. Lewis, A. J. Blake and S. T. Liddle, Nat. Chem., 2013, 5, 482–488.
7. P. A. Cleaves, D. M. King, C. E. Kefalidis, L. Maron, F. Tuna, E. J. L. McInnes, J. McMaster, W. Lewis, A. J. Blake and S. T. Liddle, Angew. Chem., Int. Ed., 2014, 53, 10412–10415.
8. D. M. King, P. A. Cleaves, A. J. Wooles, B. M. Gardner, N. F. Chilton, F. Tuna, W. Lewis, E. J. L. McInnes and S. T. Liddle, Nat. Commun., 2016, 7, 13773.
9. L. Chatelain, E. Louyriac, I. Douair, E. Lu, F. Tuna, A. J. Wooles, B. M. Gardner, L. Maron and S. T. Liddle, Nat. Commun., 2020, 11, 337.
10. J. Du, J. A. Seed, V. E. J. Berryman, N. Kaltsoyannis, R. W. Adams, D. Lee and S. T. Liddle, Nat. Commun., 2021, 12, 5649.
11. L. Barluzzi, R. Scopelliti and M. Mazzanti, J. Am. Chem. Soc., 2020, 142, 19047–19051.
12. J. Du, I. Douair, E. Lu, J. A. Seed, F. Tuna, A. J. Wooles, L. Maron and S. T. Liddle, Nat. Commun., 2021, 12, 4832.
13. N. Tsoureas, A. F. R. Kilpatrick, C. J. Inman and F. G. N. Cloke, Chem. Sci., 2016, 7, 4624–4632.
14. W. J. Evans, S. A. Kozimor and J. W. Ziller, Science, 2005, 309, 1835–1838.
15. A. R. Fox, P. L. Arnold and C. C. Cummins, J. Am. Chem. Soc., 2010, 132, 3250–3251.
16. S. Fortier, G. Wu and T. W. Hayton, J. Am. Chem. Soc., 2010, 132, 6888–6889.
17. C. Camp, J. Pecaut and M. Mazzanti, J. Am. Chem. Soc., 2013, 135, 12101–12111.
18. L. Maria, I. C. Santos, V. R. Sousa and J. Marcalo, Inorg. Chem., 2015, 54, 9115–9126.
19. J. Du, D. M. King, L. Chatelain, E. Lu, F. Tuna, E. J. L. McInnes, A. J. Wooles, L. Maron and S. T. Liddle, Chem. Sci., 2019, 10, 3738–3745.
20. C. T. Palumbo, L. Barluzzi, R. Scopelliti, I. Zivkovic, A. Fabrizio, C. Corminboef and M. Mazzanti, Chem. Sci., 2019, 10, 8840–8849.
21. C. T. Palumbo, R. Scopelliti, I. Zivkovic and M. Mazzanti, J. Am. Chem. Soc., 2020, 142, 3149–3157.
22. X. Li, Y. Rosello, Y.-R. Yao, J. Zhuang, X. Zhang, A. Rodriguez-Fortea, C. de Graff, L. Echegoyen, J. M. Poble and N. Chen, Chem. Sci., 2021, 12, 282–292.
23. L. Barluzzi, F.-C. Hsueh, R. Scopelliti, B. E. Atkinson, N. Kaltosyannis and M. Mazzanti, Chem. Sci., 2021, 12, 8096–8104.
24. M. Keener, R. Scopelliti and M. Mazzanti, Chem. Sci., 2021, 12, 12610–12618.
25. S. L. Staun, G. Wu, W. W. Lukens and T. W. Hayton, Chem. Sci., 2021, 12, 15519–15527.
26. M. D. Straub, L. M. Moreau, Y. S. Qiao, E. T. Ouellette, M. A. Boreen, T. D. Lohrey, N. S. Settineri, S. Hohloch, C. H. Booth, S. G. Minasian and J. Arnold, Inorg. Chem., 2021, 60, 6672–6679.
27. N. Jori, M. Keener, T. Rajeshkumar, R. Scopelliti, L. Maron and M. Mazzanti, Chem. Sci., 2023, 14, 13485–13494.
28. K. Li, T. Rajeshkumar, Y. Zhao, T. Wang, L. Maron and C. Zhu, Chem. Commun., 2024, 60, 2966–2969.
29. L. Barluzzi, N. Jori, T. He, T. Rajeshkumar, R. Scopelliti, L. Maron, P. Oyala, T. Agapie and M. Mazzanti, Chem. Commun., 2022, 58, 4655–4658.
30. X. Xin, I. Douair, T. Rajeshkumar, Y. Zhao, S. Wang, L. Maron and C. Zhu, Nat. Commun., 2022, 13, 3809.
31. Organometallics, ed. C. Elschenbroich, Wiley-VCH, 3rd edn, 2005, pp. 504–505.
32. N. G. Connelly and W. E. Geiger, Chem. Rev., 1996, 96, 877–910.
33. D. Patel, F. Tuna, E. J. L. McInnes, J. McMaster, W. Lewis, A. J. Blake and S. T. Liddle, Dalton Trans., 2013, 42, 5224–5227.
34. D. M. King, J. McMaster, F. Tuna, E. J. L. McInnes, W. Lewis, A. J. Blake and S. T. Liddle, J. Am. Chem. Soc., 2014, 136, 5619–5622.
35. D. M. King, B. E. Atkinson, L. Chatelain, M. Gregson, J. A. Seed, A. J. Wooles, N. Kaltsoyannis and S. T. Liddle, Dalton Trans., 2022, 51, 8855–8864.
36. K. R. D. Johnson and P. G. Hayes, Chem. Soc. Rev., 2013, 42, 1947–1960.
37. M. J. Chalkley, P. H. Oyala and J. C. Peters, J. Am. Chem. Soc., 2019, 141, 4721–4729.
38. B. M. Gardner and S. T. Liddle, Chem. Commun., 2015, 51, 10589–10607.
39. N. Wiberg, H.-W. Häring and U. Schubert, Z. Naturforsch. B, 1980, 35, 599–603.
40. C. J. Windorff and W. J. Evans, Organometallics, 2014, 33, 3786–3791.
41. J. A. Seed, L. Birnoschi, E. Lu, F. Tuna, A. J. Wooles, N. F. Chilton and S. T. Liddle, Chem, 2021, 7, 1666–1680.
42. J. H. Osborne, A. L. Rheingold and W. C. Trogler, J. Am. Chem. Soc., 1985, 107, 7945–7952.

---

Footnote

† Electronic supplementary information (ESI) available: Full experimental and computational details. CCDC 2373740 (2) and 2373741 (4). For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d4cc03846k

---