Monomeric tri-coordinated bis(ferrocenyl)haloalumanes

Abstract

The reactions of the sterically demanding ferrocenyl lithium dimer (Fc*Li)₂ (Fc* = 2,5-bis(3,5-di-t-butyl-phenyl)-1-ferrocenyl) with aluminum trihalides (AlCl₃, AlBr₃, and AlI₃) to furnish the corresponding monomeric bis(ferrocenyl)haloalumanes are reported. In the case of the reaction with AlI₃, an unexpected intramolecular 1,1′-aluminum migration in the ferrocenyl moiety was found to occur. Their monomeric structures with a tri-coordinated aluminum atom show affinitive Al⋯Fe interactions.

---

A variety of organoaluminum compounds have shown potential as Lewis-acidic catalysts for the transformation of small molecules.¹ In most cases, however, trivalent organoaluminum compounds can be expected to exist as oligomeric structures or tetra-coordinated compounds with an intra-/inter-molecular coordination due to the stabilization of the vacant p-orbital of the aluminum atom upon coordination, which reduces the intrinsic electrophilicity and/or Lewis acidity. The oligomeric structures of trivalent organoaluminum halides with intermolecularly bridging halogen atoms or Al–C σ-bonds should be of great interest from a coordination-chemistry viewpoint. In particular, ferrocenyl-substituted trivalent aluminum compounds are known to exhibit unique geometries with Al-bridging oligomeric structures and short Fe⋯Al distances.^2–5 For example, the reaction of FcLi (Fc = ferrocenyl) generated in situ with an equimolar amount of R₂AlCl (R = CH₃ or CH₂CH₃) afforded the corresponding dialkylferrocenylalumanes (FcAlR₂) in the form of Al-bridging dimer I.³ When FcLi was treated with an excess of R₂AlCl, the generated dialkylferrocenylalumane formed a complex with another molecule of R₂AlCl to yield Fc(AlR₂)₂Cl (II).³ Treatment of FcLi with chlorodimethylalumane (ClAlMe₂) followed by addition of pyridine afforded the corresponding ferrocenyldimethylalumane·pyridine complex (FcMe₂Al·pyridine; III).^3,4 As seen in these examples, the isolable organo-/halo-alumanes can be expected to exhibit in most cases tetra-coordinate geometries. However, in some cases, tri-coordinated bis(ferrocenyl)alumanes were obtained, especially when the tri-coordinated aluminum moiety is surrounded by cluster-like structures (Fig. 1).^4,5

Fig. 1 Examples of hitherto reported ferrocenylalumanes.

We aim to synthesize monomeric ferrocenylalumanes that are expected to exhibit ambiphilic properties of high reactivity and thermodynamic stability. That is, these properties arise from the combination of high electrophilicity, attributed to the vacant 3p orbital on the aluminum center, and thermodynamic stability, provided by the intramolecular electronic stabilization from the electron-donating ferrocenyl group. We have already reported the isolation of the sterically demanding ferrocenyl lithium dimer (Fc*Li)₂ (Fc* = 2,5-bis(3,5-di-t-butyl-phenyl)-1-ferrocenyl),⁶ and its synthetic application in the isolation of the corresponding monomeric bis(ferrocenyl)-germylene and -stannylene (Fc*₂E:, E = Ge and Sn).⁷ These previous results on the application of the Fc* group prompted us to attempt the isolation of monomeric, tri-coordinated ferrocenylalumanes by the introduction of the sterically demanding ferrocenyl group to the aluminum center. The group of Braunschweig has reported the isolation of a tetra-coordinated ferrocenyldiiodoalumane stabilized by substitution with an Fc* group and the coordination of an N-heterocyclic carbene (NHC), Fc*AlI₂(NHC), which was obtained from the reaction of Fc*Li with (NHC) → AlH₃ followed by iodination with iodomethane.⁸ Here, we present the synthesis of bis(ferrocenyl)haloalumanes that bear sterically demanding ferrocenyl groups to avoid self-dimerization/oligomerization (Fig. 2).

Fig. 2 Targeted compounds: monomeric bis(ferrocenyl)haloalumanes.

Treatment of the ferrocenyl lithium dimer (Fc*Li)₂ (1)⁶ with 2 eq. of AlX₃ (X = Cl and Br; i.e., Fc*Li : AlX₃ = 1 : 1) at r.t. afforded insoluble orange solids, which were converted to the corresponding dihaloalumane·thf complexes Fc*AlX₂·(thf) (2a: X = Cl; 2b: X = Br), by addition of a small amount of THF (Scheme 1). Although their isolation and purification were very difficult due to their lability and contamination with a small amount of inseparable Fc*H, the products could be identified⁹ based on NMR and mass spectral data; moreover, the solid-state structure of 2b·(thf) was determined unequivocally by single-crystal X-ray diffraction (SC-XRD) analysis.¹⁰ Thus, the insoluble orange solids obtained by mixing 1 and AlX₃ are most likely Li⁺[Fc*AlX₃]⁻ complexes. When a toluene solution of 1 was heated with an equimolar amount of AlCl₃ (i.e., Fc*Li : AlCl₃ = 2 : 1) at 60 °C for 17 h, bis(ferrocenyl)chloroalumane Fc*₂AlCl (3a) was obtained in 76% yield. Alternatively, treatment of 1 with the insoluble solid obtained from the reaction of 1 with 2 eq. of AlCl₃ also afforded 3a. Judging from the ¹H NMR spectrum, the addition of THF to 3a furnished 3a·(thf).

Scheme 1 Formation of ferrocenyldihaloalumane·thf complexes 2a (X = Cl) and 2b (X = Br).

In the expectation of obtaining the bromine analogue, bis(ferrocenyl)bromoalumane Fc*₂AlBr (3b), the reaction of 1 with an equimolar amount of AlBr₃ was examined, which unexpectedly furnished not only 3b but also 4b in 79% and 13% NMR yields, respectively. The expected product (3b) was identified based on its ¹H NMR spectrum and its similarity to that of 3a. The unexpected product (4b) was identified as another type of bis(ferrocenyl)bromoalumane, i.e., Fc*(Fc*′)AlBr (Scheme 2), based on the spectroscopic and SC-XRD analyses of its single crystals, which were obtained from a careful recrystallization of the reaction mixture from hexane (Fig. 3).

Scheme 2 Synthesis of bis(ferrocenyl)haloalumanes.

Fig. 3 (a) Molecular structure of 2b·(thf) in the crystalline state with thermal ellipsoids at 50% probability; hydrogen atoms are omitted for clarity. (b) Depiction of the geometry of the core moiety of 2b·(thf). Selected bond lengths (Å): Al–Br1, 2.3196(5); Al–Br2, 2.3057(5).

Compound 4b could be formed via 1,1′-Al-migration on one of the Fc* substituents. Furthermore, treatment of ferrocenyl lithium dimer 1 with AlI₃ afforded the corresponding bis(ferrocenyl)iodoalumanes 3c and 4c, similar to the case of the reaction of 1 with AlBr₃; however, the reaction is not very clean, as evident from several small signals due to unknown by-products in the ¹H NMR spectra of the reaction mixture. Notably, the formation ratio of 3c and 4c in the reaction mixture was found to depend on the reaction conditions (Table S1‡). Using a higher reaction temperature (60 °C) and/or longer reaction time increased the formation ratio of 3c relative to 4c. The use of a slightly higher/lower amount of AlI₃ increased/decreased the ratio of 4c relative to 3c, indicating that a small amount of residual AlI₃ promotes the 1,1′-Al-migration on the ferrocenyl moiety as a Lewis acid. When a small amount of AlBr₃ was added to the mixture of 3b and 4b (5 : 3), no significant change in their ratio was evident from the ¹H NMR spectra after leaving the sample to stand for 6 days at room temperature. Accordingly, it seems feasible to discard the possibility that 3b and 4b are transformed into each other in the presence of AlBr₃. Overall, the results of the reactions suggest that AlX₃ can promote the formation of 1,1′-Al-migrated products 4, while AlX₃ seems unable to directly transform 3 into 4.¹¹

When isolated 3b was treated with an equimolar amount of AlBr₃ in C₆D₆, ferrocenyldibromoalumane dimer (Fc*′AlBr₂)₂ (5b) was formed quantitatively (Scheme 3).^9,10 According to Scheme 1, the initially formed Li⁺[Fc*AlBr₃]⁻ would generate a small amount of Fc*Li and AlBr₃ as well as Fc*AlBr₂ and LiBr in situ; it would thus be possible that 5b reacts with Fc*Li generated from Li⁺[Fc*AlBr₃]⁻ to furnish Fc*Fc*′AlBr (4b). Thus, the transformation of 3b to 4b promoted by a small amount of AlBr₃ should most likely be interpreted in terms of intermediate 5b, which would be formed by the reaction of 3b with AlBr₃, although the detailed mechanism for the formation of 5b from 3b is still unclear at present.

Scheme 3 Reaction of 3b with AlBr₃ to yield dibromoalumane dimer 5b, which could subsequently be converted to 4b.

The molecular structures of bis(ferrocenyl)haloalumanes 3a, 4b, and 4c in the crystalline state were determined by SC-XRD analyses (Fig. 4). The obtained structural parameters¹⁰ are summarized in Table 1 together with those theoretically optimized (B3PW91-D3(BJ)/def2TZVPP) for the bis(ferrocenyl)haloalumanes Fc*(R)AlX (X = Cl(a), Br(b), and I(c); R = Fc*(3) and Fc*′(4)).⁹ The experimentally observed structural parameters were reproduced well by the theoretical calculations, indicating that the observed structural features should be due to the intrinsic nature of the compounds with negligible perturbation from packing forces. All of the obtained bis(ferrocenyl)haloalumanes exhibit monomeric tricoordinated structures in the crystalline states due to the sterically demanding ferrocenyl groups on the aluminum center. In all cases (3a, 4b, and 5b), the sum of the bond angles around the aluminum center is ∼360° (Table 1); however, the central aluminum atoms deviate from the Cp planes (α) and approach one of the Fe atoms of the ferrocenyl groups. Notably, the shortened Al⋯Fe distances should not be attributed to crystal-packing forces, given that this characteristic structural feature was also reproduced by the theoretical calculations in the gas phase (Table 1). Generally, the deviation of the central atom (E) from the Cp-plane in a ferrocenyl-substituted compound (α = 180° – (∠Cp_cent–Cp_ipso–E)) is known as the “dip-angle”,^2b,12 which indicates the electrophilicity/Lewis acidity at the central atom,¹³ since the deviation should most likely be interpreted in terms of a 3-center-2-electron bond among the C(Cp_ipso), Fe, and E atoms (e.g., α in FcBMe₂ = 13.0°, FcSi(tBu)Me⁺ = 44.8°, and FcSnMe₂⁺ = 41.6°).¹²

Fig. 4 Molecular structures and the depictions of the central moieties of (a) 3a, (b) 4b, and (c) 4c in the crystalline state with thermal ellipsoids at 50% probability; hydrogen atoms are omitted for clarity.

Table 1 Structural parameters of bis(ferrocenyl)haloalumanes

X		Al–C/Å	Fe⋯Al/Å	Al–X/Å	a /°	b /°
a Dip-angle (α = 180° − (∠Cpcent–Cpipso–E)). b Dihedral angle between the Cp and Al planes. c Optimized at the B3PW91-D3(BJ)/def2TZVPP level.
Exp.
3a	Fc*	1.957(5)	3.623(1)	2.129(2)	0.06	64.6
Cl	Fc*	1.948(4)	3.387(1)		12.5	42.1
4b	Fc*	1.944(4)	3.605(1)	2.294(1)	1.27	57.5
Br	Fc*′	1.904(3)	2.974(1)		29.6	26.7
4c	Fc*	1.952(7)	3.614(3)	2.510(3)	2.26	59.7
I	Fc*′	1.904(10)	2.988(4)		29.1	26.2
Calc.^c
3a	Fc*	1.924	3.629	2.128	3.34	64.6
Cl	Fc*	1.925	3.253		16.9	42.1
4a	Fc*	1.930	3.626	2.134	5.77	57.5
Cl	Fc*′	1.900	3.029		26.6	26.7
3b	Fc*	1.925	3.646	2.294	4.32	59.7
Br	Fc*	1.927	3.292		15.1	26.2
4b	Fc*	1.929	3.629	2.297	6.07	64.6
Br	Fc*′	1.899	3.036		26.3	42.1
3c	Fc*	1.926	3.661	2.518	5.09	57.5
I	Fc*	1.929	3.315		14.2	26.7
4c	Fc*	1.929	3.633	2.516	6.13	59.7
I	Fc*′	1.900	3.037		26.3	26.2

---

The structural features of 3a–c and 4a–c should be discussed based on both the experimental and theoretical results. The haloalumanes that bear two Fc* groups exhibit almost C₂-symmetric structures, similar to those of Fc*₂Ge: and Fc*₂Sn:.⁷ While the Al-centers of 3a–c are sterically congested as in the case of Fc*₂Ge: and Fc*₂Sn:, those of 4a–c have enough space to exhibit effective conjugation with the Cp moiety of the Fc*′ group. The Al–C(Fc*) distances are almost the same (e.g., 3a (SC-XRD): Al–C(Fc*) = 1.957(5) and 1.948(4) Å) with slightly different Al⋯Fe distances (e.g., 3a (SC-XRD): Al⋯Fe = 3.623(1) and 3.387(1) Å). Conversely, unsymmetrically substituted bis(ferrocenyl)haloalumanes 4a, 4b, and 4c (Fc*(Fc*′)AlX; X = Cl, Br, and I) showed unsymmetrical structural features. For example, 4b showed clearly different Al–C(Fc*/Fc*′) distances (1.944(4)(Fc*) and 1.904(3)(Fc*′) Å) and remarkably different Al⋯Fe distances (3.605(1) (Fc*) and 2.974(1) (Fc*′) Å), where the Fc*′ group approaches the Al atom with shortened Al–C/Al–Fe distances, causing a large dip-angle α in the Fc*′ moiety (α(Fc*′) = 29.6° and α(Fc*) = 1.27° for 4b). Moreover, the dihedral angles between the Cp(Fc*′) and Al-planes (β) are significantly smaller (e.g., 4b: β in Al–Cp(Fc*′) = 26.7°) than those between the Cp(Fc*) and Al-planes (e.g., 4b: β in Al–Cp(Fc*) = 57.5°), indicating effective conjugation between the π(Cp of Fc*′)-orbital and the vacant p-orbital of the Al atom relative to those in the Al–Cp(Fc*) moiety. The slightly longer Al–X bonds in the Fc*(Fc*′)AlX systems relative to those in the corresponding Fc*₂AlX systems should most likely be assessed in terms of effective π-coordination from π(Cp–Fc*′) electrons to the vacant p-orbital of the Al atom due to the lower steric hindrance, which would weaken the donating resonance contribution of the lone pair of the X atom to the vacant p-orbital of the Al atom.

Conclusions

Tricoordinated bis(ferrocenyl)haloalumane derivatives were obtained from the reaction of the sterically demanding ferrocenyl lithium dimer (Fc*Li)₂ (Fc* = 2,5-bis(3,5-di-t-butyl-phenyl)-1-ferrocenyl) with aluminum trihalides (AlCl₃, AlBr₃, and AlI₃). Some of the obtained compounds were structurally characterized by single-crystal X-ray diffraction analysis. We found that the bis(ferrocenyl)bromoalumane and bis(ferrocenyl)iodoalumane undergo an intramolecular 1,1′-Al migration at the Fc* ligand to yield the corresponding unsymmetrically substituted isomers Fc*(Fc*′)AlX (X = Br and I). These isolated bis(ferrocenyl)haloalumanes exhibit planar tricoordinated structures, suggesting remarkable electrophilicity/Lewis acidity due to the vacant 3p-orbital at the Al center. Further investigations into the reactivity of the obtained bis(ferrocenyl)haloalumanes are currently in progress in our laboratories.

Author contributions

The project was designed and conducted by T. S. Experimental work was carried out by T. A., and theoretical calculations were performed by K. S. and T. S., who also partially collected the analytical data. All authors contributed to writing the manuscript.

Data availability

The data supporting this article have been included as part of the ESI.‡ Crystallographic data for 2b·(thf), 3a, 4b, 4c, and 5b can be obtained viahttps://www.ccdc.cam.ac.uk at the Cambridge Crystallographic Data Centre (CCDC) under reference numbers CCDC 2401680 (2b·(thf)), 2401681 (3a), 2401682 (4b), 2401683 (4c), and 2403232 (5b).

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was financially supported by MEXT JSPS KAKENHI grants 24K08388, 23H01943, 22K18332, and 21KK0094, by the Collaborative Research Program of the Institute for Chemical Research at Kyoto University (2024-14), a project subsidized by the New Energy and Industrial Technology Development Organization (NEDO), and by JST CREST grant JPMJCR19R4. We would also like to thank Mr Toshiaki Noda for the expert manufacturing of custom-tailored glassware. We acknowledge the generous assistance of SPring-8 with X-ray diffraction measurements under proposal numbers BL02B1: 2023A1539, 2023A1771, 2023A1785, 2023A1794, 2023A1859, 2023A1925, 2023B1806, 2023B1878, 2023B1675, and 2024A1857. Computational time was generously provided by the Supercomputer Laboratory at the Institute for Chemical Research (Kyoto University) and the Research Center for Computational Science, Okazaki, Japan (projects: 24-IMS-C377/24-IMS-C397).

References

1. For a review, see: H. Yamamoto and H. Nozaki, Angew. Chem., Int. Ed., 2003, 17, 169–175.
2. (a) C. L. Lund, J. A. Schachner, J. W. Quail and J. Müller, Organometallics, 2006, 25, 5817–5823; (b) J. A. Schachner, C. L. Lund, J. W. Quail and J. Müller, Organometallics, 2005, 24, 785–787.
3. (a) B. Wrackmeyer, E. V. Klimkina, T. Ackermann and W. Milius, Inorg. Chim. Acta, 2009, 362, 3941–3948; (b) B. Wrackmeyer, E. V. Klimkina, T. Ackermann and W. Milius, Inorg. Chem. Commun., 2007, 10, 743–747; (c) R. D. Rogers, W. J. Cook and J. L. Atwood, Inorg. Chem., 1979, 18, 279–282.
4. (a) B. Wrackmeyer, E. V. Klimkina and W. Milius, Eur. J. Inorg. Chem., 2009, 2009, 3155–3162; (b) B. Wrackmeyer, E. V. Klimkina and W. Milius, Eur. J. Inorg. Chem., 2009, 2009, 3163–3171; (c) F. Voigt, A. Fischer, C. Pietzsch and K. Jacob, Z. Anorg. Allg. Chem., 2001, 627, 2337–2343.
5. (a) H. Braunschweig, G. K. B. Clentsmith, S. Hess, T. Kupfer and K. Radacki, Inorg. Chim. Acta, 2007, 360, 1274–1277; (b) B. Bagh, N. C. Breit, K. Harms, G. Schatte, I. J. Burgess, H. Braunschweig and J. Muller, Inorg. Chem., 2012, 51, 11155–11167; (c) B. Bagh, G. Schatte, J. C. Green and J. Muller, J. Am. Chem. Soc., 2012, 134, 7924–7936; (d) H. Braunschweig, C. Burschka, G. K. B. Clentsmith, T. Kupfer and K. Radacki, Inorg. Chem., 2005, 44, 4906–4908; (e) J. L. Atwood and A. L. Shoemaker, J. Chem. Soc., Chem. Commun., 1976, 536–537; (f) W. Clegg, E. Crosbie, S. H. Dale-Black, E. Hevia, G. W. Honeyman, A. R. Kennedy, R. E. Mulvey, D. L. Ramsay and S. D. Robertson, Organometallics, 2015, 34, 2580–2589.
6. T. Sasamori, Y. Suzuki and N. Tokitoh, Organometallics, 2014, 33, 6696–6699.
7. (a) Y. Suzuki, T. Sasamori, J.-D. Guo and N. Tokitoh, Chem. – Eur. J., 2018, 24, 364–368; (b) K. Iijima, K. Sugamata, P. K. Majhi and T. Sasamori, Eur. J. Inorg. Chem., 2024, 27, e202400198.
8. D. Dhara, F. Fantuzzi, M. Harterich, R. D. Dewhurst, I. Krummenacher, M. Arrowsmith, C. Pranckevicius and H. Braunschweig, Chem. Sci., 2022, 13, 9693–9700.
9. Chemical data of the newly obtained compounds and details of experimental procedures as well as theoretical calculations are shown in the ESI.‡.
10. Supplementary crystallographic data for 2b·(thf), 3a, 4b, 4c, and 5b were deposited at the Cambridge Crystallographic Data Centre (CCDC) under reference numbers CCDC 2401680 (2b·(thf)), 2401681 (3a), 2401682 (4b), 2401683 (4c), and 2403232 (5b).
11. Calculated relative energies (kcal mol⁻¹) of Fc*Fc*′AlX (4) vs. the corresponding Fc*₂X (3): 4a (X = Cl): ΔE_zero = +0.49, ΔG = +0.82; 4b (X = Br): ΔE_zero = −0.18, ΔG = +0.23; 4c (X = I): ΔE_zero = −0.65, ΔG = −0.23; all calculated at the B3PW91-D3(BJ)/def2TZVPP level.
12. (a) K. Müther, P. Hrobárik, V. Hrobáriková, M. Kaupp and M. Oestreich, Chem. – Eur. J., 2013, 19, 16579–16594; (b) J. Lukasser, H. Angleitner, H. Schottenberger, H. Kopacka, M. Schweiger, B. Bildstein, K. H. Ongania and K. Wurst, Organometallics, 1995, 14, 5566–5578; (c) K. Venkatasubbaiah, L. N. Zakharov, W. S. Kassel, A. L. Rheingold and F. Jäkle, Angew. Chem., Int. Ed., 2005, 44, 5428–5433; (d) B. E. Carpenter, W. E. Piers, M. Parvez, G. P. A. Yap and S. J. Rettig, Can. J. Chem., 2001, 79, 857–867; (e) M. Scheibitz, M. Bolte, J. W. Bats, H.-W. Lerner, I. Nowik, R. H. Herber, A. Krapp, M. Lein, M. C. Holthausen and M. Wagner, Chem. – Eur. J., 2005, 11, 584–603; (f) H. Braunschweig, C.-W. Chiu, D. Gamon, M. Kaupp, I. Krummenacher, T. Kupfer, R. Müller and K. Radacki, Chem. – Eur. J., 2012, 18, 11732–11746; (g) J. A. Gamboa, A. Sundararaman, L. Kakalis, A. J. Lough and F. Jäkle, Organometallics, 2002, 21, 4169–4181.
13. It is difficult to discuss the Lewis acidity of the bis(ferrocenyl)haloalumanes because the complexation with a base would be perturbed by the steric hindrance. On the other hand, the electrophilicity can be estimated based on their low-lying LUMO levels, which are shown in Fig. S24.‡.

---

Footnotes

† Dedicated to Prof. Dr Shigeru Nagase on the occasion of his 77th birthday.

‡ Electronic supplementary information (ESI) available. CCDC 2401680–2401683 and 2403232. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d4dt03233k

---