Skip to main content
NCBI home page
Chemical Science logoLink to Chemical Science
. 2026 Sep 24. Online ahead of print. doi: 10.1039/d6sc04717c

C–H magnesiation of unactivated arenes by a bulky β-diketiminato magnesium n-butyl complex

Louis J Morris a,†,✉, Agustín Morales a, Andrew S S Wilson a, Michael S Hill a,✉, Claire L McMullin a,✉
PMCID: PMC13635820  PMID: 42835394

Abstract

The bulky monomeric n-butyl magnesium complex [(AdBDI)MgnBu] (1Ad; AdBDI = HC(C(Ad)Ndipp)2, dipp = 2,6-diisopropylphenyl; Ad = 1-adamantyl) deprotonates benzene to provide direct access to the corresponding phenylmagnesium derivative [(AdBDI)MgPh] (3Ad) which, in turn, can be utilised in palladium-catalysed Kumada biaryl cross-coupling. By contrast, when the less sterically demanding methyl-substituted ligand, MeBDI (HC(C(Me)Ndipp)2) is used, the deprotonation reaction is slower and competitive with β-hydride elimination of 1-butene. The enhanced steric demand of 1Ad results in the deprotonation of benzene, even in the presence of more acidic 1,2,3,4-tetramethylcyclopentadiene (Cp4MH). In contrast, Cp4MH is deprotonated by [(MeBDI)Mg(µ-nBu)]2 ([1Me]2) in benzene to yield [(MeBDI)Mg(η5-Cp4M)] (4). Compound 1Ad also mediates the kinetically undiscriminating magnesiation of the C(sp3)–H and meta- and para-C(sp2)–H bonds of toluene, ortho-xylene, and meta-xylene at 120 °C, albeit para-xylene is deprotonated exclusively in the C(sp3)–H position while 1,3-di-tertbutylbenzene is selectively deprotonated at the 5-position of the aromatic ring. Both 1Ad and [1Me]2 mediate the directed ortho-magnesiation of anisole. The mechanism of benzene- and toluene deprotonation and a possible origin for the contrasting reactivity of 1Ad and [1Me]2 has been investigated computationally.

A bulky β-diketiminate ligand enables the synthesis of aryl magnesium complexes via direct C–H deprotonation of arenes by the corresponding n-butyl magnesium reagent.graphic file with name d6sc04717c-ga.webp

Introduction

Grignard's ubiquitous aryl magnesium derivatives play a leading role as nucleophilic reagents in synthetic chemistry. In addition to the direct delivery of an aryl nucleophile to electrophilic substrates, the transition metal-catalysed Kumada cross-coupling of arylmagnesium and organohalide provides an especially powerful protocol for the formation of C–C bonds. Almost invariably, however, the organomagnesium reagents themselves must be prepared through the heterogeneous Grignard reaction of metallic magnesium with an aryl halide. Magnesiation of C–H bonds provides an alternative synthetic approach to accessing Mg–Ar reagents. Following early examples of the ortho-magnesiation of activated arenes by the Hauser base, (TMP)MgBr, and diamide Mg(TMP)2 (TMP = 2,2,6,6-tetramethylpiperidyl),1 the related heteroleptic β-diketiminato amide-, n-butyl-, and hydride-derivatives [(MeBDI)Mg(TMP)] (MeBDI = HC(C(Me)Ndipp)2, dipp = 2,6-diisopropylphenyl), [(MeBDI)MgnBu·thf] (1Me·thf), [(MeBDI)Mg(µ-nBu)]2 ([1Me]2), and [(MeBDI)Mg(µ-H)]2 ([2Mg,Me]2) have been more recently utilised to magnesiate comparatively acidic heterocycles such as benzoxazole (Scheme 1a), benzothiozole, N-methyl benzimidazole, N-methylimidazole, pyrazine, triazoles and substituted pyridines.2–4 Notably, however, magnesiation of the less acidified C–H bonds (i.e. pKa > 40) of hydrocarbon arenes has only been accomplished in heterometallic systems, where the combination of magnesium with a group 1 metal alkyl, amide, or halide generates alkali metal magnesiate species with substantially increased basicity compared to the monometallic magnesium precursor.5,6 For example, ‘Turbo’ Hauser bases (TMP)2Mg·2LiCl7,8 and (TMP)MgCl·LiCl9,10 have been utilised in C–H metalation–functionalisation protocols (Scheme 1b), whilst advances in the chemistry of amido/alkyl alkali metal magnesiates derived from Na/K–Mg nBu/TMP have led to the synthesis and isolation of discrete molecular heterometallic aryl complexes via a variety of selective C(sp2)–H metalation reactions with benzene, toluene and naphthalene, as well as various heteroatom-substituted benzene derivatives (Scheme 1c).5,11–17

Scheme 1. (a) Monometallic C–H magnesiation using β-diketiminato magnesium bases;2 (b) arene functionalisation via directed metalation with a ‘Turbo’ Hauser base (DG = directing group; E+ = electrophile);9,10 (c) metallation of benzene and toluene by a sodium magnesiate.12 dipp = 2,6-di-isopropylphenyl.

Scheme 1

Open in a new tab

The strictly monometallic magnesiation of aromatic hydrocarbons is restricted to photo-activated Mg(i) dimers [LMg–MgL] or in situ-generated Mg(i) radicals. Jones and co-workers reported that UV irradiation of the Mg(i) dimer [(MeBDI)Mg]2 in benzene provided access to [(MeBDI)MgPh] (3Me) and hydride complex [2Mg,Me]2via a dimeric cyclohexadienyl intermediate [((MeBDI)Mg)2(µ2,µ4-C6H6)] (Scheme 2a).18 The same reaction in toluene, ortho-xylene, meta-xylene, or para-xylene resulted in the respective isolation of C(sp2)–Mg products [(MeBDI)Mg(C6H4-3-Me)], [(MeBDI)isoMg(C6H4-2,3-Me2)], [(MeBDI)Mg(C6H4-3,5-Me2)], and [(MeBDI)Mg(C6H4-2,5-Me2)], alongside [2Mg,Me]2.18 Photolysis of the related guanidinate derivative [(iPr2NC(Ndipp)2)Mg]2 in toluene, ortho-xylene, meta-xylene, or mesitylene provided dimeric aryl-hydride complexes [(iPr2NC(Ndipp)2)2Mg2(µ-Ar)(µ-H)] via selective C(sp2)–H activation, providing a mixture of magnesiated isomers in the case of toluene and xylenes, in preference to the more kinetically accessible and weaker C(sp3)–H bonds (Scheme 2a).19 Additionally, Crimmin and co-workers have reported the palladium-catalysed magnesiation of benzene by [(MeBDIMes)Mg]2 (MeBDIMes = HC(C(Me)NMes)2; Mes = 2,4,6-trimethylphenyl) to provide [((MeBDIMes)Mg)2(µ-Ph)(µ-H)].20 Palladium-21 and calcium22 catalysts have also been utilised to mediate the C–H alumination of benzene, toluene and xylenes through heterometallic cooperativity.

Scheme 2. (a) Photolytic magnesiation of arenes by Mg(i) dimers;18,19 (b) magnesiation of benzene by transient or ‘masked’ Mg–Ca compounds;27 (c) arene functionalisation strategies involving β-diketiminato alkaline-earth hydride complexes ([2Ae]2): nucleophilic alkylation of benzene by calcium and strontium n-alkyl complexes,28,29 biaryl cross-coupling via calcium phenyl complex32 and zincation of hydrocarbon arenes by [2Ca]2 and [Zn(TMP)2].35 Note: not all byproducts and reaction stoichiometries are shown.

Scheme 2

Open in a new tab

Harder and co-workers reported the isolation of [(MeBDIdipep)MgPh·(κ1-TMEDA)] (MeBDIdipep = HC(C(Me)Ndipep)2, dipep = 2,6-di-isopentylphenyl; TMEDA = N,N,N′,N′-tetramethylethylenediamine) alongside [((MeBDIdipep)Mg)2(µ2,µ4-C6H6)] when [(MeBDIdipep)MgI(OEt2)] was reduced by sodium in benzene/TMEDA solution.23 Related to this work, the same group reported that attempts to synthesise compounds with Mg–Ca bonds by salt metathesis reactions of the unique Mg0 complex24 [(tBuBDIdipep)MgNa]2 with CaII reagents led to intramolecular C(sp3)–H activation25,26 or benzene C–H activation to yield [(tBuBDIdipep)MgPh] and [(MeBDIdipep)Ca(µ-H)]2via a presumed heterometallic cyclohexadienyl intermediate [((tBuBDIdipep)Mg)(µ2,µ4-C6H6)(Ca(MeBDIdipep))] (Scheme 2b).27 Isolable heterometallic naphthalene and anthracene derivatives, [((tBuBDIdipep)Mg)(µ2,µ4-napth/anth)(Ca(MeBDIdipep))], also produce [(tBuBDIdipep)MgPh] and [(MeBDIdipep)Ca(µ-H)]2 upon reaction with benzene (Scheme 2b).27 Notably, each of these examples (Scheme 2a and b) generates a stoichiometric equivalent of magnesium or calcium hydride in addition to the magnesium aryl moiety. Thus, a selective monometallic C–H deprotonation–magnesiation protocol remains to be achieved for wholly hydrocarbon arenes.

A major focus of our own research has been the reactivity of a base-free β-diketiminato calcium hydride complex, [(MeBDI)Ca(µ-H)]2 ([2Ca]2). Insertion of n-alkenes into the Ca–H bonds provides heteroleptic calcium n-alkyls, [(MeBDI)Ca(µ-R)]2, which react with benzene-d6via nucleophilic substitution to provide alkylated benzene RC6D5 derivatives under elimination of deuteride as a leaving group (Scheme 2c).28 Harder and co-workers reported that the strontium analogue [(MeBDIdipep)Sr(µ-H)]2 ([2Sr]2) and its n-alkyl derivatives display an even higher degree of nucleophilic reactivity towards benzene.29 Whilst unactivated n-alkenes also insert into the Mg–H bonds of β-diketiminato magnesium hydride complex [2Mg,Me]2 at elevated temperatures,30 nucleophilic arene alkylation akin to that observed for the heavier congeners was not observed. Arene deprotonation by [2Ae]2 or the corresponding alkyl derivatives has not been previously observed under the conditions studied. Heteroleptic calcium aryl complexes such as [(MeBDI)Ca(µ-Ph)]2 could, however, be accessed via reductive transmetallation of HgAr2,31,32 and employed in the uncatalysed biphenyl cross-coupling with aryl bromides (Scheme 2c).32 Extending this transmetallation approach to diorganozinc reagents provides access to alkyl calcium complexes33 including the previously inaccessible methyl derivative [(MeBDI)Ca(µ-Me)]2via organo-hydride calcium zincate intermediates.34 In a further recent advance,35 we found that reacting the diamidozinc reagent [Zn(TMP)2] with [2Ca]2 in aromatic solvents benzene, toluene or mesitylene results in C(sp2)–H metalation and transmetallation of the MeBDI ligand to provide access to aryl zinc complexes [(MeBDI)ZnAr], the phenyl variant of which was utilised in the palladium-catalysed Negishi biaryl cross-coupling via a telescoped one-pot procedure (Scheme 2c). Whilst the same reaction in methylcyclohexane provided the highly basic calcium amide, [(MeBDI)Ca(TMP)], C(sp2)–H metalation chemistry in this system was ultimately deduced to be dependent on the synergistic action of both calcium and zinc via a hitherto unknown mechanism.

Here, we show that an n-butyl magnesium complex, [(AdBDI)MgnBu] (1Ad), supported by a very bulky β-diketiminate ligand, AdBDI (HC(C(Ad)Ndipp)2; Ad = 1-adamantyl) deprotonates benzene, toluene, xylenes, and anisole to yield the corresponding aryl- and benzyl-magnesium products at 100–120 °C without requisite addition of any further heterometallic component. AdBDI counterintuitively promotes both faster reaction rate and higher selectivity compared to the significantly less bulky MeBDI, such that, whilst [1Me]2 also readily mediates the directed ortho-metallation of anisole at 100 °C, magnesiation of benzene is extremely sluggish and competitive with β-hydride elimination. The use of [(AdBDI)MgPh] (3Ad) obtained via benzene deprotonation as a source of nucleophilic phenyl in Kumada biaryl cross-coupling is also described.

Results and discussion

Magnesiation of benzene and 1,2,3,4-tetramethylcyclopentadiene

Our serendipitous entry-point to C–H magnesiation chemistry was provided by attempts to prepare 1,2,3,4-tetramethylcylopentadienyl (Cp4M) magnesium derivatives via protonolysis of n-butyl magnesium β-diketiminates [1Me]2 and 1Ad with 1,2,3,4-tetramethylcylopentadiene (Cp4MH). Compound 1Ad was straightforwardly synthesised by protonolysis of MgnBu2 with β-diimine, AdBDIH, at 60 °C (see SI for synthesis and characterisation of AdBDIH and 1Ad) and is structurally similar to other bulky monomeric BDI-magnesium n-butyl complexes [(RBDIAr)MgnBu] (1Ar,R; Ar, R = dipp, tBu;36 dipep, tBu;37 dipp, iPr38). One molar equivalent (relative to magnesium) of Cp4MH was added to a benzene-d6 solution of 1Ad or [1Me]2 and the sealed J. Young's NMR tube was heated to 100 °C (Scheme 3). In the case of [1Me]2 (Scheme 3a), periodic analysis by in situ1H NMR spectroscopy revealed the gradual disappearance of both the methylene resonance of Cp4MH (δ 2.57 ppm) and backbone methine resonance of [1Me]2 (δ 4.92 ppm) over the course of 6 days and appearance of a new compound (4), characterised by two singlets of equal relative intensity at δ 5.81, 4.70 ppm. Gradual evaporation of the reaction solvent at room temperature provided colourless needle-shaped single crystals suitable for X-ray diffraction analysis, the results of which confirmed the formation of the cyclopentadienyl magnesium derivative [(MeBDI)MgCp4M] (4), which was isolated in 78% yield and fully characterised.

Scheme 3. (a) reaction of [1Me]2 with Cp4MH to give 4 and (b) the reaction of 1Ad with benzene(-d6) in the presence or absence of Cp4MH to give 3Ad(-d5). Isolated yields are shown with 1H NMR spectroscopic yields in parentheses.

Scheme 3

Open in a new tab

The same reaction procedure starting from 1Ad (Scheme 3b) resulted in a similar disappearance of the BDI-backbone methine signal (δ 5.67 ppm) and the magnesium-bound methylene multiplet (δ −0.62 ppm) in the 1H NMR spectrum and near-quantitative formation of a single new BDI-containing species (3Ad-d5) with a methine resonance at δ 5.74 ppm after 7 days at 100 °C. Whilst further scrutiny of the 1H NMR spectrum confirmed the evolution of n-butane (δ(CH3) 0.86 ppm), no evidence for the consumption of Cp4MH was observed and no resonance attributable to deprotonated Cp4M was detected. Recrystallisation of the crude product from a saturated n-hexane/toluene solution at room temperature provided colourless block-shaped single crystals, which were isolated in 42% yield. Subsequent X-ray diffraction analysis revealed the structure of 3Ad-d5 as the magnesium phenyl-d5 complex, [(AdBDI)Mg(C6D5)], formed by deprotonation of the reaction solvent. The same product was obtained in quantitative 1H NMR spectroscopic yield when the reaction was repeated in the absence of Cp4MH. When 1Ad was heated in proteo-benzene for 7 days in either the presence or absence of Cp4MH, the non-deuterated complex [(AdBDI)MgPh] (3Ad) was isolated as a pale-yellow solid in 79% yield after removal of volatiles (quantitative by NMR). The ortho-CH protons of the magnesium-bound phenyl moiety in compound 3Ad display a characteristic doublet at δ 6.17 ppm in the 1H NMR spectrum, which integrates 2H relative to the BDI-backbone methine (δ 5.74 ppm), whilst the meta- and para-CH protons give rise to a multiplet (3H) at δ 7.05 ppm. Resonances at the same chemical shifts were observed in the corresponding 2H NMR spectrum of 3Ad-d5. Notably, benzene-d6 solutions of 3Ad gradually transform into the isotopically labelled compound 3Ad-d5 in quantitative 1H NMR spectroscopic yields after 3 days at 100 °C, due to exchange with the NMR solvent.

The crystal structure of compound 4 contains a pseudo-3-coordinate magnesium centre in a distorted trigonal pyramidal geometry with the magnesium essentially co-planar to the BDI ligand backbone (Fig. 1a). Although disorder with respect to rotation about the Cp(centroid)–Mg axis prevents detailed discussion of distances and angles, the bulky Cp4M ligand clearly adopts an η5-coordination mode, with identical Mg1–Cp(plane) and Mg1–Cp(centroid) distances within error (2.055(3), 2.056(3) Å, respectively; calculated for the major component of the disordered Cp4M ligand). The Mg1–Cp(centroid) distance is expectedly greater than those of [(MeBDI)MgCp] (2.0463(9) Å)39 and [(MeBDIMes)MgCp] (2.0391(8), 2.0377(9) Å), but shorter than that in the even more sterically congested derivative [(MeBDI)MgCp*] (2.0781(6) Å).40

Fig. 1. Displacement ellipsoid plots (30% level) of (a) 4 and (b) 3Ad. Minor component of the disordered Cp4M in 4, and of the disordered isopropyl and adamantyl moieties in 3Ad are omitted. Hydrogen atoms are also omitted for clarity. Distances (Å) and angles (°) for 4:* Mg1–N1 2.0670(12), Mg1–N2 2.0708(11), Mg1–C30A 2.331(4), Mg1–C31A 2.364(4), Mg1–C32A 2.396(5), Mg1–C33A 2.412(10), Mg1–C34A 2.400(6), Mg1–Cp(centroid) 2.056(3), N1–Mg1–N2 92.55(4), N1–Mg1-(centroid) 135.27(10), N2–Mg1-(centroid) 132.13(10). *Distances and angles involving the Cp4M moiety refer to major component only. Distances (Å) and angles (°) for 3Ad: Mg1–N1 2.0162(10), Mg1–N2 2.0091(10), Mg1–C48 2.1052(13), C48–C49 1.402(2), C48–C53 1.404(2), C49–C50 1.3926(19), C50–C51 1.376(2), C51–C52 1.375(2), C52–C53 1.390(3), N1–Mg1–C48 133.39(5), N2–Mg1–N1 94.74(4), N2–Mg1–C48 131.72(5).

Fig. 1

Open in a new tab

The crystal structure of 3Ad (Fig. 1b) also presents a 3-coordinate distorted trigonal planar coordination geometry at magnesium (Σangles(Mg) = 359.84°), with a Mg1–C48 distance (2.1052(13) Å) that is essentially the same as that in the less sterically hindered derivative 3Me (2.095(3) Å).41 Additionally, the phenyl ligand is almost co-planar with the BDI ligand (C53–C48–Mg1–N2 = 5.30(16)°; C49–C48–Mg1–N1 5.25(16)°), such that the ortho-CH atoms point towards the centroids of the flanking dipp phenyl rings. Due to the effect of the bulky backbone adamantyl groups, the dipp groups are pushed forwards, resulting in larger Cipso–N–C angles (129.02(9), 127.57(10)°) compared to 3Me (120.39(11)°) and significantly smaller fold angles between the mean planes of the dipp rings (3Ad = 104.7(5)°, 3Me = 126.6(1)°). The bending angles between the N–Cipso bonds and the dipp centroids are highly asymmetric (centroid-C24–N1 = 170.45(10), centroid-C36–N2 = 174.59(10)°), but smaller on-average (172.52(10)°) compared to 3Me (174.30(14)°). The Cipso–N–C angles in 3Ad are also larger than those in the similarly 3-coordinate [(tBuBDIdipep)MgPh] (3tBu,dipep; average Cipso–N–C = 126.52°),24 whilst the fold angle between the dipp rings and the centroid-Cipso–N angles are both more acute than those observed in 3tBu,dipep (average fold angle = 114.44°; average centroid-Cipso–N = 174.44°), probably due to the bulkier dipep groups in the latter compound.

In contrast to 1Ad, when a benzene-d6 solution of [1Me]2 was heated to 100 °C in the absence of Cp4MH, negligible change to the 1H NMR spectrum was observed after 1 week, and phenyl complex 3Me-d5 was not observed. Heating the same solution at an increased temperature of 120 °C for 2 weeks, however, resulted in conversion of most of the starting complex into two new BDI-containing species, which were identified as the known phenyl 3Me-d5 (39% 1H NMR spectroscopic yield) and hydride [2Mg,Me]2 (48%) complexes by comparison to the reported NMR data (Scheme 4).41,42 The presence of 1-butene in the same 1H NMR spectrum suggests that [2Mg,Me]2 is formed through β-hydride elimination of [1Me]2. To identify whether the magnesium hydride [2Mg,Me]2 reacts with the solvent, a separately synthesised sample of [2Mg,Me]2 was also heated at 120 °C in benzene-d6 for 2 weeks. Subsequent analysis of the reaction mixture by 1H NMR spectroscopy confirmed the selective formation of 3Me-d5 in 38% spectroscopic yield.

Scheme 4. Reaction of [1Me]2 and [2Mg,Me]2 with benzene-d6. 1H NMR spectroscopic yields shown in parentheses.

Scheme 4

Open in a new tab

The divergence in reactivity of 1Ad compared to [1Me]2 and [2Mg,Me]2, which further contrasts to their heavier congeners, [(MeBDI)Ca(µ-R)]2 (R = n-alkyl) and [2Ca]2, is striking. Whilst the bulky, monomeric magnesium n-butyl derivative 1Ad reacts as a base, deprotonating benzene under forcing conditions via C–H activation, the analogous reaction for [1Me]2 is apparently less kinetically facile and competitive with β-hydride elimination to generate [2Mg,Me]2, which also slowly deprotonates benzene to provide 3Me. By contrast, [2Ca]2 readily undergoes H/D exchange at 60 °C in benzene-d6via nucleophilic aromatic substitution.43 Comparable nucleophilic behaviour is observed for [(MeBDI)Ca(µ-R)]2, yielding [(MeBDI)Ca(µ-D)]2 and C6D5R at 60 °C rather than solvent deprotonation.28

Steric profile of backbone-substituted β-diketiminate ligands

To begin to rationalise the faster, and more selective, reaction of 1Ad with benzene compared to the less sterically hindered, yet dimeric [1Me]2, the steric bulk of 1Ad was compared to other crystallographically characterised monomeric 1Ar,R derivatives (Ar, R = dipp, tBu;36 dipep, tBu;37 dipp, iPr38), and to the computationally optimised structure of the monomerised form of [(MeBDI)MgnBu] (1Me) (see SI for computational details). Comparing the crystal structure of 1Ad (Fig. 2a) with the calculated structure of 1Me, it is apparent that the bulky adamantyl groups force the flanking dipp groups forwards, constraining the binding pocket around the magnesium centre (Fig. 2c). This can be quantified by comparing the Cipso–N–C angles (1Ad = 126.31(8), 127.81(8)°), which are similar to those of 1dipp,tBu (126.4(4), 127.1(4)°)36 and 1dipep,tBu (126.13(9), 126.09(10)°),37 but significantly greater than that of 1dipp,iPr (120.72(12), 120.78(12)°),38 and 1Me (120.02, 120.05°). Additionally, the fold-angle between the mean planes of the dipp phenyl rings is significantly smaller for 1Ad (112.53(4)°) compared to 1Me (131.51°) and 1dipp,iPr (125.89(6)°), but slightly greater than that of 1dipp,tBu (110.4(2)°). It should, however, be noted that the fold angle is also influenced by bending of the Cipso–N bond out of the mean plane of the dipp ring. The dipp centroid-Cipso–N angles for 1Ad are smaller on average (centroid-C24–N1 = 176.93(9)°, centroid-C36–N2 = 171.66(8)°, average = 174.30°) compared to 1Me (176.25, 176.32°), but highly asymmetric, possibly due to attractive dispersion forces between the dipp group and the n-butyl chain. Visual inspection of the crystal structure of 1Ad in a space filling representation (Fig. 2b) indicates that the magnesium centre remains relatively accessible, especially if its coordination environment should distort to a more pyramidal geometry. A more extensive comparison of the buried volume (%Vbur) and steric mapping using the SambVca web tool44 across the series of n-butyl complexes 1Ad and 1Ar,R shows that steric crowding of the metal centre in AdBDIdipp (1Ad, %Vbur = 65.7%, XRD) is comparable to tBuBDIdipp (1dipp,tBu, %Vbur = 67.4%, XRD)36 and tBuBDIdipep (1dipep,tBu, %Vbur = 66.0%, XRD),37 but significantly greater than MeBDIdipp (1Me, %Vbur = 60.3%, DFT), with iPrBDIdipp (1dipp,iPr, %Vbur = 62.6%, XRD)38 intermediate in the series. The distribution of steric bulk around the metal centre (i.e. the shape of the coordination ‘pocket’) is similar across the series, but the bulky backbone groups clearly cause the the N-aryl substituents to project further forwards (Fig. 2d; SI Section 3).

Fig. 2. (a) Displacement ellipsoid plot (30%) of the crystal structure of 1Ad; (b) space filling representation of the crystal structure of 1Ad; (c) overlay of the crystal structure of 1Ad (red) with the calculated structure of 1Me (blue); (d) steric contour maps generated by the SambVca web tool44 from the crystal structure of 1Ad and the calculated structure of 1Me including hydrogen atoms and using the magnesium centre to define the centre of a sphere of radius 3.5 Å. Hydrogen atoms are omitted from (a) for clarity. Disordered solvent (methylcyclohexane) in the crystal structure of 1Ad is omitted. Selected distances (Å) and angles (°) for 1Ad: Mg1–N1 2.0257(9), Mg1–N2 2.0266(9), Mg1–C48 2.1074(12), N1–Mg1–N2 95.29(4), N1–Mg1–C48 132.81(4), N2–Mg1–C48 131.76(4), C24–N1–C1 126.31(8), C36–N2–C13 127.81(8). See Table S13 in the SI for further comparison of distances and angles for magnesium n-butyl complexes 1Ar,R.

Fig. 2

Open in a new tab

Calculation of reaction pathway by DFT

To gain further mechanistic insight, reaction pathways for the reaction of 1Ad and [1Me]2 with benzene (Fig. 3) and the competing β-hydride elimination (Fig. 4) were calculated using DFT methods at the BP86-D3BJ(SMD = C6H6)/def2-TZVPP//BP86/def2-SVP level. Starting from monomeric n-butyl complexes 1R (R = Ad, Me), deprotonation of benzene was calculated to proceed via an encounter complex, AR (AAd = 3.0 kcal mol−1, AMe = 2.7 kcal mol−1), the transition state TS(AR–BR) (R = Ad, 38.9 kcal mol−1; R = Me, 32.9 kcal mol−1), and dissociation of n-butane from intermediate BR to provide 3Ad (−5.7 kcal mol−1) or 3Me (−7.6 kcal mol−1). Whilst the transition state TS(AR–BR) is 6 kcal mol−1 lower for the methyl-substituted system, the dimeric n-butyl complex [1Me]2 was calculated to reside 2.3 kcal mol−1 lower on the potential energy surface relative to monomeric 1Me. Thus, when the prerequisite dissociation of dimeric [1Me]2 is accounted for, the total calculated barrier to benzene deprotonation is comparable for both the adamantyl and methyl cases.

Fig. 3. DFT-calculated potential energy surface (BP86-D3BJ(SMD = C6H6)/def2-TZVPP//BP86/def2-SVP) for the deprotonation and mangesiation of benzene by [1Me]2 or 1Ad to give respective phenyl complexes 3Me and 3Ad, with a depiction of the calculated transition state TS(AAd–BAd).

Fig. 3

Open in a new tab

Fig. 4. DFT-calculated potential energy surface (BP86-D3BJ(SMD = C6H6)/def2-TZVPP//BP86/def2-SVP) for the β-hydride elimination of 1-butene from [1Me]2 or 1Ad.

Fig. 4

Open in a new tab

Closer inspection of the calculated intrinsic reaction coordinate for R = Ad revealed an inflection point at 20.1 kcal mol−1, prior to TS(AAd–BAd) (38.9 kcal mol−1). Although the structure at this point (A′Ad) is not a local minimum, it is informative to note that the BDI ligand backbone is twisted (NCCN torsion = 16.7°) and the magnesium centre projected some 1.219 Å from the mean plane of the ligand backbone, whilst the incoming benzene molecule remains separated from the metal by more than 7 Å (see Table S14 in the SI). The substantial structural distortion accounts for more than half of the total energy barrier, and results in far greater accessibility of the n-butyl base and magnesium centre to the substrate. In TS(AAd–BAd) (Fig. 3), the magnesium centre is displaced from the mean plane of the β-diketiminate ligand backbone by 0.714 Å, whilst the ligand backbone itself is somewhat twisted, as indicated by the NCCN torsion angle of 11.9°. For R = Me, significantly less structural reorganisation is required; the magnesium centre is relatively less crowded for TS(AMe–BMe) (Fig. S291), such that it is only projected from the BDI mean plane by 0.490 Å, and the NCCN torsion angle is essentially nil. In the transition state TS(AR–BR), the Mg–CBu bond is elongated (R = Ad, 2.334 Å; Me, 2.323 Å) compared to the preceding intermediate AR (R = Ad, 2.139 Å; Me, 2.135 Å), such that the proton is directly transferred from benzene to the carbanionic carbon of the n-butyl chain with a near-linear CAr–H–CBu arrangement (Fig. 3 and S291). The geometry of TS(AR–BR) closely resembles that calculated for the deprotonation of benzene by [(MeBDI)Ca(TMP)].35

The calculated reaction coordinate was not found to involve a Mg⋯π interaction, and TS(AR–BR) also does not involve any interaction of the benzene π system with the metal centre. The lack of such an interaction is in marked contrast to the nucleophilic alkylation of benzene by congeneric BDI–calcium, strontium, and ytterbium(ii) alkyl complexes. In these cases, poly-hapto π-coordination was calculated to play a crucial role in activating the arene substrate towards nucleophilic attack and stabilising the resulting Meisenheimer intermediate.28,29,45 The contrasting reaction outcomes and calculated mechanisms are consistent with the smaller size of magnesium, which causes the metal centre to be less sterically acessible to the π-face of benzene, besides leading to a more charge dense and less polarisable (harder) metal centre that is less able to interact strongly with the (soft) π-electron density of aromatic rings. Further, the significantly less polarised metal–carbon bond compared to the heavier congeners is expected to result in relatively attenuated nucleophilicity.

The occurrence of proton transfer during transition state TS(AR–BR) would be expected to yield a large primary KIE. In contradiction to this assumption, however, periodic 1H NMR analysis of the pseudo first-order reaction of 1Ad with benzene-d6 and with proteo-benzene at 102 °C revealed a very small secondary kinetic isotope effect (KIE) of Inline graphic (Fig. S264 and S265 in the SI). The significant contribution that structural reorganisation makes to the total energy barrier of TS(AR–BR)via distorted structure A′Ad (Table S14) would be expected to provide a relatively small KIE value. Given the very small magnitude of the experimental KIE and considering the dimeric mechanisms previously calculated for related strontium-29 and ytterbium-mediated45 nucleophilic alkylation reactions, therefore, an alternative mechanism was considered (see Tables S15 and S16 in the SI), where endergonic dimerization prior to deprotonation would be consistent with a secondary KIE. Attempts to computationally optimise a dimeric structure for 1Ad (i.e. [(AdBDI)Mg(µ-nBu)]2) were, however, unsuccessful. Whilst transition states involving two magnesium centres in a benzene deprotonation step were found, they are not connected to dimeric reactants or products in the IRC, and the energy barriers associated with them are prohibitively high (R = Me, 49.2 kcal mol−1; R = Ad, 52.0 kcal mol−1; see SI, Tables S15 and S16). Noting previous examples of very bulky BDI ligands adopting a ring-opened κ1-amidoimine binding mode,37 an alternative energy pathway via ring-opened two-coordinate intermediate [(κ1-AdBDI)MgnBu] was also calculated (see Fig. S307 and S308 in the SI) but found to be kinetically unviable.

Starting from monomeric n-butyl complex 1R, β-hydride elimination (Fig. 4) was calculated to traverse transition state TS(1R–ER) (R = Ad, 34.5 kcal mol−1; R = Me, 30.0 kcal mol−1), before dissociation of 1-butene from magnesium hydride-butene intermediate ER to provide a monomeric terminal hydride, GR (R = Ad, 11.7 kcal mol−1; R = Me, 10.3 kcal mol−1). Subsequent dimerization of GMe to form the experimentally observed magnesium hydride dimer [2Mg,Me]2 leads to a near-thermoneutral process (−1.4 kcal mol−1), whilst dimerization of GAd is much less advantageous, and the corresponding hypothetical dimeric hydride [2Mg,Ad]2 was calculated to reside at 6.0 kcal mol−1. Taking into account the endergonic (2.3 kcal mol−1) dissociation of dimeric [1Me]2, the total barrier to formation of [2Mg,Me]2 and 1-butene via β-hydride elimination is competitive with that for benzene deprotonation and formation of 3Me and n-butane. For 1Ad, the β-hydride elimination pathway is calculated to be endergonic by 6.0 kcal mol−1. Formation of 3Rvia benzene deprotonation by monomeric hydride GR was calculated to be endergonic (R = Ad, 4.0 kcal mol−1; R = Me, 3.5 kcal mol−1; relative to GR), traversing a transition state of 30.9 kcal mol−1 (R = Ad) or 30.1 kcal mol−1 (R = Me; see Fig. S309 in the SI).

Overall, the computational study suggests that whilst benzene deprotonation and β-hydride elimination pathways are energetically competitive for both 1Ad and 1Me, the latter process is thermodynamically disfavoured for the bulkier derivative due to the lower stability of the hypothetical µ-hydride dimer ([2Mg,Ad]2). The preference for monomeric structures exerted by the bulkier AdBDI ligand causes the β-hydride elimination pathway to be thermodynamically disfavoured. Although these results do not account for the lower reaction temperature and significantly faster reaction rate for deprotonation of benzene by 1Ad compared to [1Me]2, or the unusually low KIE, the combined experimental and computational data emphasise that even subtle differences in kinetics afforded by variations in the ligand scaffold can result in highly divergent reaction outcomes.

Deprotonation and magnesiation of toluene

The remarkable kinetic discrimination of 1Ad for benzene over the hindered, yet more acidic, Cp4MH prompted an investigation of the broader scope of arene deprotonation. Heating a toluene-d8 solution of 1Ad and Cp4MH to 100 °C for 8 days in a sealed J. Young's NMR tube resulted in 89% conversion of the starting complex to a mixture of several BDI-containing products, according to in situ1H NMR spectroscopy, and the deposition of colourless block-shaped single crystals at the meniscus. A subsequent X-ray diffraction experiment provided a structure solution which, albeit hampered by significant disorder, was fully consistent with para- and meta-magnesiation of the reaction solvent to provide the magnesium tolyl complexes [(AdBDI)Mg(C6D4-3-CD3)] (5-m-d7) and [(AdBDI)Mg(C6D4-4-CD3)] (5-p-d7) as a disordered co-crystalline mixture (Fig. 5a. Occupancy 0.76 : 0.24, respectively, see SI for details). Further unassigned electron density in the difference map, however, hinted at the possible presence of a third minor co-crystalline component and analysis of the isolated crystals by 1H NMR spectroscopy in benzene-d6 was consistent with this. The reaction was, thus, repeated in proteo-toluene at the higher temperature of 120 °C in the absence of Cp4MH. After 13 days, volatiles were removed under vacuum. Analysis of the crude product mixture by 1H and 13C NMR spectroscopy was consistent with the quantitative consumption of 1Ad to form a 1 : 2 : 2 mixture of the tolyl derivatives 5-p, 5-m, and the benzyl complex [(AdBDI)Mg(CH2C6H5)] (5′; Fig. 6, Scheme 5). Compound 5′ was also separately synthesised from [(AdBDI)MgBr] (13, vide infra) and benzyl potassium. Surprisingly, the only previous example of a crystallographically characterised three-coordinate benzyl magnesium complex is the potassium magnesiate [(tmeda)K-η6-C6H5CH2-Mg(N(SiMe3)2)].46 In the 1H NMR spectrum (Fig. 6), compound 5-m displays two characteristic doublets at δ 5.83 (J = 2.0 Hz, 1H) and 6.04 (J = 6.9 Hz, 1H) ppm, which correspond to the protons in the 2- and 6-positions of the tolyl rings, respectively. The analogous ortho-CH protons in 5-p appear as a single doublet at δ 6.08 (J = 7.6 Hz, 2H) ppm. Each of these species also displays a respective singlet at δ 2.16 and 2.08 ppm for the tolyl-methyl environment. The respective ortho-, para- and meta-benzyl-resonances of compound 5′ can be identified in the 1H NMR spectrum by a set of mutually coupled doublet of doublets, triplet of triplets, and apparent triplet signals at δ 6.21 (J = 7.9, 1.0 Hz, 2H), 6.75 (J = 7.3, 1.3 Hz, 1H), 6.95 (J = 7.6 Hz, 2H) ppm. The magnesium-bound methylene group gives a singlet (2H) in the 1H NMR spectrum at δ 1.07 ppm. Accordingly, 1Ad deprotonates toluene at the meta- and para-C(sp2)–H positions to provide the statistical mixture of tolyl isomers, as well as the benzylic C(sp3)–H position to form the benzyl derivative. The hindered ortho-C(sp2)–H remains inert under these conditions.

Fig. 5. Displacement ellipsoid plots (30%) of (a) the Mg1-containing molecule in the X-ray structure solution of the disordered co-crystal of 5-m/5-p, with the minor component (5-p) shown semi-transparent; (b) compound 5′. Hydrogen atoms, disordered solvent molecules, and the Mg2-containing molecule of 5-m/5-p are omitted for clarity. Distances and angles for 5′: Mg1–N1 2.0182(13), Mg1–N2 2.0131(14), Mg1–C48 2.1160(19), C48–C49 1.477(3), N1–Mg1–C48 127.46(7), N2–Mg1–N1 95.41(5), N2–Mg1–C48 136.53(7), C49–C48–Mg1 118.40(13). Distances and angles for 5-m/5-p are not discussed due to disorder of the tolyl moiety.

Fig. 5

Open in a new tab

Fig. 6. Selected regions of the 1H NMR spectrum (500 MHz, benzene-d6) showing the product mixture of 5-m, 5-p, and 5′ obtained from the reaction of 1Ad with toluene after removal of volatiles and dissolution in benzene-d6.

Fig. 6

Open in a new tab

Scheme 5. Reaction of 1Ad with toluene. 1H NMR spectroscopic conversions are shown in parentheses. Compounds 5-p/5-m/5′ can be isolated in a combined 45% crystalline yield as a mixture of isomers.

Scheme 5

Open in a new tab

The disordered co-crystalline mixture of isomers in the crystal structure of 5-m/5-p precludes detailed structural discussion (Fig. 5a), besides noting the planar three-coordinate magnesium centre, near-coplanarity of the tolyl ring and BDI ligand backbone, and unremarkable Mg–C distances (respectively 2.130(3), 2.1353(19) Å for 5-m and 5-p) compared to previously reported examples of three-coordinate BDI–magnesium aryl derivatives.18,41 The Mg–C distance in 5′ (Fig. 5b; 2.1160(19) Å) is significantly but expectedly shorter than that in [(tmeda)K-η6-C6H5CH2-Mg(N(SiMe3)2)] (2.187(2) Å).46 The Mg–C bond in 5′ is also contracted relative to the related 4-coordinate BDI–magnesium benzyl complex [(MeBDIdipp)Mg(CH2Ph)(thf)] (2.1325(18) Å).47

The DFT-calculated reaction pathway for toluene deprotonation by 1Ad is consistent with the experimentally observed regioselectivity (Table 1). The Gibbs energies for the respective ortho-, meta-, para-, and benzyl C(sp3)-magnesiated products plus n-butane were calculated to be −4.2, −6.6, −7.0, and −7.3 kcal mol−1. In the encounter complex, CMe, which precedes C(sp3)–H deprotonation, the magnesium is already significantly displaced from the BDI ligand plane and, thus, significantly higher in energy (15.1 kcal mol−1) compared to C (4.0 kcal mol−1), which precedes C(sp2)–H deprotonation with the magnesium centre coplanar with the ligand. The geometries of both subsequent transition states, however, are similar, with magnesium centres projecting out of the planes of the BDI ligands whilst a proton is directly transferred from toluene to the n-butyl groups via a linear C–H–C arrangement. The transition states for C(sp2)–H deprotonation were located at 42.8 kcal mol−1 for the ortho position, and at 40.0, 39.9 kcal mol−1 for the meta- and para-positions, respectively. Relative to intermediate CMe, the transition state for C(sp3)–H deprotonation (TS(CMe-DMe)) is lower (24.8 kcal mol−1) than those for C(sp2)–H deprotonation. Due to the already distorted geometry of CMe (15.1 kcal mol−1), however, the overall energy barrier to C(sp3)–H deprotonation viaTS(CMe-DMe) of 39.9 kcal mol−1, is essentially identical to that for meta- or para-C(sp2)–H deprotonation. Thus, whilst ortho-magnesiation is kinetically disfavoured, there is no kinetic discrimination between deprotonation of the remaining C–H bonds, reflecting the essentially statistical mixture of isomers observed experimentally. The significantly higher energy barrier compared to benzene deprotonation is consistent with the slower reaction and higher temperature required.

Table 1. DFT-calculated Gibbs energies (BP86-D3BJ(SMD = toluene)/def2-TZVPP//BP86/def2-SVP level, in kcal mol−1) for the formation of 5-o/m/p or 5′ and nBuH through the C(sp2)–H deprotonation of toluene by 1Ad; depictions of the calculated transition states for deprotonation of the methyl, meta- or para-C–H bonds of toluene.

graphic file with name d6sc04717c-u1.jpg
Open in a new tab

Deprotonation and magnesiation of xylenes and 1,3-di-tertbutylbenzene

Consistent with the regioselectivity observed for the reaction of 1Ad with toluene, ortho-xylene is slowly deprotonated at 120 °C to provide a mixture of the meta-magnesiated xylyl derivative [(AdBDI)Mg(C6H3-3,4-Me2)] (6) and ortho-methyl benzyl derivative [(AdBDI)Mg(CH2C6H4-2-Me)] (6′). After 10 days at 120 °C, 6 and 6′ were determined to form in respective 1H NMR spectroscopic conversions of 62% and 38% according to analysis of the crude product mixture, which was isolated free of 1Ad in 94% yield (Scheme 6). Single crystals suitable for X-ray diffraction analysis were obtained from the NMR tube of the reaction mixture at 100 °C and isolated in 14% yield. The resulting structure solution was refined as a disordered co-crystalline mixture of 6 and 6′ with the two respective components in a 0.8 : 0.2 occupancy ratio (Fig. 7a).

Scheme 6. Reaction of 1Ad with xylenes or 1,3-di-tertbutylbenzene. Reaction duration was 10 days, except the reaction with 1,3-di-tertbutylbenzene, which was stopped after 8 days. Yields refer to isolated product after crystallisation unless otherwise specified, with 1H NMR spectroscopic yields in parentheses. IIsolated yield of non-crystalline product mixture. IIIsolated yield of crystalline product as 80 : 20 co-crystalline mixture of 6/6′ after reaction at 100 °C for 11 days. IIIIsolated yield of crystalline product as 90 : 10 co-crystalline mixture of 7/7′.

Scheme 6

Open in a new tab

Fig. 7. Displacement ellipsoid plots (30% level) of (a) the co-crystal of 6 (occupancy 0.8) and 6′ (occupancy 0.2, semi-transparent); (b) 7 (occupancy 0.1, semi-transparent) and 7′ (occupancy 0.9); (c) 8; (d) 9. For clarity, hydrogen atoms are omitted. The minor components of disordered magnesium-bound xylyl or benzyl moieties are shown semi-transparent, whilst for other disordered residues, the minor component is omitted. Note that in the major component (6) of the disordered co-crystal of 6/6′, C55 is also split across two positions with respect to rotation of the xylyl moiety about the Mg1–C48 bond; only the major component of this disordered feature is shown. Crystal structure of pure 7 (without component of 7′) is shown in the SI (Fig. S277). Distances (Å) and angles (°) for 8: Mg1–N1 2.0155(11), Mg1–N2 2.0126(12), Mg1–C48 2.1174(15), C48–C49 1.482(2), C52–C55 1.516(3), N1–Mg1–C48 137.01(7), N2–Mg1–N1 94.68(4), N2–Mg1–C48 127.70(6), C49–C48–Mg1 120.10(11). Distances (Å) and angles (°) for 9: Mg1–N1 2.0184(14), Mg1–N2 2.0121(14), Mg1–C48 2.1027(16), N1–Mg1–C48 133.26(6), N2–Mg1–N1 94.86(6), N2–Mg1–C48 131.80(7), C49–C48–Mg1 119.19(12), C49–C48–C53 115.76(14), C53–C48–Mg1 124.95. Distances and angles for 6/6′ and 7/7′ are not reported here due to disorder.

Fig. 7

Open in a new tab

For meta-xylene, the benzylic C(sp3)–H bonds are less hindered, and formation of the meta-methyl benzyl derivative [(AdBDI)Mg(CH2-C6H4-3-CH3)] (7′) is relatively favoured compared to the meta-magnesiated xylyl product [(AdBDI)Mg(C6H3-3,5-Me2)] (7). After 10 days at 120 °C and removal of volatiles under vacuum, analysis of the crude reaction mixture by 1H NMR spectroscopy indicated >95% conversion of 1Ad to 7 and 7′ in respective spectroscopic yields of 20% and 80% (Scheme 6). Pale-yellow block-shaped single crystals were obtained after recrystallisation from n-hexane and isolated in 66% yield. The results of a subsequent X-ray diffraction experiment provided a structure solution that was refined as a disordered co-crystal of 7′ and 7 in two components of relative occupancy 0.9 : 0.1 (Fig. 7b), whilst a few colourless plate-shaped single crystals of pure 7 suitable for X-ray diffraction analysis were obtained from the walls of the J. Young's NMR tube above the meniscus of the reaction mixture at 100 °C (See SI, Fig. S277).

The C(sp2)–H bonds of para-xylene are kinetically inaccessible under similar conditions, and heating 1Ad at 120 °C in this solvent for 10 days exclusively provided the para-methyl benzyl derivative [(AdBDI)Mg(CH2-C6H4-4-CH3)] (8) in >95% spectroscopic conversion according to 1H NMR spectroscopy (Scheme 6). Pale-yellow rod-shaped single crystals of 8 suitable for X-ray diffraction analysis were obtained in 66% yield after recrystallisation from n-hexane/toluene (Fig. 7c).

Finally, 1,3-di-tertbutylbenzene – which lacks acidic C(sp3)–H bonds – yields the 3,5-di-tert-butyl phenyl complex [(AdBDI)Mg(C6H3-3,5-tBu2)] (9) as the sole product after 8 days at 120 °C (Scheme 6). Compound 9 was isolated in 51% yield as colourless block-shaped crystals suitable for X-ray diffraction analysis (Fig. 7d), which deposited directly from the reaction mixture.

The X-ray crystal structures of compounds 6–9 are unremarkable besides providing structural verification of the reaction products (Fig. 7). Similarly to compounds 3R and 5-m/5-p, however, in each of the C(sp2)–H activated products, the benzene ring is approximately co-planar with the BDI ligand backbone. Although detailed discussion of distances and angles are limited by disorder in most cases, the Mg–C distances of the aryl derivatives (6 = 2.049(4) Å; 7 = 2.1016(18) Å; 9 = 2.1027(16) Å) fall within a relatively narrow range and are similar to previously reported examples of three-coordinate BDI-magnesium aryl derivatives,18,41 albeit significantly shorter than 5-m/5-p (2.130(3), 2.1353(19) Å). The Mg–C distances in the benzyl derivatives are also mutually similar (5′ = 2.1160(19); 7′ = 2.1223(16); 8 = 2.1174(15) Å).

Deprotonation and magnesiation of anisole

In a further utilisation of the magnesium n-butyl moiety as a strong base, anisole solutions of compounds 1Ad and [1Me]2 were heated at 100 °C. The methoxy group exerts a directing and kinetically activating effect, such that the ortho-magnesiated anisyl derivatives 10Ad and 10Me were selectively obtained after 7 days or 16 hours, respectively (Scheme 7). Recrystallisation from n-hexane/benzene or n-hexane/THF provided the respective anisole- and THF-adducted derivatives [(AdBDI)Mg(κ1C-C6H4-2-OCH3)(κ1O-O(CH3)C6H5)] (10Ad·anisole) and [(MeBDI)Mg(κ2C,O-C6H4-2-OMe)(thf)] (10Me·thf). This mode of reactivity contrasts with that of hydridocalcium derivative [2Ca]2, which reacts with anisole via nucleophilic substitution of the CH3 group, eliminating methane and forming a calcium phenoxide complex.48 The related cationic aza-macrocycle-supported calcium hydride [(Me4TACD)Ca(µ-H)2Ca(thf)(Me4TACD)]2+ (Me4TACD = N,N′,N″,N‴-1,4,7,10-tetraazacyclododecane) does, however, mediate the directed ortho-metallation of anisole and 1,3-di(methoxy)benzene.49 The X-ray crystal structure of 10Ad·anisole (Fig. 8b) contains a distorted tetrahedral four-coordinate (τ4 = 0.826)50 magnesium centre coordinated by the chelating BDI, κ1-C-bound anisyl, and neutral O-bound anisole ligands. By contrast, the anisyl ligand in the less bulky derivative 10Me·thf chelates magnesium in a κ2-C,O fashion (Fig. 8a), such that the magnesium centre is five-coordinate, adopting a square-pyramidal geometry (τ5 = 0.156; value calculated for the major component of the disordered anisyl moiety)51 with THF bound in the apical site and a shorter Mg–C distance (2.137(3) Å for the major component of 10Me·thf; 2.178(5) Å for 10Ad·anisole).

Scheme 7. Reaction of compounds 1Ad and [1Me]2 with anisole. Isolated yields are shown with 1H NMR spectroscopic yields in parentheses.

Scheme 7

Open in a new tab

Fig. 8. Displacement ellipsoid plots (30% level) of the crystal structures of 10Me·thf (left) and 10Ad·anisole (right). Only the major component of the disordered anisyl moiety in 10Me·thf is shown, dipp groups are shown as wireframes, and hydrogen atoms are omitted for clarity. Distances (Å) and angles (°) for 10Me·thf: Mg1–O2 2.0542(14), Mg1–N1 2.0740(15), Mg1–N2 2.0740(14), Mg1–O1A 2.729, Mg1–C30A 2.137(3), Mg1–O1B 2.461(10), Mg1–C30B 2.179(13), O2–Mg1–N1 104.03(6), O2–Mg1–N2 102.66(6), O2–Mg1–O1A 94.34(7), O2–Mg1–C30A 101.28(8), O2–Mg1–O1B 94.9(2), O2–Mg1–C30B 98.7(3), N1–Mg1–O1A 160.38(7), N1–Mg1–C30A 113.36(12), N1–Mg1–O1B 95.1(2), N1–Mg1–C30B 146.3(4), N2–Mg1–N1 92.48(6), N2–Mg1–O1A 90.13(7), N2–Mg1–C30A 138.89(10), N2–Mg1–O1B 158.5(2), N2–Mg1–C30B 106.5(4), C30A–Mg1–O1A 55.07(12), C30B–Mg1–O1B 58.1(5), C31A–O1A–Mg1 83.10(19), C31A–O1A–C36A 117.5(3), C36A–O1A–Mg1 158.16(19), C31A–C30A–Mg1 109.9(2), C31B–C30B–Mg1 101.5(10), C35B–C30B–Mg1 142.3(13). Distances (Å) and angles (°) for 10Ad·anisole: Mg1–O2 2.208(3), Mg1–N1 2.070(3), Mg1–N2 2.106(3), Mg1–C48 2.178(5), N1–Mg1–O2 107.85(13), N1–Mg1–N2 95.58(13), N1–Mg1–C48 115.89(16), N2–Mg1–O2 108.37(13), N2–Mg1–C48 127.29(15), C48–Mg1–O2 100.99(14).

Fig. 8

Open in a new tab

Magnesium-mediated dehydroborylation of benzene

Compound 3Ad was utilised as a nucleophilic phenyl-transfer reagent in a one-pot procedure, starting from 1Ad and benzene (Scheme 8). After heating at 100 °C for 1 week, removal of volatiles under vacuum, and redissolution in benzene-d6, the as-obtained solution of 3Ad was added to MeOBpin and a bright yellow solution was immediately obtained. Analysis by 1H and 11B NMR spectroscopies confirmed successful phenyl transfer to generate the commonly employed Suzuki cross-coupling precursor, PhBpin, by comparison to the NMR data of a commercial sample. The fate of the magnesium depends on the number of equivalents of MeOBpin employed. A strict 1 : 1 molar ratio of 3Ad and MeOBpin generates the dimeric methoxide [(κ2-AdBDI)Mg(µ-OMe)2Mg(κ1-AdBDI)] (11), which was isolated as colourless plate-shaped crystals after recrystallisation from n-hexane/benzene and separately synthesised and characterised after a similar reaction of 1Ad with MeOBpin (Fig. 9a). Under conditions of excess MeOBpin, the bright yellow boronate derivative [(AdBDI)Mg(µ-OMe)2(Bpin)] (12) was preferentially formed. Compound 12 was also separately synthesised from 1Ad and MeOBpin and crystallographically characterised after recrystallisation from n-hexane/toluene (Fig. 9b).

Scheme 8. One-pot, two-step dehydroborylation of benzene to yield PhBpin (quantitative spectroscopic yield by 1H NMR) and 11 and/or 12. Compounds 11 and 12 can be isolated in respective 55% and 54% crystalline yield from the direct reaction of 1 with 1 or >2 equiv. MeOBpin.

Scheme 8

Open in a new tab

Fig. 9. Displacement ellipsoid plots (30%) of the crystal structures of (a) 11 and (b) 12. Hydrogen atoms are omitted for clarity, except for those bound to C35 and C50 in compound 11. Solvent molecules present in the crystal structure of 11 are also omitted for clarity. Distances (Å) and angles (°) for 11: Mg1–O1 1.9571(14), Mg1–O2 1.9762(15), Mg1–N1 2.0253(16), Mg1–N2 2.0602(15), Mg2–O1 1.9430(15), Mg2–O2 1.9243(14), Mg2–N3 2.0069(16), Mg2–H50B 2.2195(8), Mg2–H50A 2.0937(8), Mg2–H35C 2.4624(8), N1–C1 1.343(2), N2–C13 1.330(2), N3–C48 1.380(2), N4–C60 1.293(2), C1–C12 1.405(2), C12–C13 1.427(2), C59–C60 1.451(3), C59–C48 1.386(2), O1–Mg1–O2 83.03(6), O1–Mg1–N1 113.11(6), O1–Mg1–N2 131.94(6), O2–Mg1–N1 121.05(6) O2–Mg1–N2 112.88(6), N1–Mg1–N2 96.36(6), O1–Mg2–N3 132.07(6), O2–Mg2–O1 84.77(6), O2–Mg2–N3 142.95(7), Mg2–O1–Mg1 95.96(6), Mg2–O2–Mg1 945.94(6). Distances (Å) and angles (°) for 12: Mg1–O1 1.9723(9), Mg1–O2 1.9590(9), Mg1–N1 2.0100(10), Mg1–N2 2.0680(10), O1–B1 1.523(16), O2–B1 1.5318(16), O3–B1 1.4304(17), O4–B1 1.4220(17), N1–C1 1.3720(15), N2–C13 1.3060(15), C1–C12 1.3856(16), C12–C13 1.4576(15), O1–Mg1–N1 119.16(4), O1–Mg1–N2 120.53(4), O2–Mg1–O1 69.05(4), O2–Mg1–N1 137.50(4), O2–Mg1–N2 115.06(4), N1–Mg1–N2 96.61(40), O1–B1–O2 94.10(9), O3–B1–O1 110.71(11), O3–B1–O2 115.28(11), O4–B1–O1 116.68(11), O4–B1–O2 112.07(11), O4–B1–O3 107.80(10).

Fig. 9

Open in a new tab

The crystal structures of compounds 11 and 12 merit brief discussion, as they illustrate the steric limits of the extremely bulky AdBDI ligand (Fig. 9). In each case, the strongly oxophilic magnesium centres bind to two µ-OMe groups. In compound 11, this results in a dimeric structure which is unable to accommodate two chelating AdBDI ligands. This causes one AdBDIMg chelate to ring-open, such that Mg2 is three-coordinate trigonal planar (Σangles(Mg) = 359.79°) and bound via N3 (Mg2–N3 = 2.0069(16) Å) to an AdBDI unit as an open-chain enamido-imine with localised C59–C48 (1.386(2) Å) and C60–N4 (1.293(2) Å) double bonds (cf. C59–C60, C48–N3 single bonds = 1.541(3), 1.380(2) Å), whilst Mg1 is four-coordinate and chelated by a typical κ2-β-diketiminate with delocalised π bonding (N1–C1, N2–C13, C1–C12, C12–C13 = 1.343(2), 1.330(2), 1.405(2), 1.427(2) Å). The coordination sphere of the planar three-coordinate Mg2 centre appears to be further augmented by close contacts with the C50-bound hydrogen atoms of the nearby adamantyl group (Mg2–H50B, 2.2195(8); Mg2–H50A, 2.0937(8) Å), and the C35-bound hydrogen atoms belonging to an isopropyl group of the Mg1-bound AdBDI ligand (Mg2–H35C = 2.4624(8) Å). This structure is retained in benzene-d6 solution, such that the backbone methine resonances of the two dissimilar AdBDI ligands are observed in the 1H NMR spectrum as singlets of equal (1H) relative intensity at δ 5.79 (κ2-AdBDI), 4.41 (κ1-AdBDI) ppm. Further, the κ1-bound AdBDI ligand displays two inequivalent dipp and adamantyl environments in the 1H and 13C NMR spectra, whereas individual chemically equivalent dipp and adamantyl environments are observed for the κ2-bound AdBDI ligand. A similar closed-open combination of two BDI ligands has been observed for the magnesium(i) dimer supported by an extremely bulky tBuBDIdipep ligand.37

Although the AdBDI ligand of compound 12 appears to be conventionally coordinated to the magnesium centre in a κ2-fashion, closer inspection reveals a significant twisting of the normally planar BDI ligand backbone, as shown by the N1–C1–C13–N2 torsion angle of 25.68(9)°. Further, the N2–C13 (1.3060(15) Å) and C1–C12 (1.3856(16) Å) distances are significantly shorter than the N1–C1 (1.3720(15) Å) and C12–C13 distances 1.4576(15) Å), indicating localised N Created by potrace 1.16, written by Peter Selinger 2001-2019 C and C Created by potrace 1.16, written by Peter Selinger 2001-2019 C double bonds and a β-enamidoimine form rather than a delocalised β-diketiminate (Fig. 9b). Additionally, the magnesium centre is significantly projected above the mean N1–C12–N2 plane by 1.1785(11) Å to accommodate the tetra-alkoxyboronate moiety. Hence, the extremely congested coordination environment enforces a loss of π-delocalisation, and a twisted ligand conformation. This situation can be viewed as an intermediate en route towards the chain-opened form observed in 11.

Utilisation of magnesium phenyl and anisyl complexes in palladium-catalysed Kumada coupling

We have previously reported that the dimeric calcium phenyl complex [(MeBDI)Ca(µ,η1:η1-Ph)]2 reacts directly with bromobenzene to provide biphenyl and [((MeBDI)Ca)2(µ,η1:η6-Ph)(µ-Br)] (Scheme 2c).32 By contrast, 1Ad does not react directly with bromobenzene, and application of elevated temperature solely results in phenyl-exchange with the benzene-d6 solvent. We35 and Crimmin and co-workers52 have previously demonstrated that a telescoped palladium-catalysed aryl–aryl coupling may be achieved via BDI–zinc aryl complexes. Compound 3Ad was, thus, similarly generated in situ by reaction of 1Ad with benzene, followed by removal of volatiles, redissolution in benzene-d6 and addition of bromobenzene (1 equiv.) and catalytic [Pd(PCy3)2] (4–5 mol%). After 3 days at 60 °C, 93% spectroscopic yield of biphenyl was confirmed by in situ1H NMR spectroscopy and the product additionally identified by GC-MS (Scheme 9). The magnesium bromide by-product, [(AdBDI)MgBr] (13), crystallised directly from the reaction mixture and was fully characterised (see SI). Although similar monomeric 3-coordinate iodide species have been reported,37,42,53 compound 13 represents the first example of a monomeric three-coordinate BDI magnesium bromide complex. The one-pot telescoped protocol was extended towards the cross-coupling of 3Ad with 2-ethyl bromobenzene, as well as 4-methoxy bromobenzene, which respectively provided the cross-coupled biphenyl derivatives, 2-ethyl-1-biphenyl and 4-methoxy-1-biphenyl, in quantitative 1H NMR spectroscopic yields after 2 days at 60 °C (Scheme 9). Further, 3Ad-d5 was employed to generate the deuterated biphenyl derivate 4-MeO-C6H4-1-C6D5 in quantitative yield via a telescoped one-pot procedure starting from 1Ad and benzene-d6. Reacting [1Me]2 with anisole followed by removal of volatiles and palladium-catalysed cross-coupling of the as-formed 10Me with bromobenzene yielded 2-methoxy-1-biphenyl in quantitative 1H NMR spectroscopic yield after 24 hours at 60 °C, with elimination of the known dimeric magnesium bromide complex [(MeBDI)Mg(µ-Br)]2 (Scheme 9).41 Whilst palladium-catalysed Kumda coupling is well established using Grignard reagents, without needing a ‘specialised’ ligand such as AdBDI, the ability to access magnesium aryl reagents via direct C–H activation may enable the development of new synthetic approaches.

Scheme 9. One-pot, two-step palladium-catalysed Kumada cross-coupling of benzene with aryl bromides via3Ad (top) to give 13 and biphenyl products (middle); one-pot, two-step palladium-catalysed Kumada cross-coupling of anisole with bromobenzene via10Me to give 2-methoxybiphenyl and [(MeBDI)Mg(µ-Br)]2 (bottom).*Three-days reaction time for the reaction with bromobenzene. Yields given were determined by integration of the in situ1H NMR spectrum against hexamethylbenzene as an internal standard.

Scheme 9

Open in a new tab

Conclusions

Although thermodynamically consistent with the pKas of the relevant C–H bonds, the C(sp2) and C(sp3) deprotonation of unactivated arenes by alkyl magnesium reagents is kinetically challenging. The magnesium n-butyl complex 1Ad with the bulky AdBDI ligand acts as a robust and sterically hindered base, enabling the deprotonation of benzene, toluene, xylenes, and 1,3-di-tertbutylbenzene under forcing conditions (100–120 °C) without competing side reactions such as ligand degradation or β-hydride elimination. The reaction of dimeric n-butyl complex [1Me]2 – which is supported by the less sterically demanding MeBDI ligand – with benzene is not only much slower than that of 1Ad but also suffers from β-hydride elimination of 1-butene as a side reaction. Although the reason for the greater kinetic facility of 1Ad for arene deprotonation compared to [1Me]2 remains to be resolved, its resistance to competing β-hydride elimination may be attributed to lower stability of the hypothetical µ-hydride dimer, [(AdBDI)Mg(µ-H)]2 ([2Mg,Ad]2). Whilst the steric demand of the AdBDI ligand enables selective deprotonation of benzene in the presence of the substantially more acidic, yet hindered, 1,2,3,4-tetramethylcyclopentadiene, and leaves ortho-C(sp2)–H bonds in the studied alkyl benzenes intact, little kinetic discrimination is observed between the various sterically accessible C(sp2)–H and C(sp3)–H bonds of toluene, ortho- and meta-xylene. By contrast, both 1Ad and [1Me]2 readily mediate the directed ortho-magnesiation of anisole. The anisyl and phenyl complexes were harnessed in a telescoped “one-pot” Kumada biaryl cross-coupling procedure, utilising the reaction solvent as the aryl source.

Compared to the facile alkylation of benzene by dimeric β-diketiminato calcium-, ytterbium-, and strontium n-alkyl complexes, the C–H deprotonation chemistry of 1Ad is striking. These contrasting observations are, however, consistent with lower nucleophilicity of the Mg–C bond compared to the heavier congeners. Being smaller, harder (less polarisable), and less electropositive, magnesium is additionally less able to interact strongly with the π-electron density of aromatic rings. Such interactions are a regular feature of heavier alkaline earth organometallics and have been suggested to be crucial in the nucleophilic alkylation chemistry of these elements. The ability to directly derive aryl magnesium reagents from the parent hydrocarbon in cross-coupling chemistry potentially opens new possibilities in C–H functionalisation protocols, and future efforts to develop soluble and kinetically discriminating alkaline-earth reagents of exceptional basicity may be fruitful in this regard.

Author contributions

L. J. M.: conceptualisation, investigation, formal analysis, methodology, data curation, visualisation, writing – original draft, writing – review and editing. A. M.: investigation, formal analysis, methodology, data curation, visualisation, writing – review and editing. A. S. S. W.: conceptualisation, investigation. M. S. H.: funding acquisition, project administration, resources, writing – review and editing. C. L. M.: funding acquisition, project administration, resources, writing – review and editing.

Conflicts of interest

There are no conflicts to declare.

Supplementary Material

SC-OLF-D6SC04717C-s001
SC-OLF-D6SC04717C-s002

Acknowledgments

The authors acknowledge the Leverhulme Trust (RPG-2023–104, “Alkaline Element Z-type Ligands for Tunable Base Metal Catalysis”), EPSRC (EP/X01181X/1, “Molecular s-block Assemblies for Redox-active Bond Activation and Catalysis: Repurposing the s-block as 3d-elements”), the University of Bath's Research Computing Group (https://doi.org/10.15125/b6cd-s854) and the Anatra HTC for their support in this work. The authors also wish to thank Dr M. Mahon for useful crystallographic discussions and Dr K. Proctor (University of Bath) for assistance with GC-MS analysis.

Data availability

CCDC 2557915–2557930 contain the supplementary crystallographic data for this paper.54a–p

Other data supporting this manuscript can be found in the supplementary information (SI). Supplementary information: experimental details, NMR spectra, chromatograms, mass spectra, details of steric mapping analysis, crystallographic details, and computational details, and Cartesian coordinates of the DFT optimised structures. See DOI: https://doi.org/10.1039/d6sc04717c.

References

  1. Eaton P. E. Lee C. H. Xiong Y. J. Am. Chem. Soc. 1989;111:8016–8018. doi: 10.1021/ja00202a054. [DOI] [Google Scholar]
  2. Baillie S. E. Blair V. L. Bradley T. D. Clegg W. Cowan J. Harrington R. W. Hernán-Gómez A. Kennedy A. R. Livingstone Z. Hevia E. Chem. Sci. 2013;4:1895–1905. doi: 10.1039/C3SC22326D. [DOI] [Google Scholar]
  3. Okokhere-Edeghoghon B. Hill M. S. Mahon M. F. Polyhedron. 2024;264:117255. doi: 10.1016/j.poly.2024.117255. [DOI] [Google Scholar]
  4. Davin L. McLellan R. Hernán-Gómez A. Clegg B. Kennedy A. R. Stepek I. Mertens M. Hevia E. Chem. Commun. 2017;53:3653–3656. doi: 10.1039/C6CC09675A. [DOI] [PubMed] [Google Scholar]
  5. Mulvey R. E. Mongin F. Uchiyama M. Kondo Y. Angew. Chem., Int. Ed. 2007;46:3802–3824. doi: 10.1002/anie.200604369. [DOI] [PubMed] [Google Scholar]
  6. Robertson S. D. Uzelac M. Mulvey R. E. Chem. Rev. 2019;119:8332–8405. doi: 10.1021/acs.chemrev.9b00047. [DOI] [PubMed] [Google Scholar]
  7. Rohbogner C. J. Clososki G. C. Knochel P. Angew. Chem., Int. Ed. 2008;47:1503–1507. doi: 10.1002/anie.200703382. [DOI] [PubMed] [Google Scholar]
  8. Clososki G. C. Rohbogner C. J. Knochel P. Angew. Chem., Int. Ed. 2007;46:7681–7684. doi: 10.1002/anie.200701487. [DOI] [PubMed] [Google Scholar]
  9. Haag B. Mosrin M. Ila H. Malakhov V. Knochel P. Angew. Chem., Int. Ed. 2011;50:9794–9824. doi: 10.1002/anie.201101960. [DOI] [PubMed] [Google Scholar]
  10. Lin W. Baron O. Knochel P. Org. Lett. 2006;8:5673–5676. doi: 10.1021/ol0625536. [DOI] [PubMed] [Google Scholar]
  11. Martínez-Martínez A. J. Armstrong D. R. Conway B. Fleming B. J. Klett J. Kennedy A. R. Mulvey R. E. Robertson S. D. O'Hara C. T. Chem. Sci. 2014;5:771–781. doi: 10.1039/C3SC52816B. [DOI] [Google Scholar]
  12. Hevia E. Gallagher D. J. Kennedy A. R. Mulvey R. E. O'Hara C. T. Talmard C. Chem. Commun. 2004:2422–2423. doi: 10.1039/B410293B. [DOI] [PubMed] [Google Scholar]
  13. Andrews P. C. Kennedy A. R. Mulvey R. E. Raston C. L. Roberts B. A. Rowlings R. B. Angew. Chem., Int. Ed. 2000;39:1960–1962. doi: 10.1002/1521-3773(20000602)39:11<1960::AID-ANIE1960>3.0.CO;2-D. [DOI] [PubMed] [Google Scholar]
  14. Armstrong D. R. Kennedy A. R. Mulvey R. E. Rowlings R. B. Angew. Chem., Int. Ed. 1999;38:131–133. doi: 10.1002/(SICI)1521-3773(19990115)38:1/2<131::AID-ANIE131>3.0.CO;2-9. [DOI] [Google Scholar]
  15. Baillie S. E. Bluemke T. D. Clegg W. Kennedy A. R. Klett J. Russo L. de Tullio M. Hevia E. Chem. Commun. 2014;50:12859–12862. doi: 10.1039/C4CC05305B. [DOI] [PubMed] [Google Scholar]
  16. Martínez-Martínez A. J. Kennedy A. R. Mulvey R. E. O'Hara C. T. Science. 2014;346:834–837. doi: 10.1126/science.1259662. [DOI] [PubMed] [Google Scholar]
  17. Martínez-Martínez A. J. Justice S. Fleming B. J. Kennedy A. R. Oswald I. D. H. O'Hara C. T. Sci. Adv. 2017;3:e1700832. doi: 10.1126/sciadv.1700832. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Jones D. D. L. Douair I. Maron L. Jones C. Angew. Chem., Int. Ed. 2021;60:7087–7092. doi: 10.1002/anie.202017126. [DOI] [PubMed] [Google Scholar]
  19. Mullins J. C. Yuvaraj K. Jiang Y. Van Trieste III G. P. Maity A. Powers D. C. Jones C. Chem. – Eur. J. 2022;28:e202202103. doi: 10.1002/chem.202202103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Garçon M. White A. J. P. Crimmin M. R. Chem. Commun. 2018;54:12326–12328. doi: 10.1039/C8CC06392C. [DOI] [PubMed] [Google Scholar]
  21. Hooper T. N. Garçon M. White A. J. P. Crimmin M. R. Chem. Sci. 2018;9:5435–5440. doi: 10.1039/C8SC02072H. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Brand S. Elsen H. Langer J. Grams S. Harder S. Angew. Chem., Int. Ed. 2019;58:15496–15503. doi: 10.1002/anie.201908978. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Gentner T. X. Rosch B. Ballmann G. Langer J. Elsen H. Harder S. Angew. Chem., Int. Ed. 2019;58:607–611. doi: 10.1002/anie.201812051. [DOI] [PubMed] [Google Scholar]
  24. Rösch B. Gentner T. X. Eyselein J. Langer J. Elsen H. Harder S. Nature. 2021;592:717–721. doi: 10.1038/s41586-021-03401-w. [DOI] [PubMed] [Google Scholar]
  25. Stecher H. Thum S. Maurer J. Mai J. Schmidt M. A. Langer J. Harder S. Angew. Chem., Int. Ed. 2025;64:e202518538. doi: 10.1002/anie.202518538. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Thakur S. K. Langer J. Roig N. Patro A. G. Schmidt M. A. Alonso M. Harder S. Angew. Chem., Int. Ed. 2026;65:e25777. doi: 10.1002/anie.202525777. [DOI] [PubMed] [Google Scholar]
  27. Mai J. Rösch B. Patel N. Langer J. Harder S. Chem. Sci. 2023;14:4724–4734. doi: 10.1039/D3SC00909B. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Wilson A. S. S. Hill M. S. Mahon M. F. Dinoi C. Maron L. Science. 2017;358:1168–1171. doi: 10.1126/science.aao5923. [DOI] [PubMed] [Google Scholar]
  29. Rösch B. Gentner T. X. Elsen H. Fischer C. A. Langer J. Wiesinger M. Harder S. Angew. Chem., Int. Ed. 2019;58:5396–5401. doi: 10.1002/anie.201901548. [DOI] [PubMed] [Google Scholar]
  30. Garcia L. Dinoi C. Mahon M. F. Maron L. Hill M. Chem. Sci. 2019;10:8108–8118. doi: 10.1039/C9SC02056J. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Pearce K. G. Dinoi C. Schwamm R. J. Maron L. Mahon M. F. Hill M. S. Adv. Sci. 2023;10:2304765. doi: 10.1002/advs.202304765. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Pearce K. G. Dinoi C. Hill M. S. Mahon M. F. Maron L. Schwamm R. S. Wilson A. S. S. Angew. Chem., Int. Ed. 2022;61:e202200305. doi: 10.1002/anie.202200305. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Pearce K. G. Mahon M. F. Hill M. S. Dalton Trans. 2025;54:8903–8907. doi: 10.1039/D5DT00958H. [DOI] [PubMed] [Google Scholar]
  34. Pearce K. G. Neale S. E. McMullin C. L. Mahon M. F. Hill M. S. Chem. Commun. 2024;60:7882–7885. doi: 10.1039/D4CC02930E. [DOI] [PubMed] [Google Scholar]
  35. Pearce K. G. Morales A. Hill M. S. Mahon M. F. McMullin C. L. J. Am. Chem. Soc. 2025;147:29554–29567. doi: 10.1021/jacs.5c10735. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Balasanthiran V. Chisholm M. H. Choojun K. Durr C. B. Wambua P. M. Polyhedron. 2016;103:235–240. doi: 10.1016/j.poly.2015.02.024. [DOI] [Google Scholar]
  37. Rösch B. Gentner T. X. Eyselein J. Friedrich A. Langer J. Harder S. Chem. Commun. 2020;56:11402–11405. doi: 10.1039/D0CC05200K. [DOI] [PubMed] [Google Scholar]
  38. Burnett S. Bourne C. Slawin A. M. Z. van Mourik T. Stasch A. Angew. Chem., Int. Ed. 2022;61:e202204472. doi: 10.1002/anie.202204472. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Thum K. Pahl J. Eyselein J. Elsen H. Langer J. Harder S. Chem. Commun. 2021;57:5278–5281. doi: 10.1039/D1CC01691A. [DOI] [PubMed] [Google Scholar]
  40. Li B. Huse K. Wölper C. Schulz S. Chem. Commun. 2021;57:13692–13695. doi: 10.1039/D1CC05617D. [DOI] [PubMed] [Google Scholar]
  41. Dove A. P. Gibson V. C. Hormnirun P. Marshall E. L. Segal J. A. White A. J. P. Williams D. J. Dalton Trans. 2003:3088–3097. doi: 10.1039/B303550F. [DOI] [Google Scholar]
  42. Bonyhady S. J. Jones C. Nembenna S. Stasch A. Edwards A. J. McIntyre G. J. Chem. – Eur. J. 2010;16:938–955. doi: 10.1002/chem.200902425. [DOI] [PubMed] [Google Scholar]
  43. Hill M. S. Mahon M. F. Wilson A. S. S. Dinoi C. Maron L. Richards E. Chem. Commun. 2019;55:5732–5735. doi: 10.1039/C9CC02418B. [DOI] [PubMed] [Google Scholar]
  44. Falivene L. Cao Z. Petta A. Serra L. Poater A. Oliva R. Scarano V. Cavallo L. Nat. Chem. 2019;11:872–879. doi: 10.1038/s41557-019-0319-5. [DOI] [PubMed] [Google Scholar]
  45. Richardson G. M. Douair I. Cameron S. A. Bracegirdle J. Keyzers R. A. Hill M. S. Maron L. Anker M. D. Nat. Commun. 2021;12:3147. doi: 10.1038/s41467-021-23444-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Zhang X.-Y. Zhai D.-D. Liu Y.-F. Guan B.-T. J. Organomet. Chem. 2022;961:122254. doi: 10.1016/j.jorganchem.2022.122254. [DOI] [Google Scholar]
  47. Bailey P. J. Coxall R. A. Dick C. M. Fabre S. Henderson L. C. Herber C. Liddle S. T. Loroño-González D. Parkin A. Parsons S. Chem. – Eur. J. 2003;9:4820–4828. doi: 10.1002/chem.200305053. [DOI] [PubMed] [Google Scholar]
  48. Wilson A. S. S. Hill M. S. Mahon M. F. Dinoi C. Maron L. Tetrahedron. 2021;82:131931. doi: 10.1016/j.tet.2021.131931. [DOI] [Google Scholar]
  49. Schuhknecht D., Doctoral thesis, RWTH Aachen University, 2020 [Google Scholar]
  50. Yang L. Powell D. R. Houser R. P. Dalton Trans. 2007:955–964. doi: 10.1039/B617136B. [DOI] [PubMed] [Google Scholar]
  51. Addison A. W. Rao T. N. Reedijk J. van Rijn J. Verschoor G. C. J. Chem. Soc., Dalton Trans. 1984:1349–1356. doi: 10.1039/DT9840001349. [DOI] [Google Scholar]
  52. Garçon M. Mun N. W. White A. J. P. Crimmin M. R. Angew. Chem., Int. Ed. 2021;60:6145–6153. doi: 10.1002/anie.202014960. [DOI] [PubMed] [Google Scholar]
  53. Martin J. Langer J. Wiesinger M. Elsen H. Harder S. Eur. J. Inorg. Chem. 2020:2582–2595. doi: 10.1002/ejic.202000524. [DOI] [Google Scholar]
  54. (a) CCDC 2557915: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqdr [DOI]; (b) CCDC 2557916: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqfs [DOI]; (c) CCDC 2557917: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqgt [DOI]; (d) CCDC 2557918: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqhv [DOI]; (e) CCDC 2557919: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqjw [DOI]; (f) CCDC 2557920: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqkx [DOI]; (g) CCDC 2557921: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqly [DOI]; (h) CCDC 2557922: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqmz [DOI]; (i) CCDC 2557923: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqn0 [DOI]; (j) CCDC 2557924: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqp1 [DOI]; (k) CCDC 2557925: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqq2 [DOI]; (l) CCDC 2557926: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqr3 [DOI]; (m) CCDC 2557927: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqs4 [DOI]; (n) CCDC 2557928: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqt5 [DOI]; (o) CCDC 2557929: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqv6 [DOI]; (p) CCDC 2557930: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqw7 [DOI]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Citations

  1. (a) CCDC 2557915: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqdr [DOI]
  2. (b) CCDC 2557916: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqfs [DOI]
  3. (c) CCDC 2557917: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqgt [DOI]
  4. (d) CCDC 2557918: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqhv [DOI]
  5. (e) CCDC 2557919: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqjw [DOI]
  6. (f) CCDC 2557920: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqkx [DOI]
  7. (g) CCDC 2557921: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqly [DOI]
  8. (h) CCDC 2557922: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqmz [DOI]
  9. (i) CCDC 2557923: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqn0 [DOI]
  10. (j) CCDC 2557924: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqp1 [DOI]
  11. (k) CCDC 2557925: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqq2 [DOI]
  12. (l) CCDC 2557926: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqr3 [DOI]
  13. (m) CCDC 2557927: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqs4 [DOI]
  14. (n) CCDC 2557928: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqt5 [DOI]
  15. (o) CCDC 2557929: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqv6 [DOI]
  16. (p) CCDC 2557930: Experimental Crystal Structure Determination, 2026, 10.5517/ccdc.csd.cc2rvqw7 [DOI]

Supplementary Materials

SC-OLF-D6SC04717C-s001
SC-OLF-D6SC04717C-s001.pdf (32.2MB, pdf)
SC-OLF-D6SC04717C-s002
SC-OLF-D6SC04717C-s002.cif (26MB, cif)

Data Availability Statement

CCDC 2557915–2557930 contain the supplementary crystallographic data for this paper.54a–p

Other data supporting this manuscript can be found in the supplementary information (SI). Supplementary information: experimental details, NMR spectra, chromatograms, mass spectra, details of steric mapping analysis, crystallographic details, and computational details, and Cartesian coordinates of the DFT optimised structures. See DOI: https://doi.org/10.1039/d6sc04717c.

Articles from Chemical Science are provided here courtesy of Royal Society of Chemistry

RESOURCES