Molecular early main group metal hydrides: synthetic challenge, structures and applications

Abstract

Within the general area of early main group metal chemistry, the controlled synthesis of well-defined metal hydride complexes is a rapidly developing research field. As group 1 and 2 metal complexes are generally highly dynamic and lattice energies for their [MH]_∞ and [MH₂]_∞ salts are high, the synthesis of well-defined soluble hydride complexes is an obvious challenge. Access to molecular early main group metal hydrides, however, is rewarding: these hydrocarbon-soluble metal hydrides are highly reactive, have found use in early main group metal catalysis and are potentially also valuable molecular model systems for polar metal hydrides as a hydrogen storage material. The article focusses specifically on alkali and alkaline-earth metal hydride complexes and discusses the synthetic challenge, molecular structures, reactivity and applications.

---

Sjoerd Harder

Sjoerd Harder obtained his BSc, MSc and PhD degrees at the University of Utrecht in the Netherlands. After several post-doc stays at the University of Erlangen-Nürnberg (P. v. R. Schleyer), University of California at Berkeley (A. Streitwieser), University of Konstanz (H.-H. Brintzinger) he finalized his habilitation in 1998 at the University of Konstanz. He became an associate professor at the University of Duisburg-Essen (2004) and Chair in Inorganic Chemistry at the University of Groningen in his home country The Netherlands (2010). Within two years he accepted an offer at the University of Erlangen-Nürnberg and is currently moving back to Germany.

---

1 Introduction

The element hydrogen holds a number of records: it is the first and lightest element in the periodic table, having the most simple atomic structure. Although it is hardly occurring in its elemental form on earth, it represents the most abundant element in the universe's chemical mass. Due to its high energy content it is currently in the picture as the energy carrier of the future.¹

Among its unique and peculiar properties is also the fact that it has an unusually high electronegativity for its position at the top of the group 1 elements (its Allred–Rochow's electronegativity of 2.2 is between those of B and C). As the 1s orbital has the correct energy for overlapping with orbitals of most other elements, hydrogen has the ability to combine with a plethora of chemical elements. As such, hydrogen acts like a chameleon and can have physical properties and reactivities varying from H⁺ and H-radical to H⁻. These various roles of the hydrogen ligand are especially underscored in its transition-metal chemistry which is exploited in a rich application in catalysis.²

The current feature article discusses hydrogen as a ligand in early main group metal complexes. Polar hydride compounds have found a rapidly increasing interest as storage materials for molecular hydrogen.³ Although previous reviews on early main group metal hydrides mainly deal with inorganic salt-like ion-pair compounds,⁴ this article focusses on molecular early main group metal hydride compounds and exclusively treats the group 1 and 2 metals. From a synthetic point of view, access to highly polar hydrides that are soluble in common organic solvents is challenging but also rewarding: this class of compounds fulfills an important role as intermediates in the rapidly developing field of early main group metal catalysis.⁵ This article will describe the historical development of this field, group 1 and group 2 separately treated, in a chronological order. Emphasis will be on complexes that have been unequivocally characterized by single crystal X-ray diffraction.

2 Saline early main group metal hydrides

Although these purely inorganic compounds are not part of the focus of this review, a short introduction on this class of salt-like compounds will help in putting the molecular main group hydrides in a better perspective. Group 1 and 2 metal hydrides form simple salt-like structures consisting of metal cations and hydride anions (Table 1).^6,7,10

Table 1 Selected data for saline early main group metal hydrides^6,7,10

Group 1 MH	Δχ^a	M–H^b, Å	Lattice energy^c, kJ mol^−1	Group 2 MH2	Δχ^a	M–H^b, Å	Lattice energy^c, kJ mol^−1
a Electronegativity differences based on Allred–Rochow values. b Average experimental values (ref. 6), except for BeH2 (ref. 10). c Calculated values taken from ref. 7.
LiH	1.2	2.04	858	BeH2	0.7	1.41	3205
NaH	1.2	2.45	782	MgH2	1.0	1.95	2791
KH	1.3	2.86	699	CaH2	1.2	2.32	2410
RbH	1.3	3.03	674	SrH2	1.2	2.49	2250
CsH	1.3	3.20	648	BaH2	1.2	2.67	2171

---

All group 1 metal hydrides are of the NaCl structural type in which each M⁺ cation is surrounded by six H⁻ anions and vice versa each H⁻ anion by six M⁺ cations. In contrast, the series of group 2 metal hydrides possesses different structures. The heavier metal hydrides MH₂ (M = Ca, Sr, Ba) have at room temperature a PbCl₂-type structure (cotunnite) in which the metal cations have the high coordination number of nine, whereas H⁻ anions have short contacts to four or five metal cations. The room temperature form of MgH₂ has a TiO₂-type structure (rutile) in which Mg²⁺ binds to six H⁻ anions and each H⁻ anion to three Mg²⁺ cations. The structure of its lightest group member, BeH₂, has been shrouded in mystery for a long time (in fact, even its existence has been subject of discussion).⁸ The originally proposed one-dimensionally polymeric structure of [BeH₂]_∞, a model often found in student manuals, consists of a linear array of tetrahedral BeH₄²⁻ ions connected by Be²⁺ cations, i.e. edge-sharing BeH₄ tetrahedra (1).⁹ More recent accurate structural studies show a three-dimensional polymeric network of corner sharing BeH₄ tetrahedra in which each Be is four-coordinate and each hydride ligand connects two Be centers (2).¹⁰

The nature of the metal–hydride bond is naturally dependent on electronegativity differences which for group 1 metal hydrides hardly vary (Δχ = 1.2–1.3) but point to highly ionic character. It should, however, be noticed that also cation size plays a dominant role: the smaller cations are able to effectively polarize only weakly bound 1s² electrons of the H⁻ ion. The free H⁻ ion is estimated to have an ionic radius of 2.08 Å but is highly compressed in early main group salt structures (ionic radii of 1–1.4 Å).⁶ Compressibility increases along the row Cs < Li and is partially due to increased covalent character. Especially in BeH₂ (Δχ = 0.7) covalent contributions can be high and this was the reason to propose the wrong linear structure (1) which was explained by 3-center-2-electron bridging of H⁻ ligands (vide supra). The heavier group 2 hydride complexes are largely ionic in nature, however, some covalent character also has been found for MgH₂.¹¹

Noteworthy are also the very high lattice energies for early main group metal hydrides (Table 1). These decrease along group 1 from Li to Cs and greatly determine their chemical reactivity. Whereas NaH–CsH salts are increasingly reactive towards moisture, the salt with the highest lattice energy, [LiH]_∞, is essentially inert. Below red heat temperatures it is unaffected by O₂ or HCl.¹² Neither does it react with tertiary alcohols at room temperature.¹³ Group 2 metal hydrides show even larger lattice energies. The high lattice energy for MgH₂ seriously affects its practical use as a promising hydrogen storage material and is the reason why high temperatures of >300 °C need to be applied for the redox reaction: MgH₂(s) → Mg⁰(s) + H₂(g).¹⁴ Commercially available CaH₂ reacts only with water but not with other common organic functional groups (halides, ketones, cyanides). This makes it the drying agent of choice for a large array of organic solvents.

The reactivity of such highly stable saline metal hydrides can be strongly increased by taking them out of their crystal lattice and bringing them into a soluble form. Commercially available solutions of LiH/BH₃ and LiH/AlH₃ are probably the best sold reduction reagents,¹⁵ however, these should rather be viewed as ‘ate’ complexes Li⁺BH₄⁻ and Li⁺AlH₄⁻ in which the hydride ligands are strongly bound to the main group 3 element and Li⁺ only acts as a charge balancer. Although these compounds show polar hydride reactivity, their nature and structures are only remotely related to saline group 1 and 2 hydrides and therefore not part of this review.

The high lattice energies encountered also make it a real challenge to synthesize early main group metal hydride complexes. Whereas for most organolithium compounds large varieties of adducts have been structurally characterized, e.g. (PhLi·Et₂O)₄, (PhLi·TMEDA)₂ or (PhLi·PMDTA)₂, hardly any hydride complexes are known. For the alkali metals, complexes of the form L⁰–M⁺–H⁻ (in which L⁰ is a neutral ligand) should be feasible. For the alkaline-earth metals complexes of the type L⁰–M²⁺(H⁻)₂ or L⁻–M⁺–H⁻ could exist. However, facile ligand exchange reactions that are typical in early main group chemistry and high lattice energies of the insoluble salts [MH]_∞ and [MH₂]_∞ make the isolation of such ligand-stabilized compounds very challenging. Although the smaller metal cations display the largest lattice energies, metal hydride complexes with larger metal cations are the most difficult to obtain. This is due to the rapidly decreasing M–H bond strengths and therefore faster ligand exchange processes in going down the groups. Also, the large coordination sphere of these heavier metal cations gives rise to intermolecular interaction and easier ligand exchange. However, with the right combination of ligands soluble early main group metal hydride complexes can be obtained. As high lattice energies of these saline metal hydrides do not allow for synthesis of their complexes starting from the metal hydride salt itself, their synthetic protocol is usually based on the in situ formation of a molecular metal hydride in the presence of suitable stabilizing ligands.

3 Mixed-metal hydride complexes

A variety of transition metal complexes with incorporated group 1 or 2 metal hydrides have been reported.^16–24 On account of the high electropositivity of early transition metals and lanthanides, their mixed-metal complexes can often be described as aggregates containing a hydride ligand.

Floriani et al. reported a Zr^IV complex that can act as a “carrier for NaH or KH in toluene”.^16,17 Two disc-like molecules consisting of a fourfold deprotonated porphyrinogen and Zr^IV sandwich a (MH)₂ dimer (M = Na, K) (3). Although such compounds may react hydridic and thus could be seen as a hydrocarbon-soluble form of an alkali metal hydride, it would be more appropriate to describe these as zirconate complexes (LZrH⁻) that are bridged by two M⁺ ions. This vision is mainly based on estimated bond strengths: the Zr–H bond is not only more covalent, and therefore stronger than the M–H bonds, but also has an electrostatic component that is based on attraction to a highly Lewis-acidic Zr^IV center (instead of to the much weaker M⁺ Lewis acid). Similar ambiguities can be found in for example (Cp*₂ZrH₂)(LiH).¹⁸ This complex crystallizes as a trimer (4) which can be described either as three Cp*₂ZrH₂ molecules trapping a cyclic LiH-trimer or as three anionic zirconates held together by three Li⁺ ions: [(Cp*₂ZrH₃⁻)(Li⁺)]₃. The report correctly describes this complex as a zirconate.¹⁸

Several more examples of such cases can be found in Andersen's (Cp*₂Ti^IIIH₂⁻)(Li⁺·TMEDA) and Mach's (Cp*₂Ti^IIIH₂⁻)₂(Mg²⁺),¹⁹ Piers's [(ArN=ScCl·THF)₃H](Li⁺),²⁰ Hou's sphere-like complex Cp*₆H₁₅K₃Sm^III₆THF₆,²¹ Nöth's [(tBuCH₂O)₅Al₃H₅⁻](Li⁺)²² or Robinson's self-assembled organometallic sphere that resembles the seams of a tennis ball and “entraps” NaH.²³ In all these cases, the hydride anions are strongly bound to highly charged cations and consequently should be assigned to them. These species should be described preferably as containing an ‘ate’ anion.

Some exceptions or borderline compounds may be found in following examples. Gambarotta's samarium(III) complex consisting of a fourfold deprotonated porphyrinogen (like in 3), Sm^III and a Li₂H⁺ ion coordinated to the opposite site of the macrocycle (5, the “through-ring” Sm–H bond distance of 2.25(7) Å is rather long).²⁴ Also Veith recently reported that Li–Al hydride represents an example in which LiH could be described as being trapped in an organometallic complex (Fig. 1).²⁵ Although the complex has suggestively been reported as a cubane-like (LiH)₄ tetramer that is stabilized by interactions with three (Me₃Si)₂NAlH molecules, it would be better to describe this species as three aluminate anions (Me₃Si)₂NAlH₂⁻ that are bridged by Li⁺ connectors. The remaining Li⁺H⁻ unit in the middle (the diagonal of the Li₄H₄ cube), however, can be regarded as a truly encapsulated LiH unit (there are no contacts of this particular H⁻ ion to Al^III).

Fig. 1 The crystal structure of [(Me₃Si)₂NAlH₂⁻](Li⁺)₃(LiH) by Veith et al.;²⁵ carbon atoms on Me₃Si are not shown for clarity.

4 Group 1 metal hydride complexes

Very active forms of LiH can be obtained simply by deprotonation of H₂ by tBuLi but this process needs forcing conditions (200 bar).²⁶ Polar (co)solvents like TMEDA or THF drastically lower the energy barrier and allow conversion of nBuLi into LiH even at atmospheric pressures.²⁷ This method has been extended by Brandsma et al. to the heavier group 1 metals. Deprotonation of H₂ with “superbase mixtures” gave highly active, but not necessarily soluble forms of NaH, KH and even CsH (eqn (1)).^27–30Harder et al. demonstrated that highly active potassium hydride can also be obtained by reaction of a benzylpotassium derivative with PhSiH₃ (eqn (2)).³¹ The in situ generated “KH” was immediately converted in the course of catalytic alkene hydrosilylation. It is generally accepted that the formation of the hydride proceeds through anionic hypervalent intermediates with a five coordinate Si center (eqn (2)).³²

In general, the in situ prepared alkali metal hydride species precipitate from solution in their cubic crystal lattices. However, these remaining suspensions are extremely finely dispersed and highly reactive (the presence of nanoparticles should not be excluded, vide infra). The precipitation of these particles nicely demonstrates the challenge of finding solubilizing ligands that prevent conversion of molecular metal hydrides into highly stable crystalline salts. Some well-established examples of metal hydride incorporation into organometallic complexes or clusters are described below.

First well-defined lithium complexes with interstitial hydride have been found by Wheatley et al.³³ who demonstrated the favourable stability of clusters in which the H⁻ ion resides in an octahedral hole spanned by 6 Li⁺ ions (Fig. 2).^34–36 These clusters are prepared by a standard route in which β-H elimination from tBuLi is held responsible for in situ formation of H⁻ (eqn (3)). It is not easy to prove the presence of the light hydride ion in a Li₆H⁻ unit with standard X-ray structural methods. The hydride anion also often remains NMR silent. However, a recent neutron diffraction study resolved the ambiguity presented by these species.³⁶ Considering the high molecular weight of the total cluster (1072), the lithium hydride content is rather low.

Fig. 2 The cationic cluster {[(2-pyridyl)NPh]Li₈H}⁺ by Wheatley et al.³³ (Ph groups only partly shown for clarity); slight deviation from projection along an approximate threefold axis.

A most spectacular example of high lithium hydride incorporation can be found in the cluster (tBuOLi)₁₆(LiH)₁₇ which was characterized using X-ray diffraction by Thomas et al. (Fig. 3).³⁷ This gigantic cluster has been prepared by in situ generation of LiH through photostimulated β-H elimination of tBuLi in the presence of the bulky lithium alkoxide tBuOLi (eqn (4)). Crystals were obtained after keeping a pentane solution for several weeks at −90 °C. Although yields are not given, the procedure is reported as being reproducible.

Fig. 3 Crystal structure of (tBuOLi)₁₆(LiH)₁₇ by Thomas et al.³⁷ (tBuO groups only partly shown for clarity); the crystallographic mirror plane is close to the plane of projection.

The crystal structure shows a rather complicated aggregate with crystallographic mirror-symmetry (Fig. 3). Although it is difficult to describe its structure, numerous distinctive cubic features, typical for group 1 metal salts, can be recognized easily. The aggregate can be thought to consist of three fused partial cubic units with tBuO on the corners and Li⁺ and H⁻ ions filling the inner positions; such a tentative complete cubic unit would be of composition (tBuO)₈Li₁₃H₆. The formation of [LiH]_∞ is apparently disrupted by the incorporation of the sterically demanding tBuO groups.

Very recently, Stasch could obtain a simpler more defined LiH-containing complex.³⁸ The LiH is formed here in situ from nBuLi and PhSiH₃ similar to the reaction discussed for KH formation (eqn (2)). Crystals can be obtained in a reproducible yield of 62% and were characterized by X-ray diffraction (Fig. 4). The hydride ions have been located and Li–H distances vary from 1.96(4)–1.99(4) Å (which is close to the 2.04 Å value in cubic LiH). The ¹H NMR of a benzene solution shows one broad resonance for all hydride ligands. Above 60 °C, the complex is quite unstable and loses the encapsulated (LiH)₄ by precipitation of saline lithium hydride.

Fig. 4 Crystal structure of (LiH)₄ sandwiched between two [Ph₂P(Ph)NLi] dimers by Stasch³⁸ (only the C_ipso atoms of the Ph groups are shown).

5 Group 2 metal hydride complexes

Early work focusses on the syntheses of soluble beryllium hydride complexes of the group 2 metals.^39–46 Although BeH₂ has the highest lattice energy, Be²⁺ being the smallest cation also has the smallest coordination sphere and is easy to saturate with small ligands like hydrides. Moreover, the considerable metal–ligand bond strengths prevent ligand exchange and thus formation of insoluble BeH₂.

Early reports, mainly by G. E. Coates et al., describe the synthesis of hydrocarbon-soluble beryllium hydride compounds. Solutions of RBeH (R = Me, Et, tBu, nPe, Ph) could be obtained by mixing R₂Be, BeCl₂ and LiH in boiling ether.^39–41 A more efficient method for in situ generation of hydrides is based on ligand exchange with Me₃SnH which after reactions with R₂Be (R = Me, Et) gives RBeH that has been isolated as amine adducts.^42,43 An elegant method for hydride formation is β-hydride elimination: iPr₂Be thermally decomposes to iPrBeH and H₂C=CMe₂.⁴⁴ It should be mentioned that the characterization of these compounds is limited to methods available in the 60–70's. An early crystal structure of a beryllium hydride was reported for the beryllate (Et₂BeH⁻)Na⁺·Et₂O that crystallized as a dimer with H ligands bridging the Be centers (Be–H = 1.49 Å) but also with short Na–H contacts of 2.40 Å.⁴⁵ The first crystal structure of a “true” beryllium hydride complex was reported already as early as 1983 by Coates et al.: [Me₂NCH₂CH₂N(Me)BeH] (Fig. 5a).⁴⁶ Interestingly, the amide anions form the bridges between the Be centers whereas the H⁻ ligands reside at a terminal position (Be–H = 1.38(1) Å). Parkin et al. reported the crystal structure of the scorpionate complex Tp^tBuBeH which features a Be–H distance of 1.23(7) Å (Fig. 5b).⁴⁷ As the lattice energy for [BeH₂]_∞ is enormous (Table 1), this complex is likely thermodynamically less stable. However, the strength of the Be–H bond, which is expected to have considerable covalent character, precludes any ligand exchange via Schlenk equilibria. Moreover, the very condensed coordination sphere of Be would not allow for the formation of the coordination complex (Tp^tBu)₂Be. The complex Tp^tBuBeH is well-soluble in benzene and the hydride ligand shows a singlet ¹H NMR resonance at 5.00 ppm.

Fig. 5 (a) Crystal structure of the first beryllium hydride complex by Coates et al.⁴⁶ (b) Crystal structure of Tp^tBuBeH by Parkin et al.⁴⁷

Apart from beryllium hydride complexes, also the magnesium hydride chemistry was developed at an early stage. Although the existence of compounds like RMgH has been the subject of interest and controversy,⁴⁸ extensive work by Ashby et al. describes the syntheses of a large variety of magnesium hydride complexes. Their synthetic method is based on ligand exchange between R₂Mg and ether-soluble LiAlH₄. Although this protocol seems troublesome and outcomes might be unsure (two different anions are mixed with three different metal cations) the following compounds are claimed on the basis of elemental analysis and IR studies: PhMgH, PhMg₂H₃, MeMg₂H₃ and CpMgH.^49,50 Additional NMR and MW determination studies on CpMgH·THF led us to propose a dimeric structure with bridging H⁻ ions and terminal Cp⁻ and THF ligands.⁵¹ Also complexes of the type XMgH, in which X = Cl, Br, OR and NR₂, have been subject of intensive research but no clear-cut structural characterization has been presented.^52–55

Access to soluble forms of calcium hydride has been more troublesome, probably because organocalcium chemistry only started to develop since the last two decades.^5,56–60 Therefore, first examples describe synthetic routes based on highly active Ca vapour. Mixed with organics like benzene, these produced PhCaH species through C–H activation.^61,62 The compounds have been characterized by quench experiments but, as for the early magnesium hydrides, no NMR or structural proof is available.

First definite proof for a well-defined magnesium hydride complex was found in the form of an “inverse crown ether”, i.e. the hydride anions are encapsulated in a ring of alternating amide and metal ions (Fig. 6).^63,64 The synthesis is based on the thermal decomposition of the magnesate complexes [(iPr₂N)₃Mg⁻] M⁺ (M⁺ = Na⁺, K⁺) in which a hydride anion is generated in situ by β-hydride elimination in the iPr₂N⁻ anion (eqn (5)). The Mg–H distances in the resulting Na/Mg and K/Mg complexes are 1.89(2) and 1.94(2) Å, respectively, and both show similar ¹H NMR resonances at 3.76 and 3.78 ppm, respectively.

Synthesis of well-defined alkaline-earth metal hydride complexes of the heavier group 2 metals (Ca, Sr, Ba), which would be highly dynamic, is expected to be more challenging. First attempts to synthesize such complexes by Hanusa et al. were based on borane extraction from a calcium boron hydride precursor: (Me₃Si)₃CpCa(HBEt₃), however, no transfer of the hydride from B to Ca was observed.⁶⁵ Similarly, H. W. Roesky et al. reported that this approach also fails for magnesium or strontium hydride complexes.^66,67

Fig. 6 Crystal structure of the “inverse crown ether” complex encapsulating two hydride anions by Mulvey et al.^63,64

Harder et al. set out to prepare first defined calcium hydrides.⁶⁸ The rationale behind their approach was to make L–Ca–R complexes in which L is a bulky ligand and R a reactive group that could be converted in situ into a hydride ligand. As KN(iPr)₂ had previously been shown to be prone to thermal decomposition into KH and Me₂C=N−iPr,^63,64 (9-Me₃Si-fluorenyl)CaNiPr₂, which crystallizes as a dimer, was heated in benzene solution (eqn (6)) but insignificant decomposition was observed. In a second approach, a similar but monomeric benzylic calcium compound was treated with PhSiH₃ (eqn (7)). Although formation of benzyl-SiH₂Ph proves exchange of the benzyl anion for H⁻, only the homoleptic calcium complex with two fluorenyl ligands could be isolated, implying that the Schlenk equilibrium is operative here. As strongly coordinating multidentate ligands could stabilize a heteroleptic calcium hydride, it seemed that the bulky scorpionate ligand Tp^tBu would be a good candidate to prevent formation of (Tp^tBu)₂Ca. The observation that attempted synthesis of homoleptic (Tp^tBu)₂Ca was not possible even under extreme conditions⁶⁹ strengthened this idea. Reaction of Tp^tBuCaN(SiMe₃)₂ with PhSiH₃, however, gave the homoleptic complex (Tp^tBu)₂Ca.⁶⁸ The crystal structure gave insight into how two bulky scorpionate ligands can coordinate to a single Ca²⁺ center: one of the ligands showed the normal tridentate coordination mode (three Ns bind the metal) whereas the other is only bound bidentate (two Ns bind the metal) and displays an additional B–H⋯Ca²⁺ interaction, i.e. the “sting of the scorpion”.

Use of the bulky β-diketiminate ligand DIPP-nacnac, (2,6-iPr₂C₆H₃)NC(Me)C(H)C(Me)N(2,6-iPr₂C₆H₃), finally led to a stable heteroleptic calcium hydride. Reaction of (DIPP-nacnac)CaN(SiMe₃)₂·THF with phenylsilane in hexane (eqn (8)) gave after cooling a batch of large colourless crystals of the heteroleptic dimer: [(DIPPnacnac)CaH·(THF)] (Fig. 7). The crystal structure shows bridging H⁻ ligands which could be located (average Ca–H distance: 2.15(3) Å). The ¹H NMR spectrum shows for the hydride ligand a relatively sharp singlet at 4.45 ppm. The hydride complex dissolves well in benzene and shows a remarkable stability. Forcing reflux conditions do only after prolonged reaction times lead to ligand exchange into (DIPPnacnac)₂Ca and insoluble [CaH₂]_∞. This is rather surprising in the light of the fact that homoleptic (DIPPnacnac)₂Ca is highly stable on account of intramolecular C–H⋯π bonding.⁷⁰

Fig. 7 Crystal structure of the first calcium hydride complex by Harder et al.⁶⁸

Using the same DIPPnacnac ligand and a similar synthetic protocol, DIPPnacnacMgnBu + PhSiH₃ → DIPPnacnacMgH + nBuSiH₂Ph, Jones et al. could isolate an analogue dimeric magnesium hydride. As this particular synthesis was run in the absence of polar THF, a solvent-free dimer [DIPPnacnacMgH]₂ with bridging H⁻ ligands crystallized.⁷¹ Crystallization in the presence of THF, however, led to the THF-adduct [DIPPnacnacMgH·THF] which has a similar structure to that of the calcium hydride shown in Fig. 7. Coordination of THF hardly affects the Mg–H bond distances which are 1.96(3) and 1.95(2) Å in the THF-free complex and the THF-adduct, respectively. Instead, THF-coordination elongated the Mg–N bonds by circa 0.1 Å. Also the ¹H NMR shifts are similar: 4.03 and 4.21 ppm, respectively.

Harder et al. found an efficient alternative for the synthesis of [DIPPnacnacMgH]₂: overnight thermal decomposition of DIPPnacnacMgNH(iPr)BH₃ gave the product in 89% isolated yield.⁷² This route represents a good alternative to the silane route which for Mg is slow and low-yielding (2 days; 40%).

In later work, Jones et al. reported a dimeric magnesium hydride complex with bulky tBu substituents in the backbone (6).⁷³ Although these remote substituents induce more steric bulk at the metal center by “pushing” DIPP groups towards the metal, the Mg–H distances in this complex are shorter, i.e. 1.84(5) Å compared to 1.96(3) Å, but within standard deviation; the reported ¹H NMR shift is 3.83 ppm. The shortest Mg–H distance of 1.75(7) Å was found in a monomeric magnesium hydride complex (7) which was obtained by addition of DMAP to a solution of the dimer; the reported ¹H NMR shift is 4.65 ppm. Hitherto, this is the only example of a complex with a terminal Mg–H bond.

A unique redox reaction between a dimeric Mg^I complex and AlH₃ (in the presence of an N-heterocyclic carbene) yielded a somewhat less sterically shielded magnesium hydride dimer (eqn (9)) with mesityl substituents in the nacnac ligand (Mg–H = 1.88(2) Å).⁷⁴ Although a crystal structure could be obtained, the species is less defined in solution and decomposes through ligand exchange in the homoleptic species. This nicely illustrates how a slight decrease in the steric bulk of the supporting ligand results in considerable loss of stability. This more open magnesium hydride is also highly reactive with silicone grease and a dimer with bridging H⁻ and Me₂HSiO⁻ ions could be characterized⁷⁴ (in addition, a dimeric complex with H⁻ and I⁻ ions was found).⁷³

Earlier work by Michalczyk describes the synthesis of MgH₂ adducts with bidentate ligands by reaction of nBu₂Mg with PhSiH₃ in the presence of DME or TMEDA. This route, however, led to insoluble, amorphous powders.⁷⁵ Hill et al. attempted to stabilize the MgH₂ unit with a strongly coordinating, bulky N-heterocyclic carbene. The in situ preparation of MgH₂ from a Mg amide precursor was found to be incomplete and they isolated a mixed amide/hydride cluster (eqn (10)).⁷⁶ The cluster can be described as a C₂-symmetric hexanuclear compound with an adamantane-like Mg₄H₆-core and terminal amide and carbene ligands (Fig. 8). The Mg–H bonds range from 1.86(1)–1.90(1) Å and the ¹H NMR shows one broad signal at 2.52 ppm for the hydride ligands, indicating exchange on the NMR time scale. The complex is described as being robust and so far defies further exchange of amide ligands for hydride anions.

Fig. 8 Crystal structure of a tetranuclear magnesium hydride/amide complex by Hill et al.;⁷⁶ the DIPP substituents at the N-heterocyclic carbene ligand are only partially shown for clarity.

Based on the successful use of β-diketiminate (nacnac) ligands in the stabilization of group 2 metal hydride clusters, Harder et al. prepared a tetranuclear magnesium hydride cluster by using a coupled nacnac ligand (eqn (11)).⁷⁷ The Mg₄H₄-core of the structure (Fig. 9a) shows a similar adamantane-like geometry as Hill's tetranuclear magnesium hydride cluster with Mg–H bond distances varying from 1.77(1) to 1.85(1) Å. The Mg²⁺ ions are situated at the corners of a hypothetical tetrahedron of which four vertices are bridged by hydride ligands and two opposite vertices by the bis-nacnac ligands. Interestingly, there are two different hydride positions, H1 and H2, which give two triplets in the ¹H NMR spectrum at low temperature (2.85 and 3.64 ppm, ²J_H–H = 8.5 Hz). At higher temperature these signals coalesce into a singlet (T_coal = 25 °C, ΔG^‡ = 56(1) kJ mol⁻¹).

Fig. 9 Crystal structure of bis(nacnac)-stabilized magnesium hydride clusters by Harder et al.; partially shown for clarity.^77,78

By using bis-nacnac ligands with a larger para-phenylene bridge (eqn (12)), Harder et al. could obtain a significantly extended magnesium hydride cluster in a reproducible yield of circa 50%.⁷⁸ The larger span of the bridge allows for incorporation of two MgH₂ units (these are likely formed in situ by the reaction of an excess of nBu₂Mg with PhSiH₃). The crystal structure of this octanuclear magnesium hydride complex, containing a total of ten hydride ligands, is shown in Fig. 9b; Mg–H bond distances range from 1.72(3) to 1.90(3) Å (average 1.81(3) Å). Although the cluster is crystallographically C₂-symmetric, the propeller-like orientation of the ligands induces near C₃-symmetry. The Mg–H framework (Fig. 9c) is similar to a three-spoked wagon wheel in which three different kinds of hydride positions are present: (a) at the axis, (b) on the spokes, and (c) at the outside. These three different hydrides can be observed in ¹H NMR as a heptet (H(a), 0.55 ppm, ²J_H–H = 4.5 Hz), double doublet (H(b), 1.72 ppm, ²J_H–H = 4.5 and 5.2 Hz) and a triplet (H(c), 2.60 ppm, ²J_H–H = 5.2 Hz), respectively. These coupling patterns can not only be observed in solutions in aromatic solvents but also in THF. Moreover, the three signals with H–H coupling can even be observed at 100 °C. This is not only strong evidence for preservation of the solid state structure in solution but also demonstrates its unusual static behaviour.

The observed hydride–hydride coupling represents the first examples of He–Mg–H magnetic coupling in magnesium hydride species. These multinuclear clusters serve as molecular model systems for bulk MgH₂ as a hydrogen storage material (vide infra).

Very recent work on group 2 metal hydride chemistry relates to the isolation and crystallographic characterization of MeBeH stabilized by a bulky N-heterocyclic carbene (NHC).⁷⁹ This complex was obtained by the “silane method” (eqn (13)), a procedure that proved to be highly successful for isolation of calcium and magnesium hydride complexes.^{68,71,73,76–78} In contrast to the earlier discussed [Me₂NCH₂CH₂N(Me)BeH]₂ (Fig. 5a), in dimeric [MeBeHNHC]₂ the hydride ligands are bridging with Be–H distances of 1.55(2) Å (the ¹H NMR shift is 2.33 ppm). Attempted isolation of NHC stabilized BeH₂ by further reaction of the product with PhSiH₃ led to ring opening of the carbene ligand.

Hitherto, no examples of well-defined strontium and barium hydride complexes have been reported. As pointed out earlier, challenges to isolate and stabilize group 2 metal hydride complexes rapidly increase with metal size. The latest work on group 2 metal hydride chemistry is a report by Okuda et al. on the second example of a calcium hydride complex.⁸⁰ Using the “silane method” a calcium hydride complex was obtained that is stabilized by a tetradentate cyclene-derived monoanionic ligand (eqn (14)). The complex crystallizes as a trimer but one of the hydride ligands is lost to the silane byproduct, forming a non-coordinating hypervalent silicate (Fig. 10). The trimeric nature of this complex is likely due to the fact that the cyclene anion is a less bulky ligand than DIPPnacnac⁻ and the amide is an excellent bridging group. The trimeric bonding scheme allows for bridging amide Ns but only leaves place for two bridging H⁻ ions at the two faces of the Ca₃ triangle. The third H⁻ ion is incorporated in the silicate which should be considered an intermediate on the pathway for formation of calcium hydrides (see eqn (2)). Three different complexes, all with the same trimeric calcium hydride cation but with different silicate anions, have been isolated. The Ca–H bond distances range from 2.03(8) to 2.50(9) Å (average: 2.33(9) Å); their ¹H NMR chemical shift is found at 3.99 ppm.

Fig. 10 Crystal structure of a cationic trimeric calcium hydride complex stabilized by tetradentate ligands (Okuda et al.);⁸⁰ the non-coordinating anion is either Ph₃SiH₂⁻, Me₃SiN-SiPh₃⁻ or Ph(H)C-SiHPh₂⁻.

6 Relevance and applications in the field

Apart from being fundamentally important, early main group metal hydride complexes are also highly relevant species in various fields, e.g. in chemical reduction, as versatile synthons in organometallic chemistry, in catalysis or in hydrogen storage.

Early work on hydrocarbon-soluble magnesium hydrides like X–Mg–H (in which X is a carbanion, alkoxide or amide) was motivated by application in selective reduction. Highly bulky amides like [Me₃Si(tBu)N] MgH show excellent stereoselectivities in the addition to cyclic ketones.⁸¹ Apart from selectivities also activities of these reagents are clearly boosted by their good solubility.

Similar observations have been made by Harder et al. who investigated the reactivities of the first soluble calcium hydride complex [DIPPnacnacCaHTHF]₂ (Fig. 7). Whereas solid CaH₂ hardly reacts with any organic functional group and is primarily used as a drying agent for organic solvents containing a variety of functional groups, the solubilized calcium hydride displays smooth reactivity to a large array of substrates (Scheme 1) and gives access to a series of new calciumorganyl complexes mostly in crystalline purity.⁸² For this reason [DIPPnacnacCaH·THF]₂ has been nicknamed “worker-bee” in calcium chemistry.

Scheme 1 Reactions of the Ca–H functionality (for simplicity represented as a monomer).

One of the motivations for the syntheses and investigation of molecular group 2 metal hydrides originates from the proposal by Harder et al. that such complexes could be key intermediates in catalytic reactions like alkene hydrosilylation (Scheme 2, cycle A).³¹ Each step in the cycle has been separately proven in a stoichiometric protocol. The fact that for alkene hydrosilylation homoleptic R₂Ca complexes are more efficient catalysts than heteroleptic LCaR species suggests that also solubilized (CaH₂)_n aggregates could be active catalysts. Indeed, experimental observations hint the possible existence of larger hydride-rich clusters which are stabilized by ligands in the outer-shell (Scheme 2, route B).³¹ It is likely that undefined nanoclusters are formed prior to precipitation of [CaH₂]_∞. Although hydride complexes can also be used as catalysts in ketone hydrosilylation, it is unlikely that the metal hydride complex is the “true” catalyst in this process.⁸³

Scheme 2 Metal hydride based intermediates in catalysis.

The existence of group 2 metal hydride complexes led to the question whether such species could be used in hydrogenation catalysis (Scheme 2, cycle C). Harder et al. could show that [DIPPnacnacCaH·THF]₂ (Fig. 7) is catalytically active in alkene hydrogenation with molecular H₂ (20 bar).⁸⁴ Also in situ generated calcium, strontium and potassium hydride species were found to be active catalysts. These represent the first non-transition metal catalysts for alkene hydrogenation.

Groups of Harder, Hill and Jones have shown that molecular alkaline-earth metal hydride complexes are excellent precursors in the syntheses of metal amidoborane complexes.^72,85–90 These species have found renewed interest as hydrogen storage materials.⁹¹ Harder et al. found that the DIPPnacnac ligand used to stabilize molecular hydrides also serves as an excellent template for the stabilization of the thermal decomposition products of group 2 metal amidoborane complexes.^72,85,86,88 It has been proposed that molecular metal hydride complexes are important intermediates in the elimination of hydrogen upon thermal decomposition of these metal amidoboranes.

Magnesium hydride complexes are also catalytic intermediates for condensation of ammonia–borane derivatives (Scheme 2, cycle D). Substituents on N determine the course of this reaction. Harder et al. introduced a catalytic route for the synthesis of HB[NH(DIPP)]₂ from (DIPP)NH₂BH₃,⁹² whereas Hill et al. found that two smaller Me substituents on N gave the ring-like product (Me₂NBH₂)₂.⁸⁷ Similarly, heavier metal hydrides like solvated SrH₂ complexes have been proposed as intermediates in such reactions but hitherto there is no precedence for well-defined complexes.⁸⁹

Hill et al. observed that Jones's molecular magnesium hydride, [DIPPnacnacMgH]₂, dearomatizes pyridine by a 1,2-addition (which later isomerizes to the 1,4-addition product),⁹³ a reaction that is well-established for group 1 metal hydrides.⁹⁴ This is the basis for catalytic dearomatization–hydroboration of pyridine derivatives (Scheme 2, route E).⁹⁵ There are indications that the magnesium hydride is a recurrent species in this cycle. This strongly contrasts with the earlier reported Ca-mediated hydroboration of alkenes in which the calcium hydridoborate complex is the most stable species and disproportionation into a hydride and borane seems unlikely.⁹⁶ The magnesium-hydride mediated catalytic hydroboration has very recently been extended to addition–hydroboration of aldehydes and ketones.⁹⁷

Solid saline [MgH₂]_∞ is currently receiving tremendous attention as a hydrogen storage material. Although this material is not very rich in hydrogen (7.7 wt%), it has the main advantage of reversibly releasing or taking up hydrogen: MgH₂ ⇆ Mg + H₂.⁹⁸ However, its relatively high thermodynamic stability is directly translated to a rather high temperature for hydrogen release (ca. 300 °C). The enormous stability of saline [MgH₂]_∞ is mainly due to the rather high lattice energy for the rutile salt structure of [MgH₂]_∞: the experimentally determined value of ΔH is 2718 kJ mol⁻¹.⁹⁹ Recent calculations by de Jong et al. have shown that there is a strong influence of cluster size on the thermodynamic stability of (MgH₂)_n clusters.¹⁴ Aggregates with cluster size n < 19 show a sudden drop in stability due to loss in lattice energies (percentually more ions are situated at the surface of the cluster). The calculated temperatures for hydrogen release rapidly decrease with cluster size: for Mg₉H₁₈ a decomposition temperature of circa 200 °C has been calculated. Therefore, the question arises whether molecular magnesium hydride complexes could function as model systems for hydrogen storage materials. Although one could envision a possible reversible reaction between [DIPPnacnacMgH]₂ and [DIPPnacnacMg^I]₂ + H₂, the dimeric magnesium hydride complex is described as an extremely stable compound (dec. 300 °C) and also the dimeric Mg^I complex does not react with molecular hydrogen.⁷¹ However, Jones et al. found that [DIPPnacnacMgH]₂ can be chemically reduced to the Mg^I dimer with K. The resulting Mg^I species can be chemically oxidized again to [DIPPnacnacMgH]₂ by reaction with AlH₃ (cf.eqn (9)).⁷⁴

Harder et al.⁷⁸ recently found that the larger magnesium hydride cluster (Fig. 9b) releases molecular hydrogen already at 200 °C (note that this is comparable to the temperature predicted for Mg₉H₁₈).¹⁴ Gas quantification by Töpler pump measurements quite accurately showed release of all five equivalents of hydrogen. Decomposition experiments with dimeric [DIPPnacnacMgH]₂ and a tetranuclear magnesium hydride (Fig. 9a) showed that there is indeed a relationship between cluster size and decomposition temperature. In contrast to earlier findings, the dimer decomposed already at 150 °C, whereas the tetranuclear cluster needed a slightly higher temperature of 175 °C.⁷⁷ Hitherto, the nature of the decomposition products is unknown, but on account of the reddish colour it is speculated that Mg^I species could be prevalent. The reversibility of hydrogen storage in molecular magnesium hydride clusters is currently under investigation.

7 Conclusions

Although their synthesis and stabilization is challenging, early main group metal hydride complexes are now an established class of compounds. The arsenal of synthetic methods is steadily growing: whereas PhSiH₃ functions as the most frequently used hydride source, molecular H₂ or boranes and stannanes also can be used, as can soluble metal hydride compounds like LiAlH₄ or AlH₃ (and to a certain extent KH). Other methods include the β-hydride elimination route (e.g. for tBu⁻, iPr₂N⁻ or H₃B(iPr)NH⁻). Today, a variety of structures, from monomer, dimer, tetramer to higher aggregates, have been identified by single crystal X-ray diffraction.

These metal hydride complexes not only function as reagent of choice in selective reductions but also frequently are proposed to be key intermediates in catalytic cycles in which early main group metal compounds are operative. A thorough understanding of structure, reactivity and stability is therefore immediately of relevance for further development of the field of early main group metal catalysis.

In the light of the growing interest for hydrogen as a future energy source, fundamental understanding of early main group metal hydride complexes should be a valuable asset. Molecular magnesium hydride clusters indeed have been shown to release hydrogen at much lower temperatures than [MgH₂]_∞ salt itself. Such clusters can be accurately studied at the molecular level and not only exact bonding information but also the first H–Mg–H NMR coupling constants have been observed. It is likely that larger magnesium hydride clusters could function as model systems that facilitate the study of such hydrogen storage systems.

As this new compound class is still relatively new, there is ample room for further investigations on their reactivity, catalytic reactions and other applications. Especially the heavier metal hydrides (Rb, Cs, Sr and Ba), which are most challenging to stabilize, will be the subject of further development in the field. It is anticipated that the novel early main group metal hydride complexes will find increasing interest by the community and that further applications in various fields of chemistry will arise.

Notes and references

1. L. Schlapbach and A. Züttel, Nature, 2001, 414, 353.
2. G. S. McGrady and G. Guilera, Chem. Soc. Rev., 2003, 32, 383.
3. W. Grochala and P. P. Edwards, Chem. Rev., 2004, 104, 1283.
4. S. Aldridge and A. J. Downs, Chem. Rev., 2001, 101, 3305.
5. (a) S. Harder, Chem. Rev., 2010, 110, 3852; (b) A. G. M. Barrett, M. R. Crimmin, M. S. Hill and P. A. Procopiou, Proc. R. Soc. A, 2010, 466, 927.
6. A. F. Holleman, E. Wiberg and N. Wiberg, Lehrbuch der Anorganischen Chemie, Berlin, 2007, 102nd edn., de Gruyter.
7. Handbook of Chemistry and Physics, ed. R. C. Weast and M. J. Astle, CRC Press, Florida, 1981, 61st edn., p. D-94.
8. W. M. Mueller, J. P. Blackledge and G. G. Libowitz, Metal Hydrides, Academic Press, 1968.
9. G. J. Brendel, E. M. Marlett and L. M. Niebylski, Inorg. Chem., 1978, 17, 3589.
10. G. S. Smith, Q. C. Johnson, D. K. Smith, D. E. Cox, R. L. Snyder, R.-S. Zhou and A. Zalkin, Solid State Commun., 1988, 67, 491.
11. T. Noritake, M. Aoki, S. Towata, Y. Seno and Y. Hirose, Appl. Phys. Lett., 2002, 81, 2008.
12. F. A. Cotton and G. Wilkinson, Advanced Inorganic Chemistry, 1988, New York, John Wiley & Sons, p. 119.
13. C. A. Brown, J. Org. Chem., 1974, 39, 3913.
14. R. W. P. Wagemans, J. H. van Lenthe, P. E. de Jongh, A. J. van Dillen and K. P. de Jong, J. Am. Chem. Soc., 2005, 127, 16675.
15. H. C. Browne and S. Krishnamurthy, Tetrahedron, 1979, 35, 567.
16. D. Jacobi, C. Floriani, A. Chiesi-Villa and C. Rizzoli, J. Am. Chem. Soc., 1993, 115, 3595.
17. D. Jacobi, C. Floriani, A. Chiesi-Villa and C. Rizzoli, J. Am. Chem. Soc., 1993, 115, 7025.
18. N. Etkin, A. J. Hoskin and D. W. Stephan, J. Am. Chem. Soc., 1997, 119, 11420.
19. (a) W. W: Lukens Jr, P. T. Matsunaga and R. A. Andersen, Organometallics, 1998, 17, 5240; (b) S. I. Troyanov, V. Varga and K. Mach, Chem. Commun., 1993, 1174.
20. K. D. Conroy, W. E. Piers and M. Parvez, Organometallics, 2009, 28, 6228.
21. Z. Hou, Y. Zhang, O. Tardiff and Y. Wakatsuki, J. Am. Chem. Soc., 2001, 123, 9216.
22. H. Nöth, A. Schlegel, J. Knizek and H. Schwenk, Angew. Chem., Int. Ed. Engl., 1997, 36, 2640.
23. Y. Wang, B. Quillian, P. Wei, C. S. Wannere, P. v. R. Schleyer and G. H. Robinson, Organometallics, 2006, 25, 3286.
24. T. Dubé, S. Gambarotta and G. Yap, Organometallics, 2000, 19, 121.
25. M. Veith, P. König, A. Rammo and V. Huch, Angew. Chem., Int. Ed., 2005, 44, 5968.
26. E. C. Ashby and R. D. Schwartz, Inorg. Chem., 1971, 10, 355.
27. P. A. A. Klusener, L. Brandsma, H. D. Verkruijsse, P. v. R. Schleyer, T. Friedl and R. Pi, Angew. Chem., Int. Ed., 1986, 25, 465.
28. R. Pi, T. Friedl, P. v. R. Schleyer, P. A. A. Klusener and L. Brandsma, J. Org. Chem., 1987, 52, 4299.
29. P. A. A. Klusener, Synthetic Applications of Strongly Basic Reagents, Dissertation, University of Utrecht, 1989.
30. E. R. Burkhardt, C. P. Suddon, J. Brüning and D. F. Rouda, Patent WO, 0071552 (A1) 2000.
31. F. Buch, J. Brettar and S. Harder, Angew. Chem., Int. Ed., 2006, 45, 2741.
32. J. L. Brefort, R. Corriu, C. Guérin and B. Henner, J. Organomet. Chem., 1989, 370, 9.
33. D. R. Armstrong, W. Clegg, R. P. Davies, S. T. Liddle, D. J. Linton, P. R. Raithby, R. Snaith and A. E. H. Wheatley, Angew. Chem., Int. Ed., 1999, 38, 3367.
34. D. R. Armstrong, W. Clegg, R. P. Davies, S. T. Liddle, D. J. Linton, P. R. Raithby, R. Snaith and A. E. H. Wheatley, Angew. Chem., Int. Ed., 1999, 38, 3367.
35. S. R. Boss, M. P. Coles, R. Haigh, P. B. Hitchcock, R. Snaith and A. E. H. Wheatley, Angew. Chem., Int. Ed., 2003, 42, 5593.
36. S. R. Boss, J. M. Cole, R. Haigh, R. Snaith, A. E. H. Wheatley, G. J. McIntyre and P. R. Raithby, Organometallics, 2004, 23, 4527.
37. D. Hoffmann, T. Kottke, R. J. Lagow and R. D. Thomas, Angew. Chem., Int. Ed., 1998, 37, 1537.
38. A. Stasch, Angew. Chem., Int. Ed., 2012, 51, 1930.
39. U. Blindheim, G. E. Coates and R. C. Srivastava, J. Chem. Soc., Dalton Trans., 1972, 2302.
40. G. E. Coates, D. L. Smith and R. C. Srivastava, J. Chem. Soc., Dalton Trans., 1973, 618.
41. N. A. Bell and G. E. Coates, J. Chem. Soc. A, 1966, 1069.
42. N. A. Bell and G. E. Coates, J. Chem. Soc., 1965, 692.
43. G. E. Coates and M. Tranah, J. Chem. Soc. A, 1967, 236.
44. G. E. Coates and F. Glockling, J. Chem. Soc., 1954, 22, 615.
45. G. W. Adamson, N. A. Bell and H. M. M. Shearer, Acta Crystallogr., 1981, B37, 68.
46. N. A. Bell, G. E. Coates, M. L. Shearer and H. M. M. Shearer, Chem. Commun., 1983, 828.
47. R. Han and G. Parkin, Inorg. Chem., 1992, 31, 983.
48. G. E. Coates and J. A. Heslop, J. Chem. Soc. A, 1968, 514.
49. E. C. Ashby and A. B. Goel, Inorg. Chem., 1977, 1441.
50. E. C. Ashby and A. B. Goel, J. Org. Chem., 1977, 42, 22.
51. A. B. Goel and E. C. Ashby, J. Organomet. Chem., 1981, 214, C1.
52. E. C. Ashby and A. B. Goel, J. Am. Chem. Soc., 1977, 99, 310.
53. E. C. Ashby and A. B. Goel, Inorg. Chem., 1978, 17, 1862.
54. E. C. Ashby and A. B. Goel, Inorg. Chem., 1979, 18, 1306.
55. E. C. Ashby, S. A. Noding and A. B. Goel, J. Org. Chem., 1980, 45, 1028.
56. T. P. Hanusa, Coord. Chem. Rev., 2000, 210, 329.
57. J. S. Alexander and K. Ruhlandt-Senge, Eur. J. Inorg. Chem., 2002, 11, 2761.
58. M. Westerhausen, Coord. Chem. Rev., 2008, 252, 1516.
59. (a) F. G. N. Cloke, P. B. Hitchcock, M. F. Lappert, G. A. Lawless and B. Royo, Chem. Commun., 1991, 724; (b) C. Eaborn, S. A. Hawkes, P. B. Hitchcock and J. D. Smith, Chem. Commun., 1997, 1961.
60. (a) F. Feil and S. Harder, Organometallics, 2000, 19, 5010; (b) S. Harder, F. Feil and A. Weeber, Organometallics, 2001, 20, 1044; (c) S. Harder, F. Feil and K. Knoll, Angew. Chem., Int. Ed., 2001, 40, 4261.
61. K. Mochida, Y. Hiraga, H. Takeuchi and H. Ogawa, Organometallics, 1987, 6, 2293.
62. J. P. Dunne, M. Tacke, C. Selinka and D. Stalke, Eur. J. Inorg. Chem., 2003, 1416.
63. D. J. Gallagher, K. W. Henderson, A. R. Kennedy, C. T. O'Hara, R. E. Mulvey and R. B. Rowlings, Chem. Commun., 2002, 376.
64. P. C. Andrikopoulos, D. R. Armstrong, A. R. Kennedy, R. E. Mulvey, C. T. O'Hara and R. B. Rowlings, Eur. J. Inorg. Chem., 2003, 3354.
65. M. J. Harvey, T. P. Hanusa and M. Pink, Chem. Commun., 2000, 489.
66. J. Prust, K. Most, I. Müller, E. Alexopoulos, A. Stasch, I. Úson and H. W. Roesky, Z. Anorg. Allg. Chem., 2001, 627, 2032.
67. S. P. Sarish, A. Jana, H. W: Roesky, T. Schulz, M. John and D. Stalke, Inorg. Chem., 2010, 49, 3816.
68. S. Harder and J. Brettar, Angew. Chem., 2006, 118, 3554.
69. The thallium route, 2 Tp^tBuTl + CaI₂ → (Tp^tBu)₂Ca gave exclusively the product Tp^tBuCaI: M. H. Chisholm, J. C. Gallucci and K. Phomphrai, Inorg. Chem., 2004, 43, 6717.
70. S. Harder, Organometallics, 2002, 21, 3782.
71. S. P. Green, C. Jones and A. Stasch, Angew. Chem., Int. Ed., 2008, 47, 9079.
72. J. Spielmann, D. F.-J. Piesik and S. Harder, Chem.–Eur. J., 2010, 16, 8307.
73. S. J. Bonyhady, C. Jones, S. Nembenna, A. Stasch, A. J. Edwards and G. J. McIntyre, Chem.–Eur. J., 2010, 16, 938.
74. S. J. Bonyhady, D. Collis, G. Frenking, N. Holzmann, C. Jones and A. Stasch, Nat. Chem., 2010, 2, 865.
75. M. J. Michalczyk, Organometallics, 1992, 11, 2307.
76. M. Arrowsmith, M. S. Hill, D. J. MacDougall and M. F. Mahon, Angew. Chem., Int. Ed., 2009, 48, 4013.
77. S. Harder, J. Intemann and J. Spielmann, mansucript in preparation.
78. S. Harder, J. Spielmann, J. Intemann and H. Bandmann, Angew. Chem., Int. Ed., 2011, 50, 4156.
79. M. Arrowsmith, M. S. Hill, G. Kociok-Köhn, D. J. MacDougall and M. F. Mahon, Angew. Chem., Int. Ed., 2012, 51, 2098.
80. P. Jochmann, J. P. Davin, T. P. Spaniol, L. Maron and J. Okuda, Angew. Chem., Int. Ed., 2012, 51, 2098.
81. E. C. Ashby, J. J. Lin and A. B. Goel, J. Org. Chem., 1978, 43, 1564.
82. J. Spielmann and S. Harder, Chem.–Eur. J., 2007, 13, 8928.
83. J. Spielmann and S. Harder, Eur. J. Inorg. Chem., 2008, 1480.
84. J. Spielmann, F. Buch and S. Harder, Angew. Chem., Int. Ed., 2008, 47, 9434.
85. J. Spielmann, G. Jansen, H. Bandmann and S. Harder, Angew. Chem., Int. Ed., 2008, 47, 6290.
86. J. Spielmann and S. Harder, J. Am. Chem. Soc., 2009, 131, 5064.
87. D. J. Liptrot, M. S. Hill, M. F. Mahon and D. J. MacDougall, Chem.–Eur. J., 2010, 16, 8508.
88. J. Spielmann and S. Harder, Dalton Trans., 2011, 40, 8314.
89. P. Bellham, M. S. Hill, D. J. Liptrot, D. J. MacDougall and M. F. Mahon, Chem. Commun., 2011, 47, 9060.
90. C. Jones, S. J. Bonyhady, S. Nembenna and A. Stasch, Eur. J. Inorg. Chem., 2012 early view.
91. (a) Z. T. Xiong, C. K. Yong, G. T. Wu, P. Chen, W. Shaw, A. Karkamkar, T. Autrey, M. H. V. K. Diyabalanage, R. P. Shrestha, T. A. Semelsberger, B. L. Scott, M. E. Bowden, B. L. Davis and A. K. Burrell, Angew. Chem., Int. Ed., 2007, 46, 8995; (b) O. Jones, S. R. Johnson, P. P. Edwards and W. I. F. David, Nat. Mater., 2008, 7, 138.
92. J. Spielmann, M. Bolte and S. Harder, Chem. Commun., 2009, 6934.
93. M. S. Hill, D. J. MacDougall and M. F. Mahon, Dalton Trans., 2010, 39, 11129.
94. (a) K. Ziegler and H. Zeiser, Chem. Ber., 1930, 63, 1847; (b) D. R. Armstrong, R. E. Mulvey, D. Barr, R. Snaith and D. Reed, J. Organomet. Chem., 1988, 350, 191; (c) W. Clegg, L. Dunbar, L. Horsburgh and R. E. Mulvey, Angew. Chem., Int. Ed., 1996, 35, 753.
95. M. Arrowsmith, M. S. Hill, T. J. Hadlington, G. Kociok-Köhn and C. Weetman, Organometallics, 2011, 30, 5556.
96. S. Harder and J. Spielmann, J. Organomet. Chem., 2012, 698, 7.
97. M. Arrowsmith, T. J. Hadlington, M. S. Hill and G. Kociok-Köhn, Chem. Commun., 2012, 4567.
98. C. Wu and H.-M. Cheng, J. Mater. Chem., 2010, 20, 5390.
99. C. M. Stander and R. A. Pacey, J. Phys. Chem. Solids, 1978, 39, 829.

---

Footnotes

† This article is part of the ChemComm ‘Frontiers in Molecular Main Group Chemistry’ web themed issue.

‡ Current address: Anorganische Chemie, Friedrich-Alexander Universitaet Erlangen, Egerlandstrasse 1, 91058 Erlangen, Germany. E-mail: sjoerd.harder@chemie.uni-erlangen.de; Tel: +49 9131 8527350.

---