Modern organometallic and coordination chemistry of beryllium

G Model

ARTICLE IN PRESS

CCR-111961; No. of Pages 9

Coordination Chemistry Reviews xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Coordination Chemistry Reviews journal homepage: www.elsevier.com/locate/ccr

Review

Modern organometallic and coordination chemistry of beryllium Kalon J. Iversen, Shannon A. Couchman, David J.D. Wilson, Jason L. Dutton ∗ Department of Chemistry, La Trobe Institute for Molecular Science, La Trobe University, Melbourne, Victoria, Australia

Contents 1. 2.

3.

4.

Introduction and general considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Be(II) chemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Carbon based ligands with Be(II) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.1. Be(II)–carbone chemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.2. Be(II)–NHC chemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.3. Be(II)–allyl chemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Be(II) with nitrogen based ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.1. Be(II) with neutral monodentate N-ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.2. Be(II) with anionic chelating N-ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.3. Be(II) with the phthalocyanine ligand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.4. Light emitting N-ligated Be(II) compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3. Be(II) with phosphine based ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Low oxidation state Be compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Be(I) compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Be(0) compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary and outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

a r t i c l e

i n f o

Article history: Received 30 September 2014 Received in revised form 19 November 2014 Accepted 19 November 2014 Available online xxx

00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00

a b s t r a c t Recent developments in the coordination and organometallic chemistry of beryllium are reviewed. The primary focus is on synthetic as well as theoretical reports relevant to the synthetic chemist since 2008 on molecular Be complexes bound to C, N and P ligands. © 2014 Elsevier B.V. All rights reserved.

Keywords: Beryllium Organometallic Coordination chemistry N-ligands C-ligands P-ligands

1. Introduction and general considerations Due to its relatively high electronegativity, beryllium is unique in comparison with the heaver alkaline earth elements with respect to the chemistry that can be accessed. However, research into

∗ Corresponding author. Tel.: +61 0394793213. E-mail address: [email protected] (J.L. Dutton).

beryllium chemistry is suppressed by the extreme toxicity of Be metal, Be-containing salts and beryllium compounds, which arises primarily via inhalation of dust containing Be compounds [1]. The danger is sufficiently acute as to prevent most research groups from investigating synthetic beryllium chemistry in any significant extent. There are even recent papers specifically devoted to the synthesis of beryllium dihalide starting materials less prone to producing dust [2]. A survey of the Cambridge Structural Database shows a total of 59 compounds containing C–Be bonds, 116

http://dx.doi.org/10.1016/j.ccr.2014.11.008 0010-8545/© 2014 Elsevier B.V. All rights reserved.

Please cite this article in press as: K.J. Iversen, et al., Coord. Chem. Rev. (2014), http://dx.doi.org/10.1016/j.ccr.2014.11.008

G Model

ARTICLE IN PRESS

CCR-111961; No. of Pages 9

K.J. Iversen et al. / Coordination Chemistry Reviews xxx (2014) xxx–xxx

2

Table 1 Number of Cambridge Structural Database hits for Be, Mg, Zn and Al compounds with C, N and P bonds of any type.

Be Mg Zn Al

C

N

P

59 647 1386 3504

116 1334 15,953 3188

4 39 126 195

Scheme 1. Synthesis of 2 via direct coordination reaction using carbodiphosphorane 1.

compounds containing N–Be bonds and 4 compounds containing P–Be bonds. In Table 1, the number of hits for the same bonds in Mg, Zn and Al compounds is compared, as these atoms possess chemistry that is most in common with Be. Clearly, exploration of Be chemistry is highly suppressed in comparison. Nevertheless, there have been some major advances in the organometallic and coordination chemistry of Be in recent years, both from an experimental and a theoretical/computational perspective. This review covers advances made since Dehnicke’s review on Be chemistry in 2008 (and relevant earlier examples not discussed in Dehnicke’s review) in Be compounds using C-, Nand P-based ligands [3]. We also would direct the reader interested in the alkaline earth elements to Hanusa’s recent review in Comp. Inorg. Chem., covering developments for the entire group over approximately the last 30 years [4]. Bonds containing Be exhibit a higher degree of covalent character compared to other s-block elements due to the higher Pauling electronegativity of Be (1.57, cf. Mg 1.31, Ca 1.00, Li 0.98, Na 0.93). For example, the metallocene Cp2 Mg has the expected sandwich structure, but in Cp2 Be one of the Cp rings binds via a covalent ␴ interaction, albeit in a highly fluxional manner [5]. Bridging groups (i.e. Cl, H, CH3 ) are also a common feature in Be chemistry, which is similar to that seen in the chemistry of Zn and Al, which are also elements often featuring electron deficient environments and having an electronegativity similar to that of Be (Zn 1.65, Al 1.61). Be has a convenient NMR handle, with 9 Be having a spin of 3/2 and a 100% natural abundance. 2. Be(II) chemistry 2.1. Carbon based ligands with Be(II) 2.1.1. Be(II)–carbone chemistry Carbones (e.g. carbodiphosphorane 1) are ligands that have found increasing interest in recent years following Frenking’s identification of these ligands as compounds containing carbon in a formal zero oxidation state [6,7]. Carbones exhibit two lone pairs of electrons about the central carbon atom and hence they can act as two-electron donors (Lewis bases) towards two Lewis acids, or as a four-electron Lewis base to a single acidic centre. Be(II) is an attractive element centre with which to explore coordination chemistry of carbones, as X–Be–X Lewis acids (e.g. BeCl2 ) are formally four electron species, and can therefore accept a further four electrons. Dehnicke and coworkers have reported on the reaction of BeCl2 with carbodiphosphorane 1 (Scheme 1) [8]. In solvents typically used in such reactions (e.g. THF), only protonated 1 was isolated. Arene solvents such as toluene and benzene presented issues since BeCl2 is insoluble in such solvents. However, use of

2-bromofluorobenzene allowed for facile isolation of 2. Exposure of the isolated compound to a wide variety of organic solvents resulted in the production of either singly or doubly protonated carbodiphosphorane, indicating that solvent choice is a crucial consideration in this system. ˚ which is The Be–C bond distance in 2 was found to be 1.74 A, typical of a Be–C single bond. Theoretical studies indicated the ␲ symmetric lone pair was only very weakly involved in the Be–C bonding, which may explain the ready decomposition of the complex as the Be remains electron deficient. A theoretical study of the analogous BeH2 complex (3) predicted the Be–C bond strength (28 kcal/mol) to be substantially weaker than in 2 (43 kcal/mol) [9]. Theoretical analysis of this hypothetical BeH2 adduct indicated that the H atoms are highly hydridic, with partial charges of −0.36 e. In comparison, charges of only −0.02 are calculated for the isoelectronic and isolated [BH2 ]+ adduct of the carbodiphosphorane (4), which indicates that Be–H bonds in adducts will potentially exhibit chemistry quite different to that of related compounds having hydrogen bound to more electronegative elements. The [BH2 ]+ adduct was also predicted to have substantially more double bond character in the B–C bond as compared to the BeH2 adduct, which may indicate it is the reduced electronegativity of Be that is suppressing formation of a double bond in adducts with the carbone.

2.1.2. Be(II)–NHC chemistry Since Arduengo’s isolation of a free N-heterocyclic carbene (NHC) in 1991 there has been substantial interest in using NHCs as ligands for elements across the entire periodic table. Beryllium is no exception, but as is typical for Be, there is a limited volume of examples. Nonetheless, some of the higher profile reports in Be–C chemistry from the last two years are associated with the use of NHCs as ligands. The first NHC adduct of Be was reported by Herrmann in 1995 where the reaction of three equivalents of Me2 NHC with BeCl2 gave a cationic tris–NHC adduct of [BeCl]+ as a Cl− salt (5; Scheme 2) [10]. This report was also an early indication that NHC–metal adducts do not require ␲-backbonding to be stable complexes, as Be2+ lacks any electrons to contribute in this respect. Direct reaction of BeCl2 with a bulkier 2,6-diisopropylphenyl substituted NHC gave the simple three-coordinate NHC–BeCl2 adduct 6 [11]. The first NHC complex of a diorganoberllylium was the NHC adduct of Ph2 Be reported in 2006, which gives a three-coordinate species 7 [12]. In this compound, the C–Be bond for the NHC was found to be longer than the C–Be bonds to the phenyl rings, with ˚ respectively. distances of 1.80 and 1.75 A, Robinson and co-workers used 6 as a starting compound to generate an NHC stabilized analogue of beryllium borohydride by reaction of 8 with two equivalents of LiBH4 (Scheme 3) [11]. The geometry about the Be atom in 8 is unusual in that it features dual bridging hydrogen atoms between the Be and B atoms along both bond axes where the chlorides were displaced. Reaction of 8 with [Na]2 [Fe(CO)4 ] resulted in an unusual complete reduction of one NHC ring in the isolated product and formation of an NHC–B bond (9), although the fate of the Be atom in this reaction was not specified.

Please cite this article in press as: K.J. Iversen, et al., Coord. Chem. Rev. (2014), http://dx.doi.org/10.1016/j.ccr.2014.11.008

G Model CCR-111961; No. of Pages 9

ARTICLE IN PRESS K.J. Iversen et al. / Coordination Chemistry Reviews xxx (2014) xxx–xxx

3

Scheme 4. Insertion of “BeH2 ” into the C–N bond of an NHC reported by Hill and co-workers.

Scheme 2. Synthesis of NHC complexes of BeCl2 and BePh2 (Dipp = 2,6diisopropylphenyl).

An intriguing reactivity of Be in NHC complexes was recently described by Hill, with insertion of Be into the C–N bond of an NHC ring (Scheme 4) [13]. In the first instance, metathesis of the chloride atoms in 6 by MeLi gave the dimethyl Be complex 10. Reaction with PhSiH3 gave the mixed methyl/hydride bridged compound 11. Hill and co-workers identified compound 11 as significant, since it is the first example of a crystallographically characterized organoberyllium hydride. Heating 11 resulted in the ring expansion of the NHC and insertion of the Be atom into the endocyclic C–N bond of the ring, giving 12. Hill proposed that insertion occurred though a transient NHC–BeH2 intermediate, generated either from using H3 SiPh as an external hydride source, or heating 11 to a sufficiently high temperature to decompose it into NHC–BeH2 and Me2 Be(THF)2 . It has been shown theoretically that the likely intermediate is a bis–NHC complex NHC–BeH2 –NHC, as the kinetic barrier for the insertion of this species is much lower [14]. This class of ring insertion reaction into NHCs [15] has subsequently been observed experimentally for systems involving Si, B and Zn [16–20], and has been the subject of detailed theoretical studies [14,21–24]. This class of reaction has been identified by all experimentalists as being significant in the context of the

decomposition of NHC–metal containing catalysts, which was demonstrated in a catalytic situation with the Zn example [20]. 2.1.3. Be(II)–allyl chemistry Hanusa and co-workers reported the use of bulky allyl ligand 13 to stabilize the first monomeric Be–allyl species 14 as an etherate adduct via a metathesis reaction between BeCl2 and two equivalents of potassium allyl (Scheme 5) [25]. Consistent with the aforementioned preference of Be for ␴ bonding, the crystal structure exhibited a Be centre sigma bound to two allyl ligands with Et2 O completing a trigonal planar arrangement about the Be. In solution the compound was observed to be fluxional via NMR spectroscopy. Computational investigations on solvent-free bis–allyl complex showed the ␲-bound complex to be 12–17 kJ/mol lower in energy that the ␴-bound form. Explicit incorporation of the Et2 O solvate into the computational model resulted in production of an optimized geometry similar to that observed in the solid state. Finally, if the reaction was carried out with equimolar amounts of BeCl2 and allyl 13, then 9 Be NMR studies of the initial reaction mixture allowed the observation of a Schlenk equilibrium between 14 and 15, which is the first instance of such an observation in organoberyllium chemistry. Theoretical studies of Be2+ interacting with open chain pentadienyl anions (16) predicted the lowest-energy conformation to be ␴-bound to a one ligand, with the other ligand bound in a sickle Á − 5 mode to the Be atom [26].

2.2. Be(II) with nitrogen based ligands

Scheme 3. Synthesis and reactivity of NHC stabilized Be(BH4 )2 .

2.2.1. Be(II) with neutral monodentate N-ligands The use of relatively “hard” N-based ligands is fairly common in Be chemistry, with several recent examples incorporating both neutral and anionic nitrogen donors. Neumüller and Dehnicke have recently reported a variety of examples where a mixture of Be powder, I2 and a nitrogen-based ligand resulted in in situ formation of BeI2 , followed by formation of a Be coordination complex as an iodide or polyiodide salt (Scheme 6). Examples include pyridine, imidiazole and carbodiimide complexes (e.g. 17–19) [27–29]. The

Please cite this article in press as: K.J. Iversen, et al., Coord. Chem. Rev. (2014), http://dx.doi.org/10.1016/j.ccr.2014.11.008

G Model CCR-111961; No. of Pages 9 4

ARTICLE IN PRESS K.J. Iversen et al. / Coordination Chemistry Reviews xxx (2014) xxx–xxx

Scheme 5. Formation of organometallic Be(II)–allyl compounds.

carbodiimide adduct was able to catalyze the cyclodimerization of further added equivalents of carbodiimide giving 22, and the pyridine adduct was shown to activate CH2 Cl2 giving dication 20 and the bispyridine adduct of BeCl2 (21). The reaction of Be0 with Br2 in MeCN solvent gives the bisacetonitrile adduct of BeBr2 , analogous to the formation of 17–19, although without displacement of the halide [30].

Condensation of NH3 onto BeCl2 was shown to result in formation of highly water sensitive tetrakis-NH3 ligated Be dication 23, which was the first crystallographic confirmation of the long considered NH3 /BeCl2 system (Scheme 7) [31]. Careful structural and multinuclear NMR studies also shed light on the hydrolysis products of 23, involving formation of several mixed bridging OH and terminal NH3 ligated di- and tri-beryllium species.

Scheme 6. Oxidative synthesis and reactivity of Be(II)–N coordination complexes reported by Neumüller and Dehnicke [27–30].

Please cite this article in press as: K.J. Iversen, et al., Coord. Chem. Rev. (2014), http://dx.doi.org/10.1016/j.ccr.2014.11.008

G Model CCR-111961; No. of Pages 9

ARTICLE IN PRESS K.J. Iversen et al. / Coordination Chemistry Reviews xxx (2014) xxx–xxx

5

Scheme 7. Formation of homoleptic tetrakis-NH3 Be(II) dication via condensation of NH3 onto BeCl2 .

2.2.2. Be(II) with anionic chelating N-ligands Introduction of anionic chelating N-based ligands to Be follows the typical route used for other elements via metathesis reactions with BeX2 halides. The ␤-diketiminate ligands have attracted much attention in main group chemistry, and the spectacular example of an Mg–Mg bond containing Mg in the +1 oxidation state from Stasch and Jones has led to increased attention for this ligand in s-block chemistry [32]. Hill and Jones independently reported the syntheses of threecoordinate BeCl and BeI species 25X via reaction with either Li or K diaryl ␤-diketiminate ligands 24M (Scheme 8) [33,34]. Derivatives with alkyl groups at Be (26R) have been synthesized via reaction of the protonated ␤-diketiminate ligand with alkyl lithium bases (e.g. MeLi) in the presence of BeCl2 . The alkyl species were further reacted with water or alcohols to give Behydroxides or Be-alkoxides (27) under very mild conditions in high yields (Scheme 9). The hydroxide species is the first example of a three-coordinate Be-hydroxide. Beryllium-amides (28) have been synthesized from the alkyl species in a similar manner by direct reaction with H2 NR reagents, although in this case heating to 110 ◦ C for 3–5 days was required. The ␤-diketiminate Be–I complex 25I has been shown to controllably ring open THF, giving alkoxide compound 29. This observation is significant as THF is often a preferred solvent for main group chemistry, including Be [35]. Schulz and co-workers have described the use of the 1tris(pyrazolyl)borate ligand (30) as a versatile platform for functionalization of L–Be–X species. Reaction of the potassium salt of this tridentate ligand with BeX2 halides (X = Cl, Br, I) gave tetrahedral L-Be-X compounds 31X in high yields [36]. The X group could be converted to H, F and N3 using straightforward reagents as illustrated in Scheme 10. The work of Schulz and co-workers demonstrated the suitability and efficacy of 9 Be NMR for analyzing the bonding environment in Be compounds, which is enhanced by only requiring minimal exposure to these potentially harmful compounds.

Scheme 8. Synthesis of halo and alkyl 3-coordinate Be–B–diketiminate complexes (Ar = aryl group).

Scheme 9. Reactivity of Be(II) ␤-diketiminate complexes with water, alcohols, amines and THF.

2.2.3. Be(II) with the phthalocyanine ligand Janczak and co-workers have investigated the chemistry of Be2+ with the macrocyclic phthalocyanine ligand (Scheme 11). The synthesis is standard for phthalocyanine complexes, by reaction of an excess of 1,2-cyanobenzene with Be metal at elevated

Scheme 10. Synthesis and metathesis chemistry of 1-tris(pyrazolyl)borate Be(II) complexes.

Please cite this article in press as: K.J. Iversen, et al., Coord. Chem. Rev. (2014), http://dx.doi.org/10.1016/j.ccr.2014.11.008

G Model CCR-111961; No. of Pages 9 6

ARTICLE IN PRESS K.J. Iversen et al. / Coordination Chemistry Reviews xxx (2014) xxx–xxx

Scheme 11. Synthesis and N/O competition reactions of phthalocyanine–Be(II) compounds.

temperatures giving 34 [37]. From the perspective of coordination chemistry, the Be in the phthalocyanine remains Lewis acidic and forms bonds in the axial position when reacted with a variety of pyridine ligands (35) [38–40]. In all cases the axial Be–N bond from ˚ is slightly shorter than the equatorial the pyridine (average 1.86 A) ˚ However, Be–N bonds from the phthalocyanine (average 1.90 A). the pyridine ligand is easily displaced by water, giving 36 even with only trace amounts of moisture. This indicates that an N-based pyridine donor does not compete effectively with harder O-based ligands at Be centres. On displacement the water remains intact as an aqua ligand and is not deprotonated by the displaced pyridine [41], which instead acts as a hydrogen-bond acceptor in the crystal structures (including from water).

2.2.4. Light emitting N-ligated Be(II) compounds Beryllium compounds with ligands containing mixed phenoxide and neutral N-donor (pyridine, amine or imidazole) sites demonstrate emissive properties. This has recently been exploited to generate blue emitting OLEDs containing Be (37) [42], PHOLEDS that can have their emissive wavelength tuned based on ligand substitution (38) [43], and emissive compounds that allow Be2+ to be sequestered and imaged in cells (39) [44]. In all cases the incorporation of Be into the chelate is achieved via the simple reaction of BeSO4 with the ligand (Scheme 12). The first two examples demonstrate the potential of Be in materials chemistry, while the last example 39 is important as the specific mechanisms by which Be is so toxic is unknown, and this compound allows for Be to be

Scheme 12. Synthesis of emitting N/O–Be(II) coordination compounds.

Please cite this article in press as: K.J. Iversen, et al., Coord. Chem. Rev. (2014), http://dx.doi.org/10.1016/j.ccr.2014.11.008

G Model CCR-111961; No. of Pages 9

ARTICLE IN PRESS K.J. Iversen et al. / Coordination Chemistry Reviews xxx (2014) xxx–xxx

Scheme 13. Synthesis of dppm–BeCl2 complex via direct coordination reaction.

effectively tracked in cells. The ligand used to sequester Be2+ in 39 was found to be highly selective for Be2+ in the presence of other metals. 2.3. Be(II) with phosphine based ligands There are limited examples of phosphine ligands being used in Be chemistry, with only four “hits” on the Cambridge Structural Database for compounds containing Be–P bonds. While for coordination chemistry it could be assumed that phosphine ligands (soft) and the Be2+ cation (very hard) are poorly matched, recent examples show that this is simply not the case. The simple reaction of two equivalents of the dppm ligand with BeCl2 in CH2 Cl2 gives compound 40, with two phosphine ligands bound in a monodentate fashion to the Be centre (Scheme 13) [45]. The other P atom on each ligand remains free. The calculated bond dissociation energy for the first ligand was 71 kJ/mol, while for the second ligand it was 100 kJ/mol, which is consistent with a relatively weak dative bond. The second recent example of Be–P coordination chemistry involves an interesting reaction between (Cy3 P)2 –Pd and BeCl2 , which results in production of Be(II) coordination compound 41 with precipitation of elemental palladium (Scheme 14) [46]. The transfer of the “soft” phosphine from Pd to very hard Be2+ is counterintuitive with respect to Hard-Soft Acid–Base theory. The reaction of BeCl2 with the platinum analogue (Cy3 P)2 –Pt gave a compound in which the electron-rich Pt centre formed a dative bond to an intact BeCl2 unit (42) [47]. Reaction of this compound

Scheme 14. Reactivity of low coordination Pd/Pt0 complexes with BeCl2 .

7

with 4-methylpyridine gave the pyridine adduct of BeCl2 (43) and the free Pt complex. Conversely, metathesis reaction with MeLi allowed for functionalization of the Be centre, giving a methylated organometallic Be compound (44) still bound to the Pt. The Be–M bonding in (R3 P)2 M–BeCl2 complexes has been analyzed in detail by Frenking and co-workers for M = Ni, Pd, Pt [48]. It was found that the metal in these compounds was donating four electrons to the Be centre and that the Be–M ␴ bond had characteristics consistent with a donor–acceptor complex (a dative Be–M bond). 3. Low oxidation state Be compounds 3.1. Be(I) compounds A major recent advance in s-block chemistry was the aforementioned report of homoatomically bound Mg(I) complexes by Stasch and Jones (e.g. 46; Scheme 15), using the ␤-diketiminate ligand. The guanidinate ligand has also been used in Mg(I)–Mg(I) compounds [32]. These unique molecules have found use as strong but controllable, and more importantly, soluble reducing agents for the formation of low-oxidation state main group compounds [49]. In general, there has been ongoing interest in unusual homoatomic bonds between main group elements (i.e. beyond B, C, N and P catenation) since West’s early 1980s work on Si Si double bonds [50]. For some time now, the group of power have used extremely bulky anionic ligands to stabilize and explore the reactivity of main group homoatomic single, double and triple bonds [51]. Most recently, NHC ligands have been used to investigate the chemistry of homoatomic bonds for p-block elements of the form NHC–E–E–NHC [52–54]. Because the NHC ligands are neutral, the homoatomic fragments bound between them contain elements in the formal 0 oxidation state. These electron rich compounds have displayed new bonding (i.e. a B–B triple bond) [55] and reactivity. Alkaline earth compounds of the type R–Ae–Ae–R with the alkaline earth element in the +1 oxidation state were reviewed in 2010; here we will limit discussion to examples specific to Be since that time, or not covered in the previous review [56]. The possibility of extending these ideas to Be for both Be(I) and Be(0) diatomic complexes has been investigated primarily using theoretical methods. The Be analogue of 46 (47) was calculated to be a stable entity, even more stable than the Mg analogue as assessed by calculated E–E dissociation energy (250 kJ/mol for Be, 188 kJ/mol for Mg; Scheme 15) and oxidation potential [57]. It is possible to conclude that 47 will be a less reactive species than 46 with respect to reduction chemistry, however given the risk of working with Be as compared to Mg, perhaps it is not an attractive target. It should be noted that attempted reductions of

Scheme 15. Theoretical homoleptic dissociation reaction of homoatomic Mg and Be ␤-diketiminate compounds to determine E–E bond strength.

Please cite this article in press as: K.J. Iversen, et al., Coord. Chem. Rev. (2014), http://dx.doi.org/10.1016/j.ccr.2014.11.008

G Model CCR-111961; No. of Pages 9

ARTICLE IN PRESS K.J. Iversen et al. / Coordination Chemistry Reviews xxx (2014) xxx–xxx

8

␤-diketiminate–BeX species (X = halide) have thus far not resulted in successful isolation of compounds such as 47, instead leading to side reactions involving H-atom abstraction from the “backbone” methyl groups [33]. Schaefer and co-workers carried out a theoretical investigation of the reduced diimine Ae–Ae bound compounds for all the alkaline earth metals in the formal +1 oxidation state (e.g. 48 for Be) [58]. As with the ␤-diketiminate analogues, the Be–Be bonding was found to be the strongest of the group (225 kJ/mol, c.f. Mg 167 kJ/mol,

Ca 100 kJ/mol), and slightly smaller in magnitude for Be–Be dissociation compared to Cp–Be–Be–Cp (280 kJ/mol) [59]. The Be–Be dissociation energies for these two compounds are similar to the Zn–Zn dissociation energy (292 kJ/mol) in the isolated Zn(I) compound Cp*–Zn–Zn–Cp* first reported in 2004 [60]. Interestingly, in Schaefer’s study on reduced diimine ligands, explicit incorporation of Na+ counterions into the model resulted in the diberyllium compound displaying substantially higher stability.

A homoatomic Be(I) dimer bound by the open-chain pentadienyl anion described in the report of 16 [26] has been the subject of theoretical studies (49) [61]. Results indicate a covalently bound Be–Be unit with each Be bound to one open pentadienyl ligand in an Á − 5 ␲-bonding fashion that is mostly ionic in nature.

ligands (e.g. 50 for L = NHC) [62]. In all cases the electron pair on the Be atom occupied a p-orbital in an sp2 hybridized Be (Fig. 1). Substantial ␲-backbonding to the ligands was noted in all cases, which results in relatively large bond dissociation energies of 320 (L = CO), 180 (L = Me3 P) and 426 kJ/mol (L = H2 NHC), respectively. Parameswaran and co-workers indicated that these hypothetical compounds are unique in that they are second row centred species that are isostructural with a borane (three-coordinate planar), but isoelectronic with an amine (three-coordinate bearing a lone pair of electrons).

Our group considered ligand-stabilized diberyllium as a species of potential interest. Diberyllium itself is a molecule which has attracted long-term sustained interest in its bonding. Despite what is taught to undergraduate students in introductions to bonding, recent studies have shown that the dimer does indeed form transiently with a Be–Be bond distance of 2.5 A˚ and an association energy of about 10 kJ/mol [63]. The lowest unoccupied ␴ symmetric orbital for Be2 is bonding with respect to the Be–Be bond therefore it was hypothesized that occupation of this orbital with ligands would strengthen the Be–Be bond. This has previously been demonstrated theoretically using CO as a ligand, but dissociation of CO from Be2 was predicted to be relatively energetically neutral [64]. Adding two NHCs to the Be2 (51) resulted in a linear geometry reminiscent of the C2 and B2 analogues [55,65–67], with a Be–Be bond distance of 1.95 A˚ and a Be–Be WBI of 1.3 [57]. Some Be–Be ␲ bonding is evident in the HOMO of the molecule (Fig. 1). The dissociation energy of the species into Be2 and two NHCs was found to be high at 250 kJ/mol. The dissociation of the analogous Mg2 –NHC complex (52) is predicted to be only 13 kJ/mol. While 51 is likely to be a relatively inaccessible complex, it does illustrate a situation where a Be compound is possible where the Mg analogue is clearly not viable, highlighting the possible opportunities that arise in studying Be chemistry. The interaction of a single beryllium atom with the parent carbene:CH2 has been investigated recently by Schaefer using theoretical methods [68]. Mg–CH2 has been observed experimentally, while Be–CH2 has not. The minimum energy species for Be–CH2 was found to be a triplet, with singly-occupied MOs (SOMOs) based on the Be-–C bond and as a non-bonding orbital centred on the Be. The isomer HBe–CH, which has been observed, was calculated to be slightly higher in energy (although a minimum on the potential energy surface) had a significantly stronger Be-C bond, which may explain the experimental observation of the higher-energy isomer.

4. Summary and outlook 3.2. Be(0) compounds Zero-oxidation state Be compounds bound by various ligands have been considered computationally for both Be1 and Be2 . Parameswaran and co-workers have examined the structure and bonding of tris-ligated Be(0) species with CO, phosphine and NHC

It is clear that chemistry of beryllium diverges significantly from that of the other alkaline earth elements, with potential transformations and compounds that are exclusive to Be. As outlined in this review, it has been demonstrated that coordination compounds of Be can be easily formed, and once formed, manipulated using

Please cite this article in press as: K.J. Iversen, et al., Coord. Chem. Rev. (2014), http://dx.doi.org/10.1016/j.ccr.2014.11.008

G Model CCR-111961; No. of Pages 9

ARTICLE IN PRESS K.J. Iversen et al. / Coordination Chemistry Reviews xxx (2014) xxx–xxx

9

Fig. 1. Highest occupied molecular orbital (HOMO) of compounds 50 (Be(NHC)3 ) and 51 (Be2 (NHC)2 ).

reactions familiar to inorganic/organometallic chemists to generate new Be compounds. Experimental beryllium chemistry may well be experiencing a renaissance, which is further encouraged by a new era whereby academic laboratories are heading towards industrial level standards. The increased general laboratory standards may allow for more academic labs to pursue Be chemistry in a responsible manner, carefully respecting and managing the risks involved. However, it is also clear that theoretical studies will continue to play an important role in exploring the chemistry of Be due to the practical issues facing chemists interested in synthetic Be chemistry. Acknowledgements We thank La Trobe University and the La Trobe Institute for Molecular Science for their funding of our work in this area. J. Dutton thanks the ARC for a Discovery Early Career Researcher Award (DE130100186). References [1] B.L. Scott, T.M. McCleskey, A. Chaudhary, E. Hong-Geller, S. Gnanakaran, Chem. Commun. (2008) 2837–2847. [2] D. Himmel, I. Krossing, Z. Anorg. Allge. Chem. 632 (2006) 2021–2023. [3] K. Dehnicke, B. Neumüller, Z. Anorg. Allge. Chem. 634 (2008) 2703–2728. [4] T.P. Hanusa, E.J. Bierschenk, L.K. Engerer, K.A. Martin, N.R. Rightmire, Comp. Inorg. Chem. II 1 (2013) 1133–1187. [5] R. Fernández, E. Carmona, Eur. J. Inorg. Chem. 2005 (2005) 3197–3206. [6] W. Petz, G. Frenking, Top. Organomet. Chem. 30 (2010) 49–92. [7] R. Tonner, F. Öxler, B. Neumüller, W. Petz, G. Frenking, Angew. Chem. Int. Ed. 45 (2006) 8038–8042. [8] W. Petz, K. Dehnicke, N. Holzmann, G. Frenking, B. Neumüller, Z. Anorg. Allge. Chem. 637 (2011) 1702–1710. [9] M.A. Celik, G. Frenking, B. Neumüller, W. Petz, ChemPlusChem 78 (2013) 1024–1032. [10] W.A. Herrmann, O. Runte, G. Artus, J. Organomet. Chem. 501 (1995) C1–C4. [11] R.J. Gilliard, M.Y. Abraham, Y. Wang, P. Wei, Y. Xie, B. Quillian, H.F. Schaefer III, P.v.R. Schleyer, G.H. Robinson, J. Am. Chem. Soc. 134 (2012) 9953–9955. [12] J. Gottfriedsen, S. Blaurock, Organometallics 25 (2006) 3784–3786. [13] M. Arrowsmith, M.S. Hill, G. Kociok-Köhn, D.J. MacDougall, M.F. Mahon, Angew. Chem. Int. Ed. 51 (2012) 2098–2100. [14] K.J. Iversen, D.J.D. Wilson, J.L. Dutton, Organometallics 32 (2013) 6209–6217. [15] K.J. Iversen, D.J.D. Wilson, J.L. Dutton, Dalton Trans. 43 (2014) 12820–12823. [16] D. Schmidt, J.H.J. Berthel, S. Pietsch, U. Radius, Angew. Chem. Int. Ed. 51 (2012) 8881–8885. [17] T. Wang, D.W. Stephan, Chem. Eur. J. 20 (2014) 3036–3039. [18] S.M.I. Al-Rafia, R. McDonald, M.J. Ferguson, E. Rivard, Chem. Eur. J. 18 (2012) 13810–13820. [19] S. Inoue, D. Frankz, Chem. Asian J. 9 (2014) 2083–2087. [20] S.K. Bose, K. Fucke, L. Liu, P.G. Steel, T.B. Marder, Angew. Chem. Int. Ed. 53 (2014) 1799–1803. [21] M.R. Momeni, E. Rivard, A. Brown, Organometallics 32 (2013) 6201–6208. [22] K.J. Iversen, D.J.D. Wilson, J.L. Dutton, Dalton Trans. 42 (2013) 11035–11038. [23] R. Fang, L. Yang, Q. Wang, Organometallics 33 (2014) 53–60. [24] M.-D. Su, Inorg. Chem. 53 (2014) 5080–5087. [25] S.C. Chmely, T.P. Hanusa, W.W. Brennessel, Angew. Chem. Int. Ed. 49 (2010) 5870–5874.

[26] S. Morales-Meza, M.E. Sanchez-Castro, M. Sanchez, J. Mol. Model 19 (2013) 5153–5158. [27] B. Neumüller, K. Dehnicke, Z. Anorg. Allge. Chem. 636 (2010) 515–517. [28] B. Neumüller, K. Dehnicke, Z. Anorg. Allge. Chem. 636 (2010) 1516–1521. [29] B. Neumüller, K. Dehnicke, Z. Anorg. Allge. Chem. 636 (2010) 1767–1771. [30] B. Neumüller, K. Dehnicke, Z. Anorg. Allge. Chem. 636 (2010) 1438–1440. [31] F. Kraus, S.A. Baer, M.R. Buchner, A.J. Karttunen, Chem. Eur. J. 18 (2012) 2131–2142. [32] S.P. Green, C. Jones, A. Stasch, Science 318 (2007) 1754–1757. [33] M. Arrowsmith, M.S. Hill, G. Kociok-Köhn, D.J. MacDougall, M.F. Mahon, I. Mallov, Inorg. Chem. 51 (2012) 13408–13418. [34] S.J. Bonyhady, C. Jones, S. Nembenna, A. Stasch, A.J. Edwards, G.J. McIntyre, Chem. Eur. J. 16 (2010) 938–955. [35] M. Arrowsmith, M.R. Crimmin, M.S. Hill, G. Kociok-Köhn, Dalton Trans. 42 (2013) 9720–9726. [36] D. Naglav, D. Bläser, C. Wölper, S. Schulz, Inorg. Chem. 53 (2014) 1241–1249. [37] R. Kubiak, A. Waskowska, M. Sledz, A. Jezierski, Inorg. Chim. Acta 359 (2006) 1344–1350. [38] J. Janczak, R. Kubiak, J. Mol. Struc. 935 (2009) 53–59. [39] R. Kubiak, J. Janczak, M. Sledz, E. Bukowska, Polyhedron 27 (2008) 3044–3052. [40] J. Janczak, R. Kubiak, J. Mol. Struc. 1030 (2012) 149–156. [41] R. Corbo, T.P. Pell, B.D. Stringer, C.F. Hogan, P.J. Barnard, D.J.D. Wilson, J.L. Dutton, J. Am. Chem. Soc. 136 (2014) 12415–12421. [42] T. Peng, K. Ye, Y. Liu, L. Wang, Y. Wu, Y. Wang, Org. Electron. 12 (2011) 1914–1919. [43] C.S. Oh, C.W. Lee, J.Y. Lee, Chem. Commun. 49 (2013) 3875–3877. [44] H.V.K. Diyabalanage, K. Ganguly, D.S. Ehler, G.E. Collis, B.L. Scott, A. Chaudhary, A.K. Burrell, T.M. McCleskey, Angew. Chem. Int. Ed. 47 (2008) 7332–7334. [45] G. Frenking, N. Holzmann, B. Neumüller, K. Dehnicke, Z. Anorg. Allge. Chem. 636 (2010) 1772–1775. [46] H. Braunschweig, K. Gruß, Z. Naturforsch. 66b (2011) 55–57. [47] H. Braunschweig, K. Gruss, K. Radacki, Angew. Chem. Int. Ed. 48 (2009) 4239–4241. [48] P. Parameswaran, G. Frenking, J. Phys. Chem. A. 114 (2010) 8529–8535. [49] A. Stasch, C. Jones, Dalton Trans. 40 (2011) 5659–5672. [50] R. West, Science 225 (1984) 1110–1114. [51] E. Rivard, P.P. Power, Inorg. Chem. 46 (2007) 10047–10064. [52] Y. Wang, G.H. Robinson, Dalton Trans. 41 (2012) 337–345. [53] D.J.D. Wilson, J.L. Dutton, Chem. Eur. J. 19 (2013) 13626–13637. [54] Y. Wang, G.H. Robinson, Inorg. Chem. 50 (2011) 12326–12337. [55] H. Braunschweig, R.D. Dewhurst, K. Hammond, J. Mies, K. Radacki, A. Vargas, Science 336 (2012) 1420–1422. [56] S. Krieck, L. Yu, M. Reiher, M. Westerhausen, Eur. J. Inorg. Chem. 2010 (2010) 197–216. [57] S.A. Couchman, N. Holzmann, G. Frenking, D.J.D. Wilson, J.L. Dutton, Dalton Trans. 42 (2013) 11375–11384. [58] S. Li, X. Yang, Y. Liu, Y. Zhao, Q. Li, Y. Xie, H.F. Schaefer Iii, B. Wu, Organometallics 30 (2011) 3113–3118. [59] Y. Kan, J. Mol. Struc. THEOCHEM 894 (2009) 88–92. [60] I. Resa, E. Carmona, E. Gutiérrez-Puebla, Á. Monge, Science 305 (2004) 1136–1138. [61] J.-H. Sheu, M.-D. Su, J. Organomet. Chem. 696 (2011) 1221–1227. [62] P. Parameswaran, S. De, Dalton Trans. 42 (2013) 4650–4656. [63] J.M. Merritt, V.E. Bondybey, M.C. Heaven, Science 324 (2009) 1548–1551. [64] K.K. Sunil, J. Am. Chem. Soc. 114 (1992) 3985–3986. [65] J.L. Dutton, D.J.D. Wilson, Angew. Chem. Int. Ed. 51 (2012) 1477–1480. [66] L. Jin, M. Melaimi, L. Liu, G. Bertrand, Org. Chem. Front. 1 (2014) 351–354. [67] Y. Li, K.C. Mondal, P.P. Samuel, H. Zhu, C.M. Orben, S. Panneerselvam, B. Dittrich, B. Schwederski, W. Kaim, T. Mondal, D. Koley, H.W. Roesky, Angew. Chem. Int. Ed. 53 (2014) 4168–4172. [68] Y. Qiu, A.Y. Sokolov, Y. Yamaguchi, H.F. Schaefer, J. Phys. Chem. A 117 (2013) 9266–9273.

Please cite this article in press as: K.J. Iversen, et al., Coord. Chem. Rev. (2014), http://dx.doi.org/10.1016/j.ccr.2014.11.008