Bonding in Barium Boryloxides, Siloxides, Phenoxides and Silazides: A Comparison with the Lighter Alkaline Earths

Article abstract of DOI:10.1002/chem.202101687

Barium complexes ligated by bulky boryloxides [OBR₂]^? (where R=CH(SiMe₃)₂, 2,4,6-ⁱPr₃-C₆H₂ or 2,4,6-(CF₃)₃-C₆H₂), siloxide

Full text of DOI:10.1002/chem.202101687

Accepted Article
Accepted Article
Title: Bonding in Barium Boryloxides, Siloxides, Phenoxides and
Silazides. A comparison with the lighter alkaline earths
Authors: Erwann Le Coz, Joanna Hammoud, Thierry Roisnel, Marie
Dallon, Vincent Dorcet, Samia Kahlal, Jean-Francois
Carpentier, Jean-Yves Saillard, and Yann Sarazin
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.202101687
Link to VoR: https://doi.org/10.1002/chem.202101687
01/2020

10.1002/chem.202101687
Chemistry - A European Journal
Bonding in Barium Boryloxides, Siloxides, Phenoxides and Silazides.
A comparison with the lighter alkaline earths.
Erwann Le Coz, Joanna Hammoud, Thierry Roisnel, Marie Cordier, Vincent Dorcet, Samia Kahlal,
Jean-François Carpentier, Jean-Yves Saillard,* and Yann Sarazin*
Univ Rennes, CNRS, ISCR (Institut des Sciences Chimiques de Rennes) – UMR 6226, F-35000 Rennes,
France
Corresponding authors: yann.sarazin@univ-rennes1.fr
jean-yves.saillard@univ-rennes1.fr
ORCID
JF. Carpentier 0000-0002-9160-7662
M. Cordier
T. Roisnel
JY. Saillard
Y. Sarazin
0000-0003-0917-2725
0000-0002-6088-4472
0000-0003-4469-7922
0000-0003-1121-0292
† Electronic supplementary information (ESI) available: full experimental details, including NMR
spectra; crystallographic data for CCDC 2032704-2032712, 2032967, 2033281, 2033287 and 2033317;
details of DFT computations.
1
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
Keywords
Barium  alkaline-earth metals  O-based ligands  DFT  Bonding analysis
Abstract
Barium complexes ligated by bulky boryloxides [OBR2]– (where R = CH(SiMe3)2, 2,4,6-iPr3-C6H2 or
2,4,6-(CF3)3-C6H2), siloxide [OSi(SiMe3)3]–, and/or phenoxide [O-2,6-Ph2-C6H3]–, have been prepared.
A diversity of coordination patterns is observed in the solid state for both homoleptic and heteroleptic
complexes, with coordination numbers ranging between 2 and 4. The identity of the bridging ligand in
heteroleptic dimers [Ba(μ2-X1)(X2)] depends largely on the given pair of ligands X1 and X2.
Experimentally, the propensity to fill the bridging position increases according to [OB{CH(SiMe3)2}2)]–
< [N(SiMe3)2]– < [OSi(SiMe3)3]– < [O(2,6-Ph2-C6H3)]– < [OB(2,4,6-iPr3-C6H2)2]–. This trend is the
overall expression of 3 properties: steric constraints, electronic density and σ- and π-donating capability
of the negatively charged atom, and ability to generate Ba∙∙∙F, Ba∙∙∙C(π) or Ba∙∙∙H-C secondary
interactions. The comparison of the structural motifs in the complexes [Ae{μ2-
N(SiMe3)2}(OB{CH(SiMe3)2}2)]2 (Ae = Mg, Ca, Sr and Ba) suggest that these observations may be
extended to all alkaline earths. DFT calculations highlight the largely prevailing ionic character of
ligand-Ae bonding in all compounds. The ionic character of the Ae-ligand bond encourages bridging
coordination, whereas the number of bridging ligands is controlled by steric factors. DFT computations
also indicate that in [Ba(μ2-X1)(X2)] heteroleptic dimers, ligand predilection for bridging vs terminal
positions is dictated by the ability to establish secondary interactions between the metals and the ligands.
2
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
Introduction
The heavy alkaline earths calcium, strontium and barium (Ae = Ca, Sr, Ba) are gaining increasing
recognition as suitable alternatives in molecular catalysis to more expensive and often more toxic
transition metals.[1-4] Versatile catalysts built on these three metals have shown for example excellent
performances in a panel of reactions such as the polymerisations of styrene[5] and cyclic esters,[6] or the
hydrogenation, hydroamination and hydrophosphination of unsaturated bonds.[7-18] Recently, they have
also proved competent in catalysed dehydrocouplings for the selective formation of amidoboranes,[19-21]
silazanes,[22-26] borasiloxanes[27-28] and asymmetric disiloxanes.[29] Mechanistic and kinetic investigations
have revealed that for a same ligand environment, and all parameters being otherwise equal, kinetics in
these catalysed reactions usually increase according to Ca < Sr < Ba.[1-4] A notable exception is the
cyclohydroamination of aminoalkenes, where the opposite trend has been consistently observed.[7,30] On
the whole, barium hence displays the highest turnover numbers. However, it also is with this metal that
the high reactivity specific of the heavy alkaline earths is most difficult to tame.[31]
Heavy alkaline-earth metals form d0 complexes where bonding is ruled by electrostatic and steric
factors, with little, or no, orbital contribution. In solution, heteroleptic complexes [{Lig}AeX], where
{Lig}‒ is a monoanionic ancillary ligand and X‒ is often an amide, an alkyl or a hydride, often engage
in ligand redistribution equilibria that generate ill-controlled mixtures unsuitable for catalysis. Stopping
these phenomena, which are most troublesome for the largest and most electropositive element barium,
still constitutes a challenge in modern alkaline-earth molecular chemistry. Yet, thanks to the use of
carefully devised ligands, e.g. β-diketiminates,[12,32] iminoanilides,[12] carbazolates,[33-34] phenoxides,[6,35-
36]
or tris(pyrazolyl)borates,[37] the number of stable, heteroleptic barium complexes has been growing
steadily in recent years. Beyond homogenous catalysis, these ligands have led to breakthroughs such as
the recent preparation of soluble barium-hydrides.[31,37-38]
Despite the overall importance of alkoxides as inorganic ligands,[40] well-defined Ba-alkoxide
remain scarce due to the large size, propensity for high coordination numbers, and high oxophilicity and
electropositivity of the Ba2+ dication. Conventional alcohols, e.g. tBuOH, yield polymeric and insoluble
Ba-alkoxides.[35]
Rare
examples
of
soluble
compounds
include
[Ba{κ4-
N(CH2CH2O)(CH2CH2OH)2}2],[41] [Ba2(μ2-OCPh3)3(OCPh3).(thf)3][42] and [Ba{κ3-O(CH2CH2O)2}{κ3-
O(CH2CH2OH)2}2].[43] The addition of CF3 groups in α position to the alkoxide give fluoroalkoxides
with enhanced stability and solubility, as in [Ba{OC(CF3)3}2.(dme)3].[44]
Oxygen-based derivatives of barium alkoxides exist. A number of Ba-phenoxides have been
described since the report on [Ba(O-2,6-tBu2-4-Me-C6H2)(thf)3],[45] e.g. the dimer [Ba(μ-I)(O-2,6-tBu2-
[36]
4-Me-C6H2).(thf)3]2 or the dinuclear [Ba2(μ-O-2,6-Ph2-C6H3)3(O-2,6-Ph2-C6H3)].[46] Siloxides have
produced base-stabilised Ba species, such as Caulton’s seminal [Ba2(μ2-OSitBu3)3(OSitBu3).thf],[42] and
[Ba(OSiPh3)2(15-crown-5).thf][47] or [Ba3(OSiPh2OSiPh2O)3.(tetraglyme)2].[48] We recently used the
bulky
silanol
(Me3Si)3SiOH
to
generate
3
the
unsolvated
Ba-siloxide
[Ba2{μ2-
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
OSi(SiMe3)3}3{OSi(SiMe3)3}] (12, Figure 1), and its heteroleptic congener [Ba{μ2-
OSi(SiMe3)3}{N(SiMe3)2}]2 (22); both compounds are dimeric in the solid state.[29] Besides, we have
also reported the preparation and structural characterisation of the first barium boryloxides, notably the
two-coordinate monomer [Ba(OB{CH(SiMe3)2}2)2] (3) and its amido, dimeric parent [Ba{μ2-
N(SiMe3)2}(OB{CH(SiMe3)2}2)]2 (42, Figure 1).[27] We had reasoned that the presence of boron or
silicon atoms would lower the π-donating ability of the O atoms in these ligands, and that, combined
with the steric bulk imparted by multiple SiMe3 groups, they would generate soluble, low-coordinate
barium complexes.
Figure 1. Solvent-free, low-coordinate barium siloxides and boryloxides;[27,29] R = SiMe3.
Although we succeeded in preparing low-coordinate complexes that proved valuable
dehydrocoupling precatalysts,[27-29] our curiosity was aroused by the variety of coordination patterns in
complexes 12-42. The homoleptic boryloxide 3 is monomeric, while 12 is a dimer with three bridging
siloxides and a single terminal one. In the heteroleptic siloxide/amide dimer 22, the siloxides are bridging
the two metals and the amido moieties occupies terminal position, as expected with alkaline earths and
other oxophilic elements. On the other hand, the opposite situation is (uncharacteristically) observed for
the boryloxide/amide 42.
As part of our continuing efforts to devise low-coordinate and reactive barium complexes suited
to molecular catalysis, we are reporting here on several new homoleptic barium boryloxides. In addition,
a number of heteroleptic barium complexes that combine bulky boryloxo, siloxo, phenoxo or amido
ligands are also presented. The structural motifs in these compounds are analysed on the basis of X-ray
diffraction crystallography, NMR spectroscopy and DFT computations, with the aim of providing an
appraisal of the criteria that enable the formation of potentially useful barium precatalysts. Besides, the
structural comparison in a homologous series of heteroleptic Ae-boryloxide/amide complexes is also
provided. It highlights the differences in the bonding patterns involving magnesium, calcium, strontium
or barium. In this context, we note that no strontium boryloxide has been reported to date, whereas the
dimer [Ca{μ-OB(CH(SiMe3)2)2}{N(C6F5)2}]2 is the sole known calcium boryloxide.[49] A handful of
magnesium congeners have been prepared before, notably the dimeric [Mg{OB(mes)2}R.thf]2 (R = Me,
Bu; mes = mesityl = 2,4,6-Me3-C6H2)[50] and the β-diketiminato [{BDIDiPP}MgOBPh2.(L)], where L is
either a neutral organoboronic ester or N,N-dimethylaminopyridine.[51]
4
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
Results and Discussion
Synthesis of homoleptic barium boryloxides
Prior to this work, existing barium boryloxides were limited to 3, 42 and to the solvated 3∙toluene and
3∙(thf)2.[27,29,52] We set out to prepare new complexes using the borinic acids {(Me3Si)2CH}2BOH,[53]
(mes)2BOH,[54-55] (Tripp)2BOH,[54-55] {Fmes}2BOH[56] and (C6F5)2BOH[57-59] (Tripp = 2,4,6-iPr3-C6H2,
Fmes = 2,4,6-(CF3)3-C6H2, Scheme 1). These known compounds were prepared on multigram scales
and high yields upon modification of published protocols. All are crystalline colourless solids with good
solubility in organic solvents. Their identity was confirmed by NMR spectroscopy and X-ray diffraction
crystallography (see the Supporting Information, Figures S49-S53). They are monomeric in the solid
state, except for (C6F5)2BOH which forms a cyclic trimer with a non-planar B3O3 core; yet, it is
monomeric in solution in toluene-d8 at room temperature.[60] Their respective chemical shifts in 11B
NMR (δ11 53.8, 50.8, 52.2, 43.5 and 40.5 ppm in benzene-d6) are in the range expected for trivalent
B
boron atoms. The borinic acids with lower electronic density at boron, {Fmes}2BOH and (C6F5)2BOH,
are the most shielded.
Scheme 1. Optimised syntheses of borinic acids.[53-59]
5
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
Following the synthesis of [Ba(OB{CH(SiMe3)2}2)2] (3),[27] a monomer stabilised by intra- and
intermolecular Ba∙∙∙H-C interactions, we were interested in obtaining other low-coordinate Ba-
boryloxides to assess the role of secondary interactions towards the stabilisation of electrophilic barium
complexes. Our initial attempts started off with (mes)2BOH, but it failed to react with [Ba{N(SiMe3)2}2]2
or [Ba{N(SiMe3)2}2.(thf)2]. Beyond the fact that this borinic acid is prone to condensation with loss of
H2O and concomitant formation of (mes)2B-O-B(mes)2, the several reactions that were carried out only
afforded the oxocluster [Ba4(O){OB(mes)2}6] as the sole isolable product, probably due to the release
of water in the reaction medium. This cluster, a colourless crystalline material, was authenticated by
XRD analysis (Figure S54). The reactivity of (mes)2BOH observed here is different from that with more
robust magnesium precursors, although the reported synthesis of [Mg{OB(mes)2}Me.(thf)]2 was said to
be erratic and to give low yields.[50] We were also unable to obtain Ba-boryloxides from the electron-
poor (C6F5)2BOH. Regardless of the conditions, all attempted reactions generated intractable mixtures
of compounds which contained various borates, as indicated by 11B NMR spectroscopy (multiple
resonances in the range ‒5 to ‒20 ppm). On the other hand, treatment of [Ba{N(SiMe3)2}2.(thf)2] in
toluene with two equivalents of (Tripp)2BOH returned the thf-free dimer [Ba{μ2-
OB(Tripp)2}{OB(Tripp)2}]2 in 74% yield (52, Scheme 2). The identical reaction with {Fmes}2BOH
produced the thf-adduct [Ba{μ2-OB(Fmes)2}{OB(Fmes)2}.thf]2 (62.(thf)2), whereas the unsolvated
[Ba{μ2-OB(Fmes)2}{OB(Fmes)2}]2 (62) was isolated near quantitatively following the stoichiometric
reaction of {Fmes}2BOH with thf-free [Ba{N(SiMe3)2}2]2. Both these complexes are sparingly soluble
in aromatic and aliphatic hydrocarbons.
The colourless complexes 52, 62 and 62.(thf)2 were characterised by NMR spectroscopy and XRD-
diffraction analysis performed on single crystals. The NMR data for 52 confirm the absence of thf in the
complex. The 1H NMR spectrum recorded in benzene-d6 shows broad, ill-defined resonances that testify
to dynamic phenomena in solution in this solvent, assumed to result from the rotation of iPr groups or
coordination-dissociation of aromatic molecules, rather than a monomer-dimer equilibrium for which
no evidence could be found. The 1H NMR spectrum recorded in thf-d8 only shows sharp resonances and
confirms the identity and purity of 52. The 11B spectrum in thf-d8 features a single resonance at 47.1
ppm. Due to their limited solubility, the NMR data for 62 and 62.(thf)2 were both recorded in thf-d8, and
the two sets of NMR resonances in this solvent are hence unsurprisingly identical. The 19F NMR spectra
only exhibit two sharp singlets at δ19F ‒56.58 and ‒63.98 ppm (ortho- and para-CF3 groups,
respectively). A unique resonance appears at δ11B 34.1 ppm in the 11B NMR spectra. There is no
indication that the dimeric structure of these three complexes is maintained in thf-d8, and we could not
detect any evidence for the presence of Ba∙∙∙F or Ba∙∙∙C(π) interactions (see below).
6
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
Scheme 2. Synthesis of homoleptic barium boryloxides.
The molecular solid-state structure of 52 is depicted in Figure 2. The complex is a centro-
symmetric O-bridged dimer with a rhomboidal Ba2O2 core. Each barium is three-coordinate in a
distorted trigonal pyramidal geometry (Σθ(Ba1) = 320.03(7)°). This pyramidalisation results from the
presence of Ba∙∙∙C(π) η3-interactions, with Ba-C interatomic distances in the range 3.0854(15)-
3.3157(16) Å to three flanking C-atoms in a Tripp moiety carried by the bridging ligand. Only one
aromatic group engages in the η3-interaction with each barium centre. Correspondingly, the Ba1-O1i-
B11 angle for the B and B-atom bearing the interacting Tripp group (105.40(9)°) is much narrower than
the two other Ba-O-B angles (143.11(10)° and 148.28(11)°), while the angle O1i-B1i-C2i is also
substantially lower than O1i-B1i-C17i (112.54(13)° and 122.23(13)°). The C2i-C3i and C2i-C7i
interatomic distances (1.423(2) and 1.414(2) Å) for the C2i, C3i and C7i involved in the π-coordination
with Ba1 are a little stretched compared to the other C=C bonds in the aromatic fragment (1.397(2)-
1.402(2) Å). Both boron atoms are in ideal trigonal planar environments.
The molecular structure of 62 displayed in Figure 3 shows this complex to also exist as an O-
bridged dimer with three-coordinate barium atoms. The complex is stabilised by a pattern of secondary
Ba∙∙∙F interactions, with corresponding interatomic distances in the range 2.932(2)-3.284(3) Å well
below the sum of Van der Waals radii for Ba and F (2.67 and 1.47 Å). Each Ba engages in up to seven
intramolecular contacts within a dimeric motif, as well as an intermolecular interaction with a
neighbouring molecule (Ba1-F40iii = 2.957(3) Å). The presence of the intermolecular contacts leads to
7
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
the formation of one-dimensional coordination polymers in the solid state, although there is no
spectroscopic indication that this arrangement is maintained in solution.
Figure 2. ORTEP representation of the molecular structure of [Ba{μ2-OB(Tripp)2}{OB(Tripp)2}]2 (52). Non-
interacting toluene molecule, H atoms and Me groups not represented for the sake of clarity. Ellipsoids at the 50%
probability level. Ba∙∙∙C(π) interactions depicted with dashed lines. Selected interatomic distances (Å) and angles
(°): Ba1-O1 = 2.5694(11), Ba1-O1i = 2.8100(11), Ba1-O2 = 2.4167(11), Ba1-C2i = 3.0854(15), Ba1-C3i =
3.1994(15), Ba1-C7i = 3.3157(16); O1-Ba1-O1i = 81.19(4), O2-Ba1-O1 = 123.86(4), O2-Ba1-O1i = 114.98(4),
Ba1-O1-B1 = 143.11(10), Ba1-O1i-B1i = 105.40(9), Ba1-O1-Ba1i = 98.82(4), Ba1-O2-B51 = 148.28(11), O1i-
B1i-C2i = 112.54(13), O1i-B1i-C17i = 122.23(13). Symmetry transformations used to generate equivalent atoms:
i, -x, -y, -z, T = [1, 0, 0].
The structure of the dimeric solvate 62.(thf)2 shares some common features with that of 62, namely
the presence of multiple intramolecular Ba∙∙∙F contacts (Figure S46). However, due to the coordination
of thf molecules, each metal is now four-coordinate, and does not feature intermolecular Ba∙∙∙F
interaction; as a result, in contrast with 62, 62.(thf)2 shares does not form a coordination polymer. As
suggested by Ruhlandt-Senge over a decade ago,[61] Ae∙∙∙F-C secondary interactions such as in 62 and
62.(thf)2 constitute an efficient tool (along other types of secondary interactions, e.g. anagostic Ae∙∙∙H
or, as above in 52, Ae∙∙∙C(π)) towards the stabilisation of low-coordinate alkaline-earth complexes. It
has been used effectively to obtain a variety of highly electrophilic Ae species in recent years.[6,49,62,63]
8
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
Figure 3. ORTEP representation of the molecular structure of [Ba{μ2-OB(Fmes)2}{OB(Fmes)2}]2 (62). H atoms
omitted for clarity. Ellipsoids at the 30% probability level. Selected interatomic distances (Å) and angles (°): Ba1-
O1 = 2.435(3), Ba1-O51 = 2.623(3), Ba1-O51i = 2.657(3), Ba1-F19 = 3.218(4), Ba1-F40iii = 2.957(3), Ba1-F47 =
3.104(3) , Ba1-F68 = 3.284(3), Ba1-F76i = 3.164(3), Ba1-F88i = 2.933(2), Ba1-F90 = 2.939(3); Ba1-O1-B2 =
171.9(3), Ba1-O51-Ba1i = 104.45(9). Symmetry transformations used to generate equivalent atoms: i, ‒x, ‒y, ‒z,
T = [1, 0, 0]; ii, -x, -y, -z, T = [0, 0, 0].
Synthesis of heteroleptic barium complexes
The synthesis of barium complexes was extended to heteroleptic boryloxides and other related species,
namely phenoxides and siloxides (Scheme 3), to complement the first examples of these compounds
recently introduced: the amido-bridged dimer [Ba{μ2-N(SiMe3)2}(OB{CH(SiMe3)2}2)]2 (42),[27] and the
dinuclear [Ba{μ2-OSi(SiMe3)3}{N(SiMe3)2}]2 (22) bridged by a siloxide.[29] The ambition was to assess
which of the bulky boryloxides, the siloxide (Me3Si)SiO‒ and the regular hexamethyldisilazide
(Me3Si)2N‒ would provide suitable electronic density and/or steric shielding to generate solvent-free
barium and perhaps, by extrapolation, other alkaline-earth complexes.
All reactions described here were carried out in toluene at room temperature. The boryloxide-
bridged [Ba{μ2-OB(Tripp)2}{N(SiMe3)2}]2 (72) was isolated in 67% yield following the equimolar
reaction of [Ba{N(SiMe3)2}2.(thf)2] with (Tripp)2BOH. Treatment of 22 with two equivalents of
{(Me3Si)2CH}2BOH afforded the quantitative and selective formation of the siloxo-bridged [Ba{μ2-
OSi(SiMe3)3}(OB{CH(SiMe3)2}2)]2 (82) upon release of HN(SiMe3)2. The analogous reaction between
22 and the aromatic borinic acid (Tripp)2BOH yielded the dimer [Ba{μ2-OB(Tripp)2}{OSi(SiMe3)3}]2
(92), where the two metals are bridged by the boryloxides. Finally, the mixed siloxide/phenoxide [Ba{μ2-
O(2,6-Ph2-C6H3)}{OSi(SiMe3)3}]2 (102) was obtained in 70% yield upon protonolysis of 22 with two
9
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
equivalents of 2,6-Ph2-C6H3-OH. On the other hand, our attempts to synthesise heteroleptic
boryloxide/phenoxide species starting from 42 or 72 were unsuccessful. All four heteroleptic complexes
72-102 were obtained as colourless solids, and were recrystallised as single crystals suitable for XRD
analysis. Compound 82 is very soluble in all organic solvents, but 72 and especially 92 and 102 have
limited solubility in hydrocarbons, including aromatic ones.
Scheme 3. Synthesis of boryloxo/siloxo/phenoxo/amido heteroleptic barium complexes 72-102.
10
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
The NMR spectroscopic data recorded for crystalline materials in benzene-d6 (for 72 and 82) and
thf-d8 (for 92 and 102) are unremarkable. Most 11B and 29Si resonances could be located, except in two
cases, for 92 and 102, most probably due to high fluxionality of the systems in solution; low temperature
NMR provided no assistance. A summary of representative NMR data for these complexes, as well as
those for the homoleptic complexes described herein and for complexes 12-42, is collated in Table 1.
Table 1. Multinuclear NMR data for complexes 12-102. Data for 12-42 are taken from references [27-29].
Complex
Solvent
19F NMR
(ppm)
-
11B NMR
29Si NMR
(ppm)
(ppm)
-
[Ba2{μ2-OSi(SiMe3)3}3{OSi(SiMe3)3}]
[Ba{μ2-OSi(SiMe3)3}{N(SiMe3)2}]2
12
22
benzene-d6
benzene-d6
‒19.37
‒27.02
‒18.60
‒19.22
‒19.27
‒24.69
‒27.74
‒3.44
-
-
[Ba(OB{CH(SiMe3)2}2)2]
3
benzene-d6
-
45.8
[Ba{μ2-N(SiMe3)2}(OB{CH(SiMe3)2}2)]2
[Ba{μ2-OB(Tripp)2}{OB(Tripp)2}]2
[Ba{μ2-OB(Fmes)2}{OB(Fmes)2}]2
42
52
62
benzene-d6
thf-d8
-
43.7
47.1
34.1
‒3.47
-
-
-
thf-d8
‒56.58
‒63.98
[Ba{μ2-OB(Tripp)2}{N(SiMe3)2}]2
72
82
benzene-d6
benzene-d6
-
-
47.5
43.7
‒18.57
‒3.38
[Ba{μ2-OSi(SiMe3)3}(OB{CH(SiMe3)2}2)]2
‒19.86
‒27.84
[Ba{μ2-OB(Tripp)2}{OSi(SiMe3)3}]2
92
thf-d8
thf-d8
-
-
not detected
‒20.67
‒36.67
[Ba{μ2-O(2,6-Ph2-C6H3)}{OSi(SiMe3)3}]2
102
-
not detected
The molecular solid-state structure of [Ba{μ2-OB(Tripp)2}{N(SiMe3)2}]2 (72) shows the complex
to be a centro-symmetric boryloxide-bridged dimer (Figure 4). Each barium is three-coordinate, with a
11
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
distorted trigonal pyramidal environment (Σθ(Ba1) = 325.97(8)°) that reflects an η1-interaction with the
ipso C-atom (Ba1-C41 = 3.1380(18) Å; the other Ba1-C(π) interatomic distances are 3.497(2) Å or
greater) in a neighbouring Tripp fragment. The Ba-O distances in 72 (2.5932(13)-2.6789(13) Å) are in
the range of those in other related complexes having boryloxides in bridging positions, e.g. 52 and 62.
The O21-B22-C41 angle involving the carbon atom (C41) that interacts with Ba1 is somewhat narrower
that the O21-B22-C23 with the non-interacting C atom, 119.07(17)° and 125.04(18)°, respectively. It
must be noted that by contrast with 42 where the amides bridge the two metal atoms, in 72 they occupy
terminal positions. Hence, the nature of the boron-bound alkyl/aryl groups clearly affects the π-donating
character of the oxygen atom in different boryloxides.
Figure 4. ORTEP view of the molecular structure of [Ba{μ2-OB(Tripp)2}{N(SiMe3)2}]2 (72) . H atoms omitted
for clarity. Ellipsoids at the 30% probability level. Selected interatomic distances (Å) and angles (°): Ba1-N1 =
2.5512(17), Ba1-O21 = 2.6789(13), Ba1-O21i = 2.5932(13), Ba1-C41 = 3.1380(18), O21-B22 = 1.334(2), N1-Si1
= 1.6825(18), N1-Si2 = 1.6803(19); N1-Ba1-O21 = 140.52(5), N1-Ba1-O21i = 109.53(5), O21-Ba1-O21i =
75.92(4). B22-O21-Ba1 = 105.69(11), B22-O21-Ba1i = 149.24(12), Ba1-O21-Ba1i = 104.08(4), O21-B22-C23 =
125.04(18), O21-B22-C41 = 119.07(17). Symmetry transformations: i, -x, -y, -z, T = [2, 0, 0].
[Ba{μ2-OSi(SiMe3)3}(OB{CH(SiMe3)2}2)]2 (82) forms a Cs-symmetric dimer bridged by the two
siloxides (Figure 5), with a symmetry plane going through the barium and two terminal oxygen atoms.
The asymmetric unit contains three independent and similar molecules, only one of which is discussed
here. Unlike in for instance 52 and 72, the Ba2O2 core is not planar, as Ba2 protrudes out of the best plane
defined by Ba1, O21 and O51 presumably due to the steric constraints imposed by the very bulky
ligands. Each barium is three-coordinate, but the geometry about Ba1 and Ba2 is very different due to
12
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
the fact that one of the terminal boryloxides (at Ba2) is tilted towards the core of the dimer, while the
other (at Ba1) is pointing away from it. Hence, Ba1 rests in a trigonal planar geometry (Σθ(Ba1) =
357.54(19)°), while the geometry around Ba2 is pyramidalised (Σθ(Ba2) = 314.62(19)°). Overall, the
geometrical arrangement in 82 appears to be intermediate between those in 22 (a Cs-symmetric dimer
with 3 bridging siloxides and a single terminal one,[29] see Figure 1) and, for instance, the Ci-symmetric
52. The Ba-O interatomic distances are unremarkable and compare well with those in the barium
boryloxides and siloxides described previously here and elsewhere.[27-29] Complex 82 exhibits four short
Ba-H interatomic distances in the range 2.493(6)-2.777(7) Å (not depicted) that may indicate the
presence of Ba∙∙∙H-C anagostic interactions contributing to its stability; however, no spectroscopic
evidence for the existence of such interactions could be detected.
Figure 5. ORTEP view of the molecular structure of [Ba{μ2-OSi(SiMe3)3}(OB{CH(SiMe3)2}2)]2 (82). Only one
of the three independent but very similar molecules in the asymmetric unit depicted. H atoms omitted for clarity.
Ellipsoids at the 30% probability level. Selected interatomic distances (Å) and angles (°): Ba1-O1 = 2.431(3), Ba1-
O21 = 2.549(3), Ba1-O51 = 2.635(3), Ba2-O5 = 2.551(3), Ba2-O31 = 2.386(3), Ba2-O21 = 2.584(3), O1-B2 =
1.331(6), O31-B32 = 1.326(6); O1-Ba1-O21 = 124.71(11), O1-Ba1-O51 = 156.57(11), O21-Ba1-O51 = 76.26(11),
O21-Ba2-O31 = 120.37(11), O21-Ba2-O51 = 77.14(10), O31-Ba2-O51 = 117.11(12), B2-O1-Ba1 = 164.8(3),
B32-O31-Ba2 = 170.20(4), O1-B2-C3 = 120.5(5), O1-B2-C10 = 122.6(5), C3-B2-C10 = 116.8(4), O31-B32-C33
= 120.3(4), O31-B32-C40 = 120.3(4), C33-B32-C40 = 119.3(4).
[Ba{μ2-OB(Tripp)2}{OSi(SiMe3)3}]2 (92) forms a centro-symmetric dimer bridged by Oboryloxide
atoms, with a planar Ba2O2 rhomboidal core and terminal siloxides (Figure 6). The geometry around
13
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
each barium atom is trigonal pyramidal (Σθ(Ba1) = 314.9(6)°), with deviation from the expected trigonal
planar arrangement due to the presence of Ba···C(π) η3-interactions with flanking aromatic groups. The
corresponding Ba-C interatomic distances are in the range 3.074(10)-3.265(10) Å. Note that the Ba1-
O1-Si1 angle of 172.0(6)° is particularly wide, which suggests substantial sp-hybridisation at O1 due to
the participation of the lone pairs of electrons to the formation of the O1-Ba1 and O1-Si1 bonds. This
angle and the Ba1-O1 bond length (2.377(10) Å) match the pertaining ones in other dimeric complexes
having the bulky siloxide OSi(SiMe3)3 in terminal positions, e.g. as in 12 (2.414(7) Å and 173.50(5)°).
Figure 6. ORTEP view of the molecular structure of [Ba{μ2-OB(Tripp)2}{OSi(SiMe3)3}]2 (92). H atoms omitted
for clarity. Ellipsoids at the 30% probability level. Selected interatomic distances (Å) and angles (°): Ba1-O1 =
2.377(10), Ba1-O11 = 2.778(10), Ba1-O11i = 2.560(10), Ba1-C12 = 3.074(10), Ba1-C13 = 3.198(10), Ba1-C17 =
3.265(10); O1-Ba1-O11 = 130.0(3), O1-Ba1-O11i = 109.9(3), O11-Ba1-O11i = 75.0(4), Si1-O1-Ba1 = 172.0(6).
Symmetry transformations used to generate equivalent atoms: i, -x, -y, -z, T = [ 1, 0, 0].
The molecular structure of [Ba{μ2-O(2,6-Ph2-C6H3)}{OSi(SiMe3)3}]2 (102) depicted in Figure 7
shows it to be a centro-symmetric dimer bridged through the Ophenoxide atoms which, like 92, has two
siloxides in terminal positions. The main structural features of these two complexes are surprisingly
similar due to the presence of flanking aromatic moieties in the bridging groups. Hence, in 102, the
bariums atoms are also in a three-coordinate trigonal pyramidal geometry (Σθ(Ba1) = 297.73(8)°)
imposed by the additional presence of multiple secondary Ba···C(π) η3-interactions (within 3.199(2)-
3.321(2) Å) with each of the four phenyl substituents in ortho positions. Similar Ba∙∙∙C(π) interactions
have already been observed, e.g. in the structurally unusual homoleptic Ba-bis(phenoxide) [Ba2{μ2-
O(2,6-Ph2-C6H3}3{μ2-O(2,6-Ph2-C6H3}].[46] The values of the Ba1-O1 interatomic distance and Ba1-O1-
14
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
Si1 angle (2.3564(14) Å and 166.68(9)°, respectively) are commensurate with those in 92 and 12 (see
above). The Ba1-O23 and Ba1-O23i bond lengths from Ba to bridging Ophenoxide atoms (2.6728(14)-
2.7152(14) Å) are similar to those measured in other barium phenoxides.[64,65]
Figure 7. ORTEP view of the molecular structure of [Ba{μ2-O(2,6-Ph2-C6H3)}{OSi(SiMe3)3}]2 (102). H atoms
omitted for clarity. Ellipsoids at the 30% probability level. Selected interatomic distances (Å) and angles (°): Ba1-
O1 = 2.3564(14), Ba1-O23 = 2.6728(14), Ba1-O23i = 2.7152(14), Ba-C14 = 3.321(2), Ba1-C15 = 3.199(2), Ba1-
C16 = 3.2031(19), Ba1-C24i = 3.120(2), Ba1-C25i = 3.252(2), Ba1-C29i = 3.273(2); O1-Ba1-O23 = 111.75(5),
O1-Ba1-O23i = 107.11(5), O23-Ba1-O23i = 78.87(4), Ba1-O1-Si1 = 166.68(9). Symmetry transformations used
to generate equivalent atoms: i, -x, -y, -z, T = [ 1, 0, 0].
Summary of crystallographic data for barium complexes
In combination with our earlier contributions,[27-29] the results disclosed herein indicate that a range of
soluble barium boryloxides, siloxides and phenoxides 12-102 are synthetically available. They feature a
diversity of structural patterns in the molecular solid-state, with the main ones (nuclearity, structural
type, coordination number and presence of non-covalent interactions) highlighted in Table 2. Most of
these compounds exist as dimeric species in the solid state, where two different bridging modes for the
Ba2O4 core have been identified: a classical and largely prevailing one with two bridging and two
terminal ligands, and a minor one, with three bridging groups and a single terminal ligand as in 12 and
its thf adduct 12.thf.[29] This latter, unusual arrangement is reminiscent of that found in the molecular
structure of the alkoxide [Ba2(μ2-OCPh3)3(OCPh3)(thf)3].[42] Other than this complex, there are very few
examples of discrete homometallic barium alkoxides indexed in the CCDC database[66] for simple
alkoxo groups,[67] even though the structures of several fluoroalkoxides, as in
15
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
[Ba{OC(CF3)3}2(dme)3],[42] and those of heterobimetallic species, e.g. [Zr2Ba(μ2-OtBu)6(OtBu)4],[68]
have been described. Taken collectively, these data also point to the importance of Ba∙∙∙F, Ba∙∙∙C(π) and
even Ba∙∙∙H-C secondary interactions towards the stability of these complexes; these non-covalent
interactions are broken upon coordination of Lewis bases, as in the solvated 12.thf, 3.(thf)2 and 62.thf.
Table 2. Summary of solid-state structural data for barium siloxides, boryloxides and phenoxides 12-102.
Complex
structural
motif
C
coordination
number
3 & 4
secondary
interactions
Ba∙∙∙H-C
ref.
[Ba2{μ2-OSi(SiMe3)3}3{OSi(SiMe3)3}]
[Ba2{μ2-OSi(SiMe3)3}3{OSi(SiMe3)3}.thf]
[Ba{μ2-OSi(SiMe3)3}{N(SiMe3)2}]2
[Ba(OB{CH(SiMe3)2}2)2]
12
[29]
[29]
[29]
[27]
[28]
12.thf
22
C
4 & 4
-
B
3
2
2
4
3
3
3
4
3
3
3
3
-
3
A
A
A
B
Ba∙∙∙H-C
Ba∙∙∙C(π)
-
[Ba(OB{CH(SiMe3)2}2)2.toluene]
[Ba(OB{CH(SiMe3)2}2)2.(thf)2]
3.tol
3.(thf)2
42
[27]
b
[Ba{μ2-N(SiMe3)2}(OB{CH(SiMe3)2}2)]2
[Ba{μ2-OB(Tripp)2}{OB(Tripp)2}]2
[Ba{μ2-OB(Fmes)2}{OB(Fmes)2}]2
[Ba{μ2-OB(Fmes)2}{OB(Fmes)2}.thf]2
[Ba{μ2-OB(Tripp)2}{N(SiMe3)2}]2
[Ba{μ2-OSi(SiMe3)3}(OB{CH(SiMe3)2}2)]2
[Ba{μ2-OB(Tripp)2}{OSi(SiMe3)3}]2
[Ba{μ2-O(2,6-Ph2-C6H3)}{OSi(SiMe3)3}]2
-
b
b
b
b
b
b
b
52
B
Ba∙∙∙C(π)
Ba∙∙∙F
-
62
B c
B
62.thf
72
B
Ba∙∙∙C(π)
Ba∙∙∙H-C
Ba∙∙∙C(π)
Ba∙∙∙C(π)
82
B d
B
92
102
B
a
Arrangement of the Ba(X1)2, Ba2(μ-X1)2(X2)2 or Ba2(μ-X1)3(X2) core, without taking coordinated solvent
molecule(s) into account. b This work. c Coordination polymer in the solid-state. d Cs-symmetric dimer.
Examination of the data for the six heteroleptic complexes obtained so far, namely 22, 42 and 72-
102, indicate that the relative position (bridging vs terminal) of a given ligand in dimeric complexes
changes with its identity. Hence, the propensity of a given anion to occupy the bridging position
increases according to the empirical relationships displayed in Figure 8. No information can be provided
about (Fmes)2BO‒, as we failed to obtain structurally characterised heteroleptic complexes with this
ligand. The boryloxide {(Me3Si)2CH}2BO‒ seems to be unable to generate dinuclear species, supposedly
16
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
owing to its considerable steric bulk. Besides, we have so far not succeeded in our attempts to
discriminate between (Tripp)2BO‒ and 2,6-Ph2-C6H3-O‒. The empirical classification provided in Figure
8 is established on the sole basis of crystallographic data. It most probably is the expression of a
composite of different properties for these ligands, such as (i) steric features, (ii) electronic density and
overall σ- and π-donor capability of the negatively charged atom (O‒ or N‒), and (iii) the ability to
generate Ba∙∙∙F, Ba∙∙∙H-C or Ba∙∙∙C(π) secondary interactions. The particularly large size of the Ba2+
cation (rionic = 1.35 Å), and the resulting ability to accommodate bulky ligands and additional secondary
interactions, must also be a contributing factor. Although it seems reasonable to hypothesise so, it is
unclear at this stage to which extent the relationships displayed in Figure 8 can be extrapolated to the
comparatively smaller calcium and strontium (rionic = 1.00 and 1.18 Å).
Figure 8. Relative bridging ability of O-based monoanionic ligands established from the structural data for the
dinuclear heteroleptic barium complexes 22, 42 and 72-102. Br = bridging and term = terminal positions.
Comparison between alkaline earths.
The heteroleptic complexes [Ae{μ2-N(SiMe3)2}(OB{CH(SiMe3)2}2)]2 (Ae = Mg, 112; Ca, 122; Sr, 132)
were prepared in 53-84% yields by reaction of [Ae{N(SiMe3)2}2]2 with {(Me3Si)2CH}2BOH. In
conjunction with the barium complex [Ba{μ2-N(SiMe3)2}(OB{CH(SiMe3)2}2)]2 (42),[27] they afford a
complete series of homologous heteroleptic dimers. The molecular structures of the three new
complexes were determined by XRD analysis performed on single crystals. Like 42, they all form dimer
bridged by the amido groups, while the boryloxides are located in terminal positions. Although not
identical (vide infra), the structural arrangement is similar across the four complexes, hence only the
strontium complex 132 is detailed here; the ORTEP representations 112 and 122 can be found in the
Supporting Information (Figures S47-S48) with a list of representative metric parameters.
Complex 132 forms a C1-symmetric dimer with a flat, rhomboidal Sr2N2 core (Figure 9). The
geometry about the two Sr atoms differs slightly. The two B1-O1-Sr1 and B2-O2-Sr2 angles close to
perfectly linear (178.90(16)° and 173.51(19)°) suggest the participation of the lone pairs at O to the
formation of sp-hybridised orbitals, and all six B1-O1-Sr1-B2-O2-Sr2 atoms are nearly aligned. The
geometry about each strontium is trigonal planar (Σθ(Sr1) = 359.99(11)°; Σθ(Sr2) = 359.79(11)°). The
‒
bridging position of the two N(SiMe3)2 ligands is unusual. To our knowledge, the sole other example
17
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
of such heteroleptic dimer is [Sr{μ-N(SiMe3)2}{N(C6F5)2}]2,[69] where the Sr-Nbr interatomic distances,
in the range 2.598(2)- 2.634(2) Å, match well those in 132 (2.601(2)-2.648(2) Å). The two boron atoms
are also in a perfect trigonal planar environment (Σθ(B1) = 360.0(3)°; Σθ(B2) = 359.9(3)°).
Figure 9. ORTEP view of the molecular structure of [Sr{μ2-N(SiMe3)2}(OB{CH(SiMe3)2}2)]2 (132). H atoms
omitted for clarity. Ellipsoids at the 50% probability level. Selected interatomic distances (Å) and angles (°): Sr1-
N1 = 2.648(2), Sr1-N2 = 2.634(2), Sr1-O1 = 2.2751(18), Sr2-N1 = 2.601(2), Sr2-N2 = 2.635(2), Sr2-O2 =
2.2727(18), O1-B1 = 1.351(3), O2-B2 = 1.346(3); Sr1-N1-Sr2 = 95.19(7), Sr1-N2-Sr2 = 94.71(7), N1-Sr1-N2 =
84.59(7), N1-Sr1-O1 = 139.43(6), N2-Sr1-O1 = 135.97(7), N1-Sr2-N2 = 85.49(7), N1-Sr2-O2 = 132.46(7), N2-
Sr2-O2 = 141.84(6), B1-O1-Sr1 = 178.90(16), B2-O2-Sr2 = 173.51(19), O1-B1-C1 = 121.8(2), O1-B1-C8 =
121.6(2), C1-B1-C8 = 116.6(2), O2-B2-C15 = 121.8(2), O2-B2-C22 = 121.1(2), C15-B2-C22 = 117.0(2).
The main structural features are very similar across the three amide-bridged complexes 112, 122
and 132 (all C1-symmetrical, although the metric parameters is geometrical arrangements about the two
metals atoms in both 112 and 122 are almost identical, resulting in a pseudo-C2 symmetry), and they
resemble those in the known barium congener 42. The most representative data are compiled in Table 3.
While the eight B, O, Met and N atoms are coplanar or very nearly so in 112-132 (see for instance the
extremely short distances between the oxygen or boron atoms and the best average plane defined by the
Ae2N2 planar core in these complexes), the two O-atoms and two B-atoms protrude out of the Ba2N2
planar core in the Ci-symmetric barium complex 42. This last compound is also the only one where the
geometry about the metal atoms shows noticeable pyramidalisation (Σθ(Ba) = 348.16(9)°). The Ae-to-
O and Ae-to-N interatomic distances globally increase as expected on account of the differences between
the ionic radii of the Mg2 (0.74 Å), Ca2+ (1.00 Å), Sr2+ (1.18 Å) and Ba2+ (1.35 Å) cations. The Ae-N-
Ae’ angle also widens as a result of increasing metal size. A particularly striking contrast between the
18
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
four complexes is the largely different B-O-Ae angles: while they are nearly perfectly linear for the
calcium and strontium complexes 122 and 132, they are much less obtuse for the magnesium and barium
derivatives 112 (157.40(3)° and 158.30(3)°) and 42 (160.00(15°).
Table 3. Summary of structural data across [Ae{μ2-N(SiMe3)2}(OB{CH(SiMe3)2}2)]2 (Ae = Mg, Ca, Sr, Ba).
Ae
Mg (112)
Ca (122)
Sr (132)
C1
Ba (42)[27]
Symmetry
C1 (pseudo-C2)
C1 (pseudo-C2)
Ci
Mg1
Mg2
Ca1
Ca2
Sr1
Sr2
Ba1 and Ba1i
Ae-N (Å)
2.107(4)
2.117(4)
2.122(4)
2.132(4)
2.4399(11) 2.4471(11) 2.634(2)
2.4467(11) 2.4499(11) 2.648(2)
2.601(2)
2.635(2)
2.7210(16)
2.8912(17)
Ae-O (Å)
B-O (Å)
1.857(3)
1.341(5)
91.87(15)
1.863(3)
1.351(5)
91.05(14)
2.1209(9)
2.1228(10) 2.2727(18) 2.2751(18) 2.3999(14)
1.3353(17) 1.3376(18) 1.351(3)
1.346(3)
85.49(7)
132.46(7)
141.84(6)
1.332(3)
82.34(5)
121.64(5)
144.18(5)
N-Ae-N’ (°)
N-Ae-O (°)
87.44(4)
87.21(4)
135.47(4)
137.32(4)
84.59(7)
135.97(7)
139.43(6)
131.46(16) 130.92(16) 136.22(4)
136.67(15) 137.88(15) 136.30(4)
Σθ(Ae)
360.00(27) 359.84(26) 359.96(11) 360.00(7)
157.4(3) 158.3(3) 178.65(9) 178.92(10) 178.90(16) 173.51(19) 160.00(14)
0.0493(30) 0.1389(30) 0.0342(9) 0.0368(9) 0.0529(18) 0.1647(18) 0.9776(14)
0.0384(49) 0.1263(49) 0.0268(15) 0.0481(15) 0.0655(29) 0.1487(29) 1.0758(22)
88.50(14)-88.51(14) 92.55(4)-92.79(4) 94.71(7)-95.19(7) 97.66(5)
359.99(11) 359.79(11) 348.16(9)
B-O-Ae (°)
dO-plane (Å)a
dB-plane (Å)b
Ae-N-Ae’ (°)
a Distance from the oxygen atoms to the best average plane defined by the rhomboidal Ae2N2 core. b Distance from
the boron atoms to the best average plane defined by the rhomboidal Ae2N2 core.
Computational analysis
We first examined the bonding within an homogeneous series of simple dinuclear homoleptic complexes
with structural motif B (see Table 2), namely, [Ae{μ2-N(SiMe3)2}(N(SiMe3)2)]2, for which X-ray
structural data are available for Ae = Mg, Ca, Sr and Ba.[70-73] For this purpose, DFT calculations were
performed at the PBE0/Def2-TZVP-D3 level of theory (see Supporting Information for computational
details). Relevant averaged computed data are provided in Table 4, together with some experimental
metrical values. The four optimised geometries (left side of Figure 10), are in a very good agreement
with their X-ray counterparts. Note that, in the four structures, the terminal nitrogen atoms (Nt) retain
their trigonal planar bonding mode (Σθ(Nt) = 360°; Table 4). This is in line with the planarity of
N(SiR3)3,[74] where hyperconjugation exists between the nitrogen lone pair and the three silyl groups.
The Nt planarity also allows for π donation to the Ae metal. However, the structures of the four
19
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
complexes differ substantially with respect to the rotational conformations of their terminal disilazides.
It is particularly noteworthy that in the crystal, the Ba and Mg complexes adopt different exact point
symmetry groups: in [Ba{μ2-N(SiMe3)2}(N(SiMe3)2)]2 (Ci symmetry), the two planes defined by the two
sets of (NtSi2) atoms remain roughly parallel to the BaNbrBaNbr rhombus (Nbr = bridging N-atom; Nt =
terminal N-atom), while for the magnesium derivative (C2 symmetry), the two (NtSi2) planes form angles
of 75.8° and 57.4° with the (BaNbr)2 core, respectively, and an angle of 46.8° between each other.[70] The
DFT-optimised structures reproduce nicely these structural differences (Figure 10). The
crystallographically determined molecular structures of the Ca and Sr species have no exact symmetry
element in the crystal. That of Ca exhibits a rotational conformation of the terminal amido groups rather
similar to that of Mg, but with the latter two terminal Nt(SiMe3)groups bent away from the Ae-Ae’ axis.
This unsymmetrical structure was also found as the most stable by DFT (Figure 10). With strontium,
the DFT-optimised geometry of [Sr{μ2-N(SiMe3)2}(N(SiMe3)2)]2 was found to be of C2 symmetry, but
this time with the C2 axis lying along the Nbr-Nbr vector; hence, the four Ae and Nt atoms are not perfectly
aligned. Although of a different symmetry group, this structure is quite similar to that of the Ba congener.
As a matter of fact, optimising [Sr{μ2-N(SiMe3)2}(N(SiMe3)2)]2 assuming Ci symmetry destabilises it
by only 1.4 kcal/mol.
Contrarily to Nt atoms, the Ae centre, which adopts a perfect trigonal planar coordination in the
case of Mg, shows increasing pyramidalisation when descending group 2 and drops to Σθ(Ba) = 350.8°
in the case of Ba (Table 4). As a result, the two Nt atoms progressively move out of the AeNbrAeNbr
plane; this out-of-plane shift is particularly noticeable in the case of Sr and Ba (~ 0.6 and 0.9 Å,
respectively). Gradual pyramidalisation at Ae atoms in the [Ae{μ2-N(SiMe3)2}(N(SiMe3)2)]2 series is
reminiscent of the bending observed in the heavier AeX2 (X = H or halide) species;[75] it is attributed to
metal core polarisation and/or participation of the valence d-orbitals of Ae elements.[75-79] The latter has
been described as resulting from a pseudo-Jahn-Teller effect from post Hartree-Fock calculations.[80]
The atomic charges obtained from a natural atomic orbital (NAO) population analysis (Table 4)
indicate strong ionicity, with Ae-charges varying within the narrow range 1.82-1.87. The corresponding
valence NAO occupations indicate that the valence s orbital very largely dominates the (weak) covalent
component of the bonding in the case of Mg. On the other hand, with the heavier Ae elements, the
valence (n-1)d orbitals are also significantly involved. Note also that with Ca, Sr and Ba, some core
orbitals are also slightly depopulated to the benefit of their valence counterpart. From the atomic
charges, it emerges that the weakly covalent interaction increases in the order Mg < Ba ≈ Sr < Ca. This
is overall consistent with the analysis of the Ae-N Wiberg bond index (WBI) values, also listed in Table
4: Mg < Sr < Ba < Ca. This trend is dominated by the involvement of the valence d(Ae) orbitals. Partial
occupation of the d(Ae) AOs is not the single factor inducing Ae pyramidalisation,[75-79] since it is
maximal for Ae = Ca which pyramidalises only moderately. Part of the (n-1)d(Ae) population might
also come from π-donation of the Nt lone pair, this effect working against pyramidalisation at Ae. At
any rate, all these orbital effects are small and bonding in these complexes remains chiefly ionic. Finally,
20
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
a significant number of Ae∙∙∙H-C interactions exist in each of the computed complexes. For example, in
[Ba{μ2-N(SiMe3)2}(N(SiMe3)2)]2, eight anagostic Ba∙∙∙C-H contacts shorter than the sum of the H and
Ba van der Waals radii (3.20 Å) are seen for each metal.
[Ae{μ2-N(SiMe3)2}(N(SiMe3)2)]2
[Ae{μ2-N(SiMe3)2}(OB{CH(SiMe3)2}2)]2
Ae = Mg
Ae = Ca
Ae = Sr
Ae = Ba
Figure 10. DFT-optimised structures of the homoleptic series [Ae{μ2-N(SiMe3)2}(N(SiMe3)2)]2 (left) and [Ae{μ2-
N(SiMe3)2}(OB{CH(SiMe3)2}2)]2 (112-132 and 42; right). Ae = Mg, Ca, Sr, Ba. Hydrogen atoms omitted for clarity.
Colour of atoms: Green = Ae; blue = N; red = O; beige = Si; pink = B; grey = C.
21
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
Table 4. Selected averaged computed data across [Ae{μ2-N(SiMe3)2}(N(SiMe3)2)]2 (Ae = Mg, Ca, Sr, Ba). Wiberg
bond indices are given in brackets. Experimental averaged metrical values[70-73] are reported in italics for
comparison. Nt and Nb refer to the terminal and bridging ligand, respectively.
Ae
Mg
Ca
Sr
Ba
a
b
Symmetry
C2
C1
C2
Ci
2.128 [0.025]
2.454 [0.045]
2.609 [0.040]
2.803 [0.042]
Ae-Nb (Å)
Ae-Nt (Å)
2.152
2.4786
2.635
2.822
1.967 [0.038]
2.274 [0.097]
2.416 [0.080]
2.552 [0.088]
1.976
2.274
2.433
2.577
95.7
95.9
88.6
89.2
84.3
83.9
82.7
81.9
Nb-Ae-Nb’ (°)
Nb-Ae-Nt (°)
Σθ(Ae) (°)
132.1
132.2
134.6
134.8
135.6
136.8
134.1
134.8
360.0
360.0
357.8
358.8
355.6
357.5
350.8
351.5
360.0
360.0
359.7
359.7
359.8
359.9
360.0
360.0
Σθ(Nt) (°)
0.000
0.000
0.556
0.444
0.634
0.478
0.949
0.915
DNt-plane (Å)c
Ae-Nb-Ae’ (°)
84.3
84.1
91.3
90.7
95.7
96.1
97.3
98.1
1.88
3s0.10 3p0.01 3d0.01
–2.03
1.82
4s0.08 4p0.01 3d0.09
–1.93
1.84
5s0.07 5p0.01 4d0.07
–1.94
1.87
6s0.07 6p0.01 5d0.07
–1.92
Ae
Atomic
charges
Nb
Nt
–1.97
–1.88
–1.88
–1.88
a C2 axis parallel to the Ae-Ae’ vector. b C2 axis parallel to the Nb-Nb’ vector. c Distance of Nt to the (AeNb)2
mean plane.
We now shift our attention to the analysis of the set of heteroleptic complexes [Ae{μ2-
N(SiMe3)2}(OB{CH(SiMe3)2}2)]2 (Ae = Mg, Ca, Sr, Ba, i.e. 112-132 and 42); relevant computed data are
provided in Table 5. The four optimised geometries (Figure 10, right) were found to be rather floppy, a
recognised feature of Ae-centred molecules,[79] with their lowest energy structure being of (or very close
to) Ci symmetry. Their optimised metrical data are in good agreement with their X-ray counterparts
(Tables 3 and 5). As in the homoleptic series, the geometry about Ae pyramidalises when going down
group 2, to reach Σθ(Ba) = 343.7° in the case of the barium complex 42. The experimental trend along
this series for the O and B out-of-plane shifting mentioned above is thus nicely reproduced by DFT. As
seen for the homoleptic series of amides, the covalent component of the Ae-ligand bonding involves
essentially the valence s(Ae) orbital in the case of Mg, and both s(Ae) and d(Ae) orbitals for the heavier
elements. The experimental tendency for larger B-O-Ae angles in the cases of Ca and Sr is also
reproduced by DFT. It is associated with larger B-O Wiberg indices, suggesting a larger tendency of the
22
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
oxygen atoms for sp hybridisation, and hence (slightly) larger π-donation to Ae. Taken collectively, the
data of Table 5 suggests increasing covalency in the order Mg < Sr < Ba < Ca. However, as for the
homoleptic series discussed above, the covalent component of the bonds remains weak compared to the
ionic contribution. Several Ae∙∙∙H-C anagostic interactions are also present in all complexes of the
heteroleptic set. For example, for the barium complex 42, eight Ba∙∙∙H-C contacts shorter than 3.20 Ǻ
are seen for each Ba atom; this is fully consistent with the crystallographically established molecular
structure for this complex.[27] Several additional Ba∙∙∙H contacts slightly larger than 3.20 Å are also
detected.
Table 5. Selected computed data for the series [Ae{μ2-N(SiMe3)2}(OB{CH(SiMe3)2}2)]2 (Ae = Mg, Ca, Sr, Ba,
112-132 and 42). Wiberg bond indices are given into brackets.
Ae
Mg (112)
Ca (122)
Sr (132)
Ba (42)
Symmetry
Pseudo-Ci
Ci
Ci
Ci
2.103 [0.029]
2.110 [0.030]
2.403 [0.048]
2.516 [0.034]
2.557 [0.040]
2.670 [0.027]
2.717 [0.044]
2.851 [0.031]
Ae-N (Å)
Ae-O (Å)
B-O (Å)
1.858 [0.042]
1.343 [1.022]
91.8
2.117 [0.082]
1.336 [1.059]
84.9
2.260 [0.065]
1.332 [1.065]
84.3
2.391 [0.070]
1.330 [1.054]
83.2
N-Ae-N’ (°)
130.2
138.0
126.9
148.0
125.4
145.1
121.9
138.7
N-Ae-O (°)
Σθ(Ae) (°)
B-O-Ae (°)
dO-plane (Å)a
dB-plane (Å)b
Ae-N-Ae’ (°)
360.0
154.5
0.029
0.176
88.2
359.8
174.1
0.126
0.145
95.1
354.8
168.7
0.626
0.750
95.7
343.7
158.2
1.173
1.351
96.8
1.88
3s0.11 3p0.01 3d0.01
–2.03
1.85
4s0.08 4p0.01 3d0.09
–1.94
1.87
5s0.08 5p0.01 4d0.07
–1.94
1.88
6s0.07 6p0.01 5d0.07
–1.93
Ae
Atomic
charges
N
O
B
–1.31
1.18
–1.27
1.16
–1.28
1.16
–1.28
1.18
a Distance of O to the (AeN)2 mean plane. b Distance of B to the (AeN)2 mean plane.
In order to provide a rationalisation of the empirical bridging ability proposed in Figure 8, the
whole Ba heteroleptic series [Ba(μ2-X1)(X2)]2, namely 22, 42 and 72-102, was also calculated along with
their hypothetical topological isomers [Ba(μ2-X2)(X1)]2 (denoted 22*, 42* and 72*-102*), where X1 and
X2 are chosen among N(SiMe3)2, OB{CH(SiMe3)2}2, OSi(SiMe3)3, O-2,6-Ph2-C6H3 and OB(Tripp)2 and
X1 ≠ X2. The corresponding optimised geometries are given in Figures S55-56; those for complexes 22,
23
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
42 and 72-102 are in good agreement with their experimental X-ray analogues. In particular, the Ba∙∙∙H-
C and Ba∙∙∙C(π) secondary interactions are well reproduced in the optimised structures. Consistently
with the fact they were not observed experimentally, the topological isomers of 22*, 42*, 72*, 82*, 92*,
and 102* were found systematically less stable by 17.5, 13.1, 24.2, 5.5, 10.8 and 7.1 kcal/mol,
respectively. Therefore, the DFT simulations reproduce the empirical trend displayed in Figure 8. As
shown above, the nature of the Ba-O or Ba-N bonds in these complexes is largely ionic in character. It
follows that the O and N atomic charge on the “free ligand” is likely to be an important parameter in the
prediction of the bonding strength and mode: the more negative the charge, the stronger the bond, but
also the more favoured the bridging position over the terminal one. A negatively charged atom is indeed
expected to interact in a stronger way with two metal cations, as compared with a single one, provided
the metal-ligand distances do not differ importantly. The absolute values of the computed N and O
natural atomic charges, and the potential bridging preference with them, are increasing in the order:
[O(2,6-Ph2-C6H3)]– (–0.72) < [OB(Tripp)2]– (–0.97) < [OB{CH(SiMe3)2}]– (–1.07) < [OSi(SiMe3)3]– (–
1.29) < [N(SiMe3)2]– (–1.75). Obviously, this trend does not follow that given in Figure 8; it is only
consistent with the configurations of [Ba{μ2-N(SiMe3)2}(OB{CH(SiMe3)2}2)]2 (42) and [Ba{μ2-
OSi(SiMe3)3}(OB{CH(SiMe3)2}2)]2 (82). Although it is only marginal, one may then wonder about the
potential ability of covalent bonding in these complexes. The ligand donating abilities are related to the
energies and localisations of their lone-pair orbitals. Whereas a terminal ligand needs a single donating
orbital (or a combination thereof) of -type, a bridging ligand needs to use two donating orbitals, one
of -type and one of π-type. The more available the π-type orbital, the more privileged the bridging
position over the terminal one. It turns out that all the investigated ligands have their HOMO containing
the π-type lone pair, as illustrated in their Kohn-Sham MO diagrams (Figure S57). All of them have
their -type lone pair lying much lower, except for [N(SiMe3)2]– in which the two donating orbitals are
close in energy. From their π-type HOMO energies, the bridging preference increases as:
[OB{CH(SiMe3)2}]– (–1.77 eV)  [OB(Tripp)2}]– (–1.70 eV) < [OSi(SiMe3)3]– (–1.55 eV) <
[N(SiMe3)2]– (–1.37 eV) < [O(2,6-Ph2-C6H3)]– (–1.27 eV). Again, and as expected owing to the weak
covalent character in these complexes, no clear agreement with the empirical trend in Figure 8 is seen.
From these considerations, it follows that the secondary interactions play an important role in the
isomer preference. Unsurprisingly, the Ba∙∙∙C(π) bonding in 72, 92 and 102 plays a key role in favouring
the bridging position of the [OB(Tripp)2}]– and [O(2,6-Ph2-C6H3)]– ligands. The corresponding
optimised Ba∙∙∙C distances of 3.0-3.3 Å are only moderately larger than that in BaCp2 (3.0-3.1 Ǻ) Å.[81]
The case of [Ba{μ2-OSi(SiMe3)3}{N(SiMe3)2}]2 (22), which does not possess any π system, is
particularly telling: whereas the siloxide [OSi(SiMe3)3]– is less negatively polarised than [N(SiMe3)2]–,
it still prefers the bridging position between the two Ba atoms. The optimised geometries of 22 and its
topological isomer [Ba{μ2-N(SiMe3)2}{ OSi(SiMe3)3}]2 (22*, Figure S55) differ in their compactness,
for 22 is rather spherical, while 22* is much more oblong. Consistently, the dispersion energy computed
24
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
through Grimme‘s empirical corrections[82] for 22 is 9.0 kcal/mol lower than in 22*. This difference is
indicative of larger van der Waals and steric stabilisations in 22. Another part of these secondary non-
bonding interaction energies is likely to be contained into the DFT electronic energy. On the other hand,
in complexes 42 and 82, the dispersion energy actually favours the isomers 42* and 82* not seen
experimentally. One may then infer that the preference for the actual structures of 42 and 82 is dictated
by the direct ligand-to-Ba bonding, which is dominated by ionicity.
In summary, the ability of N- and O-based ligands for bridging two Ba2+ cations is a composite
of several contributions. With sufficiently non-bulky ligands, since the Ba-N or Ba-O bond is mainly
ionic, the more negatively polarised (on N or O) ligand prefers occupying the bridging position, thus
maximizing electrostatic interactions. In the case of homoleptic dinuclear complexes of such “small”
ligands, since the bridging position is preferred, the number of bridging ligands will be maximised to
[29]
three, as actually seen in 12 and, for instance, in Ba2Cl4.[75,79] In these two examples, we found the
triply bridged structural motif C (Table 2) to be more stable that the “regular” doubly bridged motif B
by 13.6 and 9.2 kcal/mol respectively. Interestingly, the quadruply bridged structure [Ba(µ2-Cl)2]2 of
D4h symmetry is also a local energy minimum for Ba2Cl4.[75] At our level of calculations, it is 15.0
kcal/mol less stable than the C3v Ba2(µ2-Cl)3Cl minimum. With four bridging chlorides, the Ba-Cl
electrostatic interactions are maximised, but the repulsions between the ligands disfavour this
configuration. With bulkier ligands, secondary metal-ligand bonding interactions are expected to come
into the play. For example, Ba∙∙∙H-C anagostic interactions (~ 3 Å) are present in all our optimised
structures and their (relative) weakness can be compensated by their large number. More specific but
stronger interactions can also occur, such as the above-mentioned Ba∙∙∙C(π) more covalent interactions
with ligands that contain a π-system (e.g. a C6 aromatic ring), but also electrostatic Ba∙∙∙F interactions
in the case of fluorinated ligands. We have shown on model calculations that the magnitude of a single
Ca∙∙∙F bonding energy could reach 9-14 kcal/mol.[49] Similar values can be expected for barium. It should
be noted that the possibility for such secondary interactions could occur in both bridging and terminal
positions of the ligand, making the induction of the bridging vs. terminal choice also dependant of other
factors, such as the shape and flexibility of the co-ligand. Indeed, ligand∙∙∙ligand interactions play also
a crucial role in the structural determinism of these complexes. They result from a compromise between
steric repulsions and van der Waals bonding (dispersion forces), in particular through H∙∙∙H interactions.
The latter tend to favour pseudo-spherical shapes (as in 22) to the detriment of more open structures. The
main learning to take away is that the bridging ability of a specific ligand depends of course on its
various intrinsic features but also, in a significant part, on the way these characteristics can be expressed
in relation with those of its other co-ligand. Therefore, it seems wise to consider the trend given in Figure
8 as relative to the considered ligand interplay. Introducing other ligand combinations may partly
contradict this relative classification.
25
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
Concluding comments
Building on our earlier syntheses of low-coordinate barium-boryloxides and barium-siloxides,[27,29] we
have prepared here several homoleptic and heteroleptic barium complexes supported by various
monoanionic O-based ligands. The complexes form dinuclear species that have been structurally
characterised by X-ray diffraction crystallography. The siloxide, boryloxide and even phenoxide ligands
are shown to occupy in turn bridging or terminal positions, depending on the specific case and ligand
combination. Extended to the structural data already available for complexes bearing these O-based
ligands, including for heteroleptic compounds that also contain the [N(SiMe3)2]– amide, Table 2
highlights a diversity of coordination patterns, some of them being unexpected. The experimentally-
established propensity to occupy bridging positions, as opposed to terminal ones, in heteroleptic dimers,
increases with [OB{CH(SiMe3)2}2)]– < [N(SiMe3)2]– < [OSi(SiMe3)3]– < [O(2,6-Ph2-C6H3)]– <
[OB(2,4,6-iPr3-C6H2)2]–. In the solid state, most of these complexes display secondary Ba∙∙∙H-C, Ba∙∙∙F
and Ba∙∙∙C(π) secondary interactions which not only enhances profoundly their stability as demonstrated
on several occasions,[31,49] but also their structures in the solid state. In these complexes bearing
“simple”, monodentate ligand, the coordination of Lewis bases, such as thf, suppresses partly or entirely
these secondary interactions. This needs not be always the case, as we have shown earlier that
combinations of multiple Ca∙∙∙H-C, Ca∙∙∙F and Ca∙∙∙C(π) secondary interactions within sterically
congested calcium complexes supported by functionalised aminoether-fluoroalkoxide effectively
prevented the coordination of thf onto the metal.[83]
The comparison of crystallographic data obtained for the homologous set of heteroleptic
complexes [Ae{μ2-N(SiMe3)2}(OB{CH(SiMe3)2}2)]2 (Ae = Mg, Ca, Sr and Ba) shows that structural
features are generally well reproduced across the whole series, although some subtle variations are
observed, notably with the smaller magnesium. It therefore feels legitimate to assume that the main
conclusions drawn from the detailed examination of the structures of barium complexes can be
extrapolated to, at least, calcium and strontium; this is also confirmed by DFT computations.
Our DFT calculations on series of homo- and heteroleptic dinuclear complexes of barium and
related alkaline earths, including magnesium, stress on the largely prevailing ionic character of ligand-
Ae bonding. The minor covalent component of the bonding is mainly controlled by the valence s atomic
orbital of the alkaline earth, together with the valence d atomic orbital in the case of the heavier Ca, Sr
and Ba elements. The participation of d atomic orbitals favours non-planarity of the coordination mode
for the heavier Ae metals, associated with an increased floppiness of the complexes. The pre-eminence
of the ionic character of the Ae-ligand bond favours bridging coordination, the number of bridging
ligands (most often two) being regulated by steric factors. In the case of heteroleptic species, the relative
ligand preference for bonding vs. terminal positions is not determined by the ligand abilities to make
stronger/weaker direct terminal or bridging bonds with the metal, but rather by the secondary (but non-
negligible) interactions that can occur between the Ae metal and the ligand. This is well exemplified by
26
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
the compounds discussed herein, which exhibit a range of anagostic (and numerous) Ae∙∙∙H, covalent
Ae∙∙∙C(π) and/or ionic Ae∙∙∙F interactions.
Considering the importance of well understood and controlled structure-reactivity relationships
for the design of efficient alkaline-earth homogeneous catalysts, and in particular for those constructed
with oxygen-based ligands, the data and analysis presented herein should constitute a useful guideline
to assist synthetic organometallic main group chemists in their future endeavours. Comforted with this
knowledge and with the further confirmation of the key role of secondary interactions in alkaline-earth
chemistry, we now intend to prepare the simplest and, arguably, most versatile of O-ligated complexes:
soluble, well-defined barium alkoxides and their lighter congeners.
Acknowledgements
The authors thank the ANR for funding (ANR-17-CE07-0003-01) and for the provision of a PhD
studentship to J. H. The GENCI (Grand Equipment National de Calcul Intensif) is acknowledged for
HPC resources (Project A0050807367).
Conflicts of interest
There are no conflicts of interest to declare.
Author Contributions
E. Le Coz and J. Hammoud performed synthetic experimental work and data analysis. T. Roisnel, M.
Cordier and V. Dorcet performed the XRD analyses. J.-Y. Saillard and S. Kahlal carried out all
computational work and took part to the writing of the draft. J.-F. Carpentier participated to the design
of experiments and analysis of experimental data. Y. Sarazin was the lead investigator; he participated
to the design of experiments, analysis of experimental data and writing of the draft.
Notes and references
[1] S. Harder, Chem. Rev. 2010, 110, 3852.
[2] M. R. Crimmin, M. S. Hill, Top. Organomet. Chem. 2013, 45, 191.
[3] M. S. Hill, D. J. Liptrot, C. Weetman, Chem. Soc. Rev. 2016, 45, 972.
[4] Early Main Group Metal Catalysis: Concepts and Reactions, ed. S. Harder, Wiley-VCH Verlag
GmbH & Co, 2020.
[5] S. Harder, F. Feil, K. Knoll, Angew. Chem. Int. Ed. 2001, 40, 4261.
[6] Y. Sarazin, B. Liu, T. Roisnel, L. Maron, J.-F. Carpentier, J. Am. Chem. Soc. 2011, 133, 9069.
[7] M. R. Crimmin, I. J. Casely, M. S. Hill, J. Am. Chem. Soc. 2005, 127, 2042.
[8] F. Buch, J. Brettar, S. Harder, Angew. Chem. Int. Ed. 2006, 45, 2741.
[9] J. Spielmann, F. Buch, S. Harder, Angew. Chem. Int. Ed. 2008, 47, 9434.
27
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
[10] A. G. M. Barrett, C. Brinkmann, M. R. Crimmin, M. S. Hill, P. Hunt, P. A. Procopiou, J. Am.
Chem. Soc. 2009, 131, 12906.
[11] P. Jochmann, J. P. Davin, T. P. Spaniol, L. Maron, J. Okuda, Angew. Chem. Int. Ed. 2012, 51,
4452.
[12] B. Liu, T. Roisnel, J.-F. Carpentier, Y. Sarazin, Angew. Chem. Int. Ed. 2012, 51, 4943.
[13] H. Bauer, M. Alonso, C. Fischer, B. Rösch, H. Elsen, S. Harder, Angew. Chem. Int. Ed. 2018, 57,
15177.
[14] A. S. S. Wilson, C. Dinoi, M. S. Hill, M. F. Mahon, L. Maron, Angew. Chem. Int. Ed. 2018, 57,
15500.
[15] F. M. Younis, S. Krieck, T. M. Al-Shboul, H. Görls, M. Westerhausen, Inorg. Chem. 2016, 55,
4676.
[16] H. Bauer, M. Alonso, C. Färber, H. Elsen, J. Pahl, A. Causero, G. Ballmann, F. de Proft, S.
Harder, Nat. Catal. 2018, 1, 40
[17] J. Martin, C. Knüpfer, J. Eyselein, C. Färber, S. Grams, J. Langer, K. Thum, M. Wiesinger, S.
Harder, Angew. Chem. Int. Ed. 2020, 59, 9102.
[18] L. Zhao, X. Shi and J. Cheng, ACS Catal. 2021, 11, 2041.
[19] A. G. M. Barrett, M. R. Crimmin, M. S. Hill, P. B. Hitchcock, P. A. Procopiou, Organometallics
2007, 26, 4076.
[20] J. Spielmann, S. Harder, J. Am. Chem. Soc. 2009, 131, 5064.
[21] M. S. Hill, M. Hodgson, D. J. Liptrot, M. F. Mahon, Dalton Trans. 2011, 40, 7783.
[22] D. J. Liptrot, M. S. Hill, M. F. Mahon, A. S. S. Wilson, Angew. Chem. Int. Ed. 2015, 54, 13362.
[23] J. Spielmann, G. Jansen, H. Bandmann, S. Harder, Angew. Chem. Int. Ed. 2008, 47, 6290.
[24] M. S. Hill, D. J. Liptrot, D. J. MacDougall, M. F. Mahon, T. P. Robinson, Chem. Sci. 2013, 4,
4212.
[25] C. Bellini, J.-F. Carpentier, S. Tobisch, Y. Sarazin, Angew. Chem. Int. Ed. 2015, 54, 7679.
[26] C. Bellini, C. Orione, J.-F. Carpentier, Y. Sarazin, Angew. Chem. Int. Ed. 2016, 55, 3744.
[27] E. Le Coz, V. Dorcet, T. Roisnel, S. Tobisch, J.-F. Carpentier, Y. Sarazin, Angew. Chem. Int. Ed.
2018, 57, 11747.
[28] E. Le Coz, Z. Zhang, T. Roisnel, L. Cavallo, L. Falivene, J.-F. Carpentier, Y. Sarazin, Chem. Eur.
J. 2020, 26, 3535.
[29] E. Le Coz, S. Kahlal, J.-Y. Saillard, T. Roisnel, V. Dorcet, J.-F. Carpentier, Y. Sarazin, Chem.
Eur. J. 2019, 25, 13509.
[30] S. Tobisch, Chem. Eur. J. 2015, 21, 6765.
[31] P. M. Chapple, Y. Sarazin, Eur. J. Inorg. Chem. 2020, 3321.
[32] T. X. Gentner, B. Rösch, K. Thum, J. Langer, G. Ballmann, J. Pahl, W. A. Donaubauer, F.
Hampel, S. Harder, Organometallics 2019, 38, 2485.
28
This article is protected by copyright. All rights reserved.

10.1002/chem.202101687
Chemistry - A European Journal
[33] P. M. Chapple, S. Kahlal, J. Cartron, T. Roisnel, V. Dorcet, M. Cordier, J.-Y. Saillard, J.-F.
Carpentier, Y. Sarazin, Angew. Chem. Int. Ed. 2020, 59, 9120.
[34] P. M. Chapple, M. Cordier, V. Dorcet, T. Roisnel, J.-F. Carpentier, Y. Sarazin, Dalton Trans.
2020, 49, 11878.
[35] K. G. Caulton, M. H. Chisholm, S. R. Drake, K. Folting, J. Chem. Soc., Chem Commun. 1990,
1349.
[36] K. F. Tesh, D. J. Burkey, T. P. Hanusa, J. Am. Chem. Soc. 1994, 116, 2409.
[37] X. Shi, C. Hou, C. Zhou, Y. Song, J. Cheng, Angew. Chem. Int. Ed. 2017, 56, 16650.
[38] X. Shi, G. Qin, Y. Wang, L. Zhao, Z. Liu, J. Cheng, Angew. Chem. Int. Ed. 2019, 58, 4356.
[39] For a soluble Ba-H cluster supported by polyamines, see: M. Wiesinger, B. Maitland, C. Färber,
G. Ballmann, C. Fischer, H. Elsen, S. Harder, Angew. Chem. Int. Ed. 2017, 56, 16654.
[40] D. C. Bradley, R. C. Mehrotra, I. P. Rothwell, A. Singh, Alkoxo and Aryloxo Derivatives of
Metals, 2001, Academic Press.
[41] O. Poncelet, L. G. Hubert-Pfalzgraf, L. Toupet and J.-C. Daran, Polyhedron 1991, 10, 2045.
[42] S. R. Drake, W. E. Streib, K. Folting, M. H. Chisholm, K. G. Caulton, Inorg. Chem. 1992, 31,
3205.
[43] J. Utko, S. Szafert, L. B. Jerzykiewicz, P. Sobota, Inorg. Chem. 2005, 44, 5194.
[44] W. D. Buchanan, M. A. Guino-o, K. Ruhlandt-Senge, Inorg. Chem. 2010, 49, 7144.
[45] K. F. Tesh, T. P. Hanusa, J. C. Huffman, C. J. Huffman, Inorg. Chem. 1992, 31, 5572.
[46] G. B. Deacon, C. M. Forsyth, P. C. Junk, J. Organomet. Chem. 2000, 607, 112.
[47] W. A. Wojtczak, M. J. Hampden-Smith, E. N. Duesler, Inorg. Chem. 1996, 35, 6638.
[48] J. A. Darr, S. R. Drake, D. J. Williams, A. M. Z. Slawin, J. Chem. Soc., Chem Commun. 1990,
866.
[49] H. Roueindeji, A. Ratsifitahina, T. Roisnel, V. Dorcet, S. Kahlal, J.-Y. Saillard, J.-F. Carpentier,
Y. Sarazin, Chem. Eur. J. 2019, 25, 8854.
[50] S. C. Cole, M. P. Coles, P. B. Hitchcock, Organometallics 2004, 23, 5159.
[51] A.-F. Pécharman, M. S. Hill, M. F. Mahon, Angew. Chem. Int. Ed. 2018, 57, 10688.
[52] M. P. Coles, Coord. Chem. Rev. 2016, 323, 52.
[53] S. Al-Hashimi, J. D. Smith, J. Organomet. Chem. 1978, 153, 253.
[54] H. C. Brown, V. H. Dodson, J. Am. Chem. Soc. 1957, 79, 2302.
[55] H. Chen, P. P. Power, S. C. Shoner, Inorg. Chem. 1991, 30, 2884.
[56] V. C. Gibson, C. Redshaw, W. Clegg, M. R. J. Elsegood, Polyhedron 1997, 16, 2637.
[57] R. D. Chambers, T. J. Chivers, Chem. Soc. 1964, 4782.
[58] R. D. Chambers, T. J. Chivers, Chem. Soc. 1965, 3933.
[59] D. J. Parks, W. E. Piers, G. P. A. Yap, Organometallics 1998, 17, 5492.
[60] T. Beringhelli, G. D’Alfonso, D. Donghi, D. Maggioni, P. Mercandelli, A. Sironi,
Organometallics 2003, 22, 1588.
29
This article is protected by copyright. All rights reserved.