Challenges in the synthetic routes to Mn(BH4)2: Insight into intermediate compounds

Article abstract of DOI:10.1039/c4dt03807j

We have studied the reaction of MnCl₂ with MBH₄ (M = Li⁺, Na⁺, K⁺) in Et₂O. Crystal structures of two new intermediates, named [{M(Et₂O)₂}Mn₂(BH₄)₅] (M = Li⁺, Na⁺), were elucidated by X-ray diffraction. Mn(BH₄)₂ in a mixture with LiBH₄ or NaBH₄ forms upon the solvent removal in a vacuum. [{M(Et₂O)₂}Mn₂(BH₄)₅] contains 2D layers formed by Mn and BH₄ groups, linked through the alkali metal atoms coordinated to Et₂O. The loss of the solvent molecules leads to the segregation of the partially amorphous or nanocrystalline LiBH₄/NaBH₄ and a creation of the 3D framework of the crystalline Mn(BH₄)₂. While using LiBH₄ led to Mn(BH₄)₂ contaminated with LiCl, presumably due to an efficient trapping of the latter salt by the [Mn(BH₄)₂-Et₂O] system, the reaction with NaBH₄ produced chlorine-free Mn(BH₄)₂ accompanied with NaBH₄. Using KBH₄ led to the formation of K₂Mn(BH₄)₄ as a second phase. Two pyridine-containing solvomorphs, [Mn(py)₃(BH₄)₂] and [Mn(py)₄(BH₄)₂]·2py, were isolated in pure form. However, Mn(BH₄)₂ partly decomposes upon removal of pyridine molecules. This journal is

Full text of DOI:10.1039/c4dt03807j

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Challenges in the synthetic routes to Mn(BH4)2:
insight into intermediate compounds†
Cite this: Dalton Trans., 2015, 44,
6571
Nikolay A. Tumanov,a Damir A. Safin,a Bo Richter,b Zbigniew Łodziana,c
Torben R. Jensen,b Yann Garciaa and Yaroslav Filinchuk*a
We have studied the reaction of MnCl2 with MBH4 (M = Li+, Na+, K+) in Et2O. Crystal structures of two
new intermediates, named [{M(Et2O)2}Mn2(BH4)5] (M = Li+, Na+), were elucidated by X-ray diffraction.
Mn(BH4)2 in a mixture with LiBH4 or NaBH4 forms upon the solvent removal in a vacuum. [{M(Et2O)2}-
Mn2(BH4)5] contains 2D layers formed by Mn and BH4 groups, linked through the alkali metal atoms co-
ordinated to Et2O. The loss of the solvent molecules leads to the segregation of the partially amorphous
or nanocrystalline LiBH4/NaBH4 and a creation of the 3D framework of the crystalline Mn(BH4)2. While
using LiBH4 led to Mn(BH4)2 contaminated with LiCl, presumably due to an efficient trapping of the latter
salt by the [Mn(BH4)2–Et2O] system, the reaction with NaBH4 produced chlorine-free Mn(BH4)2
accompanied with NaBH4. Using KBH4 led to the formation of K2Mn(BH4)4 as a second phase. Two pyri-
dine-containing solvomorphs, [Mn(py)3(BH4)2] and [Mn(py)4(BH4)2]·2py, were isolated in pure form.
However, Mn(BH4)2 partly decomposes upon removal of pyridine molecules.
Received 11th December 2014,
Accepted 10th February 2015
DOI: 10.1039/c4dt03807j
www.rsc.org/dalton
nificant amounts of LiCl, limiting the use of the resulting
product.
Introduction
Borohydrides of alkali and alkali-earth metals are of interest as
Another route to Mn(BH4)2 is the metathesis reaction
hydrogen storage materials, and they are even more intriguing between MnCl2 and alkaline borohydrides in Et2O.13,14 In par-
as precursors to di-1 or even trimetallic borohydrides2 in order ticular, Schouwink et al. applied this method to obtain pure
to tune the temperatures of dehydrogenation.3,4 Up to now, Mn(BH4)2 by reacting MnCl2 and LiBH4 in Et2O (eqn (1)).14
monometallic borohydrides have been obtained by both
solvent based and mechanochemical syntheses,5 while mixed-
metal borohydrides have usually been produced using
mechanochemical synthesis.6 Very recently it was found that
the latter borohydrides can also be obtained in solutions.7
Although borohydrides have been synthesized and character-
ized for many metals,3,7–11 the synthesis of manganese boro-
hydride, Mn(BH4)2, obtained via ball-milling of MnCl2 with
LiBH4 in a 1 : 2 ratio, has been reported only recently.12 The
crystal structure of Mn(BH4)2 was successfully determined
from powder X-ray diffraction data. However, the suggested
mechanochemical approach suffers from the presence of sig-
MnCl2 þ 2LiBH4 ! MnðBH4Þ2 þ 2LiCl
ð1Þ
However, this approach was not reproducible – only one
synthesis was reported to be successful, whereas other
attempts failed due to the presence of LiCl in the final
product. The same method was earlier applied by Černý et al.
to obtain Mg(BH4)2, having a crystal structure similar to that of
Mn(BH4)2.15 However, the resulting product was also contami-
nated by LiCl, Li2MgCl4 and LiBH4. Pure Mg(BH4)2 was
obtained using the reaction of Et3NBH3 and MgH2.16,17
Because of the similarity between the structures, one may
attempt to obtain pure Mn(BH4)2 by following the synthetic
procedures to Mg(BH4)2. In particular, the latter borohydride
can be obtained by reacting MgBu2 either with Al(BH4)3
(eqn (2)) or BH3·Me2S (eqn (3)).18,19 Applying both synthetic
procedures to Mn(BH4)2 requires manganese dialkyl precur-
sors, which are either highly unstable under ambient con-
ditions20 and/or commercially unavailable. Furthermore, the
use of Al(BH4)3 is not practical because of its high instability
and explosiveness.
aInstitute of Condensed Matter and Nanosciences, Molecules, Solids & Reactivity
(IMCN/MOST), Université Catholique de Louvain, Place L. Pasteur 1, 1348 Louvain-
la-Neuve, Belgium. E-mail: yaroslav.filinchuk@uclouvain.be
bCenter for Materials Crystallography (CMC), Interdisciplinary Nanoscience Center
(iNANO), and Department of Chemistry, Aarhus University, Langelandsgade 140,
DK-8000 Århus C, Denmark
cINP Polish Academy of Sciences, Department of Structural Research, ul.
Radzikowskiego 152, 31-342 Kraków, Poland
†CCDC 985621–985623. For crystallographic data in CIF or other electronic
format see DOI: 10.1039/c4dt03807j
3MgBu2 þ 2AlðBH4Þ3 ! 3MgðBH4Þ2 þ 2AlBu3
ð2Þ
This journal is © The Royal Society of Chemistry 2015
Dalton Trans., 2015, 44, 6571–6580 | 6571

View Article Online
Paper
Dalton Transactions
3MgBu2 þ 8H3BꢀSMe2 ! 3MgðBH4Þ2 ꢀ2Me2S þ 2BBu3 ꢀSMe2
ð3Þ
Furthermore, for comparison with the light Et2O we also
studied the reaction between MnCl2 and NaBH4 in a highly
donor solvent, pyridine. Despite it allows one to isolate pure
complexes of Mn(BH4)2, its desolvation leads to a partial
decomposition of Mn(BH4)2.
With all these in mind, we have drawn our attention to
obtain pure Mn(BH4)2 by synthesizing first its solvate, followed
by removing solvent molecules from the structure. Unlike
Mn(BH4)2, solvates of Mg(BH4)2 have been extensively
studied.18,19,21–23 These results are of importance as a guide-
line to manganese-containing derivatives because of the struc-
Results and discussion
tural and property similarity of the corresponding analogues In agreement with Schouwink et al.,14 as well as it was also
of Mn2+ and Mg2+ (salts, borohydrides, etc.).24 Thus, we can established for Mg(BH4)2,29 our numerous experiments
suggest the same order of increasing donor strength of the revealed that the metathesis reaction between MnCl2 and
solvent molecules for Mn(BH4)2 as it was found for Mg(BH4)2: LiBH4 in Et2O (eqn (1)) was not reproducible to deliver pure
Et2O < THF < NH2tBu ≈ piperidine < pyridine < benzylamine Mn(BH4)2. Note, that this approach is based on the fact that
< DMSO.21 In order to facilitate solvent removal from the struc- both MnCl2 and LiCl are poorly soluble in Et2O, while
ture as well as to avoid decomposition of the final Mn(BH4)2, Mn(BH4)2 is efficiently soluble in this solvent.
the solvent molecules should be small and form weak inter-
Our results demonstrate the formation of Mn(BH4)2, con-
actions with the neighboring molecules in the structure. The taining about 35 wt% of LiCl upon reacting MnCl2 with LiBH4
last members of this series lack these properties and, hence, in Et2O. We believe that, although LiCl is extremely poorly
are unsuitable for our strategy. The most suitable solvents soluble in pure Et2O (Table 1), it can be efficiently trapped by
seem to be Et2O and THF.
the [Mn(BH4)2–Et2O] system, for example via formation of
Makhaev et al. reported the synthesis of Mn(BH4)2·THF, more complex intermediates, thus increasing the solubility.
NaBH4·Mn(BH4)2·3THF (solid), Mn(BH4)2·2THF (yellow oil) Another reason might come from the formation of very small
and Mn(BH4)2·3THF (solid).25 The structure of the latter particles of LiCl during the reaction, which cannot be filtered
complex was established. However, it was found that out even after using a PTFE membrane filter (22 µm pores).
Mn(BH4)2·3THF transforms to Mn(BH4)2·2THF, followed by the
Thus, using LiBH4, as a borohydride source for the syn-
complete decomposition of manganese borohydride during thesis of Mn(BH4)2 in Et2O, seems to be non-reliable due to
heating.26,27 This makes the THF solvates unsuitable for our the formation of LiCl, which significantly contaminates the
approach.
resulting product. With this in mind, we have focused on
In this work we report the synthesis of Mn(BH4)2 in Et2O alternative borohydride sources such as NaBH4 and KBH4.
using MnCl2 and MBH4 (M = Li+, Na+, K+) to examine the influ- Both precursors, as well as NaCl and KCl, are much less
ence of the nature of the alkali metal on the purity of the final soluble in Et2O compared to the corresponding lithium ana-
product. We also made a detailed study of intermediate com- logues (Table 1). Thus, using NaBH4 and KBH4 increased the
pounds formed during the reaction. Results on using an even reaction time significantly. It is worth noting that before our
smaller dimethyl sulfide, Me2S, as a solvent in synthesis of investigations LiBH4, being the most reactive alkali-metal
Mn(BH4)2 will be reported elsewhere.28 It should be noted that borohydride, was the only source of borohydride to produce
the main difference in using Et2O and Me2S as reaction media Mn(BH4)2. Thus, our studies on using both NaBH4 and KBH4
is that the former solvent contains the “hard” oxygen donor to react with MnCl2 in solutions were challenging. However,
atom, which is more suitable to efficiently solvate “hard” one would expect similar reaction properties of NaBH4 and
metal cations such as Li+, Na+ and K+. The latter solvent, con- KBH4 as for LiBH4 to produce Mn(BH4)2.
taining the “soft” sulfur donor atom, would efficiently solvate
Slow evaporation of the solvent from the reaction mixture of
“soft” d-metal cation, viz. Mn2+. Herein we present our results MnCl2 with MBH4 (M = Li+, Na+, K+) in Et2O afforded yellow
using the first, Et2O-based, synthetic approach. This in-depth oil, which might be the result of the formation of mono- and/
study will open an avenue for the synthesis of Mn(BH4)2. or polymetallic borohydrides with a high content of Et2O. The
Table 1 Solubility of the reported compounds at 25 °Ca
LiCl
KCl
LiBH4
NaBH4
KBH4
MnCl2
Et2O
0.77 mg/100 g30
0.6 mg/100 g31
2.7 g/100 g31
—
1.4 M,31 7.53 wt%32
Insoluble,31,34 0.02 g/100 g35
Insoluble6
Poorly soluble36
4.5 wt%33
Dioxane
Pyridine
—
—
—
—
—
—
Poorly soluble36
—
14.0 g/100 g37
10.0 g/100 g38
4.3 g/100 g40
0.002 g l−1 (ref. 38)
3.4 g/100 g39
THF
—
11.43 M,31 22.2 g/100 g41
0.07 g/100 g,31 0.1 g/100 g39
0.04 wt%42
Insoluble6
Soluble36
a No data were found for NaCl.
6572 | Dalton Trans., 2015, 44, 6571–6580
This journal is © The Royal Society of Chemistry 2015

View Article Online
Dalton Transactions
Paper
solvent molecules might both coordinate to the metal centers Mn2+ atoms coordinates four tetrahedral BH4 anions via
and serve as lattice solvents trapped within voids upon crystal- edges with the formation of a slightly distorted Mn(BH4)4 tetra-
lization. Makhaev et al. reported that transition metal boro- hedron. In the Li-containing compounds, the orientation of
−
−
hydrides form solvates with polyethers of the general formula the BH4 groups cannot be reliably determined from the
NaMn+(BH4)n+1·xSolv (M = V, Fe, Mn; Solv = dimethoxyethane, powder X-ray diffraction data. However, besides the orientation
−
bis(2-methoxyethyl) ether, THF).43,44
of the BH4 groups, other distances in the lithium structure
All our attempts to crystallize the collected oils in a glass can be determined quite precisely. The Mn–B bond distances
capillary by cooling down to 100 K have failed. Further slow are 2.280(7)–2.581(7) Å [2.329–2.497 Å; hereinafter the bond
drying of these oils under vacuum yielded yellow crystalline distances for the DFT-optimized structures are given in the
materials. Crystals obtained in NaBH4-based synthesis were brackets] and 2.403(11)–2.475(9) [2.383–2.452] Å in the Li- and
suitable for single-crystal X-ray diffraction. According to single- Na-based structures, respectively. These values are comparable
crystal and powder X-ray diffraction data, materials formed with those for pure Mn(BH4)2 and other compounds contain-
−
from the LiBH4- and NaBH4-based syntheses are the corres- ing Mn2+ and BH4 fragments. Li+ or Na+ atoms are co-
−
ponding dimetallic borohydride solvates [{M(Et2O)2}- ordinated via edges to two BH4 groups with the Li–B and
Mn2(BH4)5] (M = Li+, sample Li_solv; Na+, sample Na_solv). Na–B distances being 2.555(6) [2.582–2.566] and 2.700(11)
The bulk material of the Li_solv sample also contained LiCl. [2.748] Å, respectively. Tetrahedral coordination spheres of Li+
Further slow drying of these two samples led to the decompo- and Na+ are further filled with two oxygen atoms from two
sition of dimetallic borohydrides with the formation of Et2O molecules with the Li–O and Na–O bond lengths
Mn(BH4)2 and the corresponding MBH4 (M = Li+, sample 2.0030(14) [1.995–2.003] and 2.316(10) [2.377] Å, respectively.
Li_dry; Na+, sample Na_dry). However, only peaks for In both compounds, the BH4 group serves as a bridging
−
Mn(BH4)2 and LiCl, and Mn(BH4)2 and traces of NaBH4 were ligand with respect to the two metal atoms, as in the solvent-
found in the powder X-ray diffraction patterns of Li_dry and free bimetallic borohydrides of transition metals (Zn2+, Mn2+,
Na_dry, respectively. This might be explained by the complete, Sc3+, Cd2+, Y3+).1,14,45–50
for M = Li+, and partial for M = Na+, amorphization of the
Since the positions of the hydrogen atoms are difficult to
released MBH4, as well as by the very small size of the segregat- determine unambiguously from X-ray diffraction, we made
ing MBH4 particles. The KBH4-based synthesis produced a DFT calculations on the experimental structures, clarifying the
−
mixture of Mn(BH4)2 and K2Mn(BH4)4, with the latter com- coordination modes of BH4 groups. Both compounds are
pound being a minor product (∼4 wt%) (sample K_dry). K2Mn- insulators with a band gap of 2.1 and 1.9 eV for [{Li(Et2O)2}-
(BH4)4 has been previously obtained from the mechanochem- Mn2(BH4)5] and [{Na(Et2O)2}Mn2(BH4)5], respectively. Opti-
ical synthesis.14 Thus, although 25% excess of MnCl2 was used mized structures clearly show bidentate coordination of all
−
in the reactions with MBH4 (M = Li+, Na+, K+), a strong affinity BH4 groups with respect to metal atoms in both compounds.
to form bimetallic structures was observed in all three cases.
The symmetry of the DFT-optimized structure of [{Li(Et2O)2}-
According to X-ray diffraction data, [{Li(Et2O)2}Mn2(BH4)5] Mn2(BH4)5] is lower, the space group is Cc instead of C2/c. This
and [{Na(Et2O)2}Mn2(BH4)5] crystallize in the monoclinic space happens due to small displacements of atoms, which do not
groups C2/c and C2, respectively (Table 2). The volume of the change significantly the entire structure. Rietveld refinement
unit cell differs by roughly a factor of two; however, the topo- in the space group Cc gives a good fit, but involves nearly twice
logy of the two structures is identical. Reconstruction of the as many refined parameters. Therefore, for the experimental
reciprocal space sections for [{Na(Et2O)2}Mn2(BH4)5] did not structure we finally decided to keep the space group C2/c.
show cell doubling in any direction, while the powder data for
The Mn-based tetrahedra form 2D layers in the bc and ac
the Li-containing analogue clearly reveal superstructure peaks. planes for [{Li(Et2O)2}Mn2(BH4)5] and [{Na(Et2O)2}Mn2(BH4)5],
The asymmetric unit in both compounds contains one Mn2+ respectively (Fig. 1). These 2D layers are linked through the Li-
cation, two BH4− anions in general positions and one on the 2 or Na-containing tetrahedra producing 3D structures (Fig. 1).
axis, one Et2O molecule in a general position, and Li+ or Na+ Molecules of Et2O are located between the 2D layers formed by
cations in a special position. In both structures, each of the Mn-based polyhedra (Fig. 1), which could explain the relatively
Table 2 Unit cell parameters of the reported compounds
[{Li(Et2O)2}Mn2(BH4)5]
[{Na(Et2O)2}Mn2(BH4)5]
[Mn(py)3(BH4)2]
[Mn(py)4(BH4)2]·2py
[Mg(py)4(BH4)2]·2py22
Space group
Z
C2/c
4
C2
2
C2/c
4
Ccca
4
Ccca
4
a/Å
b/Å
c/Å
21.7106(3)
7.72944(12)
14.02422(19)
110.8649(14)
2199.09(6)
985621
14.5133(6)
7.4487(3)
10.7119(5)
104.338(5)
1121.96(9)
985622
12.3618(4)
15.5064(5)
9.2859(4)
101.126(4)
1746.53(11)
985623
11.516(2)
14.938(4)
17.263(8)
90
11.604(2)
15.114(2)
17.190(2)
90
3014.7(7)
186648
β/°
V/Å3
CCDC
2970(2)
This journal is © The Royal Society of Chemistry 2015
Dalton Trans., 2015, 44, 6571–6580 | 6573

View Article Online
Paper
Dalton Transactions
Li-based polyhedra alternate with two different orientations,
which results in the double cell for this compound, compared
to the Na-containing analogue. But even taking into account
this difference, the structures remain topologically similar. A
2D honeycomb topology of the Mn-based tetrahedra suggests
that the magnetic ordering might be complex at low tempera-
tures.51 But magnetic ordering cannot be directly detected by
X-ray diffraction, and it is not present at the temperatures of
interest – thus it is unrelated to the present study.
We suggest that evaporation of the molecules of Et2O
breaks the coordination environment around the Li+ or Na+
atoms, leading to a collapse of the structure with the formation
of Mn(BH4)2 and corresponding MBH4 (M = Li+, Na+).
The Raman spectra of Li_solv, Li_dry, Na_dry and K_dry
−
(Fig. 2) contain characteristic bands for the BH4 groups, seen
also in the structure of Mn(BH4)2.16 Furthermore, the Raman
spectrum of Li_solv exhibits a set of characteristic bands for
the Et2O molecules: medium bands at 1360 and 1450 cm−1 as
well as number of intense bands at 2820–3020 cm−1, corres-
ponding to the CH vibrations of the ethyl groups; and a weak
band at about 1090 cm−1, which is typical for the COC asym-
metric stretches. It should be noted that no bands for the Et2O
molecules were found in the Raman spectra of Li_dry, Na_dry
and K_dry testifying the complete evaporation of the solvent
molecules.
The thermal properties of Li_solv, Li_dry, Na_dry and K_dry
under an argon atmosphere were studied by means of the TG
analysis (Fig. 3). Li_solv is stable up to about 35 °C and decom-
poses in three clearly (despite the mass drift) defined steps.
The first remarkable step up to about 110 °C corresponds to
the evolution of Et2O. The second mass loss up to about
160 °C is due to the decomposition of Mn(BH4)2.28,52 It should
be noted that this decomposition step was found in the TG
data of all compounds (Fig. 3). However, the temperature-
dependent in situ powder X-ray diffraction data, collected for
Li_solv, exhibit peaks exclusively for LiCl after about 105 °C
accompanied with the new very broad humps due to the for-
mation of the either amorphous or liquid phase containing
Fig. 1 The 2D layer formed by the MnB4-based coordination tetrahedra
in the structures of [{M(Et2O)2}Mn2(BH4)5] (M = Li+, Na+) (a). Crystal
packing of [{Li(Et2O)2}Mn2(BH4)5] (b) and [{Na(Et2O)2}Mn2(BH4)5] (c).
Polyhedra around the Mn atoms are shown in yellow, while those
around the Li and Na atoms are given in blue. Hydrogen atoms were
omitted.
easy release of the solvate molecules from the structures in a
vacuum. The 2D layers formed by Mn-based polyhedra are
almost the same in both structures, but the layers of alkali
metal based polyhedra are different. In the structure of [{Na-
(Et2O)2}Mn2(BH4)5, all Na-based polyhedra are oriented in the
same way, while in the structure of [{Li(Et2O)2}Mn2(BH4)5], the
Fig. 2 Raman spectra of Li_solv, Li_dry, Na_dry and K_dry. The spec-
trum of Na_solv is similar to that of Li_solv.
6574 | Dalton Trans., 2015, 44, 6571–6580
This journal is © The Royal Society of Chemistry 2015

View Article Online
Dalton Transactions
Paper
Fig. 4 In situ synchrotron X-ray diffraction on Li_solv measured from
30 to 130 °C (λ = 0.82742 Å). [{Li(Et2O)2}Mn2(BH4)5] decomposes at
about 105 °C. Peaks from the LiCl phase are marked with asterisks.
dissolved Mn(BH4)2 and LiBH4 (Fig. 4). The observed discre-
pancies between the TG and powder X-ray diffraction data
might be explained by different experimental conditions. In
particular, the latter experiments were done in a capillary
sealed by wax, which can be considered as a “closed” system,
while TG studies were performed in a crucible under a flow of
argon, thus, in an “open” system. The latter system favors an
efficient removal of the evaporated Et2O, while it is blocked in
the former system. This, in turn, leads to dissolving of the crys-
talline borohydride residues, released during the decompo-
sition (Table 1). However, LiCl remains intact under applied
experimental conditions due to its, first, poor solubility in
Et2O (Table 1) and, second, high thermal stability.
The TG for Li_dry is very similar to that of Li_solv after the
latter compound completely losses Et2O (Fig. 3). The TG data
for Na_dry and K_dry, with the first step of mass loss due to
the decomposition of Mn(BH4)2, show the same features as
Li_dry (Fig. 3). However, since the sample K_dry contains also
K2Mn(BH4)2, its TG is, obviously, complicated by the decompo-
sition pathway of this bimetallic borohydride.14
Various solvates of the same borohydride could be used to
produce new polymorphs. In particular, porous polymorphs,
which are even more promising for gas storage. For example,
porous γ-Mg(BH4)2 was successfully obtained from the
Mg(BH4)2·0.5Me2S solvate.19 However, the formation of a
stable borohydride from its solvate requires that the parent
structure comprises a rigid 3D framework formed by coordi-
nation polyhedra around the metal atoms, or at least the pres-
ence of rigid 3D fragments. As we described above, both
dimetallic solvates exhibit exclusively 2D Mn-BH4 layers. The
loss of the solvent molecules leads to the segregation of the
partially amorphous or nanocrystalline LiBH4/NaBH4 and the
Fig. 3 TG analyses of Li_solv, Li_dry, Na_dry and K_dry performed
under a dynamic argon atmosphere. The TG data for Na_solv are similar
to those of Li_sov. All mass loss steps were found to be endothermic.
This journal is © The Royal Society of Chemistry 2015
Dalton Trans., 2015, 44, 6571–6580 | 6575

View Article Online
Paper
Dalton Transactions
Fig. 6 Molecular structure of [Mn(py)4(BH4)2]·2py.
Fig. 5 Molecular structure of [Mn(py)3(BH4)2].
pyridine molecules (Fig. 6). The Mn2+ cation coordinates four
pyridine molecules and two BH4 anions forming a distorted
creation of the new 3D framework of the crystalline Mn(BH4)2.
The behaviour of the mononuclear pyridine complexes of
Mn(BH4)2, obtained in this work, illustrates that the formation
of stable borohydrides from their solvates is difficult in the
absence of rigid metal-borohydride frameworks present
already in the parent structure.
−
octahedron.
Both pyridine-containing solvomorphs contain mono-
nuclear Mn complexes, which are well isolated from each
other. The departure of the solvent molecules leads to a com-
plete structural rearrangement accompanied by Mn(BH4)2
decomposition and gas release. The structures contain no
polymeric chains to form relatively stable frameworks, and
they turned out to be unsuitable as precursors for obtaining
neither porous nor even dense Mn(BH4)2 polymorphs.
However, both pyridine complexes described in this work, as
well as the previously reported magnesium analogue,22 are an
intriguing starting point for the possible formation of porous
metal organic borohydride frameworks using polypyridine
ligands as organic linkers aiming for different topologies and
porosity, like for example described in the work of Ingleson
et al.54
The reaction of MnCl2 with NaBH4 in pyridine produces at
least two solvates, [Mn(py)3(BH4)2] and [Mn(py)4(BH4)2]·2py.
According to single-crystal X-ray diffraction data the former
complex crystallizes in the monoclinic C2/c space group and
exhibits a heteroleptic mononuclear complex (Fig. 5 and
−
Table 2). The asymmetric unit contains one BH4 anion, one
molecule of pyridine in general and one in special positions,
and one Mn2+ cation on the 2-fold axis, which results in a
molecular complex with the C2 symmetry. The Mn2+ atom
−
coordinates two tetrahedral BH4 anions via the edges at Mn–
B distances of 2.483(2) Å. The coordination sphere of the
metal center is completed by three nitrogen atoms from three
pyridine molecules. The Mn–N bond distances are 2.246(3)
and 2.2823(19) Å. The coordination environment of Mn can be
described either as square-pyramidal or trigonal bipyramidal
depending on the parameter τ = (β − α)/60, where α and β are
the two largest bond angles around the metal ion. An ideal
square pyramidal arrangement is described by the value of τ =
0, while an ideal trigonal bipyramidal arrangement has the
value of τ = 1.53 The largest bond angles in the coordination
sphere of Mn2+ are 175.92(8) and 127.72(11)°. This gives
τ value of 0.8, which is significantly closer to that of the ideal
trigonal bipyramidal around the metal atom. The equatorial
Conclusions
In summary, we have studied the synthetic pathway to
Mn(BH4)2 by reacting MnCl2 and alkali metal borohydrides in
diethyl ether. The two new solid intermediates [{M(Et2O)2}-
Mn2(BH4)5] (M = Li+, Na+) were isolated and characterized. We
found that pure Mn(BH4)2 could not be obtained by this
approach because of the inclusion of LiBH4 or NaBH4 leading
to the stable bimetallic solvates that were isolated from the
reaction mixture. Moreover, using LiBH4 is less favorable due
to the efficient trapping of LiCl by the [Mn(BH4)2–Et2O]
system. Using NaBH4 allows one to obtain Mn(BH4)2 with
some of the starting borohydride, while KBH4 leads to
Mn(BH4)2 contaminated by K2Mn(BH4)4.
We have also obtained two new pyridine solvates of
Mn(BH4)2, by reacting MnCl2 with NaBH4 in pyridine. Both
solvomorphs consist of mononuclear complexes of Mn, that
are not suitable as precursors for pure Mn(BH4)2, as it partly
decomposes upon removal of pyridine molecules.
−
plane of the bipyramid is formed by both BH4 anions and
one of the pyridine molecules.
Unfortunately, we were unable to obtain single crystals of
[Mn(py)4(BH4)2]·2py suitable for structure determination.
However, we were successful in determining the space group
and the unit cell parameters. Based on the obtained data, we
suggest that [Mn(py)4(BH4)2]·2py is isostructural to the pre-
viously reported complex [Mg(py)4(BH4)2]·2py21 (Table 2).
Thus, the manganese derivative comprises a discrete hetero-
leptic mononuclear complex [Mn(py)4(BH4)2] and two lattice
6576 | Dalton Trans., 2015, 44, 6571–6580
This journal is © The Royal Society of Chemistry 2015

View Article Online
Dalton Transactions
Paper
powder X-ray diffraction pattern did not exhibit peaks for
LiBH4, which is, probably, due to the amorphization of the
latter compound or due to very small size of its particles.
Reaction of MnCl2 with NaBH4 in Et2O. A pink suspension
Experimental
Materials
All experiments were performed under anhydrous conditions
using Schlenk techniques and argon as a protecting gas. All of MnCl2 (2.52 g, 20.0 mmol) and NaBH4 (1.21 g, 32.0 mmol)
materials were used as received and without further purifi- in Et2O (120 mL) was vigorously stirred for several days until a
cation: MnCl2 (Sigma-Aldrich, anhydrous, 99.99%), LiBH4 clear yellow solution with white precipitate appeared. The reac-
(Sigma-Aldrich, ≥95%), NaBH4 (Alfa Aesar, ≥97%), (Sigma- tion proceeded much slower than the analogous reaction with
Aldrich, ≥99.9%), Et2O (VWR, H2O < 11 ppm), pyridine LiBH4 due to the lower solubility of NaBH4 in Et2O. The result-
(Sigma-Aldrich, anhydrous, 99.8%).
ing mixture was filtered using a microporous glass filter. The
filtrate was slowly concentrated and the formation of pale
yellow oil was observed. Further slow drying of this oil in a
Analytical methods
Raman spectra were recorded using a FTIR Nicolet Magna 860 vacuum at room temperature resulted in yellow crystals
(sample Na_solv in Table 3), according to the single-crystal
with a Raman unit and a Nd:YVO4 (λ = 1064 nm) laser.
Samples were placed in glass capillaries under an inert atmos- and powder X-ray diffraction data, of [{Na(Et2O)2}Mn2(BH4)5].
phere and sealed with vacuum grease. Thermogravimetric ana- The solid material was further dried in a vacuum at room
temperature for about 2.5 days. The resulting pale yellow
lysis was done using a SDT 2960 Simultaneous DTA-TGA
instrument under a dynamic argon atmosphere (100 mL powder (sample Na_dry in Table 3) was characterized, accord-
min−1) from ambient temperature to 600 °C with a 5 °C min−1 ing to the powder X-ray diffraction data, as a mixture of Mn
(BH4)2 (94 wt%) and NaBH4 (6 wt%). Based on the compo-
sition of the initial sample, the final product should contain
about 30.9 wt% of NaBH4. The observed discrepancy is, most
likely, due to a partial amorphization of NaBH4 or due to very
small size of its particles.
heating rate.
Synthesis
Reaction of MnCl2 with LiBH4 in Et2O. A pink suspension
of MnCl2 (2.52 g, 20.0 mmol) and LiBH4 (0.70 g, 32.0 mmol) in
Et2O (120 mL) was vigorously stirred for about 24 h. The result-
ing mixture was filtered using first a microporous glass filter
and then the PTFE membrane filter (22 µm pores). The filtrate
was slowly (∼2 h) concentrated in a vacuum at room tempera-
ture. During concentration, a fine crystalline white powder of
LiCl was precipitated alongside with the formation of pale
yellow oil. Further concentration yielded a pale yellow solid
material (sample Li_solv, sample designations and phase
content are summarized in Table 3), which was characterized,
according to the powder X-ray diffraction data, as a mixture of
[{Li(Et2O)2}Mn2(BH4)5] (79.4 wt%) and LiCl (20.6 wt%). The
solid material was further dried in a vacuum at room tempera-
ture for about 2.5 days. The resulting pale yellow powder
(sample Li_dry in Table 3) was characterized, according to the
powder X-ray diffraction data, as a mixture of Mn(BH4)2
(65 wt%) and LiCl (35 wt%). Based on the composition of the
initial sample, the final product should contain Mn(BH4)2
(61.5 wt%), LiCl (30.7 wt%) and LiBH4 (7.9 wt%). However, the
Reaction of MnCl2 with KBH4 in Et2O. A pink suspension of
MnCl2 (2.52 g, 20.0 mmol) and KBH4 (1.73 g, 32.0 mmol) in
Et2O (120 mL) was vigorously stirred for several days until a
clear yellow solution with white precipitate appeared. The reac-
tion proceeded much slower than the analogous reaction with
LiBH4 due to the lower solubility of KBH4 in Et2O. The result-
ing mixture was filtered using a microporous glass filter. The
filtrate was slowly concentrated and the formation of pale
yellow oil was observed. Further slow drying of this oil in a
vacuum at room temperature resulted in a pale yellow solid,
which was further dried in a vacuum at room temperature for
about 2.5 days. The resulting pale yellow powder (sample
K_dry in Table 3) was characterized, according to the powder
X-ray diffraction data, as a mixture of Mn(BH4)2 (96 wt%) and
K2Mn(BH4)4 (4 wt%).
Reaction of MnCl2 with NaBH4 in pyridine. A dark brown
suspension of MnCl2 (2.52 g, 20.0 mmol) and NaBH4 (1.21 g,
32.0 mmol) in pyridine (py, 120 mL) was vigorously stirred for
two days. The resulting mixture was filtered using the PTFE
membrane filter (22 µm pores). The filtrate was slowly concen-
trated in a vacuum at room temperature until colorless crystals
were formed. These crystals were stable in the mother liquor at
room temperature for few hours, but quickly decomposed at
room temperature when isolated from the solution, even
under inert conditions. At least two pyridine-containing com-
pounds, [Mn(py)3(BH4)2] and [Mn(py)4(BH4)2]·2py, were
obtained, according to the single-crystal X-ray diffraction data.
Table 3 Phase composition of samples
wt%,
experimental
X-ray data
wt%, calculated
from the initial
composition
Sample
name
Phases
Li_solv
Li_dry
[{Li(Et2O)2}Mn2(BH4)5]
LiCl
Mn(BH4)2
79.4
20.6
65
—
61.5
7.9
LiBH4
0
LiCl
35
100
94
6
96
30.7
—
69.1
30.9
—
Single crystal X-ray crystallography
Na_solv
Na_dry
[{Na(Et2O)2}Mn2(BH4)5]
Mn(BH4)2
NaBH4
Mn(BH4)2
K2Mn(BH4)4
The X-ray data were collected at 150(2) K on a MAR345 image
plate detector using Mo Kα radiation (Rigaku UltraX 18
rotation anode, Xenocs Fox3D focusing multilayer mirror). The
K_dry
4
This journal is © The Royal Society of Chemistry 2015
Dalton Trans., 2015, 44, 6571–6580 | 6577

View Article Online
Paper
Dalton Transactions
data were integrated using the CrysAlisPro software.55 The tations since the symmetry of the BH4 group is incompatible
implemented multi-scan absorption correction was applied. with an inversion center. The orientation of this group was
The structures were solved by direct methods using the constrained using distance restrains to make all Mn–H dis-
SHELXS-9756 program and refined by full-matrix least squares tances equal. Overall, 52 distance constraints and 25 angle
on |F|2 using SHELXL-97.56 Non-hydrogen atoms were refined constraints were applied. The second phase, LiCl, was
anisotropically, while hydrogen atoms of Et2O were localized modeled, with a final weight fraction of 20.54(10)%. The back-
from a differential Fourier map, but were placed at calculated ground was described by a linear interpolation of the selected
positions in a riding mode with temperature factors fixed at points. The final discrepancy factors were: RB = 5.05%, RF
=
1.2 times Ueq of the parent atoms and 1.5 times Ueq for the 6.30%, Rp = 10.52% and Rwp = 11.29%. The refinement profile
methyl groups. The B–H and H–H distances in the BH4 groups is shown in Fig. 7.
were constrained to form ideal tetrahedrons. Inspection of the
differential Fourier map revealed the presence of positional
disorder in Et2O. Attempts to model this disorder with several
positions of the Et2O molecule resulted in an insignificant
improvement of the model, so we modeled the disorder with
one position, but with higher atomic displacements.
Laboratory powder X-ray crystallography
Powder X-ray diffraction data were collected on a MAR345
image plate detector using Mo Kα radiation (Rigaku UltraX 18
rotation anode, Xenocs Fox3D focusing multilayer mirror). The
two-dimensional diffraction images were integrated using the
ESRF Fit2D56 program. The integrated patterns were treated
using the Fullprof suite software.60
Synchrotron powder X-ray crystallography
The synchrotron radiation powder X-ray diffraction data were
collected at the Swiss–Norwegian beamline BM1A at the Euro-
pean Synchrotron Radiation Facility (ESRF) (Grenoble, France)
using a PILATUS 2M pixel detector. The wavelength was
0.82742 Å, and the sample-to-detector distance was 430 mm.
These parameters along with image plate tilt angles were cali-
brated using a standard LaB6 sample. The Li_solv sample was
sealed in a thin-walled glass capillary under an argon atmos-
phere and measured at 100(2) K. Two-dimensional diffraction
images from this sample were integrated using the ESRF Fit2D
program.57 The pattern was indexed using DICVOL0458 in a
centered monoclinic crystal system. Analysis of systematic
absences suggested the C2/c and Cc space groups. Based on
the assumed composition, Mn(BH4)2·x(C2H5)2O, and an
approximate volume of the Mn(BH4)2 unit being 113 Å3, and
the volume of the Et2O molecule being 125 Å3, we suggested
the composition of 3Mn(BH4)2·2Et2O. However, attempts to
solve the structure applying this composition by global optim-
ization in direct space in the FOX59 program failed. The occu-
pancy of all but one Mn atoms was close to 0. Therefore, we
suggested the presence of Li atoms along with Mn in the struc-
ture. The structure was finally solved and refined in the C2/c
space group with a composition of [{Li(Et2O)2}Mn2(BH4)5]. In
the final refinement, antibump restraints were 3 Å for the O–O
and B–O contacts, 1.8 Å for O–H, 2.8 Å for B–C and Mn–C,
2.2 Å for H–H, and of 3.3 Å for B–B distances. The final refine-
ment was done with the C2/c space group using the Rietveld
method in the Fullprof suite software.60 Coordinates of the
Et2O molecule were refined, with its orientation and configur-
ation being fixed. The bond lengths in the Et2O and BH4
groups were constrained to be close to the following values:
d(C–C) 1.50 Å, d(C–O) 1.425 Å, d(C–H)CH3 0.97 Å, d(C–H)CH2
0.96 Å, d(B–H) 1.13 Å, d(H–H)BH4 1.841 Å. The bond angles
were constrained using soft distance and angle constraints.
The torsion angles of the CH3 groups in Et2O were constrained
using distance constraints to prevent an eclipsed confor-
mation. One of the BH4 groups was found on the inversion
DFT calculations
The structures were optimized by a conjugated gradient
method within plane wave formulation of spin polarized
density functional theory (DFT).61 The electronic configuration
of 1s22s1 for Li, 2p63s1 for Na, 3d64s1 for Mn, 2s22p2 for C,
2s22p4 for O, 2s22p1 for B, and 1s1 for H was represented by
projected augmented wave potentials.62 The gradient corrected
(GGA) functional63 was applied. The initial structure from the
Rietveld refinement was optimized with respect to the internal
atomic positions, the unit cell shape and volume were fixed.
The on-site Coulomb repulsion64 U = 4 was applied to the
manganese d electrons. Since the manganese atoms are
arranged in honeycomb-like 2D networks (Fig. 1) the ground
state magnetic ordering could be nontrivial. This is despite the
fact of obvious antiferromagnetic superexchange interaction
−
between each adjacent Mn2+ linked by the BH4 anion. For
example the lowest energy state with a collinear antiferro-
Fig. 7 Rietveld refinement plot for [{Li(Et2O)2}Mn2(BH4)5]. Red crosses
and black line show the experimental and calculated data, respectively.
Blue line is the difference profile, and green marks indicate Bragg
center, and this group was disordered over two 50/50% orien- positions.
6578 | Dalton Trans., 2015, 44, 6571–6580
This journal is © The Royal Society of Chemistry 2015

View Article Online
Dalton Transactions
Paper
magnetic arrangement of sub-layers with up and down Nature and Universe (Danscatt). We acknowledge funding
oriented BH4− tetrahedra is only 110 and 118 meV per formula from the European Community’s Seventh Framework Pro-
unit more stable than the ferromagnetic one for [{Li(Et2O)2}- gramme, the Fuel Cells and Hydrogen Joint Undertaking (FCH
Mn2(BH4)5] and [{Na(Et2O)2}Mn2(BH4)5], respectively. More- JU), project BOR4STORE (303428). We are grateful to the Carls-
over, it does not affect the structural configuration of the light berg Foundation for support. Z. Łodziana acknowledges CPU
atoms in the lattice. Noncollinear ordering of the magnetic allocation at the PLGrid Infrastructure and support by a grant
moments requires larger superstructures, but around the from Switzerland through the Swiss Contribution to the
room temperature they will not be present. We have checked enlarged European Union.
that the total ground state energy for noncollinear spin
arrangement varies on the scale of single or tenth of meV.
Interestingly, in both compounds Mn formally is in the oxi-
dation state 2+, however it is tetrahedrally coordinated by four
−
BH4 anions. This coordination environment enforces a high
spin arrangement of the Mn d-electrons. The calculated mag-
Notes and references
netic moments of Mn are 4.33μB for the ferromagnetic, and
4.29 or 4.28μB for the antiferromagnetic orders in [{Li-
(Et2O)2}Mn2(BH4)5] and [{Na(Et2O)2}Mn2(BH4)5], respectively.
In order to avoid complications related to the fact that numeri-
cal values of the magnetic moments depend on particular
choice of the Hubbard U,65 the possibility of induction of the
dipole moments/ferroelectricity due to noncentrosymmetry,
noncollinear spin arrangements on the honeycomb lattices51
overall resulting in exotic magnetic configurations, the anti-
ferromagnetic ordering of Mn magnetic moments, as
described above, was assumed. The crystal structures are
almost the same regardless of the ferro and antiferro spin
ordering. Each system was relaxed until forces exerted on
atoms were below 0.01 eV A−1. For [{Li(Et2O)2}Mn2(BH4)5] with
1 D. Ravnsbaek, Y. Filinchuk, Y. Cerenius, H. J. Jakobsen,
F. Besenbacher, J. Skibsted and T. R. Jensen, Angew. Chem.,
Int. Ed., 2009, 48, 6659.
2 R. Černý, P. Schouwink, Y. Sadikin, K. Stare, L. Smrčok,
B. Richter and T. R. Jensen, Inorg. Chem., 2013, 52, 9941.
3 L. H. Rude, T. K. Nielsen, D. B. Ravnsbaek, U. Bösenberg,
M. B. Ley, B. Richter, L. M. Arnbjerg, M. Dornheim,
Y. Filinchuk, F. Besenbacher and T. R. Jensen, Phys. Status
Solidi A, 2011, 208, 1754.
4 Y. Nakamori, K. Miwa, A. Ninomiya, H. Li, N. Ohba,
S. Towata, A. Züttel and S. Orimo, Phys. Rev. B: Condens.
Matter, 2006, 74, 045126.
5 H. Hagemann and R. Cerný, Dalton Trans., 2010, 39, 6006.
6 J. Huot, D. B. Ravnsbæk, J. Zhang, F. Cuevas, M. Latroche
and T. R. Jensen, Prog. Mater. Sci., 2013, 58, 30.
7 T. Jaroń, P. A. Orłowski, W. Wegner, K. J. Fijałkowski,
P. J. Leszczyński and W. Grochala, Angew. Chem., Int. Ed.,
2015, 54, 1236.
−
a static disorder of BH4 groups two different ordered con-
−
figurations of BH4 were assumed. The above mentioned
assumptions are sources of errors; however we do not expect
them to be larger than ∼0.1 eV per formula unit, and they are
of minor importance for the crystal structure.
CCDC 985621 ([{Li(Et2O)2}Mn2(BH4)5]), 985622 ([{Na-
(Et2O)2}Mn2(BH4)5]) and 985623 ([Mn(py)3(BH4)2]) contain the
supplementary crystallographic data. DFT optimized structures
of [{Li(Et2O)2}Mn2(BH4)5] and [{Na(Et2O)2}Mn2(BH4)5] are avail-
able as ESI.†
8 Y. Filinchuk, D. Chernyshov and V. Dmitriev, Z. Kristallogr.,
2008, 223, 649.
9 M. B. Ley, L. H. Jepsen, Y.-S. Lee, Y. W. Cho, J. M. Bellosta
von Colbe, M. Dornheim, M. Rokni, J. O. Jensen, M. Sloth,
Y. Filinchuk, J. E. Jørgensen, F. Besenbacher and
T. R. Jensen, Mater. Today, 2014, 17, 122.
10 F. Pendolino, J. Phys. Chem. C, 2012, 116, 1390.
11 L. H. Jepsen, M. B. Ley, Y.-S. Lee, Y. W. Cho, M. Dornheim,
J. O. Jensen, Y. Filinchuk, J. E. Jørgensen, F. Besenbacher
and T. R. Jensen, Mater. Today, 2014, 17, 129.
Acknowledgements
The authors acknowledge FNRS (CC 1.5169.12, PDR T.0169.13, 12 R. Černý, N. Penin, H. Hagemann and Y. Filinchuk, J. Phys.
EQP U.N038.13) for financial support. We acknowledge the Chem. C, 2009, 113, 9003.
Fonds Spéciaux de Recherche (UCL) for the incoming post- 13 G. Monnier, Ann. Chim., 1957, 13, 14.
doctoral fellowship co-funded by the Marie Curie actions of 14 P. Schouwink, V. D’Anna, M. B. Ley, L. M. Lawson Daku,
the European Commission granted to N. A. Tumanov and WBI
(Belgium) for the postdoctoral position allocated to D. A. Safin.
B. Richter, T. R. Jensen, H. Hagemann and R. Černý,
J. Phys. Chem. C, 2012, 116, 10829.
We thank ESRF for the beamtime allocation at the SNBL and 15 R. Cerný, Y. Filinchuk, H. Hagemann and K. Yvon, Angew.
V. Dyadkin for help, and A. Railliet for the help with thermo- Chem., Int. Ed., 2007, 46, 5765.
gravimetric measurements. Part of this work was supported by 16 Y. Filinchuk, R. Černý and H. Hagemann, Chem. Mater.,
the COST Action MP1103 “Nanostructured materials for solid- 2009, 21, 925.
state hydrogen storage”, by the Danish Council for Strategic 17 K. Chłopek, C. Frommen, A. Léon, O. Zabara and
Research, via the research project HyFillFast, the Danish M. Fichtner, J. Mater. Chem., 2007, 17, 3496.
National Research Foundation, Center for Materials Crystallo- 18 P. Zanella, L. Crociani, N. Masciocchi and G. Giunchi,
graphy (DNRF93), and by the Danish Research Council for
Inorg. Chem., 2007, 46, 9039.
This journal is © The Royal Society of Chemistry 2015
Dalton Trans., 2015, 44, 6571–6580 | 6579

View Article Online
Paper
Dalton Transactions
19 Y. Filinchuk, B. Richter, T. R. Jensen, V. Dmitriev, 42 U. Mirsaidov, A. Kurbobekov and M. Khikmatov,
D. Chernyshov and H. Hagemann, Angew. Chem., Int. Ed.,
2011, 50, 11162.
20 M. Tamura and J. Kochi, J. Organomet. Chem., 1971, 29,
111.
21 M. Bremer, G. Linti, H. Nöth, M. Thomann-Albach and
G. E. W. J. Wagner, Z. Anorg. Allg. Chem., 2005, 631, 683.
22 M. Bremer, H. Nöth and M. Warchhold, Eur. J. Inorg.
Chem., 2003, 111.
Russ. J. Inorg. Chem., 1981, 26, 1538.
43 V. D. Makhaev and K. N. Semenenko, Bull. Acad. Sci. USSR
Div. Chem. Sci., 1978, 27, 2520.
44 V. D. Makhaev, A. P. Borisov, N. G. Mozgina, G. N. Boyko
and K. N. Semenenko, Neorg. Mater., 1978, 14, 1726.
45 R. Černý, G. Severa, D. B. Ravnsbæk, Y. Filinchuk,
V. D’Anna, H. Hagemann, D. Haase, C. M. Jensen and
T. R. Jensen, J. Phys. Chem. C, 2010, 114, 1357.
23 G. L. Soloveichik, M. Andrus and E. B. Lobkovsky, Inorg. 46 D. B. Ravnsbæk, L. H. Sørensen, Y. Filinchuk,
Chem., 2007, 46, 3790.
24 R. Černý, N. Penin, V. D’Anna, H. Hagemann, E. Durand
and J. Růžička, Acta Mater., 2011, 59, 5171.
F. Besenbacher and T. R. Jensen, Angew. Chem., Int. Ed.,
2012, 51, 3582.
47 T. Jaroń and W. Grochala, Dalton Trans., 2010, 39, 160.
25 V. D. Makhaev, A. P. Borisov, T. P. Gnilomedova, 48 T. Jaroń, W. Koźmiński and W. Grochala, Phys. Chem.
E. B. Lobkovskii and A. N. Chekhlov, Bull. Acad. Sci. USSR
Div. Chem. Sci., 1987, 36, 1582.
Chem. Phys., 2011, 13, 8847.
49 T. Jaroń and W. Grochala, Dalton Trans., 2011, 40, 12808.
26 V. D. Makhaev, A. P. Borisov, T. P. Karpova and 50 T. Jaroń, W. Wegner and W. Grochala, Dalton Trans., 2013,
L. A. Petrova, Russ. J. Inorg. Chem., 1995, 40, 1579.
27 V. D. Makhaev, Priv. Commun., 2014.
28 B. Richter, N. A. Tumanov, D. B. Ravnsbæk, Y. Filinchuk
and T. R. Jensen, Dalton Trans., 2015, 44, 3988.
29 G. L. Soloveichik, M. Andrus, Y. Gao, J.-C. Zhao and
S. Kniajanski, Int. J. Hydrogen Energy, 2009, 34, 2144.
42, 6886.
51 P. Panissod and M. Drillon, in Magnetism: Molecules to
Materials IV: Nanosized Magnetic Materials, ed. J. S. Miller
and M. Drillon, Wiley-VCH Verlag GmbH, Weinheim, 2003,
p. 243.
52 R. Liu, D. Reed and D. Book, J. Alloys Compd., 2012, 515, 32.
30 S. K. Sharov, L. P. Portnov, L. P. Ivanov and A. I. Gorbunov, 53 A. W. Addison, R. T. Nageswara, J. Reedijk, J. van Rijn and
Russ. J. Phys. Chem., 1987, 61, 1405. G. C. Verschoor, J. Chem. Soc., Dalton Trans., 1984, 1349.
31 H. C. Brown, Y. M. Choi and S. Narasimhan, Inorg. Chem., 54 M. J. Ingleson, J. P. Barrio, J. Bacsa, A. Steiner,
1982, 21, 3657.
G. R. Darling, J. T. A. Jones, Y. Z. Khimyak and
M. J. Rosseinsky, Angew. Chem., Int. Ed., 2009, 48, 2012.
55 CrysAlis PRO, Agilent Technologies, 2013.
32 D. A. Brandreth and M. C. Molstad, J. Chem. Eng. Data,
1962, 7, 449.
33 V. Mikheeva and E. Troyanovskaya, Bull. Acad. Sci. USSR 56 G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystal-
Div. Chem. Sci., 1971, 20, 2497. logr., 2008, 64, 112.
34 H. C. Brown, Y. M. Choi and S. Narasimhan, Inorg. Chem., 57 A. P. Hammersley, S. O. Svensson, M. Hanfland, A. N. Fitch
1981, 20, 4454. and D. Hausermann, High Pressure Res., 1996, 14, 235.
35 H. I. Schlesinger, H. C. Brown, H. R. Hoekstra and 58 A. Boultif and D. Louër, J. Appl. Crystallogr., 2004, 37, 724.
L. R. Rapp, J. Am. Chem. Soc., 1953, 75, 199.
36 M. H. Jones, W. Levason, C. A. McAuliffe and M. J. Parrott,
J. Chem. Soc., Dalton Trans., 1976, 1642.
59 V. Favre-Nicolin and R. Černý, J. Appl. Crystallogr., 2002, 35,
734.
60 J. Rodríguez-Carvajal, Phys. B, 1993, 192, 55.
37 V. A. Abakshin, G. A. Krestov and O. V. Eliseeva, J. Gen. 61 G. Kresse and J. Furthmüller, Phys Rev B: Condens. Matter,
Chem. USSR, 1987, 57, 1431. 1996, 54, 11169.
38 A. V. Levkin and A. Y. Tsivadse, Russ. J. Inorg. Chem., 1986, 62 P. E. Blöchl, Phys. Rev. B: Condens. Matter, 1994, 50, 17953.
31, 161.
63 J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett.,
1996, 77, 3865.
39 B. D. James and M. G. H. Wallbridge, Prog. Inorg. Chem.,
1970, 11, 99.
64 S. L. Dudarev, G. A. Botton, S. Y. Savrasov, C. J. Humphreys
and A. P. Sutton, Phys. Rev. B: Condens. Matter, 1998, 57,
1505.
40 D. Seebach, A. Thaler and A. K. Beck, Helv. Chim. Acta,
1989, 72, 857.
41 H. C. Brown, S. Narasimhan and Y. M. Choi, J. Org. Chem., 65 C. Franchini, R. Podloucky, J. Paier, M. Marsman and
1982, 47, 4702.
G. Kresse, Phys. Rev. B: Condens. Matter, 2007, 75, 195128.
6580 | Dalton Trans., 2015, 44, 6571–6580
This journal is © The Royal Society of Chemistry 2015