M(BH3NH2BH2NH2BH3)-the missing link in the mechanism of the thermal decomposition of light alkali metal amidoboranes

Article abstract of DOI:10.1039/c4cp03296a

We report a novel family of hydrogen-rich materials-alkali metal di(amidoborane)borohydrides, M(BH₃NH₂BH₂NH₂BH₃). The title compounds are related to metal amidoboranes (amidotrihydroborates) but have higher gravimetric H content. Li salt contains 15.1 wt% H and discharges very pure H₂ gas. Differences in thermal stability between amidoboranes and respective oligoamidoboranes explain the release of the ammonia impurity (along with H₂) during the thermal decomposition of light alkali amidoboranes, LiNH₂BH₃, NaNH₂BH₃ and NaLi(NH₂BH₃)₂, and confirm the mechanism of the side decomposition reaction. This journal is

Full text of DOI:10.1039/c4cp03296a

PCCP
View Article Online
View Journal
PAPER
M(BH NH BH NH BH ) – the missing link in the
3
2
2
2
3
mechanism of the thermal decomposition of light
Cite this:DOI: 10.1039/c4cp03296a
alkali metal amidoboranes†
a
a
a
b
b
K. J. Fijalkowski,* T. Jaron, P. J. Leszczynski, E. Magos-Palasyuk, T. Palasyuk,
´
´
c
a
M. K. Cyranski and W. Grochala*
´
We report a novel family of hydrogen-rich materials – alkali metal di(amidoborane)borohydrides,
M(BH NH BH NH BH ). The title compounds are related to metal amidoboranes (amidotrihydroborates)
but have higher gravimetric H content. Li salt contains 15.1 wt% H and discharges very pure H gas.
Differences in thermal stability between amidoboranes and respective oligoamidoboranes explain the
release of the ammonia impurity (along with H ) during the thermal decomposition of light alkali
amidoboranes, LiNH BH , NaNH BH and NaLi(NH BH , and confirm the mechanism of the side
decomposition reaction.
3
2
2
2
3
Received 24th July 2014,
Accepted 8th September 2014
2
DOI: 10.1039/c4cp03296a
2
2
3
2
3
2
3 2
)
www.rsc.org/pccp
derivatives. The M(B3N2) phases may formally be regarded to
be the derivatives of metal borohydrides, M[B(H) (NH BH ) ]
Introduction
2
2
3 2
Protonic-hydridic materials (XH YH ) comprise an important (analogous to the known M[B(H) (CN) ], M[B(H) (CH ) ] or
x
y
x
4Àx
x
3 4Àx
1
,2
family of highly efficient hydrogen stores in the solid state.
The simultaneous presence of acidic H (attached to a more larger gravimetric hydrogen capacity than MAB phases and they
electronegative element, X) and basic H (attached to a more fulfill the DOE gravimetric target in excess [Li(B3N2), 15.1 wt% H;
M[B(H) (C H ) ] salts). The M(B3N2) compounds have much
x 2 5 4Àx
+
À
electropositive element, Y) results in low energy barrier and, Na(B3N2), 12.6 wt% H], which is crucial when designing the storage
concomitantly, in low temperature H
2
desorption from these system (consisting of both: storage material and various mechanical
materials. Systems based on X = nitrogen and Y = boron elements) to contain more than 7.5 wt% of hydrogen (Fig. 1).
constitute the most researched protonic-hydridic hydrogen
Here, we show that the M(B3N2) derivatives constitute the
missing link in the mechanism of the thermal decomposition
3
,4
5,6
7
stores. These encompass amide-borohydrides, NH BH ,
4
4
8
,9
ammonia borane (NH BH ), as well as its derivatives, such as of light alkali metal amidoboranes [LiAB, NaAB and NaLi(AB)2].
3
3
1
0–21
metal amidoboranes, MNH
these materials suffer from unfavorable thermodynamics, containing [M(NH
which precludes the re-absorption of H by the discharged is responsible for the presence of ammonia impurity in the H
residue of the thermal decomposition. Despite emitting impure gas evolved from MABs.
2
BH
3
(MAB phases).
Most of The transformation of MABs to the corresponding ammonia
3
)](B3N2) and then the formation of M(B3N2)
2
2
1
7,22,23
Moreover, the M(B3N2) salts
H
2
gas, the large gravimetric H content (up to 24.4 wt% for thermally decompose above 140–160 1C and release more than
NH BH ) renders them attractive to be H storage media for 50% of their total H content up to 200 1C. Importantly, Li salt
4
4
mobile technologies due to their off-board regeneration.
Here, we describe a novel family of hydrogen rich materials –
yields H gas that is virtually free from any impurities.
2
metal di(amidoborane)borohydrides, M(BH
3
2 2 2 3
NH BH NH BH )
(M(B3N2) phases), as exemplified by lithium and sodium
a
Center of New Technologies, University of Warsaw, ul. Zwirki i Wigury 93,
2-089 Warsaw, Poland. E-mail: w.grochala@cent.uw.edu.pl,
karol.fijalkowski@cent.uw.edu.pl
0
b
c
Institute of Physical Chemistry, Polish Academy of Sciences, ul. Kasprzaka 44/52,
1-224 Warsaw, Poland
0
Faculty of Chemistry, University of Warsaw, ul. Pasteura 1, 02-093 Warsaw,
Poland
Fig. 1 Comparison of total theoretical hydrogen capacity of light alkali
metal M(B3N2) phases and respective amidoboranes, M = Li, Na, K without
taking into account weight of other elements of the storage system.
†
Electronic supplementary information (ESI) available. See DOI: 10.1039/
c4cp03296a
This journal is ©the Owner Societies 2014
Phys. Chem. Chem. Phys.

View Article Online
Paper
PCCP
Results and discussion
The first MB
quarters of a century ago, when Schlesinger and Burg in 1938
reported a light porous solid of a general formula NaB N,
obtained in the reaction of diammoniate of diborane with
x y
N salt with x 4 y was probably synthesized three
2
H
8
24
metallic sodium. The product has been incompletely characterized
and these results have not been repeated by others. In 2009, we
observed evolution of ammonia-along with H -during the thermal
2
22
decomposition of alkali metal amidoboranes. Elimination of a
volatile nitrogen compound suggested the formation of a solid
residue, which was richer in boron as compared to the parent 1 : 1
stoichiometry of metal amidoboranes. At that time, the formation
2
2,23
of the ‘‘NB
which is in accordance with the claim from Schlesinger and Burg.
Probably Li(B3N2) and Na(B3N2) compounds were first
2
’’ metal derivatives in the solid state was suggested,
11
Fig. 2 Comparison of B NMR spectra in THF-d
8
of M(B3N2) phases
coloured lines) and metal amidoboranes (black lines): lithium salts (top)
24
(
and sodium salts (bottom).
2
5
observed by Evans in 2011. The work, however, missed the
basic characterization data, such as correct stoichiometry and
crystal structure, essential for the chemical identity of the
compounds. Experimental data contained in investigations by
(
BH ) (BH ) constituents of the anions. The spectra also confirm
the presence of small amount of LiAB impurity in the sample of
Li(B3N2).
2 1 3 2
2
6
Ryan, 2011, did not present new structural evidence but
suggested the formation of a new crystalline compound, which
is built of lithium cations and five-membered anionic chains
with the BNBNB skeleton. The presence of pentameric moieties
has recently been suggested for the sodium compound by W. C.
All alkali metal M(B3N2) phases are characterized by similar
and characteristic Raman (Fig. 3) and FTIR spectra (ESI†),
which are easily distinguishable from the spectra of respective
metal amidoboranes (MABs). In general, the sharp doublet in
the Raman spectra assigned to the symmetric NH stretching
modes appears at smaller wavenumbers for M(B3N2) phases
than for the respective MABs. Simultaneously, the most intense BH
stretching bands fall at higher wavenumbers for M(B3N2) phases
than for the respective MABs. These features suggest that the N–H
bonds are considerably weaker for M(B3N2) phases than for MABs,
while the opposite holds true for the B–H bonds. Although the
stiffening of the B–H modes is in qualitative agreement with the
trend predicted from our DFT calculations for the isolated
2
7
11
Ewing et al., 2013. The reported B NMR spectrum revealed
the presence of (BH (BH constituents of the anions. The
2
)
1
3 2
)
authors, however, did not present any data on the crystal
structure of the studied material.
We have now conducted a systematic study of the two
1
1
lightest alkali metal M(B3N2) phases (M = Li or Na) using
B
NMR, single crystal and powder X-ray diffraction, IR absorption,
Raman scattering, thermogravimetry and differential scanning
calorimetry. We have been able to solve the crystal structures of
the M(B3N2) phases (M = Li or Na) and provide complete
physicochemical characterization in relation to the hydrogen
storage properties. The comparative analysis of the M(B3N2)
phases with their MABs analogues has been also carried out.
M(B3N2) phases were conveniently obtained in the reaction
of ammonia borane with respective metal hydride in molar
ratio 3 : 1 in anhydrous THF (see ESI†):
À
À
(
NH BH ) and (BH NH BH NH BH ) anions in the gas phase,
2 3 3 2 2 2 3
the softening of the N–H bonds observed in experiment is in
disagreement with the computed trend (see ESI†), and it likely
originates from the solid state effects (differences of packing,
MH + 3NH BH - M(BH NH BH NH BH ) + NH m + 2H m
3
3
3
2
2
2
3
3
2
(1)
Mechanochemical synthesis followed by thermal desorption
2
5
3
of excess NH is an alternative approach (ESI†).
Alkali metal M(B3N2) phases are white low-density solids
which are sensitive to moisture. Li(B3N2) and Na(B3N2) are
2
3
stable at room temperature in contrast to respective MABs, as
discussed below. Li(B3N2) is well soluble in THF, in contrast to
Na(B3N2). However, the solubility of both compounds is sufficient
11
to obtain the B NMR spectra (Fig. 2). The spectra are different
from those of the respective metal amidoboranes (see ESI†) and
they reveal the presence of a characteristic quarter at ca. À21 ppm
Fig. 3 Comparison of Raman spectra of M(B3N2) phases (coloured lines)
and metal amidoboranes (black lines): lithium salts (top) and sodium salts
À1
À1
representing [BH
representing [BH
3
] groups and a twice weaker triplet at ca. À8 ppm
(
bottom). The NH (3200–3400 cm ) and BH (1900–2500 cm ) stretching
] moieties, thus confirming the presence of the regions are highlighted.
2
Phys. Chem. Chem. Phys.
This journal is ©the Owner Societies 2014

View Article Online
PCCP
Paper
interactions of anions with cations, or secondary interactions, Table 1 Comparison of lattice parameters for three alkali metal M(B3N2)
phases
such as possible dihydrogen bonding).
One sharp band appearing for the HNH bending vibration at
571 cm is seen in the Raman spectrum of Li(B3N2); the Compound System
Space
group a [Å]
À1
3
1
c [Å]
V [Å ]
Z
presence of a single band suggests that the NH groups have
a
2
Li(B3N2)
Tetragonal P4%2c 4.02 (1)
16.95 (5) 273.9 (12)
2
2
b
identical local symmetry (thus, only the in-phase HNH bending Na(B3N2) Hexagonal P6
3
/m 4.3392 (2) 17.853 (6) 291.11 (18)
À
is Raman-active). Indeed, the (BH
3
NH
2
BH
2
NH
2
BH
3
)
moiety in
a
b
Compare ref. 25 and 26. Compare ref. 25.
the crystal structure of Li(B3N2) is found in the n-alkane-like
quasi-linear geometry with two crystallographically equivalent
NH units.
2
The samples of alkali metal M(B3N2) phases were studied
with powder X-ray diffraction (PXD) (Fig. 4).
Lattice parameters of Li(B3N2) determined in this work are
in accordance with the values obtained for the compounds
2
5
26
reported by Evans and by Ryan (Table 1 and ESI†). This
result suggests that chemical identity of all three crystalline
samples may be the same, even though Evans suggested a
different stoichiometry for this compound. Evans proposed
+
À
the chemical formula of Li(NH ) (BH NH BH ) containing
3
3
2
3
2
5
the ammonia molecule weakly bound to alkali metal cation.
Results of our spectroscopic measurements (Raman or IR), how-
À1
ever, did not reveal any significant bands in the 3350–3400 cm
+
region, which is characteristic for Li(NH
3
) cations. Moreover, we
observed that Li(B3N2) releases pure hydrogen on heating, which
+
3
corroborates the absence of the Li(NH ) cation in material. Based
on the combination of NMR, FTIR, and Raman spectra, as well as
PXD we have been able to confirm the stoichiometry of
2
6
Li(BH NH BH NH BH ) which was previously suggested by Ryan.
3
2
2
2
3
Fig. 5 Crystal structure of alkali metal M(B3N2) phases as determined by
The crystal structure of Li(B3N2) consists of Li layers which are PXD: H – grey, B – red, N – blue, Li – green, Na – yellow. Li(B3N2) top and
+
À
3 2 2 2 3
NH BH NH BH ) anions which
Na(B3N2) – bottom. Crystallographic positions with partial population are
linked via five member chain (BH
coordinate lithium cations using the hydrogen atoms of terminal
BH ] groups only (Fig. 5). It should be recalled that the N atoms of
the (BH anions do not have any lone pairs
marked with partly-colored balls.
[
3
À
3
NH
2
BH
2
NH
2
BH
3
)
(
the N H coordination). The shortest Li–H distance is 2.0(2) Å
1 4
available for coordination to metal centers; thus, only the basic
hydride anions at boron may be used for bonding. Lithium cations
are found in tetrahedral coordination of four hydrogen atoms,
resembling those found for the orthorhombic form of LiBH , and
they are different from the first coordination sphere of Li in LiAB
(
as compared with the corresponding distance of 1.98(1) Å in Pbca
28
29
LiAB and 1.98(1) Å in Pnma LiBH
length of the (BH
4
). The average B–N bond
anion is ca. 1.59(6) Å
À
3
NH
2
BH
2
NH
2
BH
3
)
4
10
(as compared with 1.55 Å for LiAB ), whereas the average angles
between the B–N bonds (1151) are close to the tetrahedral angle.
À
The crystal structure contains disordered (BH NH BH NH BH )
3
2
2
2
3
anions lying along the z axis of the unit cell. Importantly, the inter-
anion dihydrogen bonds are absent in the crystal structure of
Li(B3N2).
3
The volume of Li(B3N2) per one formula unit, 136.95(6) Å ,
is smaller by only 3.3% than the expected volume of this
3
compound (141.65 Å ), which may be derived from the follow-
ing equation:
V[Li(BH
3
NH
2
BH
2
NH
2
BH
3
)] = 3V[Li(NH
2
BH
3
)] À V[LiNH
2
] À V[LiH]
(
2)
The PXD for Na(B3N2) yields a diffraction pattern which can
2
5
be indexed using the trigonal cell proposed by Evans. We have
determined the crystal structure of Na(B3N2) in a centrosymmetric
3
primitive hexagonal P6 /m cell with Z = 2 (Table 1 and ESI†). There
Fig. 4 Comparison of PXD patterns of M(B3N2) phases (coloured lines)
are three important similarities between crystal structures of
Na(B3N2) and Li(B3N2): first, Na(B3N2) also consists of metal
and respective amidoboranes (black lines): lithium (top) and sodium salts
(
bottom).
This journal is ©the Owner Societies 2014
Phys. Chem. Chem. Phys.

View Article Online
Paper
PCCP
cation layers separated by layers containing the (BH
3
NH
2
BH
2
-
À
NH
2
3
BH )
anions. Second, the anions are substantially disordered
along the z axis. Third, metal cations are found in the homoleptic
environment of hydrogen atoms of the terminal [BH ] groups. As a
3
consequence, sodium cations exhibit a coordination with 8 H atoms
2
(
=12 Â ), and the closest Na–H distance is 2.22(9) Å. This may be
3
compared to the shortest Na–H separation of ca. 2.33 Å, which is
seen for NaAB (cf. ESI†) and a much larger value of ca. 2.58 Å, which
is typical of the disordered room-temperature rocksalt form of
30
4
NaBH . Obviously, the metal–hydrogen bond lengths that are listed
here for both M(B3N2) salts are not exact because of the inherent
problems related with solving disordered crystal structures of com-
pounds that are built of low-Z elements using laboratory X-ray
31
radiation source.
The density functional theory as well as the Møller–Plesset
2
Fig. 7 The TGA/DSC profiles and well as the ion current for m/z = 2 (H )
and 16 (NH
3
) measured for the thermal decomposition of Li(B3N2) at
perturbation theory calculations for the isolated (BH
3
NH
2
BH
2
-
À1
1
0 1C min scanning rate.
À
NH
2
BH
3
)
anion in the gas phase (Fig. 6) indicate that the C2v
form. The relative energy of the
form is less stable than the C
1
17,22,23
C1 isomer with respect to the C one (corrected for the energy
2v
with ammonia.
After several days of storing the sample of
of zero-point vibrations) is around À0.15 Æ 0.01 eV. However,
because the energy difference is not significantly large, it may
be anticipated that the interactions of anions with cations as
well as the differences of packing lead to the small preference
light alkali amidoboranes at RT they are contaminated with the
traces of M(B3N2) phases (for PXD patterns see ESI†). Contamina-
tion of LiAB and NaAB samples by the traces of Li(B3N2) and
Na(B3N2) may be observed in PXD patterns reported in our previous
3
2
17,33
10,26,34
of the C2v quasi-alkane form in the solid state. These calcula-
tions also show that the hydride anions of the terminal BH
reports
and independently by other groups
(ESI†). Further-
3
more, the heating of light alkali amidoborane samples to ca. 50 1C
groups are the most basic (i.e. more negatively charged) H
atoms of the anion, which provides further confirmation of
their capability to coordinate with metal cations. Note that the
B–N bond lengths of the C2v form are close to the experimen-
tally determined value of 1.59(6) Å.
results in an accelerated formation of respective M(B3N2) phases
(which can be detected by PXD) accompanied by the evolution of
ammonia (Fig. 8):
3
M(NH BH ) - [M(NH )](BH NH BH NH BH ) + 2[M]
(3)
2
3
3
3
2
2
2
3
We investigated the thermal stability of alkali M(B3N2)
phases under inert argon atmosphere (at À35 1C, at room
temperature and on heating to 200 1C, Fig. 7) in comparison [M(NH
)](BH
NH
BH
NH
BH
) - M(BH
NH
BH
NH
BH
) + NH
m
3
3
3
2
2
2
3
3
2
2
2
3
to those of respective amidoboranes. It was noticed that LiAB,
NaAB and NaLi(AB)2 are unstable at room temperature and
spontaneously decompose releasing hydrogen contaminated
(4)
À
Fig. 6 The calculated geometry of the (BH
the C2v form (top) and C
3
NH
2
BH
2
NH
2
BH
3
)
anion in
Fig. 8 The simplified mechanism of two competing reactions during the
thermal decomposition of alkali metal amidoboranes: direct evolution of
1
form (bottom) as optimized at the 6-311++G**/
B3LYP (bold) or /MP2 (italics) level of theory. Zero-point vibrational
energy-corrected relative energy is in eV.
H
2
(top) and formation of M(B3N2) phases with the evolution of NH
3
(bottom).
Phys. Chem. Chem. Phys.
This journal is ©the Owner Societies 2014

View Article Online
PCCP
Paper
where [M] in eqn (3) stands for unbalanced alkali metal, thermal decomposition of MAB phases is currently unknown.
indicating that the true pathway for the decomposition reaction One of the assumed scenarios involve the short-lived inter-
is more complex than the one proposed here. One possibility to mediates containing related but shorter B2N backbone that are
2
4
balance eqn (3) is the formation of amorphous M (NHBH ) phases: similar to NaB H N reported in the 1930s, but more experi-
2
3
2
8
mental evidence is needed to confirm it.
5
M(NH
2
BH
3
) - [M(NH
3
)](BH
3
NH
2
BH
2
NH
2
BH
3
) + 2M
2
(NHBH
3
) + H
2
m
(5)
However, we choose not to speculate further about the
Experimental
Synthetic procedures
mechanism of the side reaction due to the largely amorphous
nature of the products as well as insufficient amount of
experimental data.
Alkali metal M(B3N2) phases and amidoboranes were synthesized
Li(B3N2) and Na(B3N2) salts are considerably more thermally by the direct reaction of alkali metal hydride with ammonia borane
stable than their amidoborane analogues; they decompose in dry THF under the argon gas atmosphere. Reagents of the
2
5
exothermally at temperatures over 140–160 1C (Fig. 7 and ESI†).
highest commercially available purity have been used: LiH, LiNH2,
Crystalline LiBH and NaBH were identified in the solid resi- NaH (95%, Sigma Aldrich) and NH BH (98%, JSC Aviabor). THF
4
4
3
3
dues after the dehydrogenation of the respective M(B3N2) (99.9%, Sigma Aldrich) was firstly dried over yttrium borohydride or
phases, suggesting the following tentative mechanism for the sodium hydride and then distilled.
1
7
thermal decomposition:
Synthesis of M(B3N2) phases (eqn (8)) and amidoboranes
eqn (9) and (10)) were performed according to the following
reaction equations:
(
M(BH NH BH NH BH ) - MBH
3
2
2
2
3
4
+ 2/n (NHBH)
n
2
+ 2H m
(6)
The expected mass loss according to eqn (6) is 5.0% (for Li salt).
The first sudden mass loss observed for the Li sample at 140 1C is
ca. 5 wt%, which is in a crude agreement with the simplified
mechanism described by eqn (6). Additional mass loss is observed
for temperatures up to 200 1C, which suggests that the resulting
MH + 3NH BH - M(BH NH BH NH BH ) + NH m (8)
3
3
3
2
2
2
3
3
MNH + NH BH - MNH BH + H m
(9)
2
3
3
2
3
2
MH + NH
3
BH
3
- MNH
2
BH
3
+ H
2
m
(10)
(
NHBH) also decomposes to some extent releasing H :
n
2
1
/n (NHBH) - BN + H m
(7)
n
2
Reactions were performed in THF solution at room temperature
with continuous stirring for 24 h or in a disc mill by mechano-
synthetic method. The solid products were washed several times
with fresh portions of THF and allowed to dry; they were analyzed
without further purification. Samples were stored under argon
Importantly, Li(B3N2) releases very pure H
from ammonia or any other N–B–H impurities. This feature
together with the increased total H content) represents a major
advantage of Li(B3N2) phases over the pristine LiAB, which
2
gas, which is free
(
2
2,34,35
atmosphere in Labmaster DP MBRAUN glovebox (O o 1.0 ppm;
2
yields hydrogen severely contaminated with ammonia.
H
2
O o 1.0 ppm) at À35 1C. All further analyses were performed
under inert atmosphere or in vacuum.
Conclusions
Thermal decomposition was investigated using an STA 409
simultaneous thermal analyzer from Netzsch, in the tempera-
ture range of À10 1C to +350 1C. STA 409 allows for the
simultaneous thermogravimetric analysis (TGA), differential
scanning calorimetry (DSC) and evolved gas analysis (QMS).
The samples were loaded into alumina crucibles. High purity
We have synthesized and characterized two novel protonic-hydridic
compounds, the lithium and sodium M(B3N2) derivatives con-
À
3
taining the (BH NH
2 2 2 3
BH NH BH )
anions. These lightweight
salts constitute novel H-rich materials of high gravimetric and
volumetric hydrogen capacity (12.6–15.1 wt% H); they are stable
at room temperature. Li salt releases pure H gas upon heating
in the 140–200 1C temperature range, which is of immense
importance for practical applications. This-along with the
increased total H content-represent a major advantage of
6
N argon was used as a carrier gas. The evolved gases were
2
analyzed with a QMS 403C A ¨e olos mass spectrometer from
Pfeiffer-Vacuum. Transfer line was preheated to 200 1C to avoid
the condensation of residues.
Li(B3N2) phases over the pristine LiAB, which yields H
2
severely
2
2,34,35
Infrared absorption spectroscopy
contaminated with ammonia.
Metal borohydrides consti-
tute the only crystalline products of the thermal decomposition All substrates, products and thermally decomposed samples
of the M(B3N2) salts. were characterized with infrared absorption spectroscopy in
The spontaneous formation of M(B3N2) phases in the KBr pellets using a Vertex 80v vacuum FTIR spectrometer from
samples of LiAB, NaAB and NaLi(AB)2 at room temperature Bruker. The samples sealed in a 0.6 mm thick quartz capillary
2
3
changes the N : B balance of the solids from 1 : 1 to : 1, which is under Ar were also characterized by Raman spectroscopy using
related to the evolution of ammonia impurity (along with H
2
)
a custom designed setup for micro-Raman spectroscopy based
from pristine MAB salts. The exact mechanism of formation of on monochromator Jobin Yvon THR1000 with CCD detection.
the M(B3N2) phases as a side reaction taking place during Excitation line 632.8 nm was used.
This journal is ©the Owner Societies 2014
Phys. Chem. Chem. Phys.

View Article Online
Paper
PCCP
Powder X-ray diffraction
Budzianowski for preliminary solving the crystal structure of
NaAB (to be published elsewhere).
PXD patterns of solids (sealed under argon inside 0.6 mm thick
quartz capillaries) were measured using two diffractometers:
(a) Panalytical X’Pert Pro diffractometer with linear PIXcel
Medipix2 detector (parallel beam; the CoKa1 and CoKa2 radia- Notes and references
tion intensity ratio of ca. 2 : 1, l B 1.789 Å), denoted here as
1
W. Grochala and P. P. Edwards, Chem. Rev., 2004, 104,
283.
Ł. Maj and W. Grochala, Adv. Funct. Mater., 2006, 16, 2061.
C. W. Hamilton, R. T. Baker, A. Staubitzc and I. Manners,
Chem. Soc. Rev., 2009, 38, 279.
CoK ; (b) Bruker D8 Discover diffractometer with 2D Vantec
detector (parallel beam; the CuKa1 and CuKa2 radiation inten-
a
1
2
3
sity ratio of ca. 2 : 1, l B 1.5406 Å), denoted here as CuK . All
a
the PXD results are shown in copper scale.
The diffraction signals of M(B3N2) phases, M = Li or Na,
3
6
4 A. J. Churchard, E. Banach, A. Borgschulte, R. Caputo,
J.-C. Chen, D. Clary, K. J. Fijalkowski, H. Geerlings,
R. V. Genova, W. Grochala, T. Jaron, J. C. Juanes-Marcos,
B. Kasemo, G. J. Kroes, I. Ljubic, N. Naujok, J. K. Norskov,
R. A. Olsen, F. Pendolino, A. Remhof, L. Romanszki,
A. Tekin, T. Vegge, M. Zach and A. Zuttel, Phys. Chem. Chem.
Phys., 2011, 13, 16955.
5 F. E. Pinkerton, G. Meisner, M. Meyer, M. Balogh and
M. J. Kundrat, J. Phys. Chem. B, 2005, 109, 6.
6
were indexed using X-cell program (Accelrys). Because of the
significant disorder and presence of peaks from unidentified
impurities, the successful structure solution was only possible
after the proper identification of the anions from spectroscopic
methods (M = Li or Na) and information on the BNBNB back-
bone from the single crystal diffraction (M = Na). In both the
cases, the structure solution was attempted in several unit cells
3
7
using a real-space method implemented in program FOX.
Finally, because the unit cells of lower symmetry did not lead to
the significant improvement of fit, we chose the high-symmetry
solutions, which very efficiently reproduced the experimental
Z. Xiong, G. Wu, J. Hu and P. Chen, Adv. Mater., 2004,
6, 1522.
1
3
8
7 R. W. Parry, D. R. Schultz and P. R. Girardot, J. Am. Chem.
Soc., 1958, 80, 1.
8
patterns and refined the starting models in Jana2006. Several
restraints were necessary to retain the realistic geometry of the
anions for M = Li or Na; all the N–H and B–H distances and
angles related to hydrogen atoms were restrained to 1.10(1) Å,
and 109.5(5)1 (M = Li) or 109.47(5)1 (M = Na), respectively; for
M = Na B–N distances were also restrained to 1.60(1) Å; the
atomic displacement parameters (ADP) of H atoms were set as
S. G. Shore and R. W. Parry, J. Am. Chem. Soc., 1955, 77,
084–6085.
J. Baumann, E. Baitalow and G. Wolf, Thermochim. Acta,
005, 430, 9.
6
9
2
1
0 Z. Xiong, C. K. Yong, G. Wu, P. Chen, W. Shaw, A. Karkamkar,
T. Autrey, M. O. Jones, S. R. Johnson, P. P. Edwards and
W. I. F. David, Nat. Mater., 2008, 7, 138.
1.2–1.5 of ADP of adjacent heavier atom. The weak scattering of
X-ray by hydrogen atoms allows for only rough determination of
their positions, therefore all the distances involving hydrogen
atoms should be treated as an approximate approach. The
pseudo-Voigt peak shape function was used; the background
has been manually modeled. Further details of the crystal
structure investigation(s) may be obtained from the Fachinfor-
mationszentrum Karlsruhe, 76344 Eggenstein-Leopoldshafen
1
1 H. V. K. Diyabalanage, T. Nakagawa, R. P. Shrestha, T. A.
Semelsberger, B. L. Davis, B. L. Scott, A. K. Burrell, W. I. F.
David, K. R. Ryan, M. Owen Jones and P. P. Edwards, J. Am.
Chem. Soc., 2010, 132, 11836.
12 J. Luo, X. Kang and P. Wang, Energy Environ. Sci., 2012, 6,
1018–1025.
1
3 H. V. K. Diyabalanage, R. P. Shrestha, T. A. Semelsberger,
B. L. Scott, M. E. Bowden, B. L. Davis and A. K. Burrell,
Angew. Chem., Int. Ed., 2007, 46, 8995.
(Germany), on quoting the depository CSD numbers: 428007 –
Li(B N H ), 428008 – Na(B N H ).
3
1
2
12
3
2
12
1
B NMR spectra of M(B3N2) phases and respective amido-
boranes were obtained using a NMR UnityPlus 200 MHz
VARIAN spectrometer with BF :C –O–C as an external
standard and H NMR measurements were performed with
TMS. THFd (Aldrich, 99.5 atom% D), dried over metallic
14 J. Spielmann, G. Jansen, H. Bandmann and S. Harder,
Angew. Chem., Int. Ed., 2008, 47, 6290.
3
2
H
5
2 5
H
1
15 Q. Zhang, Ch. Tang, Ch. Fang, F. Fang, D. Sun, L. Ouyang
and M. Zhu, J. Phys. Chem. C, 2010, 114, 1709.
8
1
1
6 R. V. Genova, K. J. Fijalkowski, A. Budzianowski and
W. Grochala, J. Alloys Compd., 2010, 499, 144.
7 K. J. Fijalkowski, R. V. Genova, Y. Filinchuk, A. Budzianowski,
M. Derzsi, T. Jaro n´ , P. Leszczy n´ ski and W. Grochala, Dalton
Trans., 2011, 40, 4407.
sodium, was used as a solvent.
Acknowledgements
18 Y. Zhang, K. Shimoda, T. Ichikawa and Y. Kojima, J. Phys.
To Mr Jerzy Antoni Fijalkowski on his birthday. This research
Chem., 2010, 114, 14662.
was funded from 0122/IP3/2011/71 grant ‘‘Iuventus Plus’’ of the 19 W. Li, L. Miao, R. H. Scheicher, Z. Xiong, G. Wu, C. M. Araujo,
Polish Ministry of Science and Higher Education. E.M.-P. and
T.P. gratefully acknowledge the support of the Polish National
A. Blomqvist, R. Ahuja, Y. Feng and P. Chen, Dalton Trans.,
2012, 41, 4754.
Science Center (project no. 2011/01/M/ST3/00855) (program 20 X. Kang, J. Luo, Q. Zhang and P. Wang, Dalton Trans., 2011,
‘Harmonia’’). The authors would like to thank Dr Armand 40, 3799.
‘
Phys. Chem. Chem. Phys.
This journal is ©the Owner Societies 2014

View Article Online
PCCP
Paper
21 H. Wu, W. Zhou, F. E. Pinkerton, M. S. Meyer, Q. Yao, 32 The Rietveld fit of the crystal structure of Na(B3N2) is
S. Gadipelli, T. J. Udovic, T. Yildirim and J. J. Rush, Chem.
Commun., 2011, 47, 4102.
of somewhat poorer quality than that for its lithium
analogue. It could be due to the presence of a fraction
2
2
2 K. J. Fijalkowski and W. Grochala, J. Mater. Chem., 2009,
1
of the C -like anionic moieties in the crystal structure
19, 2043.
model, which improve fit parameters. We have not tried
this due to a generally weak diffraction was obtained from
this sample.
3 K. J. Fijalkowski, R. Jurczakowski, W. Kozminski and
W. Grochala, Phys. Chem. Chem. Phys., 2012, 14, 5778.
2
2
2
2
4 H. I. Schlesinger and A. B. Burg, J. Am. Chem. Soc., 1938, 60, 290. 33 E.g., in our paper, we have noticed an unknown impurity on
5 I. C. Evans, PhD dissertation, University of Birmingham, 2011.
6 K. R. Ryan, PhD dissertation, University of Oxford, 2011.
NaLi(AB)
2
, which was denoted as ‘‘P1 phase’’ (ref. 17) which
has now been identified as Na(B3N2).
7 W. C. Ewing, P. J. Carroll and L. G. Sneddon, Inorg. Chem., 34 C. Wu, G. Wu, Z. Xiong, W. I. F. David, K. R. Ryan, M. O.
2
013, 52, 10690.
8 H. Wu, W. Zhou and T. Yildirim, J. Am. Chem. Soc., 2008,
30, 14834.
Jones, P. P. Edwards, H. Chu and P. Chen, Inorg. Chem.,
2010, 49, 4319.
35 Cf.: Y. S. Chua, P. Chen, G. Wu and Z. Xiong, Chem.
Commun., 2011, 47, 5116, and references therein.
36 M. A. Neumann, J. Appl. Crystallogr., 2003, 36, 356.
2
2
3
3
1
9 J.-Ph. Souli ´e , G. Renaudin, R. Cern ´y and K. Yvon, J. Alloys
Compd., 2002, 346, 200.
ˇ
ˇ
0 R. S. Kumar and A. L. Cornelius, Appl. Phys. Lett., 2005, 37 V. Favre-Nicolin and R. Cern ´y , J. Appl. Crystallogr., 2002,
7, 261915. 35, 734.
1 Crystallographic density of both M(B3N2) salts is quite low, 38 V. Petricek, M. Dusek and L. Palatinus, Jana2006. Structure
8
À3
À3
about 0.96 g cm for M = Li and 1.09 g cm for M = Na, but
in the range typical for the H-rich salts of these metals.
Determination Software Programs, Institute of Physics, Praha,
Czech Republic, 2006.
This journal is ©the Owner Societies 2014
Phys. Chem. Chem. Phys.