Tuning hydrogen storage properties and reactivity: Investigation of the LiBH4NaAlH4 system

Article abstract of DOI:10.1016/j.jpcs.2010.03.023

Tuning the hydrogen storage properties of complex metal hydrides is of vast interest. Here, we investigate the hydrogen release and uptake pathways for a reactive hydride composite, LiBH₄-NaAlH₄ utilizing in situ synchrotron radiation powder X-ray diffraction experiments. Sodium alanate transforms to sodium borohydride via a metathesis reaction during ball milling or by heating at T～95 °C. NaBH₄ decomposes at ～340 °C in dynamic vacuum, apparently directly to solid amorphous boron and hydrogen and sodium gas and the latter two elements are lost from the sample. Under other conditions, T=400 °C and p(H₂)=～1 bar, NaBH₄ only partly decomposes to B and NaH. On the other hand, formation of LiAl is facilitated by dynamic vacuum conditions, which gives access to the full hydrogen contents in the LiBH₄-NaAlH₄ system. Formation of AlB₂ is observed (T～450 °C) and other phases, possibly AlB_x or Al_1-xLi_xB₂, were observed for the more Li-rich samples. This may open new routes to the stabilization of boron in the solid state in the dehydrogenated state, which is a challenging and important issue for hydrogen storage systems based on borohydrides.

Full text of DOI:10.1016/j.jpcs.2010.03.023

ARTICLE IN PRESS
Journal of Physics and Chemistry of Solids 71 (2010) 1144–1149
Contents lists available at ScienceDirect
Journal of Physics and Chemistry of Solids
journal homepage: www.elsevier.com/locate/jpcs
Tuning hydrogen storage properties and reactivity: Investigation of the
LiBH4–NaAlH4 system
Dorthe B. Ravnsbæk, Torben R. Jensen n
Interdisciplinary Nanoscience Center (iNANO) and Department of Chemistry, University of Aarhus, Langelandsgade 140, DK-8000 Aarhus C, Denmark
a r t i c l e i n f o
a b s t r a c t
Tuning the hydrogen storage properties of complex metal hydrides is of vast interest. Here, we
investigate the hydrogen release and uptake pathways for a reactive hydride composite, LiBH4ꢀNaAlH4
utilizing in situ synchrotron radiation powder X-ray diffraction experiments. Sodium alanate transforms
to sodium borohydride via a metathesis reaction during ball milling or by heating at Tꢁ95 1C. NaBH4
decomposes at ꢁ340 1C in dynamic vacuum, apparently directly to solid amorphous boron and
hydrogen and sodium gas and the latter two elements are lost from the sample. Under other conditions,
T¼400 1C and p(H2)¼ ꢁ1 bar, NaBH4 only partly decomposes to B and NaH. On the other hand,
formation of LiAl is facilitated by dynamic vacuum conditions, which gives access to the full hydrogen
contents in the LiBH4ꢀNaAlH4 system. Formation of AlB2 is observed (Tꢁ450 1C) and other phases,
possibly AlBx or Al1ꢀxLixB2, were observed for the more Li-rich samples. This may open new routes to
the stabilization of boron in the solid state in the dehydrogenated state, which is a challenging and
important issue for hydrogen storage systems based on borohydrides.
Keywords:
A. Inorganic compounds
C. X-ray diffraction
D. Phase transitions
D. Thermodynamic properties
& 2010 Elsevier Ltd. All rights reserved.
1. Introduction
enthalpy [13,14]. The latter principle (iii) forms the basis for the
work presented in this paper, where the combined use of two
interesting hydrogen storage materials are explored, namely
lithium borohydride LiBH4 and sodium alanate, NaAlH4.
In vacuum, LiBH4 decomposes above 375 1C into LiH, B and H2
according to Eq. (1).
Developments of new energy concepts are essential for the
future of the modern industrialized society and hydrogen has been
suggested as a future carrier of renewable energy. Unfortunately,
the storage of hydrogen remains an unsolved and extremely
challenging task for a future hydrogen economy, especially for
utilization of hydrogen in mobile applications [1,2]. Solid state
hydrogen storage as light element hydrides appears to be the only
potential solution, which may provide sufficiently high energy
density and efficiency [3]. However, no single metal hydride has
yet been identified, which fulfills all criteria concerning energy
density, efficiency and safety for mobile applications. Therefore,
research has now been concentrated on three major strategies: (i)
design and characterization of completely novel materials, e.g.
M1–M2–BH4, where M1 typically is an alkali metal and M2 is a d-
block metal, e.g. Zn or Sc [4–7], (ii) utilization of additives, which
can catalyze hydrogen release and uptake, e.g. the use of TiCl3 in
NaAlH4 [8,9] and (iii) design of so-called reactive hydride
composites (RHCs). The basic idea behind RHC systems is that
two or more hydrides may react and form a new intermediate that
may facilitate thermodynamic and kinetic properties of hydrogen
release and uptake [10–12]. Lately, di- and hexaborides have
received much attention due to the discovery that light complex
metal borohydrides can be hydrogenated with reduced reaction
LiBH4(s)-LiH(s)+B(s)+3/2H2(g)
(1)
However, hydrogen uptake requires elevated temperatures
and pressures, T¼600 1C and p(H2)¼155 bar, also illustrated by
the relatively large enthalpy change for reaction (1),
D
H¼68.6 kJ/
mol H2 [15,16]. The reaction scheme (1) also indicates that three-
quarters of the hydrogen is accessible under moderate condition,
i.e. rm¼13.4 w/w%. On the other hand, sodium alanate, NaAlH4,
can reversibly store hydrogen at moderate conditions, according
to the two-step reaction shown in reaction schemes (2) and (3).
3NaAlH4(s)-Na3AlH6(s)+2Al(s)+3H2(g)
Na3AlH6(s)-Al(s)+3NaH(s)+3/2H2(g)
(2)
(3)
The storage capacity is limited, and again, three-quarters
of the hydrogen is accessible under moderate condition,
i.e. rm¼5.5 w/w%, however the reactions have a lower enthalpy
change,
DH¼56.5 kJ/mol H2 [3].
The composite 2LiBH4ꢀAl has been found to decompose under
formation of AlB2, and this system is more easily rehydrogenated
at T¼400 1C and p(H2)¼100 bar, as compared to LiBH4 [17–21].
n Corresponding author. Tel.: +45 8942 3894; fax: +45 8619 6199.
E-mail address: trj@chem.au.dk (T.R. Jensen).
0022-3697/$ - see front matter & 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jpcs.2010.03.023

ARTICLE IN PRESS
D.B. Ravnsbæk, T.R. Jensen / Journal of Physics and Chemistry of Solids 71 (2010) 1144–1149
1145
Unfortunately, the hydrogen storage capacity is reduced con-
siderably to rm¼11.4 w/w% by addition of the inert metal Al.
The basic idea in this research project is to explore the utilization
of other metal hydrides as Al source. Careful selection of the
components in the RHC system can provide relatively high
gravimetric hydrogen storage capacity, which will be the weighted
average of the individual components, e.g. the system 2LiBH4ꢀ
MgH2 has gravimetric H2 storage capacity, rm¼14.4 w/w%. Here,
we investigate the mixed complex hydride system, LiBH4ꢀNaAlH4,
i.e. NaAlH4 is utilized as Al source.
then cooled to 130 1C (
1 h in order to rehydrogenate NaAlH4. Afterwards the sample was
cooled to RT ( T/
30 s per powder X-ray diffraction pattern. The FIT2D program was
used to remove diffraction spots from the single crystal sapphire
tube and to transform raw data to powder patterns [26].
D
T/
D
t¼10 1C/min) and kept at 130 1C for
D
D
t¼10 1C/min). The X-ray exposure time was
3. Results and discussion
3.1. Preparation of the reactive hydride composite systems
Fig. 1 shows PXD data measured at RT for the three samples
after preparation as described in Table 1. The sample prepared by
hand-mixing in mortar (s3) shows diffraction from the reactants,
i.e. LiBH4 and NaAlH4. In contrast, the samples prepared by ball
milling reveal diffraction from both LiAlH4 and NaBH4, hence a
metathesis reaction (4) occurs [27].
2. Experimental
2.1. Sample preparation
Samples of lithium borohydride LiBH4 and sodium alanate
NaAlH4 in molar ratios of 1:0.5 and 1:1.5, respectively, were
prepared using ball milling as preparative method. Titanium-
diboride, TiB2 (2 mol%) was added to both samples prior to BM.
The samples were ball milled for 120 min under inert conditions
(argon atmosphere) using a Fritsch Pulverisette no. 4 at 390 rpm
and a sample to ball ratio of approximately 1:38 using WC vessel
(80 mL) and WC balls (10 mm). Furthermore, a sample of LiBH4
and NaAlH4 in molar ratio 1:1.5 was prepared by thoroughly
mixing the reactants in a mortar for approximately 10 min,
see Table 1. The chemicals used were LiBH4 (Z90%, Aldrich),
NaAlH4 (Z90%, Aldrich) and TiB2 (Aldrich). All handling and
manipulation of the chemicals were performed in an argon-filled
glove box with circulation purifier, p(O2, H2O) o0.1 ppm.
LiBH4(s)+NaAlH4(s)-LiAlH4(s)+NaBH4(s)
(4)
Furthermore, the reactant being in excess as compared to the
1:1 reaction indicated in Eq. 4 is also observed, i.e. LiBH4 for
sample s1 and NaAlH4 for sample s2. Diffraction from small
amounts of TiB2 is also observed for samples s1 and s2.
3.2. Reactivity of the hydride composites
In situ SR-PXD is utilized to study the reaction pathways of
both the hydrogen release and the uptake in the three samples of
LiBH4ꢀNaAlH4 described in Table 1. Fig. 2 shows the data for the
2.2. In situ time resolved synchrotron radiation powder X-ray
diffraction (SR-PXD)
Data were collected at beamline I711 at the synchrotron MAX
II, Lund, Sweden in the research laboratory MAX-Lab with a
MAR165 CCD detector system [22]. The selected wavelengths
˚
were 1.097 and 1.098 A. The sample cell was specially developed
for studies of gas/solid reactions and allows high pressure and
temperature to be applied [23–25]. The powdered samples were
mounted in a sapphire single crystal tube (Al2O3, o.d. 1.09 mm, i.d.
0.79 mm) in an argon-filled glove box p(O2, H2O) o1 ppm. The
temperature was controlled with a thermocouple placed in the
sapphire tubeꢁ1 mm from the sample. A gas supply system was
attached to the sample cell, which allowed the change in gas and
pressure via a vacuum pump during X-ray data acquisition. The
system was flushed with argon and evacuated three times before
the valve to the sample was opened prior to the X-ray experiment.
Dehydrogenation reactions were studied in dynamic vacuum
(sample connected to vacuum pump) or at p(Ar) ꢁ1 bar in a
closed system, i.e. p(H2) increases slightly as the sample desorbs
hydrogen. For dehydrogenation, samples were heated from RT
to 500 1C
conducted at p(H2) ꢁ110 bar. The samples were heated from RT
to 400 1C ( T/
(
D
T/
D
t¼10 1C/min). Hydrogenation reactions were
Fig. 1. PXD data for (a): LiBH4+NaAlH4, 1:0.5 (s1) after BM, (b): LiBH4+NaAlH4,
1:1.5 (s2) after BM and (c): LiBH4+NaAlH4, 1:1.5 (s3) after physical mixing in a
D
D
t¼10 1C/min). The temperature was kept constant
˚
˚
J
mortar (a and b:
l
¼1.097 A, and c:
l
¼1.098 A). Symbols: LiAlH4, ’ NaBH4, m o-
n
LiBH4,K NaAlH4 and TiB2.
at 400 1C for 1 h in order to rehydrogenate LiBH4. The sample was
Table 1
Investigated samples prepared either by hand-mixing in a mortar (M) or by ball milling (BM). The theoretical gravimetric H2 storage capacity,
sample and the products obtained after preparation are listed. For samples s1 and s2, the metathesis reaction, shown in Eq. (4), takes place.
rm, is calculated for each
No.
Molar ratio of LiBH4:NaAlH4
TiB2 (mol%)
Preparation
qm (w/w%)
Reaction products
s1
s2
s3
1:0.5
1:1.5
1:1.5
2
2
0
BM
BM
M
11.9
9.5
LiBH4, LiAlH4, NaBH4.
LiAlH4, NaBH4, NaAlH4.
LiBH4, NaAlH4.
9.8

ARTICLE IN PRESS
1146
D.B. Ravnsbæk, T.R. Jensen / Journal of Physics and Chemistry of Solids 71 (2010) 1144–1149
After cooling to RT only LiAl, AlB2 and small amounts of LiH are
observed. No sodium-containing compounds are observed. How-
ever, metallic Na might have formed as a dehydrogenation
product from NaBH4 due to the use of dynamic vacuum (NaH is
reported to decompose at ꢁ425 1C [30]). Metallic Na (Tm¼98 1C,
Tb¼883 1C) might evaporate at the relative high temperature and
low pressure used for dehydrogenation.
The dehydrogenation of the sample containing LiBH4ꢀNaAlH4
in the molar ratio 1:1.5 (s2) proceeds through a somewhat
different reaction pathway due to the presence of excess NaAlH4.
A dehydrogenation was first conducted in p(Ar) ꢁ1 bar (Fig. 3),
which was followed by a dehydrogenation of the same sample in
dynamic vacuum (Fig. 4). At RT this sample contains LiAlH4,
NaBH4, NaAlH4 and TiB2. The formation of Li3AlH6 and Al from
LiAlH4 is observed at ꢁ110 1C (see Eq. 5) during the first
dehydrogenation conducted in p(Ar) ꢁ1 bar, Fig. 3. At this
temperature, a decrease in the diffracted intensity from NaAlH4
is also observed and in the temperature range 120–200 1C only
Li3AlH6, NaBH4 and TiB2 are observed by SR-PXD. At ꢁ200 1C,
LiNa2AlH6 forms simultaneously with the formation of LiH and Al.
The fact that these compounds are formed and that no significant
change in the diffracted intensity from NaBH4 is observed
suggests that molten NaAlH4 reacts with Li3AlH6 according
to Eq. 7.
Fig. 2. In situ SR-PXD data for the dehydrogenation of LiBH4–NaAlH4 1:0.5 (s1) in
˚
dynamic vacuum and heated from RT to 500 1C,
D
W
T/
D
t¼10 1C/min (
l
¼1.097 A).
J
Symbols: LiAlH4,’ NaBH4, m o-LiBH4, . h-LiBH4,
Li3AlH6, & Al+LiH,
AlB2,
n
~ LiAl and TiB2.
molar ratio 1:0.5 (s1). At RT LiAlH4, NaBH4, LiBH4 and TiB2 are
observed and at ꢁ105 1C the phase transformation from
orthorhombic LiBH4, denoted as o-LiBH4 to hexagonal LiBH4,
denoted as h-LiBH4 is observed simultaneously with the
formation of Li3AlH6 and Al from LiAlH4 according to Eq. (5).
Li3AlH6(s)+2NaAlH4(l)-LiNa2AlH6(s)+2LiH(s)+2Al(s)+3H2(g) (7)
LiNa2AlH6 decomposes at ꢁ290 1C into Al, LiH and NaH as well
as H2 gas. Upon further heating diffraction from both NaBH4 and
NaH vanish at ꢁ340 1C, as was also observed for NaBH4 during
dehydrogenation of sample s1. No major changes in the diffracted
intensity are observed in the temperature range 340–500 1C.
After cooling the sample (Fig. 3) to RT diffractions from Al, LiH,
NaH and NaBH4 are observed (see Fig. 4). Observation of the latter
two compounds suggests that these compounds re-crystallize
upon cooling of the sample. However, the desired dehydrogena-
tion product LiAl, which allows utilization of the full hydrogen
content of LiBH4 is not observed. Therefore, a second dehydro-
genation was conducted in dynamic vacuum, see Fig. 4. At ꢁ300
3LiAlH4(s)-Li3AlH6(s)+2Al(s)+3H2(g)
(5)
Li3AlH6 is observed to decompose at ꢁ180 1C under the
formation of more Al and LiH. It should be noted that the latter
compounds are isostructural and have similar unit cell para-
meters, i.e. overlapping Bragg reflections. However, differences in
the relative peak intensities allow identification of the two
compounds. Upon further heating, the melting of h-LiBH4 is
observed at ꢁ260 1C and the diffraction from NaBH4 vanishes at
ꢁ340 1C, which is significantly lower than the reported decom-
position temperature, Tdec ꢁ400 1C [28]. A Bragg reflection is
observed at 2
an unidentified phase (see Fig. 2), which shows a shift towards
lower diffraction angle 2
y
ꢁ351 in the temperature range 375–475 1C from
y
(at ꢁ425 1C), i.e. larger unit cell
volume. This is only observed for sample s1 possibly due to excess
of LiBH4 and may be due to formation of AlBx or Al1ꢀxLixB2 [29].
Formation of the latter compound might explain the observation
that all LiH is consumed during formation of LiAl. The reactions
observed here may only proceed patially and in this case Li is in
excess relative to Al. The disappearance of diffraction at 2
y
ꢁ351
ꢁ201,
may be related to the broad hump in the background at 2
y
which occurs at T ꢁ450 1C.
At ꢁ450 1C, the alloy LiAl forms from Al and LiH, which is
clearly visible by the disappearance of the peak at 2
This suggests that all Al and LiH are consumed by this process,
which also releases hydrogen (see Eq. (6)). The peak at 2
y
ꢁ271.
y
ꢁ311
contains presumably only diffraction from TiB2 at RT however,
the intensity increases at ꢁ180 1C due to the formation of
Al and LiH. The compounds, TiB2, AlB2, Al and LiH, all have
overlapping Bragg reflections at 2
exhibit quite similar diffraction patterns. The peak at 2
shows a dip in intensity at ꢁ450 1C when LiAl forms, possibly
due to the replacement of LiH and Al by some AlB2, described
in Eq. 6.
y
ꢁ311. In fact TiB2 and AlB2
ꢁ311
y
Fig. 3. In situ SR-PXD data for the first dehydrogenation of LiBH4–NaAlH4 1:1.5
(s2) in p(Ar) ꢁ1 bar (closed system, i.e. p(H2) will increase slightly as the sample
˚
desorbs hydrogen). Heated from RT to 500 1C,
D
T/
D
t¼10 1C/min (
l
¼1.097 A).
NaH, &
J
Symbols:
LiAlH4, ’ NaBH4, K NaAlH4,
n
W
Li3AlH6,
LiNa2AlH6,
2LiH(s)+2B(s)+3Al(s)-AlB2(s)+2LiAl(s)+H2(g)
(6)
Al+LiH and TiB2.

ARTICLE IN PRESS
D.B. Ravnsbæk, T.R. Jensen / Journal of Physics and Chemistry of Solids 71 (2010) 1144–1149
1147
Fig. 5. In situ SR-PXD data for the rehydrogenation of LiBH4ꢀNaAlH4 1:1.5 (s2)
conducted in p(H2) ꢁ110 bar and heated from RT to 400 1C, temperature kept
constant at 400 1C for 1 h, cooled from 400 to 130 1C, temperature kept constant at
Fig. 4. In situ SR-PXD data for the second dehydrogenation of LiBH4ꢀNaAlH4 1:1.5
(s2) in dynamic vacuum and heated from RT to 500 1C,
DT/Dt¼10 1C/min
130 1C for 1 h and finally the sample was cooled to RT. Heating and cooling rates of
˚
(
l
¼1.097 A). The temperature was kept constant at 500 1C for 15 min. Symbols:
˚
D
T/
D
t¼10 1C/min (
l
¼1.097 A). Symbols: m o-LiBH4, . h-LiBH4, & Al+LiH,
AlB2,
n
n
’ NaBH4,
NaH, & Al+LiH,
AlB2, ~ LiAl and TiB2.
~ LiAl and TiB2.
and ꢁ400 1C diffractions from NaH and NaBH4, respectively, are
observed to vanish and only Al and LiH are observed. LiAl forms
from Al and LiH at ꢁ450 1C, which is in accordance with the
observations for s1. Simultaneously, with the formation of LiAl, a
diffracted intensities at 2y ꢁ271 and ꢁ311 indicate that LiH and
AlB2 are consumed as described in Eq. 9.
LiAl(s)+1/2H2(g)-LiH(s)+Al(s)
(8)
(9)
slow increase in the diffracted intensity of the peak at 2
y ꢁ311 is
observed, which suggests that AlB2 is formed. The formation of
AlB2 also gives an increase in intensity in reflections assigned to
TiB2. Since no LiBH4 is present in this sample, AlB2 may form from
a reaction with NaBH4, which suggests formation of NaH. Sodium
hydride decomposes rapidly and molten sodium is formed, which
might evaporate due to the high temperature and low pressure.
Neither Na nor NaH is observed in the PXD data at RT. In contrast,
LiH is still present after the formation of LiAl. Upon cooling to RT,
LiAl, LiH, AlB2 and TiB2 are observed.
2LiH(s)+AlB2(s)+3H2(g)-2LiBH4(l)+Al(s)
Hence, the rehydrogenation occurs via two coupled but
distinct reactions. The first step (Eq. (8)) is relatively fast, while
the second (Eq. (9)) proceeds for approximately 40 min at 400 1C.
Upon cooling, h-LiBH4 is observed to crystallize from the melt at
ꢁ260 1C but no further changes are observed during cooling and
dwelling at 130 1C. At ꢁ100 1C h-LiBH4 transforms to o-LiBH4.
Fig. 1 shows PXD data for s1, s2 and s3 measured before
dehydrogenation, whereas Fig. 6 shows PXD data measured for
the dehydrogenated and the rehydrogenated samples. All PXD
data shown in Figs. 1 and 6 are measured at RT.
The rehydrogenation experiment for the LiBH4ꢀNaAlH4
(1:1.5) s3 prepared by hand-mixing in a mortar also shows
formation of h-LiBH4 at ꢁ260 1C during cooling, as well as the
h- to o-LiBH4 phase transformation (see Fig. 6). Hence, the
LiBH4ꢀNaAlH4 system shows partial reversibility. Fig. 6 shows
the PXD data measured (at RT) for the three samples s1, s2 and s3
for the dehydrogenated state and after rehydrogenation.
Formation of LiAl was not observed for s3. Further experi-
mental work is needed to clarify whether this is due to phase
separation and insufficient mixing of the reactants, contamination
by air or the fact that the catalyst TiB2 was not added to this
sample. Apparently, formation of LiAl alloy is facilitated by
dynamic vacuum as LiAl did not form in the closed system, i.e.
p(Ar) ꢁ1 bar and 0op(H2)o2 bar. On the other hand, sodium
escapes from the sample as vapor at reduced pressures. Therefore,
the LiBH4ꢀNaAlH4 system was only partially reversible in this
study, i.e. the formation of LiBH4 was observed during hydro-
genation but not NaAlH4. Importantly, formation of AlB2 is
observed (Tꢁ450 1C), which stabilizes boron in the solid state
and may be further optimized to eliminate the possible release of
diboran (B2H6) [31]. For the more Li-rich sample (s1) other phases
were also observed, possibly AlBx or Al1ꢀxLixB2 [29]. This may
The LiBH4ꢀNaAlH4 (1:1.5) sample prepared by mortar (s3)
was also studied by in situ SR-PXD and the dehydrogenation
resembles that of the corresponding sample prepared by BM (s2).
Due to the different preparation methods both reactants are
present at RT (see Fig. 1c). However, the metathesis reaction
shown in Eq. (4) and formation of Li3AlH6 (Eq. (5)) occur at
ꢁ95 1C and the compounds NaBH4, Li3AlH6, LiH and Al are
formed. LiNa2AlH6 forms upon further heating, and decomposes
into LiH, NaH and Al as observed for sample s2. Furthermore,
diffractions for NaH and NaBH4 also vanish at ꢁ300 and ꢁ400 1C,
respectively. An unidentified phase is observed at ꢁ480 1C, which
may be an oxide impurity formed by contamination by air. This
may explain that formation of LiAl is not observed in this case,
even though the dehydrogenation is carried out in dynamic
vacuum. At RT after dehydrogenation, Al, LiH and the unidentified
phase are observed.
3.3. Rehydrogenation and reversibility
The samples prepared by BM (s1 and s2) exhibit similar
behavior during the rehydrogenation experiments, namely, the
diffracted intensity from LiAl decreases while LiH and Al form at
ꢁ200 1C, according to Eq. 8 (shown for sample s2 in Fig. 5).
However, the consumption of AlB2 initiates at T ꢁ280 1C. The

ARTICLE IN PRESS
D.B. Ravnsbæk, T.R. Jensen / Journal of Physics and Chemistry of Solids 71 (2010) 1144–1149
Table 2
1148
Comparison of hydrogen storage properties for selected hydrides and reactive
hydride composites.
rm (w/w%) rV (kg H2/m3) Tdec (1C)
Ref.
LiBH4–MgH2 (1:0.5)
LiBH4–Al (1:0.5)
14.4
11.4
119.8
106.3
111.4
104.2
122.5
113.1
96.7
270–440a [32]
320–400a [33]
LiBH4–NaAlH4 (1:0.5) 11.9
105–450a This work
110–450a This work
LiBH4–NaAlH4 (1:1.5)
LiBH4
9.8
18.4
10.6
10.5
7.4
4380b,c
4400b
4125d
4210d
279e
[34]
[34]
[34]
[34]
[35]
NaBH4
LiAlH4
NaAlH4
94.8
MgH2
7.7
110.2
a
Decomposition reaction performed in vacuum.
Experimental values.
b
c
Calculated from thermodynamic data, Tdec¼254 1C [15].
Three reaction steps involved.
d
e
Calculated from thermodynamic data for p(H2)¼1 bar.
Reactive hydride composites can have high hydrogen storage
capacities and for practical purposes more useful thermodynamic
properties. This is illustrated in Table 2 by comparison of some
hydrides with selected RCH systems. Unfortunately, the
composites may suffer from poor kinetics and in some cases
limited reversibility, which call for continued research and
optimization of these systems.
4. Conclusion
In this work, the LiBH4ꢀNaAlH4 system (1:0.5 and 1:1.5) was
investigated using in situ synchrotron radiation powder X-ray
diffraction (SR-PXD). Several (coupled) chemical reactions are
involved in hydrogen release and uptake. Formation of AlB2 is
observed (T ꢁ450 1C), which stabilizes boron in the solid state
and may be optimized further to eliminate the possible release of
diboran (B2H6). The alloy LiAl forms at reduced pressures
(dynamic vaccum) and facilitates release of the full hydrogen
content. The LiBH4ꢀNaAlH4 system is partially reversible at the
conditions used here and rehydrogenation of LiBH4 was per-
formed under moderate conditions, T¼400 1C and p(H2)¼110
bar. In contrast, sodium alanate was not observed to form during
dehydrogenation. Apparently, sodium escapes from the hydrogen
storage system as vapor, which may prevent rehydrogenation of
sodium alanate. This calls for further experimental work in order
to optimize the conditions for the coupled chemical reactions
observed in this study.
Acknowledgement
Dr. Yngve Cerenius is thanked for valuable discussions
regarding the synchrotron measurements. The Synchrotron
radiation facility MAX-II at MAX-lab is thanked for beamtime.
The Danish research councils are thanked for support from the
program DanScatt.
Fig. 6. SR-PXD data measured at RT. (a) and (b): LiBH4ꢀNaAlH4, 1:0.5 (s1)
prepared by BM. Data measured after dehydrogenation and after rehydrogenation,
respectively. (c) and (d): LiBH4ꢀNaAlH4, 1:1.5 (s2) prepared by BM. Data
measured after the second dehydrogenation and after rehydrogenation, respec-
tively. (e) and (f): LiBH4ꢀNaAlH4, 1:1.5 (s3) prepared by mortar. Data measured
after dehydrogenation and after rehydrogenation, respectively. The dehydrogena-
References
[1] L. Schlapbach, A. Zu¨ttel, Nature 414 (2001) 353–358.
[2] J.A. Ritter, A.D. Ebner, J. Wang, R. Zidan, Mater. Today 9 (2003) 18.
[3] S. Orimo, Y. Nakamori, J.R. Eliseo, A. Zuttel, C.M. Jensen, Chem. Rev. 107
(2007) 4111–4132.
tions were performed in dynamic vacuum. Symbols: m o-LiBH4, & Al+LiH,
AlB2,
n
˚
˚
~ LiAl and TiB2 (a to d:
l¼1.097 A, and e and f: l¼1.098 A).
[4] D. Ravnsbæk, Y. Filinchuk, Y. Cerenius, H.J. Jakobsen, F. Besenbacher,
J. Skibsted, T.R. Jensen, Angew. Chem. Int. Ed. 48 (2009) 6659–6663.
[5] D.B. Ravnsbæk, L.H. Sørensen, Y. Filinchuk, D. Reed, D. Book, H.J. Jakobsen,
F. Besenbacher, J. Skibsted, T.R. Jensen, J. Eur. Inorg. Chem. (2010), in press,
doi:10.1002/ejic.201000119.
open new routes to the stabilization of boron in the solid state in
the dehydrogenated state, which is a challenging and important
issue for hydrogen storage systems based on borohydrides.

ARTICLE IN PRESS
D.B. Ravnsbæk, T.R. Jensen / Journal of Physics and Chemistry of Solids 71 (2010) 1144–1149
1149
[6] H. Hagemann, M. Longhini, J.W. Kaminski, T.A. Wesolowski, R. Cerny, N.
Penin, M.H. Sørby, B.C. Hauback, G. Servera, C.M. Jensen, J. Phys. Chem. A 112
(2008) 7551–7555.
[7] Y. Filinchuk, D. Chernyshov, V. Dmitriev, Z. Kristallogr. 223 (2008) 649.
[8] B. Bogdanovic, M. Felderhof, S. Kaskel, A. Pommerin, K. Schlichte, F. Schuth,
Adv. Mater. 15 (2003) 1012–1015.
[21] O. Friedrichs, J.W. Kim, A. Remhof, F. Buchter, A. Borgshulte, D. Wallacher,
Y.W. Cho, M. Fichtner, K.H. Oh, A. Zu¨ttel, Phys. Chem. Chem. Phys. 11 (2009)
1515–1520.
˚
[22] Y. Cerenius, K. Sta˚ hl, L.A. Svensson, T. Ursby, A Oskarsson, J. Albertsson,
A. Lijas, J. Synchrotron Radiat. 7 (2000) 203–208.
[23] B.S. Clausen, G. Steffensen, B. Fabius, J. Villadsen, R. Feidenhans´l, H. Topsøe,
J. Catal. 132 (1991) 524–535.
[9] B.C. Hauback, Z. Kristallogr. 223 (2008) 636–648.
[10] J.J. Vajo, S.L. Skeith, F. Mertens, J. Phys. Chem. B 109 (2005) 3719.
[11] G. Barkhordarian, T. Klassen, M. Dornheim, R. Bormann, J. Alloys Compd. 440
(2007) L18.
[12] G. Barkhordarian, T.R. Jensen, S. Doppiu, U. Bo¨senberg, A. Borgschulte,
R. Gremaud, Y. Cerenius, M. Dornheim, T. Klassen, R. Bormann, J. Phys. Chem.
C 112 (2008) 2743–2749.
[24] J.A. Rodriguez, J.C. Hanson, A.I. Frenkel, J.Y. Kim, M. Perez, J. Am. Chem. Soc.
124 (2002) 346.
[25] D. Ravnsbæk, L. Mosegaard, J.E. Jørgensen, T.R. Jensen, Proceedings of the
29th Risø International Symposium on Materials Science: Energy
Materials—Advances in Characterization, Modelling and Application, 2008,
pp. 349–356.
[13] E. Ro¨nnebro, E.H. Majzoub, J. Phys. Chem. B 111 (2007) 12045–12047.
[14] U. Bo¨senberg, S. Doppiu, L. Mosegaard, G. Barkhordarian, N. Eigen,
A. Borgschulte, T.R. Jensen, Y. Cernius, O. Gutfleich, T. Klassen, M. Dornheim,
R. Bormann, Acta Mater. 55 (2007) 3951.
[15] M.S. Smith, G.E. Bass, J. Chem. Eng. Data 8 (1963) 341.
[16] Ph. Mauron, F. Buchter, A. Borgschulte, O. Friedrichs, A. Remhof, M. Bielmann,
Ch.N. Zwicky, A. Zu¨ ttel, Phys. Chem. B. 112 (2008) 906.
[17] X.-D. Kang, P. Wang, L.-P. Ma, H.-M. Cheng, Appl. Phys. A 89 (2007) 963–966.
[18] S.-A. Jin, J.-H. Shim, Y.W. Cho, K.-W. Yi, O. Zabarac, M. Fichtner, Scr. Mater. 58
(2008) 963–965.
[26] A.P. Hammersley, Internal Report (1997) ESRF-97-HA02T.
[27] Q. Shi, X. Yu, R. Feidenhans’l, T. Vegge, J. Phys. Chem. C 112 (2008) 18244–
18248.
[28] A. Zu¨ttel, S. Rentsch, P. Fischer, P. Wenger, P. Sudan, Ph. Mauron,
Ch. Emmenegger, J. Alloys Compd. 356–357 (2003) 515–520.
[29] T. Sun, Y.G. Zhao, R. Fan, X.P. Zhang, B.G. Liu, Y.H. Xiong, P.J. Li,
J. Superconductivity Incorporating Novel Magn. 17 (2004) 473–480.
[30] W. Grochala, P.P. Edwards, Chem. Rev. 104 (2004) 1283–1315.
[31] J. Kostka, W. Lohstroh, M. Fichtner, H. Hahn, J. Phys. Chem. C 111 (2007)
14026–14029.
[19] A. Remhof, O. Friedrichs, F. Buchter, Ph. Mauron, J.W. Kim, K.H. Oh,
A. Buchsteiner, D. Wallacher, A. Zu¨ttel, J. of Alloys and Compd. 484 (2009)
654–659.
[20] J.W. Kim, O. Friedrichs, J.-P. Ahn, D.H. Kim, S.C. Kim, A. Remhof, H.-S. Chung,
J. Lee, J.-H. Shim, Y.W. Cho, A. Zu¨ttel, K.H. Oh, Scr. Mater. 60 (2009)
1089–1092.
[32] J.J. Vajo, S.L. Skeith, F. Mertens, J. Phys. Chem. B 109 (2005) 3719–3722.
[33] J. Yang, A. Sudik, C. Wolverton, J. Phys. Chem. C 111 (2007) 19134.
[34] A. Zu¨ttel, S. Rentsch, P. Fischer, P. Wenger, P. Sudan, Ph. Mauron,
Ch. Emmenegger, J. Alloys. Compd. 356–357 (2003) 515–520.
[35] J.F. Stampfer, C.E. Holley, J.F. Suttle, J. Am. Chem. Soc. 82 (1960)
3504–3508.