A new dehydrogenation mechanism for reversible multicomponent borohydride systems - The role of Li-Mg alloys

Article abstract of DOI:10.1039/b607869a

A new dehydrogenation mechanism for LiBH₄-MgH₂ mixtures revealed that magnesium destabilised the LiBH₄ resulting in complete dehydrogenation of the borohydride phase and the formation of a Li-Mg alloy. The Royal Society of Chemistry 2006.

Full text of DOI:10.1039/b607869a

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
A new dehydrogenation mechanism for reversible multicomponent
borohydride systems—The role of Li–Mg alloys{
ab
X. B. Yu, D. M. Grant and G. S. Walker*
a
a
Received (in Cambridge, UK) 2nd June 2006, Accepted 19th July 2006
First published as an Advance Article on the web 8th August 2006
DOI: 10.1039/b607869a
A new dehydrogenation mechanism for LiBH
4
–MgH
2
mix-
decomposition temperature of ball milled MgH . Further heating
2
tures revealed that magnesium destabilised the LiBH
4
resulting
led to a second decomposition peak at 405 uC. This led to a region
of steady hydrogen desorption in the temperature range of 430–
in complete dehydrogenation of the borohydride phase and the
formation of a Li–Mg alloy.
5
80 uC. The TG results suggest that the weight loss of LiBH4–
MgH mixture is in three steps. The first step is around 360 uC,
2
Using hydrogen as an energy carrier is an environment friendly
approach that produces almost no local emission of pollutants
from power generators such as fuel cells. However, for the last
several decades the efficient and safe storage and transportation of
hydrogen have been major concerns in the use of hydrogen as an
energy carrier and there is significant interest in solid state
corresponding to the first hydrogen desorption peak, with a weight
loss of 5.7 wt%, which is comparable with the theoretical hydrogen
capacity of MgH in the mixture (5.8 wt%). The weight loss of the
2
second and the third steps are 2.65 wt% and 0.85 wt%, respectively,
this combined weight loss is in good agreement with the hydrogen
capacity from the LiBH (3.6 wt%). The total weight loss observed
4
1,2
hydrogen storage materials. A design target for automobile
fueling has been set by the U.S. Department of Energy at 6.5 wt%.
Because it is unlikely that capacities of .6 wt% can be obtained in
for the LiBH –MgH mixture is equivalent to their theoretical
4
2
hydrogen capacities, suggesting that all the hydrogen was released
from the mixture below 600 uC. The TG data for the
dehydrogenation of ball milled LiBH4 is given in Fig. 1 for
comparison and clearly shows that the dehydrogenation tempera-
ture for the borohydride has been significantly reduced.
3–7
transition metal-based materials, intense interest has developed
2
2
4 2
in complex hydrides such as alanates (AlH ), amides (NH ),
2 8–14
4
).
and borohydrides (BH
For LiBH , full decomposition to Li + B + 2H
8.3 wt% hydrogen, exhibiting promising prospects for on-board
applications. However, the main evolution of gas starts at 380 uC
4
2
can yield
The fact that the first weight loss is equivalent to the hydrogen
1
capacity of MgH
indicative that the MgH2 and LiBH4 decomposed separately.
Therefore, the improved dehydrogenation kinetics for the LiBH
2 4 2
in the LiBH –MgH mixture is strongly
1
5
and only releases half the hydrogen below 600 uC. Furthermore,
to reverse the reaction is a significant challenge because of the
extremely rigorous reaction conditions required (high pressure and
4
was likely due to the Mg formed. To investigate the effect of Mg
on the dehydrogenation of LiBH a ball milled LiBH –Mg sample
4
4
1
5
high temperature). Z u¨ ttel et al. reported that SiO
a catalyst for the dehydrogenation of LiBH
temperature of hydrogen evolution to 300 uC. Pinkerton et al.
also reported that LiBH could react with LiNH to form
Li BN . This quaternary hydride released 10 wt% hydrogen
above y250 uC. However, neither of the above reactions have
2
may be used as
with a mass ratio of 1 : 4 was prepared. Fig. 2 shows the TG-MS
4
, lowering the
results with a hydrogen desorption peak at 405 uC, which was the
16
same temperature as that for the decomposition of LiBH
LiBH –MgH samples, indicating that the Mg metal ball milled
with LiBH resulted in a similar effect on the decomposition of
4
in the
4
2
4
2
3
2 8
H
4
17
been shown to be reversible. More recently, Vajo et al. found
that LiBH may be reversibly dehydrogenated and rehydrogenated
4
with a reduced reaction enthalpy by the addition of MgH
thought that the formation of MgB stabilized the dehydrogenated
state and effectively destabilized the LiBH . According to the
above hypothesis, the formation of MgB should be accompanied
with the decomposition of LiBH . However, our results presented
2
. They
2
4
2
4
here reveal that the decomposition of the borohydride resulted in
the formation of a Li–Mg alloy, prior to the appearance of MgB2.
The MS and TG profiles in Fig. 1 show the hydrogen desorp-
tion properties of ball milled LiBH –MgH (mass ratio, 1 : 4). The
4
2
first hydrogen desorption peak appeared at 354 uC, similar to the
a
School of Mechanical, Materials and Manufacturing Engineering,
University of Nottingham, University Park, Nottingham, UK NG7 2RD.
E-mail: Gavin.Walker@Nottingham.ac.uk
Shanghai Institute of Microsystem and Information Technology,
b
Fig. 1 (a) TG and MS results for the evolution of H
MgH milled for 1 h. (b) TG results for LiBH . Both experiments used a
heating rate of 10 uC min
2 4
from LiBH –
Shanghai 200050, P. R. China. E-mail: yuxuebin@hotmail.com
Electronic supplementary information (ESI) available: Material prepar-
ing and measurement, ESI 1, ESI 2, ESI 3. See DOI: 10.1039/b607869a
{
2
4
21
.
3
906 | Chem. Commun., 2006, 3906–3908
This journal is ß The Royal Society of Chemistry 2006

View Article Online
Table 1 Comparison of d-spacings for LiBH
1
6
4
–MgH
: 4) heated to 360 uC (S1), 405 uC (S2), 440 uC (S3), 500 uC (S4) and
00 uC (S5) with Mg, Li0.184Mg0.816 (ICSD: 104740), Li, Li0.3Mg0.7
2
(mass ratio,
(
ICSD: 104741)
˚
d-Spacings (A)
Sample
(100)
(002)
(101)
(110)
(200)
(211)
Mg
Li0.184Mg0.816
S1
S2
S3
2.779
2.764
2.778
2.777
2.772
2.765
—
2.606
2.566
2.603
2.603
2.591
2.566
—
2.452
2.434
2.450
2.451
2.444
2.434
—
—
—
—
—
—
—
2.482
2.487
2.488
—
—
—
—
—
—
1.755
1.756
1.757
—
—
—
—
—
—
1.433
1.436
1.437
S4
Li
Li0.3Mg0.7
S5
—
2.768
—
2.572
—
2.438
Fig. 2 TG and MS results for the evolution of H
2
from LiBH –Mg
4
2
and H during the first dehydrogenation step. There appears to be
21
milled for 1 h. Heating rate was 10 uC min
.
little change in the XRD pattern after heating to 405 uC, however,
when heated to 440 uC and 500 uC, a shift to higher 2h values can
be observed in the position of the XRD reflections. The pattern at
LiBH as did MgH . The hydrogen released by 600 uC gave a total
4
2
18
.
weight loss of 3.35 wt%, lower than the 3.66 wt% expected from
500 uC matches the d-spacing for Li0.184Mg0.816
The small
the LiBH fraction of the sample. The lower wt% of hydrogen
4
amount of MgO was caused by oxidation during loading. No
released is likely to be due to partial decomposition of LiBH
4
MgB was formed, but after further heating to 600 uC, the XRD
2
18
during ball milling. The strong correlation of the effect of Mg on
the decomposition of LiBH4 with that in the LiBH –MgH
pattern can be seen to consist of Li0.30Mg0.70
,
Li0.184Mg0.816,
4
2
MgO, and MgB , showing that MgB only formed at the higher
2
2
1
9
samples shows that the main reaction in LiBH –MgH
4
2
includes
temperature. Table 1 presents the d-spacing data comparing the
experimental results with the reference data for Mg, Li0.184Mg0.816
two processes: one is the decomposition of MgH , and the other is
2
,
the reaction of Mg with LiBH
desorption from LiBH
XRD results for the as prepared LiBH
4
, which facilitates the hydrogen
Li0.30Mg0.70 and Li.
4
.
The 1 : 4 LiBH –MgH sample dehydrogenates in three steps.
4
2
4
–MgH
2
and after
The first step is the dehydrogenation of the MgH phase and
2
heating to temperatures up to 600 uC are shown in Fig. 3. The
XRD pattern for the as prepared sample only had a weak diffuse
results in the formation of Mg metal. The LiBH4 phase
dehydrogenates at 405 uC. Unfortunately because of the low
content and poor crystallinity of this phase it was not possible to
follow the loss of this phase by XRD. As there was no apparent
change to the Mg phase during the 2nd dehydrogenation step, this
indicates that the Mg acts as a catalyst. Assuming that the 2nd and
3rd dehydrogenation steps were due to the decompostion of the
borohydride phase, the TG data shows that 76% of the hydrogen
was produced during the 2nd dehydrogenation step and 24%
during the 3rd. This corresponds well with the expected 3 : 1 ratio
2
pattern corresponding to MgH . No pattern was detected for the
borohydride phase because this was in much lower concentration
and would also be nanocrystalline/disordered after the ball milling
process. After heating to 360 uC the XRD pattern corresponds to
2
Mg metal, suggesting that only MgH decomposed forming Mg
4
for hydrogen evolved from the dehydrogenation of LiBH yielding
LiH and B and the subsequent dehydrogenation of the LiH. LiH
does not normally dehydrogenate until temperatures above 600 uC
(ESI 1{), thus the presence of Mg also destabilises the LiH phase.
At temperatures from 440 uC and higher the formation of Li–Mg
alloy was evidenced in the XRD data. The Li–Mg alloy must be
forming directly from the dehydrogenation of the LiH as this
reaction is occuring at temperatures well above the vaporisation
temperature of Li. The Li–Mg mole fraction for this system is
0.23 Li. This is within a two phase region consisting of a HCP a
phase with Li content of 0.184 and BCC b phase with a Li content
2
0
of 0.30. Therefore as the LiH decomposes, an a phase with
increasing Li content forms until saturation is reached whence the
b phase forms. A solid state reaction between Mg and B has
occured by 600 uC forming MgB , which will also deplete the
2
available Mg for the Li–Mg system, resulting in further formation
of the b alloy. This reaction mechanism was supported by the fact
that the ratio of the 2nd and 3rd dehydrogenation steps remained
Fig. 3 XRD patterns for the LiBH
dehydrogenation to 360 uC (S1), 405 uC (S2), 440 uC (S3), 500 uC (S4) and
00 uC (S5).
4 2
–MgH mixture before (S0) and after
6
4 2
the same for different ratios of LiBH –MgH (ESI 2{) and that the
This journal is ß The Royal Society of Chemistry 2006
Chem. Commun., 2006, 3906–3908 | 3907

View Article Online
4 4 2
LiBH . The total weight loss for the rehydrogenated LiBH –MgH
sample was about 8.5 wt% (inserted TG curve, Fig. 4), comparing
well with the first dehydrogenation cycle (9.2 wt%). The XRD
results of the rehydrogenated sample also revealed the formation
of LiBH (ESI 3{). The effect of multiple cycles is currently being
4
investigated.
In summary, it has been demonstrated that the hydrogen
desorption in LiBH
MgH decomposes to Mg and H
dehydrogenation of LiBH forming LiH and B, but the presence
of Mg also expedites the dehydrogenation of LiH resulting in the
formation of Li–Mg phases. MgB was only formed about 500 uC.
Dehydrogenated LiBH –MgH mixture can be rehydrogenated at
00 bar hydrogen pressures and 400 uC and the reversibility of the
reaction does not depend on the formation of MgB
4
–MgH
2
mixtures include three steps: first,
2
2
, then the Mg catalyses the
4
2
4
2
1
2
.
The authors would like to acknowledge the support of the
EPSRC who funded this work and collaborative arrangements
with the Shanghai Institute of Microsystems and Information
Technology, CAS. Author Xuebin Yu also acknowledges the
support of the Shanghai Rising-Star Program (project number:
05QMX1463).
Fig. 4 MS and TG results (inserted) for initial and rehydrogenated
21
4 2
LiBH –MgH with a heating rate of 10 uC min .
LiH phase in a ball milled control experiment of LiH–MgH (mass
2
ratio 1 : 9 corresponding to a mole fraction of 0.27 Li, also within
the two phase region) started dehydrogention just above 400 uC
and resulted in the same Li–Mg phases being formed (ESI 1{).
Notes and references
1
2
3
4
L. Schlapbach and A. Z u¨ ttel, Nature, 2001, 414, 353.
W. Grochala and P. P. Edwards, Chem. Rev., 2004, 104, 1283.
R. Li, J. M. Wu and X. L. Wang, J. Alloys Compd., 2000, 311, 40.
B. K. Singh, A. K. singh and O. N. Srivastava, Int. J. Hydrogen Energy,
1996, 21, 111.
4
Based on the above analysis, the dehydrogenation for LiBH –
MgH
2
mixture is postulated to be:
0
*360
C
MgH DCCA MgzH2
(1)
(2)
2
5 T. Mouri and H. Iba, Mater. Sci. Eng., A, 2002, 329–331, 346.
6 M. Tsukahara, K. Takahashi, T. Mishima, T. Sakai, H. Miyamura,
N. Kuriyama and I. Uehara, J. Alloys Compd., 1995, 226, 203.
0
*
405 C, Mg
0
:30 LiBH4 DCCA 0:30 LiHz0:30 Bz0:45 H2
7
X. B. Yu, Z. Wu, F. Li, B. J. Xia and N. X. Xu, Appl. Phys. Lett., 2004,
4, 3199.
8 B. Bogdanovi c´ , M. Felderhoff, S. Kaskel, A. Pommerin, K. Schlichte
8
w420 0
C
Mgz0:30 LiHz0:30 B DCCA
and F. Sch u¨ th, Adv. Mater., 2003, 15, 1012.
(
3)
9
S. Orimo, Y. Nakamori and A. Zuttel, Mater. Sci. Eng., B, 2004, 108,
1.
0 Z. Xiong, G. Wu, J. Hu and P. Chen, Adv. Mater., 2004, 16, 1522.
0
:37 Li0:184Mg0:816z0:15 MgB z0:78 Li0:30Mg0:70z0:15 H2
2
5
1
The total reaction equation is:
0
0
11 P. Wang and C. M. Jensen, J. Phys. Chem. B, 2004, 108, 15827.
12 A. M. Seayad and D. M. Antonelli, Adv. Mater., 2004, 16, 765.
3
60 C{600 C
MgH z0:30 LiBH4 DCCA
2
1
1
1
3 P. Chen, Z. Xiong, J. Lou, J. Lin and K. L. Tan, Nature, 2002, 420, 302.
4 C. M. Jensen and K. Gross, J. Appl. Phys A, 2001, 72, 213.
5 A. Z u¨ ttel, P. Wenger, S. Rentsch, P. Sudan, Ph. Mauron and
Ch. Emmenegger, J. Power Sources, 2003, 118, 1.
0
:37 Li0:184Mg0:816z0:15 MgB z0:78 Li0:30Mg0:70z1:60 H2
2
In addition we have shown that the dehydrogenation LiBH
4
–
MgH can be reversed without the formation of MgB . A LiBH –
4
2
2
16 F. E. Pinkerton, G. P. Meisner, M. S. Meyer, M. P. Balogh and
MgH sample (mass ratio, 1 : 4) which had been dehydrogenated
2
M. D. Kundrat, J. Phys. Chem. B, 2005, 109, 6.
17 J. J. Vajo, S. L. Skeith and F. Mertens, J. Phys. Chem. B, 2005, 109,
under a vacuum up to 400 uC (for which XRD results showed no
3719.
MgB phase (ESI 3{)) was rehydrogenated at 400 uC under 100 bar
2
1
8 F. H. Herbstein and B. L. Averbach, Acta Metallurgica, 1956, 4, 407.
Reference data from the Inorganic Crystal Structure Database (http://
cds.dl.ac.uk/icsd/) ICSD-104740 (Li0.184Mg0.816), and ICSD-104741
hydrogen pressure. Fig. 4 shows the TG-MS results for hydrogen
4 2
evolution from the rehydrogenated LiBH –MgH sample (the MS
curve for an initial dehydrogenation is also shown for compar-
ison). The first hydrogen desorption peak corresponds to the
(Li
9 In Vajo’s results, MgB
our results. The main reasons might result from their different
experimental conditions most notably the presence of TiCl and H
atmosphere. It is possible that the presence of TiCl and H accelerated
the formation of MgB at lower temperatures.
0 Constitution of binary alloys, ed. M. Hansen, McGraw-Hill, New York,
2nd edn., 1958, pp. 897–898.
0 3
. Mg0.7).
1
2
was formed below 500 uC, lower than that for
decomposition of MgH located at 340 uC, which is slightly lower
2
3
2
than that of the initial sample, suggesting that the dehydrogena-
3
2
tion–rehydrogenation process has further improved the kinetics of
+
, potentially due to lattice defects from the presence of Li in
2
2
MgH
2
21
the MgH lattice as postulated by Johnson et al. The second
2
2
1 S. R. Johnson, P. A. Anderson, P. P. Edwards, I. Gameson,
J. W. Prendergast, M. Al-Mamouri, D. Book, I. R. Harris,
J. D. Speight and A. Walton, Chem. Commun., 2005, 22, 2823.
hydrogen desorption peak at 405 uC, coinciding with that of the
initial sample, corresponds to the hydrogen desorption from
3
908 | Chem. Commun., 2006, 3906–3908
This journal is ß The Royal Society of Chemistry 2006