Effects of Al-based additives on the hydrogen storage performance of the Mg(NH2)2-2LiH system

Article abstract of DOI:10.1039/c3dt32165g

The Mg(NH₂)₂-2LiH composite is a promising hydrogen storage material due to its relatively high reversible hydrogen capacity (～5.6 wt%) and suitable thermodynamic properties that allow hydrogen sorption conducting at temperatures below 90 °C. However, the presence of a severe kinetic barrier inhibits its low-temperature operation. In the present work, Li₃AlH₆ was introduced to the Mg(NH ₂)₂-2LiH system. Experimental results show that a 3.2% mol Li₃AlH₆-modified Mg(NH₂)₂-2LiH sample released hydrogen at a rate ca. 4.5 times as fast as that of the Li ₃AlH₆-free sample at 140 °C. The enhancement of desorption kinetics was simultaneously demonstrated by activation energy (E _a) of ca. 96.3 ± 9 kJ mol^-1 which was significantly decreased by 31 kJ mol^-1 from that of the Li₃AlH ₆-free sample. The interaction of Li₃AlH₆ and Mg(NH₂)₂ during ball milling results in the formation of LiAl(NH)₂, LiNH₂ and Mg₃N₂. LiAl(NH)₂ was actually the active species for the enhancement of dehydrogenation/re-hydrogenation kinetics of the system.

Full text of DOI:10.1039/c3dt32165g

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Effects of Al-based additives on the hydrogen storage
performance of the Mg(NH ) –2LiH system†
Cite this: Dalton Trans., 2013, 42, 5524
2 2
a,b,c
a
a
a
a
Hujun Cao, b Yao Zhang,* Jianhui Wang, Zhitao Xiong, Guotao Wu,
a,d
Jieshan Qiu and Ping Chen*
2 2
The Mg(NH ) –2LiH composite is a promising hydrogen storage material due to its relatively high reversi-
ble hydrogen capacity (∼5.6 wt%) and suitable thermodynamic properties that allow hydrogen sorption
conducting at temperatures below 90 °C. However, the presence of a severe kinetic barrier inhibits its
low-temperature operation. In the present work, Li
Experimental results show that a 3.2% mol Li AlH -modified Mg(NH
at a rate ca. 4.5 times as fast as that of the Li AlH
3
AlH
6
was introduced to the Mg(NH
2 2
) –2LiH system.
3
6
2 2
)
–2LiH sample released hydrogen
3
6
-free sample at 140 °C. The enhancement of desorption
) of ca. 96.3 ± 9 kJ mol− which was
1
kinetics was simultaneously demonstrated by activation energy (E
significantly decreased by 31 kJ mol−1 from that of the Li
AlH
and Mg(NH during ball milling results in the formation of LiAl(NH)
a
Received 18th September 2012,
Accepted 24th January 2013
3
6
-free sample. The interaction of Li
3
AlH
was
6
2
)
2
2
, LiNH and Mg . LiAl(NH)
2
3
N
2
2
DOI: 10.1039/c3dt32165g
actually the active species for the enhancement of dehydrogenation/re-hydrogenation kinetics of the
system.
www.rsc.org/dalton
1
H2 of the LiNH –LiH system. Correspondingly, the hydro-
2
1. Introduction
gen desorption plateau pressure increases from 1 bar at ca.
Various amide–hydride composites have been developed for 280 °C for LiNH –LiH to above 30 bar at 200 °C for Mg-
2
1
hydrogen storage in recent years. Systems such as LiNH –LiH,
(NH ) –2LiH composite. Moreover, the dehydrogenation temp-
2 2
2
2
–4
5
6–8
LiNH
CaH2,
2
–LiBH
4
,
LiNH
2
–LiAlH
4
,
LiNH
2
–Li
3
AlH
6
,
LiNH
2
–
2 2
erature for the Mg(NH ) –2LiH composite over 1 bar equili-
9
,10
11
12
LiAl(NH ) –LiH,
LiAlH4, Mg(NH ) –Li AlH , and Mg(NH ) –LiH
extensively studied. Among these systems, the Mg(NH
Ca(NH ) –2LiH,
Mg(NH ) – brium H pressure is below 90 °C derived from the van’t Hoff
2 4
2 2
2 2
2
1
3
14
15–19
20
were plot. However, the majority of hydrogen desorption can only
2 2
–LiH be achieved at temperatures above 180 °C even if Mg(NH ) –
2 2
3
6
2 2
2
)
2
1
5,21
(
1 : 2 in molar ratio) composite has been regarded as a promi- 2LiH composite has been milled intensively,
showing the
sing candidate for onboard application due to its favorable presence of a severe kinetic barrier in hydrogen release. Inves-
reaction thermodynamics and relatively high capacity (5.6 wt% tigations indicate that the kinetic barrier was probably resulted
1
5,20
H2).
The hydrogen desorption from the composite pro- from the relocation of atoms across amide–imide and imide–
ceeds according to the reaction (1).
hydride phase boundaries, and also from the mass transport
2
2
along the imide layer. Numerous attempts have been made
to overcome the barrier in recent years. Doping additives
appears to effectively tune the hydrogenation/dehydrogenation
kinetics, and therefore, lowers the operating temperature.
Partial substitution of Mg or Li by Na in the Mg(NH ) –2LiH
MgðNH2Þ þ 2LiH $ Li2MgðNHÞ þ 2H2 5:6 wt% H2 ð1Þ
2
2
The dehydrogenation enthalpy of the reaction (1) is about
9 kJ mol− -H
1
20
−1
3
2
,
which is significantly lower than 66 kJ mol -
2
2
composite leads to ca. 10 °C decrease in desorption tempera-
a
2
3
24
Dalian National Laboratory for Clean Energy, Dalian Institute of Chemical Physics,
ture. Wang et al., found that introducing 20 wt% of tri-
phenyl phosphate can effectively inhibit the aggregation of
dehydrogenation products and depress the crystallization of
Chinese Academy of Sciences, Dalian 116023, P. R. China. E-mail: pchen@dicp.ac.cn,
zhangyao@dicp.ac.cn; Fax: +86-411-84685940; Tel: +86-411-84379905
b
Carbon Research Laboratory, Liaoning Key Lab for Energy Materials and Chemical
Engineering, State Key Lab of Fine Chemicals, Dalian University of Technology,
Dalian 116024, P. R. China
2 2
Mg(NH ) in the hydrogenation/dehydrogenation cycles, which
further lead to a kinetic improvement. Recently, potassium
compounds were found by the same authors as effective
additives which can remarkably reduce the desorption peak
temperature to ca. 132 °C which is nearly 50 °C lower
c
University of Chinese Academy of Sciences, Beijing 100049, P. R. China
d
International Institute for Carbon-Neutral Energy Research (I2CNER),
7
†
44 Moto-oka, Nishi-ku, Fukuoka, 819-0395, Japan
Electronic supplementary information (ESI)
0.1039/c3dt32165g
available. See DOI:
2
5
1
than the pristine one. Other materials such as LiBH , TiN,
4
5
524 | Dalton Trans., 2013, 42, 5524–5531
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
TiF , TaN, Li N, VCl
Paper
3
3
3
, graphite-supported Ru nanoparticles, release test was carried out on a custom built Sievert-type
single-walled carbon nanotubes, NaBH , NaOH etc. have apparatus.
4
also been tested as candidates with observable kinetic
improvement.
The hydrogen desorption and absorption capacities were
measured by means of volumetric method on an automatic
2
6–34
In this work, Li AlH6 was doped into Mg(NH ) –2LiH Sievert-type apparatus (Hy-Energy scientific instruments PCT
3
2 2
system to form a Mg(NH
2
)
2
–2LiH–0.1Li
3
AlH
6
composite and pro-2000). Initial pressures in the sample chamber for hydro-
effectively enhanced the kinetics of hydrogen desorption and gen desorption and absorption were 0.01 and 60 bar, respect-
absorption. Our investigations show that 3.2 mol% Li AlH - ively. The ramping rate was fixed at 2 °C min−
1
.
3
6
doped Mg(NH
.5 times as fast as that of the Li
re-hydrogenated at 140 °C also faster than that of the Li AlH - was used to protect the samples from air contamination. FTIR
2
)
2
–2LiH sample released hydrogen at a rate ca.
XRD measurements were conducted on a PANalytical X’pert
4
3
AlH -free sample at 140 °C, diffractometer (Cu Kα, 40 kV, 40 mA). A self-made sample cell
6
3
6
free sample, and had an dehydrogenation activation energy measurements were conducted on a Varian 3100 FTIR spectro-
−
1
a
(E ) of ca. 96.3 ± 9 kJ mol which is significantly decreased by photometer (Excalibar Series) in DRIFT mode (diffuse reflec-
1 kJ mol− from that of the Li AlH -free sample. The reaction tance infrared Fourier transform). Magic-angle spinning (MAS)
1
3
3 6
2
7
7
mechanisms were subsequently investigated and discussed in solid-state Al and Li NMR experiments were performed at
this paper.
room temperature on a Bruker AVANCE 500 MHz NMR spec-
trometer (11.7 T). 1 M aqueous solution of aluminum nitrate
and equivalent lithium chloride at 0 ppm were chosen as refer-
2
7
7
ences for Al and Li NMR, respectively.
2. Experimental
3 6 4
Li AlH was prepared by ball milling ca. 1 g mixture of LiAlH
(
97%, Alfa-Aesar) and LiH (98%, Alfa-Aesar) (molar ratio 1 : 2)
3
. Results and discussions
in a SPEX 8000 high-energy mill, of which the XRD pattern
was shown in Fig. S1.† To alleviate temperature ramping
during ball milling, every 15 minutes running is followed by
3
0
.1 Dehydrogenation properties of Mg(NH ) –2LiH–
2 2
.1Li AlH composite
3
6
an interval of 5 minutes pause. Mg(NH ) was synthesized by The TPD measurements revealed that the dehydrogenation
2
2
reacting metallic Mg power (99%, Sigma-Aldrich) with purified peak temperature of the Li
NH at 300 °C. 3.3 mol% Li AlH
was added to the mixture of 15 °C from that of Li AlH -doped sample (186 °C, shown in
Mg(NH ) –2LiH (equivalent to molar ratio of Li AlH : Mg- Fig. 1). The peak temperatures of hydrogen release from the
3 6
AlH -doped sample lowered by
3
3
6
3
6
2
2
3
6
(
NH
on a Retsch planetary ball-mill (PM400) under Ar atmosphere 4 °C after 4 cycles, respectively. Moreover, ammonia releasing
for 36 h. Li AlH -free Mg(NH ) –2LiH sample was also pre- from the Li AlH -doped sample was below the detection limit
pared under the same conditions. ca. 0.17 MPa and 0.05 MPa of the instrument (100 ppm), but it can be obviously detected
equivalent to 1.25 and 0.34 H atoms per Mg(NH
–2LiH) of from the Li AlH -free sample by MS.
gas were evolved from the Li AlH -doped sample and the
The XRD patterns and IR spectra of Li
-free sample, respectively, after ball milling. Analyses collected at different desorption temperatures are given in
by cobalt nitrate solution, which changes from transparent to Fig. 2. After ball milling for 36 h, the sample exhibits poor
2 2
)
3 6
= 1 : 10) following by ball milling at a rate of 200 rpm Li AlH -free and doped samples were increased by ∼24 °C and
3
6
2 2
3
6
(
2
)
2
3
6
AlH -doped samples
6
3
6
3
3 6
Li AlH
red in color only when it meets ammonia, as well as MS crystallinity; only a few weak peaks belonging to Mg
N can be
3 2
revealed that hydrogen was the only gaseous product. LiAl-
2 4
(NH ) with purity of 99.8% (tested by thermogravimetric
analysis, Fig. S3†) was synthesized by ball milling LiAlH in
4
1
1,35
liquid ammonia.
and LiAl(NH as described in the literature. Its XRD pat-
terns as well as FTIR spectra are shown in Fig. S1 and S2.†
Mg and AlN were purchased from Sigma-Aldrich with
2 4
LiAl(NH) was prepared by milling LiAlH
3
6
2 4
)
3 2
N
purity 99.5% and 99.9%, respectively. All handlings were per-
formed in a MBRAUN glove box filled with high-purity argon
(
O
2
< 10 ppm, H
0 mg sample was loaded in a Al
vimetric measurement on Netzsch TG-DTA apparatus
Netzsch, Germany) in a ramping rate of 2 °C min . A temp-
2
O < 0.1 ppm) to avoid air contamination.
1
2 3
O
crucible for thermogra-
a
−1
(
erature-programmed desorption was performed to study the
thermal decomposition properties of samples on a custom
built TPD and mass spectrometer (MS, Hiden) combined
system. About 20 mg sample was tested each time. Tempera-
Fig. 1 TPD curves of Li
3
AlH
6
-doped sample (a), pure Li
3
AlH
6
(b), Li
3 6
AlH -free
sample (c), Li AlH -doped sample after 4 cycles (d) and Li
3
6
3
AlH -free sample
6
−
1
ture was raised at a ramping rate of 2 °C min . The volumetric after 4 cycles (e).
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 5524–5531 | 5525

View Article Online
Paper
Dalton Transactions
3 6
Fig. 2 XRD patterns (A) and FTIR spectra (B) of the Li AlH -doped samples after ball-milling (BM), desorption at 173 °C, 245 °C and subsequent soak.
3 6 3 6
Fig. 3 Isothermal hydrogen desorption (A) and sorption (B) at 140 °C for the Li AlH -doped sample and the Li AlH -free sample.
detected (shown in Fig. 2A). The diffraction peaks of Li
NH) , Mg and AlN can be easily detected after the sample Li
was dehydrogenated at 245 °C. The diffraction peaks of Mg- sample was also faster than that of the Li AlH -free sample. As
2
Mg- the Li
3
AlH
6
-doped sample was ca. 4.5 times more than the
(
2
3
N
2
3
AlH
6
-free one. The sorption kinetics of the Li
3
AlH -doped
6
3
6
2 2
(NH ) cannot be easily found from the fully hydrogenated shown in Fig. 3B, soaked hydrogen capacity can reach ca. 70%
sample. However, FTIR measurements clearly show the appear- in the doped-sample in 240 minutes. However, only 50% could
ance of typical N–H vibrations of Mg(NH ) at 3327 and be absorbed in the Li AlH -free sample even when we pro-
2
2
3
6
1
3
274 cm− in the re-hydrogenated sample (see Fig. 2B). When longed the time to 600 minutes.
the Li
3
AlH
6
-doped sample dehydrogenated from 173 °C to
a
Furthermore, we compared the activation energies (E ) of
−
1
2
45 °C, a newly formed peak centered at 3174 cm which the Li AlH -doped sample with that of the Li AlH -free one.
3
6
3
6
3
7
belongs to the N–H stretch in Li
2
Mg(NH)
2
gradually Kissinger’s method was employed to determine the activation
increased in intensity.
energy (eqn (2)):
Fig. 3 exhibits the isothermal (140 °C) dehydrogenation/
re-hydrogenation performances of the Li AlH -doped sample.
It can be found that the hydrogen release from the Li AlH -
2
3
6
d½lnðβ=Tm Þ=dð1=T Þꢀ ¼ ꢁE =R
m
a
ð2Þ
3
6
doped sample was accelerated remarkably from that of the
Here Tm is the peak temperature at which the maximum
Li AlH -free sample, especially in the beginning stage. After an reaction-rate was attained, β is the heating rate, E is the acti-
3
6
a
initial 100 minutes, 70% of hydrogen was desorbed from the vation energy and R is the gas constant. TPD technique was
Li AlH -doped sample, while less than 15% of the capacity can applied to collect the maximum reaction-rate temperatures at
3
6
-free sample. In the same figure, various heating rates (2, 4, 6, and 8 °C min− , respectively, as
1
be liberated from the Li
3
AlH
6
2
the tangent slopes (1.16% and 0.26% of extent of the Li AlH - shown in Fig. 4B). The dependency of ln(β/Tm ) upon 1/Tm was
3
6
doped sample and Li AlH -free sample, respectively) of the plotted in Fig. 4A. The slope of the fitted line was used to
3
6
initial linear parts indicate that the dehydrogenation rate for determine the value of E /R. Thus, the activation energy for
a
5
526 | Dalton Trans., 2013, 42, 5524–5531
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
3 6 3 6 2
Fig. 4 Kissinger plots (A) and TPD profiles (B) for the Li AlH -free sample (a), Li AlH -doped sample (b), and LiAl(NH) -doped sample (c).
hydrogen desorption from the Li AlH -doped sample is ca. (NH ) . The achieved 0.63 mol H was produced most probably
3
6
2 2
2
1
+
−
9
6.3 ± 9 kJ mol− , which is about a 24% reduction from that of due to the coordination of H from 0.3Mg(NH ) with H from
2 2
−
1
5
the Li
3
AlH
6
-free one (127.0 ± 12 kJ mol ).
3 6 2 4
0.1Li AlH as in some previous work, e.g. LiNH –LiAlH . We
thus suggested a molar ratio of 1 : 3 for the interaction
between Li AlH and Mg(NH ) .
3
6
2 2
3
.2 Behaviors of Li
To understand the behavior of Li
of Mg(NH –2LiH composite, the interaction between Li
and Mg(NH2)2 was investigated. As mentioned above, weak
reflection peaks of Mg appeared in the XRD pattern of the
Li AlH
3
AlH
6
in the composite
AlH in the dehydrogenation
AlH
Li3AlH6 þ 3MgðNH2Þ ! LiAlðNHÞ þ 2LiNH2 þ Mg3N2 þ 6H2
3
6
2
2
2
)
2
3
6
ð3Þ
3
N
2
The Li
3
AlH
6
–3Mg(NH ) composite was thus prepared by
2 2
3
6
-doped sample after milling for 36 h (see in Fig. 2A). ball milling and was investigated by XRD, FTIR and NMR tech-
The Al resonance in Li AlH -doped sample after ball milling niques. For the sample milled for 10 h, the diffraction peaks
27
3
6
for 36 h belongs to LiAl(NH)
between LiH with Li AlH or Mg(NH )
3 6 2 2
2
(see Fig. S4†). Since interaction belonging to the LiNH
can be hardly observed pattern (in Fig. 5A) at the 2θ of 30.7, 41.1 and 51.4°. Minor
during the process of ball milling, the major hydrogen release Mg N peaks appear in the post-26 h sample, and can be
2
phase can be observed in the XRD
3
2
3 6
should come from the reaction between Li AlH and Mg- clearly identified after ball milling for 36 h. The FTIR profiles
3 6 2 2
Fig. 5 XRD patterns (A) and FTIR spectra (B) of the Li AlH –3Mg(NH ) sample at different ball-milling stages (2, 10, 26, and 36 h, respectively).
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 5524–5531 | 5527

View Article Online
Paper
Dalton Transactions
5
AlN (113 ppm, which formed in the Li
3 6
AlH -doped sample
after dehydrogenation to 250 °C). It is likely that LiAl(NH) is
2
the only Al-containing product of the reaction between Li
and 3Mg(NH . Based on those evidences presented by XRD
Fig. 5A), IR (Fig. 5B), NMR (Fig. 6) and the amount of
hydrogen (6.05 mol per Li AlH –3Mg(NH ) released during
the ball milling process, the interaction between Li AlH
and Mg(NH ) during the ball milling can be described by the
3 6
AlH
2 2
)
(
3
6
2 2
)
3
6
2
2
reaction (3).
The volumetric release tested on a Sievert-type apparatus
shows that the Li AlH -doped sample can liberate about
3
6
5
2
.2 wt% H within the temperature range of 110–390 °C at a
−
1
ramping rate of 2 °C min (Fig. 7). The dehydrogenation may
be roughly divided into two steps, i.e., the first step ends up at
ca. 250 °C, which is also the end temperature of hydrogen
2
7
Fig. 6
Al MAS NMR spectra of (a) Li
3
AlH
6
–Mg(NH
2
)
2
after ball milling, (b)
. # denotes the
release from the Li
Table 1 summarizes all the reactions occurring in the dehydro-
genation of the Li AlH -doped sample. About 1.55 wt% H is
released during the ball milling process and 0.3 equiv. Mg-
NH ) is consumed following the reaction (3). The first-step
3 6
AlH -free sample (as shown in Fig. 1).
Li
3
AlH
6
-doped sample after volumetric release, and (c) LiAl(NH)
2
spinning sidebands.
3
6
2
(
2
2
3 6
dehydrogenation of the Li AlH -doped sample taking place
within the temperature range of 120 and 250 °C follows the
eqn (1) in which the residual 0.7 equiv. Mg(NH2)2 further
reacts with 1.4 equiv. LiH to generate 1.4 equiv. H
Moreover, about 0.1 equiv. H can be released from the inter-
action between LiAl(NH) and LiH (eqn (4)), which was verified
2
(3.62 wt%).
2
2
by the XRD pattern and IR spectra in Fig. 8. The Li
AlN phases in the LiAl(NH) –2LiH sample after dehydrogena-
tion at 250 °C can be observed. Therefore, in total 1.5 equiv.
(3.89 wt%) can be achieved from the ternary mixture in
theory, which has been confirmed by the achieved value
.62 wt% hydrogen through volumetric release measurement.
2
NH and
2
H
2
3
3 6
Fig. 7 Volumetric release curve of the Li AlH -doped sample under 0.01 bar.
LiAlðNHÞ þ LiH ! Li2NH þ AlN þ H2
ð4Þ
2
The second-step dehydrogenation might be attributed to
shown in Fig. 5B indicate that the intensities of the N–H
vibration of Mg(NH ) at 3327 and 3274 cm were clearly
2 2
weakened with the increase of the ball milling time. On the
2
the reaction between LiH with LiNH and/or imide. According
−
1
2
2,40
to the coordinated solid-state reaction mechanism,
the
δ−
negatively charged (H ) LiH tended to interact with the posi-
δ+
other hand, the emerging signals at around 3310 and
tively charged H from LiNH
2 2 2
or Li Mg(NH) to form molecu-
1
3
254 cm− agreed well with the N–H vibrations of LiNH
2
.
lar hydrogen.
2
7
Fig. 6 shows the solid-state Al NMR spectra of the
Li AlH –3Mg(NH ) sample after ball milling for 36 h, the
3
6
2 2
3
.3 Roles of LiAl(NH) in the composite
2
Li
3
AlH
6
-doped sample after dehydrogenation and the prepared
LiAl(NH)2 sample. The spectrum of the Li AlH –3Mg(NH )
In order to understand the effects of intermediate LiAl(NH)2
sample exhibits a symmetric peak at around 117 ppm which is on the dehydrogenation process of the Mg(NH ) –2LiH–
3
6
2 2
2
2
3
8
identical to the chemical shift of LiAl(NH)
2
(see Fig. 6). 0.1Li AlH composite, we prepared three composites contain-
3 6
However, it differs from that of metallic Al (1640 ppm) and ing Mg N , AlN or LiAl(NH) (i.e. Mg(NH ) –2LiH–0.1Mg N ,
3
9
3
2
2
2 2
3 2
3 6
Table 1 The hydrogen released and the proposed reaction pathway for the Li AlH -doped sample
Step Product phases Reaction
2
H atoms released wt% of H released
In ball milling LiAl(NH)
2
, 2LiNH
2
2
, Mg N
3 2
Li
Mg(NH
LiAl(NH)
3
AlH
6
+ 3Mg(NH
2
)
2
→ LiAl(NH)
+ 2LiH ↔ Li Mg(NH) + 2H
+ LiH → Li NH + AlN + H
↔ Li NH + H
2
+ 2LiNH
2
3
+ Mg N
2
+ 6H
2
1.2
2.8
0.2
0.4
5.2
1.55
3.62
0.26
0.52
6.72
T < 250 °C
Li
Li
Li
2
2
2
Mg(NH)
NH, AlN
NH
)
2 2
2
2
2
2
2
2
T > 250 °C
LiH + LiNH
2
2
2
In total
5
528 | Dalton Trans., 2013, 42, 5524–5531
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Fig. 10 Temperature-programmed volumetric desorption/sorption curves of
3 6
the Li AlH -doped sample in the first four cycles.
2
Fig. 8 XRD pattern (A) and FTIR (B) spectra of LiAl(NH) –LiH sample after de-
hydrogenation at 250 °C.
2
7
7
Fig. 9 (A) TPD curves of (a) Mg
c) Li AlH -doped sample, (d) 7Mg(NH
sample, (B) Mg(NH –2LiH–xLiAl(NH) (x = 0, 0.1, 0.2, 0.3).
3
N
2
-doped sample, (b) LiAl(NH)
2
-doped sample,
3 6
Fig. 11 The Al NMR spectra (A) and Li NMR spectra (B) of the Li AlH -
(
3
6
2
)
2
–2LiNH –20LiH, and (e) AlN-doped
2
doped sample after ball milling (BM), after volumetric release and subsequent
sorption below 170 °C.
2
)
2
2
Mg(NH
2
)
2
–2LiH–0.1AlN and Mg(NH
2
)
2
–2LiH–0.1LiAl(NH)
2
)
the onset and peak temperatures of the Mg(NH
2
)
2
–2LiH–
under the same conditions as that of Mg(NH ) –2LiH– 0.3LiAl(NH)
sample were below 100 and 153 °C, respectively.
2
2
2
0
.1Li
3
AlH
6
composite. Another composite of 7Mg(NH
2
)
2
–
2
These evidences indicate that LiAl(NH) may be the active
2
LiNH
2
2 2
–20LiH was also made to mimic the composition of the species in improving the desorption kinetics of Mg(NH ) –
post-milled Mg(NH ) –2LiH–0.1Li AlH sample. TPD curves of 2LiH composite.
2
2
3
6
those composites were shown in Fig. 9A, in which the LiAl-
NH)
-doped sample exhibits the same kinetic behaviors as volumetric release and subsequent soak measurements on the
that of the Li AlH -doped sample. However, a shoulder peak Li AlH -doped samples for four cycles to investigate its reversi-
was observed bility. Fig. 10 indicates that about 2.7 equiv. H per Mg(NH
-doped sample, which was formed in the ball 2LiH can be reversibly desorbed/absorbed at temperatures
Furthermore, we performed the temperature-programmed
(
2
3
6
3
6
arising from the reaction between LiH and LiNH
in the Li AlH
2
2 2
) –
3
6
milling process of Li AlH and Mg(NH2)2 as reaction (3). It below 170 °C. If counting the 1.25 equiv. H detached during
3
6
seems that the LiAl(NH)
2
species improves the kinetics of the the ball-milling process, in total 3.95 equiv. H atoms per Mg-
Mg(NH
2 2
)
2 2
–2LiH composite. Therefore, a few samples of Mg- (NH ) –2LiH were released during the process of the first de-
(
NH ) –2LiH–xLiAl(NH) , where x = 0, 0.1, 0.2, 0.3, were pre- sorption. The hydrogen sorption/desorption capacities exhibit
2
2
2
2
7
pared to verify the effects of LiAl(NH)
of Mg(NH
–2LiH system. Fig. 9B shows that the hydrogen and Li NMR spectra of post-milled, dehydrogenated and re-
desorption temperature was lowered with the amount of hydrogenated Li AlH -doped samples are shown in Fig. 11.
increasing. Notably, when x was increased to 0.3, Chemical shift of Al was observed at 117 ppm, which is close
2
on the dehydrogenation little degradation at least within the initial four cycles. Al
7
2 2
)
3
6
2
7
LiAl(NH)
2
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 5524–5531 | 5529

View Article Online
Paper
Dalton Transactions
2
7
to the Al resonance of LiAl(NH)
2
(Fig. 11A), verifying the
7 O. Dolotko, H. Zhang, O. Ugurlu, J. W. Wiench, M. Pruski,
L. S. Chumbley and V. Pecharsky, Acta Mater., 2007, 55,
3121–3130.
8 J. Lu, Z. Z. Fang, Y. J. Choi, H. Y. Sohn, C. Kim, R. C. Bowman
and S. J. Hwang, J. Power Sources, 2008, 185, 1354–1358.
9 G. Wu, Z. Xiong, T. Liu, Y. Liu, J. Hu and P. Chen, Inorg.
Chem., 2007, 46, 517–521.
existence of LiAl(NH)2 in the cycles. Based on the control
experiments (Fig. 9), the activation energy (Fig. 4) and the
NMR spectra (Fig. 11) information, we can inferred that LiAl-
NH) is the active species in improving kinetics performances
of desorption/sorption of Li
(
2
3 6
AlH -doped sample. The possible
reason for improvement of hydrogen storage properties may be
the weakening of N–H bond in Mg(NH ) by LiAl(NH) . To the 10 H. Wu, J. Am. Chem. Soc., 2008, 130, 6515–6522.
2
2
2
best of our knowledge, the reaction firstly takes place between 11 R. Janot, J. B. Eymery and J. M. Tarascon, J. Phys. Chem. C,
LiAl(NH) and LiH. Hydrogen was released due to more 2007, 111, 2335–2340.
extended N–H bond in LiAl(NH) than in Mg(NH ) . The N–H 12 K. Tokoyoda, S. Hino, T. Ichikawa, K. Okamoto and
2
2
2 2
bond in Mg(NH
the broken N–H bond in LiAl(NH)
2
)
2
will be weakened by the interaction with
H. Fujii, J. Alloys Compd., 2007, 439, 337–341.
13 J. Lu, Z. Z. Fang, H. Y. Sohn and R. C. Bowman, J. Phys.
Chem. C, 2007, 111, 16686–16692.
2
.
14 D. Liu, A. Sudik, J. Yang, P. Ferro and C. Wolverton, J. Phys.
Chem. C, 2012, 116, 1485–1492.
4. Conclusions
15 Z. Xiong, G. Wu, J. Hu and P. Chen, Adv. Mater., 2004, 16,
1
522–1525.
3 6 2 2
Li AlH -doped Mg(NH ) –2LiH composite was investigated sys-
1
1
6 W. Luo, J. Alloys Compd., 2004, 381, 284–287.
7 H. Y. Leng, T. Ichikawa, S. Hino, N. Hanada, S. Isobe and
H. Fujii, J. Phys. Chem. B, 2004, 108, 8763–8765.
8 T. Ichikawa, K. Tokoyoda, H. Y. Leng and H. Fujii, J. Alloys
Compd., 2005, 400, 245–248.
9 Y. Nakamori, G. Kitahara, K. Miwa, N. Ohba, T. Noritake,
S. Towata and S. Orimo, J. Alloys Compd., 2005, 404–406,
tematically in the present work. Several improvements and
phenomena can be observed: (1) the rate of dehydrogenation
from the doped sample is 4.5 times as much as the pristine
one at 140 °C and the sorption kinetics were also notably
1
1
improved; (2) Li
milling and yields an active species LiAl(NH)
.4 wt% hydrogen can be reversibly absorbed and desorbed in
the Mg(NH –2LiH–0.1Li AlH sample at temperatures below
70 °C; (4) LiAl(NH) is the active species actually improving
the kinetics of Mg(NH ) –2LiH–0.1Li AlH .
3
AlH
6
prefers to react with Mg(NH
2 2
) in ball
2
; (3) around
3
396–398.
2
)
2
3
6
2
2
2
2
2
2
0 Z. Xiong, J. Hu, G. Wu, P. Chen, W. Luo, K. Gross and
J. Wang, J. Alloys Compd., 2005, 398, 235–239.
1 Y. Liu, K. Zhong, K. Luo, M. Gao, H. Pan and Q. Wang,
J. Am. Chem. Soc., 2009, 131, 1862–1870.
2 P. Chen, Z. Xiong, L. Yang, G. Wu and W. Luo, J. Phys.
Chem. B, 2006, 110, 14221–14225.
3 Y. Liu, J. Hu, Z. Xiong, G. Wu and P. Chen, J. Mater. Res.,
1
2
2
2
3
6
Acknowledgements
We acknowledge the financial support from National Natural
Science foundation of China (Grant Nos. 50901070, 20971120
and 21273229), the National Basic Research Program of China
2007, 22, 1339–1345.
4 J. Wang, J. Hu, Y. Liu, Z. Xiong, G. Wu, H. Pan and
P. Chen, J. Mater. Chem., 2009, 19, 2141–2146.
5 J. Wang, T. Liu, G. Wu, W. Li, Y. Liu, C. M. Araújo,
R. H. Scheicher, A. Blomqvist, R. Ahuja, Z. Xiong, P. Yang,
M. Gao, H. Pan and P. Chen, Angew. Chem., Int. Ed., 2009,
(Grant No. 2010CB631304) and CAS-JSPS collaborative
funding.
48, 5828–5832.
2
6 J. Hu, Y. Liu, G. Wu, Z. Xiong, Y. S. Chua and P. Chen,
Chem. Mater., 2008, 20, 4398–4402.
References
1
2
3
4
P. Chen, Z. Xiong, J. Luo, J. Lin and K. L. Tan, Nature, 2002, 27 J. Hu, M. Fichtner and P. Chen, Chem. Mater., 2008, 20,
20, 302–304. 7089–7094.
F. E. Pinkerton, G. P. Meisner, M. S. Meyer, M. P. Balogh 28 Q. Wang, Y. Chen, G. Niu, C. Wu and M. Tao, Ind. Eng.
and M. D. Kundrat, J. Phys. Chem. B, 2005, 109, 6–8. Chem. Res., 2009, 48, 5250–5254.
P. A. Chater, W. I. F. David, S. R. Johnson, P. P. Edwards 29 L. P. Ma, H. B. Dai, Z. Z. Fang, X. D. Kang, Y. Liang,
4
and P. A. Anderson, Chem. Commun., 2006, 2439–2441.
P. J. Wang, P. Wang and H. M. Cheng, J. Phys. Chem. C,
2009, 113, 9944–9949.
T. Noritake, M. Aoki, S. Towata, A. Ninomiya, Y. Nakamori
and S. Orimo, Appl. Phys. A: Mater. Sci. Process., 2006, 83, 30 R. R. Shahi, T. P. Yadav, M. A. Shaz and O. N. Srivastva,
77–279. Int. J. Hydrogen Energy, 2010, 35, 238–246.
Z. Xiong, G. Wu, J. Hu, Y. Liu, P. Chen, W. Luo and 31 L. P. Ma, H. B. Dai, Y. Liang, X. D. Kang, Z. Z. Fang,
2
5
6
J. Wang, Adv. Funct. Mater., 2007, 17, 1137–1142.
P. J. Wang, P. Wang and H. M. Cheng, J. Phys. Chem. C,
2008, 112, 18280–18285.
Y. Kojima, M. Matsumoto, Y. Kawai, T. Haga, N. Ohba,
K. Miwa, S. I. Towata, Y. Nakamori and S. Orimo, J. Phys. 32 Y. Chen, P. Wang, C. Liu and H. M. Cheng, Int. J. Hydrogen
Chem. B, 2006, 110, 9632–9636. Energy, 2007, 32, 1262–1268.
5
530 | Dalton Trans., 2013, 42, 5524–5531
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
3
3
3
3 C. Liang, Y. Liu, Y. Jiang, Z. Wei, M. Gao, H. Pan and 37 J. Hu, Y. Liu, G. Wu, Z. Xiong and P. Chen, J. Phys. Chem.
Q. Wang, Phys. Chem. Chem. Phys., 2011, 13, 314–321. C, 2007, 111, 18439–18443.
4 C. Liang, Y. Liu, Z. Wei, Y. Jiang, F. Wu, M. Gao and 38 T. Ono, K. Shimoda, M. Tsubota, S. Kohara, T. Ichikawa,
H. Pan, Int. J. Hydrogen Energy, 2011, 36, 2137–2144.
K. Kojima, M. Tansho, T. Shimizu and Y. Kojima, J. Phys.
Chem. C, 2011, 115, 10284–10291.
5 J. B. Eymery, L. Truflandier, T. Charpentier, J. N. Chotard,
J. M. Tarascon and R. Janot, J. Alloys Compd., 2010, 503, 39 J. W. Wiench, V. P. Balema, V. K. Pecharsky and M. Pruski,
94–203. J. Solid State Chem., 2004, 177, 648–653.
6 G. Karl and C. A. Fremont, U. S. Patent, 20070231254A1, 40 P. Chen, Z. Xiong, J. Luo, J. Lin and K. L. Tan, J. Phys.
007. Chem. B, 2003, 107, 10967–10970.
1
3
2
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 5524–5531 | 5531