Materials Transactions, Vol. 52, No. 4 (2011) pp. 627 to 634
Special Issue on Advanced Materials for Hydrogen Energy Applications
#2011 The Japan Institute of Metals
Hydrogen Storage Properties of the Mg(NH3 )6 Cl2 -LiH Combined System
Yongfeng Liu, Ruijun Ma, Runlai Luo, Kun Luo, Mingxia Gao and Hongge Pan*
State Key Laboratory of Silicon Materials & Department of Materials Science and Engineering, Zhejiang University,
Hangzhou 310027, P. R. China
Metal ammine complexes (MACs) were recently regarded as one of promising materials for reversible hydrogen storage due to their high
hydrogen content. In this work, a first attempt is conducted to elucidate the hydrogen storage reversibility by combining Mg(NH3 )6 Cl2 with LiH.
It is found that hydrogen is gradually evolved from the combined system during ball milling. After 24 h of milling, approximate 3.5 mass% of
hydrogen, equivalent to 6 mol of H2 molecules, is released from the Mg(NH3 )6 Cl2 -18LiH mixture. Additional 3.4 mass% of hydrogen is further
desorbed from the combined system milled for 24 h with a two-step reaction in heating process. Totally 12 mol of H2 molecules are librated
from the Mg(NH3 )6 Cl2 -18LiH mixture along with the formation and consumption of Mg(NH2 )2 and LiNH2 in the ball milling and subsequent
heating process. The resultant products consist of Li2 Mg(NH)2 , Li2 NH, LiH and LiCl after dehydrogenation at 310 C. Further hydrogenation
experiment indicates that 6 mol of H2 molecules are reversibly stored in the Mg(NH3 )6 Cl2 -12LiH mixture.
[doi:10.2320/matertrans.MA201013]
(Received October 1, 2010; Accepted November 16, 2010; Published December 29, 2010)
Keywords: solid-state reaction, hydrogen storage, metal ammine complexes, dehydrogenation
1.
Introduction
To establish safe and efficient hydrogen storage technologies is necessary for utilizing hydrogen as a fuel of the
future.1) In contrast to the compressed hydrogen and the
liquefaction hydrogen, hydrogen storage in the solid state
is the most promising alternative.2–5) In the past decades,
the complex hydrides consisting of light elements, e.g.,
alanates,6–10) amides,11–15) borohydrides16–20) and ammonia
borane (AB),21–23) have been attracting extensive attention as
the potential hydrogen storage materials due to their high
gravimetric and volumetric hydrogen storage densities. In
particular, significant efforts have been made in recent years
with metal amide-hydride combined systems since Chen
et al. reported that lithium nitride, Li3 N could absorb/desorb
reversibly 11.4 mass% of hydrogen in 2002.11) A variety of
metal amide-hydride combinations have been designed and
developed for their hydrogen storage performances.11–15)
However, most of them still suffer from the high operating
temperature for practical applications.
More recently, a novel way of storing hydrogen was
proposed in the form of metal ammine complexes (MACs)
formulated as M(NH3 )n Xm (M = metal cations like Mg, Ca,
Cr, Ni, and Zn; X = anions like Cl or SO4 ) by ammoniaintermediate. Christensen et al. revealed that Mg(NH3 )6 Cl2
could store 9.1 mass% of hydrogen in the form of ammonia
with a three-step decomposition reaction as:24)
Mg(NH3 )6 Cl2 ! Mg(NH3 )2 Cl2 þ 4NH3
! MgNH3 Cl2 þ 5NH3
! MgCl2 þ 6NH3
ð1Þ
Interestingly, Mg(NH3 )6 Cl2 can be compacted into shaped
objects essentially without any void space, which offers a
very high volumetric storage density similar to that of liquid
ammonia ( 109 kgH2 /m3 ).24) For use as a potential hydrogen carrier, ammonia must first be decomposed into nitrogen
*Corresponding
author, E-mail: [email protected]
and hydrogen. However, hydrogen is delivered only at
temperature above 300 C by combining with a present-day
ammonia decomposition catalyst.25) Thus, the challenge
remains in the rapid and efficient conversion of ammonia
into hydrogen.
It is well known that hydrogen can be easily obtained
by reacting metal amide or ammonia with metal hydrides.11–15,26–28) Yamamoto et al. found that NH3 could
react with alkali metal hydrides MH (M = Li, Na, and K) in
an exothermic nature to release hydrogen (H2 ) at room
temperature.26) Moreover, an ultrafast reaction was reported
for NH3 and LiH at above 300 C, which is in the order of
microsecond.27) However, the reaction between MgH2 and
NH3 occurred in heating process with a much slower rate
than that between LiH and NH3 .28) Mechanistic investigations suggested that the chemical driving force between
amides and hydrides originated possibly from the strong
affinity between Hþ in amides and H in hydrides, which
results in lower dehydrogenation temperatures owing to the
favorable energy relationship.12) Recently, a dramatically
improved dehydrogenation performance was obtained
through substituting one or two H atoms in the NH3 group
of ammonia borane (AB) by alkali or alkaline-earth metals.22,29) Xiong et al. reported that 10:9 mass% hydrogen
was librated from LiNH2 BH3 obtained by ball milling the
mixture of NH3 BH3 and LiH at significantly lower temperatures with respect to NH3 BH3 itself.22) More interestingly,
the reaction between NH3 BH3 and LiH in THF delivers more
than 14 mass% of hydrogen at a temperature as low as
40 C.30) Significantly, combining compounds containing
NHx group with metal hydrides results in interesting hydrogen storage systems with improved thermodynamics and
kinetics.
The motivation of this work is to obtain the reversible
hydrogen storage in the combinations of metal ammine
complex and metal hydride. Since H ions in the NH3 group of
Mg(NH3 )6 Cl2 are positive charge (Hþ , electron acceptor)
and those in LiH possess negative charge (H , electron
628
Y. Liu et al.
donor), the direct interaction between Mg(NH3 )6 Cl2 and LiH
is expected to occur to generate hydrogen. Tsubota et al.
reported recently a Mg(NH3 )6 Cl2 -LiH mixture with a molar
ratio of 1 : 6 which liberates 4:9 mass% of hydrogen at far
lower temperatures than the decomposition temperatures
of ammonia.31) Unfortunately, no reversible reaction was
observed under up to 5 MPa H2 at their experiment. Moreover, the underlying mechanism for hydrogen desorption
from the combination of Mg(NH3 )6 Cl2 and LiH are not
fully understood. In this study, we mix equimolar Hþ in
Mg(NH3 )6 Cl2 and H in LiH by means of ball milling
to investigate in detail the dehydrogenation process in the
mechanochemical reaction and thermal decomposition.
The Mg(NH3 )6 Cl2 -18LiH combination is developed by ball
milling the corresponding chemicals. A stiochiometric
amount of hydrogen is released after ball milling and thermal
decomposition. Structural characteristics of the combined
system are identified at different dehydrogenation stages to
elaborate reaction details. The clarification of the mechanisms of hydrogen release and uptake in the Mg(NH3 )6 Cl2 LiH combined system is very helpful for designing the novel
hydrogen storage system with improved thermodynamics and
kinetics.
2.
apparatus, and gradually heated from 25 to 350 C at 2 C/min
for hydrogen desorption (initially in vacuum) and to 300 C
at 1 C/min for hydrogen absorption (initial H2 pressure
being 10 MPa). Differential Scanning Calorimetry (DSC)
measurements (Netzsch DSC 200 3F) were performed at a
heating rate of 10 C/min under a high pure argon (purity
>99:9999%) flow.
2.3 Structure characterization
The phase structures of the samples were characterized
by using X-ray powder diffraction (XRD). The data were
collected on a X’Pert PRO (PANalytical, The Netherlands)
X-ray diffractometer with Cu K radiation ( ¼ 0:15418 nm)
at 40 kV and 30 mA from 10 to 70 (2) with step increments
of 0.02 , and analyzed by referring to PDF-2 database of the
JCPDS-ICDD and published literatures.32,33) All the samples
were protected in a homemade container under Ar atmosphere during data collection. N-H vibrations in all samples
were identified by a Bruker Tensor 27 Fourier infrared
spectrometer (FTIR, Germany). The powdery sample was
first mixed with KBr at a weight ratio of 1 : 30 and then was
cold-pressed into a pellet with a diameter of 13 mm. FTIR
spectra were acquired in transmission mode at a resolution
of 4 cm1 .
Experimental Section
3.
2.1 Sample preparation
The starting chemicals, anhydrous MgCl2 (purity 99%)
and LiH (purity 95%) were purchased from Alfar Aesar and
Fluka, respectively, and used as-received. Mg(NH3 )6 Cl2
was synthesized by reacting anhydrous MgCl2 under 7 bar
ammonia at ambient temperature. The X-ray powder diffraction (XRD) examination confirmed the formation of
Mg(NH3 )6 Cl2 . The Mg(NH3 )6 Cl2 -18LiH mixture was prepared by ball milling the corresponding chemicals on a
planetary ball mill (QM-1SP2) at 550 rpm. The milling jar
with a volume of around 180 ml was equipped with a gas
valve to monitor the inner pressure change during ball
milling. The ball-to-sample weight ratio was about 60 : 1. All
sample handling was carried out in a MBRAUN glove box
filled with pure argon to prevent chemicals from contacting
air and moisture (H2 O: <1 ppm, O2 : <1 ppm). The ammonia
concentration was determined by using a Thermo Orion
410P-19 pH/ion meter (USA) equipped with an NH3 selective electrode. After ball milling, the gaseous products
were slowly introduced from the milling jar to 100 mL of
distilled water. NH3 gas will be absorbed in water upon
contacting with water and its concentration in water was
measured by the NH3 -selective electrode.
2.2 Property evaluation
The temperature-dependence of hydrogen desorption was
carried out on a homemade temperature-programmed desorption (TPD, with pure Ar as carrier gas) system combined
with a gas chromatograph (GC). About 40 mg post-milled
samples were loaded and tested. The sample was heated at a
ramp of 2 C/min. Hydrogen absorption/desorption curves
were measured by the volumetric method with a homemade
Sieverts-type apparatus. The sample container with about
200 mg powder samples was connected to the Sieverts-type
Results and Discussion
3.1 Preparation and characterization of Mg(NH3 )6 Cl2
Mg(NH3 )6 Cl2 was synthesized by reacting anhydrous
MgCl2 with 7 bar ammonia for 24 h at ambient temperature.
Figure 1 shows XRD patterns and FTIR spectra of MgCl2
before and after amination. Apparently, a new set of
diffraction peaks is developed with the disappearance of
the characteristic diffraction peaks of the anhydrous MgCl2
after complete amination. The XRD pattern of the aminated
product exhibits a single-phase structure having the same
peak positions as reported by Long et al.,32) implying the
formation of Mg(NH3 )6 Cl2 . Calculations based on XRD
results led to a cubic crystal structure (Space No.: Fm3m)
with a ¼ 1:021 nm, very close to the results reported by
Hwang et al. (a ¼ 1:0189 nm).33) Moreover, FTIR examination demonstrates that four absorbances are clearly observed
at 3361, 3156, 3053 and 2817 cm1 for the aminated sample,
which is in the typical N-H vibration range of the ammine
ions.34) Combined with XRD analyses, these absorbances
should originate from the NH3 group in Mg(NH3 )6 Cl2 .
The phase purity of the as-prepared Mg(NH3 )6 Cl2 sample
was further determined by measuring its weight loss in the
thermal decomposition process. Figure 2(a) shows the TPD
curve of the as-prepared sample. With increasing temperature, ammonia is gradually evolved from Mg(NH3 )6 Cl2 , and
three desorption peaks are observed over the temperature
range 80–310 C with peak temperatures at 113 C, 185 C,
and 300 C, respectively, indicating a three-step decomposition process, which is in good agreement with the previous
report.24) The quantity of ammonia released is further
measured as shown in Fig. 2(b). Clearly, ammonia release
starts sluggishly at around 80 C followed by three decomposition steps. The first-step ammonia release occurs in the
temperature range 80–130 C to liberate about 32.5 mass%
Hydrogen Storage Properties of the Mg(NH3 )6 Cl2 -LiH Combined System
(a)
∆
∆
∆
Intensity (a.u.)
∆ MgCl2(NH3)6
MgCl2(NH3)6
∆ ∆
∆
∆∆∆ ∆ ∆
∆
(b)
∆
*
* MgCl2
*
Transmittance (a.u.)
∆
2817
3053
MgCl2
*
*
*
*
3361 3165
**
10 20 30 40 50 60 70 80 90 4000
2θ (°)
Fig. 1
629
3600
3200
2800
2400
-1
Wavenumber (cm )
XRD patterns (a) and FTIR spectra (b) of MgCl2 before and after amination.
0
Volumetric release
1
(b)
2
3
4
5
6
0.4
0.3
Mg(NH3)6Cl2-18LiH
Mg(NH3)6Cl2
0.2
0.1
0.0
0
50
100
150
200
250
300
350
Temperature, T /°C
Fig. 2 TPD (a) and NH3 desorption (b) curves of the as-prepared
Mg(NH3 )6 Cl2 sample.
of ammonia, which is equivalent to 4 mol of NH3 molecules.
As the temperature was further elevated to 140–240 C and
240–320 C, another two decomposition steps proceed one
after the other to release 9.2 mass% of ammonia or 1 equiv.
NH3 molecule for each step. We believe therefore that the
thermal decomposition process can be still described well by
reaction 1. On the other hand, totally 50.9 mass% of ammonia
is detached from the as-prepared Mg(NH3 )6 Cl2 sample in the
temperature range 80–310 C, which corresponds to 98.5%
of the theoretical weight loss (51.8 mass%). Thus, the purity
of the as-prepared Mg(NH3 )6 Cl2 sample is about 98.5%.
3.2
Pressure increase, P/MPa
No. of NH3 released Intensity (a.u.)
0.5
(a)
TPD
Milling investigation on the Mg(NH3 )6 Cl2 -LiH
(1 : 18) mixture
For obtaining the reversible hydrogen desorption from the
metal ammine complexes, a Mg(NH3 )6 Cl2 -18LiH combined
system was designed based on the interaction between Hþ
and H .12) Typically, 0.985 g of Mg(NH3 )6 Cl2 ( 0:005
mol) and 0.72 g of LiH ( 0:09 mol) were mixed and then
0
2
4
6
8
10
12
14
16
18
20
22
24
26
Milling time, t /h
Fig. 3 Dependence of pressure increase in the milling jar on the ball
milling time for Mg(NH3 )6 Cl2 -18LiH and Mg(NH3 )6 Cl2 .
ball milled at 550 rpm. For comparison, 1:7 g of
Mg(NH3 )6 Cl2 was alone loaded into a milling jar for ball
milling under same conditions. The pressure within milling
jar was examined at room temperature after ball milling.
Figure 3 presents the time dependence of the pressure
increase in the milling jar for Mg(NH3 )6 Cl2 -18LiH and
Mg(NH3 )6 Cl2 . The pressure in the milling jar with the
Mg(NH3 )6 Cl2 -18LiH mixture increases gradually with milling time while it remains unchanged in that only with
Mg(NH3 )6 Cl2 even though after 24 h of ball milling. This
phenomenon indicates that the chemical reaction occurs
between Mg(NH3 )6 Cl2 and LiH to generate H2 . By applying
the equation of state, the pressure was converted to the
corresponding H2 molecules per Mg(NH3 )6 Cl2 -18LiH. As
shown in Table 1, it can be seen that hydrogen was quickly
released in the initial 12 h of milling. When the milling time
was extended from 12 to 24 h, the dehydrogenation reaction
was almost stopped as no obvious pressure increase was
detected. The sample milled for 24 h released totally
630
Y. Liu et al.
∆ Mg(NH3)6Cl2
#
(b)
Ο LiH # LiCl
# Ο
Ο #
#
# Ο
Ο #
#
#
Ο #
#
Ο
#
Ο
Ο
BM 24h
BM 24h
Intensity (a.u.)
#
#
Ο
#
#
∆
# Ο
Ο
#
#
∆ ∆ ∆Ο Ο
∆
∆ ∆#
∆
∆
∆
∆
∆
Ο
#
Ο
Ο
BM 12h
BM 8h
BM 4h
Ο
BM 2h
Ο
Transmittance (a.u.)
(a)
∆ ∆ ∆ ∆
Ο
BM 1h
Ο
2θ (°)
Mg(NH3 )6 Cl2 -18LiH
Mg(NH3 )6 Cl2 -12LiH
BM 4h
BM 2h
3303
3248
3272 3258
BM 1h
3365
3600 3500 3400 3300 3200 3100 3000
-1
XRD patterns (a) and FTIR spectra (b) of the Mg(NH3 )6 Cl2 -18LiH samples milled for different periods.
Milling
time
(h)
Pressure
increase
(MPa)
H2 molecule
released
(mol)
H/M
(mass%)
0
0
0
0
1
0.043
0.6
0.34
2
0.209
2.8
1.65
4
0.333
4.5
2.64
8
0.428
5.7
3.34
12
0.441
5.9
3.49
24
0.442
5.9
3.50
24
0.439
5.9
4.02
0:442 MPa of hydrogen, which corresponds to 5.9 mol of
H2 molecules per Mg(NH3 )6 Cl2 -18LiH or 3.5 mass% of
hydrogen. It should be mentioned that one of the potential
problems for the nitrogen-containing systems is the coevolution of NH3 . In the present study, the generation of NH3
was measured by NH3 -selective electrode. After 24 h of ball
milling, the concentration of NH3 in the gaseous product
is only about 2.8 ppm, which is possibly due to the presence
of excessive LiH. The rather low NH3 concentration is in
favor of practical applications in the polymer electrolyte
membrane fuel cells (PEMFCs).35)
3.3
3297 3242
Wavenumber (cm )
Table 1 Hydrogen Release in Dependence of Ball Milling for the mixture
of Mg(NH3 )6 Cl2 and LiH.
Samples
BM 8h
3326
Ο
10 20 30 40 50 60 70 80 90
Fig. 4
BM 12h
Dehydrogenation mechanisms in ball milling process
For elucidating the reaction process occurring in the ball
milling process, samples at different milling stages were
collected from the milling jar for XRD and FTIR examinations. Figure 4(a) shows XRD patterns of the Mg(NH3 )6 Cl2 18LiH samples milled for different periods. After 1 h of
milling, the sample consists mainly of the starting chemicals,
Mg(NH3 )6 Cl2 and LiH, since only a few of hydrogen
( 0:6 mol of H2 molecule) is liberated. With the ball
milling proceeding, the LiH phase is still discernable with
considerable intensities. However, the characteristic diffraction peaks of Mg(NH3 )6 Cl2 gradually weaken, and even
disappear after 4 h of milling, which may be due to the
chemical reaction between Mg(NH3 )6 Cl2 and LiH and/or the
amorphization of Mg(NH3 )6 Cl2 caused by energetic milling.
In the meantime, a new phase of LiCl can be unambiguously
identified. When the milling time was extended to 4 h, LiCl
and LiH are the only phases detected by XRD. With further
prolonging the milling time to 24 h, the intensity of the
typical diffraction peaks of the LiH phase gradually decreases, indicating the consumption of LiH. Moreover, the XRD
pattern of LiCl seems almost unchanged in addition to the
slight broadening of the diffraction peaks, implying that the
crystalline size of LiCl is possibly decreased because of the
longtime energetic milling treatment.
Changes in the composition and structure of the
Mg(NH3 )6 Cl2 -18LiH mixture during ball milling were
further verified by FTIR analyses as shown in Fig. 4(b). It
is seen from Fig. 4(b) that the sample milled for 1 h exhibits a
very broad absorbance peaked at around 3365 cm1 , which
should originate from the N-H vibration of the ammine ions
in Mg(NH3 )6 Cl2 . After 2 h of ball milling, two new
absorbances are detected at 3272 and 3258 cm1 with very
weak intensities. Taking into account the existed elements
in the present study, the absorbance at 3272 cm1 can be
assigned to Mg(NH2 )2 , and that at 3258 cm1 should belong
to LiNH2 . For the sample milled for 4 h, the absorbances
at 3365 and 3272 cm1 are still discernable, indicating the
presence of Mg(NH3 )6 Cl2 and Mg(NH2 )2 . However, the
absorbance at 3258 cm1 assignable to LiNH2 shifts toward
lower wavenumber to 3248 cm1 along with the appearance
of another peak at 3303 cm1 . The lowering wavenumber
of the N-H vibration indicates the decreased covalent
nature of N-H bonds, which is possibly due to the partial
substitution of Cl for H in LiNH2 caused by energetic ball
milling. Moreover, milling the sample to 8 h results in that
Hydrogen Storage Properties of the Mg(NH3 )6 Cl2 -LiH Combined System
the doublet N-H vibration of LiNH2 further decreases to
3242/3297 cm1 . When the ball milling time was extended
to 12 h, the N-H vibration of LiNH2 remains unchanged at
3242/3297 cm1 . At the same time, the typical N-H vibration
of Mg(NH2 )2 can be clearly identified at 3272/3326 cm1
along with the weaken absorbance at 3365 cm1 . After 24 h
of ball milling, the absorbance at 3365 cm1 is almost
not detectable, indicating a complete consumption of
Mg(NH3 )6 Cl2 . There are two doublet absorbances at 3242/
3297 and 3272/3326 cm1 with considerable intensities,
which manifests that Mg(NH2 )2 and LiNH2 should exist in
the post-24 h milled sample.
Obviously, the binary Mg(NH3 )6 Cl2 -18LiH combined
system releases about 6 mol of H2 molecules to convert
into the mixture of LiCl, LiH, LiNH2 and Mg(NH2 )2
during ball milling. Therefore, the chemical process occurring in the milling process can be described by the following
reaction:
BM 24h
MgCl2 (NH3 )6 þ 18LiH ! Mg(NH2 )2 þ 4LiNH2 þ 2LiCl þ 12LiH þ 6H2
3.4
Dehydrogenation process of the sample milled for
24 h
As shown in reaction 2, there are 1 mol of Mg(NH2 )2 ,
4 mol of LiNH2 and 12 mol of LiH in the solid residue after
24 h of milling. It is well known that both Mg(NH2 )2 and
LiNH2 can react with LiH to generate a large number of
H2 owing to strong affinity between Hþ in amides and H
in hydride.12) Hence, the dehydrogenation behaviors of the
Mg(NH3 )6 Cl2 -18LiH mixture milled for 24 h were further
investigated in the thermal decomposition process by means
of TPD and volumetric method. The results are shown in
Fig. 5. Obviously, hydrogen desorption from the Mg(NH3 )6 Cl2 -18LiH sample milled for 24 h is a two-step reaction with
peak temperatures at 170 C and 240 C, respectively. Quantitative measurement of hydrogen desorption revealed that
the first-step reaction took place in the temperature range of
130–190 C to evolve about 1.18 mass% of hydrogen, which
corresponds to 2.0 mol of H2 molecules. As the temperature
was further elevated from 190 to 310 C, the second dehydrogenation reaction proceeded and another 2.24 mass% of
hydrogen or 3.8 mol equiv. of H2 molecules were detached.
In total, approximately 3.42 mass% of hydrogen or 5.8 mol
equiv. of H2 molecules were released from the Mg(NH3 )6 Cl2 -18LiH sample milled for 24 h at 130–310 C. Together
with 5.9 mol of H2 molecules detached during ball milling,
11.7 mol of H2 molecules were evolved from the starting
mixture of Mg(NH3 )6 Cl2 -18LiH, which is about 6.86 mass%
of hydrogen. Significantly, hydrogen is directly obtained
Intensity (a.u.)
170 °C
240 °C
Volumetric release
0
H/M (mass%)
TPD
(a)
(b)
1.18 wt% 1.9H2
-1
-2
3.42 wt% 5.6H2
-3
-4
0
50
100
150
200
250
300
350
Temperature, T /°C
Fig. 5 TPD curve (a) and volumetric release (b) of the Mg(NH3 )6 Cl2 18LiH sample milled for 24 h.
631
ð2Þ
from the metal ammine complex (Mg(NH3 )6 Cl2 ) through
combining with the binary alkali metal hydride (LiH).
3.5
Dehydrogenation mechanisms in thermal decomposition process
In order to get the clear insight on the dehydrogenation
reaction of the Mg(NH3 )6 Cl2 -18LiH sample milled for 24 h
at elevated temperatures, XRD patterns and FTIR spectra
of the samples dehydrogenated at 180 C and 310 C were
examined as shown in Fig. 6. It can be seen that LiCl and LiH
are still observed in the sample dehydrogenated at 180 C by
means of XRD. Moreover, the diffraction peaks of LiNH2
and Li2 Mg(NH)2 are also visible. After dehydrogenated at
310 C, the LiCl phase seems unchanged and the newly
developed Li2 Mg(NH)2 phase persists. Moreover, no deviation was found for the peak position of Li2 Mg(NH)2 with
respect to the dehydrogenated sample at 180 C, implying that
no further reaction occurs for Li2 Mg(NH)2 as the temperature
was elevated from 180 to 310 C although Nakamura et al.
reported previously that it possibly reacted with LiH to
release more hydrogen at above 200 C.36) The peak intensity
of the LiH phase is obviously weakened with the disappearance of the LiNH2 phase, indicating the consumption of
LiNH2 and LiH. Correspondingly, Li2 NH can be identified
from the peaks at 30.1 , 35.0 , 50.3 and 59.9 (2). FTIR
examination shows that in addition to the N-H vibration
of amide ions at 3242/3297 cm1 assignable to LiNH2 as
described above, a new absorbance at 3174 cm1 is detected
with strong intensity for the sample dehydrogenated at
180 C. As the sample was heated up to 310 C, only a broad
absorbance is discernable at 3167 cm1 , which locates at the
typical N-H vibration range of imide. Careful observation
reveals that this broad absorbance is composed of two peaks
at 3162 and 3173 cm1 . It is well known that the typical
N-H vibrations of Li2 NH and Li2 Mg(NH)2 are at 3163 and
3174 cm1 , respectively.15,37) As a result, it can be believed
that two phases of Li2 NH and Li2 Mg(NH)2 are involved in
the solid residues after dehydrogenated at 310 C. According
to dehydrogenation measurements and structural analyses,
we can give a picture of the chemical process in heating
process of the Mg(NH3 )6 Cl2 -18LiH combined system. As
described above, the starting sample of Mg(NH3 )6 Cl2 -18LiH
converts to the mixture of Mg(NH2 )2 -4LiNH2 -2LiCl-12LiH
after 24 h of milling. As the post-milled sample was heated
from room temperature to 180 C, Mg(NH2 )2 first reacts with
LiH to release H2 and produce Li2 Mg(NH)2 .
632
Y. Liu et al.
*
&
# #:
(a)
LiCl Ο: LiH ⊥: LiNH2
Intensity (a.u.)
&
&
⊥
#
&* Ο
Ο
&&
*
#
& Ο
#
#
Dehy at 180°C
&⊥
⊥
#
&
Ο
Ο &&
⊥
#&
⊥
&⊥
# # #&Ο
#
Ο
Transmittance (a.u.)
Dehy at 310°C
*
&
#
#
3167
Dehy at 180°C
3242
10 20 30 40 50 60 70 80 90 3600
Fig. 6
3297
#
2θ (°)
(b)
Dehy at 310°C
&: Li2MgN2H2 *: Li2NH
3400
3174
3200
3000
-1
Wavenumber (cm )
XRD patterns (a) and FTIR spectra (b) of the post-24 h milled Mg(NH3 )6 Cl2 -18LiH samples at different dehydrogenation stages.
Mg(NH2 )2 þ 2LiH ! Li2 Mg(NH)2 þ 2H2
ð3Þ
In this case, 1 mol of Mg(NH2 )2 can react with 2 mol of
LiH to liberate 2 mol of H2 , which consists well with the
experimental result as shown in Fig. 5. With further
increasing the temperature to 310 C, the reaction between
LiNH2 and LiH may take place to yield Li2 NH and H2 as
described by the following reaction:
LiNH2 þ LiH ! Li2 NH þ H2
ð4Þ
Since there are 4 mol of LiNH2 in the sample milled for
24 h, theoretically 4 mol of H2 can be desorbed from
reaction 4. In the present study, 2:24 mass% of hydrogen
is released in the temperature range of 180–310 C, which
is equivalent to 3.7 mol of H2 molecule. It is close to the
theoretical dehydrogenation amount of 4 mol of H2 . On the
other hand, it should be mentioned that no further reaction
between Li2 Mg(NH)2 and LiH was observed at 180–310 C
as the XRD pattern of Li2 Mg(NH)2 remains unchanged
(Fig. 6).
Therefore, the dehydrogenation process of the post-24 h
milled sample can be expressed as follows:
130{180 C
Mg(NH2 )2 þ 4LiNH2 þ 6LiH ! Li2 Mg(NH)2 þ 4LiNH2 þ 4LiH þ 2H2
180{310 C
ð5Þ
! Li2 Mg(NH)2 þ 4Li2 NH þ 6H2
Thus, the overall reaction between Mg(NH3 )6 Cl2 and LiH with a molar ratio of 1 : 18 is:
BM 24h
Mg(NH3 )6 Cl2 þ 18LiH ! Mg(NH2 )2 þ 4LiNH2 þ 2LiCl þ 12LiH þ 6H2
130{200 C
! Li2 Mg(NH)2 þ 4LiNH2 þ 2LiCl þ 10LiH þ 8H2
ð6Þ
200{350 C
! Li2 Mg(NH)2 þ 4Li2 NH þ 2LiCl þ 6LiH þ 12H2
Totally, the hydrogen release from the above reaction
amounts to 12 mol of H2 molecules, which corresponds to
about 7.0 mass% of hydrogen.
3.6
Reversibility of hydrogen storage in Mg(NH3 )6 Cl2 LiH combined system
In general, a thermodynamically reversible hydrogen
storage process should possess an endothermic nature for
hydrogen desorption.38) Figure 7 shows the DSC curve of the
sample milled for 24 h. It can be seen that hydrogen
desorption from the Mg(NH3 )6 Cl2 -18LiH sample milled for
24 h exhibits an endothermic nature, implying the dehydro-
genation reaction is thermodynamically reversible. Two
endothermic peaks was observed at 174 C and 235 C,
respectively, in the DSC curve, which consists well with
the results of the TPD curve, further confirming that the
hydrogen desorption from the Mg(NH3 )6 Cl2 -18LiH sample
milled for 24 h is a two-step reaction.
To ascertain the hydrogen storage reversibility and
simultaneously obtain higher hydrogen storage capacity,
the mixture of Mg(NH3 )6 Cl2 -LiH with a molar ratio of 1 : 12
was re-designed since 6 mol of LiH are excessive and not
involved in the dehydrogenation reaction of the Mg(NH3 )6 Cl2 -18LiH sample in the temperature range 25–310 C as
Hydrogen Storage Properties of the Mg(NH3 )6 Cl2 -LiH Combined System
633
0.4
0
endo
-1
0.0
H/M (mass%)
DSC (mW/mg)
0.2
-0.2
174 °C
-0.4
-0.6
-2
Dehydrogenation
Rehydrogenation
-3
-4
4.0 wt% 5.9H2
-0.8
235 °C
0
50
100
150
200
-5
250
300
0
350
50
Fig. 7
100
150
200
250
300
450
Temperature, T /°C
Temperature, T /°C
DSC curve of the post-24 h milled Mg(NH3 )6 Cl2 -18LiH sample.
shown in reaction 5. Thus, the Mg(NH3 )6 Cl2 -12LiH mixture
was prepared under same conditions. The hydrogen desorption performance of the Mg(NH3 )6 Cl2 -12LiH mixture in
the ball milling and subsequent heating process was further
investigated. Theoretically, hydrogen released from the
Mg(NH3 )6 Cl2 -12LiH mixture amounts to 8.2 mass%, a
1.2 mass% increase relative to the Mg(NH3 )6 Cl2 -18LiH
sample ( 7:0 mass%), which is well proved by our
experimental result as total hydrogen desorption amount
of about 8.02 mass% was obtained in the ball milling and
subsequent heating process (Table 1 and Fig. 8). The
dehydrogenated Mg(NH3 )6 Cl2 -12LiH mixture was subsequently hydrogenated under 10 MPa of hydrogen. Figure 8
also shows the hydrogenation curve of the sample dehydrogenated at 310 C. It is seen that hydrogen absorption starts at
about 65 C, and the amount of hydrogen absorbed gradually
increased with elevating the temperature. As the sample
was heated to 300 C, the hydrogen charged amounts to
4.0 mass%, which is equivalent to 5.9 mol of H2 molecules.
The hydrogen uptake exhibits in part reversibility. In general,
the released hydrogen during ball milling is rather difficult
to be recharged.39,40) On the other hand, it is well known that
reactions 3 and 4 are completely reversible at 300 C under
Fig. 8 Dehydrogenation and re-hydrogenation curves of the dehydrogenated Mg(NH3 )6 Cl2 -12LiH sample.
10 MPa of hydrogen.11–13) As a result, we believe that in the
present study, the reversible hydrogen storage in the postmilled Mg(NH3 )6 Cl2 -12LiH combined system should be
attributed to the chemical reactions between Mg(NH2 )2 ,
LiNH2 and LiH as described by reactions 3 and 4.
4.
Conclusions
Mg(NH3 )6 Cl2 was synthesized by reacting anhydrous
MgCl2 with ammonia, and the Mg(NH3 )6 Cl2 -18LiH combination was prepared by ball milling the corresponding
chemicals. Hydrogen storage properties of the Mg(NH3 )6 Cl2 -18LiH combination were systematically investigated.
It was found that the starting sample of Mg(NH3 )6 Cl2 18LiH mixture liberated 5:9 mol of H2 molecules to
convert to the mixture of Mg(NH2 )2 -4LiNH2 -2LiCl-12LiH
after ball milling for 24 h. In the subsequent heating
process, Mg(NH2 )2 and LiNH2 reacted successively with
LiH to release additional 2 and 4 mol of H2 molecules,
respectively. The overall dehydrogenation reaction occurring
in ball milling and heating process can be expressed as
follows:
BM 24h
MgCl2 (NH3 )6 þ 18LiH ! Mg(NH2 )2 þ 4LiNH2 þ 2LiCl þ 12LiH þ 6H2
130{180 C
! Li2 Mg(NH)2 þ 4LiNH2 þ 2LiCl þ 10LiH þ 8H2
180{310 C
! Li2 Mg(NH)2 þ 4Li2 NH þ 2LiCl þ 6LiH þ 12H2
Further hydrogenation examination showed that 6 mol of
H2 molecules could be recharged into the dehydrogenated
Mg(NH3 )6 Cl2 -12LiH mixture at 300 C under 10 MPa of
hydrogen, exhibiting good reversibility for hydrogen storage.
Acknowledgements
The authors would like to acknowledge the financial
supports from the National Natural Science Foundation
of China (50701040, 51025102 and 50871100), from
the Ministry of Science and Technology of China
(2009AA05Z106 and 2010CB631304) and from the Speci-
alized Research Fund of the Doctoral Program of Higher
Education of China (20070335005).
REFERENCES
1) L. Schlapbach and A. Züttel: Nature 414 (2001) 353–358.
2) A. Züttel: Mater. Today 6 (2003) 24–33.
3) U. Eberle, M. Felderhoff and F. Schüth: Angew. Chem. Int. Ed. 48
(2009) 6608–6630.
4) P. Chen and M. Zhu: Mater. Today 11 (2008) 36–43.
5) C. Liu, F. Li, L. P. Ma and H. M. Cheng: Adv. Mater. 22 (2010) E28–
E62.
6) B. Bogdanović and M. Schwickardi: J. Alloy. Compd. 253–254 (1997)
634
Y. Liu et al.
1–9.
7) S. Orimo, Y. Nakamori, J. R. Eliseo, A. Züttel and C. M. Jensen: Chem.
Rev. 107 (2007) 4111–4132.
8) F. Fang, S. Y. Zheng, G. R. Chen, G. Sang, B. He, S. Q. Wei and L. D.
Sun: Acta Mater. 57 (2009) 1959–1965.
9) Y. F. Liu, F. H. Wang, Y. H. Cao, M. X. Gao, H. G. Pan and Q. D.
Wang: Energy Environ. Sci. 3 (2010) 645–653.
10) T. Sun, B. Zhou, H. Wang and M. Zhu: J. Alloy. Compd. 467 (2009)
413–416.
11) P. Chen, Z. T. Xiong, J. Z. Luo, J. Y. Lin and K. L. Tan: Nature 420
(2002) 302–304.
12) Z. T. Xiong, G. T. Wu, J. J. Hu and P. Chen: Adv. Mater. 16 (2004)
1522–1525.
13) W. F. Luo: J. Alloy. Compd. 381 (2004) 284–287.
14) H. Y. Leng, T. Ichikawa, S. Hino, N. Hanada, S. Isobe and H. Fujii:
J. Phys. Chem. B 108 (2004) 8763–8765.
15) Y. F. Liu, K. Zhong, K. Luo, M. X. Gao, H. G. Pan and Q. D. Wang:
J. Am. Chem. Soc. 131 (2009) 1862–1870.
16) A. Züttel, P. Wenger, S. Rentsch, P. Sudan, Ph. Mauron and Ch.
Emmenegger: J. Power Sources 118 (2003) 1–7.
17) J. J. Vajo, S. L. Skeith and F. Mertens: J. Phys. Chem. B 109 (2005)
3719–3722.
18) X. B. Yu, D. M. Grant and G. S. Walker: J. Phys. Chem. C 112 (2008)
11059–11062.
19) P. J. Wang, L. P. Ma, Z. Z. Fang, X. D. Kang and P. Wang: Energy
Environ. Sci. 2 (2009) 120–123.
20) H. W. Li, K. Kikuchi, T. Sato, Y. Nakamori, N. Ohba, M. Aoki, K.
Miwa, S. Towata and S. Orimo: Mater. Trans. 49 (2008) 2224–2228.
21) V. M. Parvanov, G. K. Schenter, N. J. Hess, L. L. Daemen, M. Hartl,
A. C. Stowe, D. M. Camaioni and T. Autrey: Dalton Trans. 33 (2008)
4514–4522.
22) Z. T. Xiong, C. K. Yong, G. T. Wu, P. Chen, W. Shaw, A. Karkamkar,
T. Autrey, M. O. Jones, S. R. Johnson, P. P. Edwards and W. I. F.
David: Nature Mater. 7 (2008) 138–141.
23) J. Z. Zhao, J. F. Shi, X. W. Zhang, F. Y. Cheng, J. Liang, Z. L. Tao and
J. Chen: Adv. Mater. 22 (2010) 394–397.
24) C. H. Christensen, R. Z. Sorensen, T. Johannessen, U. J. Quaade, K.
Honkala, T. D. Elmøe, R. Køhler and J. K. Nørskov: J. Mater. Chem. 15
(2005) 4106–4108.
25) A. Klerke, C. H. Christensen, J. K. Nørskov and T. Vegge: J. Mater.
Chem. 18 (2008) 2304–2310.
26) H. Yamamoto, H. Miyaoka, S. Hino, H. Nakanishi, T. Ichikawa and Y.
Kojima: Int. J. Hydrogen Energy 34 (2009) 9760–9764.
27) Y. H. Hu and E. Ruckenstein: J. Phys. Chem. A 107 (2003) 9737–9739.
28) T. Markmaitree, W. Osborn and L. L. Shaw: J. Power Sources 180
(2008) 535–538.
29) J. H. Luo, X. D. Kang and P. Wang: ChemPhysChem 11 (2010) 2152–
2157.
30) Z. T. Xiong, Y. S. Chua, G. T. Wu, W. L. Xu, P. Chen, W. Shaw, A.
Karkamkar, J. Linehan, T. Smurthwaite and T. Autrey: Chem.
Commun. 43 (2008) 5595–5597.
31) M. Tsubota, S. Hino, H. Fujii, C. Oomatsu, M. Yamana, T. Ichikawa
and Y. Kojima: Int. J. Hydrogen Energy 35 (2010) 2058–2062.
32) G. M. Long, P. H. Ma, Z. M. Wu, M. Z. Li and M. C. Chu: Thermo.
Acta 412 (2004) 149–153.
33) I. C. Hwang, T. Drews and K. Seppelt: J. Am. Chem. Soc. 122 (2000)
8486–8489.
34) A. D. Allen and C. V. Senoff: Canad. J. Chem. 45 (1967) 1337–1341.
35) F. A. Uribe, S. Gottesfeld and T. A. Zawodzinski: J. Electrochem. Soc.
149 (2002) A293–A296.
36) Y. Nakamura, S. Hino, T. Ichikawa, H. Fujii, H. W. Brinks and B. C.
Hauback: J. Alloy. Compd. 457 (2008) 362–367.
37) W. F. Luo and S. Sichafoose: J. Alloy. Compd. 407 (2006) 274–281.
38) G. Sandrock: J. Alloy. Compd. 293–295 (1999) 877–888.
39) J. J. Hu, G. T. Wu, Y. F. Liu, Z. T. Xiong, P. Chen, K. Murata, K.
Sakata and G. Wolf: J. Phys. Chem. B 110 (2006) 14688–14692.
40) Y. F. Liu, J. J. Hu, G. T. Wu, Z. T. Xiong and P. Chen: J. Phys. Chem. C
111 (2007) 19161–19164.