Synthesis of calcium alanate and its dehydriding performance enhanced by FeF3 doping

Article abstract of DOI:10.1016/j.jallcom.2010.09.155

Ca(AlH₄)₂ was synthesized by ball-milling the mixture of NaAlH₄ and CaCl₂ in a molar ratio of 2:1 and under a hydrogen atmosphere of 1 MPa. The results indicate that the reactants have entirely transformed to Ca(AlH₄)₂ with a byproduct of NaCl after ball-milling for 48 h. Investigations of dehydriding behavior of the as-prepared Ca(AlH₄)₂ sample show that approximately 5.2 wt.% of hydrogen is desorbed during the first two dehydrogenation reactions of Ca(AlH₄)₂, which exhibit an exothermic event at 148 °C and an endothermic event at 267 °C, respectively. The high temperature dehydrogenation at 267 °C mainly concentrates on the thermolysis of CaAlH₅ intermediate. FeF₃-doped Ca(AlH₄) ₂ system represents an improved dehydriding performance, the dehydrogenation temperature of CaAlH₅ intermediate is decreased about 43 °C. After FeF₃ doping, the apparent activation energy of CaAlH₅ is reduced from 153.4 kJ/mol (undoped) to 88.3 kJ/mol (doped), it renders a possibility to realize the rehydrogenation of CaAlH₅. The catalytic effect is attributed to a fluorine transfer reaction that occurred to generate CaF₂ and Fe catalysts.

Full text of DOI:10.1016/j.jallcom.2010.09.155

Journal of Alloys and Compounds 509 (2011) 590–595
Contents lists available at ScienceDirect
Journal of Alloys and Compounds
journal homepage: www.elsevier.com/locate/jallcom
Synthesis of calcium alanate and its dehydriding performance enhanced
by FeF3 doping
Changxu Li, Xuezhang Xiao, Lixin Chen∗, Kun Jiang, Shouquan Li, Qidong Wang
Department of Materials Science and Engineering, Zhejiang University, Hangzhou 310027, People’s Republic of China
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 14 July 2010
Received in revised form 25 August 2010
Accepted 7 September 2010
Available online 20 October 2010
Ca(AlH4)2 was synthesized by ball-milling the mixture of NaAlH4 and CaCl2 in a molar ratio of 2:1 and
under a hydrogen atmosphere of 1 MPa. The results indicate that the reactants have entirely transformed
to Ca(AlH4)2 with a byproduct of NaCl after ball-milling for 48 h. Investigations of dehydriding behavior
of the as-prepared Ca(AlH4)2 sample show that approximately 5.2 wt.% of hydrogen is desorbed dur-
ing the first two dehydrogenation reactions of Ca(AlH4)2, which exhibit an exothermic event at 148 ◦C
and an endothermic event at 267 ◦C, respectively. The high temperature dehydrogenation at 267 ◦C
mainly concentrates on the thermolysis of CaAlH5 intermediate. FeF3-doped Ca(AlH4)2 system repre-
sents an improved dehydriding performance, the dehydrogenation temperature of CaAlH5 intermediate
is decreased about 43 ◦C. After FeF3 doping, the apparent activation energy of CaAlH5 is reduced from
153.4 kJ/mol (undoped) to 88.3 kJ/mol (doped), it renders a possibility to realize the rehydrogenation of
CaAlH5. The catalytic effect is attributed to a fluorine transfer reaction that occurred to generate CaF2 and
Fe catalysts.
Keywords:
Calcium alanate
Dehydriding performance
Iron fluoride
Ball-milling
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
of 5.9 wt.% hydrogen can be considered for practical applications
[9].
Great efforts have being paid to find an efficient means of storing
hydrogen as a fuel on board, but there is still a long way between
the targeted and currently available technologies. Since Bogdanovic
and Schwickardi [1] discovered that sodium alanate doped with Ti
could be reversibly dehydrogenated and rehydrogenated, alanates
became one of the most promising hydrogen storage materials,
generally characterized by high hydrogen capacity and less than
perfect thermodynamic and kinetic properties [2–5]. Current stud-
ies have extended to alternative complex hydride systems in order
to clarify the hydrogen absorption/desorption mechanism and
screen for potential materials of more practical value [6–7]. LiAlH4
has a theoretical capacity of 10.6 wt.% H2, the direct hydrogenation
of LiAlH4 is difficult to be achieved because of its unstable ther-
modynamics. The decomposition of KAlH4 occurs through three
endothermic events at temperatures of 294, 311, and 347 ◦C with
the release of 3.8 wt.% hydrogen, and its catalyzed sample exhibits
dehydrogenation temperatures dropped by nearly 50 ◦C, however,
which are still rather high when compared to sodium and lithium
alanates [8]. Ca(AlH4)2 with a total hydrogen capacity of 7.9 wt.%
has been far less studied. Its decomposition process mainly consists
of four-step reactions, of which the first two steps with the release
Ca(AlH4)2 → CaAlH5 + Al + 3/2H2
CaAlH5 → CaH2 + Al + 3/2H2
CaH2 + 4Al → Al4Ca + H2
(1)
(2)
(3)
(4)
CaH2 + Al4Ca → 2Al2Ca + H2
Calcium alanate was first synthesized with CaH2 and AlCl3 in
THF in the 1950s, mainly used as a reducing agent for a long
time [10]. Later works about the synthesis of calcium alanate
went through a metathesis reaction of CaCl2 and NaAlH4 by
wet-chemical or mechanochemistry method, thermal dissocia-
tion process as well as thermodynamic and kinetic properties of
Ca(AlH4)2 were also illustrated [11–13]. Recently, Kabbour et al.
[14] reported a direct synthesis path of Ca(AlH4)2 by ball-milling
the mixture of AlH3 and CaH2. Besides, it was reported that CaAlH5
as an important intermediate could also be synthesized by using
AlH3 and CaH2 [15]. However, the preparation of AlH3 is still a chal-
lenge. Another way of preparing alkaline earth alanates through
hydrogenation of alkaline earth-Al alloy was proved to be invalid
for Ca(AlH4)2 as well as Mg(AlH4)2, in despite of its feasibility in
forming Sr2AlH7 and Ba2AlH7 [16,17]. The structural properties of
Ca(AlH4)2 and CaAlH5 have been determined by first-principle den-
sity function theory (DFT) calculation in the past few years. Crystal
structure of Ca(AlH4)2 is confirmed to be CaB2F8-type and that of
CaAlH5 consists of corner sharing [AlH6] octahedral that form heli-
∗
Corresponding author. Tel.: +86 571 8795 1152; fax: +86 571 8795 1152.
E-mail address: lxchen@zju.edu.cn (L. Chen).
0925-8388/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2010.09.155

C. Li et al. / Journal of Alloys and Compounds 509 (2011) 590–595
591
cal chains [18–21]. From the foregoing, attempts to improve the
properties of Ca(AlH4)2 and CaAlH5, such as doping or element
substitute, are also important. Better understanding of Ca(AlH4)2
system renders an important perspective to investigate other com-
plex hydrides with high capacity.
In this work, we attempt to illustrate the synthetic process
and catalytic modification of Ca(AlH4)2. Furthermore, the effect of
FeF3-doping on the dehydriding properties of Ca(AlH4)2 has been
investigated in an attempt to understand the role played by FeF3
catalyst on the enhancement of the dehydriding properties.
2. Experiments
Synthesis operations of Ca(AlH4)2 sample were conducted using commercial
NaAlH4 (with purity ≥90%, Aldrich) and CaCl2 (with purity ≥96%, Alpha Aesar)
that were used without further purification. To prevent any contamination, whole
synthesis process was done in a glovebox filled with high pure argon atmosphere
(99.999% purity) with the content of water and oxygen below 5 ppm. Prior to milling,
NaAlH4 and CaCl2 were mixed in a molar ratio of 2:1 according to the reaction:
[2NaAlH4 + CaCl2 → Ca(AlH4)2 + 2NaCl] with a total sample amount of 2.74 g. The
mixture was transferred into a sealed stainless steel vial equipped with a gas valve,
under 1 MPa of pure hydrogen atmosphere. The ball to powder ratio was about 40:1.
Milling was carried out on a planetary ball mill (QM-3SP4, Nanjing) at 400 rpm with
different milling time. Every 0.4 h a pause (0.1 h) was set to counterbalance the
temperature rise during milling. The sample of 10 wt.% FeF3-doped calcium alanate
was prepared by milling FeF3 (with purity ≥98%, Aldrich) and the synthetic prod-
uct (previously milled for 48 h) for 3 h under 0.5 MPa hydrogen pressure with ball
to powder ratio of 60:1, the milling rate was reduced to 200 rpm to prevent the
untimely dissociation.
Structural characterization was determined by a Rikagu D/Max-RA X-ray diffrac-
tometer with Cu K␣ radiation at 40 kV and 30 mA from 10◦ to 90◦ (2ꢀ) with step
increments of 0.05◦, a homemade container filled with argon was used to prevent
the samples against contamination. IR spectrum of the sample was recorded in the
range of 400–4000 cm−1 at ambient condition in air using a Bruker Tensor 27 FTIR
spectrometer. The dehydriding processes of the samples were carried out on a cal-
ibrated Sievert’s type apparatus, details are described in the results part. For the
differential scanning calorimetry and thermogravimetric (DSC/TG), measurements
were conducted on a Netasch DSC 449F3 at different heating rate from 30 to 450 ◦C.
X-ray photoelectron spectroscopy (XPS) was used to monitor the change of FeF3
phase during the catalytical process. XPS analysis was carried out on a VG Escalab
Mark II spectrometry with CAE 50 eV and Mg K␣ radiation.
Fig. 1. XRD patterns of 2NaAlH4 + CaCl2 samples milled for different milling times.
Table 1
The frequencies (cm−1
[8,10,22].
)
of [AlH4]− group in different chemical environments
Complex
Ca(AlH4)2
1798
Mg(AlH4)2
1835
NaAlH4
1675
KAlH4
1785
LiAlH4
1757
Frequencies
1615
the results show that main contributions to the Al–H stretching
vibrations are found at 1798 cm−1. In the fingerprint, the band at
695 cm−1 is due to Ca–Al–H bending mode. The results from Fig. 2
together with Fig. 1 indicate that the reactants have transformed
to Ca(AlH4)2 and NaCl after ball-milling for 48 h.
3.2. Dehydrogenation performance of the as-synthesized sample
The decomposition performance and desorbed hydrogen
amount of the as-synthesized Ca(AlH4)2 + 2NaCl sample milled for
48 h were studied by DSC/TG measurement (Fig. 3). It can be seen
that the dissociation process of the as-synthesized sample is mainly
divided into four stages (as reactions (1)–(4)).
The first step of decomposition is an exothermic process. DFT
calculations indicated that the enthalpy value of the first step was
−9 kJ/mol H2 at the decomposition temperature of 127 ◦C [20]. TG
3. Results and discussions
3.1. The synthesis of Ca(AlH4)2
The synthesis of calcium alanate by ball-milling the mixture of
2NaAlH4 + CaCl2 was based on a metathesis reaction as follows:
2NaAlH4 + CaCl2 → Ca(AlH4)2 + 2NaCl
(5)
The microstructures of as-synthesized samples at different ball-
milling stages were characterized by XRD measurement. Fig. 1
shows the XRD patterns of the samples milled for 3–48 h. For com-
parison, the XRD pattern of mixed 2NaAlH4 + CaCl2 sample that is
hand-milled using an agate mortar and pestle for 5 min is also pre-
sented in Fig. 1. It can be seen that some diffraction peaks from
NaAlH4 and CaCl2 disappear after ball milling for 3 h, the peak from
NaCl emerges concurrently. Ca(AlH4)2 might be present in X-ray
amorphous state initially, small diffraction peaks from Ca(AlH4)2
are found until the ball-milling time extends to 12 h, and are inten-
sified with the increasing milling time. After milling for 48 h, the
diffraction peaks of NaAlH4 and CaCl2 disappear completely, the
product is a mixture of Ca(AlH4)2 and NaCl.
The characterization of Ca(AlH4)2 can also be performed by
IR measurement. As it is known that the FTIR spectrum consists
mainly of two wide bands, of which the ∼1800 cm−1 band corre-
sponds to the Al-H stretching vibrations. The characteristic band of
Al-H bond changes with its chemical environment. Table 1 shows
the frequencies (cm−1) of [AlH4]− group in different chemical
environments. Combined with the spectra profile of the metathe-
sis reaction product after ball-milling for 48 h shown in Fig. 2,
Fig. 2. FTIR spectra of the metathesis reaction product after ball-milling for 48 h,
the band at 1798 cm−1 corresponds to the Al-H stretching vibrations.

592
C. Li et al. / Journal of Alloys and Compounds 509 (2011) 590–595
Fig. 5. TPD curves of the as-prepared and FeF3-doped samples of Ca(AlH4)2 + 2NaCl
Fig. 3. DSC/TG curves of the as-synthesized Ca(AlH4)2 + 2NaCl sample milled for
at a heating rate of 5 ◦C/min.
48 h at a heating rate of 10 ◦C/min.
Al phases. After the temperature arises to 450 ◦C, the formed Ca–Al
alloys mainly consist of Al4Ca and Al2Ca.
curve shows that the amount of released hydrogen is about 2.4 wt.%.
The value is based on the ratio of the evaluated amount of des-
orbed hydrogen to the amount of Ca(AlH4)2 calculated from the
reaction. The second step of decomposition begins at nearly 210 ◦C,
corresponds to the decomposition of CaAlH5 to CaH2 and Al phase.
CaAlH5 is viewed as a promising on-board hydrogen storage reac-
tion owing to its good thermodynamics (+32 kJ/mol H2 according
to the experimental value reported [11]). This step lasts until about
2.8 wt.% H2 is desorbed. The last two steps involve the formation
of Ca–Al alloys. According to the in situ XRD analysis reported [12],
Al4Ca formed first in the third step and a portion of it transferred to
Al2Ca in the fourth step. At 450 ◦C, the phase composition includes
a mixture of Al2Ca and Al4Ca. The formation process of Ca–Al alloys
needs higher temperature and is considered to be irreversible and
impractical.
The first two steps of the decomposition of Ca(AlH4)2 are consid-
ered for hydrogen storage application. However, the decomposition
temperatures for reactions (1) and (2) are relatively high in com-
parison with the targeted value, and the thermolysis process of
Ca(AlH4)2 and CaAlH5 takes time. Therefore, it is necessary to
find ways to improve the thermodynamic and kinetic properties
of Ca(AlH4)2 system, clarify the main factors that controlling the
behavior of dehydrogenation, which may contribute to the fur-
ther understanding of thermodynamic and kinetic limitations in
a number of high capacity complex hydrides.
3.3. Influence of FeF3 doping on the dehydrogenation
performance of Ca(AlH4)2
Fig. 4 shows the XRD patterns of the decomposition products
of the as-synthesized Ca(AlH4)2 + 2NaCl sample at different tem-
peratures. It is shown that CaAlH5 remains stable when heating
to 200 ◦C, the pattern of CaAlH5 is in good agreement with the
Rietveld analysis results of Iosub et al. [15] and Weidenthaler et
al. [19]. At 320 ◦C, CaAlH5 is decomposed completely to CaH2 and
Fig. 5 records the courses of thermal dissociations of the as-
prepared and FeF3-doped samples of Ca(AlH4)2 + 2NaCl by Sievert’s
apparatus. Samples are heated from room temperature to 250 ◦C
at 5 ◦C/min. As evident from the figure, the starting point of
decomposition for both is 70 ◦C, the difference is that FeF3-doped
sample experiences a shorter incubation stage before fast dehydro-
genation. The temperature for fast-speed dehydriding is 90 ◦C for
FeF3-doped sample compared to 110 ◦C for the as-prepared sample.
Besides, FeF3-doped sample represents an obscure change in dehy-
drogenation kinetics that reflects the transition from reactions (1)
to (2) when compared with the as-prepared sample, and the total
time-consuming of the decomposition from Ca(AlH4)2 to CaH2 has
been shortened by about 30 min. The results suggests that FeF3-
doping is able to improve the dehydrogenation performance of
Ca(AlH4)2 system effectively.
It is noteworthy that the desorbed hydrogen amount does not
almost decline after 10 wt.% FeF3 is doped. In Fig. 5, the desorbed
hydrogen amounts of the as-prepared and FeF3-doped samples
are 5.5 and 5.48 wt.% respectively, in contrast with the theoretical
hydrogen amount of 5.9 wt.%. The overflow of desorbed hydro-
gen might have been due to the reason that FeF3 participates in
the decomposition reactions, which brings additional hydrogen
released.
In order to better understand the role of FeF3 that improves
the dehydrogenation performance, the thermal analyses were per-
formed by the DSC methods. A series of measurements were
conducted at different heating rates of 3, 5, 8 and 10 ◦C/min for the
as-prepared and FeF3-doped samples. According to the Kissinger
Fig. 4. XRD patterns of the decomposition products of the as-synthesized
Ca(AlH4)2 + 2NaCl sample when heating to different temperatures: (a) room tem-
perature; (b) 200 ◦C; (c) 320 ◦C; (d) 450 ◦C.

C. Li et al. / Journal of Alloys and Compounds 509 (2011) 590–595
593
Table 2
Activation energy (Ea) of CaAlH5, K3AlH6 and Na3AlH6.
Intermediate
CaAlH5
Ea (kJ/mol)
As-prepared
Doped
161 [15], 153.4
88.3
K3AlH6
As-prepared
Doped
131 20 [8]
105 2 [8]
Na3AlH6
As-prepared
Doped
120 [24]
97–100 [24]
for the as-prepared and FeF3-doped samples. It is observed that the
activation energy of the as-prepared sample is 153.4 kJ/mol, which
is in good agreement with the published result in other literature
[15]. With respect to the FeF3-doped sample, the value of activa-
tion energy is 88.3 kJ/mol. The variation of activation energy poses
an interesting approach to investigate the mechanism that hinders
the dehydrogenation and rehydrogenation process.
It has been confirmed that CaAlH5 is thermodynamically
reversible by DSC analyses and DFT calculations, the reaction
enthalpy of CaAlH5 is 32 kJ/mol, which is in the targeted window
suitable for near-ambient hydrogen storage [12,19,20]. However,
the rehydrogenation of CaAlH5 has not been realized experimen-
tally due to its slow kinetics property. Table 2 lists the values of
activation energy of some intermediates for comparison. It is inter-
esting to see that the activation energy tends to be higher with
the increasing of the size of metal cation. The value of CaAlH5
is much higher than those of K3AlH6 and Na3AlH6 which are
technically reversible [8,24], this indicates that mass transport is
much more difficult in the dehydrogenation and rehydrogena-
tion processes of CaAlH5. Fortunately, FeF3-doped sample exhibits
favorable value compared to those of doped K3AlH6 and Na3AlH6.
High activation energy may be the main factor that control-
ling the hydriding/dehydriding kinetics, and this result suggests
a possibility to realize the rehydrogenation of CaAlH5. However,
rehydrogenation attempt under 100 bar H2 and 160 ◦C for 24 h gave
no sign of pressure change. It is suggested that unilateral kinetic
enhancement is not enough for rehydriding. Further investigations
about the improvements from thermodynamic and kinetic opti-
mization of CaAlH5 are underway.
Fig. 6. DSC curves of the as-prepared and FeF3-doped sample of Ca(AlH4)2 + 2NaCl
at a heating rate of 5 ◦C/min.
relation of reaction (6) [23]:
ꢀ
ꢁ
C
Tp2
Ea
In
= −
+ A
(6)
RTp
where Ea is the activation energy for the transformation, R is the gas
constant and A is a constant, C is heating rate and Tp is transforma-
tion peak temperature that changed with different heating rate (C),
usually Tp shifts towards the higher temperature side with increas-
ing heating rate. Fig. 6 shows the DSC curves of the as-prepared and
FeF3-doped sample at a heating rate of 5 ◦C/min. It is found that the
decomposition process of FeF3-doped sample remains four steps.
In addition, the peak intensity for the first step changes and disso-
ciation temperature of the second step is significantly dropped by
about 43 ◦C compared with that of the as-prepared sample.
As is known that high temperature dehydrogenation mainly
concentrates on the thermolysis of CaAlH5 intermediate. In the
above, FeF3 doping makes for great reduction of decomposition
temperature of CaAlH5. In order to further understand the role of
FeF3 played on the dehydrogenation process of CaAlH5, Fig. 7 gives
the activation energies of CaAlH5 decompositions calculated from
the data performed at different heating rates (3, 5, 8 and 10 ◦C/min)
3.4. The mechanism of FeF3 doping for enhanced kinetics
To verify the structural changes during the dehydrogena-
tion process of the FeF3-doped sample of Ca(AlH4)2 + 2NaCl, and
hence to understand the mechanism of FeF3 doping for enhanced
Fig. 7. Kissinger plots of CaAlH5 decomposition reactions of the as-prepared (a) and FeF3-doped (b) samples.

594
C. Li et al. / Journal of Alloys and Compounds 509 (2011) 590–595
Fig. 9. XPS curves of (a) the decomposition products of FeF3-doped sample at 250 ◦
and (b) commercial FeF3.
C
Fig. 8. XRD patterns of the decomposition products of the FeF3-doped sample of
Ca(AlH4)2 + 2NaCl at different decomposition temperatures: (a) room temperature;
(b) 160 ◦C; (c) 250 ◦C.
dehydriding performance of Ca(AlH4)2 system, X-ray diffraction
measurements were carried out. As shown in Fig. 8, when heat-
ing to 160 ◦C, Ca(AlH4)2 decomposes into CaAlH5 and Al phases,
FeF3 is visible and no other fluoride is detected, which indicate
that FeF3 is stable during the first step decomposition reaction. At
320 ◦C, a new phase confirmed as CaF2 appears following the dis-
appearance of FeF3. Based on XRD measurements and DSC analysis
mentioned above, it can be conclude that FeF3 might react with
CaAlH5 according to the following reaction:
Besides, the generated transitional metal Fe, like other transition-
metal additives such as Ti or Ce, might be responsible for the
improved kinetics of dehydrogenation. It is generally accepted that
mass transport is the rate-limiting step during the decomposi-
tion of alanates, and dehydriding kinetics should be controlled by
the diffusion of some species [30–32]. Hence, the highly dispersed
transition-metal Fe might act as active catalyst to facilitate the dif-
fusion process and dehydrogenation reaction.
CaAlH5 + 2/3FeF3 → CaF2 + 2/3Fe + Al + 5/2H2
(7)
4. Conclusions
The formation heats (ꢁHf), of CaF2, FeF3, and CaAlH5 are −1228,
1042.3 and 31.6 kJ/mol, respectively [12,25], the total enthalpy
change of reaction (7) is −501.5 kJ/mol. This indicates that reac-
tion (7) is thermodynamically favorable. This result best explains
the phenomenon reflected in Fig. 5 that desorbed hydrogen amount
of Ca(AlH4)2 system does not almost decline after 10 wt.% FeF3 is
doped. The generation of CaF2 leads to additional liberation of H2,
which allows the desorbed hydrogen amount changes in a limited
range.
The synthesis and characterization of Ca(AlH4)2 were investi-
gated. The thermolysis of Ca(AlH4)2 involves four steps, of which
the first two steps have a total amount of 5.2 wt.% desorbed
hydrogen. CaAlH5 as an important intermediate, has moderate
decomposition enthalpy (+32 kJ/mol) and a theoretical gravimetric
density of 4.2 wt.%, and is thought to be a promising near-ambient
hydrogen storage material, but the decomposition of CaAlH5 to
CaH2 begins at nearly 210 ◦C and its re-hydrogenation is still a chal-
lenge though DSC analysis and thermodynamic data have shown
this possibility. The FeF3-doped Ca(AlH4)2 exhibits similar thermol-
ysis process except for significant reduction of dehydrogenation
temperature (∼43 ◦C) and activation energy (∼65 kJ/mol) for the
decomposition of CaAlH5, making it perhaps a realized reversible
hydrogen storage material that needs to be confirmed in follow-
ing works. At last, the mechanism of FeF3 doping for the enhanced
dehydrogenation performance of Ca(AlH4)2 system has been dis-
cussed. Plausible reasons include the reaction of CaAlH5 and FeF3,
and the catalyst effects are rendered by the generated CaF2 and
transitional metal Fe catalyst.
The absence of Fe phase in Fig. 8(c) is probably due to two rea-
sons: First, the Fe phase is overlapped with the Al phase in XRD
pattern, which induces that the characteristic peaks of Fe phase
is obscured; Second, small amount of Fe phase might be highly-
dispersed and in an extra fine form. However, the existence of Fe
could be confirmed by XPS measurements. Fig. 9 shows the XPS
spectra of the yields of reaction (7) as compared to the commercial
FeF3 powder. It can be observed that, when the reaction is com-
pleted, Fe2p1/2, Fe2p3/2 and F1s peaks from FeF3 disappear, a bulge
region around 710 eV is visualized with a few peaks, of which the
peak at 706.6 eV is attributed to Fe2p3/2 in Fe, and the others are
attributed to Fe oxides [26], which may be formed in the testing
process. The weakness of peak intensity is probably duo to the rea-
son that XPS spectra only detect the very limited penetration depth
of the sample, in which very small amount of Fe disperses.
Reports about the improvement of hydriding/dehydriding prop-
erties by the intelligent manipulation of additional reactions exist
in a great number [27,28]. It provides an available means to meet
temperature and pressure requirements for complex hydrides as
reversible hydrogen storage materials for mobile applications. In
our work, the fluoride doping results in a substitution of H− by
F− anion in the hydride lattice and hence probably a favorable
thermodynamics adjustment, F− acts as a functional anion [29].
Acknowledgments
This work was financially supported by the National Basic
Research Program of China under grant no. 2007CB209701 and
2010CB631300, the National Natural Science Foundation of China
under grant no. 50871099 and 50631020, the China Postdoctoral
Science Foundation under grant no. 200902622, the Program for
New Century Excellent Talents in University under grant no. NCET-
07-0741, and the University Doctoral Foundation of the Ministry of
Education under grant no. 20090101110050.

C. Li et al. / Journal of Alloys and Compounds 509 (2011) 590–595
595
References
[16] Q.A. Zhang, E. Akiba, J. Alloys Compd. 460 (2008) 272–275.
[17] X.F. Liu, K. Asano, K. Sakaki, Y. Nakamura, H. Enoki, E. Akiba, J. Phys. Chem. C
112 (2008) 17423–17426.
[18] O.M. Lovvik, Phys. Rev. B 71 (2005) 144111.
[19] C. Weidenthaler, T.J. Frankcombe, M. Felderhoff, Inorg. Chem. 45 (2006)
3849–3851.
[20] C. Wolverton, V. Ozolins, Phys. Rev. B 75 (2007) 064101.
[21] T. Sato, M.H. Sorby, K. Ikeda, S. Sato, B.C. Hauback, S. Orimo, J. Alloys Compd.
487 (2009) 472–478.
[22] J.R. Fernandez, A.F. Aguey-Zinsou, M. Elsaesser, X.Z. Ma, M. Dornheim, T.
Klassen, R. Bormann, Int. J. Hydrogen Energy 32 (2007) 1033–1040.
[23] H.E. Kissenger, Anal. Chem. 29 (1957) 1702–1706.
[24] M. Monteferrante, S. Bonella, S. Meloni, E. Vanden-Eijnden, G. Ciccotti, Sci.
Model. Simul. 15 (2008) 187–206.
[25] D.R. Lide, CRC Handbook of Chemistry and Physics, 89th ed., CRC Press/Taylor
and Francis, Boca Raton (internet version 2009).
[26] T. Yamashita, P. Hayes, Appl. Surf. Sci. 254 (2008) 2441–2449.
[27] J.J. Vajo, F. Mertens, C.C. Ahn, R.C. Bowman, B. Fultz, J. Phys. Chem. B 108 (2004)
13977–13983.
[28] X.L. Fan, X.Z. Xiao, L.X. Chen, K.R. Yu, Z. Wu, S.Q. Li, Q.D. Wang, Chem. Commun.
44 (2009) 6857–6859.
[29] L.C. Yin, P. Wang, X.D. Kang, C.H. Sun, H.M. Chen, Phys. Chem. Chem. Phys. 9
(2007) 1499–1502.
[30] D. Blanchard, H.W. Brinks, B.C. Hauback, J. Alloys Compd. 416 (2006) 72–79.
[31] A. Peles, C.G. Van de, Walle, Phys. Rev. B 76 (2007) 214101.
[32] H. Gunaydin, K.N. Houk, V. Ozolinsˇ, Proc. Natl. Acad. Sci. U.S.A. 105 (2008)
3673–3677.
[1] B. Bogdanovic, M. Schwickardi, J. Alloys Compd. 253–254 (1997) 1–9.
[2] X.Z. Ke, I. Tanaka, Phys. Rev. B 71 (2005) 024117.
[3] B. Bogdanovic, M. Felderhoff, A. Pommerin, F. Schüth, N. Spielkamp, Adv. Mater.
18 (2006) 1198–1201.
[4] C.P. Blade, B.P.C. Herijgers, J.H. Bitter, K.P. de Jong, J. Am. Chem. Soc. 130 (2008)
6761–6765.
[5] T. Sun, B. Zhou, H. Wang, M. Zhu, Int. J. Hydrogen Energy 33 (2008) 2260–2267.
[6] M. Resan, M.D. Hampton, J.K. Lomness, D.K. Slattery, Int. J. Hydrogen Energy 30
(2005) 1413–1416.
[7] F.H. Wang, Y.F. Liu, M.X. Gao, K. Luo, H.G. Pan, Q.D. Wang, J. Phys. Chem. C 113
(2009) 7978–7984.
[8] J.R. Ares, K.F. Aguey-Zinsou, F. Leardini, I.J. Ferrer, J.F. Fernandez, Z.X. Guo, C.
Sánchez, J. Phys. Chem. C 113 (2009) 6845–6851.
[9] M. Mamatha, C. Weidenthaler, A. Pommerin, M. Felderhoff, F. Schüth, J. Alloys
Compd. 416 (2006) 303–314.
[10] W. Schwab, K. Wintersberger, Z. Fur Naturforsch. Sect. B: J. Chem. Sci. 8 (1953)
690–691.
[11] M. Fichtner, C. Frommen, O. Fuhr, Inorg. Chem. 44 (2005) 3479–3484.
[12] A. Mamatha, B. Bogdanovic, M. Felderhoff, A. Pommerin, W. Schmidt, F. Schüth,
J. Alloys Compd. 407 (2006) 78–86.
[13] K. Komiya, N. Morisaku, Y. Shinzato, K. Ikeda, S. Orimo, Y. Ohki, et al., J. Alloys
Compd. 446 (2007) 237–241.
[14] H. Kabbour, C.C. Ahn, S.J. Hwang, R.C. Bowman Jr., J. Graetz, J. Alloys Compd.
446 (2007) 264–266.
[15] V. Iosub, T. Matsunaga, K. Tange, M. Ishikiriyama, Int. J. Hydrogen Energy 34
(2009) 906–912.