How does TiF4 affect the decomposition of MgH2 and its complex variants?-An XPS investigation

Article abstract of DOI:10.1039/c7ta03081a

Magnesium hydride and its complex variants, i.e. Mg(AlH₄)₂ and Mg(BH₄)₂ are known for their high hydrogen capacity. The theoretical capacities of these three are 7.6 wt%, 9.3 wt% and 14.4 wt%, respectively. In spite of having very attractive H₂ capacities, their high operating temperature, i.e. more than 300 °C and sluggish kinetics are big issues to be solved before these can be realized for a practical hydrogen storage system. There are several efforts devoted to reduce the working temperature and enhance the sorption kinetics using various additives. Ti-based additives have always been interesting contenders in enhancing the kinetics as well as altering thermodynamics thus reducing the working temperature for various hydrides. Recently, TiF₄ has shown superior catalytic activity on the decomposition of Mg(AlH₄)₂. This motivates us to study its effect on the decomposition of all the above-mentioned Mg-based hydrides. It is observed that the addition of 10 wt% TiF₄ to the above materials greatly influences the decomposition temperature. The decomposition temperature for all three reactions is shifted to the lower temperature side. The oxidation state of the catalyst surface has been investigated using the XPS technique and then the detailed mechanism associated with this improvement is proposed herein.

Full text of DOI:10.1039/c7ta03081a

View Article Online
View Journal
Journal of
Materials Chemistry A
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: A. Jain, S.
Agarwal, S. Kumar, S. Yamaguchi, H. Miyaoka, Y. Kojima and T. Ichikawa, J. Mater. Chem. A, 2017, DOI:
10.1039/C7TA03081A.
This is an Accepted Manuscript, which has been through the
Volume
4 Number 1 7 January 2016 Pages 1–330
Royal Society of Chemistry peer review process and has been
accepted for publication.
Journal of
Materials Chemistry A
Accepted Manuscripts are published online shortly after
Materials for energy and sustainability
www.rsc.org/MaterialsA
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 2050-7488
PAPER
Kun Chang, Zhaorong Chang et al.
Bubble-template-assisted synthesis of hollow fullerene-like
MoS2 nanocages as a lithium ion battery anode material
rsc.li/materials-a

Page 1 of 9
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C7TA03081A
Graphical Abstract
The addition of 10wt% TiF4 to magnesium hydride and its complex variants greatly influence the decomposition temperature.
The detailed mechanism associated with this improvement is proposed herein.

Journal of Materials Chemistry A
Page 2 of 9
View Article Online
DOI: 10.1039/C7TA03081A
Journal Name
RSCPublishing
ARTICLE
How does TiF4 affect the decomposition of MgH2
and its complex variants? – An XPS investigation
Cite this: DOI: 10.1039/x0xx00000x
Ankur Jaina*, Shivani Agarwalb, Sanjay Kumara,c, Shotaro Yamaguchid, Hiroki
Miyaokaa, Yoshitsugu Kojimaa, Takayuki Ichikawaa,e#
Received 00th January 2016,
Accepted 00th January 2016
Magnesium hydride and their complex variant i.e. Mg(AlH4)2 and Mg(BH4)2 are known for
their high hydrogen capacity. The theoretical capacities of these three are 7.6wt%, 9.3wt% and
14.4wt%, respectively. In spite of having very attractive H2 capacities, their high operating
DOI: 10.1039/x0xx00000x
temperature i.e. more than 300 °C and sluggish kinetics are big issues to be solved before these
www.rsc.org/
can be realized for a practical hydrogen storage system. There are several efforts devoted to
reduce the working temperature and enhance the sorption kinetics using various additives. Ti
based additives have always been interesting contenders to enhance the kinetics as well as
altering thermodynamics thus reducing the working temperature for various hydrides. Recently
TiF4 has shown the superior catalytic activity on the decomposition of Mg(AlH4)2. This
motivates us to study its effect on the decomposition of all the above mentioned Mg based
hydrides. It is observed that the addition of 10wt% TiF4 to the above materials greatly
influence the decomposition temperature. The decomposition temperature for all the three
reactions is shifted to the lower temperature side. The oxidation state of catalyst surface has
been observed using XPS technique and then the detailed mechanism associated with this
improvement is proposed herein.
Tel&Fax: +81ꢀ(0)82ꢀ424ꢀ5744. Eꢀmail: *ankur.j.ankur@gmail.com (A. Jain);
#tichi@hiroshimaꢀu.ac.jp (T.Ichikawa)
Introduction
MgH2 can store 7.6wt% hydrogen while its complex variants
hold 9.3 and 14.9 wt% H2, respectively. They undergo through
decomposition by the following reactions:
Hydrogen has been considered as one of the best future energy
carrier, as it has almost 3 times higher gravimetric energy
density than the presently available petroleum based fuels [1].
Its efficient storage still possesses serious issues which need to
be solved before establishing its selfꢀdriven infrastructure. In
last few decades, several efforts have been carried out in search
of suitable material that can store hydrogen safely and
efficiently [2ꢀ5]. As a result, several materials e.g. LaNi5,
MmNi5, TiꢀCrꢀV, MgH2, LiH, alanates, borohydrides, amides,
and silanides [6ꢀ17] have been explored and found suitable for
various practical stationary / moving applications. However,
thermodynamic or kinetic issues associated with them need to
be solved before making them ideal to store hydrogen.
Magnesium has been claimed as the most promising material to
store hydrogen either as binary hydride (MgH2) [4, 18] or its
complex hydrides i.e. Mg(AlH4)2 [19ꢀ24] and Mg(BH4)2 [19,
25ꢀ35].
MgH2
→
Mg + H2
(1)
(2)
(3)
Mg(AlH4)2→MgH2 + 2Al + 3H2→½Al3Mg2 + ½Al + H2
Mg(BH4)2 → MgH2 + 2B + 3H2 → MgB2 + H2
Although these all possess very attractive hydrogen capacity,
their high working temperature hinders their use for practical
applications. The requirement of high temperature is due to
both the thermodynamic and kinetic reasons. Several catalysts
have been adopted to enhance the kinetics of magnesium
hydride including transition metals [36], Nb2O5 [37], ZrCl4 [38],
hydrogen storage alloys such as ZrCrNi [39] etc. Following the
report by Bogdanovic et al. [12], Ti based additives especially
TiCl3 emerged as an efficient catalyst for the kinetic
enhancement as well as thermodynamic alteration of metal
hydrides as well as complex hydrides. The formation of AlxTiy
on the surface of alanate improve the dissociation of hydrogen
thus support kinetic enhancement [40], while the other studies
suggest the substitution of Ti for Na/Li or Al atoms in alanate
unit resulting in the weakening of AlꢀH bind thus support
thermodynamic alteration [41]. Recent studies show the
superiority of Fꢀ anion over Clꢀ in Ti based catalysts [42ꢀ45]. It
is observed that Fꢀ anion actively participates in the
dehydrogenation process by forming active MꢀF (M=Na, Mg,
Ca etc.) in contrast to its analog Clꢀ in TiCl3, where it forms
aNatural Science Centre for Basic Research and Development, Hiroshima
University, HigashiꢀHiroshima 739ꢀ8530, Japan.
bJaipur Engineering College and Research Centre, Jaipur 302022, India
cMaterials Processing and Corrosion Engineering Division, BARC Mumbaiꢀ
400085, India
dGraduate School of Advanced Science of Matter, Hiroshima University,
Japan
eGraduate School of Engineering, Hiroshima University, 1ꢀ4ꢀ1 Kagamiyama,
HigashiꢀHiroshima 739ꢀ8527, Japan
This journal is © The Royal Society of Chemistry 2016
J. Name., 2016, 00, 1-7 | 1

Page 3 of 9
ARTICLE
Journal of Materials Chemistry A
View Article Online
Journal Name
DOI: 10.1039/C7TA03081A
inactive MꢀCl compounds. It is elucidated that TiF3 doped with heating rate if other conditions remains same and this can be
MgH2 perform 10 times better than TiCl3 doped MgH2 [42, 45]. used to calculate activation energy of reaction by plotting a curve
The same behaviour holds true when it is doped with alanate between ln K and 1/RTp using following equation,
e.g. NaAlH4, Na3AlH6, Na2LiAlH6, LiBH4, Ca(BH4)2 etc. [43,
ꢃꢄꢅꢆ
ꢊ ln ꢃꢇꢋꢌ
(4)
44]. More recently it is observed that TiF4 enhances the
ln ꢀ ꢁ ꢂ
ꢇꢈꢉ
ꢄꢅꢆ
reaction kinetics even more in comparison to TiF3 [20, 21, 46ꢀ
49]. However, the mechanism of this improvement is still
unclear and under active debate. This motivates us to study the
reaction mechanism of TiF4 doped MgH2 and its complex
variants Mg(AlH4)2 and Mg(BH4)2. In the present study, TiF4
has been added as catalyst to MgH2, Mg(AlH4)2 and Mg(BH4)2
using mechanical milling. The reaction mechanism has been
established by structural investigation using Xꢀray diffraction
(XRD), Thermogravimetry (TG) coupled with Mass
spectroscopy (MS) and decomposition kinetics in terms of
activation energy using Kissinger equation. Xꢀray
photoelectron spectroscopy (XPS) technique is employed to see
the surface interaction of catalyst with the material and track
the changes of chemical state of catalyst during milling and
decomposition process.
2
where K = β/Tp ; β is heating rate, Tp is peak temperature, R is gas
constant and Edes is activation energy of desorption.
To confirm AlꢀH and BꢀH bonding in the samples, Fourier
Transform Infrared Spectroscopy (FTIR) was performed using
Perkin Elmer Spectrum One unit. KBr was used as diluting agent
mixed in 1:10 ratio and the IR scan was performed in diffuse
reflection cell at a resolution of 1 cmꢀ1. The surface state of catalyst
as well as the material, XPS technique was employed using Thermo
Fisher Scientific ESCALAB 250Xi unit with AlꢀKα (1486.6 eV) Xꢀ
ray source. The powder sample was spread on carbon tape on the
sample holder for the data recording.
Results and discussion
The addition of TiF4 significantly improves the
decomposition of MgH2 and its complex variants as can be seen
from TG/MS results shown in figure 1. As received MgH2
desorbs 7.2 wt% H2 with MS peak temperature at 420°C. The
decomposition temperature is found to be shifted towards lower
Experimental
Commercially available MgH2 (95%, Wako pure chemical
Industry Ltd), Mg(BH4)2 and TiF4 (>95%, Sigma Aldrich) were used
as starting material. Mg(AlH4)2 was prepared by 2h milling of
LiAlH4 and MgCl2 according to following reaction:
side (335
addition of TiF4, the temperature is greatly reduced to 210
with onset temperature of 150 C. This shows dramatic
°C) after 2h milling under Ar atmosphere. With the
°
C
°
2LiAlH4 + MgCl2 → Mg(AlH4)2 + 2LiCl
improvement in the desorption temperature of MgH2. Figure 1b
shows TG/MS profile of as prepared and TiF4 added
Mg(AlH4)2. The weight loss during heating is normalised to
Mg(AlH4)2 only as LiCl presented in the sample behave as dead
mass and doesn’t play any role or affect decomposition
mechanism. It is clear from the spectra that as prepared sample
desorbs 9wt% H2 in a twoꢀstep process with peak temperatures
The milling was performed under 0.1MPa Ar using Fritsch P7 ball
milling apparatus with 1h milling and 30 min rest pattern with 400
rpm milling speed. The mixture of Mg(AlH4)2 and LiCl was used for
all the studies as LiCl does not affect the reaction mechanism at all
as reported earlier [22]. The catalyst TiF4 was added in all the
samples by milling in 0.1MPa Ar atmosphere. The milling time was
kept 2 h for MgH2 while 30 min for Mg(AlH4)2 and Mg(BH4)2
samples. The samples were handled without exposing them to air /
moisture throughout the experiments reported in this manuscript.
High purity Ar (>99.999%) filled glove box (Miwa MFG, MPꢀ
P60W) was used for the sample preparation for milling, XRD, IR,
and XPS measurements. Structural experiment was performed by
XRD on polimide sheet covered samples using Rigaku RINTꢀ2000 40
at 158 °C and 280 °C. These two steps correspond to 6.6wt%
and 2.4 wt% hydrogen as can be observed from TG curve. As
described in equation (1), these are in agreement with the
literature [23]. Addition of TiF4 enables Mg(AlH4)2 to be
decomposed immediately with the temperature rise right from
°
C. It is noteworthy here that the decomposition peak at
instrument. The thermal decomposition behavior was tracked using 100°C has a long tail in contrast to two clear peaks in
Thermogravimetry (Rigaku TG 8120) coupled with TDMS (Anelva, Mg(AlH4)2, which suggests continuous transformation of
MꢀQA200TS) system placed in Ar filled glove box to avoid
Mg(AlH4)2 to the final desorbed products with the elimination
contamination even before / after TG experiment. The scan was of intermediate step. It is quite obvious as Mg(AlH4)2 first
performed under 0.1 MPa Ar flow at different heating rates in order decomposes into MgH2 but TiF4 added MgH2 starts
to calculate activation energy using Kissinger’s equation [50]. decomposing at around 150°C only as seen from figure 1a.
According to Kissinger, the peak positions in thermal analyses vary
Fig. 1 TG (upper) / H2ꢀMS (lower) profile of (a) MgH2; (b) Mg(AlH4)2; and (c) Mg(BH4)2 with and without TiF4 addition.
2 | J. Name., 2016, 00, 1-5
This journal is © The Royal Society of Chemistry 2016

Journal of Materials Chemistry A
Page 4 of 9
View Article Online
Journal Name
DOI: 10.1039/C7TAA03R0T8I1CALE
Fig.
2
Kissinger plot for (a) MgH2; (b) Mg(AlH4)2; and (c) Mg(BH4)2 with and without TiF4 addition.
Thus formed MgH2 immediately decomposes into Mg and H2 of earlier report where it was believed that LiCl doesn’t have
in the presence of added TiF4. These processes are overlapped any effect on the decomposition of Mg(AlH4)2. The activation
with each other thus eliminate two peaks and form a sharp peak energy of TiF4 doped Mg(AlH4)2 is found to be comparable as
corresponding to first step followed by long tail. The amount of reported by Pang et al. [20]. Dramatic change in activation
desorbed hydrogen is found to be 6wt%, which is slightly less energy of Mg(BH4)2 is also observed which is lowered down to
than the theoretical value of 10 wt% TiF4 added sample i.e. 140 kJ molꢀ1 by TiF4 addition in compared to 274 kJ molꢀ1 for
8.4wt%. This must be due to the loss of hydrogen content pure Mg(BH4)2. It is to be noted here that the above activation
during milling process, which will be confirmed and discussed energies for Mg(AlH4)2 and Mg(BH4)2 are only for first step. It
at a later stage in this manuscript. The results are in good is not possible to calculate the activation energy of other steps
agreement with those reported by Pang et al. [20]. The TG/MS as the peaks are not well separated. Moreover, the steps are
profile of Mg(BH4)2 with and without catalyst is shown in controlled by their previous decomposition steps, so it is not
fig.1c. TG profile shows a total weight loss of 14.6 wt%, which reasonable to be calculated using Kissinger equation for the
occurred in three major steps as shown from corresponding H2ꢀ higher peaks.
MS profile. The onset temperature for weight loss is observed
at around 160 C, which slowly continues until the first main
desorption peak at 300
C followed by 2nd and 3rd peaks at 350
C and 370
C, respectively. After completion of 3rd step at
400°C, the weight loss continues to further reduce slowly up to
500 C making total weight loss of 14.6 wt%, which is very
near to theoretical value of Mg(BH4)2. The onset temperature
160 C in this work must be during the phase transformation of
αꢀMg(BH4)2 to βꢀMg(BH4)2 as reported by Hanada et al. [25],
although the temperature is slightly different i.e. 183 C in this
report. Addition of TiF4 ignites the phase transformation at
lower temperature i.e. 110 C and the weight loss started to
To track the decomposition pathway of all the samples, XRD
measurements were performed at different steps as observed
from TG/MS experiment. Figure 3a shows XRD profile of
MgH2ꢀ10%TiF4 before and after H2 desorption. Asꢀmilled
MgH2ꢀ10%TiF4 sample shows the existence of MgH2 phase
only. The profile does not show any peaks corresponding to
TiF4 due to small amount (~2 mol %) / overlapping of the peaks
with the main MgH2 phase. However, the XRD profile of
desorbed samples shows the existence of MgF2 phase as
observed from the presence of small characteristic peaks at
around 27.1° and 41°. This suggests that TiF4 reacts with MgH2
to form MgF2 and release hydrogen in addition to the possible
formation of TiH2, even though we do not get any traces of
TiH2 in XRD profile. Thus possible reaction would be as
follows:
°
°
°
°
°
°
°
°
occur immediately after this phase transformation. Interestingly,
the main desorption occurred in two steps in contrast to 3 steps
for pristine Mg(BH4)2. It seems that first two peaks are merged
MgH2+0.02TiF4
→
0.96Mg
+
0.04MgF2
+
and shifted to lower temperature to shape first peak at 285
°C
0.02TiH2+0.098H2(4)
The formation of MgF2 and TiH2 is in agreement with the
report by Xie et al. [42] on TiF3 doped MgH2 system.
followed by 2nd peak at 358
°
C. The total weight loss is
observed as 13.0 wt%, which is near to the theoretical value i.e.
13.1 wt% for Mg(BH4)2ꢀ10% TiF4.
Figure 3b shows XRD profiles of as prepared Mg(AlH4)2
and the samples heated at different temperatures. It is observed
that 2h milling is enough to transform LiAlH4 and MgCl2 into
Mg(AlH4)2 and LiCl mixture. Heating up to 200 °C
decomposes Mg(AlH4)2 into MgH2 and Al. Further heating up
to 400 °C allows MgH2 to react with Al and turned into H2 and
intermetallic phases i.e. Mg37Al61 and Mg17Al12 in present study.
These MgꢀAl intermetallic phases are slightly different from
that reported in previous literatures [22, 23].
The XRD profile of TiF4 added Mg(AlH4)2 shown in figure
3c suggests slight decrease in the peak intensity corresponding
to Mg(AlH4)2 phase, which further reduced with heating up to
145°C and completely vanished at 185°C. This must be due to
decomposition of Mg(AlH4)2 during milling with TiF4 as this
additive is significantly effective for the decomposing reaction.
This approves the TG result shown in fig. 1b, where the total
weight loss is observed significantly less than the theoretical
capacity. No peaks corresponding to TiF4 are visible in the
In order to quantify this improvement in desorption
behaviour, Kissinger equation is employed to calculate the
activation energy. To calculate this, TG/MS experiments were
performed at different heating rate of 1, 2, 5 and 10°C/min and
the peak temperature is noted from MS signal as shown in the
inset of figure 2(a). Figure 2(a) – (c) shows the Kissinger plot
for MgH2, Mg(AlH4)2, and Mg(BH4)2 with and without catalyst.
The activation energy for as received MgH2 is found to be 170
kJ molꢀ1, which reduces to 137 kJ molꢀ1 for 2h milled MgH2
without catalyst. This value is further reduced to 71 kJ molꢀ1 for
TiF4 added MgH2, which is much lower than those reported for
TiF3 doped MgH2 [42]. Similarly the addition of TiF4 lowers
the activation energy down to 84 kJ molꢀ1 from 95kJ molꢀ1 for
as prepared Mg(AlH4)2. The activation energy of as prepared
Mg(AlH4)2 is found to be slightly less than that reported
previously by Pang et al [20], which may be caused by the
presence of LiCl as reported by Xiao et al [24]. It is in contrast
This journal is © The Royal Society of Chemistry 2016
J. Name., 2016, 00, 1-7 | 3

Page 5 of 9
ARTICLE
Journal of Materials Chemistry A
View Article Online
Journal Name
DOI: 10.1039/C7TA03081A
profile similar to MgH2 case. It is noteworthy here that we do decomposition
not observe any peaks corresponding to MgH2 during
Fig. 3 XRD profile of (a) MgH2ꢀ10%TiF4, (b) Mg(AlH4)2, (c) Mg(AlH4)2ꢀ10%TiF4, (d) Mg(BH4)2, (e) Mg(BH4)2ꢀ10%TiF4; and (f)
FTIR spectra of Mg(BH4)2 (Black Solid Line) and Mg(BH4)2–10% TiF4 (Red dotted line) at different decomposition steps;
Different phases are indexed by symbols (shown in brackets) as follows: MgH2 no. 01ꢀ076ꢀ2727 (°), Mg no. 01ꢀ073ꢀ3812 (*);
MgF2 no. 00ꢀ041ꢀ1443 (∆); Mg(AlH4)2 (♦); LiCl no. 01ꢀ071ꢀ4664 (
Mg17Al12 no. 01ꢀ073ꢀ1148 ( ); MgB2 no. 01ꢀ075ꢀ3263 (|).
ꢀ
); Al no. 01ꢀ071ꢀ4622 (+); Mg37Al61 no. 01ꢀ076ꢀ3843 (∇);
↓
in contrast to as prepared Mg(AlH4)2, however well crystalline confirmed by surface analysis by XPS at a later stage in this
Al phase could easily be identified. This suggests that the main manuscript.
reaction in TiF4ꢀadded Mg(AlH4)2 follows similar path to
Figure 3d and 3e shows the XRD profile of Mg(BH4)2 with
Mg(AlH4)2 but MgH2 starts decomposing immediately after its and without TiF4, whereas FTIR of Mg(BH4)2 is shown in
formation and forms an amorphous MgꢀAl compound. Further figure 3f. As purchased Mg(BH4)2 has some impurities of
heating to high temperature i.e. 400°C crystallizes the solvents used during the preparation, so heat treatment was
intermetallic compounds Mg37Al61 and Mg17Al12. In addition, performed at 150°C for 24 h under 0.1MPa H2. The XRD
peaks corresponding to MgF2 are also visible in the sample profile of annealed sample confirms the presence of αꢀ
Mg(BH4)2 (bottom stack of fig. 3d), which is converted to βꢀ
heated up to 400 °C, which suggests the reduction of TiF4 into
Ti metallic phase or the possible formation of intermetallic Mg(BH4)2 phase by heating up to 225
°
C during TG/MS
phase with aluminum as suggested earlier in several reports on experiment with a total weight loss of 2.2 wt%. The presence of
alanates doped with Ti based catalysts [40, 46]. This will be IR band at around 2300 cmꢀ1 confirms the presence of BH4 unit
4 | J. Name., 2016, 00, 1-5
This journal is © The Royal Society of Chemistry 2016

Journal of Materials Chemistry A
Page 6 of 9
View Article Online
Journal Name
DOI: 10.1039/C7TAA03R0T8I1CALE
– characteristic peak for Mg(BH4)2. Further heating up to 325 which suggest a strong electrochemical interaction of catalyst
with MgH2 and forming Ti2+ chemical states of titanium. The
°
C, Mg(BH4)2 decomposes to some amorphous phase with a
peaks corresponding to Ti are at very similar position as
mentioned in recent work by Jangir et al [49]. However, the
above work didn’t consider the effect of Fꢀ on the improvement
of sorption properties of MgH2. To investigate the effect of F
anion, the F1s spectra (fig. 4b) has also been analysed where a
marginal peak shift of F1s electron (685.42 eV to 685.88
total weight loss of 6.9 wt%, thus generating first major MS
peak. The IR spectra of sample heated up to 325 C shows a
°
broad peak in 2000ꢀ2600 cmꢀ1 region, thus suggesting the
presence of BꢀH bonding in the sample. This confirms the
presence of higher boranate with less content of hydrogen i.e.
MgB12H12. It is in agreement with the earlier reported
dehydriding reaction of Mg(BH4)2 [26]. Further heating up to
360°C and 440°C gradually decompose MgB12H12 into MgH2
and finally Mg and B. Thus formed Mg and B react together,
start forming MgB2 at 360°C slowly as indicated from the
characteristic peak at 42.3° and converted completely
crystalline MgB2 at 500°C. TiF4 addition accelerate the
formation of MgB2 at temperature as low as 335 °C. Moreover,
we do not observe the XRD peaks corresponding to MgH2,
which must be due to instant decomposition with its generation.
The peaks corresponding to Mg are also very weak intense,
suggesting high rate of MgB2 formation. Moreover, IR profile
of sample heated up to 335°C suggest no formation of higher
boranate or its very fast decomposition in TiF4 added sample.
This promoted formation of MgB2 and elimination of
MgB12H12 formation may facilitate easy rehydrogenation of this
complex system, as the formation of MgB12H12 has been
reported as limiting step during hydrogenation process [26, 27].
Although the formation of MgF2 is likely to occur during
decomposition process, unfortunately, we do not observe any
peaks corresponding to MgF2 in contrast to MgH2 and
Mg(AlH4)2 systems.
To confirm the above mechanism in more details, the changes
in chemical states of the catalyst (TiF4) in the milled as well as
in desorbed MgH2, Mg(BH4)2 and Mg(AlH4)2 samples have
been studied using XPS. The XPS of TiF4 under similar ball
milling condition has also been conducted to compare the
results. Fig. 4 shows Ti2p and F1s spectra for TiF4 added MgH2,
where the binding energy peaks of Ti 2p1/2 and 2p3/2 electron in
4+ valence state are at 465.45 and 459.74 eV, respectively. In
the case of TiF4ꢀadded MgH2 system, a peak shift towards
lower energy (464.16 and 458.51) is observed in both the peaks
with the existence of new peak at 453.52 eV, which clearly
indicates the change in the chemical state of titanium from Ti4+
to Ti3+/Ti2+ states. The peak fitting is shown in fig. S1
(supplementary data), which shows the Ti2+/Ti3+ ratio to be
increased from 0.5 for as milled sample to 2 for desorbed
sample, which confirms the proposed reaction shown in
equation 4. The peak at 464.16 eV further reduced its intensity
with a developed broad peak at 453.52 for desorbed sample,
Fig. 4 XPS spectra for (a) Ti2p bands and (b) F1s bands of
MgH2 sample milled with TiF4; the spectra of pure TiF4 is also
shown at the lower part of figure for the comparison.
eV) for ball milled and desorbed samples were observed as
compared to BE peak position of F1s electron in TiF4. In
addition, the peak corresponding to F1s for desorbed sample is
found to be broadened in comparison to as milled sample or
pure TiF4. This broadening suggests the existence of different
Fꢀ containing species i.e. MgF2 and one mixed TiꢀMgꢀF phase
as suggested by Ma et al. [45]. This TiꢀMgꢀF phase acts as
catalytic active site and enhances dissociation / recombination
of hydrogen atom.
Figure 5 shows the XPS spectra for Al2p, Ti2p and F1s
bands for Mg(AlH4)2 samples. The Al2p spectra of as prepared
Mg(AlH4)2 has shown a single peak at 73.66 eV, which can be
ꢀ
assigned to Al3+ state of aluminum in AlH4 complex unit. The
o
XPS analysis of the sample dehydrogenated up to 200 C has
shown an additional peak at 70.87 eV, which must be
originated from Al0 state due to partial decomposition of
alanate into magnesium hydride and metallic aluminum.
o
Further heating to 400 C decompose alanate completely, but
the
presence
of
peak
This journal is © The Royal Society of Chemistry 2016
J. Name., 2016, 00, 1-7 | 5

Page 7 of 9
ARTICLE
Journal of Materials Chemistry A
View Article Online
Journal Name
DOI: 10.1039/C7TA03081A
Fig. 5 XPS spectra for (a) Al2p bands; (b) Ti2p bands; and (c) F1s bands of Mg(AlH4)2 sample milled with TiF4; the spectra of
pure TiF4 is also shown at the lower part of figure (b) and (c) for the comparison.
Fig. 6 XPS spectra for (a) B1s bands; (b) Ti2p bands; and (c) F1s bands of Mg(BH4)2 sample milled with TiF4; the spectra of
pure TiF4 is also shown at the lower part of figure (b) and (c) for the comparison.
at 73.66 eV (Al3+) must be due to partial oxidation of Al element has the ability to be diffused away from fluorine (F)
surface, thus forming Al2O3. A slight shift from 70.87 eV to leaving Fꢀ sphere more positive. Consequently, F1s electron
70.64 eV at this temperature is due to the formation of AlꢀMg required more energy (comparatively) to be knocked out its 1s
intermetallic. Since Mg is more electropositive, so in the case electron. Thus the binding energy of F1s electron will be in the
of MgꢀAl alloy, the electron density will be shifted towards Al. order of MgꢀF > TiꢀF > AlꢀF. Since the binding energy of
Consequently, the binding energy of Al2p of MgꢀAl alloy will
F1s for as milled TiF4 sample is shifted towards lower energy,
be lower than metallic Al. The intermetallic phase has also been thus suggesting that Fꢀ anion should be bonded with Al by
ꢀ
confirmed by XRD analysis. The Al2p spectra of TiF4 added substituting hydronium ion of AlH4 complex and consequently
Mg(AlH4)2 had shown two peaks after ball milling. The XPS magnesium alanates have modified to Mg(AlH4ꢀδFδ)2. The
result suggests that the decomposition of Mg(AlH4)2 into modified alanate (Mg(AlH4ꢀδFδ)2) which could have low
metallic aluminum has been occurred even during ball milling. activation barrier for dissociation and required less energy to
The result is further confirmed by the thermogravimetric (TG) remove H atom as suggested by Yin et al. [27] for Na3AlH6.
analysis (fig. 1b), which clearly has shown the substantial less After decomposition at 400°C, Fꢀ anion react with Mg to form
hydrogen released than the theoretical expected value.The Ti2p MgF2, which drive the XPS peak towards higher binding
spectra (fig. 5b) suggest the reduction of Ti4+ state to Ti1+ and energy. Thus collectively on the basis of above results,
Ti0 as observed from the substantial shift of doublet peaks to following two reasons should be responsible together for the
458.10 eV and 452.62 eV for as milled sample in compared to improvement in Mg(AlH4)2 dissociation: (1) formation of TiꢀAl
465.45 eV and 459.74 eV of pure TiF4. The peaks are quite clusters which enhance the recombination of H2 and (2)
broad and comprise of different peaks corresponding to formation of Mg(AlH4ꢀδFδ)2 by Fꢀ anion substitution which
different states of Ti as evidenced by peak fitting shown in fig. requires less energy to dissociate.
S2 (supplementary data). The molar ratio of Ti0 and Ti1+
The XPS analysis of B1s bands for Mg(BH4)2 heated up to
containing phase is found to be increased with the heating as different temperatures is presented in Fig. 6a. The boron is
ꢀ
evidenced by the peak area as well as peak intensity covalently bonded with hydrogen in BH4 , therefore the peak of
corresponding to Ti0 state. This suggests the interaction the 1s electron is substantially shifted to higher energy
between metallic Ti and metallic Al during milling and forms compared to elemental boron. The broadened peak of 1s
some TiꢀAl clusters. These TiꢀAl clusters must facilitate fast electron attributed to the covalent bonding in which the
dissociation and recombination of hydrogen on the surface of electrons are in diffused state. The B1s peak gradually shifted
material, thus reducing the dissociation temperature of to lower energy side when Mg(BH4)2 was heated up to 200, 325,
Mg(AlH4)2 similar to other alanates [40, 46, 51, 52]. Heating to 360 and 500°C. The sample partially decomposed below 325°C,
higher temperature, the peaks continue shifting towards lower thus originating only a small peak at 187.8 eV, which
energy marginally and become relatively sharper which must be corresponds to elemental boron and in good agreement with the
due to the formation of TiꢀAl phase with a wellꢀdefined database. The sample heated up to 325°C shows significantly
electronic band structure. Fig. 5c presents F1s spectra of TiF4 high intense peak corresponds to elemental boron, although the
added Mg(AlH4)2. Interestingly, it is observed that the F1s peak peak at 192.13 eV still remained significantly high which must
for the TiF4 doped Mg(AlH4) is marginally shifted to lower be due to the presence of covalently bonded boron in MgB12H12
energy i.e. 684.82 eV from 685.42 eV for pure TiF4, then it as confirmed from IR spectra shown in fig. 3f. Further heating
started moving towards higher energy i.e. 685.2 eV and 686.1 up to 360 and then 500°C gradually decreases the peak at
eV for the sample heated up to 200 oC and 400 oC, respectively. 192.13 eV and then completely vanishes, which suggest
This can be justified on the basis of Pauling electronegativity of complete decomposition of Mg(BH4)2. A marginal shift in the
Al (1.61), Ti (1.54), and Mg (1.31). The binding energy of peak at 187.8 to 188 eV is observed for sample heated up to
F1s electron will be influenced by the electrochemical binding 500°C, which suggests the formation of magnesium diboride
of these metals. Since the electron of the more electronegative (MgB2) which has been confirmed by XRD (fig. 3d) as well.
6 | J. Name., 2016, 00, 1-5
This journal is © The Royal Society of Chemistry 2016

Journal of Materials Chemistry A
Page 8 of 9
View Article Online
Journal Name
ARTICLE
DOI: 10.1039/C7TA03081A
The upper panel of fig. 6a shows the B1s spectra of TiF4 added desorption kinetics. Whereas, the decomposition of TiF4 added
Mg(BH4)2 just after milling. The spectra seems very similar to Mg(BH4)2 has been found to be accelerated by the formation of
annealed Mg(BH4)2, thus suggesting no interaction of TiF4 with Ti1ꢀxMgx(BH4)2 during initial heating then the formation of
Mg(BH4)2 during milling. This is also confirmed from Ti2p TiB2, TiH2, and TiꢀMgꢀF species at elevated temperature acts as
spectra shown in fig. 6b, which shows very similar doublet at catalyst and provide exit gate for hydrogen. Although Fꢀ anion
465.45 eV and 459.74 eV to that of pure TiF4. The heating of acts as functional anion in contrast to its analogous inactive
this sample up to 200°C shows a substantial shift in both peaks Clꢀ anion and significantly improves the decomposition
towards lower energy (460.1 and 454.0 eV), actually both peaks behaviour of all the three hydrides, the formation of MgF2 at
are quite broad and consist of several peaks corresponding to the end of process is still an issue need to be solved, as it act as
different chemical states of Ti (fig. S3, supplementary data). dead mass and lower the hydrogen content slightly.
This indicates clear reduction of Ti4+ state into mixed Ti2+/Ti1+
state of titanium and possible formation of titanium hydride
Acknowledgements
(TiH2) or titanium boride (TiB2) in addition to substituted
borohydride i.e. Ti1ꢀxMgx(BH4)2. The formation of above
compounds has been reported in earlier report on the
decomposition of TiCl3 added Mg(BH4)2 [28]. Further heating
up to 335°C, sharpen the doublets and positioned at 459.85 and
453.95 eV which must be due to the growth of TiH2 phase.
Heating at higher temperature i.e. 500°C again broadened the
peak at 453.95 eV, while decrease the intensity of peak at
459.85 eV, which should be attributed to the presence of Ti0
state of metallic titanium as a result of TiH2 decomposition in
addition to TiB2, which is quite stable up to elevated
temperature. The formation of boride species during
decomposition of Mg(BH4)2 is quite common for many
transition metal based additives e.g. NiCl2, NiF2, CoCl2, CoF3
etc [29, 30]. In addition to TiB2, MgF2 is also formed as
confirmed from the F1s spectra shown in fig. 6c. It is observed
that there is no significant shift in the peak during heating up to
335°C, however, heating up to 500°C, gives rise a shift of F1s
peak to higher energy i.e. 686.5 eV from 685.42 eV, thus
suggesting the formation of MgF2. At the same time, the peak is
broadened which shows a possibility of two Fꢀ species i.e.
MgF2 and some TiꢀMgꢀF phase similar to TiF4 added MgH2
sample. It is interesting to note that we do not observe
substitution effect of Fꢀ anion in Mg(BH4)2 in contrast to
Mg(AlH4)2, where a clear indication of Mg(AlH4ꢀδFδ)2 was
observed. It is also in contrast to the LiBH4 case where the
formation of modified LiBH4 i.e. LiBH4ꢀxFx was reported [47].
It concludes that Fꢀ anion doesn’t play any significant active
role for decomposition of Mg(BH4)2 at first, however,
accelerate the decomposition of intermediately formed MgH2
phase. Thus, collectively, the decomposition of TiF4 added
Mg(BH4)2 is first governed and accelerated by the formation of
Ti1ꢀxMgx(BH4)2 during initial heating followed by TiB2, TiH2,
and TiꢀMgꢀF species which acts as catalyst and provide exit
gate for hydrogen at later step.
One of the author Sanjay Kumar acknowledge financial
support from JSPS under overseas postdoctoral fellowship
programme (P15078).
Notes and references
1. I.P. Jain, Int. J. Hydrogen Energy, 2009, 34, 7368.
2. L. Schlapbach, A. Züttel, Nature, 2001, 414, 353.
3. A. Züttel, Mater. Today, 2003, 6, 24.
4. B. Sakintuna, F.L.ꢀDarkrim, M. Hirscher, Int. J. Hydrogen Energy,
2007, 32, 1121.
5. P. Chen, Mater. Today, 2008, 11, 36.
6. G. Liang, J. Huot, R. Schulz, J. Alloys Compd., 2001, 320, 133.
7. A. Jain, R.K. Jain, I.P. Jain, J. Power Sources, 2006, 159, 132.
8. H. Itoh, H. Arashima, K. Kubo, T. Kabutomori, K. Ohnishi, J. Alloys
Compd., 2005, 404ꢀ406, 417.
9. S. Agarwal, A. Jain, P. Jain, M. Jangir, I.P. Jain, J. Phys. Chem. C,
2013, 117, 11953.
10. D. Vyas, P. Jain, G. Agarwal, A. Jain, I.P. Jain, Int. J. Hydrogen
Energy, 2012, 37, 16013.
11. A. Jain, E. Kawasako, H. Miyaoka, T. Ma, S. Isobe, T. Ichikawa, Y.
Kojima, J. Phys. Chem. C, 2013, 117, 5650.
12. B. Bogdanović, M. Schwickardi, J. Alloy Compd., 1997, 253ꢀ254, 1.
13. A. Borgschulte, A. Jain, A. J. RamirezꢀCuesta, P. Martelli, A. Remhof,
O. Friedrichs, R. Gremaud, A. Züttel, Faraday Discuss., 2011, 151,
213.
14. P. Chen, Z. Xiong, J. Luo, J. Lin, K.L. Tan, Nature, 2002, 420, 302.
15. T. Ichikawa, N. Hanada, S. Isobe, H. Leng, H. Fujii, , J. Phys. Chem.
B, 2004, 108, 7887.
16. J.ꢀN. Chotard, W.S. Tang, P. Raybaud, R. Janot, Chem. Eur. J., 2011,
17, 12302.
17. A. Jain, T. Ichikawa, S. Yamaguchi, H. Miyaoka, Y. Kojima, Phys.
Chem. Chem. Phys., 2014, 16, 26163.
18. I.P. Jain, C. Lal, A. Jain, Int. J. Hydrogen Energy, 2010, 35, 428.
19. I.P. Jain, P. Jain, A. Jain, J. Alloys Compd., 2010, 503, 303.
20. Y. Pang, Y. Liu, X. Zhang, M. Gao, H. Pan, Int. J. Hydrogen Energy,
2013, 38, 13343.
21. Y. Pang, Q. Li, Scr. Mater., 2017, 130, 223.
22. Y. Kim, E.ꢀK. Lee, J.ꢀH. Shim, Y.W. Cho, K.B. Yoon, J. Alloys
Compd., 2006, 422, 283.
Conclusions
23. M. Fichtner, O. Fuhr, O. Kircher, J. Alloys Compd., 2003, 356ꢀ357,
418.
24. X. Xiao, T. Qin, Y. Jiang, F. Jiang, M. Li, X. Fan, S. Li, H. Ge, Q.
Wang, L. Chen, Progress in Natural Science: Materials International,
2017, 27, 112.
25. N. Hanada, K. Chlopek, C. Frommen, W. Lohstroh, M. Fichtner, J.
Mater. Chem., 2008, 18, 2611.
26. H.ꢀW. Li, K. Kikuchi, Y. Nakamori, N. Ohba, K. Miwa, S. Towata, S.
Orimo, Acta Mater., 2008, 56, 1342.
27. G. Severa, E. Rönnbro, C.M. Jensen, Chem. Comm., 2010, 46, 421.
28. D. Matsumura, T. Ohyama, Y. Okajima, Y. Nishihata, H.ꢀW. Li, S.ꢀI.
Orimo, Mater. Trans., 2011, 52, 635.
29. I. Saldan, S. Hino, T.D. Humpheries, O. Zavorotynska, M. Chong,
C.M. Jensen, S. Deledda, B.C. Hauback, J. Phys. Chem. C, 2014, 118,
23376.
30. O. Zavorotynska, I. Saldan, S. Hino, T.D. Humphries, S. Deledda, B.C.
Hauback, J. Mater. Chem. A, 2015, 3, 6592.
The mechanism of decomposition pathway of MgH2 and its
complex variants i.e. Mg(AlH4)2 and Mg(BH4)2 has been
established. The role of TiF4 addition for the improvement of
decomposition properties have been understood using XRD and
XPS. It is demonstrated that Ti4+ is reduced into lower
oxidation state i.e. Ti3+ / Ti2+ in all the cases. In addition, Fꢀ
anion plays important role to improve the dehydrogenation
properties. A strong chemical interaction between TiF4 and
MgH2 resulted in the formation of TiH2 and TiꢀMgꢀF species
which enhance the sorption kinetics of MgH2 and reduce the
desorption temperature. In case of Mg(AlH4)2, the formation of
TiꢀAl clusters on the surface and formation of Mg(AlH4ꢀδFδ)2
by Fꢀ anion substitution together enhance the recombination of
H2 thus reduce the dissociation temperature and enhance the
This journal is © The Royal Society of Chemistry 2016
J. Name., 2016, 00, 1-7 | 7

Page 9 of 9
ARTICLE
Journal of Materials Chemistry A
VJieowuArrtnicalelONnlainme e
DOI: 10.1039/C7TA03081A
31. Y. Yan, A. Remhof, D. Rentsch, A. Zuttel, Chem. Comm., 2015, 51,
700.
41. M. Fichtner, P. Canton, O. Kircher, A. Léon, J. Alloys Compd., 2005,
404ꢀ406, 732.
32. O. Zavorotynska, S. Deledda, G. Li, M. Matsuo, S.ꢀi. Orimo, B.C.
Hauback, Angew.Chem. Int. Ed., 2015, 54, 10592.
33. O. Zavorotynska, S. Deledda, B.C. Hauback, Int. J. Hydrogen Energy,
2016, 41, 9885.
34. O. Zavorotynska, A. ElꢀKharbachi, S. Deledda, B.C. Hauback, Int. J.
Hydrogen Energy, 2016, 41, 14387.
35. M. Chong, M. Matsuo, S.ꢀi. Orimo, T. Autrey, C.M. Jensen, Inorg.
Chem., 2015, 54, 4120.
36. M. Pozzo, D. Alfè, Int. J. Hydrogen Energy, 2009, 34, 1922.
37. T. Kimura, H. Miyaoka, T. Ichikawa, Y. Kojima, Int. J. Hydrogen
Energy, 2013, 38, 13728.
42. L. Xie, Y. Liu, Y.T. Wang, J. Zheng, X.G. Li, Acta Mater., 2007, 55,
4585.
43. X.ꢀD. Kang, P. Wang, H.ꢀM. Cheng, Scr. Mater., 2007, 56, 361.
44. L.ꢀC. Yin, P. Wang, X.ꢀD. Kang, C.ꢀH. Sun, H.ꢀM. Cheng, Phys.
Chem. Chem. Phys., 2007, 9, 1499.
45. L.ꢀP. Ma, X.ꢀD. Kang, H.ꢀB. Dai, Y. Liang, Z.ꢀZ. Fang, P.ꢀJ. Wang, P.
Wang, H.ꢀM. Cheng, Acta Mater., 2009, 57, 2250.
46. Y. Liu, F. Wang, Y. cao, M. Gao, H. Pan, Q. Wang, Energy Environ.
Sci., 2010, 3, 645.
47. Suwarno, Int. J. Techno., 2011, 2, 28.
48. C.B. Minella, S. Garroni, C. Pistidda, M.D. Baró, O. Gutfleisch, T.
Klassen, M. Dornheim, J. Alloys Compd., 2015, 622, 989.
49. M. Jangir, A. Jain, S. Yamaguchi, T. Ichikawa, C. Lal, I.P. Jain, Int. J.
Hydrogen Energy, 2016, 41, 14178.
50. H.E. Kissinger, Anal. Chem., 1957, 29, 1702.
51. J. Voss, Q. Shi, H.S. Jacobsen, M. Zamponi, K. Lefmann, T. Vegge, J.
Phys. Chem. B, 2007, 111, 3886.
38. S. Kumar, A. Jain, S. Yamaguchi, H. Miyaoka, T. Ichikawa, A.
Mukherjee, G.K. Dey, Y. Kojima, Int. J. Hydrogen Energy, 2017, 42,
6152.
39. S. Agarwal, A. Aurora, A. Jain, I.P. Jain, Amelia Montone, Int. J.
Hydrogen Energy, 2009, 34, 9157.
52. J. Liu, Y. Han, Q. Ge, Chem. – Eur. J., 2009, 15, 1685.
40. B. Bogdanović, U. Eberle, M. Felderhoff, F. Schüth, Scr. Mater.,
2007, 56, 813.
8 | J. Name., 2016, 00, 1-5
This journal is © The Royal Society of Chemistry 2016