Improving hydrogen sorption kinetics of MgH2 by mechanical milling with TiF3

Article abstract of DOI:10.1016/j.jallcom.2006.05.103

MgH₂ + TiF₃ system was mechanically prepared and investigated with regard to hydrogen absorption/desorption performance. It was found that the sorption kinetics of MgH₂ can be markedly improved by mechanical milling with 4 mol% TiF₃, in particular at reduced operation temperatures. Even at a moderate temperature range (313-373 K), the hydrogen absorption can be largely completed within about 25 s. On the desorption aspect, the catalytic enhancement arising upon adding TiF₃ allows the sample to desorb near 5 wt.% hydrogen within 600 s at 573 K. Furthermore, such enhancement in kinetics was observed to persist in the absorption/desorption cycles. Preliminary XRD examination was also performed to understand the observed catalytic effect of TiF₃ additive.

Full text of DOI:10.1016/j.jallcom.2006.05.103

Journal of Alloys and Compounds 432 (2007) L1–L4
Letter
Improving hydrogen sorption kinetics of MgH by
2
mechanical milling with TiF3
∗
Lai-Peng Ma, Ping Wang , Hui-Ming Cheng
Shenyang National Laboratory for Materials Science, Institute of Metal Research, Chinese Academy of Sciences, Shenyang 110016, PR China
Received 24 April 2006; received in revised form 25 May 2006; accepted 25 May 2006
Available online 7 July 2006
Abstract
MgH + TiF
2
3
system was mechanically prepared and investigated with regard to hydrogen absorption/desorption performance. It was found that
can be markedly improved by mechanical milling with 4 mol% TiF , in particular at reduced operation temperatures.
the sorption kinetics of MgH
Even at a moderate temperature range (313–373 K), the hydrogen absorption can be largely completed within about 25 s. On the desorption
aspect, the catalytic enhancement arising upon adding TiF allows the sample to desorb near 5 wt.% hydrogen within 600 s at 573 K. Furthermore,
such enhancement in kinetics was observed to persist in the absorption/desorption cycles. Preliminary XRD examination was also performed to
understand the observed catalytic effect of TiF additive.
2006 Elsevier B.V. All rights reserved.
2
3
3
3
©
Keywords: Magnesium; Hydrogen storage; Kinetics; TiF3
1
. Introduction
mechanical milling MgH2 with a small amount of mesoporous
Nb2O , more than 5 wt.% hydrogen could be restored within
5
As a hydrogen storage medium, MgH2 is highly appreci-
30 s at 423 K. What is more striking, about 4.5 wt.% hydrogen
could be absorbed within 15 s even at room temperature [29].
Quite recently, in our investigation on sodium alanate sys-
tem, we found that TiF3 possessed pronounced catalytic effect
on the reversible de-/hydrogenation of NaAlH4. In compari-
son with the conventional TiCl3, this novel dopant precursor
resulted in further enhanced H-capacity, de-/hydriding kinetics
as well as reduced operation temperature and pressure condi-
tions [30–32]. Interestingly, in the study of Mg-based materials,
ated due to its high hydrogen capacity and low cost. However,
the practical application of MgH2 for onboard hydrogen stor-
age is largely hampered by its high thermodynamic stability
and sluggish sorption kinetics [1]. After decades of extensive
research efforts, partial success has been achieved in improving
the kinetic performance of MgH2 while leaving the problematic
thermodynamics untouched. It was found that the addition of H-
storage alloys, transition metals and/or their oxides, fluorides,
and even some non-metal materials through mechanical milling
might all accelerate the absorption/desorption processes of mag-
nesium [2–26]. However, most of the literature reports focus on
the property improvement at relatively high temperatures (usu-
ally higher than 473 K). Only few reports concern the substantial
kinetics improvement at a moderate operation temperature lower
than 373 K, which is of practical interests for onboard H-storage
application. Liang et al. claimed that the mechanically pre-
pared MgH2 + 5 mol%Ti and MgH2 + 5 mol%V systems exhib-
ited rapid absorption kinetics at temperatures ranging from room
temperature to 373 K [27,28]. Hanada et al. reported that, by
−
it has been demonstrated that both Ti hydride and F contribute
to the catalytic enhancement of the absorption/desorption pro-
cesses of MgH2 [12,24,25,27]. As a source of both species, TiF3
is therefore of great interest for preparing catalytically enhanced
Mg-based H-storage materials.
In the present study, the MgH2 + TiF3 system was mechani-
cally prepared and examined with respect to hydrogen sorption
kinetics. As expected, this system possesses enhanced sorption
kinetics, in particular the absorption performance at a moderate
temperature range.
2
. Experimental
∗
Corresponding author. Tel.: +86 24 2397 1622; fax: +86 24 2389 1320.
The starting material MgH2 was prepared by mechanical milling magne-
E-mail address: pingwang@imr.ac.cn (P. Wang).
sium powder (purity > 99.9%, 300 mesh) under hydrogen atmosphere with an
0
925-8388/$ – see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2006.05.103

L2
L.-P. Ma et al. / Journal of Alloys and Compounds 432 (2007) L1–L4
initial pressure of ∼1 MPa, followed by hydrogenation for 12 h at 623 K. The
process was repeated for three times to achieve a hydrogenation ratio of ∼80%,
as determined by using the volumetric method [26]. The TiF3 powder was pur-
chased from Sigma–Aldrich Chemical Corp. and was used as received. The
MgH2 + 4 mol%TiF3 mixture was mechanically milled under Ar atmosphere
for 10 h by using a Fritsch P7 planetary mill with stainless steel vial and balls
at 400 rpm. The ball-to-powder weight ratio was about 40:1. For comparison,
MgH2 was mechanically milled under identical conditions. All the sample han-
dlingwasperformedinanAr-filledglove-boxequippedwithpurificationsystem,
in which the typical H2O/O2 levels are below 1 ppm.
The samples were characterized by X-ray diffraction (XRD, Rigaku D/max
2
400) with Cu K␣ radiation. Hydrogen storage properties of the samples, with a
typical amount of around 200 mg, were examined by using a carefully calibrated
Sievert-type apparatus. Absorption/desorption measurements were performed
at desired temperatures with an initial hydrogen pressure of 2.0 and 0.01 MPa,
respectively. The hydrogen supply (with an initial purity of 99.999%) was fur-
ther purified by using a hydrogen storage alloy system to minimize H2O/O2
contamination. To allow a practical evaluation of the hydrogen storage capa-
bility of the materials, the weight of the catalyst was taken into account in the
determination of H-capacity.
Fig. 2. Hydrogen absorption curves of MgH2 + 4 mol%TiF3 at moderate tem-
peratures: (a) 373 K, (b) 340 K, and (c) 313 K.
3
. Results and discussion
at moderate temperature. As shown in Fig. 2, even at 313 K,
about 2 wt.% hydrogen could be recharged within 25 s. When
the operation temperature was increased to 373 K, the restored
hydrogen amount within 25 s increased to 3.4 wt.%. Such kinetic
performance is comparable to the highest level of the literature
results [27–29].
Mechanical milling with TiF3 has resulted in marked
improvement on the absorption kinetics of Mg. As seen in Fig. 1,
the sample with 4 mol% TiF3 additive could absorb 5 wt.%
hydrogen within about 30 s at 573 K, with an average absorption
rate 20 times higher than that of pure Mg. Moreover, in great
contrast to the high kinetics-temperature dependence in pure
Mg, the variation of operation temperature from 573 to 423 K
was observed to produce no detectable influence on the absorp-
tion kinetics of MgH2 + 4 mol%TiF3. Within the investigated
period, the only effect of lowered temperature was that the H-
capacity decreased from over 5 wt.% at 573 K to about 3.8 wt.%
at 423 K. Here it should be noted that the samples with TiF3
additive exhibited their optimal H-absorption/desorption per-
formance from the 1st cycle, thus eliminating the pre-activation
generally required for pure Mg.
The pronounced catalytic enhancement arising upon adding
TiF was also observed on the H-desorption aspect. Fig. 3 gives
3
the typical H-desorption curves of the MgH samples with and
2
without TiF additive at varied temperatures. For the pure MgH
2
3
sample, appreciable H-desorption could be achieved only when
the operation temperature was increased up to 623 K (curve c).
While for the MgH + 4 mol%TiF sample, near 5 wt.% hydro-
2
3
gen was desorbed within 600 s at 573 K (curve e). Even at a
lowered temperature of 553 K, the sample with TiF additive
3
could release about 2.8 wt.% hydrogen within 600 s, much faster
Of particular interest, it was found that the fast absorption
kinetics of the sample milled with TiF3 was well maintained
than that of pure MgH at 623 K. Currently, without definite evi-
dence on the favorable thermodynamic adjustment, the observed
2
accelerated desorption rate and reduced operation temperature
Fig. 1. Hydrogen absorption curves (with an initial hydrogen pressure of
2
.0 MPa) of mechanically milled MgH2 + 4 mol%TiF3 and MgH2 at varied tem-
Fig. 3. Hydrogen desorption curves of MgH2 + 4 mol%TiF3 and MgH2 at varied
temperatures with an initial H-pressure of 0.01 MPa: (a) 573 K, (b) 593 K, and
(c) 623 K for MgH2; (d) 553 K, (e) 573 K for MgH2 + 4 mol%TiF3.
peratures: (a) 573 K, (b) 473 K and (c) 423 K for MgH2 + 4 mol%TiF3; (d) 573 K
and (e) 473 K for MgH2.

L.-P. Ma et al. / Journal of Alloys and Compounds 432 (2007) L1–L4
L3
Fig. 4. Hydrogen absorption/desorption curves of MgH2 + 4 mol%TiF3 in the
st and 16th cycle at 573 K.
1
Fig. 5. XRD patterns for: (a) as-milled MgH2, (b) as-milled MgH2 + 4
mol%TiF3, (c) MgH2 + 4 mol%TiF3 after dehydrogenation at 573 K, and (d)
MgH2 + 4 mol%TiF3 after hydrogenation at 573 K. A small amount of MgO
was detected, which may come from air contamination during XRD measure-
ments.
are attributed to kinetic improvement arising upon adding TiF3.
Further investigations on the thermodynamic properties (i.e.
pressure–composition–isotherms measurements) of the materi-
als are therefore required to clarify this point.
(1) [34]:
The mechanically prepared MgH2 + 4 mol%TiF3 was further
examined with respect to the cycling performance. Whereas the
materials possess fast kinetics at moderate temperature, an oper-
ation temperature of 573 K was applied in absorption/desorption
cycles to evaluate the performance stability at elevated tem-
perature [33]. As seen in Fig. 4, the absorption/desorption
kinetics persisted well even after 15 cycles. However, the hydro-
gen capacity was observed to gradually decrease, and down to
2
2
1
MgH2 + TiF3 → MgF2 + TiH2 + H2,
3
3
3
◦
ꢀ
rG = −197 kJ/mol
(1)
m
In view of the instant high-energy environment produced by
collision between vial and balls and among the balls, it is highly
expected that this reaction can occur, at least partially, during
milling process. The failure to detect the in situ formed phases in
the as-milled samples should be ascribed to the defective nanos-
tructure and the peak locations close to those of host hydride.
More exactly, the weak diffraction peaks from the in situ formed
phases might be masked by large and broad neighboring peaks
of Mg hydride. After being subjected to de-/hydrogenation pro-
cesses, both the in situ formed and host hydride phases obtained
an increased structure ordering. As a result, the previously invis-
ible phases became visible, and previously weak peaks became
stronger.
The combined property examination and phase characteriza-
tion suggest a possible correlation between the in situ formed
MgF2 and/or TiH2 and property improvement achieved in the
MgH2 + 4 mol%TiF3 system. From literature, both phases pos-
sess catalytic activity on the absorption/desorption processes
of MgH2. It has been proposed that, in the MgH2 samples
milled with transition metal (Fe, Ni) fluorides, the formed MgF2
replaces the original surface oxide layer and provides a reactive
and protective fluorinated surface for hydrogen uptake [24,25].
Ti species in the catalytic form of titanium hydride were reported
to be excellent in enhancing the absorption/desorption kinetics
of host MgH2 among those investigated transition metals [27].
In the present material system, both MgF2 and TiH2 phases were
involved. However, the attempt of adding both phases simulta-
neously into the host MgH2 did not result in a catalytic enhance-
ment similar to that obtained by adding TiF3, especially at mod-
erateoperationtemperatures(resultsnotshownhere). Therefore,
4.6 wt.% in the 16th cycle within the investigated timescale.
Fortunately, further investigation has found that the problematic
capacity loss upon cycling could be partially solved by adjust-
ing the composition of the system, the details of which will be
presented in a forthcoming paper.
In our preliminary efforts to understand the catalytic effect
of TiF3 on the sorption kinetics of MgH2, XRD examination
was performed. Fig. 5 presents the XRD patterns of the as-
milled MgH2 + 4 mol%TiF3 sample as well as those after dehy-
drogenation and hydrogenation at 573 K. For comparison, the
profile of the pure MgH2 sample milled under identical con-
ditions is also included. It was observed that the diffraction
patterns of the as-milled samples with or without TiF3 addi-
tive were quite similar. Both were characterized by the weak
and broad diffraction peaks, indicating the decreased grain size
and the increased lattice strain [12,19–21,24,25]. Additionally,
the intensive milling led to a partial transformation of MgH2
from a low-pressure ␤-phase (tetragonal) to a high-pressure
metastable ␥-phase (orthorhombic) [7,20,25,26], which disap-
peared in the following de-/hydrogenation processes. For the as-
milled MgH2 + 4 mol%TiF3 sample, no diffraction peaks from
TiF3 or any other reaction products could be definitely identi-
fied. But the situation changed after the sample was subjected to
de-/hydrogenation processes. As shown in the patterns (c) and
(
d), MgF2 and TiH2 could then be identified. Judging from the
thermodynamic potentials of the various possible reaction path-
ways, the formation of these two phases should follow reaction

L4
L.-P. Ma et al. / Journal of Alloys and Compounds 432 (2007) L1–L4
the observed kinetic enhancement should not be safely attributed
to the role of MgF2 and TiH2 phases (or even a possible syner-
getic catalytic function of the individual catalysts). This leaves
open the question of the dominant catalytic mechanism respon-
sible for the kinetic enhancement. It is worth noting that for the
[7] A. Bassetti, E. Bonetti, L. Pasquini, J. Grbovic, M. Vittori Antisari, Eur.
Phys. J. B 43 (2005) 19.
[
8] S. Boufricha, J.P. Dodelet, D. Guay, J. Hout, R. Schulz, J. Mater. Res. 16
2001) 2893.
9] A. Zaluska, L. Zaluski, J.O. Strm-Olsen, J. Alloys Compd. 288 (1999)
17.
[10] T. Ichikawa, N. Hanada, S. Isobe, H.Y. Leng, H. Fujii, Mater. Trans. 46
2005) 1.
(
[
2
MgH2 + Nb2O and MgH2 + Nb systems, which possess fast H-
5
(
sorption kinetics at moderate temperatures, a newly identified
ternary magnesium–niobium oxide has been suggested to act as
a catalytically active phase [16,35]. This finding may provide
a ready reference for the current MgH2 + TiF3 system. In this
regard, further investigations are currently underway.
[
11] Y. Kojima, Y. Kawai, T. Haga, Mater. Res. Soc. Symp. Proc. 837 (2004)
N3.5.1.
12] J. Charbonnier, P. de Rango, D. Fruchart, S. Miraglia, L. Pontonnier, S.
Rivoirard, N. Skryabina, P. Vulliet, J. Alloys Compd. 383 (2004) 205.
[
[13] N. Hanada, T. Ichikawa, H. Fujii, J. Phys. Chem. B 109 (2005) 7188.
[
[
[
14] N. Bazzanella, R. Checchetto, A. Miotello, Appl. Phys. Lett. 85 (2004)
212.
15] P. Wang, A.M. Wang, H.F. Zhang, B.Z. Ding, Z.Q. Hu, J. Alloys Compd.
13 (2000) 218.
16] O. Friedrichs, F. Aguey-Zinsou, J.R. Ares Ferna’ndez, J.C. Sa’nchez-
Lo’pez, A. Justo, T. Klassen, R. Bormann, A. Ferna’ndez, Acta Mater.
5
4
. Conclusions
3
By mechanical milling with TiF3, the hydrogen sorption
kinetics of MgH2 could be markedly improved, in particular
on the absorption half-cycle at near room temperature. Fur-
thermore, the catalytically enhanced kinetic performance was
found to persist in the absorption/desorption cycles. But the cat-
alytic mechanism has not yet been well established. In addition,
this system still suffered from cycling capacity loss. Further
exploration on the composition–structure–property correlation
may lead to a comprehensive mechanism understanding and
make possible optimal preparation conditions to pursue further
improved sorption kinetics and capacity.
54 (2006) 105.
[
[
17] J.L. Bobet, F.J. Castro, B. Chevalier, Scripta Mater. 52 (2005) 33.
18] S. Rivoirard, P. de Rango, D. Fruchart, J. Charbonnier, D. Vempaire, J.
Alloys Compd. 356–357 (2003) 622.
19] N. Hanada, T. Ichikawa, S. Orimo, H. Fujii, J. Alloys Compd. 366 (2004)
269.
[
[
[
20] F.J. Castro, J.L. Bobet, J. Alloys Compd. 366 (2004) 303.
21] N. Hanada, T. Ichikawa, H. Fujii, J. Alloys Compd. 404–406 (2005)
716.
[
[
22] W. Oelerich, T. Klassen, R. Bormann, J. Alloys Compd. 315 (2001) 237.
23] H. Imamura, M. Kusuhara, S. Minami, M. Matsumoto, K. Masanari, Y.
Sakata, K. Itoh, T. Fukunaga, Acta Mater. 51 (2003) 6407.
[
[
24] A.R. Yavari, A. LeMoulec, F.R. de Castro, S. Deledda, O. Friedrichs, W.J.
Botta, G. Vaughan, T. Klassen, A. Fernandez, A. Kvick, Scripta Mater. 52
Acknowledgements
(
2005) 719.
The financial support for this research from the Hundred
Talents Project of Chinese Academy of Sciences is gratefully
acknowledged. L. Ma thanks Mr. Cheng-Zhang Wu, Mr. Xiang-
Dong Kang and Mr. Yong Chen for their helpful discussions.
25] S. Deledda, A. Borissova, C. Poinsignon, W.J. Botta, M. Dornheim, T.
Klassen, J. Alloys Compd. 404 (2005) 409.
[26] C.Z. Wu, P. Wang, X.D. Yao, C. Liu, D.M. Chen, G.Q. Lu, H.M. Cheng,
J. Phys. Chem. B 109 (2005) 22217.
[
27] G. Liang, J. Huot, S. Boily, A. Van Neste, R. Schulz, J. Alloys Compd. 292
1999) 247.
28] G. Liang, J. Huot, S. Boily, A. Van Neste, R. Schulz, J. Alloys Compd. 291
1999) 295.
(
References
[
(
[
[
[
1] R.C. Bowman, B. Fultz, MRS Bull. 27 (2002) 688.
2] P. Mandal, O.N. Srivastava, J. Alloys Compd. 205 (1994) 111.
3] G. Liang, S. Boily, J. Huot, A. Van Neste, R. Schulz, J. Alloys Compd. 268
[29] N. Hanada, T. Ichikawa, S. Hino, H. Fujii, J. Alloys Compd. in press.
[30] P. Wang, X.D. Kang, H.M. Cheng, Appl. Phys. Lett. 87 (2005) 071911.
[31] P. Wang, X.D. Kang, H.M. Cheng, Chemphyschem 6 (2005) 2488.
[32] P. Wang, X.D. Kang, H.M. Cheng, J. Alloys Compd., in press.
[33] Z. Dehouche, R. Djaozandry, J. Hout, S. Boily, J. Goyette, T.K. Bose, R.
Schulz, J. Alloys Compd. 305 (2000) 264.
(
1998) 302.
4] P. Wang, A.M. Wang, H.F. Zhang, B.Z. Ding, Z.Q. Hu, Acta Mater. 49
2001) 921.
5] P. Palade, S. Sartori, A. Maddalena, G. Principi, S. Lo Russo, M. Lazarescu,
G. Schinteie, V. Kuncser, G. Filoti, J. Alloys Compd. 415 (2006) 170.
6] A. Maddalena, M. Petris, P. Palade, S. Sartori, G. Principi, E. Settimo, B.
Molinas, S. Lo Russo, Int. J. Hydrogen Energy, in press.
[
[
[
(
[
[
34] I. Barin, Thermochemical Data of Pure Substances, 3rd ed., Wiley–VCH
Verlag GmbH, Weinheim, 1995.
35] H. Gijs Schimmel, J. Huot, L.C. Chapon, F.D. Tichelaar, F.M. Mulder, J.
Am. Chem. Soc. 127 (2005) 14384.