Method for preparing Ti-doped NaAlH4 using Ti powder: Observation of an unusual reversible dehydrogenation behavior

Article abstract of DOI:10.1016/j.jallcom.2004.02.006

Titanium powder can be directly used as dopant in the preparation of catalytically enhanced Ti-doped NaAlH₄ upon mechanical milling under an atmosphere of hydrogen. The hydrogen storage performance of NaAlH₄ that was doped through th

Full text of DOI:10.1016/j.jallcom.2004.02.006

Journal of Alloys and Compounds 379 (2004) 99–102
Method for preparing Ti-doped NaAlH4 using Ti powder:
observation of an unusual reversible dehydrogenation behavior
P. Wang∗, C.M. Jensen∗
Department of Chemistry, University of Hawaii, Honolulu, HI 96822, USA
Received 13 January 2004; accepted 12 February 2004
Abstract
Titanium powder can be directly used as dopant in the preparation of catalytically enhanced Ti-doped NaAlH4 upon mechanical milling
under an atmosphere of hydrogen. The hydrogen storage performance of NaAlH4 that was doped through this method was found to be highly
dependent on the milling time. A sample mechanically milled under hydrogen for 10 h was initially observed to discharge >3.3 wt.% of
hydrogen within 1 h at 150 ◦C. Rehydrogenation was accomplished at 120 ◦C under about 12 MPa H2 within 10 h. The hydrogen capacity of
this material was found to gradually decrease to 2.8 wt.% after eight cycles. The dehydriding kinetics and cycling properties of NaAlH4 that
is doped through this novel method is clearly distinct from the material that is obtained through the utilization of Ti(III) or Ti(IV) dopant
precursors and milling under argon. This suggests that the nature of active Ti-species may differ in these two varieties of doped NaAlH4.
© 2004 Elsevier B.V. All rights reserved.
Keywords: Alanate; Hydrogen storage; Kinetics; Titanium powder
1. Introduction
cling capacities of only 3–4 wt.% have been achieved under
practical conditions. This is far from the 5.6 wt.% hydrogen
that is theoretically available according to Eq. (1)
3
In 1997, Bogdanovic and Schwickardi reported that
the elimination of hydrogen from solid NaAlH4 could be
markedly accelerated and rendered reversible under mod-
erate conditions upon mixing the hydride with a few mole
percent of selected transition metal based precursors [1].
This was a revolutionary finding in the area of metal hy-
drides as hydrogen cycling at moderate temperatures was
unprecedented for saline hydrides. Following this break-
through, there have been intensifying efforts aimed at the
utilization of NaAlH4 and related hydrides with high gravi-
metric hydrogen as practical hydrogen storage materials
[2–24]. We subsequently discovered an improved doping
method that greatly promotes the kinetics of the reversible
dehydrogenation of NaAlH4 and stabilizes its hydrogen
cycling capacity [2–4]. Further efforts on the development
of Ti-doped NaAlH4 have given rise to hydrogen storage
materials that stand as the current state-of-the-art [5–11].
However, despite this progress, the reversible hydrogen cy-
NaAlH4 → NaH + Al + H2
(1)
2
The doping process results in the formation of by-products
in the kinetically enhanced hydride. For example, doping
with TiCl3 results in the formation of NaCl [9,10]. The
by-products are present as segregated crystallites that are dis-
persed throughout the substituted hydride [9,10]. Although
they do not seem to interfere with the dehydriding or hydrid-
ing processes, they constitute “dead weight” that becomes
increasingly problematic as the doping level of the hydride is
raised. Thus despite inducing improved kinetic and thermo-
dynamic properties, current methods for doping the hydride
results in the lowering of the gravimetric hydrogen capacity
to an unacceptable level.
Clearly, elimination of the doping side-products would
greatly improve the hydrogen capacity of these materials.
In this regard, it was recently reported that active Ti-dopant
was introduced into NaAlH4 upon ball-milling with the Ti(0)
precursor, Ti13·6THF [13,14]. We have now developed a
method whereby off-the-shelf titanium powder can be uti-
lized directly for the introduction of Ti dopants into the
NaAlH4 host. We herein report the details of this method of
∗
Corresponding authors. Tel.: +1-808-956-2769;
fax: +1-808-956-5908.
E-mail addresses: jensen@gold.chem.hawaii.edu (C.M. Jensen),
pwang@hawaii.edu (P. Wang).
0925-8388/$ – see front matter © 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2004.02.006

100
P. Wang, C.M. Jensen / Journal of Alloys and Compounds 379 (2004) 99–102
preparation of Ti-doped NaAlH4 and the hydrogen storage
behavior of this material.
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
a
2. Experimental
Titanium powder (99.98%, ∼325 mesh) and raw NaAlH4
were purchased from Aldrich Chemical Co. and Albemarle
Corp., respectively. The purification of NaAlH4 was ac-
complished under an atmosphere of nitrogen using standard
Schlenk technique. Aluminum powder was removed from
the raw hydride via Soxlet extraction with dry, oxygen-free
tetrahydrofuran (THF). Final purification was accomplished
through recrystallization from THF/pentane.
Mixtures of NaAlH4 and titanium powder were mechan-
ically milled by using a Fritsch 6 Planetary mill at 400 rpm
in a stainless steel bowl and grinding balls. The milling
was performed under hydrogen or argon atmosphere, with
an initial pressure of 0.8 and 0.1 MPa, respectively. The
ball-to-powder ratio varied between 30:1 and 40:1.
b
c
0
30
60
90
120
150
180
Time, min
Fig. 1. First H-desorption profiles of different samples at 150 ◦C. (a)
NaAlH4 + 7 mol% Ti mechanically milled under H2 atmosphere for 5 h;
(b) NaAlH4 + 7 mol% Ti mechanically milled under Ar atmosphere for
5 h; (c) NaAlH4 mechanically milled under H2 atmosphere for 5 h.
Hydrogen absorption/desorption behavior was monitored
with a carefully calibrated Sievert’s type apparatus (LESCA
Co., Japan). Precise pressure measurement and temperature
controlling were accomplished by using a high-precision
pressure transducer and silicon oil bath, respectively. A typ-
ical cyclic experiment entailed desorption at 150 ◦C and
absorption at 120 ◦C with an initial pressure condition of
<1 Torr and ∼12 MPa, respectively. The weight of elemental
Ti was taken into account in the determination of H-capacity.
quired to obtain the maximum kinetic enhancement for the
hydride mechanically doped with Ti powder. As seen in
Fig. 3, increasing the milling time from 1 to 10 h clearly
improves the dehydrogenation kinetics observed at 150 ◦C.
It is notable that the sample milled for 10 h was observed
to evolve ∼3.3 wt.% hydrogen in about 1 h. This value
approaches the theoretical amount of hydrogen that can
be obtained from NaAlH4 containing 7 mol% Ti in the
dehydrogenation reaction seen in Eq. (2)
NaAlH4 → 1 Na3AlH6 + Al + H2
(2)
2
3
3
3. Results
However, this material undergoes the subsequent decompo-
sition of Na3AlH6 to NaH only at a very low rate, even at
150 ◦C.
We have found that a pronounced enhancement of the
dehydrogenation kinetics of NaAlH4 can be achieved upon
mechanical milling the hydride with Ti powder under hy-
drogen atmosphere. As seen in Fig. 1, over 3.3 wt.% of hy-
drogen was discharged in about 2 h at 150 ◦C from a sample
of NaAlH4 that was ball-milled with 7 mol% Ti powder un-
der hydrogen for 5 h. The achieved kinetics enhancement is
markedly superior to that of NaAlH4 milled under Ar atmo-
sphere. For comparison, NaAlH4 that was ball-milled under
hydrogen for 5 h without Ti powder was also examined. As
seen in Fig. 1, hydrogen must play a critical role in the gen-
eration of the active titanium species from Ti powder.
It is also noteworthy that the dehydrogenation kinetics of
NaAlH4 that is doped through this novel method are signif-
icantly slower than those arising upon doping the hydride
with Ti(III) or Ti(IV) dopant precursors [2–11]. The differ-
ing kinetic behaviors can be seen in Fig. 2 which compares
the first dehydriding profiles between a sample of NaAlH4
that was doped with TiF3 by standard milling procedure
[2–11] and a sample prepared by the present method.
In contrast to the standard procedure for preparing
Ti-doped NaAlH4 through mechanical milling with Ti(III)
or Ti(IV) precursors [2–11], prolonged milling time is re-
The hydrogen recharging of this material was accom-
plished at 120 ◦C under about 12 MPa H2 within 8–10 h.
4
NaAlH4+4mol%TiF3
NaAlH4+4mol%Ti
3
2
1
0
0
2
4
6
8
10
12
Time, h
Fig. 2. Comparison of the first H-desorption profile at 150 ◦C between
NaAlH4 + 4 mol% Ti milled under H2 atmosphere for 10 h and NaAlH4
+ 4 mol% TiF3 milled under N2 for 0.5 h.

P. Wang, C.M. Jensen / Journal of Alloys and Compounds 379 (2004) 99–102
101
the presence of a Ti–Al alloy in the doped hydride [23] and
substitution of Ti into the bulk hydride lattice [19,24] have
been reported. However, neither a catalytically active Ti–Al
alloy nor evidence for well characterized sites of lattice sub-
stitution has yet been obtained. Thus the nature of the active
Ti-species in the doped hydride has not been firmly estab-
lished.
The results obtained in this study suggest other possibili-
ties as to the nature of the active Ti-species in doped NaAlH4.
Doping through our alternative method clearly confers re-
versible dehydrogenation behavior and a pattern of cycling
stability that are quite distinct from Ti-doped NaAlH4 aris-
ing from milling with Ti(III) and Ti(IV) dopant precursors.
This finding raises the additional possibility that more than
one Ti-species can contribute to the kinetic enhancement
of the dehydrogenation of NaAlH4. This possibility is also
suggested by the discontinuity in the relationship between
level of Ti doping and the rate of dehydrogenation that is
observed at higher (>2 mol%) doping levels [10].
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
10h
5h
2h
1h
0
50
100
150
200
Time, min
Fig. 3. First 150 ◦C H-desorption profiles of NaAlH4 + 7 mol% Ti me-
chanically milled under H2 atmosphere for different periods.
The observations of reduced kinetic enhancement and cy-
cling instability suggest that the Ti-dopant not be as tightly
incorporated in the doped hydride that is produced by our
alternative method. One possibility is that in this material,
Ti-species occupy interstitial rather than lattice positions.
By material design, a suitable interstitial position in host
lattice could be created that would better accommodate the
Ti-species such that improved reversible dehydrogenation
performance could be achieved. Following this approach, we
have obtained some promising results that will be presented
in forthcoming publication.
4
1
2
3
4-6
3
7
8
2
1
0
0
2
4
6
8
10
12
14
5. Conclusion
Time, h
Fig. 4. 150 ◦C H-desorption profiles of NaAlH4 + 4 mol% Ti (the first
Mechanical milling of NaAlH4 with off-the-shelf Ti pow-
der under hydrogen atmosphere has been found to pro-
duce Ti-doped NaAlH4 that exhibits kinetically enhanced,
reversible dehydrogenation behavior. The Ti-doped hydride
that was obtained in the present study from the alternative
doping procedure has practical hydrogen storage properties
that are inferior to the doped hydride that is produced by me-
chanical milling with Ti(III) or Ti(IV) precursors under an
inert atmosphere. However, our findings establish a possibil-
ity of developing a high-performance Ti-doped NaAlH4 for
hydrogen storage. This process would eliminate the prob-
lems that plague the standard method of preparation of the
doped hydride: capacity loss due to the generation of inac-
tive by-products; and the evolution of gaseous impurities that
are highly detrimental to fuel-cell operation. Furthermore,
our results suggest that more than one Ti-species and/or Ti
in more than one structural environment can kinetically en-
hance the dehydrogenation of NaAlH4. In view of the quite
limited understanding on the role(s) of Ti-species played in
the reversible dehydrogenation of alanates, the researchers
developing alanate-based materials for hydrogen storage ap-
plications should maintain a wide perspective.
eight cycles) milled under H2 atmosphere for 10 h.
However, the initial hydrogen capacity can not be fully re-
stored. As seen in Fig. 4, the hydrogen capacity gradually
decreases with increasing the cycle number and down to
2.8 wt.% after eight cycles. Additionally, the kinetics of the
first decomposition step were also observed to gradually de-
crease.
4. Discussion
The nature of the active Ti-species in Ti-doped NaAlH4
has been a subject of interest, speculation, and controversy.
It was initially suggested that the remarkable enhancement
of the hydrogen cycling kinetics in Ti-doped NaAlH4 was
due to surface-localized catalytic species consisting of tita-
nium metal or a Ti–Al alloy [1,5,8]. Alternatively, we have
hypothesized that doping involves the substitution of tita-
nium into the bulk of the hydride [19]. Evidence for both

102
P. Wang, C.M. Jensen / Journal of Alloys and Compounds 379 (2004) 99–102
[11] B. Bogdanovic, G. Sandrock, MRS Bull. 27 (2002) 712.
Acknowledgements
[12] G. Sandrock, K. Gross, G. Thomas, C. Jensen, D. Meeker, S. Takara,
J. Alloys Comp. 330/332 (2002) 696.
[13] M. Fichtner, O. Fuhr, O. Kircher, J. Rothe, Nanotechnology 14 (2003)
778.
[14] B. Bogdanovic, M. Felderhoff, S. Kaskel, A. Pommerin, K. Schlichte,
F. Schüth, Adv. Mater. 15 (2003) 1012.
[15] A. Zaluska, L. Zaluski, J.O. Ström-Olsen, J. Alloys Comp. 298
(2000) 125.
The authors are very grateful to Mr. D. Kempton for his
assistance in constructing the experimental apparatus. The
financial support for this research from the Office of Hydro-
gen, Fuel Cells, and Infrastructure Technologies of the US
Department of Energy is greatfully acknowledged.
[16] K.J. Gross, S. Guthrie, S. Takara, G.J. Thomas, J. Alloys Comp. 297
(2000) 270.
References
[17] V.P. Balema, K.W. Dennis, V.K. Pecharsky, Chem. Commun. (2000)
1665.
[18] G.J. Thomas, K.J. Gross, N.Y.C. Yang, C. Jensen, J. Alloys Comp.
330–332 (2002) 702.
[19] D. Sun, T. Kiyobayashi, H.T. Takeshita, N. Kuriyama, C.M. Jensen,
J. Alloys Comp. 337 (2002) L8.
[20] G.P. Meisner, G.G. Tibbetts, F.E. Pinkerton, C.H. Olk, M.P. Balogh,
J. Alloys Comp. 337 (2002) 254.
[21] M.P. Balogh, G.G. Tibbetts, F.E. Pinkerton, G.P. Meisner, C.H. Olk,
J. Alloys Comp. 350 (2003) 136.
[22] M. Fichtner, O. Fuhr, J. Alloys Comp. 345 (2002) 286.
[23] B. Bogdanovic, M. Felderhoff, M. Germann, M. Härtel, A. Pom-
merin, F. Schüth, C. Weidenthaler, B. Zibrowius, J. Alloys Comp.
350 (2003) 246.
[24] T. Kiyobayashi, S.S. Srinivasan, D. Sun, C.M. Jensen, J. Phys. Chem.
A 107 (2003) 7671.
[1] B. Bogdanovic, M. Schwickardi, J. Alloys Comp. 253–254 (1997) 1.
[2] C.M. Jensen, R. Zidan, N. Mariels, A. Hee, C. Hagen, Int. J. Hy-
drogen Energy 24 (1999) 461.
[3] R.A. Zidan, S. Takara, A.G. Hee, C.M. Jensen, J. Alloys Comp. 285
(1999) 119.
[4] C.M. Jensen, R.A. Zidan, US Patent 6,471,935 (2002).
[5] B. Bogdanovic, R.A. Brand, A. Marjanovic, M. Schwickardi, J. Tölle,
J. Alloys Comp. 302 (2000) 36.
[6] C.M. Jensen, K.J. Gross, Appl. Phys. A 72 (2001) 213.
[7] B. Bogdanovic, M. Schwickardi, Appl. Phys. A 72 (2001) 221.
[8] K.J. Gross, G.J. Thomas, C.M. Jensen, J. Alloys Comp. 330–332
(2002) 683.
[9] K.J. Gross, G. Sandrock, G.J. Thomas, J. Alloys Comp. 330–332
(2002) 691.
[10] G. Sandrock, K. Gross, G. Thomas, J. Alloys Comp. 339 (2002)
299.