Effect of mechanically induced modification on TiH2 thermal stability

Article abstract of DOI:10.1016/j.jallcom.2011.01.048

Interrupted thermal desorption and X-ray diffraction techniques were used to study the effect of graphite and boron additives on non-equilibrium decomposition of mechanically activated commercial TiH₂. The phase transformation sequence is described as a number of consecutive reactions corresponding to desorption peaks. The process is compared to non-equilibrium decomposition of commercial TiH₂. The mechanical pre-treatment with additives significantly eases and accelerates decomposition of TiH₂ to αTi(H), but hinders the stage of αTi(H) → αTi transformation. Only a small portion of pure αTi was detected for as-milled TiH₂ and TiH₂/B powders after the final TPD. No αTi was formed in the case of as-milled TiH₂/C.

Full text of DOI:10.1016/j.jallcom.2011.01.048

Journal of Alloys and Compounds 509S (2011) S754–S758
Contents lists available at ScienceDirect
Journal of Alloys and Compounds
journal homepage: www.elsevier.com/locate/jallcom
Effect of mechanically induced modification on TiH2 thermal stability
O.S. Morozovaa,∗, T.I. Khomenkoa, Ch. Borchersb, A.V. Leonovc
a Semenov Institute of Chemical Physics RAS, Kosygin st. 4, 119991 Moscow, Russia
b Institute of Material Physics, University of Goettingen, Friedrich-Hund-Platz 1, 37077 Goettingen, Germany
c Lomonosov Moscow State University, Chemical Department, Leninskie Gory, 119899 Moscow, Russia
a r t i c l e i n f o
a b s t r a c t
Article history:
Interrupted thermal desorption and X-ray diffraction techniques were used to study the effect of graphite
and boron additives on non-equilibrium decomposition of mechanically activated commercial TiH2.
The phase transformation sequence is described as a number of consecutive reactions corresponding
to desorption peaks. The process is compared to non-equilibrium decomposition of commercial TiH2.
The mechanical pre-treatment with additives significantly eases and accelerates decomposition of TiH2
to ␣Ti(H), but hinders the stage of ␣Ti(H) → ␣Ti transformation. Only a small portion of pure ␣Ti was
detected for as-milled TiH2 and TiH2/B powders after the final TPD. No ␣Ti was formed in the case of
as-milled TiH2/C.
Received 11 August 2010
Received in revised form
30 December 2010
Accepted 7 January 2011
Available online 20 January 2011
Keywords:
TiH2 as-milled
TiH2/B and TiH2/C nanocomposites
Ball-milling treatment
High-resolution transmission electron
microscopy
© 2011 Elsevier B.V. All rights reserved.
Interrupted temperature programmed
desorption
X-ray diffraction
1. Introduction
This study is focused on the mechanism of non-equilibrium
decomposition of ball-milled commercial TiH2 compared to that
The major requirements for metal hydrides as hydrogen storage
materials are the high hydrogen mass percentage, chemical stabil-
ity and low-temperature fast hydrogenation and dehydrogenation
kinetics [1]. The charging–discharging kinetics can be significantly
improved by ball milling treatment of the powders in the presence
of different additives [2–4]. Non-equilibrium heating conditions
in the temperature programmed desorption (TPD) regime accel-
erate the metal hydride decomposition procedure. As was shown
by the example of TiH2, ball-milling with graphite, boron and h-
BN additives [5,6] drastically decreases Ti-hydride decomposition
temperature in TPD regime due to (1) powder size degradation and
(2) appearance of new occupation sites available for H atoms in
TiH2 lattice modified by C, B, and N interstitial atoms, respectively.
However, this non-equilibrium metal hydride decomposition is a
poorly studied multi-stage process, the single stages of which can
be accelerated or retarded by different pretreatment procedures
[7].
of original Ti-hydride [8], and on the effect of boron and
graphite additives on this process. The phase transformation
sequence is depicted as a number of consecutive and parallel
reactions.
2. Experimental
Commercial TiH2 (Aldrich, 99% pure, 325 mesh, S = 0.32 m2/g), boron (amor-
phous, purity: 98.5%, S = 5.4 m2/g) and graphite (highly oriented pyrolytic, purity:
99%, S = 3 m2/g) powders were used. 16.6 mass% of boron or graphite was added
to TiH2. Milling was carried out for 66 min at the room temperature in a flow
mechanochemical reactor with an average power intensity of 1 W/g in He flow. The
TiH2 decomposition was studied by temperature-programmed desorption (TPD)
technique. Peaks were separated according to Lorentzian TPD curve fitting. The
experimental details are published elsewhere [5]. The sample was heated up to
the chosen temperature and then rapidly cooled by transferring the reactor from
the heater to ice water. A fresh powder was used in each TPD run. XRD patterns
were recorded after each TPD experiment using a Dron-3 diffractometer with Cu
K␣ radiation. Quantitative X-ray phase analysis was performed applying a fitting
procedure [9].
3. Results and discussion
∗
Corresponding author at: Department of Kinetics and Catalysis, Semenov
Significant TiH2 powder size degradation accompanied by par-
tial hydrogen loss was observed during the mechanical treatment
(Fig. 1). TiH2, TiH2/C and TiH2/B powders lost approximately
Institute of Chemical Physics RAS, Kosygin st. 4, 119991 Moscow, Russia.
Tel.: +7 495 939 7195; fax: +7 495 651 2191.
E-mail address: om@polymer.chph.ras.ru (O.S. Morozova).
0925-8388/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2011.01.048

O.S. Morozova et al. / Journal of Alloys and Compounds 509S (2011) S754–S758
S755
Fig. 1. SEM micrograph: (a) commercial TiH2; (b) TiH2 milled for 66 min; (c) TiH2/B milled for 66 min; (d) TiH2/C milled for 66 min.
∼5 at.%, ∼6 at.%, and ∼14 at.%, respectively, after 66 min of milling.
As a result, TiH1.9, TiH1.88/C and TiH1.72/B emerged. Accordingly, the
specific surface area increased from 0.32 m2/g to 6 m2/g, 38.2 m2/g
and 6 m2/g for TiH2, TiH2/C and TiH2/B, respectively. Average grain
size (from Scherer formula) was 13–15 nm. Fig. 2 shows TPD curves
recorded for TiH2 original and as-milled powders. Desorption in the
low-temperature region is a special feature of as-milled powders:
H2 release was detected at 300–350 K, which is remarkably lower
than original TiH2 onset temperature. Hydrogen desorption dur-
ing TPD was accompanied by structural transformations of TiH2
up to total decomposition. Step-by-step correlation between both
processes was studied by XRD, see Fig. 3.
The decomposition of original TiH2 was recently studied in
detail [8]. According to XRD data (Fig. 3a), the total decomposi-
tion of original TiH2 was divided into three parts corresponding to
three peaks that could be fitted to the TPD spectrum (Fig. 2a): (1)
␧TiH2 → ␦TiH2−x transition at Tmax = 743 K, (2) ıTiH2−x → ␣Ti(H) at
Tmax = 817 K and (3) ␣Ti(H) → ␣Ti at Tmax = 925 K. Finally, original
␧TiH2 was totally decomposed to ␣Ti. According to Ti–H phase dia-
gram, the presence of ␤Ti (the high-temperature phase) should be
postulated. But, it was not detected by XRD.
gen content [10]. Further heating stimulated both ␣Ti(H) formation
and progressive decrease of hydrogen content in ␦TiH2−x. The
high-temperature peak with Tmax = 795 K is mainly reflecting the
decomposition ␣Ti(H) → ␣Ti. After final heating, the sample con-
sists of about 83 wt.% of ␣Ti–6 at.% H and ␣Ti. The data obtained are
in a good agreement with [7].
Figs. 2c and 3c show TPD spectrum and corresponding XRD
patterns of TiH2/B powder. TiH2 decomposition started at about
300 K. The major low-temperature process was ␦TiH2−x → ␣Ti(H)
accompanied by a decrease of hydrogen content in both ␦TiH2−x
and ␣Ti(H) lattices. Thus, heating to 689 K led to H2 loss of ∼
47 mol.%. As a result, about 37 wt.% of hexagonal solid solution
␣Ti–11.4 at.% H (a = 0.298 nm, c = 0.481 nm) was detected. Such a
hydrogen concentration exceeds the equilibrium concentration of
8.4 at.% H [10] by far. Besides, ␦TiH2−x lattice parameter decreased
from 0.444 nm to 0.442 nm due to a depletion of hydrogen in Ti-
hydride. Heating to 742 K led to about 66 mol.% H2 loss and further
␦TiH2−x → ␣Ti(H) transformation accompanied by narrowing and
large-angle shift of XRD peaks of both phases. The hydrogen con-
tent in ␣Ti(H) solid solution decreased to 9–10 at.% (a = 0.297 nm,
c = 0.481 nm), which still exceeds the equilibrium concentration.
The high-temperature TPD peak (Tmax = 759 K) reflects the decom-
position ␦TiH2−x → ␣Ti(H) → ␣Ti. The major phase present after
heating to 822 K was a solid solution ␣Ti–7.4 at.% H (a = 0.297 nm;
c = 0.477 nm). Only traces of ␦TiH2−x were detected. The high tem-
perature slope reflects the transformation ␣Ti(H) → ␣Ti: the final
product of heating to 945 K contained traces of ␣Ti and two hexag-
onal phases characterized by equal parameters a = 0.296 nm and
Figs. 2b and 3b show TPD spectrum and XRD patterns recorded
for as-milled TiH2. The low-temperature part of TPD curve was
attributed to the fast transformation ␦TiH2−x → ␣Ti(H). Approxi-
mately 50% of H2 was desorbed during the heating to 723 K. As
a result, ∼50 wt.% of ␦TiH2−x was decomposed to a solid solution
␣Ti–9 at.% H. The lattice constant of residual ␦TiH2−x decreased
from a = 0.444 nm to a = 0.441 nm indicating a decrease in hydro-

S756
O.S. Morozova et al. / Journal of Alloys and Compounds 509S (2011) S754–S758
Fig. 2. TPD spectra recorded for (a) commercial TiH2 [8] and as-milled: (b) TiH2, (c) TiH2/B, and (d) TiH2/C. Vertical arrows point to temperatures of heating
interruption.
different c = 0.471 nm and c = 0.478 nm, respectively. The first phase
was attributed to ␣Ti–1.6 at.% H. Since hydrogen evolution was vir-
tually finished, the second hexagonal phase was attributed to a
solid solution ␣Ti(B), which was confirmed by XES technique [6].
Residual hydrogen was totally released after heating to 1073 K.
The hexagonal phase ␣Ti–1.6 at.% H disappeared, and the hexag-
onal solid solution ␣Ti(B) and a rather dispersed cubic phase TiB
(a = 0.421 nm) were detected.
XRD data, the final product formed at 900 K consisted of two
hexagonal phases (␣Ti – 8 at.% H and ␣Ti(C)) and a highly dis-
persed cubic phase Ti(C,H) in an amount of ∼32 wt.%. Here we
also suppose the carburization of Ti-hydride during hydrogen
evolution from Ti-hydride bulk, which may be a result of TiH2
lattice modification by C atoms under mechanical treatment
[5]. Due to graphite excess, a carbon peak is present in XRD
patterns.
Figs. 2d and 3d show TPD and XRD spectra for TiH2/C powder.
TiH2 decomposition also started at ∼300 K by the transforma-
tion ␦TiH2−x → ␣Ti(H), a solid solution with a non-equilibrium
H content up to 9 at.%. At 774 K, the major phase was a solid
solution ␣Ti(H). Only traces of ␦TiH2−x were detected. The high-
temperature slope of TPD peak with Tmax = 774 K reflected two
processes: the evolution of residual hydrogen and the forma-
tion of Ti(C,H) solid solutions. According to mass balance and
As was mentioned above, the presence of ␤Ti (the high-
temperature phase) should be postulated according to Ti–H phase
diagram. But, it was not detected by XRD in the samples cooled on
different TPD stages.
Table 1 summarizes the data obtained in TPD experiments and
the effective activation energies Ea of H2 desorption calculated
for the first and the last (high-temperature) peaks responsible
for initial ␦TiH2−x decomposition and ␣Ti(H) → ␣Ti transforma-
Table 1
TPD parameters calculated on a base of Lorentzian fitting of TPD spectra.
Sample
TiH2
Peak
H2 evolution [mol/g TiH2]
Tmax [K]
Ea [kJ/mol]
1
2
3
3.9 × 10−3 (19.5 mol.%)
1.23 × 10−2 (61.6 mol.%)
3.8 × 10−3 (18.9 mol.%)
743
817
920
219.5 6.5
229.7 7.2
216.1 7.3
First
Last
2.04 × 10−3 (10.9 mol.%)
607
787
107.9 5.2
126.3 2.9
TiH2 as-milled
TiH2/B as-milled
TiH2/C as-milled
1.13 × 10−2 (56.4 mol.%)
First
Last
2.28 × 10−3 (14.3 mol.%)
5.6 × 10−3 (35.2 mol.%)
596
759
65.4 1.5
170.4 7.3
First
Last
5.46 × 10−3 (∼29 mol.%)
614
774
35.7 0.5
209.2 11.9
4.88 × 10−3 (∼26 mol.%)

O.S. Morozova et al. / Journal of Alloys and Compounds 509S (2011) S754–S758
S757
Fig. 3. XRD patterns recorded for TiH2, TiH2/B and TiH2/C powders original, as-prepared and after TPD experiments depicted in Fig. 2: (a) commercial TiH2, original and
heated to785 K (low-temperature slop), 823 K (Tmax of the major peak) and 925 K (final decomposition) [8]; (b) as-milled TiH2 original and heated to 723 K (low-temperature
slop), 795 K (∼Tmax of the major peak) and 917 K (final decomposition); (c) as-milled TiH2/B original and heated to 689 K (just after the low-temperature Tmax), 742 K (nearby
∼Tmax of the high-temperature peak) and 945 K (final decomposition); (d) as-milled TiH2/C original and heated to 745 K (after the low-temperature Tmax), 774 K (∼Tmax of
the high-temperature peak) and 900 K (final decomposition).
tion, respectively. The processes include diffusion of H atoms from
the bulk to the surface, surface recombination of H atoms and
H2 desorption itself. Nevertheless, our estimations are in a good
agreement with experiment: Ea corresponding to the first pro-
cess (at around 600 K) drastically decreases in the order TiH2, TiH2
as-milled, TiH2/B, TiH2/C. Ea corresponding to the second process
significantly decreases for as-milled sample, but increases again
for TiH2/B, TiH2/C. This effect is likely to be related to the growth
of the fraction of highly distorted intergrain regions in nanocom-
posite TiH2/B and TiH2/C powders, where H mobility is much
lower than in the crystalline grains [11] or caused by difficulties
in H atoms recombination on the surfaces blocked by boron or
carbon.
cally decreased and (3) B and C atoms modified activated ␦TiH2−x
lattice of to give ␣Ti(B) or Ti-carbide-like phases. In contrast, it sup-
presses ␣Ti(H) → ␣Ti decomposition due to high concentration of
structural defects and strong affinity of hydrogen to boron and car-
bon, reflected by a pronounced increase in effective Ea of hydrogen
desorption.
Acknowledgements
This work was partly supported by Russian Foundation of Basic
Research (Project No. 10-03-00942-а) and Ministry of Science of
Russian Federation (Contract No 02.740.11.5176).
References
4. Conclusions
[1] A. Zaluska, L. Zaluski, J.O. Ström-Olsen, Appl. Phys. A 72 (2001) 157–165.
[2] R.A. Varin, S. Li, A. Calka, D. Wexler, J. Alloys Compd. 373 (2004) 270–276.
[3] H. Imamura, N. Takashi, T. Fujinaga, J. Alloys Compd. 253–254 (1997) 34–37.
[4] C.Z. Wu, P. Wang, X. Yao, C. Liu, D.M. Chen, G.Q. Lu, H.M. Cheng, J. Alloys Compd.
414 (2006) 259–264.
[5] E.Z. Kurmaev, O.S. Morozova, T.I. Khomenko, Ch. Borchers, S.N. Nemnonov, Y.
Harada, T. Tokushima, H. Osawa, T. Takeuchi, S. Shin, J. Alloys Compd. 395
(2005) 240–246.
The mechanical activation of TiH2 and, in particular, mechanical
activation with graphite and boron additives enhances Ti-hydride
decomposition kinetics to hexagonal solid solution T(H) because:
(1) one stage of this process ␧TiH2 → ␦TiH2−x was eliminated, (2)
initial Ea of hydrogen desorption at low temperatures dramati-

S758
O.S. Morozova et al. / Journal of Alloys and Compounds 509S (2011) S754–S758
[6] O.S. Morozova, T.I. Khomenko, A.V. Leonov, Ch. Borchers, E.Z. Kurmaev, A.
Moewes, J. Alloys Compd. 483 (2009) 309–312.
[7] V. Bhosle, E.G. Baburaj, M. Miranova, K. Salama, Mater. Sci. Eng. A356 (2003)
190–199.
[9] E.V. Shelekhov, T.A. Sviridova, Met. Sci. Heat Treat. 42 (2000) 309.
[10] F.D. Manchester, A. San-Martin, The H–Ti (hydrogen–titanium) system, in: F.D.
Manchester (Ed.), Phase Diagrams of Binary Hydrogen Alloys, ASM Interna-
tional, USA, 2000, pp. 238–258.
[8] Ch. Borchers, T.I. Khomenko, A.V. Leonov, O.S. Morozova, Termochim. Acta 493
(2009) 80–84.
[11] A.Ye. Yermakov, N.V. Mushnikov, M.A. Uimin, V.S. Gaviko, A.P. Tankeev,
A.V. Skripov, A.V. Solonin, A.L. Buzlukov, J. Alloy Compd. 425 (2006) 367–
372.