Novel hydrogen storage properties of MoS2 nanotubes

Article abstract of DOI:10.1016/S0925-8388(03)00114-2

Multi-wall MoS₂ nanotubes were prepared using a catalyzed thermal decomposition of (NH₄)₂MoS₄. The as synthesized nanotubes were treated with 5 M KOH solution at 50°C for 1 h. These two nanotubes were characterized by XRD, SEM, TEM/HRTEM, BET, XPS, CV and charge-discharge measurements. The results show that MoS₂ nanotubes can store relatively large amounts of hydrogen by gaseous and electrochemical storage. The adsorption-desorption reactions are highly reversible at the temperature of 25°C, implying that these nanotubes are promising materials for hydrogen storage.

Full text of DOI:10.1016/S0925-8388(03)00114-2

Journal of Alloys and Compounds 356–357 (2003) 413–417
L
www.elsevier.com/locate/jallcom
Novel hydrogen storage properties of MoS2 nanotubes
*
J. Chen , S.L. Li, Z.L. Tao
Institute of New Energy Material Chemistry, Nankai University, Tianjin 300071, PR China
Received 1 September 2002; accepted 15 November 2002
Abstract
Multi-wall MoS2 nanotubes were prepared using a catalyzed thermal decomposition of (NH4)2MoS4. The as synthesized nanotubes
were treated with 5 M KOH solution at 50 8C for 1 h. These two nanotubes were characterized by XRD, SEM, TEM/HRTEM, BET,
XPS, CV and charge–discharge measurements. The results show that MoS2 nanotubes can store relatively large amounts of hydrogen by
gaseous and electrochemical storage. The adsorption–desorption reactions are highly reversible at the temperature of 25 8C, implying that
these nanotubes are promising materials for hydrogen storage.
2003 Elsevier B.V. All rights reserved.
Keywords: MoS2; Nanotube; Hydrogen storage; Characterization
1. Introduction
chemical hydrogen storage in MoS2 nanotubes with 60%
purity. In this study, we report the synthesis of MoS2
nanotubes with 90% purity by a low-temperature catalytic
reaction and compare their gaseous and electrochemical
hydrogen storage properties.
Effective reversible hydrogen storage plays an important
role in practical use of hydrogen energy. One method,
which has been rigorously studied, involves the utilization
of metals, alloys, complexes, and ‘controversial’ advanced
carbon materials [1]. However, among them, no one has
been found to meet all the demands required for commer-
cial electric-vehicle applications. Apparently, researchers
in the hydrogen storage field must face the challenge on
how to look into new material combinations that exhibit
high volumetric/gravimetric hydrogen capacity as mag-
nesium-based hydrides and fast kinetics at room tempera-
tures as LaNi5 system.
Graphite is the most stable form of carbon under
ambient conditions. Notwithstanding, graphite nanoparti-
cles have been unstable and they close into fullerenes [2]
and nanotubes [3]. A large amount of work on the
experimental and theoretical studies of hydrogen adsorp-
tion has been carried out to measure the storage capacity of
carbon nanotubes [4–10]. Using the paradigm of carbon
fullerenes, Tenne and co-workers [11,12] reported that
nanoparticles of 2-D layered transition-metal dichal-
cogenides collapse into fullerene-like cages and nanotubes.
In our previous paper [13], we described the electro-
2. Experimental
Our production for yielding MoS2 nanotubes consists of
three steps [14]: the preparation of needle-like
(NH4)2MoS4 from (NH4)6Mo12O39?12H2O, the ball mil-
ling of (NH4)2MoS4 in H2 atmosphere (0.2 MPa with a
Fuji Auto mixer at 800 rpm for 1 h), and the sintering at
400 8C of the as ball-milled sample in a floating hydrogen/
thiophene (C4H4S) atmosphere (volume ratio of
H2:C4H4S519:1). The gas chromatograph analysis shows
that C4H4S is taking the role of catalyst. The gas–solid
reaction which occurred during the sintering process is:
(NH4)2MoS4 1 H2 → MoS2 1 2NH3 1 2H2S
(1)
After the reaction was completed, the resulting black solid
was treated by 5 M KOH solution at 50 8C for 1 h. Then,
the solid was filtered, washed with deionized water several
times, and finally dried in a vacuum at 80 8C for 1 h.
The untreated and treated samples were analyzed by
XRD (Rigaku INT-2000, 40 kV/150 mA), SEM (Jeol
*
Corresponding author.
E-mail address: chenabc@nankai.edu.cn (J. Chen).
0925-8388/03/$ – see front matter
doi:10.1016/S0925-8388(03)00114-2
2003 Elsevier B.V. All rights reserved.

414
J. Chen et al. / Journal of Alloys and Compounds 356–357 (2003) 413–417
JSM-5600, 15 kV); TEM and HRTEM (Jeol JEM 3000F,
300 kV), XPS (ESCA-3400, Shimadzu Electron), and BET
adsorption–desorption (Shimadzu-Micromeritics ASAP-
2010).
The gaseous hydrogen adsorption/desorption amounts
of polycrystalline MoS2, untreated and treated MoS2
nanotubes (about 0.5 g) were determined using an auto-
mated Sieverts-type apparatus (Advanced Materials Corpo-
ration) in the temperature of 25 8C. The volume in the
hydriding/dehydriding process had been carefully cali-
brated before the actual measurement. Ultra-high purity
hydrogen (99.999%) was used. Temperature was con-
trolled to 2560.1 8C with a circulating water bath around
the specimen reactor. Complete isotherms were determined
during hydrogen absorption and desorption for the above
three samples. However, in brief only the data corre-
sponding to the adsorption process were described.
The MoS2 electrodes were prepared by the following
method. In each case, 0.2 g MoS2 (untreated or treated
nanotubes or polycrystalline powders), 0.1 g nickel pow-
der, and 0.02 g polytetrafluoroethylene were thoroughly
mixed and then pressed into a pellet (diameter 1.0 cm)
under the pressure of 200 kg cm22, which was sandwiched
in nickel foam and attached to a nickel tape for connection.
The MoS2 electrode was tested in an open cell by using
a sintered Ni(OH)2 /NiOOH counter electrode and a Hg/
HgO reference electrode in 5 M KOH solution at 25 8C.
Discharge capacities were measured at various current
densities by using an automatic galvanostatic charge–
discharge unit interfaced with a computer. The electro-
chemical performance was investigated by means of
Solartron SI 1260 Potentionstat Analyzer with 1287 Inter-
face and an Arbin charge–discharge unit. The discharge
capacity of the MoS2 nanotubes in any electrode was
based on the amount of the active material (MoS2) not
including the weight of the additives in the electrode.
Fig. 1. XRD patterns of ball milled (NH4)2MoS4 (a) before and (b) after
low-temperature catalytic thermal reaction.
pattern in Fig. 1b coincides well with that of polycrystal-
line MoS2 (ICDD-JCPDS Card No. 39-1492).
The SEM observations show that polycrystalline MoS2
powder consists of micrometer particles (Fig. 2a), but the
samples without and with KOH treatment are characterized
by a large quantity of wirelike nanostructures with typical
lengths of several hundreds of nanometers (Fig. 2b and c).
Fig. 3a is a typical TEM image of the sample without
KOH treatment, showing that the product consists of many
nanotubes. Fig. 3b shows the HRTEM image, which
demonstrates that the nanotube tip is completely open. The
outer diameter of a typical hollow tube is |25 nm, while
the inner diameter is |10 nm. The average distance
between each two neighbouring fringes (c/2) is 0.63 nm,
which corresponds to the interlayer (002) d-spacing of the
2H–MoS2 lattice. It is noted that for the sample with KOH
treatment, more defects were detected around the nanotube
walls (Fig. 3c). Based on the SEM, TEM, and HRTEM
analyses, we estimated the purity of MoS2 nanotubes in
our products to be more than 90 wt%. The chemical
composition of the nanotube was analyzed by energy
dispersive X-ray spectroscopy (EDXS), giving an atomic
Mo/S ratio of 1.0:2.0. The XPS analysis shows that the
chemical valence of molybdenum in the nanotubes is 14
owing to the fact that a binding energy of 228.9 eV, which
can be assigned to Mo3d5 / 2 in MoS2, is detected. There-
fore, the as synthesized sample is MoS2 nanotubes.
3. Results and discussion
3.1. Sample characterization
The XRD patterns of the ball milled (NH4)2MoS4
before and after thermal reaction are shown in Fig. 1.
Comparison of these two diffraction peaks shows that their
features look different. In the XRD pattern of Fig. 1a for
the sample before thermal reaction, the characteristic peak
at 2u517.28 indicates that the phase is orthorhombic
structure (ICDD-JCPDS Card No. 48-1662). The peak
broadening is owing to the very fine grain size and defects
produced during the high-energy ball milling process.
However, the diffraction peaks corresponding to
(NH4)2MoS4 in Fig. 1b disappeared, and new peaks
appeared with relatively strong intensities. The XRD
BET measurements by the nitrogen gas adsorption/
desorption method show that the specific surface area
(SSA) of polycrystalline MoS2 is only 3.6 m2 g21, where-
as the SSA values of untreated and treated MoS2
nanotubes are 22 and 28 m2 g21, respectively. These
results suggest that the SSA of MoS2 nanotubes is much

J. Chen et al. / Journal of Alloys and Compounds 356–357 (2003) 413–417
415
Fig. 2. SEM images of MoS2: (a) polycrystalline, (b) nanotubes without KOH treatment, and (c) nanotubes with KOH treatment.
larger than that of the polycrystalline material and that the
KOH treatment is effective in increasing the SSA, possibly
owing to the production of defects.
The container with the powders was then weighed. There
exists little weight increase (about 0.15%), revealing that
most hydrogen could be desorbed at 25 8C, but some still
remained.
3.2. Gaseous storage
3.3. Electrochemical storage
Fig. 4 shows that the hydrogen storage capacities for the
above three samples from the highest capacity to the
lowest are in the order of MoS2 nanotubes with KOH
treatment.MoS2 nanotubes without KOH treatment.
polycrystalline MoS2. This order illustrates that higher
specific surface areas are responsible for the improved
hydrogen adsorption behavior. When adsorption equilib-
rium was achieved, the hydrogen in the system was vented.
We previously investigated reversible hydrogen adsorp-
tion/desorption through cyclic voltammetric analysis [13].
The results are briefly described. First, a cathodic peak
appeared with the peak position at 20.955 V vs. Hg/HgO,
attributed to hydrogen reduction on the Ni/MoS2 site,
whereas an anodic peak was observed at 20.645 V vs.
Hg/HgO, being assigned to the hydrogen oxidation.
Second, the values of peak currents of hydrogen reduction/
oxidation for MoS2 nanotubes were much higher than that
of the polycrystalline MoS2 electrode. Third, the peak
similarity between the hydrogen reduction and oxidation
indicates that the electrochemical hydrogen adsorption/
desorption proceeds reversibly. Fourth, the peak currents
for hydrogen reduction and oxidation are quite sensitive to
SSA, and they increase as the SSA increases. Conse-
quently, it appears that the electrochemical charge–dis-
Fig. 3. (a) TEM and (b) HRTEM images of MoS2 nanotubes without
KOH treatment, and (c) HRTEM image of KOH-treated MoS2 nanotubes.
Fig. 4. The hydrogen adsorption amount vs. pressure of polycrystalline
MoS2, and nanotubes without and with KOH treatment at 25 8C.

416
J. Chen et al. / Journal of Alloys and Compounds 356–357 (2003) 413–417
Fig. 5. The electrode charge–discharge curves of MoS2 nanotubes (with
KOH treatment) at 25 8C.
Fig. 7. Cycle life of the MoS2 nanotube (with KOH treatment) electrode.
charge mechanism occurring in MoS2 nanotubes is some-
where in between the carbon nanotubes (physical process)
and metal hydride electrodes (chemical process), and
consists of the charge transfer reaction and diffusion step.
Fig. 5 shows the charge–discharge curves of MoS2
nanotube (with KOH treatment) electrode at the 5th cycle
and the discharge current density of 50 mA g21. The
absolute value of the electrode potential increased gradual-
ly during the charging process, while it decreased in the
discharging step. A maximum discharge capacity of 262
mAh g21, corresponding to about 1.0 wt% hydrogen, was
achieved. This capacity is about 83% of that obtained in
the gaseous storage.
mA g21, the electrode discharge capacity was 262
mAh g21. As a comparison, at the discharge current
density of 200 mA g21, the electrode discharge capacity
was 182 mAh g21, which showed a decrease of about
30%.
Fig. 7 shows the change of discharge capacity for the
MoS2 nanotube (with KOH treatment) electrode during
100 cycles. After the test to 100 consecutive cycles of
charging and discharging (100% DOD at 100 mA g21), the
capacity of the nanotube electrode decreased by only about
5%, showing a stable electrode property.
Fig. 6 shows the discharge capacity vs. discharge current
densities for the MoS2 nanotube (with KOH treatment)
electrode. When the discharge current density was 50
4. Conclusions
We have demonstrated that MoS2 nanotubes, which
were synthesized by heating ball-milled (NH4)2MoS4 in
hydrogen/thiophene atmosphere at a relatively low tem-
perature of 400 8C, can reversibly store hydrogen by
gaseous and electrochemical methods. The KOH treatment
of the MoS2 nanotubes produced larger specific surface
area due to the production of more defects. The KOH
treated MoS2 nanotubes showed the highest gaseous
storage capacity of 1.2 wt% hydrogen and the highest
electrochemical discharge capacity of 262 mAh g21 at
25 8C. Our new results show that the alkaline-treated
nanotubes with much higher specific surface areas may be
one reason for the improved hydrogen adsorption/desorp-
tion behavior. There is an indication that this kind of
nanotube (especially the light-transition metal disulfide
nanotubes such as TiS2 nanotubes) may be important for
hydrogen storage. The precise mechanism of the hydrogen
adsorption/desorption is still not understood and further
studies are needed prior to possible applications in effec-
tive energy storage and high-energy batteries.
Fig. 6. The discharge capacity vs. discharge current densities for the
MoS2 nanotube (with KOH treatment) electrode.

J. Chen et al. / Journal of Alloys and Compounds 356–357 (2003) 413–417
417
[3] S. Iijima, Nature 354 (1992) 56.
[4] A.C. Dillion, K.M. Jones, T.A. Bekkadahl, C.H. Kiang, D.S.
Bethune, M.J. Heben, Nature 386 (1997) 377.
Acknowledgements
This work was supported by the Scientific Research
Foundation for the Returned Overseas Chinese Scholars
from State Education Ministry (Grant No. 2002247) and
the Specially Appointed Professorial Award from Nankai
University (Grant No. 20010907) (China). The authors
acknowledge the help and discussions from Dr Q. Xu, Dr
K. Tanaka, Dr N. Kuriyama and Dr. T. Sakai in National
Institute of AIST Kansai Center (Japan).
[5] A.C. Dillion, M.J. Heben, Appl. Phys. A 72 (2001) 133.
[6] V. Meregalli, M. Parrinello, Appl. Phys. A 72 (2001) 143.
[7] R.G. Ding, G.Q. Lu, Z.F. Yan, M.A. Wilson, J. Nanosci. Nanotech-
nol. 1 (2001) 7.
[8] C.C. Ahn, Y. Ye, B.V. Ratnakumar, C. Witham, R.C. Bowman Jr., B.
Fultz, Appl. Phys. Lett. 73 (1998) 3378.
[9] M. Hirscher, M. Becher, M. Haluska, U. Dettlaff-Weglikowska, A.
Quintel, G.S. Duesberg, Y.M. Choi, P. Downes, M. Hulman, S. Roth,
I. Stepanek, P. Bernier, Appl. Phys. A 72 (2001) 129.
[10] G.G. Tibbetts, G.P. Meisner, C.H. Olk, Carbon 39 (2002) 2291.
[11] R. Tenne, L. Margulis, M. Genut, G. Hodes, Nature 360 (1992) 444.
[12] L. Margulis, G. Salitra, R. Tenne, Nature 365 (1993) 113.
[13] J. Chen, N. Kuriyama, H.T. Yuan, H.T. Takeshita, T. Sakai, J. Am.
Chem. Soc. 123 (2001) 11813.
References
¨
[1] L. Schlapbach, A. Zuttel, Nature 414 (2001) 353.
[14] J. Chen, S.L. Li, Q. Xu, K. Tanaka, Chem. Commun. (2002) 1722.
[2] H.W. Kroto, J.R. Heath, S.C. O’Brien, R.F. Curl, R.E. Smalley,
Nature 318 (1985) 162.