CC BY
13H YDROGEN E N ERG Y A N D T RAN SPORT
Hydrogen storage
ENHANCEMENT OF HYDROGEN STORAGE IN COMPOSITE MATERIALS BY NANOSTRUCTURING
J. Kleperis , L. Grinberga, G. Vaivars*, J. Klavins
Member of International Editorial Board
Institute of Solid State Physics of University of Latvia Kengaraga Street, 8, Riga, LV-1063, Latvia Fax: +371-7132778; e-mail: kleperis@latnet.lv
* SA Institute of Advanced Material Chemistry, Univ. Western Cape Private Bag X17, Bellville, 7535, Cape Town, South Africa
To enhance the volume of stored hydrogen in solid hydrogen carriers, the volume and the surface of material must be effectively used. Physicochemical properties are strongly influenced by the composition of material and the dimensions of grains used in the composite. In our work the hydrogen absorption of materials with selected surface and bulk properties were tested. The gravimetric hydrogen uptake of AB5 and AB5 glass phase samples was measured at different pressure steps. AB5 was used as a catalyst and bulk material, but silica based Pyrex glass was used for high surface composite formation.
Introduction
There is an urgent need for the alternative fuel and renewable energy carriers due to the limited supply of fossil fuels on the Earth. Hydrogen is the most abundant element on the Earth; moreover, the chemical energy per weight of hydrogen (142 MJ/kg) is at least three times larger than that of other chemical fuels [1]. Nowadays, research work of scientists of the entire world is focused on the transfer to the hydrogen economy; however, hydrogen storage is still a key problem remaining to be solved. Recent studies demonstrate that solid materials can potentially be utilized to solve the storage problem by reversible ab- and desorption of large amounts of hydrogen [2]. An efficient storage media for hydrogen is desirable for the widespread application of fuel cells and the adoption of hydrogen as an energy source. The U. S. Department of Energy (DOE) has set a target for 2010 of 6.5 % by weight for hydrogen storage for new adsorbent materials [2]. Although several metal hydrides and composite materials are capable of meeting this target, the high desorption temperatures, slow absorption/desorption rates and small cycling capacity limit the widespread application of current metal hydrides. Recent advances in composite carbon materials have been of interest to material researchers [2]. Although ini-
tial hydrogen storage reports on carbon nanotubes indicated a high potential of these materials at moderate temperatures and pressures, the results were irreproducible and have implied the necessity of high pressure and/or cryogenic conditions [2].
Nanostructuring of materials and enhancement of surface absorption capability are two main factors to increase the amount of sorbed hydrogen. One way to combine the effectiveness of hydrogen absorption in metal hydrides and the desirable weight/volume proportion is to make composite material from alloy forming hydride and appropriate support material. The presence of hydride forming alloys in composite does not necessarily mean that the support material is inactive. Recently the role of residual metals in subsequent hydrogen uptake of carbon nanotubes was reported [3, 4]. Hydrogen spillover from metals to carbon and oxide surfaces is showed by M. Boudart et all [5, 6]. The spillover of hydrogen involves a transfer of electrons to acceptors within the support; this process modifies the chemical nature of the support and can also activate a previously inactive material and/or induce subsequent hydrogen physisorp-tion [7]. The hydrogen spillover concepts was established in the 1950s by N. Kobozev [8], and has been reviewed several times [9, 10]. Dissociation of hydrogen molecule on a metal and subsequent spillover of atomic hydrogen to its support is highly
Доклад на Первом Всемирном конгрессе «Альтернативная энергетика и экология» WCAEE-2006, 21—25 августа 2006 г., Волга, Россия.
Paper at the First International Congress "Alternative energy and ecology" WCAEE-2006, August 21—25, Volga, Russia.
dependent upon the chemical bridges formed at the interface, either carbon bridges or proton acceptors [5-7]. Hydrogen spillover can be assessed in a number of ways, but perhaps the most common is simple calculation of the hydrogen to metal ratio, either the surface metal or total metal content. When spillover occurs, the relation H:Mgurface will typically exceed unity. In the case of materials that form hydrides, this relation will exceed the stoichiometric ratio of the hydride.
The objective of this work was to better understand how hydrogen spillover affects hydrogen storage in alloy forming hydrides and on oxide support materials. Mixing a catalytic material with a previously inert secondary material is used to demonstrate hydrogen spillover. This secondary spillover requires intimate contact between the two unlike materials and there may be an energy barrier to transfer hydrogen from one material to another. In this work, we have explored the possibility to use the spill-over effect to enhance catalytic activity and in that way the kinetics and the amount of absorbed hydrogen. The materials with selected surface and bulk properties were milled and tested for hydrogen absorption. LaNi5 and AB5 type (nickel-lanthanum mishmetal) alloy was used as catalyst, and bulk material for hydrogen storage and silica based Pyrex glass — as material with developed surface for adsorption of spilled atomic hydrogen from AB5 grain surfaces.
Experimental
A commercial powder of LaNi5 and AB5 (lan-thanumm rich mischmetal alloy, where A = La, Ce, Pr, Nd and B = Ni, Co, Mn, Al, Cr, purchased from China under tradename 7-10) were used as hydride forming alloys. Structural properties of the samples were studied by X-ray Diffractometer System X-STOE Theta/theta, using KaCu radiation, and the diffraction patterns were analyzed by appropriate software of STOE system. Sample morphology and elemental ratio was determined by Scanning Electron Microscope (SEM) EVO 50 XVP. For alloy 710 (AB5) the constitution was determinate next:
La0.56Ce0.31Nd0.1Pr0.03Ni3.98Co0.47Mn0.38Al0.15Cr0.03.
Samples were prepared by mixing alloy AB5 with grinded Pyrex glass that afterwards was milled for 80 minutes at 15 Hz. The constitution of Pyrex glass used in this work was next (in wt. %): SiO2 — 80.5%; Al2O3 — 2.2%; Cl — 0.10%; B2O3 — 12.9%; K2O — 0.4%; MgO — 0.05%; Na2O — 3.8 %; CaO — 0.10 %; Fe2O3 — 0.04 %. Only one alloy (LaNi5) was milled in Retsch M301 ball mill in tungsten carbide crucible and balls, using weight proportion 10:1 (ball : sample). The composition of samples (composites) used in this work is decoded in the Table 1.
Table 1
Composition of the composite samples used in this work
C1 Glass + AB5 (19:1, weight parts)
C2 Glass + AB5 (9:1, weight parts)
C3 Glass + AB5 (4:1, weight parts)
C4 Glass + LaNi5 milled (4.2:0.4, weight parts)
C5 Glass + AB5 (1:3.7, weight parts)
Pre-treatment of samples for high pressure balance measurements was carried out 3 times by vacuuming the sample chamber, heating it up to 250 °C at 20 atm of hydrogen, cooling and vacuuming again. After last treatment procedure the sample chamber was vacuumed at the room temperature, and measurements were started, when weight/pressure stabilizes. In the volumetric method, the final step in the hydrogenation procedure was the heating of sample in a hydrogen atmosphere (2 atm) up to 170 °C, waiting for pressure stabilization and next cooling by rate 30 degrees per minute, until the room temperature was reached.
Equipment for gravimetric analysis was constructed on the base of Sartorius high-pressure balance in steel pressure container (RIS0 National Laboratory, Denmark). Applied pressures varied from low vacuum (10-3) up to 20 atmospheres at the temperature range from room temperature (RT) to 250 °C. The changes of weight AmH during hydrogenation of an alloy AB5 is due insertion of atomic hydrogen into the alloy, therefore an amount of absorbed hydrogen is calculated for hydrogen gas at normal conditions (NC = 1 atm, RT).
Volumetric method was constructed in the Institute of Solid State Physics of University of Latvia (Latvia), equipped with pressure sensors ECO-1 from Greisinger Electronic GmbH and a copper container with an internal volume of 36 cm3. The changes of the weight during hydrogenation of alloys and composites were calculated from the pressure changes initiated by the absorption/adsorption of hydrogen in the sample. It can be used at constant temperatures and at constant pressure in the container. Calculations were done according to the equation of an ideal gas, where the mass of stored hydrogen gas AmH is proportional to the changes of pressure Ap.
To determine the amount of permanently stored hydrogen in the samples, an automatic differential thermogravimeter DTG-60 (Shimadzu Corporation) with pressure regulator FC-60A in an argon atmosphere (flow 50-75 ml/min) was used. The temperature varied from room temperature up to 500 °C.
Results and discussions
Before measurements, powders of alloys AB5 and LaNi5 were characterized by SEM and X-Ray diffraction analysis. It can be seen on the scanning electron microscope pictures (Figure 1) that starting alloys consist from grains with large (1-2 mm) and small (1-2 |im) dimensions, but after milling the average dimensions were about 1 |im and smaller.
XRD pattern of both alloys shows, that they belong to single phase LaNi5 hexagonal CaCu5-type structure (space group P6/mmm). Only XRD peak intensities of both alloys vary just due to small difference between amounts of compounds in alloys. In order to compare the hydrogen absorption kinetics and amount of absorbed hydrogen, the sample powders of alloy AB5 and composite C5 were pre-treated with sorbtion/desorbtion cycles as described above. Measured weight changes
of the alloy AB5 and composite C5 at 295 K are plotted in the Figure 2, recalculations shows that absorbed hydrogen amount in the alloy AB5 reaches 1,25 wt. %.
(a) Small LaNi5 particles on large grain
1.8 1.6 1.4 1.2 1
0.8 0.6 0.4 0.2 0
1 atm 2 atm 10 atm 20 atm
A\ C5 \ ,----•------
^......
100
200 time, minutes
300
400
showed that the diffraction peaks of hydrogenat-ed alloy AB5 were shifted to the smaller angles comparing with the starting alloy, indicating that the a-phase of hydride is completely converted into the P-phase and the lattice parameters and the cell volume of the hydride is larger than that of the starting alloy (Table 2).
But for the composite C5 the observed shift of XRD peaks after hydrogenation was even larger than that for hydrogenated AB5 alloy! Also corresponding lattice parameters and cell volume for hexagonal P6/mmm symmetry was larger of hy-drogenated composite sample C5 as for fully hy-drogenated alloy AB5 (Table 2). Referring to the [11,12], it could be assumed, that the gamma hydride phase (y) is forming, when the alloy AB5 is mixed in composite with Pyrex glass. In the g phase the hydrogen atoms prefer the interstitial sites closer to Ni atoms in an elementary cell, producing inhomogeneous distribution and large lattice distortions [11]. Nakamura et al [12] found, that the y-phase contains more hydrogen atoms per unit cell than the P-phase.
Structural parameters
Sample a, Â c, Â V, Â3
ab5 5,0083 4,0567 88,12
AB5 hydrogenated 5,326 4,234 104,0
C5 hydrogenated 5,369 4,2754 106,78
(b) LaNi5 after milling — small grains stick together
Fig. 1. SEM pictures of the alloy LaNi5 before (a) and after (b) milling
Fig. 2. Change of weight during hydrogenation of alloy AB5 (A1) and composite C5
For the composite sample C5 the maximal observed absorbed hydrogen reached 1,76 wt.%. From the obtained results and the graph is con-eluded that hydrogenation kinetisc wasn't improved with addition of a glass phase. However, a large plateau region was observed that presents the progress of a change of the solid solution phase (a-phase, the beginning of plateau) ^ solid solution and hydride phases (a-phase + P-phase, plateau) ^ hydride phase (P-phase, the end of the plateau) during hydriding procedure, as expected.
Unexpected result was observed after hydrogenation of the composite C5. The XRD results
Table 2
For the present it is not clear the mechanism, which secures the more deeply hydrogenation of an alloy in the presence of inorganic material (glass phase).
The calculations of absorbed hydrogen for the sample C5 shows that it is more than for pure alloy sample AB5. That could be explained with hydrogen spillover and absorption in alloy-glass mixture. The measurements made with composites C1-C4, using a volumetric method at a pressure range of 0.5-2.5 atmospheres, showed only a small amount of hydrogen uptake (Table 3). Nevertheless the same composite samples after milling intensively at 15 Hz (80 minutes) showed fast hydrogen uptake during heating, and more expressed uptake during cooling. Table 3 summarizes the amount of stored hydrogen measured for the different samples using gravimetric, volumetric and TGA methods. At present it isn't clear while weakly sorbed hydrogen on Pyrex glass don't gives the noticeable weight changes of composite samples.
Metal alloys, which forms hydrides, can serve as a catalyst that binds hydrogen weakly and dissociates H2 readily. In combination with the inert matrix, the small size of the hydride particles maximizes the available surface area of the hydride for both catalysis and hydrogen absorption. The high specific surface area of the matrix provides a large area for hydrogen absorption.
The present study shows that the presence of non-absorbable admixture (glass) is changing the hydrogen absorption kinetics in composite. We tried to explain these changes using simplified scheme
0
Table 3
Measured amounts of stored hydrogen in alloys AB5, LaNi5 and their composites
Sample Sample weight, g Stored hydrogen, measured by gravimetric method, wt% Amount of stored hydrogen, measured by volumetric method Weight loss (%), TGA method
Ap, meas. wt%, calc.
AB5 raw 0,341 1.25
LaNi5 raw 0.340 1.37
C1 milled 4,863 0.05 0.003 +0.102
C2 milled 4,912 0.20 0.013 -0.214
C3 milled 4,988 0.22 0.014 -0.231
C4 milled 4,24 0.35 0.024 -0.26
C5 raw 0,340 1.76
of potential energy of hydrogen molecule/atoms on the surface of composite (Figure 3, according to A. Zuttel [2]). The physically adsorbed hydrogen molecule has overcome an activation barrier for dissociation on the surface of metal cluster. The formed hydrogen atoms has two options: 1) to diffuse in the bulk of metal (hydride is formed with characteristic hydrogen metal bond) or 2) to spill over from the metal surface to neighbor carrier material - oxide or carbon. The height of the activation barrier for spilled hydrogen atoms is less as for chemisorbed hydrogen atoms, and can arise from specific surface states characteristic to nanosize metal clusters. The smaller are grains, the larger is surface and lower activation barrier for spillover.
300
200
S
Ü2
100
-100
Atomic hydrogen sorbed in metal hydride
^ d
Fig. 3. When hydrogen molecule approaches the surface of composite material, the potential energy of hydrogen atoms depends from the distance and material, consisting from small alloy LaNi5 clusters and support material (picture substituted from [2])
The nanostructuring of traditional and new materials is perspective approach that will provide the material properties different from the bulk material properties. In nanomaterials the surface interactions will dominate rather than bulk, and slow hydrogen diffusion will be replaced by quick spillover onto developed surfaces. Considerable promise for hydrogen storage could be demonstrated using hydrogen spill over effect onto different carriers: nanosize glass particles, nanotubes and zeolites. Typically hydrogen spillover is observed in hydrogen-catalyzed reactions on supported metal cata-
lysts [13], when hydrogen molecules (H2) dissociate on the metal part of the catalyst (H2 ^ 2Ho, Ed = 435.99 kJ/mol). Some hydrogen atoms remain attached to the metal, whilst others diffuse to the support and are said to spillover. Spillover hydrogen has often been inferred from hydrogen adsorption and reactivity studies. However, no direct measurements have been reported, except recently reported results from inelastic neutron scattering [14]. These experiments are confirming the presence of the spillover hydrogen on the Pt/Ru carbon-supported catalyst from surface vibration states of the C-H bond through the H riding modes. The effect of the catalyst and carrier material on hydrogen spillover is not understood, and much research is performed worldwide for better understanding, mostly in catalysis reactions.
For hydrogen spillover to occur, it is well established that appropriate contact between the metal and carbon must be present [5, 6]. The electron transport and electric conduction processes would be important parts in hydrogen spillover processes. Therefore dry mixing of the catalyst AB5 and the glass substrate did not enhance hydrogen storage in our case (samples C1-C4). As discussed previously, the hydrogen spillover is certainly a function of the substrate-catalyst contact, and this can be optimized via different methods — using carbon, conductive polymer and other materials as substrates, high-temperature pre-treatment of samples before hydrogenation etc. The next experiments are planned with carbon, xerogels and aerogels, which normally absorb only a small amount of hydrogen, but a combination with the hydride alloy could produce an unexpected synergistic effect, with the composition capable of storing surprisingly large amounts of hydrogen, more than the mathematically combined capacity of the xero-gel and the hydride alloy separately.
Conclusions
Hydrogen uptake in the composite alloy/glass (larger in sample C5, smaller in samples C1-C4) could be explained with spillover over the glass phase due to the AB5 catalytic properties. The following mechanism can be deduced: the hydrogen chemisorbs to surface sites found on the AB5 (mostly Ni sites). Bridges between the catalyst and glass particles allow the chemisorbed hydrogen to mi-
0
grate onto the glass surface. Desorption occurs directly from the relatively lower energy glass sites without migration back to the catalyst. Hydrogen spillover depends upon the glass-catalyst contact. The contact between particles changes with the quality of the mixing and milling, as well as the position of AB5 grains in the mixture.
It was observed from the X-ray diffraction patterns, that during the hydrogenation of the composite sample C5 (AB5 alloy + grinded glass), the beta phase occurred much faster than in the AB5 sample and the gamma phase appears that was not observed in the pure alloy ABg. The results from the gravimetric and volumetric hydrogen uptake measurements in both AB5 alloys and composite samples showed that for the AB5 alloy it is possible to achieve the predicted amount of absorbed hydrogen, but for the composite sample the amount of absorbed hydrogen can be greater.
Acknowledgments
Author (L. G.) acknowledges the European Union European Social Fund for support. Authors (L. G. and J. K.) acknowledge the NORSTORE project (Nordic Energy Research Project) for possibility to work in the RIS0 National Research Centre (Denmark). The Latvian State Research Program in Material Science VPP-05 is acknowledged tightly for support in this research.
References
1. SchlapbachL., Zuttel A. Nature 414. 2001. P. 353.
2. Zuttel A. Materials Today. September, 2003, P. 24.
3. Lueking A. D., Yang R. T. Applied Catalysis A: General 265. 2004. P. 259-268.
4. Lueking A. D., Yang R. T. AIChE J. 49. 2003. P. 1556.
5. Boudart M., Aldag A. W., Vannice M. A. // J. Catal. 18. 1970. P. 46.
6. Vannice M. A., Boudart M., Fripiat J. J. // J. Catal. 17. 1970. P. 359.
7. Roland U., Braunschweig T., Roessner F. // J. Mol. Catal. A Chem. 127. 1997. P. 61.
8.KobozevN. // J. Phys. Chem. (Russia). 1952. Vol. 26. P. 112.
9. Kleperis J., Lusis A. // Z. Phys. Chemie, Vol. 181. 1993. P. 321.
10. Conner W. C., Falconer J. L. // Chem. Rev. 95. 1995. P. 759.
11. Sohmura T. et al. // J. Phys. F: Met. Phys., 8. 1978. P. 2061.
12. Nakamura Y., NomiyamaT., AkibaE. // J. Alloys and Comp., 413. 2006. P. 42.
13. Chen B., Falconer J. L. // J. Catalysis, Vol. 134, 1992. P. 737.
14. Mitchell P. C. H., Ramirez-Cuesta A. J., Parker S. F., Tomkinson J. Web publication: http:/ /www.isis.rl.ac.uk/isis2002/highlights/ 19_HydrogenSpillover.htm (2002).
Pfwid.nj; 10 Of «Aîdltftie [в th*
chI A pi jmtrttfy
8th ANNUAL MIDDLE EAST REFINING CONFERENCE
4-5 February 2007, Abu Dhabi, UAE
The World Refining Association is delighted to announce their 8th Annual Middle East Refining Conference will take place from 4-5 February 2007 in Abu Dhabi.
Co-hosted with TAKREER, the event will present a complete and up to date overview of the Middle Eastern refining sector focusing upon strategic development programmes, which continue to proceed at a rapid pace, ongoing policy developments to consistently upgrade, integrate, modernise and expand refining facilities, enabling them to remain competitive whilst at the same time meeting international standards.
Continuing to examine the key issues that are crucial to the future development of the region, the 8th Annual Conference will attract key organisations and senior decision makers involved in the refining industry, providing attendees with exceptional networking, business and training opportunities.
MIDDLE EAST FUELS SYMPOSIUM
Co-hosted with ADNOC DISTRIBUTION, the World Refining Association are pleased to announce the inaugural MIDDLE EAST FUELS SYMPOSIUM taking place from 6-7 February 2007 in Abu Dhabi which has been established to address the important and growing issue of clean fuel production within the region.
High on the agenda of the Symposium is the gathering of <5 experts in their fields to facilitate increased communica- * tion and networking opportunities within the region be- s tween all levels of managers and employees, allowing 3 you to Listen, Question, Meet and Exchange Ideas di- | rectly with them. &
Make the most of your visit to Abu Dhabi and the 8th 1
Annual Middle East Refining Conference by extending £
your stay and network with fuel executives and discov- |
er the latest developments and business opportunities S
that this exciting and fast growing sector can bring to g
your business. S Discounts available for attending both events.
71 ie World Refining Association, 25 St George's Road, Cheltenham,
Gloucestershire GL50 3DT
+44 (0)1242 529 090 or +44 (0)207 067 1B00
+44 ( 0)1242 582 157 or +44 ( 0)207 430 9513
www.wraconferences.com. c.pallen@theenergyexchange.co.uk
http://www.wraconferences.com/wra113a_overview.html
