Academia.eduAcademia.edu

New sorbent materials for the hydrogen storage and transportation

2007, International Journal of Hydrogen Energy

International Journal of Hydrogen Energy 32 (2007) 5015 – 5025 www.elsevier.com/locate/ijhydene New sorbent materials for the hydrogen storage and transportation L.L. Vasiliev a,∗ , L.E. Kanonchik a , A.G. Kulakov a , D.A. Mishkinis a , A.M. Safonova b , N.K. Luneva b a Luikov Heat & Mass Transfer Institute, National Academy of Sciences, P. Brovka, 15, 220072, Minsk, Belarus b Institute of General and Inorganic Chemistry Minsk, Republic Belarus Surganova, 9, 220606, Minsk, Belarus Received 2 September 2005; accepted 12 July 2007 Available online 14 September 2007 Abstract Activated carbon fiber was chosen as an efficient gas (ammonia, methane, hydrogen) sorption material to design a gas storage system. To increase gas sorption capacity a complex compound (activated carbon fiber + chemicals) was applied. The application of a heat pipe in gas accumulator enables one to control the temperature of sorbent bed and provide optimum operational conditions. 䉷 2007 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights reserved. Keywords: Carbon; Ammonia; Methane; Hydrogen; Heat pipe; Sorption; Metal hydride/chloride 1. Introduction Hydrogen is the ideal alternative for fossil fuel systems. From environmental point of view, it is the cleanest fuel known until recently, and from economic point of view hydrogen technology will be able to revolutionize the transport and energy market. Hydrogen is the smallest and lightest molecule known. Among the advantages of hydrogen are its low density and small volume heat of combustion, which leads to the necessity of hydrogen storage in large-sized vessels and cylinders. According to the standard “The US Department of Energy (DOE) Hydrogen Plan” the modern onboard system of hydrogen storage should have capacity of storage 6–7 wt.% at average pressure (2–6 MPa) and temperature 273–363 K; capacity of 5 kg, and specific energy of 7.2 MJ/kg. None of the existing methods of hydrogen storage is efficient in terms of energy density, of neither a volume nor a mass and gas release rate at same time. From the application point of view it is vital to find the most effective way to store hydrogen and then, to replace the current fuel systems. ∗ Corresponding author. Luikov Heat & Mass Transfer Institute, National Academy of Sciences, P. Brovka, 15, 220072, Minsk, Belarus. Tel.: +375 172 84 21 33; fax: +375 172 84 21 33. E-mail address: [email protected] (L.L. Vasiliev). Hydrogen storage by solid materials is the most recent system proposed [1]. Initially the research was based on cryogenic systems [2], which are unprofitable from an economic point of view. Recently, the studies have focused in the search of the high capacity adsorbent to be used at room temperature. Due to the high density of the adsorbed phase (it corresponds to density of pressurized gas at 60–80 MPa) the gas storage capacity in the vessel with sorbent can increase significantly. Since it is close to the liquid density, the volumetric capacity of adsorption system is predicted to be higher then for compression system. Activated carbons, activated carbon fibers and graphite nanofibers are perspective candidates for hydrogen adsorption storage. Many metals and alloys can also reversibly adsorb large amounts of hydrogen. However, none of them is known for the mobile storage pressure-temperature range with H in the range of 15–24 kJ/mol hydrogen. Our approach is to make composite materials containing two or more different components, in an effort to compensate the limitations of each. Activated carbon fiber with chemicals can be considered as a promising material for hydrogen storage. The objective of this work is to analyze hydrogen storage in several porous carbon-based materials with different porous structures to propose perspective activated carbons (carbon fibers) and metal hydrides compositions for high performance hydrogen storage system. Another interrelated work objective is 0360-3199/$ - see front matter 䉷 2007 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.ijhydene.2007.07.029 5016 L.L. Vasiliev et al. / International Journal of Hydrogen Energy 32 (2007) 5015 – 5025 Nomenclature a av c C Cg Ca E G gi Ks0 M Mi m N P qst Q r and z R R0 r0 and r1 R Rp 2S T W0 v a Va adsorption capacity, kg/kg volume capacity of hydrogen storage using physisorption, ml(STP)/g density of free gas, kg/m3 solid sorbent specific heat capacity, J/(kg-K) specific heat capacity of free gas, J/(kg-K) specific heat capacity of adsorbed gas, J/(kg-K) activation energy, J/kg gas output from the cylinder, kg/s, g/s gas output from the elementary cell, used for computer modeling, kg/s pre-exponent constant in the equation of kinetic of sorption, s−1 mass of the gas in the cylinder, kg mass of the gas in the calculated cell, kg dynamic coefficient of filling the cylinder number of calculated cells in the cylinder pressure, Pa latent isosteric heat of sorption, J/kg heat flow, W cylindrical coordinates, m external radius of the cylinder shell, m internal radius of the heat pipe, m internal and external radii of the annular layer of the sorbent, m gas constant, J/(kg-K) mean radius of the particles, mm fins step, m, mm temperature, K, ◦ C maximum microporous specific volume, m3 /kg component of the velocity vector, m/s specific volume of adsorbed medium, m3 /kg partial molar adsorption volume development of thermally regulated adsorption storage system for dual-fuel (hydrogen and natural gas) automobile. In fact, such gases as methane, ammonia, methanol can be considered as hydrogen storages by itself also. 2. Carbon materials as a storage medium for the most current gases Activated carbon is well known as one of the best adsorbents for gases [3]. In contrast to the chemisorption in metal hydrides [4], the phenomenon of physical adsorption is essentially accumulation of the undissociated hydrogen molecules on a surface of microporous carbon fibers or particles. This property is due to ability of carbon to be prepared in a very fine powdered or fiber form with highly porous structure and due to specific interactions between carbon atoms and gas molecules. The total amount of adsorbed hydrogen strongly depends on the pore geometry and pore size distribution as well as the Vg zg molar gas-phase volume coefficient of gas compressibility Greek letters  ε 2  s  v  heat transfer coefficient, W/(m2 -K) porosity determined as a part of the volume occupied by the free gas (not bound by adsorption) fin thickness, m, mm effective thermal conductivity of the sorbent layer, W/(m-K) density of the solid sorbent, kg/m3 total density of the free and adsorbed gases in the cylinder, kg/m3 volume density of storage, ml(STP)/ml time, s Subscript a cr e env eq hp 0 s f t adsorbate critical state finite value environment equilibrium conditions heat pipe initial value sorbent fin transfer Abbreviations HPP STP WAC heat pipe panels standard of temperature (273 K) and pressure (0.1 MPa) wood-based active carbon storage pressure and temperature. Recently many improvements have been accomplished to obtain microporous carbonaceous materials with extremely high adsorbing properties for different gases [5]. Adsorption of methane and hydrogen usually takes place in micropores. Macropores have no practically influence on the adsorption capacity, as they are only important for the gas compression and for adsorption/desorption reaction rates [6]. Due to its high surface area and abundant pore volume, activated carbon is considered as good adsorbent. For conventional activated carbon, the hydrogen uptake is proportional to its surface area and pore volume, while, unfortunately, a high hydrogen adsorption capacity (4–6 wt.%) can be only obtained at very low cryogenic temperatures according to theoretical calculations. To gain the goal of the problem—efficient ammonia, methane, and hydrogen storage and transportation it is necessary to develop a high performance microporous adsorbent material and an advanced system of the vessel thermal control. L.L. Vasiliev et al. / International Journal of Hydrogen Energy 32 (2007) 5015 – 5025 5017 Fig. 1. Active carbon material made from waste wood (IGIC NASB): (a) Image multiplied by 30 times; (b) by 1000 times. Fig. 2. Active carbon fiber “Busofit”: (a) Image multiplied by 50 times; (b) by 1000 times; (c) by 50 000 times. Thus, the cheap activated carbon fabricated by special thermal treatment of impregnated raw (wood, sawdust, cellulose, straw, paper for recycling, peat, etc.) is attractive for modern sorption technologies. The use of specific organic and non-organic compounds as raw impregnates offers production activated carbons with controlled porous structure and high yield (up to 50 wt.%). Developed advanced technology allows to produce the homogeneous carbon adsorbents with benzene pore volume 0.3–0.6 cm3 /g (70–80%—volume of micropores), nitrogen surface area up to 1500 m2 /g, iodine adsorption capacity 40–70 wt.% and methane adsorption capacity up to 160 mg/g (3.5 MPa, 293 K). Impregnated cellulose—containing raw for manufacture special activated carbon materials for ammonia, methane and hydrogen storage systems with high microporosity, surface area and narrow micropore size distribution is the attractive host material for adsorption of different gases. The activated carbon fiber “Busofit” and activated wood-based carbon particles ([7], the Institute of General and Inorganic Chemistry, NASB) fabricated in Belarus are perspective materials for gas storage systems (Figs. 1–2). “Busofit” is a universal adsorbent, which is efficient to adsorb different gases (H2 , N2 , O2 , CH4 , and NH3 ). Fig. 2 shows the texture of the active carbon fiber filament. The carbon fiber refers to microporous sorbents with a developed surface and a complicated bimodal structure. The material can be performed as a loose fibers bed or felt or as monolithic blocks with binder to have a good thermal conductivity along the filament. In our experiments some samples of activated carbon “Busofit” obtained by the new technology were investigated. The surface area of the commercially available “Busofit” was measured with “Micromeritics AccuSorb 2100” and BET Sorbtometer NOVA and varied from 1140 up to 1570 m2 /g. Now it is clear, that ammonia, methane and hydrogen storage vessels filled with “Busofit” have certain advantages (for example, methane storage capacity near 170 v/v). To be commercially profitable the adsorption storage is required to have at last 150 v/v. It can be considered as a typical microporous adsorbent with pore diameter near 1–2 nm and at the same time as material with high gas permeability. The micropore distribution is performed mostly on the carbon filament surface. Nowadays a program was undertaken to examine the parameters of an active carbon fiber to optimize both the mass uptake of ammonia, methane and hydrogen and the carbon density. “Busofit” has such advantages as • high rate of adsorption and desorption; • uniform surface pore distribution (0.6–1.6 nm); • small number of macropores (100–200 nm) with its specific surface area 0.5 m2 /g; • small number of mesopores with its specific surface area 50 m2 /g. 5018 L.L. Vasiliev et al. / International Journal of Hydrogen Energy 32 (2007) 5015 – 5025 The total volume V, associated with an active carbon adsorbent may be split up into its components: V = Vc + Vv + Vvoid + V , where Vc is the volume of the carbon atoms of which the adsorbent is composed; V the micropores volume; Vv the meso- and macropores volume; Vvoid the space inside the vessel free from adsorbent bed. This latter Vvoid can be eliminated by making the solid block of adsorbent. 3. Experimental research of texture and hydrogen-sorption capacity for the activated carbon materials Activated carbon fibers (commercially available “BusofitAYTM” and modified grades of “Busofit-AYTM”) and granular activated carbon (additionally activated in Porous Media Laboratory of Luikov Heat & Mass Transfer Institute and commercially available) have been used in this work (Table 1). Porous texture of the different materials was all characterized using nitrogen (N2 ) physisorption at 77 K and up to a pressure of 0.1 MPa. From the nitrogen physisorption data, obtained with the High Speed Gas Sorption Analyser NOVA 1200, the BET-surface area, total pore volume, microporous volume and t-volume were derived. The BET surface area (SBET ) is the surface area of the sorbent according to the model formulated by Brunauer et al. [8] for planar surfaces. The micropore volume is defined as the pore volume of the pores < 2 nm. Microporous volumes calculated from the application of the Dubinin–Radushkevich equation to the N2 adsorption isotherms at 77 K. The mean pore size of each sample obtained from N2 adsorption was determined by applying Dubinin–Radushkevich equation. The hydrogen sorption isotherms were measured with the High Speed Gas Sorption Analyser NOVA 1200 at 77 K in the pressure range 0–0.1 MPa. Table 1 summarizes the results of the N2 and H2 physisorption measurements of the materials analysed. All samples are highly micro-and mesoporous carbon materials. In our experiments four samples of carbon “Busofit-AYTM” (1–4) and three samples of wood-based activated carbon (5–7) obtained by new technology were investigated. The activated carbon 207C (8) is made in the Great Britain from coconut shell. Samples 9 and 10—granular activated carbons, specially developed for effective storage of methane. According to the offered technology some samples from “Busofit–AYTM” have been prepared by selective thermal processing at high temperature 850 ◦ C. In this way some of the carbon atoms are removed by gasification, which yields a very porous structure. Numerous pores, cracks were formed in the carbon material increasing a specific surface area due to the growth of micropore volume. Additional activation of a sample 1 was carried out at the presence of oxygen. As follows from Table 1 the increase of time of activation from 2 h until 8 h in an atmosphere of carbonic gas promotes increase to sorption capacity almost in 1.5 times (samples 2 and 4). The atmosphere of carbonic gas appeared more preferably oxygen for growth of a specific surface and sorption capacity—time of activation of samples 1 and 2 was identical. To increase the adsorbent capacity and the bulk density of material we compressed active carbon fiber together with a binder. Briquet “Busofit” disks have a high effective thermal conductivity and a large surface area. Wood-based carbons (5–7) were produced by controlled pyrolysis of waste wood and special activation. As seen in Table 1, the greatest values of a surface area and micropore volume among carbon fibrous materials has “Busofit-M8”; from wood-based activated carbons it is stand out “WAC 3-00”, from granular carbons—“Sutcliff”. The approach of research institutes AGLARG [9] was used for an operative estimation of gas sorption capacity for carbon sorbents. According to it micropore volume and the specific surface area have been chosen as determining parameters. To obtain the function approximating dependence of hydrogen sorption capacity on carbon materials from value of a specific surface area (at pressure 0.1 MPa and temperature 77 K), we used our experimental data (Table 1) and an experimental database (Table 2) of group of institutes—Inorganic Chemistry and Catalysis, Debye Institute, Utrecht University [10]. Table 1 Textural characteristics and hydrogen-sorption capacities at 77 K and 0.1 MPa for the researched carbon materials No. Sorbent av (ml/g) a (wt%) SH (m2 /g) SBET (m2 /g) SDR (m2 /g) VDR (ml/g) Vt (ml/g) ´ RDR (Å) 1 2 3 4 5 6 7 8 9 10 Busofit 191-5 Busofit-M2 Busofit-M4 Busofit-M8 WAC 97-03 WAC 19-99 WAC 3-00 207C Norit sorbonorit-3 Sutcliff 199.9 203.9 225.1 252.9 115 172.1 221.1 209.2 193.8 236.6 1.76 1.79 1.98 2.23 1.01 1.51 1.95 1.84 1.71 2.08 462 465 536 571 271 393 575 502 458 527 1691 1702 1715 1939 715 1005 1383 1300 1361 1925 2496 2507 2547 2985 1050 1486 2142 1944 2044 2864 0.887 0.89 0.9 1.04 0.37 0.53 0.74 0.69 0.73 1.02 0.234 0.43 0.42 0.27 0.33 0.44 0.22 0.37 0.26 0.254 49.9 41.5 42 51 33.4 41.7 50 41 50 53.6 Note: av , volume capacity of hydrogen storage using physisorption, ml/g; a, capacity of hydrogen storage using physisorption, wt%, g/100 g; SH , BET surface area determined on hydrogen, m2 /g; SBET , BET surface area determined on nitrogen, m2 /g; SDR , surface area, determined on Dubinin–Radushkevich method, m2 /g; VDR , micropore volume, determined on Dubinin–Radushkevich method, ml/g; Vt —mesopore volume, determined on t-method, ml/g; RDR , size of pore, ´. determined on Dubinin–Radushkevich method, Å 5019 L.L. Vasiliev et al. / International Journal of Hydrogen Energy 32 (2007) 5015 – 5025 Table 2 Textural characteristics and hydrogen-sorption capacities at 77 K and 0.1 MPa for carbonaceous materials [10] Sorbent SBET (m2 /g) VDR (ml/g) av (ml/g) av,meso (ml/g) av,micro (ml/g) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Synthetic graphite Large-diameter CNF Activated graphite 100 Medium-diameter CNF1 ACF 400 ACF 1200 AC Norit 990721 AC Norit ROZ 3 Activated graphite 300 Medium-diameter CNF2 AC Norit SX 2 ACF 500 AC Norit UOK A AC Norit SX 1 AC Norit SX 1G AIR AC Norit GSX AC Norit SX plus AC Norit SX 1 G AC Norit 990293 AC Norit Darco KB Hyperion CNF 7 49 119 120 883 899 988 287 287 65 841 988 1195 922 1030 933 1051 1176 2029 1462 238 0.00 0.01 0.02 0.00 0.34 0.37 0.43 0.05 0.05 0.00 0.27 0.40 0.47 0.31 0.36 0.26 0.35 0.40 0.92 0.42 0.00 0 6 14 12 143 184 142 36 36 7 150 142 188 168 171 161 165 187 238 146 22 0 2 6 11 1 1 2 6 16 7 17 15 10 18 16 27 21 20 7 54 22 0 4 8 1 142 183 140 28 19 0 133 127 178 150 155 134 144 167 231 92 0 400 600 300 450 av, ml/g av, ml/g No. 200 100 0 0 500 1000 1500 SBET, m2/g 2000 150 0 2500 0 0.5 1 1.5 VDR, ml Fig. 3. Volume capacity of hydrogen storage for carbon sorbents vs. BET surface area at pressure 0.1 MPa and 77 K: •—experimental data (Table 1), a continuous line—the linear approximation obtained by authors; —experimental data (Table 2), a dashed line—the linear approximation given in [10]. A test matrix of about 20 different carbon samples, including commercial fibers and fiber composites, nanofibers with conical or tubular graphite planes was assembled. The sorbents were chosen to represent a large variation in surface areas and micropore volumes. Both non-porous materials, such as graphites and micro- and mesoporous sorbents, such as activated carbons, were selected. Characterization via N2 adsorption at 77 K was conducted on the majority of the samples; for this a Quantachrome Autosorb-1 system was used. The results of the N2 and H2 physisorption measurements are shown in Table 2. In the table CNF is used to designate carbon nanofibers, ACF is used for activated carbon fibers and AC for activated carbon. For carbon samples we have found linear relationship between BET surface area and volume sorption capacity of hydrogen at 77 K and 0.1 MPa as: av = 0.0783 SBET + 84.02. 300 (1) Fig. 3 shows all experimental data and variants of approximation. It is possible to see that our linear correlation fairly good Fig. 4. Volume capacity of hydrogen storage for carbon sorbents vs. micropore volume (determined on Dubinin–Radushkevich method) at pressure 0.1 MPa and 77 K: •—experimental data (Table 1), a continuous line—the linear approximation obtained by authors; —experimental data (Table 2), a dashed line—the linear approximation given in [10]. describes all results except for the literature data which correspond to materials with low values hydrogen sorption capacity (av < 50 ml/g). Influence of micropore volume on sorption capacity of hydrogen for various carbon materials under the chosen conditions (77 K, 0.1 MPa) is well described by the following linear correlation: av = 119.12 VDR + 115.41. (2) From Fig. 4 it can be concluded that this correlation does not apply to the carbon samples with low values sorption capacity of hydrogen (av < 50 ml/g). “Hydrogen” sorbent should have maximum high value of a specific surface area and maximum narrow distribution of micropores. As follows from Table 1, sorbents “Busofit-M8”, “Sutcliff” have micropore volume more than 1 ml/g. They are also the best storage systems for hydrogen (accordingly 253 and 237 ml(STP)/g) due to physisorption. Our results demonstrate 5020 L.L. Vasiliev et al. / International Journal of Hydrogen Energy 32 (2007) 5015 – 5025 that a large capacity of adsorbed hydrogen by physisorption under chosen conditions is obtained with sorbents containing a large volume of micropores and a high BET surface area. Optimization of sorbent and sorption conditions is expected to lead to storage capacity of 500–600 ml(STP)/g, close to targets set for mobile applications. 4. The analysis of isotherms of gas sorption on carbon materials 4.1. Experimental set-up The analysis of isotherms of gas at the temperature interval 233–293 K and pressure interval 0.1–6 MPa was realized by the gravimetric control of the sample (200–400 g) during adsorption/desorption cycle. The experimental set-up is shown in Fig. 5. The full amount of the adsorbate was measured. The cylindrical experimental rig of 47 mm inner diameter and 540 mm length was used to simulate full-scale conditions of the experiment. This experimental rig was made from the stainless steel and served as a simulator of the real vessel in the ratio 1:50. Thermocouples were attached to the sample through the sorbent bed (12.5 mm thick). Pressure sensor was used for pressure measurements in the experimental chamber. 4.2. Hydrogen and methane isotherms for a high surface activated carbons The rate of the adsorption/desorption of different gases (ammonia, methane, hydrogen) on the surface of “Busofit” can be evaluated by the isotherms analysis at different temperatures of the sorbent bed. In order to study the sorption capacity of the adsorbent it is necessary to know the quantity of gas adsorbed on each point of the cycle. There is a general need to have a good fit of experimental isotherms and temperature and to extrapolate some isotherms beside the experimental field. For the carbon fiber “Busofit” the approach of Dubinin is well adapted and allows linking quite simply the physical properties of “Busofit” to the capacity of adsorption of the carbon fiber. Methane isotherms evolution during the cycle of adsorption/desorption of “Busofit AYTM” [11,12] is shown in Fig. 6. Based on these data we can conclude that “Busofit-AYTM” is competitive to best activated carbons with methane adsorption capacity 0.113–0.135 kg/kg at 273 K. As an example, Fig. 7 shows the experimental isotherms of the adsorption and desorption of hydrogen on carbon fiber “Busofit-M8” and wood-based activated carbon “WAC 3-00” at the nitrogen temperature. Absence of an appreciable hysteresis confirms that reversible physisorption exclusively takes place with all investigated materials. Physisorption on microporous carbons can be described with the Dubinin–Radushkevich equation [13]     R T ln(Psat /P ) 2 W0 aeq = . (3) exp − va E The theory of microporous volume filling, worked out by Dubinin, is widely used for quantitative characteristic of adsorptive properties and basic varieties of porous structure. Equilibrium state Eq. (3) includes the saturation pressure Psat . Since the hydrogen sorption isotherms are measured within the Fig. 5. Experimental apparatus: 1—AG cylinder, 2—electronic balance, 3—insulated chamber, 4—heat exchanger, 5—thermostat, 6, 7—computer with software, 8, 11—pressure sensors, 9, 15—20—valves, 10—calibrated volume, 12—vacuum pump, 13—helium and hydrogen vessels, 14, 15—reducer, 21—flow meter, 22, 24—fans, 23—ventilation gate, 25—thermocouples, 26—perforated pipe, 27—sorbent bed, 28—heat exchanger. 5021 L.L. Vasiliev et al. / International Journal of Hydrogen Energy 32 (2007) 5015 – 5025 av, ml (STP )/g 600 500 400 300 200 100 0 Bus-M8 Bus-M8 207C Bus-M2 Sutcliff WAC 77 153 193 293 K Fig. 8. Volume capacity of hydrogen storage using physisorption at pressure 6 MPa for active carbon materials and different temperatures. Fig. 6. Methane sorption isotherms for active carbon fiber “Busofit”: experimental data—points; calculated data (Dubinin–Radushkevich equation)—lines. av, ml (STP)/g 300 250 200 150 Bus-M8. ads. Bus-M8. des. WAC 3-00 ads. WAC 3-00 des. 100 50 0 0.0 0.4 0.8 1.2 P, MPa Fig. 7. Isotherms of hydrogen adsorption and desorption on carbon materials at temperature 77 K, measured by the High Speed Gas Sorption Analyser NOVA 1200. Table 3 The empirical coefficients of Dubinin–Radushkevich equation for hydrogen sorption on the carbon materials Material W0 (ml/g) E (J/g) Busofit-M2 Busofit-M4 Busofit-M8 Sutcliff WAC 3-00 207 C 369 376 482 453 270 343 1783 1922 1710 1699 3782 1969 temperature and pressure intervals comprising the regions of supercritical states of the adsorptive (Tcr =33.24 K Pcr =1.298 Ì Pà), the notion of saturation pressure loses its physical meaning. In the work [14] the saturation pressure is determined by the formula: Psat = Pcr (T /Tcr )2 . (4) As a results of the experiments, we obtained hydrogen sorption isotherms for different carbon materials and empirical coefficients for the Dubinin–Radushkevich Eq. (3), presented in Table 3. A summary of results on volumetric capacity of storage of hydrogen for the investigated materials and the influence of temperature can be seen in Fig. 8. The storage capability is increasing for lower temperatures. It is visible, that the best sorbent in the field of nitrogen temperature—the activated fiber ”Busofit-M8”, having major effective porosity (0.78), bulk density of 500 kg/m3 and an advanced surface area (2985 m2 /g). At pressure 6 MPa its capacity of a storage due to a physical sorption achieves 482 ml/g. Apparently, that after-treatment of a carbon fiber by carbonic gas within 8 h has given in modification both microstructural, and the surface characteristics. Finally, the hydrogen sorption of a carbon fiber at cryogenic temperatures was improved almost twice, in comparison with samples “Busofit-M2” and “Busofit-M4” whose activation was prolonged 2 and 4 h. It is observed that the H2 uptakes are very low at room temperature and pressure 6 MPa for all carbon samples, but the wood-based material of large surface area has the best sorption characteristics—storage capacity equal of 80 ml(STP)/g. Fig. 9 compares the behaviour of the adsorption isotherms at different temperature levels for two of the more promising samples: steam activated “Busofit-M8” and wood-based carbon “WAC 3-00”. The shape of the isotherms in the two cases is dissimilar. The isotherms for the 77 and 153 K exhibit a classical type 1 isotherm shape indicating a microporous material. The isotherms at room temperature exhibit a much less pronounced curvature (more like type II isotherm). As is seen from plots (Figs. 6 and 9) experimental data fit the calculated adsorption values (Dubinin–Radushkevich equation) with an error sufficient for practical purposes. 5. Active carbon fiber and chemical components To minimize a void space and increase the adsorbent capacity of the active carbon fiber we need to compress “Busofit” together with a binder. If we combine the carbon material with metal hydride/chloride microcrystals disposed in the same volume, we can solve the problem of efficient gas storage and transportation. The sorbent complex compound microstructure obtained in the Luikov Heat & Mass Transfer Institute has been studied using scanning electro microscope (SEM). The characterization 5022 L.L. Vasiliev et al. / International Journal of Hydrogen Energy 32 (2007) 5015 – 5025 1 500 300 1 2 400 3 av, ml (STP)/g av, ml (STP)/g 2 300 3 200 200 100 4 100 4 0 0 0 2 4 6 0 P, MPa 2 4 6 P, MPa Fig. 9. Hydrogen adsorption isotherms for active carbon fiber “Busofit-M8 ” (a), wood-based cardon “WAC 3-00” (b) and different temperatures (1–77, 2–153, 3–193, 4–293 K): experimental data—points, calculated data (Dubinin–Radushkevich equation)—lines. Fig. 10. “Busofit” filaments covered by CaCl2 microcristals. Image multiplied by 5000 times (a) and 10 000 times (b). should reveal a porous structure with a uniform distribution of salts without formation of agglomerates or distribution of particles with different dimensions [15]. Fig. 10 shows that the structure of the salt around the fiber is porous which is favorable for heat and mass transfer. The results of the experimental analysis of the sorption capacity of the active carbon fiber “Busofit” and fiber “Busofit”+chemicals are presented in the Table 4; the data are obtained for the room temperature, except methane (253 K). Fig. 11 shows ammonia sorption isotherms for the active carbon fiber “Busofit” and complex compound “Busofit+CaCl2 ”. For “Busofit” it is typical a micropore distribution mostly on the filament surface. A set of microcrystals of metal hydrides/chlorides is attached on the filament surface also. Combination “Busofit”+metal hydrides have some particularities to compare with the combination “Busofit”+metal chlorides. Due to the high density of metal hydrides, the systems tend to be very compact, minimizing precious volume and envelope. How- Table 4 Full sorption capacity of a sorbent bed Full sorption capacity of the sorbent bed “Busofit” Methane, T = 253 K, P = 6 MPa Ammonia (g/g) Hydrogen (wt%) Methanol (g/g) 182 v/v 0.35 1.5 0.55 “Busofit”+CaCl2 or “Busofit”+ LaNi4.5 Al0.29 Mn0.21 0.62–1.03 3.2 ever, the relatively low hydrogen capacity of metal hydrides tends to make the systems heavy. Thus, metal hydride applications are best suited for systems where power and volume are more critical than weight. A hydrogen compressor, or storage reservoir based on a reversible solid sorption will be attractive if the process can store high gas density. For that, the reactor must contain a large quantity of reactive mixture and the 5023 L.L. Vasiliev et al. / International Journal of Hydrogen Energy 32 (2007) 5015 – 5025 0.35 0.6 0.30 0.5 0°C -adsorption 0°C -desorption 20°C -adsorption 20°C -desorption 30°C -adsorption 30°C -desorption 40°C -adsorption 40°C -desorption 0.20 0.15 0.10 0.05 a, kg/kg a, kg/kg 0.25 0.4 20°C 40°C 0.3 0.2 0.1 0.0 0.00 0 2 4 6 8 10 12 0 2 -5 4 6 8 10 12 -5 P.10 , Pa P.10 , Pa Fig. 11. Ammonia sorption isotherms for “Busofit” (a) and “Busofit+CaCl2 ” (b). Table 5 Metal hydride materials for the hydrogen storage and transportation Metal hydride Suitable as sorbent Technical storage capacity (wt%) Temperature (K) LaNi4.1 Al0.52 Mn0.37 LaNi4.5 Al0.29 Mn0.21 LmNi4.91 Sn0.15 Ti0.9 Zr 0.10 V0.43 Fe0.09 Cr 0.09 Mn1.5 Ti0.99 Zr 0.01 V0.43 Fe0.09 Cr 0.05 Mn1.5 Yes Yes Yes Yes Yes 0.7–0.9 0.8–1.0 1.0 1.0 1.0 430–490 373–433 279–303 283–323 268–303 kinetics of the gas–solid reaction will be sufficiently fast. Heat and mass transfer must be high, which implies good porosity and high thermal conductivity of the reactor material. In the case of gas-salt reaction, the use of a salt-binder mixture (carbon, exfoliated graphite, graphite intercalation compounds and intercalated or impregnated carbon fibers) enhanced the energy and power performances of the system. Another solution is the use of impregnated carbon fabrics saturated by metal hydrides. Successful application in practical systems requires particular properties of the hydride alloys. Good alloy candidates (Table 5) must have good temperature and pressure interval. His parameter dictates, along with the density of the alloy and the total uptake of hydrogen, the total energy density of the material. High energy densities allow for volume and/or weight minimization for a given storage system requirement. The plateau pressures must also be in the range of the operational constraints of the system. Metal hydride reaction beds inside the storage vessel can be operated temperature controlled (by heat pipe heat transfer devices), or pressure controlled (isobaric). To keep the temperature of an absorbing bed constant, the pressure has to raise due to the increasing absorbed gas concentration that result in a higher pressure at a given temperature. To keep the pressure of an absorbing bed constant, the temperature of the hydride must decrease. In absorption the reactor operates as absorber in which a solid material (metal chloride/hydride) reacts with the gas to produce an exothermic reaction: S(sol) + nG(gas) → SGn(sol) , Hreact < 0. (5) Practically, this chemical reaction is equilibrium, and for example, in the case of transition metal chloride MCl2 (S) and ammonia (G), the following reaction is observed: M(NH3 )2 Cl2(sol) + 4NH3(gas) → M(NH3 )6 Cl2(sol) . (6) The enthalpy of reaction Hreact. in (5) is about 50 kJ/NH3 mol. for heating. The low thermal conductivity of the solid salt (about 0.1 W/(m K)) and the very high expansion factor of the salt S during its reaction with the gas G, are two important coefficients being able to reduce the sorption capacity. Heat has to be evacuated outside during the chemical reaction of absorption, and gas diffusion has not to be slackened to reach the solid. Conversely, in desorption phase, the reactor operates as regenerator in an endothermic reaction. The use of additives (active carbon fiber) to metal chloride/hydride has two functions: • the increase of the total sorption capacity (adsorption+ absorption), • the increase of the thermal conduction of the reactant and of the heat transfer coefficient at the interface between reactant and wall, • the maintenance of high porosity of the medium during the solid-gas reaction. The addition to metal chloride/hydride of carbon powders, graphite compounds, expanded graphite, activated carbons, carbon fibers, carbon fabrics have enhanced the performances of the gas storage vessel. It is particularly convenient to 5024 L.L. Vasiliev et al. / International Journal of Hydrogen Energy 32 (2007) 5015 – 5025 use carbon fabrics because they possess the following main qualities: jT jz • 2D high thermal conduction (if the precursor fibers have good thermal conductivity), • gas permeability into the inter-fibers space, • no discontinuities of the thermal conductivity between the reactor core and the heat exchanger wall. − The metal hydride+“Busofit” gas storage vessel can also be considered as a heat sink for the transport vehicle, or the source of energy for the air conditioning system. The heat sink absorbs and rejects metabolic, sun and equipment heat loads generated during the car transportation. Table 5 contains some of the low temperature metal hydrides. 6. Heat transfer in the sorbent bed Carbon material can be used as a compact sandwich with cylindrical or flat heat pipes, applied as thermal control systems. The dynamic model of the sorbent bed has the components [12]: (1) Dubinin–Radushkevich equation (3) of the state of gas. (2) The equation of energy: jT jT r(εcC g + C + aC a ) + rcvC g j jr     j ja jT j jT = r + r + rq st  , jr jr jz jz j where the isosteric heat of desorption is:   j ln P qst = R T . j ln T /a=const (7) (8) (3) The equation of continuity: j j (εc + a) + (rcv) = 0. j jz (9) (4) The equation of kinetic of sorption:    da E aeq − a , = Kso exp − d R T (10) r where Ks0 = 15Ds0 /Rp2 , Rp is radii of the particle. Ds0 is constant necessary to determine the coefficient of a surface diffusion, Ds = Ds0 exp[E/(R T )]. The solution was found for the fixed gas flow from the SSH vessel 2 S d d  r1 (εc + a)r dr dz = −g/N (11) r0 with boundary conditions: P |=0 = P0 , T (r, z)|=0 = T0 (r, z) = Tenv , (12) − jT jz = 0, z=0 jT jr jT jr = 0, z=S = env (T − Tenv ), (13) r=R = r=R0 Qhp , 2 R0 SN or T |r=R0 = Thp , (14) where Qhp is the heat flow used to heat one cylinder of the vessel and Thp the gas vessel wall temperature. The first condition (14) corresponds to the situation where the power Qhp is known, and the second condition corresponds to the situation where the heating power is not limited and the temperature Thp is set. This temperature is maintained at the inner surface of the heat pipe due to the contact of its evaporation zone with a large heated body, such as an engine. To solve the set of equations (3, 7–10) with boundary conditions (12–14) the method of finite elements was chosen. The suggested simple model gives us a possibility to obtain the field of temperature and gas concentrations during charge–discharge procedure of the gas vessel. Successful application in practical systems requires particular properties of the hydride alloys. Suitable alloys must also have reasonable kinetics of absorption and desorption. Rapid kinetics of desorption are required for operation during the cooling mode. If it is extremely difficult to recharge the alloy to its fully hydride state due to unfavourable absorption kinetics, it may be impractical to regenerate the material after use. 7. Heat pipe heat exchangers The efficient system to perform a sorbent bed thermal control during its charging/discharging is heat pipe heat exchanger. “Busofit” can be used as a compact sandwich with flat, or cylindrical heat pipes being in good thermal contact with the surrounding, or source of energy (for example, gas flame, electric heater). Flat heat pipe panels (HPP) have some advantages over conventional cylindrical heat pipes, such as geometry adaptation, ability for much localized heat dissipation and the production of an entirely flat isothermal surface. The liquid–vapour system formed in capillary channels inside the heat pipe panel is capable to generate self-sustained thermally driven oscillations. Thin layer (1–2 mm) of the sorbent composite (for example, active carbon powder+salts) between mini-fins on the outer side of the heat pipe panel ensures an advanced heat and mass transfer during the cycle adsorption/desorption. HPP is considered as a system including some channels of capillary dimension, which are bent in a serpentine manner and the ends are joined. This device is filled partially with working fluid. The flow instabilities inside of this device are produced due to the heat input in one part of it and heat output from the another part by heating multi-channels (H = 2 mm, L = 5 mm) at one end and simultaneously cooling the other end thus resulting in pulsating fluid. This heat input and output stimulates a heat transfer, as a combination of sensible and latent heat portions. The flow instabilities are a superposition of various underlying effects. L.L. Vasiliev et al. / International Journal of Hydrogen Energy 32 (2007) 5015 – 5025 5025 References Fig. 12. Heat pipe panels (pulsating heat pipe) for sorbent bed thermal control. In our experiments as an experimental set-up an aluminium multi-channel panel was chosen. The main parameters of flat aluminium heat pipe panel, developed in the Luikov Institute are: HPP width—70 mm, HPP height—7 mm, HPP length—700 mm, evaporator length—98 mm, condenser length—500 mm, mass—0.43 kg. Propane was applied to fill the HPP and it is interesting as a working fluid with the point of view of its compatibility with all heat pipe envelopes and wick materials (aluminium, steel, stainless steel, copper, AL2 O3 ). The typical heat pipe panel for sorbent bed thermal control is shown on Fig. 12. 8. Conclusions To improve the parameters of the gas storage system the activated carbon fiber and metal hydride/chloride with the binder was suggested as a new solid sorption composite. Some sorption capacity data for new materials were obtained. The use of a thermo-chemical material based on carbon nanofibers in a solid sorption reactor promises a good perspective for design a new type of gas storage tanks. Acknowledgments This work was sponsored by the National Academy of Sciences of Belarus, the State program of applied scientific researches “Hydrogen”, project no. 12. [1] Hynek S, Fuller W, Bentley J. Hydrogen storage by carbon sorption. Int J Hydrogen Energy 1997;22:601–10. [2] Carpetis C, Peschka W. A study on hydrogen storage by use of cryoadsorbents. Int J Hydrogen Energy 1980;5:539–54. [3] Cheng HM, Yang QH, Liu C. Hydrogen storage in carbon nanotubes. Int J Hydrogen Energy 2002;27:193–202. [4] Kuznetsov AV, Vafai K. Analytical comparison and criteria for heat and mass transfer models in metal hydrides packed beds. Int J Hydrogen Energy 1995;38:2873–84. [5] Vasiliev LL, Mishkinis DA, Antukh AA, Vasiliev Jr. LL. Solar–gas solid sorption refrigerator. Adsorption 2001;7:149–61. [6] Vasiliev LL, Kanonchik LE, Mishkinis DA, Rabetsky MI. Adsorbed natural gas storage and transportation vessels. Int J Therm Sci 2000;39:1047–55. [7] Vasiliev LL, Kulakov AG, Mishkinis DA, Safonova AM, Luneva NK. Activated carbon for gas adsorption. In: Proceedings of III international of the symposium “Fullerene and Semifullerene Structures in the Condensed Media”, Minsk, Belarus; 2004, pp. 110–5. [8] Brunauer S, Emmertt PH, Teller E. J Am Chem Soc 1938;60:309–19. [9] Adsorbed Natural Gas Research Conducted by Atlanta Gas Light Adsorbent Group (AGLARG). Final report (1990–1993) Gas Research Institute. Atlanta GRI-95/0068; 1994. [10] Nijkamp NG, Raaymakers JEMJ, van Dillen AJ, de Jong KP. Hydrogen storage using physisorption—materials demands. Appl Phys A Mater Sci Process A 2001;72:619–23. [11] Vasiliev LL, Mishkinis DA, Kanonchik LE, Khrolenok VV, Zhuravlyov AS. Activated carbon ammonia and natural gas adsorptive storage, Ext Abstracts of 23rd biennial conference on carbon, vol. 1. “Carbon ’97”, Philadelphia, 1997. p. 334–5. [12] Vasiliev LL, Kanonchik LE, Mishkinis DA, Rabetsky MI. A new method of methane storage and transportation. Int J Environ Conscious Des Manuf 2000;9:35–62. [13] Dubinin MM. The potential theory of adsorption of gases and vapors for sorbents with energetically nonuniform surfaces. Chem Rev 1960;60: 235–41. [14] Agarwal RK. High pressure adsorption of pure gases on activated carbon: analysis of adsorption isotherms by application of potential theory and determination of heats and entropies of adsorption, PhD. Dissertation, Syracuse; 1988. [15] Vasiliev LL, Kulakov AG, Mishkinis DA. Activated carbon for gas adsorption. In: Proceedings of international conference “solid state hydrogen storage—materials and applications”, Hyderabad, India; 2005. p. 10–8.