Hydrogen storage in Chabazite zeolite frameworks

RESEARCH PAPER PCCP Laura Regli,a Adriano Zecchina,*a Jenny G. Vitillo,a Donato Cocina,a Giuseppe Spoto,a Carlo Lamberti,a Karl P. Lillerud,b Unni Olsbyeb and Silvia Bordigaa www.rsc.org/pccp Hydrogen storage in Chabazite zeolite frameworks a Dipartimento di Chimica IFM and NIS Centre of Excellence, Via P. Giuria 7, I-10125 Torino, Italy. E-mail: [email protected]; Fax: þ39 0116707855; Tel: þ39 0116707860 b Department of Chemistry, University of Oslo, P.O. Box 1033, N-0315 Oslo, Norway Received 28th June 2005, Accepted 15th July 2005 First published as an Advance Article on the web 3rd August 2005 We have recently highlighted that H-SSZ-13, a highly siliceous zeolite (Si/Al ¼ 11.6) with a chabazitic framework, is the most efficient zeolitic material for hydrogen storage [A. Zecchina, S. Bordiga, J. G. Vitillo, G. Ricchiardi, C. Lamberti, G. Spoto, M. Bjørgen and K. P. Lillerud, J. Am. Chem. Soc., 2005, 127, 6361]. The aim of this new study is thus to clarify both the role played by the acidic strength and by the density of the polarizing centers hosted in the same framework topology in the increase of the adsorptive capabilities of the chabazitic materials towards H2. To achieve this goal, the volumetric experiments of H2 uptake (performed at 77 K) and the transmission IR experiment of H2 adsorption at 15 K have been performed on H-SSZ-13, H-SAPO-34 (the isostructural silico-aluminophosphate material with the same Brønsted site density) and H-CHA (the standard chabazite zeolite: Si/Al ¼ 2.1) materials. We have found that a H2 uptake improvement has been obtained by increasing the acidic strength of the Brønsted sites (moving from H-SAPO-34 to H-SSZ-13). Conversely, the important increase of the Brønsted sites density (moving from H-SSZ-13 to H-CHA) has played a negative role. This unexpected behavior has been explained as follows. The additional Brønsted sites are in mutual interaction via H-bonds inside the small cages of the chabazitic framework and for most of them the energetic cost needed to displace the adjacent OH ligand is higher than the adsorption enthalpy of the OH


H2 adduct. From our work it can be concluded that proton exchanged chabazitic frameworks represent, among zeolites, the most efficient materials for hydrogen storage. We have shown that a proper balance between available space (volume accessible to hydrogen), high contact surface, and specific interaction with strong and isolated polarizing centers are the necessary characteristics requested to design better materials for molecular H2 storage. DOI: 10.1039/b509124a 1. Introduction There are still major problems that must be solved before an energy system with hydrogen as the energy carrier can replace today’s hydrocarbon fuel based engines. A crucial point is that all the aspects related to hydrogen production, transport, storage and fuel cell technology have to be clarified simultaneously to allow hydrogen economy to succeed. Concerning the point of hydrogen storage, as at room temperature and atmospheric pressure hydrogen occupies about 3000 times the volume of gasoline providing the same amount energy, it must be mechanically compressed, liquefied or subjected to some other more sophisticated storage method to be used for practical purpose.1–3 At the present time prototype hydrogen vehicles use space demanding tanks with compressed gas.4–6 On the other hand, hydrogen liquefaction is attractive on the energy density ground, but about 30% of the energy available has to be spent for the required cryogenic environment. An additional complication is the weight of the extensively insulated tanks. Large effort has been devoted in these years to develop new materials for hydrogen storage. Materials interesting as hydrogen reservoirs should provide a secure, light and cheap storage, possess an high adsorption capacity and the fastest possible recharging kinetics.2 Many storage strategies have been followed in recent years, aiming at fulfilling these requirements. The most important ones are: (1) storage in metals and alloys;2,7–9 (2) storage in complex hydrides (alanates, borohydrides);2,10 (3) storage by trapping (i.e. in clathrates);11,12 (4) storage in microporous materials (carbons,2,13 metallorganic frameworks polymers,14 zeolitic materials15). Storage in micro- porous materials involves the interaction of molecular hydrogen with the internal surfaces of micropores, governing the stability of the molecular adducts and the optimal temperature of storage. The forces involved in this process are essentially physical (dispersive) and the enthalpies involved are low. To increase the interaction energy, the presence of ions gives a relevant contribution: studies of H2 interactions with cations in the gas phase report high interaction energies.16 Unfortunately, cations need counterions to be stabilized and, when counterions are considered the interaction energy decrease substantially. A well known class of microporous materials containing isolated and exposed cations17 (already widely employed in catalysis and gas separation18), are zeolites and zeotypes. Due to their non-negligble weight, these materials cannot be the complete solution for H2 storage, however they can represent model systems to investigate the interaction of H2 with both dispersive and electrostatic forces. Strength points of these materials are the availability of a large variety of frameworks and chemical compositions combined with low cost, good mechanical and thermal stabilities and the possibility to perform reproducible, systematic fundamental studies. All these aspects give relevance to the results obtained on this class of materials which, until now, have shown H2 adsorptive properties inferior to those of the other microporous matrices19–32 (in terms of moles of H2 per gram of hosting matrix at given pressures and temperatures). At the present time many data about adsorptive properties of zeolites towards H2 have been reported, however a systematic study is not available yet. In this work, by combining volumetric measurements (performed at 77 K in the 1  107–0.92 bar range) and transmission infrared (IR) spectroscopy (at 15 K and in the 1  107–0.013 This journal is & The Owner Societies 2005 Phys. Chem. Chem. Phys., 2005, 7, 3197–3203 3197 bar range), we report the results obtained on three zeolites and zeotype materials with chabazitic frameworks in comparison with data coming from the literature obtained on others high surface area microporous materials. The choice of the three investigated materials will be explained in the next paragraph. We have recently highlighted that H-SSZ-13, the highly siliceous zeolite (Si/Al ¼ 11.6) with chabazitic framework, is the most efficient zeolitic material for hydrogen storage.15 Three main factors can influence the H2 uptake capabilities of an acidic zeolitic material: (i) the framework topology; (ii) the strength of the acidic Brønsted sites; (iii) the average density of the acidic Brønsted sites (in terms of number of sites per zeolite cavity). In this work, by keeping fixed point (i), we have faced points (ii) and (iii). Point (ii) has been considered by investigating a H-SAPO-34 zeotype, the SAPO material isostructural to the chabazite zeolite. For H-SAPO-34 the synthesis has been driven to obtain a chemical composition of Si : Al : P ¼ 1 : 6 : 5, guaranteeing the same proton density as for SSZ-13. As the Brønsted sites hosted in the H-SAPO-34 are slightly less acidic than those hosted in the H-SSZ-13 framework, this samples allows thus to investigate the effect of point (ii) by keeping fixed points (i) and (iii). The effect of point (iii) has been investigated by studying a common H-CHA zeolite with Si/Al ¼ 2.1. 2. Experimental Three zeolites with Chabazite topology have been studied. (i) Zeolite SSZ-13 (Si/Al ¼ 11.6) was obtained by following an hydrothermal preparation procedure described elsewhere.33 (ii) H-SAPO-34 sample (Si : Al : P ¼ 1 : 6 : 5, a composition guaranteeing the same proton density as for SSZ-13) was synthesized and calcined in accordance with standard procedures.34 (iii) A standard H-CHA with high aluminium loading (Si/Al ¼ 2.1) has been also considered for comparison. This sample has been obtained by conversion of a NH4-Y zeolite as described in ref. 35. XRD analysis confirm the true CHA structure for the three samples. H2 adsorption measurements were performed with a Micromeritics ASAP 2010 sorption analyzer at 77 K in the 1  107– 0.92 bar range. Surface area have been obtained on the same instrument by N2 adsorption at 77 K; accessible microporous volume has been estimated by the t-plot (Harkins and Jura thickness equation) of the N2 adsorption data. The infrared spectroscopic measurements were performed by using an home made cryogenic set-up which allows: (a) in situ high-temperature activation of the sample under high vacuum condition (in these experiments an activation in vacuum at 773 K has been adopted); (b) to perform FTIR adsorption experiments in transmission mode at fixed temperature, as low as ca. 15 K (estimated at the sample level), and variable H2 pressure; (c) to record IR spectra in the 300–15 K temperature interval, monitoring simultaneously the gas-phase equilibrium pressure of the adsorbed species. A detailed description of the cryogenic set-up (consisting of a properly modified closed circuit liquid helium Oxford CCC 1204 cryostat) is given elsewhere.36 The spectra were acquired at a resolution of 1 cm1 by averaging 128 interferograms on a Bruker Equinox-55 FTIR spectrometer whose sample compartment was modified ad hoc to accommodate the cryogenic IR set-up. All volumetric and spectroscopic measurements have been preceded by activation in high vacuo at 773 K in order to remove physisorbed water and impurities. H2 adsorption– desorption experiments have been performed to check the reversibility of the phenomenon in the whole range of investigated pressures. 3. Results and discussion 3.1. 77 K volumetric data on H-SSZ-13, H-SAPO-34 and H-CHA materials Adsorptive properties of a large variety of microporous materials, changing in framework and composition, have been widely investigated by volumetric measurements performed Table 1 Volumetric measurements obtained for a wide variety of high surface area materials (zeolites; zeotypes; carbons; MOFs). When not specified, see third column, the H2 uptake experiments have been performed at liquid nitrogen temperature Materials VSB-1 Aerogel NaY (Si/Al ¼ 2.4) Zeolite L H-Y (Si/Al ¼ 2.7) VSB-5 MCM-41 Ferrierite H-Mordenite (Si/Al ¼ 7.0) (Mg,Na,K)-ETS-10 H-ZSM-5 (Si/Al ¼ 40) H-ZSM5 (Si/Al ¼ 16) Silicalite NaX (Si/Al ¼ 1.4) (Na,K)-ETS-10 H-SAPO-34 H-Chabazite (Si/Al ¼ 2.1) NaX (Si/Al ¼ 1.05) NaA H-SSZ-13 (Si/Al ¼ 11.6) MOF-5 (IRMOF-1) MOF-5 NAC Activated carbon AC Norit 990293 Activated carbon AC Norit GSX (highly micro- and mesoporous carbon material) SWNT 3198 Framework type (IZA code) FAU LTL FAU FER MOR MFI MFI MFI FAU CHA CHA FAU LTA CHA Phys. Chem. Chem. Phys., 2005, 7, 3197–3203 Pressure/ bar H2 uptake (mass%) 0.79 0.92 0.57 1.00 0.95 0.79 1.00 1.00 0.66 0.92 0.92 0.66 0.92 0.55 0.92 0.92 0.92 0.61 1.05 0.92 1.02 1.02 0.11 0.32 0.37 0.53 0.56 0.57 0.58 0.58 0.60 0.67 0.71 0.72 0.73 0.79 0.80 1.09 1.10 1.22 1.24 1.28 1.13 1.23 1.00 (295 K) 1.00 (295 K) 1.00 (295 K) BET area/m2 g1 183 533 Ref. 638 3362 565 BET 747 Langmuir 20,21 15 23 37 25 20,21 37 37 25 15 15 25 15 23 15 This work This work 23 22 15 14 59 2.10 1.42 2200 933 38 38 0.54 380 38 344 500 1017 344 324 418 474 337 547 490 This journal is & The Owner Societies 2005 at 77 K and in the 0–1 bar range. An overview of the data is reported in Table 1, which collects both literature20–23,25,37,38 and new data obtained in our laboratory. As it can be seen from the data reported in the table, there is a large dispersion of values related to the adsorptive properties and, even more surprising, there are big differences within the same class of materials, such as zeolites or microporous carbons. For instance, if we compare values within the zeolites and zeotypes, from the less efficient material (VSB-1) to the most active one (H-SSZ-13) there is a difference of more than one order of magnitude. Conversely, similar properties in respect to H2 adsorption, seems to characterize very different materials such as ferrierite (H2 uptake of 0.58 wt% at 1 bar at 77 K)37 and SWNT (H2 uptake of 0.54 wt% at 1 bar 77 K)38 or H-SSZ-13 (H2 uptake of 1.28 wt% at 1 bar at 77 K) and MOF-5 (H2 uptake of 1.23 wt% at 1 bar 77 K). This observation indicates that H2 adsorption in microporous materials is governed by concomitant effects. For carbonaceous materials, Nijkamp et al.37 have shown that there is a linear relationship between the microporous volume and the hydrogen uptake. Data reported in Table 1 suggest that, beside a high surface area, a peculiar framework topology (characterized by small nanocavities) combined with the presence of highly dispersed exposed polarizing species are relevant to increase the value of H2 uptake. In this respect, among zeolitic materials, diluted proton exchanged chabazitic frameworks seem to show the most favorable conditions to entrap and liquefy hydrogen at liquid nitrogen temperature. Fig. 1a compares hydrogen isotherms obtained for: H-SSZ13, its aluminophosphate homologue H-SAPO-34 and H-CHA samples. Even if the general behavior of all the three zeolites is very similar, the uptake measured for H-SSZ-13 is higher than that found for the other two materials, especially at medium at high H2 loadings. Conversely, at very low H2 equilibrium pressures, the highest H2 uptake is measured for H-CHA sample (see inset of Fig. 1a). At about 1 bar none out of the three samples shows a plateau in the H2 uptake (Fig. 1a). This suggests that the H2 uptake capability of the chabazitic framework is far from saturated at those conditions (77 K, 1 bar). This means the possibility of substantial additional increase of H2 storage with a moderate increase of the working pressure. Note that a pressure increase of even one or two orders of magnitude could represent practical operating conditions in applications. In a previous paper15 we have already extensively described the peculiar properties of the chabazitic framework in respect to H2 adsorption, underlining the importance of the framework topology of these materials. The chabazitic framework can be described as a layer of double six-rings that are interconnected by four-membered rings. The resulting structure is characterized by small cages delimited by four eight-ring windows that maximize the accessible volume and the interac- Fig. 1 (a) Hydrogen gas sorption isotherms on H-SSZ-13 (full black squares), H-CHA (full gray dots) and in H-SAPO-34 (open black squares) collected at 77 K in the 1  107–0.92 bar range. Inset: enlargement of the low-pressure region (in the 1  107–0.004 bar range). (b) Schematic representation of Chabazite structure. tion surfaces. In Fig. 1b a picture of the chabazitic framework is reported (for sake of simplicity only T positions are shown). Another important peculiarity of volumetric data obtained in case of chabazitic frameworks is represented by the shape of the curves in the low pressure range, as it is reported in the inset of Fig. 1a. We observe that the curves have a steep behavior, which suggest the presence of strong polarizing centers. 3.2. IR spectroscopy of adsorbed H2 probe: a powerful tool for adsorption site discrimination An evaluation of the number, distribution and strength of specific interaction sites can be obtained by low-temperature IR spectroscopy of adsorbed molecules.39,40 The possibility of using H2 as a probe molecule for surface characterization is the fundamental key to further information on the site


H2 interaction. However, as (i) the corresponding adsorption enthalpy values (DH) are usually very low, (ii) the equilibrium between molecules in the gas phase and in the adsorbed state is driven by the Gibbs function (DG ¼ DH  TDS), and (iii) adsorption processes require a relevant decrease of the entropy of the system (DS), a significant fraction of molecules in the adsorbed state can be obtained only once the entropic term (TDS) is minimized by suitably low temperatures. This makes the use of our ad hoc conceived IR experimental setup mandatory to approach the problem (see experimental and ref. 36). Once adsorbed on polarizing centers, the H2 molecule is perturbed with disruption of the local symmetry and IR activation of its stretching mode. The resulting red-shift of the frequency and the intensity of the band associated to the perturbed hydrogen is proportional to the strength of the interaction.23–25,29,41–43 Moreover, when the polarizing species is a hydroxyl group, additional information on the interaction is available by the observation of n(O–H) bands which are also strongly affected by the interaction.25 Results of FTIR experiments obtained for the three chabazitic frameworks are discussed in subsections 3.3, 3.4 and 3.5. Corresponding spectra are reported in Fig. 2–4: parts (a) and (b) referring to n(O–H) and n(H–H) regions, respectively. In any part of the figures, the spectra have been divided in two series of curves which correspond to low (1  107–0.0025 bar) and to high (0.0030–0.013 bar) H2 equilibrium pressure (PH2): bottom and top parts, respectively. Fig. 2 IR spectra of H2 adsorbed at 15 K on H-SSZ-13 previously outgassed in vacuo at 773 K for 1 hour. The spectral sequence was obtained at decreasing H2 equilibrium pressures. The bottom set of data refers to the low PH2 spectra (1  107–0.0025 bar). The bold spectrum refers to the activated sample prior contact with H2, while the dashed one has been collected at PH2 ¼ 0.0025 bar. The top set of data refers to the high PH2 spectra, from 0.0030 bar (dashed spectrum) to 0.013 bar (dotted spectrum). Parts (a) and (b) refer to the n(O–H) and n(H–H) stretching region, respectively. The spectra reported in (b) have been background subtracted. This journal is & The Owner Societies 2005 Phys. Chem. Chem. Phys., 2005, 7, 3197–3203 3199 Fig. 3 As Fig. 2 for H-SAPO-34 sample. 3.3. 15 K IR spectroscopy of H2 adsorbed on H-SSZ-13 material A new set of low temperature IR experiments done on the HSSZ-13 zeolite and on the isostructural H-SAPO-34 material using different probes molecules (CO dosed at 70 K,44 H2O and CH3OH, both dosed at 300 K45) highlighted a proton distributions among all out of the four crystallographic independent sites of the chabazitic framework. The new data consistently suggest that the most abundant family of protons occupies sites H(1), H(2), and H(4). As for H-SSZ-13, protons hosted in these three sites are almost indistinguishable on spectroscopic grounds, giving rise to an adsorption in the n(O–H) stretching region at 3616 cm1 defined as the high frequency (HF) component. The second, and less abundant, family occupies site H(3). This proton is therefore only exposed to one cage and is characterized by a n(O–H) mode at 3584 cm1, labelled as the low frequency (LF) component.15,44–46 These features are clearly visible in the IR spectrum of the H-SSZ-13 activated zeolite (before H2 dosage); see bold spectrum in the bottom part of Fig. 2a, where all the vibrational properties of hydroxyls present on the sample can be appreciated. The IR spectrum of all H-SSZ-13 in the n(O–H) region is characterized by two groups of bands: the first group (3750– 13700 cm1) is assigned to n(O–H) modes of weakly acidic sites (external silanols),47,48 the second group of bands (3620–3570 cm1) is due to the vibrations of two families of strong acidic sites already defined as HF and LF band.49 All the remaining spectra reported in Fig. 2 report increasing doses of H2 adsorbed on H-SSZ-13 at 15 K. From the bottom part of Fig. 2a, it is evident that rising doses of H2 produce a gradual decrease in intensity of the bands associated to strong Brønsted sites (in particular HF component) and a parallel increase of a new broad band with a maximum at 3548 cm1, associated with the H-bonded structures caused by formation of the OH


H2 adduct characterized by Dn(O–H) ¼ 64 cm1). The appearance of an isosbestic point suggests that in this situation the ratio between adsorbent and adsorbate is 1 : 1. By following the spectroscopic evolution it is evident that, up to an H2 partial pressure of 0.0025 bar (dashed curves in Fig. 2a), all HF sites are interacting with H2, while LF sites and external silanols are only marginally involved in the interaction with H2. By considering the H2 point of view, the consequence of the interaction with H-SSZ-13 can be observed by looking at curves reported in Fig. 2b, where the n(H–H) vibration region is reported (free H2 Raman frequency ¼ 4161 cm1). In correspondence with increasing doses of H2 a red shifted broad and complex absorption appears with two principal components at 4109 and 4090 cm1. These signals grow, in a parallel way, in correspondence of erosion of bands associated to acid OH, suggesting that H2 is perturbed by acid sites of H-SSZ-13. We assign the band at 4109 cm1 to the formation of O–H


H2 adducts on the more abundant HF family of sites and the 4090 cm1 component on the less abundant LF sites. This is in agreement with the comparable intensity of the two bands, as the n(H–H) extinction coefficient increases proportionally with the entity of H–H bond perturbation.41,50 Note that 4090 cm1 band, associated with the disappearance of the band due to LF component, represents the lowest value observed for H2 adsorbed on proton–exchanged zeolites,15 confirming the very strong character of these acidic sites. At higher PH2, the growth of a component at 4135 cm1 is associated to H2 interacting with silanols,25,51 as it is accompanied by the perturbation of the band at 3740 cm1 (see the corresponding spectra in Fig. 2a). Coming back to the n(O–H) stretching region (Fig. 2a), we observe that in the high PH2 spectra (upper set of spectra) also the LF component (3584 cm1) is completely consumed, giving rise to a new red-shifted component at 3536 cm1. The reason why this component is eroded only at higher PH2 may be explained by considering the different location of sites associated to this band. As evidenced by a previous work,44 devoted to the study of CO adsorption at low temperature, we have found that the LF component is associated to OH groups on the six-membered rings forming the internal top or bottom of the barrel shaped cage. These sites, being partially coordinated to a framework oxygen, are less available to readily interact with probe molecules. Finally, as already mentioned, at high PH2, hydrogen condenses inside the nanovoids forming a new phase. Under these circumstances, silanol groups are also interacting with hydrogen, as shown by the erosion of the band with a maximum at 3740 cm1 and the parallel growth of a broader absorption at 3718 cm1. 3.4. 15 K IR spectroscopy of H2 adsorbed on H-SAPO-34 material Fig. 4 3200 As Fig. 2 for H-CHA sample. Phys. Chem. Chem. Phys., 2005, 7, 3197–3203 The IR spectrum of pretreated H-SAPO-34 in the OH stretching region, represented by bold curve in the bottom of Fig. 3a, is dominated by a doublet at 3636 (HF) and 3612 cm1 (LF).52 The HF component has the highest intensity and both bands can be ascribed to hydroxyl groups with a strong Brønsted character. Similarly to what has been observed for H-SSZ-13 (see the begin of Section 3.3), the presence of a doublet has been explained in terms of the same two distinct families of OH sites.46,53,54 Minor components at 3748, 3742 and 3676 cm1 are, respectively assigned to Al–OH, Si–OH and P–OH species located outside the microcrystals. After the first H2 dosages, (spectra reported in the low part of Fig. 3a) we observe the erosion of the HF band, accompanied by the growth of a broader component centered around 3578 cm1. This new This journal is & The Owner Societies 2005 band is quite broad and its left hand side extends till to be over imposed with the LF component of the original spectrum, a fact that makes the fate of the LF band difficult to be established. The presence of two components at 4095 and 4104 cm1 in the n(H–H) region (Fig. 2b), suggests that H2 interacts with two types of acidic sites, with the consequence of erosion also of signal at 3612 cm1 associated to a less exposed family of acid sites. However, as discussed previously in case of H-SSZ-13, as H2 extinction coefficient increases proportionally with the entity of molecule perturbation; therefore, despite the intensity of the two bands (4095 and 4104 cm1) is similar, the abundance of the species associated to the signal at 4095 cm1 is much lower than that related to the component at 4104 cm1. The small entity of the phenomenon agrees with the fact that it is hardly visible from the spectral evolution of the O–H stretching region. The last spectrum of the first series of curves has been reported as dashed line and corresponds to PH2 ¼ 0.0025 bar. Concerning other weak acid sites of sample (Al–OH, P–OH and Si–OH) it is clearly visible that they are not involved in interaction with H2 in these range of pressure (bottom spectra in Fig. 3a). The situation changes upon increasing PH2 (upper curves in Fig. 3a). In particular, in the 0.0030–0.013 bar range, we observe some relevant changes: (i) a complete erosion of components associated with Al–OH, P–OH, Si–OH species; (ii) the disappearance of the shoulder at 3603 cm1 associated to LF sites, (iii) a red-shift of the band due to hydrogenperturbed Brønsted sites, originally at 3578 and now at 3565 cm1. This fact can be explained by considering that in these conditions all LF sites are engaged by hydrogen and that a new phase due to liquid hydrogen inside the zeolitic cavities is formed. The spectroscopic evolution in the H2 stretching region produced by low temperature (15 K) and relatively high pressure, is reported in the upper part of Fig. 3b. We observe the growth of two components at 4135 and 4128 cm1, associated mainly with H2 interacting with the zeolite walls and with weak acid sites such as Al–OH, P–OH and Si–OH. The substantial increase of intensity of the signal at 4095 cm1, in agreement with the disappearance of the shoulder at 3612 cm1 associated to LF component, and the progressive decreasing of the band at 4104 cm1 with parallel formation of a broad absorption centered at 4115 cm1, suggest that in presence of high H2 loadings (liquid H2), the specific interaction between the most exposed Brønsted sites and hydrogen (band at 4104 cm1) is replaced by a non-specific interaction involving more hydrogen molecules. 3.5. 15 K IR spectroscopy of H2 adsorbed on H-CHA material On the reciprocal way, multi-centers interactions becomes even more relevant in the last sample, characterized by an high aluminium loading (Si/Al ¼ 2.1, to be compared with the value of 11.6 for the H-SSZ-13 material, see Section 2). In this case a single H2 molecule is expected to interact with more than one vicinal OH group. The results obtained for this material are reported in Fig. 4, parts (a) and (b) for the n(O–H) and n(H–H) stretching region, respectively. The IR spectrum of pretreated H-CHA in the OH stretching region, represented by bold curve in the bottom part of Fig. 4a, is very complex and it is difficult to distinguish single components, with the only exception of the peak associated to external silanols, clearly visible at 3750 cm1. This band has a broad and intense component at lower frequency for which it is not possible to define a maximum. In the region 3680–3550 cm1 a composite strong adsorption appears. The complexity of this band has been previously explained in terms of the copresence of the well-known two families of Brønsted sites characteristic of this framework (HF and LF species) and acidic groups associated to partially extraframework species.55 Due to the very broad character of the absorption, it is very difficult to indicate the exact number and position of all the components; however a clear shoulder is observed at 3665 cm1, a maximum at 3620 cm1 and a second shoulder at about 3580 cm1. The component at higher frequency has been assigned to Brønsted sites associate to partially extraframework Al species,55 while the other two components have been associated to the well-known HF and LF components. Note however the broad character of all these absorptions and the huge low frequency tail of the spectrum, exhibiting an important absorption even below 3500 cm1. These peculiarities confer to the spectrum of the activated H-CHA sample in vacuo an unusual shape if compared with the other two samples (see the bold line spectra in the bottom parts of Fig. 2a and 3a). This feature suggests that, in a framework characterized by small cages and by a high number of acidic sites like H-CHA, several OH groups are already mutually engaged via H-bonding interactions of different strength. This hypothesis is confirmed if we consider the effect of increasing doses of H2 at low temperature. The spectra of Fig. 4 are reported following the same procedure as described for Fig. 2 and 3. Upon H2 interaction we observe a modest erosion of the complex absorption associated to the strong Brønsted sites and the growth of a band with maximum at 3557 cm1 (Fig. 4a). The number of sites that seem to be involved is not very high as testified by the modest intensity of the component at 3557 cm1, while the strength of the OH


H2 interaction seems very similar to what it has been observed for the other zeolites. In particular, if we consider the shift in relation with the most evident maximum (3620 cm1), a value of 63 cm1 is obtained. At low hydrogen coverages it seems that all the families of strong Brønsted sites are involved, but as a minor fraction, being most of them left unperturbed. By looking at the hydroxyl region it is not possible to discriminate the fate of each single component. Unfortunately, the n(H–H) region of Fig. 4b, is not more informative, as it shows a broad band, with a maximum at 4095 cm1, characterized by an obvious tail on the high frequency side. By moving to the spectra associated to a higher PH2, top parts of Fig. 4a and 4b, the main phenomenon is the interaction of hydrogen with low acidic sites, as testified by the erosion of the silanol component at 3750 cm1 and the parallel growth of a band at 3700 cm1. In the mean time, in the n(H–H) region, we observe the formation of a complex absorption centered at about 4130 cm1, already assigned to liquid hydrogen inside the zeolite cages interacting with the oxygen of the walls and with the silanols. The component at 4095 cm1, dominating the low PH2 spectra, progressively undergoes a redshift upon increasing PH2 down to 4086 cm1. This shift is accompanied by a modest intensity increase, most probably mainly due to an increase of the extinction coefficient of the H–H mode, rather than to an increase of the engaged OH


H2 sites). A similar red-shift has been already observed for the other two zeolites. As a general comment we must say that H-CHA, characterized by an high number of potential polarizing sites (about five times than the H-SSZ-13 sample), do not show an increased activity towards H2. Moreover, it seems that most of the Brønsted sites are not involved in the interaction with hydrogen, compare the bold spectrum in the bottom part of Fig. 4a with the highest PH2 one reported in the top part of the same figure. This spectroscopic evidence is explained by the fact that, in the activated sample, several OH groups are already mutually engaged via H-bonding interactions of different strength. Owing to the relatively low adsorption enthalpy of the OH


H2 adduct, dihydrogen is able to engage only the minor fraction of almost isolated Brønsted groups and, via a ligand displacement reaction (OH


OH þ H2 - OH


H2 þ OH), those engaged in weak H-bonding. For all OH groups mutually engaged in stronger H-bonding the ligand displacement reaction is energetically disfavored and the corresponding OH


H2 adduct is not formed. This spectroscopic feature well This journal is & The Owner Societies 2005 Phys. Chem. Chem. Phys., 2005, 7, 3197–3203 3201 strong polarizing site in any cage is a favorable condition. The increase of the strong Brønsted site density (H-CHA sample) does not have a positive effect on the H2 storage ability of the material because the additional Brønsted sites are in mutual interaction via H-bonds inside the small cages of the chabazitic framework and for most of them the energetic cost needed to displace the adjacent OH ligand is higher than the adsorption enthalpy of the OH


H2 adduct. A parallel IR experiment, where CO has been used as a stronger base than H2, confirms this thesis. Acknowledgements Fig. 5 IR spectra of CO adsorbed at liquid nitrogen temperature on H-CHA previously outgassed in vacuo at 773 K for 1 h. The spectral sequence was obtained at decreasing the CO equilibrium pressures from 0.002 bar down 1  107 bar. The bold spectrum refers to the activated sample prior contact with CO, while the dotted one has been collected at highest coverage. Parts (a) and (b) refer to the n(O–H) and n(C–O) stretching region, respectively. Please note that, owing to the much greater perturbation induced by CO, the n(O–H) stretching region has been reported down to 3050 cm1, corresponding to a much broader interval with respect to that reported in parts (a) of Fig. 2–4. The spectra reported in part (b) have been background subtracted. agrees with the lower H2 uptake shown by H-CHA with respect to H-SSZ-13 material, see Fig. 1. To validate this hypothesis the H-CHA sample has also been investigated using a probe molecule with a higher proton affinity like CO (140 kcal mol1),56 to be compared with a value lower than 100 kcal mol1 for H2.57 Fig. 5 reports the IR spectra of different CO equilibrium pressures (PCO) dosed at liquid nitrogen temperature (around 100 K) on H-CHA sample. The n(O–H) and n(C–O) stretching regions are reported in parts (a) and (b), respectively. In this case, a much greater erosion of the complex bands due to strong Brønsted acidic sites is observed (even at lower equilibrium pressures and higher adsorption temperature). This means that the ligand displacement reaction (OH


OH þ CO - OH


CO þ OH) can occur for a significant higher number of Brønsted sites when a stronger base like CO is used instead of H2. If the same experiment is performed with N2 (proton affinity ¼ 118 kcal mol1, spectra not reported for brevity)56 an intermediate erosion occurs. In the case of CO two main maxima are observed for the n(O–H) mode of the OH


CO adducts at about 3470 and 3320 cm1. The corresponding shifts (in relation with the most evident maximum at 3620 cm1) are of 150 and 300 cm1, respectively. A rapid inspection of the n(C–O) stretching region (Fig. 5b) confirms the multi-center interactions that CO molecules also have with Brønsted sites hosted in the H-CHA cavities. At low PCO, broad component appears around 2174 cm1, that progressively shifts down to 2163 cm1 at higher PCO, where also the typical band of liquidlike CO trapped inside the zeolitic cages appears around 2138 cm1.58 4. Conclusions Phys. Chem. Chem. Phys., 2005, 7, 3197–3203 References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 In this work we have shown that proton exchanged chabazitic frameworks represent, among zeolites, promising materials where an appreciable amount of hydrogen is stored. A proper balance between available space (volume accessible to hydrogen), high contact surface, and strong interaction with polarizing centers are the necessary characteristics requested to design better materials for molecular hydrogen storage. Within the series of the considered materials, H-SSZ-13 has shown the best behavior in terms of hydrogen uptake, suggesting that in case of matrices with small cages, the presence of an isolated 3202 S. I. Zones is acknowledged for the private communication concerning the SSZ-13 synthesis. The friendly and stimulating discussion with Gabriele Ricchiardi (University of Torino) and with Morten Bjørgen (Univesrity of Oslo) is gratefully acknowledged. We thank Francesca Bonino for the IR spectra of CO adsorbed on H-CHA sample (Fig. 5). Laura Regli acknowledges the Regione Piemonte for financing her Ph.D. grant. 25 26 27 28 A. Züttel, Naturwissenschaften, 2004, 91, 157. L. Schlapbach and A. Züttel, Nature, 2001, 414, 353. G. D. Berry and S. M. Aceves, Energy Fuels, 1998, 12, 49. L. Schlapbach, MRS Bull., 2002, 27, 675. R. S. Irani, MRS Bull., 2002, 27, 680. J. Wolf, MRS Bull., 2002, 27, 684. R. C. Bowman Jr. and B. Fultz, MRS Bull., 2002, 27, 688. J.-M. Joubert, M. Latroche and A. Percheron-Guégan, MRS Bull., 2002, 27, 694. E. Akiba and M. Okada, MRS Bull., 2002, 27, 699. W. Grochala and P. P. Edwards, Chem. Rev., 2004, 104, 1283. W. L. Mao, H.-k. Mao, A. F. Goncharov, V. V. Struzhkin, Q. Guo, J. Hu, J. Shu, R. J. Hemley, M. Somayazulu and Y. Zhao, Science, 2002, 297, 2247. H. Lee, J.-W. Lee, D. Y. Kim, J. Park, Y.-T. Seo, H. Zeng, I. L. Moudrakovski, C. I. Ratcliffe and J. A. Ripmeester, Nature, 2005, 434, 743. A. Züttel and S. Orimo, MRS Bull., 2002, 27, 705. J. L. C. Rowsell, A. R. Millward, K. S. Park and O. M. Yaghi, J. Am. Chem. Soc., 2004, 126, 5666. A. Zecchina, S. Bordiga, J. G. Vitillo, G. Ricchiardi, C. Lamberti, G. Spoto, M. Bjørgen and K. P. Lillerud, J. Am. Chem. Soc., 2005, 127, 6361. C. H. Wu, J. Chem. Phys., 1979, 71, 783. C. Lamberti, S. Bordiga, F. Geobaldo, A. Zecchina and C. Otero Areán, J. Chem. Phys., 1995, 103, 3158. M. Grayson and D. Eckroth, Kirk-Othmer Encyclopedia of Chemical Technology, John Wiley & Sons, Inc., New York, 3rd edn, 1981, vol. 15, pp. 638–669. H. W. Langmi, A. Walton, M. M. Al-Mamouri, S. R. Johnson, D. Book, J. D. Speight, P. P. Edwards, I. Gameson, P. A. Anderson and I. R. Harris, J. Alloy Compd., 2003, 356, 710. P. M. Forster, J. Eckert, J. S. Chang, S. E. Park, G. Ferey and A. K. Cheetham, J. Am. Chem. Soc., 2003, 125, 1309. N. Guillou, Q. Gao, P. M. Forster, J. S. Chang, M. Nogues, S. E. Park, G. Ferey and A. K. Cheetham, Angew. Chem. Int. Ed., 2001, 40, 2831. F. Stéphanie-Victoire, A. M. Goulay and E. Cohen de Lara, Langmuir, 1998, 14, 7255. V. B. Kazansky, V. Y. Borovkov, A. Serich and H. G. Karge, Microporous Mesoporous Mater., 1998, 22, 251. S. Bordiga, E. Garrone, C. Lamberti, A. Zecchina, C. Otero Areán, V. B. Kazansky and L. M. Kustov, J. Chem. Soc., Faraday Trans., 1994, 90, 3367. M. A. Makarova, V. L. Zholobenko, K. M. Alghefaili, N. E. Thompson, J. Dewing and J. Dwyer, J. Chem. Soc., Faraday Trans., 1994, 90, 1047. E. Garrone, V. B. Kazansky, L. M. Kustov, J. Sauer, I. N. Senchenya and P. Ugliengo, J. Phys. Chem., 1992, 96, 1040. A. Zecchina, C. Otero Areán, G. Turnes Palomino, F. Geobaldo, C. Lamberti, G. Spoto and S. Bordiga, Phys. Chem. Chem. Phys., 1999, 1, 1649. C. Otero Areán, O. V. Manoilova, B. Bonelli, M. R. Delgado, G. Turnes Palomino and E. Garrone, Chem. Phys. Lett., 2003, 370, 631. This journal is & The Owner Societies 2005 29 G. Spoto, E. Gribov, S. Bordiga, C. Lamberti, G. Ricchiardi, D. Scarano and A. Zecchina, Chem. Commun., 2004, 2768. 30 A. I. Serykh and V. B. Kazansky, Phys. Chem. Chem. Phys., 2004, 6, 5250. 31 K. W. Chapman, P. D. Southon, C. L. Weeks and C. J. Kepert, Chem. Commun., 2005, 3322. 32 S. S. Kaye and J. R. Long Jr., J. Am. Chem. Soc., 2005, 127, 6506. 33 S. I. Zones, US Pat., 4,544,538, 1985 and personal communication. 34 M. B. Lok, C. A. Messina, R. L. Patton, R. T. Gajek, T. R. Cannan and E. M. Flanigen, US Pat., 4,440,871, example 35 and 36, 1984. 35 C. G. Coe, T. R. Gaffney and R. S. Srinivasan, US Pat. 4,925,460, July 20 1989. 36 G. Spoto, E. N. Gribov, G. Ricchiardi, A. Damin, D. Scarano, S. Bordiga, C. Lamberti and A. Zecchina, Prog. Surf. Sci., 2004, 76, 71. 37 M. G. Nijkamp, J. E. M. J. Raaymakers, A. J. van Dillen and K. P. de Jong, Appl. Phys. A, 2001, 72, 619. 38 H. G. Schimmel, G. J. Kearley, M. G. Nijkamp, C. T. Visserl, K. P. de Jong and F. M. Mulder, Chem. Eur. J., 2003, 9, 4764. 39 A. Zecchina, D. Scarano, S. Bordiga, G. Spoto and C. Lamberti, Adv. Catal., 2001, 46, 265. 40 A. Zecchina and C. Otero Arean, Chem. Soc. Rev., 1996, 25, 187. 41 X. Solans-Monfort, V. Branchadell, M. Sodupe, C. M. ZicovichWilson, E. Gribov, G. Spoto, C. Busco and P. Ugliengo, J. Phys. Chem. B, 2004, 108, 8278. 42 E. N. Gribov, S. Bertarione, D. Scarano, C. Lamberti, G. Spoto and A. Zecchina, J. Phys. Chem. B, 2004, 108, 16174. 43 Y. Jeanvoine, J. G. Angyan, G. Kresse and J. Hafner, J. Phys. Chem. B, 1998, 102, 5573. 44 S. Bordiga, L. Regli, D. Cocina, C. Lamberti, M. Bjørgen and K. P. Lillerud, J. Phys. Chem. B, 2005, 109, 2779. 45 S. Bordiga, L. Regli, C. Lamberti, A. Zecchina, M. Bjørgen and K. P. Lillerud, J. Phys. Chem. B, 2005, 109, 7724. 46 L. Smith, A. K. Cheetam, L. Marchese, J. M. Thomas, P. A. Wright, J. Chen and E. Gianotti, Catal. Lett., 1996, 41, 13. 47 S. Bordiga, I. Roggero, P. Ugliengo, A. Zecchina, V. Bolis, G. Artioli, R. Buzzoni, G. Marra, F. Rivetti, G. Spano and C. Lamberti, J. Chem. Soc., Dalton Trans., 2000, 3921. 48 S. Bordiga, P. Ugliengo, A. Damin, C. Lamberti, G. Spoto, A. Zecchina, G. Spano, R. Buzzoni, L. Dalloro and F. Rivetti, Top. Catal., 2001, 15, 43. 49 The LF band appears in the bold spectrum of Fig. 2a around 3600 cm1. This value represents an over-estimation of its actual frequency due to the effect of the low frequency tail of the stronger HF component. The true value can be appreciated only when a probe molecule (H2 here or CO in ref. 44) consumes preferentially the HF component allowing the LF one to emerge at 3584 cm1, see higher coverage spectra in the bottom part of Fig. 2b. 50 J. G. Vitillo, A. Damin, A. Zecchina and G. Ricchiardi, J. Chem. Phys., 2005, 122, 114311. 51 I. N. Senchenya, B. Civalleri, P. Ugliengo and G. Garrone, Surf. Sci., 1998, 412/413, 141. 52 As was the case for the LF component in the H-SSZ-13 material, see ref. 49, also for H-SAPO-34 the spectrum of the activated sample (bold spectrum in the bottom part of Fig. 3a) reports an apparently blue-shifted value (3612 cm1). Upon erosion of the dominant HF component by the CO molecule the actual position of the HF component is observed at 3603 cm1, see ref. 44. 53 A. Frache, E. Gianotti and L. Marchese, Catal. Today, 2003, 77, 371. 54 L. Smith, A. K. Cheetam, R. E. Morris, L. Marchese, J. M. Thomas, P. A. Wright and J. Chen, Science, 1996, 271, 799. 55 S. Koertel, J. H. Lunsford and H. Knözinger, J. Phys. Chem. B, 2001, 105, 3917. 56 C. Pazé, S. Bordiga, C. Lamberti, M. Salvalaggio, A. Zecchina and G. Bellussi, J. Phys. Chem. B, 1997, 101, 4740. 57 S. C. Goh and J. B. Swan, Phys. Rev. A, 1981, 24, 1624. 58 (a) S. Bordiga, E. Escalona Platero, C. Otero Areán, C. Lamberti and A. Zecchina, J. Catal., 1992, 137, 179; (b) C. Lamberti, C. Morterra, S. Bordiga, G. Cerrato and D. Scarano, Vib. Spectrosc., 1993, 4, 273; (c) A. Zecchina, S. Bordiga, C. Lamberti, G. Spoto, L. Carnelli and C. Otero Areán, J. Phys. Chem., 1994, 98, 9577; (d) S. Bordiga, C. Lamberti, F. Geobaldo, A. Zecchina, G. Turnes Palomino and C. Otero Areán, Langmuir, 1995, 11, 527. 59 S. Bordiga, J. G. Vitillo, G. Ricchiardi, L. Regli, D. Cocina, A. Zecchina, B. Arstad, M. Bjørgen and K. L. Lillerud, J. Phys. Chem. B, 2005, 109, in press. This journal is & The Owner Societies 2005 Phys. Chem. Chem. Phys., 2005, 7, 3197–3203 3203