A critical property of electrode materials is the ability to conduct Li through the host lattice. Li-ion mobility can be directly probed with INS and QENS, however only a few neutron studies report the direct measurement of Li dynamics mostly due to its moderate incoherent neutron-scattering cross section. Typically, each material has to be considered in order to determine whether the signal originates from Li, magnetism or other atoms. Usually, the hopping diffusion in Li-ion insertion electrodes is relatively slow compared to the timescale of INS and QENS in which case the local mobility is observed. As a consequence few studies exist that probe the Li motion directly. A different approach is to probe the Li diffusion in electrode materials with diffraction. Both the anisotropic contribution in the Debye-Waller factor and the deviation from harmonicity due to thermal motions at elevated temperatures can indicate the directionality of Li motion, which in turn may allow the identification of diffusion pathways. Both are illustrated here: anisotropy in the ADPs in combination with a maximum-entropy method (MEM) was used to identify the Li-ion trajectory in the positive-electrode material LiFePO4, and an — harmonicity of the ADPs revealed the Li-ion trajectory in the negative electrode material Li4Ti5O12.

Li12C60 fulleride is a good example where Li diffusion was studied using QENS and INS, quantifying the diffusional motion of Li-ions across a phase transition proposing a localized jump-diffusion model in the octahedral voids of the Li12C60 structure. This accounts for the changes in the vibrational density of states near the phase transition and results in a model of the dynamical behaviour [141]. Another QENS study revealed the diffusion coefficient of Li in a highly-oriented pyrolytic graphite electrode at high temperatures, deriving an activation energy of 0.35 eV [142]. Interestingly, the diffusion coefficient obtained is similar to that obtained using electrochemical methods despite the diffusion lengths measured by the two techniques differing by a factor of 15,000. Li diffusion is more frequently deter­mined indirectly using neutrons, and an example of this is the studies of anion dynamics to shed light on Li diffusion. Li-containing metal hydride systems have been investigated, such as in LiBH4 and LiAlH4, where translational modes of Li are linked with BH4 in the high-temperature form of LiBH4 [143]. Additionally, the disappearance of Li-containing lattice vibrations near phase transitions in these compounds is thought to be associated with the delocalisation of Li that enhances diffusivity. In this way, hydrogen-containing group dynamics can provide infor­mation on Li dynamics.

One QENS study describes alkali-ion diffusion (including for Li) in alkali — containing silicate melts [144], of interest for cathode materials based on silicates. This study used the decoupling of the incoherent (below 60 ps) and coherent neutron-scattering as a signature for Li-ion diffusion along channels in the immobile Si-O network. The relaxation times for Li were a factor of two smaller than for Na, indicating that Li-ion diffusivity is a factor of two larger, in agreement with con­ductivity data. QENS experiments have also been performed on single crystals of 7Li2MnCl4 (an inverse spinel-type structure), revealing a lack of anisotropy in the local Li motion [145]. Li-ions at 8a tetrahedral sites were shown to visit neigh­bouring 16c interstitial sites and jump back, but longer-range translational motion was outside the timescale used for the measurement.

Significant insight into Li diffusion can be gained from diffraction. Diffusion pathways can be identified by the anisotropy in ADPs in combination with MEMs. The exceptionally high discharge rate [47] observed in LiFePO4 indicates that ionic mobility in the LiFePO4 matrix is unusually fast. This has raised the question of how this is possible by the small polarons that are strongly localized at Fe sites in phase-separated LiFePO4 and FePO4 [146]. Morgan et al. [27] used the nudged elastic band method in calculations that show high Li ion mobility occurs in tunnels along the [010] direction, but reveal that hopping between tunnels is unlikely, confirmed by calculations of Islam et al. [108]. Fast one-dimensional conduction along the b-axis in the LiFePO4 Pnma structure was predicted by atomistic mod­elling [27, 108] and the first experimental proof of the diffusion trajectory came from Nishimura et al. [147] using NPD in combination with the MEM. To enhance the sensitivity towards Li 7LiFePO4 was prepared using 7Li-enriched Li2CO3 as the raw material. In this study the ADPs readily show the direction of the Li-trajectory towards adjacent Li-sites, with green ellipsoids in Fig. 7.13 representing the refined Li vibration (displacement parameters) and indicating preferred diffusion towards the face-shared vacant tetrahedra. This suggests a curved trajectory in the [010] direction, consistent with atomic modelling [27, 39].

To relate further the vibrational motions with diffusion, the material with the overall composition Li0.6FePO4 was heated to approximately 620 K. In this com­position Li06FePO4 forms a solid solution at a relatively low temperature, * 500 K, due to the unusual eutectoid as shown in the phase diagram in Fig. 7.13. This is confirmed as single phase by neutron diffraction. Thereby a large number of Li defects are introduced, that in combination with the higher thermal energy, enhances Li motion. Note that the Li trajectory in the solid solution should rep­resent both end members because the crystal symmetry does not change upon heating and Li insertion. In the refinement of the Li0.6FePO4 structure no reliable solution using harmonically-vibrating Li could be found. To evaluate the dynamic

disorder of the Li the MEM was used to estimate the nuclear-density distribution from neutron diffraction. By considering the entropy the most probable distributions of nuclear species can be evaluated, making it possible to evaluate not only the missing and overlapping reflections, but also the more complicated nuclear densi­ties. This approach applied to neutron diffraction data of Li06FePO4 at 620 K leads to the three-dimensional nuclear distribution of Li (Fig. 7.13). The observed dif­fusion along the [010] direction is consistent with the shape of the anisotropic thermal motions shown in Fig. 7.13 and atomistic modelling [27, 39]. Note that the Fe, P, and O atoms remain at their normal positions. The data show that the Li-ions move from one octahedral 4a site to the next via the intermediate tetrahedral vacant site. Along this trajectory the sites do not face-share with other occupied polyhedra. This is in contrast to, for instance, diffusion along the [001] direction where the intermediate octahedral position shares two faces with PO4 tetrahedra which will lead to higher activation energies.

Laumann et al. [148] investigated Li migration in commercial spinel Li4Ti5O12 using variable-temperature neutron diffraction. At 900 °C a marked deviation is observed in the linear dependence of the cell volume, O position, and anisotropic displacement parameters. Refinement of the Li occupancies resulted in almost complete 8a site occupation below 900 °C. However, at 900 °C a Li deficiency of approximately 14 % was observed, which was interpreted as the result of anhar — monic motions and migration of the Li-ions. Therefore, in the fitting procedure one isotropic anharmonic ADP was refined. Examination of the nuclear density revealed negative scattering-length density peaks next to the 16c site. In this way Li-ion occupancy at the 32e site was discovered and subsequent refinement of Li at the 32e sites results in the probability density shown in Fig. 7.14. This makes it possible to formulate the diffusion pathway. Rather than occupying the 16c as an intermediate site between two 8a sites, which introduces an unacceptably long Li-O bond, Li passes from the 8a site through the face of the surrounding O tetrahedron to the nearby 32e site. This is followed by switching to the adjacent 32e site where it is bonded to another O atom, and from where it can hop to the next tetrahedral

8a position. Effectively, this mechanism results in a number of short jumps along the [111] direction between adjacent 8a sites. The energy barriers can be approximated by assuming Boltzmann statistics for single-particle motion resulting in the one — particle potential shown in Fig. 7.14. These findings are consistent with nuclear magnetic resonance measurements indicating that the 16c site forms the saddle point of the barrier between two 8a sites [149]. In this way NPD is able to reveal the details of the three-dimensional long-range diffusion pathway in spinel Li4Ti5Oi2.

The majority of QENS studies on proton-conducting perovskites have been per­formed with the use of either time-of-flight or backscattering methods [58]. These methods give access to the picosecond timescale, extended to * 1 nanosecond in some cases, in which local diffusional-dynamics have been observed, although data have also been interpreted in terms of translational diffusion [59, 60]. The first work was done by Hempelmann et al. [32, 33] for SrYb0.05Ce0.95O2.9-75, where rotational — diffusional motion of the — OH group was observed. These results gave support for molecular-dynamics simulations, which suggested that the proton-conduction mechanism in hydrated perovskites involves proton jumps between neighbouring oxygens and rotational diffusion of the — OH group between proton transfers [28-31]. Later, Groh et al. [61] reported on localized diffusional proton-dynamics in BaZr085M0.15O2.925 (M = Y, In, and Ga), Pionke et al. [59] reported the proton self­diffusion constant for protons in Ba[Ca0.39Nb0.6i]O2.9i, and similarly, Wilmer et al. [62] presented results for BaY0.i0Zr0.90O2.95. Braun et al. [60], reported two dif­ferent activation energies for proton diffusion in BaY0.i0Zr0.90O2.95 at different temperature ranges, Colomban et al. [63] reported a change in local proton — dynamics across a structural phase transition of (Ba/Sr)Zrj_xLnxO3_d, whilst Karlsson et al. [64] reported a relatively-small difference in the activation energy for local proton-dynamics depending on the choice of dopant atom in BaM0.i0Zr0.90 O2.95 (M = Y and Sc). This collection of examples illustrates the success of time — of-flight and backscattering methods to study the local diffusional proton-dynamics in proton-conducting oxides. However, to reach the long time-scale of several nanoseconds needed to study the long-range translational proton-diffusion on an atomic length-scale (* 1-30 A), another QENS method, namely neutron spin-echo (NSE) spectroscopy, is required.

Hydrogen Storage Materials

Juergen Eckert and Wiebke Lohstroh

Abstract An eventual realization of a Hydrogen Economy requires working solutions in three fundamental areas, namely hydrogen production, hydrogen storage, and fuel cells, in addition to the development of an extensive, new infra­structure. While neutron scattering experiments and the associated techniques of analysis have been of utility in all three of these research areas, they have had by far the most significant impact on the development and understanding of materials for hydrogen storage applications. This chapter examines some of these contributions.

While the bulk PEM is at the heart of a working fuel cell, it is also a critical component in the catalyst layer of the membrane electrode assembly (MEA). The polyelectrolyte is typically used as a binder in the electrode(s) where it is in contact with other components including platinum and carbon. These materials can co-exist along with pores (filled with O2, H2O, etc.) in the electrodes to form what is known as the triple-phase interface, or boundary. This term refers to the comingled interfaces of (i) carbon/platinum (C/Pt) particles and pores, (ii) C/Pt particles and polyelectrolyte, and (iii) polyelectrolyte and pores (Fig. 10.4). It has been shown that in these composite electrodes the polyelectrolyte is heterogeneously dispersed and can be confined to films on the order of 2-10 nm thick. It is crucial to the development of such materials for fuel-cell applications to understand how the polyelectrolyte structures at these interfaces impact water transport, proton trans­port, electrochemical reactions, and how certain forms of degradation occur at the

triple-phase boundaries. Furthermore, the structures at interfaces between Nafion® and additive nano-particles may serve to enhance, or improve, properties such as water transport and water retention. Although the structural properties of bulk PEMs have been the focus of many studies using SANS, fewer studies have focused on the thin film and interfacial structural aspects of these materials. Therefore, researchers have employed in situ NR techniques to investigate the structural characteristics of PEMs at interfaces with a variety of materials [38—42].

To date, there have been a limited number of relevant studies that have used NR to investigate the structure of PEM materials. Most of this work has been carried out on Nafion® thin films deposited on various substrates including smooth glassy carbon (GC) [41, 43], sputtered Pt [40, 41, 43], electrochemically oxidized Pt (PtO) [41, 43], SiO2 [38, 44, 45], and gold (Au) [38, 44, 45].

Wood et al. report results from Nafion® thin films spin-coated onto glassy carbon (GC), platinum (Pt) and platinum oxide (PtO) surfaces used to experimentally model the PEMFC electrode interface and by annealing at 140 and 210 °C simulate the decal electrode-preparation method developed by Wilson and Gottesfeld [41, 43, 46, 47]. The films were exposed to 10 % relative humidity H2O and D2O vapour as well as saturated D2O vapour and were found to have different multi-layer structures depending on the substrate. In composite structures of Nafion®/Pt/GC different behaviour was found depending on the relative humidity. At low relative humidity (^10 %) in either H2O or D2O the scattering results were fitted with a single-layer model consisting of hydrated Nafion® with thicknesses on the order of 61-62 nm. The SLD determined for films exposed to 10 % relative humidity H2O and D2O were relatively high (SLDNafion®H2Q = 4.59 x 10-6 A-2; SLDNaflon®D2O = 4.80 x 10-6 A-2) when compared to the value calculated for “dry” Nafion® (SLDNafion®dry = 4.16 x 10-6 A-2) with a known mass density of 1.98 g. cm-3. One would expect that, given the SLD of H2O (-0.56 x 10-6 A-2), the SLD of the hydrated Nafion® film should be lower than that of a dry film. Assuming the water content is unaffected by the isotope the two reported SLD values can be used to calculate that the water volume-fraction at 10 % relative humidity for Wood et al.’s films is approxi­mately 3.2 % by volume and that the SLD of the dry Nafion® would be 4.76 x 106 A-2.

This corresponds to a mass density of 2.27 g. cm-3, which is about 15 % greater than the reported bulk density of 1.98 g. cm-3. One possible explanation for this high density is that the density of the films is higher than that found in bulkNafion® because of the thermal-processing procedure used to prepare the samples, which can increase their crystallinity. However, data for Nafion® on glassy carbon surfaces in ambient air could be fitted as a single layer with SLD =4.12 x 10-6 A-2, which is much lower than for Nafion® in the same conditions on Pt. Another explanation might be the relatively narrow Q-range of the data. When Nafion® on Pt was exposed to saturated D2O vapour the reflectivity curve was best modelled using a two-layer heterogeneous Nafion® film with an approximately 7.5 nm-thick “hydrophobic” layer at the Nafion®/ Pt interface, followed by a thicker (ca. 62 nm) hydrated Nafion® film. This hydro­phobic layer manifests as a “dip” in the SLD profile as shown in Fig. 10.5a. This is in contrast to the work by Murthi and Dura which demonstrated that when Nafion® films that are spin-coated onto Pt or Au are exposed to H2O vapour there is a thin water-rich layer that forms at the polymer/metal interface.

When Nafion® was in direct contact with the GC substrate a more complex scenario evolved (Fig. 10.5b). For the Nafion®/GC systems exposed to a D2O- saturated environment, a three-layer heterogeneous model was used to describe the scattering. In this case, the researchers determined that there was a thin, rough layer (ca. 9 nm thick with 6 nm roughness) sandwiched between two thicker layers. The layer at the Nafion®/vapour interface was the thickest (ca. 57.7 nm) followed by the layer at the Nafion®/GC interface (ca. 26.5 nm). While the water content of each layer was not directly reported, a calculation using the SLD for each layer shows that the layer closest to the GC contained ca. 50 % water by volume. The thick layer at the Nafion®/vapour interface contained about 37 % water, and the middle layer was relatively water depleted at about 24 % water.

Of particular interest are the cyclic-voltammetry results obtained when the Naf — ion®/Pt/GC systems were electrochemically converted to Nafion®/PtO/Pt/GC (Fig. 10.5c). It was reported that although the initial potential-cycle showed no measurable Pt oxidation, subsequent cycles showed clear indications of Pt oxidation and PtO reduction. Once the PtO was formed, the structure was once again probed under saturated D2O conditions. From analysis of the scattering curves Wood and co-workers [41, 43] found that with the PtO layer the interface became more “hydrophilic” compared to the previous Nafion®/Pt interfacial layer. Also of sig­nificance was that, after conversion of the Pt surface to PtO, there was less D2O uptake in the Nafion®/PtO/Pt/GC system. Based on these results a vision of the development of the polymeric structure, specifically for Nafion®, near an interface was formed. Due to strong interactions of the polymer chains with the substrate it was proposed that the typically-isotropic structure reported for bulk Nafion® was modified and becomes anisotropic at the interface. According to Wood et al., the first layer acts as a template and affects the long-range structural properties of the Nafion® thin film. Murthi et al. also showed that the water uptake in Nafion® thin films on metal substrates was measurably lower than that reported for bulk films [40].

Murthi et al. [40] examined Nafion® thin films (59 nm) spin cast onto 6 nm sputtered Pt, using deposition procedures similar to Wood et al., and annealed at

500 Ю00 1500 2000 2500

Thickness (A)

60 °C for one hour or more. From X-ray reflectivity data (to Qmax = 0.7 A 0 it was determined that prior to Nafion® deposition a 0.7 nm PtO surface-oxide layer had formed, presumably by air exposure occurring between deposition steps. Neutron reflectivity data for samples formed in controlled humidity H2O vapour and in liquid water was obtained to measure the water uptake. H2O was chosen over D2O in order to provide a greater contrast between the water domains and the hydro­phobic domains. The data for samples under a controlled humidity-environment at relative humidity values between 0 and 97 % were fitted using a single-layer model and the water content determined from the SLD profiles. In liquid water, a two — layer model was required to describe the data comprised of a thin 16 nm-thick

hydrophilic layer next to the PtO with a water content X (moles H2O per mole SO3-) of 21 and an outer layer of 76 nm with X = 10.2. Similar results were obtained from films prepared on gold substrates.

Dura and co-workers showed a very interesting effect in Nafion® films cast on SiO2 substrates. In contrast to scattering profiles fitted with simple, single-layer, or two-layer models, NR results revealed alternating lamellar layers of water-rich and Nafion®-rich domains that are induced at the interface of hydrated Nafion® films and native silicon-oxide substrates. A cartoon depiction can be seen in Fig. 10.6 which shows the silicon substrate, the native silicon-oxide, and a five-lamella structure. These structures were evidenced by the presence of a peak in the NR curve at approximately Qz = 0.21 A-1 for Nafion® on SiO2 equilibrated at 97 % relative humidity. The lamellar morphology was confirmed by off-specular scat­tering and the location of the lamellar structures at the interface with the SiO2, as opposed to the vapour-polymer interface, was confirmed by comparison of the scattering on thick thermal oxides. The position and intensity of the NR peak were shown to be highly dependent on the hydration level of the film. Detailed mea­surements and analysis including transverse Q-scans led to the conclusion that the

structures were indeed two-dimensional sheets, or lamellae, lying parallel to the substrate surface. The first layer was found to be water-rich at nearly 100 % H2O. The next two water rich layers were found to contain progressively less water, leading to a bulk-like swollen Nafion® layer.

He et al. [39] have used in situ NR to study the structure and kinetic absorption of water in thin films of sulfonated polyphenylene. Films of thickness ranging from 13 to 57 nm cast on oxidized-silicon substrates were exposed to D2O vapour. Typically, NR data collection is too slow to obtain kinetic data, but by limiting the acquisition to the low-Q regime time-averaged data at 10 min intervals could be collected. The NR curves were modelled using a three-layer model. It was deter­mined that there were D2O-rich layers at both the vapour/polymer and polymer/ SiOx interfaces. The kinetic studies revealed that the D2O mass uptake scaled with time1/2 at early times and diverged at later stages. At early stages of water adsorption the effective diffusivity was found to be significantly slower compared to diffusion in the bulk polyelectrolyte.

Our group is investigating the fundamental origins of these lamellar structures in Nafion® and their potential impact on interfacial transport by using a variety of techniques including NR, grazing-incidence SAXS (GISAXS), and quartz-crystal microbalance (QCM). The aim of this work is to investigate the role that specific interactions play on lamellae formation. This is being done by using substrates with tunable chemical characteristics. From our initial studies, it appears that the interfacial structures are generally absent from hydrophobic surfaces, but that in highly hydrophilic substrates there is a strong tendency to form interfacial-lamellar struc­tures [48]. More specifically, we have recently demonstrated that the wettability of the substrate, i. e. hydrophobic or hydrophilic, is a factor governing interfacial structure.

Although a clear and complete picture of how interfacial structure and con­finement in thin polyelectrolyte films influences materials properties, such as water and proton transport, is yet to be gained, it is clear that NR techniques have a significant role to play. In summary, NR has shown that the structure at the interface between a PEM and a substrate depends largely on the surface chemistry, film processing, and even electrochemistry. These factors are certain to have an impact on the transport at these interfaces. Moreover, the confinement of the PEM to a thin film, which certainly has technological relevance, reduces the transport coefficients and can even impact the solubility of water in these materials.

The first report on proton conduction in perovskite-structured oxides dates back to 1981, when Iwahara and co-workers [18] showed that the perovskites SrCeO3 and BaCeO3 exhibit high proton-conductivities in hydrogen-containing atmospheres and may be applicable as electrolytic membrane in electrochemical cells such as fuel cells, steam electrolyzers, and gas sensors. Although BaCeO3 has remained a benchmark for high proton-conductivity in perovskites, many other related mate­rials have been developed and found to be proton conducting. Examples include more complex perovskites such as Ba3Caii8Nbi,82O873 [19—21], rare earth oxides such as Er2O3 [22], and pyrochlore-type oxides such as La2Zr2O7 [23, 24].

Commercial electrolytes typically contain a Li salt dissolved in an organic solvent and are often composed of two components: one for the dissolution of the salt and another that assists in the formation of a protective layer on the anode to prevent continuous electrolyte-reduction and self-discharge, e. g. ethylene carbonate. These electrolyte systems, being non-aqueous and highly air-sensitive, tend to be flam­mable and can turn from liquid to gas at elevated temperatures (Fig. 7.15). The electrolyte also determines the cathode and anode materials that can be used by limiting the applicable voltage range which is associated with the HOMO of the cathode and LUMO of the anode [4]. The key factors that determine a good electrolyte are ionic conductivity, flammability and chemical stability, and appli­cable voltage windows.

To overcome the safety and long-term reliability issues of using organic elec­trolytes, research has been directed to aqueous electrolyte systems with Li salts. Unfortunately, voltage limitations have hampered significant development of aqueous electrolytes, but these safe electrolyte-systems have found niche use in medical applications. In addition to aqueous electrolytes, liquid electrolytes based on ionic liquids have attracted significant attention.

Fig. 7.15 An example of a LiFePOdlgraphite battery containing 1:1 mol. % ethylene carbonate:dimethyl carbonate heated to 90°C where the dimethyl carbonate (organic solvent) has boiled, expanding the casing of the battery

Ю3 T~’ (К’1)

Fig. 7.16 A collection of conductivity data of pertinent electrolytes used for commercial and research-scale Li-ion batteries. Reprinted with permission from (N. Kamaya, K. Homma, Y. Yamakawa, M. Hirayama, R. Kanno, M. Yonemura, T. Kamiyama, Y. Kato, S. Hama, K. Kawamoto, A. Mitsui, Nat. Mater. 10, 682 (2011)) [150]. Nature Publishing Group

Apart from electrolytes in the liquid state, semi-solid or solid-state electrolytes such as gel and solid polymer electrolytes continue to be a preferred option in overcoming safety and leakage issues. Neutron scattering work has been undertaken on ceramic and glass-ceramic solid state Li-ion conducting electrolytes. Some of these electrolytes feature Li-ion conductivities that can be as good as commercial organic electrolytes as elegantly demonstrated for Li10GeP2S12 [150]. This is the first solid-state electrolyte that shows conductivity that matches that of commer — cially-available liquid electrolytes (Fig. 7.16).

NSE offers a unique opportunity to obtain information about dynamical processes on different timescales, e. g. from the elementary processes of the proton-conduction mechanism occurring on the picosecond timescale to the long-range translational diffusion of protons occurring on the nanosecond timescale, simultaneously. Despite these advantages, the first application of NSE to investigate proton-conducting ceramics was relatively recent, in 2010 when Karlsson et al. [65] reported proton dynamics in hydrated BaY010Zr0.90O2.95. Figure 9.10a shows the intermediate­scattering function, I(Q, t), at different temperatures (521-650 K) at the Q-value of 0.3 A-1, within the time-range of 0.2-50 ns. The I(Q, t) is characterized by a decay with time, which is related to the proton motions in the material. In particular, from the shape of this decay and how it depends on temperature and momentum transfer, hQ, information about the mechanistic detail of proton motions, such as the timescale, activation energy, and spatial geometry, can be obtained.

From Fig. 9.10a, it can be seen that the I(Q, t) is described well by a single exponential function (solid lines) with a relaxation time s and a relaxation rate s-1 that exhibits a Q2-dependence (inset), which indicates that the relaxational decay is related to long-range translational diffusion. To further justify a result obtained using a single exponential function, the authors modelled the scattering function, Icalc.(Q, t), using a kinetic model based on first-principles calculations. Figure 9.10b shows these results for momentum transfers Q = 0.3, 0.5, 2.0 A-1, as well as for the long-range diffusion limit Q! 0, at a temperature T = 563 K. In the latter case, the scattering function is given by a single exponential with a characteristic relaxation rate s-1(Q) = DQ2, where D is the diffusion constant. Since the

7calc.(Q, t)s are plotted against Q2t they collapse onto a single curve in the long — range regime and this can be seen in Fig. 9.10b up to at least Q = 0.5 A-1. These results suggest that the long-range translational diffusion of protons occurs and on this basis, a diffusion coefficient assuming a s-1(Q) = DQ2 dependence was extracted. Temperature-dependent results are shown in the Arrhenius plot in Fig. 9.10c, from which it is evident that the diffusion constant is consistent between different temperatures, showing that the analysis is physically reasonable. Included in this figure is also the diffusion constant extracted from conductivity experiments and derived from first-principles calculations. It is evident that the diffusion con­stant obtained from NSE and conductivity experiments are comparable, which implies that already on a length-scale as short as * 20 A the effect of potential local traps or other “imperfections’’ in the structure that can be expected to affect the proton dynamics, has averaged out. That is, there are no new features revealed on a larger length-scale that have not been experienced by the proton on the shorter length-scale probed by NSE. However, by extending the Q range to higher Q values it should be possible to observe the crossover from single-exponential behaviour at low Q values, typical for long-range proton diffusion, to a more complex behaviour at larger Q values, suggesting that several processes are taking place. Further work along these lines is likely to give answers to questions like how the type and concentration of dopant atoms correlate with the macroscopic proton-conductivity, which as discussed above is a topic of some controversy.

Several researchers claim that the dopant atoms act as localized trapping-centres where the proton spends an extended time before it diffuses further throughout the material. This view was first introduced by Hempelmann et al. [32, 33] on the basis of their QENS data for SrYb0 05Ce0.95O2.975, which could be described by a so — called “two-state” model, suggesting that the proton migration takes place through a sequence of trapping and release events (Fig. 9.11 (left)). This view later found support from muon spin relaxation experiments [66] and computational studies of proton dynamics in perovskite-type oxides [34, 67-70], and most recently from a combined thermogravimetric and a. c. impedance spectroscopy study [71], as well as from luminescence spectroscopy measurements [72]. Converse to this picture, Kreuer et al. [73, 74] proposed that the dopant atoms may affect the proton transport in a more nonlocal fashion (Fig. 9.11 (right)). This suggestion is based on con­ductivity data for Y-doped BaCeO3 [73, 74], which shows that the proton conductivity increases with dopant level, but not as a result of a decrease of the pre­exponential factor, D0, in the expression for the diffusivity D = D0exp(—Ea/kBT) as anticipated by the two-state model, but rather as a result of an increased activation energy, Ea [73, 74].[13] Moreover, Mulliken population analyses of the electron densities at the oxygens showed that the additional negative charge introduced by

the dopants is distributed rather homogeneously over the oxygen lattice, which results in stronger bonding of the protons with increased proton-transfer barriers in general [73].

Irrespective of whether the dopant atoms influence the proton diffusion in a spatially restricted way or more non-locally, further QENS investigations using time-of-flight, backscattering, and spin-echo methods, is likely the only way to experimentally elucidate the mechanistic detail of proton dynamics in proton­conducting perovskites. Such information is crucial for the development of strat­egies for the strategic design of new materials with conductivities beyond the current state-of-the-art materials and hence are critical for future breakthroughs in the development of intermediate-temperature fuel-cell technology.

Hydrogen can readily be stored in large quantities as a cryogenic liquid for sta­tionary applications, e. g. for use as a rocket fuel, since the size and weight of the storage tanks are not limited in practice. This is, however, not the case for mobile applications, for which the US Department of Energy (DOE) has established some well-known criteria for a working system that would provide a driving range of 300 miles (480 km) for a hydrogen fuelled car. These include both gravimetric and volumetric storage capacities, operating conditions, and several other important factors. The requirements for a hydrogen storage system for mobile applications can be summarized in a qualitative way as follows:

(i) Appropriate thermodynamics (favourable enthalpies of hydrogen absorp­tion and desorption).

(ii) Fast kinetics (quick uptake and release).

J. Eckert (H)

Department of Chemistry, University of South Florida, Tampa, FL, USA e-mail: juergen@usf. edu

W. Lohstroh

Heinz Maier-Leibnitz Zentrum (MLZ), Technische Universitat Mflnchen,

Garching, Germany

e-mail: wiebke. lohstroh@frm2.tum. de

© Springer International Publishing Switzerland 2015 205

G. J. Kearley and V. K. Peterson (eds.), Neutron Applications in Materials for Energy, Neutron Scattering Applications and Techniques,

DOI 10.1007/978-3-319-06656-1_8

(iii) High storage capacity.

(iv) Effective heat transfer.

(v) High gravimetric and volumetric densities (light in weight and conserva­tive in use of space).

(vi) Long cycle lifetime for hydrogen absorption/desorption.

(vii) High mechanical strength and durability of material and containers.

(viii) Safety under normal use and acceptable risk under abnormal conditions.

Most of the considerable effort to realize such a working system has been focused on increasing the capacities of storage materials in terms of wt.%. The ultimate requirement specified by the US DOE (updated to 7 wt.% for the entire system, not just the storage medium) requires that the hydrogen molecules be packed much more closely than they are in liquid H2. This can, of course, more readily be accomplished if hydrogen is stored in atomic form, which occupies much less space. In these cases hydrogen is rather strongly bound, as, for example, in chemical compounds, which does, however, make desorption kinetics and revers­ibility more difficult if not impossible.

The types of materials that are candidate hydrogen-storage systems (Fig. 8.1) may briefly be summarized as follows:

(1) Hydrogen in metals, such as FeTiHx and LaNi5H6, that were actively inves­tigated in the early 1980 and tested for vehicular use. Here H2 dissociates at the metal surface and forms a solid solution with the metal. The hydrogen gas can be released at elevated temperatures. These materials can have very high volumetric capacities, but their gravimetric capacities are low because of the high densities of the metal hosts.

(2) In so-called complex or light metal hydrides, where hydrogen is essentially covalently bonded to metals such as Al or Li in the form of a chemical compound. Elevated temperatures and/or a catalyst are required for a release of hydrogen, and on-board regeneration is not readily possible.

(3) Chemical hydrides, such as hydrocarbons, ammonia, or amino compounds, which have the highest hydrogen content (both volumetric and gravimetric), but where the release of hydrogen is only by a chemical reaction, and not all of the hydrogen content is available under reasonable conditions. On-board regeneration does not seem feasible.

(4) Molecular hydrogen adsorbed on surfaces, such as carbons of various types, or inside large pore materials, such as zeolites, clathrates, and metal-organic frameworks (MOFs). High volumetric and gravimetric capacities have been achieved in cases where the surface area is very large, but at operating conditions (low temperature and elevated pressures) that are not currently considered to be suitable for vehicular applications, on account of the weak interaction (physisorption) with the host material.

(5) Molecular hydrogen stored as a gas or liquid. These are straightforward, established technologies, and therefore are typically used in cars used to test hydrogen fuel-cells, or simply for combustion in a normal engine.

Targets for the capacities of a hydrogen storage system have been somewhat modified by the US DOE from those of the original ‘Freedom Car’ program. The emphasis has shifted towards system targets that must be met, i. e. the weights of the containers, and all associated plumbing, valves etc. must be included. This can easily be a factor of two, so that a material needs to have an intrinsic capacity of 12 wt.% recoverable H2 to meet the system target (for 2015) of 6 wt.%.

Prospects for a practical hydrogen storage system based on the DOE guidelines basically depend on two critical, but very difficult developments (Fig. 8.1): (1) in the case of sorption-based systems binding energies of hydrogen to the host must be substantially increased so that the necessary capacities can be achieved at room temperature and modest pressures, while (2) for chemical or light metal hydrides desorption conditions must be improved to the point where all available hydrogen is released at or below 100 °C, and, more importantly, on-board regeneration becomes facile. DOE guidelines, for example, also specify a refueling rate of 2 kg H2/min in addition to the better-known targets on gravimetric and volumetric capacities. Pro­gress in these areas ultimately depends on materials synthesis, namely the ability to design and to produce (on a large scale, at low cost) the storage material, which overcomes the current limitations. While much progress in synthesis can be made by applying chemical principles in conjunction with standard thermodynamic mea­surements (i. e. adsorption isotherms), the use of advanced, molecular-level char­acterization methods should provide additional essential details on the interactions of hydrogen with the host material, and this is the type of information which serves
both as input and all-important validation of the extensive computational studies, which are also being employed in the search for suitable hydrogen storage media.

Perhaps the most powerful experimental methods for molecular-level studies of hydrogen in materials involve the use of neutrons because of their outstanding sensitivity to hydrogen. Neutrons can provide both the rotational and vibrational dynamics of the adsorbed hydrogen, be it H2, H, or one of many other forms in the complex or chemical hydrides, as well as diffusive motions through the material. The dynamics of the molecule are affected by its surroundings and can therefore be taken as an indirect measure of its interaction. The microscopic diffusion of hydrogen is critical for the desired rapid loading, and release, but has not been studied in great detail to date.

In order to understand the performance of PEM fuel cells, it is necessary to not only understand the structure but also the water dynamics in these materials over a large range of humidities, temperatures, and processing conditions. Proper water man­agement is critical to optimal fuel-cell performance. Many studies have focused on understanding the bulk-water transport and there has been a considerable effort to understand how this transport is related to the nanostructure of the membrane.

While bulk-water transport is an important aspect of fuel-cell operations, it is important to keep in mind that this macroscopic property is governed by the nano­structure of the membrane and, ultimately, the local-water dynamics within this structure. Researchers have used neutron spectroscopic techniques to investigate the local-water dynamics within the nanoscale ionic aggregates present in the material. In addition to understanding the dynamics and transport, the static and dynamic water concentration-gradient that is present across the MEA during fuel-cell operation is also a central piece of data for proper water management. Current neutron imaging techniques do not have the spatial resolution to map the water gradient across the thickness of the MEA during fuel-cell operation, but researchers have been able to use SANS techniques to elucidate this information in a clever way.

While the focus has mainly been on water, little effort has been spent on understanding the relationship between the water dynamics (both local and mac­roscopic) on the polymer-chain dynamics and fluctuations within, and of, Nafion®’s complex morphological features. Page and co-workers have used QENS and NSE techniques to address these issues.

The following sections summarize the work, to date, using QENS, SANS, and NSE to study the transport and dynamics in PEM materials, particularly Nafion®.

The basic chemical formula of perovskite-type oxides is ABO3, where A represents a relatively large cation with oxidation state +1, +2, or +3, and B is a cation with oxidation state between +1 and +7. For proton-conducting perovskites, A is usually +2 (e. g. Sr2+ or Ba2+) and B is usually +4 (e. g. Ce4+ or Zr4+). The “ideal” perovskite structure is defined by a cubic lattice of corner-sharing BO6 octahedra and 12-fold coordinated A site ions, see Fig. 9.3a. A cubic structure is, however, typically observed only if the sizes of the cations are compatible with the sizes of their respective interstices, and if not, the perovskite adopts a structure of lower symmetry. The deviation from cubic symmetry may be quantified with the

Goldschmidt factor, tG, which has a value of one or close to one for a cubic perovskite [25].[12]