Hydrogen Incorporation in Zintl Phases and Transition Metal Oxides– New Environments for the Lightest Element in Solid State Chemistry Reji Nedum Kandathil - DocsLib

sign in sign up

<<

Home , Calcium disilicide, Calcium hydride, Oxyhydride

!"# "$%!! & '( ) ' *"+,%

- . / %0 . '1 / %0 / / 1 20 %-0 / ' 1 / % . '3 ' 4 '53'6' '6' %4%7. 1 ' ' / ' / 5 . %-. '5$ 8' 39(':466 1 1 '0 '1 %; ' +: $: 1 %< '66 5$ 81 % 1"!!= 660 /5$ 80 >/1 %- 3 4

. 3 4% "!! ' 0$ 1 / 0$ %- ?@A 0$1 1 1 3B4' 0B: % ,- 2$ / ,- 2$/0/1 % B0'0 '0 'C 08B,081 1 / % 1 DE 1 3E:4' " 0B:%- 0 ,- 2$/0/1 1'1 1 %< 2 ,- 2$%- 2 ' ' ,- 2$/0F3/41 / !%+ !%!G @!% 3/4H@ 1 1 %

!"# IJJ %% J K 9 II I I "8"GLL

<,B #L "#+8 #L ! <,B #L "#+8 # !+

! " # $ %""&

'"!+ "

Hydrogen incorporation in Zintl phases and transition metal oxides– new environments for the lightest element in solid state chemistry Reji Nedum Kandathil

Hydrogen incorporation in Zintl phases and transition metal oxides – new environments for the lightest element in solid state chemistry

Reji Nedum Kandathil

Doctoral Thesis 2017

Department of Material and Environmental Chemistry Arrhenius Laboratory, Stockholm University S-106 91 Stockholm, Sweden

Faculty opponent: Prof. Michael A. Hayward Department of Chemistry, University of Oxford, UK

Evaluation committee: Prof. Yvonne Brandt Andersson Department of Chemistry- Inorganic Chemistry, Uppsala University

Doc. Cesar Pay Gomez Department of Chemistry- Structural Chemistry, Uppsala University

Doc. Mikael Kritikos AB Sandvik Coromant

Substitute: Prof. Kalman J Szabo Department of Organic Chemistry, Stockholm University

©Reji Nedum Kandathil , Stockholm University 2017

ISBN print 978-91-7649-789-0 ISBN PDF 978-91-7649-790-6

Printed by Universitetsservice US-AB, Stockholm 2017 Distributor: Department of Materials and Environmental Chemistry, Stockholm University

To my beloved family

Abstract

This PhD thesis presents investigations of hydrogen incorporation in Zintl phases and transition metal oxides. Hydrogenous Zintl phases can serve as important model systems for fundamental studies of hydrogen- metal interactions, while at the same time hydrogen-induced chemical structure and physical property changes provide exciting prospects for materials science. Hydrogen incorporation in transition metal oxides leads to oxyhydride systems in which O and H together form an anionic substructure. The H species in transition metal oxides may be highly mobile, making these materials interesting precursors toward other mixed anion systems.

Zintl phases consist of an active metal, M (alkali, alkaline earth or rare earth) and a more electronegative p-block metal or semimetal component, E (Al, Ga, Si, Ge, etc.). When Zintl phases react with hydrogen, they can either form polyanionic hydrides or interstitial hydrides, undergo full hydrogenations to complex hydrides, or oxidative decomposition to more E-rich Zintl phases. The Zintl phases investigated here com- prised the CaSi2, Eu3Si4, ASi (A= K, Rb) and GdGa systems which were hydrogenated at various temperature, H2 pressure, and dwelling time conditions. For CaSi2, a regular phase transition from the conventional 6R to the rare 3R took place and no hydride formation was observed. In contrast, GdGa and Eu3Si4 were very susceptible to hydrogen uptake. Already at temperatures below 100 ºC the formation of hydrides GdGaH2-x and Eu3Si4H2+x was observed. The magnetic properties of the hydrides (antiferromagnetic) differ radically from that of the Zintl phase precursor (ferromagnetic). Upon hydrogenating ASi at temperatures around 100 °C, silanides ASiH3 formed which contain discrete complex - ion units SiH3 . The much more complicated β – α order-disorder phase transition in ASiH3 was evaluated with neutron powder diffraction (NPD), 2H NMR and heat capacity measurements.

A systematic study of the hydride reduction of BaTiO3 leading to perovskite oxyhydrides BaTiO3-xHx was done. A broad range of reducing agents including NaH, MgH2, CaH2, LiAlH4 and NaBH4 was employed and temperature and dwelling conditions for hydride reduction exam- ined. Samples were characterized by X-ray powder diffraction (XRPD), thermal gravimetric analysis and 1H NMR. The concentration of H that can be incorporated in BaTiO3-xHx was found to be very low, which is in contrast with earlier reports. Instead hydride reduction leads to a high concentration of O vacancies in the reduced BaTiO3. The highly O- deficient, disordered, phases – BaTiO3-xHy□(x-y) with x up to 0.6 and y in a range 0.05 – 0.2 and (x-y) > y – are cubic and may represent interesting materials with respect to electron and ion transport as well as cataly- sis.

Keywords: Zintl phases, metal hydrides, transition metal oxyhydrides, XRPD, NPD, Rietveld refinement.

List of Papers

This thesis is based on the following papers

PAPER I: The 3R polymorph of CaSi2. Reji Nedumkandathil, Daryn Benson, Jekabs Grins, Kristina Spektor and Ulrich Häussermann. Journal of Solid State Chemistry 222, 18- 24 (2015) DOI: 10.1016/j.jssc.2014.10.033 My contribution: Synthesis planning, synthesis and structural characterization of the compounds; XRPD data collection, contribution to Rietveld analysis; SEM imaging; contributions towards manuscript writing and figure preparation.

PAPER II: Hydrogenation Induced Structure and Property Changes in GdGa. Reji Nedumkandathil, Verina Kranak, Robert Johansson, Jonas Ångström, Oliver Balmes, Mikael S. Andersson, Per Nordblad, Ralph H. Scheicher, Martin Sahlberg and Ulrich Häussermann. Journal of Solid State Chemistry 239, 184- 191 (2016) DOI: 10.1016/j.jssc.2016.04.028 My contribution: Synthesis planning, synthesis, hydrogenation and structural characterization of the compounds; XRPD and HTXRD data collection, contribution to Rietveld analysis; contributions towards manuscript writing and figure preparation.

PAPER III: Hydrogen Induced Structure and Property Changes in Eu3Si4 Reji Nedumkandathil, Mikael S. Andersson, Per Nordblad, Robert Johansson, Jekabs Grins and Ulrich Häussermann In manuscript My contribution: Synthesis planning, synthesis and structural characterization of the compounds; Rietveld refinement of XRPD and synchrotron data; contributions towards manuscript writing and figure preparation

PAPER IV: Investigation of the order-disorder rotator phase transition in KSiH3 and RbSiH3 Reji Nedumkandathil, Aleksander Jaworski, Andreas Fischer, Carin Österberg, Yuan-Chih Lin, Maths Karlsson, Jekabs Grins, Andrew J. Pell, Mattias Edén and Ulrich Häussermann The Journal of Physical Chemistry C (2017) DOI: 10.1021/acs.jpcc.6b12902 My contribution: Synthesis planning, synthesis and structural characterization of the compounds; NPD data collection and analysis by GSAS refinement; contributions towards manuscript writing and figure preparation.

PAPER V: Hydride Reduction of BaTiO3 - Oxyhydride vs O-Vacancy formation Reji Nedumkandathil, Aleksander Jaworski, Jekabs Grins, Diana Bernin, Maths Karlsson, Carin Österberg, Alexandra Neagu, Cheuk-wai Tai, Andrew J. Pell, and Ulrich Häussermann In manuscript My contribution: Synthesis planning, synthesis and structural characterization of the compounds; XRPD, HTXRD, TGA data collection, contributions towards manuscript writing and figure preparation

Reprints were made with permission from the publishers

Contents

1. Introduction ...... 1

2. Materials and methods ...... 9 2.1 Synthesis ...... 9 2.1.1 Starting materials for Zintl phases and Zintl phase hydrides ...... 9 2.1.2 Preparation of Zintl phases ...... 10 2.1.3 Experimental set up for Zintl phase hydrogenations ...... 11 2.1.4 Starting materials for synthesizing transition metal oxyhydrides...... 13 2.1.5 Experimental setup for hydride reduction synthesis ...... 14 2.1.6 Significance of in-situ measurements (in-situ XRD) ...... 15 2.2 Characterization techniques ...... 16 2.2.1 X-ray powder diffraction ...... 17 2.2.2 Thermogravimetry ...... 21 2.2.3 Raman ...... 22 2.2.4 Nuclear magnetic resonance spectroscopy ...... 23 2.3 The role of neutrons, NPD and QENS ...... 24 2.3.1 Quasielastic neutron scattering ...... 27

3. Zintl phases and Zintl phase hydrides ...... 29

3.1 CaSi2 (paper I) ...... 29 3.2 GdGa (paper II) ...... 32

3.3 Eu3Si4 (paper III) ...... 39

4. Investigation of the order-disorder phase transition in KSiH3 and RbSiH3...... 45

4.1 Structural relationships between α and β-ASiH3 ...... 46 4.2 Phase transition probed by NPD ...... 50 4.3 Phase transition probed by heat capacity measurements ...... 52 4.4 Coexistence of α and β at low temperatures...... 54 4.5 Conclusion ...... 57

5. Hydride reduction of BaTiO3...... 59

5.1 Reduction of BaTiO3 with CaH2 ...... 61 5.2 1H NMR investigations...... 65 5.3 Other reducing agents ...... 69 5.4 Conclusions ...... 69 6. Conclusions and future perspectives ...... 71

7. Populärvetenskaplig Sammanfattning ...... 73

8. Acknowledgements ...... 75

9. Bibliography ...... 79

markable in several respects. It represents a defect/vacancy-free solid solution of O2- and H- ions which commonly form the octahedral environment around Ti (in a mixed IV/III oxidation state). During this thesis, the hydride reduction of BaTiO3 was investigated with various metal hydrides as reducing agents. At variance with the original reports it was found that only a small concentration of H replaces O in the anionic substructure of BaTiO3. Instead, highly O-deficient, disordered, phases BaTiO3-xHy□(x-y) with x up to 0.6 and y in a range 0.05 – 0.2 and (x-y) > y were obtained.

8 2. Materials and methods

2.1 Synthesis

2.1.1 Starting materials for Zintl phases and Zintl phase hydrides All materials were stored inside an argon filled glovebox with both moisture content and oxygen content less than 0.1 ppm. The purity and supplier of these materials are given in Table 2.1.

Table 2.1: Material specifications Starting material Purity Supplier (CAS) Calcium (Ca) 99.5 % Sigma-Aldrich (7440-70-2) Gadolinium (Gd) 99.99 % Sigma-Aldrich (7440-54-2) Gallium (Ga) 99.999 % Sigma Aldrich (7440-55-3) Europium (Eu) 99.9 % ABCR (7440-53-1) Potassium (K) 99.9 % Sigma Aldrich (7440-09-7) Rubidium (Rb) 99.9 % Sigma Aldrich (7440-17-7) Silicon (Si) 99.999 % Sigma Aldrich (7440-21-3) H2 gas 99.99999 % AGA (1333-74-0) D2 gas 99.997 % AGA (7782-39-0)

All the metals were cut into small pieces. Eu was stored inside mineral oil and this oil was washed away thoroughly with n-hexane before the Eu metal was cut into pieces.

9 2.1.2 Preparation of Zintl phases

Most of the reagents for Zintl phase synthesis and the actual Zintl phases are air and moisture sensitive. Hence, the Zintl phases were prepared inside an argon filled glovebox from LC Technology Solutions Inc. (O2 and moisture < 0.1ppm). Arc melting was the common method to produce Zintl phases here. A continuous high voltage and low current dis- charge as an electric arc is generated between two electrodes. This arc can be used either to weld or melt conductive materials (Figure 2.1a).

Figure 2.1a: Sample preparation inside glove box and arc melting apparatus inside glove box for synthesizing Zintl phases

Water cooled Cu hearth and the welding W tip acted as electrodes. The arc melting was done using Centorr/Vacuum industries model 5SA single arc furnace that was installed in the glove box. At times arc melting is a little challenging since especially alkali and rare earth metals may be volatile at high temperatures. In these cases, a slight excess from the stoichiometric proportions were applied to account for the loss. The reaction conditions and nominal compositions were altered in order to attain a uniform phase without impurities.

CaSi2 The Ca metal was chopped into small pieces and Si was ground in an agate mortar. A slight excess of Si was used here (Ca/Si =1:2.02). Both reactants were pressed into a single pellet (~200 mg) and arc melted three to four times by flipping between each melting. The 6R-phase of CaSi2 was formed with a small impurity of Si. This Zintl phase is stable in air.

10 GdGa Pieces of Gd and Ga were placed one over the other in the arc furnace and arc melted inside the glove box. A 2% excess of Gd was taken con- tributing the total mass of a sample to be ~1g. Samples were melted five to six times and flipped between each melting to get good homogeneity. Arc melted samples were then sintered at 900 °C in an evacuated fused silica ampoule for 10 – 20 days, following the recipe according to [22], to “homogenize” the sample. However, GdGa samples before and after this annealing appeared to be identical.

Eu3Si4 Stoichiometric mixtures of Eu (pieces) and Si (powder) were pressed in- to a pellet and melted five to six times and flipped between each melting to ensure homogeneity. Slight changes in the stoichiometry or insuffi- cient re-melting lead to the presence of EuSi and EuSi2 in the product. Most homogeneous samples Eu3Si4 were obtained when pre-reacting stoichiometric amounts of Eu and Si - enclosed in a Ta ampoule - in an RF furnace for 2 h at about 1000 ºC into a mixture of Eu3Si4, EuSi, and EuSi2, and subsequently arc-melting this mixture.

ASi (A= K, Rb) The Zintl phases KSi and RbSi (ASi) were prepared from stoichiometric amounts of alkali metals and Si. Here, of course, arc melting is not a suitable method. Instead, the metal component was placed inside a stainless-steel tube ampoule and Si powder was pressed into pellet and placed on top of the metal. Ampoules were sealed by arc welding and subsequently heated inside a box furnace at a rate of 5 °/min to 600 °C, held for 30 h and then cooled to room temperature at a rate of 5 °/min.

2.1.3 Experimental set up for Zintl phase hydrogenations

Most Zintl phases and their hydrides are air and moisture sensitive. Hence, all steps of preparation for Zintl phase hydrogenation and the re- covery of Zintl phase hydride products was performed inside an argon filled glovebox from LC Technology Solutions Inc. (O2 and moisture < 0.1ppm). For hydrogenations, Zintl phases (obtained e.g. from arc melting) were ground into powders. Typically, batches of ~200 mg were pressed into pellets which were placed inside a corundum (Al2O3) crucible. These crucibles were then placed inside a custom-made stainless steel autoclave reactor. Autoclaves were sealed, taken outside the glove box and connected to a H2 gas manifold. At the manifold autoclaves were evacuated and subsequently pressurized with H2 gas. Thereafter,

11 er dwelling time (5 days). These synthesis experiments targeted materials for neutron scattering investigations. So, large batches (3.5 g KSiD3 and 4.5 g RbSiD3, 2.5 g KSiH3 and 2 g RbSiH3) had to be prepared. ASi, ASiH3 and ASiD3 are highly air/moist sensitive. Outside the glove box these materials, especially ASi, can ignite suddenly!

2.1.4 Starting materials for synthesizing transition metal oxyhydrides.

Barium titanate with an average particle size of 500 nm was used as the starting material for synthesizing oxyhydrides through metal hydride reduction. The metal hydrides used in this thesis are CaH2, NaH, MgH2, NaBH4 and NaAlH4. The purity and supplier of these materials are given in Table 2.2.

Table 2.2: Material specifications Starting material Purity Supplier (CAS) Barium Titanate 99.9 % ABCR (BaTiO3) (12047-27-7)

Calcium hydride 99.99 % Aldrich (CaH2) (7789-78-8)

Sodium hydride 95 % Aldrich (NaH) (7646-69-7)

Magnesium hy- 96.5 % Aldrich dride (MgH2) (7693-27-8)

Sodium borohy- 98 % ABCR dride (NaBH4) (16940-66-2)

Sodium aluminum 93 % Aldrich hydride (NaAlH4) (13770-96-2)

13

2.1.5 Experimental setup for hydride reduction synthesis In a hydride reduction a ternary transition metal oxide is reacted with a metal hydride (typically CaH2, NaH, or LiH) at relatively low temperature conditions (150 – 600 °C). The transition metal in the oxide may get reduced, and reduction is then accompanied with O vacancy formation. The expected product for BaTiO3 from metal hydride reduction, however, is the oxyhydride phase BaTiO3-xHx, according to

BaTiO3 + 2x/n MHn → BaTiO3-x + 2x/n MOn/2 + x H2

That is, upon Ti(IV) reduction hydride ions are inserted into BaTiO3, and the product has no O vacancies [21].

The starting material BaTiO3 was dried at 200 °C overnight in order to remove surface water. The sample preparations here too were performed under Ar atmosphere inside the glove box. In a typical synthesis, about 1.4 g of BaTiO3 was mixed with metal hydride by grinding the materials together in an agate mortar. The molar proportions BaTiO3 + n H, with n = 0.2, 0.6, 1.2 and 1.8 and H = NaH, 0.5 CaH2, 0.5 MgH2, 0.25 NaAlH4 and 0.25 NaBH4 were used. BaTiO3/metal hydride mixtures were then pelletized and the pellet sealed inside a stainless-steel ampoule, which in turn was placed in a fused silica jacket. Silica jackets were then placed in a vertical tube furnace, outside the glove box, and connected to a vacuum line. The empty space of the furnace was filled with silica insulation wool and a K-type thermocouple was introduced parallel to the silica jacket to monitor temperature close to the location of the metal ampoule. The silica jacket was then evacuated and ampoules heated for 1 – 7 days. Reaction temperatures were var- ied from 500 to 700 °C (Figure 2.2). Products were washed with 0.1 M CH3COOH to remove excess metal hydride and metal oxide formed during hydride reduction, and dried at 120 °C under dynamic vacuum. For washing, the pellets were crushed and sonicated with 50 mL acidic washing agent for 15 min and then centrifuged. The procedure was repeated 3 times. As a last step, samples were treated with 20 mL pure methanol. Selected reduced BaTiO3 samples were subjected to a post-treatment for which pellets were heated in a corundum crucible in a stainless-steel autoclave at 600 °C under a pressure of 30 – 50 bar of hydrogen gas during 24 h.

14

2.2.1 X-ray powder diffraction XRPD is an important tool for phase identification and quantification, unit cell determination and structure solution and refinement of both Zintl phase materials and oxyhydrides materials. According to Bragg’s law (eq. 2.1), crystal lattice can be seen as a series of crystallographic planes with Miller indices hkl. Incident X-rays have a wavelength λ, usually which is in the range of the interatomic distance within the solids. X-ray photons are scattered when they interact with atoms on the planes which are separated by an interplanar distance dhkl. A constructive interference of the coherent X-rays yields Bragg’s law.

Figure 2.4: A graphical model of Bragg’s law.

Bragg’s law ݊ߣ ൌ ʹ݀ ߠ (2.1) where n is the diffraction order (usually n=1) , d is the interplanar distance and θ is the scattering angle. The XRPD pattern is usually measured as a 1D function of 2θ, and the peak positions are defined by the crystalline lattice dimensions and the intensities are related to the type of scattering atoms, their positions within the unit cells and thermal vibra- tions assigned with them. An example of BaTiO3 powder patterns are shown in Figure 2.5.

17

The purpose of Rietveld refinement is to match the calculated intensities as close as possible with the observed ones, thereby minimising the function ϕ.

௡ ௢௕௦ ௖௔௟௖ ଶ (2.2) ߶ ൌ෍ ܹ௜ሺܻ௜ െܻ௜ ሻ ௜ୀଵ

Qualities of refinement are judged by a set of parameters, namely

profile residual, Rp = ∑ |y −y |/ ∑ y i c,i i weighted profile residual, Rwp = [∑ w |y −y |2/ ∑ w y 2]1/2 i i c,i i i expected profile residual, R = [(n − p)/ ∑ w y 2]1/2 exp i i goodness of fit, χ2 = [R /R ]2 wp exp Bragg residual, R = ∑ |I −I |/ ∑ I B obs cal obs R factor RF = ∑ |Fobs − Fcal |/ ∑ Fobs

Besides these structural parameters, some non-structural ones such as background, half width, zero shifts etc. are also refined. Phase fractions of each mixture can be extracted by this method. Also, effects caused by preferred orientations and absorption can be minimised effectively. Sometimes only Le Bail fits [26] were performed in order to extract unit cell parameters from the peak positions without refining peak intensities. Rietveld refinements and Le Bail fits were performed using the Full Prof package [27]. For a full Rietveld refinement the refined parameters in- cluded the background, the scale factor, the zero shift, unit cell parameters, peak shape function, atomic coordinates, and the thermal displace- ment coefficients. In some rare cases (e.g. paper 1) issues from anisotropic peak broadening were solved by using spherical harmonic func- tions [28]. A Rietveld refinement pattern is shown in Figure 2.7 Finally, temperature dependent (in situ) XRPD can be used to study the thermal stability of samples with respect to phase changes. In situ XRPD measurements were recorded on a PANalytical X’Pert Pro diffractome- ter using an Anton Paar XRK900 reaction chamber upon heating from room temperature to 900 °C under air or vacuum and then subsequent cooling back to room temperature. For paper II, powder samples were mounted on a Si wafer and XRPD was measured within the 2θ range 5 – 90 ° under dynamic vacuum while heating up to 450 °C. In paper V, the oxidation of oxyhydrides back to perovskite was studied by heating the furnace to 900 °C while the sample was mounted on an Au wafer. Some

20

Figure 2.9: Weight increase with H content in TGA analysis of BaTiO3-xHx synthesized using CaH2

For an oxyhydride oxidation the molar mass increase will be M O2- MH2 = 16 g/mol – 1.08g/mol ~ 15 g/ atom If ‘m’ is the mass of BaTiO3 after oxidation and ‘n’,- is the number of moles of BaTiO3, then the weight of the sample after oxidation will di- rectly give ‘n’ which is same as for the oxyhydride.

n BaTiO3-xHx + n 0.75x O2 → n BaTiO3 + n 0.5x H2O (2.4)

If Δm being the weight gain during oxidation,

୼୫ (2.5) x = ቂ ቃൈ݊ ଵହ௚

2.2.3 Raman Raman spectra were measured using a Labram HR 800 spectrometer. The instrument is equipped with an 80 mm focal length spectrograph and an air cooled (-70 °C) charge-coupled device (CCD) detector. The samples were typically enclosed in a glass capillary and were excited with a double frequency Nd:YAG laser (532 nm). The Raman spectra were collected at room temperature with an exposure time of 60–300s and an accumulation number of 1–3 (exposure and accumulation largely depend on the signal to noise ratio of the spectra and the laser sensitivity of the sample). Using an 1800 lines/mm grating and a 50X Mitutoyo ob-

22

encing. 2H NMR spectra were recorded upon cooling in the temperature range from 300 to 200 at 10 K intervals. A measurement was also performed on KSiD3 at 290 K upon heating from 200 K. Before starting each final acquisition, the sample was held for 30 min to equilibrate after reaching the desired temperature (stabilized within ±0.1 K). A powder of Pb(NO3)2 was used for the temperature calibration. The 1H MAS NMR experiments in paper V were performed at the 1 magnetic field B0= 14.1T (600.12 MHz H Larmor frequency) and MAS

rate νr =60.00 kHz on a Bruker Avance-III spectrometer equipped with 1.3 mm MAS probe head. Acquisitions involved rotor-synchronized, double-adiabatic spin-echo sequence with 90 o 1.2μs excitation pulse followed by two 50.0μs tanh/tan high-power adiabatic pulses (SHAPs) [30] [31] with 5 MHz frequency sweep. All pulses operated at the nutation fre-

quency νnut=208 kHz. 4096 signal transients with 5s relaxation delays were accumulated for each sample. Shifts were referenced with respect to tetramethylsilane (TMS).

2.3 The role of neutrons, NPD and QENS Neutron scattering is one of the most powerful and versatile experimental methods to study the structure and dynamics of materials on the atomic and nanometer scale. In contrast to the interference of X-rays with electrons, the scattering processes for neutrons are mainly through interaction with the atomic nucleus and unpaired electrons. Thermal neutrons have a wavelength (around 1.8 Å) similar to inter-atomic distances, and energy (around 25 meV) similar to elementary excitations in solids. Therefore, simultaneous information on the structure and dynamics of materials is obtained. The neutron scattering cross section varies in an almost random fashion between elements and even between different isotopes of the same element as shown in Figure: 2.11. Thus they can be used to study light isotope like hydrogen, which is almost invisible to X- rays. With neutrons, the large difference in scattering between usual hydrogen (1H) and deuterium, (2H) can be used to change the contrast in scattering. Neutron powder diffraction (NPD) and X-ray diffraction (XRPD) methods are complimentary to each other For XRPD the scattering power are proportional to the atomic number, Z of a specific atom while that for neutrons varies non-linearly with the atomic numbers [32]

24

3. Zintl phases and Zintl phase hydrides

This chapter summarizes the findings from studying the hydrogenation behavior of the Zintl phases CaSi2, GdGa and Eu3Si4. These Zintl phases were prepared from the elements by arc melting stoichiometric amounts of elements. Hydrogenation experiments were performed at temperatures ranging from room temperature up to 700 °C and using pressures from 5 up to 80 bar. Most of these experiments were terminated after 24 h and the products analyzed by XRPD. For GdGa also the time aspect of hydrogenation processes was investigated.

3.1 CaSi2 (paper I)

Calcium disilicide (CaSi2) is an archetypical Zintl phase which was already discovered in 1863. Its rhombohedral crystal structure consists of puckered hexagon layers formed by the Si atoms which are intercalated by the Ca atoms. According to the Zintl concept Si is regarded as re- - 2+ - duced Si species (i.e. Ca (Si )2) which is isoelectronic to As. Indeed, the puckered hexagon layers relate to the elemental structure of grey arsenic. However, the stacking of layers in CaSi2 differs: The translational period constitutes of six layers (6R-CaSi2) whereas the rhombohedral structure of grey arsenic is repeated after three layers (Figure 3.1)

Figure 3.1: Structures of rhombohedral grey arsenic and 6R- CaSi2

29 The expectations of hydrogenating CaSi2 are outlined below: CaSi2 is the most Si rich phase in the binary Ca-Si system [33] (referring to ambient pressure) (Figure 3.2).

Figure 3.2: Phase diagram of Ca-Si system [34]

Recently a compound CaSi6 was reported from high pressure synthesis. CaSi6 has a clathrate-like structure: Four-bonded Si atoms built an open tetrahedral framework with large cavities that encapsulate the Ca atoms as shown in Figure3.3 [35].

Figure 3.3: Clathrate-like structure of CaSi6. Si and Ca atoms are depicted as red and grey circles

30 True clathrates are afforded by high pressure synthesis in the systems [36] [37] Ba-Si (Ba8Si46) and Na-Si (Na8Si46 and Na24Si136) . At the same time, metastable clathrates Na6Si46 and K8Si46 can be obtained from hydrogenating the Zintl phases NaSi and KSi according to e.g. 46 NaSi + [38] 19 H2 = 38 NaH + Na8Si46 . H-induced oxidative decompositions of Zintl phases may represent an interesting route for the synthesis of new, metastable, materials because of the typically mild (low temperature) conditions. Metal hydride by-products can be easily removed from the typically air and moist stable clathrate materials by simple washing. The question was whether CaSi2 may undergo oxidative decomposition upon hydrogenation to CaSi6 (or other clathrate-like materials) which would represent a much more convenient synthesis compared to the application of high pressures. The results from the hydrogenation experiments are compiled in Figure 3.4. When sintering CaSi2 in a hydrogen atmosphere at temperatures above 200 °C, a structural change from the common 6R polymorph to the 3R form is observed. The phase transformation is quantitative, and so obtained 3R-CaSi2 could not be reverted back to the 6R form by subsequent temperature treatments. It was concluded that CaSi2 does not undergo the expected H-induced decomposition. However, the observation of the 6R-to-3R phase transition was surprising. 3R-CaSi2 is typically observed in epitaxial films on Si substrates where it grows on initially [39] [40] [41] formed 6R-CaSi2 . The phase diagram of the Ca-Si system, which has been recently reinvestigated, does not contain temperature [33] polymorphism of CaSi2 . According to this phase diagram, 6R-CaSi2 decomposes peritectoidal at 1050 °C into melt and Ca14Si19. The structure of 3R-CaSi2, displayed in Figure 3.4 was refined from synchrotron XRPD data. It consists of only two distinct positions, one Ca and one Si atoms, respectively. Si atoms on the site 6c (0, 0, z) in space group R-3m form puckered hexagon layers with Si–Si distances of 2.42 Å. These are oriented and stacked in an ABC fashion as in grey arsenic shown in Figure3.1. Ca atoms on the site 3a (0, 0, 0) are sandwiched between two trigonal pyramids of Si atoms and thus attain an 8- coordination with two different Ca-Si distances, 3.09 Å (×6) and 3.14 Å (×2). The stacking sequence of 6R-CaSi2 is usually specified as AA’BB’CC’ where prime and non-prime denoted layers are formed by Si1 and Si2, respectively [42] [43]. The A’B’C’ sequence of Si1 layers cor- responds to the arrangement of Si atoms in the 3R structure. The A, B, C layers of Si2 atoms are rotated by 180o (around their normal) with respect to the B’, A’, and C’ layers, respectively [44] [43].

31

Figure 3.4: Crystal structure of 3R and 6R CaSi2 (Space group R-3m). Si and Ca atoms are depicted as red and grey circles

3 The molar volume of CaSi2 in 6R-CaSi2 (65.7 Å ) is slightly smaller 3 compared to the 3R form (67.5 Å ), which suggests that 3R-CaSi2 is a high temperature modification. However, this is not clear. Also unclear remains the nature of the hydrogen initiated transition of the 6R to the 3R phase. One may speculate that the 3R phase is stabilized by small concentrations of incorporated hydrogen and/or hydrogen induced defects. Recently Castillo et al. obtained 3R-CaSi2 from regular sintering [29] (that is, without H2 atmosphere) of 6R-CaSi2 at 800 °C .

3.2 GdGa (paper II) The great majority of rare earth (RE) metals form monogallides REGa which all crystallize with the orthorhombic CrB structure featuring linear zigzag chains of Ga atoms as shown in Figure 3.5a [45] [46] [47]. RE atoms form trigonal prisms around the Ga atoms, and slabs of trigonal prisms are stacked along the [010] direction. According to the Zintl con-

32 cept and assuming trivalent RE, Ga will correspond to Ga3-, that is, the electron count per chain atom is six. The perception is then that the chain is built from singly bonded Ga species, each carrying two lone pairs Figure 3.5b. The REGa compounds display interesting magnetic properties [48] [49] [50] [51] [52] [53] [54]. In general they are ferromagnets and across the lanthanide series the Curie temperature TC increases from 32 K for PrGa to a maximum of ~190 K for GdGa and then decreases to 15 K for TmGa [55] [56] [57]. The ferromagnetic properties of REGa are a consequence of the coupling of magnetic moments arising from localized 4f electrons, which is mediated via conduction electrons (Ruderman-Kittel- Kasuya-Yosida (RKKY) coupling mechanism [58] [59] [60]). The validity of RKKY coupling, however, contradicts the simple description of REGa compounds as Zintl phases [61], because all valence electrons would be localized as bonds and lone pairs within the polyanionic Ga zigzag chains.

Figure 3.5: Orthorhombic CrB structure featuring linear zigzag chains of Ga atoms. Ga and Gd atoms are depicted as red and grey circles

To reconcile RKKY coupling within the Zintl concept one may assume that Ga is just reduced to Ga2- species, which form -bonded zigzag chains in a charge imbalanced (electron excess) Zintl phase RE3+Ga2-e- as shown in Figure 3.5. Hydrogen incorporation in REGa may give a possibility to probe the electronic structure of REGa systems and, thus, to provide insight into the interplay between chemical bonding in the polyanion and magnetic properties. In particular, itinerant electrons may be localized as H- through the formation of interstitial hydrides [19] [62]. This should result in a change of the strength and possibly also the sign of the magnetic interaction. Provided that hydrogen uptake and desorp-

33 tion are easily reversible, one could even envision switchable magnetic properties with systems REGa/REGaHx. The recent investigation of the hydrogenation behavior of NdGa gave a [63] surprising result . NdGa transformed into a hydride NdGaH2-x (x ≈ 0.33 by incorporating H into two distinct positions, H1 and H2 (Figure 3.6). Interstitial H1 atoms are exclusively coordinated by Nd atoms, whereas H2 atoms are inserted between two Ga atoms of neighbouring zigzag chains and, thus, part of a novel two-dimensional polyanion featuring linear –H2-Ga-H2-Ga-H2- chains with alternating short and long Ga-H distances. H2 deficiency (from the refinement of neutron powder diffraction data of a deuterated sample an occupancy of about 0.66 was deduced for the H2 position) results in the fragmentation of chains. For x = 0.33, arrangements with five-atom moieties Ga-H-Ga-H-Ga were found to be energetically most favourable. Upon hydrogenating NdGa the magnetic interaction changes from ferromagnetic to antiferromagnetic in NdGaH2-x. Also, hydrogenation of NdGa is reversible. At temperatures between 450 °C and 500 °C, NdGa2+x desorb H and NdGa is quan- titatively reformed [63].

Figure: 3.6: (a) Cmcm CrB-type structure of rare earth (RE) monogallides. RE and Ga atoms are depicted as grey and red circles, respectively. RE atoms are arranged as slabs of trigonal prisms hosting linear zigzag chains of Ga atoms. (b)The crystal structure of NdGaH1.66 and H atoms are depicted as green circles.

34

Table 3.1: Lattice parameters

Empirical formula GdGa GdGaH1.66 20 bar/100 °C space group Cmcm Cmcm

unit cell (Å) a = 4.340(3) a = 3.9867(7) b = 11.012(2) b =12.024(2) c = 4.105(3) c = 4.1009(6)

Gd 4c (0,y,1/4) 0.1414(3) 0.151(2)

Ga 4c (0,y,1/4) 0.4245(5) 0.445(5)

V (Å3) 196.2(4) 196.6(5)

RF 2.05 2.40

Rwp 2.65 3.06

When hydrogenating GdGa at harsher conditions, at temperatures above 500 °C and/or pressures above 50 bar, GdH2 could be identified in the XRPD of reaction products, which indicates H-induced oxidative decomposition of GdGa [19]. The complexity of the hydrogenation behavior of GdGa suggested to perform in situ diffraction XRPD experiments at a synchrotron facility (MAX IV, Lund). In these experiments GdGa was heated to 500 °C in a H2 atmosphere of 20 bar and subsequently cooled. The results are shown in Figure 3.8. Upon heating XRPD patterns changed abruptly after 35 min (at ~370 °C), and more subtly and gradually upon cooling at temperatures below 200 °C. The diffraction patterns were indexed in a C-centered orthorhombic cell. The structure change upon heating corre- sponds to an expansion of the b axis and a contraction of the a axis by 7.5 and 6.5%, respectively. The structure change upon cooling corre- sponds to a further increase of b and a further decrease of a. During heating and cooling the c parameter varies only slightly. In Figure 7 are inserted the values from DFT simulations for phases GdGaH1.0, GdGaH1.33 and GdGaH1.66. There is a close correspondence of the parameters of the product after cooling with the ones of the simulated GdGaH1.66 structure.

36

Figure 3.8: a) Densitometric view of the in-situ XRPD patterns (λ = 0.991 Å) recorded during heating and cooling of GdGa using 20 bar H2 pressure. b) Rela- tive change of lattice parameters as a function of time (red). DFT-calculated changes for GdGaH, GdGaH1.33 and GdGaH1.66 are inset as green, blue, and orange symbols, respectively. Figure adapted from Paper II.

The main findings of the In-situ experiments are: (i) Hydrogenation pro- ceeds already at low temperatures, which is seen in the gradual change of lattice parameters. Above ~350 °C kinetic barriers are absent, as indicated in the sudden change of lattice parameters. Hydrogen absorption at these high temperatures leads to the formation of a hydride phase GdGaH~1.3 which is stable to up to 500 °C. This hydride phase cannot be recovered in ex situ experiments and would have gone unnoticed without the in situ study. (ii) At lower temperatures (below 200 °C) additional H is absorbed leading to a hydride GdGaH1.66, which is also the final product in synthesis experiments using H2 pressures above 15 bar. Conse- quently, GdGaH1.66 is thermally labile. When heating GdGaH1.66 in dynamic vacuum (~ 10-5 bar) changes are apparent already at 100 °C. At temperatures above 300 °C GdGa hydride decomposes into GdH2 and another, poorly crystalline, product. Thus, in contrast with NdGa the hydrogenation of GdGa is not reversible.

37 GdGa is a ferromagnetic with a reported TC of ~190 K. Additionally there is a spin reorientation at 85 K [48] [64] [65] [66] [67]. This behavior was confirmed, as shown in Figure 3.9.

Figure 3.9: Zero field cooled M vs T curves of (a) GdGa and (b) GdGa(H)1.66 at H= 4 kA/m. Figure adapted from Paper II.

The magnetization behavior is drastically changed after the hydrogenation as GdGaH1.66 shows an antiferromagnetic transition at about 45 K. The behavior at higher temperatures was fit to the Curie-Weiss law (C/(T-CW)). The derived Curie-Weiss constant CW and the effective number of Bohr magnetons (eff) are given in Table 3.2. Both GdGa and GdGaH1.66 have an effective magneton number close to 8, which is the expected number for Gd3+. So, hydrogen incorporation in GdGa is not associated with a change of the oxidation state of Gd. The change of sign of CW after hydrogenation proofs that GdGa transforms from a ferro- magnet to an antiferromagnet upon hydrogenation.

Table 3.2: Effective number of Bohr magnetons (eff) and Curie-Weiss con- stants (CW) obtained from Curie-Weiss fits to the high temperature susceptibil- ity data.

Sample eff CW (K)

GdGa 8.4 190

GdGaH1.66 7.9 -13 .

38 3.3 Eu3Si4 (paper III) [68] Eu3Si4 is another interesting case of a ferromagnetic RE Zintl phase . Its body centered orthorhombic crystal structure (space group Immm, #71) is depicted in Figure 3.10.

Figure 3.10: a) Crystal structure of Eu3Si4 with polyanionic ribbons of planar hexagon rings projected approximately along the a (left) and the b (right) direction. b) Crystal structure of Eu3iSi4 emphasizing the arrangement of Eu2 atoms into arrays of edge condensed tetrahedra. c) Crystal structure of Eu3iSi4 emphasizing the trigonal prismatic coordination of Eu2 around Si atoms. d) Relation of the CrB and Eu3Si4 structures. Eu and Si atoms are depicted as grey and red circles, respectively. Figure adapted from Paper III

39 Si atoms form hexagon rings, which are condensed into one-dimensional ribbons that run along the b direction. The Si hexagon rings are composed of two kinds of Si atoms, Si1 and Si2. According to the Zintl concept, three-bonded (3b) Si1 atoms carry a formal charge of -1, and two- bonded (2b) Si2 atoms a charge of -2. A polyanionic strand is composed of two Si1 and two Si2 and, thus, carries a formal charge of -6. This charge is balanced by Eu cations, provided they are in a +2 state. Weit- zer et al. confirmed the divalent character of Eu from magnetic measurement, which showed a magnetic moment in agreement with a 4f7 electron configuration above 117 K when Eu3Si4 is in the paramagnetic state. Importantly, there are also two kinds of Eu atoms. Eu1 atoms are sandwiched between Si hexagon rings. Eu2 atoms form arrays of edge condensed tetrahedra, which separate Si hexagon ribbons in the c direction. Both Eu atoms together provide a trigonal prismatic coordination to Si atoms. There is an apparent close relationship to the orthorhombic CrB structure (space group Cmcm #63), which is adopted by the REGa compounds (Figure 3.10b). The Eu3Si4 structure can be considered as built up from slabs of CrB structure (with a thickness b/2) with alternating [010] and [0-10] orientation. Condensing slabs at layers of common RE atoms will then connect zigzag chains into hexagon rings. [68] Eu3Si4 displays two magnetic transitions at 117 and 40 K . The first one was attributed to a ferromagnetic ordering of the Eu2 atoms. This transition may relate to the ferromagnetic transition in REGa compounds because RE atoms in REGa have the same structural arrangement as the [55] [56] Eu2 atoms in Eu3Si4. Also, TC of REGa are at similar temperatures [57] . The second transition in Eu3Si4 is caused by a ferromagnetic ordering of the Eu1 atoms, resulting in a net ferrimagnetic ground state. As for REGa, the Zintl concept for Eu3Si4 is at variance with its ferromagnetic properties. The results from the hydrogenation experiments, which were restricted to pressures of 30 bar, are compiled in Figure 3.11. Eu3Si4 was found to be remarkably susceptible to H uptake. Almost quantitative hydride formation was already observed at room temperature during 24 experiments. At 100 °C and 200 °C a phase pure hydride was obtained. This hydride was assigned a composition Eu3Si4H2+x. The diffraction pattern of Eu3Si4H2+x could be readily indexed in a body-cantered orthorhombic cell with lattice parameters that closely resembled those of the starting material Eu3Si4 (see Table 3.3).

40

Table 3.3: Crystallographic information obtained from Rietveld refinements.

SG: Immm a (Å) b (Å) c (Å) V (Å3) R R 2 Bragg F χ

0.410 Eu3Si4 4.61499(9) 3.95949 (8) 18.2361 (5) 333.23 (1) 4.49 3.71

Eu3Si4- 4.40055(5) 3.97166 (5) 19.8098 (3) 346.225(8) 4.8 3.68 0.831 H2 200°C

Eu3Si4-

H2 300°C 4.40247(7) 3.97478 (9) 19.8646 (5) 347.61 (1) 5.35 4.06 0.941 Phase 1- 64(1) %

Eu3Si4-

H2 300°C 4.4391 (1) 3.9778 (1) 19.5784 (8) 345.72 (2) 7.22 6.01 0.941 Phase 2- 35(7)%

The high neutron capture cross section of Eu prohibits the application of neutron diffraction for structural analysis of Eu3Si4H2+x. For identify- ing possible locations for H atoms the established pattern of REGaH2-x was followed. The Eu3Si4 structure can be considered built from slabs of CrB structure and the arrangement of Eu2 atoms in “Si” free layers cor- responds to an array of edge sharing tetrahedra (Figure 3.12). The centers of these tetrahedra are occupied in REGaH2-x by one kind of H atoms (H1). Assuming the same scenario for Eu3Si4, one arrives at an interstitial hydride Eu3Si4H2 (Figure 3.12). The structure parameters of the Eu3Si4H2 model were subsequently optimized by DFT calculations. Eu3Si4H2 is stable by -0.46 eV/H atom with respect to Eu3Si4 and gaseous H2. Deviations between the lattice parameters of the DFT optimized Eu3Si4H2 structure and the ones extracted from XRPD patterns pointed to the presence of additional H in interstitials also involving Si atoms. Again guided by REGaH2-x, H was placed between two Si2 atoms of neighbouring hexagon rings as a possible location for additional H. This position centers at the same time Eu2Eu1 triangles and would relate to the H2 position in REGaH2-x. Subsequent DFT modelling of compositions Eu3Si4H3 and Eu3Si4H4 showed considerably better agreement to the experimental unit cell volumes. However, the ordered monoclinic model structures shown in Figure 3.13 did not provide a good match to

42

tiferromagnetic interactions; the latter attains a value of ~8 B which is typical for compounds containing Eu2+ 4f7 ions

It is clear that a lot more investigations are necessary in order to understand the hydrogenation behavior of Eu3Si4. As a first step, one could think of volumetric measurements for establishing the actual H content of Eu3Si4H2+x and the homogeneity range of this phase. Also, it would be interesting to understand the two-phase behavior of the hydrogenation product at 300 °C. Monitoring hydrogenation reactions by in-situ XRPD using synchrotron radiation, similar to that performed for GdGa, may help there. Further, it would be interesting to gain more information on the magnetic property changes. This holds as well for the RE- Ga/REGaH2-x systems. What is the origin of the antiferromagnetic behavior?

44

K the hydrogen storage capacity is 4.3 wt.% and there has been efforts to improve the hydrogen adsorption/desorption kinetics by finding catalysts [70]. The advantageous thermodynamic characteristics of the absorption- desorption process was attributed to an unusually low entropy variation for 2ASiH3 = 2ASi + 3H2. Typically, entropies of desorption for metal hydrides are 120 - 130 J/K mol (H2), corresponding to the gain of rotational and translational entropy of the released gaseous hydrogen (i.e. the standard entropy change of hydrogen). Accordingly, because of Tdes = Hdes/Sdes, the endothermic desorption enthalpies of metal hydride considered suitable hydrogen storage materials are typically around 50 kJ/mol H2. For ASiH3, however, desorption enthalpies are in a range 20 – 30 kJ/mol H2, which is compensated by low entropies of desorption, 55 – 70 J/K mol (H2 ) for A = K and Rb. The peculiarity with ASiH3 obtained from the hydrogenation of ASi is the presence of structural disorder. This so called form crystallizes in a + - NaCl-type arrangement of A and SiH3 ions which implies that pyrami- - dal anions SiH3 are distributed in random orientations in the crystal structure. Disorder in α-ASiH3 will imply that these materials have high molar entropies, which in turn provides a natural explanation for the low desorption entropies. The kind of disorder (static or dynamic), however, was debated for a long time. Recent quasielastic neutron scattering (QENS) experiments could then settle this question by confirming the - [71] presence of rotational-dynamical disorder of SiH3 in α -ASiH3 . The dynamical nature of the disorder is remarkable because of the low temperatures (near room temperature) it occurs. When cooling ASiH3 from room temperature, phase transition to ordered (β) forms take place. Tang et al. established the structures of -KSiH3 and -RbSiH3 from neutron [20] powder diffraction data of ASiD3 measured at 1.5 K. . -KSiH3 crystallizes with an orthorhombic Pnma structure whereas -RbSiH3 adopts the monoclinic P21/m KClO3 structure. During this thesis transition the β – α order-disorder transition of KSiH3 and RbSiH3 was carefully investigated and comprehensively characterized. It was especially interesting to see how the evolution to a - “rotator” phase with more or less freely rotating SiH3 moieties expresses itself in diffraction and various spectroscopic experiments.

4.1 Structural relationships between α and β-ASiH3

The structures of high-temperature -KSiH3 and low-temperature - KSiH3 are compared in Figure 4.2. A close relationship of the orthorhombic -phase structure to the cubic NaCl structure is seen when con- sidering planes {110} in the NaCl structure. In such planes rows of near-

46 est neighbor coordinated ions run parallel along the cubic unit cell axis directions. Corresponding planes in the orthorhombic -phase structure are the (mirror) planes (0,y,0), y = ¼, ¾. Instead of straight rows, ions are here arranged as zigzag chains, yielding a hexagon pattern with mu- tually 3-coordinated ions. The different arrangement of ions in these planes is the major difference between the two structures. The NaCl structure (-phase) is completed by stacking {110} planes with a spacing of a/√2. Atoms achieve their mutual 6-coordination by two additional nearest neighbors situated in each plane adjacent a central one in the stack. The -phase structure is completed exactly the same way. The stacking of layers occurs in the b direction with a spacing of b/2. This results in a mutual 7-coordination of ions as seen Figure 4.2c

Figure 4.2: Comparison of the crystal structure of cubic α-KSiH3 and orthorhombic β-KSiH3. (a) {110} plane in the NaCl structure (left) and corresponding plane (0, 1/4, 0) in the β-KSiH3 structure (right). Gray circles represent K cations and red circles SiH− moieties. Gray bars mark corresponding rows of ions in both structures. (b) Stacking of {110} planes in the NaCl structure (left) and (0, y, 0), y = 1/4, 3/4 planes in the β-KSiH3 structure (right). Dark and light colors differentiate atoms on different heights. (c) drawn lines are now emphasizing polyhedra around anions, octahedra for the NaCl structure (left) and monocapped trigonal prisms for β-KSiH3 (right, H atoms are depicted as green circles).

47

− The actual local coordination of SiH3 in both structures is then shown in Figure 4.3.

Figure 4.3a: (a) Two views of the monocapped trigonal prismatic coordination − environment of SiH3 ions in the β-KSiH3 structure. H ligands are coordinated by three K ions each. The resulting HSiK3 tetrahedra share edges (left). (b) Two views of the octahedral coordination environment of SiH3 ions in the α-KSiH3 − (and α-RbSiH3) structure. Gray circles represent K cations, red circles SiH moieties and H atoms are depicted as green circles

− In -KSiH3 SiH3 units are encapsulated in mono-capped trigonal prisms defined by 7 K+ cations. Each H ligand is coordinated by three K+ ions giving a quasi-tetrahedral environment. H coordination involves 6 out of the 7 K+ ions around a Si atom (Figure 4.3 a( right)). The three + HSiK3 tetrahedra share common edges. The seventh K ion around is - situated on the pseudo three-fold axis of the SiH3 anion, coordinating its “lone pair” site. The mono-capped trigonal prism formed by the alkali metal ions is also the building block of the monoclinic -RbSiH3 structure. The difference between the orthorhombic -KSiH3 and monoclinic -RbSiH3 structure is a different linking of these monocapped trigonal prisms.

48

the C3 axis of the moiety and jumps related to the 8-fold reorientations of the moiety along the body diagonals of the cubic unit cell, yielding 24 jump locations [71].

4.4 Coexistence of α and β at low temperatures.

The order-disorder (–) transition in ASiH3 displays a significant hysteresis. Upon cooling, transitions in the Cp measurements are observed at 274 K and 241 K for KSiH3 and RbSiH3, respectively (that is 26 and 38 K lower, respectively). Looking at the relation between the low and high temperature structure (Figure 4.2) it is clear that the - transition has to be reconstructive and, thus, will proceed by a nucleation-and-growth mechanism. Hysteresis has to be attributed to a lack of equilibrium in the

phase transition. It is also noticeable that in the Cp measurements, values upon heating and cooling coincide reproducibly only at temperatures considerably below the lower transition temperature. This might indicate the presence of non-equilibrium situations at temperatures below the hysteresis region.

Figure 4.8 shows 2H NMR spectra collected from static powders of ASiD3 upon cooling from 300 K to 200 K. At 300 K, the presence of - rapid dynamics of the SiD3 moieties produces one narrow resonance around 1.5 ppm for both -phases, with a full width at half maximum (FWHM) height of around 8 ppm. When the systems are cooled below the phase transition temperature (260 K for KSiD3 and 240 K for RbSiD3), the NMR spectra show the characteristic feature of a quadru- polar-broadened peak shape. The progressive broadening of the -KSiD3 NMR peak shape observed when the temperatures decreases from 240 to - 200 K are attributed to slow dynamics. At 200 K SiD3 motion becomes essentially arrested. The finding of slow dynamics in the β-phases in agreement with the report by Tang et al. seeing dynamics on a time scale 108 – 1010 jumps/s in fixed window scan neutron scattering experiments [72]. Note that the associated mobility is several orders of magnitude low- - er compared to the rapid (sub-picoseconds) dynamics of SiH3 in 15% expanded -ASiH3. According to Tang et al. it is expected that for - [72] RbSiD3 dynamics is frozen below 150 K .

54

2 Figure 4.8: H NMR spectra recorded from static powders of RbSiD3 (left panel) and KSiD3 (right panel) recorded at the as-indicated temperatures on cooling from 300 to 200 K.

The NMR spectra in Figure 4.8 reveal another important finding, namely the presence of -phase well below the phase transition upon cooling. From NMR spectra deconvolutions, the amount of -phase observed at 240 K for RbSiD3 and 270 K for KSiD3 was estimated to 20% and 3%, respectively. These values drop to 7% and 2% at 220 K (RbSiD3) and 250 K (KSiD3). This indicates that the non-equilibrium behavior between - and -phase extends beyond the range of hysteresis. There could be several reasons for this. For example, the transformation kinetics -to-, and/or the relative stability (Gibbs energy difference, G) between the polymorphs, can depend on the size of -phase

55 domains. Then -phase domains will grow at the expense of -phase during the phase transition until -phase domains obtain a critical size below which the transformation kinetics are considerably slower or, al- ternatively, G is decreased relative to the -phase. As a consequence, the -phase can appear substantially undercooled and a complete transformation will only occur after prolonged annealing at low T. Note that in the NPD experiment, performed upon heating, the -phase was only recognized at temperatures above 300 K, in agreement with the phase transition temperature determined by the Cp measurement. Therefore, it has to be assumed that undercooled-phase possesses a domain size too small for coherent scattering, i.e. crystallites are sub-Bragg (2–4 nm) sized and can only express short and medium range order. The -to- phase transition seen in diffraction and Cp experiments should then correspond to the sudden growth of -phase domains. So there can be undercooled -phase upon cooling (remnant phase) and upon heating (premonitory).

Figure 4.9 shows QENS spectra for ASiH3. The instrument used for the QENS experiments senses only fast dynamics associated with the - phase. The - phase transition upon heating is seen by the sudden broadening of the elastic line. This occurs between 300 and 305 K for KSiH3, and 270 and 285 K for RbSiH3, in agreement with the observations from the NPD and Cp measurements. Upon cooling, QENS spectra remain considerably broadened also at temperatures below the transition. This reflects the presence of remnant non-equilibrium -phase domains, as established from the NMR experiments. A small broadening is also present in the spectra upon heating before the phase transition, which has to be attributed to the presence of premonitory -phase. Thus, the QENS spectra mirror exactly the results from the 2H NMR investigation.

56

Figure 4.9: Temperature dependence of the fit to the QENS spectra for KSiH3 −1 (a, b) and for RbSiH3 (c, d) at Q = 1.9 Å , measured with 4.8 Å neutrons

4.5 Conclusion

The β-α order-disorder phase transition in ASiH3 is rather complicated and required the application of multiple techniques in order to obtain a comprehensive understanding. It is clear that the α phases are dynami- cally disordered, but there is also (much) slower dynamics present in the low temperature phases, down to about 100 K below the actual phase transition.

57

58 5. Hydride reduction of BaTiO3.

Hydride reductions of transition metal oxides has been established as a convenient low temperature route for obtaining highly reduced materials with unusual coordination environments, electronic structures, and magnetic properties [73] [74]. The method has been especially fruitful when using perovskite oxides or Ruddlesden-Popper phases as precursors. Fig- ure 5.1 shows the principle with two prominent examples, both repre- senting lanthanum nickel oxides. The reduction of LaNiO3 and La3Ni2O7 to LaNiO2 and La3NiO6, respectively, is accompanied with the removal of planes of O ions, and there is a strict topotactic relationship between the precursor oxide and its reduced form.

Figure5.1: Topochemical reductions of (layered) nickel perovskites. White and orange spheres represent A-site cations and oxide ions. B-site cations lie in the centre of the polyhedra. [73] (Figure copyright permission obtained from The Chemical Society of Japan (CSJ))

59

Metal hydrides are strong reducing agents. The standard potential is - specified as -2.25 V for the pair H /H2. However, values for actual poten- tials may be radically different in diffusion controlled solid state reactions. As a characteristic feature, metal hydride reductions work at com- paratively low temperatures (reaction temperatures are in a range between 150 and 600 °C) and popular reducing agents are the metal hydrides CaH2, NaH, and LiH although with the latter frequently side reactions seem to occur [75]. The side product of metal hydride reductions are the corresponding metal oxides, e.g. according to LaNiO3 + 2 NaH = LaNiO2 + Na2O + H2. The side product is then removed by washing samples with weakly acidic agents, e.g. with a mixture of NH4Cl and methanol. When H2 gas is produced, the reaction is thought to proceed as a one-electron process. H2 may also act reducing, and therefore “two- electron” processes producing protons could also occur, according to LaNiO3 + NaH = LaNiO2 + NaOH. However, mechanisms or processes behind metal hydride reductions are not well investigated. Some insights come from the works of Kobayashi et al, [76], Hayward et al. [77], and [75] Hermden et al. studying the reductions SrFeO2.5 to SrFeO2, LaNiO3 to LaNiO2, and Sr2MnO4 to Sr2MnO4−x. Sometimes metal hydride reduction leads to a simultaneous incorporation of H-, that is, instead of the formation of O vacancies H- replaces O2- in the oxygen substructure of the transition metal oxide. This oxyhydride formation is very rare. It has been first reported in 2002 [78]. Hayward at al. observed that LaSrCoO4 reduction with CaH2 yielded LaSrCoO3H0.7. This was a spectacular finding at this time, which led to a perspective comment in Science [79]. Conceptually and chemically transition metal oxyhydrides are very different from Zintl phase hydrides presented in the previous parts of this thesis. It is commonly believed that strongly reducing H- and O2- should be incompatible in forming a common anion substructure. It was then very surprising when in 2012 Kobayashi et al. reported that the reaction of CaH2 with the archetypical perovskite BaTiO3 affords cubic BaTiO3- [21] xHx with large amounts (x < 0.6) of hydrogen incorporated . The oxyhydride BaTiO3-xHx is remarkable in many respects. Ti attains a mixed III/IV oxidation state. According to the stoichiometry, the Ti(III) concentration should be equal to the H content. The color of BaTiO3-xHx is dark blue to black. The compound is stable in air and water. Further, BaTiO3-xHx can be heated to approximately 400 °C, above which hydrogen is released. When present, oxygen is scavenged and BaTiO3 is re- tained. In inert gas atmospheres containing D2 a hydride exchange H/D occurs at hydrogen release temperatures. Also, oxynitrides BaTiO3-δNy may be prepared by heating BaTiO3-xHx under N2 flow at 400 – 600 °C [80]. These observations led to the conclusion that the hydride species in

60 BaTiO3-xHx is labile and that the material represents a versatile precursor toward more mixed anion compounds. During this thesis, a rather detailed investigation of the hydride reduction of BaTiO3 was undertaken.

5.1 Reduction of BaTiO3 with CaH2

The investigation started with attempting to reproduce the results reported by Kobayashi et al. These authors prepared BaTiO3-xHx by reacting BaTiO3 with a particle size of 170 nm with a large excess of CaH2 (3 M) at temperatures between 500 and 580 °C for 4-7 days in pyrex or fused silica tubes [21] [81]. Here reactions were performed at 600 °C in welded stainless steel ampoules. The average particle size of the BaTiO3 starting material was 500 nm, which is somewhat larger than the material used by Kobayashi et al. Figure 5.2 shows the XRPD patterns of products that were obtained with increasing CaH2 concentrations, from 0.1 M to 0.9 M, during 2 days experiments at 600 °C and Table 5.1 present the results from the refinement of XRPD patterns.

Figure 5.2: a) XRPD patterns of products obtained from hydride reduction of BaTiO3 at 600 ºC and during 2 days, using various concentrations of CaH2. b) Rietveld plots for the XRPD patterns of 0.6, 1.2 and 1.8 H showing the evolution of two-phase mixture with increasing .

61 The product obtained with 0.1 M CaH2 had a pale blue color and re- mained tetragonal whereas with higher concentrations cubic products were obtained. A higher degree of reduction with increasing CaH2 concentration was recognizable by a deepening of the color to dark blue, almost black, in agreement with the observation of Kobayashi et al. The unit cell volume of the reduced products is very similar to that of the starting material and slightly increasing with increasing CaH2 concentration. A closer inspection of the XRPD patterns of 0.6 M and 0.9 M reduced samples revealed that reflections were pronouncedly anisotropic which hinted to phase heterogeneity. These patterns were refined as a mixture of two cubic phases. The unit cell parameter of the minority phase was larger by about 0.014 Å and its weight fraction was about 11 %. Higher concentrations than 1.8 H led to a drastic broadening of reflections and a diminished crystallinity.

Table 5.1. Synthesis products from hydride reduction with CaH2 during 2 day experiments at 600 °C. conc. of Product/ lattice parame- Volume x from 3 CaH2 fraction (wt.%) ters (Å) (Å ) TG 0 tetragonal a = 3.9964(1), 64.379(2) 0 c = 4.0310(1)

0.1 M tetragonal a = 3.9971(1), 64.324(3) 0.03 c = 4.0260(1)

0.3 M cubic 4.0051(6) 64.247(2) 0.10

0.6 M cubic-I / 89(1) 4.0096(1) 64.461(2) 0.24 cubic-II / 11(1) 4.0227

0.9 M cubic-I 89(1) 4.0138(1) 64.662(2) 0.34 cubic-II / 11(1) 4.0288

TGA under flowing air will monitor the reaction

BaTiO3-xHx + 0.75x O2 → BaTiO3 + 0.5x H2O (5.1)

and thus, provide a convenient way to determine the H content of BaTiO3-xHx because the substantial weight increase associated with this reaction makes the determination of x fairly accurate. Figure 5.3 depicts the TGA traces for CaH2 reduced samples. All samples show initially a small weight loss (0.1-0.15 %). This is attributed to surface hydroxyl and

62 water coordinated to these hydroxyl groups. A small weight loss starting off at low temperatures and amounting to 0.15 % at 700 °C is also observed for the starting material. The subsequent weight increase for the reduced BaTiO3 samples should be due to oxidation, which for all samples is completed above 700 °C. The associated x values according to equation (1) are contained in Table 5.2. The value x = 0.34 for the sample reduced with the highest CaH2 concentration (1.8 M) is considerably smaller than the maximum value reported by Kobayashi et al. (x ≈ 0.6).

Figure 5.3: TGA traces for products obtained from hydride reduction of Ba- TiO3 at 600 ºC and during 2 days, using various concentrations of CaH2

Figure 5.4 shows the lattice parameter variation as a function of temperature from a XRPD experiment in which 0.6 M reduced sample was heated in air to 900 °C and subsequently cooled. In agreement with the TGA experiment, oxidation occurs between 500 and 600 °C. Above 600 °C the lattice parameters during heating and cooling coincide. Upon cooling the phase transition into tetragonal BaTiO3 occurs below 200 °C. The reported temperature of the tetragonal-to-cubic phase transition for BaTiO3 is near 120 °C (cubic BTO). The lattice parameter of the cubic high temperature form of BaTiO3, 4.0037 Å at 200 °C, is clearly smaller than that of reduced samples (by about 0.07 Å).

63 periment is clearly different. The XRPD pattern shows significantly broader reflections and also the presence of an impurity phase (Ti3O) which is indicative of the onset of decomposition of BaTiO3. The unit cell parameters are increased with respect to the products obtained after shorter times. Peculiar is the TGA trace, exhibiting significant weight loss (~0.4%) between room temperature and 350 ºC, after which a weight increase in excess of 3% occurs.

Table 5.2. Synthesis products from hydride reduction with 0.6 M CaH2 at 600 °C at varying reaction time. reaction Product/ fraction lattice param- volume x from time (wt.%) eters (Å) (Å3) TG 1-d cubic-I / 91(2) 4.0093(2) 64.446(4) 0.24 cubic-II / 9 (2) 4.0219(1) 2-d cubic-I / 87(2) 4.0079(1) 64.380(4) 0.24 cubic-II / 13 (2) 4.0205(1) 4-d cubic-I / 89(2) 4.0094(1) 64.451(4) 0.26 cubic-II / 11(2) 4.0221(1) 7-d cubic-I / 70(1) 4.0173(1) 64.834(2) 0.51 cubic-II / 28(1) 4.0275 (1)

Ti3O / 2(1)

To summarize, it was possible to reproduce the hydride reduction of [21] BaTiO3 with CaH2 as reported by Kobayashi et al. . At the same time it was noticed that products became heterogeneous and considerably less crystalline with increasing CaH2 concentration.

5.2 1H NMR investigations. It appears obvious that 1H NMR should be used for the detailed analysis of BaTiO3-xHx. In their initial paper on BaTiO3-xHx Kobayashi et al. showed a spectrum with a single sharp peak at 4.4 ppm [21]. They concluded that all H species are in the same chemical state and environment, which fit the expectation of the phase being a defect-free solid solution in which O2- is randomly replaced by H-. Figure 5.6 show the 1H NMR spectrum of the starting material before and after heating to 900 °C in air in a TG apparatus. There is a distinct sharp signal at 1 ppm and two more groups of protons with chemical shifts of 4.8 and 6.5 ppm.

65

1 Figure 5.6: H NMR spectrum of the BaTiO3 starting material before and after heating to 900 oC in air flow (during a TGA experiment).

All these protons are attributed to surface OH species. One can exclude the presence of hydroxyl in the structure of the starting material because of the great similarity of the 1H spectra before and after heating. Unlike surface OH, structural OH will not reform upon exposure to ambient atmosphere. The different chemical shifts can be associated with different surface sites. In order to obtain an estimate of the H content in the Ba- TiO3 starting material the strength of the proton signal of the sample was related to that of an adamantane (C10H16) sample with the same volume. The BaTiO3 sample had 0.94% proton signal compared to the adamantane sample. From the densities and molecular weights of adamantane 3 (1.08 g/cm , 136.23 g/mol) and BaTiO3 (6.02 g/cm3, 233.2 g/mol), a molar ratio H: BaTiO3 ≈ 0.046 can then be estimated. This ratio relates well to the TGA weight loss of 0.15%. If one assumes that this weight loss stems from the condensation of surface hydroxyl. one obtains a molar ratio H2O: BTO ≈ 0.02 (that is H: BTO = 0.04). Figure 5.7 shows the 1H NMR spectrum of the 0.1 M and 0.6 M reduced samples. The one of the 0.6 M reduced sample shows a broad resonance with a maximum at -18 ppm.

66

1 Figure 5.7: H NMR spectra of 0.2 and 1.2 H CaH2 reduced samples (600 ºC, 2 days experiments). The inset shows proton signal intensity with respect to the starting material.

This resonance is missing in the spectrum of the sample reduced with 0.1 M CaH2 which is still tetragonal. Consequently, this resonance is attributed to hydridic H on the O position in the perovskite structure. The signals at positive shifts are in the same ppm range as observed for the starting material and, thus, are attributed to surface OH. Interestingly, 1 the H signal strengths of 0.1 M reduced sample and the BaTiO3 starting material are very similar. Therefore, both materials should contain similar proton concentrations. Really surprising was to see that the 1H signal strength for the 0.6 M reduced sample is only about 4 times stronger than that of the 0.1 M reduced one. Assuming that the concentration of protic surface hydroxyl is comparable for 0.1 M and 0.6 M reduced samples, one can estimate the molar ratio between hydridic H and BaTiO3 in the latter sample to be approximately 0.12. There is a large discrepancy with the TGA experiment (Figure 5.3) which gave an x value of 0.24 when referring to BaTiO3-xHx. The discrepancy can only be resolved if one assumes a simultaneous presence of O vacancies.

67

The TGA weight increase for the reaction

BaTiO3-x + x/2 O2 → BaTiO3 (5.2)

will be only slightly different (by about 6.5 % higher) compared to reaction (1). Therefore, TGA under flowing air will hardly be able to dis- criminate BaTiO3-x and BaTiO3-xHx as products of the hydride reduction. The combined 1H NMR and TGA results suggest that hydridic H and O vacancies are present in roughly equal concentrations in 1.2 M CaH2 reduced samples. One could formulate the product as BaTiO3-xHy□(x-y) with x ≈ 0.25 and y ≈ 0.12. 1 Figure 5.8 shows the H NMR spectra of products from 0.6 M CaH2 reductions during different reaction times. Samples obtained after 1, 2, and 4 days have very similar spectra.

1 Figure 5.8: NMR spectra of products obtained from reduction with CaH2 at 600 °C during 1, 2, 4, and 7 d experiments

This is not surprising because XRPD and TGA investigations gave very similar results for those samples. The sample after 7-days is clearly different. The TGA weight increase is twice that of the other samples, shown in Table 5.2, and the cubic lattice parameters of the products are larger by about 0.06 Å . The 1H signal in the spectrum of the 7-days sample is more asymmetric compared to the other spectra and its maximum (at around -50 ppm) is moved to more negative shifts. Importantly, when analyzing the relative strengths and shift contributions to the 1H

68 NMR of the 7-days reduced sample, one obtains a very similar H- content (negative shift contribution) compared to 1, 2, and 4-days samples. The positive shift contribution is considerably higher. This is attributed to a strong surface modification during the long reduction time. Long reduction reaction times (and subsequent washing) may create additional surface OH sites, which will also explain the pronounced TGA weight loss behavior of this sample at low temperatures. The 7-days sample has more defects than incorporated H and its composition is estimated as BaTiO3-xHy□(x-y) with x ≈ 0.50 and y ≈ 0.12.

5.3 Other reducing agents

Apart from CaH2 reactions, hydride reduction of BaTiO3 was also performed with NaH, MgH2, and NaBH4 and NaAlH4. It was found that NaH act only weakly reducing. A conversion into a cubic product was not observed at the conditions applied (600 oC and up to 1.8 M NaH). This was attributed to the low decomposition temperature of NaH, mak- - ing instead of H gaseous H2 the reducing species. MgH2 and NaAlH4 reduction gave very similar products. These agents act more strongly reducing compared to CaH2. NaBH4 is peculiar. Lattice parameters of reduced products were virtually independent from the concentration of ap- 1 plied NaBH4. Further, H NMR showed that only a very small concentration of H was in incorporated in BaTiO3 whereas TGA showed com- paratively large weight increases. This indicated that NaBH4 reduction primarily led to the creation of O vacancies.

5.4 Conclusions

It appears that hydride reduction of BaTiO3 yields complex, heterogeneous, materials, due to the simultaneous presence of vacancies and H in the anion substructure. The overall concentration of hydridic H is actual- ly rather low. The formation of O-deficient phases BaTiO3-xHy□(x-y) is surprising and at variance with the results of Kobayashi et al. It is unclear why the hydride reduction experiments with BaTiO3 resulted in a different product as described by Kobayashi et al. Yet, highly O- deficient BaTiO3-xHy□(x-y) is an interesting material. Reduced BaTiO3-x, with applications in electroceramics, electrocatalysis and electronics, has been described and studied earlier [82] . Here is general agreement that the crystal symmetry changes from tetragonal to cubic for small values of x, x < 0.02, before transforming to the hexagonal 6H-perovskite for x > 0.02 [83] [84]. The hexagonal form can maintain a maximum x of about [85] 0.15 . Such reduced forms of BaTiO3 are typically synthesized at high

69 temperatures (T > 1000 C) using a flowing mixture of 5% H2/95% N2 [84] . Therefore, the existence of cubic phases BaTiO3-xHy□(x-y) with x values exceeding 0.5 is surprising. Clearly, the incorporation of H plays an important role in stabilizing these highly O- deficient variants of BaTiO3

70 6. Conclusions and future perspectives

The thesis dealt with two conceptually different classes of hydrogen containing solid state compounds, Zintl phase hydrides and transition metal oxyhydrides. The aim was to investigate the flexibility of hydrogen to adapt to various environments. Zintl phase hydrides were obtained by hydrogenating Zintl phases which are chemical compounds formed by an active metal component (i.e. alkali, alkaline earth, rare earth) and a more electronegative p-element metal/semimetal component. Zintl phases can respond in various ways to hydrogenation, by forming Zintl phase hydrides (usually distinguished between interstitial and polyanionic ones), oxidative decomposition, and full hydrogenation into complex metal hydride systems. The systems investigated here covered CaSi2, for which hydrogenation resulted in a changed crystal structure without incorporation of hydrogen, GdGa and Eu3Si4, for which hydrogenation produced mixed interstitial/polyanionic hydrides GdGaH2-x and Eu3Si4H2+x, respectively, and ASi (A = K, Rb) for which a full hydrogenation – - to silanides ASiH3 with discrete SiH3 moieties – was observed. It seems very difficult to predict the outcome of Zintl phase hydrogenations. Also conditions, especially hydrogenation temperature, play an important role. Another important issue with Zintl phase hydrogenations are intermediate products, or Zintl phase hydrides that are only stable in a pressurized hydrogen atmosphere. To detect such species, in situ investigations at synchrotron facilities are very helpful. Generally, the hydrogenation behavior of Zintl phases can only be conclusively studied by in situ experiments. In Zintl phase hydrides, hydrogen experiences a mixed metal environment. Structures and properties of Zintl phase hydrides are markedly different compared to the Zintl phase precursors. In this thesis drastic changes in the magnetic properties of rare earth Zintl phases were observed upon hydrogenation. GdGa and Eu3Si4 are both ferromagnets with rather high Curie temperatures. The hydrides GdGaH2-x and Eu3Si4H2+x were found to be antiferromagnetic. However, the hydrogenation induced magnetic property change is far from understood. Hydrogen incorporation could give a unique possibility to probe the electronic structure of rare earth Zintl phase systems and provide insight into the interplay between chemical bonding in the polyanion and magnetic properties. In particular, itinerant electrons may be localized as H- through the formation of interstitial hydrides. This should express in a change of the strength and possibly also the sign of the magnetic interac-

71 tion. In practice, these kinds of investigations are hampered by difficulties in determining actual hydrogen concentrations and the crystallographic location of H in Zintl phase hydrides. In transition metal oxyhydrides, O2- and H- ions forms commonly an anion substructure. Transition metal oxyhydrides can be obtained by hydride reduction reactions between active metal hydrides (e.g. LiH, NaH, CaH2) and transition metal oxides. However, more typical for hydride reductions is to produce reduced metal oxide systems via O vacancy formation. The investigated metal hydride reduction of BaTiO3 delivered complex, heterogeneous, materials BaTiO3-xHy□(x-y) with x up to 0.6, y in a range 0.05 – 0.2 and (x-y) > y, rather than homogeneous solid solutions BaTiO3Hx. The anion substructure contains O vacancies and H. The overall concentration of hydridic H is rather low. The existence of cubic phases BaTiO3-xHy□(x-y) with x values exceeding 0.5 is surprising. Clearly, the incorporation of H plays an important role in stabilizing these highly O-deficient variants of BaTiO3. Materials properties (electron and ion transport, catalytic properties) are not known. It will be interesting to investigate them in the future. Also it would be desirable to understand mechanisms behind the hydride reduction of transition metal oxides better. .

72 7. Populärvetenskaplig Sammanfattning

Denna doktorsavhandling presenterar undersökningar av väteinkorporering i Zintl-faser och övergångsmetalloxider. Vätehaltiga Zintl-faser kan användas som viktiga modellsystem vid bedrivandet av grundläggande studier hos väte-metallinteraktioner, samtidigt som väteinducerade förändringar i kemisk struktur och fysikaliska egenskaper bådar om spännande prospekt inom materialvetenskapen. Väteinkorporering i övergångsmetallhydrider leder till bildandet av ett oxyhydridsystem, där O och H tillsammans bildar en anjonisk substruktur. Vätespecien i övergångsmetalloxider kan vara mycket mobila, vilket gör dessa typer av material till intressanta prekursorer för andra bland-anjoniska system.

Zintl-faser består av en aktiv metall, M (alkalisk, alkalisk jordartsmetall eller sällsynt jordartsmetall), och en mer elektronegativ p-blocksmetall eller semimetallkomponent, E (Al, Ga, Si, Ge, etc.). När Zintl-faser reagerar med väte, bildar de antingen polyanjoniska hydrider eller interstitiella hydrider, totala hydrogeneringar för att bilda komplexa hydrider, eller oxidativ dekompositionering till mer E-haltiga Zintl-faser. Zintl-faserna som undersöktes här bestod av CaSi2, Eu3Si4, ASi (A = K, Rb) och GdGa-system som hydrogenerades vid olika temperaturer, H2-tryck och uppehållstider. För CaSi2 så förekom en regelbunden fasövergång från den konventionella 6R till den sällsynta 3R och ingen hydridbildning observerades. I kontrast till detta så visade GdGa och Eu3Si4 en stor benägenhet för hydridupptag. Redan vid temperaturer under 100 ιC så kunde hydridbildning av GdGa2-x och Eu3Si4H2+x observeras. De magnetiska egenskaperna av hydriderna (antiferromagnetiska) åtskiljer sig drastiskt från Zintl-fasprekursorerna (ferromagnetiska). Hydrogenering av ASi vid temperaturer kring 100 ιC ledde till bildandet a ASiH3, som innehåller diskreta jonkomplexenheter - SiH3 . Den komplicerade β-α ordning-oordning fasövergången i ASiH3 utvärderades med NPD, NMR och med värmekapacitetsmätningar.

En systematisk studie på hydridreduktionen av BaTiO3 som leder till bildandet av perovskitoxyhydriden BaTiO3-xHx utfördes. Ett brett utbud av reduktionsmedel som inkluderar NaH, MgH2, CaH2, LiAlH4 och NaBH4 tillämpades och temperaturer och uppehållvillkor för hydridreduktionerna undersöktes. Prover karaktäriserades genom pulverröntgendiffraktion, termogravimetrisk analys och 1H NMR. Koncentrationen av H som går att

73 inkorporera i BaTiO3-xHx upptäcktes vara väldigt låg, vilket motsäger det som tidigare har rapporterats. Hydridreduktionen leder istället till en hög koncentration av O-vakanser i den reducerade BaTiO3. De kraftigt O-fattiga, oordnade, faserna – BaTiO3-xHy(x-y), med x upp till 0,6 och y inom intervallet 0,05 – 0,2 och (x-y)> y – är kubiska och kan vara intressanta material med avseende på elektron- & jontransport och även för katalys.

74 8. Acknowledgements

It has been my greatest pleasure to work with Ulrich Häussermann who was a great inspiration to me through out. Ulrich, I have learned a lot from you and both scientifically and personally and I am very grateful for all the interesting and fruitful moments I spend with you. The joy and enthusiasm which you have for the research was contagious and motiva- tional to me even in the tough times of my PhD. Your encouragement and guidance was the driving force in refining the researcher in me. You supported me not only academically but also emotionally in the rough roads of finishing this thesis. I could not have imagined having a better advisor and mentor for my PhD studies.

I would like to extend special appreciation and thanks to my co- supervisor Jekabs, for his valuable advice and all the insightful discus- sions and suggestions. You have been really supportive from the day one when I started working at MMK. Your willingness to help and be there always was something which motivated me.

I was quite a beginner in working with metal hydrides, when I came to MMK. Thank you Verina for giving me insights about the interesting world of hydrides. You were so kind and helpful and I should mention your explicit skill of explaining things. I would like to appreciate all the efforts you put in. I am also grateful to Daryn who helped me with all the theoretical calculations.

Dozens of people have helped me during my work at SU. I am grateful to Aleksader, Diana, Andrew Pell and Mattias Edén for all their help with NMR. Thank you so much Alexandra and Cheuk-Wai Tai for helping me with TEM.

I greatly acknowledge the support received through the collaboration work with Matt’s group at Chalmers University. Thank you, Karin, for all those neutron diffraction evaluations. I also extend my gratitude to Robert, Jonas and Mikael from Uppsala University for all the magnetic measurements studies.

75 I would like to express my gratitude to all the MMK professors for their great support. Thank you, Xiaodong Zou, Osamu Terasaki, Arnold Ma- liniak, Sven Hövmöller, Peter Oleynikov, Matts Johansson, Kjell Jans- son, Niklas Hedin, Junliang Sun, Feifei Gao, Zoltán Bacsik, Lynn McCusker, Mattias Edén, German Salazar Alvarez, James Zhijian Shen, Cheuk- Wai Tai, Aji Mathew and Lars Eriksson. I am grateful to Sven Hövmöller and Xiaodong Zou for reading through my thesis and provid- ing appropriate suggestions for improvement.

A very special thanks to Ann Britt who is so kind and supportive. You are really a wonderful person, always with a beautiful smile.

I was lucky enough to be a part of a great group. The group has been a source of friendship, inspiration well as collaboration. I had a great time with you Sumit, Daniel and Kristina. You were the best people I could work with. Sumit and Kristina a big thanks for all the quality time we spent. Thank you, for all your help and support. Daniel, you were my Stockholm encyclopaedia and the first person I ran into whenever I had some queries or need some help. Thank you for the beautiful mid- summer days’ celebrations, the home-made cakes, games, movies and all the wonderful moments which we all had together. Thank you buddies. Johanna Nylén, thank you for all your kind help. Emil, Michae, Marie and Isthvan it’s been a pleasure working with you guys at lab.

All the members of the Department of MMK have contributed to both my professional and personal time at Stockholm University. I would like to thank all my friends and colleagues at MMK. Thank you so much Magda, you are a really good friend who never failed to surprise mr with your delicious apple pies and cupcakes. Thanking you all my collegues- Abtin, Alexandra, Arnaud, Arto, Baltzar, Bin, Bojan, Changjiu, Christi- na, Dariusz, Diana, Dickson, Duan, Ekaterina, Elina, Fabian, Fei, Gholamhasan, Hani, Haoquan, Henrik, Imran, Inna, Iwan, Junzhong, Kamran, Ken, Korneliya, Leifeng, Neda, Ning, Michael, Panagiotis, Peng, Przemek, Santhosh, Taimin, Tom, Valentina, Yang, Yanhang, Yifeng, Yuan, Yulia, Yunchen, Yunxiang. Special thanks to Renny for all your support and help. I would like to extend my gratitude to all the administrative staff for the for their cooperation: Ann Britt Rönnell, Tatiana Bulavina, Camilla Berg, Hanna Nyholm, Helmi Frejman, Paula Jokela, Anette Lindberg, Ann Loftsjö, Karin Sandberg, Anna- Karin Olsson, Pia Raninen and Daniel Emanuelsson. Thanks to Rolf Eriksson for managing all sorts of computer issues. Big thanks to Pelle for helping with the reactor design and for extending great support along with Hans Erik.

76 A very special thanks to BaPaas and their Beevis for all the fun and beautiful moments we had. Stockholm would not have been this enjoya- ble without you guys. “Lida” and “Pistah” will always be cherished. I thank all the people of MiSt who made Stockholm a second home. Very special thanks to Utsa, Louisa and Andy for the wonderful moments we spend together.

I have an amazing family, which is always the source of my energy. Thank you so much Amma and Achan for your unconditional love and support. I would not have made it this far without you. I would also like to thank my sister who was encouraging and supportive in all my pur- suits. Last but certainly not the least, I would like to acknowledge my life partner, Deepthi for all her love, patience, support and unwavering belief in me. She has patiently endured many long hours alone when I worked on my dissertation. Thank you so much for believing in me and instilling confidence in me.

77

78 9. Bibliography

[1] J. Rigden, Hydrogen: The Essential Element, Harward University Press:Cambridge,MA, 2003. [2] M. W. Cronyn, “The Proper Place for Hydrogen in the Periodic Table,” Journal of Chemical Education, vol. 80, no. 8, p. 947, 2003. [3] F. A. Cotton and G. Wilkinson, Advance Inorganic Chemistry, A comprehensive Text, John Wiley & Sons, Inc., 1980. [4] P. Atkins, Shriver and Atkins' Inorganic Chemistry, Oxford: Oxford University Press, 2010. [5] M. B. Ley, L. H. Jepsen, Y.-S. Lee, Y. W. Cho, J. M. Bellosta von Colbe, M. Dornheim, M. Rokni, J. O. Jensen, M. Sloth, Y. Filinchuk, J. E. Jørgensen, F. Besenbacher and T. R. Jensen, “Complex hydrides for hydrogen storage – new perspectives,” Materials Today, vol. 17, no. 3, pp. 122-128, 2014. [6] J. N. Huiberts, R. Griessen, J. H. Rector, R. J. Wijngaarden, J. P. Dekker, D. G. de Groot and N. J. Koeman, “Yttrium and lanthanum hydride films with switchable optical properties,” Nature, vol. 380, no. 6571, pp. 231-234, 1996. [7] J. B. Condon and T. Schober, “Proton conductors and metal hydrides,” Solid State Ionics, vol. 77, pp. 299-304, 1995. [8] E. Aitken, Intermetallic Compounds ed J H Wetbrook, Wiley, 1967. [9] M. B. Ley, L. H. Jepsen, Y.-S. Lee, Y. W. Cho, J. M. Bellosta von Colbe, M. Dornheim, M. Rokni, J. O. Jensen, M. Sloth, Y. Filinchuk, J. E. Jørgensen, F. Besenbacher and T. R. Jensen, “Complex hydrides for hydrogen storage – new perspectives,” Materials Today, vol. 17, no. 3, pp. 122-128, 2014. [10] H. Grove, L. H. Rude, T. R. Jensen, M. Corno, P. Ugliengo, M. Baricco, M. H. Sørby and B. C. Hauback, “Halide substitution in Ca(BH4)2,” RSC Advances, vol. 4, no. 9, pp. 4736-4742, 2014. [11] S. Takagi, Y. Iijima, T. Sato, H. Saitoh, K. Ikeda, T. Otomo, K. Miwa, T. Ikeshoji, K. Aoki and S. Orimo, “True Boundary for the Formation of Homoleptic Transition-Metal Hydride Complexes,” Angewandte Chemie International Edition, vol. 54, no. 19, pp. 5650-5653, 2015.

79 [12] E. Callini, Z. Ö. K. Atakli, B. C. Hauback, S. Orimo, C. Jensen, M. Dornheim, D. Grant, Y. W. Cho, P. Chen, B. Hjörvarsson, P. d. Jongh, C. Weidenthaler, M. Baricco, M. Paskevicius, T. R. Jensen, M. E. Bowden, T. S. Autrey and A. Züttel, “Complex and liquid hydrides for energy storage,” Applied Physics A, vol. 122, no. 4, p. 353, 2016. [13] K. Yvon and G. Renaudin, Hydrides: Solid State Transition Metal Complexes, John Wiley & Sons, Ltd, 2011. [14] Q. Xie and R. Nesper, “Structural and Electronic Characterization of Eu2LiSi3, Eu2LiGe3 and EuxSr2−xLiGe3 Mixed Crystals,” Zeitschrift für anorganische und allgemeine Chemie, vol. 632, no. 10-11, pp. 1743- 1751, 2006. [15] S. M. Kauzlarich, Structure, and Bonding of Zintl Phases and Ions: Selected Topics and Recent Advances, vol. 632, Wiley, 1996, pp. 1743- 1751. [16] H. Schäfer, “On the Problem of Polar Intermetallic Compounds: The Stimulation of E. Zintl's Work for the Modern Chemistry of Intermetallics,” Annual Review of Materials Science, vol. 15, no. 1, pp. 1-42, 1985. [17] R. Nesper, “Structure and chemical bonding in zintl-phases containing lithium,” Progress in Solid State Chemistry, vol. 20, no. 1, pp. 1-45, 1990. [18] H. Schäfer, B. Eisenmann and W. Müller, “Zintl Phases: Transitions between Metallic and Ionic Bonding,” Angewandte Chemie International Edition in English, vol. 12, no. 9, pp. 694-712, 1973. [19] U. Häussermann, V. F. Kranak and K. Puhakainen, “Hydrogenous Zintl Phases: Interstitial Versus Polyanionic Hydrides,” in Zintl Phases: Principles and Recent Developments, Springer Berlin Heidelberg, 2011, pp. 143-161. [20] W. S. Tang, J.-N. Chotard, P. Raybaud and R. Janot, “Enthalpy– Entropy Compensation Effect in Hydrogen Storage Materials: Striking Example of Alkali Silanides MSiH3 (M = K, Rb, Cs),” The Journal of Physical Chemistry C, vol. 118, no. 7, pp. 3409-3419, 2014. [21] Y. Kobayashi, O. J. Hernandez, T. Sakaguchi, T. Yajima, T. Roisnel, Y. Tsujimoto, M. Morita, Y. Noda, Y. Mogami, A. Kitada, M. Ohkura, S. Hosokawa, Z. Li, K. Hayashi, Y. Kusano, J. e. Kim, N. Tsuji, A. Fujiwara, Y. Matsushita, K. Yoshimura, K. Takegoshi, M. Inoue, M. Takano and H. Kageyama, “An oxyhydride of BaTiO3 exhibiting hydride exchange and electronic conductivity,” Nature Materials, vol. 11, no. 6, pp. 507-511, 2012. [22] A. E. Dwight, J. W. Downey and R. A. Conner, “Equiatomic compounds of Y and the lanthanide elements with Ga,” Acta

80 Crystallographica, vol. 23, no. 5, pp. 860-862, 1967. [23] T. R. Jensen, T. K. Nielsen, Y. Filinchuk, J.-E. Jørgensen, Y. Cerenius, E. M. Gray and C. J. Webb, “Versatile in situ powder X-ray diffraction cells for solid–gas investigations,” Journal of Applied Crystallography, vol. 43, no. 6, pp. 1456-1463, 2010. [24] V. Höglin, J. Ångström, M. S. Andersson, O. Balmes, P. Nordblad and M. Sahlberg, “Sample cell for in-field X-ray diffraction experiments,” Results in Physics, vol. 5, pp. 53-54, 2015. [25] H. Rietveld, “A profile refinement method for nuclear and magnetic structures,” Journal of Applied Crystallography, vol. 2, no. 2, pp. 65- 71, 1969. [26] A. Le Bail, H. Duroy and J. L. Fourquet, “Ab-initio structure determination of LiSbWO6 by X-ray powder diffraction,” Materials Research Bulletin, vol. 23, no. 3, pp. 447-452, 1988. [27] J. Rodriguez-Carvajal, “Fullprof: A Program for Rietveld Refinement and Pattern Matching Analysis,” Abstract of the Satellite Meeting on Powder Diffraction of the XV Congress of the IUCr, Toulouse, France, p. 127, 16-19 July 1990. [28] M. Järvinen, “Application of symmetrized harmonics expansion to correction of the preferred orientation effect,” Journal of Applied Crystallography, vol. 26, no. 4, pp. 525-531, 1993. [29] S. M. Castillo, Z. Tang, A. P. Litvinchuk and A. M. Guloy, “Lattice Dynamics of the Rhombohedral Polymorphs of CaSi2,” Inorganic Chemistry, vol. 55, no. 20, pp. 10203-10207, 2016. [30] . T. L. Hwang, . P. C. M. van Zijl and M. Garwood, “Fast Broadband Inversion by Adiabatic Pulses,” Journal of Magnetic Resonance, vol. 133, pp. 200-203, 1998. [31] G. Kervern, G. Pintacuda and L. Emsley, “Fast Adiabatic Pulses for Solid-State NMR of Paramagnetic Systems,” Chemical Physics Letters, vol. 435, p. 157 – 162, 2007. [32] B. Willis and C. Carlile, Experimental Neutron Scattering, Oxford University Press, 2009. [33] P. Manfrinetti, M. L. Fornasini and A. Palenzona, “The phase diagram of the Ca–Si system,” Intermetallics, vol. 8, no. 3, pp. 223-228, 2000. [34] T. Massalski, H. Okamoto, P. Subramanian and L. Kacprzak, “Binary Alloy Phase Diagrams, II Ed.,” ASM International, 1990, pp. 951-954. [35] A. Wosylus, Y. Prots, U. Burkhardt, W. Schnelle and U. Schwarz, “High-pressure synthesis of the electron-excess compound CaSi6,” Science and Technology of Advanced Materials, vol. 8, no. 5, pp. 383- 388, 2007.

81 [36] S. Yamanaka, E. Enishi, H. Fukuoka and M. Yasukawa, “High-Pressure Synthesis of a New Silicon Clathrate Superconductor, Ba8Si46,” Inorganic Chemistry, vol. 39, no. 1, pp. 56-58, 2000. [37] O. O. Kurakevych, T. A. Strobel, D. Y. Kim, T. Muramatsu and V. V. Struzhkin, “Na-Si Clathrates Are High-Pressure Phases: A Melt-Based Route to Control Stoichiometry and Properties,” Crystal Growth & Design, vol. 13, no. 1, pp. 303-307, 2013. [38] W. S. Tang, J.-N. Chotard, P. Raybaud and R. Janot, “Hydrogenation properties of KSi and NaSi Zintl phases,” Physical Chemistry Chemical Physics, vol. 14, no. 38, pp. 13319-13324, 2012.

[39] G. Vogg, M. S. Brandt, M. Stutzmann and M. Albrecht, “From CaSi2 to siloxene: epitaxial silicide and sheet polymer films on silicon,” Journal of Crystal Growth, vol. 203, no. 4, pp. 570-581, 1999. [40] G. Vogg, N. Zamanzadeh-Hanebuth, M. S. Brandt, M. Stutzmann and M. Albrecht, “Preparation and Characterization of Epitaxial CaSi2 and Siloxene Layers on Silicon,” Monatshefte für Chemie / Chemical Monthly, vol. 130, no. 1, pp. 79-87, 1999. [41] G. Vogg, M. Brandt and M. Stutzmann, “Elastic Properties of the Layered Zintl-Phase CaSi2,” physica status solidi (a), vol. 185, no. 2, pp. 213-219, 2001. [42] S. Fahy and D. R. Hamann, “Electronic and structural properties of CaSi2,” Physical Review B, vol. 41, no. 11, pp. 7587-7592, 1990. [43] P. H. Tobash and S. Bobev, “Synthesis, structure and electronic structure of a new polymorph of CaGe2,” Journal of Solid State Chemistry, vol. 180, no. 5, pp. 1575-1581, 2007.

[44] J. Evers, “Transformation of three-connected silicon nets in CaSi2,” Journal of Solid State Chemistry, vol. 28, no. 3, pp. 369-377, 1979. [45] O. Schob and E. Parthé, “AB compounds with Sc, Y and rare earth metals. I. Scandium and yttrium compounds with CrB and CsCl structure,” Acta Crystallographica, vol. 19, no. 2, pp. 214-224, 1965. [46] A. E. Dwight, J. W. Downey and R. A. Conner, “Equiatomic compounds of Y and the lanthanide elements with Ga,” Acta Crystallographica, vol. 23, no. 5, pp. 860-862, 1967. [47] P. Villars and L. D. Calvert, Pearson's Handbook of Crystallographic Data for Intermetallic Compounds, ASM International: Materials Park, OH, 1991. [48] J. Y. Zhang, J. Luo, J. B. Li, J. K. Liang, Y. C. Wang, L. N. Ji, Y. H. Liu and G. H. Rao, “Magnetic properties and magnetocaloric effect of GdGa compound,” Journal of Alloys and Compounds, vol. 469, no. 1-2, pp. 15-19, 2009.

82 [49] J. Chen, B. Shen, Q. Dong, F. Hu and J. Sun, “Large reversible magnetocaloric effect caused by two successive magnetic transitions in ErGa compound,” Applied Physics Letters, vol. 95, no. 13, p. 132504, 2009. [50] J. Chen, B. G. Shen, Q. Y. Dong and J. R. Sun, “Giant magnetocaloric effect in HoGa compound over a large temperature span,” Solid State Communications, vol. 150, no. 3-4, pp. 157-159, 2010. [51] X. Zheng, J. Chen, J. Shen, H. Zhang, Z. Xu and W. Gao, “Large refrigerant capacity of RGa (R = Tb and Dy) compounds,” Journal of Applied Physics, vol. 111, no. 7, p. 07A917, 2012. [52] T. Gao, K. Nishimura, T. Matsumoto, T. Namiki and Y. Isikawa, “Magnetic properties and magnetocaloric effect in TmGa and HoGa single crystals,” Solid State Communications, vol. 158, pp. 1-4, 2013. [53] Z.-J. Mo, J. Shen, L.-Q. Yan, C.-C. Tang, J. Lin and J.-F. Wu, “Low field induced giant magnetocaloric effect in TmGa compound,” Applied Physics Letters, vol. 103, no. 5, p. 052409, 2013. [54] X. Zheng, J. Chen, L. Wang, R. Wu, F. Hu and J. Sun, “Magnetic properties and magnetocaloric effects of GdxEr1−xGa (0 ≤ x ≤ 1) compounds,” Journal of Applied Physics, vol. 115, no. 17, p. 17A905, 2014. [55] B. Barbara, N. Nyugen and E. Siaud, C.R. Acad. Sci. B., vol. 274, p. 1053–1056, 1972. [56] H. Fujii, N. Shohata, T. Okamot and E. Tatsumoto, “Magnetic Properties of Rare Earth Gallium Compounds RGa,” Journal of the Physical Society of Japan, vol. 31, no. 5, p. 1592, 1971. [57] N. Shohata, “Magnetic Properties of Rare Earth Gallium Intermetallic Compounds,” Journal of the Physical Society of Japan, vol. 42, no. 6, pp. 1873-1880, 1977. [58] M. A. Ruderman and C. Kittel, “Indirect Exchange Coupling of Nuclear Magnetic Moments by Conduction Electrons,” Physical Review, vol. 96, no. 1, pp. 99-102, 1954. [59] T. Kasuya, “A Theory of Metallic Ferro- and Antiferromagnetism on Zener's Model,” Progress of Theoretical Physics, vol. 16, no. 1, pp. 45- 57, 1956. [60] K. Yosida, “Magnetic Properties of Cu-Mn Alloys,” Physical Review, vol. 106, no. 5, pp. 893-898, 1957. [61] S. Sevov, “Zintl Phases,” in J. H. Westbrook, R. L. Fleischer (eds.). Intermetallic Compounds. Principles and Practice., John Wiley & Sons,Ltd, United States, 2002, pp. 113-132. [62] U. Häussermann, “Coexistence of hydrogen and polyanions in

83 multinary main group element hydrides,” Zeitschrift für Kristallographie International journal for structural, physical, and chemical aspects of crystalline materials, vol. 223, no. 10, p. 628–635, 2009. [63] J. Ångström, R. Johansson, T. Sarkar, M. H. Sørby, C. Zlotea, M. S. Andersson, P. Nordblad, R. H. Scheicher, U. Häussermann and M. Sahlberg, “Hydrogenation-Induced Structure and Property Changes in - the Rare-Earth Metal Gallide NdGa: Evolution of a [GaH]2 Polyanion Containing Peierls-like Ga–H Chains,” Inorganic Chemistry, vol. 55, no. 1, pp. 345-352, 2016. [64] N. N. Delyagin, V. I. Krylov and I. N. Rozantsev, “The magnetic spin- reorientation transitions in the RGa (R=rare earth) intermetallic compounds studied by measurements of the hyperfine interactions of the 119Sn probe atoms,” Journal of Magnetism and Magnetic Materials, vol. 308, no. 1, pp. 74-79, 2007. [65] X. B. Liu and Z. Altounian, “Exchange interaction in GdGa from first- principles,” Physica B: Condensed Matter, vol. 406, no. 3, pp. 710-714, 2011. [66] A. Leithe-Jasper and K. Hiebl, “Magnetic Properties of the Pseudobinary Solid Solution GdGa1-xGex (0 ≤ x ≤ 1),” physica status solidi (a), vol. 155, no. 1, pp. 223-231, 1996. [67] R. A. Susilo, J. M. Cadogan, D. H. Ryan, N. R. Lee-Hone, R. Cobas and S. Muñoz-Pérez, “Spin-reorientation in GdGa,” Hyperfine Interactions, vol. 223, no. 1-3, pp. 257-266, 2014. [68] F. Weitzer, Y. Prots, W. Schnelle, K. Hiebl and Y. Grin, “The novel silicide Eu3Si4: structure, chemical bonding, magnetic behavior and electrical resistivity,” Journal of Solid State Chemistry, vol. 117, no. 6, pp. 2115-2121, 2004. [69] J.-N. Chotard, W. S. Tang, P. Raybaud and R. Janot, “Chotard, J.-N.; Tang, W. S.; Raybaud, P.; Janot, R. Potassium,” Chemistry – A European Journal, vol. 17, no. 44, pp. 12302-12309, 2011. [70] R. Janot, W. S. Tang, D. Clémençon and J.-N. Chotard, “Catalyzed KSiH3 as a reversible hydrogen storage material,” Journal of Materials Chemistry A, vol. 4, no. 48, pp. 19045-19052, 2016. [71] C. Österberg, H. Fahlquist, U. Häussermann, C. M. Brown, T. J. Udovic – and M. Karlsson, “Dynamics of Pyramidal SiH3 Ions in ASiH3 (A = K and Rb) Investigated with Quasielastic Neutron Scattering,” The Journal of Physical Chemistry C, vol. 120, no. 12, pp. 6369-6376, 2016. [72] W. S. Tang, M. Dimitrievska, J.-N. Chotard, W. Zhou, R. Janot, A. V. Skripov and T. J. Udovic, “Structural and Dynamical Trends in Alkali- Metal Silanides Characterized by Neutron-Scattering Methods,” The

84 Journal of Physical Chemistry C, vol. 120, no. 38, pp. 21218-21227, 2016. [73] T. Yamamoto and H. Kageyama, “Hydride Reductions of Transition Metal Oxides,” Chemistry Letters, vol. 42, no. 9, pp. 946-953, 2013. [74] M. A. Hayward, “Topochemical reactions of layered transition-metal oxides,” Semiconductor Science and Technology, vol. 29, no. 6, p. 064010, 2014. [75] B. C. Hernden, J. A. Lussier and M. Bieringer, “Topotactic Solid-State Metal Hydride Reductions of Sr2MnO4,” Inorganic Chemistry, vol. 54, no. 9, pp. 4249-4256, 2015. [76] Y. Kobayashi, Z. Li, K. Hirai, C. Tassel, F. Loyer, N. Ichikawa, N. Abe, T. Yamamoto, Y. Shimakawa, K. Yoshimura, M. Takano, O. J. Hernandez and H. Kageyama, “Gas phase contributions to topochemical hydride reduction reactions,” Journal of Solid State Chemistry, vol. 2017, pp. 190-193, 2013. [77] M. A. Hayward, M. A. Green, M. J. Rosseinsky and J. Sloan, “Sodium Hydride as a Powerful Reducing Agent for Topotactic Oxide Deintercalation: Synthesis and Characterization of the Nickel(I) Oxide LaNiO2,” Journal of the American Chemical Society, vol. 121, no. 38, pp. 8843-8854, 1999. [78] M. A. Hayward, E. J. Cussen, J. B. Claridge, M. Bieringer, M. J. Rosseinsky, C. J. Kiely, S. J. Blundell, I. M. Marshall and F. L. Pratt, “The Hydride Anion in an Extended Transition Metal Oxide Array: LaSrCoO3H0.7,” Science, vol. 295, no. 5561, pp. 1882-1884, 2002. [79] K. Poeppelmeier, “A Mixed Oxide-Hydride Perovskite,” Science, vol. 295, no. 5561, pp. 1849-1849, 2002. [80] N. Masuda, Y. Kobayashi, O. Hernandez, T. Bataille, S. Paofai, H. Suzuki, C. Ritter, N. Ichijo, Y. Noda, K. Takegoshi, C. Tassel, T. Yamamoto and H. Kageyama, “Hydride in BaTiO2.5H0.5: A Labile Ligand in Solid State Chemistry,” Journal of the American Chemical Society, vol. 137, no. 48, pp. 15315-15321, 2015. [81] T. Sakaguchi, Y. Kobayashi, T. Yajima, M. Ohkura, C. Tassel, F. Takeiri, S. Mitsuoka, H. Ohkubo, T. Yamamoto, J. e. Kim, N. Tsuji, A. Fujihara, Y. Matsushita, J. Hester, M. Avdeev, K. Ohoyama and H. Kageyama, “Oxyhydrides of (Ca,Sr,Ba)TiO3 Perovskite Solid Solutions,” Inorganic Chemistry, vol. 51, no. 21, pp. 11371-11376, 2012.

[82] H.-I. Yoo, C.-R. Song and D.-K. Lee, “BaTiO3−δ: Defect Structure, Electrical Conductivity, Chemical Diffusivity, Thermoelectric Power, and Oxygen Nonstoichiometry,” Journal of Electroceramics, vol. 8, no. 1, pp. 5-36, 2002.

85 [83] M. Schrader, D. Mienert, T.-S. Oh, H.-I. Yoo and K. D. Becker, “An optical, EPR and electrical conductivity study of blue barium titanate, BaTiO3−δ,” Solid State Sciences, vol. 10, no. 6, pp. 768-775, 2008. [84] T. Kolodiazhnyi, “Insulator-metal transition and anomalous sign reversal of the dominant charge carriers in perovskite BaTiO3−δ,” Physical Review B, vol. 78, no. 4, p. 045107, 2008. [85] D. C. Sinclair, J. M. S. Skakle, F. D. Morrison, R. I. Smith and T. P. Beales, “Structure and electrical properties of oxygen-deficient hexagonal BaTiO3,” Journal of Materials Chemistry, vol. 9, no. 6, pp. 1327-1331, 1999.

86

Web Analytics