Thesis Reference

Thesis

Experimental and theoretical studies of boron and hydrogen containing compounds in relation to potential hydrogen storage and ionic conduction applications

SHARMA, Manish

Abstract

This thesis deals with the fundamental studies of some materials containing boron-hydrogen bonds which can potentially be used either as the hydrogen storage materials (M(BH4)2, M=Alkaline earth metal), as the solid electrolytes for batteries (Na2B12H12) or as reducing agents for CO2 (Mg(BH4)2). First part of thesis deals with borohydrides (BH4-). Synthesis and characterization of halide-free Sr(BH4)2, Ba(BH4)2 and Eu(BH4)2 is reported. Crystallographic study of these compounds helped in identifying several new phases and a new species metal borohydride hydride (M2(BH4)H3). In depth study of B-H bond breaking is reported via isotope exchange reaction in Ca(BH4)2.A practical example of borohydride as reducing agent is reported by showing the reduction of CO2 with gamma-Mg(BH4)2. The second part of the thesis focuses on closoboranes derived from the B12H122- ion. Compounds of this family have recently attracted great interest as solid ionic conductors for Li and Na ions.Results of DFT calculations on isolated B12H122- anions and halogen (F, Cl or Br) substituted anions were analysed in detail. Synthesis of Na2B12(SCN)H11 is [...]

Reference

SHARMA, Manish. Experimental and theoretical studies of boron and hydrogen containing compounds in relation to potential hydrogen storage and ionic conduction applications. Thèse de doctorat : Univ. Genève, 2017, no. Sc. 5101

DOI : 10.13097/archive-ouverte/unige:96376 URN : urn:nbn:ch:unige-963769

Available at: http://archive-ouverte.unige.ch/unige:96376

Disclaimer: layout of this document may differ from the published version.

1 / 1

UNIVERSITÉ DE GENÈVE FACULTÉ DES SCIENCES

Section de Chimie et Biochimie Professeur Hans Hagemann

Département de Chimie Physique

Experimental and Theoretical Studies of Boron and Hydrogen Containing Compounds In Relation to Potential Hydrogen Storage and Ionic Conduction Applications

THÈSE

présentée à la Faculté des sciences de l’Université de Genève

pour obtenir le grade de Docteur ès Sciences, mention Chimie

par

Manish Sharma d’Inde

Thèse N° 5101

GENÈVE

Atelier d’impression ReproMail

2017

Acknowledgements

I would like to express my sincere gratitude towards my supervisor, Prof. Hans Hagemann, for giving me the opportunity to study at the University of Geneva and for his explanations and scientific discussions. He created a very liberal work environment and encouraged me to think independently. He gave full freedom from the choice of projects and collaborations. For me Hans is a perfect example to learn the work-life balance. He emphasized on having a good social life outside office and encouraged to play sports, visit museums, music festivals and to travel.

I would like to thank Prof. H. W. Li and Dr. A. Remhof for accepting and evaluating my thesis.

I am grateful to Dr. L. M. L. Daku for helping me to conduct computational calculations discussed in the thesis and for being the part of the thesis review committee.

Thanks to Prof. A. Hauser for his support and valuable suggestions throughout my PhD. I admire him for his art of explaining complex concepts in a lucid way.

I would like to thank all the collaborators, in particular Prof. R. Černý, E. Didelot, Dr. D. Jeannerat, Dr. M. Pupier, Prof. T. Burgi, Dr. M. Chekini, Prof. Y. Filinchuk, F. Morelle, Prof. T. Jensen, Dr. B. Richter, Prof. H. W. Li, Dr. L. He, Dr. A. Remhof and Dr. E. Roedern for synergic work.

I am thankful to Ma’am C. Ludy and I. Garin for helping me with bureaucratic formalities. N. Amstutz and P. Barman for providing technical support and to Dominique Lovy for IT support.

I would like to acknowledge all the present and former members of Hauser-Hagemann group and my friends whom I met during my PhD: Teresa, Pablo, Romain, Angelina, Daniel, Andrea, Pradip, Jacob, Enza, Yolanda, Antoine, Quinchao, Jiji, Alexandra, Jan, Elia, Jakob, Igor, Ani, Rania, Mahshid, Martin and Roberto. They have been my closest friends in this period, sharing with me good and bad moments.

Finally, I would like to thank my wife and my parents for having supported me during these years.

Résumé

Ce travail de thèse porte sur les études fondamentales de certains matériaux contenant des liaisons bore-hydrogène qui peuvent potentiellement être utilisé comme des matériaux pour stockage de l'hydrogène, comme électrolyte solide dans les batteries, ou comme agents réducteurs pour le CO2.

La première partie de cette thèse sera consacrée aux études expérimentales des borohydrures métalliques de type M(BH4)n.

Le premier chapitre expérimental aborde la synthèse et la caractérisation des composés dépourvus d’halogénures Sr(BH4)2, Ba(BH4)2 et Eu(BH4)2. La synthèse de ces composés a été effectuée à partir de l'hydrure métallique et de Et3N-BH3. Lorsqu’ils sont chauffés à des températures élevées, ces composés sont plus stables que ceux rapportés dans la littérature obtenus par différentes voies de synthèse et impliquant des impuretés d'halogène. L'étude cristallographique de ces composés a permis d'identifier plusieurs nouvelles phases. Au cours des réactions de décomposition thermique, la formation d'un composé métal hydrure-borohydrure

M2H3(BH4), avec M = Sr et Eu, a été observée pour la première fois. Le produit Eu(BH4)2 s'est révélé être un excellent luminophore émettant dans le bleu à 463 nm. De plus, plusieurs méthodes de synthèse infructueuses ou qui ont abouti à des produits impurs sont également discutées.

Dans le chapitre subséquent, la réaction d'échange hydrogène-deutérium ainsi que la réaction inverse sont étudiées pour le composé Ca(BH4)2 en fonction de la température et de la pression de deutérium (resp. d'hydrogène). La progression de cette réaction est suivie par la spectroscopie IR. Cette réaction d'échange d'isotopes fournit l'énergie d'activation minimale nécessaire à la rupture de la liaison B-H sans induire d'autres changements structurels dans le composé. On constate que l'énergie d'activation pour Ca(BH4)2 est significativement plus grande que pour le composé analogue au Mg, ce qui implique que l'ion métallique joue un rôle important dans la rupture de la liaison B-H.

Le dernier chapitre consacré aux borohydrures, étudie la réaction de Mg(BH4)2 avec le CO2. La consommation de CO2 et la formation de B2H6 sont contrôlées en utilisant la spectroscopie IR à phase gazeuse. Ces expériences montrent que le γ-Mg(BH4)2 poreux réagit facilement avec le CO2 même à température ambiante, tandis que la phase stable de α-Mg(BH4)2 réagit uniquement à température élevée. Cette réaction peut initier d'autres études concernant l’utilisation des borohydrures pour la réduction du CO2 ainsi que pour la régénération du carburant.

2- La deuxième partie de cette thèse porte sur les dérivés closoboranes de l'ion B12H12 . Les composés issus de cette famille ont récemment attiré un grand intérêt en tant que conducteurs ioniques solides pour les ions Li et Na. [1]

iii

2- Les résultats des calculs DFT sur les anions B12H12 isolés et les anions substitués par les halogènes (F, Cl ou Br) ont été analysé en détail. Le but de cette étude est de prédire les données spectroscopiques pour les composés halogénés, puisque ces composés partiellement substitués sont difficile à préparer et à purifier. Des tendances systématiques dans diverses propriétés, comme par exemple pour l'étirement de la liaison B-H par rapport au nombre d'atomes d'hydrogène, sont observés. La stabilité thermodynamique de divers isomères est également discutée dans cette partie.

Le dernier chapitre expérimental se concentre sur la synthèse de certains substitués des closoboranes. Un mélange de Na2B12(SCN)nH(12-n) avec (n =1 et 2) a été synthétisé. La mesure de la conductivité ionique de cet échantillon montre qu’à température ambiante, la conductivité est

1000 fois meilleure par rapport au composé Na2B12H12 pur. La bromation complète et partielle de Na2B12H12 est également rapportée. Cependant, la bromation partielle donne un mélange de différents produits. Et finalement, des résultats préliminaires sur d'autres systèmes sont également inclus dans cette partie.

[1] Hansen, B. R. S.; Paskevicius, M.; Li, H.-W.; Akiba, E.; Jensen, T. R., Coord. Chem. Rev. Metal boranes: Progress and applications, 2016, 323, 60-70.

Summary

This thesis deals with the fundamental studies of some materials containing boron-hydrogen bonds which can potentially be used either as the hydrogen storage materials, as the solid electrolytes for batteries or as reducing agents for CO2.

In the first part of thesis, experimental studies on metal borohydrides M(BH4)n are reported.

The first experimental chapter presents the synthesis and characterization of halide-free

Sr(BH4)2, Ba(BH4)2 and Eu(BH4)2 starting from the metal hydride and Et3N-BH3. These compounds are more stable upon heating compared to other preparations reported in the literature involving halogen impurities. The crystallographic study of these compounds allowed identifying several new phases. During the thermal decomposition reactions, the formation of a metal hydride-borohydride M2H3(BH4) for M = Sr and Eu was observed for the first time.

Eu(BH4)2 has been shown to be an excellent phosphor emitting at 463 nm (blue colour). Several unsuccessful or partially successful reaction schemes which end up in impure products are also discussed.

In the next chapter, the hydrogen-deuterium exchange reaction as well as the reverse reaction is studied for Ca(BH4)2 as a function of temperature and deuterium (resp. hydrogen) pressure. The progress of this reaction is monitored using IR spectroscopy. This isotope exchange reaction provides the minimum activation energy of breaking the boron hydrogen bond without inducing further structural changes in the compound. It is observed that the activation energy for Ca(BH4)2 is significantly larger than for the analogous Mg compound, which implies that the metal ion plays an important role in the breaking of the B-H bond.

The last chapter dedicated to borohydrides explores the reaction of Mg(BH4)2 with CO2. The consumption of CO2 and formation of B2H6 is monitored using gas phase IR spectroscopy. These experiments show the porous γ-Mg(BH4)2 reacts readily with CO2 even at room temperature, while the stable α-Mg(BH4)2 phase reacts only upon heating. This reaction may initiate further studies to use borohydrides for CO2 reduction and possible fuel regeneration.

2- The second part of the thesis focuses on closoboranes derived from the B12H12 ion. Compounds of this family have recently attracted great interest as solid ionic conductors for Li and Na ions.[1]

2- Results of DFT calculations on isolated B12H12 anions and halogen (F, Cl or Br) substituted anions were analysed in detail. The focus of this study was the prediction of spectroscopic data for the halogenated compounds, as partially substituted compounds are not easily prepared and purified. Systematic trends in various properties, for instance, B-H stretching vs the number of hydrogen atoms are reported. The thermodynamic stability of various isomers is also discussed.

The last experimental chapter focuses on the synthesis of some substituted closoboranes. A mixture of Na2B12(SCN)nH(12-n) with (n =1 and 2) was synthesized. The measurement of the ionic conductivity of this sample shows that the conductivity at room temperature is much better (1000 fold) compared to pure Na2B12H12. Complete and partial bromination of Na2B12H12 is also reported. The partial bromination gives however mixture of different products. Further preliminary results on other systems are also included.

[1] Hansen, B. R. S.; Paskevicius, M.; Li, H.-W.; Akiba, E.; Jensen, T. R., Coord. Chem. Rev, Metal boranes: Progress and applications, 2016, 323, 60-70.

Table of Contents Chapter 1 Introduction ...... 1 1.1 Introduction ...... 3 1.2 References ...... 10 Chapter 2 Experimental & Computational Techniques ...... 13 2.1 Structural characterization ...... 15 2.1.1 Fourier Transform Infrared Spectroscopy ...... 16 2.1.1a Attenuated Total Reflectance (ATR) ...... 16 2.1.2 Raman Spectroscopy ...... 17 2.1.3 Powder XRD ...... 17 2.1.4 DSC ...... 17 2.1.5 NMR ...... 17 2.2 Computational characterization ...... 18 2.2.1 Calculation on crystals ...... 18 2.2.2 Orthogonalized Plane Waves (OPWs) method ...... 19 2.2.3 Pseudopotential ...... 20 2.2.4 Kinetic energy cutoff and Monkhorst-Pack grid ...... 22 2.2.5 Details of the periodic calculations performed ...... 23 2.3 References ...... 24 Chapter 3 Synthesis, Structure and Decomposition of Metal Hydrides and Borohydrides ..... 27 3.1 Introduction ...... 29 3.2 Synthesis and Discussion ...... 31

3.2.1a Synthesis of EuH2 ...... 31 3.2.1b Results and Discussion...... 31

3.2.2a Synthesis of M(BH4)2 (M=Sr, Ba, Eu) ...... 32 3.2.2b Results and Discussions ...... 32 3.2.3 Alternative Synthesis Methods ...... 46

3.2.3 a, b & c Synthesis of (BaBH4)2 ...... 46 Results and Observations ...... 46

3.2.3 d, e & f Synthesis of Sr(BH4)2 ...... 47 Results and Observations ...... 47 3.3 Conclusions ...... 49 3.4 Supporting Information ...... 50

vii

3.5 References ...... 52 Chapter 4 Isotope Exchange Reactions ...... 55 4.1 Introduction ...... 57 4.2 Experimental Section ...... 59

4.2A. Deuterium exchange reactions in Ca(BH4)2 via solid gas reaction ...... 59

4.2B. Deuterium exchange reactions in Na2B12H12 via solid gas reaction ...... 60

4.2C. Deuterium exchange reactions in Na2B12H12 in DCl/D2O solution ...... 60 4.3 Results and Discussion ...... 62

4.3A. Solid Gas Reaction in Ca(BH4)2 ...... 62

4.3 B. Deuteration by Solid Gas Reaction in Na2B12H12 ...... 70

4.3 C. Deuterium exchange reactions in Na2B12H12 in DCl/D2O solution ...... 74 4.4 Conclusions ...... 81 4.5 References ...... 82

Chapter 5 Reduction of CO2 using γ-Mg(BH4)2 ...... 85 5.1 Introduction ...... 87 5.2 Experimental Procedure ...... 89 5.2.1 Room Temperature, in-situ IR measurement using glass cell ...... 89 5.2.2 Room temperature, in-situ IR measurements using metallic cell ...... 90 5.2.3 High temperature (100 °C), in-situ IR measurements using metallic cell ...... 90

5.2.4 Ex-situ measurements (α- Mg(BH4)2) ...... 90 5.3 Results and Observations ...... 91 5.3.1 Results for experimental section 5.2.1 ...... 93 5.3.2 Results for experimental section 5.2.2 and 5.2.3 ...... 96 5.3.3 Results for experimental section 5.2.4 ...... 99 5.4 Conclusions ...... 100 5.5 References ...... 102 2- Chapter 6 Theoretical calculations on halogenated closoboranes B12H12-nXn (X=F,Cl,Br) .... 105 6.1 Introduction ...... 107 6.2 Computational Details ...... 108 6.3 Results and Discussions ...... 109 6.3.1 Relative Stability ...... 110 6.3.2 Isomerism ...... 111 6.3.3 Bond Length ...... 112

viii

6.3.4 Vibrational Spectra ...... 113 6.3.5 NMR spectra ...... 118 6.4 Conclusions ...... 119 6.5 Supporting Information ...... 120 6.6 References ...... 122 2- Chapter 7 Synthesis of Various Substituted Closoboranes (B12XnH(12-n) ) ...... 125 7.1 Introduction ...... 127 7.2 Synthesis, Results and Discussions ...... 128

7.2.1a Synthesis of ((C2H5)3NH)2B12H12)...... 128 7.2.1b Results and Observation...... 128

7.2.2a Synthesis of CaB12H12 ...... 130 7.2.2b Results and Observations ...... 130

7.2.3a Bromination of Na2B12H12 ...... 132 7.2.3b Results and Observations ...... 132 7.2.4 a Synthesis of thiocyanated closoborane ...... 138 Scheme 1 ...... 138 Scheme 2 ...... 138 7.2.4 b Results and Observation ...... 138 7.3 Conclusions ...... 142 7.4 References ...... 143 Chapter 8 Conclusions ...... 145 8.1 Conclusions ...... 147 8.2 References ...... 153

Chapter 1

Introduction

1.1 Introduction

Motivation Anthropogenic factors have caused serious climatic changes.1-4 One of the main reasons of these changes is the consumption of the fossil fuels which generates greenhouse gases.5-6 Renewable sources of energy such as solar radiation, wind, and tides provide green energy and are the key to overcome the challenge of climate change.7-9 In order to use renewable sources efficiently they need to be supported with ideal energy storage systems. Hydrogen can be used as a fuel to meet our mobile energy needs (fuel in automobiles). Hydrogen is regarded as a green fuel as it produces water upon combustion.10 Another way to store energy is the use of batteries, which can be designed for mobile or stationary applications.

The volumetric energy capacity of hydrogen is low. At the ambient conditions 1 kg of hydrogen gas occupies a volume of 11m3. In order to store hydrogen efficiently, the main methods which are available are: (i) storing hydrogen at cryogenic temperature, (ii) storing hydrogen under high pressure, (iii) physisorption of hydrogen on high surface materials, (iv) metal hydrides (incorporation of hydrogen in the host metallic structures) and (v) complex hydrides, chemisorption of hydrogen (vi) storage via chemical reactions, like metals with water.11 Each of the above listed methods has its own set of challenges to overcome before an ideal hydrogen storage system can be developed.

Batteries provide an efficient way to store electrical energy in the form of chemical energy, which can subsequently be released without any gaseous exhaust.12-13 In this context, it is essential to investigate the path to design economical, safe and rechargeable batteries.14-15 Among the available rechargeable batteries, lithium ion batteries (LIBs) present high gravimetric and volumetric energy density which is essential for portable (cell phone, laptop) and mobile (cars) applications. The available big resources of lithium are concentrated in few places, currently 75% of the world production of lithium comes from Chile, Australia and China. The potential strong increase of production of batteries for automobile applications in the near future makes it essential to look for an economical and geopolitically neutral alternative for lithium.

Sodium ion batteries (SIBs) are interesting candidates to be investigated in this context, as sodium is present in abundance all throughout the globe. However, there are currently still

several challenges to overcome to produce these batteries for mobile applications. One important component in a battery is the electrolyte. Currently, Natrium Superionic CONductors (NASICON) are well known electrolytes but they exhibit reasonable ionic conductivity only at high temperature. The development of solid sodium electrolytes with liquid-like ion conductivity at ambient temperature would be a big step towards developing safe all-solid state sodium-ion batteries for large scale energy storage applications.

This thesis deals with the fundamental studies of some materials containing boron-hydrogen bonds which can potentially be used as hydrogen storage materials, as solid electrolytes for sodium ion batteries or as reducing agents for CO2.

The different research projects of this thesis are part of a long term research of our group on boron hydrogen compounds. After more than a decade of investigation the synthesis, structural - and vibrational properties of different compounds with BH4 , the topic of ongoing projects has been expanded as shown in the figure 1.1.

Figure 1.1: Schematic representation of research projects on boron hydrogen related compounds in Prof. Hans-Hagemann’s lab. Dark green boxes represents the work covered in this thesis.

Diversity in structures and applications of boron-hydrides 16 Boron hydrogen compounds BxHy were first studied by Alfred Stock. He decomposed magnesium boride with acid, to obtain the first boron hydride B4H10 (liquid) whose physical and chemical properties were thoroughly investigated. Heating of B4H10 yielded diborane B2H6 (gas) 17 and the thermal decomposition of diborane resulted in the formation of B10H14 (crystal). Apart from regular 2-center-2- electron covalent bonds, boron can also form 3-center-2-electron (3c- 18-20 2e) covalent bonds (with hydrogen and boron itself). The structure of B2H6 contains two 3c- 2e bonds, this structure of diborane was initially proposed by Christopher Longuet-Higgins.21 These fascinating bonding properties led to many experimental and theoretical studies on boron- hydrogen compounds, with many experiments being performed more than 50 years ago. For instance, the volatile compound U(BH4)4 was considered for the separation and purification of 22-23 uranium in the Manhattan project. Some compounds like liquid B5H9 or solid B10H14 have been considered as potential rocket fuel about 50 years ago because of their high energy content; however their toxicity is a serious impediment for this application.24 Today, over 25 neutral n- 25 boron-hydrogen compounds and an even larger number of borane anions BxHy are known. The structures of some of these compounds can be classified according to the Wade rule as 2- 26-28 arachno- (BnHn+6), nido- (BnHn+4) and closo- (BnHn ) boranes.

- This thesis focuses mainly on two types of compounds, namely tetrahydroborates (BH4 ) 2- and dodecahydro-closo-dodecaborates (B12H12 ). Tetrahydroborates are referred as borohydrides and dodecahydro-closo-dodecaborates have been referred as closoboranes throughout in the thesis.

Metal borohydrides have various properties. They find well established applications in organic chemistry as reducing agents.29-30 The reducing nature of borohydrides was first investigated by Schlesinger and H.C Brown.30 Since then borohydrides are used as reducing agents in organic chemistry (H. C. Brown was awarded with Nobel Prize due to his work in the field of borane- organoborane in 1979). Since about 15-20 years, they are studied for their potential as light hydrogen storage material (for mobile applications) and more recently as fast ion conductors for new types of batteries.31-35

Hydrogen desorption properties of borohydrides were already investigated in the middle of last century.36-37 A revived interest in metal hydrides and borohydrides came after Bogdanovic

demonstrated a reversible hydrogen release and uptake in titanium doped aluminium hydride. 38 Since then many new metal borohydrides have been synthesized and characterized. However, with some transition metals (eg Zn), instead of hydrogen the release of diborane has been observed. A brief summary of the research done on metal borohydrides in the last decade is presented in a recent review.39

The thermal decomposition of metal borohydrides goes through a series of intermediate steps n+ 2- 40 and can result in the formation of higher boranes like M (B12H12 )n/2 (where M is the metal). Closoboranes have recently gained an increased interest as potential electrolyte for all solid state batteries. Na2B12H12 shows a dramatic high superionic conductivity (~0.1 S/cm) after an order- disorder phase transition at high temperature (~540 K).41-42 Tang et al have studied the structural changes and phase transitions by anion and cation modifications of closoboranes M2B12X12 (M = alkali metal, X=H, Cl, Br).43-44

Outline of the thesis - This thesis is divided into two sections. The first section deals with borohydrides (BH4 ) and the 2- second section deals with closoboranes (B12H12 ). In the beginning of the first section, different synthesis schemes to synthesize metal borohydrides are discussed. The following chapters in this section show the results and observations for the different projects done on the materials synthesised in the synthesis part. These projects show the versatile nature of applications of borohydrides in different fields like hydrogen storage and reduction of CO2 gas.

The figure above shows the different aspects considered in this thesis. In the first part, high temperature decomposition reactions, the activation energy for the breaking of the boron- hydrogen bond monitored by isotopic substitution reactions, and finally the use of (γ-Mg(BH4)2) as reducing agent for carbon dioxide are studied. In the second part, closoboranes and substituted closoboranes as potential solid state electrolytes are studied both experimentally and theoretically. These different projects are described in more detail in the following paragraphs.

Chapter 3 of this thesis is dedicated to the “Synthesis, Structure and decomposition of Halide Free Metal Borohydride” in this regard. A detailed study of the synthesis of several borohydrides is reported while for europium, the synthesis of europium hydride starting from pure europium metal is also discussed. The compounds discussed in this section cannot be used as hydrogen

storage materials as the mass of the cation (metal) is large, but these materials can also be used to synthesize mixed cationic compounds. Studying these compounds also imparts us a better understanding of the dehydrogenation process in borohydrides.

In order to investigate the hydrogen release process from borohydrides, it is essential to understand boron-hydrogen bond breaking. Chapter 4 of this thesis, “Isotope exchange reaction in Ca(BH4)2” is an attempt to study the kinetics and thermodynamics of boron hydrogen breaking without causing any major structural degradation due to the thermal decomposition in Ca(BH4)2. Deeper understanding of this step would help us to better understand and control the dehydrogenation step. We have used infrared spectroscopy as a tool to monitor the isotope exchange reaction. The distinct bands for B-H and B-D stretching, facilitates the calculation of amount of H or D in the reaction sample and thus kinetic parameters for the reaction can also be calculated.

In a collaborative work with Prof. Y. Fillinchuk (Université Catholique de Louvain) and Prof. H. W. Li (International Research Center for Hydrogen Energy, Kyushu University, Japan), we have investigated reducing nature of a porous phase of magnesium borohydride γ-Mg(BH4)2. The kinetics of the reduction reaction is studied using gas phase infrared spectroscopy. These results are so far unpublished and summarized in chapter 5 of this thesis.

Furthermore, in a research project of our laboratory in collaboration with Prof. T. Jensen (Aaarhus University), we have studied the destabilization of alkali borohydrides by the addition - of BF4 which led to a dramatic lowering of the decomposition temperature and the formation of 2- 45 B12H12 . This led us to investigate theoretically the structure and spectroscopic properties of 2- various B12HnX(12-n) (X = F, Cl or Br and n = 1-3, 9-12) anions. The results of this computational study are presented in chapter 6 of the thesis. The results for fluoride mixed closoboranes have been published in the reference [33], while the results on bromide and chloride mixed closoboranes are still unpublished.46

Chapter seven of this thesis deals with developing a system based on closoboranes which can exhibit ionic conductivity in the range of few mS/cm at room temperature. Due to the high interest in the area of potential efficient solid ionic conductors, we have started a series of

experiments to synthesize and purify halide-hydride mixed closoboranes. Preliminary experimental results on the synthesis of substituted closoboranes are presented in chapter 7.

In order to avoid numerous repetitions, we have summarized in chapter 2 entitled “Experimental Techniques” the spectroscopic techniques and computational details which have been used throughout the entire thesis, while more specific details are given in each chapter separately.

1.2 References

[1] Vitousek, P. M.; Mooney, H. A.; Lubchenco, J.; Melillo, J. M., Science Human Domination of Earth's Ecosystems, 1997, 277, 494-499.

[2] Nature Clim. Change The human factor, 2012, 2, 555-555, doi:10.1038/nclimate1657.

[3] Min, S.-K.; Zhang, X.; Zwiers, F. W.; Hegerl, G. C., Nature Human contribution to more-intense precipitation extremes, 2011, 470, 378-381.

[4] Syvitski, J. P. M.; Kettner, A. J.; Overeem, I.; Hutton, E. W. H.; Hannon, M. T.; Brakenridge, G. R.; Day, J.; Vorosmarty, C.; Saito, Y.; Giosan, L.; Nicholls, R. J., Nature Geosci Sinking deltas due to human activities, 2009, 2, 681-686.

[5] Canadell, J. G.; Le Quéré, C.; Raupach, M. R.; Field, C. B.; Buitenhuis, E. T.; Ciais, P.; Conway, T. J.; Gillett, N. P.; Houghton, R. A.; Marland, G., Proceedings of the National Academy of Sciences Contributions to accelerating atmospheric CO2 growth from economic activity, carbon intensity, and efficiency of natural sinks, 2007, 104, 18866-18870.

[6] Lehmann, J., Nature A handful of carbon, 2007, 447, 143-144.

[7] Chu, S.; Cui, Y.; Liu, N., Nat Mater The path towards sustainable energy, 2017, 16, 16-22.

[8] Gielen, D.; Boshell, F.; Saygin, D., Nat Mater Climate and energy challenges for materials science, 2016, 15, 117-120.

[9] Streimikiene, D.; Sivickas, G., Environ Int The EU sustainable energy policy indicators framework, 2008, 34, 1227-40.

[10] Nat. Energy Hydrogen on the rise, 2016, 1, 16127.

[11] Züttel, A., Mater. Today Materials for hydrogen storage, 2003, 6, 24-33.

[12] Palomares, V.; Serras, P.; Villaluenga, I.; Hueso, K. B.; Carretero-Gonzalez, J.; Rojo, T., Energy Environ. Sci. Na-ion batteries, recent advances and present challenges to become low cost energy storage systems, 2012, 5, 5884-5901.

[13] Dietrich, P.; Wokaun, A.; Bauer, C., Aqua Gas Chemical energy storage. A contribution to the energy strategy 2050, 2012, 92, 22-27.

[14] Armand, M.; Tarascon, J. M., Nature Building better batteries, 2008, 451, 652-657.

[15] Tarascon, J. M.; Armand, M., Nature Issues and challenges facing rechargeable lithium batteries, 2001, 414, 359-367.

[16] Johnson, W. C., J. Chem. Educ. Hydrides of Boron and Silicon (Stock, Alfred), 1934, 11, 256.

[17] Wiberg, E., Pure Appl. Chem. Alfred Stock and the renaissance of inorganic chemistry, 1977, 49, 691-700.

[18] Eberhardt, W. H.; Crawford, B., Jr.; Lipscomb, W. N., J. Chem. Phys. The valence structure of the boron hydrides, 1954, 22, 989-1001.

[19] Price, W. C., J. Chem. Phys. The Structure of Diborane, 1947, 15, 614-614.

[20] Price, W. C., J. Chem. Phys. The Absorption Spectrum of Diborane, 1948, 16, 894-902.

[21] Longuet-Higgins, H. C.; Bell, R. P., J. Chem. Soc. 64. The structure of the boron hydrides, 1943, 250-255.

[22] Ghiassee, N.; Clay, P. G.; Walton, G. N., Journal of Inorganic and Nuclear Chemistry Thermal decomposition of U(BH4)4, 1981, 43, 2909-2913.

[23] Paine, R. T.; Schonberg, P. R.; Light, R. W.; Danen, W. C.; Freund, S. M., Journal of Inorganic and Nuclear Chemistry Photochemistry of U(BH4)4 and U(BD4)4, 1979, 41, 1577-1578.

[24] Agarwal, J. P., Wiley-VCH, Weinheim, Germany High Energy Materials: Propellants, Explosives and Pyrotechnics, 2010.

[25] Greenwood, N. N.; Earnshaw, A., Chemistry of Elements. VCH Verlagsgesellschaft: 1988; p 1700 pp.

[26] Wade, K., Nat Chem Bonding with boron, 2009, 1, 92-92.

[27] Mingos, D. M. P., Accounts of Chemical Research Polyhedral skeletal electron pair approach, 1984, 17, 311-319.

[28] Jemmis, E. D.; Balakrishnarajan, M. M., Journal of the American Chemical Society Polyhedral Boranes and Elemental Boron: Direct Structural Relations and Diverse Electronic Requirements, 2001, 123, 4324-4330.

[29] Zhu, Y.; Hosmane, N. S., Coord. Chem. Rev. Nanocatalysis: Recent advances and applications in boron chemistry, 2015, 293–294, 357-367.

[30] Schlesinger, H. I.; Brown, H. C.; Finholt, A. E.; Gilbreath, J. R.; Hoekstra, H. R.; Hyde, E. K., J. Am. Chem. Soc. Sodium Borohydride, Its Hydrolysis and its Use as a Reducing Agent and in the Generation of Hydrogen1, 1953, 75, 215-219.

[31] Schlapbach, L.; Zuettel, A., Nature Hydrogen-storage materials for mobile applications, 2001, 414, 353-358.

[32] Matsuo, M.; Orimo, S.-i., Adv. Energy Mater. Lithium Fast-Ionic Conduction in Complex Hydrides: Review and Prospects, 2011, 1, 161-172.

[33] Li, C.; Peng, P.; Zhou, D. W.; Wan, L., Int. J. Hydrogen Energy Research progress in LiBH4 for hydrogen storage: A review, 2011, 36, 14512-14526.

[34] He, T.; Pachfule, P.; Wu, H.; Xu, Q.; Chen, P., Nat. Rev. Mater. Hydrogen carriers, 2016, 1, 16059.

[35] Ley, M. B.; Jepsen, L. H.; Lee, Y.-S.; Cho, Y. W.; Bellosta von Colbe, J. M.; Dornheim, M.; Rokni, M.; Jensen, J. O.; Sloth, M.; Filinchuk, Y.; Joergensen, J. E.; Besenbacher, F.; Jensen, T. R., Mater. Today (Oxford, U. K.) Complex hydrides for hydrogen storage - new perspectives, 2014, 17, 122-128.

[36] Fedneva, E. M.; Alpatova, V. I.; Zh, V. I. M., Neorgan. Khim. Thermal stability of lithium borohydride, 1964, 9, 1519-20.

[37] Stasinevich, D. S.; Egorenko, G. A., Zh. Neorg. Khim. Thermographic study of borohydrides of alkali metals and magnesium at pressures up to 10 atm, 1968, 13, 654-8.

[38] Bogdanović, B.; Schwickardi, M., J. Alloys Compd. Ti-doped alkali metal aluminium hydrides as potential novel reversible hydrogen storage materials1, 1997, 253–254, 1-9.

[39] Paskevicius, M.; Jepsen, L. H.; Schouwink, P.; Cerny, R.; Ravnsbaek, D. B.; Filinchuk, Y.; Dornheim, M.; Besenbacher, F.; Jensen, T. R., Chem. Soc. Rev. Metal borohydrides and derivatives - synthesis, structure and properties, 2017, 46, 1565-1634.

[40] Welchman, E.; Thonhauser, T., Journal of Materials Chemistry A Decomposition mechanisms in metal borohydrides and their ammoniates, 2017, 5, 4084-4092.

[41] Udovic, T. J.; Matsuo, M.; Unemoto, A.; Verdal, N.; Stavila, V.; Skripov, A. V.; Rush, J. J.; Takamura, H.; Orimo, S.-i., Chem. Commun. Sodium superionic conduction in Na2B12H12, 2014, 50, 3750-3752.

[42] Verdal, N.; Her, J.-H.; Stavila, V.; Soloninin, A. V.; Babanova, O. A.; Skripov, A. V.; Udovic, T. J.; Rush, J. J., J. Solid State Chem. Complex high-temperature phase transitions in Li2B12H12 and Na2B12H12, 2014, 212, 81-91.

[43] Udovic, T. J.; Matsuo, M.; Tang, W. S.; Wu, H.; Stavila, V.; Soloninin, A. V.; Skoryunov, R. V.; Babanova, O. A.; Skripov, A. V.; Rush, J. J.; Unemoto, A.; Takamura, H.; Orimo, S.-i., Advanced Materials Exceptional Superionic Conductivity in Disordered Sodium Decahydro-closo-decaborate, 2014, 26, 7622-7626.

[44] Tang, W. S.; Udovic, T. J.; Stavila, V., J. Alloys Compd. Altering the structural properties of A2B12H12 compounds via cation and anion modifications, 2015, 645, Supplement 1, S200-S204.

[45] Rude, L.; Filso, U.; D'Anna, V.; Spyratou Stratmann, A.; Richter, B.; Hino, S.; Zavorotynska, O.; Baricco, M.; Sørby, M. H.; Hauback, B. C.; Hagemann, H.; Besenbacher, F.; Skibsted, J.; Jensen, T. R., Phys. Chem. Chem. Phys. Hydrogen-fluorine exchange in NaBH4-NaBF4, 2013, 15, 18185-18194.

[46] Sharma, M.; Sethio, D.; D'Anna, V.; Hagemann, H., Int. J. Hydrogen Energy Theoretical study of B12HnF(12-n)2- species, 2015, 40, 12721-12726.

Chapter 2 Experimental &

Computational Techniques

2.1 Structural characterization

The structural characterizations discussed in the thesis were performed using FTIR, Raman, XRD, DSC, NMR and mass spectrometry.

FTIR spectroscopy has been systematically used to check not only the purity of the samples - (residual solvent removal etc), but also to study to local structure of BH4 based on the expertise acquired in our laboratory in the last 15 years. Raman spectra complement these analysis. Also, during isotope exchange, the frequency of vibrations for different isotopes changes, as the frequency is inversely proportional to square root of mass. Thus IR/Raman spectra were also used to monitor the progress of isotope exchange reactions in various projects by monitoring the area under B-H and B-D stretching peaks.

Differential Scanning Calorimetry coupled to thermogravimetry (DSC/TG) was used to check the thermal stability of synthesised products. The temperature at which products thermally decomposes or show any phase transformation were obtained using DSC.

Temperature dependent synchrotron X-ray powder diffraction (SR-XPD) analysis was used to analyse the crystal structure and their decomposition products. SR-XPD data used for the crystal structure solutions and refinements were collected between RT and 773 K at the Swiss- Norwegian Beamlines of ESRF (European Synchrotron Radiation Facility, Grenoble, France) and analysed by Radovan Cerny and Emilie Didelot (Department of Quantum Matter Physics of the University of Geneva). Periodic density functional theory (DFT) calculations were applied to identify the exact position of hydrogen for Ba(BH4)2.

DFT calculations on isolated species in the gas phase also helped to characterize halogen substituted closoboranes. All the possible positional isomers for each compound were also characterized with the help of DFT calculations. Experimentally synthesizing halogen substituted 2- closoboranes B12H(12-n)Xn in pure phase for X = F, Cl and Br and n =0-3 and 9-12 is challenging. The halogen gases X2 (X=F, Cl, and Br) are highly reactive and may rapidly form a multitude of compounds (mono, di or tri halo substituted) compared to pseudo halogen substituted (–SCN) closoboranes. Thus pseudo-halogen substituted closoboranes were synthesized to form mono-thiocyanate substituted compounds. NMR and MS spectra were used to characterize the substitution. 11B {1H} NMR also helped to find out the progress of deuteration in Na2B12H12 during reaction with concentrated DCl.

Mass spectrometry is helpful to analyse the different compounds formed during bromination of sodium closoboranes.

2.1.1 Fourier Transform Infrared Spectroscopy

Attenuated Total Reflection Fourier Transform Infrared Spectroscopy (ATR-FTIR) was performed on solid samples. Biorad Excalibur instrument equipped with a Specac Golden Gate heat-able ATR set-up was used for FTIR. The spectral resolution was set to 1 cm-1. Samples were loaded on ATR setup in an inert atmosphere of Nitrogen. Powdered sample was pressed between diamond crystal and bridge clamped sapphire anvil to ensure optimum optical contact of the powder.

2.1.1a Attenuated Total Reflectance (ATR)

ATR-FTIR is based on attenuation effect of light when infrared light is internally reflected at the interface of diamond and sample (from high refractive index medium to low refractive index medium). Light penetrates into the rarer medium as an evanescent wave. The penetration depth of evanescent wave is in the order of microns (0.2 – 5 µm).

An infrared beam is directed to a diamond crystal (optical dense medium, refractive index 2.417 for yellow light) at the specific angle to create the total internal reflection. As a result of the internal reflection, an evanescent wave is created beyond the surface of the diamond. This evanescent wave interacts with the sample which is in close contact with diamond crystal (due to pressure by a spring loaded sapphire anvil), thus sample absorbs parts of the light wave (bands of the infrared spectrum). Reflected wave reaching the detector Evanescent wave is passed back to IR beam and finally to a detector.1

Figure 2.1: Schematic representation of a single reflection ATR system, similar to the one used for FTIR mentioned in the thesis.

2.1.2 Raman Spectroscopy

Raman spectra were recorded with a homemade setup. The laboratory assembled Raman Instrument consists of a solid state 50mW 488nm laser, a Kaiser Optical Holospec f/1.8 monochromator equipped with the Princeton Instruments liquid nitrogen cooled CCD camera. The instrument is controlled by a program developed by D. Lovy at the department of Chime Physique at the University of Geneva. All measurements were performed in the backscattering conﬁguration. The samples were contained in standard melting point capillaries which were ﬁlled in the glove box and then sealed using vacuum grease.

2.1.3 Powder XRD

Powder X-ray diffraction patterns were obtained on a STOE STADI P diffractometer in Debye-

Scherrer geometry with monochromated Cu-Kα1 radiation. The phase composition was determined by the Rietveld method using the software FullProf2.

2.1.4 DSC

Differential scanning calorimetry data were measured using a NETZSCH STA449 F3 instrument. The measurements were performed under an inert atmosphere of nitrogen with a purge rate of 20ml/min. The samples were contained in Al2O3 (or Aluminium, maximum temperature = 600 °C) crucibles with a lid to prevent exposure to atmosphere while mounting. Samples were loaded in crucibles in an inert atmosphere of nitrogen. The heating rate was set at 10 °C/min.

2.1.5 NMR

11B NMR experiments were recorded at 298 K on a Bruker AVANCE III HD-NanoBay 300 MHz (and 400 MHz) spectrometer, equipped with a 5 mm BB(F)-H-D probe, at a frequency of 300.13 MHz (400.13 MHz).

2.2 Computational characterization

2.2.1 Calculation on crystals

The theoretical description of the electronic structure of a crystalline system relies on the use of the Bloch’s theorem, which states that the eigenfunctions of the Hamiltonian of the system can be written as the product of a plane wave and a function with the periodicity of the Bravais lattice of the system. A wavefunction describing an electronic state of the system thus reads:

풊풌.풓 휓푛풌(풓) = 풆 푢푛풌(풓) (2.1) where 푢푛풌(풓) is a periodic function with the periodicity of the Bravais lattice and thus verifies for any lattice vector R,

푢푛풌(풓 + 푹) = 푢푛풌(풓), (2.2) k is a vector of the reciprocal space and the quantum number n is the so-called band index used to distinguish the many solutions to the Schrödinger equation at a given k. The name of this quantum number is due to the fact that to each solution is associated an eigenvalue n which is a continuous function of k, termed “band” with band index n. The set of functions 푛풌 = 푛(풌) defines the band structure of the system.

Practically, given that for any reciprocal lattice vector K: 휓푛,풌+푲(풓) = 휓푛,풌(풓), one lets k take values in the first Brillouin zone only so as to achieve a unique description of the electronic structure. Furthermore, the functions 풖풏풌(풓) being periodic, it can be expanded in a basis set of plane waves, which allows the use of efficient Fourier transform methods.

However, the use of a plane-wave basis set poses a challenge because an enormous number of plane waves are necessary to represent the strongly localized core orbitals and the inner nodal structures of the valence wavefunctions, thus possibly making the use of plane waves computationally prohibitive. In order to overcome this challenge, the ‘frozen core approximation’ can be used. According to this approximation, the contribution of the core states to the chemical bonding and to the properties of the system is negligibly small and thus one can ignore their variation; that is, keep them frozen to their values in a reference system.

Based on this approximation, several methods have been developed to efficiently treat the electronic structures of crystalline systems using plane waves. This includes the plane-wave pseudopotential methods, which have been used in this work, or the Projector Augmented Waves (PAWs) method, which all have been preceded by the Orthogonalized Plane Wave (OPWs) method.3-4

2.2.2 Orthogonalized Plane Waves (OPWs) method

The OPWs method was developed in 1940 by Herring.5 In this method, the valence states are built by using plane waves which have been made orthogonal to the core states, the so-called

OPWs. An OPW 휙푘 thus reads:

풊풌.풓 훟풌 = 풆 + ∑풋 풃풌풋풖풋 (풓) (2.3) where 푢푗 is the wavefunction of the j-th core level, (2.4) and 푏푘푗 is a constant which is defined such that 휙푘 is orthogonal to all the core levels, thus:

∗ 풊풌.풓 풃풌풋 = − ∫ 풅풓 풖풋 (풓)풆 (2.5)

푢푗 is an eigenfunction of the Schrodinger equation for the core states

ퟏ 휵ퟐ풖 + (푬 − 푽 )풖 = ퟎ (2.6) ퟐ 풋 풋 풋 풋

퐸푗 is the total energy associated to 푢푗 and 푉푗 denotes the potential of the system.

Using the OPW method, the wavefunction of a valence state (휓푣) of a periodic crystal is described by a linear combination of OPWs. It is the sum of a smooth wavefunction (휓̌풗) and a sum over the localized wavefunctions of the core states (∑푗 퐵푗푢푗 (푟)): 푣 ̌ 푣 휓 (풓) = 휓 (풓) + ∑푗 퐵푗푢푗 (풓) (2.7)

All the quantities of the relation (2.7) can be expressed in terms of the original OPWs rewritten

푂푃푊 1 푖풒.풓 푖풒.풓 푥푞 (풓) = 훺{푒 − ∑푗⟨푢푗|풒⟩ 푢푗(풓)}, where ⟨푢푗|풒⟩ = ∫ 푑풓 푢푗(풓)푒 (2.8, 2.9)

by the Fourier transforms:

푣 푂푃푊 휓 (풓) = ∫ 푑풒 푐(풒)휒풒 (풓) (2.10)

풗 휓̌ (풓) = ∫ 푑풒 푐(풒)푒푖풒.풓 (2.11)

퐵푗 = ∫ 푑풒 푐(풒)⟨푢푗|풒⟩ (2.12)

The simplest approach is to assume that the core states in the molecules or solids are the same as in the atoms. That is, to choose them to be atomic core orbitals.6

2.2.3 Pseudopotential

The basic idea in pseudopotential methods is to replace the strong Coulomb potential of the nucleus of an atom by a much softer potential which reproduces the complicated effect of both the nucleus and the core electrons. The wavefunctions of the resulting pseudo-atom also referred to as pseudo-wavefunctions are nodeless.

The pseudopotential approximation was introduced by Hans Hellmann in 1934.7 Further development in this field was done by Antonick,8-9 and Philips and Kleinmann10 (PKA). Assuming that the true valence wavefunction 휓푣(푟) solution of

ퟏ 퐻휓푣(푟) = [− 휵ퟐ + 푽] 휓푣(푟) = 휀푣휓푣(푟) (2.13) 푖 ퟐ 푖 푖 푖

reads:

푣 ̌푣 푐 ̌푣 푐 휓푖 (풓) = 휓푖 (풓) − ∑푗 ⟨휓푗 |휓푖 ⟩ 휓푗 (풓) (2.14)

푣 where 휓̌푖 is a smooth wavefunction and where the sum ensures the orthogonality to the core 푐 ̌푣 functions 휓푗 , they showed that the wavefunctions 휓푖 are solutions of

푣 ퟏ 푣 푣 퐻푃퐾퐴휓̌ (푟) = [− 휵ퟐ + 푽푷푲푨] 휓̌ (푟) = 휀푣휓̌ (푟) (2.15) 푖 ퟐ 푖 푖 푖

where 푉푃퐾퐴 is the sought pseudopotential given by 푉푃퐾퐴 = 푉 + 푉푅 . 푉푅 is a repulsive non-local operator that results from the condition of orthogonality; it acts on the pseudo-wavefunction 푣 휓푖 (푟) with the effect shown below:

푅 푣 푣 푐 푐 푣 푐 푉 휓푖 (푟) = ∑푗(휀푖 − 휀푗 ) ⟨휓푗 |휓푖 ⟩ 휓푗 (푟) (2.16)

The above two equations show that the pseudopotential acting on the valence electrons is the sum of a long-range local attractive potential and a short-range repulsive non-local potential and that it is thus softer than the original potential V.

There is no uniqueness in the definition of pseudopotentials. They are derived either empirically or by ab-initio methods. Emprical pseudopotentials were proposed by Cohen and Bergstresser.11 For deriving empirical pseudopotentials, their parameters are varied to obtain a good agreement with the properties of a series of systems. Cohen and Bergstresser used spectroscopic data to construct the pseudopotentials. The ab-initio pseudopotentials are constructed using atomic all- electron calculations. There should be a good agreement between the valence properties calculated using pseudopotentials and all electron methods. In the calculation related with

Ba(BH4)2, described in the thesis ab initio norm-conserving and ultra-soft pseudopotentials were used.

A norm-conserving pseudopotential (NCPP) should be same as the atomic potential outside the 12 “core region” of radius Rc (Rc = chosen core radius). For the pseudopotential to be transferable and reproduce the scattering properties of the real potentials, the logarithmic derivative of the all- 푃푆 electron 휓푙(푟) and nodeless pseudo- 휓푙 (푟) wavefunctions should agree at Rc. Inside the sphere 푃푆 of radius Rc, the pseudopotential and radial pseudo-orbital 휓푙 differ from their all-electron 푃푆 counterparts but the integrated charge (Ql) should be the same for 휓푙 and for the all-electron radial orbital 휓푙 for a valence state. The conservation of Ql insures that the total charge in the core region is correct and the normalized pseudo-orbital is equal to the true orbital outside of Rc.

Ultrasoft pseudopotentials (USPP) allow to perform calculations with the smallest possible cutoff energy for the plane-wave basis set.4 This requirement becomes important if the study involves the second row and 3d elements because a large number of plane waves are required to describe the localized 2p and 3d valence states.13 This approach was developed independently by

Vanderbilt14 and Bloch15. An USPP differs from NCPP mainly due to (a) USPP allow us to choose more than one reference energy per quantum state (l). It guarantees a good transferability over a wide energy range for larger cut off radii rc. (b) USPP allow to choose much larger value for rc as the only restriction is the matching of pseudo and the all-electron function for r > rc.

The ultra-soft and norm-conserving radial pseudo wavefunctions for the 2p state of oxygen are shown in Figure 2.2. An all-electron wavefunction is also shown for the comparison.14

Figure 2.2: 2p radial wavefunction for oxygen (solid line), and corresponding norm-conserving and ultrasoft pseudo wave functions (dotted and dashed line respectively).14

2.2.4 Kinetic energy cutoff and Monkhorst-Pack grid

An infinite plane-wave basis set cannot be used and it must be truncated. The expansion of the wavefunction is limited to plane waves characterized by the reciprocal lattice vectors contained within a sphere of the radius defined by the kinetic energy cutoff, Ecut. A larger value of Ecut corresponds to more flexible basis set and thus to more accurate results but at the expense of a higher computational cost. The choice of Ecut is made carefully in order to ensure that the results obtained for the quantities of interest are close to those that may be obtained in the complete basis limit.

Many calculations in crystals involve integrating periodic functions of a wave vector over either the entire Brillouin zone (BZ) or over specified portions. For this purpose, methods have been developed to give special points of the BZ that allow for efficient integrations. Chadi and Cohen gave a method to compute functions at carefully selected points in the Brillouin zone.16 The most widely used method was proposed by Monkhorst and Pack.17 It gives a uniform set of points determined by eq (2.17).

3 2푛푖−푁푖−1 푘푛1,푛2,푛3 = ∑푖 퐺푖 (2.17) 2푁푖 where the Gi are the primitive vectors of the reciprocal lattice.

2.2.5 Details of the periodic calculations performed

Density functional theory18-19 was applied to the characterisation of the structural and vibrational properties of three candidate structures of Ba(BH4)2. Periodic DFT calculations were thus performed with the PBE functional20 using the Quantum Espresso program package, which is based on pseudopotentials and planewaves21. The calculations consisted in geometry optimisations followed by the determination of the phonons at Γ within density functional perturbation theory22. We employed ultrasoft pseudopotentials23 and a 5×4×4 Monkhorst-Pack grid17, and expanded the wavefunctions and the charge density in plane waves up to a kinetic- energy cutoff of 80 Ry and 800 Ry, respectively.

2.3 References

[1] Kazarian, S. G.; Chan, K. L. A., Analyst ATR-FTIR spectroscopic imaging: recent advances and applications to biological systems, 2013, 138, 1940-1951.

[2] Rodriguez-Carvajal, J., Physica B (Amsterdam) Recent advances in magnetic structure determination by neutron powder diffraction, 1993, 192, 55-69.

[3] Reitz, W., Mater. Manuf. Processes A Review of: "Electronic Structure: Basic Theory and Practical Methods" by R. M. Martin, 2006, 21, 123.

[4] Martin, R. M., Electronic structure:basic theory and practical methods. Cambridge University Press: 2004.

[5] Herring, C., Phys. Rev. A New Method for Calculating Wave Functions in Crystals, 1940, 57, 1169- 1177.

[6] Callaway, J., Phys. Rev. Orthogonalized Plane Wave Method, 1955, 97, 933-936.

[7] Hellman, H., J. Chem. Phys. New approximation method in the problem of many electrons, 1935, 3, 61.

[8] Antoncik, E., Czech. J. Phys. A new formation of the method of nearly free electrons, 1954, 439.

[9] Antoncik, E., Phys. Chem. Solids Approximate formulation of the orthogonalized plane-wave method, 1959, 10, 314-20.

[10] Phillips, J. C.; Kleinman, L., Phys. Rev. New method for calculating wave functions in crystals and molecules, 1959, 116, 287-94.

[11] Cohen, M. L.; Bergstresser, T. K., Phys. Rev. Band structures and pseudopotential form factors for fourteen semiconductors of the diamond and zinc-blende structures, 1966, 141, 789-96.

[12] Hamann, D. R.; Schlüter, M.; Chiang, C., Phys. Rev. Lett. Norm-Conserving Pseudopotentials, 1979, 43, 1494-1497.

[13] Kresse, G.; Hafner, J., J. Phys.: Condens. Matter Norm-conserving and ultrasoft pseudopotentials for first-row and transition elements, 1994, 6, 8245-57.

[14] Vanderbilt, Phys Rev B Condens Matter Soft self-consistent pseudopotentials in a generalized eigenvalue formalism, 1990, 41, 7892-7895.

[15] Blochl, Phys Rev B Condens Matter Generalized separable potentials for electronic-structure calculations, 1990, 41, 5414-5416.

[16] Chadi, D. J.; Cohen, M. L., Phys. Rev. B Electronic structure of mercury cadmium telluride (Hg1- xCdxTe) alloys and charge-density calculations using representative k points, 1973, 7, 692-9.

[17] Monkhorst, H. J.; Pack, J. D., Phys. Rev. B Special points for Brillouin-zone integrations, 1976, 13, 5188-5192.

[18] Hohenberg, P.; Kohn, W., Phys. Rev. Inhomogeneous Electron Gas, 1964, 136, B864-B871.

[19] Kohn, W.; Sham, L. J., Phys. Rev. Self-Consistent Equations Including Exchange and Correlation Effects, 1965, 140, A1133-A1138.

[20] Perdew, J. P.; Burke, K.; Ernzerhof, M., Phys. Rev. Lett. Generalized Gradient Approximation Made Simple, 1996, 77, 3865-3868.

[21] Giannozzi, P.; Baroni, S.; Bonini, N.; Calandra, M.; Car, R.; Cavazzoni, C.; Ceresoli, D.; Chiarotti, G. L.; Cococcioni, M.; Dabo, I.; Dal Corso, A.; de Gironcoli, S.; Fabris, S.; Fratesi, G.; Gebauer, R.; Gerstmann, U.; Gougoussis, C.; Kokalj, A.; Lazzeri, M.; Martin-Samos, L.; Marzari, N.; Mauri, F.; Mazzarello, R.; Paolini, S.; Pasquarello, A.; Paulatto, L.; Sbraccia, C.; Scandolo, S.; Sclauzero, G.; Seitsonen, A. P.; Smogunov, A.; Umari, P.; Wentzcovitch, R. M., Journal of physics. Condensed matter : an Institute of Physics journal QUANTUM ESPRESSO: a modular and open-source software project for quantum simulations of materials, 2009, 21, 395502.

[22] Baroni, S.; de Gironcoli, S.; Dal Corso, A.; Giannozzi, P., Rev. Mod. Phys. Phonons and related crystal properties from density-functional perturbation theory, 2001, 73, 515-562.

[23] Vanderbilt, D., Phys. Rev. B Soft self-consistent pseudopotentials in a generalized eigenvalue formalism, 1990, 41, 7892-7895.

Chapter 3 Synthesis, Structure and Decomposition of Metal

Hydrides and Borohydrides

3.1 Introduction

This chapter is divided into three different sections. The first section deals with the synthesis of

EuH2 from europium metal. The second section presents a way to synthesize halide- free strontium, barium and europium borohydrides (M(BH4)2) starting with their respective hydrides as precursors. A detailed study of the decomposition pathways of these borohydrides was made and tentative decomposition reactions for all three borohydrides are proposed. The last section deals with the synthesis of several borohydrides where the pure product wasn’t obtained.

The synthesis of metal borohydrides is either carried out by mechano-chemical processes such as ball milling or by wet chemistry1. During ball milling, the reaction conditions can be tuned by controlling the factors like milling time, breaks during the milling, milling frequency, vial and ball composition and their size, powder to ball weight ratio, milling atmosphere and pressure of the gas selected for milling2. Although ball milling appears to be a simple technique to produce a large variety of borohydrides, undesired products may also be formed as a result of various 2-4 competing reactions occurring simultaneously . For eg, LiZn2(BH4)5 is formed rather than

Zn(BH4)2 when LiBH4 is milled with ZnCl2, the presence of chloride impurities may form mixed 5-8 anion compounds such as the solid solution of LiBH4 with other lithium halides .

Exchange reactions in solution can be one of the approaches to synthesize borohydrides. In general Na/Li borohydride is reacted with the metal halide to form metal borohydride with sodium/lithium chloride as a side product. For example, NaBH4 reacts with LiX (X=Cl,Br,I) in 9 isopropylamine to give LiBH4 . LiBH4 and MgCl2 mixtures in diethyl ether give Mg(BH4)2,

LiBH4 with MnCl2 in toluene-dimethylsulfide gives Mn(BH4)2, LiBH4 with AlCl3 in toluene 10-13 forms Al(BH4)3 and LiCl is side product in each case .

Monometallic borohydrides may be used as precursors for synthesizing bi- or tri-metallic borohydrides. For example, Sr(BH4)2 and Eu(BH4)2 are used for heterovalent doping in recently + discovered garnet-borohydrides, Li conductors and Eu(BH4)2 is of interest for luminescent perovskites-borohydrides.14-15 They are often synthesized via halide containing synthetic routes. ------Furthermore, anion-substitution BH4 ↔ X (X = Cl , Br . I ) can readily take place in borohydrides. The presence of halide ions changes the crystal symmetry and physical properties like melting point and hydrogen release temperature, as well as ionic conduction properties.16-21

- Fluoride ions are much smaller than BH4 . Mixing MBF4 with MBH4 can result at higher - temperatures in the formation of BHxF(4-x) ions before undergoing further decomposition reactions. This again changes the chemical and physical properties of borohydrides due to change in bond strength arising from substitution of H- with highly electronegative F- ion.22-23

24-25 Several synthesis schemes of Eu(BH4)2 were reported recently. Humphries et al. synthesized

Eu(BH4)2 from EuCl3 and LiBH4 in diethyl ether, the product was washed with dimethyl sulfide 25 to remove LiCl. The crystal structure and synthesis of Sr(BH4)2 via metathesis reaction between 26 LiBH4 and SrCl2 was also reported recently. Similarly, Ba(BH4)2 was prepared by a chloride 27 containing synthetic route. While α-PbO2 type polymorph has been reported as ambient structure for Sr(BH4)2 and Eu(BH4)2, two polymorphs isostructural to their chloride analogues 27 have been reported for Ba(BH4)2.

Crystal structures were characterized by the crystallography laboratory of the Department of Quantum Matter Physics of the University of Geneva, the contribution of Prof. R. Černý and PhD student E. Didelot is greatly acknowledged.

3.2 Synthesis and Discussion

3.2.1a Synthesis of EuH2

Commercially purchased europium ingots were introduced into an autoclave in an inert atmosphere of argon in a glove box. The autoclave was brought out of the glove box and was evacuated. Hydrogen was introduced into the autoclave, and then the autoclave was placed into a pre-heated furnace at 330 °C. The pressure of hydrogen was adjusted to 60 bars at 330 °C, and the europium sample was kept under these conditions for 5 days. After this treatment, the europium ingots did not pulverize and pure EuH2 was not obtained. These ingots were then first crushed using mortar and pestle and further pulverized using a ball mill. 1 gram of this powder was again kept in an autoclave under a hydrogen pressure of 60 bars at 330 °C for 3 days (a drop in hydrogen pressure was observed during this second heating, indicating the reaction of the metal with the hydrogen gas).

3.2.1b Results and Discussion

The synthesis was monitored using IR spectra at all the steps. The IR spectrum of the final product (EuH2) is shown in figure 3.1. This spectrum agrees with the previously reported 28 spectrum of EuH2.

Figure 3.1: IR spectra to monitor the progress of reaction for the EuH2 synthesis. The red curve shows the IR spectrum of the sample after 8 days of hydrogenation revealing a peak around 1000 cm-1 which matches with the previously reported IR spectrum of EuH2.

3.2.2a Synthesis of M(BH4)2 (M=Sr, Ba, Eu)

The finally successful synthesis scheme is similar to the synthesis of Mg(BH4)2 reported by 29 Chlopek et al. MH2 was activated by ball milling under inert conditions (Ar atmosphere, 12 cycles, 600 rpm for 6 min each with a pause of 2 min). Freshly ball milled MH2 was mixed in excess (of the stoichiometric amount, Table 3.1) of tri-ethyl amine borane. This mixture was heated overnight under reflux at a temperature of 100 °C and then at a temperature of 145 °C for 5 hours.

MH2 + 2 (C2H5)3N-BH3  M(BH4)2 + 2 N(C2H5)3 (3.1)

In the case of EuH2, a change in sample colour from dark red to white was observed. The mixture was cooled down to room temperature and was washed with cyclohexane. The residue was dried under dynamic vacuum up to the temperature of 100 °C for few hours. The solid product was periodically analysed with IR spectroscopy to monitor the presence of the residual solvent. All the starting materials were commercially available. EuH2 used for the preparation of

Eu(BH4)2 was kindly provided by Dr. Holger Kohlmann.

Table 3.1 Sample prepared by reaction of pre-milled hydride with tri-ethyl amine borane. Sample Hydride Tri-ethylamine Product gram mmol borane

S1 SrH2, 0.5 5.5 1.7 ml, 11.5 mmol Sr(BH4)2

S2 BaH2, 0.5 3.6 1.1 ml, 7.5 mmol Ba(BH4)2

S3 EuH2, 0.6 3.9 1.2 ml, 8.1mmol Eu(BH4)2

3.2.2b Results and Discussions

The purity of the final products was confirmed using IR spectra. Figure 3.2 shows IR spectra for

M(BH4)2 (M=Eu, Sr or Ba). These spectra confirm the complete absence of any organic solvent or water.

B-H B-H bending stretching

C-H H O stretching 2

Figure 3.2: IR spectra of freshly synthesized Sr(BH4)2 , Ba(BH4)2 and Eu(BH4)2. The range for O-H stretching in water and C-H stretching in organic solvents is highlighted. The absence of C-H stretching and O-H stretching bands confirms that all these samples are properly dried and are free from any organic solvent as impurity.

In order to probe the thermal stability of these samples, in situ IR spectroscopy was performed along with heating (Figure 3.3). The samples were heated up to a temperature of 280 °C and the IR spectra did not show any evidence of thermal decomposition. The thermal stability of these samples was also confirmed by X-ray diffraction and differential scanning calorimetry.

B-H bending B-H stretching

Figure 3.3: IR spectra of Eu(BH4)2 (upper), Sr(BH4)2 (middle), Ba(BH4)2 (bottom) from room temperature to 280 °C. The B-H stretching and bending frequency regions are highlighted. The absence of any new peaks on heating or no significant deformation of the peaks observed at 25 °C indicates that these samples are thermally stable up to 280 °C.

- 30 The IR spectra shown in figure 3.3 are typical for isolated BH4 ions in the solid. No strong - 4, 8, 18 splitting of the B-H stretching mode as observed for Sc(BH4)4 is found. Some weak additional bands below 1000 cm-1and underlying B-H deformation modes from 1000-1400 cm-1 are probably related to borate species present as impurities.

As discussed later in section 3.2.3a Ba(BH4)2 washed and recrystallized with water was also stable. It was observed that Ba(BH4)2 can be recrystallized with trace amounts of water as shown -1 in figure 3.4. O-H and N-H stretching bands are observed above 3000 cm , and the NH2 or OH2

deformation bands around 1598 cm-1 and 1630 cm-1 respectively. There is a possibility of partial hydrolysis leading to a formation of borate, which brings the difference in spectra close to 1000 wavenumbers.

B-H B-H bending stretching

C-H N-H stretching stretching H2O

Figure 3.4: IR spectra of Ba(BH4)2 recrystallized from water (bottom curve), middle curve illustrates IR spectra of ethanolamine and top curve show Ba(BH4)2 with ethanolamine. The frequency range of O-H,

N-H, C-H, B-H stretching and B-H bending is highlighted. Ba(BH4)2 could be recrystallized from water without being decomposed.

SR-XPD (Rietveld plot for all the samples are shown below in figures 3.5-3.13.) reveals that as synthesized Eu(BH4)2 and Sr(BH4)2 exist in a single crystalline phase identified as o-Eu(BH4)2, o-Sr(BH4)2. On heating o-Eu(BH4)2 transforms to t-Eu(BH4)2 at 395 °C and o-Sr(BH4)2 transforms to t-Sr(BH4)2 at 450 °C (o = orthorhombic, t = tetragonal). Chloride containing

Eu(BH4)2 and Sr(BH4)2 were studied by B. C. Hauback & coworkers and T. R. Jensen & co- workers and both the compounds had lower thermal stability compared to their pure phase.25-26 Similar results of lowering of decomposition temperature are obtained on mixing chloride (or 31 other halides) with borohydrides. t-Eu(BH4)2 melts at 425 °C while diffraction peaks for t-

Sr(BH4)2 disappears at 490 °C. A trigonal borohydride (P-3m1) borohydride-hydride M2(BH4)H3 was observed above 395 °C for Eu(BH4)2, and above 410 °C for Sr(BH4)2.

2 Figure 3.5: Rietveld plot for sample S3, o-Eu(BH4)2 at 235 °C (Rwp = 5.2459 %, χ = 2472.4431).

2 Figure 3.6: Rietveld plot for sample S3 at 395 °C (Rwp = 4.5432 %, χ = 1859.4402). Regions with peaks of unidentified phases were excluded from the refinement. Several regions in the observed diffractogram contained unidentified impurity peaks. These regions were excluded for the refinement and highlighted in the grey.

2 Figure 3.7: Rietveld plot for sample S3, Eu2(BH4)H3 at 425 °C (Rwp = 1.9956 %, χ = 293.6898). Regions with peaks of unidentified phases were excluded from the refinement. Several regions in the observed diffractogram contained unidentified impurity peaks. These regions were excluded for the refinement and highlighted in the grey.

2 Figure 3.8 Rietveld Plot for sample S1, o-Sr(BH4)2 at 70 °C (Rwp = 10.8206 %, χ = 5361.9126). Regions with peaks of unidentified phases were excluded from the refinement.

2 Figure 3.9: Rietveld plot for sample S1 at 490 °C (Rwp = 5.2751 %, χ = 1503.9866). Regions with peaks of unidentified phases were excluded from the refinement. Several regions in the observed diffractogram contained unidentified impurity peaks. These regions were excluded for the refinement and highlighted in the grey.

2 Figure 3.10: Rietveld plot for sample S1, Sr2(BH4)H3 at 500 °C (Rwp = 6.8891 %, χ = 3237.8928). Regions with peaks of unidentified phases were excluded from the refinement. Several regions in the observed diffractogram contained unidentified impurity peaks. These regions were excluded for the refinement and highlighted in the grey.

2 Figure 3.11: Rietveld Plot for sample S2, o1-Ba(BH4)2 at 210 °C (Rwp = 5.8686 %, χ = 1047.3322).

2 Figure 3.12: Rietveld Plot for sample S2, o2-Ba(BH4)2 at 430 °C (Rwp = 4.3890 %, χ = 631.9717). Regions with peaks of unidentified phases were excluded from the refinement. Several regions in the observed diffractogram contained unidentified impurity peaks. These regions were excluded for the refinement and highlighted in the grey.

2 Figure 3.13: Rietveld Plot for sample S2, t-Ba(BH4)2 at 500 °C (Rwp = 5.3919 %, χ = 1840.1650). Regions with peaks of unidentified phases were excluded from the refinement. Several regions in the observed diffractogram contained unidentified impurity peaks. These regions were excluded for the refinement and highlighted in the grey.

The high thermal stability is confirmed by calorimetry. The DSC curves of europium and strontium borohydrides show a strong endothermic peak above ca 400°C on heating which corresponds to the decomposition of borohydrides in both samples (figure 3.14). The second endothermic peak, better resolved in Sr(BH4)2 corresponds to the decomposition of the borohydride-hydride in this sample and to the melting in Eu(BH4)2.

Figure 3.14: DSC curve for Sr(BH4)2,Ba(BH4)2 and Eu(BH4)2. Curves for Sr(BH4)2 (red) and Ba(BH4)2 (black) are shifted on y axis to clearly show the decomposition in each of the samples. DSC curves show that samples are stable up to 400 °C and show phase transitions at a temperature close to 400°C.

In order to confirm the space group of Ba(BH4)2, periodic DFT calculations were performed. Variable-cell and fixed-cell optimizations were done in the three candidate space groups, namely Immm, Imm2, and Pnnm. Phonon calculations were performed after structure optimization. For the first two settings, more than three imaginary frequencies were obtained, while none was obtained for Pnnm. Moreover Ba(BH4)2 in the Pnnm space group is found to be energetically more stable than the other two (Imm2 by 30.9 kJ/mol and Immm by 42.6 kJ/mol).

For Ba(BH4)2 in the Pnnm space group, two types of structural relaxation were carried out. First, the atomic positions were optimized with the cell parameters kept fixed to their experimental values. In the second type of optimisation, both the cell parameters and the atomic positions were relaxed (variable-cell optimisation). The relaxation of the cell parameters led to a 4.5% decrease in the volume, while the energy decreased by 4.6 kJ/mol. For the computation of IR and Raman spectra from theory, calculation using norm-conserving pseudopotentials was performed. Figure 3.15 compares experimental and theoretical results for both IR and Raman spectroscopy for the bending mode region between 1000 and 1400 cm-1. The stretching mode are subject to strong Fermi resonances as observed and analysed previously for the alkali borohydrides.32 The

complete experimental spectra (from 900 to 2700 cm-1) and a list of calculated frequencies are given in the supporting information (figure 3.16 and table ST3.1). The calculated frequencies of the bending modes are systematically lower than the experimental values, but the splitting is similar.

Figure 3.15: Experimental vibrational spectra (thicker curve, Raman on top and IR on bottom) of

Ba(BH4)2 compared with calculated vibrational spectra for the Pnnm structural model (thinner curve). Axis on the left shows the Raman intensity and axis on the right show IR intensity. The calculated bending frequencies are about 20-30 cm-1 lower than the experimental ones, however the splitting (indicative of local symmetry) are similar.33

Figure 3.16: Experimental IR and Raman spectra of Ba(BH4)2. The deformation bands are compared to the DFT calculated spectra in figure 3.15.

- In the Pnnm polymorph, the barium atom is surrounded by six BH4 units in an octahedral coordination. Each octahedron shares all six vertices and two edges building straight chains along the c-axis. Each chain share vertices with four other chains (figure 3.17). The distances Ba- - B are 3.09 and 3.18 Å from the DFT optimized structure. Each BH4 tetrahedron is surrounded by three Ba atoms in nearly planar triangular coordination.

Ba(BH4)2 synthesized with scheme 3.2.3b contains a new polymorph o1-Ba(BH4)2. On heating it transforms to o2-Ba(BH4)2 at 395 °C. This polymorph is isostructural to o-Eu(BH4)2 and o-

Sr(BH4)2. On further heating to 445 °C it transforms to t-Ba(BH4)2 and on heating, to 460 °C a cubic polymorph also appears. Ba(BH4)2 is stable up to 500 °C (maximum temperature of the

SR-XPD experiment). Ex-situ heating of Ba(BH4)2 to 750 °C lead to the formation of BaB6 along with two unidentified phases.

The DSC heating trace for Ba(BH4)2 (figure 3.14) presents two endothermic peaks around 400 °C and 450°C, in agreement with the structural phase transitions observed by X-ray diffraction.

Based on the above observations, the decomposition of europium and strontium borohydride can be tentatively described as:

2 M(BH4)2 M2(BH4)H3 + B3H9 (M = Sr or Eu) (3.2)

The B3H9 decomposition product can be understood as a higher borane anion like B12H12 and released hydrogen or as a diborane B2H6. As the ratio M/B is higher in the borohydride-hydride

M2(BH4)H3, it is not probable that metal closoboranes or other salts containing higher boranes anions are formed. It seems likely that diborane is formed during the thermal decomposition followed by the decomposition of diborane at temperatures higher than ~300 °C.

Figure 3.17: View of the refined structure of o1-Ba(BH4)2, Pnnm, CaCl2 structure type, a) along the c- axis b) octahedra sharing corners and edges. Each barium atom is surrounded by six borohydride tetrahedrons.

Similarly, as the B/Ba ratio in BaB6 is higher than in Ba(BH4)2, it seems unlikely that boron would be present in other decomposition products. It is possible that during our experiments the metallic Ba reacted with the container to yield the unknown phases observed.

3 Ba(BH4)2 BaB6 + 12 H2 + 2 Ba, or (3.3)

3 Ba(BH4)2 BaB6 + 2 BaH2 + 10 H2 (3.4)

Divalent Eu ions in solids find many applications for luminescent materials. As borohydrides are inherently reducing materials, one does not expect to observe Eu3+ ions, as one can for instance 2+ in Eu-doped oxides. The luminescence of Eu in the perovskites CsEu(BH4)3 and Eu-doped 8 CsCa(BH4)3 was observed at 485 nm at room temperature. The excitation and emission spectra 14 of Eu(BH4)2 is shown in figure 3.18 and was also reported previously. There is a single emission band at 463 nm (FWHM 55 nm), corresponding to the CIE colour coordinates of x=

0.145, y= 0.134 (figure 3.19). The Stokes shift (0.25 eV) is typical for a normal f-d transition of 2+ 34 Eu . Thus the emission can already be excited at 400-410 nm, which could make Eu(BH4)2 or

Eu-doped Sr(BH4)2 useful blue phosphors for trichromatic light sources.

Figure 3.18: Photoluminescence spectra of Eu(BH4)2. Blue curve show the excitation spectrum monitored at 463 nm and red curve shows the emission spectra excited at 350 nm. Eu(BH4)2 can be excited even at 400-410 nm.

Figure 3.19: CIE colour co-ordinate diagram. The black ring marks the color emitted by Eu(BH4)2.

3.2.3 Alternative Synthesis Methods

3.2.3 a, b & c Synthesis of (BaBH4)2 a) Ba(BH4)2: In order to synthesize solvent free Ba(BH4)2, the product obtained after the reaction of BaH2 with tri-ethyl amine borane was divided into two parts. The first part was subjected to dynamic heating which gave pure Ba(BH4)2 as discussed above. The second part was dissolved in water to remove the (co-crystallized) organic solvent. It was observed that

Ba(BH4)2 is stable in water and can be recrystallized to form Ba(BH4)2 with trace amounts of water. b) Ba(BH4)2: BaCl2 is reacted with NaBH4 in ethanolamine (or monoethanol amine, MEA) as a solvent to form Ba(BH4)2 according to the following reaction.

ethanolamine BaCl2 + 2NaBH4 Ba(BH4)2 + 2NaCl (3.5)

A mixture of 0.189 g of NaBH4 and 3 mL of MEA was made. Another mixture of 0.520 g BaCl2 and 1.5 mL MEA was made. Both mixtures were stirred separately (for 3 hours) until clear solutions were obtained. These solutions were combined and stirred overnight at room temperature. The solution was filtered and the filtrate was dried over vacuum for 24 hours at a temperature of 60 °C. c) Ba(BH4)2: A 1:2 mixture of BaCl2 and NaBH4 was ball milled. IR analysis of the product confirmed a reaction as the spectra did not match the initial spectra of NaBH4. But the best results were obtained with a mixture of the ratio (BaCl2 : NaBH4) 1:1 and 2:1. Replacing NaBH4 with LiBH4 in above mentioned experiments resulted in sharper peaks for the product.

Results and Observations

Figure 3.4 illustrates that Ba(BH4)2 can be recrystallized from water. Metal borohydrides typically decompose in water with the evolution of hydrogen gas and the formation of metal borate. The IR absorption around 1600 cm-1 in Figure 3.4 signifies the presence of trace amounts of water. Ba(BH4)2 synthesized by the scheme presented in section 3.2.3b, contains ethanolamine. On vacuum drying, ethanolamine couldn’t be removed thus pure Ba(BH4)2 wasn’t obtained by this reaction procedure.

Reaction scheme 3.2.3c is a metathesis reaction between alkali metal borohydride and BaCl2 via ball milling. The expected products were Ba(BH4)2 and alkali metal (Na, Li) chloride.

Laboratory X-ray powder diffraction data of the sample prepared with LiBH4 showed diffraction peaks compatible with the room temperature BaCl2 structure, but no estimate of the potential chloride content could be obtained. This reaction probably did not reach completion.

3.2.3 d, e & f Synthesis of Sr(BH4)2 d) Sr(BH4)2: Synthesis of Sr(BH4)2 was attempted starting with a mixture of Sr(iOPr)2 with borane tetrahydrofuran according to the following reaction:

. Sr(iOPr)2 + 2 BH3 THF Sr(BH4)2 + 2 iPrOH (3.6)

A suspension of Sr(iOPr)2 (0.8g) in 15mL of dry THF was heated to reflux with stirring (under

N2). While stirring under reflux was continued, 50 mL of a 1M solution of borane in THF was added slowly in a period of 3h. The solution was filtered to give a white powder whose IR spectra showed characteristic vibrational stretching of B-H bond around 2300 cm-1.

e) Sr(BH4)2: A mixture of 0.79g (0.005mol) of SrCl2 and 0.21g (0.01mol) of LiBH4 was ball milled using the following ball milling parameters: 10min milling, 5min break, 600 rotations per minute, 30 times.

f) Sr(BH4)2: SrCl2 and NaBH4 were dissolved in EtOH/MeOH at ice cold temperature. The addition of NaBH4 even to a cold solution of SrCl2 evolves H2 thus decomposing BH4. Ball milling of SrH2 and NaBH4 did not show any reaction. The result was same when SrCl2 was ball milled with NaBH4. But the reaction of SrCl2 and NaBH4 in ethanolamine or N, N- dimethylformamide yields solvated Sr(BH4)2. On heating between 100-150°C under dynamical vacuum, the product starts to decompose and hence no pure Sr(BH4)2 is obtained even by this method.

Results and Observations

Scheme 3.2.3d presents a reaction scheme for synthesizing Sr(BH4)2 from Sr(iOPr)2. The product obtained was inspected by IR spectroscopy. Figure 3.20 shows the IR spectrum of this white powder in comparison with IR spectra of Sr(iOPr)2 and dry THF. In our attempt to remove THF

from the powder, it was heated up to 200 °C which resulted in decomposition of the borohydride, as no sign of B--H stretching was found in IR spectra after this heating.

Sr(BH4)2 Sr(iOPr)2

THF Intensity (a.u)

1000 1500 2000 2500 3000 3500 4000 -1 Wavenumber (cm )

Figure 3.20: Comparison of IR spectra of Sr(BH4)2 containing THF with Sr(iOPr)2 and THF. Sr(BH4)2 (red curve) show the presence of THF. Upon vacuum drying to remove the THF solvent we observed that

Sr(BH4)2 was decomposed.

Scheme 3.2.3e and 3.2.3f presents a metathesis reaction between SrCl2 and LiBH4 via ball milling. After the ball milling, an IR spectrum of the product is measured showing the formation -1 of a new band between 1000 and 1300cm which could correspond to Sr(BH4)2. Unreacted

LiBH4 was also present with Sr(BH4)2.

3.3 Conclusions

Successful synthesis schemes for EuH2 and halide free strontium, barium and europium borohydride are reported. EuH2 can be used as starting material for future luminescence studies. The halide free compounds show higher thermal stability than previously reported samples obtained from chloride based syntheses. The thermal decomposition route of all three borohydrides is rather complex involving unidentified phases but leads in the case of M = Sr and

M = Eu to M2(BH4)H3, a first compound containing hydride in two anions, as a simple hydride and as a complex borohydride. It is likely that other compounds in the family of M2(BH4)H3 can be prepared and studied. It will be interesting to see to which extent their formation could take place during the dehydrogenation reaction or whether they can be reversibly converted back to

M(BH4)2. This work is currently under investigation in our laboratory.

Several partially successful schemes to synthesize Sr(BH4)2 and Ba(BH4)2 are also reported. Metathesis between metal halide and metal borohydride via ball milling provides an easy way for the preparation of certain metal borohydrides, but at the same time, it gives a challenge of purification (removal of metal halide). If the solvent is used then removal of the solvent is a crucial step too. Sr(BH4)2 thermally decomposed during removal of iso-propyl from Sr(BH4)2.

Metal borohydride hydrides are reported as a first decomposition intermediate in europium and strontium borohydride. The crystal structures of several polymorphs are characterized and put into a context of crystal chemistry of homoleptic metal borohydrides.

3.4 Supporting Information

Table ST3.1: Vibrational active modes in Pnnm structural model of Ba(BH4)2 obtained from GGA and LDA calculations. All internal modes are subject to factor group splittings due to the presence of 4 molecules in the elementary cell, thus each internal motion is split into 4 components. This is schematically visualized in the table alternating between bold and normal numbers. Wavenumber (cm-1) Comments GGA & Ultrasoft LDA & Norm-conserving Pseudopotentials pseudopotentials -3.82 -0.54 3 Acoustic modes -1.50 -0.36 (formally with zero -1.21 -0.12 frequency)

44.53 60.55 Optical lattice modes 48.17 66.93 + Librations 51.50 67.89 70.70 99.86 72.79 133.29 85.48 145.77 109.67 173.48 130.06 178.10 149.89 178.48 157.11 199.61 157.46 200.37 158.04 203.61 169.43 208.12 174.15 208.75 180.83 219.09 195.49 221.54 207.84 233.61 209.48 236.39 211.57 240.70 249.69 293.25 275.01 299.96 277.51 305.82 323.36 329.87 345.17 398.70 355.99 409.73 357.75 410.81 372.06 421.25 1043.88 1003.72 υ4 (T2) 1047.35 1009.91 1051.00 1012.12 1053.18 1016.62 1082.10 1059.07 1084.57 1059.71

1085.79 1063.05 1092.66 1071.54 1093.23 1071.86 1096.44 1071.89 1100.33 1071.93 1102.50 1077.77 1212.27 1191.43 υ2 (E) 1213.53 1192.51 1221.96 1199.66 1222.08 1201.27 1225.83 1210.95 1228.57 1214.11 1235.42 1220.00 1240.56 1225.40

2323.23 2280.40 υ1 (A1) 2326.16 2283.00 2327.67 2283.40 2330.26 2287.65 2371.51 2348.91 υ3 (T2) 2372.49 2349.63 2376.28 2351.61 2384.91 2355.02 2385.35 2365.59 2385.84 2366.39 2386.65 2366.91 2386.79 2367.17 2386.88 2367.66 2394.15 2368.29 2402.51 2377.68 2403.24 2380.15 - 35 The internal modes of the tetrahedral BH4 are labelled according to the book by K. Nakamoto .

3.5 References

[1] Hagemann, H.; Cerny, R., Dalton Trans. Synthetic approaches to inorganic borohydrides, 2010, 39, 6006-6012.

[2] Huot, J.; Ravnsbaek, D. B.; Zhang, J.; Cuevas, F.; Latroche, M.; Jensen, T. R., Prog. Mater. Sci. Mechanochemical synthesis of hydrogen storage materials, 2013, 58, 30-75.

[3] Jaron, T.; Orlowski, P. A.; Wegner, W.; Fijalkowski, K. J.; Leszczynski, P. J.; Grochala, W., Angew. Chem., Int. Ed. Hydrogen Storage Materials: Room-Temperature Wet-Chemistry Approach toward Mixed-Metal Borohydrides, 2015, 54, 1236-1239.

[4] Hagemann, H.; Longhini, M.; Kaminski, J. W.; Wesolowski, T. A.; Cerny, R.; Penin, N.; Sørby, M. H.; Hauback, B. C.; Severa, G.; Jensen, C. M., J. Phys. Chem. A LiSc(BH4)4: A Novel Salt of Li+ and Discrete Sc(BH4)4− Complex Anions, 2008, 112, 7551-7555.

[5] Ravnsbaek, D.; Filinchuk, Y.; Cerenius, Y.; Jakobsen, H. J.; Besenbacher, F.; Skibsted, J.; Jensen, T. R., Angew. Chem., Int. Ed. A Series of Mixed-Metal Borohydrides, 2009, 48, 6659-6663, S6659/1- S6659/9.

[6] Ravnsbaek, D. B.; Soerensen, L. H.; Filinchuk, Y.; Reed, D.; Book, D.; Jakobsen, H. J.; Besenbacher, F.; Skibsted, J.; Jensen, T. R., Eur. J. Inorg. Chem. Mixed-Anion and Mixed-Cation Borohydride KZn(BH4)Cl2: Synthesis, Structure and Thermal Decomposition, 2010, 1608-1612.

[7] Arnbjerg, L. M.; Ravnsbaek, D. B.; Filinchuk, Y.; Vang, R. T.; Cerenius, Y.; Besenbacher, F.; Jorgensen, J.-E.; Jakobsen, H. J.; Jensen, T. R., Chem. Mater. Structure and Dynamics for LiBH4-LiCl Solid Solutions, 2009, 21, 5772-5782.

[8] Cerny, R.; Severa, G.; Ravnsbæk, D. B.; Filinchuk, Y.; D'Anna, V.; Hagemann, H.; Haase, D.; Jensen, C. M.; Jensen, T. R., J. Phys. Chem. C NaSc(BH4)4: A Novel Scandium-Based Borohydride, 2010, 114, 1357-1364.

[9] Brown, H. C.; Choi, Y. M.; Narasimhan, S., Inorg. Chem. Addition compounds of alkali metal hydrides. 22. Convenient procedures for the preparation of lithium borohydride from sodium borohydride and borane-dimethyl sulfide in simple ether solvents, 1982, 21, 3657-61.

[10] Soloveichik, G. L.; Andrus, M.; Gao, Y.; Zhao, J. C.; Kniajanski, S., Int. J. Hydrogen Energy Magnesium borohydride as a hydrogen storage material: Synthesis of unsolvated Mg(BH4)2, 2009, 34, 2144-2152.

[11] Tumanov, N. A.; Safin, D. A.; Richter, B.; Lodziana, Z.; Jensen, T. R.; Garcia, Y.; Filinchuk, Y., Dalton Trans. Challenges in the synthetic routes to Mn(BH4)2: insight into intermediate compounds, 2015, 44, 6571-6580.

[12] Richter, B.; Ravnsbaek, D. B.; Tumanov, N.; Filinchuk, Y.; Jensen, T. R., Dalton Trans. Manganese borohydride; synthesis and characterization, 2015, 44, 3988-3996.

[13] Lascola, R.; Knight, D. A.; Mohtadi, R.; Sivasubramanian, P.; Zidan, R., Int. J. Hydrogen Energy Synthesis and structural characterization of stabilized aluminum borohydride adducts with triethylenediamine, 2013, 38, 13368-13380.

[14] Schouwink, P.; Ley, M. B.; Tissot, A.; Hagemann, H.; Jensen, T. R.; Smrcok, L.; Cerny, R., Nat. Commun. Structure and properties of complex hydride perovskite materials, 2014, 5, 5706.

[15] Brighi, M.; Schouwink, P.; Sadikin, Y.; Černý, R., J. Alloys Compd. Fast ion conduction in garnet- type metal borohydrides Li3K3Ce2(BH4)12 and Li3K3La2(BH4)12, 2016, 662, 388-395.

[16] Lee, J. Y.; Lee, Y.-S.; Suh, J.-Y.; Shim, J.-H.; Cho, Y. W., J. Alloys Compd. Metal halide doped metal borohydrides for hydrogen storage: The case of Ca(BH4)2-CaX2 (X = F, Cl) mixture, 2010, 506, 721-727.

[17] Mosegaard, L.; Moller, B.; Jorgensen, J.-E.; Filinchuk, Y.; Cerenius, Y.; Hanson, J. C.; Dimasi, E.; Besenbacher, F.; Jensen, T. R., J. Phys. Chem. C Reactivity of LiBH4: In Situ Synchrotron Radiation Powder X-ray Diffraction Study, 2008, 112, 1299-1303.

[18] Cerny, R.; Ravnsbæk, D. B.; Severa, G.; Filinchuk, Y.; D'Anna, V.; Hagemann, H.; Haase, D.; Skibsted, J.; Jensen, C. M.; Jensen, T. R., J. Phys. Chem. C Structure and Characterization of KSc(BH4)4, 2010, 114, 19540-19549.

[19] Ravnsbaek, D. B.; Rude, L. H.; Jensen, T. R., J. Solid State Chem. Chloride substitution in sodium borohydride, 2011, 184, 1858-1866.

[20] Olsen, J. E.; Sorby, M. H.; Hauback, B. C., J. Alloys Compd. Chloride-substitution in sodium borohydride, 2011, 509, L228-L231.

[21] Grove, H.; Rude, L. H.; Jensen, T. R.; Corno, M.; Ugliengo, P.; Baricco, M.; Sorby, M. H.; Hauback, B. C., RSC Adv. Halide substitution in Ca(BH4)2, 2014, 4, 4736-4742.

[22] Corno, M.; Pinatel, E.; Ugliengo, P.; Baricco, M., J. Alloys Compd. A computational study on the effect of fluorine substitution in LiBH4, 2011, 509, S679-S683.

[23] Yin, L.; Wang, P.; Fang, Z.; Cheng, H., Chem. Phys. Lett. Thermodynamically tuning LiBH4 by fluorine anion doping for hydrogen storage: A density functional study, 2007, 450, 318-321.

[24] Olsen, J. E.; Frommen, C.; Jensen, T. R.; Riktor, M. D.; Sorby, M. H.; Hauback, B. C., RSC Adv. Structure and thermal properties of composites with RE-borohydrides (RE = La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Er, Yb or Lu) and LiBH4, 2014, 4, 1570-1582.

[25] Humphries, T. D.; Ley, M. B.; Frommen, C.; Munroe, K. T.; Jensen, T. R.; Hauback, B. C., J. Mater. Chem. A Crystal structure and in situ decomposition of Eu(BH4)2 and Sm(BH4)2, 2015, 3, 691- 698.

[26] Ravnsbaek, D. B.; Nickels, E. A.; Cerny, R.; Olesen, C. H.; David, W. I. F.; Edwards, P. P.; Filinchuk, Y.; Jensen, T. R., Inorg. Chem. Novel Alkali Earth Borohydride Sr(BH4)2 and Borohydride- Chloride Sr(BH4)Cl, 2013, 52, 10877-10885.

[27] Ravnsbaek, D. B., Ph.D. Thesis Synthesis, structure and properties of novel metal borohydrides, 2011.

[28] Saitoh, H.; Machida, A.; Matsuoka, T.; Aoki, K., Solid State Commun. Phase diagram of the Eu–H system at high temperatures and high hydrogen pressures, 2015, 205, 24-27.

[29] Chlopek, K.; Frommen, C.; Leon, A.; Zabara, O.; Fichtner, M., J. Mater. Chem. Synthesis and properties of magnesium tetrahydroborate, Mg(BH4)2, 2007, 17, 3496-3503.

[30] D’Anna, V.; Spyratou, A.; Sharma, M.; Hagemann, H., Spectrochim. Acta Mol.Biomol. Spectrosc. FTIR spectra of inorganic borohydrides, 2014, 128, 902-906.

[31] Rude, L.; Filso, U.; D'Anna, V.; Spyratou Stratmann, A.; Richter, B.; Hino, S.; Zavorotynska, O.; Baricco, M.; Sørby, M. H.; Hauback, B. C.; Hagemann, H.; Besenbacher, F.; Skibsted, J.; Jensen, T. R., Phys. Chem. Chem. Phys. Hydrogen-fluorine exchange in NaBH4-NaBF4, 2013, 15, 18185-18194.

[32] Carbonnière, P.; Hagemann, H., J. Phys. Chem. A Fermi Resonances of Borohydrides in a Crystalline Environment of Alkali Metals, 2006, 110, 9927-9933.

[33] D'Anna, V.; Lawson Daku, L. M.; Hagemann, H., J. Phys. Chem. C Quantitative Spectra-Structure Relations for Borohydrides, 2015, 119, 21868-21874.

[34] Dorenbos, P., J. Phys. Condens. Matter Anomalous luminescence of Eu2+ and Yb2+ in inorganic compounds, 2003, 15, 2645-2665.

[35] Nakamoto, K., Applications in Inorganic Chemistry. In Infrared and Raman Spectra of Inorganic and Coordination Compounds, John Wiley & Sons, Inc.: 2008; pp 149-354.

Chapter 4

Isotope Exchange Reactions

4.1 Introduction

This chapter is divided into two parts. In the first part, a detailed study of the deuteration of

Ca(BH4)2 via a solid-gas reaction is presented. The second part reports experiments on the deuteration of Na2B12H12 in solution. This first section of the chapter has been published in reference [1].1 The results of the second part are not well understood and the problem encountered in deciphering the results is presented.

Among the alkali metal and alkaline earth metal borohydrides, calcium borohydride (Ca(BH4)2) -1 2,3 is of prime interest due to its favourable dehydrogenation enthalpy (32 kJ mol ). Ca(BH4)2 has been subjected to various theoretical and experimental investigations to study its dehydrogenation pathway, the effect of catalysts on dehydrogenation and the reversibility 4-8 9 (rehydrogenation). Ca(BH4)2 contains 11.6 hydrogen by weight. For desorption of the hydrogen (H2) from Ca(BH4)2 breaking of the B-H bond is potentially one of the first steps involved in the dehydrogenation (and the last one during the reverse reaction) of borohydrides. A detailed kinetic and thermodynamic study of this process is required to improve the understanding of dehydrogenation. The deuteration of borohydrides is a process which involves breaking of B-H bonds and formation of B-D bonds without having to consider the chemical and structural changes associated with a thermal decomposition reaction. Since this process is also involved in dehydrogenation (and rehydrogenation) of borohydrides, thermodynamic and kinetic study of this step can aid in better understanding of dehydrogenation (and rehydrogenation) process.

- The isolated borohydride ion (BH4 ) has tetrahedral symmetry. Four normal vibrational modes are expected; out of which two modes are IR active (both are triply degenerate). The symmetry is lowered in crystalline lattice of alkali (or alkaline earth) metal borohydride owing to site- symmetry and crystal field effects.10-11 This causes degenerate levels to split and activation of - inactive vibrations. Partial deuterium substitution of hydrogen in BH4 ion further causes - - - - lowering of symmetry (BH4 (Td), BH3D (C3v), BH2D2 (C2v), BHD3 (C3v)). These changes can be monitored by vibrational spectroscopy as the vibrational frequency of molecular groups with H and D varies hugely because of the large mass difference between H and D.12-13 Thus IR spectroscopy was chosen to be the main technique to calculate hydrogen and deuterium content in a sample during deuteration (or rehydrogenation).

Isotope exchange in borohydrides have been studied in past for borohydrides of Li, Na, K and Mg (T ≥ 200, 350, 500, 132 °C respectively).14-17 We have reported the isotope exchange 17 between Mg(BH4)2 and D2 gas at 40 bar and found an activation energy of 50 kJ/mol. A recent 11 publication on Mg(BH4)2 have also concluded similar results. In this work, we study the reaction Ca(BH4)2 + D2  Ca(BD4)2 and the reverse reaction in detail. It is shown that Ca(BH4)2 can also be deuterated at a pressure as low as 1 bar. A detailed analysis of temperature and pressure dependence of the reaction is presented.

18-20 + Na2B12H12 was recently found to be a solid state Na-superionic conductor. It consists of Na 2- 2- ions and B12H12 polyanion. A detailed computational study of B12H12 polyanion is presented in chapter 6. It shows a dramatic high superionic conductivity (~0.1 S/cm) after an order-disorder phase transition at high temperature (~540 K).20 In this regard, it would be interesting to prepare specifically deuterated Na2B12DnH(12-n) to probe the structure in more detail. Na2B12H12 was deuterated by two methods, a) Solid gas reaction between Na2B12H12 and D2 gas b) Reaction between solutions Na2B12H12 in D2O and DCl in D2O. Complete deuteration in solution was reported in brief previously.21-23 The reported results state that lowering of pH increases the rate of deuteration. Thus we have performed solution state deuteration in acidic condition (DCl). All the chemicals used were purchased commercially and used without any treatment (Na2B12H12-

Katchem, DCl-Sigma Aldrich and D2O-Armar AG). The progress of the deuteration reaction is monitored by IR, in-situ IR, and in-situ NMR.

4.2 Experimental Section

4.2A. Deuterium exchange reactions in Ca(BH4)2 via solid gas reaction

Synthesis of Ca(BH4)2. Ca(BH4)2 was prepared by heating a commercial sample of -2 Ca(BH4)2.2THF progressively up to 130 °C under vacuum (5 x 10 mbar). X-Ray diffraction of the de-solvated material shows that the phase composition of the resulting powder is a mixture of

α, β, and γ phase of Ca(BH4)2 with 47.7(8), 45.6(7) and 6.7(5) mole % respectively. All experiments were performed with samples of this batch. After full deuteration, this composition changed to 30.03(5), 58(8) and 11.7(4) mole % respectively. This change is likely due to the temperature applied and the stabilization of the high-temperature β-phase.

The deuteration of solid samples was carried out in a sealed autoclave (figure 4.1). A vial

(ceramic crucible) containing 150 mg of Ca(BH4)2 was placed in an autoclave in an inert atmosphere of argon (in a glove-box). The autoclave was evacuated and deuterium gas was introduced into the autoclave. The deuterium pressure was monitored with a pressure gauge which is attached to the autoclave. The autoclave was introduced into the preheated furnace. Different experiments were performed by varying temperature at a constant pressure (termed as temperature dependent deuteration experiments) and by varying pressure at a constant temperature (termed as pressure dependent deuteration experiments).

The autoclave was removed from the furnace after regular intervals of time. Excess pressure was removed and it was allowed to cool down to room temperature. Small samples were taken out for IR characterization after moving the autoclave into the glove box. The experiment was continued by repeating the same procedure.

In one set of experiments (first set) the furnace temperature was kept constant at 200 °C and the pressure was varied (1, 5, 10 and 20 bar). In another set of experiments (second set), the pressure was kept constant at 30 bar and the temperature was varied (140, 170 and 200 °C). Both sets of experiments were repeated for Ca(BD4)2 which was formed by the complete deuteration of

Ca(BH4)2.

Depending upon temperature and pressure conditions 퐶푎(퐵퐻4)2 was either partially or completely deuterated. T,P Ca(BH4)2 + xD2 ↔ Ca(BH(4-x)Dx)2 + xH2 (1≤x≤4) (4.1)

Figure 4.1: Image of an autoclave for high-pressure solid gas reaction. A known amount of Ca(BH4)2 was put into the ceramic crucible inside the autoclave. The autoclave was tightly sealed with the help of screws and then evacuated. Deuterium/Hydrogen gas was inserted in the autoclave and the pressure of the gas was measured with the help of the barometer included in the setup.

4.2B. Deuterium exchange reactions in Na2B12H12 via solid gas reaction

A solid gas reaction in an autoclave was carried out using the method described above (for

CaBH4). 28 mg Na2B12H12 was kept in a ceramic crucible. Deuterium at 45 bar was introduced at room temperature and the autoclave was heated to 300 °C for 24 hours.

x T,P x Na B H + D → Na B H D + H (4.2) 2 12 12 2 2 2 12 (12 − x) x 2 2

4.2C. Deuterium exchange reactions in Na2B12H12 in DCl/D2O solution

Two solutions 1 M, Na2B12H12 in D2O and 10 M DCl, in D2O (commercially obtained 35 wt%

DCl/D2O) were made. Both solutions were mixed together in the proportions given in the table below (table 4.1) and the final volume of the solution(s) thus obtained was adjusted to 0.75 ml.

Table 4.1: Volume and concentration of different solutions used for deuteration of Na2B12H12 Final Concentration Vol. Na2B12H12 Vol. DCl Vol D2O Na2B12H12 DCl 0.19 ml 0.5 ml 0.06 ml 0.25 M 6.67 M

0.19 ml 0.25 ml 0.31 ml 0.25 M 3.33 M

0.19 ml 0.13 ml 0.42 ml 0.25 M 1.67 M

4.3 Results and Discussion

4.3A. Solid Gas Reaction in Ca(BH4)2

It is essential, to begin with a pure unsolvated borohydride in order to rule out any effects due to the solvent. Thus vacuum drying of the borohydride is a crucial step. Figure 4.2 shows different stages during drying of Ca(BH4)2 monitored by IR spectra. The sample contained water and THF as an impurity (solvent). THF could be removed by heating the sample to 130 °C at low pressure (~10-2 mbar), but due to the presence of water heating up directly to 130 °C would result in decomposition of borohydride. Thus it is essential to remove the water before heating to a higher temperature. As shown in figure 4.2, water is removed by heating the Ca(BH4)2 sample first to 50 °C for 1 hour and THF is removed in subsequent steps once the sample is free from water.

THF H2O H2O

Figure 4.2: IR spectra illustrating the different steps of Ca(BH4)2 drying. The initial sample contains THF and H2O as impurities. Upon heating the sample gradually under vacuum, water can be removed at a temperature of 50 °C, after which upon further heating the sample to 100 °C for a longer period of time the THF can be removed.

The powder patterns of initial and completely deuterated phases are shown in figure 4.3. The slight shift of the lattice parameters to higher angle is due to the relatively heavier deuterium

nucleus. Lattice parameters of deuterated samples are systematically smaller than those of their hydrogenated analogues.24-26 Importantly, these powder diffraction data demonstrate that the crystallinity is not affected by the isotope exchange; hence the kinetics of the reaction is expected to be comparable in both phases.

Figure 4.3: Powder diffraction data shown for initial Ca(BH4)2 and fully deuterated Ca(BD4)2. Inset: The smaller lattice of Ca(BD4)2 is manifested in a clear shift of Bragg peaks to higher angles.

During the solid gas reaction between Ca(BH4)2 powder and deuterium gas, IR spectra were measured at regular intervals of 3 hours to monitor the progress of the reaction. Figure 4.4 compares IR spectra of Ca(BH4)2 after 3, 6 and 9 hours of reaction at 1 bar pressure and 200 °C. The intensity at 1750 cm-1 (B-D stretch) is normalized to highlight the relative intensity change at 2330 cm-1 (B-H stretch). The band around 2330 cm-1 corresponds to the B-H stretching modes, while the band around 1750 cm-1 corresponds to the B-D stretching mode.27-28 Figure 4.5 18 compares the IR spectra of selectively labelled Ca(BH3D)2 and Ca(BD3H)2 to highlight the

spectral differences for different isotopomers. The wide separation of stretching frequency in these two cases helps in easy identification of each species. The relative intensities of these peaks can be related to the mole fraction of hydrogen and deuterium present in the sample17.

B-D B-H stretching stretching

Figure 4.4: IR spectra showing the B-H and B-D stretching region at a different time during Ca(BH4)2 deuteration at 1 bar and 200 °C. The spectra were normalized at 1700 cm-1 to highlight the relative intensity changes. The band centered around 2300 cm-1 corresponds to B-H stretching modes while the band centered around 1700 cm-1 corresponds to B-D stretch. The ratio of areas of both the bands gives an estimate of the mole fraction of deuterium and hydrogen present in the sample.

Database spectra27 Samples from Deuterium exchange experiments

Figure 4.5: Comparison of the IR spectra of Ca(BH3D)2, Ca(BHD3)2 and Ca(BD4)2 (from top to bottom, in the first column) taken from the FT-IR database of borohydrides 27 with IR spectra of experimental samples (second column) with increasing deuterium content.

For the first set of experiments, IR spectra were measured at different reaction times (3h, 6h, 9h) for different pressures (1, 5, 10, 20 bar). The mole fraction of hydrogen in the sample was calculated by taking the ratio of a relative intensity of B-H and B-D stretching bands. This mole ratio was used to plot the curves showing the progress of the reaction with time (figure 4.6). The rate constant of deuteration was obtained by comparing the curves with the first order rate -kt equation [H] = [H]0 e . Figure 4.6 reveals that as pressure was increased the rate constant of the reaction also increases. This behaviour is consistent with the reaction scheme in which an intermediate is in equilibrium with the reactants.

(4.3)

The pressure above which the increase in pressure wouldn’t affect the reaction rate significantly can be estimated by analysing the variation of the initial reaction speed estimated from the

exchanged amount after 3h. Variation of effective rate constant k was plotted (figure 4.6) against deuterium pressure P(D2).

1 1 K +K = + 2 −1 (4.4) K K2 K1K2P(D2)

Figure 4.6: Progress of the Ca(BH4)2 deuteration reactions as a function of time at 200 °C and different pressures. The reaction rate increases as the pressure is increased. It should be noted that deuteration starts at a pressure as low as 1 bar at 200 °C (deuteration at a pressure lower than 1 bar wasn’t investigated).

The variation of the initial rate constant (measured after 3h) with deuterium pressure is shown in figure 4.7. The solid red line corresponds to a fit of these data sets using the equation (1.1) with

1/K2 = 3.05, (K2 + K-1)/K1 = 2.70. It appears that for the pressures above ca 30-35 bar the rate constant does not increase anymore. In our previous study17 on the isotope exchange in

Mg(BH4)2, the reaction rate did not change when increasing the pressure from 40 to 80 bar, which is then consistent with the present observations.

Equation (4.4) implies a reversible first reaction step. In a typical gas –solid reaction, the initial step is the adsorption followed by further reactions (such as the dissociation of the hydrogen molecule before the diffusion into the solid). Assuming that the slow step is the dissociation reaction, the amount of hydrogen available for diffusion is dictated by the pressure dependent

adsorption equilibrium until the entire surface is covered with hydrogen at pressures about 30 bars. Alternate models could be proposed, such as adsorption followed by the diffusion of hydrogen molecules.

Figure 4.7: Variation of the effective rate constant with pressure. PLimiting refers to the maximum pressure after which further increase in pressure wouldn’t affect the reaction rate.

In the second set of experiments, the pressure was kept constant at 35 bar and the temperature was varied. The rate constants at different temperatures were calculated as mentioned above. The activation energy for this reaction was calculated (using the Arrhenius equation k = A e-E/RT) and found to be 82.1±2.7 kJ/mol as shown in figure 4.8. This value is, as expected, significantly lower than the activation energy for the first dehydrogenation step for Ca(BH4)2 at ca 370°C which was reported to be 225 kJ/mol29. I. Llamas-Jansa et al30 calculated the activation energy to be 184(14), 192(3), 230(1) kJ/mol respectively for γ, α and β polymorphs of Ca(BH4)2.

The same procedure was repeated for the reverse reaction and the corresponding activation energy is observed to be 98±8.4 kJ/mol. Figure 4.9 illustrates the IR spectra before and after the deuteration reaction as well as after the rehydrogenation reactions showing that this exchange reaction is fully reversible.

Figure 4.8: Arrhenius plot for the deuteration reaction at 35 bar and 130, 170 and 200 °C. As expected, the activation energy of this isotope exchange process is significantly lower than the activation energy of the dehydrogenation of Ca(BH4)2 leading to CaH2 and other species.

The difference of the values of the activation energy obtained for Mg(BH4)2 and Ca(BH4)2 is quite important. It is important to note in this context the difference of crystal structures, in the 31 3 sense that the structure of Mg(BH4)2 is relatively open (volume per formula unit ca 115 Å ) and 32,26 potentially offers easier diffusion paths for hydrogen (and deuterium) compared to Ca(BH4)2 which has a rather compact crystal structure (volume per formula unit less than 113 Å3). It must also be noted that the Ca2+ ions are much larger than the Mg2+ ions. Hydrogen adsorption (at low 33 temperatures has been observed in the γ phase of Mg(BH4)2 . It appears thus that the higher activation energy observed for Ca(BH4)2 may be related to a higher barrier for the diffusion of hydrogen in the crystal, as compared to Mg(BH4)2.

B-H stretching

B-D stretching

Figure 4.9: IR spectra showing complete conversion of Ca(BH4)2 to Ca(BD4)2 and complete rehydrogenation leading back to Ca(BH4)2 formation (200 °C, 20bar). Bands corresponding to B-H and B-D stretching are highlighted. The black curve is the initial sample which does not show any B-D stretch while after deuteration it converts to the green curve which does not show and B-H stretching, showing that all the hydrogen has been replaced by deuterium. Similarly, this deuterated sample was hydrogenated back to Ca(BH4)2 which does not show any remaining B-D stretching peak.

4.3 B. Deuteration by Solid Gas Reaction in Na2B12H12

The IR spectrum (figure 4.10) after 24 hours of reaction shows the presence of both B-H and B-

D bonds in the product obtained. Na2B12H12 was thus partially deuterated to Na2B12H(12-x)Dx and the deuterated mole fraction is estimated to be 0.55.

B-H stretching

B-D stretching

Figure 4.10: IR spectra of the product obtained after reaction of Na2B12H12 with D2. Band centred at 2500 cm-1 corresponds to B-H stretching and band centred at 1700 cm-1 corresponds to B-D stretching. The red curve is the IR spectra of starting material and green curve is the IR spectrum of the product obtained after reaction with deuterium gas in an autoclave. Band around 1000 cm-1 corresponds to ring breathing (B-B vibration mode).

The trace for Na2B12H(12 − x)Dx clearly shows the presence of characteristic B-H stretching -1 -1 band of Na2B12H12 at 2480 cm along with a newly band (B-D stretching) at 1850 cm . The B12 icosahedral cage breathing absorption band at 1075 cm-1 shifts to 925 cm-1 as observed by E.L Muetterties and co-workers.22 As this sample is partially deuterated thus this band is somewhat spread over a larger range with a statistical distribution of different isotopomers (D4, D5, and D6 etc.) Another interesting feature is that strong band at 715 cm-1 and the weaker band at 750 cm-1 -1 (in Na2B12H12) is not observed in the deuterated sample as this band shifts below 600 cm . A similar observation was reported in ref [22].22 These features could be observed clearly in the 2- 2- figure 4.11 where computationally calculated spectra of B12H12 and B12D12 are compared. B-H

and B-D stretching frequency could also be observed in Raman spectrum of the sample shown in figure 4.12.

B-B cage vibration

2- 2- Figure 4.11: Computationally calculated IR and Raman spectra of B12H12 and B12D12 . Top two curves are Raman spectra and bottom two curves are IR spectra (deuterated sample black curve). Pure B-H vibrational frequency changes by a factor of 0.707 on deuteration. Smaller separation refers to B-B vibrational motion or combination of B-B and B-H vibrational modes.

B-D stretching B-H stretching

Figure 4.12: Experimental Raman spectra of the product obtained after reaction of Na2B12H12 with D2. B- H, and B-D stretching modes are observed around 2500 cm-1 and 1800 cm-1 respectively.

For the further analysis, the deuterated sample was dissolved in D2O and its NMR spectrum was measured. The NMR spectra are shown in figure 4.13 illustrate the presence of both B-H and B-

D bonds. For Na2B12H12, a doublet is observed, corresponding to the coupling of each boron atom with one hydrogen atom. With deuterium, the coupling is about 6 times smaller and probably not resolved. One expects therefore to observe a superposition of a doublet and a broader singlet.

11 Figure 4.13: B NMR spectra of the product obtained after reaction of Na2B12H12 with D2 after dissolving the product in D2O. As the number of hydrogen atoms is reduced, the intensity of the doublet resulting from boron hydrogen coupling decreases and a triplet (resulting from B-D coupling) which is not resolved in this spectrum is added.

Deconvolution of the graph of Na2B12HnD(12-n) in three peaks (2 similar belonging to B-H coupling differing by 1.3 ppm and a peak of B-D coupling close to B-H coupling at the low field), also shows that B-H mole fraction is close to 0.51.

In order to investigate any further deuteration, a small amount of DCl was added to the sample in NMR tube. The NMR spectrum (figure 4.14) showed the clear shift towards low field B-H coupling peak indicating more deuteration.

Figure 4.14: 11B NMR spectra of the product obtained by dissolving the product of the reaction between

Na2B12H12 and D2 in DCl/D2O. As the time progresses, the intensity of the doublet due to boron hydrogen coupling decreases and indicates the progress of deuteration.

4.3 C. Deuterium exchange reactions in Na2B12H12 in DCl/D2O solution

Infra-Red measurements were used to monitor the progress of the reactions in all the three different solutions. In order to simplify the discussion of the IR spectra of Na2B12H12 in DCl/D2O solution, it is essential to know the IR spectra of DCl and D2O individually. Figure 4.15 shows the IR spectra of D2O, pure DCl (35 wt % in DCl) and Na2B12H12 in D2O. Figure 4.16 shows the progress of deuteration in 3.34 M DCl solution. With time, the band at 1850 cm-1 corresponding to B-D stretching increases in intensity which signifies an increase in deuterium/hydrogen ratio.

Similar graphs for 1.67 M and 6.67 M DCl solutions with 0.25 M Na2B12H12 are shown in figure 4.17 and 4.18.

Figure 4.15: Figure illustrating the IR spectra of D2O, pure DCl (35 wt % in DCl) and Na2B12H12 in D2O. 2- Comparison of the curves shows that it is difficult to analyse the B-H stretching of B12H12 ion in D2O as O-D stretching band is also observed in same frequency range (around 2500 cm-1). B-H bending vibrational band around 1050 cm-1 is observed only in highly concentrated solutions.

Once the B-D peak in IR spectra ceased to grow, a sample of each of these solutions was analysed by Raman and NMR spectroscopy (figure 4.19 and 4.20). In NMR spectra each of these samples shows only one peak (Sol A corresponds to 6.67 M DCl, Sol B to 3.34 M DCl and Sol C to

1.67 M DCl in D2O). Similar to the IR spectra, the Raman spectra also show both B-H and B-D stretching peaks.

Figure 4.16: IR spectra showing the progress of deuteration in Na2B12H12 in 3.34 M DCl/D2O solution. The band around 1250 cm-1 and 2500 cm-1 corresponds to O-D bending and stretching modes -1 -1 respectively. O-D stretching band at 2500 cm eclipses the B-H stretching of Na2B12H12 at 2470 cm .

Figure 4.17: IR spectra showing the progress of deuteration in Na2B12H12 in 6.67 M DCl/D2O solution.

Figure 4.18: IR spectra showing the progress of deuteration in Na2B12H12 in 1.67 M DCl/D2O solution.

Figure 4.19: 11B NMR spectra of all the three solutions discussed above. NMR spectrum confirms the complete deuteration of Na2B12H12 in all the three solutions (Sol A = 6.67 M DCl, Sol B =3.34 M DCl and Sol C= 1.67 M DCl in D2O).

Figure 4.20: Raman spectra of solutions of Na2B12H12 with different concentration of DCl.

A fraction of each of these solutions was dried under vacuum, and the corresponding powder was expected to show the completely deuterated product (according to NMR as NMR was taken after a time delay of 24 hours). But surprisingly the powder obtained from each of these 3 solutions show both B-H as well as B-D stretching bands in their IR spectra (figure 4.21). Solutions A and B were vacuum dried by heating the solutions to 100 °C while solution C was dried by evaporating D2O over time. Na2B12H12 is highly deliquescent and absorbs moisture from the air, thus a broad band around 3500 cm-1 can be observed.

The apparently contradictory indications from NMR and IR spectra couldn’t be explained. To rule out any possibility of chlorination, Na2B12H12 was dissolved in HCl /H2O solution for a period of 5 days. Initial and final NMR spectra did not show any changes. Thus any possibility of chlorination was ruled out. In order to probe the deuteration in more detail, in-situ NMR spectroscopy was used to monitor the deuteration. Figure 4.22 in supporting information show 11 the NMR spectra of Na2B12H12 in D2O and in DCl solution. On addition of acid B NMR peaks shift towards lower chemical shift

B-D stretching B-H stretching

Figure 4.21: IR spectra of the powder obtained by drying the three solutions after the confirmation of complete deuteration by NMR spectra (Sol A = 6.67 M DCl, Sol B =3.34 M DCl and Sol C= 1.67 M DCl in D2O).

Figure 4.22: NMR spectra of Na2B12H12 in pure D2O and DCl. With the increase in pH, peaks shift by 0.65 ppm.

11B NMR spectrum of the solution was taken every 20 minutes for two hours and every hour for next 60 hours. Figure 4.23 illustrates the time dependent boron NMR spectra. It is observed that the intensity of B-H doublet started to decrease while B-D peak (triplet, but can be seen as singlet because of a very narrow splitting range) started to emerge in between the doublet as time elapsed. After 16 hours only a single peak (expected to be corresponding to B-D) was left, which did not change during further measurements. These data alone suggest that after 16 hours deuteration was complete.

11 Figure 4.23: In-situ B NMR spectra of Na2B12H12 in DCl/D2O. As the time progresses, the intensity of the doublet due to boron hydrogen coupling decreases and indicates the progress of deuteration. After 1000 min (~16 hours) of reaction, all the hydrogen seems to be replaced by deuterium. In order to check for the possibility of chlorination Na2B12H12 was kept in a solution of HCl and no chlorination was observed.

A drop of solution was put on heatable ATR cell of FTIR and was heated to 100°C to remove all

D2O from Na2B12H12. IR spectra (figure 4.24) show however the presence of both B-D and B-H bonds.

Figure 4.24: IR spectra of the Na2B12H12 in DCl solution (Top, 25 °C). The solution was heated to 100 °C to dry and final spectra also show the presence of B-H and B-D bonds. As the temperature increases,

DCl/D2O evaporates and a powder is left behind. Broad O-D stretching band (also narrow B-D band but eclipsed due to O-D stretching band) observed around 2500 cm-1 keeps on decreasing with increase in temperature. At 100°C no broad O-D stretching is observed, the band at 2500 cm-1 at 100 °C can be assigned to B-H stretching.

4.4 Conclusions

The reversible isotopic exchange reaction between Ca(BH4)2 and D2 has been studied. Complete deuteration can be easily achieved. It is important to note that deuterated borohydrides can be interesting deuterium storage materials34 and also efficient neutron moderators. The pressure and temperature dependent study show a two-step reaction process.

The first step depends on the deuterium pressure, but at pressures above 35 bars, the rate constant does not change anymore. The activation energy estimated for the deuteration reaction is found to be around 82 kJ/mol, which is larger than the previously found value for Mg(BH4)2 of 51 ± 15 kJ/mol. The activation energy for the reverse reaction is 98 kJ/mol. It appears thus that the nature of the metal ion plays an important role in this reaction.

11 Deuteration of Na2B12H12 was monitored via in-situ B-NMR, IR and Raman spectroscopy. Some of the results obtained by NMR and IR appear to contradictory and we are still unable to interpret the reason of discrepancy in the results obtained by different methods. According to the vibrational spectra, only a partial isotopic replacement could be achieved so far. It must be 2- - further studied whether the protonation of B12H12 to form B12H13 leads to the observed features in the NMR spectra.

Further experiments are necessary to shed more light on the deuteration of sodium closo-borane. Insitu NMR or time dependent IR spectroscopy data can in principle be used to determine the kinetic parameters for the deuteration reaction in Na2B12H12.

4.5 References

[1] Sharma, M.; Sethio, D.; D'Anna, V.; Fallas, J. C.; Schouwink, P.; Cerny, R.; Hagemann, H., J. Phys. Chem. C Isotope Exchange Reactions in Ca(BH4)2, 2015, 119, 29-32.

[2] Miwa, K.; Aoki, M.; Noritake, T.; Ohba, N.; Nakamori, Y.; Towata, S.-i.; Zuttel, A.; Orimo, S.-i., Phys. Rev. B: Condens. Matter Mater. Phys. Thermodynamical stability of calcium borohydride Ca(BH4)2, 2006, 74, 155122/1-155122/5.

[3] Gu, J.; Gao, M.; Pan, H.; Liu, Y.; Li, B.; Yang, Y.; Liang, C.; Fu, H.; Guo, Z., Energy Environ. Sci. Improved hydrogen storage performance of Ca(BH4)2: a synergetic effect of porous morphology and in situ formed TiO2, 2013, 6, 847-858.

[4] Kim, J.-H.; Shim, J.-H.; Cho, Y. W., J. Power Sources On the reversibility of hydrogen storage in Ti- and Nb-catalyzed Ca(BH4)2, 2008, 181, 140-143.

[5] Wang, L.-L.; Graham, D. D.; Robertson, I. M.; Johnson, D. D., J. Phys. Chem. C On the Reversibility of Hydrogen-Storage Reactions in Ca(BH4)2: Characterization via Experiment and Theory, 2009, 113, 20088-20096.

[6] Rongeat, C.; D'Anna, V.; Hagemann, H.; Borgschulte, A.; Züttel, A.; Schultz, L.; Gutfleisch, O., J. Alloys Compd. Effect of additives on the synthesis and reversibility of Ca(BH4)2, 2010, 493, 281-287.

[7] Ravnsbaek, D. B., Ph.D. Thesis Synthesis, structure and properties of novel metal borohydrides, 2011.

[8] Kim, Y.; Hwang, S.-J.; Shim, J.-H.; Lee, Y.-S.; Han, H. N.; Cho, Y. W., J. Phys. Chem. C Investigation of the Dehydrogenation Reaction Pathway of Ca(BH4)2 and Reversibility of Intermediate Phases, 2012, 116, 4330-4334.

[9] Riktor, M. D.; Sorby, M. H.; Chlopek, K.; Fichtner, M.; Buchter, F.; Zuttel, A.; Hauback, B. C., J. Mater. Chem. In situ synchrotron diffraction studies of phase transitions and thermal decomposition of Mg(BH4)2 and Ca(BH4)2, 2007, 17, 4939-4942.

[10] Nakamoto, K.; Editor, Infrared and Raman Spectra of Inorganic and Coordination Compounds, Part A: Theory and Applications in Inorganic Chemistry, Sixth Edition. John Wiley & Sons, Inc.: 2009; p 419 pp.

[11] Zavorotynska, O.; Deledda, S.; Li, G.; Matsuo, M.; Orimo, S.-i.; Hauback, B. C., Angew. Chem. Int. Ed Isotopic Exchange in Porous and Dense Magnesium Borohydride, 2015, 54, 10592-10595.

[12] Remhof, A.; Gremaud, R.; Buchter, F.; Lodziana, Z.; Embs, J. P.; Ramirez-Cuesta, T. A. J.; Borgschulte, A.; Zuttel, A., Zeitschrift Fur Physikalische Chemie-International Journal of Research in Physical Chemistry & Chemical Physics Hydrogen Dynamics in Lightweight Tetrahydroborates, 2010, 224, 263-278.

[13] Borgschulte, A.; Jain, A.; Ramirez-Cuesta, A. J.; Martelli, P.; Remhof, A.; Friedrichs, O.; Gremaud, R.; Zuttel, A., Faraday Discuss. Mobility and dynamics in the complex hydrides LiAlH4 and LiBH4, 2011, 151, 213-230.

[14] Brown, W. G.; Kaplan, L.; Wilzbach, K. E., J. Am. Chem. Soc. The Exchange of Hydrogen Gas with Lithium and Sodium Borohydrides, 1952, 74, 1343.

[15] Than, C.; Morimoto, H.; Andres, H.; Williams, P. G., J. Labelled Compd. Radiopharm. Tritium and deuterium labelling studies of alkali metal borohydrides and their application to simple reductions, 1996, 38, 693-711.

[16] Mesmer, R. E.; Jolly, W. L., J. Am. Chem. Soc. The Exchange of Deuterium with Solid Potassium Hydroborate, 1962, 84, 2039-2042.

[17] Hagemann, H.; D'Anna, V.; Rapin, J.-P.; Yvon, K., J. Phys. Chem. C Deuterium-Hydrogen Exchange in Solid Mg(BH4)2, 2010, 114, 10045-10047.

[18] Udovic, T. J.; Matsuo, M.; Unemoto, A.; Verdal, N.; Stavila, V.; Skripov, A. V.; Rush, J. J.; Takamura, H.; Orimo, S.-i., Chem. Commun. Sodium superionic conduction in Na2B12H12, 2014, 50, 3750-3752.

[19] Her, J.-H.; Zhou, W.; Stavila, V.; Brown, C. M.; Udovic, T. J., J. Phys. Chem. C Role of Cation Size on the Structural Behavior of the Alkali-Metal Dodecahydro-closo-Dodecaborates, 2009, 113, 11187-11189.

[20] Verdal, N.; Her, J.-H.; Stavila, V.; Soloninin, A. V.; Babanova, O. A.; Skripov, A. V.; Udovic, T. J.; Rush, J. J., J. Solid State Chem. Complex high-temperature phase transitions in Li2B12H12 and Na2B12H12, 2014, 212, 81-91.

[21] Leites, L. A.; Bukalov, S. S.; Kurbakova, A. P.; Kaganski, M. M.; Gaft, Y. L.; Kuznetsov, N. T.; Zakharova, I. A., Spectrochim. Acta, Part A Vibrational spectra of icosabedral closoborate anions B12X2−12 (X = H, D, Cl, Br, I), 1982, 38, 1047-1056.

[22] Muetterties, E. L.; Balthis, J. H.; Chia, Y. T.; Knoth, W. H.; Miller, H. C., Inorg. Chem. Chemistry of boranes. VIII. Salts and acids of B10H102- and B12H122, 1964, 3, 444-51.

[23] Muetterties, E. L.; Merrifield, R. E.; Miller, H. C.; Knoth, W. H., Jr.; Downing, J. R., J. Am. Chem. Soc. Chemistry of boranes. III. Infrared and Raman spectra of B12H12- and related anions, 1962, 84, 2506-8.

[24] Dammak, H.; Antoshchenkova, E.; Hayoun, M.; Finocchi, F., J. Phys.: Condens. Matter Isotope effects in lithium hydride and lithium deuteride crystals by molecular dynamics simulations, 2012, 24, 435402/1-435402/6.

[25] Buchter, F.; Łodziana, Z.; Remhof, A.; Friedrichs, O.; Borgschulte, A.; Mauron, P.; Züttel, A.; Sheptyakov, D.; Barkhordarian, G.; Bormann, R.; Chłopek, K.; Fichtner, M.; Sørby, M.; Riktor, M.; Hauback, B.; Orimo, S., J. Phys. Chem. B Structure of Ca(BD4)2 β-Phase from Combined Neutron and Synchrotron X-ray Powder Diffraction Data and Density Functional Calculations, 2008, 112, 8042-8048.

[26] Buchter, F.; Lodziana, Z.; Remhof, A.; Friedrichs, O.; Borgschulte, A.; Mauron, P.; Zuttel, A.; Sheptyakov, D.; Palatinus, L.; Chlopek, K.; Fichtner, M.; Barkhordarian, G.; Bormann, R.; Hauback, B. C., J. Phys. Chem. C Structure of the Orthorhombic γ-Phase and Phase Transitions of Ca(BD4)2, 2009, 113, 17223-17230.

[27] D'Anna, V.; Spyratou, A.; Sharma, M.; Hagemann, H., Spectrochim. Acta, Part A FT-IR spectra of inorganic borohydrides, 2014, 128, 902-906.

[28] Fichtner, M.; Chlopek, K.; Longhini, M.; Hagemann, H., J. Phys. Chem. C Vibrational spectra of Ca(BH4)2, 2008, 112, 11575-11579.

[29] Mao, J.; Guo, Z.; Poh, C. K.; Ranjbar, A.; Guo, Y.; Yu, X.; Liu, H., J. Alloys Compd. Study on the dehydrogenation kinetics and thermodynamics of Ca(BH4)2, 2010, 500, 200-205.

[30] Hammersley, A. P.; Svensson, S. O.; Hanfland, M.; Fitch, A. N.; Häusermann, D., High Pressure Res. Two Dimensional detector systems: from real detector to idealised image of two-theta scan. , 1996, 14, 235-248.

[31] Filinchuk, Y.; Cerny, R.; Hagemann, H., Chem. Mater. Insight into Mg(BH4)2 with Synchrotron X- ray Diffraction: Structure Revision, Crystal Chemistry, and Anomalous Thermal Expansion, 2009, 21, 925-933.

[32] Filinchuk, Y.; Roennebro, E.; Chandra, D., Acta Mater. Crystal structures and phase transformations in Ca(BH4)2, 2009, 57, 732-738.

[33] Filinchuk, Y.; Richter, B.; Jensen, T. R.; Dmitriev, V.; Chernyshov, D.; Hagemann, H., Angew. Chem. Int. Ed. Porous and Dense Magnesium Borohydride Frameworks: Synthesis, Stability, and Reversible Absorption of Guest Species, 2011, 50, 11162-11166.

[34] Jaron, T.; Grochala, W., J. Nucl. Mater. Y(BD4)3, an efficient store of deuterium, and impact of isotope effects on its thermal decomposition, 2012, 420, 307-313.

Chapter 5

Reduction of CO2 using

γ-Mg(BH4)2

5.1 Introduction

Radiation coming from Sun to earth contains a broad spectrum of wavelengths (from long wavelength infrared to visible and short wavelength UV rays). As these solar radiations enter the earth atmosphere, some of these are reflected back due to small particles in earth’s atmosphere, clouds or by earth’s surface. Some of the radiations are absorbed by earth’s atmosphere by stratospheric ozone, water vapour and by earth itself. Energy absorbed by earth is re-emitted, these terrestrial radiations are in part absorbed by the atmospheric gases like CO2, water vapour,

CH4, and N2O. Since these gases absorb (and later emit) radiation in IR range, these gases are called as greenhouse gases (or heat trapping gases). Radiations (long wavelength or IR) absorbed by these gases are emitted in all the directions even back towards earth. The net effect of this process is that most of the outgoing radiations stay within the atmosphere itself causing heat to be trapped inside the atmosphere.1

Among the various heat trapping (greenhouse) gases, CO2 remains in the atmosphere for the longest time. There are natural cycles which fix and release these greenhouse gases. It takes about a decade for CH4 to leave the atmosphere, about a century for N2O and several centuries 2-4 for CO2. Since the industrial revolution, the amount of CO2 in the atmosphere has dramatically increased, which is a direct sign of the excessive combustion of fossil fuels. This increase in CO2 is linked with increase in global temperature and thus there is a need to reduce the accumulation 5-6 of CO2 in the atmosphere.

In this chapter, we investigate a porous complex metal hydride (Mg(BH4)2) as a reducing agent to reduce CO2. Several experiments were carried to measure the kinetic parameters of CO2 reduction. The kinetics of the reduction of CO2 using a metal borohydride should be mainly dependent on two processes:

1) The affinity of the cation (of metal in borohydride) for CO2. 2) Ease of Metal Borohydride to loose hydrogen.

Affinity of the cation (of metal borohydride) for CO2

The affinity of CO2 for different metal ions is dependent on the pressure of CO2. At 1 bar pressure, the affinity is reported to be in following order: Ref7:7 Mg > Co > Ni > Zn ( at 296 K, metal ions connected by multifunctional ligands that act as linkers based on both the tetra-anionic form 2,5-dioxido-1,4-benzene-dicarboxylate (DOBDC)). 8 Ref8: Mg ~Ni~Co~Fe> Zn>Mn>Cu (at 298 K for M2DOBDC, where M=metal).

Ease of Metal Borohydride to loose hydrogen

The ease of metal borohydride to loose hydrogen has been related to the electronegativity of the cation9. Thus metal borohydrides, with metal ions such as Co, Ni, Fe or Al (electronegativity 1.9,1.9, 1.8, 1.5 respectively) will release hydrogen upon heating at much lower temperatures compared to the borohydrides of K, Na or Li (electronegativity <1) or Mg (electronegativity=1.2). It should be noted that for several transition metals, the release of toxic diborane instead of hydrogen is observed.

By combining these two effects, the highest rate of CO2 reduction should be achieved either with

Mg(BH4)2 or with Fe/Ni/Co (BH4)x. Recently Bordiga and co-workers have published a study 10 confirming γ- Mg(BH4)2 as an efficient reducing agent to reduce CO2. Zuttel and co-workers 11 have also investigated the origin of catalytic activity of metal hydride in reduction of CO2.

5.2 Experimental Procedure

IR spectra were measured from 500 cm-1 to 5000 cm-1. The spectral resolution of the instrument was set to 0.5 cm-1 for gas phase IR spectra and to 1 cm-1 for solid phase IR spectra. Other details for the respective experiments are given below. IR measurements were done at room temperature unless temperature is specified.

5.2.1 Room Temperature, in-situ IR measurement using glass cell

Room temperature in situ FTIR measurements were conducted in a homemade glass cell (figure 5.1) of volume 52 cm3. At the top and the bottom, this glass cylinder (at the left and the right end, in the image) had CaF2 windows which are IR transparent in the range of interest. 1 mmol

(54 mg) of Mg(BH4)2 is placed inside this glass cell under an inert atmosphere of N2. The cell is evacuated and then CO2 is introduced. To monitor the progress of the reaction, IR spectra were collected every 10 minutes for first 10-15 hours of the reaction. The experiments were repeated to check the reproducibility of data. In one of the experiments, the total pressure was monitored, by connecting one of the openings to a barometer.

Initially, commercial Mg(BH4)2 (bought from Sigma Aldrich) was used in this experiment and later, experiments were repeated with γ-Mg(BH4)2.

Figure 5.1: Experimental setup for gas phase IR. One of the taps is connected to vacuum pump to evacuate the cell after evacuation tap connected to the vacuum pump is closed while the second tap connected to CO2 supply (balloon) is opened. IR transparent CaF2 windows are used at the left and right end.

5.2.2 Room temperature, in-situ IR measurements using metallic cell

The room temperature experiments were repeated using a metallic cell. The volume of the cell was measured to be 125.6 cm3. ZnSe windows were used in this cell. The same procedure as described above was followed to monitor the reaction via in-situ IR spectra.

5.2.3 High temperature (100 °C), in-situ IR measurements using metallic cell

The same procedure was also repeated for Mg(BH4)2 and CO2 at 100 °C in a heatable metallic cell. Mg(BH4)2 was introduced into the cell at room temperature, after which cell was evacuated.

The cell was heated to 100 °C and then CO2 was introduced and the measurements were performed as explained above.

5.2.4 Ex-situ measurements (α- Mg(BH4)2)

In another set experiment (ex-situ), 1 mmole of α-Mg(BH4)2 was introduced in two separate flasks (50 cc RB flask) and CO2 was introduced in the flasks. One flask was maintained at RT while another was heated to 120 °C.

5.3 Results and Observations

The full range (500 - 5000 cm-1) gas phase IR spectra of the product of the reaction between

CO2(gas) and γ-Mg(BH4)2 (solid) is shown in figure 5.2. The only gases present are CO2 and diborane B2H6. Bands occurring in specified ranges (corresponding to CO2 and B2H6) were monitored to check the progress of the reaction (between CO2 and Mg(BH4)2). IR for CO2 contains a doubly degenerate O=C=O bending band around 667 cm-1 and a C=O asymmetrical -1 -1 stretching band around 2350 cm . A weak band at 3700 cm labelled CO2 in figure 5.2 is a combination band formed by the combination of asymmetric stretch and IR inactive symmetric stretch occurring at 1388 cm-1.

CO B H CO2 2 2 6 CO2 B2H6

Figure 5.2: Gas phase IR spectrum of the product of the reaction between CO2(gas) and γ-Mg(BH4)2

(solid). In order to measure the kinetic parameters of the reaction a part of the CO2 spectrum ranging from 2200-2300 cm-1 was zoomed in and the average area of 25 different peaks in this range was used to quantify the amount of CO2 present. To monitor the progress of the reaction, IR spectra was taken at regular intervals of time and the area under these peaks was measured.

-1 -1 The range corresponding to B2H6 between 800 cm to 1100 cm was used to monitor the -1 -1 evolution of B2H6 gas, while bands between 2220 cm to 2300 cm were used for calculating the amount of CO2 consumed during the reaction. Figure 5.3 show these bands. In order to obtain kinetic parameters from the IR spectra, the following procedure was followed:

a) Curves were reduced by certain constant value to bring the baseline close to zero.

Figure 5.3: Gas phase IR spectrum of the evolution of B2H6 (upper) and depletion of CO2 (lower) with time. In each of the spectra, upper curve is at time t=0 and the lowest curve is at time t=240 minutes. Note -1 that as the time increases, the intensity of the band at 980 cm (corresponding to B2H6) increases confirming the evolution of diborane while bands between 2220 to 2290 cm-1 decreases in intensity indicating that some of the CO2 is consumed during the reaction.

a) aa

b) For CO2 peaks maxima was recorded for 21 different peaks (for B2H6 peak maxima for 15 different peaks were taken). These maxima were stored in a new wave (column/vector).

c) Maxima_vector is normalized by dividing with the intensity of CO2 bands at time t=0. d) Average of the ratio for each wave is taken by summing up the ratio of the 21 peaks and dividing by 21. e) A plot of average vrs time is plotted and rate constant is obtained via best fit.

5.3.1 Results for experimental section 5.2.1

XRD pattern for the commercial Mg(BH4)2 show that sample is mostly amorphous and some γ-

Mg(BH4)2 phase is also present. To measure kinetic parameters, intensities over 25 different peaks of CO2 between the range of 2240 to 2280 were averaged to calculate the amount of CO2 present in the cell. A curve of CO2 vs time was plotted (figure 5.4) to find out rate constant for the CO2 degradation. Similarly, intensity over 11 peaks was averaged to find out the amount of

B2H6 and to finally calculate the rate constant for B2H6 evolution (figure 5.5). From the curve, it can be inferred that reaction completes in 150-170 minutes. Rate constant for the CO2 depletion -1 -1 is 0.027(1) min while it 0.043(2) min for B2H6 evolution. Similar values of the rate constant were measured for γ-Mg(BH4)2. Three experiments were performed with γ-Mg(BH4)2 and rate constants are summarised in table 5.1.

Table 5.1: Rate constants (unit: min-1) for different experiments. The following abbreviations are used, GC- Glass Cell, MC-Metallic Cell, Com-Commercial.

Sample 1/τ CO2 consumption 1/τ B2H6 formation Com. Mg(BH4)2 0.027(1) 0.043(2) Exp1. γ Mg(BH4)2 0.023(4) 0.041(18) Exp2. γ Mg(BH4)2 0.019(2) 0.046(3) Exp3. γ Mg(BH4)2/GC 0.021(9) - Exp4. γ Mg(BH4)2/MC - - Exp5. γ Mg(BH4)2/MC (100 °C) 0.106(24) -

Figure 5.4: Curve illustrating the rate at which CO2 is reacting with γ-Mg(BH4)2. Rate constant is 0.027 -1 min , the initial amount of CO2 is normalized to 1.

Figure 5.5: Curve illustrating the rate at which B2H6 is evolving on the reaction between CO2 and γ- -1 Mg(BH4)2. Rate constant was observed to be 0.043 min , the final amount of B2H6 was normalized to 1 and the initial amount was assumed to be 0.

After the completion of the reaction, the cell was opened in an inert atmosphere and the powder was collected. IR spectra of the solid powder revealed that B-H stretching of Mg(BH4)2 were diminished while new bands around 1600 and 1400 cm-1 appeared. On comparing the IR spectra it was observed to match with IR spectra of magnesium formate. Figure 5.6 shows the IR spectrum of the residue (marked as MgBH4_Rdue_72h) after 72 hours of reaction. This spectrum is compared with IR spectra of pure MgBH4 and Mg(HCOO)2.

Figure 5.6: Solid state IR spectrum of the solid residue after the reaction compared with IR spectra of pure Mg(BH4)2 and Mg(HCOO)2.

CO2 is probably reduced to the formate ion. The IR spectra of the solid residue show bands similar to those of Mg(HCOO)2 but shifted to higher frequencies. This suggests the possibility of a mixed ionic species Mg(HCOO)x which generates a local pressure leading to an upward shift of the vibrational frequencies. The overall reaction can be summarized as:

Mg(BH4)2 + 2 CO2 Mg(HCOO)2 + B2H6 (5.1)

If this reaction is correct then the pressure of the cell should decrease with the progress of the reaction. In-order to check the pressure drop, the cell was connected to a barometer and the pressure was recorded. Figure 5.7 shows that pressure decreased with the progress of the reaction. The rate of pressure drop also decreases as the time progresses.

Figure 5.7: Curve illustrating pressure inside the cell as a function of time during the progress of the reaction. The decrease in pressure certainly indicates that more than one gaseous molecule from the reactant side is consumed to generate one gaseous molecule in product side.

5.3.2 Results for experimental section 5.2.2 and 5.2.3

The experiments were repeated at room temperature in the metallic heatable IR gas cell. Initial experiments did not show any rise of the diborane partial pressure, thus the cell was carefully rechecked for leaks and fresh samples were used. It is possible that part of the diborane reacts with the metallic walls of the cell.

At 100 °C, the evolution of B2H6 could be seen as soon as CO2 is introduced into the cell (note that CO2 is introduced after heating the cell containing Mg(BH4)2 to 100 °C). IR spectra of the solid residue obtained after 18 hours of the reaction show the complete disappearance of B-H stretching peak confirming the reaction kinetics to be faster at high temperature. The peaks corresponding to B2H6 did not increase with time (figure 5.8) thus to monitor the reaction only the depletion of CO2 was measured.

Peak corresponding to B2H6

Peak corresponding to CO2

Figure 5.8: Curves showing IR spectra of the progress of reaction between Mg(BH4)2 and CO2 in the metallic cell. It can be observed that the band of B2H6 does not increase in intensity with time (compared to the reaction in the glass cell).

-1 The rate constant for CO2 depletion at 100 °C is found to be 0.106(24) min . This rate constant at 100 °C is four times as faster as the rate constant at 25 °C (rate constants for different reactions are compared in table 3.1). The reaction at 100 °C seems to be complete in 40 minutes as compared to 150-160 minutes at room temperature.

Figures 5.9 and 5.10 illustrate the IR spectra of the sample after being in contact with CO2 (under specified conditions) in comparison to gamma Mg(BH4)2 (starting material) and Mg(HCOO)2 (expected product). It can be observed that at high-temperature characteristic B-H stretching and -1 -1 bending peak of Mg(BH4)2 around 2400 cm and 1200 cm respectively, almost disappears. While new peaks matching with the peaks of magnesium formate appears. Peaks of magnesium formate also appear in an experiment conducted at room temperature but B-H stretching and bending peaks does not completely disappear stating reaction did not complete. Thus the reaction is faster at high temperature.

Mg(BH4)2 Mg(HCOO) B-H bending 2 Mg(BH4)2 C-O stretching (1567 cm-1) B-H stretching COO rocking (1377 cm-1)

Figure 5.9: Spectra of residues obtained after the reaction between CO2 and γ-Mg(BH4)2 under different conditions and in different cells. The red curve shows the spectra of the residue obtained after reaction at 100 °C. There is no band at 2380 cm-1, suggesting a complete decomposition of the borohydride.

5.3.3 Results for experimental section 5.2.4

The reaction with α-Mg(BH4)2 is very slow compared to γ-Mg(BH4)2. IR spectra of the solid were measured after 45 hours. α-Mg(BH4)2 reacted at room temperature presents still very strong B-H stretching bands (around 2300 cm-1), and only a weak carbonyl stretching band can be seen -1 around 1600 cm . The sample reacted with CO2 at 120°C presents practically no more BH4, and formate bands are clearly formed. Figure 5.10 shows the spectra of both samples in comparison with spectra of magnesium formate and pure magnesium borohydride.

B-H stretching

Figure 5.10: IR spectra of the solid residue after reaction between α-Mg(BH4)2 and CO2 at room -1 temperature and 120 °C. The B-H stretching band observed around 2370 cm in Mg(BH4)2 is absent in the sample heated to 120 °C, suggesting that α-Mg(BH4)2 also reduces CO2 at higher temperature. Red curve matches with the green curve (MgFormate) suggesting that magnesium formate is the product of the reaction between CO2 and α-Mg(BH4)2 at 120 °C. The red curve also show some bands around 2800- 3000 cm-1 these bands can be assigned to C-H stretching in magnesium formate.

5.4 Conclusions

γ- Mg(BH4)2 reacts at ambient temperature with 1 bar of CO2. In α-Mg(BH4)2 there is no significant change in IR at room temperature even after 45 hours, indicating a very slow reaction. However upon heating at 120°C the reaction takes place, leading to a final solid residue similar to the product of γ- Mg(BH4)2 at room temperature. This result shows that the diffusion of CO2 in the solid (which is very easy in the porous structure of γ- Mg(BH4)2) is a critical limiting factor of the reactivity of Mg(BH4)2 with CO2.

Rate of reaction between CO2 and γ- Mg(BH4)2 at room temperature was observed to be faster in 10, 12-13 comparison to α-Mg(BH4)2 and previously reported results for KBH4 and alanates. KBH4 was found to react only at temperature greater than 90 °C forming potassium formylhydroborates, K[HxB(OCHO)4-x] (x=1-3) as the main product along with evolution of hydrogen, methanol and carbon monoxide. Alanates are reported to react at temperature greater than 120 °C yielding CH4, H2 and metal oxides as the major products. It was also observed that in TiCl3-doped alanate system hydrogen desorption increases but CO2 reduction was hindered.

We observe for the reaction of γ- Mg(BH4)2 that the rate of formation of B2H6 is faster than the consumption of CO2. Since these rate constants do not match it appears that the reaction follows a complex multistep path (eg CO2 adsorption on Mg(BH4)2 and formation of B2H6 in addition to other steps). Most of the reaction takes place within first three hours. The IR spectra of the solid residue show bands similar to those of Mg(HCOO)2 but shifted to higher frequencies suggesting the possibility of a mixed ionic species containing magnesium and formate. The reaction of γ-

Mg(BH4)2 with CO2 appears to be 4 times faster at 100 °C than at room temperature.

100

101

5.5 References

[1] Forster, P.; V. Ramaswamy; P. Artaxo; T. Berntsen; R. Betts; D.W. Fahey; J. Haywood; J. Lean; D. C. Lowe; G. Myhre; J. Nganga; R. Prinn; G. Raga; M. Schulz; Dorland, R. V., Cambridge University Press Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on climate change, 2007.

[2] Inman, M., Carbon is forever, 2008.

[3] Trenberth, K. E.; Jones, P. D.; Ambenje, P.; Bojariu, R.; Easterling, D.; Tank, A. K.; Parker, D.; Rahimzadeh, F.; Renwick, J. A.; Rusticucci, M.; Soden, B.; Zhai, P., Cambridge University Press Surface and Atmospheric Climate Change. In: Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, 2007.

[4] Mackenzie, F. T., Nat. Geosci. The Long Thaw: How Humans are changing the Next 100,000 Years of Earth's Climate by David Archer, 2009, 2, 85.

[5] Mikkelsen, M.; Jorgensen, M.; Krebs, F. C., Energy Environ Sci The teraton challenge. A review of fixation and transformation of carbon dioxide, 2010, 3, 43-81.

[6] Petit, J. R.; Jouzel, J.; Raynaud, D.; Barkov, N. I.; Barnola, J. M.; Basile, I.; Bender, M.; Chappellaz, J.; Davis, M.; Delaygue, G.; Delmotte, M.; Kotlyakov, V. M.; Legrand, M.; Lipenkov, V. Y.; Lorius, C.; Pepin, L.; Ritz, C.; Saltzman, E.; Stievenard, M., Nature Climate and atmospheric history of the past 420,000 years from the Vostok ice core, Antarctica, 1999, 399, 429-436.

[7] Caskey, S. R.; Wong-Foy, A. G.; Matzger, A. J., J. Am. Chem. Soc. Dramatic Tuning of Carbon Dioxide Uptake via Metal Substitution in a Coordination Polymer with Cylindrical Pores, 2008, 130, 10870-10871.

[8] Queen, W. L.; Hudson, M. R.; Bloch, E. D.; Mason, J. A.; Gonzalez, M. I.; Lee, J. S.; Gygi, D.; Howe, J. D.; Lee, K.; Darwish, T. A.; James, M.; Peterson, V. K.; Teat, S. J.; Smit, B.; Neaton, J. B.; Long, J. R.; Brown, C. M., Chem. Sci. Comprehensive study of carbon dioxide adsorption in the metal- organic frameworks M2(dobdc) (M = Mg, Mn, Fe, Co, Ni, Cu, Zn), 2014, 5, 4569-4581.

[9] Ohba, N.; Miwa, K.; Aoki, M.; Noritake, T.; Towata, S.-i.; Nakamori, Y.; Orimo, S.-i.; Zuttel, A., Phys. Rev. B: Condens. Matter Mater. Phys. First-principles study on the stability of intermediate compounds of LiBH4, 2006, 74, 075110/1-075110/7.

[10] Vitillo, J. G.; Groppo, E.; Bardaji, E. G.; Baricco, M.; Bordiga, S., Phys. Chem. Chem. Phys. Fast carbon dioxide recycling by reaction with [gamma]-Mg(BH4)2, 2014, 16, 22482-22486.

[11] Kato, S.; Matam, S. K.; Kerger, P.; Bernard, L.; Battaglia, C.; Vogel, D.; Rohwerder, M.; Züttel, A., Angew. Chem. Int. Ed. The Origin of the Catalytic Activity of a Metal Hydride in CO2 Reduction, 2016, 55, 6028-6032.

[12] Picasso, C. V.; Safin, D. A.; Dovgaliuk, I.; Devred, F.; Debecker, D.; Li, H.-W.; Proost, J.; Filinchuk, Y., International Journal of Hydrogen Energy Reduction of CO2 with KBH4 in solvent-free conditions, 2016, 41, 14377-14386.

102

[13] Hugelshofer, C. L.; Borgschulte, A.; Callini, E.; Matam, S. K.; Gehrig, J.; Hog, D. T.; Züttel, A., The Journal of Physical Chemistry C Gas–Solid Reaction of Carbon Dioxide with Alanates, 2014, 118, 15940-15945.

103

104

Chapter 6 Theoretical calculations on halogenated closoboranes 2- B12H12-nXn (X=F,Cl,Br)

105

106

6.1 Introduction

2- The closoborane ion B12H12 possesses a fascinating icosahedral structure. Compounds such as

M2B12H12 and their derivatives have recently attracted high interest owing to their superionic conductivity.1-2 Upon heating alkali closoboranes, a phase transition leading to a superionic phase can be observed with conductivities reaching 0.79 S/cm at 550 K for (Li,Na)2B12H12. Other classes of compounds such as metal borohydride closoboranes are also studied in this context.3 The onset of superionic behaviour can be lowered by altering cation to anion ratio, very 4-5 recently high conductivities have been reported for NaCB11H12 and related species. 2- 2- Halogenated derivatives of B12H12 (B12HnX(12-n) where n=0-12 and X=F, Cl, Br) may also present interesting superionic properties.4 The synthesis of fully halogenated closoboranes 2- 6-8 B12X12 has been reported by several authors , however to date partially halogenated species have only been studied sporadically.

2- - 9-13 B12H12 has been observed as an intermediate during BH4 decomposition. Polynuclear borane anion salts have been studied for applications in portable hydrogen storage systems.14 2- Species such as B12HnX(12-n) (where X is halogen and n is a number between 0-12) can be - thought as a probable intermediate if halides or BX4 are used for anion exchange. The synthesis 2- 2- 2- 15-16 of pure salt with B12F12 , B12Cl12 or B12Br12 ion has been reported recently. Our group has - previously reported that isolated BHnF(4-n) (n=0,4) species can be formed from mixtures of 17 11 NaBH4 and NaBF4. Heating these mixtures to 300°C led to the formation of Na2B12H12. Vibrational properties of metal borohydrides and metal borofluorides have been discussed by 18 O. Zavorotynska et al. Highly fluorinated Li2B12HnF(12-n) has been reported to be a good electrolyte for lithium batteries.19

2- 2- In order to investigate whether only B12H12 or partly halogenated B12HnX(12-n) (X=F, Cl, Br) - - species can be formed during the decomposition of BH4 /BX4 mixtures, we have studied in this 2- work different mixed B12HnX(12-n) ions using DFT calculations. This work addresses in particular structural, energetic (thermodynamic) and spectroscopic (vibrational and NMR) properties. The results are compared with available experimental data. The results discussed for the fluorinated species are published in ref[20] and the results for chlorinated and brominated closoboranes will soon be published.20

107

6.2 Computational Details

All calculations were performed using Density Functional Theory (DFT) on isolated B12HnX(12- 2- n) (X=F, Cl, Br) ions Calculations were performed with the B3LYP functional and 6-31G(d,p) basis sets with the Gaussian09 program.21 After optimizing the structure, NMR and vibrational spectra (in the harmonic approximation) were calculated.

108

6.3 Results and Discussions

2- B12X12 (or, X = F, Cl or Br) is an icosahedron. The negative charge is delocalized over the entire icosahedron which imparts high stability to this dianion.

Partial substitution of hydrogen with halogens lowers the symmetry and changes the bond 2- lengths. In this study, we have considered 9 different compositions of B12HnX(12-n) (n = 0-3,6,9- 12) for X = F, Cl and Br. For each of these compositions, all the possible positional isomers were 2- investigated except for B12H6X6 (X=F, Cl, Br). Figure 6.1 illustrates the numbering of the boron atoms, the same numbers apply for the substituted X atoms. All species studied are stable (i.e. no imaginary vibrational frequency was found).

2- Figure 6.1: B12Br12 ion showing the numbering of the different boron atoms as used in the discussion. Boron atoms are shown in light pink and bromine atoms are shown in dark red. Boron atoms are numbered in cyclic order.

109

6.3.1 Relative Stability

Using the total energies obtained from the DFT calculations (at 0K), the energy difference for the 2- 2- 2- formation of B12HnX(12-n) ion from B12H12 and B12X12 ions can be calculated via the reaction:

푛 2- 12−푛 2- 2- B12H12 + B12X12 → B12HnX(12-n) (6.1) 12 12

The result is illustrated in Figure 6.2 for X = F, Cl, and Br. This figure shows that for X = Cl and 2- Br, the species B12H6X6 is the most stable, while this is not the case for B12HnF12-n. The formation of brominated closoboranes appears to be more favourable than with the lighter halogen atoms. In the case of the bromide, the calculated reaction enthalpy exceeds 100 kJ/mol.

Figure 6.2 Calculated energy (at 0 K) for reaction 1 for n varying between 0-12 for X = F, Cl and Br. For X = Cl and Br, two competing contributions are observed: the B-X bond is energetically more stable than the B-H bond, but there is also a steric repulsion between adjacent X ions. This results in the stabilization of the half substituted species.

Fluoride rich compounds with n=1-6 appear to be more stable than the corresponding mixture of 2- 2- B12H12 and B12F12 . Increasing the number of fluorine atoms in the molecule facilitates the formation of the stronger B-F bond (the bond dissociation energy for B-F and B-H at 0 K is 778 kJ/mol and 331 kJ/mol respectively22) but this increase in the number of fluorine atom increases also the repulsion between two F atoms. Hence the most stable species is the one containing either 8 or 9 fluorine atoms. Experimental studies on Li2B12HnF(12-n) have shown that it is

110

difficult to prepare compounds with more than 9 fluoride atoms19. The thermal decomposition of 23 these Li2B12HnF(12-n) salts takes place well above 400°C . Li2B12F12 has been reported to be very resistant against strong acids and bases.23

A similar trend was observed for chlorinated and brominated species but in these two cases, most stable species is with n=6. B-Br and B-Cl bonds are relatively less strong than the B-F bond (B- Br = 433 kJ/mol, B-Cl 531 kJ/mol)24 thus the most stable species for X=Cl and Br is different 2- than X=F in B12HnX(12-n) .

6.3.2 Isomerism

2- B12HnX(12-n) has an icosahedral structure and various positional isomers are possible for n ranging between 2 and 10. For the values of n = 2, 3, 9 and 10, all different positional isomers were studied. For a given value of n, the energy difference between different positional isomers was found to be 3-7 kJ/mol for chlorinated species and 6-11 kJ/mol for brominated species.

2- In the case of B12H10X2 , the 1,7 positional isomer is the most stable, while the 1,12 positional 2- isomer (figure 6.3) is the most stable in B12H2X10 . A similar observation was also made by 2- 25 2- Lepsik M., et al. for B12H10(SCN)2 . The relative energy of the 1,2 isomer of B12H10X2 increases from X=F-Br and thus reflects the effect of mutual repulsion of the halogens which increases with increasing size of the halogen. Table 6.1 summarizes the relative energy of different isomers of the candidates with n=10 and n=2. The XBBH dihedral angle also increases 2- from F to Br. Table.6.2 and figure 6.4 illustrate the bond angles of 1, 2 isomers of B12H10X2 .

2- 2- Figure 6.3 Most stable isomers of B12H2Cl10 and B12H3Cl9 . The 1, 7 isomer is most stable for 2- 2- B12H10Cl2 while the 1, 12 isomer is most stable for B12H2Cl10 .

111

2- The most stable positional isomer of B12H9X3 is the one substituted on the boron atoms 1,3 (or 2- 5) and 7. For B12H3X9 , the most stable positional isomer has hydrogen atoms on boron 1, 2 and 12 as illustrated in Figure 6.3.

Table 6.1 summarizes the relative energies of di- and deca- halogen-substituted isomers (energy relative to the most stable isomer for which E = 0 kJ/mol). Composition X = F [kJ/mole ] X = Cl [kJ/mole] X = Br [kJ/mole] 2- B12H10X2 (1,2) 2.8 6.1 8.7 2- B12H10X2 (1,7) 0 0 0 2- B12H10X2 (1,12) 2.6 0.7 0.2

2- B12X10H2 (1,2) 5.5 2.7 5.0 2- B12X10H2 (1,7) 2.1 0 1.1 2- B12X0H2 (1,12) 0 0 0

2- Table 6.2 Summarizing the bond angle ∠BBX and ∠XBB for 1, 2 isomer of B12H12X2 . ∠BBX is the angle between atom 1,2 and 3 and the ∠XBB is angle between atom 3, 2 and 4 (atom 1-4 as marked in the figure on the right). 2- B12H10X2 (1,2) ∠ BBX ∠ XBB X=F 121.1 122.1 X=Cl 121.9 121.1 X=Br 122.5 120.4

Figure 6.4 Illustrating the atom marking scheme for table 6.2. 6.3.3 Bond Length

2- The average B-H and B-X bond lengths in B12HnX(12-n) increases with increasing number of H 2- atoms (or increasing n) in B12HnX(12-n) . Figure 6.5 compares the average bond lengths of B-B, -2 B-H and B-Cl in B12HnCl(12-n) . The relative increase of bond length is larger for the B-Br bonds than for the B-Cl bonds, while the changes of the B-H bonds are similar for both Cl and Br 2- substituted B12H12 . Note that the corresponding changes for the fluoride substituted species are

112

2- 2- much smaller. The available experimental data for B12X12 and B12H11X with X= Cl, Br confirm these computed trends.26-28

-2 Figure 6.5 Average B-B, B-H, and B-Cl bond lengths in B12HnCl(12-n) . As the number of hydrogen decreases (from right to left) B-X and B-H bond length decreases, while B-B bond length doesn’t follow a clear trend.

2- The average calculated bond lengths of the B12X12 ions (X=H, F, Cl, and Br) are summarized in table 6.3. These data show good agreement with experimental data from crystallographic studies.

2- Table 6.3 Average bond lengths (in Å) of B-B and B-H bonds in B12X12 (X=H, F, Cl or Br). Species B-B B-B Exp B-X B-X Exp 2- 26 26 B12H12 1.786 1.78 (X=H/F)1.201 1.20 2- 27 27 B12F12 1.793 1.77 1.393 1.37-1.38 2- 28 28 B12Cl12 1.792 1.78 1.815 1.79 2- 28 28 B12Br12 1.795 1.78 1.980 1.95 6.3.4 Vibrational Spectra

2- B12X12 (X= H, F, Cl or Br) is a perfect icosahedron (Ih point group). There are 66 vibrational degrees of freedom. 3 IR active (3T1u) and 6 Raman active (2Ag + 4Hg) modes are expected. Experimental spectra of these species have been reported previously in solution and in solids.20, 29 2- For B12H12 , the calculated frequencies are in good agreement with the observed frequencies; the calculated frequencies are summarized in table 6.4.

113

2- Table 6.4 Calculated vibrational frequencies and their symmetry for B12H12 . Frequencies marked with the * symbol are low-intensity bands observed in the Raman spectrum and arise from different boron isomers. Calculated Experimental Mode Frequency (cm-1) Frequency (cm-1) symmetry

* 516 (x5) 530 Hu 30 572 (x5) 584 Hg * 652 (x4) 663 Gg 30-31 702 (x3) 720 T1u 734 (x4) Gu 30 737 (x1) 743 Ag 30 753 (x5) 770 Hg 758 (x3) T2u * 862 (x4) 872 Gu 937 (x4) Gg 30 941 (x5) 949 Hg 943 (x5) Hu 957 (x3) T1g 30-31 1063 (x3) 1070 T1u 2476 (x3) T2u 30 2483 (x5) 2475 Hg 30-31 2504 (x3) 2470-2480 T1u 30 2542 (x1) 2541 Ag

2- All the IR and Raman active frequencies obtained from computational calculations for B12X12

(X=F, Cl, Br) are presented in table 6.5. It could be observed that highest energy T1u (IR active),

Hg and Ag modes (Raman active) do not change in energy upon substitution by different halogen in the same proportion as their mass changes. This indicates that these are vibrational modes related to boron-boron motion. While lower energy T1u mode is decreased by more than half on substituting F by Br indicating these are the modes related to B-X motions. Table 6.5 also presents some experimental data for comparison with computationally calculated data and experimentally observed frequencies for IR and Raman spectra are in good agreement with calculated frequencies. For Na2B12Br12 calculated frequencies can be compared to the experimentally synthesized sample discussed in the next chapter (figure 7.7).

114

2- Table 6.5 Summarizing the calculated IR and Raman active modes for B12X12 (X=F, Cl, and Br) and experimental data available. 2- 2- 2- IR Raman B12F12 B12Cl12 B12Br12 mode mode Calc. Exp. Calc. Exp. Calc. Exp. Freq. Freq. Freq. Freq. Freq. Freq. (cm-1) (cm-1) (cm-1) (cm-1) (cm-1) (cm-1) T1u 237 154 105 T1u - 724 525 425 7 T1u 1216 1013 1030 956 980,1000 7 32 - Hg 181 125 126 85 33 32 - Ag 421 431 293 303 187 33 - Hg 383 393 286 185 - Hg 774 686 648 - Hg 1164 964 913 - Ag 1285 1025 946

2- In crystals, the symmetry of the B12H12 ion is lowered, leading to new selection rules. The -1 splitting of the Hg modes (calculated at 572, 753, 941 and 2483 cm ) is clearly observed in the -1 Raman spectrum of Li2B12H12.4H2O (figure 6.6). For instance, the 572 cm mode splits into 2 bands at 574 and 587 cm-1. Caputo et al34 have performed periodic DFT calculations on monoclinic Na2B12H12. The calculated frequencies (Table ST6.1, in supplementary information) confirm the significant splitting which in this case is similar to those seen in Raman spectrum of 31 Li2B12H12.4H2O. Further, the average calculated frequencies in Na2B12H12 are slightly higher 2- than those of the free B12H12 ion, which reflects a “chemical pressure” effect generated by the 4- 4- crystalline surroundings, as observed and calculated for crystals containing FeH6 and RuH6 ions35.

2- Note that for B12H12 , the presence of different boron isotopes can also lead to a breakdown of selection rules. The effect of isotopic substitution (of boron atoms) has been reported previously36. Experimentally this leads to the observation of very weak additional bands in the Raman spectrum, as shown in figure 6.7. The corresponding frequencies are marked with the label (*) symbol in table 6.4.

Substituting one or several hydrogen atoms by halogens leads to lower symmetries and new selection rules. The effect of different symmetries is illustrated in Figure 6.8 which compare the 2- calculated Raman spectra of the 3 positional isomers of B12H10X2 (X = Cl, Br). The 1, 12 isomer has D5d symmetry (like ferrocene), while the 1, 7 and 1, 3 isomers have C2v symmetry.

115

Figure 6.6: Raman spectrum of Li2B12H12.4H2O. Splitting of various modes can be observed clearly. 2- -1 B12H12 as perfect icosahedra should have a Hg vibrational mode at 572 cm . In a crystal, the local symmetry is lower (no fivefold axis!) and splitting of bands are expected. In this case, we observe two bands at 574 and 587 cm-1 respectively.

Thus, in the B-H stretching region, the B-H stretching modes transform for the 1, 12 isomers as

A1g+E1g+E2g+A2u+E1u+E2u, corresponding to 3 Raman active and 2 IR active vibrations (and one inactive) around 2550 cm-1. In the case of the two other isomers, 9 IR active B-H stretching modes are predicted, leading to much more complex spectra as shown in Figure 6.8.

2- Figure 6.7: Parallel polarized Raman spectra of B12H12 in water. The inset reveals the weak additional bands arising from the presence of 10B boron isotopes.

116

2- Figure 6.8 Raman spectra of different isomers of B12H10X2 species in a) B-H bending region b) B-H stretching vibration region.

If the average frequency of the B-H stretching modes is plotted against the number of H atoms in 2- B12HnX(12-n) species, an empirical relation between frequency and bond length (number of H atoms) can be observed (figure 6.9).

2- Figure 6.9 Average B-H stretching frequency of B12HnX(12-n) for varying n with respect to B-H bond length.

117

6.3.5 NMR spectra

NMR shielding tensors were calculated using the Gauge-Independent Atomic Orbital (GIAO) 37 11 method. Chemical shifts were calculated with respect to BF3.OEt2 for B by subtracting the 2- shielding tensor of the given molecule (B12HnX(12-n) ) from the shielding tensor of BF3.OEt2.

The chemical shifts for the B atoms attached to halogen vary systematically as shown in figure 2- 2- 2- 6.10 from -12.34 (-7.26) ppm in B12Cl12 (B12Br12 ) to +0.12 (+0.63) ppm in B12H11Cl 2- (B12H11Br ). For substitution with fluorides, a variation from -18.3 to + 8.3 ppm was obtained.

It is interesting to note that the strongest changes are observed for the fluorides. We made attempts to correlate the NMR shifts with charges on boron atom but no direct correlation was observed, indicating that several other factors play a role in the difference in slopes of the chemical shift values for different halogenated closoboranes.

11 2- Figure 6.10 Trend of B-NMR shift with n, in B12HnX(12-n) for X = F, Cl and Br. No correlation between NMR shifts and charges on boron atom was observed.

118

6.4 Conclusions

2- The DFT calculations on different B12HnX(12-n) species show systematic trends in several 2- properties like bond length or NMR shift vs the number of H atoms. For B12HnF(12-n) the presence of 1-3 fluoride ions leads to a destabilization in a similar way as previously calculated - 11 for BH4-xFx ions. The presence of fluorides may facilitate the kinetics of the thermal release of 2- hydrogen (by lowering the necessary activation energy). B12H3F9 is predicted to be most stable 2- species among fluoride containing closoboranes while B12H6X6 (X=Cl, Br) is predicted to be most stable for chloride and bromide containing species. Probably the size of halide ion and B-X bond energy plays an important role in determining the stability. The 1,7 isomer is most stable 2- out of 3 isomers which exist for B12H10X2 (X=F, Cl or Br as well as SCN). The change in symmetry on substituting halogens at different positions (positional isomers) has significant effects in the vibrational spectra.

These calculations done on isolated ions produces various interesting results and correlations, and the structural and spectroscopic data agree well with available literature data. However, these calculations do not take into account the effect of the crystalline environment. Further experimental studies on the thermal decomposition of mixed MBH4/MBF4 samples as well as on the controlled synthesis of partially halogenated closoboranes are in progress.

119

6.5 Supporting Information

34 Table ST6.1 Comparison of DFT calculated vibrational frequencies for the isolated ion and Na2B12H12 .

2- 2- B12H12 calc DFT Ih Cs2B12H12 obs Na2B12H12 DFT B12H12 Allis38 Caputo34 exp 516 (x5) Hu 531 (IR) 532, 533 (530) 535, 536 536, 537 543, 544 549, 550 572 (x5) Raman Hg 581 (R) 574, 577 584 586 (R) 578, 579 580, 582 586, 587 590, 592 652 (x3) Gg 663, 664 663 672, 674 674, 675 681, 681 702 (x3) IR T1u 708,718 (IR) 704,704 720 709,712 724,731 734 (x4) Gu 757 (IR) 743,744 748,750 753,754 755,757 737 (x1) Raman Ag 747 747, 751 743 753 (x5) Raman Hg 762 754, 756 770 787 757, 758 774, 776 784, 789 807, 808 758 (x3?) IR vw T2u 760,762 768,768 784,785 862 (x4) Gu 860 859, 860 (872) 872, 873 874, 874 899, 901 937 (x4) Gg 910, 912 918, 920 920, 922 932, 933 941 (x5) Hu 950 929, 932 949 934, 935

120

940, 942 985,987 1002,1002 943 (x5) Raman Hg 972 939, 940 947, 948 959, 962 967, 967 981, 981 957 (x3) Raman vw T1g 990,990 999,1000 1008,1009 1063 (x3) IR T1u 1057 1060,1061 1070 1073 1073, 1073 1089,1090 2476 (x3) T2u 2506,2508 2511,2512 2513,2519 2483 (x5) Raman Hg 2507,2509 2475 2511,2515 2517,2518 2523,2524 2530,2539 2504 (x3) IR T1u 2520,2522 2470-2480 2525,2530 2545,2550

2542 (x1) Raman Ag 2555,2559 2541

121

6.6 References

[1] Verdal, N.; Her, J.-H.; Stavila, V.; Soloninin, A. V.; Babanova, O. A.; Skripov, A. V.; Udovic, T. J.; Rush, J. J., J. Solid State Chem. Complex high-temperature phase transitions in Li2B12H12 and Na2B12H12, 2014, 212, 81-91.

[2] He, L.; Li, H.-W.; Nakajima, H.; Tumanov, N.; Filinchuk, Y.; Hwang, S.-J.; Sharma, M.; Hagemann, H.; Akiba, E., Chem. Mater. Synthesis of a Bimetallic Dodecaborate LiNaB12H12 with Outstanding Superionic Conductivity, 2015, 27, 5483-5486.

[3] Sadikin, Y.; Brighi, M.; Schouwink, P.; Cerny, R., Adv. Energy Mater. Superionic Conduction of Sodium and Lithium in Anion-Mixed Hydroborates Na3BH4B12H12 and (Li0.7Na0.3)3BH4B12H12, 2015.

[4] Tang, W. S.; Udovic, T. J.; Stavila, V., J. Alloys Compd. Altering the structural properties of A2B12H12 compounds via cation and anion modifications, 2015, 645, Supplement 1, S200-S204.

[5] Tang, W. S.; Unemoto, A.; Zhou, W.; Stavila, V.; Matsuo, M.; Wu, H.; Orimo, S.-i.; Udovic, T. J., Energy & Environmental Science Unparalleled lithium and sodium superionic conduction in solid electrolytes with large monovalent cage-like anions, 2015, 8, 3637-3645.

[6] Derendorf, J.; Ke; Knapp, C.; Ruhle, M.; Schulz, C., Dalton Trans. Alkali metal-sulfur dioxide complexes stabilized by halogenated closo-dodecaborate anions, 2010, 39, 8671-8678.

[7] Knoth, W. H.; Miller, H. C.; Sauer, J. C.; Balthis, J. H.; Chia, Y. T.; Muetterties, E. L., Inorg. Chem. Chemistry of boranes. IX. Halogenation of B10H102- and B12H122, 1964, 3, 159-67.

[8] Tiritris, I.; Schleid, T., Z. Anorg. Allg. Chem. Single crystals of the dodecaiodo-closo-dodecaborate Cs2[B12I12].2CH3CN (≡{Cs(NCCH3)}2[B12I12]) from acetonitrile, 2001, 627, 2568-2570.

[10] Orimo, S.-I.; Nakamori, Y.; Ohba, N.; Miwa, K.; Aoki, M.; Towata, S.-i.; Zuttel, A., Appl. Phys. Lett. Experimental studies on intermediate compound of LiBH4, 2006, 89, 021920/1-021920/3.

[11] Yan, Y.; Li, H.-W.; Maekawa, H.; Aoki, M.; Noritake, T.; Matsumoto, M.; Miwa, K.; Towata, S.-i.; Orimo, S.-i., Mater. Trans. Formation process of [B12H12]2- from [BH4]- during the dehydrogenation reaction of Mg(BH4)2, 2011, 52, 1443-1446.

[12] Hansen, B. R. S.; Ravnsbaek, D. B.; Reed, D.; Book, D.; Gundlach, C.; Skibsted, J.; Jensen, T. R., J. Phys. Chem. C Hydrogen Storage Capacity Loss in a LiBH4-Al Composite, 2013, 117, 7423-7432.

[13] Hansen, B. R. S.; Ravnsbaek, D. B.; Skibsted, J.; Jensen, T. R., Phys. Chem. Chem. Phys. Hydrogen reversibility of LiBH4-MgH2-Al composites, 2014, 16, 8970-8980.

[14] Safronov, A. V.; Jalisatgi, S. S.; Lee, H. B.; Hawthorne, M. F., Int. J. Hydrogen Energy Chemical hydrogen storage using polynuclear borane anion salts, 2011, 36, 234-239.

122

[15] Hill, M.; Baron, P.; Cobry, K.; Goll, S. K.; Lang, P.; Knapp, C.; Scherer, H.; Woias, P.; Zhang, P.; Krossing, I., ChemPlusChem Direct Fluorination of Cyclic Carbonates and closo-K2[B12H12] in a Slug- Flow Ministructured Reactor, 2013, 78, 292-301.

[16] Peryshkov, D. V.; Popov, A. A.; Strauss, S. H., J. Am. Chem. Soc. Direct Perfluorination of K2B12H12 in Acetonitrile Occurs at the Gas Bubble-Solution Interface and Is Inhibited by HF. Experimental and DFT Study of Inhibition by Protic Acids and Soft, Polarizable Anions, 2009, 131, 18393-18403.

[17] Rude, L.; Filso, U.; D'Anna, V.; Spyratou Stratmann, A.; Richter, B.; Hino, S.; Zavorotynska, O.; Baricco, M.; Sørby, M. H.; Hauback, B. C.; Hagemann, H.; Besenbacher, F.; Skibsted, J.; Jensen, T. R., Phys. Chem. Chem. Phys. Hydrogen-fluorine exchange in NaBH4-NaBF4, 2013, 15, 18185-18194.

[18] Zavorotynska, O.; Corno, M.; Damin, A.; Spoto, G.; Ugliengo, P.; Baricco, M., J. Phys. Chem. C Vibrational Properties of MBH4 and MBF4 Crystals (M = Li, Na, K): A Combined DFT, Infrared, and Raman Study, 2011, 115, 18890-18900.

[19] Ivanov, S. V.; Casteel, W. J.; Pez, G. P.; Ulman, M. Polyfluorinated boron cluster anions for lithium electrolytes. US20060204843A1, 2006.

[20] Sharma, M.; Sethio, D.; D'Anna, V.; Hagemann, H., Int. J. Hydrogen Energy Theoretical study of B12HnF(12-n)2- species, 2015, 40, 12721-12726.

[21] Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery Jr., J. A.; Peralta, J. E.; Ogliaro, F.; Bearpark, M. J.; Heyd, J.; Brothers, E. N.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A. P.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, N. J.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, Ö.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, Gaussian, Inc.: Wallingford, CT, USA, 2009.

[22] Darwent, B. D., Bond Dissociation Energies in Simple Molecules. GPO: 1970; p 52 pp.

[23] Ivanov, S. V.; Miller, S. M.; Anderson, O. P.; Solntsev, K. A.; Strauss, S. H., J. Am. Chem. Soc. Synthesis and stability of reactive salts of dodecafluoro-closo-dodecaborate(2-), 2003, 125, 4694-5.

[24] Barrow, R. F., Trans. Faraday Soc. Dissociation energies of the gaseous monohalides of boron, aluminum, gallium, indium, and thulium, 1960, 56, 952-8.

[25] Lepšík, M.; Srnec, M.; Plešek, J.; Buděšínský, M.; Klepetářová, B.; Hnyk, D.; Grüner, B.; Rulíšek, L., Inorg. Chem. Thiocyanation of closo-Dodecaborate B12H122−. A Novel Synthetic Route and Theoretical Elucidation of the Reaction Mechanism, 2010, 49, 5040-5048.

[26] Tiritiris, I.; Schleid, T., Z. Anorg. Allg. Chem. The dodecahydro-closo-dodecaborates M2[B12H12] of the heavy alkali metals (M = K+, Rb+, NH4+, Cs+) and their formal iodide adducts M3I[B12H12] (≡ MI·M2[B12H12]), 2003, 629, 1390-1402.

123

[27] Peryshkov, D. V.; Strauss, S. H., J. Fluorine Chem. K2B12F12: A rare A2X structure for an ionic compound at ambient conditions, 2010, 131, 1252-1256.

[28] Boere, R. T.; Derendorf, J.; Jenne, C.; Kacprzak, S.; Kessler, M.; Riebau, R.; Riedel, S.; Roemmele, T. L.; Ruehle, M.; Scherer, H.; Vent-Schmidt, T.; Warneke, J.; Weber, S., Chem. - Eur. J. On the Oxidation of the Three-Dimensional Aromatics [B12X12]2- (X=F, Cl, Br, I), 2014, 20, 4447-4459.

[29] Sethio, D.; Lawson Daku, L. M.; Hagemann, H., Int. J. Hydrogen Energy A theoretical study of the spectroscopic properties of B2H6 and of a series of Bx species (x = 1-12, y = 3-14, z = 0-2): From BH3 to B12H122, 2016, 41, 6814-6824.

[30] Muetterties, E. L.; Merrifield, R. E.; Miller, H. C.; Knoth, W. H., Jr.; Downing, J. R., J. Am. Chem. Soc. Chemistry of boranes. III. Infrared and Raman spectra of B12H12- and related anions, 1962, 84, 2506-8.

[31] Muetterties, E. L.; Balthis, J. H.; Chia, Y. T.; Knoth, W. H.; Miller, H. C., Inorg. Chem. Chemistry of boranes. VIII. Salts and acids of B10H102- and B12H122, 1964, 3, 444-51.

[32] Kessler, M.; Knapp, C.; Zogaj, A., Organometallics Cationic Dialkyl Metal Compounds of Group 13 Elements (E = Al, Ga, In) Stabilized by the Weakly Coordinating Dianion [B12Cl12]2–, 2011, 30, 3786-3792.

[33] Peryshkov, D. V.; Goreshnik, E.; Mazej, Z.; Strauss, S. H., J. Fluorine Chem. Co-crystallization of octahedral and icosahedral fluoroanions in K3(AsF6)(B12F12) and Cs3(AsF6)(B12F12). Rare examples of salts containing fluoroanions with different shapes and charges, 2010, 131, 1225-1228.

[34] Caputo, R.; Garroni, S.; Olid, D.; Teixidor, F.; Surinach, S.; Baro, M. D., Phys. Chem. Chem. Phys. Can Na2[B12H12] be a decomposition product of NaBH4?, 2010, 12, 15093-15100.

[35] Hagemann, H.; D'Anna, V.; Lawson Daku, L. M.; Gomes, S.; Renaudin, G.; Yvon, K., J. Phys. Chem. Solids Structural and vibrational properties of Ca2FeH6 and Sr2RuH6, 2011, 72, 286-289.

[36] Nogi, N.; Tanaka, S., J. Solid State Chem. Ab-initio calculations of Raman, IR-active vibrational modes in isotopically modified B12 icosahedral clusters, 2006, 179, 2927-2933.

[37] Wolinski, K.; Hinton, J. F.; Pulay, P., J. Am. Chem. Soc. Efficient implementation of the gauge- independent atomic orbital method for NMR chemical shift calculations, 1990, 112, 8251-8260.

[38] Allis, D. G.; Hudson, B. S., J. Phys. Chem. A Inelastic Neutron Scattering Spectrum of Cs2[B12H12]: Reproduction of Its Solid-State Vibrational Spectrum by Periodic DFT, 2006, 110, 3744- 3749.

124

Chapter 7 Synthesis of Various Substituted Closoboranes 2- (B12XnH(12-n) )

125

126

7.1 Introduction

Closoboranes have recently gained interest as potential electrolyte for all solid state batteries due to properties such as low toxicity, high electrochemical stability ( > 5V vs Li/Li+ electrode) and lower weight density compared to oxides and sulphides type solid electrolytes.1-4

5 Since the discovery of superionic Na conduction in the high-temperature phase of Na2B12H12, a considerable effort has been made to achieve superionic conduction at lower temperatures, ideally at room temperature. Several groups have recently achieved very promising solid ionic 6 7 conductivities in mixed carboranes, in mixed LiNaB12H12, and in mixed Na2B10H10-Na2B12H12 8 9 systems, but also in “perturbed” iodide substituted Na2B12H12.

Sodium ion batteries are considered to be very promising due to the large abundance of sodium compared to lithium which will help to check the cost as and when demand increases.10-11 In this regard, it is interesting to investigate these materials. Chemical substitutions of hydrogen by other species leads to changes in local symmetry, ionic size, but also changes in chemical and electrochemical stability which must be investigated in view of new battery applications.

The previous chapter provides computationally calculated spectroscopic data for various 2- halogenated closoboranes B12H12-nXn (X=F, Cl, Br). In this chapter, we deal with the experimental synthesis of various substituted closoboranes.

The synthesis of fully halogenated compounds such as Na2B12Br12 has been reported previously, and some of these can be purchased commercially. The synthesis of the completely substituted closoboranes (for ex: Na2B12Br12) poses in principle fewer challenges compared to the partially 12 substituted species (Na2B12HnBr(12-n)), as mixtures of different compounds are always formed. To date, the method for purification of different species formed as side products of the reaction has not been developed. In this chapter, we report the preliminary results of some synthetic approaches.

127

7.2 Synthesis, Results and Discussions

7.2.1a Synthesis of ((C2H5)3NH)2B12H12)

This synthesis is inspired from an already reported synthesis by Pitochelli and co-workers as 13 early as 1960. 122.2 mg B10H14 (1 mmol) was mixed with 293.7 µl (C2H5)3N-BH3 (2 mmol, density = 0.783 g/ml) and the mixture was heated to 190 °C for 15 hours. Initially, the formation of yellow liquid was observed which on evaporation of solvent turned to light yellow paste. This paste was washed with di-ethyl ether. On vacuum drying, at room temperature, a yellow coloured powder was obtained. 190 °C B10H14 + 2 (C2H5)3N-BH3 → (C2H5)3NH)2B12H12 + 3H2 (7.1)

7.2.1b Results and Observation

Proton decoupled boron NMR (11B {1H}) of the resulting product show only single peak indicating the presence of identical boron in the product. The peak positions of proton-coupled and decouple 11B NMR (figure 7.1) were found to be in good agreement with previously reported NMR spectra.14 For further investigation, 1H NMR of the product was measured it confirmed the presence of -CH3 and -CH2- groups, spectrum is shown in figure 7.2.

11 11 Figure 7.1: B (top, in red colour) and B (bottom in black colour) NMR spectra for (NEt3H)2B12H12.

128

1 Figure 7.2: H NMR spectrum of (NEt3H)2B12H12. Triplet at 1.2 ppm corresponds to –CH3 group, doublet at 3.1 ppm corresponds to –CH2– group. Origin of peaks at 2.6 ppm and 3.3ppm is yet to be figured out.

The IR spectrum of the product shows the presence of the characteristic B-H stretching peak at 2470 cm-1.15 Figure 7.3 shows the IR spectra of the product in comparison to tri-ethyl amine borane complex. Peaks in the range of 2800-3000 cm-1 can be assigned to -C-H stretching.

B-H C-H stretching stretching

Figure 7.3: IR spectrum of (NEt3H)2B12H12 (red curve) in comparison with amine borane complex

(Et3NBH3, green curve). IR spectrum of the product just before drying (black curve) is also shown in the figure.

129

Further optimization of this reaction scheme is required as some of the peaks in proton NMR spectra of the product (figure 7.2) can’t be explained. This reaction scheme on changing the precursor (if Et3N-BH3 is replaced by NEt3) can obtain decahydro-decaborane ( B10H14 + Et3N + 12, 16 → B10H10 + 4Et3NH ). (NEt3H)2B12H12 can be used as a precursor for various cation exchange reactions, as (NEt3H)2B12H12 is soluble in most organic solvents.

7.2.2a Synthesis of CaB12H12

The synthesis of calcium closoborane is inspired from the synthesis of MB12H12 (Mg, Ca) reported by Liqing He and co-workers.7, 17 All the chemicals used for the reaction were purchased commercially and were used without any further purification.

Ca(BH4)2 (0.5 mmol) was mixed with 62 mg B10H14 (0.5 mmol) in pentadecane in a crucible. This crucible was inserted into an autoclave which was tightly sealed. The autoclave was kept in a pre-heated oven at 160 °C (for 4 hours). The mixture turned yellow and the yellow powder settled down after some time. pentadecane,160 °C Ca(BH4)2 + B10H14 → CaB12H12 + 5 H2 (7.2)

7.2.2b Results and Observations

11 2- IR and B NMR spectra of the product show the presence of the B12H12 group. Figure 7.4 shown below compares the IR spectrum of the final product with the reactants and the solvent. The final product seems to be a mixture of borohydride and closoborane as the characteristic vibrational peaks for borohydride anion (broad band around 2300 cm-1) and closoborane anion (band at 2470 cm-1) can be observed in the spectrum.

The product was handled in open atmosphere, thus peaks of water (moisture) can be observed in the product. Borohydrides or closoboranes quickly absorb moisture from the atmosphere. Another feature which can be noticed from the IR spectrum of the product is the absence of peaks in the lower frequency range (around 600cm-1) which are present in the IR spectrum of 2- B10H14. Thus it can be inferred that B10H14 is consumed during the reaction to form B12H12 .

2- Boron NMR spectrum of the product shown in figure 7.5 indicates the presence of B12H12 group along with other types of borane-hydride compounds. A doublet in 11B NMR and a singlet 11 1 2- in B { H} NMR at δ=15.5 ppm indicates the presence of B12H12 . The NMR chemical shifts at

130

δ=21 and 32 ppm are not assigned yet. This reaction should be optimized to yield CaB12H12 as a pure single product or a method to separate CaB12H12 needs to be devised.

Instead of Ca(BH4)2 deuterated borohydride Ca(BD4)2 was also used in one of the reactions to synthesize partially deuterated closoborane, but the reaction gave a mixture of products which could not be separated. B-H stretching B-H stretching (borohydrides) (closoboranes)

C-H stretching H2O

H2O

Figure 7.4 IR spectrum of the product (red curve) obtained after the reaction of Ca(BH4)2 with B10H14 in pentadecane. For comparison, the IR spectra of Ca(BH4)2 (black curve), B10H14 (green curve) and pentadecane (yellow-green curve) are also shown.

131

11 11 1 Figure 7.5 B NMR and B { H} NMR for the reaction between Ca(BH4)2 and B10H14. Product seems to be a mixture of borane-hydrides.

7.2.3a Bromination of Na2B12H12

Commercially purchased Na2B12H12 was mixed with a stoichiometric excess of bromine in a teflon crucible. The bromine bottle was kept in an ice bath for an hour before it was opened.

Bromine and Na2B12H12 were mixed under a fume hood. This crucible was placed in an autoclave which was tightly sealed. The autoclave was kept in an oven which was pre heated to 160 °C for 16 hours. The autoclave was allowed to cool down and opened. A reddish brown coloured powder was obtained (due to the presence of bromine). By controlling the stoichiometry of bromine added to Na2B12H12 and varying the time of the reaction, different mixtures of partially brominated compounds were also obtained. These mixtures could not be purified to obtain pure products.

7.2.3b Results and Observations

The spectroscopic data of the fully brominated closoborane matches well with the commercial sample. Figure 7.6 shows the boron NMR of the synthesized product in comparison to commercial sample. The small shift is due to the difference in the solvent (commercial sample in

CD3CN δ = -12.73, and synthesized sample in D2O δ = -13.09). This NMR spectrum indicates

132

that only one boron species is present in the product and thus the presence of partially substituted species can be ruled out. IR spectrum of the product (figure 7.7) shows a flat plateau around 2400-2600 cm-1 where B-H stretching should be observed if B-H is present.

11 Figure 7.6: B NMR spectra of as synthesized and commercial sample of Na2B12Br12.

H2O B-H stretching (closoboranes)

Figure 7.7: IR spectrum of as synthesized Na2B12Br12 as compared with commercial sample and

Na2B12H12. No B-H stretching band is observed confirming the complete reaction of Na2B12H12.

133

In order to confirm the presence of the B-Br bond infrared spectra between 200 cm-1 to 700 cm-1 were measured. The IR spectrum is shown in the supporting information figure 7.8. This IR measurement was performed on a Bruker VERTEX 80v Fourier transform infrared (ATR-FTIR) spectrometer. The band for B-Br stretching was observed around 445 cm-1, this band is in good 2- agreement with the peak calculated theoretically for the B12Br12 anion (mentioned in the previous chapter). Raman measurements in solution (solvent H2O) were also performed. The Raman spectrum of the product is shown in figure 7.9, it shows the boron-boron vibration of -1 boron cage in Na2B12Br12 icosahedron close to 1000 cm and the very strong B-Br band at 190 cm-1. The Raman spectrum is also in good agreement with the calculated Raman spectrum for 2- B12Br12 .

Figure 7.8: IR spectrum of experimentally synthesized Na2B12Br12 (red curve) in comparison with 2- computationally calculated IR spectrum for B12Br12 di-anion.

2- While analysing the calculated IR spectra for B12HnBr(12-n) , it was found that as the number (n) 2- of hydrogen atoms in B12HnBr(12-n) decreases, the B-H stretching peak shifts towards higher frequency (figure 7.10). In order to confirm this observation, the synthesis was repeated by varying the bromine to closoborane ratio. This bromine to Na2B12H12 ratio (hydrogen in closoborane) was varied 2:12, 4:12 and 12:12 in three different reactions. The product of the three reactions consisted of the mixtures of brominated closoboranes. The stretching frequencies

134

of the products for all the three reactions are plotted in figure 7.11. As anticipated from the DFT calculations, we observe that the B-H stretching band shifts towards higher wavenumber, as the bromine to closoborane ratio is increased.

Figure 7.9: Raman spectrum of the solution of experimentally synthesized Na2B12Br12 in H2O (green 2- curve) in comparison with computationally calculated Raman spectrum for B12Br12 di-anion.

Figure 7.10: Computational IR spectra of different partially brominated closoborane showing the B-H shift towards higher frequencies as the number of bromines increases.

135

The figure 7.12 compares the mass spectra of Na2B12Br12 with the one for the synthesis products with a 3:1 ratio of bromine (Br2) and Na2B12H12. Electron spray mass spectra (ESI-MS) were 2- obtained on a Shimadzu (LC10A, API 150 Ex) instrument. m/Z peaks for B12Br12 is expected at 1088/2 = 544. A single peak at m/z=543 is observed for as synthesized product (black curve). Presence of single peak confirms single product, spectra of the product is also compared with the product of another bromination reaction in which the partially substituted product was obtained.

For the partially substituted product, the amount of bromine (Br2) and Na2B12H12 was in the 3:1 stoichiometric ratio.

Figure 7.11: B-H stretching vibration of the products obtained from the reaction of bromine with

Na2B12H12. Bromine was added in 3 different ratios. The ratio of Na2B12H12: Br2 is 1:1 for the black curve, 1:2 for the green curve and 1:6 for the red curve.

136

Figure 7.12: Mass spectrum of fully brominated Na2B12Br12 compared to the mass spectrum of the product which contains a mixture of partially substituted closo-boranes.

137

7.2.4 a Synthesis of thiocyanated closoborane

Scheme 1

This reaction scheme was adapted from already published work by Lepsik and co-workers.18

Two solutions were prepared by dissolving 210 mg (1.1mmol) Na2B12H12 in 0.5 ml H2O and dissolving 850 mg (≈ 10.5 mmol) NaSCN in 0.8 ml H2O. Both solutions were mixed and heated to 60 °C. Another solution containing 730 mg CuCl2 (≈ 5.5 mmol) in 1.5 ml H2O was prepared and added dropwise to the mixed solutions. The precipitate was filtered off and the solution was dried to give the powder. This powder was further washed with dry acetone to remove NaCl and acetone was dried after filtration.

Na2B12H12 + (SCN)2 → Na2B12H11SCN + HSCN (7.3)

Na2B12H11SCN + (SCN)2 → Na2B12H10(SCN)2 + HSCN (7.4)

Scheme 2

This reaction scheme was reported by H. G. Srebny and W. Preetz in 1984.19 323 mg (1 mmol)

Pb(SCN)2 was dissolved in 5 ml dry DCM. 1 mmol of Br2 was added to this solution. The solution was filtered after which precipitate was discarded and Na2B12H12 was added to the filtrate. The solution was stirred overnight at before vacuum drying.

Pb(SCN)2 + Br2 → PbBr2 + (SCN)2 (7.5)

7.2.4 b Results and Observation

Boron NMR of the reaction scheme 2 shows the presence of the mixture of thiocyanates and thus further characterizations were performed on the product obtained from reaction scheme 1. 11B NMR spectrum of product obtained from reaction scheme 1 agrees well with the chemical shifts observed by Lepsik and co-workers for mono thiocyanated closoborane (-9(B1), -13.9 (B2- 2- 18 B6),-14.6 (B7-B11),-16.6 (B12) for B12H11SCN ). We expect the final product to be mixed 2- with di-thiocyanated closoborane (B12H10(SCN)2 ) due to an additional chemical shift observed 2- at -15.1 ppm which was reported for B12H10(SCN)2 . The NMR spectrum is shown in figure 7.13.

138

The first reaction was repeated several times in order to optimize it to obtain the pure product. It was observed that reaction depends upon the concentration of the Na2B12H12 and NaSCN solution and the rate of addition of CuCl2 solution. The solutions of Na2B12H12 and NaSCN should be highly concentrated and CuCl2 should be added slowly. We expect that the addition of

CuCl2 to NaSCN solution gives rise to the formation of (SCN)2 which will preferential react with

Na2B12H12 in a concentrated solution while in a dilute solution thiocyanogen will preferentially form a precipitate of poly-thiocyanogen. A slow addition of CuCl2 is important to ensure the

(SCN)2 already produced is reacted with Na2B12H12. A similar observation related to the concentration of the solution is also reported by D. Tudela.20 We couldn’t optimize the reaction scheme to obtain as pure product either mono or di thiocyanated closoborane.

The IR spectrum of the product is shown in figure 7.14. The characteristic peak for the thiocyanate is observed at 2140 cm-1. The band for B-H stretching is found at slightly higher frequencies, this observation is similar to halogenated closoboranes where we observed the B-H stretching bands to shift at higher frequencies upon addition of halogen. The IR spectrum agrees well with the previously reported spectrum and with the DFT calculated spectrum.19, 21

Figure 7.13: 11B NMR spectrum for the product of reaction described in scheme 7.2.4(1).

139

Figure 7.14: IR spectrum (green curve, upper) for the product of reaction described in scheme 7.2.4(1).

For comparison, the IR spectrum of pure Na2B12H12 (red curve, lower) is also shown. Computationally calculated spectra for mono and di thiocyanate closoborane (black and blue curves respectively) are also plotted.

Thiocyanated closoboranes were predicted to be a potential electrolyte for metal ion batteries.22 In order to investigate the conductivity of our product, conductivity was measured using Electrochemical Impedance Spectroscopy (EIS). The sample was pressed into pellets of 12 mm diameter and thickness between 0.4-0.5 mm by applying a pressure of 0.9 GPa. These pellets were loaded into air tight stainless steel cells and contacted between two spring loaded molybdenum electrodes. Ionic conductivity was measured using a Biologic VMP3 multi-channel potentiostat in the frequency range from 1 MHz to 1 Hz with the voltage amplitude of 50 mV. Temperature dependent measurements were performed by placing the cell in an oven and equilibrating the cell’s temperature for 2 hours at each temperature (between 0-100 °C). Pellet resistance was extracted from the last measurement at each temperature at the intercept of the x- axis of fitted semi-circles in Nyquist plots. The temperature dependence of the conductivity for the product is shown in the figure 7.15.

The room temperature conductivity of Na+ ion in partially thiocyanated closoboranes (or closoboranes mixed thiocyanated closoborane) is in the order of 10-4 S/cm. This conductivity is

140

3 10 fold higher than room temperature conductivity of Na2B12H12. The curves on the right of 7 figure 7.15 illustrate previously reported values for Na/Li closoboranes. Pure Na2B12H12 attains conductivity in the range of 10-4 S/cm at the temperature > 400K. Another interesting point to be observed is that the conductivity in pure sodium closoborane increases after a phase transformation at 529 K. In order to find the possibility of such phase change in the thiocyanated closoboranes, they were subjected to calorimetric measurements. No such phase change is observed in the DSC measurements up to 750 K shown in figure 7.16

Figure 7.15: Temperature dependent Na+ conductivity for the product obtained by reaction scheme 7.2.4(1) (on the left). Shaded region corresponds to the near room temperature region (300 ± 5 K). The ionic conductivity of pure Na and Li closoborane reported in reference [7] (on the right).7

Figure 7.16: DSC curve of the thiocynated product (on the left). On the right is the DSC curve of the Na/Li closoborane taken from reference [23] for comparison.23

141

7.3 Conclusions

Several synthesis schemes and challenges in the purification after initial synthesis of metal closoboranes are reported. Partially reacted salts obtained as side products of the reaction show typically similar solubility as the target closoborane compounds in most of the solvents and thus it’s a challenge to obtain the pure product. Working with organic solvents using (NEt3H)2B12H12 can be helpful as some inorganic side products could be precipitated out.

We have shown the synthesis of pure Na2B12Br12 starting from Na2B12H12 and bromine using an autoclave. Partially brominated closoboranes have also been synthesized. Further purification techniques like chromatography need to be optimized for the purification of single phase compounds from the mixture of products obtained. This work is currently in progress in our laboratory. Ion exchange chromatography is also reported to be used for the separation of earlier 2- 24 reported synthesis of mono-fluoro substituted closoborane (pure B12H11F ). For bromination we have used pure bromine thus limiting the effect of solvent as discussed by Strauss and co- 2- 25-26 workers for the fluorination of B12H12 .

Thiocyanated closoboranes are shown to be better ionic conductors than pure Na2B12B12. No phase change is observed for thiocyanated closoborane on heating up to 750 K. Further investigations to understand the reason of higher conductivity in these compounds are required and are currently being carried out by our group.

142

7.4 References

[1] Matsuo, M.; Nakamori, Y.; Orimo, S.-i.; Maekawa, H.; Takamura, H., Appl. Phys. Lett. Lithium superionic conduction in lithium borohydride accompanied by structural transition, 2007, 91, 224103/1- 224103/3.

[2] Matsuo, M.; Orimo, S.-i., Adv. Energy Mater. Lithium Fast-Ionic Conduction in Complex Hydrides: Review and Prospects, 2011, 1, 161-172.

[3] Unemoto, A.; Matsuo, M.; Orimo, S.-i., Adv. Funct. Mater. Complex Hydrides for Electrochemical Energy Storage, 2014, 24, 2267-2279.

[4] Mohtadi, R.; Orimo, S.-i., Nat. Rev. Mater. The renaissance of hydrides as energy materials, 2016, 2, 16091.

[5] Udovic, T. J.; Matsuo, M.; Unemoto, A.; Verdal, N.; Stavila, V.; Skripov, A. V.; Rush, J. J.; Takamura, H.; Orimo, S.-i., Chem. Commun. Sodium superionic conduction in Na2B12H12, 2014, 50, 3750-3752.

[6] Tang, W. S.; Yoshida, K.; Soloninin, A. V.; Skoryunov, R. V.; Babanova, O. A.; Skripov, A. V.; Dimitrievska, M.; Stavila, V.; Orimo, S.-i.; Udovic, T. J., ACS Energy Lett. Stabilizing Superionic- Conducting Structures via Mixed-Anion Solid Solutions of Monocarba-closo-borate Salts, 2016, 1, 659- 664.

[7] He, L.; Li, H.-W.; Nakajima, H.; Tumanov, N.; Filinchuk, Y.; Hwang, S.-J.; Sharma, M.; Hagemann, H.; Akiba, E., Chem. Mater. Synthesis of a Bimetallic Dodecaborate LiNaB12H12 with Outstanding Superionic Conductivity, 2015, 27, 5483-5486.

[8] Duchene, L.; Kuhnel, R. S.; Rentsch, D.; Remhof, A.; Hagemann, H.; Battaglia, C., Chem. Commun. A highly stable sodium solid-state electrolyte based on a dodeca/deca-borate equimolar mixture, 2017, 53, 4195-4198.

[9] Sadikin, Y.; Schouwink, P.; Brighi, M.; Łodziana, Z.; Černý, R., Inorg. Chem. Modified Anion Packing of Na2B12H12 in Close to Room Temperature Superionic Conductors, 2017, 56, 5006-5016.

[10] Yabuuchi, N.; Kubota, K.; Dahbi, M.; Komaba, S., Chem. Rev. Research Development on Sodium- Ion Batteries, 2014, 114, 11636-11682.

[11] Janek, J.; Zeier, W. G., Nat. Energy A solid future for battery development, 2016, 1, 16141.

[12] Kaszynski, P., Collect. Czech. Chem. Commun. Four decades of organic chemistry of closoboranes: a synthetic toolbox for constructing liquid crystal materials. A review, 1999, 64, 895-926.

[13] Pitochelli, A. R.; Hawthorne, F. M., J. Am. Chem. Soc. THE ISOLATION OF THE ICOSAHEDRAL B12H12-2 ION, 1960, 82, 3228-3229.

[14] Ghanta, S. R.; Rao, M. H.; Muralidharan, K., Dalton Trans. Single-pot synthesis of zinc nanoparticles, borane (BH3) and closo-dodecaborate (B12H12)2- using LiBH4 under mild conditions, 2013, 42, 8420-8425.

143

[15] Leites, L. A.; Bukalov, S. S.; Kurbakova, A. P.; Kaganski, M. M.; Gaft, Y. L.; Kuznetsov, N. T.; Zakharova, I. A., Spectrochim. Acta, Part A Vibrational spectra of icosahedral closoborate anions B12X122- (X = hydrogen, deuterium, chlorine, bromine, iodine), 1982, 38A, 1047-56.

[16] Hawthorne, M. F.; Pitochelli, A. R., J. Am. chem. Soc. THE REACTIONS OF BIS- ACETONITRILE DECABORANE WITH AMINES, 1959, 81, 5519-5519.

[17] He, L.; Li, H.-W.; Tumanov, N.; Filinchuk, Y.; Akiba, E., Dalton Trans. Facile synthesis of anhydrous alkaline earth metal dodecaborates MB12H12 (M = Mg, Ca) from M(BH4)2, 2015, 44, 15882- 15887.

[18] Lepšík, M.; Srnec, M.; Plešek, J.; Buděšínský, M.; Klepetářová, B.; Hnyk, D.; Grüner, B.; Rulíšek, L., Inorg. Chem. Thiocyanation of closo-Dodecaborate B12H122−. A Novel Synthetic Route and Theoretical Elucidation of the Reaction Mechanism, 2010, 49, 5040-5048.

[19] Srebny, H. G.; Preetz, W., Z. Anorg. Allg. Chem. Darstellung und Charakterisierung von Thiocyanatderivaten der Hydroboratanionen B10H102− und B12H122−, 1984, 513, 7-14.

[20] Tudela, D., J. Chem. Educ. The Reaction of Copper(II) with Thiocyanate ions, 1993, 70, 174.

[21] Cataldo, F., Polyhedron 13C NMR and FT-IR spectra of thiocyanogen, S2(CN)2, selenocyanogen, Se2(CN)2, and related compounds, 2000, 19, 681-688.

[22] Fang, H.; Jena, P., J. Phys. Chem. C B12(SCN)12–: An Ultrastable Weakly Coordinating Dianion, 2017, 121, 7697-7702.

[23] Tang, W. S.; Udovic, T. J.; Stavila, V., J. Alloys Compd. Altering the structural properties of A2B12H12 compounds via cation and anion modifications, 2015, 645, Supplement 1, S200-S204.

[24] Thomsen, H.; Haeckel, O.; Krause, U.; Preetz, W., Z. Anorg. Allg. Chem. Preparation and spectroscopic characterization of the monofluorohydro-closo-borates [B6H5F]2- and [B12H11F]2, 1996, 622, 2061-2064.

[25] Ivanov, S. V.; Lupinetti, A. J.; Solntsev, K. A.; Strauss, S. H., J. Fluorine Chem. Fluorination of deltahedral closo-borane and -carborane anions with N-fluoro reagents, 1998, 89, 65-72.

[26] Peryshkov, D. V.; Popov, A. A.; Strauss, S. H., J. Am. Chem. Soc. Direct Perfluorination of K2B12H12 in Acetonitrile Occurs at the Gas Bubble-Solution Interface and Is Inhibited by HF. Experimental and DFT Study of Inhibition by Protic Acids and Soft, Polarizable Anions, 2009, 131, 18393-18403.

144

Chapter 8

Conclusions

145

146

8.1 Conclusions

In order to design materials with desired properties, it is essential to have a fundamental understanding of how and why the properties of any material vary with respect to its structure, phase, and the presence of impurities (or dopants). Thus the understanding of the correlation between these fundamental aspects and the properties of the system is the key to ultimate material design. Future research in the design of functional materials demands the use of computational chemistry to accurately predict the properties of the materials in order to accelerate the material design process. Thus it becomes essential to have a holistic approach including Synthesis, Characterization, Analysis of properties and computatioNal chemistry (SCAN approach) to excel in the field of material design.

My research work focused on analysing the fundamental aspects of materials with potential commercial applications. Different projects concerning the boron-hydrogen compounds with potential application related to hydrogen storage or ionic conduction were undertaken. In order to learn the material design process, the SCAN approach was applied. Different synthesis techniques like Schlenk line techniques, solvothermal synthesis, high-pressure solid-gas reactions, mechano-chemical synthesis and wet lab synthesis were explored. As a result several materials Ba(BH4)2, Sr(BH4)2, Na2B12(SCN)nH(12-n), Na2B12Br12 etc., were successfully synthesized.

In this thesis, we have addressed some chemical aspects related to the synthesis of alkaline earth and europium borohydrides (M(BH4)2).

In combination with previous results for M=Ca and M=Mg, all these compounds can be 1-2 efficiently prepared by the reaction of premilled metal hydride and Et3N-BH3. The starting hydride EuH2 has been prepared directly from the metal and hydrogen gas. This synthesis allows to obtain, very pure samples, after careful removal of residual organic solvents. The crystal structures of several new phases of Ba(BH4)2, Sr(BH4)2 and Eu(BH4)2 have been determined (with R. Cerny and E. Didelot, Lab. Crystallography, Univ. of Geneva). These compounds may also be used as starting materials for bimetallic compounds such as the perovskite-type 3 RbSr(BH4)3.

147

Previously reported syntheses of Sr(BH4)2, Ba(BH4)2 and Eu(BH4)2 had been carried out with corresponding metal halides (in particular chlorides) as starting reagents.4-5 The presence of halide impurities in the product can potentially lead to the formation of solid solutions M(BH4)2- xClx which may present different decomposition pathways, making it more challenging to understand the decomposition pathways of these compounds.

We have observed that our M(BH4)2 samples present a higher decomposition temperature 4, 6 compared to previously reported results. Ba(BH4)2 was found to be stable up to 500 °C, after which it decomposes (between 500-750 °C) to BaB6. A new class of compounds, metal hydride borohydride Sr2H3(BH4) and Eu2H3(BH4) were also observed the first time and characterized. It will be interesting to see whether these intermediate compounds can be reversibly converted back to M(BH4)2. A tentative thermal decomposition pathway is discussed for Eu, Sr and Ba borohydrides.

Following the previous results for Mg(BH4)2, we have studied the isotope exchange reaction for 7 Ca(BH4)2 with D2 gas as well as the reverse reaction for Ca(BD4)2 with H2 gas. A complete isotope exchange can be realized, which may also find applications for labelling in organic chemistry, as Ca(BH4)2 is a commercially available reducing agent. Fully deuterated M(BD4)n compounds may also be used as neutron absorbers.

In the second part of the thesis, we have studied exchange reactions of the type

2− 2− 퐵12퐻12 → 퐵12퐻12−푛푋푛

The synthesis of this type of compounds is motivated by the potential structural implications of partial chemical substitution on the closoborane cage in view of ionic conductivity applications.

. - - With X= SCN and Br . With X = SCN , a mixture of Na2B12(SCN)nH(12-n) with (n =1 and 2) was - synthesized, while for X = Br , complete or partial bromination of Na2B12H12 could be achieved.

2- Controlling the halogenation for partial functionalization of B12H12 is challenging because of the possibility of multiple substitution in the products formed. We have obtained a mixture of partially substituted products from the reaction between bromine and Na2B12H12. The separation of a specific desired compound from this type of mixture requires the development of

148

purification techniques like ion exchange chromatography which has been used in past for 2- 8 separation of B12H11F . These challenges will be part of our future research activities.

In order to study the purity, structural and chemical properties of our samples, many experimental techniques have been used: vibrational spectroscopy, NMR (11B and 1H) spectroscopy, differential scanning calorimetry (DSC), powder X-ray diffraction, mass spectrometry and electrochemical impedance spectroscopy (conducted at EMPA Dübendorf).

In particular, infrared (IR) spectroscopy has been used extensively. It is now a routine experiment in our lab to perform temperature dependent experiments up to 300 °C on air or water sensitive solid samples which are loaded in the glove box. During this thesis, several other studies which are not included in this thesis were undertaken: decomposition reactions of LiBH4 in the presence of LiBF4 (in collaboration with T. Jensen, Aarhus University, Denmark), structural phase transitions of CsAlH4 (with M. Felderhoff, MPI Kohlenforschung, Germany), - - study of BH4 /NH2 systems (with Y. Filinchuk, University de Louvain, Belgium) and confined

LiBH4 (with Petra de Jongh, Utrecht University, Netherlands).

The progress of the isotope exchange in Ca(BH4)2 with D2 gas as well as the progress of the reverse reaction was monitored using IR spectroscopy and allowed to characterize the kinetics and activation energy for this process.

Gas phase IR spectroscopy was used to monitor the consumption of CO2 and formation of diborane for the reaction of CO2 with γ- Mg(BH4)2 or α-Mg(BH4)2.

Additional experiments using Raman spectroscopy allowed to obtain additional spectra to compare with the theoretical predictions from periodic DFT calculations for Ba(BH4)2.

Computational chemistry was used as a predictive tool to predict the structure and vibrational 2- properties in different projects. The DFT calculation on isolated B12H12 anion and halogen (F, Cl or Br) substituted anions gave the spectroscopic data for the compounds which are not trivial to synthesize in pure form. These theoretical spectroscopic data could be used as a reference to compare experimental results and to analyse different trends. For instance, we have recently proposed a correlation between B-H bond lengths and B-H stretching frequencies in closoborane using the computationally calculated spectroscopic data.

149

The experimental and theoretical studies in this work provided several new insights into the fundamental properties of boron-hydrogen compounds.

a) Isotope exchange reactions The reaction of alkaline earth borohydrides with deuterium gas takes place much faster than reactions with alkali borohydrides, implying that the metal is involved in the slow

step of this reaction. Comparison of the activation energy for this reaction in Mg(BH4)2

and Ca(BH4)2 shows that it is about half for the Mg compound. Further experiments are needed to analyze whether this big difference is related to the easier diffusion of hydrogen (respectively deuterium) in the crystal or whether it is related to the nature of the metal ion (Lewis acidity).

b) Reaction of Mg(BH4)2 with CO2

Our experiments reveal that the highly porous modification γ- Mg(BH4)2 reacts already at

room temperature with CO2 forming B2H6 and magnesium formate, while the more stable

modification α-Mg(BH4)2 reacts only at higher temperatures. Previous results in the

literature have shown that KBH4 was found to react only at temperatures above 90 °C to

form potassium formylhydroborates, K[HxB(OCHO)4-x] (x=1-3) as the main product along with the evolution of hydrogen, methanol and carbon monoxide.9 Alanates are

reported to react at temperature greater than 120 °C yielding CH4, H2 and metal oxides as 10 the major products. It is interesting to note that the formation of B2H6 is clearly seen

only at room temperature. The reaction of CO2 with Mg(BH4)2 at room temperature could

in principle allow to recycle CO2 into reduced (and potentially combustible) species. From the fundamental side, these observations suggest that the activated complex 2+ involves the Mg ion directly associated to a reducing species together with CO2, and this could possibly lead to the design for an electrochemically catalytically active system 2+ based on Mg for the heterogeneous reduction of CO2 on larger scales.

2- c) Solid ionic conductors derived from B12H12 Impedance spectroscopy has shown that partially thiocyanated closo-borane is a promising candidate for room temperature solid state ionic conductors, as this material shows a thousand fold increase in room temperature conductivity compared to

Na2B12H12. This preliminary result shows that there is room for improvement of the ionic

150

conductivity for Na+ or Li+ (and eventually Mg2+) ions by modulating the crystalline 2- - packing by partial or full substitution of hydrogen in B12H12 by species such as SCN or halogens. These chemical modifications may also alter the electrochemical stability, as 2- 11 was recently shown theoretically for B12(SCN)12 . Another source for improved conductivity arises from additional disorder using mixed anion species such as 12 Na2(B10H10)1-x(B12H12)x. These promising results suggest to explore even more complex

systems like Na2(B10H10-nXn)1-x(B12H12-mXm)x, provided one can prepare efficiently the pure selectively substituted ingredients, which is still a synthetic challenge to date.

Personal remarks The ease of experimentation and amount of information which can be obtained from temperature dependent infrared spectroscopy (and DFT calculations) allowed me to be the part of various international collaborations (most of these contributions are not mentioned in the thesis). I had the pleasure to work with the research group from Japan, Germany, Netherland, Denmark, and Belgium. Each collaborating group had its own area of expertise and approach to solve the problem, thus collaboration brought in a great learning opportunity. Coherently combining the different results together required the knowledge of different disciplines and at the same time, it was stimulating and challenging. These collaborations allowed me to design an efficient strategy in order to solve the problems using a multidisciplinary approach.

I would like to conclude my Ph.D. thesis by making some final remarks describing the impact of last four years of research on me. During my doctoral studies, I have learned the art of solving a complex problem effectively and economically. All my projects focused towards solving futuristic problems. They instilled in me the habit of thinking one step ahead. Several failures taught me perseverance and trying harder next time. I have been extremely benefitted in terms of learning from the projects I undertook. After finishing each project I had a bigger skill set, to begin with, the next project. In future, I would continue to learn more starting with what I have already learned so far.

151

List of publications during PhD:

 Sharma, M.; Didelot, E.; Spyratou, A.; Lawson Daku L. M.; Cerny, R.; Hagemann, H.,Halide

Free M(BH4)2 (M= Sr, Ba and Eu): Synthesis, Structure and Decomposition, Inorg Chem, 2016, 55, 7090-97. 2-  Sharma, M.; Sethio, D.; D’Anna, V.; Hagemann, H., Theoretical Study of B12HnF(12-n) species, Int. J. Hydrogen Energy, 2015, 40, 12721-12726.  Sharma, M.; Sethio, D.; D’Anna, V.; Fallas, J. C.; Schouwink, P.; Cerny, R.; Hagemann, H.,

Isotope Exchange Reactions in Ca(BH4)2. J. Phys. Chem. C 2015, 119, 29-32.  Chekini, M.; Guenee, L.; Marchionni, V.; Sharma, M.; Burgi, T., Twisted and tubular silica structures by anionic surfactant fibers encapsulation, J. Colloid Interface Sci. 2016, 477, 166- 175.  Morelle, F.; Jepsen, L. H. ; Jensen, T. R. ; Sharma, M.; Hagemann, H.; Filinchuk, Y., Reaction

pathway in Ca(BH4)2-NaNH2 and Mg(BH4)2-NaNH2 hydrogen rich systems, J. Phys. Chem. C, 2016, 8428-8435.  He, L.; Li, H.W.; Nakajima, H.; Tumanov, N.; Filinchuk, Y.; Hawang S.J.; Sharma, M.;

Hagemann, H.; Akiba, E., Synthesis of a Bimetallic Dodecaborate LiNaB12H12 with outstanding superionic conductivity, Chem. Mater, 2015, 27, 5483-5486.  D'Anna, V.; Spyratou, A.; Sharma, M.; Hagemann, H., FT-IR spectra of inorganic borohydrides. Spectrochim. Acta, Part A 2014, 128, 902-906.

152

8.2 References

[1] Chlopek, K.; Frommen, C.; Leon, A.; Zabara, O.; Fichtner, M., J. Mater. Chem. Synthesis and properties of magnesium tetrahydroborate, Mg(BH4)2, 2007, 17, 3496-3503.

[2] Hagemann, H.; Cerny, R., Dalton Trans. Synthetic approaches to inorganic borohydrides, 2010, 39, 6006-6012.

[3] Moller, K. T.; Ley, M. B.; Schouwink, P.; Cerny, R.; Jensen, T. R., Dalton Transactions Synthesis and thermal stability of perovskite alkali metal strontium borohydrides, 2016, 45, 831-840.

[4] Ravnsbaek, D. B.; Nickels, E. A.; Cerny, R.; Olesen, C. H.; David, W. I. F.; Edwards, P. P.; Filinchuk, Y.; Jensen, T. R., Inorg. Chem. Novel Alkali Earth Borohydride Sr(BH4)2 and Borohydride- Chloride Sr(BH4)Cl, 2013, 52, 10877-10885.

[5] Humphries, T. D.; Ley, M. B.; Frommen, C.; Munroe, K. T.; Jensen, T. R.; Hauback, B. C., J. Mater. Chem. A Crystal structure and in situ decomposition of Eu(BH4)2 and Sm(BH4)2, 2015, 3, 691-698.

[6] Grube, E.; Olesen, C. H.; Ravnsbaek, D. B.; Jensen, T. R., Dalton Transactions Barium borohydride chlorides: synthesis, crystal structures and thermal properties, 2016, 45, 8291-8299.

[7] Hagemann, H.; D'Anna, V.; Rapin, J.-P.; Yvon, K., J. Phys. Chem. C Deuterium-Hydrogen Exchange in Solid Mg(BH4)2, 2010, 114, 10045-10047.

[8] Thomsen, H.; Haeckel, O.; Krause, U.; Preetz, W., Z. Anorg. Allg. Chem. Preparation and spectroscopic characterization of the monofluorohydro-closo-borates [B6H5F]2- and [B12H11F]2, 1996, 622, 2061-2064.

[9] Picasso, C. V.; Safin, D. A.; Dovgaliuk, I.; Devred, F.; Debecker, D.; Li, H.-W.; Proost, J.; Filinchuk, Y., International Journal of Hydrogen Energy Reduction of CO2 with KBH4 in solvent-free conditions, 2016, 41, 14377-14386.

[10] Hugelshofer, C. L.; Borgschulte, A.; Callini, E.; Matam, S. K.; Gehrig, J.; Hog, D. T.; Züttel, A., The Journal of Physical Chemistry C Gas–Solid Reaction of Carbon Dioxide with Alanates, 2014, 118, 15940-15945.

[11] Fang, H.; Jena, P., J. Phys. Chem. C B12(SCN)12–: An Ultrastable Weakly Coordinating Dianion, 2017, 121, 7697-7702.

[12] Duchene, L.; Kuhnel, R. S.; Rentsch, D.; Remhof, A.; Hagemann, H.; Battaglia, C., Chem. Commun. A highly stable sodium solid-state electrolyte based on a dodeca/deca-borate equimolar mixture, 2017, 53, 4195-4198.

153