ANTONZUBAYER -Astudybyneutrons,muonsandx-rays- IonDynamics&MagneticOrderin2DHoneycombMaterials

(1)

DEGREE PROJECT IN ENGINEERING PHYSICS, SECOND CYCLE, 30 CREDITS

STOCKHOLM, SWEDEN 2020

Ion

D

ynamics &

M

agnetic

O

rder in 2D

H

oneycomb

M

aterials

-

A study by neutrons, muons and xrays

-ANTON ZUBAYER

KTH ROYAL INSTITUTE OF TECHNOLOGY SCHOOL OF ENGINEERING SCIENCES

(2)

(3)

Ion Dynamics & Magnetic

Order in 2D Honeycomb

Materials

A study by neutrons, muons and xrays

-ANTON ZUBAYER

Master’s in Engineering Physics Date: July 1, 2020

Supervisor: Dr. Nami Matsubara Examiner: Prof. Martin Månsson School of Engineering Sciences Department of Applied Physics

Sustainable Materials Research & Technologies (SMaRT)

Swedish title: Jondynamik & magnetism i 2D bikakematerial - En studie via neutroner, muoner och röntgenstrålning

(4)

(5)

iii

Abstract

Arguably one of the world’s most pressing challenges is the storage of enor-mous quantities of energy produced through large power plants, especially as fossil fuel-based plants are being dismantled and renewable energy plants constructed. Where do we store during high season and how to we distribute during low season? Storing the energy in massive batteries is arguably the best solution, although today’s lithium (Li) based batteries pose various strug-gles regarding resource limitation, price, availability, extraction difficulties, toxicity, conflict elements, environmental aspects and mining ethics. On the other hand, other alkali elements like sodium (Na) and potassium (K), do not face those issues. A novel series of honeycomb layered cathode materials have been investigated by Masese et.al., using various structural and electrochem-ical measurements. However, more detailed studies had not yet been carried

out, hence this thesis focuses on advanced characterization of Li2Ni2TeO6,

Na2Ni2TeO6and K2Ni2TeO6. The neutron powder diffraction (NPD) and muon

spin spectroscopy (µ+SR) were carried out at cutting edge large-scale research

laboratories with emphasis on the J-PARC (Japan) facility with complemen-tary measurements at PSI (Switzerland) and at ISIS (UK). The detailed investi-gation elucidated additional evidence of the potential of the novel Na/K based cathode materials and also revealed an enhanced fundamental understanding of the compounds by comprehensive studies at low and high temperatures

(T ≈ 2 − 550 K). XRD/NPD revealed complementary structural information,

showcasing the honeycomb layering of Na2Ni2TeO6 and K2Ni2TeO6, while

Li2Ni2TeO6was disordered. Ion diffusion probed with µ

+

SR further proved the potential of Na/K as a battery cathode, by demonstrating activation ener-gies (207 meV and 155 meV respectively) comparable with the current battery

cathodes LiCoO2 (145 meV). Magnetic measurements using MPMS/PPMS

and low temperature µ+SR revealed an antiferromagnetic behaviour for all

compounds below their respective Néel temperatures, yet Li2Ni2TeO6

demon-strated a 3D magnetic order compared to the 2D order of the two other com-pounds. Investigation of low temperature NPD pattern revealed interesting magnetic spin properties. By high-temperature NPD, a phase transition from

RT hexagonal to HT orthorhombic phase was observed at T ≈ 450 K.

Conclu-sively, the research revealed interesting fundamental characteristics and great battery potential for honeycomb layered Na/K-ion cathode materials.

Keywords: neutron scattering, muons, x-rays, battery materials,

(6)

iv

Sammanfattning

Världens, förmodligen, allra brådskande utmaning är lagringen av enorma mängder av energi som producerats genom stora kraftverk, speciellt när kraft-verk baserade på fossila bränslen rustas ner och kraftkraft-verk baserade på förny-bar energi byggs upp. Vart lagrar vi under hög säsong och hur distribuerar vi under låg säsong? Dagens överflödiga energi från hög säsong går till spil-lo. Att lagra energi i massiva batterier är troligen den bästa lösningen. Dock står dagens litium (Li) baserade batterier inför många hinder relaterade till re-surs begränsningar, pris, tillgänglighet, utvinningssvårigheter, giftighet, kon-fliktmaterial, miljö aspekter och gruvdriftsetik. Natrium (Na) och kalium (K) har inte dessa problem och därav har en ny serie av bikakeskiktade katodma-terial forskats på utav Masese et.al., med hjälp av strukturella och elektro-kemiska mätningar. Dock, mer avancerade karakteriseringar hade ej utförts,

därav handlar min avhandling om avancerad karakterisering av Li2Ni2TeO6,

Na2Ni2TeO6och K2Ni2TeO6. Neutron diffraktion och muon experimenten

ut-fördes i de högteknologiska partikelacceleratorfaciliteterna J-PARC, PSI och ISIS. Den detaljerade undersökningen verkställdes för att belysa ytterligare bevis för potentialen av de nya natrium och kalium baserade katod materialen och dessutom att avslöja en förbättrad fundamental förståelse av kompositio-nerna genom undersökning av ett brett spektrum av låga och höga tempera-turer (∼ 2K till ∼ 550K). X-ray och neutron diffraktion uppenbarade kom-plementära och tydliga strukturella resultat, som påvisade bikakeskiktningen

av Na2Ni2TeO6och K2Ni2TeO6, medan Li2Ni2TeO6var oordnad. Magnetiska

mätningar, m.h.a. MPMS/PPMS och magnetisk µSR, exponerade ett antiferro-magnetiskt beteende för alla kompositioner under var sin Néel temperatur, men

Li2Ni2TeO6demonstrerade ett 3D magnetiskt uppträdande jämfört med 2D

be-teendet som de två andra påvisade. Undersökningen av låg temperatur neutron diffraktion avslöjade intressanta magneto-spin egenskaper. Genom hög tempe-ratur neutron diffraktion så skedde en fas transition från RT hexagonal till HT

orthorhombisk fas runt T = 450K. Jondiffusions µ+SR stärkte potentialen av

natrium och kalium ytterligare, genom att uppdaga en aktiveringsenergi (207

meV och 155 meV) jämförbart med dagens katod LiCoO2 (145meV).

Sam-manfattningsvis, så påvisade forskningen intressanta fundamentala karaktärer samt enorm batteripotential för bikakeskiktad natrium och kalium katodmate-rial.

Nyckelord: neutronspridning, muoner, röntgen, batterimaterial,

(7)

Papers

This MSc thesis has resulted in the outcome of the following research papers:

1. Magnetism and Ion Diffusion in Honeycomb Layered Oxide K2Ni2TeO6:

First Time Study by Muon Spin Rotation & Neutron Scattering

N. Matsubara, E. Nocerino, O. K. Forslund,A. Zubayer, P. Gratrex, D.

Andreica, J. Sugiyama, R. Palm, Z. Guguchia, S. Cottrell, T. Saito, A. Kalaboukhov, Y. Sassa, T. Masese, and M. Månsson

Scientific Reports (2020), submitted manuscript - arXiv:2003.05805

2. Honeycomb Layered Oxides: Structure, Energy Storage, Transport,

Topology and Relevant Insights

Godwill Mbiti Kanyolo, Titus Masese, Nami Matsubara, Chih-Yao Chen, Josef Rizell, Ola Kenji Forslund, Elisabetta Nocerino, Konstantinos Pa-padopoulos,A. Zubayer, Minami Kato, Kohei Tada, Keigo Kubota,

Hi-roshi Senoh, Zhen-Dong Huang, Yasmine Sassa, Martin Mansson, Ha-jime Matsumoto, Qiang Xu

Manuscript (2020) - arXiv:2003.03555

3. Ion Diffusion in Layered Honeycomb Oxides A2Ni2TeO6 (A = Li,

Na, K) Studied by Muons & Neutrons

A. Zubayer, N. Matsubara, R. Palm, E. Nocerino, O. K. Forslund, P.

Gratrex, Y. Sassa, K. Papadopoulos, T. Saito, S. Cottrell, A. Koda, J. Sugiyama, T. Masese, and M. Månsson

Manuscript in writing (2020)

The full versions of the first paper (manuscript) is also included i the Appendix of this thesis.

(8)

List of Figures

1.1 Lithium resources in the world clearly showing the uneven

dis-tribution between different continents and countries. [3] . . . . 3

1.2 Evolution of the Lithium price per metric ton clearly showing

the drastic increase during last decade. [9] . . . 4

2.1 Schematic view of an archetypical commercial Li-ion

recharge-able battery with LiCoO2 cathode, electrolyte/separator and

graphite anode. During charge/discharge cycles the Li+ions

move between the anode and cathode through the electroni-cally insulating electrolyte with electron transport occurring

in the outer circuit. [15] . . . 7

2.2 Crystal structure of multiple unit cells of a typical 2D layered

energy material where the alkali ion layers are sandwiched be-tween the transition metal oxide (TMO) layers. One example of such material is the well-known battery cathode material

LiCoO2 where alkali ions are lithium and TMO layers

con-sists of CO2. [3] . . . 8

2.3 (a) Crystallographic view of the P2-type-layered crystal

struc-ture of A2Ni2TeO6(A = Na and K) along the c-axis. The

pris-matic polyhedra around the K ions are shown for simplicity.

(b) A fragment of the P63/mcm structure of K2Ni2TeO6 in

the ab plane (the magnetoactive honeycomb layers) with three major K sites (K1, K2, and K3). Samples are produced by our close collaborators at the National Institute of Advanced

Industrial Science and Technology (AIST) in Japan. [10] . . . 9

3.1 J-PARC, Japan Proton Accelerator Research Complex [11].

The neutron experiments of this MSc thesis were mainly per-formed at the Materials and Life Science Experimental

Facil-ity (MLF). . . 11

(12)

x LIST OF FIGURES

3.2 (a) Schematic view of muon production where the high-energy

protons collide with the carbon target, producing a pion that then decays into a muon and a neutrino. The muons are then transported through the beamline to the experimental station

where the muon spin rotation (µ+SR) experiments are

per-formed. (b) Schematic view of the spallation process where the high-energy protons collide with the heavy (Hg or W) tar-get to produce neutrons. The energy of the neutrons are then moderated before they are transported through the neutron

beam-line (via superrmirrors) to the experimental station. . . 13

3.3 Safety procedures. . . 14

3.4 Experimental Halls 1 and 2 within the Material and Life

Sci-ence Experimental Facility (MLF) of J-PARC. . . 15

3.5 Overview of all the neutron and muon beamlines/instruments

that are available at MLF/J-PARC. For this thesis, mainly S1 line (muon) and SPICA (neutron) instruments were utilized.

[37] . . . 16

3.6 Typical He(g) or Ar(g) glovebox used for handling sensitive

battery materials. All samples measured by neutron powder diffraction and muon spin rotation measurements were sealed

inside their respective sample can/cell inside such glovebox. . 17

4.1 Bragg diffraction from two atomic planes in a material [42]. . 20

4.2 Sample preparation for x-ray diffraction (XRD) experiments.

For sensitive samples these procedures were performed inside

an Argon or Helium glovebox. . . 21

4.3 XRD Machine by Rigaku in the KEK laboratories [28]. . . 22

5.1 Measurement of magnetic susceptibility using the

Recipro-cating Sample Option (RSO) protocol with a very small am-plitude and a Superconducting QUantum Interference Device

(SQUID) magnetometer.[49] . . . 24

5.2 Schematic view of the most fundamental types of magnetic

spin ordering that can be found in materials. . . 25

5.3 Schematic temperature dependence of the magnetic

suscepti-bility χ(T ) for different types of magnetic spin order

(corre-lations). . . 26

5.4 Sample (black powder) inside a tightly packed capsule in

prepa-ration for the MPMS/PPMS measurements. . . 28

(13)

LIST OF FIGURES xi

6.1 Schematic view of the normalized form factor for x-rays vs.

neutrons (nuclear and magnetic/spin scattering) as a function

of sin θ/λ (i.e. momentum transfer, Q). . . 31

6.2 X-rays vs. neutrons in regards to cross-sectional scattering.[55]

[56] . . . 32

6.3 SPICA, Neutron Diffraction instrumentation. The figure is a

composite of [37] and our own data. . . 34

6.4 Performance of the different detector banks, data from K2Ni2TeO6. 34

6.5 Vanadium sample can preparations. . . 36

6.6 Sample mounting stick for the SPICA neutron powder

diffrac-tion instrument of J-PARC. . . 37

6.7 Sample mounting stick for the iMATERIA neutron powder

diffraction instrument of J-PARC. . . 37

6.8 Lowering of the sample sticks into the cryostat at SPICA (left)

and iMATERIA (right). . . 38

6.9 Closing and securing for SPICA (left) and iMATERIA (right).

Arrows show the path for how the shutters close. . . 38

6.10 SPICA PID- and heat controller. . . 40

7.1 Schematic view of the muon spin rotation (µ+SR) experiment.

The spin polarized muon beam arrives via the beamline and muons are implanted into the target material while a muon de-tector starts the electronic clock. Inside the material, the muon spin is subjected to the internal fields of the material and starts

to spin and/or depolarize. At an average time (τµ= 2.2 µs) the

muon decays and ejects a positron preferentially along the mo-mentary spin direction. The positron is then detected by the forward or backward detector and thereby stops the electronic

clock (i.e. detecting a signal).[Sonier] . . . . 44

(14)

xii LIST OF FIGURES

7.3 (a) Schematic view on how the implanted muon stops close to

the oxygen within a typical layered TMO battery cathode ma-terial. Here it mainly feels the nuclear magnetic dipole mo-ment of the alkali ions. If the alkali ions are static the µSR time spectrum is described by a (b) Static Kubo-Toyabe func-tion (blue curve) with only the field distribufunc-tion (∆) as fit-ting parameter. In the case the ions become mobile the µSR time spectrum changes shape and is better described by the dynamics Kubo-Toyabe function that has the additional field-fluctuation rate (ν) as fitting parameter. Here green, red and black curves symbolize an increasing dynamic contribution. (c) Temperature dependence of ν is fitted to the Arrhenius type

equation. . . 47

7.4 µ+SR sample cell preparation tools, sample pellet press and titanium sample cell. . . 48

7.5 Placing the pellet onto a cell. . . 49

7.6 Sealed titanium sample cell used for µ+SR experiments. . . . 50

7.7 Cell mounted onto a cryostat. . . 51

7.8 The cryostat with the sample inside. . . 51

7.9 Hatch controllers for S1 . . . 52

7.10 An overview of the S1 beamline in J-PARC, Japan. . . 53

7.11 Mounting the cryostat to the sample environment. . . 54

7.12 Valves for pump. . . 55

7.13 Compressor interface. . . 55

7.14 Helium tube mounted. . . 56

7.15 Alignment using laser and optics. . . 57

8.1 XRD data (blue symbols) and Rietveld fit (red line) of Li2Ni2TeO6 using the orthorhombic spacegroup ’F ddd’ (No. 70). Bragg peak positions are shown at the bottom as triangles and above that the difference between fit and data as a green solid line. . . 60

8.2 Crystal structure for Li2Ni2TeO6, where oxygen is not shown in order to get a better view och the structure. . . 61

8.3 Crystal structure for Li2Ni2TeO6viewed along the a-axis, show-ing only the polyhedral for (a) Li1 sites (b) Li2 sites (c) Li3 sites. . . 62

(15)

LIST OF FIGURES xiii

8.4 XRD data (blue symbols) and Rietveld fit (red line) of Na2Ni2TeO6

using the orthorhombic spacegroup ’P 63/mcm’ (No. 193).

Bragg peak positions are shown at the bottom as triangles and above that the difference between fit and data as a green solid

line. . . 63

8.5 Crystal structure for Na2Ni2TeO6 where oxygen is not shown

in order to get a better view och the structure. . . 64

8.6 Crystal structure for Na2Ni2TeO6: (a) 3D rendering of the unit

cell (b) 4 unit cells viewed along the c-axis visualizing a clear honeycomb structure of the nickel around tellurium. (c)

Sev-eral stacked unit cells visualizing the layered structure. . . 65

8.7 XRD data (blue symbols) and Rietveld fit (red line) of K2Ni2TeO6

using the orthorhombic spacegroup ’P 63/mcm’ (No. 193).

Bragg peak positions are shown at the bottom as triangles and above that the difference between fit and data as a green solid

line. . . 66

8.8 Crystal structure for K2Ni2TeO6where oxygen is not shown in

order to get a better view och the structure. . . 66

8.9 Crystal structure for K2Ni2TeO6: (a) 3D rendering of the unit

cell (b) 4 unit cells viewed along the c-axis visualizing a clear honeycomb structure of the nickel around tellurium. (c)

Sev-eral stacked unit cells visualizing the layered structure. . . 67

9.1 NPD data at T = 300 K (blue symbols) and Rietveld fit (red

line) of Li2Ni2TeO6from the BS-bank. Bragg peak positions

are shown as triangles and the difference between fit and data

as a green solid line. . . 70

line) of Li2Ni2TeO6from the QA-bank. Bragg peak positions

line) of Li2Ni2TeO6from the LA-bank. Bragg peak positions

9.4 NPD data at T = 300 K (blue symbols) and global Rietveld fit

(red line) of Li2Ni2TeO6from all three detector banks. Bragg

peak positions are shown as triangles and difference between

(16)

xiv LIST OF FIGURES

line) of Na2Ni2TeO6from the BS-bank. Bragg peak positions

line) of Na2Ni2TeO6from the QA-bank. Bragg peak positions

line) of Na2Ni2TeO6from the LA-bank. Bragg peak positions

(red line) of Na2Ni2TeO6from all three detector banks. Bragg

fit and data as a green solid line. . . 75

line) of K2Ni2TeO6from the BS-bank. Bragg peak positions

line) of K2Ni2TeO6from the QA-bank. Bragg peak positions

line) of Na2Ni2TeO6from the LA-bank. Bragg peak positions

(red line) of K2Ni2TeO6from all three detector banks. Bragg

fit and data as a green solid line. . . 78

9.13 Comparison of XRD and NPD regarding structure parameters

for the Li2Ni2TeO6compound. . . 81

9.14 Comparison of XRD/NPD structure parameters for Na2Ni2TeO6. 82

9.15 Comparison of XRD/NPD structure parameters for K2Ni2TeO6. 82

10.1 Magnetic susceptibility (blue) and inverse susceptibility

(17)

LIST OF FIGURES xv

(or-ange) for Na2Ni2TeO6. . . 85

(or-ange) for K2Ni2TeO6using both field cool (FC) and zero-field

cool protocols. In the field cool protocol a field is applied dur-ing cooldur-ing and in the zero-field cool, the sample is cooled

without any external field. . . 86

10.4 Comparison of the temperature dependent magnetic

suscepti-bility for Li2Ni2TeO6, N2Ni2TeO6and K2Ni2TeO6, which clearly

show the onset of antiferromagnetic order for all samples. . . . 87

10.5 Lattice of the magnetic Ni atoms in Li2Ni2TeO6. Due to the

structure and interchangeable Ni/Li positions there are many different Ni-Ni distances: 2.956 Å, 2.999 Å, 2.983 Å, 4.206

Å, 3.038 Å and 2.962 Å . . . 88

10.6 Lattice of the magnetic nickel atoms in Na2Ni2TeO6showing

the Ni-Ni (a) Inter-plane distance d1 = 5.619 Å and (b)

Intra-plane distance d2 = 3.014 Å. . . 88

10.7 Lattice of the magnetic nickel atoms in K2Ni2TeO6 showing

the Ni-Ni (a) Inter-plane distance d1 = 6.219 Å and (b)

Intra-plane distance d2 = 3.042 Å. . . 89

11.1 Data (blue symbols) and fit (black line) of the ZF µ+SR time

spectrum of Li2Ni2TeO6acquired at the base temperature (T =

2 K). Main figure show the long time domain while the inset

focus on the shorter time-scale. . . 91

11.2 Selected temperature dependent ZF µ+SR time spectra

(sym-bols) and corresponding fit (solid black lines) for Li2Ni2TeO6.

Data are displayed in the short time domain and each spectrum

is slightly offset along the y-axis for clarity of display. . . 92

11.3 Magnetic order parameter of Li2Ni2TeO6showing the

temper-ature dependence of the frequencies, ωAF n obtained from the

fits of the ZF µ+SR data. Dashed lines are guides to the eye. . 93

spectrum of Na2Ni2TeO6acquired at the base temperature (T =

1.6 K). Main figure show the long time domain while inset

(18)

xvi LIST OF FIGURES

(sym-bols) and corresponding fit (solid black lines) for Na2Ni2TeO6.

11.6 Magnetic order parameter of Na2Ni2TeO6showing the

temper-ature dependence of the frequencies, ωAF nobtained from the

fits of the ZF µ+SR data. Dashed lines are guides to the eye. . 96

spectrum of K2Ni2TeO6acquired at the base temperature (T =

2 K). Main figure show the long time domain while the inset

focus on the shorter time-scale. . . 97

(sym-bols) and corresponding fit (solid black lines) for K2Ni2TeO6.

11.9 Magnetic order parameter of K2Ni2TeO6showing the

temper-ature dependence of the frequencies, fAF1 and fAF2 vs

tem-perature extracted from the fits to the ZF µ+SR data. Here the

frequency is related to the angular frequency as: f = ω/2π.

Solid lines are guides to the eye. . . 99

12.1 2D Intensity map of the LA bank with temperatures ranging from 7.5K to 300K. The arrows pointing at magnetic peaks. The left arrow pointing at two peaks clustered together. . . 102

12.2 Temperature dependence of two magnetic Bragg peaks in K2Ni2TeO6

detected at low Q using the LA bank. . . 103 12.3 Magnetic order parameter plotted as the normalized area of

magnetic peak intensity for temperatures ranging from 7.5K to 51.2K. Here the different curves relates to peaks in the LA

banks with LA Peak 1 at Q = 0.69 Å−1and LA peak 2 at Q =

0.74 Å−1as well as the QA bank with a peak at Q = 1.84 Å−1. 104

12.4 Neutron Diffraction for the magnetic structure. Data in blue, refinement as red solid line and the atomic (black circles) as well as magnetic (red triangles) peak positions, together with the difference (green). . . 105 12.5 Three different proposed antiferromagnetic spin structure of

K2Ni2TeO6. . . 107

(19)

LIST OF FIGURES xvii

13.2 Temperature evolution of NPD data of K2Ni2TeO6, using the

BS bank. A clear ’split’ of several Bragg peaks are revealed

for T ≤ 450 K. . . 109

13.3 Individual NPD pattern for K2Ni2TeO6as a function of

temper-ature clearly showing the emergence of a new atomic structure

for T ≤ 450 K. . . 111

13.4 Bravais lattices for orthorhombic lattice system. . . 112

13.5 NPD data from K2Ni2TeO6acquired at T = 550 K data using

the #6: Monoclinic structure listed in Table 13.1. . . 113

13.6 NPD data from K2Ni2TeO6acquired at T = 550 K data using

the #4: Orthorhombic structure listed in Table 13.1. . . 114

13.7 Crystal structure of K2Ni2TeO6for the highest recorded

tem-perature (T = 550 K). (a) 3D view (b) View along the c-axis,

clearly showing the honeycomb around tellurium . . . 114

13.8 Lattice parameters a and c and the cell volume V as a function

of temperature. . . 117

14.1 µ+SR time spectra (symbols) and asymmetry fit (solid line)

for Li2Ni2TeO6at 100 K and 550 K. Here black symbols are

ZF, red LF = 5 G, and green LF = 20 G. . . 119

14.2 Temperature dependence of the fit parameters for Li2Ni2TeO6

including (a) Asymmetry (b) Field distribution width, and (c) Relaxation rate. . . 120 14.3 Temperature dependence of the ion hopping rate, ν(T ) for the

Li2Ni2TeO6compound. Dashed line is a fit to the Arrhenius

equation that yields the activation energy of the Li-ion self

diffusion process as Ea ≈ 363 meV. . . 121

for Na2Ni2TeO6at 250 K and 550 K. Here black symbols are

ZF, red LF = 5 G, green LF = 10 G and blue LF = 20 G. . . 122

14.5 Temperature dependence of the fit parameters for Na2Ni2TeO6

including (a) Asymmetry and (b) Relaxation rate. . . 123 14.6 Temperature dependence of the ion hopping rate, ν(T ) for the

Na2Ni2TeO6compound. Dashed line is a fit to the Arrhenius

equation that yields the activation energy of the Na-ion self

diffusion process as Ea ≈ 207 meV. . . 124

for K2Ni2TeO6 at 100 K and 550 K. Here black symbols are

(20)

xviii LIST OF FIGURES

14.8 Temperature dependence of the fit parameters for K2Ni2TeO6

including (a) Asymmetry and (b) Relaxation rate. Note that KT1 component are shown in blue symbols and KT2

compo-nent in red symbols. . . 126

14.9 Temperature dependence of the ion hopping rate, ν(T ) for the K2Ni2TeO6 compound. Dashed line is a fit of KT2 hopping rate to the Arrhenius equation that yields the activation en-ergy of the Na-ion self diffusion process as Ea ≈ 155 meV. Note that KT1 show a more or less temperature independent hopping rate. . . 127

16.1 Me at the S1 muon beamline of J-PARC/MLF! (13thJune 2019)133 17.1 SPICA, June 2019. From left: Professor Yasmine Sassa, Kon-stantinos Papadopulous, Dr. Nami Matsubara, me (Anton Zubayer), researcher Dr. Titus Masese, Elisabetta Nocerino, Ola Kenji Forslund and Professor Martin Månsson. . . 136

17.2 Dinner with the Director of J-PARC, Professor Naohito Saito, June 2019. . . 136

A.1 An overview of the µSR S1 in J-PARC, Japan. . . 147

A.2 The cryostat with the sample inside. . . 148

A.3 Hatch controllers . . . 149

A.4 How to mount the µSR. . . 151

A.5 Valves for compressor. . . 152

A.6 Compressor interface. . . 153

A.7 Helium tube mounted. . . 154

(21)

List of Tables

2.1 Table of compounds similar to K2Ni2TeO6. . . 10

8.1 Parameter table of Li2Ni2TeO6. . . 60

8.2 Structure parameters for Li2Ni2TeO6with a Bragg R-factor of

4.25. . . 61

8.3 Parameter table of Na2Ni2TeO6. . . 63

8.4 Structure parameters for Na2Ni2TeO6with a Bragg R-factor of

23.1. . . 64

8.5 Parameter table of K2Ni2TeO6. . . 65

8.6 Structure parameters for K2Ni2TeO6with a Bragg R-factor of

17.3. . . 67

9.1 Parameter table of Li2Ni2TeO6. . . 72

9.2 Structure parameters for Li2Ni2TeO6with a Bragg R-factors of

11.1, 9.92 and 6.56. . . 72

9.3 Parameter table of Na2Ni2TeO6. . . 75

9.4 Structure parameters for Na2Ni2TeO6 with a Bragg R-factors

of 3.64, 3.39 and 3.88. . . 76

9.5 Parameter table of K2Ni2TeO6. . . 78

9.6 Structure parameters for K2Ni2TeO6with a Bragg R-factors of

7.52, 6.10 and 4.34. . . 79

9.7 Comparison (XRD vs. NPD) of cell parameters for Li2Ni2TeO6. 80

9.8 Comparison (XRD vs. NPD) of cell parameters for Na2Ni2TeO6. 80

9.9 Comparison (XRD vs. NPD) of cell parameters for K2Ni2TeO6. 80

10.1 Comparison of the magnetic properties of the three different

compounds, as extracted from bulk magnetic measurements. . 89

11.1 Comparison of Néel temperatures for MPMS/PPMS and LT

µ+_{SR. . . 100}

(22)

xx LIST OF TABLES

13.1 Possible atomic structures as extracted from the Conograph software. Note that #0 is the room temperature hexagonal

structure that is listed only as a reference. . . 112

(23)

LIST OF TABLES 1

Abbreviation/Acronym Explanation

XRD X-Ray Diffraction

XAS X-ray Absorption Spectroscopy

MPMS Magnetic Properties Measurement System

PPMS Physical Properties Measurement System

NPD Neutron Powder Diffraction

µ+_SR _{muon Spin Rotation/Relaxation/Resonance/Research/etc.}

J-PARC Japan Proton Accelerator Research Complex

KEK High Energy Accelerator Research Organization

MLF Materials and Life Science facility

RCS Rapid Cycling Synchrotron

CROSS Comprehensive Research Organization for Science and Society

PSI Paul Scherrer Institute

ISIS neutron and muon source

SQUID Superconducting QUantum Interference Device

AC/DC Alternating Current / Direct Current

FC Field Cool

ZFC Zero Field Cool

LA/QA/BS/HA/SA Low-/90 degree-/Back Scattering/High-/Small- Angle

PID Proportional-Integral-Derivative

TOF Time Of Flight

KT Kubo-Toyabe

ZF Zero Field

LF Longitudinal Field

wTF weak Transverse Field

KTH Kungliga Tekniska Högskolan (Royal Institute of Technology)

HT high temperature LT low temperature AFM antiferromagnetic FiM ferrimagnetic FM ferromagnetic PM paramagnetic

(24)

Chapter 1 Introduction

The ever-changing world requires an ever-increasing quantity of batteries since vehicles and power plants based on fossil fuel are rapidly getting dismantled or transformed into environmental friendly and renewable electrical alternatives. From wind- and hydro-plants to electric cars, batteries are an essential part to store energy. As the efficiency of such electrically driven products increase, the demand for higher energy storage capacity also increases but at the same time the batteries have to be cost effective, yet environmental friendly, in order to compete with its rival, the fossil fuel.

During the last couple of years lithium-ion (Li-ion) batteries have taken the world by storm and even resulted in a Nobel Prize in Chemistry in 2019, given to John B. Goodenough, Stanley Whittingham and Akira Yoshino.[1] This discovery and application of Lithium-ion batteries really shifted the world to-wards a much more environmentally friendly place and has had a great impact on science and Technology itself. However, lithium and today’s transition-metals used in batteries are very limited and most of them are located and extracted from developing countries that faces political and economical con-flicts[2], as seen in Figure 1.1, for lithium. Lithium is not conveniently avail-able or evenly distributed around the world and hence quite an expensive metal, as seen in Figure 1.2.

(25)

CHAPTER 1. INTRODUCTION 3

Bolivia

40%

Chile

22%

Global Lithium Reserves

Lithium Resources [arb. units]

Max Min

Figure 1.1: Lithium resources in the world clearly showing the uneven distri-bution between different continents and countries. [3]

If we want to continue along the path of finding more convenient, sustain-able and cheaper batteries, new battery materials are necessary. Therefore, we conducted research on the potential of sodium (Na) -ions and potassium (K) -ions in batteries. Na and K reserves in the world are at least two orders of magnitude larger than of lithium.[4] The Na/K metals can be found in numer-ous countries and are more evenly distributed in the earth’s crust and they are also cheaper and easier to extract [5]. There is also a considerable concentra-tion of Na/K in the normal sea water of our great oceans, which makes Na and K available across the entire globe. In short, sodium-ion and potassium-ion batteries seems to solve the problems stated about the lithium-ion batteries. However, sodium and potassium are heavier. On the contrary, the weight does not affect stationary nor large-scale movable batteries. Power-grids beneath large power plants could make great use of huge stationary batteries in order to distribute electricity in an even manner, contrary to present high and low season distribution. These power-grids are facing a so-called reverse salient due to not being able to store large quantities of energy.[6] Regarding

electro-chemical behaviours, Na+and K+may not be the most suitable for

intercala-tion for certain carbon structures in the anode material. Therefore, a suitable anode also needs to be found for Na and K ions.[7][8]

(26)

4 CHAPTER 1. INTRODUCTION

Year

Lithium Price (USD/me

tric t

on)

Figure 1.2: Evolution of the Lithium price per metric ton clearly showing the drastic increase during last decade. [9]

In a recent article in Nature Communications by Masese et.al.[10], an

in-vestigation was carried out for the novel potassium compound K2N i2T eO6,

where it was verified that potassium-ions are a promising candidate to re-place the lithium-ions in batteries. They performed; powder X-ray diffraction (PXRD) to characterise the crystal structure, electrochemical measurements revealing high volumetric energy density and X-ray absorption spectroscopy (XAS) to check the oxidation and reduction process during potassium-ion ex-traction/insertion. Further, Masese et.al. also synthesized an entire series of novel honeycomb layered tellurates with various alkali metals, such as lithium (Li), sodium (Na) and potassium (K).

A more fundamental understanding of the materials and their potential as battery materials had to be addressed. Magnetism can elucidate various be-haviours of a compound, such as magnetic ordering and transition tempera-tures. These characteristics can for example reveal if they are viable super-conductors or showcase characteristics other than of battery potential. Hence, it was of great interest to us to analyze the magnetic properties, by investigat-ing their behaviour at low temperatures, and additionally solidify if Na-ions and K-ions are a good replacement for Li-ions in batteries, by investigating its diffusive properties. There were a multitude of different experimental ap-proaches to characterise materials that granted those answers and an improved understanding. Thus, the following four characterising techniques were exe-cuted:

(27)

CHAPTER 1. INTRODUCTION 5

• X-Ray Diffraction (XRD) measurement, to examine the quality of the sample after transportation and obtain a crystal structure.

• Magnetic measurement, as a function of temperature, using MPMS (Mag-netic Properties Measurement System), which provides us the sensitivity of a Superconducting QUantum Interference Device (SQUID) in order to analyse the magnetic susceptibility.

• Neutron Powder Diffraction (NPD) measurement, in order to investigate the crystal and magnetic spin structure with temperature dependence.

• Muon Spin Rotation (µ+SR), to obtain a better understanding of the

diffusive properties and low temperature magnetic behaviour.

The neutron powder diffraction and muon spin rotation measurements were carried out at the particle-accelerator facilities J-PARC in Japan [11], PSI in Switzerland [12] and ISIS in UK [13].

1.1 Research aims

The research was conducted to strengthen the potential of honeycomb lay-ered cathodes as battery materials and gain a fundamental understanding of the compounds.

The motivation is to further prove the potential of sodium and potassium as carrier ions for cathode materials in batteries, especially for the honeycomb tellurates. Verifying the layered structure that is essential for a good battery material, using XRD and NPD. Revealing the ion dynamics and calculate the

activation energy using µ+SR.

Another motivation for the investigation is a fundamental understanding of honeycomb layered materials. Investigating the magnetic characteristics with temperature dependence for low temperatures, using NPD, MPMS and low

temperature µ+SR. Such characteristics were for example: lattice expansion

with temperature, magnetic spin structure, antiferromagnetic behaviour and

(28)

Chapter 2 Honeycomb Layered Materials

A battery consists of an anode, a cathode and an electrolyte. Figure 2.1 shows such a configuration connected to a light bulb. When charging the battery, carrier ions travel from the cathode through the electrolyte and towards the anode. The unbalanced cathode then needs to release the excess electrons in order to stabilize. The opposite occurs during discharge. Extensive research of all three components are being made and special investigation is carried out for the next viable cathode material. In order for a battery to charge, it is crucial that ions inside the cathode material efficiently can travel towards the anode material and vice versa in order to distribute the energy to any given us-age of electricity.[14] The ion traveling efficiency is what is being thoroughly studied in order to discover new solutions. A solid state cathode material usu-ally consists of an alkali metal, one or multiple various transition metal/s and oxygen. The alkali atoms act as the travelling ions when charging or discharg-ing. As the ions reach the anode, electrons are released and travel through the electrical device’s circuit. Hence, providing the device with electricity. The transition metals act as stabilizers, which means a formation of suitable struc-tures and the capability to form a redox pair in the operation voltage window (e.g. the change of the Co oxidation state from Co3+ to Co4+). The redox-pair possibility is why these transition metals are suitable at all and the carrier ion is able to exit and come back into the cathode structure.

(29)

CHAPTER 2. HONEYCOMB LAYERED MATERIALS 7

Charge

Dis-charge

Li

+

Figure 2.1: Schematic view of an archetypical commercial Li-ion rechargeable

battery with LiCoO2 cathode, electrolyte/separator and graphite anode.

Dur-ing charge/discharge cycles the Li+ions move between the anode and cathode

through the electronically insulating electrolyte with electron transport occur-ring in the outer circuit. [15]

In present day, LiCoO2is the most common cathode compound [16], even

though lithium is very limited and cobalt is toxic. Lithium (Li) is the alkali element that serves as the moving ions and cobalt (Co) serves as the transition metal that stabilizes the compound during ion diffusion. As seen in Figure 2.2, the transition metal along with the oxygen at the many corners, can be viewed as slabs that sandwich the ions and keep them between the slabs. This layered structure is crucial for quick ion mass-transfer into and out of the cathode ma-terial along 2-dimensions. Low dimensional cathode mama-terials are the most efficient for batteries since the ions can then be "directed" towards the anode quickly when a voltage is induced, rather than moving in a transverse direction and slowing the process.

(30)

8 CHAPTER 2. HONEYCOMB LAYERED MATERIALS

Figure 2.2: Crystal structure of multiple unit cells of a typical 2D layered en-ergy material where the alkali ion layers are sandwiched between the transition metal oxide (TMO) layers. One example of such material is the well-known

battery cathode material LiCoO2where alkali ions are lithium and TMO layers

consists of CO2. [3]

Masese et.al. synthesized a series of honeycomb layered tellurates and ex-ecuted electrochemical measurements to elucidate a possible potential for this

series of compounds. Among the compounds were the series of A2Ni2TeO6,

where A = Li/Na/K.[10] Meaning the first three alkali metals in the periodic table, which are known to be carrier elements for battery materials. The hon-eycomb structure is commonly found in various parts of nature and is simply a hexagonally shaped structure. The hexagonal layers mimic the traditional honeycombs we find in nature.

(31)

CHAPTER 2. HONEYCOMB LAYERED MATERIALS 9

(a)

(b)

K/Na

Te

O

Te

Ni

Figure 2.3: (a) Crystallographic view of the P2-type-layered crystal structure

of A2Ni2TeO6 (A = Na and K) along the c-axis. The prismatic polyhedra

around the K ions are shown for simplicity. (b) A fragment of the P63/mcm

structure of K2Ni2TeO6in the ab plane (the magnetoactive honeycomb layers)

with three major K sites (K1, K2, and K3). Samples are produced by our close collaborators at the National Institute of Advanced Industrial Science and Technology (AIST) in Japan. [10]

As seen in Figure 2.3, the golden colored potassium or sodium are layered between the slabs of nickel, thus shaped in a potential cathode material. The transition metals is nickel (Ni). Tellurium is what is shaping the structure into a honeycomb formation. With a tellurium atom at the center of a hexagonal and the actual hexagonal surrounding the tellurium consists of six nickel atoms. Previous studies and electrochemical measurements have revealed great po-tential for sodium and potassium ions in this honeycomb series. Tellurium however, is a rare and toxic element, but the use of tellurium in the study is to highlight the potential of honeycomb layers with sodium and potassium. The tellurium is a great stabilizer and is therefore used in that regard only. The tellurium is easily exchangeable with for example tungsten (W), yet for investigation purposes the tellurium is a better stabilizer. The potassium

com-pound K2Ni2TeO6is difficult to synthesise due to moisture sensitivity which

can cause phase separation and impurity phase. Thus extra precautions were made for that compound.

(32)

10 CHAPTER 2. HONEYCOMB LAYERED MATERIALS

Below is a table of some similar compositions with their symmetry, lattice parameters and for some composition also their Néel temperatures:

Composition Symmetry a [Å] b [Å] c [Å] TN [K] ref

Li2Ni2TeO6 F DDD: 2 5.961 8.805 17.67 - [17] Li2Cu2TeO6 C2/m 5.644 8.607 5.352 - [18] Li4NiTeO6 C2/m 5.160 8.895 5.139 1.8 [19], [20] Na2Ni2TeO6 P63/mcm 5.207 5.207 11.15 26 [21], [22], [23] Na4NiTeO6 P6322 5.228 5.228 11.19 19 [24], [25] Na4CuTeO6 C2/c 5.667 9.917 10.45 2.5 [26] Na2Zn2TeO6 P6322 5.287 5.287 11.29 - [27] K2Co2TeO6 P6322 5.243 5.243 12.42 - [10] K2Mg2TeO6 P6322 5.286 5.286 12.54 - [10]

(33)

Chapter 3 Japan Proton Accelerator

Research Complex (J-PARC)

3.1 Overview

Neutron and muon measurements were executed at the particle-accelerator fa-cility J-PARC in Japan.[11] In order to utilize the instruments, one has to apply for beamtime and achieve acceptance before experimenting. Prior to experi-ments, safety, training and regulations are necessary to conduct any form of activity inside and to get the pass/ID-card granted. A schematic view of J-PARC can be seen in Figure 3.1.

Figure 3.1: J-PARC, Japan Proton Accelerator Research Complex [11]. The neutron experiments of this MSc thesis were mainly performed at the Materials and Life Science Experimental Facility (MLF).

(34)

12 CHAPTER 3. JAPAN PROTON ACCELERATOR RESEARCH COMPLEX (J-PARC)

The entire facility is jointly owned by J-PARC, KEK[28] and universi-ties. J-PARC, in its essence, is a high intensity proton accelerator facility. We performed the experiments in the Materials and Life Sciences experiment di-vision (MLF). The accelerator and the Rapid Cycling Synchrotron (RCS) are treated, by aiming a beam of high energy protons to concurrently implant into carbon to produce muons, and, through collision with mercury, to produce neutrons. A schematic view for the muon production is detailed in Fig. 3.2(a). The XRD and MPMS instruments were located in the CROSS laboratories next to J-PARC.[11]

Traditionally, reactor sources have been used to produce neutrons for neu-tron scattering experiments [29], e.g. at the ILL in France [30] and FRM-II in Germany [31]. However, at present time both politics and safety issues have resulted in a change and most new neutron sources are based on the so-called spallation process [32] that requires the use of a high-energy proton accelera-tor and a heavy target [see Fig. 3.2(b)]. Such facilities are still rather rare and only a handful exists in the world. However, many are currently under devel-opment, for instance in Sweden where the European Spallation Source (ESS) is currently being constructed [33]. J-PARC is indeed one of the most recent spallation sources that uses a pulsed neutron beam and a mercury (Hg) tar-get. J-PARC is in fact the accelerator facility with the highest intensity, which makes it into a one of the world’s most formidable research facilities [34].

The funding of the facility was a joint project, with the main contributors being J-PARC and KEK. The development cost estimates up to 1.9 billion dollars and still counting, since improvements, new beamlines and upgrades are being made.

(35)

CHAPTER 3. JAPAN PROTON ACCELERATOR RESEARCH COMPLEX (J-PARC) 13

Figure 3.2: (a) Schematic view of muon production where the high-energy protons collide with the carbon target, producing a pion that then decays into a muon and a neutrino. The muons are then transported through the beamline

to the experimental station where the muon spin rotation (µ+SR) experiments

are performed. (b) Schematic view of the spallation process where the high-energy protons collide with the heavy (Hg or W) target to produce neutrons. The energy of the neutrons are then moderated before they are transported through the neutron beamline (via superrmirrors) to the experimental station.

(36)

Safety is understandably one of J-PARC’s main concerns,[35] [36] and the facility enforces firm safety regulations and restrictions both for the sake of the facility and for the users. After the safety training, a dosimeter has to worn by each user, which still have to pass through devices measuring your radiation upon each exit from the experimental halls. Separate shoes also have to be worn inside the facility, since the dust on the floor may contain active dust. In labs, special coats and yellow shoes have to be borrowed and worn. Panic buttons and toxic waste boxes are situated in numerous areas and corners. Ex-tremely strict procedures have to be followed before any action, like opening a hatch or opening a shutter, can be taken. Finally, Geiger counters are imple-mented to measure samples after each experiment and no active samples are allowed to leave the facility.

(a) Safety. (b) Waste.

Figure 3.3: Safety procedures.

Figure 3.3a shows the safety counters between the restricted area of MLF and the non-restrictive area of MLF. 3.3b shows the secure garbage bins that could contain radioactive or toxic waste. Strict identification processes were mandatory. The ID-pass obtained after successful safety training was scanned when entering any sections of the facility and re-scanned in specific areas. This ID-pass was also used to enter the outer security gates, by showing you ID to respective guards. When entering laboratories, the users needed to sign in with handwritten signature, time, date, affiliation and additional informa-tion required, and sign out when leaving. Security cameras were installed in a manner that total surveillance was possible for every area in the entire facility. The neutron diffraction and µSR experiments were conducted at the

(37)

Mate-CHAPTER 3. JAPAN PROTON ACCELERATOR RESEARCH COMPLEX (J-PARC) 15

rial and Life Science Experimental Facility (MLF), as seen in Figure 3.1. The large building consists of an office, locker room, security station, laboratories, and two large experimental halls. The halls, Hall 1 and Hall 2 [Figure 3.4], are mirror images of each other and both halls contains both neutron and muon beamlines as well as the respective experimental stations.

(a) Hall 1. (b) Hall 2.

Figure 3.4: Experimental Halls 1 and 2 within the Material and Life Science Experimental Facility (MLF) of J-PARC.

3.2 Beamlines

All of the beamlines are located within the above mentioned experimental halls. The numerous beamlines all have different properties depending on the location of the beamline (which part of the target/moderator the ’see’) and the properties of the neutron guides as well as the final experimental station. In a way, the two experimental halls can be viewed as two compartments of a toolbox that contains a set of different tools that all have specific purposes. Such toolbox also covers a a very wide scope of scientific areas, covering bi-ology, chemistry, material science, engineering, fundamental physics, mag-netism and medicine. The experiments presented in this MSc thesis were con-ducted in both of the experimental halls. For the neutron powder diffraction measurements, they were executed at the SPICA beamline. For the µSR, ex-periments were conducted at the S1 beamline. As a learning experience, I also joined experiments at the HRC (single crystal inelastic neutron scattering), iMATERIA (neutron powder diffraction) and SOFIA (neutron reflectometry) beamlines. An overview of the beamlines available at MLF can be seen in Figure 3.5.

(38)

Figure 3.5: Overview of all the neutron and muon beamlines/instruments that are available at MLF/J-PARC. For this thesis, mainly S1 line (muon) and SPICA (neutron) instruments were utilized. [37]

3.3 Other Neutron & Muon Sources

As mentioned, most of the data that I collected for this MSc thesis originates from the J-PARC facility. However, complementary data were also collected at the Swiss Muon Source (SµS) at the Paul Scherrer Institute (PSI) in Switzer-land [12] as well as at the ISIS pulsed neutron and muon source in UK [13]. Such measurements were, however, collected by other members of the SMaRT research group and therefore I do not describe such facilities in great detail within this thesis. Briefly, ISIS is in similarity to J-PARC/MLF a pulsed spal-lation source having both neutron and muon beamlines. The pulsed source uniquely gives the possibility to mainly study ion diffusion using muons. On the contrary, SµS at PSI is a continuous spallation source, which gives us a much better time-resolution for the muon experiments. Such facility is hence better suited for studies of magnetically ordered samples at low temperatures.

(39)

CHAPTER 3. JAPAN PROTON ACCELERATOR RESEARCH COMPLEX (J-PARC) 17

3.4 Sample preparations using glovebox

The handling and transportation of scientific samples should generally be per-formed with great care since they can be very sensitive. This is especially true for K-ion battery materials that tend to degrade when subjected to air and moisture. In order to avoid such effects and obtain reliable and reproducible experimental results, all such samples are packed, handled and prepared in a glove box. The glovebox is usually filled with either argon or helium [38] where helium is advantageous for low temperature measurements, while ar-gon is less expensive [39]. A typical glovebox used in this thesis is shown in Figure 3.6.

(a) Typical glovebox (b) Load-lock chamber (c) Purity monitor

Figure 3.6: Typical He(g) or Ar(g) glovebox used for handling sensitive battery materials. All samples measured by neutron powder diffraction and muon spin rotation measurements were sealed inside their respective sample can/cell inside such glovebox.

The load-lock (or exchange) chamber seen in Figure 3.6b was used to trans-fer any of the tools and materials needed for preparing the sample, into the glovebox. After inserting into the chamber and closing the shutter, the cham-ber was flushed three times with helium and vacuum in order to decrease any contamination. Note however that that any airtight closed bottles or packages needed to be opened before. This is to prevent contamination in the glovebox. When "flushing" respective sample in the exchange chamber, the air/moisture is replaced with helium and vacuum environment. Once flushed three times, the inner shutter was opened and the goods then received successfully. De-pending on the glovebox, the flushing procedures were different, on the basis of if the glovebox was analogue or digital. The digital monitor that keeps track of the purity of the glovebox atmosphere is shown in Figure 3.6c. Af-ter preparation depending on sensitivity, the samples were either stored inside

(40)

the glovebox or inside specific moisture free packages sealed using a vacuum packing machine and an impulse heat sealer. In sensitive cases, the sealed packages were further stored in outer vacuum boxes for double safety.

Three samples were examined in this thesis, namely Li2Ni2TeO6, Na2Ni2TeO6

and K2Ni2TeO6. Here the K-ion compound is much more moisture sensitive

than the Li/Na analogues [10]. However, to ensure systematic measurements and data, all samples were equally handled inside a glovebox during prepa-rations for the XRD (x-ray diffraction), NPD (neutron powder diffraction), magnetic bulk measurements (MPMS/PPMS) and muon experiments.

(41)

Chapter 4 X-Ray Diffraction

X-ray diffraction, also known as XRD, is a technique often utilized for inves-tigating the crystal (atomic) structure of a given compound.[40] The raw data obtained from performing an XRD measurement comes in the form of an in-tensity spectra as a function of scattering angle. The raw data can thereafter be fitted using Le Bail fit or Rietveld fit. Consequently, it is possible to gain knowledge of the cell parameters and, if Rietveld fitted, also the atomic place-ments i.e. the crystal structure. Chemists, material scientists and physicists all over the world uses this convenient technique since in-house x-ray diffractome-ters are reasonably inexpensive and compact enough to fit into any conven-tional laboratory. However, if higher precision is needed or only a very small sample amount is available, synchrotron XRD facilities are instead available [41]. XRD is also a potent technique to inspect the quality of the sample just before the experiment at the large-scale facility. This is to avoid wasting very expensive neutron/muon beamtime on a degraded sample. The purposes for the XRD-measurements in this thesis, were thus to verify the safe transporta-tion of the samples, by comparing with the results of XRD measurements prior to the transportation, and the atomic crystal structure, also serve as a reference and comparison for the complementary neutron diffraction measurements.

4.1 Theoretical background

X-ray diffraction is the elastic scattering of x-ray photons by atoms in a peri-odic lattice. When an x-ray strikes an electron, secondary spherical waves are emitted from the electrons. The scattered monochromatic x-rays that are in phase give constructive interference, where the directions can be determined

(42)

20 CHAPTER 4. X-RAY DIFFRACTION

using the well-known Bragg’s law:

2dsinθ = nλ, (4.1)

where d is the distance between parallel atomic lattice planes, θ is the angle of between the direction of the ray and the lattice/plane, n is a positive integer and λ is the wavelength of the ray (see also Fig. 4.1).

Figure 4.1: Bragg diffraction from two atomic planes in a material [42]. X-rays with wavelengths in the same order of magnitude to the distance d are used. X-rays are defined as a pretty wide electromagnetic wavelength

spectrum (from ∼ 0.01 nm to ∼ 10 nm).[40] XRD and NPD complements

each other in numerous ways. XRD measurements can be carried out in a shorter amount of time compared to NPD. Yet, since XRD’s less penetrative characteristics, XRD is a less bulk sensitive technique. Further, heavier el-ements can be over-represented (in intensity) compared to lighter elel-ements because of the higher number of electrons. For sensitive samples (especially biological and organic materials) x-rays can also damage (’burn’) the sam-ples, which can cause unreliable results. Finally, since XRD scatters with the electron clouds, while NPD scatters with the nuclei points, the XRD has a stronger Q-dependency than NPD (see also Fig. 6.1 below in the chapter on neutron scattering). Therefore, XRD may be advantageous for low Q-ranges while NPD could investigate higher Q-ranges. [43]

4.2 Experimental setup

4.2.1 Sample Preparation

The samples were opened and poured onto a mortar and grinded into a smooth powder form as seen in Figure 4.2a. The powder was layered onto a thin

(43)

in-CHAPTER 4. X-RAY DIFFRACTION 21

dented glass plate as seen in Figure 4.2b. Only the indented part of the glass contained the powder and the thickness of the layer was made as homogeneous as possible. For the sensitive samples, all preparations were conducting inside a glovebox with controlled atmosphere. These preparations and the following measurements were performed in the CROSS [44] and KEK [28] laboratories.

(a) Mortar with powder. (b) Powder on thin indented glass plate.

Figure 4.2: Sample preparation for x-ray diffraction (XRD) experiments. For sensitive samples these procedures were performed inside an Argon or Helium glovebox.

4.2.2 Execution

XRD measurements were performed in the KEK laboratories [28] using an XRD machine manufactured by Rigaku [45] with Cu-Kα radiation having a wavelength λ = 1.54 Ångström (Å). The Rigaku SmartLab is an automatic XRD instrument that changes angle (2θ) for each iteration, thus we included angles (2θ) ranging between 5.5 and 90 degrees in a smooth fashion. The glass plate with the sample was mounted inside the XRD as seen in Figure 4.3 and then the measurement was automatically started and performed.

(44)

22 CHAPTER 4. X-RAY DIFFRACTION

Figure 4.3: XRD Machine by Rigaku in the KEK laboratories [28].

4.2.3 Data management & fitting

Data was saved as a .dat file. The data file contained to columns with data. One of the columns contain data with units in degrees (2θ) and is therefore

converted to units of Å−1 (Q) using the equation:

Q= 4π

λ sin(

2θ

2 ), (4.2)

where λ is 1.54 Å and θ is the scattering angle. The other column represents intensity in an arbitrary unit. By employing the commonly used software Full-Prof [46] conversion of the data from 2θ to Q-range was made, fitted the data, using Le Bail fit and simulated the Bragg positions. Lastly, the data was saved as a four column .XYY file of which my own MATLAB script handled in order to plot the intensity spectra. Furthermore the software VESTA [47] was im-plemented in order to visualize the crystal structure, using the .cif file created by FullProf after the fitting procedure was executed.

(45)

Chapter 5 Magnetic Bulk Characterizations

Using MPMS/PPMS

Bulk magnetic susceptibility measurements are utilize in order to extract the basic magnetic properties of a given sample. Furthermore, Super Conduct-ing QUantum Devices (SQUID) are known to reveal the smallest of magnetic fields due to their exceptional sensitivity to magnetic fields.[48] A Magnetic Properties Measurement System (MPMS) implements such a SQUID in order to measure magnetization in regards to temperature. An MPMS offers ex-ceptional sensitivity, field uniformity, reputability and reliability for magnetic measurements. The purposes of the MPMS were to elucidate the temperature dependence of the magnetic susceptibility and thereby obtain a first view on the possible magnetic order within the samples. In a way, MPMS measurements of magnetic materials can be compared to the in-house XRD measurements of the atomic structure, i.e. it is the first ’look’ of your material’s physical proper-ties. In addition to MPMS, there is a closely related setup called the Physical Properties Measurement System (PPMS). The PPMS setups also include the possibilities to measure electrical resistivity and heat capacity.

5.1 Theoretical background

AC susceptibility measurements were conducted using the Reciprocating Sam-ple Option (RSO) protocol (see Fig. 5.1), which differs from classical DC mea-surements since it implements a small amplitude periodic oscillation of the sample. This movement culminates in an oscillating AC that is detected by the Superconducting QUantum Interference Device magnetometer (SQUID).[48]

(46)

24 CHAPTER 5. MAGNETIC BULK CHARACTERIZATIONS USING MPMS/PPMS

Figure 5.1: Measurement of magnetic susceptibility using the Reciprocating Sample Option (RSO) protocol with a very small amplitude and a Supercon-ducting QUantum Interference Device (SQUID) magnetometer.[49]

A SQUID is an apparatus that consists of a superconducting ring inter-rupted at two points by a normally conducting or electrically insulating ma-terial.[50] Magnetic flux flows through the superconducting ring with a size

of an integral number multiple of the magnetic flux quantum Φ0 = 2.07 ·

10−15

Vs. When the magnetic field varies, an electric circuit current is excited in the ring to the nearest multiple of the flux quantum. In the superconducting ring there are two Josephson contacts that passes a current through the SQUID and thus a measurable voltage through the SQUID. The critical current oscil-lates as a function of the applied flux. Using the techniques of the SQUID, the MPMS is thus highly sensitive to even very tiny magnetic fields.

Magnetic spin order in materials is a very extensive topic and there is no room for a complete description. However in the most fundamental repre-sentation we can divide magnetic order in four basic states that are schemat-ically described in Fig. 5.2. In a paramagnetic (PM) phase the spins are dis-ordered, meaning that the spin points in random directions.[48] In a ferrimag-netic (FiM) phase the spins are aligned antiparallel in magferrimag-netic domains, but do not cancel due to the sizes of the spins being distinct. In a ferromagnetic (FM) phase the spins are all aligned parallel to each other in magnetic domains. Finally, in a antiferromagnetic (AFM) phase the spins are aligned antiparallel to each other.

(47)

CHAPTER 5. MAGNETIC BULK CHARACTERIZATIONS USING MPMS/PPMS 25

Paramagne!sm (PM) Ferrimagne!sm (FiM)

Ferromagne!sm (FM) An!ferromagne!sm (AFM)

Figure 5.2: Schematic view of the most fundamental types of magnetic spin ordering that can be found in materials.

Depending on how the spins are correlated (i.e. strength and sign of their magnetic coupling constants J ) determine how susceptible the spins are to an external magnetic field. Consequently, depending on the type of magnetic in-teractions and magnetic ground state the temperature dependence of the mag-netic susceptibility, χ(T ), will look different for the four types of magmag-netic order mentioned above. In Figure 5.3 the typical χ(T ) curves for PM, FM and AFM are shown. The FiM case is slightly special but in general it’s χ(T ) curve is similar to the FM case.

(48)

26 CHAPTER 5. MAGNETIC BULK CHARACTERIZATIONS USING MPMS/PPMS

Figure 5.3: Schematic temperature dependence of the magnetic susceptibility

χ_{(T ) for different types of magnetic spin order (correlations).}

In this thesis the materials mainly display AFM order. Here the physical properties of interest are their Néel, temperature, the Weiss temperature and

the effective magnetic moment, µeff. The phase transition to the

antiferromag-netic state is known as a Néel transition and is denoted as TN. The Weiss

temperature, θ is related to the strength and type of interaction between mo-ments. Its sign depicts if this interaction aids to align adjacent moments in the same direction (positive = FM) or opposite on another (negative = AFM). Usually the temperature dependence of the magnetic susceptibility is fitted by the so-called Curie-Weiss law that is described by Equation 5.1 below:

χ= C

T _{− θ}, (5.1)

where C is the Curie constant. By applying the inverse, the Weiss temperature can be isolated as seen below:

θ = T − 1

χ · C, (5.2)

By linearly fitting the PM phase of the inverse susceptibility, two properties can be obtained.[48] From the intersection of the linear fit with the x-axis, the Weiss temperature θ can be obtained. Furthermore, if the Weiss temper-ature turns out to be negative, then the sample is anti-ferromagnetic below

ANTONZUBAYER -Astudybyneutrons,muonsandx-rays- IonDynamics&MagneticOrderin2DHoneycombMaterials

Ion

D

ynamics &

M

agnetic

O

rder in 2D

H

oneycomb

M

aterials

-

A study by neutrons, muons and xrays

-ANTON ZUBAYER

Ion Dynamics & Magnetic

Order in 2D Honeycomb

Materials

A study by neutrons, muons and xrays

-ANTON ZUBAYER

Abstract

Sammanfattning

Papers

Contents

List of Figures

List of Tables

Chapter 1

Introduction

Bolivia

40%

Chile

22%

Global Lithium Reserves

1.1

Research aims

Chapter 2

Honeycomb Layered Materials

Li

(a)

(b)

K/Na

Te

Te

Ni

Chapter 3

Japan Proton Accelerator

Research Complex (J-PARC)

3.1

Overview

3.2

Beamlines

3.3

Other Neutron & Muon Sources

3.4

Sample preparations using glovebox

Chapter 4

X-Ray Diffraction

4.1

Theoretical background

4.2

Experimental setup

4.2.1

Sample Preparation

4.2.2

Execution

4.2.3

Data management & fitting

Chapter 5

Magnetic Bulk Characterizations

Using MPMS/PPMS

5.1

Theoretical background