Novel Phases in Hetero-Epitaxial and Super-Oxygenated Thin Films of Complex Oxides by Hao Zhang a Thesis Submitted in Conformity

Novel Phases in Hetero-Epitaxial and Super-Oxygenated Thin Films of Complex Oxides

Hao Zhang

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Graduate Department of Physics University of Toronto

Novel Phases in Hetero-Epitaxial and Super-Oxygenated Thin Films of Complex Oxides

Hao Zhang Doctor of Philosophy Graduate Department of Physics University of Toronto 2018

In this thesis we study structural phase transition and defect structures in complex oxide thin films induced by either heterostructuring or superoxygenation, in an effortto understand their effects on the electronic and superconducting properties. Thin filmswere chosen for the ability to tune the heteroepitaxial strain, as well as their large surface-to- volume ratio. We focus on two families of oxides, including the cuprate superconductors

Y-Ba-Cu-O (YBCO), and the Ruddlesden-Popper iridates Srn+1IrnO3n+1.

We examine the effect of heteroepitaxial strain on the superconducting critical temperature (Tc) of YBa2Cu3O7−δ (YBCO-123) thin film. Strain-induced intergrowth of

CuO defect structures is seen in La2/3Ca1/3MnO3 (LCMO)/YBCO-123 bilayer films, and can account for the reduced Tc in such bilayers. Perovskite/YBCO-123/perovskite trilayers with either ferromagnetic LCMO or paramagnetic LaNiO3 (LNO) as clamping layers show similarly strong reduction of the Tc. The Tc reduction is much milder when orthorhombic PrBa2Cu3O7−δ (PBCO) is used as clamping layers instead. These results indicate that heteroepitaxial strain, rather than long-range proximity effect, is responsible for the long length scales of Tc reduction in LCMO/YBCO-123 heterostructures.

We carried out superoxygenation experiments on YBCO-123 films in order to search for novel YBCO phases. YBCO-123 films are annealed in high-pressure oxygen, incon- junction with Cu enrichment by solid-state diffusion. The annealed films show clear evidence of phase transformation to Y2Ba4Cu7O15−δ and Y2Ba4Cu8O16, increasing with

ii higher degree of Cu enrichment. Regions of exotic phases containing multiple CuO or Y layers are also seen in high-pressure annealed films. Our results demonstrate a novel route of synthesis towards discovering more complex YBCO phases.

We also carried out superoxygenation on Sr2IrO4 (SIO) films in an attempt to metallize the iridates via hole-doping. High-pressure oxygen annealed SIO films show a progressive drop in room-temperature resistivity of up to 3 orders of magnitude, and an evolution towards metallic behavior. The post-annealed films show transformation to SrIr1−xO3, increasing with annealing time, pressure, and reduced film thickness. The evolution towards metallicity is attributed to phase transformation, interstitial oxygens, and Ir vacancies. The Ir vacancies in the most phase-transformed film appears to be structurally-ordered along the c-axis. Our results demonstrate a novel method for hole- doping and phase-transforming the iridates.

iii Acknowledgements

First and foremost, I would like to express my deepest gratitude and appreciation to my supervisor, Prof. John Wei. Throughout my PhD studies, Prof. Wei has continuously supported, guided, and encouraged me both professionally and personally, for which I am forever grateful. He has been a great mentor and role model to me, and pushed me beyond my limitations over and over again. It has been a pleasure and privilege to be a member of Prof. Wei’s research group.

I would like to express my appreciation to current and former members of the Wei group, Igor Fridman, Chris Granstrom, and Charles Zhang for giving me invaluable assistance in the lab as well as constructive scientific insights. This research has benefited from the contributions of many undergraduate students who has worked with our group, including Tianmin Liu, Alexander Su and Ben Xu.

I would also like to thank my committee members Prof. Young-June Kim and Prof. Arun Paramekanti, who has contributed time and expertise, and provided valuable guid- ance towards the completion of my thesis. I thank the careful proof-reading and constructive comments on my thesis by Prof. Hae-Young Kee, Prof. Luyi Yang, and Prof. Alain Pignolet.

I would like to acknowledge and thank our collaborators. Prof. Gianluigi Botton and Dr. Nicolas Gauquelin at McMaster University performed microscopy imaging of our thin films; Prof. David Hawthorn and Christopher McMahon at University of Waterloo took x-ray absorption spectroscopy data; Prof. Thomas Gredig and Anh Nguyen at California State University Long Beach performed x-ray reflectivity measurements; Dr. Patrick Clancy took some x-ray diffraction (XRD) data for me and taught me many technical details about XRD; Some of the XRD data were taken by Dr. Abdolkarim Danaei and Dr. Raiden Acosta at Department of Material Science, and Dr. Srebri Petrov at Department of Chemistry. Prof. Ambrose Seo at University of Kentucky provided the

Sr2IrO4 samples used in the study. I thank Prof. Oscar Bernal, Prof. Guo-Meng Zhao

iv and his students Victor Aguilar, Carlos Sanchez, and Bo Truong at California State University Los Angeles for assisting me with Hall effect measurements. Finally I would like to thank Prof. Young-June Kim for kindly letting me use his high-pressure furnace, which is a valuable instrument essential for a significant portion of my research. I am indebted to my family for the emotional support they gave me throughout the years. I would like to thank my parents for their love and support, without which I would never be able to complete this challenging endeavor. I thank my wife Fei for her unconditional love and patience, for helping me staying sane, for giving me strength, and for her understanding and support in every possible ways. Last but not least, I wish to thank my daughter Muyao for giving me motivation to finish my thesis during the last, and the most stressful year of my PhD study. I feel extremely lucky to have them all in my life.

v Contents

List of Publications vi

1 Introduction 1

1.1 Cuprate superconductors and Y-Ba-Cu-O ...... 1

1.2 Half-metallic manganites ...... 7

1.3 Superconducting pairing symmetry and proximity effect ...... 11

1.4 Ruddlesden-Popper iridates ...... 15

1.5 Thesis outline ...... 17

2 Experimental Techniques 20

2.1 Pulsed-laser deposition ...... 20

2.2 Thickness and roughness measurement by X-ray reflectivity and atomic force microscopy ...... 25

2.3 High-pressure oxygen annealing ...... 30

2.4 Electrical transport measurements ...... 32

3 Attenuation of Superconductivity in Manganite/Cuprate Heterostruc- tures by Epitaxially-Induced CuO Intergrowths 39

3.1 Introduction ...... 39

3.2 Experiment ...... 41

3.3 Results ...... 41

vi 3.4 Discussions ...... 47 3.5 Conclusion ...... 48

4 Strain-Induced Tc Reduction in Heteroepitaxial Perovskite/YBa2Cu3O7−δ/ Perovskite Trilayers 50 4.1 Introduction ...... 50 4.2 Experimental ...... 52 4.3 Results ...... 53 4.4 Discussion and outlook ...... 58 4.5 Conclusion ...... 60

5 Synthesis of High-Oxidation Y-Ba-Cu-O Phases in Superoxygenated Thin Films 61 5.1 Introduction ...... 61 5.2 Experimental ...... 62 5.3 Results and discussions ...... 64 5.4 Conclusion ...... 77

6 Phase Transformation and Hole Doping in Superoxygenated Sr2IrO4 Thin Films 78 6.1 Introduction ...... 78 6.2 Experimental ...... 80 6.3 Results and discussions ...... 80 6.4 Conclusion ...... 94

7 Conclusions and Future Perspectives 95 7.1 Conclusions ...... 95 7.2 Suggested future experiments ...... 98

Bibliography 102

vii List of Tables

1.1 Table of some common multi-layer hole-doped cuprates that demonstrate

the Tc tends to scale with the lattice complexity ...... 4 1.2 Superconductor pairing symmetries allowed by Pauli exclusion principle . 12

3.1 Comparison of the ab-plane lattice parameters between YBCO-123, YBCO- 247, LCMO, and LSAT...... 48

4.1 Bulk lattice parameters for the oxides used in the trilayer study . . . . . 52

5.1 Known and possible phases of the YBCO family of cuprates ...... 62

viii List of Figures

1.1 Doping phase diagram of the cuprate superconductors ...... 3

1.2 Lattice structures of YBCO-123, YBCO-247 and YBCO-248 ...... 5

1.3 Phase diagrams of temperature versus oxygen pressure for bulk YBCO samples ...... 6

1.4 Density of states close to Fermi energy for a half-metal ...... 8

1.5 Illustration of the double exchange mechanism in mixed-valence manganites 9

1.6 Lattice structure of the perovskite LCMO ...... 10

1.7 A typical resistance versus temperature plot of LCMO ...... 11

1.8 Illustration of long-ranged proximity effect across a superconductor/ferromagnet junction ...... 13

1.9 Lattice structures of Sr2IrO4 and perovskite SrIrO3 ...... 15

1.10 The schematic energy diagram of the Ruddlesden-Popper iridates. . . . . 16

2.1 The schematic diagram of our PLD setup...... 21

2.2 A photograph of the inside of the PLD chamber during film growth with the plume showing...... 22

2.3 Comparison of AFM images of unoptimized and optimized YBCO-123 films 24

2.4 The experimental setup of typical XRR measurements ...... 26

2.5 Comparison between measured XRR data obtained on LNO/YBCO-123/ LNO trilayer films with the simulated XRR spectra ...... 27

ix 2.6 YBCO-123 film that were etched with HCl solution for thickness measurements ...... 29

2.7 A schematic diagram of the Morris Research HPS-5015 high-pressure furnace 30

2.8 Circuit diagram for four-terminal resistivity measurement using two lock- in amplifiers...... 33

2.9 Wiring configurations for van der Pauw measurement used for thin-film resistivity measurements ...... 34

2.10 The tip of our dipper probe for measuring electrical resistance as a function of temperature...... 35

2.11 Wiring diagram for Hall effect measurement with the PPMS . . . . . 37

3.1 HAADF-STEM images of a 25 nm/50 nm bilayer LCMO/YBCO-123 film grown on (001)-oriented LSAT substrate ...... 42

3.2 STEM image of a 25 nm unilayer YBCO-123 films grown on LSAT substrate showing negligible amount of double-CuO chains ...... 43

3.3 XRD pattern of a 25 nm/50 nm bilayer LCMO/YBCO-123 film grown on (001)-oriented LSAT substrate ...... 44

3.4 resistance versus temperature plots for unilayer YBCO-123 and bilayer LCMO/ YBCO-123 films ...... 46

3.5 A schematic diagram illustrating the difference in both lattice symmetry and the ab-plane lattice structure between YBCO-123, YBCO-247, LCMO, and LSAT ...... 47

4.1 XRD patterns of unilayer thin films of YBCO-123, PBCO, LCMO, and LNO taken with θ − 2θ scans ...... 53

4.2 Resistance versus temperature of PBCO/YBCO-123/PBCO trilayer films 54

4.3 Resistance versus temperature of LCMO/YBCO-123/LCMO trilayer films 55

4.4 Resistance versus temperature of LNO/YBCO-123/LNO trilayer films . . 57

x 4.5 Plot of Tc versus YBCO-123 layer thickness in oxide/YBCO-123/oxide trilayers ...... 59

5.1 Comparison of the STEM and XRD data between HP-annealed and as- grown YBCO-123 films ...... 65 5.2 STEM image taken on a 1-atm annealed YBCO-123 film ...... 66 5.3 Comparison of XRD data between different YBCO-123 films HP-annealed with different degree of Cu enrichment ...... 67

5.4 Comparison of normalized resistance versus temperature data for two high-pressure annealed YBCO-123 films with different degree of Cuen- richment ...... 68 5.5 Corrected and normalized XAS spectra taken on both as-grown and HP- oxygen annealed YBCO-123 films ...... 70 5.6 Evidence for phase formation to YBCO-125 and YBCO-126 in HP-annealed YBCO-123 films ...... 73 5.7 Evidence for double and triple Y defect in HP-annealed YBCO-123 films 74

6.1 Comparison of ρ vs. T data for SIO-214 films annealed under different conditions ...... 82 6.2 Hall effect and magnetoresistance measurements of a 7-day HP-annealed 20 nm SIO-214 film ...... 83 6.3 HAADF-STEM images of differently annealed SIO-214 films . . . . 85 6.4 Comparison of synchrotron XRD data taken on high-pressure annealed and 1 atm-annealed iridate films ...... 87 6.5 XRD patterns of the annealed SIO-214 films together with HP-annealed SIO-214 powder serving as control ...... 88 6.6 Normalized XAS spectra taken on annealed SIO-214 films together with the data for SIO-214 powder and STO substrate ...... 91

xi Chapter 1

Introduction

In this chapter we provide a brief introduction to the main families of materials that we study in this thesis, together with the main phenomena these materials exhibit. The Y-Ba-Cu-O family of cuprate superconductors is a central part of our thesis, and they are discussed in general terms in Section 1.1 with a focus on the CuO chains which are unique to Y-Ba-Cu-O. Section 1.2 describes the manganites, which is a half metal with spin-polarized charge carriers. When the half-metallic manganites are placed in contact with Y-Ba-Cu-O superconductors as heterostructures, an exotic kind of proximity effect involving spin-triplet pairing may arise. This spin-triplet proximity is the motivation for Chapter 3 and 4 of this thesis, and the theory of proximity effect is discussed in Section 1.3. The Ruddlesden-Popper iridates are described in Section 1.4. These iridates will be the focus of the study described in Chapter 6. Finally Section 1.5 gives a overview of the thesis.

1.1 Cuprate superconductors and Y-Ba-Cu-O

Before the discovery of lanthanum barium copper oxide (LBCO) by Bednorz and Müller in

1986 [1], no superconductor with a transition temperature Tc higher than 20 K was known.

It was commonly held that it’s theoretically impossible to have a Tc greater than ∼ 28

1 Chapter 1. Introduction 2

K, according to BCS theory with electron-phonon coupling [2]. This is because electron- phonon interaction is usually not strong enough to sustain pairing at high temperature.

Bednorz and Müller’s discovery of LBCO, which exhibits a Tc at 30 K, opened a door

to a whole new class of material known as the cuprate superconductors. YBa2Cu3O7

(YBCO-123) with the Tc of 93 K was subsequently discovered by M. K. Wu et al. [3],

and was the first superconductor with a Tc above the boiling point of liquid nitrogen.

Because of the unusually high Tc of the cuprate superconductors, they were also known as high-temperature superconductors (HTSC). Many other cuprate HTSC with more

complex lattice structure and higher Tc were found later. HgBa2Ca2Cu3O8 (HBCCO-

1223) with a Tc of 133 K under ambient pressure, and 150 K under hydrostatic pressure

is the current record holder of the HTSC with the highest Tc. The remarkably high Tc of the cuprate HTSCs cannot be explained by BCS theory with phonon-mediated pairing alone, and the origin of superconductivity in these material is under active investigation ever since cuprate HTSC was discovered.

All cuprates HTSC have layered structure containing quasi-2 dimensional CuO2 planes, which are the site where superconducting pairing happens. Because the electrical coupling between planes are generally weak, cuprate HTSC have highly anisotropic electrical properties. Without doping, the CuO2 planes are antiferromagnetic Mott insulators.

When right amount of holes or electrons are introduced to the CuO2 planes from the

charge reservoir layers, which are situated between the CuO2 planes, superconductivity is induced. A doping phase diagram of the cuprate HTSC is shown in Figure 1.1. There are more types of hole-doped cuprates than electron-doped cuprates, and the maximum

Tc tends to be higher in the hole-doped cases. For hole-doped cuprates with optimal

doping, the Tc tends to scale with the lattice complexity and the number (n) of charge reservoir layers per unit cell [7–9], until some critical value of n ∼ 3 is reached as shown in Table 1.1.

Out of many families of cuprate HTSCs discovered, we focus on the Y-Ba-Cu-O Chapter 1. Introduction 3

Figure 1.1: The doping phase diagram of the cuprate superconductors. In the figure AFI stands for antiferromagnetic insulator and SC stands for superconductor. This phase diagram is adapted from review papers such as [4–6].

(YBCO) family of cuprate in most of this thesis, because YBCO is widely studied with well-established methods for growing high quality epitaxial films. An unique feature of YBCO is the CuO chains. The 1-dimensional chains run along the b-axis, and make YBCO orthorhombic with a larger b-axis lattice parameter than the a-axis. These CuO chains give YBCO its unique orthorhombicity [10], in-plane anisotropy [11–13], oxygen variability [14], and rich phase diagram. Different arrangements of the chains give rise to different phases of YBCO. Figure 1.2 shows the lattice structures ofthe three most common members of the YBCO family with the chain layers highlighted.

YBa2Cu3O7−δ (YBCO-123) has single layer of CuO chain between adjacent BaO layers, and Y2Ba4Cu8O16 (YBCO-248) has double layers of CuO chains between adjacent BaO layers. Y2Ba4Cu7O15−δ (YBCO-247) can be thought as a superlattice of YBCO-123 and Chapter 1. Introduction 4

Family n Tc (K) BSCCO n = 1 (2201) 10 Bi2Sr2Can−1CunO2n+4 n = 2 (2212) 85 n = 3 (2223) 110 TBCCO n = 1 (2201) 85 Tl2Ba2Can−1CunO2n+4 n = 2 (2212) 105 n = 3 (2223) 125 HBCCO n = 1 (1201) 95 HgBa2Can−1CunO2n+2 n = 2 (1212) 125 n = 3 (1223) 133 n = 4 (1234) 127 n = 5 (1245) 110 n = 6 (1256) 91 n = 7 (1267) 85

Table 1.1: Table of some common multi-layer hole-doped cuprates that demonstrate the Tc tends to scale with the lattice complexity and the number of charge reservoir layers up to a critical value of n ∼ 3.

YBCO-248, and consists of alternating single and double chains. Oxygen atoms in single chains are highly mobile, thus the oxygen content are tunable by changing growth and post-annealing conditions. In YBCO-123 and YBCO-247, both of which contain single CuO chains, the δ values and the doping level can be tuned from 0 to 1. This corresponds to no oxygen in the chains and filled chains respectively. The oxygen atoms are tightly bound in double CuO chains because of the inter-chain attractions, thus there is no doping variability in YBCO-248, which contains double chains alone. Pure YBCO-248 samples also do not tend to contain twin planes and other planar defects due to their stability [15]. Another consequence of the inter-chain attraction is the shrinking of the b-axis lattice parameter and the elongation of the a-axis lattice parameter. As a consequence YBCO-248 is less orthorhombic than YBCO-123, and YBCO-247 has intermediate orthorhombicity between YBCO-123 and YBCO-248. Besides these three common phases of YBCO, exotic phases of YBCO containing triple or more chains are occasionally seen in trace amount no more than 1 or 2 unit cells as defect phases. However deliberate attempt of synthesizing these other phases has not been successful in the past.

Recent studies have shown that the CuO chains can have rich physics in their own Chapter 1. Introduction 5

Figure 1.2: Lattice structures of YBCO-123, YBCO-247 and YBCO-248. These three materials are distinguished by CuO chains running along the b-axis direction. YBCO- 123 has single CuO chains and YBCO-248 has double chains. YBCO-247 has alternating single and double chains, and can be thought as a superlattice of YBCO-123 and YBCO- 248.

right, instead of merely serving as charge reservoirs for Cooper pairing in the CuO2 planes. It is believed that the chains can host charge density wave order [16–20], break charge confinement within the CuO2 planes [21, 22], as well as proximity-couple with the CuO2 planes to produce d + s pairing symmetry [23–27]. Most interestingly, it was reported that double-CuO chains can sustain superconductivity on their own when the CuO2 planes are rendered insulating by Pr-substitution in YBCO-247 [28–32]. This chain- based superconductivity appears to involve pairing of electrons instead of holes [33], Chapter 1. Introduction 6 and to be connected with Luttinger liquid physics [29]. One promising way of studying the role of the CuO chains in the superconductivity is to grow different phases of YBCO films with different arrangement of the chains.

Figure 1.3: Two phase diagrams of temperature versus oxygen pressure for bulk YBCO samples. Panel (a) and (b) are adapted from [34] and [35] respectively. The blue dashed box in panel (a) corresponds to the pressure-temperature range covered in panel (b). Both phase diagrams show that YBCO-123 is stable at low oxygen pressure, YBCO- 248 is stable at high oxygen pressure, and YBCO-247 is stable at an oxygen pressure intermediate between the first two cases.

YBCO-123 is the most commonly synthesized form of YBCO either in thin-film forms or in bulk. Other phases of YBCO can be stabilized either by changing the growth/annealing temperature and oxygen pressure, or by applying heteroepitaxial strain for thin- film samples. We will investigate both of these possibilities in this thesis. A temperature- oxygen pressure phase diagram for bulk YBCO samples is shown in Figure 1.3. At any given temperature, higher oxygen pressure encourages the formation of the phase with Chapter 1. Introduction 7

more CuO chain per unit cell. In bulk form, conversion between different phases can be kinetically limited and extremely slow, even if the thermodynamic condition for phase conversion is met. As we will discussed in more detail in chapter 5, the thermodynamics of thin films are different from that of bulk samples. Sometimes it is easier to convertbe- tween different phases and discover new phases of YBCO in thin films. Heteroepitaxial strain, which is unique to thin-film samples, is another factor that can drive phase conversion. As we will discuss in more detail in chapter 3 and 4, an anisotropic strain that compresses the b-axis relative to the a-axis will encourage the conversion from YBCO- 123 to less orthorhombic phases such as YBCO-247 and YBCO-248.

1.2 Half-metallic manganites

Half-metals are an extreme form of ferromagnets in which the spin of charge carriers is almost completely polarized. They appear as metals for charge carriers with one spin, and as insulators for charge carriers of the opposite spin, because the energy bands of the opposite spin do not cross the Fermi surface. An energy diagram for a typical half metal is shown in Fig. 1.4.

La2/3Ca1/3MnO3 (LCMO) is one of the most well-known half-metals. Depending on measurement techniques and different definition of spin polarization, between 70%to 90% of electrons can be regarded as spin polarized. The root of half-metallicity and ferromagnetism in LCMO lies at the valence of the manganese ions [36]. The parent

3+ 3+ compound LaMnO3 of LCMO consists entirely of Mn ions. When La is partially substituted by Ca2+, some Mn3+ ions lose one electron to become Mn4+ in order to conserve charge neutrality. When one Mn3+ ion and one Mn4+ ion are separated by an intermediate O2− ion, electrons can hop between the two Mn ions through the middle O ion. Such electron hopping gives rise to the double exchange mechanism [37], which is illustrated in Figure 1.5. Fig. 1.5a) shows a configuration with the Mn d electrons Chapter 1. Introduction 8

Figure 1.4: Schematic diagram of density of states for a half-metal as a function of energy close to Fermi energy (EF ).

3+ order ferromagnetically. The Mn ion has one eg electron, and this electron can hop to the bridging O2− ion provided one p electron of the same spin from the O2− ion hops to the Mn4+ ion simultaneously. The hopping electrons give LCMO its metallicity, and

they need to align with the localized Mn t2g electrons in order to preserve Hand’s rule and save energy. The amount of energy save is proportional to cos(θ/2) [38], where θ is

the angle between the spins on neighboring Mn ions. The minority spin state t2g ↓ is

positioned at a higher energy than the full t2g ↑ state by ∼ 2.5 eV due to loss of exchange energy [39, 40]. This energy gap is the reason why LCMO behaves as an insulator to electrons of the minority spin. Fig. 1.5b) shows an antiferromagnetic configuration. Electron hopping is not energetically favorable because Hund’s rule is violated. Since the electron’s ability to hop gives a kinetic energy saving thus reduces the overall energy, the Mn d electrons prefer to align parallelly in the ferromagnetic configuration [36, 41].

LCMO has a perovskite ABO3 structure as shown in Fig. 1.6a). Each B-site ion (Mn

2− in this case) are surrounded by 6 O ions forming a BO6 octahedra. Ideal perovskites are cubic, however most of the perovskite materials encountered in nature are distorted from the ideal cubic structure, depending on the so-called Goldsmith tolerance factor. Chapter 1. Introduction 9

Figure 1.5: An illustration of the double exchange mechanism in mixed-valence manganites, which is responsible for the half-metallicity and the ferromagnetism in LCMO [36, 41]. (a) shows a configuration with ferromagnetically aligned Mn ions. Electrons can hop from Mn3+ to Mn4+ through the intermediate O2−. (b) shows a configuration with antiferromagnetically ordered Mn ions. Due to violation of Hund’s rule, hopping is not allowed. The antiferromagnetic alignment is energetically unfavorable because there is an energy cost when electrons cannot hop.

√ The Goldsmith tolerance factor t is defined by t = (rA + rO)/ 2(rB + rO), in which rA,

2− rB, and rO are the average ionic radii of A- and B- site cations, and O ion respectively. A perfect cubic perovskite can only form when t = 1. The value of the tolerance factor is 0.917 for LCMO, and there is a distortion of the lattice manifested by a rotation of the MnO6 octahedra and a change of the Mn-O-Mn bond angle. As a result of this distortion, the unit cell of LCMO is enlarged, and consists of multiple ABO6 perovskite units. One unit cell of LCMO is orthorhombic with a = 5.474 Å, b = 5.459 Å, and c =

7.715 Å [42]. When LCMO film is grown on another perovskite such as SrTiO3, the a- and b-axis of the film and the substrate does not align. Instead the substrate a- axis is aligned with the (110) axis of the LCMO film as shown in Fig. 1.6b). In this alternative definition of a unit cell, a = b = c = 3.86 Å, and α = 90.15◦. This unit cell is very close to being cubic with a small rhombic distortion thus LCMO unit cells are usually referred as pseudocubic.

Another distortion of the lattice comes from the Jahn-Teller effect, which break the Chapter 1. Introduction 10

Figure 1.6: Lattice structure of the perovskite LCMO. (a) shows the ideal perovskite structure without any distortion. (b) shows the LCMO lattice structure. The lattice is distorted from ideal perovskite because of ionic size difference and Jahn-Teller effect. The unit cell size is multiplied from the perovskite unit cell due to different rotation of the MnO6 octahedra at different layers. (c) shows the in-plane projection of the MnO6 octahedra.

3+ symmetry of the eg doublet. The one eg electron from Mn is energetically favorable to be at the 3z2 − r2 orbital instead of the x2 − y2 orbital. The rhombic distortion of the MnO2 plane mentioned above is mainly due to this lift of degeneracy. When the eg electron hops from a Mn3+ ion to a Mn4+ ion, the distortion is carried along with the hopping electron, and a polaron is formed [43]. Both the Jahn-Teller effect and the double exchange mechanism are intimately linked with the colossal magnetoresistance (CMR) phenomenon observed in LCMO [36, 40, 44–47] . The resistivity of LCMO samples can vary by several orders-of-magnitude in response to a magnetic field because of the CMR effect. Without a magnetic field applied, the same mechanism responsible for CMR effect causes the metal-to-insulator transition at the ferromagnetic transition temperature TCurie. This metal-to-insulator transition gives the characteristic CMR peak in the resistance versus temperature curve of LCMO samples as shown in Fig. 1.7. Chapter 1. Introduction 11

Figure 1.7: A typical resistance versus temperature plot of LCMO showing a metal-to- insulator transition at the ferromagnetic transition temperature (TCurie) ∼ 260 K.

1.3 Superconducting pairing symmetry and proxim-

ity effect

All superconductors are characterized by a position-dependent order parameter ∆(~r),

which can be defined as ∆(~r) = VN (~r)F (~r) [48]. In this expression VN (~r) stands for the strength of electron-electron interaction, and can be positive for an attractive interaction and negative for a repulsive interaction. The condensation amplitude F (~r) is the probability amplitude of finding a Cooper pair at point ~r, and is closely related to the anomalous Green’s function [49–51]

fαβ(~r1, ~r2; t1, t2) = T < ψα(~r1, t1)ψβ(~r2, t2) > (1.1)

Here T is the time-ordering operator. ψα(~r1, t1) and ψβ(~r2, t2) are fermionic creation operators that create electrons at space coordinate ~r1 and ~r2, time t1 and t2, and with Chapter 1. Introduction 12 spin α and β respectively. Here the anomalous Green’s function can also be expressed as a function of the Matsubara frequency instead of time by a Fourier transformation [51]. The symmetry in the order parameter in orbital, spin, and time (frequency) is collectively known as the pairing symmetry of the superconductor. The pairing symmetry is constrained by the Pauli exclusion principle, which means the anomalous Green’s function must be odd under an exchange of particles [51–53]. This symmetry requirement can be satisfied by one of 4 different ways as illustrated in Table 1.2. Mostofthe commonly encountered superconductors are even in both orbital and frequency, and are odd in spin, thus are called spin-singlet superconductors. orbital spin time (frequency) material Even (s, d) Odd (singlet) Even elemental superconductors, cuprates… Even (s, d) Even (triplet) Odd MgB2, superconductor/ferromagnet heterostructures… Odd (p, f) Even (triplet) Even Sr2RuO4, UPt3… Odd (p, f) Odd (triplet) Odd unknown

Table 1.2: Superconductor pairing symmetries allowed by Pauli exclusion principle [51– 53].

When a superconductor is placed in electrical contact with a non-superconducting metal, the Cooper pairs penetrate into the non-superconductor. This phenomenon is referred as proximity effect. The length-scale of this effect varies greatly, depending on the pairing symmetry of the superconductor and the magnetic properties of the non- superconducting metal [54].

The simplest case of proximity effect involves a spin-singlet superconductor anda normal metal. The order parameter ∆ decays exponentially to zero in the metal side with a characteristic length ξN [55]. ξN =hv ¯ F /2πkBT in the clean limit (l > ξN ) and p hD/¯ 2πkBT in the dirty limit (l < ξN ), where l and vF are the mean free path and Fermi

1 velocity in the metal respectively. The diffusivity D is given by the expression 3 vF l. The induced superconducting gap in the quasiparticle density of states (DOS) of the metal can be detected via tunneling or point contact spectroscopy measurements. For example, Chapter 1. Introduction 13

Figure 1.8: Schematic illustration of long-ranged proximity effect across a superconductor/ferromagnet junction. Odd-frequency spin-triplet pairs with even-parity orbital symmetry are induced near the interface and penetrate into the ferromagnet far deeper than spin-singlet pairs, since spin-triplet pairs are not easily broken by the exchange field in the ferromagnet.

in a Ag/Pb junction, ξN has been shown to be as large as 400nm in agreement with theoretical predictions [56]. On the superconductor side, ∆ and the critical temperature

Tc is also suppressed near the interface with d∆/dx ∼ ∆∞/ξGL, where ∆∞ is the bulk order parameter far away from the interface, and ξGL is the Ginzburg-Landau coherence length. The reason for this suppression of ∆ is that as one electron from the Cooper pair enters the normal metal leaving behind the other electron inside the superconductor, the attraction within the Cooper pair weakens and the Cooper pair breaks down.

In the interface between a spin-singlet superconductor and a ferromagnet, the analysis is more complicated because of the incompatible orders between ferromagnetism and spin- singlet superconductivity. Spin-singlet superconductors consist of electron pairs with antiparallel spins, while ferromagnetism favors parallel spin alignment. The exchange coupling in the ferromagnet strongly suppress spin-singlet Cooper pair formation. If the exchange field is homogeneous, a spin-triplet component with projection Sz = 0

(Sz is the ~z component of the electron spin) along the direction of magnetization is induced, due to the spin-dependent phase shift of the two electrons within the Cooper pair [54, 57]. This spin-dependent phase shift also causes an oscillation in the singlet Chapter 1. Introduction 14

component of the order parameter alongside the exponential decay as depicted in Figure 1.8 [54]. In the superconductor/ferromagnet junction, the length-scale of proximity effect p is reduced to ξF,singlet =hv ¯ F /2Eex in the clean limit and ξF,singlet = hD/¯ 2Eex in

the dirty limit, where Eex is the exchange energy [54]. The range of proximity effect has been shown to be ∼ 5 nm for a junction between the superconductor Nb and the ferromagnet PdNi via tunneling measurement [58]. For half-metallic ferromagnet such

as La2/3Ca1/3MnO3 (LCMO), ξF,singlet is even more reduced to a theoretical value of ∼ 0.5 nm. In a ferromagnet/superconductor multilayer structure, Ginsburg Landau theory

2 shows that the Tc is reduced to Tc(ds) = Tc(∞)[1 − (πds/ξs) ] from the bulk value of

Tc(∞) [59], where ds is the thickness of superconductor layers.

Most of the ferromagnetic materials encountered in experiments tend to contain domain walls, and the presence of these domain walls make the magnetization inhomogeneous inside the ferromagnet. When the magnetization is inhomogeneous near the superconductor/ferromagnet interface, spin-rotation by the inhomogeneous magnetic field

causes the spin-triplet component to acquire a Sz = ±1 component along the local magnetization vector [57, 60]. Because the electron spins are parallel aligned, this spin-triplet

Cooper pairs with Sz = ±1 are not destroyed by exchange field unlike the spin-singlet

or the Sz = 0 spin-triplet Cooper pairs. As a result these Sz = ±1 spin-triplet Cooper p pairs can penetrate into the ferromagnet over a length scale equal to hD/¯ 2πkBT , and

is much longer than ξF,singlet as seen in Figure 1.8 [61]. For example, Keiser et al. has

shown that a supercurrent can survive inside the half-metallic ferromagnet CrO2 over ∼ 1

µm, which is much longer than ξF,singlet ∼ 1 nm. For a d-wave superconductor with even parity such as the cuprates, the spin-triplet component is odd in frequency as shown in Table 1.2. Chapter 1. Introduction 15

Figure 1.9: The lattice structures of Sr2IrO4 and perovskite SrIrO3. Panel (a) shows the three dimensional structure and panel (b) shows the projection of the IrO2 planes on the ab-plane. The Ir, Sr, and O atoms are color-labeled as green, purple and red respectively.

1.4 Ruddlesden-Popper iridates

Even though the exact mechanism of superconductivity remains an open question in the cuprates, it is known that superconductivity happens within the quasi-2 dimensional

CuO2 planes, which form a square lattice with antiferromagnetic coupling. In order to further investigate the origin of superconductivity in cuprates, it is natural question to ask if superconductivity can be retained if Cu in the CuO2 planes is replaced with other cations. The layered perovskite Sr2RuO4 is the first such Cu-free superconductor discovered that is isostructural to the cuprate superconductors [62, 63]. Sr2RuO4 contains

RuO2 planes instead CuO2 planes, and has a Tc ∼ 1.5K. Since the discovery of Sr2RuO4, many theoretical and experimental studies have since shown that replacing CuO2 with

IrO2 is another promising way of obtaining superconductivity, because of many similarities between the cuprates and the iridates [64–69].

All the Ruddlesden-Popper series of iridates Srn+1IrnO3n+1 contain IrO2 planes. The Chapter 1. Introduction 16

Figure 1.10: The schematic energy diagram of the Ruddlesden-Popper iridates. This figure is adapted from [72].

lattice structures for Sr2IrO4 (SIO-214) and SrIrO3 (SIO-113), which corresponds to n = 1 and n = ∞ respectively, are shown in Figure 1.9. In both SIO-214 and SIO-213, each

Ir atom is surrounded by 6 O atoms forming corner-sharing IrO6 octahedra. The IrO6 octahedra form square lattice in the ab-plane, and each of the octahedra rotates about the c-axis by ∼ 11◦ for SIO-214 and ∼ 14◦ for SIO-113 [70] as shown in Figure 1.9b).

Apart from the rotation of the IrO6 octahedra, SIO-214 has the same lattice structure as the cuprate La2−xSrxCuO4 (LSCO). Adjacent IrO2 planes are separated by two insulating

SrO layers, thus the electrical connection between IrO2 planes is weak, just like the CuO2 planes in LSCO. Under ambient pressure, SIO-113 takes a hexagonal structure, and the perovskite structure cannot stably form in bulk unless a hydrostatic pressure of at least 40 kbar is applied [71].

5 Because of a tetragonal distortion of the IrO6 octahedra, the five 5d electrons in both SIO-113 and SIO-214 take a low spin configure with empty eg orbital as shown in Figure 1.10 [72]. The t2g orbital is further split by spin-orbit interaction into the full Jeff = 3/2 level and the half-filled Jeff = 1/2 level. The bandwidth of the Jeff = 1/2 level is very narrow, thus small electron correlation is enough to induce a Mott metal- Chapter 1. Introduction 17

insulator transition. SIO-214 is more 2-dimensional than SIO-113, because there are more

insulating SrO layers between the IrO2 planes in SIO-214. As a result, the bandwidth is smaller in SIO-214 than in SIO-113, and SIO-214 is an antiferromagnetic Mott insulator just like undoped cuprate superconductors. On the other hand, the electron correlation

is not strong enough to split the Jeff = 1/2 level in SIO-113. SIO-113 is a correlated semimetal with rather poor metallicity.

1.5 Thesis outline

We will begin in Chapter 2 by describing the main experimental techniques used for this thesis. First the pulsed laser deposition (PLD) technique, which is the main technique for growing all our thin-film samples, is introduced with the details specific to ourown system. Next, we discuss the techniques for determining the growth rates of the films using X-ray reflectivity and atomic force microscopy. The later technique can alsogive the surface morphology of films and this information is crucial for optimizing thePLD growth conditions. The high-pressure annealing setup is described in the next section.

Finally, we cover the procedures for resistivity versus temperature measurements, with details about our home-made dipper probe and the instrumental layouts.

In Chapter 3, we report the experimental results on strain-induced phase transformation and superconductivity attenuation in LCMO/YBCO-123 thin-film heterostructures. STEM on bilayer LCMO/YBCO-123 thin films revealed double CuO-chain intergrowths which form regions with the YBCO-247 lattice structure in the YBCO-123 layer. These nanoscale 247 regions do not appear in XRD, but can physically account for the reduced

Tc of bilayer thin films relative to unilayer films with the same YBCO-123 thickness, at least down to ∼ 25 nm. By comparison the double CuO-chain defect structures are largely absent in unilayer YBCO-123 films. The CuO intergrowths is attributed tothe

bilayer heteroepitaxial mismatch, and the Tc reduction is related to the generally lower Chapter 1. Introduction 18

Tc seen in bulk YBCO-247 samples. These epitaxially-induced phase-transformation to YBCO-247 provides a microstructural mechanism, as opposed to a magnetic mechanism, for the attenuation of superconductivity in LCMO/YBCO-123 heterostructures.

Chapter 4 of the thesis builds on the concept of strain-induced attenuation of superconductivity that is developed in Chapter 3, and further extends this idea by study-

ing the dependence of the Tc of YBCO-123 films on thickness when heteroepitaxially clamped by different oxides. We study perovskite/YBCO-123/perovskite thin films using either ferromagnetic LCMO or paramagnetic LaNiO3 (LNO) as the clamping layers.

For a lattice-symmetry matched comparison, orthorhombic PrBa2Cu3O7−δ (PBCO) is also used in place of the pseudocubic perovskites. Both LCMO/YBCO-123/LCMO and

LNO/YBCO-123/LNO trilayers show strong attenuation of the superconducting Tc as

YBCO-123 layer thickness is reduced from 21.4 to 5.4 nm. The change in Tc is much milder in PBCO/YBCO-123/PBCO trilayers. This results indicate that heteroepitaxial strain, rather than long-range proximity effect, is responsible for the long length scales

of Tc attenuation observed in c-axis LCMO/YBCO-123 heterostructures.

Besides using heteroepitaxial strain, superoxygenation is another powerful technique that can induce phase transformation on oxide thin films. Chapter 5 of the thesis fo- cuses on the superoxygenation and cation enrichment of YBCO thin films. By exploit- ing the high surface-to-area ratio and the high thermodynamical reactivity of thin films, we annealed YBCO-123 thin films in ultra-high-pressure oxygen at up to 700 atmand 900 ◦C while buried under a mixture of YBCO-123 and CuO powder. Scanning transmission electron microscopy (STEM), X-ray diffraction (XRD) and X-ray absorption spectroscopy on the films show clear structural phase transformation to YBCO-247 and YBCO-248 after high-pressure annealing. More complete phase transformation is seen on films annealed with more excess CuO powder. Partially converted films have amixtureof YBCO-123, YBCO-247 and YBCO-248 that are structurally ordered, with the chain-rich

phases appear closer to the film surface. Regions of exotic phases YBa2Cu5O9−δ (YBCO- Chapter 1. Introduction 19

125), YBa2Cu6O10−δ (YBCO-126), Y2Ba2Cu3O9−δ (YBCO-223), and Y3Ba2Cu3O11−δ (YBCO-323) are also seen in the STEM images taken on some high-pressure annealed films. Similarly annealed YBCO-123 powders show no phase conversion, even whenthe annealing time is 4 times longer. These results demonstrate a novel route of synthesis towards discovering more complex phases of cuprates and other superconducting oxides.

In Chapter 6, we report the work on superoxygenation of Sr2IrO4 thin films. The motivation for this part of the work is to apply the superoxygenation technique used in YBCO films to the iridates, in order to metallize Ir-based perovskites through hole doping. The annealing temperature was kept low to minimize cation interdiffusion. 20-nm and

100-nm films annealed in 400 atm ofO2 at 400 C for 2 - 7 days show a progressive drop in room-temperature resistivity of up to 3 orders of magnitude, and an evolution from insulating to metallic behavior. Hall effect measurement at 40 K on the least resistive film shows the majority charge carriers are positive, thus confirming the SIO-214 films become hole-doped metal upon extended high-pressure annealing. STEM, XRD and XAS on post-annealed films revealed structural transformation to the SrIr1−xO3 (SIO-113) phase. The degrees of resistivity reduction and phase transformation both scale with increasing annealing time, pressure, and reduced film thickness. The evolution towards metallicity can be attributed to the phase transformation, interstitial oxygens, and Ir vacancies, with the latter two constitute hole-doping. The Ir vacancies in the most phase-transformed film appears to be structurally-ordered along the c-axis, as evidenced by the STEM data. Finally in Chapter 7, we summarize the main conclusions of the thesis and give some suggestions for potential future works based on the current results. Chapter 2

Experimental Techniques

2.1 Pulsed-laser deposition

The cuprate unilayer and heterostructure films used in this study were made in our labs using pulsed-laser deposition technique (PLD). PLD is a well-established method for growing a variety of thin films from insulators to metals to polymers, and with film quality competitive to molecular orbital chemical vapor deposition (MOCVD) and molecular beam epitaxy (MBE) [73]. During the PLD process, a high-energy laser pulse is used to ablate a small area on the target material. A thin layer of the target is vaporized and ejected as a plasma plume. The plume contains a complex mixture of species including atoms, molecules, ions and clusters, many of which have different stoichiometry as the target material. The plume interacts with background oxygen gas in the PLD chamber, before condenses on heated substrates and form thin films.

Compared with other film fabrication techniques, PLD has the advantage of offering great versatility and short turn-around time. A large class of materials can be grown with PLD. The depositing material can be switched in seconds during film growth, thus heterostructure films are easily made. Furthermore, because PLD is a highly non-equilibrium process, metastable materials, which cannot be synthesized in bulk forms, can be made

20 Chapter 2. Experimental Techniques 21 into thin films with PLD.

Figure 2.1: The schematic diagram of our PLD setup.

A diagram of our PLD set-up is shown in Figure 2.1. We used a 248 nm KrF excimer laser (Coherent Compex 201), which produces ∼ 10 ns long pulses with energy of ∼ 0.093 J per pulse. The repetition rate was kept at 2 Hz when making cuprate films and 5 Hz when making other kind of films. The laser beam first passes through an aluminum cropper, which is used to eliminate inhomogeneous part of the beam, and then gets focused by a lens into the PLD chamber on the surface of the ceramic target. The fluence of the laser pulse at the surface of the targetis ∼ 2J/cm2. The front window of

PLD chamber needs to be periodically cleaned with 10% HNO3 solution, to get rid of any ablated target material that condenses on the window. Otherwise the incident laser beam is attenuated, and a desired laser fluence cannot be achieved.

We mainly use SrTiO3 (STO) or (LaAlO3)0.3(Sr2TaAlO6)0.7 (LSAT) substrates for all film growth. Prior to deposition, the substrates were cut with diamond saw alongthe a− and b− crystalline axis, to rectangles no larger than 5 mm along either side. The substrates were then glued to the contact heater inside the PLD chamber using silver paint (Ted Pella Leitsilber 200), which is used for heat conduction. The location the heater where the substrates are glued was determined by arranging multiple 2.5 × 2.5 mm square substrates on the heater in an array, and then deposit YBCO-123 films on all of these substrates simultaneously. It is found that for substrates placed outside a circle of ∼ 1 cm radius, the Tc of the YBCO-123 films grown on these substrate starts to Chapter 2. Experimental Techniques 22 degrade. Thus the “sweet spot” on the heater was found to be this 1 cm-radius circle.

Figure 2.2: A photograph of the inside of the PLD chamber during film growth with the plume showing.

During deposition, the PLD chamber was pumped down to a background pressure of 10−6 Torr by a turbo pump before getting filled with background oxygen (99.997% pure) at 200 to 800 mTorr. The purpose of the background oxygen is twofold. The first one is to provide reactive oxygen that will be incorporated into the lattice structure of the oxide films during growth. The second purpose is to reduce the kinetic energy ofthe plume from several hundred eV to as low as 1 eV before the plume hits the substrates [74], in order to prevent damage to the film by high kinetic energy of the plume. In general the film growth rates are exponentially related to the background pressure [75]. The substrate is heated to a temperature of 700 to 850 ◦C, which is also calibrated with an infrared pyrometer. The substrate-to-target distance roughly equals the length of the visible plume ejected from the target, and equals ∼ 6 cm for most of our targets. A plume cropper is placed around the target in order to remove the edge part of the plume which may be inhomogeneous. The target is continuously rotated during deposition, so that Chapter 2. Experimental Techniques 23

the laser beam does not always hit the same spot on the target. Repeated ablation on the same spot of the target causes overheating, and excess particulates on the deposited films. We have multiple target holders, which allow us to load up to 6 different targets

into the PLD chamber at the same time. The targets can be switched in-situ during PLD growth without breaking vacuum, thus we can easily make thin-film heterostructures. A photograph of the inside of the PLD chamber during film growth is shown in Fig. 2.2, to show the relationship between the targets, the plume and the substrate. Typically it takes ∼ 20 to 50 laser pulses to grow 1 nm of film. The slow deposition time allows us to grow heterostructure films with high precision.

After growth, the PLD chamber was back-filled with 1 atm of oxygen, and the film

is slowly cooled at a rate of 12 °C/min to 300 °C to in-situ anneal and fully oxygenate the films. The films were then furnace cooled to room temperature at the laststep.

Many different factors influence the film growth conditions. Some of the mainfac- tors include the target density, laser energy and repetition rate, substrate temperature, chamber background pressure, and target-to-substrate distance. These factors constitute the PLD recipe. For each type of film we grew, extensive optimization of the PLD recipe was performed. The target-to-substrate distance and the chamber pressure are two factors with the most variability and have the most effect on the film quality. We first coarse-tune the parameters with surface morphology measurements using atomic force microscope (AFM) and room temperature resistivity measurements. For AFM measurements, we use a ThermoMicroscopes Explorer AFM operating in contact mode. Figure 2.3 compares representative AFM images on sub-optimally grown and optimal grown YBCO-123 films. Fig. 2.3a) shows a film containing a-axis intergrowth, which shows up as perpendicular, rod-like features. These a-axis intergrowth is mainly caused by lower than ideal density of the target, and was eliminated with another target with higher density. Fig. 2.3b) shows a film containing large number of island-like particulates. These particulates tend to form when the laser energy and the target-to-substrate distance is Chapter 2. Experimental Techniques 24

Figure 2.3: Comparison of AFM images of unoptimized and optimized YBCO-123 films. (a) shows a film containing a-axis intergrowth, which shows up as rod-like features that are perpendicular to each other. (b) shows island-like particulates on the film surface. (c) shows a film with optimized growth conditions. The optimal film is essentially featureless with random fluctuations mostly coming from instrumental noise.

suboptimal, and can be eliminated by tuning these parameters. When the film surface is free from any defects and looks essentially featureless on the AFM images as shown in Fig. 2.3c), and when the room temperature resistivity becomes reasonably close to the ideal values of each material, the PLD recipe is further fine-tuned using resistivity

versus temperature measurements. Small adjustments are made on the parameters, es-

pecially the laser repetition rate and substrate temperature, based on the Tc for cuprate films, the CMR peak temperature for manganite films, and the residual-resistance ratio for other metallic films such as nickelate films. Chapter 2. Experimental Techniques 25

2.2 Thickness and roughness measurement by X-ray

reflectivity and atomic force microscopy

For all the experiments involving thin-film samples, it’s crucial to know the film thickness with high precision, because one important goal of the experiments is to study the effects of changing thickness on various physical properties of the films. The interfacial roughness needs to be controlled and minimized, because roughness not only introduces uncertainty to the film thickness, it may also has detrimental effects on the film crystallinity. Our two main techniques used for determining film thickness and roughness include X-ray reflectivity (XRR) and atomic force microscopy (AFM).

XRR is a non-destructive technique for determining the thickness and surface roughness of thin films with up to Angstrom-level precision. The XRR data on our films were obtained by Professor Thomas Gredig and Ahn Nguyen from California State University Long Beach using a Rigaku Smartlab diffractometer. The setup of XRR measurements is very similar to that of X-ray diffraction (XRD) as shown in Figure 2.4. Specular reflection condition is satisfied, which means the incident X-ray beam has the sameangle θ to the surface as the reflected beam. The detector rotates at the twice the speedas the sample. Both the sample and the detector rotate about the same axis, which lies on the sample surface. The intensity of the reflected beam is measured as a function of 2θ. Experimentally the only difference between XRR and XRD is that the X-ray beam irradiates the thin-film sample at a grazing angle for XRR. Typically only 2θ < 5◦ is rel- evant for XRR analysis. For XRD, the measurements usually are taken at 2θ > 10◦.

Because of the different index of refraction for each film layers and the substrate, the X-ray beams reflected from each interfaces undergo angle-dependent phase shifts. The phase shifts result in constructive and destructive interference, and interference fringes appear in the XRR spectra as a result. Such fringes are called the Kiessig fringes, and the separation of the fringes are mainly determined from the film thicknesses. For an Chapter 2. Experimental Techniques 26

Figure 2.4: The experimental setup of typical XRR measurements. Both the sample and the detector rotate about the same axis, which lies on the sample surface. The detector rotates at the twice the speed as the sample. The same experimental setup is also used for XRD. unilayer film of thickness d, the distance ∆θ between consecutive maxima of the XRR

λ spectra is given by ∆θ ≈ 2d with small angle approximation, where λ is the wavelength of the X-ray [76]. For multi-layer films, the different periods from each layer give riseto beats in the Kiessig fringes. The XRR spectra expected from multi-layer films, taking multiple reflections into account, can be calculated using Parratt’s recursive method [77]. Parratt’s recursive method calculates the XRR spectra by iterating through the multilayer samples layer by layer, starting from the substrate and working upwards. The reflectivity from each interface rn is related from the reflectivity from the interface below rn+1 by the relationship 0 rn + rn+1 exp(iqnd) rn = 0 (2.1) 1 + rnrn+1 exp(iqnd) where r0 = qn−1−qn is the Fresnel reflectivity from the n-th interface without taking mul- n qn−1+qn 4π tiple reflection into account. The momentum transfer for the n-th layer qn = λ sin(θn) is influenced by the complex index of reflection forthe n-th layer in the multilayer. The total reflectivity of the entire multiplayer sample at the top surface isthen r0.

Interfacial and surface roughness of the films can also be quantitatively determined using XRR. The roughness of the films determines how fast XRR intensity decays exponentially as a function of angle. Films with rough interfaces are modeled as an ensemble Chapter 2. Experimental Techniques 27

Figure 2.5: Comparison between measured XRR data obtained on LaNiO3/YBa2CuO3O7−δ/LaNiO3 trilayer films with the simulated XRR spectra obtained using the GenX software. (a) shows the spectra for a 12.7 nm/6.6 nm/12.7 nm trilayer and (b) shows the spectra for a 12.8 nm/13.4 nm/12.7 nm trilayer. The RMS interfacial roughness was determined to be ∼ 4 Å for both of these trilayer films.

of smooth films having Gaussian-distributed thickness, with mean thickness of d and standard deviation of σ. Under this model the reflectivity at the particular interface is reduced by a factor of exp(−σ2q2/2) ≈ exp(−8π2θ2σ2/λ2) and is incorporated into Parratt’s recursive calculation. Electronic density of films is another parameter that can be extracted with XRR spectra, and is determined by the critical angle of total reflection and the amplitude of spectra oscillation.

To extract the thickness and roughness values of our multi-layer samples from the XRR data, the fitting program GenX is used [78]. GenX uses genetic algorithm tosearch for the optimal parameters with the best fit to experimental data. The program starts with an initial guess of the thickness and roughness values for each individual layer in the multilayer film, and generates a theoretical XRR spectra using Parratt’s algorithm. Chapter 2. Experimental Techniques 28

The thickness and roughness values are progressively refined until the difference inthe simulated spectra produced from the model and the experimental XRR data is minimized. The comparison between the measured XRR data with the simulated XRR spectra is shown in Figure 2.5 for two LaNiO3/YBa2CuO3O7−δ/LaNiO3 films with thickness 12.7 nm/6.6 nm/12.7 nm and 12.8 nm/13.4 nm/12.7 nm respectively. The RMS interfacial roughness was determined to be ∼ 4 Å for both of these trilayer films.

AFM is another commonly used technique for film thickness and roughness measurements. AFM does not provide as precise thickness resolution as XRR, as the uncertainty can be as large as several nm. Also thickness measurements using AFM is destructive. However AFM does have the advantage of providing local morphology of our films that is unobtainable with XRR. The local morphology information is indispensable for optimizing the PLD growth conditions for all the films. Another advantage of AFM over XRR is that we can do AFM measurements in-house with our own ThermoMicroscopes Explorer AFM instead of relying on collaborators, thus AFM is by far the most frequent technique used for thickness and roughness measurement.

To measure film thickness using AFM, the film is first masked by covering thesur- face with Scotch tape. The exposed part is chemically etched with either a mixture of potassium iodide and 10% of hydrochloric acid solution for LCMO films, or just 10% hydrochloric acid for all other types of oxide films. The chemical etching process takes 2 seconds. After etching, a sharp step edge is formed at the boundary of the mask as shown in Fig. 2.6a), and the height of the steps can then be measured with AFM as seen in Fig. 2.6b). Both the STO and the LSAT substrates we use do not react with etching solutions, and no step edge is formed on bare substrates even after extended soaking in the etching solutions.

We do not measure the thickness of all the films directly, instead the thicknesses of most films are determined from the growth rates of each individual materials. Thegrowth rate of each PLD target is fixed if the PLD growth conditions are fixed, thus the thickness Chapter 2. Experimental Techniques 29

Figure 2.6: YBCO-123 film that were etched with HCl solution for thickness measurements. (a) shows a photograph of the film with a sharp interface at the edge of the mask. (b) shows the AFM image taken on the etched film, and the height of the step edgecan be measured using AFM.

of a film layer made from one target is a linear function of the deposition time. Foreachof the PLD target we use, reference unilayer films with precisely measured growth time are made for growth rate measurements. The thicknesses of the reference films are determined from both XRR and AFM measurements. XRR typically provides much better precision than AFM, but on rare occasions a good fit cannot be produced, and the theoretical spectra obtained from Parratt’s algorithm deviates from the experimental XRR spectra. One of the reasons this could happen is that the fitting parameters get trapped in local minima. Only when the thickness values obtained from both XRR and AFM agree, both precision and accuracy of the thickness can be ensured. The growth rate of each material is then calculated as the thickness of the reference films divided by the deposition time. For YBCO-123 and LCMO, which are our most commonly grown materials, the growth rates are also corroborated with scanning electron microscopy (SEM) and transmission electron microscopy (TEM) measurements to further increase confidence. Chapter 2. Experimental Techniques 30

2.3 High-pressure oxygen annealing

Many of the experiments carried out in this thesis require annealing in ultra-high-pressure oxygen at up to 900◦C. Such highly oxidizing environment presents an unique experimental challenge and specially designed high-pressure furnace is required for the annealing process. We used a Morris Research HPS-5015 high-pressure furnace for most of the high-pressure annealing work.

Figure 2.7: A schematic diagram of the Morris Research HPS-5015 high-pressure furnace used for most of the high-pressure annealing works, adapted from the manual of the furnace.

A schematic diagram of the high-pressure furnace is shown in Fig. 2.7. The core of the furnace consists of a pressure vessel made from an alloy consisting of 55% nickel, 20% chromium, and 10% tungsten. This alloy is one of the few materials that does not react with high-pressure oxygen in high temperature while maintaining mechanical strength. The alloy has low thermal conductivity. Therefore the closed end of the vessel, which is in contact with the heating elements, can be heated up to 1000 ◦C, while the open end of the vessel away from the heating elements is still cool to touch. The sample being annealed is placed on a ceramic boat before inserted into the vessel. Chapter 2. Experimental Techniques 31

The void portion of the vessel is filled with a metal rod made of the same alloy as thevessel wall in order to reduce dead volume. The total volume of empty space inside the vessel is estimated to be less than 10 mL, thus a small amount of oxygen gas is consumed for each annealing run. The vessel is flushed with oxygen for 2 seconds before being back- filled with up to 2600 psi of high-purity oxygen at room temperature. As the vesselis heated, the oxygen pressure inside the vessel further increases according to ideal gas law. At 900◦C, a maximum pressure of 700 atm is achievable, and lower pressure is obtained by partially filling the vessel with lower oxygen pressure before heating up.

In addition to high-pressure annealing, annealing in 1 atm of flowing oxygen is also performed sometimes. Two main purposes of this 1-atm annealing is to guarantee the films are fully oxygenated, and to serve as a control for high-pressure annealed films. Flowing oxygen annealing is carried out with a tube furnace, which consists of a quartz tube surrounded by heating elements. Fresh oxygen gas is passed through the quartz tube at ∼ 0.1 mL/s. The flow rate is kept low to keep the temperature uniform within the tube. When underdoping of samples are needed, a turbo pump can be easily connected to the end of the quartz tube and the pressure inside the tube can be reduced to as low as ∼ 10 mTorr during annealing. Thus a full doping range of YBCO-123 can be achieved with our tube furnace.

When a bare film is annealed inside the furnaces at only 400◦C for 2 hours, the film completely sublimates. It is essential to maintain a high vapor pressure of the material constituting the film close to the film surface. This can be achieved via two ways.The first one is to put two films together tightly, with the film facing each other. Inthiscase any exposed part of the films vanishes after annealing. The second way of preventing sublimation is to bury the films underneath powders of the same material as the top layer of the films. This is our most commonly used method. The burial powder not only suppresses film sublimation during annealing, but also provides cations to thefilms via solid-state diffusion. The stoichiometry of the burial powder is precisely determined Chapter 2. Experimental Techniques 32 with a microgram balance in order to control the amount of excess cation in the powder. This cation enrichment by solid-state diffusion is a central part for some of our annealing experiments and will be further elaborated in Chapter 5.

2.4 Electrical transport measurements

The resistance of our samples were measured using either ac lock-in amplifiers or ac resistance bridge in four-point configurations. The resistance bridge used is a Lake Shore Model 370. Resistance bridge has the advantage of being simple to setup and essentially plug-and-play. However it is not as flexible as lock-in amplifiers, and does not record complete phase information of the complex impedance. For this reason the majority of the samples were measured with lock-in amplifiers, and resistance bridge is only usedas a check.

The set-up for resistance measurements with lock-in amplifiers is shown in Figure 2.8. We used two Stanford Research lock-in amplifiers for this setup, which are used to measure the voltage and current respectively. The lock-in #1 in the diagram, which is a Model SR850, supplies a constant ac voltage of 0.1 V at 197 Hz. This frequency is chosen to be a prime number, and is far away from the frequencies of commonly countered noise such as the 60 Hz ac line noise. The output voltage of the lock-in amplifier is reduced by a factor of 100 by voltage divider before feeding through the sample. The voltage drop across the sample is kept below 0.3 mV, so that the electrical energy per electron never excesses the thermal energy kBT at 4 K. The lock-in #2 used to measure current is a Model SR830 and was synchronized with lock-in #1 so that the reference frequency is the same. The current through the sample was calculated from the voltage drop across a known resistor connected in series with the sample. The phase shift of the voltage relative to current is recorded, and the resistance and reactance values can be extracted as the real and complex components of the complex impedance respectively. Chapter 2. Experimental Techniques 33

Figure 2.8: Circuit diagram for four-terminal resistivity measurement using two lock-in amplifiers.

When the phase shift becomes large, this usually indicates contact problems that need further investigation.

To measure the resistance of our thin-film samples, 4 silver contacts were sputter coated onto the surface of the samples in parallel-strip geometry. Such four-terminal measurement can eliminate contact resistance and is especially advantageous for measuring low-resistance samples such as thin-film superconductors. To measure the resistivity of the films, we used van der Pauw technique, which can accurately measure the resistivity of an arbitrarily shaped thin-film samples with uniform thickness. Aluminum contacts were either glued with silver paint, or wire-bonded to the four corners of square samples. We used great effort to minimize the size of the contact points between wires Chapter 2. Experimental Techniques 34

Figure 2.9: The wiring configurations for van der Pauw measurement used to extract the resistivity of thin film samples. The V/I values obtained from both channels are combined to calculate the resistivity of the sample.

and samples. Current is applied through two of the wires and the voltage drop is measured through the other two wires. Two different configurations of the wires are needed to obtain the resistivity value, which have the current path along one vertical edge and

Vab along one horizontal edge respectively as shown in Figure 2.9. We define Rab,cd = and Icd

Vda Rda,cb = , then the resistivity of a sample with thickness t can be found by solving the Icb van der Pauw equation numerically:

e−tπRab,cd/ρ + e−tπRda,cb/ρ = 1 (2.2)

For resistivity measurements with the van der Pauw technique, in order to rule out artifacts from sample inhomogeneity, contacts were made at different locations of the same sample including at the corners, at the edges and near the center as double check, because the corners of a thin film can become inhomogeneous due to damage from cutting. When the contacts are moved away from the outer perimeter of a square sample, the

d error introduced in resistivity value is proportional to D , where d is the distance from the contact to the perimeter of the sample, and D is the side length of the square sample

∆ρ [79, 80]. In order to have an error ρ < 1%, the contacts need to be placed no more D than 5 away from the perimeter [79, 80], which equals 1 mm for our 5 X 5 mm samples. We checked that different contact placements gave similar resistivity versus temperature Chapter 2. Experimental Techniques 35 behaviors within the error. The voltage-current (V-I) characteristics was checked for linearity for many samples, in order to confirm the samples were ohmic with well-defined resistance. Finally, V-I characteristics were measured with both 4-point and 2-point configuration, and the agreements between these two results indicate negligible contact resistance.

Figure 2.10: The tip of our dipper probe for measuring electrical resistance as a function of temperature.

The resistance or resistivity of our samples were measured as a function of temperature by placing the sample inside our home-made dipper probe, which is placed inside a liquid helium dewar. The tip of the probe, where samples are mounted, is shown in Figure 2.10. The sample is tightly held down on a piece of sapphire sample stage, Chapter 2. Experimental Techniques 36 which is placed on a piece of Cernox temperature sensor. Silver paint is used between samples and the sample stage, and between the sample stage and the temperature sensor for thermal conduction. Before placing into liquid helium dewar, the probe is first sealed and pumped down to a vacuum of ∼ 10−4 Torr by a turbo pump in order to slow down cooling and prevent condensation. We adjust the temperature of the sample by gradually lowering the dipper probe below the surface of liquid helium. The sample holder is kept cryogenically autonomous from the rest of the probe, by inserting a piece of Teflon spacer between the sample holder and the stem of the probe. Thermal exchange is provided by either helium exchange gas (at ∼ 0.1 atm) or crumbled aluminum foil that is in thermal contact with the sample holder. The cooling rate was kept below ∼ 0.1 K per second, so that the temperature is uniform across the entire sample. A piece of Manganin wire is also wrapped around the sample holder as a heater, which can be used to raise or stabilize the temperature of the samples. The temperature is recorded and controlled by a Lakeshore Model 331 temperature controller. The data recording process for resistivity versus temperature measurement is automated with a LabView program I wrote. The program keeps track of all the raw data, as well as performs real-time processing and plotting of the data to extract resistivity versus temperature curve. The program also has the capability to interface with the temperature controller to adjust the sample temperature.

To determine the sign of the charge carrier for some of our films, Hall effect measurements are performed using the Physical Property Measurement System (PPMS) man- ufactured by Quantum Design. Figure 2.11 shows the wiring diagram used during the Hall effect measurement. The sample is mounted to the sample puck with silver paint for thermal conduction. Aluminium wires are wire-bonded from the four corners of the square sample to Channel 3 of the sample puck. The wiring configuration of Channel 3is the same as one of the van der Pauw configurations, and can give an order-of-magnitude estimate of the sample resistivity as a sanity check during the Hall effect measurement. Chapter 2. Experimental Techniques 37

Figure 2.11: The wiring diagram for Hall effect measurement with the PPMS. The Hall voltage is read by Channel 1 and Channel 2; Channel 3 reads one configuration of the van der Pauw resistance.

Channel 1 and 2 of the sample puck is jumper-connected to Channel 3 using insulated copper wires. In both Channel 1 and 2, the voltage is measured along the diagonal of the samples perpendicular to the current path. Both of these channels are needed to reduce the effects of imperfect sample shapes and contact placements.

At each temperature and field setting, up to 5 different DC current at distinct am- perage is applied to the sample. The purpose of changing the magnitude of the current is to check if the sample is ohmic, and to make sure that both the resistivity and the Hall resistance does not depend on the current size. A long time delay up to 30 s is used between each current changes, before the voltage is read, so that the effect of RC time constant is minimized. At each current value, the voltage is also measured with the current reversed in order to remove the effects of thermoelectric voltage, since the thermoelectric voltage is independent of current polarity.

There is always a component of the measured voltage that is parallel to the current Chapter 2. Experimental Techniques 38 path, because samples cannot be cut into perfect squares and the contacts cannot be placed at strict diagonal lines of the samples. On our typical measurements, this parallel voltage component is more than 2 orders magnitude higher than the expected Hall voltage, and is the most significant source of error. Because the parallel voltage component is independent of the field polarity, by subtracting the voltage measured at negative field from the voltage measured at positive field, the polarity-independent parallel voltage component can be eliminated so that the Hall voltage is guaranteed to be 0 at zero field.

After the Hall voltage VH is obtained, the carrier density n and the mobility µm can be easily calculated using the formula

IB n = V tq H (2.3) V t µ = H m IBρ where I is the excitation current used, B is the applied magnetic field, ρ is the sample resistivity, t is the thickness of the thin-film sample, and q is the charge on one charge carrier. Chapter 3

Attenuation of Superconductivity in Manganite/Cuprate Heterostructures by Epitaxially-Induced CuO Intergrowths

3.1 Introduction

The lattice compatibility among transition-metal oxides enables them to be epitaxially combined in thin-film form [81]. In recent years, there have been numerous studies of heteroepitaxial thin films comprising the ferromagnetic (F) manganites and superconducting (S) cuprates, probing novel effects of the F/S interplay ranging from spin injection to proximity coupling [82–87]. An observation of particular interest is the dependence of the superconducting critical temperature (Tc) on the c-axis layer thicknesses in multilayer La2/3Ca1/3MnO3/YBa2Cu3O7−δ (LCMO/YBCO-123) films [88, 89].

39 Chapter 3. Strain-induced phase conversion in LCMO/YBCO-123 40

The length scale of this dependence indicates an extremely long-ranged F/S proximity effect which is of both technological and theoretical interest [54, 61, 90], although direct and microscopic evidence for this effect is still elusive [91, 92]. Other interfacial mechanisms, such as charge transfer [89, 93, 94], orbital reconstruction [95], spin diffusion [89], induced magnetic modulation [96, 97], and formation of charge density wave [98] are also believed to affect the superconductivity in LCMO/YBCO-123 heterostructures.

A crucial aspect of LCMO/YBCO-123 heterostructures that has not been well studied is the microscopic stoichiometry of the YBCO-123 layer. The Y-Ba-Cu-O (YBCO) compounds are exceptional among the cuprates in having CuO chains, the number of which per unit cell allows the cation stoichiometry to vary between the so-called 123, 248, and 247 phases, which we denote as YBCO-123, YBCO-248, and YBCO-247. These phases have different optimal Tc, with bulk YBCO-247 showing generally lower Tc than either YBCO-248 or fully-oxygenated YBCO-123 [34, 35, 99–102]. As shown in Figure 3.1a), the 123 and 248 phases have single and double CuO chains, respectively. Be- cause of the zigzag staggering of the double CuO chains, YBCO-248 has doubled c-axis lattice parameter, and consists of two YBCO-124 building blocks. The 247 phase consists of alternating 123 and 124 blocks. Local stoichiometric variations have often been seen in nominally-123 YBCO-123 samples, typically as intergrowths of extra CuO chains [103–105]. For sufficiently thin YBCO-123 layers, such nanoscale intergrowths could constitute significant phase inhomogeneity in the paths of conduction, thus affecting the resistively determined Tc.

To elucidate the effect of CuO intergrowths on the superconductivity inLCMO/ YBCO-123 heterostructures, in this chapter we carried out a microstructural study of bilayer LCMO/YBCO-123 and unilayer YBCO-123 thin films, in relation to their resistively measured Tc. Chapter 3. Strain-induced phase conversion in LCMO/YBCO-123 41

3.2 Experiment

The bilayer LCMO/YBCO-123 and unilayer YBCO-123 films used in our study were

epitaxially grown on (001)-oriented (La, Sr)(Al, Ta)O3 (LSAT) substrates using pulsed laser-ablated deposition (PLD). The PLD growths were done at 800 ◦C in 200 mTorr of background oxygen for YBCO-123 layers, and in 300 mTorr of O2 for LCMO layers. The repetition rates for the laser pulse was 2 and 5 Hz for YBCO-123 and LCMO layers respectively. In situ post annealing of the films were performed within the PLD chamber

◦ ◦ after growth, by slowly cooling the samples from 800 C to 300 C in 760 Torr of O2 at 11 ◦C/min. The LCMO layer was 25 nm thick, while the YBCO-123 layer was either 25 nm or 50 nm thick. LSAT was chosen for its close lattice matching with YBCO-123. Targets of LCMO and YBCO-123, each having ∼ 90% material density and > 99.9% chemical purity, were used. The microstructure of the films was characterized by Dr. Nicolas Gauquelin and Professor Gianluigi Botton at the Canadian Centre for Electron Microscopy (CCEM) using a FEI Titan 80-300 electron microscope fitted with a high- brightness field emission gun and CEOS aberration correctors for both condenser and objective lens aberrations. The microscope was operated at 200 keV in scanning mode.

XRD was also carried out on the films using the θ – 2θ method with Cu Kα radiation from a Philips PW2273/20 X-ray tube. Finally, electrical resistance of the films was measured

versus temperature using standard ac lock-in technique in the four-contact configuration.

3.3 Results

Figure 3.1 shows High-Angle Annular Dark-Field (HAADF) STEM images taken over the cross section of a 25 nm/50 nm bilayer LCMO/YBCO-123 film grown on LSAT substrate. Fig. 3.1b) shows a low-resolution image, demonstrating uniform heteroepitaxy and layer thickness. Fig. 3.1c) shows a high-resolution image near the LCMO/YBCO- 123 interface; the color labels indicate different elements as identified from the HAADF- Chapter 3. Strain-induced phase conversion in LCMO/YBCO-123 42

Figure 3.1: HAADF-STEM images of a 25 nm/50 nm bilayer LCMO/YBCO-123 film grown on (001)-oriented LSAT substrate. A low-resolution image is shown in panel (b), demonstrating uniform heteroepitaxy and layer thickness, with the LCMO/YBCO-123 interface marked by blue dotted line. A high-resolution image near the LCMO/YBCO- 123 interface is shown in panel (c), revealing intergrowths of double CuO chains which form nanoscale YBCO-247 regions. These double chains are also visible in panel (b), as indicated by arrows. The lattice structures of YBCO-123, YBCO-248 and YBCO-247 phases are shown in panel (a), with the Cu, Y, and Ba atoms color-labeled as yellow, green, and red, respectively. Chapter 3. Strain-induced phase conversion in LCMO/YBCO-123 43

Figure 3.2: STEM image of a 25 nm unilayer YBCO-123 films grown on LSAT substrate, showing much less intergrowth of double-CuO chains (indicated by arrows) than the bilayer LCMO/YBCO-123 film displayed in Figure 3.1.

STEM images in which the contrast is sensitive to the atomic number (the higher atomic number the brighter the atomic column appears). This image shows the LCMO/YBCO- 123 interface consists of Mn atoms joining Ba atoms at CuO chain sites. This type of LCMO/YBCO-123 interface is commonly reported for LCMO/YBCO-123 heterostructures grown by either PLD or sputtering [106–108]. To the right of this interface, there are three defect-free unit cells of YBCO-123 characterized by single CuO chains between Ba atoms, followed by a block of YBCO-124 characterized by double CuO chains between Ba atoms. This alternation of single and double CuO chains effectively forms a region of YBCO-247, which comprises YBCO-123 and YBCO-124 building blocks. These nanoscale YBCO-247 regions appear in patchy strips throughout the YBCO layer, as indicated by the arrows in Fig. 3.1b). Further on the right of Fig. 3.1c) is another YBCO-247 unit cell, also containing a double CuO chain. It is interesting to note the variation in the Cu atom alignment in the double CuO chains. In the YBCO-247 unit cell on the left, the Cu atoms are horizontally aligned along the c-axis; whereas, in the YBCO-247 unit cell on the right, the Cu atoms are staggered in a zigzag fashion. This difference can be interpreted in terms of micro-twinning between the two YBCO-247 unit Chapter 3. Strain-induced phase conversion in LCMO/YBCO-123 44 cells, whose a- and b-axis orientations are switched. Finally, because the YBCO-123 target used has 123 stoichiometry, the occurrence of double CuO chains is expectedly com- pensated by missing CuO chains elsewhere in the film. Such a chain-less region can be seen between the two YBCO-247 unit cells, showing two adjacent Ba layers with no CuO chain in between. These chain-less regions have tetragonal structures, and can reduce the orthorhombicity of surrounding regions, since the orthorhombicity in YBCO-123 is caused by the quasi-1d CuO chains. The tetragonal chain-less regions may also facilitate the micro-twinning seen between YBCO-247 unit cells, because a- and b-axis directions cannot be distinguished in a tetragonal lattice. It should be remarked that such double- chain and missing-chain intergrowths are significantly less in unilayer YBCO-123 films grown under similar conditions, as shown by the STEM image in Figure 3.2.

Figure 3.3: XRD pattern of a 25 nm/50 nm bilayer LCMO/YBCO-123 film grown on (001)-oriented LSAT substrate. Only peaks associated with the c-axis of either YBCO- 123, LCMO, or LSAT are visible. No peaks associated with YBCO-247 are visible, indicating that the nanoscale YBCO-247 regions seen in the high-resolution STEM image do not appear in the XRD. Chapter 3. Strain-induced phase conversion in LCMO/YBCO-123 45

To probe the pervasiveness of these nanoscale YBCO-247 regions, we carried out XRD on our LCMO/YBCO-123 thin films. Figure 3.3 shows the diffraction pattern of a 25 nm/50 nm bilayer LCMO/YBCO-123 film grown on LSAT substrate. All the major peaks are identified in terms of the c-axis lattice of either YBCO-123, LCMO, or LSAT, although the peaks for LCMO and LSAT are not distinguishable because of their close lattice parameters. The YBCO-123 (003)-peak at 2θ = 22.85◦ and the LSAT (001)-peak at 2θ = 22.99◦ are also not distinguishable. By relating the YBCO-123 (005)-and (007)-peaks with 2θ = 38.55◦ and 2θ = 55.06◦, respectively, we find the c-axis lattice parameter of our YBCO-123 film to be 11.68 Å, in agreement with values reported in the literature [109]. It should be emphasized that no peaks associated with YBCO-247 are visible in the XRD pattern within the resolution of our instrument, indicating that the nanoscale YBCO-247 regions seen in the high-resolution STEM image do not appear in the XRD.

The occurrence of double CuO-chain intergrowths in our bilayer LCMO/YBCO-123

thin films can be physically linked to their resistively measured Tc. Figure 3.4 plots

the resistance (R) versus temperature (T ) data taken on various films. To facilitate comparison, each R vs. T curve is normalized to its R value at room temperature. Both unilayer YBCO-123 films grown on LSAT show sharp superconducting transitions

with Tc near 90 K, consistent with the YBCO-123 being fully oxygenated. As a control,

we also grew unilayer YBCO-123 films on SrTiO3 (STO), which show a similar Tc also

near 90 K. These results indicate that the resistive Tc of the YBCO-123 layer is largely insensitive to the lattice mismatch with the substrate material, down to 25 nm YBCO- 123 thickness. However, the 25 nm/25 nm bilayer LCMO/YBCO-123 film shows a much

lower Tc, near 60 K, and a broader transition than any of the unilayer YBCO-123 films, indicating that the addition of an epitaxial LCMO overlayer significantly reduces the

resistive Tc in the YBCO layer. We can plausibly attribute this Tc reduction to the nanoscale YBCO-247 regions seen in our high-resolution STEM images, since YBCO- Chapter 3. Strain-induced phase conversion in LCMO/YBCO-123 46

Figure 3.4: Plot of normalized resistance versus temperature for unilayer YBCO-123 and bilayer LCMO/YBCO-123 films. All the unilayer films show sharp superconducting transitions with Tc near 90 K. The 25 nm/25 nm bilayer LCMO/YBCO-123 film shows a significantly reduced Tc near 60 K, with a broadened transition. Chapter 3. Strain-induced phase conversion in LCMO/YBCO-123 47

247 has generally shown lower Tc than either YBCO-248 or fully-oxygenated YBCO-123

[34, 100, 101, 110–112]. We note that an alternative explanation of this Tc reduction in terms of under-oxygenated YBCO-123 is not likely, since the LCMO overlayer was grown in-situ at an even higher oxygen pressure than the YBCO-123 layer [91].

3.4 Discussions

Figure 3.5: A schematic diagram illustrating the difference in both lattice symmetry and the ab-plane lattice structure between YBCO-123, YBCO-247, LCMO, and LSAT. The relative length scales between the a- and b-axes for each material, and between the materials, are exaggerated for clarity.

To explain the formation of CuO intergrowths in our LCMO/YBCO-123 films, we con- sider the lattice structures of the materials involved in the heteroepitaxy. Figure 3.5 gives a comparison of the ab-plane lattice structures between YBCO-123, YBCO-247, LCMO, and LSAT. Figure 3.5 shows a schematic diagram illustrating the differences in both lattice symmetry and lattice parameters [113–115], which are tabulated in Table 3.1. First, it is well known that all the superconducting phases of Y-Ba-Cu-O are orthorhombic due

to the CuO chain that runs along the b-axis. Because of the inter-chain attraction within the double-CuO chains, YBCO-247 has a shorter b-axes and is thus less orthorhombic than YBCO-123. Since both the LCMO overlayer and the LSAT substrate have pseu- Chapter 3. Strain-induced phase conversion in LCMO/YBCO-123 48

docubic lattices, their combined mismatch in lattice symmetry with the YBCO-123 layer would favor the formation of the less orthorhombic YBCO-247 phase. In addition to this

bilayer lattice-symmetry mismatch, the a- and b-axis lattice parameters of both LCMO and LSAT are closer to YBCO-247 than to YBCO-123. Thus the lattice-parameter mismatch, from both sides of the YBCO-123 layer, also tends to favor the formation of YBCO-247. In essence, the intergrowth of double CuO chains provides an effective mechanism for relieving the heteroepitaxial strain, imposed by both the LCMO overlayer and LSAT substrate, in the YBCO-123 layer. For the unilayer YBCO films grown on LSAT, the effect of heteroepitaxial strain is responsible for the small amount ofdouble CuO-chain intergrowth seen in unilayer films, and the effect is much smaller than inbi- layers films in which the YBCO layer is heteroepitaxially clamped by two pseudocubic materials. The strain effects from lattice-symmetry-mismatched substrates may also ex-

plain the lower Tc in ultra-thin YBCO films grown on cubic STO than in YBCO films sandwiched between orthorhombic PrBa2Cu3O7 clamping layers [116].

Compound Symmetry a (Å) b (Å) YBCO-123 Orthorhombic 3.821 3.885 YBCO-247 Orthorhombic 3.868 3.911 LCMO Pseudocubic 3.858 3.858 LSAT Pseudocubic 3.868 3.868

Table 3.1: Comparison of the ab-plane lattice parameters between YBCO-123, YBCO- 247, LCMO, and LSAT.

3.5 Conclusion

In summary, we have performed HAADF-STEM, XRD, and electrical resistance measurements on bilayer LCMO/YBCO-123 and unilayer YBCO-123 thin films grown by PLD. The STEM images on the bilayer films revealed YBCO-247 regions formed by double CuO-chain intergrowths, which we attribute to heteroepitaxial lattice mismatch of the YBCO-123 layer with both the LCMO overlayer and LSAT substrate. These Chapter 3. Strain-induced phase conversion in LCMO/YBCO-123 49 nanoscale YBCO-247 regions do not appear in XRD, but can physically explain why

Tc in the bilayer LCMO/YBCO-123 thin films is significantly lower than 90 K, atleast down to ∼ 25 nm YBCO-123 thickness. Our results suggest an alternative framework, in terms of nanoscale phase inhomogeneity induced by heteroepitaxial lattice mismatch, for understanding the dependence of Tc on layer thicknesses in LCMO/YBCO-123 multilayers. The epitaxially-induced CuO intergrowths are also potentially useful as controlled inclusions to synthesize novel phases of the Y-Ba-Cu-O family, as further discussed in Chapter 5. Chapter 4

Strain-Induced Tc Reduction in Heteroepitaxial

Perovskite/YBa2Cu3O7−δ/Perovskite Trilayers

4.1 Introduction

In thin-film heterostructures of complex oxides, heteroepitaxial strain due to interfacial lattice mismatch can significantly affect their electronic properties, by virtue of thesen- sitivity to bond lengths and angles [117, 118]. Examples include enhancement of the

superconducting critical temperature (Tc) in superconducting cuprates [119, 120], variation of the Mott gap size in iridates [121], modulation of the conductivity in nickelates [122], and tunable carrier mobility at aluminate/titanite interfaces [123]. Although the heteroepitaxial strain in oxide thin films can be relieved by dislocations, cracking and formation of vacancies [124–127], studies of various oxides including cuprates and manganites have shown that the strain can extend well into the film, as far as ∼ 200 nm from

50 Chapter 4. Tc Reduction in Epitaxially Strained YBCO-123 Films 51

the interface [128–131]. In addition to lattice-parameter mismatch, lattice-symmetry mismatch is also known to affect the physical properties of oxide heterostructures via anisotropic strain [132].

As described in the previous chapter, in c-axis thin-film heterostructures comprising the high-Tc cuprate YBa2Cu3O7−δ (YBCO-123) and the half-metallic manganite

La2/3Ca1/3MnO3 (LCMO), heteroepitaxial strain can induce CuO intergrowths in the YBCO-123 layer [133]. Attributed to lattice-symmetry mismatch between the orthorhom-

bic YBCO-123 and the pseudocubic LCMO (see Table 4.1, aa and bb stands for the percentage lattice mismatch along the a and b direction respectively), these CuO intergrowths give rise to nanoscale phase inhomogeneity that can reduce the resistively- measured Tc as the YBCO-123 layer becomes thin. This strain-based mechanism of Tc attenuation has crucial implications for the study of ferromagnet/superconductor (F/S) proximity effect in these heterostructures [83, 85–87, 89, 91, 96, 134–140], as severalof these studies have interpreted the Tc dependence on layer thickness as crucial evidence for long-range proximity effect involving spin-triplet pairing [54, 57, 61, 141–143]. Since the purported range (∼ 10 - 40 nm) of this F/S proximity effect [85, 89] is within the length scale of epitaxial strain in such heterostructures, it is questionable whether they can simply be identified with the proximity effect. In fact, although a long-range F/S proximity effect is expected from theory based on spin-triplet pairing, the actual range of the proximity effect in c-axis LCMO/YBCO-123 heterostructures is still under experimental debate [59, 91].

In this chapter, we continue on with the results from the previous chapter, and further distinguish between the effects of strain and magnetism on the superconductivity in c-axis

LCMO/YBCO-123 heterostructures. We study the Tc of perovskite/YBCO-123/perovskite trilayer thin films with either ferromagnetic LCMO or paramagnetic LaNiO3 (LNO) as the clamping layers. The trilayer geometry was chosen to symmetrize the heterostructure. For comparison with trilayers that are lattice-symmetry matched, we Chapter 4. Tc Reduction in Epitaxially Strained YBCO-123 Films 52

Material Lattice a (Å) b (Å) a/b aa(%) bb(%) YBa2Cu3O7 orthorhombic 3.821 3.885 0.984 − − PrBa2Cu3O7 orthorhombic 3.868 3.911 0.989 +1.2 +0.7 YBa2Cu3O6.35 tetragonal 3.858 3.858 1 +1 −0.7 La2/3Ca1/3MnO3 pseudocubic 3.858 3.858 1 +1 −0.7 LaNiO3 pseudocubic 3.838 3.838 1 +0.5 −1.2

Table 4.1: Bulk lattice parameters for the oxides involved in this study. Oxygen-deficient YBa2Cu3O6.35, which is tetragonal, is included for comparison. Columns 2 and 3 show a-axis and b-axis lattice parameters. Column 5 shows the ratio between the a-axis and b-axis lattice parameters. Columns 6 and 7 show the % of lattice-parameter mismatch relative to YBa2Cu3O7; positive/negative value denotes tensile/compressive strain on YBa2Cu3O7.

substitute the pseudocubic perovskite with orthorhombic PrBa2Cu3O7−δ (PBCO) which is isostructural to YBCO-123.

4.2 Experimental

The trilayer thin films used in our study were grown on c-axis oriented SrTiO3 (STO) substrates by pulsed laser-ablated deposition (PLD). STO substrates are chosen for better comparison with previous studies in the literature [85]. A KrF excimer laser was used, operating at 248 nm, 2 - 5 Hz and a fluence of 2 J/cm2. The PLD growths were done at

◦ 750 - 800 C in 200 mTorr of O2. After deposition, each film was annealed in situ by slow

◦ cooling over 45 minutes to 300 C in 760 Torr of O2 to optimally oxygenate the YBCO-123 layer. For each trilayer sample, the YBCO-123 layer was clamped symmetrically between either PBCO, LCMO or LNO layers. For our resistance measurements, thickness of the clamping layers was fixed at 10.7 nm, while the YBCO-123 layer was varied between 5.4 and 21.4 nm in ∼ 5.4 nm increments. The range of relative thickness was deliberately chosen to tune the heteroepitaxial strain on the YBCO-123 layer.

Structural characterization of our thin-film samples was done by X-ray diffraction (XRD) using θ−2θ scans. The XRD data were taken with either a Bruker D8 DISCOVER or a Rigaku MiniFlex 600 diffractometer. Figure 4.1 plot the XRD data for unilayer films Chapter 4. Tc Reduction in Epitaxially Strained YBCO-123 Films 53

Figure 4.1: XRD patterns of unilayer thin films taken with θ − 2θ scans. Panels (a), (b), (c) and (d) plot the data for YBCO-123, PBCO, LCMO and LNO respectively, showing all the expected peaks for each material, without any impurity peaks. The XRD patterns are similar for films with the same lattice structure. of YBCO-123, PBCO, LCMO and LNO, showing the expected peaks in each film with no impurity peaks. Films with the same lattice structure have similar peak patterns. The starred peaks are due to radiation contamination of W Lα and Cu Kβ. The growth rate for each material was determined with X-ray reflectivity (XRR) and checked with atomic force microscopy. Finally, the electrical resistance of each trilayer film was measured versus temperature in a 4He cryostat, using standard ac lock-in technique in the four- contact configuration.

4.3 Results

Figure 4.2 shows the resistance R versus temperature T data for PBCO/YBCO-123/ PBCO trilayers of various YBCO-123 thickness. To facilitate comparison, each R-vs.-T Chapter 4. Tc Reduction in Epitaxially Strained YBCO-123 Films 54

Figure 4.2: Normalized resistance versus temperature of PBCO/YBCO-123/PBCO trilayer films, plotted for YBCO-123 layer thickness ranging from 5.4 to 16.1 nm.Allthe films show sharp superconducting transitions with similar Tc onsets near 90 K, attesting to the lattice-symmetry matching between YBCO-123 and PBCO. The mild attenuation of Tc versus decreasing YBCO-123 thickness can be understood in terms of the lattice- parameter mismatch, which stretches the a-axis of YBCO-123 more than its b-axis, thus slightly reducing its oxygen content. For comparison, the inset shows resistance versus temperature of unilayer YBCO-123 and PBCO films that are 21.4 nm thick. curve is normalized by its room-temperature value. All the trilayers show sharp superconducting transitions, with the Tc onset (defined as the intersection of the two linear fits before and after the transition) being mildly attenuated from 94 to 91 and 88 K as the

YBCO-123 thickness is decreased from 16.1 to 10.7 and 5.4 nm. Similar Tc attenuation was seen in previous studies of PBCO/YBCO-123/PBCO trilayers [144], and can be explained by the PBCO/YBCO-123 lattice-parameter mismatch, which stretches the a-axis of YBCO-123 more than its b-axis (direction of the CuO chains), thus slightly reducing its oxygen content. The mildness of this Tc attenuation attests to the lattice-symmetry matching between PBCO and YBCO-123, both being orthorhombic, in contrast to the LCMO/YBCO-123/LCMO and LNO/YBCO-123/LNO results presented below. It is Chapter 4. Tc Reduction in Epitaxially Strained YBCO-123 Films 55 worth noting that at 5.4-nm YBCO-123 thickness, there is a slope inflection near 210 K, which can be explained using a parallel-resistor model. Namely, since PBCO is insulating, PBCO/YBCO-123/PBCO trilayers would appear insulating above/below ∼ 210 K when the YBCO-123 layer is more/less resistive than the two PBCO layers in parallel.

Figure 4.3: Normalized resistance versus temperature of LCMO/YBCO-123/LCMO trilayer films. As YBCO-123 thickness is decreased from 21.4 to 5.4 nm,the Tc onset is strongly attenuated from 83 to 35 K, and the superconducting transition broadens. This attenuation of superconductivity versus decreasing YBCO-123 thickness also occurs in LNO/YBCO-123/LNO trilayers (Figure 4.4) but not in PBCO/YBCO-123/PBCO trilayers (Figure 4.2). At 5.4-nm YBCO-123 thickness, the slope inflection near 200 K can be explained using a parallel-resistor model for the trilayer, with the LCMO layer having a CMR peak near its Curie temperature. For comparison, the inset shows resistance versus temperature of unilayer YBCO-123 and LCMO films that are 21.4 nm thick.

Figure 4.3 shows the normalized R-vs.-T data for LCMO/YBCO-123/LCMO trilayers of various YBCO-123 thickness. As the YBCO-123 thickness is decreased from 21.4 nm to

5.4 nm, the Tc onset is strongly attenuated from 83 K to 35 K, and the superconducting transition broadens. A similar Tc attenuation was seen in our study of bilayer LCMO/

YBCO-123 films grown on (La, Sr)(Al, Ta)O3 (LSAT) substrates shown in the previous Chapter 4. Tc Reduction in Epitaxially Strained YBCO-123 Films 56

chapter: the Tc of a 25-nm YBCO-123 film is significantly reduced when a 25-nm LCMO layer is grown over it [133]. This Tc reduction was directly attributed to the intergrowth of double-CuO chains, induced by heteroepitaxial strain from both the LCMO overlayer and the LSAT substrate. Since LCMO and LSAT are both pseudocubic in structure and similar in lattice parameter (< 0.25 % mismatch), it is reasonable to assume that the strain effects on YBCO-123 are similar whether it is clamped between LCMO or between LCMO and LSAT. In our present study, it is worth noting that at 5.4-nm YBCO-123 thickness there is a slope inflection near ∼ 200 K. This inflection can also be explained using a parallel-resistor model, with the LCMO layer having a colossal magnetoresistance (CMR) peak near its Curie transition. The presence of such CMR peak suggests that the LCMO layers in the trilayers are still ferromagnetic, as previous x-ray measurements have shown that there is negligible influence of superconductivity on the magnetic properties of LCMO [136].

Figure 4.4 shows the normalized R-vs.-T data for LNO/YBCO-123/LNO trilayers of various YBCO-123 thickness. Apart from the absence of a CMR peak, all the LNO/ YBCO-123/LNO samples show similar behavior as their LCMO/YBCO-123/LCMO coun-

terparts (see 4.3). That is, the Tc onset is strongly attenuated from 82 K to 42 K, and the superconducting transition broadens, as the YBCO-123 thickness is decreased from 21.4 nm to 5.4 nm. In fact, except for the thinnest (5.4 nm) YBCO-123 case, the

Tc onsets and transition widths agree to within 2 K between the LNO/YBCO-123/LNO and LCMO/YBCO-123/LCMO data. This agreement is illustrated in Figure 4.5, which plots Tc versus YBCO-123 layer thickness for all three types of trilayers measured in this study. The similarity between the LNO/YBCO-123/LNO and LCMO/YBCO-123/

LCMO data indicates that the Tc attenuation is less associated with magnetism of the perovskites than with heteroepitaxial strain between these pseudocubic perovskites and orthorhombic YBCO-123. The contrast to the PBCO/YBCO-123/PBCO data is entirely consistent with the lattice-symmetry compatibility between YBCO-123 and PBCO, as Chapter 4. Tc Reduction in Epitaxially Strained YBCO-123 Films 57

Figure 4.4: Normalized resistance versus temperature of LNO/YBCO-123/LNO trilayer films. As YBCO-123 thickness is decreased from 21.4 to 5.4 nm,the Tc onset is strongly attenuated from 82 to 42 K, and the superconducting transition broadens. Apart from the absence of a CMR peak near 200 K and a sharper transition for the 5.4 nm case, the LNO/YBCO-123/LNO trilayers show similar behavior as their LCMO/YBCO-123/ LCMO counterparts (Figure 4.3). The similarities are more clearly illustrated in Figure 4.5. For comparison, the inset shows resistance versus temperature of unilayer YBCO- 123 and LNO films that are 21.4 nm thick.

described earlier. At 5.4-nm YBCO-123 thickness, it is possible that the lower Tc onset and broader transition for LCMO/YBCO-123/LCMO versus LNO/YBCO-123/LNO are due to F/S proximity coupling between LCMO and YBCO-123. However, this length scale is well below the purported range (∼ 10 - 20 nm) of the F/S proximity effect for c-axis LCMO/YBCO-123 heterostructures [85, 89, 134, 135], thus still calling this long range into question. Chapter 4. Tc Reduction in Epitaxially Strained YBCO-123 Films 58

4.4 Discussion and outlook

A closer examination of the lattice parameters listed in Table 4.1 sheds further light on the

various dependences of Tc versus YBCO-123 thickness shown in Figure 4.5. Whereas the heteroepitaxial strain of PBCO on YBCO-123 is consistently tensile in the ab-plane, the

strain of either LCMO or LNO on YBCO-123 is anisotropic, being tensile (aa > 0) along

its a-axis but compressive (bb < 0) along its b-axis. As shown in the previous chapter, this anisotropic strain can induce nanoscale formation of the Y2Ba4Cu7O15−δ (YBCO-

247) phase, which is less orthorhombic and tends to have lower Tc than YBa2Cu3O7−δ [34, 110, 111]. Even if the YBCO-247 phase does not robustly form, the remaining

YBa2Cu3O7−δ can still relieve the anisotropic strain by losing oxygen from its CuO

chains, which run along the compressed b-axis, thereby lowering Tc. Such a strain- relieving mechanism via formation of oxygen vacancies has been observed in a variety of

oxides [126, 127]. For oxygen-deficient YBa2Cu3O6.35, which is tetragonal, its in-plane lattice parameter (3.858 Å) is in fact the same as that of LCMO. Thus it is inherently unfavorable for the YBCO-123 within LCMO/YBCO-123/LCMO trilayers to remain strongly orthorhombic and fully oxygenated. In essence, by heteroepitaxially clamping orthorhombic YBCO-123 between pseudocubic perovskites in thin-film form, one can no longer ensure that the YBCO-123 layer retains either its phase purity or oxygen stoichiometry.

Furthermore, the elastic modulus of YBCO-123 is between 90 to 110 GPa [145–148]. This is smaller than the elastic modulus of either LCMO or LNO, which is between 150 to 190 GPa [149], and 300 GPa [150] respectively. Since YBCO-123 is softer than the perovskite clamping layers, the middle YBCO-123 layers are more easily deformed by the perovskite clamping layers than the other way around. On the other hand, PBCO is softer than YBCO-123 [151], thus in PBCO/YBCO-123/PBCO trilayers the strain effect on YBCO-123 is further reduced.

The above analysis implicates the anisotropic strain due to lattice-symmetry mis- Chapter 4. Tc Reduction in Epitaxially Strained YBCO-123 Films 59

Figure 4.5: Plot of Tc versus YBCO-123 layer thickness for all three types of trilayer films measured. Top/bottom end of each error bar corresponds to onset/completion of the superconducting transition. The similarity between LNO/YBCO-123/LNO and LCMO/YBCO-123/LCMO data indicates that the Tc attenuation is less associated with magnetism of the perovskite than with heteroepitaxial strain between these perovskites and the orthorhombic YBCO-123. The contrast to the PBCO/YBCO-123/PBCO data is entirely consistent with the lattice-symmetry matching between YBCO-123 and PBCO.

match in YBCO-123, more than the ferromagnetism of LCMO, as driving the attenuation of superconductivity in c-axis LCMO/YBCO-123 heterostructures. Although strain was reported to affect the ferromagnetic modulations seen in Pr-doped LCMO/YBCO- 123 superlattices [137], the lattice-symmetry mismatch between LCMO and YBCO-123 is a crucial experimental issue that has generally been overlooked. In addition to inducing ordered defects in the YBCO-123 lattice that can affect Tc [133], this symmetry mismatch may also affect the novel types of charge-density-wave order recently observed in LCMO/YBCO-123 bilayers and superlattices [98, 152], by affecting either the electronic nematicity or the interplay between multiple order parameters [153–155]. Further studies correlating structural, transport and spectroscopic measurements are called for Chapter 4. Tc Reduction in Epitaxially Strained YBCO-123 Films 60

to elucidate these possibilities. For our present study of trilayers, the predominant effect

of strain over magnetism implies that the long length scales of Tc attenuation observed in c-axis LCMO/YBCO-123 heterostructures cannot be identified with the F/S proximity effect. To avoid the anisotropic strain, the tetragonal La2−xSrxCuO4 (LSCO) could be used instead of orthorhombic YBCO-123 [92, 156, 157]. Specifically, a comparative study of c-axis LCMO/LSCO/LCMO and LNO/LSCO/LNO trilayers, both being interfacially- matched in lattice symmetry, would enable a clearer determination of the range of the c-axis F/S proximity effect in hole-doped manganite/cuprate heterostructures.

4.5 Conclusion

In summary, we have studied c-axis perovskite/YBCO-123/perovskite trilayer thin films, using either ferromagnetic LCMO or paramagnetic LNO as the clamping perovskite, to distinguish between the effects of strain and magnetism on the superconductivity. LCMO/YBCO-123/LCMO and LNO/YBCO-123/LNO trilayers show similarly strong attenuation of Tc, as YBCO-123 layer thickness is reduced in the range of 21.4 to 5.4 nm. PBCO/YBCO-123/PBCO trilayers, which are lattice-symmetry matched, show a much milder Tc attenuation. The Tc attenuation of the perovskite/YBCO-123/perovskite trilayers may be caused by either structural phase transformation observed in the previous chapter, or strain-induced oxygen vacancies in the CuO chains. These results indicate that heteroepitaxial strain, rather than long-range F/S proximity effect, is responsible for the long length scales of Tc attenuation observed in c-axis LCMO/YBCO-123 heterostructures. Chapter 5

Synthesis of High-Oxidation Y-Ba-Cu-O Phases in Superoxygenated Thin Films

5.1 Introduction

The superconducting critical temperature Tc of hole-doped cuprate tends to scale with their lattice complexity [7–9, 158, 159]. For the Y-Ba-Cu-O (YBCO) family, best known

for YBa2Cu3O7−δ (YBCO-123) [3, 99], earlier studies have shown that solid-state reaction of powder samples in high O2 pressure can stabilize the formation of higher-oxidation

phases such as Y2Ba4Cu7O15−δ (YBCO-247) and Y2Ba4Cu8O16 (YBCO-248) [34, 35, 100, 160, 161]. These phases are distinguished by double-CuO chains that either replace or alternate with single-CuO chains, resulting in longer unit cells along the c-axis [102,

162–164]. More complex phases, with higher ratios of CuO chains to CuO2 planes, are also believed to exist but have not yet been synthesized, likely due to limitations in thermodynamic stability [165–169]. Oxidation states of the various YBCO phases are listed in Table 5.1. The total cation valence is defined as the sum of valences for all the

61 Chapter 5. Superoxygenation of Y-Ba-Cu-O thin films 62

1 cations in one block of 2 (YBa2Cu3+x/2O7+x/2) that contains one CuO2 plane, and scales with the ratio of CuO chains to CuO2 planes. In this chapter, we extend this superoxygenation concept of oxide synthesis to YBCO thin films which, due to their large surface-to-volume ratio, are thermodynamically more reactive than powders or crystals [170]. Epitaxial thin films of YBCO-123 are grown on lattice-matched perovskite substrates by pulsed laser-ablated deposition (PLD), and then post-annealed in high O2 pressure both at and above the PLD growth temperature. The high-pressure annealing is done in conjunction with Cu enrichment by solid-state diffusion, to facilitate the inclusion of extra CuO chains. Scanning transmission electron microscopy (STEM), X-ray diffraction (XRD) and X-ray absorption spectroscopy (XAS) are used to characterize the films.

Name Formula Unit Total Cation Valence CuO/CuO2 Synthesized before YBCO-123 YBa2Cu3O7 +7 0.5 Yes YBCO-247 Y2Ba4Cu7O15 +7.5 0.75 Yes (bulk only) YBCO-248 Y2Ba4Cu8O16 +8 1 Yes YBCO-249 Y2Ba4Cu9O17 +8.5 1.25 No YBCO-125 YBa2Cu5O9 +9 1.5 No YBCO-126 YBa2Cu6O10 +10 2 No

Table 5.1: Known and possible phases of the YBCO family of cuprates, listed by nominal formula unit and ratio of cations. The total cation valence (relative to one CuO2) scales with the ratio of CuO chains to CuO2 planes. For simplicity, oxygen non-stoichiometry is not shown.

5.2 Experimental

The PLD apparatus used for our experiment is equipped with a KrF excimer laser operating at 248 nm, 2 Hz and 2 J/cm2. The ceramic YBCO-123 target used is > 99.9% pure and ∼ 93% dense. Thin films of YBCO-123 were grown on c-axis faces of

◦ (LaAlO3)0.3(Sr2TaAlO6)0.7 (LSAT) substrates, at 800 C in 200 mTorr of O2. The films were 50 nm thick and 2 nm smooth, as determined by STEM and atomic force microscopy.

After deposition, each film was annealed in-situ by cooling at 11 ◦C/min to 300◦C in Chapter 5. Superoxygenation of Y-Ba-Cu-O thin films 63

760 Torr of O2 to ensure proper oxygenation of YBCO-123. The resistively-measured Tc of our as-grown films are ∼ 91±1 K, comparable to optimally-doped YBCO-123 single crystals. The as-grown films were post-annealed with a commercial high-pressure (HP)

◦ furnace, at up to 700 atm O2 and 900 C, for 12 hours before being cooled in the HP oxygen. During the HP-annealing, each film was buried under YBCO-123 powder enriched with CuO powder, to enable solid-phase diffusion of the excess Cu. For the enrichment, CuO to YBCO-123 molar ratios of 2:1 and 4:1 were used, yielding HP-annealed films that are respectively denoted as Samples A and B in this chapter. These samples were

◦ annealed in 650 atm of O2 at 800 C, which we found was the optimal temperature for driving phase conversion without causing decomposition.

XRD measurements were done using the θ - 2θ method with a Bruker D8 DISCOVER X-ray diffractometer by Dr. Srebri Petrov at Department of Chemistry. STEM andelec- tron energy loss spectroscopy (EELS) measurements were made by Dr. Nicolas Gauquelin and Professor Gianluigi Botton at the Canadian Centre for Electron Microscopy (CCEM) with a FEI Titan 80-300 microscope. The microscope is fitted with a high-brightness field emission gun and CEOS hexapole-based aberration correctors for both condenser and objective lens aberrations, and operated at 200 keV in scanning mode. The resulting high-angle annular dark-field (HAADF) images are sensitive to atomic-number contrast. The elemental analysis with HAADF was corroborated with EELS spectra, which were recorded on a high-resolution Gatan GIF Tridiem spectrometer. XAS measurements were performed at the Canadian Light Source’s REIXS beamline by Professor David Hawthorn and Christopher McMahon from University of Waterloo [171]. The analysis of the XAS data was carried out by myself. Data was taken at room temperature for both Cu L edge and O K edge on all samples, with the orientation of the linear polarization aligned along both the in-plane (E k ab) and the out-of-plane (E k c) directions. Chapter 5. Superoxygenation of Y-Ba-Cu-O thin films 64

5.3 Results and discussions

Figure 5.1 compares the STEM and XRD data between our HP-annealed and as-grown films, the former showing significant conversion of YBCO-123 to both YBCO-247 and YBCO-248, as represented by the equations:

HP O2 2YBa2Cu3O7−δ + CuO −−−−→ Y2Ba4Cu7O15−δ (5.1) HP O2 2YBa2Cu3O7−δ + 2CuO −−−−→ Y2Ba4Cu8O16

Fig. 5.1a) shows the STEM image taken on the HP-annealed Sample A. The double- CuO chains appear as dark stripes and are labelled as D, while the single-CuO chains are labelled as S. In this image we see a gradation of phases, with YBCO-248 occurring near the surface, followed by YBCO-247 below that and then YBCO-123 near the film- substrate interface. This gradation of phases can be understood in terms of a natural tendency for chain-rich YBCO phases with higher CuO/CuO2 ratio to form closer to the film surface, since it is easier for both O and Cu ions to diffuse into the top partthaninto the bottom part of the film. On the other hand, the STEM image of a typical as-grown film in Fig. 5.1b) shows pure YBCO-123 containing only single-CuO chains. Double CuO-chain intergrowths were occasionally seen in some of the as-grown films, but only in trace amounts. Figure 5.2 shows the STEM image taken on 1-atm annealed films for comparison. Trace amount of double CuO chain intergrowths can be seen. In general the 1-atm annealed films and as-grown films have similar epitaxy and film purity within individual sample variations.

To corroborate the STEM images, Fig. 5.1c) and 5.1d) show the XRD patterns for Sample A and the as-grown film respectively. In these figures the peaks associated with the LSAT substrate are not labelled, since the locations of all the substrate peaks coincide with some of the YBCO peaks. For Sample A, all the peaks associated with YBCO-123, YBCO-247 and YBCO-248 are present, indicating a mixture of phases consistent with Chapter 5. Superoxygenation of Y-Ba-Cu-O thin films 65

Figure 5.1: Comparison of the STEM and XRD data between HP-annealed and as-grown films. (a) shows STEM image of the HP-annealed Sample A, showing a gradation (bottom to top) of phase conversion from YBCO-123 to YBCO-247 and YBCO-248, with single- CuO chains labelled as S and double-CuO chains labelled as D. (b) shows STEM image of an as-grown film, showing pure YBCO-123 containing only single-CuO chains. (c)and (d) show the XRD patterns of sample A and an as-grown film, respectively. The XRD data are consistent with the phases seen by STEM. Chapter 5. Superoxygenation of Y-Ba-Cu-O thin films 66

Figure 5.2: STEM image taken on a 1-atm annealed YBCO-123 film. The film is predominately YBCO-123 with trace amount of double CuO chain intergrowth.

the STEM image shown in Fig. 5.1a). All the YBCO-247 peaks are quite weak, even though YBCO-247 constitutes the majority of the film volume according to the STEM image. The weakness of these peaks can be understood in terms of the complex unit cell of YBCO-247, which have two sets of single-CuO chains and two sets of double-CuO chains. The c-axis lattice parameter of YBCO-247 (50.68 Å) is almost double that of YBCO- 248 (27.24 Å) and more than four time that of YBCO-123 (11.67 Å), making YBCO-247 inherently more difficult to detect by XRD. Coherent YBCO-247 diffraction peakdoes not form unless very large numbers single and double CuO chains are strictly ordered in alternating fashion. Specifically, Fig. 5.1a) shows only three complete unit cells YBCO- 247 with 12 CuO chains that can contribute to Bragg diffraction. Therefore, although YBCO-247 is the majority phase by volume here, YBCO-123 is the minority phase by number of unit cells. This difficulty in detecting YBCO-247 with XRD can explain why 247 peaks do not show up in the XRD plot for LCMO/YBCO-123 bilayers in Chapter 3. For the as-grown film, on the other hand, all the expected XRD peaks associated with the c-axis of YBCO-123 are present, and there are no impurity peaks to within Chapter 5. Superoxygenation of Y-Ba-Cu-O thin films 67 the resolution of our instrument. By relating the YBCO-123 (005)- and (007)-peaks with

2θ = 38.58◦ and 2θ = 55.08◦, respectively, we find a c-axis lattice parameter of 11.67 Å for our as-grown film, consistent with fully-oxygenated YBCO-123 [109, 172].

Figure 5.3: Comparison of XRD data between different HP-annealed YBCO-123 films, along with a HP-annealed powder serving as control. (a) and (b) show the data for Samples A and B, which were HP-annealed with Cu enrichment using 2:1 and 4:1 molar ratio of CuO to YBCO-123, respectively. (c) shows the data for YBCO-123 powder that was HP-annealed after being mixed with CuO powder in 4:1 molar ratio. Whereas higher Cu enrichment promoted conversion to chain-rich phases in the thin films, there is no sign of phase conversion in the powder sample.

The conversion to chain-rich phases induced by HP annealing in our YBCO-123 films is enhanced by higher Cu enrichment. Figure 5.3 compares the XRD data between different HP-annealed films, as well as HP-annealed powder serving as a control. Fig. Chapter 5. Superoxygenation of Y-Ba-Cu-O thin films 68

Figure 5.4: Comparison of the resistance versus temperature data between Sample A and Sample B, normalized relative to their room-temperature values. Sample B shows Tc ∼ 80 K, consistent with the predominance of YBCO-248 phase as indicated by both the STEM and XRD data. Sample A shows Tc ∼ 91 K, consistent with a mixture of phases where the significant presence of YBCO-123 provides higher-Tc conduction paths.

5.3a) and 5.3b) show the data for Samples A and B, which were HP-annealed with Cu enrichment using 2:1 and 4:1 molar ratio of CuO to YBCO-123, respectively. In contrast to the mixture of 123, 247 and 248 peaks seen in sample A, almost all the 247 peaks have disappeared from sample B, indicating more thorough conversion to the chain-rich

248 phase. As a further corroboration, Figure 5.4 compares the resistance (R) versus

temperature (T ) data between Sample A and Sample B. Sample B shows Tc ∼ 80 K,

consistent with the predominance of YBCO-248 phase, which is known to have Tc ∼ 80

K because of its inherent underdoping [173]. Sample A shows Tc ∼ 91 K, consistent with a mixture of phases where the significant presence of YBCO-123 provides higher-Tc conduction paths. The extrapolated zero-temperature residual resistivity ρ0 for both sample A and sample B are close to 0 µΩ-cm, and this implies our HP-annealed films are highly crystalline. Both the XRD data and the R vs. T data suggest higher CuO chemical potential during HP annealing drives the film towards more complete conversion to chain-rich phases. Finally, we note that there is no sign of phase conversion in powder YBCO-123 samples that were similarly HP-annealed, even for 4 times as long. As evident Chapter 5. Superoxygenation of Y-Ba-Cu-O thin films 69

by the XRD data in Fig. 5.3c), YBCO-123 powder that was HP-annealed for 2 days

◦ in 650 atm of O2 at 800 C, after being mixed with CuO powder in 4:1 molar ratio, shows only YBCO-123 and CuO peaks, just as before the annealing. Since the surface- to-volume ratio of films are 0.02 nm−1, which are more than 2 orders-of-magnitude larger

than our powders, the difference between the HP-annealed film versus powder provides strong evidence that the large surface-to-volume ratio of thin films is thermodynamically favoring the phase conversion.

Figure 5.5 shows the corrected and normalized XAS spectra taken on our as-grown and HP-annealed films. The E k ab data are shifted up for clarity. The mechanism for XAS is shown in the inset. XAS measures the unoccupied density of electronic states. Core electrons of the sample absorbs X-ray photons and are excited to unoccupied energy states, leaving behind core holes. When these core holes are filled by other electrons in the sample, energy is released either as fluorescence photons or as Auger electrons. Only X-ray photons with energy close to the binding energy of the core electrons are strongly absorbed, and these energies with strong absorption are shown as absorption edges in the XAS spectra. For our experiments, the photon absorption is measured with either total fluorescence yield (TFY) mode or total electron yield (TEY) mode. The TFYmode measures the total fluorescence photon created when core holes are filled, and theTEY mode measures the total ejected Auger electron and scattered secondary electrons. Since the interaction cross section of electrons in solid is much larger than that of photons, the TEY mode is much more surface-sensitive than the TFY mode. Only the top 7 - 10 nm of samples are probed with the TEY mode [174, 175]. For comparison, TFY can probe up to the top 100 - 500 nm of samples [174]. Only TFY mode is used for the current study on high-pressure annealed cuprates.

Fig. 5.5a) shows the Cu L3 absorption edge spectra, which arises from the 2p → 3d transition [176]. In the E k ab spectra, the main peak at 931 eV is mostly associated with the Cu(2) atoms in the planes, and do not show much variation between different samples. Chapter 5. Superoxygenation of Y-Ba-Cu-O thin films 70

Figure 5.5: Corrected and normalized XAS spectra taken on both as-grown and HP- oxygen annealed YBCO-123 films. (a) shows the spectra for theCu L3 edge, with a schematic diagram of the XAS process in the inset. (b) shows the spectra for the O K edge. The XAS spectra show that there are more chain states in the annealed films. There is also evidence that the CuO2 planes are underdoped in the HP-annealed films, with the sample containing YBCO-247 being the most underdoped. Chapter 5. Superoxygenation of Y-Ba-Cu-O thin films 71

Slightly above the main peak in the E k ab spectra, there is a shoulder at 932 eV that is caused by the presence of itinerant holes in the plane [177–180]. Surprisingly, our data show that both HP-annealed films have fewer holes in the planes than does the as-grown film, because the shoulder is weakened after annealing. Sample A, which contains both YBCO-247 and YBCO-248 in additional to YBCO-123, appears to be more underdoped in the planes than Sample B, which lacks YBCO-247. Since YBCO-248 is known to be inherently underdoped with little oxygen variability [173], it is reasonable that the as- grown film is less underdoped than any of the HP-annealed films, all of which contain YBCO-248. More interestingly, our data seem to suggest that the film containing YBCO- 247 is more underdoped than the films without YBCO-247 because the shoulder at 932eV is the weakest in Sample A. This observation could explain the generally lower Tc reported

in pure YBCO-247 samples [34, 100, 101, 110–112]. On the other hand, in the E k c data, the peak near 931 eV is mostly associated with the Cu(1) atoms in the CuO chains [181]. It is clear that Sample B has the strongest peak, followed by Sample A, and the as-grown film has the weakest peak. This observation is consistent with the natural expectation that higher Cu enrichment introduces more CuO chains into the YBCO lattice.

The O K edge absorption edge spectra is shown in Fig. 5.5b). The peaks at ∼ 529 eV and ∼ 530 eV in the E k ab data represent the Zhang-Rice singlet band (ZRSB) and the upper Hubbard band (UHB) respectively, and the increase of the spectral weight in the UHB indicates a reduction of holes in the planes [181, 182]. The as-grown sample has the weakest UHB spectral weight, while Sample A, which contains large regions of YBCO-247, has the strongest UHB. This observation is again consistent with the XAS data for the Cu L3 absorption edge, which shows that the sample containing pure YBCO- 123 is the least underdoped, and that the sample containing a mixture of YBCO-123,

YBCO-248 and YBCO-247 is the most underdoped. In the E k c spectrum, the peak at

∼ 529 eV represents the O 2pz states from apical O(4) sites [181], and this peak in both HP-annealed films are broader than in the as-grown film, implying that HP annealing Chapter 5. Superoxygenation of Y-Ba-Cu-O thin films 72 increases the number of holes in the apical sites.

In some of our HP-annealed films, regions with triple-CuO and quadruple-CuO chains were observed, as shown by the high-resolution STEM image in Figure 5.6. These regions constitute exotic YBCO phases with three and four CuO chains per unit cell, correspond- ing to YBCO-125 and YBCO-126 respectively, as illustrated in Fig. 5.6c) and 5.6d) by the lattice structures drawn with VESTA software [183]. YBCO-125 and YBCO- 126 have been seen as defect phases in trace amounts before [105, 167, 168, 184], but not yet been systematically synthesized despite attempts in up to 3000 bar of O2 [185]. Evidently, the unique thermodynamics of thin-film samples facilitates the formation of YBCO phases that do not form in bulk samples. The phase conversion from YBCO-123 to YBCO-125 and to YBCO-126 can be represented by the following equations:

HP O2 YBa2Cu3O7−δ + 2CuO −−−−→ YBa2Cu5O9−δ (5.2) HP O2 YBa2Cu3O7−δ + 3CuO −−−−→ YBa2Cu6O10−δ

The multiple CuO chains in the YBCO-125 and YBCO-126 phases offer new op- portunities for investigating the rich physics of the CuO chains in Y-Ba-Cu-O. First, a comparative study of these multi-chain phases will shed further light on the proximity- induced pairing in the CuO chains [27], and also help to clarify whether the quasi-1D ribbons formed by these chains can sustain pairing on their own [28, 29, 32]. Second, the multiple CuO chains may be used to tune the coupling strength between the CuO2 planes

[186], which is known to affect the Tc in cuprates [8, 9, 112, 119, 187]. Thus, successful isolation of these exotic Y-Ba-Cu-O phases will allow us to explore novel superconductivity in the CuO chains and potentially also to enhance Tc [188].

Another novel defect structure we observed in our HP-annealed films is the intergrowth of multiple Y layers as shown in Fig. 5.7. Such Y defect was never reported in YBCO compound before. Fig. 5.7a) shows a low-resolution STEM image of a HP- annealed film. The film is predominantly YBCO-123 near the substrate, and becomes Chapter 5. Superoxygenation of Y-Ba-Cu-O thin films 73

Figure 5.6: Evidence for phase formation to YBCO-125 and YBCO-126 in HP-annealed YBCO-123 films. The high-resolution STEM image in panel (a) shows regions withthree CuO chains per unit cell. The image in panel (b) shows regions with four CuO chains per unit cell. The lattice structures of YBCO-125 and YBCO-126 are shown in (c) and (d) respectively, with Y, Ba, Cu and O color-labelled as green, red, yellow and blue. Chapter 5. Superoxygenation of Y-Ba-Cu-O thin films 74

Figure 5.7: Evidence for double and triple Y defect in HP-annealed YBCO-123 films. The STEM image in panel (a) shows regions with two and three Y layers between CuO2 planes. The high-resolution electron energy loss spectroscopy (EELS) elemental map in panel (b) show thicker Y layers. The larger inter-Y layer distance and the brighter O K spectra at these defect regions suggest the presence of O2 layers between Y layers. The lattice structures of these phases with extra Y layers are shown in (c), with Y, Ba, Cu and O color-labelled as green, red, yellow and blue.

YBCO-247 away from the substrate, as our typical HP-annealed films. Between the YBCO-123 and YBCO-247 regions, we see an unique kind of defect with either two or three Y layers inserted between the CuO2 planes, as the distance between adjacent

CuO2 planes are doubled and tripled respectively. The structure of the defect region is color-labeled, with green, red, and yellow for Y, Ba, and Cu respectively. Fig. 5.7b) shows a high-resolution atomic-resolved electron energy loss spectroscopy (EELS) elemental maps in the region containing these Y layer intergrowths. The Y atom map clearly shows that the brightness of Y layers increase significantly for the double and Chapter 5. Superoxygenation of Y-Ba-Cu-O thin films 75 triple Y layers compared with single Y layer. In in these double and triple Y layers, the spacing between consecutive Y layers is measured to be ∼ 2.8 Å, which is almost twice as large as the spacing of 1.418 Å between Y layer and the CuO2 planes. Thus we can infer there is a layer of O2 between consecutive Y layers. Even though the

O2 layers cannot be clearly seen in HAADF and EELS images because of the small atomic number of oxygen, the presence of extra O2 layers between Y atoms is strongly hinted by the EELS O K edge map, which shows a increase in brightness with increasing number of Y layers. These Y-O2-Y and Y-O2-Y-O2-Y blocks between CuO2 planes take the fluoride structure, just like the (Ho1/3Ce2/3)O2 blocks in the Pb-base cuprate

(Pb1/2Cu1/2)Sr2(Ho1/3Ce2/3)3Cu2O11+z (Pb-1232) [189]. The YBCO unit cells with the double and triple Y layers have the nominal chemical formula of Y2Ba2Cu3O9−δ (YBCO-

223) and Y3Ba2Cu3O11−δ (YBCO-323) respectively. The structures of both YBCO-223 and YBCO-323 are shown in Fig. 5.7c). Since the high-Tc cuprates that contain these fluorite-type defects are predominantly tetragonal, we can infer that the presence ofthese defects in YBCO-123 tends to weaken its orthorhombicity, thereby reducing the oxygen content and thus the hole concentration in the CuO chains. The origin of these Y defects may be related to the extra CuO chains we see elsewhere in the film, because these Y defects are only observed in HP-annealed films that contain extra CuO-chain intergrowths. When YBCO-123 converts to these chain-rich phases during annealing, the chain-rich parts of the film take extra Cu from other parts of the film to incorpo- rate into the chains, and there is excess Y and Ba left behind in these other parts of the film. These excess Y and Ba atoms can order together to form novel defect structures. Thus it is likely the growth of this Y defect can be deliberately encouraged by burying films with excess Y during annealing to provide extra Y chemical potential, just likethe excess CuO can drive the formation of double and triple CuO chains.

An interesting observation of the double Y defect is the breaking of inversion symmetry about the Y plane via the zigzag ribbon structure. This structural peculiarity Chapter 5. Superoxygenation of Y-Ba-Cu-O thin films 76

is physically significant, in that the resulting lack of a center of symmetry alongthe

c-axis could profoundly affect the electron pairing in the CuO2 layers [190, 191], either by allowing spin-singlet/triplet mixing or by introducing a Rashba spin-orbit term in the pairing Hamiltonian, especially when strong electron correlations are present. Prior studies of such non-centrosymmetric superconductors have been limited to a few intermetallic

materials with rather low Tc, such as CePt3Si, Li2Pd3B and KOs2O6 [192–194]. Fur- thermore, these YBCO-223 and YBCO-323 phase with Y defects may potentially have

higher Tc, as recent works have shown that the increased intra CuO2 bilayer distance can lead to significant Tc enhancement [188]. Our observation of these Y defect structures is potentially useful as a building block for synthesizing non-centrosymmetric high-Tc superconductors, as well as investigating the effect of tuning inter CuO2 plane coupling on

Tc.

At present, these exotic YBCO phases in our thin-film samples are not yet robust enough for their Tc to be determined resistively. To better isolate these phases, our superoxygenation technique will need to be refined. Some of the refinement strategies we are pursuing include: 1) varying the annealing pressure and temperature; 2) incorporating excess CuO directly into the PLD target; 3) varying the film thickness; and 4) using alternative substrates with different lattice parameters to tune the heteroepitaxial strain. The first two strategies are guided by thermodynamic and kinetic considerations, towards finding higher-purity regions in an expanded phase diagram. The latter two strategies are motivated by our STEM observation that more phase conversion tends to occur closer to the film surface, which suggests that ionic diffusivity and lattice strain [133] mayalso affect the phase stability. Chapter 5. Superoxygenation of Y-Ba-Cu-O thin films 77

5.4 Conclusion

In summary, we have demonstrated that high-oxidation phases of YBCO can be synthesized in epitaxial thin-film samples via superoxygenation. The YBCO-123 thin films were

◦ grown by PLD on LSAT substrate, and then annealed at up to 700 atm O2 and 900 C in conjunction with Cu enrichment by solid-state diffusion. STEM revealed clear formation of YBCO-247 and YBCO-248 phases in the high-pressure oxygen annealed films. Regions of YBCO-125 and YBCO-126 phases having triple-CuO and quadruple-CuO chains, as well as YBCO-223 and YBCO-323 having double and triple Y sandwiched between the

CuO2 plane layers, were also seen on the STEM images of some of the annealed films. XRD and XAS measurements showed that higher Cu enrichment resulted in greater phase conversion to higher-oxidation phases, and that similarly annealed YBCO-123 powders do not undergo phase conversion. As proof of concept, our results open a novel route of synthesis towards discovering more complex phases of high-Tc cuprates and other superconducting oxides. Chapter 6

Phase Transformation and Hole

Doping in Superoxygenated Sr2IrO4 Thin Films

6.1 Introduction

The 5d transition metal oxide Sr2IrO4 (SIO-214) has attracted much interest recently due to novel physics arise from the interplay between its strong spin-orbit coupling, electron correlation, and crystal field [72]. A particular important feature of SIO-214 isits similarity with the cuprate superconductor La2CuO4. SIO-214 has the same layered perovskite structure as La2CuO4, with the corner-sharing IrO6 octahedra forming a square lattice at each layer [195, 196]. Large body of theoretical and experimental works has demonstrated the analogous electronic structure [64–67] and magnetic excitation spectra [68, 69] between the iridates and the cuprates, with the roles of electron and holes swapped, because a sign difference between the two systems in one of the tight-binding parameters [64, 197, 198]. Superconductivity has been predicted in the iridate system with either electron-doping [64, 199] or hole-doping [200, 201]. Electron doping on SIO-214

78 Chapter 6. Superoxygenation of SIO-214 thin films 79 has been attempted with either cation substitution [202], surface dosing [65], or vacuum annealing [203]. A metal-to-insulator transition with up to 9 orders of magnitude drop in the low-temperature (T = 1.8 K) resistivity [203], a V-shaped gap in the density of states [204, 205], and Fermi arcs [65] that are reminiscent of d-wave superconductivity have been observed recently in electron-doped SIO-214. However conclusive evidence of superconductivity is still elusive. Despite the potential higher superconducting transition temperature predicted in the hole-doped side [200, 201], hole-doping of SIO-214 is less explored compared to electron-doping, particularly because it is difficult to introduce large amount of holes controllably and cleanly into SIO-214 [206–209].

Superoxygenation techniques have been extensively used to introduce holes into oxides such as the cuprates, the nickelates, and the ruthenates [210–215]. Because of the presence of interstitial sites [216–218], up to 0.13 oxygens per unit cell can be introduced in La2CuO4 [210] to give access to a large doping range. In the ruthenates, as much as 0.25 excess oxygens per unit cell was reported to fit in the interstitial sites of Sr2RuO4 [211]. Because of the similar ionic sizes between Ru4+ and Ir4+ (0.65 Å vs. 0.66 Å [219]), we would expect it is also relatively easy for Sr2IrO4 lattice to accommodate excess oxygens. In addition, as we have seen in the previous chapter, superoxygenation of thin films can rearrange cations to induce structural phase transition. The phase-converted films can have different doping levels and different electronic properties from the as-grown films.

In this chapter, we extend the superoxygenation technique to SIO-214 thin films, in an effort to metallize Ir-based perovskites through hole doping. Thin films were chosenfor their large surface-to-volume ratio to enhance reactivity, and the annealing temperature was kept low to minimize cation interdiffusion. Chapter 6. Superoxygenation of SIO-214 thin films 80

6.2 Experimental

Sr2IrO4 films were epitaxially grown on STO substrates with pulsed laser deposition (PLD) technique by Professor Ambrose Seo’s group from University of Kentucky as described in [220]. For this study both 20 nm and 100 nm films were used. The as-grown films were post-annealed with the commercial Morris Research HPS-5015 HP furnace,

◦ in either 400 atm or 1 atm of O2 at 400 C, for up to 7 days before being cooled in the oxygen. Annealing temperature is kept low to minimize cation interdiffusion between

film and substrate. Resistivity of the films were measured versus temperature using a Lakeshore 370 AC resistance bridge, with contacts applied to the corners of the samples in van der Pauw geometry. Hall Effect measurements were carried out with a Physical Property Measurement System (PPMS) at California State University Los Angeles as described in Section 2.4, with assistance from Bo Truong and Professor Guo-Meng Zhao.

STEM measurements were made by Professor Gianluigi Botton at the Canadian Cen- tre for Electron Microscopy (CCEM) with a FEI Titan 80-300 microscope. XRD was carried out on the films using the θ-2θ method with either Cu Kα or Mo Kα radiation by Dr. Patrick Clancy, Dr. Raiden Acosta and myself. Finally, both the total fluorescence yield (TFY) and the total electron yield (TEY) of XAS spectra for the O K edge were measured at the Canadian Light Source’s REIXS beamline at room temperature by Professor David Hawthorn and Christopher McMahon from University of Waterloo as a collaboration. Spectra with the orientation of the linear polarization aligned along both the in-plane (E kab) and the out-of-plane (E kc) directions were measured. The analysis of the XAS data was carried out by myself.

6.3 Results and discussions

Figure 6.1 compared the resistivity (ρ) vs. temperature (T ) data of our SIO-214 films annealed under different conditions. Fig. 6.1a) shows the data for 20 nm films.Com- Chapter 6. Superoxygenation of SIO-214 thin films 81

pared with the ∼ 100Ω − cm room-temperature resistivity of as-grown samples, the HP- annealed films shows progressively more drop in room-temperature resistivity from 2to 3 orders-of-magnitude when the annealing time increases from 2 to 7 days. In particular the 7-day HP annealed film becomes weakly metallic with a positive dρ/dT and room- temperature ρ of ∼ 10−3 Ω − cm. Both the value and the slope of ρ are similar to that of

SrIrO3 (SIO-113) films grown on STO [221]. Such low room-temperature ρ and metallic ρ vs. T behavior has not been observed in SIO-214 before, and is only associated with extended HP oxygen annealing. After the 7-day HP-annealed sample is exposed to air for ∼ 6 months, it becomes non-metallic and has 1 order-of-magnitude increase in resistivity. Such aging effect has been observed before on pure SIO-113 films grownon SrTiO3 [222], and was attributed to the instability of SIO-113 in air. By comparison, the 1-atm 7 days annealed film showed only ∼ 1 order-of-magnitude drop in room-temperature ρ with highly insulating ρ vs. T behavior.

Fig. 6.1b) shows the data for 100 nm films. Unlike the 20 nm film, the 100nm film is still insulating after 7 days of HP annealing, even though the room-temperature and low-temperature ρ dropped by ∼ 2 orders and ∼ 5 orders of magnitude respectively compared with as-grown samples [121]. The ratio between ρ(10K) and ρ(290K) is only ∼ 5, compared with the typical ratio of ∼ 104 for as-grown samples. Thus extended HP annealing has made 100 nm films less insulating than as-grown samples. There is some drop in room-temperature ρ after 1-atm O2 annealing, however the effect is much smaller than HP annealing. The room-temperature ρ was about the same for the 20 nm

and 100 nm films after 1-atm O2 annealing for 7 days. Finally, it is worth noting the large drop in ρ induced by HP oxygen annealing is only seen in thin films and not in powder samples, even after much longer period of annealing. The ρ vs. T data of powder

Sr2IrO4 samples annealed in HP O2 for 30 days is plotted as dashed lines in both Fig. 6.1a) and 6.1b) as controls. The resistivity at all temperature is significantly higher than any of the annealed films, even though the annealing time was more than 4 times longer Chapter 6. Superoxygenation of SIO-214 thin films 82

Figure 6.1: Comparison of ρ vs. T data for SIO-214 films annealed under different conditions. Data for HP-annealed powder is also shown as dashed lines as controls. (a) shows data for 20 nm films. The HP-annealed films show progressively lower ρ than the 1-atm annealed film as the annealing time increases. The 7-day HP annealed film exhibits metallic dρ/dT and up to 3 orders-of-magnitude drop in room-temperature ρ. The 7-day HP annealed sample becomes non-metallic with 1 order-of-magnitude increase in ρ after exposed to air for ∼ 6 months. (b) shows data for 100 nm films. The HP-annealed film is again less resistive than the 1-atm annealed film, but the drop in ρ is not as significant as the 20 nm film after HP-annealing. There is minimum drop in ρ for the HP-annealed powder, even though the annealing time is 4 times longer than the films. than the thin-film samples.

Figure 6.2 shows the Hall effect and magnetoresistance measurements of a different 7-day HP-annealed 20 nm film made from a different batch. The ρ vs. T plot is shown in Fig. 6.2a) for comparison with the sample in Figure 6.1. It is important to note that the measurement was carried out more than 1 month after the annealing procedure. During this time the room-temperature resistivity has increased by a factor of ∼ 8. Thus the larger ρ of this sample can be explained by the aging effect. It is quite likely the low-temperature resistivity upturn is also caused by the aging effect. Such upturn is commonly seen when a metallic material approaches metal-to-insulator transition, and was reported for pure SIO-113 films before [221]. A freshly HP-annealed sample Chapter 6. Superoxygenation of SIO-214 thin films 83

Figure 6.2: Hall effect and magnetoresistance measurements of another 7-day HP- annealed 20 nm film. The ρ vs. T plot is shown in panel (a) for comparison with the sample in Figure 6.1. This sample is partially aged, and showing increased resistivity compared to freshly annealed sample. A low-temperature resistivity upturn is fitted to a logarithmic line in red. Panel (b) shows the resistivity vs field plot at 5 K. The V-shaped positive magnetoresistance is commonly observed in SIO-113 samples. Panel (c) and (d) shows the transverse resistance Rxy in two different wiring configurations at 40 K. In the schematic diagram of the wiring configurations shown in the inset, the field points outof the page. Both of these two panels show the majority charge carriers are holes. Chapter 6. Superoxygenation of SIO-214 thin films 84 is fully metallic down to 4 K, and a 6-month aged sample is entirely insulating, thus a metal-to-insulator transition happens at an intermediate time. The resistivity upturn is usually attributed to two-dimensional weak localization [221, 223–225]. The resistivity would scale as log(T ) at low temperature. A log(T ) curve was fitted to the data in red, and it is almost a perfect fit. Fig. 6.2b) shows the resistivity versus field measurement taken at 5 K. The V-shaped positive magnetoresistance is commonly observed in SIO-113 samples [226, 227]. Fig. 6.2c) and 6.2d) show the Rxy vs. magnetic field (H) data taken at 40 K with two different wiring configurations. From the slopes of the graphs, both configurations clearly show that the majority carriers are holes.

The drop in resistivity on some of the HP-annealed SIO-214 films can be attributed to phase conversion to SIO-113. Figure 6.3 shows the STEM images taken on the SIO-214 films annealed with different conditions. The images were taken with high-angle annular dark-field (HAADF) mode, in which the contrast is sensitive to the atomic numbers. Some of the representative atoms were color-labeled to show the lattice structure. Fig. 6.3a) and 6.3b) show the 2-day HP annealed 20 nm film and the 7-day 1-atm annealed 100 film respectively, both of which contain pure, defect-free SIO-214.

Fig. 6.3c) shows the 7-day HP-annealed 20 nm film. Only SIO-113 can be seen from the STEM images with no trace of SIO-214 left. Such phase transformation from SIO-214 to SIO-113 can be understood in terms of the following equation in Kröger-Vink notation [228]:

0000 Sr2IrO4 + O2 −→ 2SrIrO3 + VIr (6.1)

Because the Sr:Ir ratio in SIO-214 is 1:0.5, while the ratio is 1:1 in SIO-113, SIO-113 lattice transformed from SIO-214 must have Ir vacancies. In the STEM image, the Ir vacancies show up as darker atoms at the Ir sites. Since each Ir vacancy carries -4 charge, in order to maintain charge neutrality, some of the Ir ions need to be oxidized to higher oxidation states, which constitutes hole-doping. Our collaborator Charles Zhang performed elemental analysis on similarly annealed SIO-214 films grown on GdScO3 Chapter 6. Superoxygenation of SIO-214 thin films 85

Figure 6.3: HAADF-STEM images of differently annealed SIO-214 films. (a) and(b) show the 2-day HP annealed 20 nm film and the 7-day 1-atm annealed 100 film respectively, both of which contain pure SIO-214. (c) shows the 7-day HP annealed 20 nm film. There is no trace of SIO-214 left in the STEM image and only SIO-113 can be seen. The brightness of Ir atoms is periodic along the c-axis, which can be explained by ordering of Ir vacancies. (d) shows the 7-day HP annealed 100 nm film, which mainly consists of SIO-214 with trace amount of SIO-113 defects away from the substrate. Same amount of film-substrate interdiffusion by up to 2 unit cells is seen in all films and isindependent of annealing conditions. The lattice structures of SIO-214 and SIO-113 are shown as in- sets in (a) and (c), with Ir, Sr, Ti, and O color-labelled as green, purple, orange, and red respectively. Chapter 6. Superoxygenation of SIO-214 thin films 86

and NdGaO3 substrates, using energy dispersive spectroscopy (EDX). These particular substrates were chosen because they contain neither Sr nor Ir, thus signals from the substrates do not interfere with determining the Sr/Ir ratio in the films. The EDX results suggest the HP-annealed film indeed has about 10% less Ir than Sr, giving an empirical

formula of SrIr0.9O3 for the converted SIO-113 films. However the large uncertainly on the EDX results at ∼ 10% makes exact determination of the Ir vacancies difficult. In the STEM image, the brightness of Ir atoms forms a periodic structure along the c-axis with a period of 3 unit cells, and this suggest the Ir vacancies form some kind of super-latticing along the c-axis. In the converted SIO-113, the average c-axis lattice parameter is elongated compared with the bulk lattice parameter of 3.96 Å by ∼ 4%,

and larger than the c-axis lattice parameter of pure SIO-113 film on STO by ∼ 2% [221]. Preliminary synchrotron XRD performed by our collaborator Dr. Sae Hwan Chun re- confirms both the period-of-3 supper-latticing and the 4% lattice parameter elongation along the c-axis as seen in Figure 6.4. At this point it is not clear whether such increase in the lattice parameter is due to the super-latticing, interstitial oxygens, or incomplete phase transformation from SIO-214 to SIO-113.

Fig. 6.3d) shows STEM image taken on the 7-day HP-annealed 100 nm film. Unlike the 20 nm film, not much phase transformation is seen. The film predominately consists of SIO-214 with trace amount of SIO-113 defects, which occurs away from the substrates. The brightness periodicity of Ir atoms, which is present in the 7-day HP-annealed 20 nm film, cannot be seen in the SIO-113 defect regions. The lack of the brightness periodicity in 100 nm films may be because there is too little SIO-113 converted from SIO-214, and there are not enough Ir vacancies to order. The extra Ir atoms needed for phase transformation to SIO-113 may come from SIO-214 regions, and there are Ir vacancies in both SIO-113 and SIO-214 regions. Thus the distribution of Ir vacancies is further diluted to span the entire film and not just the SIO-113 regions. Finally, the film-substrate interfaces consist of IrO2-SrO of SIO-214 joining TiO2 of STO in all the four samples. Chapter 6. Superoxygenation of SIO-214 thin films 87

Figure 6.4: Comparison of synchrotron XRD data taken on two iridate films annealed under different pressure. The high-pressure annealed film show clear evidence ofperiod- of-3 supper-latticing and lattice parameter elongation along the c-axis, both of which are seen in STEM images. There is no evidence of structural transition in the 1 atm annealed film compared to as-grown film. The data is courtesy of Dr. Sae HwanChun.

Such interfaces were recently reported on films grown by both PLD and molecular beam epitaxy (MBE) [229, 230], and were shown to be thermodynamically stable [230]. It is worth noting that on the STEM images for all the samples, there are evidences of interdiffusion across the film-substrate interface by up to two unit cells. However the degrees of interdiffusion are independent of annealing time and pressure. In particular the less resistive, extendedly annealed films do not show more interdiffusion than themore resistive films that are annealed at lower pressure or for shorter duration, indicating the superoxygenation-induced drop in resistivity is not due to cation interdiffusion with the substrate.

We carry out XRD and XAS measurements to probe the pervasiveness of the phase transformation from SIO-214 to SIO-113. Figure 6.5 shows the XRD patterns taken on our iridate films. To facilitate comparison, the 2θ values for samples measured with Mo Kα radiation are converted to 2θ values that would be obtained from Cu Kα radiation Chapter 6. Superoxygenation of SIO-214 thin films 88

Figure 6.5: XRD patterns of the annealed SIO-214 films together with HP-annealed SIO-214 powder serving as control. (a) shows the data for 20 nm films. The SIO-214 peaks weaken with one peak disappearing after 2 days of HP-annealing, and all the SIO-214 peaks are gone after 7 days of HP-annealing. (b) shows the 100 nm films, and 7 days of either 1-atm or HP annealing does not eliminate any SIO-214 peaks. In both panels, peaks labelled with asterisk are identified with Al peaks from the sample holders. There is no sign of phase conversion in the powder sample even after 30 days of HP-annealing. Chapter 6. Superoxygenation of SIO-214 thin films 89

λCu using the formula 2θCu = 2 arcsin ( sin θMo), where λCu = 1.54 Å and λMo = 0.71 Å. λMo The data for 20 nm films are shown in panel (a). In the HP, 7-day annealed film,no peak associated with SIO-214 can be identified, and this is consistent with the metallic resistivity and the lack of SIO-214 in STEM images. Because of the close lattice constants between the pseudocubic SIO-113 and the STO substrate, the peaks associated with SIO-113 cannot be distinguished from the substrate peaks. Besides the extra Al peaks coming from the aluminum sample holder, no other impurity peak can be identified. Because these XRD data were taken with low-intensity X-ray, the c-axis super-latticing that was observed in STEM and synchrotron XRD is not detectable here. On the 2-day HP-annealed film, the SIO-214 peaks are weakened, and in particular the (008) peak is undetectable. This suggests there might be partial transformation from SIO-214 to SIO-113 or degradation of crystallinity in the 2-day HP-annealed sample that is not seen by STEM. On the 7-day 1-atm annealed sample, the degree of phase conversion, if there is any, is smaller than on the 2-day HP annealed sample, as all the SIO-214 peaks remain quit strong.

Fig. 6.5b) shows the XRD data for the 100 nm films. None of the XRD plots shows the disappearance of any SIO-214 peaks, and this indicates SIO-214 still remains in the 100 nm films after annealing. On the 1-atm annealed sample, the SIO-214 peaks are weakened compared with the as-grown sample and the HP-annealed sample. This indicates either more thorough conversion toward SIO-113 or more disorders in the 1-atm annealed sample than the HP-annealed sample. The later case is more likely, since more conversion toward the metallic SIO-113 would suggest lower ρ in the 1-atm annealed sample compared with the HP-annealed sample, contradicting the resistivity measurements. Finally, the XRD pattern for the powder sample that were HP-annealed for 30 days is shown as grey lines in both panel a) and b) for comparison. All the peaks present can be identified in terms of SIO-214 and the XRD pattern is indistinguishable from the XRD pattern for as-grown SIO-214 powders. There is no evidence of any phase transformation in powder sample Chapter 6. Superoxygenation of SIO-214 thin films 90

even though the annealing time is much longer than the films samples.

Figure 6.6 compares the XAS spectra for SIO-214 films annealed under different conditions, together with the XAS spectra for annealed SIO-214 powder and STO substrates. The spectra were normalized by the intensity values at 528.5 eV and 564 eV, which are in the flat parts of the spectra before and after the main peaks. The TFYspec- tra are shifted up for clarity. According to previous thickness-dependent XAS studies on SIO-214 films, the locations of the XAS peaks do not change as the film thickness is reduced [121]. Thus the spectra for the as-grown 100 nm film shown in Panel e) is used as a control. Following dipole selection rules, the peak at ∼ 531.5 eV in the in-plane

(E k ab) spectra correspond to the planar O 2px,y hybridized with Ir 5dxy orbital, and

in the out-of-plane (E k c) spectra correspond to the planar O 2pz hybridized with Ir

5dyz,zx orbital [121]. For both the TFY and TEY curves, this peak for the out-of-plane spectra is higher than the in-plane spectra. This dichroism at ∼ 531.5 eV is common in all layered perovskite of the 214 structure [231, 232], and is attributed to the larger

number of holes in the O 2pz-Ir 5dyz,zx bands than the O 2px,y-Ir 5dxy band [232]. In the in-plane spectra, the peak at ∼ 530.8 eV is associated with apical oxygen.

Fig. 6.6a) shows the spectra for the 2-day HP annealed 20 nm film. In TFY spectra, the peaks at ∼ 531.5 eV broadens and shifts to lower energy by ∼ 0.1 eV. Similar shift has been observed on the XAS spectra for Rh-doped SIO-214 samples before [233], and was attributed to hole-doping of SIO-214. The hole-doping is likely caused by the presence of interstitial oxygens, which distort the apical oxygens [218], and can reduce the hybridization of the O 2pz orbital with Ir 5dyz,zx orbital. As a result, the peak at ∼ 530.8 eV that’s associated with apical oxygen is missing. Fig. 6.6b) shows the data for the 7-day 1-atm-annealed 100 nm film. Both the TFY and the TEY spectra are almost identical to the as-grown 100 nm film shown in Fig. 6.6e). This indicates lower pressure annealing on thicker films has minimum effect. This is in contrast to the weakened SIO-214 peaks seen in XRD, and again suggests the weaker peaks in XRD is more likely Chapter 6. Superoxygenation of SIO-214 thin films 91

Figure 6.6: Normalized XAS spectra taken on annealed SIO-214 films together with the data for SIO-214 powder and STO substrate. (a) shows the data for the 2-day HP annealed 20 nm film. The peaks at ∼ 531.5 eV broadens and shifts to lower energy by ∼ 0.1 eV, and can be a signature of hole-doping. (b) shows the data for the 7-day 1-atm- annealed 100 nm film, which is virtually indistinguishable from spectra for as-grown SIO-214 nm films. (c) shows the data on the 7-day HP annealed 20 nm film, andthe spectra is similar to the spectra for SIO-113 films with no signature for SIO-214. The small dichroism may be caused by super-latticing and tetragonal distortion. (d) shows the data for the 7-day HP annealed 100 nm film. The spectra show mixture of SIO-113 and SIO-214 peaks. The shift of the peaks at ∼ 531.5 eV to lower energy and the pre- edge at ∼ 530 eV indicates hole-doping of SIO-214 by interstitial oxygens. (e) shows the spectra for an as-grown 100 nm SIO-214 as control. (f) shows the spectra for the powder samples and STO substrate. The powder still remains SIO-214 after 30 days of annealing. Chapter 6. Superoxygenation of SIO-214 thin films 92 caused by crystalline disorders such as grain boundaries and stacking faults rather than phase conversion.

Fig. 6.6c) shows the data on the 7-day HP annealed 20 nm film. There is no signature of SIO-214 from both of the TFY and TEY spectra [72, 121], and the spectra resemble that of SIO-113 except the dichroism [234]. The peak at ∼ 531 eV from the in-plane spectra becomes stronger than from the out-of-plane spectra. This dichroism has the opposite sign as the dichroism of SIO-214, thus cannot come from SIO-214 in the film. Pure SIO-113 has negligible dichroism because of its the cubic symmetry [121, 235], as there is little distinction between the c-axis and the ab-axis of the lattice. Thus this dichroism is most likely caused by the super-latticing and tetragonal distortion seen in the STEM images. Fig. 6.6d) shows the data on the 7-day HP annealed 100 nm film. The TEY spectra become polarization-independent and resemble the spectra for SIO-113, while the TFY spectra still show significant dichroism. Since TEY spectra only probe the film surface not exceeding 10 nm [174, 175], the change in the TEY spectra indicates the surface of the 100 nm film has transformed to SIO-113 while the part below surface remains mostly SIO-214. Such phase transformation near the film surface after HP-annealing has been seen in cuprates as discussed in the previous chapter, as it is easier for oxygens to diffuse into the top part than into the bottom part of the films [215]. There are four kinds of changes on the TFY spectra on the 100 nm film after HP-annealing. The first two changes are the reduction of dichroism and the appearance of SIO-113 peak between 530.8 eV and 531.5 eV. These two changes in the spectra are caused by the partial conversion from SIO-214 to SIO-113. More interestingly, the next two changes in the spectra consist of a red-shift of the peak at ∼ 531.5 eV that are similar as seen in the 2-day HP annealed 20 nm film, and appearance of a pre-peak at ∼ 530 eV. This pre-peak has been previously observed in superoxygenated La2NiO4+δ [236], and was attributed to interstitial oxygens. Thus both of these changes in the spectra are likely to be evidence for hole-doping via the introduction of interstitial oxygens to SIO-214. Chapter 6. Superoxygenation of SIO-214 thin films 93

The XAS spectra for powder SIO-214 samples are shown in Fig. 6.6f) for comparison. Even after 30 days of extended HP-annealing, the XAS spectra for the powder sample still resembles SIO-214 more than SIO-113. The lack of the SIO-113 peak at ∼ 531 eV between the two SIO-214 peaks suggest negligible conversion to SIO-113. Compared with the as-grown powder, the height of the peak at 530.8 eV relative to the peak at 531.5 eV is strengthened. Such transfer of spectral weight is associated with increasing number of holes [237]. The peak at ∼ 531.5 eV again shifts to lower energy, which indicates hole doping. The pre-edge at ∼ 530 eV is associated with interstitial oxygens, which is seen in the HP-annealed 100 nm film as well. Thus 30 days of extended HP annealing introduced interstitial oxygen to the powder sample, but without evidence for conversion to SIO-113.

A clear trend can be seen from the data taken on all the samples. The degree of phase transformation depends on sample dimensionality, annealing pressure, and annealing time. There is more phase transformation in thinner films annealed at higher oxygen pressure for longer period of time. There is little effect on the films if the annealing pressure is not high enough. By Le Chatelier’s principle, an oxidizing environment drives equation 6.1 towards the right. It is known that during film growth, higher oxygen

4+ pressure favors Ruddlesden-Popper series (An+1BnO3n+1) with larger n [238]. Since Ir has higher electron negativity than Sr2+, SIO-113, which has higher Ir:Sr ratio than SIO-214, is more stable in more oxidizing environment. Thus it is natural for the phase transformation from SIO-214 to SIO-113 to be more complete under higher annealing pressure. The phase transformation from SIO-214 to SIO-113 involves major cation 1 ~ rearrangements, as every other IrO2 layers from SIO-214 need to translate by 2 (~a + b), and a layer of SrO also needs to be effectively ejected from every unit cell of SIO-214. At 400 ◦C, it is conceivable that such major cation rearrangements are kinetically limited, thus require long reaction time. Finally it was shown in other oxides such as the cuprates that phase transformation during annealing is much more thermodynamically active in thin films than in bulk samples [170, 215]. Since the perovskite form of SIO-113 isnot Chapter 6. Superoxygenation of SIO-214 thin films 94 stable in bulk under ambient pressure [71], it is only stabilized in thin-film forms via the compressive strain from the perovskite substrate. As the film gets thicker, the strain gradually relaxes, and there is not enough strain to stabilize SIO-113. There has even been reports showing that homogeneous SIO-113 films thicker than ∼ 40 nm cannot be made [221, 239]. As a result, there is only partial phase transformation on the thicker films. For the HP-annealed films that are not completely transformed to SIO-113, XAS data indicates hole-doping of the SIO-214 parts of the films by interstitial oxygens, which are responsible for the lower resistivity observed in these samples.

6.4 Conclusion

In summary, we have carried out a superoxygenation study of PLD-grown Sr2IrO4 thin films that range between 20 nm and 100 nm in thickness, by post-annealing theminupto

◦ 400 atm of O2 at 400 C. The room-temperature resistivity progressively drops after the high-pressure annealing by up to 3 orders of magnitude with an evolution from insulating to metallic behavior. Hall effect measurement at 40 K on the least resistive sample shows the majority charge carriers are positive, thus confirming the SIO-214 films become hole- doped metal upon extended high-pressure annealing. STEM, XRD, and XAS revealed structural transformation to the SrIr1−xO3 phase. There is more significant resistivity drop and phase transformation with increasing annealing time, pressure, and reduced film thickness. The evolution towards metallicity is attributed to the phase transformation, interstitial oxygens, and Ir vacancies. The latter two scenarios can result in hole doping. The Ir vacancies in the most phase-transformed film shows a novel structural order along the c-axis, as evidenced by the STEM data. Our results demonstrate a novel method for hole-doping and phase-transforming the iridates in thin-film form. Chapter 7

Conclusions and Future Perspectives

7.1 Conclusions

The unifying theme of this thesis is the study of structural phase transition and defect structures in complex oxide thin films, induced by either heterostructuring or superoxygenation. Thin films were chosen for their ability to tune the heteroepitaxial strain, which is a parameter not accessible to bulk samples, as well as their large surface-to-volume ratio, which significantly enhances their reactivity. The two families of oxides thatwe focus on in this thesis include the Y-Ba-Cu-O family of cuprate superconductors, and the Ruddlesden-Popper iridates Srn+1IrnO3n+1.

We first examined the effects of heteroepitaxial strain on phase transformation and the Tc of YBCO-123 thin films. In epitaxial LCMO/YBCO-123 thin-film heterostructures grown on LSAT substrate, large amount of YBCO-247 regions, which are characterized by double CuO-chains, is seen in STEM images. On the other hand, almost no such double CuO-chain defect structures is present in unilayer YBCO-123 films grown on the same substrate. Since YBCO-247 tends to show lower Tc than optimally doped YBCO-123, the presence of the YBCO-247 intergrowths is responsible for lowering the

Tc of our LCMO/YBCO-123 bilayer films. Resistance versus temperature measurements

95 Chapter 7. Conclusions and Future Perspectives 96

on our films show that the LCMO/YBCO-123 bilayer films indeed have lower Tc than the unilayer YBCO-123 films, at least down to ∼ 25 nm. The CuO intergrowths is attributed to bilayer heteroepitaxial mismatch. These epitaxially-induced CuO intergrowths provide a microstructural mechanism, as opposed to long-ranged proximity effect, for the attenuation of superconductivity in LCMO/YBCO-123 heterostructures.

The effect of heteroepitaxial strain on superconductivity of YBCO-123 thin filmswas more explicitly investigated by changing the layer thickness of YBCO-123 when it is heteroepitaxially clamped by different oxides. Trilayer films of LCMO/YBCO-123/LCMO, LNO/YBCO-123/LNO, and PBCO/YBCO-123/PBCO were grown, with the thickness of the middle YBCO-123 layer changing from 21.4 to 5.4 nm, and the clamping layers fixed at 10.7 nm. This trilayer geometry is used so that the strain is symmetrized onthe both sides of the YBCO-123 layer. Both LCMO and LNO are pseudocubic with similar lattice parameters, and both LCMO/YBCO-123/LCMO and LNO/YBCO-123/LNO trilayers show similarly strong attenuation of Tc as the YBCO-123 layer thickness is reduced. On the other hand, the change of Tc in PBCO/YBCO-123/PBCO trilayers with reducing YBCO-123 layer thickness is much smaller, since PBCO is orthorhombic and isostructural to YBCO-123. The lattice-symmetry mismatch between YBCO-123 and LCMO stands as a crucial experimental issue that was overlooked in prior studies, as one can no longer be sure that the YBCO-123 layer maintains either its bulk phase purity or oxygen stoichiometry when it is heteroepitaxially clamped between cubic perovskite thin films. This results indicate that heteroepitaxial strain, rather than magnetism and long-range proximity effect, is responsible for the long length scales of Tc attenuation observed in c-axis LCMO/YBCO-123 heterostructures.

Besides heteroepitaxial strain, structural phase conversion can also be induced in oxide thin films by the technique of superoxygenation. We presented a novel route offilm synthesis towards discovering more complex phases of the Y-Ba-Cu-O family of cuprate superconductors, using superoxygenation together with cation enrichment. We showed Chapter 7. Conclusions and Future Perspectives 97 that epitaxial YBCO-123 films annealed in ultra-high-pressure oxygen at up to 700 atm and 900 ◦C while buried under a mixture of YBCO-123 and CuO powder transforms to the more complex YBCO-247 and YBCO-248 phases. Intermediate amount of excess CuO powder during annealing converts YBCO-123 film to a film with a gradation of phases, with YBCO-248 occurring near the surface, followed by YBCO-247 below that and then YBCO-123 near the film-substrate interface. A increase in the amount excess CuO powder used during annealing drives YBCO-123 film towards more complete conversion to CuO-chain rich phases, with XRD and R vs. T data consistent with pure YBCO-248 in the annealed film. Exotic phases of Y-Ba-Cu-O, such asYBa2Cu5O9−δ (YBCO-125), and YBa2Cu6O10−δ (YBCO-126), which contains 3 and 4 CuO chains per unit cells, as well as Y2Ba2Cu3O9−δ (YBCO-223) and Y3Ba2Cu3O11−δ (YBCO-323), which contains 2 and 3 extra Y layers inserted between the CuO2 planes, are also observed in the high- pressure annealed films. Successful isolation of the YBCO-125 and YBCO-126 phases will provide a unique opportunity to explore novel roles that multiple-CuO chains can play in high Tc superconductivity, by manipulating the inter-plane coupling as well the interplay between CuO chains and CuO2 planes. On the other hand, YBCO-223 and YBCO-323 can shed light on the effect of changing intra CuO2 bilayer coupling, as well as can be used as a building block for synthesizing non-centrosymmetric high-Tc superconductors. It is important to note that all the superoxygenation-induced phase conversion can only been seen thin films, and there is no evidence of any phase conversion seen in similarly high-pressure annealed YBCO-123 powder, even for annealing time 4 times as long as the films.

Motivated by the similar structural, electrical and magnetic properties of the cuprate superconductors with the Ruddlesden-Popper iridates, the same superoxygenation technique used on the cuprates were also extended to the iridates, in an effort to metallize the iridate through hole doping. Since long annealing time is used, the annealing temperature was kept low at 400 ◦C to minimize cation interdiffusion. The high-pressure Chapter 7. Conclusions and Future Perspectives 98

annealed Sr2IrO4 films show a progressive drop in room-temperature resistivity ofupto 3 orders of magnitude, and an evolution from insulating to metallic behavior. The drop in resistivity is more significant with longer annealing time, higher pressure and reduced film thickness. On thinnest sample subject to the most extended high-pressure annealing, Hall effect measurement taken at 40 K shows the majority charge carriers are positive, consistent with a hole-doped metal. STEM, XRD and XAS revealed transformation to the

SrIr1−xO3 phase on the high-pressure annealed films. Larger degree of phase transformation is seen on extendedly annealed films that have lower room-temperature resistivity. Evidence of Ir vacancies, which appear to be structurally-ordered along the c-axis, can be seen by the STEM data on the most phase-transformed film. XAS measurements also showed evidence for hole-doping and interstitial oxygens on high-pressure annealed films

that are partially converted to SrIr1−xO3. Thus the evolution towards metallicity is attributed to phase transformation, interstitial oxygen, and Ir vacancies. Both Ir-deficiency and O-interstitials can result in hole-doping. Our results demonstrate the prospect of hole-doping the iridates by superoxygenation of thin-film samples.

7.2 Suggested future experiments

For the c-axis manganite/cuprate heterostructure project, we have shown that heteroepi-

taxial strain can strongly influence and attenuate the Tc of YBCO-123 thin films. Such at-

tenuation of Tc is independent of magnetism, and can be explained by the structural phase transition we see in manganite/cuprate heterostructures via STEM. To continue studying the possible long-ranged proximity effect in manganite/cuprate heterostructures, the am- biguity introduced by the highly mobile CuO chains and the oxygen variability of YBCO-

123 can not be overlooked. To avoid the anisotropic strain, the tetragonal La2−xSrxCuO4 (LSCO) could be used instead of orthorhombic YBCO-123 [92, 156, 157]. Specifically, a comparative study of c-axis LCMO/LSCO/LCMO and LNO/LSCO/LNO trilayers, both Chapter 7. Conclusions and Future Perspectives 99 being interfacially-matched in lattice symmetry, would enable a clearer determination of the range of the c-axis ferromagnet/superconductor proximity effect in hole-doped manganite/cuprate heterostructures.

Recently in a collaboration project, we have started resonant inelastic X-ray scattering (RIXS) measurements on LNO/YBCO-123/LNO and LCMO/YBCO-123/LCMO trilayer films, in an effort to study the effect of heteroepitaxial strain on the noveltypes of charge-density-wave order recently observed in LCMO/YBCO-123 bilayers and superlattices [98, 152]. Further studies correlating structural, transport and spectroscopic measurements are needed to unambiguously distinguish the physical effect of ferromagnet/superconductor proximity effect from that of interfacial strain.

For the superoxygenation of YBCO-123 thin films, the goal is to search for new phases of Y-Ba-Cu-O family with novel superconducting properties. We have demonstrated that high-pressure oxygen annealing of YBCO-123 thin films together with cation enrichment via solid diffusion is a promising way of discovering more complex Y-Ba-Cu-O phases. From the STEM images taken on some of the phase-converted YBCO-123 films after high-pressure oxygen annealing, we see that most of the structural phase transformation happens near the surface of the films. This suggests that the phase conversion is limited by the distance excess cations can diffuse into the film. A promising way of overcoming this limitation and isolating the exotic YBCO-125 and YBCO-126 phases is to incorpo- rate excess cation directly into the PLD target, so that excess cation ions required for phase conversion are already uniformly distributed throughout the as-grown films before annealing. Besides using excess Cu powder for cation ion enrichment, other cations can also be used for enrichment. In particular excess Y powder may provide chemical potential needed to isolate the Y-Ba-Cu-O phase that contains double Y layers. This phase is particularly interesting because the broken inversion symmetry within the CuO2 bilayers. As discussed in Chapter 3, interfacial strain is another important factor that can influence thin film phase transition during annealing in addition to cation enrichment. Chapter 7. Conclusions and Future Perspectives 100

Interfacial strain can be tuned using alternative substrates with different lattice parameters or by changing the film thickness. Work is under way to refine our film-growth and annealing recipes, towards more robust synthesis of the novel Y-Ba-Cu-O phases using this superoxygenation technique.

For high-pressure oxygen annealed iridate films, we have shown evidence of hole- doping and metallicity. The next step of this project requires more detailed magneto- transport measurements on the annealed samples. Hall effect measurements need to be performed as a function of temperature, and the carrier density and mobility needs to be extracted from the Hall effect measurements. One of the particular interesting obser- vations on the converted SIO-113 samples after annealing is the super-latticing structure along the c-axis. It is reasonable to expect the electronic structure to be anisotropic as well, unlike pristine SIO-113 films which have the pseudocubic structure. The resistivity and magnetoresistance anisotropy needs to be measured in order to better understand the implication of this super-latticing and structural anisotropy. An explanation for the origin of the Ir vacancy ordering along the c-axis requires a more precise determination of the stoichiometry of the phase-converted SIO-113 films. Energy dispersive X-ray spectroscopy (EDX), wavelength dispersive X-ray spectroscopy (WDS), and X-ray pho- toelectron spectroscopy (XPS) are the three main experimental techniques we are cur- rently investigating in order to obtain a precise determination of the stoichiometry of the annealed iridate films.

Finally, the annealing parameter space for both the cuprates and the iridates can be expanded in the next stage of the project. Hole-doping of SIO-214 via interstitial oxygens is only observed on high-pressure annealed films that are not fully converted to SIO-113, and YBCO-247 mostly converts to YBCO-248 with increasing CuO enrichment. Thus it is reasonable to reduce the annealing time/pressure and excess cation enrichment, in order to isolate these phases that are stable with less extended annealing. In the present study the cuprate films were annealed at high temperature for short duration, while Chapter 7. Conclusions and Future Perspectives 101 the iridate films were annealed at low temperature for long duration, and in both cases significant phase transformation were observed with negligible interfacial diffusion. Itis a natural next step to investigate whether more robust phase conversion can be achieved by using alternative annealing strategies on both iridate and cuprate families of oxide films. Bibliography

[1] J.G. Bednorz and K. A. Müller. Possible high-Tc superconductivity in the Ba-La-

Cu-O system. Zeitschrift fur Physik B, 64:189, 1986.

[2] W. L. McMillan. Transition temperature of strong-coupled superconductors. Phys. Rev., 167:331, 1968.

[3] M. K. Wu, J. R. Ashburn, C. J. Torng, P. H. Hor, R. L. Meng, L. Gao, Z. J. Huang, Y. Q. Wang, and C. W. Chu. Superconductivity at 93 K in a new mixed-phase

Y-Ba-Cu-O compound system at ambient pressure. Phys. Rev. Lett., 58:908, 1987.

[4] Andrea Damascelli, Zahid Hussain, and Zhi-Xun Shen. Angle-resolved photoemis-

sion studies of the cuprate superconductors. Rev. Mod. Phys., 75:473, 2003.

[5] Patrick A. Lee, Naoto Nagaosa, and Xiao-Gang Wen. Doping a Mott insulator:

Physics of high-temperature superconductivity. Rev. Mod. Phys., 78:17, 2006.

[6] Damian Rybicki, Michael Jurkutat, Steven Reichardt, Czesław Kapusta, and Jür- gen Haase. Perspective on the phase diagram of cuprate high-temperature super-

conductors. Nat. Comm., 7:11413, 2016.

[7] C. W. Chu, L. Z. Deng, and B. Lv. Hole-doped cuprate high temperature super-

conductors. Physica C, 514:290, 2015.

[8] S. Chakravarty, H.Y. Kee, and K. Volker. An explanation for a universality of tran-

102 BIBLIOGRAPHY 103

sition temperatures in families of copper oxide superconductors. Nature, 428:6978, 2004.

[9] Y. Hirata, K. M. Kojima, M. Ishikado, S. Uchida, A. Iyo, H. Eisaki, and S. Tajima. Correlation between the interlayer josephson coupling strength and an enhanced su-

perconducting transition temperature of multilayer cuprate superconductors. Phys. Rev. B, 85:054501, 2012.

[10] J. D. Jorgensen, B. W. Veal, A. P. Paulikas, L. J. Nowicki, G. W. Crabtree,

H. Claus, and W. K. Kwok. Structural properties of oxygen-deficient YBa2Cu3O7−δ.

Phys. Rev. B, 41:1863, 1990.

[11] K. Takenaka, K. Mizuhashi, H. Takagi, and S. Uchida. Interplane charge transport

in YBa2Cu3O7−y: Spin-gap effect on in-plane and out-of-plane resistivity. Phys. Rev. B, 50:6534, 1994.

[12] D. N. Basov, R. Liang, D. A. Bonn, W. N. Hardy, B. Dabrowski, M. Quijada, D. B. Tanner, J. P. Rice, D. M. Ginsberg, and T. Timusk. In-plane anisotropy of the

penetration depth in YBa2Cu3O7−x and YBa2Cu4O8 superconductors. Phys. Rev. Lett., 74:598, 1995.

[13] D. H. Lu, D. L. Feng, N. P. Armitage, K. M. Shen, A. Damascelli, C. Kim, F. Ron- ning, Z.-X. Shen, D. A. Bonn, R. Liang, W. N. Hardy, A. I. Rykov, and S. Tajima.

Superconducting gap and strong in-plane anisotropy in untwinned YBa2Cu3O7−δ. Phys. Rev. Lett., 86:4370, 2001.

[14] J. D. Jorgensen, M. A. Beno, D. G. Hinks, L. Soderholm, K. J. Volin, R. L. Hitterman, J. D. Grace, Ivan K. Schuller, C. U. Segre, K. Zhang, and M. S. Kleefisch. Oxygen ordering and the orthorhombic-to-tetragonal phase transition

in YBa2Cu3O7−x. Phys. Rev. B, 36:3608, 1987. BIBLIOGRAPHY 104

[15] E. Kaldis, P. Fischer, A.W. Hewat, E.A. Hewat, J. Karpinski, and S. Rusiecki.

Low temperature anomalies and pressure effects on the structure and Tc of the

superconductor YBa2Cu4O8 (Tc=80 K). Physica C, 159:668, 1989.

[16] H. L. Edwards, A. L. Barr, J. T. Markert, and A. L. de Lozanne. Modulations in

the CuO chain layer of YBa2Cu3O7−δ: Charge density waves? Phys. Rev. Lett., 73:1154, 1994.

[17] B. Grévin, Y. Berthier, and G. Collin. In-plane charge modulation below Tc and

charge-density-wave correlations in the chain layer in YBa2Cu3O7. Phys. Rev. Lett., 85:1310, 2000.

[18] D. J. Derro, E. W. Hudson, K. M. Lang, S. H. Pan, J. C. Davis, J. T. Markert, and A. L. de Lozanne. Nanoscale one-dimensional scattering resonances in the CuO

chains of YBa2Cu3O6+x . Phys. Rev. Lett., 88:097002, 2002.

[19] Makoto Maki, Terukazu Nishizaki, Kenji Shibata, and Norio Kobayashi. Electronic

structure of the CuO-chain layer in YBa2Cu3O7−δ studied by scanning tunneling

microscopy. Phys. Rev. B, 65:140511, 2002.

[20] A. J. Achkar, R. Sutarto, X. Mao, F. He, A. Frano, S. Blanco-Canosa, M. Le Tacon, G. Ghiringhelli, L. Braicovich, M. Minola, M. Moretti Sala, C. Mazzoli, Ruixing Liang, D. A. Bonn, W. N. Hardy, B. Keimer, G. A. Sawatzky, and D. G. Hawthorn.

Distinct charge orders in the planes and chains of ortho-III-ordered YBa2Cu3O6+δ

superconductors identified by resonant elastic x-ray scattering. Phys. Rev. Lett., 109:167001, 2012.

[21] N. E. Hussey, M. Kibune, H. Nakagawa, N. Miura, Y. Iye, H. Takagi, S. Adachi, and K. Tanabe. Magnetic field induced dimensional crossover in the normal state

of YBa2Cu4O8. Phys. Rev. Lett., 80:2909, 1998. BIBLIOGRAPHY 105

[22] N. E. Hussey, H. Takagi, Y. Iye, S. Tajima, A. I. Rykov, and K. Yoshida. Charge

confinement on the CuO2 planes in slightly overdoped YBa2Cu3O7−δ and the role

of metallic chains. Phys. Rev. B, 61:R6475(R), 2000.

[23] V. Breit, P. Schweiss, R. Hauff, H. Wühl, H. Claus, H. Rietschel, A.Erb, and G. Müller-Vogt. Evidence for chain superconductivity in near-stoichiometric

YBa2Cu3Ox single crystals. Phys. Rev. B, 52:R15727(R), 1995.

[24] Robert Gagnon, Song Pu, Brett Ellman, and Louis Taillefer. Anisotropy of heat

conduction in YBa2Cu3O6.9: A probe of chain superconductivity. Phys. Rev. Lett., 78:1976, 1997.

[25] H. Srikanth, B. A. Willemsen, T. Jacobs, S. Sridhar, A. Erb, E. Walker, and

R. Flükiger. Microwave response of YBa2Cu3O6.95 crystals: Evidence for a multi-

component order parameter. Phys. Rev. B, 55:R14733(R), 1997.

[26] F. Raffa, M. Mali, A. Suter, A. Yu. Zavidonov, J. Roos, D. Brinkmann, andK.Con-

der. Spin and charge dynamics in the cu-o chains of YBa2Cu4O8. Phys. Rev. B, 60:3636, 1999.

[27] J. H. Ngai, W. A. Atkinson, and J. Y. T. Wei. Tunneling spectroscopy of c-axis

Y1−xCaxBa2Cu3O7−δ thin-film superconductors. Phys. Rev. Lett., 98:177003, 2007.

[28] M. Matsukawa, Yuh Yamada, M. Chiba, H. Ogasawara, T. Shibata, A. Matsushita,

and Y. Takano. Superconductivity in Pr2Ba4Cu7O15−δ with metallic double chains.

Physica C, 411:101, 2004.

[29] E. Berg, T. H. Geballe, and S. A. Kivelson. Superconductivity in zigzag CuO

chains. Phys. Rev. B, 76:214505, 2007.

[30] Tatsuro Habaguchi, Yoshiaki Ōno, Hai Ying Du Gh, Kazuhiro Sano, and Yuh Yamada. Electronic states and superconducting transition temperature based on BIBLIOGRAPHY 106

the tomonaga–luttinger liquid in Pr2Ba4Cu7O15−δ. J. Phys. Soc. Jpn, 80:024708, 2011.

[31] Taiji Chiba, Michiaki Matsukawa, Junki Tada, Satoru Kobayashi, Makoto Hagi- wara, Tsuyoshi Miyazaki, Kazuhiro Sano, Yoshiaki Ōno, Takahiko Sasaki, and Jun ichi Echigoya. Effect of magnetic field on the superconducting phase inthe

electron-doped metallic double-chain compound Pr2Ba4Cu7O15−δ. J. Phys. Soc. Jpn, 82:074706, 2013.

[32] Masayoshi Kuwabara, Michiaki Matsukawa, Keisuke Sugawara, Haruka Taniguchi, Akiyuki Matsushita, Makoto Hagiwara, Kazuhiro Sano, Yoshiaki Ōno, and Takahiko Sasaki. Effect of pressure on magneto-transport properties inthe superconducting and normal phases of the metallic double chain compound

Pr2Ba4Cu7O15δ. J. Phys. Soc. Jpn, 85:124704, 2016.

[33] Akiyuki Matsushita, Kazunori Fukuda, Yuh Yamada, Fumihiro Ishikawa, Syun Sekiya, Masato Hedo, and Takashi Naka. Physical properties of a new cuprate

superconductor Pr2Ba4Cu7O15−δ. Sci. Tech. Adv. Mater., 8:477, 2007.

[34] D. E. Morris, N. G. Asmar, J. Y. T. Wei, J. H. Nickel, R. L. Sid, J. S. Scott, and

J. E. Post. Synthesis and properties of the 2:4:7 superconductors R2Ba4Cu7O15−x

(R=Y,Eu,Gd,Dy,Ho,Er). Phys. Rev. B, 40:11406, 1989.

[35] J. Karpinski, S. Rusiecki, E. Kaldis, B. Bucher, and E. Jilek. Phase diagrams

of YBa2Cu4O8 and YBa2Cu3.5O7.5 in the pressure range 1 bar≤PO2 ≤ 3000 bar. Physica C, 160:449, 1989.

[36] M. I. Katsnelson, V. Yu. Irkhin, L. Chioncel, A. I. Lichtenstein, and R. A. de Groot.

Half-metallic ferromagnets: From band structure to many-body effects. Rev. Mod. Phys., 80:315, 2008. BIBLIOGRAPHY 107

[37] Clarence Zener. Interaction between the d-shells in the transition metals. ii. ferro-

magnetic compounds of manganese with perovskite structure. Phys. Rev., 82:403, 1951.

[38] P. W. Anderson and H. Hasegawa. Considerations on double exchange. Phys. Rev., 100:675, 1955.

[39] S. Satpathy, Zoran S. Popović, and Filip R. Vukajlović. Electronic structure of the

perovskite oxides: la1−x cax mno3. Phys. Rev. Lett., 76:960, 1996.

[40] A P Ramirez. Colossal magnetoresistance. J. Phys. Condens. Matter, 9:8171, 1997.

[41] S. Blundell. Magnetism in Condensed Matter. Oxford Master Series in Condensed Matter Physics. OUP Oxford, 2001.

[42] Yongquan Guo, Sujoy Roy, and Naushad Ali. Effect of nitrogen incorporation in

La2/3Ca1/3MnO3 manganite. J. Phys. Condens. Matter., 14:181, 2002.

[43] David Emin. Polaron formation and motion in magnetic solids. Mat. Res. Soc. Symp. Proc., 494:163, 1997.

[44] J. M. D. Coey, M. Viret, and S. von Molnár. Mixed-valence manganites. Adv. Phys., 48:167, 1999.

[45] Y Tokura and Y Tomioka. Colossal magnetoresistive manganites. J. Magn. Magn. Mater., 200:1, 1999.

[46] Myron B. Salamon and Marcelo Jaime. The physics of manganites: Structure and

transport. Rev. Mod. Phys., 73:583, 2001.

[47] Y Tokura. Critical features of colossal magnetoresistive manganites. Rep. Prog. Phys., 69:797, 2006.

[48] T. M. Klapwijk. Proximity effect from an Andreev perspective. JOSC, 17:593, 2004. BIBLIOGRAPHY 108

[49] V.Z. Kresin, H. Morawitz, and S.A. Wolf. Mechanisms of Conventional and High

Tc Superconductivity. International Series of Monographs on Physics. Oxford Uni- versity Press, 1993.

[50] K.H. Bennemann and J.B. Ketterson. Superconductivity. SpringerLink: Springer e-Books. Springer Berlin Heidelberg, 2008.

[51] J. Linder and A. V. Balatsky. Odd-frequency superconductivity. 2017.

[52] A A Golubov, Y Tanaka, Y Asano, and Y Tanuma. Odd-frequency pairing in

superconducting heterostructures. J. Phys. Condens. Matter., 21:164208, 2009.

[53] Y. Tanaka, M. Sato, and N. Nagaosa. Symmetry and topology in superconductors

–odd-frequency pairing and edge states–. J. Phys. Soc. Jpn, 81:011013, 2012.

[54] F. S. Bergeret, A. F. Volkov, and K. B. Efetov. Odd triplet superconductivity and

related phenomena in superconductor-ferromagnet structures. Rev. Mod. Phys., 77:1321, 2005.

[55] G. Deutscher and P. G. de Gennes. Superconductivity, volume Vol. 2, chapter Proximity effects. Marcel Dekker, 1969.

[56] P. C. van Son, H. van Kempen, and P. Wyder. New method to study the proximity

effect at the normal-metal–superconductor interface. Phys. Rev. Lett., 59:2226, 1987.

[57] J. Linder and J. W. A. Robinson. Superconducting spintronics. Nat. Phys., 11:307, 2015.

[58] T. Kontos, M. Aprili, J. Lesueur, and X. Grison. Inhomogeneous superconductivity

induced in a ferromagnet by proximity effect. Phys. Rev. Lett., 86:304, 2001.

[59] C.L. Chien and Daniel H. Reich. Proximity effects in superconducting/magnetic

multilayers. J. Magn. Magn. Mater., 200:83, 1999. BIBLIOGRAPHY 109

[60] F. S. Bergeret, A. F. Volkov, and K. B. Efetov. Long-range proximity effects in

superconductor-ferromagnet structures. Phys. Rev. Lett., 86:4096, 2001.

[61] A. I. Buzdin. Proximity effects in superconductor-ferromagnet heterostructures.

Rev. Mod. Phys., 77:935, 2005.

[62] Y Maeno, H Hashimoto, K Yoshida, S Nishizaki, T Fujita, JG Bednorz, and F Licht-

enberg. Superconductivity in a layered perovskite without copper. Nature, 372:532, 1994.

[63] Andrew Peter Mackenzie and Yoshiteru Maeno. The superconductivity of Sr2RuO4

and the physics of spin-triplet pairing. Rev. Mod. Phys., 75:657, 2003.

[64] Fa Wang and T. Senthil. Twisted Hubbard model for Sr2IrO4: magnetism and

possible high temperature superconductivity. Phys. Rev. Lett., 106:136402, 2011.

[65] Y. K. Kim, O. Krupin, J. D. Denlinger, E. Rotenberg A. Bostwick, Q. Zhao, J. F. Mitchell, and and B. J. Kim J. W. Allen. Fermi arcs in a doped pseudospin-1/2

heisenberg antiferromagnet. Science, 345:187, 2014.

[66] Yan Liu, Li Yu, Xiaowen Jia, Jianzhou Zhao, Hongming Weng, Yingying Peng, Chaoyu Chen, Zhuojin Xie, Daixiang Mou, Junfeng He, Xu Liu, Ya Feng, Hemian Yi, Lin Zhao, Guodong Liu, Shaolong He, Xiaoli Dong, Jun Zhang, Zuyan Xu, Chuangtian Chen, Gang Cao, Xi Dai, Zhong Fang, and X. J. Zhou. Anomalous

high-energy waterfall-like electronic structure in 5 d transition metal oxide Sr2IrO4

with a strong spin-orbit coupling. Sci. Rep., 5:13036, 2015.

[67] Yue Cao, Qiang Wang, Justin A. Waugh, Theodore J. Reber, Haoxiang Li, Xiao- qing Zhou, Stephen Parham, Nicholas C. Plumb, Eli Rotenberg, Aaron Bostwick, Jonathan D. Denlinger, Tongfei Qi, Michael A. Hermele, Gang Cao, and Daniel S. Dessau. Hallmarks of the Mott-metal crossover in the hole-doped pseudospin-1/2

Mott insulator Sr2IrO4. Nat. Comm., 7:11367, 2016. BIBLIOGRAPHY 110

[68] Jungho Kim, D. Casa, M. H. Upton, T. Gog, Young-June Kim, J. F. Mitchell, M. van Veenendaal, M. Daghofer, J. van den Brink, G. Khaliullin, and B. J. Kim.

Magnetic excitation spectra of Sr2IrO4 probed by resonant inelastic x-ray scatter-

ing: Establishing links to cuprate superconductors. Phys. Rev. Lett., 108:177003, 2012.

[69] S. Bahr, A. Alfonsov, G. Jackeli, G. Khaliullin, A. Matsumoto, T. Takayama, H. Takagi, B. Büchner, and V. Kataev. Low-energy magnetic excitations in the

spin-orbital Mott insulator Sr2IrO4. Phys. Rev. B, 89:180401, 2014.

[70] K-H Kim, H-S Kim, and M J Han. Electronic structure and magnetic properties of

iridate superlattice SrIrO3/SrTiO3. J. Phys.: Condends. Matter, 26:185501, 2014.

[71] J. M. Longo, J. A. Kafalas, and R. J. Arnott. Structure and properties of the high

and low pressure forms of SrIrO3. J. Solid State Chem., 3:174, 1971.

[72] B. J. Kim, H. Jin, S. J. Moon, J.-Y. Kim, B.-G. Park, C. S. Leem, J. Yu, T. W.

Noh, C. Kim, S.-J. Oh, V. Durairai, G. Cao, and J.-H. Park. Novel Jeﬀ = 1/2

Mott state induced by relativistic spin-orbit coupling in Sr2IrO4. Phys. Rev. Lett., 101:076402, 2008.

[73] Robert Eason. Pulsed Laser Deposition of Thin Films. Wiley, 2006.

[74] K. R. Chen, J. N. Leboeuf, R. F. Wood, D. B. Geohegan, J. M. Donato, C. L. Liu, and A. A. Puretzky. Mechanisms affecting kinetic energies of laser‐ablated

materials. J. Vac. Sci. Technol. A, 14:1111, 1996.

[75] David B.Geohegan. Physics and diagnostics of laser ablation plume propagation

for high-Tc superconductor film growth. Thin Solid Films, 220:138, 1992.

[76] J. Als-Nielsen and D. McMorrow. Elements of Modern X-ray Physics. Wiley, 2011. BIBLIOGRAPHY 111

[77] L. G. Parratt. Surface studies of solids by total reflection of x-rays. Phys. Rev., 95:359, 1954.

[78] Matts Björck and Gabriella Andersson. GenX: an extensible X-ray reflectivity refinement program utilizing differential evolution. J. Appl. Crystallogr., 40:1174, 2007.

[79] D. W. Koon. Effect of contact size and placement, and of resistive inhomogeneities

on van der Pauw measurements. Rev. Sci. Instrum., 60:271, 1989.

[80] D. W. Koon, A. A. Bahl, and E. O. Duncan. Measurement of contact placement

errors in the van der Pauw technique. J. Vac. Sci. Technol. A, 60:275, 1989.

[81] S. A. Chambers. Epitaxial growth and properties of doped transition metal and

complex oxide films. Adv. Mater., 22:219, 2010.

[82] V. A. Vas’ko, V. A. Larkin, P. A. Kraus, K. R. Nikolaev, D. E. Grupp, C. A. Nordman, and A. M. Goldman. Critical current suppression in a superconductor

by injection of spin-polarized carriers from a ferromagnet. Phys. Rev. Lett., 78:1134, 1997.

[83] P. Kraus, A. Bhattacharya, and A.M. Goldman. Magnetic-field scaling of the

conductance of epitaxial cuprate-manganite bilayers. Phys. Rev. B, 64:220505, 2001.

[84] Z. Y. Chen, A. Biswas, I. Žutić, T. Wu, S. B. Ogale, R. L. Greene, and T. Venkate-

san. Spin-polarized transport across a La0.7Sr0.3MnO3/YBa2Cu3O7−x interface:

Role of Andreev bound states. Phys. Rev. B, 63:212508, 2001.

[85] Z. Sefrioui, D. Arias, V. Pe na, J. E. Villegas, M. Varela, P. Prieto, C. León, J. L. Martinez, and J. Santamaria. Ferromagnetic/superconducting proximity effect in

La0.7Ca0.3MnO3/YBa2Cu3O7−δ superlattices. Phys. Rev. B, 67:214511, 2003. BIBLIOGRAPHY 112

[86] K. Dybko, K. Werner-Malento, P. Aleshkevych, M. Wojcik, M. Sawicki, and

P. Przyslupski. Possible spin-triplet superconducting phase in the La0.7Sr0.3MnO3/

YBa2Cu3O7/La0.7Sr0.3MnO3 trilayer. Phys. Rev. B, 80:144504, 2009.

[87] C. Visani, Z. Sefrioui, J. Tornos, C. León, J. Briatico, M. Bibes, A. Barthelemy, J. Santamaria, and J. E. Villegas. Equal-spin Andreev reflection and long-range coherent transport in high-temperature superconductor/half-metallic ferromagnet

junctions. Nat. Phys., 8:539, 2012.

[88] H.-U. Habermeier, R. K. Kremer G. Cristiani, O. Lebedev, and G. van Tendeloo. Cuprate/manganite superlattices: A model system for a bulk ferromagnetic super-

conductor. Physica C, 364:298, 2001.

[89] V. Pe na, C. Visani, J. Garcia-Barriocanal, D. Arias, Z. Sefrioui, C. Leon, and J. Santamaria. Cuprate/manganite superlattices: A model system for a bulk fer-

romagnetic superconductor. Phys. Rev. B, 73:104513, 2006.

[90] I. Žutić, J. Fabian, and S. Das Sarma. Spintronics: Fundamentals and applications.

Rev. Mod. Phys., 76:323, 2004.

[91] I. Fridman, L. Gunawan, G. A. Botton, and J. Y. T. Wei. Scanning tunneling spectroscopy study of c-axis proximity effect in epitaxial bilayer manganite/cuprate

thin films. Phys. Rev. B, 84:104522, 2011.

[92] G. Koren. and T. Kirzhner. Transport and spectroscopic properties of superconduc-

tor/ferromagnet/superconductor junctions of La1.9Sr0.1CuO4/La0.67Ca0.33MnO3/

La1.9Sr0.1CuO4. Phys. Rev. B, 84:134517, 2011.

[93] T. Holden, H.-U. Habermeier, G. Cristiani, A. Golnik, A. Boris, A. Pimenov, J. Humlíc̆ek, O. I. Lebedev, G. Van Tendeloo, B. Keimer, and C. Bernhard. Prox-

imity induced metal-insulator transition in YBa2Cu3O7/La2/3Ca1/3MnO3 superlat-

tices. Phys. Rev. B, 69:064505, 2004. BIBLIOGRAPHY 113

[94] M. Varela, A.R. Lupini, V. Pena, Z. Sefrioui, I. Arslan, N.D. Browning, J. Santa- maria, and S.J. Pennycook. Direct measurement of charge transfer phenomena at ferromagnetic/superconducting oxide interfaces. 2006.

[95] J. Chakhalian, J. W. Freeland H.-U. Habermeier, G. Cristiani, G. Khaliullin, M. van Veenendaal, and B. Keimer. Orbital reconstruction and covalent bonding at an

oxide interface. Science, 318:1114, 2007.

[96] J. Chakhalian, J. W. Freeland G. Srajer, J. Strempfer, G. Khaliullin, J. C. Cezar, T. Charlton, R. Dalgliesh, C. Bernhard, G. Cristiani, H.-U. Habermeier, and B. Keimer. Magnetism at the interface between ferromagnetic and superconduct-

ing oxides. Nat. Phys., 2:244, 2006.

[97] J. Stahn, J. Chakhalian, C. Niedermayer, J. Hoppler, T. Gutberlet, J. Voigt, F. Treubel, H.-U. Habermeier, G. Cristiani, B. Keimer, and C. Bernhard. Mag-

netic proximity effect in perovskite superconductor/ferromagnet multilayers. Phys. Rev. B, 71:140509, 2005.

[98] Junfeng He, Padraic Shafer, Thomas R. Mion, Vu Thanh Tra, Qing He, J. Kong, Y.-D. Chuang, W.L. Yang, M.J. Graf, J.-Y. Lin, Y.-H. Chu, E. Arenholz, and Rui- Hua He. Observation of a three-dimensional quasi-long-range electronic supermod-

ulation in YBa2Cu3O7−x/La0.7Ca0.3MnO3 heterostructures. Nat. Comm., 7:10852, 2016.

[99] R. J. Cava, B. Batlogg, R. B. van Dover, D. W. Murphy, S. Sunshine, T. Siegrist, J. P. Remeika, E. A. Rietman, S. Zahurak, and G. P. Espinosa. Bulk supercon-

ductivity at 91 K in single-phase oxygen-deficient perovskite Ba2YCu3O9−δ. Phys. Rev. Lett., 58:1676, 1987.

[100] P. Bordet, C. Chaillout, J. Chenavas, J. L. Hodeau, M. Marezio, J. Karpinski, and BIBLIOGRAPHY 114

E. Kaldis. Structure determination of the new high-temperature superconductor

Y2Ba4Cu7O14+x. Nature, 334:596, 1988.

[101] J. Karpinski, S. Rusiecki, B. Bucher, E. Kaldis, and E. Jilek. The nonstoichiometry

of the high-Tc superconductor Y2Ba4Cu7O15±x (14 K≤ Tc ≤68 K). Physica C, 161:618, 1989.

[102] J. L. Tallon, D. M. Pooke, R. G. Buckley, and M. R. Presland F. J. Blunt.

R2Ba4Cu7O15−δ : A 92-K bulk superconductor. Phys. Rev. B, 41:7220, 1990.

[103] A. F. Marshall, K. Char, R.W. Barton, A. Kapitulnik, and S. S. Laderman. Mi-

crostructural interaction of Y2Ba4Cu8O16 stacking faults within YBa2Cu3O7−x. J. Mater. Res., 5:2049, 1990.

[104] R. Ramesh, D. M. Hwang, J. B. Barner, L. Nazar, T. S. Ravi, A. Inam, B. Dutta, X. D. Wu, and T. Venkatesan. Defect structure of laser deposited Y-Ba-Cu-O thin

films on single crystal MgO substrate. J. Mater. Res., 5:704, 1990.

[105] L. F. Fu, N. D. Browning, W. Ramadan, S. B. Ogale, D. C. Kundaliya, and

T. Venkatesan. Interface and defect structures in YBa2Cu3O7−δ and Nb:SrTiO3

heterojunction. J. Phys. D: Appl. Phys., 40:187, 2007.

[106] M. Varela, A. R. Lupini, S. J. Pennycook, Z. Sefrioui, and J. Santamaria.

Nanoscale analysis of YBa2Cu3O7−x/La0.67Ca0.33MnO3 interfaces. Solid-State Elec- tron., 47:2245, 2003.

[107] C. Visani, J. Tornos, N. M. Nemes, M. Rocci, C. Leon, J. Santamaria, S. G. E. te Velthuis, Y. Liu, A. Hoffmann, J. W. Freeland M. Garcia-Hernandez, M.R. Fitzsimmons, B. J. Kirby, M. Varela, and S. J. Pennycook. Symmetrical interfacial

reconstruction and magnetism in La0.7Ca0.3MnO3/YBa2Cu3O7/La0.7Ca0.3MnO3

heterostructures. Phys. Rev. B, 84:060405, 2011. BIBLIOGRAPHY 115

[108] V. K. Malik, I. Marozau, S. Das, B. Doggett, D. K. Satapathy, M. A. Uribe-Laverde, N. Biskup, M. Varela, C. W. Schneider, C. Marcelot, J. Stahn, and C. Bernhard.

Pulsed laser deposition growth of heteroepitaxial YBa2Cu3O7/La0.67Ca0.33MnO3

superlattices on NdGaO3 and Sr0.7La0.3Al0.65Ta0.35O3 substrates. Phys. Rev. B, 85:054514, 2012.

[109] J. J. Capponi, C. Chaillout, A. W. Hewat, P. Lejay, M. Marezio, N. Nguyen, B. Raveau, J. L. Soubeyroux, J. L. Tholence, and R. Tournier. Structure of the 100

K superconductor Ba2YCu3O7 between (5 ÷ 300) K by neutron powder diffraction.

Europhys. Lett., 3:1301, 1987.

[110] M. Kato, M. Nakanishi, T. Miyano, T. Shimizu, M. Kakihana, K. Yoshimura,

and K. Kosuge. Microscopic studies on Y2Ba4Cu7O15−δ by use of TEM and NQR

techniques. J. Solid State Chem., 139:266, 1998.

[111] A. Irizawa, T. Ohmura, T. Shibata, M. Kato, K. Yoshimura, K. Kosuge, Y. Ito, H. Michor, and G. Hilscher. Collapse of superconductivity in the Y123/Y124 su-

perlattice system Y2Ba4Cu7O15−δ. J. Phys. Soc. Jpn., 71:574, 2002.

[112] H. Sato, A. Tsukada, M. Naito, and A. Matsuda. La2−xSrxCuOy epitaxial thin films

(x = 0 to 2):structure, strain, and superconductivity. Phys. Rev. B, 61:12447, 2000.

[113] C. P. Poole Jr. Handbook of Superconductivity, page 267. Academic, 2000.

[114] Y. H. Li, K. A. Thomas, P. S. I. P. N. de Silva, L. F. Cohen, A. Goyal, M. Rajeswari, N. D. Mathur, M. G. Blamire, J. E. Evetts, T. Venkatesan, and J. L. MacManus- Driscoll. Transmission electron microscopy and x-ray structural investigation of

La0.7Ca0.3MnO3 thin films. J. Mater. Res., 13:2161, 1998.

[115] M. H. Francombe. Frontiers of Thin Film Technology and Vol. 28. Academic, 2001. p.251. BIBLIOGRAPHY 116

[116] M. Varela, W. Grogger, D. Arias, Z. Sefrioui, C. León, L. Vazquez, C. Ballesteros, K. M. Krishnan, and J. Santamaría. Effects of epitaxial strain on the growth

mechanism in YBa2Cu3O7−x thin films in YBa2Cu3O7−x/PrBa2Cu3O7 superlat-

tices. Phys. Rev. B, 66:174514, 2002.

[117] M.G. Smith, A. Manthiram, J. Zhou, J. B. Goodenough, and J. T. Mark- ert. Electron-doped superconductivity at 40 K in the infinite-layer compound

Sr1yNdyCu02. Nature, 351:549, 1991.

[118] S. J. May, J.-W. Kim, J. M. Rondinelli, E. Karapetrova, N. A. Spaldin, A. Bhat- tacharya, and P. J. Ryan. Quantifying octahedral rotations in strained perovskite

oxide films. Phys. Rev. B, 82:014110, 2010.

[119] J.-P. Locquet, J. Perret, J. Fompeyrine, E. Machler, J. W. Seo, and G. Van Ten-

deloo. Doubling the critical temperature of La1.9Sr0.1CuO4 using epitaxial strain.

Nature, 394:453, 1998.

[120] I. Bozovic, G. Logvenov, I Belca, B. Narimbetov, and I. Sveklo. Epitaxial strain

and superconductivity in La2−xSrxCuO4 thin films. Phys. Rev. Lett., 89:107001, 2002.

[121] C. Rayan Serrao, Jian Liu, J. T. Heron, G. Singh-Bhalla, A. Yadav, S. J. Suresha, R. J. Paull, D. Yi, J.-H. Chu, M. Trassin, A. Vishwanath, E. Arenholz, C. Frontera, J. Železný, T. Jungwirth, X. Marti, and R. Ramesh. Epitaxy-distorted spin-orbit

Mott insulator in Sr2IrO4 thin films. Phys. Rev. B, 87:085121, 2013.

[122] Mónica Burriel, José Santiso, Marta D. Rossell, Gustaaf Van Tendeloo, Albert

Figueras, and Gemma Garcia. Enhancing total conductivity of La2NiO4+δ epitaxial

thin films by reducing thickness. J. Phys. Chem. C, 112:10982, 2008.

[123] Bharat Jalan, S. James Allen, Glenn E. Beltz, Pouya Moetakef, and Susanne Stem- BIBLIOGRAPHY 117

mer. Enhancing the electron mobility of SrTiO3 with strain. Appl. Phys. Lett., 98:132102, 2011.

[124] J. W. Matthews and A. E. Blakeslee. Defects in epitaxial multilayers: I. misfit

dislocations. J. Cryst. Growth, 27:118, 1974.

[125] M. D. Thouless. Modeling the development and relaxation of stresses in films.

Annu. Rev. Mater. Sci., 25:69, 1995.

[126] R. F. Klie, J. P. Buban, M. Varela, A. Franceschetti, C. Jooss, Y. Zhu, N. D. Browning, S. T. Pantelides, and S. J. Pennycook. Enhanced current transport at

grain boundaries in high-Tc superconductors. Nature, 435:475, 2005.

[127] Ulrich Aschauer, Reto Pfenninger, Sverre M. Selbach, Tor Grande, and Nicola A.

Spaldin. Strain-controlled oxygen vacancy formation and ordering in CaMnO3.

Phys. Rev. B, 88:054111, 2013.

[128] T. L. Meyer, L. Jiang, S. Park, T. Egami, and H. N. Lee. Strain-relaxation and

critical thickness of epitaxial La1.85Sr0.15CuO4 films. Appl. Phys. Lett. Mater., 3:126102, 2015.

[129] H. B. Peng, B. R. Zhao, Z. Xie, Y. Lin, B. Y. Zhu, Z. Hao, H. J. Tao, B. Xu, C. Y.

Wang, H. Chen, and F. Wu. Ordered surface structure in La1−x Cax MnO3 films.

Phys. Rev. Lett., 82:362, 1999.

[130] L. S. J. Peng, X. X. Xi, B. H. Moeckly, and S. P. Alpay. Strain relaxation during

in situ growth of SrTiO3 thin films. Appl. Phys. Lett., 83:4592, 2003.

[131] Tianqi Wang, Koustav Ganguly, Patrick Marshall, Peng Xu, and Bharat Jalan.

Critical thickness and strain relaxation in molecular beam epitaxy-grown SrTiO3

films. Appl. Phys. Lett., 103:212904, 2013. BIBLIOGRAPHY 118

[132] X. L. Tan, F. Chen, P. F. Chen, H. R. Xu, B. B. Chen, F. Jin, G. Y. Gao, and W. B. Wu. Anisotropic-strain-relaxation-induced crosshatch morphology in epitax-

ial SrTiO3/NdGaO3 thin films. AIP Adv., 4:107109, 2014.

[133] H. Zhang, N. Gauquelin, G. A. Botton, and J. Y. T. Wei. Attenuation of superconductivity in manganite/cuprate heterostructures by epitaxially-induced CuO

intergrowths. Appl. Phys. Lett., 103:052606, 2013.

[134] S. Soltan, J. Albrecht, and H.-U. Habermeier. Ferromagnetic/superconducting

bilayer structure: A model system for spin diffusion length estimation. Phys. Rev. B, 70:144517, 2004.

[135] B. S. H. Pang, R. I. Tomov, C. Bell, and M. G. Blamire. Effect of ferromagnetism

on superconductivity in manganite/cuprate heterostructures. Physica C, 415:118, 2004.

[136] J. W. Freeland J. Chakhalian, H.-U. Habermeier, G. Cristiani, and B. Keimer. On

magnetic interlayer coupling and proximity effect in a La0.67Ca0.33MnO3 (10nm)/

YBa2Cu3O7 (10nm) superlattice. Appl. Phys. Lett., 90:242502, 2007.

[137] J. Hoppler, J. Stahn, Ch. Niedermayer, V. K. Malik, H. Bouyanfif, A. J. Drew, M. Rössle, A. Buzdin, G. Cristiani, H.-U. Habermeier, B. Keimer, and C. Bernhard. Giant superconductivity-induced modulation of the ferromagnetic magnetization

in a cuprate–manganite superlattice. Nat. Mater., 8:315, 2009.

[138] Yoav Kalcheim, Israel Felner, Oded Millo, Tal Kirzhner, Gad Koren, Angelo Di Bernardo, Mehmet Egilmez, Mark G. Blamire, and J. W. A. Robinson. Magnetic field dependence of the proximity-induced triplet superconductivity at ferromag-

net/superconductor interfaces. Phys. Rev. B, 89:180506, 2014.

[139] C. Visani, F. Cuellar, A. Pérez-Mu noz, Z. Sefrioui, C. León, J. Santamaría, and Javier E. Villegas. Magnetic field influence on the proximity effect at BIBLIOGRAPHY 119

YBa2Cu3O7/La2/3Ca1/3MnO3 superconductor/half-metal interfaces. Phys. Rev. B, 92:014519, 2015.

[140] B. A. Gray, S. Middey, G. Conti, A. X. Gray, C.-T. Kuo, A. M. Kaiser, S. Ueda, K. Kobayashi, D. Meyers, M. Kareev, I. C. Tung, Jian Liu, C. S. Fadley, J. Chakhalian, and J. W. Freeland. Superconductor to Mott insulator transition

in YBa2Cu3O7/LaCaMnO3 heterostructures. Sci. Rep., 6:33184, 2016.

[141] Y. Asano, Y. Tanaka, A. A. Golubov, and S. Kashiwaya. Conductance spectroscopy

of spin-triplet superconductors. Phys. Rev. Lett., 99:067005, 2007.

[142] Ya. V. Fominov, Y. Tanaka, Y. Asano, and M. Eschrig. Odd-frequency superconducting states with different types of Meissner response: Problem of coexistence.

Phys. Rev. B, 91:144514, 2015.

[143] Petra Högl, Alex Matos-Abiague, Igor Žutić, and Jaroslav Fabian. Magne-

toanisotropic Andreev reflection in ferromagnet-superconductor junctions. Phys. Rev. Lett., 115:116601, 2015.

[144] I. N. Chan, D. C. Vier, O. Nakamura, J. Hasen, J. Guimpel, S. Schultz, and Ivan K. Schuller. Thickness dependence of the superconducting transition temperature of

YBCO. Phys. Lett. A, 175:241, 1993.

[145] H. M. Ledbetter, M. W. Austin, S. A. Kim, and Ming Lei. Elastic constants and

debye temperature of polycrystalline YBa2Cu3O7−x. J. Mater. Res., 2:786, 1987.

[146] D P Almond, E Lambson, G A Saunders, and Wang Hong. An ultrasonic study of

the elastic properties of the normal and superconducting states of YBa2Cu3O7−δ.

J. Phys. F, 17:L221, 1987.

[147] M Cankurtaran, GA Saunders, KC Goretta, and RB Poeppel. Ultrasonic determi- BIBLIOGRAPHY 120

nation of the elastic properties and their pressure and temperature dependences in

very dense YBa2Cu3O7−x. Phys. Rev. B, 46:1157, 1992.

[148] R. Ravinder Reddy, M. Murakami, S. Tanaka, and P. Venugopal Reddy. Elastic

behavior of a Y-Ba-Cu-O sample prepared by the MPMG method. Physica C, 257:137, 1996.

[149] L.S. Ewe, Imad Hamadneh, Hazar A. Salama, and R. Abd-Shukor. Sound velocity

in perovskite manganites La0.67Ca0.33MnO3 with different grain sizes. Physica B: Condensed Matter, 403:2394, 2008.

[150] S. Masys and V. Jonauskas. Elastic properties of rhombohedral, cubic, and mon-

oclinic phases of LaNiO3 by first principles calculations. Comput. Mater. Sci, 108:153, 2015.

[151] Peter Lemmens, Christoph Hünnekes, Philip Fröning, Siegfried Ewert, Helga Pass-

ing, Gerolf Marbach, and Albert Comberg. Elastic properties of YBa2Cu3Ox,

(Y1−yPry)Ba2Cu3O7 and (BiPb)2Sr2Ca2Cu3O10 studied by ultrasound. J. Lessr Common Met., 164-165:1129, 1990.

[152] A. Frano, S. Blanco-Canosa, E. Schierle, Y. Lu, M. Wu, M. Bluschke, M. Mi- nola, G. Christiani, H. U. Habermeier, G. Logvenov, Y. Wang, P. A. van Aken, E. Benckiser, E. Weschke, M. Le Tacon, and B. Keimer. Long-range charge-density-

wave proximity effect at cuprate/manganate interfaces. Nat. Mater., 25:831, 2016.

[153] R. Comin, R. Sutarto, F. He, E. da Silva Neto, L. Chauviere, A. Frano, R. Liang, W. N. Hardy, D. Bonn, Y. Yoshida, H. Eisaki, A.J. Achkar, D.G. Hawthorn, B. Keimer, G. A. Sawatzky, and A. Damascelli. Symmetry of charge order in

cuprates. Nat. Mater., 14:796, 2015.

[154] S. Gerber, H. Jang, H. Nojiri, S. Matsuzawa, H. Yasumura, D. A. Bonn, R. Liang, W. N. Hardy, Z. Islam, A. Mehta, S. Song, M. Sikorski, D. Stefanescu, Y. Feng, BIBLIOGRAPHY 121

S. A. Kivelson, T. P. Devereaux, Z.-X. Shen, C.-C. Kao, W.-S. Lee, D. Zhu, and

J.-S. Lee. Three-dimensional charge density wave order in YBa2Cu3O6.67 at high

magnetic fields. Science, 350:949, 2015.

[155] L. E. Hayward, D. G. Hawthorn, R. G. Melko, and S. Sachdev. Angular fluctuations

of a multicomponent order describe the pseudogap of YBa2Cu3O6+x. Science, 343:1336, 2014.

[156] S. Das, K. Sen, I. Marozau, M. A. Uribe-Laverde, N. Biskup, M. Varela, Y. Khay- dukov, O. Soltwedel, T. Keller, M. Döbeli, C. W. Schneider, and C. Bernhard. Structural, magnetic, and superconducting properties of pulsed-laser-deposition-

grown La1.85Sr0.15CuO4/La2/3Ca1/3MnO3 superlattices on (001)-oriented LaSrAlO4

substrates. Phys. Rev. B, 89:094511, 2014.

[157] G. De Luca, G. Ghiringhelli, C. A. Perroni, V. Cataudella, F. Chiarella, C. Cantoni, A. R. Lupini, N. B. Brookes, M. Huijben, G. Koster, G. Rijnders, and M. Salluzzo. Ubiquitous long-range antiferromagnetic coupling across the interface between su-

perconducting and ferromagnetic oxides. Nat. Comm., 5:5626, 2014.

[158] N. Poccia, A. Ricci, G. Campi, M. Fratini, A. Puri, D. Di Gioacchino, A. Marcelli, M. Reynolds, M. Burghammer, N. Lal Saini, G. Aeppli, and A. Bianconi. Optimum

inhomogeneity of local lattice distortions in La2CuO4+y. Proc. Natl. Acad. Sci. USA, 109:15685, 2012.

[159] Z. Fisk and J. L. Sarrao. The new generation high-temperature superconductors.

Annu. Rev. Mater. Sci., 27:35, 1997.

[160] J. Karpinski, E. Kaldis, E. Jilek, S. Rusiecki, and B. Bucher. Bulk synthesis of the

81-K superconductor YBa2Cu4O8 at high oxygen pressure. Nature, 336:660, 1988.

[161] D. E. Morris, J. H. Nickel, J. Y. T. Wei, N. G. Asmar, J. S. Scott, U. M. Scheven, C. T. Hultgren, A. G. Markelz, J. E. Post, P. J. Heaney, D. R. Veblen, and R. M. BIBLIOGRAPHY 122

Hazen. Eight new high-temperature superconductors with the 1:2:4 structure. Phys. Rev. B, 39:7347, 1989.

[162] M. L. Mandich, A. M. DeSantolo, R. M. Fleming, P. Marsh, S. Nakahara, S. Sun- shine, J. Kwo, M. Hong, T. Boone, T. Y. Kometani, and L. Martinez-Miranda.

Superconducting properties of a 27-Å phase of Ba-Y-Cu-O. Phys. Rev. B, 38:5031, 1988.

[163] P. Marsh, R. M. Fleming, M. L. Mandich, A. M. Desantolo, J. Kwo, M. Hong, and

L. J. Martinez-miranda. Crystal structure of the 80 k superconductor YBa2Cu4O8.

Nature, 334:14, 1988.

[164] R. J. Cava, J. J. Krajewski, W. F. Peck Jr, B. Batlogg, L. W. Rupp Jr, R. M. Fleming, A. C. W. P. James, and P. Marsh. Synthesis of bulk superconducting

YBa2Cu4O8 at one atmosphere oxygen pressure. Nature, 338:328, 1989.

[165] R. Ramesh, S. Jin, and P. Marsh. Superconductor defect structure. Nature, 346:420, 1990.

[166] R. Ramesh, D. M. Hwang, T. Venkatesan, T. S. Ravi, L. Nazar, A. Inam, X. D. Wu, B. Dutta, G. Thomas, A. F. Marshall, T. H., and Geballe. Direct observation of structural defects in laser-deposited superconducting Y-Ba-Cu-O thin films.

Science, 247:57, 1990.

[167] R. Ramesh, S. Jin, S. Nakahara, and T. H. Tiefel. Phase decomposition and

structural defects in a Y-Ba-Cu-O superconductor. Appl. Phys. Lett., 57:1458, 1990.

[168] M. A. Señaris-Rodríguez, A. M. Chippindale, A. Várez, E. Morán, and M. A. Alario-

Franco. A novel “126” phase of the family of Y2Ba4Cu6+nO14+n high-temperature

superconducting materials. Physica C, 172:477, 1991. BIBLIOGRAPHY 123

[169] G. F. Voronin and S. A. Degterov. Thermodynamics of superconducting phases in

the Y-Ba-Cu-O system. Physica C, 176:387, 1991.

[170] M. Naito, H. Sato, and H. Yamamoto. Oxygen nonstoichiometry of MBE-grown

cuprate films to lose or give high-Tc superconductivity. Proc. SPIE, 3481:214, 1998.

[171] D. G. Hawthorn, F. He, L. Venema, H. Davis, A. J. Achkar, J. Zhang, R. Su- tarto, H. Wadati, A. Radi, T. Wilson, G. Wright, K. M. Shen, J. Geck, H. Zhang, V. Novák, and G. A. Sawatzky. An in-vacuum diffractometer for resonant elastic

soft x-ray scattering. Rev. Sci. Instrum., 82:073104, 2011.

[172] M. A. Beno, L. Soderholm, D. W. Capone II, D. G. Hinks, J. D. Jorgensen, J. D. Grace, Ivan K. Schuller, C. U. Segre, and K. Zhang. Structure of the single‐phase

high‐temperature superconductor YBa2Cu3O7−δ. Appl. Phys. Lett., 51:57, 1987.

[173] R. G. Buckley, D. M. Pooke, J. L. Tallon, M. R. Presland N. E. Flower, M. P. Staines, H. L. Johnson, M. Meylan, G. V. M. Williams, and M. Bowden. Ca-

and La-substitution in YBa2Cu3O7−ξ,Y2Ba4Cu7O15−ξ and YBa2Cu4O8. Physica C, 174:383, 1991.

[174] A. Krol, C. S. Lin, Z. H. Ming, C. J. Sher, Y. H. Kao, C. T. Chen, F. Sette, Y. Ma, G. C. Smith, Y. Z. Zhu, and D. T. Shaw. X-ray-absorption studies of Y-Ba-Cu- O and Bi-Sr-Ca-Cu-O films at oxygen K edge by means of fluorescence and total

electron yield: A comparison of two techniques. Phys. Rev. B, 42:2635, 1990.

[175] A. Ruosi, C. Raisch, A. Verna, R. Werner, B. A. Davidson, J. Fujii, R. Kleiner, and D. Koelle. Electron sampling depth and saturation effects in perovskite films

investigated by soft x-ray absorption spectroscopy. Phys. Rev. B, 90:125120, 2014.

[176] J. Fink, N. Nücker, E. Pellegrin, H. Romberg, M. Alexander, and M. Knupfer. Electron energy-loss and x-ray absorption spectroscopy of cuprate superconductors

and related compounds. J. Electr. Spectr. and Rel. Phen., 66:395, 1994. BIBLIOGRAPHY 124

[177] C. T. Chen, L. H. Tjeng, J. Kwo, H. L. Kao, P. Rudolf, F. Sette, and R. M. Fleming.

Out-of-plane orbital characters of intrinsic and doped holes in La2−xSrxCuO4. Phys. Rev. Lett., 68:2543, 1992.

[178] E. Pellegrin, N. Nücker, J. Fink, S. L. Molodtsov, A. Gutiérrez, E. Navas, O. Strebel, Z. Hu, M. Domke, G. Kaindl, S. Uchida, Y. Nakamura, J. Markl, M. Klauda, G. Saemann-Ischenko, A. Krol, J. L. Peng, Z. Y. Li, and R. L. Greene.

Orbital character of states at the Fermi level in La2−xSrxCuO4 and R2−xCexCuO4

(R=Nd,Sm). Phys. Rev. B, 47:3354, 1993.

[179] K. B. Garg, S. Dalela, N. L. Saini, R. K. Singhal, D. C. Jain, and C. T. Chen. Po- larised XAS study of anomalous temperature dependence of aggregation of itiner-

ant holes and pair formation in a YBa2Cu3O∼7−δ single crystal. Physica C, 399:98, 2003.

[180] D. G. Hawthorn, K. M. Shen, J. Geck, D. C. Peets, H. Wadati, J. Okamoto, S.-W. Huang, D. J. Huang, H.-J. Lin, J. D. Denlinger, Ruixing Liang, D. A. Bonn, W. N. Hardy, and G. A. Sawatzky. Resonant elastic soft x-ray scattering in oxygen-ordered

YBa2Cu3O6+δ. Phys. Rev. B, 84:075125, 2011.

[181] N. Nücker, E. Pellegrin, P. Schweiss, J. Fink, S. L. Molodtsov, C. T. Simmons, G. Kaindl, W. Frentrup, A. Erb, and G. Müller-Vogt. Site-specific and doping-

dependent electronic structure of YBa2Cu3Ox probed by O 1s and Cu 2p x-ray

absorption spectroscopy. Phys. Rev. B, 51:8529, 1995.

[182] M. B. J. Meinders, H. Eskes, and G. A. Sawatzky. Spectral-weight transfer: Break-

down of low-energy-scale sum rules in correlated systems. Phys. Rev. B, 48:3916, 1993.

[183] K. Momma and F. Izumi. VESTA 3 for three-dimensional visualization of crystal, volumetric and morphology data. J. Appl. Crystallogr., 44:1272, 2011. BIBLIOGRAPHY 125

[184] K. Hashimoto, M. Akiyoshi, A. Wisniewski, M. L. Jenkins, Y. Toda, and T. Yano.

A high-resolution electron microscopy study of structural defects in YBa2Cu4O8

superconductor. Physica C, 269:139, 1996.

[185] J. Karpinski, G. I. Meijer, H. Schwer, R. Molinski, E. Kopnin, K. Conder, M. Angst, J. Jun, S. Kazakov, A. Wisniewski, R. Puzniak, J. Hofer, V. Alyoshin, and A. Sin. High-pressure synthesis, crystal growth, phase diagrams, structural and magnetic

properties of Y2Ba4CunO2n+x, HgBa2Can−1CunO2n+2+δ and quasi-one-dimensional

cuprates. Supercond. Sci. Technol., 12:R153, 1999.

[186] R. Stern, M. Mali, J. Roos, and D. Brinkmann. Spin pseudogap and interplane

63 coupling in Y2Ba4Cu7O15:A Cu nuclear spin-spin relaxation study. Phys. Rev. B, 51:15478, 1995.

[187] E. Pavarini, I. Dasgupta, T. Saha-Dasgupta, O. Jepsen, and O. K. Andersen. Band-

structure trend in hole-doped cuprates and correlation with Tc max. Phys. Rev. Lett, 87:047003, 2001.

[188] W. Hu, S. Kaiser, D. Nicoletti, C. R. Hunt, I. Gierz, M. C. Hoffmann, M. Le Tacon, T. Loew, B. Keimer, and A. Cavalleri. Optically enhanced coherent transport

in YBa2Cu3O6.5 by ultrafast redistribution of interlayer coupling. Nat. Mater., 13:705, 2014.

[189] Takahiro Wada, Ataru Ichinose, Fujio Izumi, Akiko Nara, H. Yamauchi, Hajime- Asano, Shoji, and Tanaka. Neutron powder diffraction study of the Pb-based cop-

per oxide containing thick fluorite blocks: (Pb1/2Cu1/2)Sr2(Ho1/3Ce2/3)3Cu2O11+z.

Physica C, 179:455, 1991.

[190] Ernst Bauer and Manfred Sigrist, editors. Non-Centrosymmetric Superconductors: Introduction and Overview, volume 847. Springer-Verlag Berlin Heidelberg, 2012. BIBLIOGRAPHY 126

[191] C C Tsuei and J R Kirtley. Pairing symmetry in cuprate superconductors. Rev. Mod. Phys., 72:969, 2000.

[192] E Bauer, G Hilscher, H Michor, Ch Paul, E W Scheidt, A Gribanov, Yu Seropegin, H Noël, M Sigrist, and P Rogl. Heavy fermion superconductivity and magnetic

order in noncentrosymmetric CePt3Si. Phys. Rev. Lett., 92:027003, 2004.

[193] H. Q. Yuan, D. F. Agterberg, N. Hayashi, P. Badica, D. Vandervelde, K. Togano, M. Sigrist, and M. B. Salamon. S-wave spin-triplet order in superconductors with-

out inversion symmetry: Li2Pd3B and Li2Pt3B. Phys. Rev. Lett., 97:017006, 2006.

[194] T. Shibauchi, L. Krusin-Elbaum, Y. Kasahara, Y. Shimono, Y. Matsuda, R. D. McDonald, C. H. Mielke, S. Yonezawa, Z. Hiroi, M. Arai, T. Kita, G. Blatter, and M. Sigrist. Uncommonly high upper critical field of the pyrochlore superconductor

KOs2O6 below the enhanced paramagnetic limit. Phys. Rev. B, 74:220506(R), 2006.

[195] Q. Huang, J. L. Soubeyroux, O. Chmaissen, I. Natali Sora, A. Santoro, R. J. Cava, J. J. Krajewski, W. F. Peck, and Jr. Neutron powder diffraction study of the

crystal structures of Sr2RuO4 and Sr2IrO4 at room temperature and at 10 k. J. Solid State Chem., 112:355, 1994.

[196] K. Crawford, M. A. Subramanian, R. L. Harlow, J. A. Fernandez-Baca, Z. R. Wang,

and D. C. Johnston. Structural and magnetic studies of Sr2IrO4. Phys. Rev. B, 49:919, 1994.

[197] J.-Y. P. Delannoy, M. J. P. Gingras, P. C. W. Holdsworth, and A.-M. S. Tremblay. Low-energy theory of the t - t0 - t00 - U Hubbard model at half-filling: Interac- tion strengths in cuprate superconductors and an effective spin-only description of

La2CuO4. Phys. Rev. B, 79:235130, 2009.

[198] H. Jin, H. Jeong, T. Ozaki, and J. Yu. Anisotropic exchange interactions of spin-

orbit-integrated states in Sr2IrO4. Phys. Rev. B, 80:075112, 2009. BIBLIOGRAPHY 127

[199] H. Watanabe, T. Shirakawa, and S. Yunoki. Microscopic study of a spin-orbit-

induced Mott insulator in ir oxides. Phys. Rev. Lett., 110:027002, 2013.

[200] Yang Yang, Wan-Sheng Wang, Jin-Guo Liu, Hua Chen, Jian-Hui Dai, and Qiang-

Hua Wang. Superconductivity in doped Sr2IrO4: A functional renormalization

group study. Phys. Rev. B, 89:094518, 2014.

[201] Zi Yang Meng, Yong Baek Kim, and Hae-Young Kee. Odd-parity triplet superconducting phase in multiorbital materials with a strong spin-orbit coupling: Appli-

cation to doped Sr2IrO4. Phys. Rev. Lett., 113:177003, 2014.

[202] A. de la Torre, S. McKeown Walker, F.Y. Bruno, S. Riccó, Z. Wang, I. Gutierrez Lezama, G. Scheerer, G. Giriat, D. Jaccard, C. Berthod, T.K. Kim, M. Hoesch, E.C. Hunter, R.S. Perry, A. Tamai, and F. Baumberger. Collapse of the Mott

gap and emergence of a nodal liquid in lightly doped Sr2IrO4. Phys. Rev. Lett., 115:176402, 2015.

[203] O. B. Korneta, Tongfei Qi, L. E. De Long S. Chikara, S. Parkin, P. Schlottmann,

and G. Cao. Electron-doped Sr2IrO4−δ (0 ≤ δ ≤ 0.04): Evolution of a disordered

1 Jeﬀ = 2 Mott insulator into an exotic metallic state. Phys. Rev. B, 82:115117, 2010.

[204] Y. K. Kim, N. H. Sung, J. D. Denlinger, and B. J. Kim. Observation of a d-wave

gap in electron-doped Sr2IrO4. Nat. Phys., 12:37, 2016.

[205] Y.J. Yan, M.Q. Ren, H.C. Xu, B.P. Xie, R. Tao, H.Y. Choi, N. Lee, Y.J. Choi,

T. Zhang, and D.L. Feng. Electron-doped Sr2IrO4: An analogue of hole-doped

cuprate superconductors demonstrated by scanning tunneling microscopy. Phys. Rev. X, 5:041018, 2015.

[206] M. Ge, T. F. Qi, O. B. Korneta, D. E. De Long, P. Schlottmann, W. P. Crummett, and G. Cao. Lattice-driven magnetoresistivity and metal-insulator transition in

single-layered iridates. Phys. Rev. B, 84:100402, 2011. BIBLIOGRAPHY 128

[207] Y. Klein and I. Terasaki. Insight on the electronic state of Sr2IrO4 revealed by

cationic substitutions. J. Phys.: Condens. Matter, 20:295201, 2008.

[208] T. F. Qi, O. B. Korneta, L. Li, K. Butrouna, V. S. Cao, Xiangang Wan, P. Schlottmann, R. K. Kaul, and G. Cao. Spin-orbit tuned metal-insulator transi-

tions in single-crystal Sr2Ir1−xRhxO4 (0 ≤ x ≤ 1). Phys. Rev. B, 86:125105, 2012.

[209] J. P. Clancy, A. Lupascu, H. Gretarsson, Z. Islam, Y. F. Hu, D. Casa, C. S. Nelson, S. C. LaMarra, G. Cao, and Young-June Kim. Dilute magnetism and spin-orbital

percolation effects in Sr2Ir1−xRhxO4. Phys. Rev. B, 89:054409, 2014.

[210] J. D. Jorgensen, B. Dobrowski, S. Pei, D. G. Hinks, L. Soderholm, M. Morosin, J. E.

Schirber, E. L. Venturini, and D. S. Ginley. Superconducting phase of La2CuO4+δ:

A superconducting composition resulting from phase separation. Phys. Rev. B, 38:11337, 1988.

[211] K. Chandrasekaran, R. Vijayaraghavan, and U. V. Varadaraju. Effects of oxygen

non-stoichiometry and cationic substitutions on the properties of Sr2RuO4+δ. Mat. Chem. Phys., 56:63, 1998.

[212] Stephen J Skinner and John A Kilner. Oxygen ion conductors. Mater. Today, 6:30, 2003.

[213] S.-Y. Jeona, M.-B. Choia, H.-N. Ima, J.-H. Hwangb, and S.-J. Songa. Oxygen ionic

conductivity of La2NiO4+δ via interstitial oxygen defect. J. Phys. Chem. Solids, 73:656, 2012.

[214] Thomas Jarlborg and Antonio Bianconi. Fermi surface reconstruction of superox-

ygenated La2CuO4 superconductors with ordered oxygen interstitials. Phys. Rev. B, 87:054514, 2013. BIBLIOGRAPHY 129

[215] H. Zhang, N. Gauquelin, C. McMahon, D. G. Hawthorn, G. A. Botton, and J. Y. T. Wei. Synthesis of high-oxidation Y-Ba-Cu-O phases in superoxygenated thin films.

Phys. Rev. Materials, 2:033803, 2018.

[216] Yue Wu, Donald E. Ellis, Li Shen, and Thomas O. Mason. Point defects of

La2CuO4-based ceramics, part i: Oxygen interstitials. J. Am. Ceram. Soc., 79:1599, 1996.

[217] F. Cordero and R. Cantelli. Interstitial O and O vacancies in La2CuO4+δ during

high-temperature treatments. Physica C, 312:213, 1999.

[218] K. H. Lee and R. Hoffmann. Oxygen interstitials in superconducting La2CuO4:

their valence state and role. J. Phys. Chem. A, 110:609, 2006.

[219] James F. Shackelford and William Alexander. CRC Materials Science and Engi- neering Handbook. CRC Press, 3 edition, 2000.

[220] J. Nichols, J. Terzic, E. G. Bittle, O. B. Korneta, L. E. De Long, J. W. Brill, G. Cao,

and S. S. A. Seo. Tuning electronic structure via epitaxial strain in Sr2IrO4 thin

films. Appl. Phys. Lett., 102:141908, 2013.

[221] Abhijit Biswas, Ki-Seok Kim, and Yoon Hee Jeong. Metal insulator transitions in

perovskite SrIrO3 thin films. J. Appl. Phys., 116:213704, 2014.

[222] D. J. Groenendijk, N. Manca, G. Mattoni, L. Kootstra, S. Gariglio, Y. Huang, E. van Heumen, and A. D. Caviglia. Epitaxial growth and thermodynamic stability

of SrIrO3/SrTiO3 heterostructures. Appl. Phys. Lett., 109:041906, 2016.

[223] John H. Gruenewald, John Nichols, Jasminka Terzic, Gang Cao, Joseph W. Brill, and Sung S.Ambrose Seo. Compressive strain-induced metal–insulator transition

in orthorhombic SrIrO3 thin films. J. Mater. Res., 29:2491, 2014. BIBLIOGRAPHY 130

[224] Lunyong Zhang, Qifeng Liang, Ye Xiong, Binbin Zhang, Lei Gao, Handong Li, Y. B. Chen, Jian Zhou, Shan-Tao Zhang, Zheng-Bin Gu, Shu-hua Yao, Zhiming Wang, Yuan Lin, and Yan-Feng Chen. Tunable semimetallic state in compressive-strained

SrIrO3 films revealed by transport behavior. Phys. Rev. B, 91:035110, 2015.

[225] D. J. Groenendijk, C. Autieri, J. Girovsky, M. Carmen Martinez-Velarte, N. Manca, G. Mattoni, A. M. R. V. L. Monteiro, N. Gauquelin, J. Verbeeck, A. F. Otte,

M. Gabay, S. Picozzi, and A. D. Caviglia. Spin-orbit semimetal SrIrO3 in the

two-dimensional limit. Phys. Rev. Lett., 119:256403, 2017.

[226] J. G. Zhao, L. X. Yang, Y. Yu, F. Y. Li, R. C. Yu, Z. Fang, L. C. Chen, and C. Q.

Jin. High-pressure synthesis of orthorhombic SrIrO3 perovskite and its positive

magnetoresistance. J. Appl. Phys., 103:103706, 2008.

[227] J. Fujioka, T. Okawa, A. Yamamoto, and Y. Tokura. Correlated dirac semimetal-

lic state with unusual positive magnetoresistance in strain-free perovskite SrIrO3.

Phys. Rev. B, 95:121102, 2017.

[228] F. A. Kröger and H. J. Vink. Relations between the concentrations of imperfections

in crystalline solids. In Frederick Seitz and David Turnbull, editors, Solid State Physics, volume 3, page 307. Academic Press, 1956.

[229] S. S. A. Seo, J. Nichols, J. Hwang, J. Terzic, J. H. Gruenewald, M. Souri, J. Thomp-

son, J. G. Connell, and G. Cao. Selective growth of epitaxial Sr2IrO4 by controlling

plume dimensions in pulsed laser deposition. Appl. Phys. Lett., 109:201901, 2016.

[230] J. H. Lee, G. Luo, I. C. Tung, S. H. Chang, Z. Luo, M. Malshe, M. Gadre, A. Bhat- tacharya, S. M. Nakhmanson, J. A. Eastman, H. Hong, J. Jellinek, D. Morgan, D. D. Fong, and J. W. Freeland. Dynamic layer rearrangement during growth of

layered oxide films by molecular beam epitaxy. Nat. Mater., 13:879, 2014. BIBLIOGRAPHY 131

[231] T. Mizokawa, L. H. Tjeng, G. A. Sawatzky, G. Ghiringhelli, O. Tjernberg, N. B. Brookes, H. Fukazawa, S. Nakatsuji, and Y. Maeno. Spin-orbit coupling in the

Mott insulator Ca2RuO4. Phys. Rev. Lett., 87:077202, 2001.

[232] M. Schmidt, T. R. Cummins, M. Bürk, D. H. Lu, N. Nücker, S. Schuppler, and F. Lichtenberg. Nature of the electronic states in the layered perovskite noncuprate

superconductor Sr2RuO4. Phys. Rev. B, 53:R14761(R), 2016.

[233] C. H. Sohn, Deok-Yong Cho, C.-T. Kuo, L. J. Sandilands, T. F. Qi, G. Cao, and T. W. Noh. X-ray absorption spectroscopy study of the effect of Rh doping in

Sr2IrO4. Sci. Rep., 6:23856, 2016.

[234] J. Matsuno, K. Ihara, S. Yamamura, H. Wadati, K. Ishii, V.V. Shankar, Hae-Young Kee, and H. Takagi. Engineering a spin-orbital magnetic insulator by tailoring

superlattices. Phys. Rev. Lett., 114:247209, 2015.

[235] D. W.Jeong, W. S. Choi, S. Okamoto, J.–Y. Kim, K. W. Kim, S. J. Moon, D.–Y. Cho, H. N. Lee, and T. W. Noh. Dimensionality control of d-orbital occupation in

oxide superlattices. Sci. Rep., 4:6124, 2014.

[236] E. Pellegrin, C.T. Chen, J. Zaanen, and J. Fink. Symmetry of hole states in

La2−xSrxCuO4+δ and La2−xSrxNiO4+δ. Physica B, 208:487, 1995.

[237] C. T. Chen, F. Sette, Y. Ma, M. S. Hybertsen, E. B. Stechel, W. M. C. Foulkes, M. Schulter, S-W. Cheong, A. S. Cooper, L. W. Rupp, Jr., B. Batlogg, Y. L. Soo,

Z. H. Ming, A. Krol, and Y. H. Kao. Electronic states in La2−xSrxCuO4+δ probed

by soft-x-ray absorption. Phys. Rev. Lett., 66:104, 1991.

[238] Kazunori Nishio, Harold Y. Hwang, and Yasuyuki Hikita. Thermodynamic guiding principles in selective synthesis of strontium iridate ruddlesden-popper epitaxial

films. Appl. Phys. Lett. Mater., 4:036102, 2016. BIBLIOGRAPHY 132

[239] Seung Yup Jang, Soon Jae Moon, Byung Chul, and Jin-Seok Chung. Pld growth of

epitaxially-stabilized 5d perovskite SrIrO3 thin films. J. Korean Phys. Soc, 56:1814, 2010.