ABSTRACT Title of dissertation: MAGNETISM AND SUPERCONDUCTIVITY IN TOPOTACTICALLY MODIFIED TRANSITION METAL CHALCOGENIDES Brandon Cody Wilfong Doctor of Philosophy, 2020 Dissertation directed by: Professor Efrain Rodriguez Department of Chemistry and Biochemistry Professor Johnpierre Paglione Department of Physics Inspired by the structure of the simplest iron-based chalcogenide superconductor, FeSe, the class of tetrahedral transition metal chalcogenides (TTMCs) exhibit interesting chemical and physical properties due to its structure. This structure consists of tetrahe- drally coordinated transition metal chalcogenides stacked to form two dimensional layers held together by van der Waals forces. This structure and its associated tetrahedral co- ordination of transition metal to chalcogenide, square transition metal sublattice, van der Waals layered structure, and d-electron filling at the Fermi level yields interesting prop- erties from superconductivity to frustrated itinerant magnetism. In this dissertation work, we demonstrate that the anti-PbO type FeCh (Ch = S, Se, Te) structure offers a perfect platform for the study of superconductivity in the iron-based system as well as new physics as the class is expanded to different transition metals. Prior to this work, the binaries of the TTMC family was limited to iron, but has been expanded to cobalt. In the cobalt compound, CoSe, superconductivity in the FeSe binary is sup- pressed and a frustrated spin glass-like magnetic state emerges. Beyond the binaries, we have shown that topotactic hydrothermal synthetic routes on the iron chalcogenide system can lead to novel intercalated phases where long range magnetic order can co-exist with superconductivity in the (LiOH)FeSe system. This synthetic scheme also allows the in- tercalation of organic molecules, specifically ethylenediamine, to form organic-inorganic hybrids which can offer a new avenue for designing heterolayer compounds with complex interlayer interactions and bonding. MAGNETISM AND SUPERCONDUCTIVITY IN TOPOTACTICALLY MODIFIED TRANSITION METAL CHALOCGENDIES by Brandon Cody Wilfong Dissertation submitted to the Faculty of the Graduate School of the University of Maryland, College Park in partial fulfillment of the requirements for the degree of Doctor of Philosophy 2020 Advisory Committee: Professor Efrain Rodriguez, Chair/Co-advisor Professor Johnpierre Paglione, Co-advsior Professor Ichiro Takeuchi Professor Andrei Vedernikov Professor Nicholas Butch ?c Copyright by Brandon Cody Wilfong 2020 Foreward The following is a list of all the manuscripts that I have contributed to during my time as a graduate student those marked as under review, in submission, or in prepara- tion denote manuscripts which have yet to be officially published. An asterisk denotes manuscripts that make up the work described in this thesis. * Zhou, X.; Wilfong, B.; Vivanco, H.; Paglione, J.; Brown, C. M.; and Rodriguez, E. E. Metastable layered cobalt chalcogenides from topochemical deintercalation. J. Am. Chem. Soc. 2016, 138(50), 16432-16442. Larson, A. M.; Wilfong, B.; Moetakef, P.; Brown, C. M; Zavalij, P.; and Rodriguez, E. E. Metal?insulator transition tuned by magnetic field in Bi1.7V8O16 hollandite. J. Mater. Chem. C. 2017, 5(20), 4967-4976. Zhou, X.; Eckberg, C.; Wilfong, B.; Liou, S. C.; Vivanco, H. K.; Paglione, J.; and Rodriguez, E. E. Superconductivity and magnetism in iron sulfides intercalated by metal hydroxides. Chem. Sci. 2017, 8(5), 3781-3788. Zhou, X.; Wilfong, B.; Liou, S. C.; Hodovanets, H.; Brown; C. M., and Rodriguez, E. E. Proton and ammonia intercalation into layered iron chalcogenides. Chem. Comm. 2018, 54(50), 6895-6898. * Wilfong, B.; Zhou, X.; Vivanco, H.; Campbell, D. J.; Wang, K.; Graf, D.; Paglione, J.; and Rodriguez, E. E. Frustrated magnetism in the tetragonal CoSe analog ii of superconducting FeSe. Phys. Rev. B. 2018, 97(10), 104408. Virtue, A.; Zhou, X.; Wilfong, B.; Lynn, J. W.; Taddei, K.; Zavalij, P.; Wang, L.; and Rodriguez, E. E. Magnetic order effects on the electronic structure of KMMnS2 (M= Cu, Li) with the ThCr2Si2-type structure. Phys. Rev. Mater., 2019, 3(4), 044411. Zhou, X.; Wang, L.; Fan, X.; Wilfong, B.; Liou, S. C.; Wang, Y.; Zheng, H.; Feng, Z.; Wang, C.; and Rodriguez, E. E. (2020). Isotope Effect between H2O and D2O in Hy- drothermal Synthesis. Chem. Mater. 2020, 32(2), 769-775. * Wilfong, B.; Zhou, X.; Zheng, H.; Babra, N.; Brown, C. M.; Lynn, J. W.; Tad- dei, K.; Paglione, J.; and Rodriguez, E. E. Long range magnetic order in hydroxide layer doped (Li 1?xy FexMnyOD) FeSe. arXiv preprint arXiv:1912.09329, under review. * Wilfong, B.; Zhou, Z.; Zheng, H.; Jayathilake, R.; Campbell, D.J.; Liou, S.C.; Paglione, J.; and Rodriguez, E.E. Alkali metal-free hydrothermal synthesis of ethylenedi- amine intercalated iron chalcogenides. in preparation. Campbell, D.J.; Wilfong, B.; Zic, M.; Zavalij, P.; Rodriguez, E.E.; and Paglione, J. Preparation and Properties of KCo2As2 Single Crystals. in preparation. iii Dedication To my parents who have supported me in every aspect of my life and have given me every opportunity to pursue and achieve any goal I desired. And to my best friend, Elizabeth Cardosa who?s love and support has made everything possible. iv Acknowledgments I would like to thank both of my advisors Professor Efrain Rodriguez and Professor Johnpierre Paglione for guidance and support for this dissertation work. I would also like to thank my other committee members, Professors Ichiro Takeuchi, Andrei Vedernikov, and Nicholas Butch for their time to review this dissertation and valuable advice to im- prove this work. I want to thank the present and past members of the Rodriguez and Paglione groups for their assistance with experiments and pleasant company: Xiuquan Zhou, Huafei Zheng, Stephanie Gnewuch, Timothy Diethrich, Lahari Balisetty, Austin Virtue, Amber Lar- son, Tianyu Li, Daniel Campbell, Chris Eckberg, Connor Roncaioli, Halyna Hodovanets, Shanta Saha, I-Lin Liu. In particular, my collaboration with Dr. Xiuquan Zhou led to an incredible 3 years of learning, productivity and friendship which made this thesis work possible. I am grateful to the three undergraduate students who helped with my research in the chemistry and physics departments: Hector Vivanco, Navneeth Babra and Mark Zic. This work could not have been done without all of their help over these five years. I would like to thank Drs. Peter Zavalij, Sz-Chian Liou, and Marya Anderson at the University of Maryland user facilities for their assistance with structural, microscopy, and elemental analysis analysis. I would also like to thank Drs. Craig Brown, Jeffrey Lynn, Jose Rodriguez, Yiming Qui, Keith Taddei, and Simon Kimber for their assistance with neutron scattering and spectroscopy. Research at the University of Maryland was supported by the NSF Career DMR- 1455118, AFOSR Grant No. FA9550-14-10332, and the Gordon and Betty Moore Foun- v dation Grant No. GBMF4419. We also acknowledge support from the Maryland Nanocen- ter and Center for Nanophysics and Advanced Materials. We acknowledge the support of the National Institute of Standards and Technology, U. S. Department of Commerce, in providing the neutron research facilities used in this work. The use of the Advanced Photon Source at Argonne National Laboratory was supported by the U. S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract No. DE- AC02-06CH11357. In particular, we acknowledge the use of 11-BM and 17-BM through collaboration with Dr. Andrey Yakovenko. We acknowledge the University of Maryland supercomputing resources (http://www.it.umd.edu/hpcc) made available for conducting the research reported in this work. Lastly, I cannot be more thankful for the support of my friends and family of the past five years. Doctoral coursework and research can be intensely difficult at times and the unwavering support of my friends and family, even when they don?t know what I am talking about, has made this process not only doable but enjoyable. In particular, Elizabeth and Matt have been constants and helped make the past five years a time of learning and personal growth for which I am incredibly thankful. vi Contents Foreward . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i Dedication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.1 Iron-based superconductors - background and motivation . . . . . . . . . 1 1.2 Tetrahedral transition metal chalcogenides . . . . . . . . . . . . . . . . . 7 1.3 Topotactic chemistry and hydrothermal synthesis . . . . . . . . . . . . . 12 1.4 Objectives and Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 List of Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2. Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 2.1 Synthetic Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 2.1.1 Self-flux synthesis of ThCr2Si2-type transition metal chalcogenides 21 2.1.2 Soft chemical reductive de-intercalation . . . . . . . . . . . . . . 23 2.1.3 Hydrothermal ion exchange . . . . . . . . . . . . . . . . . . . . 25 2.1.4 Alkali metal-free ethylenediamine intercalation of iron chalco- genides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 2.2 Characterization Methods . . . . . . . . . . . . . . . . . . . . . . . . . . 29 2.2.1 Laboratory powder x-ray diffraction . . . . . . . . . . . . . . . . 30 2.2.2 Synchrotron powder x-ray diffraction . . . . . . . . . . . . . . . 31 2.2.3 Neutron powder x-ray diffraction . . . . . . . . . . . . . . . . . 32 2.2.4 DC magnetic susceptibility and magnetization . . . . . . . . . . . 35 2.2.5 AC magnetic susceptibility . . . . . . . . . . . . . . . . . . . . . 40 2.2.6 Longitudinal electrical resistivity . . . . . . . . . . . . . . . . . 41 2.2.7 Low temperature specific heat . . . . . . . . . . . . . . . . . . . 43 2.2.8 Electron microscopy . . . . . . . . . . . . . . . . . . . . . . . . 45 2.2.9 Inductively coupled plasma - atomic emission spectroscopy . . . 48 2.2.10 Thermal stability and characterization . . . . . . . . . . . . . . . 50 vii 2.2.11 Density functional theory calculations . . . . . . . . . . . . . . . 52 3. Topochemical synthesis and frustrated magnetism in CoSe - analog of supercon- ducting FeSe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 3.1 Prologue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 3.2 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 3.3 Experimental Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 3.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 3.4.1 Magnetic properties . . . . . . . . . . . . . . . . . . . . . . . . 65 3.4.2 Transport Properties . . . . . . . . . . . . . . . . . . . . . . . . 71 3.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 3.5.1 Ground state of CoSe . . . . . . . . . . . . . . . . . . . . . . . . 76 3.5.2 Anisotropy and Magnetic Direction . . . . . . . . . . . . . . . . 79 3.5.3 FeSe vs. CoSe . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 3.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 4. Long-range magnetic order in transition metal doped (LiOH)FeSe by soft chemi- cal design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 4.2 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 4.2.1 Synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 4.2.2 Magnetic and transport measurements . . . . . . . . . . . . . . . 90 4.2.3 X-ray diffraction measurements . . . . . . . . . . . . . . . . . . 90 4.2.4 Neutron diffraction measurements . . . . . . . . . . . . . . . . . 91 4.3 Evidence for long range magnetic order . . . . . . . . . . . . . . . . . . 92 4.4 Hydrothermal synthesis and crystallographic results . . . . . . . . . . . . 95 4.5 Magnetic and transport properties . . . . . . . . . . . . . . . . . . . . . 99 4.6 Effect of other transition metal dopants . . . . . . . . . . . . . . . . . . . 104 4.7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 5. Alkali metal-free hydrothermal synthesis of ethylenediamine intercalated iron chalco- genides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 5.2 Experimental Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . 120 5.3 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 5.3.1 Alkali metal free hydrothermal ethylenediamine intercalation . . 121 5.3.2 Intercalated ethylenediamine crystallography and symmetry . . . 126 5.3.3 Intercalated ethylenediamine configuration and guest-host inter- actions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132 5.3.4 Magnetic and transport properties of intercalated species . . . . . 137 5.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 6. Conclusions and Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 6.1 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 6.2 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 viii Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176 ix List of Tables 3.1 Structural parameters for KCo2Se2 and CoSe . . . . . . . . . . . . . . . 60 4.1 Lattice parameters of transition metal doped K0.85Fe1.8?xMxSe2 M = (Mn, Co, Ni, Cu, Zn) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 4.2 Structural parameters for (Li0.875(2)Fe0.062(3)Mn0.062(3)OD)FeSe . . . . . . . 98 4.3 Structural parameters for (Li0.842(1)Fe0.135(4)Co0.02(3)OD)Fe0.950(2)Se . . . . 105 4.4 Structural parameters for Li0.822(1)Fe0.176(3)OD)Fe0.964(3)Ni0.020(1)Se . . . . 106 4.5 Structural parameters for (Li0.857(1)Fe0.14(4)OD)Fe0.992(3)Se . . . . . . . . . 107 4.6 Structural parameters for (Li0.847(2)Fe0.15(1)OD)Fe0.995(2)Se . . . . . . . . . 108 x List of Figures 1.1 Crystal structures of some iron-based superconductors . . . . . . . . . . 6 1.2 Representative crystal structure of tetrahedral transition metal chalcogenides 9 1.3 Pourbaix diagram of iron . . . . . . . . . . . . . . . . . . . . . . . . . . 17 2.1 As-recovered single crystals of KxM2Ch2 compounds . . . . . . . . . . . 22 2.2 Soft chemical reductive de-intercalation of KCo2Ch2 to form CoCh . . . . 24 2.3 Hydrothermal cation exchange routes for iron chalcogenides . . . . . . . 26 2.4 Schematic of longitudinal resistivity measurement technique . . . . . . . 42 3.1 Reaction schemes for the topochemical synthesis of cobalt chalcogenides 58 3.2 Powder x-ray and neutron diffraction on KCo2Se2 and CoSe . . . . . . . 59 3.3 Comparison of crystal structure and magnetic transition temperatures for KCo2Se2 and CoSe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 3.4 Curie-Weiss analysis of CoSe at different applied fields . . . . . . . . . . 66 3.5 Normalized magnetic susceptibility at different applied fields . . . . . . . 68 3.6 AC magnetic susceptibility of CoSe . . . . . . . . . . . . . . . . . . . . 69 3.7 Longitudinal electrical resistance and magnetoresistance of CoSe single crystals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 3.8 Specific heat of single crystal KCo2Se2 . . . . . . . . . . . . . . . . . . . 74 3.9 Specific heat of a pressed pellet of CoSe . . . . . . . . . . . . . . . . . . 75 3.10 Non-spin polarized density of states for CoSe . . . . . . . . . . . . . . . 77 3.11 Magnetic anisotropy of CoSe single crystals by magnetic susceptibility measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 4.1 Synthetic scheme for (Li1?x?yFexMnyOD)FeSe . . . . . . . . . . . . . . . 86 4.2 Powder neutron diffraction results of Mn-doped (LiOD)FeSe . . . . . . . 92 4.3 Powder x-ray and neutron diffraction of (Li1?x?yFexMnyOD)FeSe . . . . . 96 4.4 Magnetic susceptibility and isothermal magnetization of (Li1?x?yFexMnyOD)FeSe single crystal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 4.5 Resistivity and specific heat measurements and extracted magnetic en- tropy of (Li1?x?yFexMnyOD)FeSe single crystal . . . . . . . . . . . . . . 111 4.6 Powder x-ray and neutron diffraction of (Li1?x?yFexNiyOD)FeSe . . . . . 112 4.7 Powder x-ray and neutron diffraction of (Li1?x?yFexCuyOD)FeSe . . . . . 113 4.8 Powder x-ray and neutron diffraction of (Li1?x?yFexZnyOD)FeSe . . . . . 114 4.9 Powder x-ray and neutron diffraction of (Li1?x?yFexCoyOD)FeSe . . . . . 115 xi 4.10 Magnetic susceptibility and isothermal magnetization of (Li1?x?yFexNiyOD)FeSe powders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116 4.11 Magnetic susceptibility and isothermal magnetization of (Li1?x?yFexCuyOD)FeSe powders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116 4.12 Magnetic susceptibility and isothermal magnetization of (Li1?x?yFexZnyOD)FeSe powders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 4.13 Magnetic susceptibility and isothermal magnetization of (Li1?x?yFexCoyOD)FeSe powders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 5.1 Pawley refinements on ground powders of (C2H8N2)xFeS after varying hydrothermal reaction times. . . . . . . . . . . . . . . . . . . . . . . . . 123 5.2 Pawley refinements on ground powders of (C2H8N2)yFeSe after varying hydrothermal reaction times. . . . . . . . . . . . . . . . . . . . . . . . . 124 5.3 TGA and SEM/EDS analysis of (C2H8N2)xFeS ground powders. . . . . . 126 5.4 TGA and SEM/EDS analysis of (C2H8N2)yFeSe ground powders. . . . . 126 5.5 Pawley refinements with powder XRD for the structures of (C2H8N2)xFeS and (C2H8N2)yFeSe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128 5.6 Electron diffraction pattern of ethyelenediamine-intercalated FeS . . . . . 129 5.7 Electron diffraction pattern of ethyelenediamine-intercalated FeSe . . . . 130 5.8 Additional electron diffraction patterns of ethyelenediamine-intercalated- FeS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 5.9 Proposed structures for ethylenediamine intercalated FeSe . . . . . . . . 137 5.10 Temperature dependence of magnetic susceptibility of (C2H8N2)xFeS and (C2H8N2)y . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 5.11 Isothermal magnetization of (C2H8N2)xFeS and (C2H8N2)yFeSe . . . . . 139 5.12 Temperature dependence of resistance of pressed pellets of ground pow- ders of C2H8N2)yFeS and (C2H8N2)xFeSe . . . . . . . . . . . . . . . . . 140 6.1 Rietveld refinement results for in-situ synthesis of (LiOH)FeS . . . . . . 148 xii List of Abbreviations 2D two-dimensional AFM antiferromagnetic BCS Bardeen-Cooper-Schrieffer CSD Cambridge Structural Database Ch chalcogenide DSC differential scanning calorimetry EDA ethylenediamine EDS energy dispersive X-ray spectroscopy DFT density functional theory FC field-cooled FM ferromagnetic GGA generalized gradient approximation ICP-AES inductively coupled plasma atomic emission spectroscopy ICSD Inorganic Crystallographic Structural Database MPMS magnetic properties measurement system NCNR NIST Center for Neutron Research NPD neutron powder diffraction PAW projector augmented wave Pn pnictide PPMS physical property measurement system PSD position sensitive detector SEM scanning electron microscope SQUID superconducting quantum interference device Tc Critical Temperature TEM transmission electron microscope TGA thermogravimetric analysis TTMC tetragonal transition metal chalcogenide VASP the Vienna Ab-initio Simulation Package XRD X-ray diffraction ZFC zero field-cooled xiii Chapter 1: Introduction 1.1 Iron-based superconductors - background and motivation The discovery of superconductivity in 1911 by Heike Kamerlingh Onnes was, like many discoveries in science, a serendipitous event.1 Onnes was conducting research on cryogenics in attempts to achieve temperatures in which helium would liquefy. Through the course of these experiment, he measured the resistance of mercury and observed a transition, called the critical temperature (Tc), from finite resistance to a seemingly zero resistance state at 4.2 K. This was the first measurement of the superconducting state of a material. Through continued work, it was determined that superconductivity is not just the observation of a zero resistance state. Magnetic measurements of samples in the superconducting state exhibited perfect diamagnetism as external magnetic flux was com- pletely expelled - this phenomena was dubbed the Meissner effect after its founder.2 Both of these phenomena have possible commercial applications. As to be expected, research in this new field exploded as zero resistance components or wires would completely rev- olutionize the world?s electronic industry and electrical grid. The main hurdle the field has faced since its discovery is that the critical temperature at which most materials be- come superconducting is very low and thus requires expensive and complex cryogenic systems to achieve their respective superconducting states. However, research into syn- 1 thesis of new superconductors, fundamental properties and theory of superconductivity has continued to be topical for the past 100+ years. The early work on superconductors encompassed work on a specific type of super- conductor which would later be called ?conventional" superconductors. The behavior of these materials could be extremely well by the microscopic theory of superconductivity described by Bardeen, Cooper and Schrieffer (BCS theory) which was awarded the No- bel Prize in 1972.3 A formal description and work through of BSC theory can be found in many texts including the seminal textbook by Tinkham.4 The key takeaways from the theory were the following: 1) electrons at the Fermi level under an attractive potential can form bound pairs called Cooper pairs,5 2) external magnetic fields decay exponentially inside of a superconductor,6,7 3) the critical temperature experiences an isotope effect,8,9 and 4) the electron-electron coupling is mediated by phonons. The experimental work leading up to BCS theory and the culmination of a microscopic theory to describe the experimental observations exemplifies the amazing conjunction of experimental and the- oretical work to describe the natural world. One problem which arose was the necessity of a phonon-mediated mechanism to describe superconductivity of these systems. This was assumed to set an upper-limit of critical temperature to the region of 30 - 50 K10 for known systems as the time. The key requirements were that the material had to have a high density of states at the Fermi level (the system must be metallic), a high Debye frequency (lighter elements preferred) and strong electron-phonon coupling. Of course, these latter two are linked in most materials, thus the critical temperature ?limit" of 30 - 50 K was imposed based on the currently known materials. However, it was later fa- mously predicted that metallic hydrogen would be a room temperature superconductor11 2 which means the superconductivity in BCS theory does not have an actual limit - its just limited by the currently available materials. This conventional thinking regarding super- conductors behaving in line with BCS theory was abruptly shook with the discovery of a new class of superconducting materials. In 1986, Bednorz and Muller discovered superconductivity in Ba-La-Cu-O ceramic system at 30 K which kicked off the field of high-temperature or ?unconventional" su- perconductors. They were awarded Nobel Prize in 1987 for their discovery. This work began an incredible amount of research activity around these new materials called the cuprate superconductors.12?15 Two of the most important materials discovered through this research have critical temperatures above the limit of liquid nitrogen which signif- icantly expanded the possibility of integrating these materials into commericialization. This is because liquid nitrogen cyrogenic systems are much cheaper and easier to man- age than liquid helium systems. These two materials are YBa2Cu3O7?x (YBCO) which has a Tc= 93 K16 and the other Bi2Sr2Can?1CunO2n+4+x (BSSCO) with a critical tempera- ture up to 108 K17 depending on which composition is measured. In general, all cuprate superconductors are built around the structural motif of Cu-O planes which consist of a checkerboard lattice of Cu2+ and O2? ions which are separated by different spacers lay- ers. The superconductivity in these systems comes from electrons within CuO2 planes. The main differences between these materials and previous ?conventional" superconduc- tors are: 1) the undoped phases exhibit insulating and antiferromagnetic long range order before doping,18 2) upon both electron or hole-doping antiferromagnetic order is sup- pressed and superconductivity is induced which suggests electronic interactions rather than electron-phonon are responsible for superconductivity emergence, 3) the observed 3 superconducting gap is not isotropic19,20 indicating non s-wave pairing. Although 30+ years have passed since their discovery, the mechanism which causes superconductivity in the cuprate systems is still not fully understood. Despite that, high temperature cuprate superconductors have begun to be integrated into commercial applications21 and still of- fers a hot-bed of research activity. The field of superconductivity was up-ended again with the discovery of another class of superconducting materials called the iron-based superconductors. The discovery of superconductivity in LaO1?xFxFeAs in 2008 by Hosono et al. opened up an entire new field of study on superconducting materials containing iron pnictide or iron chalco- genide layers.22 Historically, magnetism and superconductivity were considered antago- nistic, which caused the excitement around the possibility of superconductors built out of the prototypical ferromagnetic element to explode. Ever since, there has been a fury of work done by chemists and physicists alike attempting to understand the mechanism(s) for superconductivity in this family of compounds through guided synthesis of analo- gous compounds offering different probes into the physical system.23?27 The physics of the iron-based pnictide superconducting system is very similar to those of the cuprate superconductors whereby a high temperature antiferromagnetic state is suppressed upon doping to reveal superconductivity.27,28 However, there are a number of key difference be- tween the iron-based pnictide systems and the cuprates: 1) the undoped parent compounds are magnetic metals, 2) electronic correlations within the iron layers are moderate29,30 in comparison to the strong correlation in the cuprate systems,31 3) the electronic structure in the cuprates is single band32,33 and multi-band in the iron-based systems.34?36 The com- bination of all these has made iron-based superconductors one of the most widely studied 4 topics in condensed matter physics in the past 10+ years. The multiband nature of iron-based superconductors and the corresponding physics has been met with an abundance of theoretical work to describe the system. This has led to the coinage of the term ?Hund metal" to describe the electronics structure and corre- lations in the system.37 Unlike the single band nature of the cuprates and the description of Mott insulator physics leading to superconductivity,31?33 the multiband nature of the iron-based system requires a new description. This ?Hund metal" description is charac- terized by enhanced electron correlations, high spin configuration, and orbital character selectivity.38,39 In summary, the correlations arise from local Coulomb interactions and Hund?s rules which allows another paradigm to access strong correlations in condensed matter systems. Hund?s rules is the energy scale which dictates intra-atomic exchange (i.e. lowering the cost of placing two electrons in different orbitals as opposed to in the same orbital). Thus, from theoretical approaches, it becomes clear the when the energy scale of Hund?s rules becomes meaningful, electronic correlations can become significant and give rise to very exotic physics, such as high temperature superconductivity.40,41 The iron-based and the cuprate superconducting systems both exhibit a similar structural motif whereby two-dimensional lattices are stacked to form layered materials. In the iron pnictide and chalcogenide systems, square iron layers are the basic building blocks of the iron-based superconductors which stack to form more complicated layered structures as shown in Figure 1.1. The largest structural difference between the pnictide and chalcogenide families is that the pnictide family is built of anionic (FePn)? (Pn = pnictide) layers whereas the chalcogenides are built from neutral FeCh (Ch = chalco- genide) layers. This means that the pnictide family requires cationic interlayer species to 5 KFe Se LiOHFeSeFeSe 2 2 (a) Tc = 8 K (b) Tc = 30 K (c) Tc = 43 K Fig. 1.1: Crystal structures of FeSe (left) and KFe2Se2 (center), and LiOHFeSe (right), and their respective critical temperatures, which all demonstrate the stacking of tetrahedral FeSe layers as the basic building block of the iron-based chalcogenide superconductors. stabilize the structure, but the chalcogenide family can be made as a simple binary and ex- panded from that simple building block. The most structurally simple iron chalcogenide superconductor is FeSe which crystallizes in the anti-PbO structure (P4/nmm, Figure 1.1, left) and exhibits superconductivity at 8 K.42?45 The Tc of FeSe can be improved signif- icantly to 42-46 K from 8 K after intercalation,46?48 37 K with applied pressure49 or 65 K in the monolayer limit.50 Our group?s research has focused on the chalcogenide family for two important reasons. To start, the neutral building block of FeSe and FeS layers is a perfect candidate for chemical manipulation through intercalation and doping due to its two-dimensional van der Waals structure and neutral charge. Also, unlike the related iron pnictide phases, FeSe is a superconductor without required doping and exhibits no parent magnetic phase.42,51 Although no parent magnetically ordered phase exists, strong mag- netic fluctuations have been observed in a wide range of temperatures in FeSe through neutron and NMR spectroscopy experiments.51?54 6 The use of FeSe as a building block (and the underlying structure-property rela- tionships associated with the FeSe building block) and its subsequent chemical manipu- lation is the dominant motivation for the work described in this thesis. FeSe is the only tetrahedral transition metal chalcogenide material (TTMC), with the anti-PbO structure, that is known to be thermodynamically stable and can be synthesized through a number of methods, such as: direct elemental reaction and chemical vapor transport.55?57 The FeS analogue is metastable and cannot be made through a traditional high temperature solid-state method. However, in 2015 a new hydrothermal synthetic scheme was devised to form FeS and found superconductivity at 4 K in powder samples.58 More recently, another reaction scheme was devised to form single crystals of tetragonal FeS through topochemical de-intercalation of KxFe2?yS2, which supported the previous work with FeS formed through this method exhibiting superconductivity at 4 K.59 Thus, my work at- tempts to expand the class of TTMC materials beyond iron to understand how the TTMC structure leads to different properties for different electron filling and expand the chemical manipulations possible on the FeSe and FeS building blocks to tune the superconducting and magnetic properties of the iron-based chalcogenide superconducting family through topotactic synthetic routes. 1.2 Tetrahedral transition metal chalcogenides The study of iron-based chalcogenide superconductors is motivated largely by their structure which is built upon neutral FeCh layers with a square transition metal sublattice; it is this fundamental structure that gives rise to the exotic physical phenomena observed 7 in these materials. Thus, it is useful to take a deeper look from a structural point of view to understand the nuances of the structure and bonding in this class of materials. the fundamental unit of tetrahedral transition metal chalcogenides (TTMCs) are edge-sharing MCh4 (M = Fe, Co, Ni, Cu, Zn; Ch = S, Se, Te) tetrahedron which condense together in the plane to form extended two-dimensional layers of the MCh species. In the simplest case, these two-dimensional layers are then stacked together and held through van der Waals forces to form the anti-PbO type structure; to date only the iron and cobalt versions of these binaries have been synthesized.60?62 This main building block highlights the most important features of this structure: 1) tetrahedral chalcogenide bonding of the transition metal, 2) square transition metal sublattice, and 3) neutral two-dimensional layers held together by van der Waals forces. A representative crystal structure illustrating the stacked layers held together by van der Waals forces, square metal sublattice, and tetrahedral coordination is shown in Figure 1.2. There are many different structures which use this building block or similar building blocks which can fall into the category of TTMCs which are cataloged in the review work by Zhou and Rodriguez,63 but the remainder of the discussion will focus on the simple binary building block. In a cubic close-packing arrangement of chalcogenide ions, there are tetrahedral and octahedral interstitial holes which can hold cations. In general, octahedral sites require the cation radius to be > 41% of the anion radius (rc/ra = 0.414) and tetrahedral sites requires the cation radius to be between 41% - 23% of the anion radius (rc/ra = 0.225). If we consider the sulfide S2? anion which has an ionic radius of 1.84 ?, this means that cations with ionic radii between 0.75 - 0.42 ? may be able to order within the tetrahedral interstices. This means that larger transition metal cations are unlikely to be incorporated 8 Fig. 1.2: Representative crystal structure of tetrahedral transition metal chalcogenides illustrating the stacked layers held together by van der Waals forces, square metal sublattice, and tetrahedral coordination adapted from Rodriguez.63 into this crystal structure. Another consideration must be taken into account as well due to the 2- charge on the sulfide ion. In order to maintain a neutral layer, which is key for our desired building block, a cationic charge is needed to yield a neutral metal chalcogenide layer. Since the MCh layers are only two layers of close-packed chalcogenide anions in the (AB)-(AB) sequence (the third layer C is empty in this structure), only one tetrahedral interstice exists which requires transition metal to be in a low valent 2+ cationic state. Thus, we can assume that low-valent transition metals with small cationic radii in the case of tetrahedral coordination will preferentially form this structural building block. 9 The tetrahedral bonding also leads to other benefits with regards to the electronic structure of the MCh layers themselves. The bonding between the transition metal and chalcogenide is covalent in nature and its impact is fully discussed in a some seminal works by Roald Hoffmann and more recently by our group.63?65 The translation of the chemical bond?s impact to electronic structure is very difficult as one is a real space con- cept while the other is best viewed in reciprocal space; fortunately two very important works have helped bridge that gap in thinking.66,67 Another key feature, is that tetrahedral interstitial sites are closer together than octahedral sites within the cubic close packing array which means that interactions between the cations can be stronger and metal-metal bonding can be significant. This metal-metal bonding leads to a delocalized or wide-band description for the orbitals with character from the metal cations. Since the bands at the Fermi level are primarily metallic in character in these compounds,63 the metal-metal in- teraction plays a significant role in determining the electronic structure and properties in these compounds. Beyond the tetrahedral bonding of the transition metal, the square metallic sublat- tice itself is one of the main features of the TTMC class of materials. As previously stated, the metal-metal interactions caused by the proximity of the tetrahedral interstices in the cubic close packing of chalcogenide anions leads to bands with metallic charac- ter becoming delocalized and dispersed. The square metallic sublattice has been shown through a variety of works to be susceptible to electronic instabilities depending on elec- tron filling.68,69 These instabilities drive exotic physics and observable properties from charge density waves to superconductivity. Up until now, we have only considered the electronic effects of the transition metal sublattice, however; since the low-valent late 10 transition metals have unfilled d-orbitals, the effects of magnetism must be considered as well. To that end, there has been a significant amount of theoretical work on describing the magnetic interaction of the square lattice for configurations that would lead to mag- netic frustration and possible exotic manetic ground states. This work has been done for local moment systems70?74 and needs to be expanded to consider itinerant systems which includes metallic character with delocalized moments. Finally, the last important piece of structural makeup of the TTMC class of ma- terials is that the neutral MCh layers are held together by van der Waals forces which makes these materials perfect candidates for chemical manipulation. With FeSe and FeS displaying superconductivity at 8 K42 and 4 K,58 respectively, there has been substantial research on intercalation chemistry within the system to probe how intercalated species will affect the physical properties. This is because intercalating interlayer species can achieve two main goals: 1) increased interlayer spacing between adjacent MCh layers leads to increasingly two-dimensional electronic structure and 2) interlayer species can charge dope the MCh layers to tune the electron filling. Although the cobalt analogues to FeCh have been synthesized,61,75 reported intercalation chemistry has been limited due to the phase stability of those compounds. Currently the list of successful intercalated species is fairly extensive including simple alkali metal cations,76?78 which are typically able to be synthesized from a traditional solid state route such as self-flux or salt flux re- actions. Other intercalates can be partially charged hydroxide or amine layers46,79?86 that typically require low temperature or chimie douce techniques to synthesize. After these different intercalation schemes, the Tc of FeSe can be improved significantly to 43-46 K from 8 K. Although both hydroxide or amine intercalation can increase the Tc of FeSe, 11 only LiOH intercalation is known to increase the Tc of FeS from 4 to 8 K.87 Almost all of these intercalation chemistry methods are done at low temperatures and exploit the stability of chalcogenide anions in basic solution as well as the ability of basic aqueous and other basic solutions to oxidize iron metal to form soluble iron cations. As a note of caution: this chemistry must be done in basic conditions as acidic condi- tions may give rise to the evolution of toxic hydrogen chalcogenide gases. In any case, the means by which low temperature reaction schemes have been employed to modify TTMCs is the focus second main focus of this thesis work. 1.3 Topotactic chemistry and hydrothermal synthesis The field of solid-state chemistry and physics is dominated by materials and syn- thetic routes which employ high temperatures in order to stabilize the formation of ther- modynamically stable phases. This is for good reason, high temperatures typically al- lows for the formation of single crystals through dissolution of constituents in some flux and precipitation slowly through very precise temperature control. Single crystals are essential for in-depth physical property characterizations including resistivity, magnetic anisotropy, crystal structure determination, etc. Beyond that, traditional solid-state syn- thetic routes are often dubbed (often times with a negative connotation) ?shake-and-bake" reactions for their ease of setup and lack of sophistication. In these reactions, mechanisms are typically not considered and there is often no control over the formation of interme- diates. From a solid-state chemist?s point of view, this is in stark contrast to the organic chemist?s approach to the synthesis of a new compound or molecule. There, every step 12 of the reaction is considered and intermediates of complex reactions can be isolated and changed throughout the synthetic scheme to guide the reactions in new ways. The ex- tension of a similar, albeit less dramatic, approach to the synthesis of solid-state mate- rials offers a window into the synthesis of a wide-range of new compounds. Here, we consider the formation of metastable phases which are typically synthesized at low tem- perature as to limit the effects of thermodynamics and trap the desired metastable phases kinetically. Metastable phases are niche but are widely used for a variety of important applications.88?91 As mentioned, the stabilization of metastable phases is often done at low temper- atures in order to avoid the effect of thermodynamics which will dictate the formation of the lowest energy phase if the thermodynamic energy barrier is overcome. However, temperature control is not the only tool we have to employ, we can also consider other chemical factors such as reaction time, pressure, concentrations, redox chemistry, etc. Thus, thinking about the synthesis of metastable phases can be approached by two ways: 1) changing the thermodynamic landscape of the chemical reaction by adjusting the chem- ical environment such as solvent, pH etc, or 2) using thermodynamically stable phases as a starting point to convert to metastable phases. As imagined, the approach of 1) can be extremely difficult and time-intensive as the number of variables which can be considered and adjusted in a reaction scheme is very large. This is not to say it can?t be done, a great example of this is the synthesis of metastable tetragonal FeS, mackinawite, which can be done through a number of methods.92?94 One caveat of this example is that mackinawite was known to exist from geological mineral samples, so a synthetic route could be ra- tionalized from Mother Nature. In its mineral form, mackinawite is typically stabilized 13 by nickel impurities on the iron site and found in low temperature aqueous environments, such as marine sediments. However, in its synthetic form FeS is stabilized by a low tem- perature kinetically controlled route as to avoid the formation of more stable phases like gregite and pyrite.92?94Although mackinawite cannot be synthesized by direct solid-state reaction, it can be afforded using hydrothermal reaction under basic conditions. The com- pound itself converts to the more stable hexagonal phase when heated above 100 ?C, but is stabilized above that temperature with high pH. This demonstrates that the thermody- namic landscape can be changed by tuning the synthetic environment. However, for the synthesis of materials which do not exists in nature, a successful synthetic route can, and often times is, be much more difficult to determine. This brings us to 2), the conversion of one thermodynamically stable product to a metastable state through some chemical conversion method. This has been dubbed topotactic chemistry and is defined as a reaction scheme which retains some structural building blocks of the reactant in the product. Topotactic chemistry is essential in intercalation/de- intercalation/ion-exchange routes where a guest species is added, removed, or exchanged between host layers in order to tune the structure and physical properties. These reac- tion schemes often require low temperature and are often called chimie douce or ?soft chemistry" routes in literature.89,90,95?97 They are dubbed so as they offer a contrast to the high temperature approaches of traditional solid-state chemists. The field is not new and chimie douce reaction schemes have been used extensively in intercalation chemistry in the past 40 - 50 years to great effect.98?103 With that being said, it is still considered a niche field as it straddles the fence between the worlds of the organic chemist and that of the solid-state chemist. The future outlook of topotactic chemistry and chimie douce 14 methods is very high especially with the current emphasis on rational material design as this synthetic paradigm allows for a guided design of extended solids - much like the thinking from the organic chemistry world. When considering topotactic chemistry and chimie douce methods, the applica- tion to the work motivated in this thesis is readily apparent. As materials or solid-state chemists/physicists, we are intimately concerned with structure-property relationships - how does the crystal structure and bonding directly affect the manifestation of physical properties? The easiest example to consider is that of graphite and diamond, both made up of carbon but each have incredibly different properties due to structure and bonding within each material. We have already emphasized how TTMCs exhibit a structural motif for which we are interested: tetrahedral chalcogenide bonding to a transition metal, square transition metal sublattice, and van der Waals layers. Thus, this topotactic or chimie douce paradigm is an excellent lens to approach this problem. We are directly targeting one spe- cific building blocks, and as such, we can use topotactic reaction schemes to change the overall structure without affecting that building block. All three chapters of this thesis utilize topotactic or chimie douce methods in order to synthesize new materials. The chemistry of topotactic schemes can vary depending on the building blocks which are meant to be retained. For oxides and chalcogenides, a main tool for topotac- tic chemistry and chimie douce methods is the hydrothermal technique. Low tempera- ture topotactic orchimie douce hydrothermal techniques can be used to synthesis a wide range of materials.89,90,104?109 Hydrothermal reactions are used to mimic Nature?s own hy- drothermal reactions which exists at hydrothermal vents where typically insoluble min- erals are heated in water by volcanic activity to dissolve before being released into sur- 15 rounding cold waters to form new species.110,111 Due to the amazing chemistry present at these natural sites, there have also been intensely studied for their possible role in the origin of life.112,113 In this research, hydrothermal conditions are defined as any reaction carried out above 100 ?C and above 1 atm of pressure. Typical growths and reactions require the use of hydrothermal autoclaves which are used to maintain high temperature and pressure for the reaction media. The theory of hydrothermal growth, the growth methods, equipment, growth pa- rameters, etc are cataloged very well in the following sources;114?117 we will focus on a brief discussion of how hydrothermal synthesis is done in this research and its relationship to topotactic chemistry and TTMCs. The low temperatures often used in hydrothermal reactions fits perfectly into the paradigm of topotactic chemistry for the synthesis of new TTMC materials. To reiterate, the low temperature allows for the stabilization of kinetic phases as thermodynamic effects are limited and phases for which kinetic pathways exists can be trapped and isolated. Another benefit of hydrothermal reactions is that the redox chemistry which occurs in water is highly desirable for chalcogenides, especially for iron chalcogenides. In particular, this is true for highly alkaline hydrothermal reactions. To that end, we can consult a Pourbaix diagram which shows which species will exist in solution as a function of pH and electrochemical potential. Pourbaix diagrams for any system are well tabulated but can be calculated from electrochemical potentials for all species which should exist in aqueous solution. An example Pourbaix diagram for the speciation of Fe in H2O is shown in Figure 1.3. From these Pourbaix diagrams, it is discovered that under reducing conditions at high pH, chalcogenide ions can be stabilized as HCh? species which can act as the anion 16 Fig. 1.3: Pourbaix diagram of iron calculated at standard temperature and pressure adapted from Channei.118 building block for the desired building block. For the cation, iron, the Pourbaix diagrams shows that iron oxidizes in reducing alkaline solution to yield Fe2+ species in solution. The combination of these two species in solution under reducing and high pH conditions has been shown to cause the formation of Fe-HCh? clusters in aqueous and non-aqueous solutions with tetrahedral Fe-Ch coordination119,120 which likely condense to form the extended layers in iron-based TTMCs. The special consideration in these reactions is that under anaerobic hydrothermal conditions, water plays the role of solvent and cathode for electrochemical reactions. In this case, we can consider the simple electrochemical reaction whereby iron powder is oxidized: 17 Fe(s) + 2 OH?(aq) ??????? Fe(OH)2(s) + 2 e? E = 0.89 V (1.1) 2 H2O(l) + 2 e ? ??????? H2(g) + 2 OH?(aq) E = ?0.828 V (1.2) Fe(s) + 2 H2O(l) ??????? Fe(OH)2(s) + H2(g) E = 0.062 V (1.3) This formulation shows that iron is oxidized in solution to form Fe2+ species with a small but positive driving force which can be increased as a function of temperature. This Fe2+ species then reacts with SH? or SeH? in solution to form the Fe-HCh? clusters as the starting point for extended layers of iron chalcogenides. However, in the case where the SH? or SeH? species is not in solution, the Fe2+ species will all continue to oxidize to Fe3+ species.121 Thus, we can push the reaction toward the formation of new prod- ucts by utilizing the formation of the intermediate oxidized species. This harkens back to point 1) of how to stabilize metastable phases by changing the thermodynamic landscape of the chemical reaction by adjusting the chemical environment such as solvent, pH etc. The electrochemical considerations show that intermediate species can form due to redox potentials which can then be converted. As before, this is assuming the effects of thermo- dynamics are limited at low temperatures as thermodynamics will decide which phases forms at higher temperatures. Although this thesis focuses mainly on the topotactic ap- proach to the formation of metastable phases, this other view point offers to bring another platform for future exploration. In particular, the exploitation of the associated hydrother- mal redox chemistry with regards to the formation of desirable intermediate species that can be converted and trapped as metastable phases is underexplored. 18 1.4 Objectives and Outline The previous sections have helped illustrate the motivation leading up to this thesis work. This work is highly interdisciplinary and lies directly in between the realms of solid state chemistry and condensed matter physics. In this thesis, we describe the synthesis and characterization of three different topotactically modified transition metal chalcogenide systems and their corresponding physical properties. In Chapter 2, we discuss the synthetic techniques and characterization techniques employed throughout this work. In order to fully characterize crystal structure, elemental composition and physical/chemical properties, many different instruments and techniques were utilized. Each Chapter 3-5 focuses on a different family of materials. In Chapter 3, we discuss the synthesis and characterization of tetragonal CoSe which is isostructural to su- perconducting FeSe. This material requires a topotactic de-intercalation synthetic route and we can successfully synthesize powders and single crystals. Physical property mea- surements displays metallic spin glass type behavior below 10 K which we attribute to magnetic frustration due to competing interactions of the square metal sublattice. Chapter 4 focuses on the synthesis of transition metal doped (LiOH)FeSe. In previ- ous work, (LiOH)FeSe was shown to display magnetic order below the superconducting critical temperature. We have used transition metal doping and a topotactic hydrothermal synthetic route to increase the magnetic signal of the system and have been able to observe the formation of long range magnetic order in nominally 20% Mn-doped (LiOH)FeSe. The majority of the work focuses on the characterization of this nominally 20% Mn- 19 doped sample family, however; we are able to synthesize other transition metal doped samples from Mn - Zn with varying results due to the chemistry and properties of the other transition metals. In Chapter 5, we focus on synthesis and characterization of an ethylenediamine- intercalated iron chalcogenide family. This work echoes other work in the field on similar intercalation with one large difference. Other reaction schemes requires the addition of an alkali metal to the ethylenediamine neat solvent in order to stabilize the formation of the intercalated species. This has the detrimental effect of making the true doping level achieved by intercalation hard to identify due to the formation of other charged inter- calated species with the alkali metal. Thus, we have developed a hydrothermal method which does not require the addition of an alkali metal to stabilize the intercalation of ethylenediamine into FeSe and FeS. This work shows that the hydrothermal method may be a more universal means to intercalate other adducts as it does not require liquid or low melting temperature adducts or the co-intercalation of alkali metals. Chapter 6 summarizes the main findings of this work and discusses the future di- rections related to the work which can be explored. 20 Chapter 2: Methods 2.1 Synthetic Methods The synthetic methods discussed here are more general with regards to the synthe- sis of a wide range of layered transition metal chalcogenides. More in-depth synthetic methods and discussion of those routes are cataloged in their respective chapters. 2.1.1 Self-flux synthesis of ThCr2Si2-type transition metal chalco- genides The synthesis of single crystals of new intercalated and de-intercalated transition metal chalcogenides starts with the self-flux synthesis of the ThCr2Si2-type KxM2Ch2 (M = Fe, Co, Ni; Ch = S, Se) single crystals. The main benefit of this approach is that many different compositions of the ThCr2Si2 exist even beyond chalcogenides and these phases can be usually be formed as relatively large single crystals. These single crystals offer a great base for post synthetic modification to form new layered transition metal chalcogenides. A typical growth of KxM2Ch2 (M = Fe, Co, Ni;Ch = S, Se) single crystals uses approximately 0.25 g K chunks and the required molar ratios of metal powder and chalco- 21 Fig. 2.1: As recovered single crystals of KCo2Se2 (left), KNi2Se2 (right) grown via the double ampoule method. genide powder is calculated from the measured potassium amount. The amount of potas- sium has to be limited as potassium vapors attack the quartz tube which can cause them to break during the growth process. In general, this can be avoided if reaction vessels which are inert to alkali metal vapors such as niobium or tantalum tubes are used. The potas- sium, metal powders, and chalcogenide powders are loaded in a quartz ampoule within a inert atmosphere glovebox. As a note of safety, the reaction of potassium with chalco- genides can be extremely exothermic so these growths are usually done so that the loaded potassium does not come in contact with the chalcogenide powders. One way around this issue is to pre-react the transition metal and chalcogenide to form a stable binary; how- ever, we have found that better quality single crystals are formed through a reaction of direct elements. Once the reaction mixture has been removed from the glovebox, it is sealed under vacuum using a natural gas flame torch at a typical vacuum of ? 10?3 Torr. In the case of direct element reactions, the reaction mixture is usually frozen in liquid nitrogen before sealing to ensure that any possible reaction between the potassium and chalcogenides is limited. Once sealed, the smaller quartz ampoule is loaded inside a larger quartz ampoule and vacuum sealed again. This is done so that if the inner quartz ampoule breaks due to 22 potassium vapor attack or other factors, the reaction mixture is not fully compromised; inner quartz tube failure occurs in about 20% of growths. These double-sealed quartz ampoule reaction mixtures are loaded into a box furnace, heated to 1050 ?C at a rate of 10 ?C/hr and held at 1050 ?C for 12 hours to ensure reaction homogeneity. The very slow heating is done to ensure the melting of potassium occurs slowly to react with the chalcogenide to form a potassium polychalcogenide flux. As the reaction slowly heats the potassium polychalcogenide flux react with the transition metal to form the KxM2Ch2 (M = Fe, Co, Ni; Ch = S, Se) product as the potassium polychalcogenide flux oxidizes the transition metal. The reaction mixture is slow cooled from 1050 ?C to 550 ?C at a rate of 2 ?C/hr which assures slow crystallization. Once 550 ?C is reached, the reaction mixture is cooled the room temperature naturally by turning off the furnace. These crystals are typically recovered in air but is done in an inert gas glovebox when required. Crystals are typically 5 x 5 x 0.5 mm platelike shiny crystals of different colors depending on what transition metal was used in the reaction mixture shown in Figure 2.1. The crystals are usually air stable on the order of a day or so but are stored in an inert gas atmosphere for long term storage. 2.1.2 Soft chemical reductive de-intercalation The class of materials of tetrahedral transition metal chalcogenides with the de- scribed layered structure was previously limited to the FeSe. This is because the lay- ered structure is not thermodynamically stable for other combinations of transition metal 23 K1-xCo2Ch2 CoCh K+,, H2(g) LiOH (aq) RT, t = 24 hours Fig. 2.2: Soft chemical reductive de-intercalation at room temperature in saturated LiOH solution of KCo2Ch2 single crystals and powders to form CoCh powders and single crystals. and chalcogenides, even for FeS. Previous work in our group showed that single crys- tals of FeS would be synthesized by hydrothermal reductive de-intercalation of KFe2S2 single crystals to form FeS single crystals.59 Similar work had also been attempted to de- intercalate the KNi2Se2 compound to form the layered NiSe binary, isotructural to FeSe, but was not successful.122 My work has expanded the available TTMC?s beyond FeSe and FeS to the isostructural cobalt analogues.61 Two methods are used to form layered tetragonal CoCh (Ch = S and Se) compounds shown in Figure 2.2. KCo2Ch2 powders and/or single crystals were placed in approxi- mately 10 mL of saturated LiOH solution made by dissolving LiOH?H2O in de-ionized water. The reaction mixture was purged with argon and placed in an ultrasonic bath for an hour. The reaction mixture was then recovered via centrifuge and washed up to five times with water and ethanol which yielded black shiny powders of CoCh. Another method was employed to ensure starting crystals of KCo2Se2 were not destroyed in the ultrasonic method; at the current time single crystals of KCo2S2 could not be made. Single crystals 24 of KCo2Se2 were placed in 20 mL saturated LiOH solution in a round-bottom flask on a Schlenk line under inert gas flow for 1 day. Shiny silver crystals were recovered after 1 day after water and ethanol washes and dried under vacuum. In these reaction scheme one important note of safety must be acknowledged. Highly basic solutions must be used for chalcogenide reactions in water in order to prevent the formation of highly toxic H2S and H2Se gases. This room temperature reductive de-intercalation was found to work for the cobalt analogue whereas the iron versions, FeSe and FeS, requires hydrothermal conditions to achieve the de-intercalation. Preliminary attempts with the cobalt system under hy- drothermal conditions gave different polymorphs depending on the temperature used.61 Some work has been done to employ room temperature and hydrothermal schemes to the KNi2Ch2 system to form NiSe, isostructural to FeSe, with no success. It has been found that room temperature reactions in saturated basic solutions leave the starting KNi2Ch2 unreacted and higher temperature achieved in hydrothermal reactions give different NiCh polymorphs. The expansion of soft chemical techniques to expand the TTMC class of materials is promising avenure for future research. 2.1.3 Hydrothermal ion exchange Much like the soft chemical reductive de-intercalation reaction scheme, we can also expand the synthetic routes to include ion exchange starting with the KxM2Ch2 (M = Fe, Co, Ni; Ch = S, Se) single crystals from the self-flux synthesis. As a note, for some desired products, it is possible to complete a two-step reaction whereby the KxM2Ch2 25 Fig. 2.3: Possible hydrothermal cation exchange routes for iron chalcogenides which includes the intercalation of extended hydroxide layers and various cations.123 crystals are de-intercalated and then intercalated with another species, but this does not work for all desired intercalants. A summary of possible reaction schemes is shown in Figure 2.3 With regards to our research, the predominate use of this reaction scheme is for the synthesis of single crystals of the (Li1?xFexOH)FeSe. In this compound, FeSe layers are intercalated with an extended hydroxide layer of LiOH. It has been found through multi- ple works that some of the Li in the LiOH layers is substituted with Fe which gives the LiOH layer a partial positive charge. This partial positive charge helps stabilize the inter- 26 calation as well as charge doping the FeSe layers - increasing the superconducting critical temperature to 45 K from 8 K in FeSe.48,79,82,124 This compound can also be made from a ?bottom-up" route whereby Fe metal and a selenium source are added to a hydrothermal autoclave with excess LiOH to build the FeSe layers in the reaction although this reaction scheme only yields powders. To remedy this, we have developed a hydrothermal cation exchange method to start with single crystals of KxM2Ch2 and exchange the interlayer potassium ions for LiOH. The extreme benefit of this reaction scheme is that single crystals of (Li1?xFexOH)FeSe are afforded which enables much for expansive physical property characterization. In a typical synthesis, selected single crystals of KxM2Ch2 are added to 15 mmols Fe gran- ules/powders, 1.2 mmol selenourea, 1.2 mmol tin granules, 0.30 moles LiOH, and 40 mL H2O or D2O (required for neutron experiments). All components were added to a Teflon cup and the head space was purged with argon for 2 minutes before being tightly sealed. The autoclave was placed into an oven for 72 hrs at 120 ?C before being recovered in air and dried under vacuum. The addition of additional Fe granules/powders and selenourea in the reaction scheme is to ensure the product is not selenium or iron deficient. Often time, the starting KxM2Ch2 crystals are iron deficient;125 thus the addition of extra Fe ions in solution acts to replace vacancies as superconductivity in these compound is very dependent on stoichiometry. Single crystals produced from the self-flux method KxM2Ch2 are frequently on the size 3 x 3 mm2 or larger and the thickness can be controlled by cleaving from the growth boule. After the hydrothermal cation exchange it is often noted that the starting crystals will have flaked apart to form much thinner crystals. This method is used heavily for the 27 production of very high quality (Li1?xFexOH)FeSe single crystals, but can be expanded for the intercalation of other species are future work. 2.1.4 Alkali metal-free ethylenediamine intercalation of iron chalco- genides The previously mentioned ?bottom-up" synthetic route has the downside of only producing powders, but the large benefit is that the ability to synthesis new compounds is more robust. This is attributed to the fact that instead of actively having to exchange one cation for another, as in the hydrothermal cation exchange method, FeCh layers are built in ? situ from Fe and a chalcogenide source which form around the desired intercalated species. This allows the intercalation of much larger species between the FeCh layers. One such growth is the hydrothermal intercalation of FeCh with ethylenediamine. This reaction can be done starting with pre-reacted tetragonal FeCh, but that reaction scheme must be done solvothermally which limits desired intercalate adducts to ones that have melting points in the low temperature regime. Thus, we have developed a hydrothermal route whereby the desired adduct, ethylenediamine in this case, is mixed with water and other constituents to form the intercalated species. Again, the benefit of this scheme is its generality, where by many adducts which are miscible or soluble in water can be used as possible intercalants. For a typical growth, 4 mmol of Fe powder, 10 mmol thiourea or 5 mmol sele- nourea, 7 mmols KOH, 10 mL H2O and 3 mL ethylenediamine were added to a Teflon cup, purged with argon for 2-5 minutes and sealed in a hydrothermal autoclave. The reac- 28 tion mixture was placed in a convection oven for 2- 6 days at 120 - 160 ?C before recovery via centrifuge in air and drying under vacuum. The recovery of this reaction yielded shiny black powders and excess unreacted Fe granules or powders could easily be removed via magnet. The benefits of this reaction scheme are twofold: 1) the in ? situ formation of FeCh layers should allow for a wider range of intercalated species and is the only way to synthesize the intercalated species of the metastable sulfide analogue, 2) the addition of water allows for a wider range of intercalated adducts as long as the adduct is miscible or soluble in water. The extension of this work to include other adducts is a promising avenue for future work. 2.2 Characterization Methods Synthesis of new compounds takes the bulk of one?s research time as it can take many, many trials to achieve success in forming new compounds. However, the most important steps come next in the characterization of the structure and properties of the synthesized compounds. This section details the numerous methods used in this the- sis for structure determination, elemental composition, physical property measurements, and theoretical calculations. Each section catalogues typical usage and important notes. Technical and theoretical references are added to each section for more in-depth reading as required. 29 2.2.1 Laboratory powder x-ray diffraction The X-ray Crystallographic Center at the University of Maryland is extremely well- equipped for research activity and Dr. Peter Zavalij has been incredibly helpful in discus- sion and helping conduct various x-ray experiments. As recovered samples were ground in a mortar and pestle for approximately 10 minutes until the sample crystallite size was uniform and grinding yielded smooth motion. Since our growths are usually small in recoverable product, the ground powders were typically mounted on a very small or flat sample plate and pressed down with a glass slide to ensure the surface as flat. Powder x-ray diffraction measurements for quick phase identification were done on a Bruker D8 X-ray diffractometer with Cu K? radiation (? = 1.5406?) in Bragg-Brentano geometry from 5? - 90? 2? with a step-size of 0.020? and total counting time of 20 minutes. Ex- tended measurements for Rietveld analysis were altered to a step-size of 0.010? and a total counting time of 1 - 3 hours depending on how much powder was available for the measurement. All crystallographic analysis was done using the Bruker software suite. Bruker Diffrac.EVA was used for analyzing collected diffraction patterns and often times for quick search-and-match phase identifications to quickly identify all phase (products and impurities) in a reaction mixture. Bruker Diffrac.Suite TOPAS was used for all Lebail/Pawley and Rietveld refinements.126,127 Laboratory x-ray diffraction and subsequent Rietveld re- finement was usually paired with synchrotron x-ray diffraction, neutron powder diffrac- tion, electron diffraction and/or other elemental analysis techniques to fully understand the crystal structure and chemical composition. 30 The most frequently used references for x-ray diffraction and Rietveld refinements are collected here.128 The website curated by John Evans at Durham University is tremen- dous is explaining usage of TOPAS for all types of x-ray and neutron refinements, peak in- dexing, and more complex techniques (http://community.dur.ac.uk/john.evans/ topas_academic/topas_main.htm). Crystallographic databases were often time ex- tremely useful to identify new compounds via similar structural motifs and/or by match- ing catalogued data - the ones most frequently used for the Inorganic Crystal Structure Database (ICSD) (https://icsd.fiz-karlsruhe.de/index.xhtml;jsessionid= AA35F9A5641C059D4D0AA0C29791ACC2) and the Cambridge Structural Database (CSD) (https://www.ccdc.cam.ac.uk/structures/). The ICSD is most widely used in in- organic research, but the CSD adds the ability to search for organic, organometallic, and organic-inorganic hybrid compounds. 2.2.2 Synchrotron powder x-ray diffraction In many cases, we used synchrotron x-ray diffraction techniques to supplemental in-house laboratory x-ray diffraction. For work in this thesis, the mail-in program at 11- BM at the Argonne National Laboratory Advanced Photon Source was used exclusively, however; our group has done a series of experiments at 17-BM at Argonne National Lab- oratory Advanced Photon Source for in? situ x-ray diffraction of hydrothermal reactions. The mail-in program offers access to the 11-BM instrument which offers the highest res- olution powder diffraction instrumentation in the Americas (< 1.4 ? 10?4 ?Q/Q) with extremely fast collection times (< 1 hour collection). The constant wavelength from the 31 synchrotron beam is determined by the beamline cycle, typically between 15 - 30 keV. Diffraction data is collected from 1? to 45? 2? with a step size of 0.001? and total collec- tion time of about 1 hour. Synchrotron x-ray diffraction allows for higher resolution x-ray diffraction patterns which aids in the analysis of crystallographic structure and elemental composition. This is especially true in complex systems with complicated crystal structures or high amounts of disorder with regards to elemental composition. The wesbite for 11-BM is quite useful in describing how to obtain mail-in time as well as user guidance for sample preparation and experimental questions (https://11bm.xray.aps.anl.gov/users.html). 2.2.3 Neutron powder x-ray diffraction Neutron powder diffraction is a very powerful complementary technique to x-ray powder diffraction for two main reasons. To start, neutrons are used instead of x-rays which means since neutrons have a magnetic moment, the magnetic order of your sample may be determined as well as the crystallographic structure. Secondly, the x-ray scattering form factor scales with number of electrons so elements close to each other on the periodic table are hard to identify whereas neutron scattering is determined by the nucleus of the scattering atom which may be appreciably different for nearby elements and especially useful for light elements. One important note about neutron diffraction is that neutron flux is much less than typical x-ray flux. For example, a neutron flux of 1.1 ?106 n/s/cm2 is achievable at BT- 1 at the NIST Center for Neutron Research (NCNR) whereas the flux on the sample 32 at a synchrotron x-ray source 5 ?1011 p/s/mm2, many orders of magnitude larger. For neutron diffraction experiments, this means two things: 1) more sample is often required for neutron powder diffraction (> 1g pure sample) and 2) longer collection times are required to obtain desirable statistics. The research cataloged here has made use of three different instruments for neu- tron powder diffraction. The main instrument used was BT-1 and the NCNR with help and collaboration with Dr. Craig Brown. BT-1 is a instrument designed specifically for high-resolution neutron powder diffraction. Ground powder samples were loaded in vanadium cans in a helium glovebox and sealed with pressed indium wire. The use of vanadium can is important as vanadium nuclei has a very small coherent scattering length and is essentially transparent to incoming neutrons. Helium is used when low tempera- ture experiments are required to avoid freezing other gases. Loaded vanadium cans were typically mounted on closed-cycle refrigerators (CCRs) and cooled to base temperature, usually about 3 - 6 K. Diffraction measurements were conducted using either a Ge(311) monochroma- tor (? = 2.0709 ?) or Cu(311) monochromator (? = 1.540 ?) depending on the desired outcome of experiment. The Ge(311) monochromator offers the highest total flux on the sample which is often needed if the total amount of sample is small and it has the best resolution at low angles which may be required for crystal structures with large units cells and for magnetic phase identification. Cu(311) is a more universal choice with high resolution at all angles. A typical diffraction pattern is measured from 15? to 145? 2? with a step size of 0.05? at multiple temperature, typically above and below the observed magnetic transition temperature from other experimental methods. Col- 33 lection times have varied depending on the amount of sample but average 8-10 hours at each temperature. The BT-1 website has a lot of useful information regarding in- strument capabilities and how to acquire beamtime (https://www.nist.gov/ncnr/ high-resolution-powder-diffractometer-bt-1). Additional neutron powder diffraction experiments were conducted at BT-7 at the NCNR. This research was done in collaboration with Dr. Jeffery Lynn and his team at BT-7. BT-7 is a triple-axis spectrometer which is typically used for inelastic neutron spectroscopy. However, the instrument at BT-7 has a position sensitive dectector system (PSD) which is available for use as needed. On the typical setup at BT-7, a point detector is used which would make the collection of diffraction patterns very time-consuming. The PSD setup allows a 5? 2? range to be collected at once which is very useful for diffraction patterns of one or more diffraction peaks in close proximity. The setup here is very similar to BT-1 where ground powders were loaded in a vanadium can under helium atmosphere and cooled to base temperature on a CCR. In the case of the transition metal doped (LiOH)FeSe work, one diffraction peak in question was measured as a function of changing the temperature on the CCR setup. For this, the sample was cooled to the desired temperature and a diffraction pattern using the 5? 2? range of the PSD was collected for approximately 1 hour using a pyrolytic graphite monochromator (? = 2.359 ?) at each temperature. The BT-7 website has more information regarding the more traditional use of the triple-axis instrument for inelastic measurements (https://www.ncnr.nist.gov/ instruments/bt7_new/). Finally, additional neutron powder diffraction measurements were completed at Oak Ridge National Lab High Flux Isotope Reactor (HFIR) at the HB-2A instrument. 34 This work was done in collaboration with Dr. Simon A.J. Kimber and Dr. Keith Taddei. The sample setup and experiment at HB-2A is very similar to BT-1. Powder samples were loaded in vanadium cans and sealed under He atmosphere. The sample environment at HB-2A uses a multiple sample changer and an open ILL orange cryostat which requires frequent refills of liquid nitrogen, but the multiple sample changer enables less sample changes. HB-2A uses a Ge(115) monochromator (? = 1.54?) and has higher flux at the sample than BT-1 so typical collection times were approximately 2 hours so more samples can be scanned during experimental time. The HB-2A website catalogs the instruments capabilities, how to request time on the instrument, and how diffraction data is obtained, manipulated and analyzed (https://neutrons.ornl.gov/powder). Neutron powder diffraction is very useful but much different with regards to mag- netic diffraction and experimental setup than x-ray diffraction. Since neutron production usually requires a nuclear reaction as at NCNR and HFIR, neutron experiments are only done at national facilities which require proposals, approval and training. The most used references for neutron diffraction were the beamline website as well as the texts listed here.129,130 2.2.4 DC magnetic susceptibility and magnetization Magnetic susceptibility is a measure of a materials response to an external field. These measurements are done on both single crystals and powder samples. Powder sam- ples are typically loaded in a gel capsule, which has very small parasitic magnetic back- ground, and then loaded into a straw to fit onto the sample holder on the instrument. The 35 straw has very minimal background signal as well which can be measured separately and subtracted from the sample measurements in the case that the straw signal is contaminat- ing the signal of the material. Single crystal samples are typically mounted on a quartz paddle or brass rod which both offer very minimal background. Single crystal measure- ments must be carefully planned as to measure the sample along desired crystallographic axes as the magnetic field in the instruments are fixed with regards to direction. Magnetic susceptibility of the sample is determined by applying a known excitation field and mea- suring the magnetization through different techniques. Magnetic susceptibility per unit volume is defined as: ? ?0 M= B , where M is the magnetization of the sample, and B is the macroscopic magnetic field intensity, ? itself is a dimensionless quantity. Magnetic susceptibility can be converted to a molar quantity if the mass, molar mass and magnetic ions per unit cell is known. DC magnetization is used to observe two different phenomena in a sample. Typ- ically, the first experiment done is a measure of the temperature dependent magnetic susceptibility. This is used to determine if the sample undergoes any magnetic phase transition at a certain temperature which can be used to identify what type of magnetic order exists in the sample. This experiment consists of the collection of two temperature sweeps at the same applied field. The first is called the zero field cooled (ZFC) curve, where the sample is cooled without any applied field down to base temperature. Once base temperature is reached, an applied excitation field is applied and then the magnetiza- tion of the sample is measured as the temperature is heated to room temperature. For the second curve, field cooled (FC) magnetization data is collected from room temperature as the system is cooled to base temperature. The difference between ZFC and FC curves 36 is that in the ZFC curve the sample is allowed to cool down to base temperature without an external applied field and the sample?s magnetization will align based on its own ener- getics. However, in the FC curve, the directionality of the applied field in the instrument will cause the sample?s magnetization to align with the applied field. This is very important to understand ferromagnetic, ferrimagnetic and antiferro- magnetic ordering. In ferromagnetic order, at some critical temperature called the Curie temperature, the magnetic spins in the sample will all co-align. So, in the ZFC curve the magnitude of magnetization below the transition temperature will be different than in the FC mode. This depends on the strength of the applied excitation field in proportion to the internal field generated by the alignment of the spins. In a strong ferromagnetic material, the ZFC and FC mode will clearly and abruptly bifurcate at the transition temperature. In an antiferromagnetic or ferrimagnetic material, the spins will anti-align at a transition temperature called the Ne?l temperature. Antiferromagnetic and ferrimagnetic order can exist with many different possible arrangements and are often very difficult to solve the magnetic ordering structure. Ferrimagnetic ordering is similar to antiferromagnetic but instead of perfect anti-alignment, one set of spins will be weaker than the other. In these classes of materials, since the spins are anti-aligned and the net moment is zero or very low, the bifurcation of the ZFC and FC curves is very weak or non-existent. This type of measurement can help determine the magnetic ordering of the system.131,132 Superconductivity exhibits distinct behavior in the temperature dependent magnetic susceptibility curve. In the introduction, we discussed how superconductors exhibit the Meissner effect which is the exhibition of perfect diamagnetic behavior. Thus, in magnetic susceptibility, a transition to diamagnetic behavior is observed at the superconducting 37 critical temperature. As a note, for superconductors, magnetic susceptibility is typically reported as dimensionless (4??) which requires that the density of the sample is known which is typically determined from Rietveld refinements. The use of (4??) is done to show the shielding fraction of the sample at base temperature of the superconducting ma- terial as a perfect diamagnetic signal will yield a value of -1 at the measured temperature. In this scale, shielding fraction is just the percentage of the (4??) value from 0 to -1. This is used to determine if the sample is a bulk superconductor, which should show high shielding fraction, or filamentary, which typically shows very low shielding fraction. The behavior of magnetic susceptibility of superconductors can be very complex and a more detailed description can be found here.4 The other experiment usually performed is called isothermal magnetization. In this experiment, the magnetization of the sample is measured at fixed temperature as a function of applied field up to the max field available on the instrument. This is also very useful for determining the magnetic order of the system. Ferromagnetic systems, and sometimes ferrimagnetic systems, with a net moment can exhibit magnetic hysteresis as the field is sweeped where the measured magnetization depends on the magnetic history of the sample. Antiferromagnetic materials do not exhibit this behavior in the simplest case. When considering superconductors, they exhibit Meissner shielding up to a certain field, called the upper critical field, where the magnetization becomes non-zero. Isothermal magnetization curves are typically performed at many temperature, above and below the magnetic transition temperature in order to understand how the ordering and behavior changes with temperature.4,131,132 Throughout my work, DC magnetic susceptibility was measured by two different 38 techniques: 1) superconducting quantum interference device (SQUID) magnetometry and vibrating sample magnetometry (VSM). SQUID magnetometry measurements were com- pleted on two instruments, Quantum Design MPMS-XL and MPMS3, both of which are equipped with a 7 T magnet. SQUID magnetometry is the most widely used technique because it is incredibly sensitive to small magnetic fields on the order of 10?8 - 10 ?9 emu.133 In this technique, the sample is moved through a SQUID device which is two connected branches of Josephson junctions which form a closed loop. As the magnetized sample moves through the SQUID loop, the magnetic flux inside the loop changes so that the voltage across the SQUID loop will begin to change depending on the magnetic flux. The change in voltage is measured as a function of sample position inside the sample space which is then converted to magnetization by converting the measured voltage across the SQUID coil into magnetization. This is a simplified explanation and more in-depth references can be found here.133?135 The other technique used is VSM. VSM is typically used in our labs when higher fields are required as the VSM option is accessible of the Dynacool system which is equipped with a 14 T magnet. The use of a stronger external magnet can be needed in some systems with high superconducting upper critical fields or in systems with interest- ing magnetic ordering at higher fields. The one downside of VSM usage is that the VSM is less sensitive and usually on the order of 10?5 emu.136 The VSM technique uses Fara- day?s law of induction to measure the magnetization of the sample. The sample is placed inside a conducting coil, a magnetic field is applied to the sample, and the sample is then oscillated inside the conducting coil. If the sample is magnetized by the applied mag- netic field, the oscillating sample will cause swiftly changing flux within the coil which 39 will then induce a AC voltage through the coil. The amplitude of the induced voltage is directly proportional to the magnetization of the sample. 2.2.5 AC magnetic susceptibility AC magnetic susceptibility is very similar to VSM measurements except that in- stead of oscillating the sample through a conducting coil, a oscillating magnetic field is applied to the sample. Since the applied excitation field is oscillatory, this will induce an oscillatory voltage in the pick-up coil of the AC measurement. The big difference between the VSM measurement and the AC magnetic susceptibility measurement is that the induced AC voltage in the AC magnetic susceptibility measurement comes from the changing magnetization of the sample due to the applied field. This means that the time dynamics of the induced voltage are intrinsic to the system and describe the time dynam- ics of the magnetization in the sample. This is very useful in the study of systems where the magnetization changes in time. This is true in spin glass systems, ferromagnetic domain formation and even superconduc- tors inside the superconducting state.4,137?139 In our lab, the AC magnetic susceptibility is performed on the Quantum Design PPMS 14 T.140 The measurement can be done on powders or single crystals. Typically, powders are loaded into a Teflon bucket with an attached sample rod which seal the bucket and fits into the pick-up and drive coils which are all then loaded into the PPMS system for atmosphere and temperature control. With this setup, the lowest achievable temperature is 1.8 K on the PPMS 14 T, but members of Dr. Johnpierre Paglione?s group have designed and implemented many AC magnetic 40 susceptibility instruments for use in systems below 1.8 K in the milliKelvin regime. 2.2.6 Longitudinal electrical resistivity Longitudinal electrical resistivity is a useful technique for the measurement of the electronic properties of a material especially down to low temperature. Electrical resis- tivity measurements were performed on Quantum Design PPMS 9 T, PPMS 14 T and Dynacool 14 T systems. For a typical measurement, single crystals of the material were wired up by attaching four gold wires with DuPont 4929N silver paste. The DuPont 4929N silver paste must be mixed with a small amount of 2-butoxylethyl acetate in order to dissolve the silver paste in such a way that it is workable for wiring up the sample. The four wire setup is used to eliminate the contamination of wire resistance, although gold wire is very low, as well as contact resistances from the silver paste connection to the sample. In this setup, a known current is applied to the two outer wires and then voltage in measured across the two inner wires as shown in Figure 2.4. In this case, no current should pass through voltage wires and thus the voltage measured across the two voltage terminals should only be dependent on the material itself. Thus, longitudinal resistance can be determined by the measured voltage across the voltage lead since the applied current is known. This is then converted to resistivity, which is intrinsic to the sample and does not depend on the sample size, by dividing out geometrical components. ? tw Vxx twxx = Rxx ? l = I ? l . Here, ?xx describes the longitudinal resistivity along the x-direction which can be changed as long as the crystallographic axes are well-known, Rxx describes the resistance along the x-direction, t,w, l are geometric 41 Fig. 2.4: Schematic representation of longitudinal resistivity measurement technique and example of wired CoSe single crystal on the PPMS DC puck. parameters of the crystal which are typically measured using a microscope and attached camera, and Vxx is the measured voltage across the voltage leads in the x-direction. The samples are typically wired and then the other ends of the wires are attached to solder pads on the Quantum Design PPMS DC resistivity pucks. The puck is then loaded in the PPMS 9 T, PPMS 14 T, or Dynacool 14 T and resistivity is measured as both a function of temperature and field. Measurements done as a function of temperature are done to determine phase tran- sitions which may occur at low temperature such as a superconducting transition or the onset of a charge density wave. Measurements done as a function of field, called mag- netoresistance, are done in a similar vain to understand how the material behaves at high 42 magnetic field, such as the suppression of a superconducting transition. A systematic study of resistivity in materials can be found here.134,141 One key caveat to longitudinal resistivity measurements is that it is not a bulk char- acterization technique. Electrical transport will occur on the path of least resistance which means the sample technique as a whole is highly susceptible to sample inhomogenity and other confounding effects. This means that a result from a single experiment may be misleading. To that end, electrical resistivity measurements are always done in at least triplicate on the same sample or from the same sample batch. If interesting phenomena is observed in resistivity measurements, these are always corroborated with a bulk charac- terization technique such as magnetic susceptibility or heat capacity measurements. 2.2.7 Low temperature specific heat As mentioned, electrical resistivity measurements are often paired with specific heat measurements in order to match observed phenomena. Specific heat is a bulk characteri- zation technique which relates a known heat applied to a material to the resulting change of temperature of the material itself. Because the specific heat is related to the change in entropy of a material with respect to temperature, specific heat measurements are sus- ceptible to any change in entropy of the system. Specific heat was measured on single crystals and pressed powders of samples on the PPMS 14 T system142 using the two-tau technique.143 In this technique, the sample is placed atop a heating platform which is specially designed by Quantum Design on top of a traditional measurement puck. The sample 43 is placed onto the heating platform with low temperature Apiezon N grease to allow optimal thermal contact between the sample and the heating platform. The sample puck is then loaded into the PPMS 14 T which is set to high vaccum (< 10?4 Torr) which is required for heat capacity measurements as the sample must be as thermally isolated from the environment as possible. For the measurement, a known constant heat is applied to the heating platform for a fixed amount of time and then the cooling of the system is measured as a function of time. The cooling of the system behaves according to the differential equation: C dTtotal dt = ?Kw(T ? TB) + P(t) where Ctotal is the total heat capacity of the system, Kw is the thermal conductance of the system, T and TB are measured temperatures of the system and P(t) is the applied power to the heating platform. This is a first order differential equation with an exponential solution which is used to model the growth and decay of the sample temperature as a function of time through the Quantum Design software to extract the total heat capacity. This setup only measures the total heat capacity of the system: puck, grease, and sample. This means that in order to extract just the component of heat capacity of the sam- ple, a preliminary measurement of the puck and grease must be done, called an addenda. Addenda measurements are typically done with less temperature intervals, in order to save time, and data points are interpolated as required to match the experimental data for the sample. Heat capacity measurements are typically done at low temperature but will be done at any temperature where a phase transition has been observed. One caveat of this is that at higher temperature, the growth and decay of the temperature of the system can take very long to achieve so measurement times for a complete measurement may take on the order of days. Addenda heat capacity values are subtracted from the total heat 44 capacity values to yield the heat capacity of the sample itself which can be converted to specific heat if the mass and molar mass of the sample are known. Another complication which may arise in these measurements is that the expo- nential solution assumes a perfect coupling between the sample platform and the sample itself. This is not always the case and poor coupling can cause the modeling from the soft- ware to be very poor. The Quantum Design software does allow for the case where there is thermal link from the stage and the sample and modeling using this fitting procedure is employed when data modeled using the previously mentioned functional dependence is poor. In many case, poor sample coupling can be fixed by added more grease to the sample stage to ensure there is thermal contact between the sample stage and the sample. In one instance, a pressed pellet of a material was used as opposed to a single crystal as single crystals of the material was unable to achieve appropriate thermal coupling. 2.2.8 Electron microscopy Electron microscopy is a catch-all term for any imaging and diffraction technique which uses electrons as the source of illumination in contrast to traditional microscopy which uses light as the source of illumination. All electron microscopy characterizations were carried out in the Advanced Imaging and Microscopy Laboratory (AIM) Lab at the University of Maryland Nanocenter through collaboration with Dr. Sz-Chian Liou. Two main techniques were used scanning electron microscopy (SEM) and transmission electron microscopy (TEM). Scanning electron microscopy was performed on a Hitachi SU-70 field emission 45 scanning electron microscope and energy dispersive X-ray spectroscopy (EDS) was done using a Bruker EDS detector. Typically, an image is produced by secondary electrons which are emitted from the sample by excitation from the incoming electrons. SEM images can typically achieve resolution on the nanometer scale which is useful for deter- mining sample morphology. Most importantly, SEM techniques are used for elemental analysis using EDS techniques. In the EDS method, X-rays are produced by the sample as the high energy electrons cause the atoms of the sample to lose inner core electrons. Those inner core electrons are then replaced by outer core electrons which radiate down in energy by releasing an X-ray. These X-rays are characteristic for each element and are well-cataloged. Thus, the atomic composition of the sample can be determined by measuring which characteristic X-rays are produced. In general, EDS techniques are sen- sitive on the order of 2-5% atomic composition. In some cases, characteristic X-rays of two elements may have strong overlap in energy which can cause large errors in evalu- ating the elemental composition of a sample. Another issue which may arise from EDS analysis is that the measurement can easily be contaminated by impurities which will yield confusing elemental compositions. In the case where a sample may have impurities, the sample morphology is easily shown via SEM images and EDS is only performed on crystallites with the correct morphology. In typical usage, elemental analysis from EDS is used to corroborate Rietveld refinement data from X-ray or neutron diffraction in order to determine the elemental composition of the sample. Transmission electron microscopy was performed on a JEM 2100 LaB6 transmis- sion electron microscope with an acceleration voltage of 200 KeV. The main difference between SEM and TEM is that TEM uses transmitted electrons for imaging as opposed 46 to reflected electrons as in SEM images. The main benefit of this is TEM is not strictly limited to surface characterization as SEM is typically used for surface morphology and imaging. For our research, the TEM is employed to produce electron diffraction images of a powder sample. Electron diffraction has been useful for systems with very compli- cated crystal structures. The main reason is that the wavelength of the electrons used are 0.25 ? whereas typical X-ray diffractometers using Cu K? radiation has a wavelength of 1.54 ? and neutrons are typically even longer wavelength. A shorter wavelength means that the Ewald sphere for meeting Bragg conditions in the reciprocal lattice is much larger which means many more Bragg reflections can be identified. Another huge benefit of TEM electron diffraction is that because electrons can be focused down to a very small size, this allows for nanometer size crystallites to be in- dividually used for diffraction experiments. For our research, using hydrothermal reac- tion often times produces very small crystallites on the nanometer scale. Thus, electron diffraction offers the ability to do diffraction on a single crystallite in order to help de- termine the crystal structure. Albeit, electron diffraction patterns alone are typically not enough to accurately determine a crystal structure or lattice parameters and are often used as a complementary technique to traditional X-ray or neutron diffraction techniques. An- other caveat, is that small crystallites are typically required for TEM and electron diffrac- tion as the sample must be thin enough to allow electrons to transmit without significant absorption and reflection. On the whole, electron microscopy is a very useful set of techniques to be used when single crystal samples are not able to be synthesized. SEM imaging allows for sam- ple morphology to be determined which is extremely useful in determining different im- 47 purities phases. This is especially true in our research where our desired products are lay- ered phases which are usually very easily identified in SEM images. TEM has also been particularly useful to determine the symmetry present in our intercalated sample using electron diffraction. In particular, in systems with complex intercalated species, electron diffraction is useful to image the four fold symmetry of the transition metal chalcogenide layers to show they remain intact through intercalation. 2.2.9 Inductively coupled plasma - atomic emission spectroscopy Inductively coupled plasma - atomic emission spectroscopy (ICP-AES) is a elemen- tal composition technique which is done to determine metal concentrations in a product to extremely high levels of accuracy. This work was done in collaboration with Dr. Marya Anderson and the Environmental Engineering Laboratories in the Department of Civil and Environmental Engineering at the University of Maryland. Data was collected us- ing a Shimadzu ICPE-9000 spectrometer which utilizes an argon plasma source to ionize sample solutions. For ICP-AES analysis, a sample solution is ionized using an argon plasma and characteristic emission lines are collected using a spectrometer within the in- strument depending on which elements were ionized in the argon plasma. The intensity of each characteristic emission for each element is then matched with a known calibration of standards. The standard and sample preparation is the most difficult part of ICP-AES analy- sis. Standard preparation requires the preparation of typically 5-7 standards of known concentration made from commercially purchased elemental solutions (purchased from 48 FLUKA). These standards are acquired at a concentration of 100 ppm and must be diluted down to form a calibration curve, typically in the range of 0.2, 0.5, 1, 2, 5, 10, 20, 100 ppm. This full range of a standards allows for samples of a wide variety of concentra- tions to be properly characterized. Sample preparation is done through the dissolution or digestion of a bulk sample and dilution down to concentrations within the calibration curve are required. To start, a bulk sample, crystal or powder, is massed and then dis- solved in concentrated nitric acid until the entire sample is fully solvated; this is done in a hood for safety. One important note, ICP-AES is known as a destructive method whereby the sample must be dissolved and cannot be recovered for the analysis. Thus, it is only done when completely necessary to accurately determine the elemental composition when other techniques such as SEM-EDS and Rietveld refinement of powder diffraction data is not sufficient. The sample solution is now diluted to 2% nitric acid which is required for proper ionization in the argon plasma for this particular instrument. The initial mass of the bulk sample and the dissolution is used to determine the approximate concentration of each element in the solution to ensure it falls within the range of the calibration curve. This process is done for each standard and each sample and at least 15 mL of each sample solution is required for proper usage on the instrument. This is because each measurement is done in triplicate as the instrument takes 5 mL aliquots from the sample solution and feeds them into the argon plasma torch for spectroscopic analysis. The in- strument automatically runs through each sample up to a total of 60 samples. Once data has been collected, the software requires the user to select which characteristic emission wavelengths are chosen for each element and to ensure the calibration curve is accurate. Once those are chosen, the software matches each emission intensity for each element in 49 each sample to the calibration curve and determines the concentration of each element in the sample. This data can then be used to determine the relative atomic percents of each element in a sample to determine the stoichiometry. For ICP-AES analysis typically er- rors are in the < 0.1 ppm which equates to 0.00001 atomic percent. However, larger errors typically arise from weighing the bulk sample and determining the dilution of each sam- ple which typically puts an error bar of < 0.01 atomic percentage depending on accuracy of the balance used and volumetric glassware for serial dilutions. One caveat of the ICP-AES technique is that it is not able to determine the concen- tration of non metallic elements such as nitrogen, oxygen, halides and noble gases within the sample. This is even true for some metalloid elements depending on the digestion method used. All together, ICP-AES analysis is typically used for very accurate metal concentration determination in a sample and is used in conjunction with SEM-EDS and Rietveld analysis techniques to understand the stoichiometry and elemental composition of a material. 2.2.10 Thermal stability and characterization Thermal stability measurements on samples were performed using two different techniques: thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC). Both techniques are used to determine how thermally stable a material is with regards to heating in different environments. For our research, this has implications for determining the composition of a material as well as understanding how a material may be used in future research. Thermogravimetric analysis and differential scanning calorimetry was 50 conducted on a Mettler-Toledo TGA/DSC 3+ with high temperature furnace; typically samples were heated from room temperature to 800 ?C in both an air and inert gas atmo- sphere to see how they behaved in different environments. Thermogravimetric analysis measures the mass of a sample very accurately as a function of time as a heating profile is applied to the sample. The change of mass as a function of time and temperature describes different thermal phenomena in the sample such as phase transformations and decomposition. One piece of very useful information which can be gleaned from TGA analysis with regards to our research is to show how intercalated species behave in a compound. This is especially true for organic intercalates which will decompose at relatively low temperatures, which can be shown as an abrupt and significant mass change in a sample at some characteristic temperature which de- scribes the temperature at which the sample de-intercalates the organic molecule and the organic molecule decomposes through oxidation. Differential scanning calorimetry is a similar technique but is used specifically to determine the phase transformation of a material. In this technique, the sample and a reference sample are heated at the same rate. The temperature of each sample is monitored as a function of time as the sample will allow heat to flow differently than the reference sample. At some temperature, the heat flow will change abruptly which indicates the sample is going through a phase transformation. We use this information to understand the thermal stability of new compounds and this informs us about what possible chemical manipulations and post synthetic modifications are available for a particular material. 51 2.2.11 Density functional theory calculations Density functional theory (DFT) calculations are performed by our group in order to match experimental results to theoretical calculations. Through density functional theory caculation code, we are able to predict the stability of a compound depending on its crystal structure and lattice parameters. All density functional theory (DFT)144,145 calculations were performed by using the Vienna Ab-initio Simulation Package (VASP)146?149 software package with potentials using the projector augmented wave (PAW)150 method. The exchange and correlation functional were treated by the generalized gradient approximation (PBE-GGA).151 The cut-off energy, 450 eV, was applied to the valance electronic wave functions expanded in a plane-wave basis set. A Monkhorst-Pack152 generated 23?23?17 k-point grid was used for the Brillouin-zone integration to obtain accurate electronic structures. I would like to thank University of Maryland supercomputing resources (http://www.it.umd.edu/hpcc) made available for conducting DFT calculations in this thesis. A typical workflow for our group with regards to DFT calculations is as follows. A new material is synthesized and the structure and properties are measured through the long list of characterization techniques listed above. We then will complete density functional theory calculations on the compound in order to determine how well our results match those predicted by theory. It must be noted that density functional theory, especially those performed by VASP, which is a pseudopotential method, are not exact solutions. They are always approximations; structural properties and dynamics are typically fairly close to experimental values, but electronic and magnetic properties may vary greatly 52 between experiment and theory. This is because, especially in our research, the hardest aspects of the physical system to account for is the electron-electron correlations which are typically non-negligible and significant. To start, a measured crystal structure is arranged in the VASP software and a pre- liminary calculation is done in order to determine the optimal lattice parameters for the system using a series of fixed lattice calculations. These calculations yield some energy for the system as calculated, and that energy is then plotted as a function of lattice param- eters or volume of the unit cell. These energy values as a function of volume is then fit with an equation of state to determine the optimal lattice parameters for the system. As a last step for structural calculations, the optimal lattice parameters are used for a structural calculations where the ions in the unit cell are allowed to relax which as a whole allows the lattice parameters to change as well as atomic coordinates. Once this calculation is done, the optimal structure has been calculated for the system at hand. This structure is then used for all electronic property calculations. The first step for electronic property calculations is to perform a self-consistent field calculation. Here, the optimal fixed structure is used iteratively to determine optimal de- scription of the electron density functional for the system. Once this is optimized, the fixed structure and fixed electron density functional is used to calculate density of states as well as band structure. VASP software allows for angular momentum decomposition within the calculations so that each orbital contribution can be determined in the den- sity of state and band structure calculations. This is particularly useful for our materials as transition metal d-orbitals are the most significantly contributing at the Fermi level. Throughout these processes, the effect of electron spin degree of freedom can be consid- 53 ered or not which equates to considering if the system should have magnetic order or not. As a caveat, due to the way the calculations are done starting with structure, complex magnetic order of the antiferromagnetic or ferrimagnetic type require the construction of a supercell of the original unit cell. This is a more advanced technique and in cases where magnetic order is the direct aim of the calculations collaborations are usually required. In conclusion, our group uses DFT calculations as a method to confirm experi- mental findings in our materials. It is a very powerful technique which allows the user to understand the ground state properties of a material. As noted, the theory has many approximations which means the technique can be used with appreciable accuracy for structural properties but electronic and magnetic properties may vary between experi- ment and theory. Another note of caution, all calculations begin with an accurate de- scription of the structure of the material; any error in the structure will lead to huge errors in the electronic property calculations. There are many references which describe how to use VASP and related software listed here.153,154 In particular, VASP software page https://www.vasp.at/wiki/index.php/The_VASP_Manual and the materials explorer function on the Materials Project https://materialsproject.org/ offers a wide-range of functionality and starting points for DFT calculations. 54 Chapter 3: Topochemical synthesis and frustrated magnetism in CoSe - analog of superconducting FeSe The work described within this chapter was published in Journal of the Ameri- can Chemical Society 2016, 138, pg. 16432. Xiuquan Zhou, Hector Vivanco, John- pierre Paglione, Craig M. Brown, and Efrain Rodriguez were contributing authors of the manuscript. B.W., X.Z. and H.V. prepared the samples, X.Z. performed MPMS mea- surements and DFT calculations, B.W. collected resistivity data, X.Z., B.W., and C.M.B. collected the neutron data. as well as Physical Review B 2018, 97, pg. 104408. Xi- uquan Zhou, Hector Vivanco, Daniel J. Campbell, Kefeng Wang, Dave Graf, Johnpierre Paglione and Efrain Rodriguez were contributing authors for the manuscript. B.W., X.Z., and H.V. contributed to crystal growth, B.W. collected MPMS measurements, AC sus- ceptibility measurements, and heat capacity measurements, X.Z. performed DFT calcu- lations, and D.J.C., K.W. and D.G. collected the high field resistivity data. 3.1 Prologue The majority of the work described in this chapter was enabled by preceding work on the synthesis of the metastable CoSe which is isostructural to FeSe in the iron chalco- genide superconducting family. Superconductivity is very sensitive to doping and chem- 55 ical effects so from an experimental point of view, the next step would be to substitute iron for other transition metals in order to investigate how these dopants affect supercon- ductivity in the system. It was found that the all transition metal substitutions are limited to ? 10% substitution on the iron site due to phase transformation or the appearance of multiple competing phases and all the dopants except manganese were shown to destroy superconductivity with minimal doping.155?158 As mentioned, FeSe is the only tetrahedral metal chalcogenide to crystallize in this structure type. Thus, when doping the system be- yond some limit, thermodynamics will govern the formation of other phases which moves away from the desired structural motif. This is where the requirement for a new reaction pathways, kinetically controlled, is needed in order to overcome thermodynamics and stabilize the phase desired. Superconductivity in FeSe was reported in 2008, but was not reported in the sulfide analogue, FeS, until 2015. This is because anti-PbO type FeS (mackinawite) is metastable and previous chimie douce synthetic methods used resulted in the observation of ferro- magnetic semiconducting behavior.159,160 However, in 2015 a new hydrothermal synthetic scheme was devised to form FeS and found superconductivity at 5 K;58 the discrep- ancies in these works can be attributed to the instability of mackinawite which trans- forms to Fe3S4 under adverse conditions. More recently, another reaction scheme was devised to form single crystals of tetragonal FeS through topochemical de-intercalation of KxFe2?yS2, which supported the previous work with FeS formed through this method exhibiting superconductivity at 4 K.59 This de-intercalation of potassium ions between layers of FeS utilized the thermodynamic stability of KxFe2?yS2, which can be formed through a solid-state self flux method,161 to access the desired structural motif to form 56 anti-PbO mackinawite. This led to a general realization of the possibility to target other extended solids through the de-intercalation of AxM2Ch2, where A = alkali metal, M = transition metal, and Ch = chalcogenide to form other anti-PbO type tetrahedral transition metal chalcogenide. Here, we target the synthesis of the cobalt analogues to superconducting FeSe and FeS through a kinetically controlled topochemical de-intercalation of the thermodynami- cally stable KxCo2Ch2 phases. However, in general, the ICSD lists 1865, 40 being chalco- genides, compounds that crystallize in the ThCr2Si2 structure which could offer an expan- sive playground for future utilization of chimie douce techniques to form new materials. The corresponding phases, KCo2Se2 and KCo2S2, were synthesized by Greenblatt et al. and their group has developed extensive methods to synthesize these ternary/quaternary chalcogenide.162?164 Their work reported ferromagnetism in KCo2Se2 and KCo2S2 with Curie temperatures, TC = 78 and 127 respectively. Thus, the targeting of CoSe and CoS, isostrutural to FeSe, fulfills two routes of investigation: 1) how does changing metal ox- idation states and electronic configuration affect the physical properties from KCo2Ch2 to CoCh, and 2) how do the properties in these metastable cobalt chalcogenides relate to superconductivity in the iron-based system. So, in total, this preliminary work aims to in- vestigate a kinetically controlled route to form metastable tetrahedral cobalt chalcogenide phases, the character of the magnetic interactions on the layered Co square lattice, and how the physical properties relate to superconductivity in the analogous iron system. Figure 3.1 illustrates the collection of synthetic routes encountered through attempts to synthesize metastable tetragonal CoSe and CoS phases. In Figure 3.1, the successful topochemical de-intercalation process is shown. 57 K1-xCo2Ch2 (1) CoCh (2) m-Co1-xCh (4) K+,, H2(g) LiOH (aq) RT, t = 24 hours ? T > 300?C n s T ? (Lien)Co Ch (s) , e ay > d 15 2 2 Li = 7 0? Co9Ch, t C 8 (3) ?C 70 ? ? Fig. 3.1: Reaction schemes for the topochemical synthesis of cobalt chalcogenides. Thermody- namically stable KCo2Ch2 is converted to tetragonal CoCh through a room temperature de-intercalation reaction. Higher temperature cause the formation of cobalt pentlandite and monoclinic CoCh. Possible avenues for intercalation are shown so LienCoSe can be formed where en = ethylenediamine. In this process, KCo2Ch2 reacts with water to form KOH and evolve hydrogen gas through the extraction of interlayer potassium ions to form CoCh depicted below: 1 KCo2Ch2(s) + H2O(l) ?? 2CoCh(s) + KOH(aq) + H2(g) (3.1)2 Due to the observed evolution of hydrogen gas, the reaction rate could be semi- quantitatively studied to understand how the de-intercalation reaction proceeds. Analysis confirmed that the de-intercalation reaction involved the hydration of K+ ions to form KOH accompanied by a transfer of electrons between Co1.5+ ions and H2O to release H2 gas. The reaction scheme in Figure 3.1 illustrates the necessity to complete the de- intercalation under strongly basic conditions in order to ensure the formation and stability of Se2? ions so that no selenium is lost to the formation of selenide compounds.165 Powder x-ray diffraction (PXRD) was performed in order to compare the crystal 58 structure of KCo2Ch2 and CoCh; Figure 3.2 shows PXRD and neutron powder diffraction (NPD) patterns for KCo2Se2 and CoSe. (a) Ycalc eg NPD (T = 3 K) I4/mmm (c) Yobs - Ycalc Ycalc ? = 2.079 ? a = 3 .832 ? Yobs Yobs - Ycalc c = 13.848 ? a = 3 .717 ? Yobs Rwp = 2 .92% c = 5 .275 ?Rwp = 5 .32% 20 40 Magnet8ic0 Phase (b) P4/nmm (d) Rwp = 24?.7 (0d%eg a = 3 .717 ? c = 5 .330 ? Rwp = 2 .11% 10 20 30 40 50 60 70 20 40 60 80 100 120 140 2? (deg.) 2? (deg.) Fig. 3.2: Powder X-ray diffraction patterns of (a) KCo2Se2) at room temperature and (b) CoSe room temperature. (c) Neutron powder diffraction (NPD) pattern of CoSe at 3 K (d) NPD pattern with attempted magnetic refinement of the ferromagnetic contributions. Tick marks representing the phases are shown below the calculated, observed, and differences curves from Rietveld analysis. As seen from the x-ray and neutron data (details listed in Table 3.1), de-intercalated CoSe can be matched very well with the anti-PbO structural model with neither diffrac- tion pattern showing contributions from remaining KCo2Se2 phase. Elemental analysis utilizing energy dispersive spectroscopy (EDS) and inductively-coupled plasma atomic emission spectroscopy (ICP-AES) techniques illustrate a composition of Co0.98?0.02Se with no residual potassium remaining in the selenide analogue. However, for the sulfide analogue, EDS measured 4.5 at.% remaining potassium owning to slightly incomplete de-intercalation which could be remedied through longer reaction times. In total, crys- tallographic and elemental analysis show the successful de-intercalation of KCo2Ch2 to form anti-PbO type tetragonal CoCh phases. 59 Relative Intensity (arb. units) Relative Intensity (arb. units) KCo2Se2 (298 K, PXRD), I4/mmm, Rwp = 2.917% a = 3.832(2) ?, c = 13.848(3) ? atom Site x y z Occ. U 2iso (? ) K1 2a 0 0 0 0.94(6) 0.109(22) Co1 4d 0 0.5 0.25 0.96(6) 0.060(11) Se1 4e 0 0 0.359(3) 1 0.019(4) Co-Se (?) Se-Co-Se (?) Se-Co-Se (?) Co-Co (?) anion height (?) 2.442(6) 103.4(4) 112.6(2) 2.710(3) 1.509(3) CoSe (298 K, PXRD) , P4/nmm , Rwp = 2.102% a = 3.717(3) ?, c = 5.330(3) ? atom Site x y z Occ. U 2iso (? ) Co1 2a 0 0 0 1 0.012(3) Se1 2c 0 0.5 0.265(5) 1 0.010(3) Co-Se (?) Se-Co-Se (?) Se-Co-Se(?) Co-Co (?) anion height (?) 2.332(2) 111.382(63) 105.8(2) 2.6284(3) 1.412(3) CoSe (3 K, NPD) , P4/nmm , Rwp = 5.318% a = 3.716(6) ?, c = 5.275(1) ? atom Site x y z Occ. U 2iso (? ) Co1 2a 0 0 0 1 0.0026(8) Se1 2c 0 0.5 0.269(2) 1 0.0020(5) Co-Se (?) Se-Co-Se (?) Se-Co-Se(?) Co-Co (?) anion height (?) 2.339(5) 111.632(36) 105.232(68) 2.6280 1.412(3) Tab. 3.1: Structural parameters for ground single crystals of KCo2Se2 and CoSe. Structures are for room temperature PXRD data and 3 K NPD data. All relevant bond angles and distances from the refinements are given. Standard uncertainties given in parantheses indicate one standard deviation. To further understand the stability of these phases, extensive reaction condition experiments were executed to probe the appropriate thermal window for the formation of these phases. The remaining structures in Figure 3.1, cobalt seleno-pentlandite (Co9Se8: 3) and monoclinic cobalt selenide (Co3Se4: 4) illustrate the formed products through different temperature regimes. It was found that any hydrothermal reaction between 100 - 200 ?C led to the co-existence of tetragonal CoSe and cobalt seleno-pentlandite with increasing formation of the monoclinic phase at elevated temperatures. To understand this in more detail, temperature-dependent x-ray diffraction was performed on ground single 60 crystals of tetrgonal CoSe from 27 - 600 ?C in an Argon atmosphere. It was found that at 200 ?C tetragonal CoSe completely transformed to the cobalt seleno-pentlandite phase. This remained the major product to 400 ?C where it began to convert to the monoclinic phase which remained the major product up to 600 ?C and when cooled back to room temperature. This illustrates that these tetragonal CoCh phases have a limited thermal window of stability which can be exploited for future chemical manipulations. We now move on to describe the extensive physical property measurements employed in order to understand the interesting magnetism present in CoSe. 3.2 Introduction The iron-based superconducters are composed of Fe2+ square lattices stacked to form layered materials. For example, the simple FeSe superconductor contains stacked layers of Fe2+ centers tetrahedrally-coordinated to selenide anions. Remarkably, its Tc of 8 K,42 can be increased to 65 - 100 K when isolated as a single layer.50 Therefore, it is the square sublattice of d-cations that may hold the key to understanding the physical properties of these systems. In this article, we have completely replaced the Fe2+ cations in FeSe with Co2+, and studied its magnetization, magnetotransport and specific heat properties to further explore the physics of metal square lattices. In addition to crystal structure, the relationship between magnetism and supercon- ductivity is of paramount importance for these layered chalcogenides. In the iron pnictide superconductors (e.g. BaFe2As2 and LaOFeAS), suppression of the parent antiferromag- netic (AFM) phase can lead to the emergence of superconductivity.24,166 However, no 61 KCo2Se2 CoSe I4/mmm, TC = 78 K P4/nmm, Tg ~ 10 K Fig. 3.3: Crystal structures of KCo2Se2 and CoSe. long-range magnetic ordering has been observed in any of the FeSe or FeS superconduc- tors. Although antiferromagnetism was found in Fe1+xTe, the origin of its magnetism is different from that of the pnictides, and it is largely influenced by the amount of intersti- tial iron.167?169 Thus, it is less clear how magnetism and superconductivity interact in the FeCh (Ch = chalcogenide) systems compared to their pnictide counterparts. Currently, one key issue is that isostructural systems to FeCh are limited due to syn- thetic challenges. Previously, we have overcome this challenge by topochemical means to convert KFe2S2 to superconducting FeS.59 Using a similar method, we successfully prepared two new FeCh analogues, tetragonal CoSe and CoS.61 The ferromagnetic order- ing from 78 K in KCo2Se 1642 to 10 K in CoSe61 was suppressed by de-intercalation of potassium cations to form pure CoSe as shown in Figure 3.3. These new Co-based phases are promising for understanding the Fe-based superconductors due to their structural and electronic proximity. Much of the work performed to understand the magnetism in iron pnictides has been done with those that adopt the ThCr2Si2 structure-type (?122?-system). This structure- 62 type allows for a wider range of substitutions on the metal, anion and interlayer cation sites to study doping effects.27,82,87,170 There has been extensive work on the cobalt ana- logues to ?122? iron pnictides, ACo2Pn2, with various interlayer alkali or alkali-earth cations (A).171?178 The observed magnetism in these pnictides was largely tuned by size and electronic effects from changing the CoPn layer distances. An intriguing question is: can CoSe be tuned into a superconductor like FeSe? By directly comparing their band structures, CoSe chould share similar electron-hole pock- ets with FeSe if the electron filling level is reduced.61 Therefore, it may be possible to tune CoSe into a superconductor by increasing the Co oxidation state to form d6 cations isoelectronic to Fe2+. In order to investigate this, two fundamental factors must be under- stood: 1) the character of the magnetic interactions within the Co square lattice, and 2) how its magnetism compares to other FeCh based superconductors. Here, we have per- formed extended magnetic and transport characterizations to understand the magnetism within CoSe and its proximity to superconductivity in related FeSe. 3.3 Experimental Methods Single crystals and powders of CoSe were synthesized following the previous method in literature.61 Crystals of CoSe were lustrous silver with high degree of layered morphol- ogy. Temperature dependent DC (direct current) magnetic susceptibility measurements were carried out using a Quantum Design Magnetic Susceptibility Measurement System (MPMS) on powder samples of tetragonal CoSe. Field-cooled (FC) and zero field-cooled 63 (ZFC) measurements were taken from 1.8 K to 300 K with various applied magnetic field strengths. Magnetic hysteresis measurements were carried out using a PPMS DynaCool utilizing a vibrating sample magnetometer (VSM) taken at a series of temperatures with applied magnetic field between H = ?14 T on single crystals of CoSe mounted on a quartz paddle via Ge 7031 varnish. AC (alternating current) magnetic susceptibility was measured with a 14 T Quan- tum Design Physical Property Measurement System (PPMS-14) on powder samples of tetragonal CoSe. Zero field-cooled measurements were taken from 35 K to 1.8 K with an AC-field of 10 Oe and AC-frequencies of 10 Hz to 10 kHz. Due to the instrument setup, a residual DC field within the PPMS-14 ranged from 40 Oe to 100 Oe. Electrical transport measurements were preformed using a 9 T Quantum Design Physical Property Measurement System (PPMS-9) with single crystals of CoSe mounted on a Quantum Design AC transport puck. Electrical resistivity was measured using the four-probe method with gold wire and contacts made with silver paste. The temperature and field dependence of longitudinal electrical resistivity was measured in a range from 300 K to 1.8 K with applied fields up to 9 T. Electrical transport measurements at fields up to 31 T were performed at the DC Field Facility of the National High Magnetic Field Laboratory in Tallahassee, Florida. Angular dependence measurements at base temperature of the He-3 system (500-600 mK) were made by rotating the sample plane (ab-plane) from perpendicular (0 degrees) to parallel (90 degrees) to the applied field. Temperature dependent magnetotransport was measured for applied field both perpendicular and parallel to the sample plane between base temperature and 12 K. 64 Heat capacity measurements were preformed using the PPMS-14. Heat capacity measurements on tetragonal CoSe single crystals yielded poor results due to low thermal contacts arising from the micaceous nature of the CoSe flakes. Consequently, a pressed pellet of CoSe ground single crystals was used for the heat capacity measurements per- formed with the relaxation technique.143,179,180 All density functional theory (DFT)144,145 calculations were performed by using the Vienna Ab-initio Simulation Package (VASP)146?149 software package with potentials using the projector augmented wave (PAW)150 method. The exchange and correlation functional were treated by the generalized gradient approximation (PBE-GGA).151 The cut-off energy, 450 eV, was applied to the valance electronic wave functions expanded in a plane-wave basis set. A Monkhorst-Pack152 generated 23?23?17 k-point grid was used for the Brillouin-zone integration to obtain accurate electronic structures. 3.4 Results 3.4.1 Magnetic properties Our previous work demonstrated the suppression of ferromagentism from 78 K in KCo Se 1642 2 to 10 K in CoSe61. However, due to the very low ordering moment as well as the proximity to the iron-based superconductors, a more detailed investigation of the magnetism and electronic properties has been undertaken. Figure 3.4 shows the temperature dependence of inverse FC magnetic susceptibility for ground single crystal samples of CoSe at various applied DC fields. The inverse susceptibility was fit in the paramagnetic range from 100 K to 300 K to the Curie-Weiss 65 Curie-Weiss 1000 0.01 T 0.20 T 1.00 T 5.00 T 500 500 0 0 50 100 0 0 100 200 300 T (K) Fig. 3.4: Inverse magnetic susceptibility of CoSe vs. temperature measured in applied field of 100 Oe. The inverse magnetic susceptibility is fit from 100 K to 300 K to the Curie-Weiss law plus a temperature-independent term. The inset shows inverse magnetic susceptibilities for different applied DC fields (0.2 T, 1 T and 5 T) to emphasize the change in slope near 82 K. law: C ?mol = ?0 + ? (3.2)T ?CW where ? = 3.52 ? 10?4 emu0 Oe?mol accounts for parasitic paramagnetic and diamagnetic con- tributions, C = 0.1579 emu?KOe?mol denotes the Curie constant, and ?CW = -87.29 K is the Weiss constant. A strongly negative Weiss constant empirically indicates predominant antifer- romagnetic fluctuations. The frustration parameter, f , for a magnetic system is defined as the ratio of the ab- solute value of the Weiss constant and the observed ordering temperature from magnetic 66 ??1 (Oe ?molmol emu ) susceptibility:181 | ? f CW | = (3.3) TC We obtain a frustration parameter of approximately 8.7, indicating strong suppression the magnetic ordering temperature. The inset of Figure 3.4 displays the inverse susceptibility behavior with different applied fields; it is shown that the paramagnetic regime (> 100 K) does not change, but the deviation at approximately 82 K shows differing behavior with applied DC field. Empirically, in the frustrated regime (Tc < T < |?CW |) increasing field drives the system toward increasing antiferromagnetic fluctuations as the slope of ??1(T ) decreases. Our earlier work showed that the magnetic susceptibility of CoSe exhibited a fer- romagnetic transition at 10 K, but the discontinuity at 10 K was not a classic example of a ferromagnetic transition. In order to explore this, the temperature dependence of mag- netic susceptibility was measured at different fields to see how the transition was altered. Insets of Figure 3.5 shows the magnetic susceptibility from 10 K to 1.8 K at various ap- plied fields. At low fields, 0.01 T, the transition at 10 K is clear from the bifurcation of the zero-field cooled (ZFC) and field-cooled (FC). However, as the field is increased, the tran- sition temperature is suppressed until at high fields, > 2 T, there is a complete suppression of the ZFC-FC splitting indicative of complete suppression of long-range ferromagnetic ordering. The closing of the normal ZFC-FC splitting at the proposed ferromagnetic transition is a hallmark of spin glass behavior as opposed to classic ferromagnetism.137 Without a 67 FC (a) ZFC (b) (c) Fig. 3.5: Magnetic susceptibility of CoSe vs. temperature measured in various applied fields. The insets show the zoomed region close to the transition temperature; ZFC (Zero field- cooled) and FC (field-cooled) curves are shown by arrows which indicate the irreversibil- ity of the magnetic ordering in the system at low fields. The bifurcation of ZFC-FC curves at low applied field (a) = 0.01 T and (b) = 0.2 T is destroyed with high applied fields (c) = 1 T and (d) = 5 T turning the system into a paramagnetic state with no irreversibility. sufficiently applied field, spins are able to ?freeze" in the random orientation of spin glass yielding net magnetization opposing the applied field in the ZFC process. With a stronger field, the ?freezing" is destroyed as the spins are forced to align with the applied field. We can rule out superparamagnetism as a possible explanation as we have observed remanent magnetization and magnetic hysteresis for CoSe which would not occur in a superpara- magnetic material.61 In order to observe the glassy character in CoSe, we performed AC magnetic susceptibility measurements to probe the time-dependence of the magnetization around the transition temperature. AC-susceptibility measurements use an applied field with a time-dependent wave- 68 Normalized (a) 0.04 0.03 0.02 (b) 1e?3 100 Hz 2.4 1 kHz 10 kHz 1.8 1.2 0.6 0.0 0 5 10 15 20 T (K) (c) ?4 ?5 ?6 ?7 Vogel-Fulcher Fit ?8 6.0 6.2 6.4 6.6 6.8 7.0 T (K) Fig. 3.6: AC magnetic susceptibility measured with various driving frequencies. The applied AC field was 10 Oe and the residual DC applied field due to internal instrumentation was 40 Oe to 100 Oe. (a) The real parts of magnetic susceptibility (??) and (b) the imaginary parts (???) parts. (c) Temperature dependence of ?? peaks at various driving frequencies (100 to 1200 Hz) and a fit with the Volger-Fulcher law. form to produce a time-dependent response in the material. It can therefore probe spins fluctuating with time such as in spin glasses or strongly frustrated systems.137 Figure 69 ?00 ( emu ? 0 ( emu ) ln ? Oe ? ) Oe ?molmol 3.6 shows the real (??) and imaginary (???) parts of magnetic susceptibility as a function of temperature near the transition. Frequency dependence in ?? appears below 10 K, and accompanying non-zero peaks in ??? indicate some out-of-phase contributions to the mag- netic susceptibility. Thus, time-dependence in the magnetic domain size arises below 10 K, and any mangetic ordering appears dynamic down to base temperature. A fit to the non-zero ??? peaks with the Arrhenius law would be simple yet inad- equate for canonical spin glasses and spin-glass-like materials. The transition into the glassy state is more than a simple thermal activation process, and magnetic moments can also be strongly interacting.137 A more phenomenological approach that incorporates different regimes of coupling above and within the glassy state uses the Vogel-Fulcher law:137,139 ( ) ? = ?0 ? E exp a? (3.4)kB(T f T0) where T f is the temperature of the ??? peaks, ?0 = 1/?0 is the characteristic relaxation time, and kB the Boltzmann constant. The added parameter, T0, describes the ?ideal glass temperature? where the coupling of the system effectively changes to give rise to new phenomena.137,139,182 Our modelling of the AC susceptibility data with the Vogel-Fulcher law is shown in Figure 3.6c. The temperature values for T f were fit by Gaussian curves in the range from 5-12 K. The fit yields parameters: ?0 = 0.67 ? 1.61 s, Ea = 12.75 ? 10.47 K, and T0 = 8.74?0.89 K. The large degree of uncertainty in the relaxation time and activation energy comes from the high correlation between ?0 and T0 parameters and narrow temperature 70 range of the ??? peaks. The lack of meaningful values from the initial Vogel-Fulcher fit led us to perform additional analysis using a dynamical scaling model. Dynamical scaling relates the re- laxation time of an observable to a correlation length that scales with a power law near the transition temperature. We consider scaling of the frequency-dependent transition temperature from the ??? peaks such that:183 (T ? T )?zvc f ? = ?? (3.5) T f where Tc is the critical temperature, ?? the critical relaxation time, and zv the critical exponent. Our fit yields ?? = 0.064 ? 0.008 and zv = 5.47 ? 0.21, which fall into the general range of spin-glass and glassy-like materials183. Substituting the value of ?? for the the characterstic relaxation time in the Vogel- Fulcher law, we obtain more precise values for the activation energy (E?a = 14.84 ? 0.59 K) and ideal glass temperature (T ?0 = 8.91 ? 0.09 K). The obtained critical relaxation temperature is significantly higher than canonical spin-glass materials, but compatible with Monte-Carlo modeling of a 3D Ising spin glass.184,185 3.4.2 Transport Properties To further probe the dynamics of the transition within CoSe, we have employed more electronic transport measurements. Figure 3.7a shows the temperature dependence of electrical resistivity for CoSe. For a truly ordered material, one would anticipate a noticeable change in the resistivity near the critical point. However, no such anomaly 71 occurs in the resistivity measurements. This lack of an anomaly could be understood on the basis of weak ferromagnetism as the observed moment of CoSe via neutron diffraction is very small.61 We observe positive magnetoresistance for all applied field directions (Figure 3.7b), which does not occur in typical ferromagnets. The positive magnetoresistance can be interpreted in two ways: 1) the spins have no fixed direction and are randomly distributed as would be the case for a glass-like material, or 2) the spins are fixed but their associated moments are so small that their contribution to scattering is negligible. The complete angular dependence of the resistance versus field direction at 31 T (Figure 3.7c) shows two-fold symmetry, which is due to the geometry of the four-probe longitudinal measurements. Angular measurements in other planes are not possible due to sample morphology. CoSe crystals are highly layered and micaecous so that only allow the ab-plane is available as the wiring surface. We performed specific heat measurements from 1.8 - 150 K (Figure 3.9) on a pressed pellet of CoSe obtained through the potassium de-intercalation route. The mi- caecous nature of the single crystals caused poor thermal coupling between the sample and the heating platform, and we therefore utilized a pressed pellet of CoSe. For compar- ision, we also performed specific heat measurements of KCo2Se2 single crsytals, known from previous studies to exhibit a clear ferromagnetic transition below 80 K.164,186 Our own heat capacity measurements of KCo2Se2 confirm a clear transition at 78 K shown in Figure 3.8. The temperature dependence of the specific heat for CoSe, however, shows no anomaly near 78 K. The inset of Figure 3.9 shows a zoomed in region around the transition observed 72 1.0 0 T 2 T 4 T 8 T 0.5 0.132 0.130 5 10 0 100 200 300 Temperature (K) 8 -0.52? 34.76? 6 65.00? 90.20? 4 2 0 0 10 20 30 1e?6 Field (T)3.65 3.60 3.55 Exp. Data 3.50 Sine Fit 0 30 60 90 120 150 Angle (deg.) Fig. 3.7: Electrical transport measurements of CoSe single crystals obtained through de- intercalation of KCo2Se2. (a) Temperature dependence of longitudinal resistivity at var- ious applied fields with inset around the transition temperature. (b) Normalized longi- tudinal magnetoresistance up to 31 T with different applied field directions by sample rotation. (c) Angular dependence of longitudinal magnetoresistance at an applied field of 31 T. The magnetoresistance is fit with a sinusoidal dependence to the field angle. in magnetic susceptibility measurements with a fit to a conventional specific heat model. There is no apparent discontinuity in the specific heat in this region, which indicates a 73 Magnetoresistance (?) Resistivity (m? ? cm) Magnetoresistance (%) 100 80 75 70 65 70 80 90 50 0 0 20 40 60 80 100 T (K) Fig. 3.8: Specific heat as a function of temperature measured on a single crystal of KCo2Se2 mea- sured from 1.8 - 100 K with no applied field. Specific heat shows a distinct disconti- nuity at 79 K which corresponds to previous reports of the ferromagnetic transition in KCo2Se2; this is used to further prove the lack of KCo2Se2 phase within our CoSe sam- ples signifying that complete de-intercalation is achievable for this system. lack of a distinct phase transition. This result either supports a glass-like material,137,139 or that the magnetic ordering does not change the energy scale due to the low ordering moment of a weak itinerant ferromagnet.187,188 The low temperature region of the specific heat, T < 15 K, was fit to a general model to extract electronic and vibrational contributions.134 Cp = ?T + ?T 3 + cT 5 (3.6) where the ?-term accounts for electronic contributions and ?/c-terms for vibrational con- tributions. The fit yields a ? = 15.7 mJ mol?1 K?2, significantly larger than in the iron-based analogues FeSe and FeS (5.4 and 5.1 mJ mol?1 K?2, respectively).44,59 This could indicate stronger electron correlations in the cobalt system. However, recent angle- 74 C ( Jp )mol ?K Fig. 3.9: Temperature dependent specific heat of a pressed pellet of CoSe from 150 K to base temperature. Upper inset shows the temperature dependence near the transition temper- ature as well as a fit to a specific heat model accounting for electronic and vibrational components in the range 1.8 K to 15 K. resolved photoemission spectroscopy (ARPES) work on related KCo2Se2 indicated weaker electron correlations in the cobalt system than in the KFe2Se2 analogue.189 A possible ex- planation for the larger ? in CoSe than in FeSe is that it arises from spin fluctuations present in a weak intinerant ferromagnet.190,191 We can use the parameter ? = 6.2 ? 10?4 mJ mol?1 K?4 to calculate the Debye temperature, ?D, for CoSe by the relation:134 (12?4nR)1/3 ?D = (3.7)5? where R is the universal gas constant. This fit yields a ?D = 232 K. We added the T 5 75 term since the T 3 contribution is generally only applicable up to ? /50 = 4.6 K.192D The resulting c is ?5.9 ? 10?7 mJ mol?1 K?6, two orders of magnitude lower than the iron analogue. 3.5 Discussion 3.5.1 Ground state of CoSe Despite the structural simplicity of CoSe, its magnetic ground state is less straight- forward. Initial temperature dependence of magnetic susceptibility indicated a ferromag- netic transition at 10 K corroborated by powder neutron diffraction work.61 When consid- ering itinerant systems, it is often useful to evaluate Stoner?s criterion for ferromagnetism in the system where the enhanced susceptibility ?S is given by:193 ?P ?P ?S = 1 ? =F 1 ? (3.8)Dband(E f )Is/2 where Dband(E f ) is the density of states at the Fermi level, Is is the Stoner factor for Co (? 0.9 eV) divided by two to account for the two Co atoms per unit cell, and ?P is Pauil paramagnetic susceptibility. The denominator allows us to formulate the Stoner?s criterion such that F = Dband(E f )Is/2. We performed DFT calculations on CoSe which yielded Dband(E f ) = 7.33 states/eV for non-spin dependent calculations shown in Figure 3.10 Thus, this result leads to F = 3.29 > 1 which indicates that CoSe should have a ferromagnetic ground state. The structurally related KCo2Se2 exhibits a ferromagnetic transition at approxi- 76 12 Total DOS Co s Co p 10 Co d Se s 8 Se p Se d 6 4 2 0 6 5 4 3 2 1 0 1 2 Energy (eV) Fig. 3.10: Non-spin polarized density of states for CoSe decomposed for Co and Se atoms and their corresponding molecular orbitals. mately 78 K measured on single crystals. The previous work used anisotropic single crys- tal measurements of magnetization to show that the magnetic moment resides completely in the ab-plane for the CoSe layers.186 No neutron diffraction work has been reported to date on KCo2Se2, but the fairly large spacing between CoSe layers and anisotropic suscep- tibility indicates that the moment is likely in the ab-plane. When we remove the interlayer potassium ions and reduce the CoSe interlayer spacing from ? 6.92 ? in KCo2Se2 to 5.33 ? in CoSe, we can consider how these adjacent planes may begin to interact. Figure 3.4 showed that the magnetic susceptibility of CoSe displayed Curie-Weiss behavior above 100 K yielding a strongly negative Weiss constant, ?CW = ?87.29 K. Although CoSe is an itinerant electron system, we can minimally consider a square lattice Heisenberg model,174,194 for which similar models have been applied extensively to the FeSe system,195?198 to yield: 77 Density of States (DOS) ? (J1 + J2)?CW = (3.9)kB where J1 and J2 describe the nearest-neighbor and next-nearest-neighbor interactions on the square lattice, respectively. In this case we see that J1 + J2 = 87.29 K = 7.53 meV and that the exchanges should be antiferromagnetic based on the inverse susceptilibity data. At approximately 80 K, ??1(T ) increases its slope so that the Weiss field changes to a positive value, possibly indicative of increasingly ferromagnetic fluctuations in this lower temperature regime. Interestingly, specific heat, magnetization, AC susceptibility, and resistivity show no anomalous changes in near 80 K. Considering the ferromagnetic-like transition at 10 K shown in magnetic susceptibility measurements and the antiferromagnetic Weiss field at high temperature, we postulate that the ferromagnetic ordering at 78 K in KCo2Se2 is suppressed down to 10 K for CoSe. The suppressed ordering may arise from geometric frustration, vacancies on the Co sites, or competing interactions between magnetic Co2+ ions. In the case of CoSe, we can eliminate two of these possibilities: vacancy ordering and geometric frustration. Elemental analysis from previous work showed that the percentage of Co vacancies did not exceed 2%, within error of that amount. Not enough to significantly suppress order- ing. Geometric frustration occurs in systems where magnetic sublattices cannot arrange in a unique lowest energy ordered state, such as in an antiferromagnetic triangular lattice. CoSe contains a square lattice of cobalt atoms that cannot host this type of geometric frustration. Theoretical work, however, on square lattices have found frustration when 78 the nearest neighbor and next-nearest neighbor magnetic interactions compete. This has been termed interaction frustration.72,199 3.5.2 Anisotropy and Magnetic Direction Our previous results from powder neutron diffraction indicated that the magnetic moments are aligned along the c-axis, contrary to the ordering in the related ?122?- phase.61 A possible reason for a difference in moment direction between the two systems could be could be due to closer CoSe layers in the CoSe than in KCo2Se2. To under- stand the anisotropy present in the system, we performed single crystal magnetization measurements similar to the work done on KCo2Se2 by Yang et. al. In Figure 3.11a, we see that magnetic susceptibility in the ab-direction is about four to five times larger than the susceptibility in the c-direction. This suggests a fair amount of anisotropy, but not as large as in KCo2Se2, where there is an a order of magnitude difference between the two field directions.186 Unexpectedly, the anisotropy in the field dependence of the magnetization for KCo2Se2 did not hold for CoSe (Figure 3.11b). We see that for both field directions the magnetization does not saturate up to 14 T and ap- proaches a moment value of 0.1 ?B. The itinerant nature of the magnetism leads to an unsaturated magnetization. An important insight from these measurements is obtained by Arrott plot analysis. From Landau theory, we can expand the free energy, F(H,T,M), of a magnetic system in the order parameter, M, corresponding to magnetization: F = F0 + aM2 + bM4. Minimiz- ing the free energy with respect to magnetization we arrive at:193 79 (a) (b) (c) H || ab (d) H || c Fig. 3.11: Magnetic measurements of CoSe crystals mounted on a quartz paddle with orientations relative to the applied field direction as listed. a) Temperature dependent magnetic susceptibility for a 100 Oe field applied in two different orientations. b) Field dependent magnetization for both field orientations. c) Arrott plots constructed from M(H) curves from 2 - 12 K for H ? ab, which indicate a ferromagnetic transiton within the 8 - 10 K range. d) Arrott plots constructed for H ? c-axis showing no spontaneous magnetization in the c-direction for any temperature. 1( H ) a(2 T ? T )M (T,H) c= ? (3.10) b M(T,H) b Tc 80 We can plot M2 vs. H/M, known as an Arrott plot, to obtain linear relationships be- tween M2(H) curves at different set temperatures. From Equation 3.10, as the temperature approaches Tc, the M2(H) curves approach zero. Positive y-intercept values correspond to spontaneous magnetization at those temperatures. From Figure 3.11c for H ? ab-plane, the critical temperature appears to be in the 8 - 10 K range, as linear extrapolations of the M2(H) curves yield a zero y-intercept between 8 K and 10 K, which corresponds to transi- tion temperature in the powder measurement. However, for H ? c-axis (Figure 3.11d), no M2(H) curves yield positive extrapolations back to the y-axis. Therefore, no spontaneous ferromagnetic moment orders along with c-axis. The Arrott plot analysis matches previous reports for KCo2Se2, where the moment is claimed to lie solely in the ab-plane.186 However, what causes the difference both the in ordering temperature and strength of the ferromagnetism between the two systems? The removal of potassium ions between the layers affects a number of factors: 1) cobalt oxidizes from Co1.5+ in KCo2Se2 to Co2+ which means a removal of electron carriers, 2) CoSe layer distances are reduced from 6.92 ? in KCo2Se2 to 5.33 ? in CoSe which may cause more effective exchange between the moments in adjacent ab-planes, and 3) the Co-Co distance shrinks from 2.710(3) ? in KCo2Se2 to 2.6284(3) ? in CoSe, which causes more orbital overlap between Co centers. 3.5.3 FeSe vs. CoSe Currently, Fe and Co are the only transition metals that have been able to form the anti-PbO structure which is closely related to the parent ThCr2Si2 structure. The 81 ThCr2Si2 hosts over 1,500 structures and a wide-range of physical phenomena. The anti- PbO phases are structurally simpler and can be used as the building blocks to systemati- cally explore the physics within this structure type and, in general, metal square lattices.63 Unconventional superconductivity has emerged in the FeCh systems with the pair- ing mechanism for this phenomena still to be understood. With the close proximity of magnetism and superconductivity in the iron system, we need to understand the salient differences between CoSe and FeSe. Previous work directly compared the band structures of FeSe and CoSe and showed they differed by just a rigid band shift corresponding to the extra electron added by cobalt as compared to iron.61 This shift moved the Fermi level away from the nesting of hole and electron pockets evident in the FeSe superconductor, which could to be key to realizing superconductivity in this system. Since band structure measurements have yet to be conducted on CoSe, we can directly compare the results of recent studies on KCo2Se2 and AFe Se .200,201y 2 ARPES studies have shown that going from AFeySe2 to KCo2Se2 (i.e. electron charge doping) changes the 3d orbital that contributes the most at the Fermi level. ARPES work on the AFeySe2 series showed that the 3dxy orbitals contribute the most at the Fermi level. The Se 4pz orbitals also contribute to allow superexchange interactions. However, for KCo2Se2 the most significant orbital is the 3dx2?y2 which would change the interactions allowed between adjacent Co atoms.189 This change in geometry of the d-orbital likely is the mechanism for tuning away superconductivity to frustrated magnetism in CoSe. Extensive work has been performed to understand the magnetic fluctuations in FeSe which are integral in understanding the mechanism responsible for superconductivity in the iron-based superconductors. As previously stated, the interesting interplay of mag- 82 netism in this system seems to stem from the electronic instabilities that accompany the square lattice formation.63 Recent inelastic neutron diffraction work and theoretical work has shown that within the FeSe layers there is strong frustration between different mag- netic ordered states (stripe vs. N?el), which causes FeSe to not exhibit a true long-range magnetically order state.52,197,198 The magnetic ordering in CoSe appears to suffer from similar frustration via the square lattice motif, although single crystal inelastic neutron spectroscopy measurements would shed further light on this hypothesis. 3.6 Conclusion The synthesis of isostructural CoSe has allowed extensive characterization of the magnetic and transport properties of the system to understand its proximity to the iron- based superconducting analogues. Magnetic measurements have shown a transition rem- iniscent of ferromagnetism at 10 K with low applied fields that is fully suppressed at high fields. AC-susceptibility shows non-zero out-of-phase contributions, and such time dissi- pative magnetization below 10 K is indicative of a spin glass. Our more detailed analysis of the AC-susceptibility matches the behavior of CoSe to a spin glass, and we a possible explanation is the physics of interaction frustration present in square lattices. Our Arrott plot analysis of the magnetization data reveals that the moment in CoSe lies within the ab-plane much like in related KCo2Se2. However, even if these two sys- tems have similar anisotropy, the transition temperature is vastly different, having been suppressed from 78 K to 10 K in CoSe. Therefore, the amount of electron doping and density of states at the Fermi level can be used to tune the magnetic interactions in the Co 83 square sublattice. Resistivity measurements indicate a metallic state in CoSe with no significant anisotropic magnetoresistance and no discontinuity at the 10 K transition. Heat capacity measure- ments indicate no observable transition at 10 K either, but low temperature analysis re- veals an enhanced Sommerfeld coefficient due to strong spin fluctuations at low tempera- tures. The lack of a discernable transition within transport measurements further corrob- orates the glassy character at low temperatures due to interaction frustration. Comparing CoSe to FeSe, we now see that the nature of the d-orbital occupany near the Fermi level vastly tunes the ground state from a metal with weak and competing magnetic interactions (CoSe) to a superconductor (FeSe). Future work on the CoSe system includes inelastic neutron spectroscopy to shed further light on the nature of the exchange interactions leading to interaction frustration. Chemical manipulation to charge dope CoSe would also be an important step in further expanding the phase diagram of these metal square lattices. There has been some previous cobalt doping studies on FeSe but the amount of substitution on cobalt has been limited to less than 20% due to phase stability with increased cobalt content.157,158 However, the topochemical de-intercalation route should be able to expand the solid solution of cobalt- doped FeSe available to directly observe how superconductivity evolves into frustrated magnetism. 84 Chapter 4: Long-range magnetic order in transition metal doped (LiOH)FeSe by soft chemical design The work described within this chapter was submitted to in Physical Review Ma- terials Brandon Wilfong, Xiuquan Zhou, Navneeth Babra, Huafei Zheng, Johnpierre Paglione, Craig M. Brown, Jeffery Lynn, Keith M. Taddei and Efrain Rodriguez were contributing authors of the manuscript. B.W., X.Z., H.Z., and N.B. prepared the samples, B.W. and X.Z. performed MPMS measurements, B.W. collected resistivity data, X.Z., B.W., C.M.B., K.M.T, and J.W.L. collected the neutron powder diffraction data. 4.1 Introduction The binary FeSe, with a superconducting critical temperature (Tc) of 8 K,42 pro- vides an excellent template to study exotic physical phenomena in iron-based supercon- ductors due to its simple structure, ease of chemical manipulation and relatively high superconducting critical temperature. Amazingly, the Tc of FeSe can be improved signifi- cantly to 42-46 K from 8 K after intercalation,46?48 37 K with applied pressure,49 or 65 K in the monolayer limit.50 The well-studied intercalated compound (Li1?xFexOH)FeSe (Tc = 42-44 K) consists of a tetragonal layer of partially charged Li1?xFexOH (x ? 0.1-0.2) between the FeSe layers. Such a structure is stabilized by the partial charge transfer due to 85 Fe doping on the Li site as well as hydrogen bonding from the LiOH layer to the Se atoms in the FeSe layers.47,202,203 The Fe substitution in the insulating hydroxide layer not only plays a crucial role in the enhancement of Tc, but also can induce exotic physical phenom- ena such as coexistence of magnetic order and superconductivity.80,81,124 Although many reports have shown signatures of magnetic order in this compound, there are no definitive signs to understand the true nature of its magnetism. The magnetic transition is intrinsic to the system with different reports attributing this transition to ferromagnetic ordering, canted antiferromagnetic ordering, and spin glassiness.80,81,124,204,205 Despite their differ- ences, all of these studies point toward the hydroxide layer as the source of the magnetic ordering. K Fe (Li Fe Mn0.85 1.8-xMnxSe2 1-x-y x yOD)FeSe (FeSe)?- Mn 120 oC (Li 3 days 1-x-y Fex MnyOD)?+ (FeSe)?- Fig. 4.1: Synthetic scheme and results of targeted (Li1?x?yFexMnyOD)FeSe by converting Mn- doped KFe1.8?zMnzSe2 hydrothermally. In contrast to the uncertainties in (Li1?xFexOH)FeSe, there is a history of the obser- vation of the coexistence of magnetism and superconductivity, mostly in rare-earth ele- ment containing compounds. The first examples of long-range magnetic order coexisting with superconductivity was observed in the ternary Chevrel phases RMo6S8 and RRh4B4 86 (R = lanthanide) where magnetic order arose from the lanthanide ion sublattice which was isolated from the superconducting sublattice. In these systems, both antiferromag- netism and ferromagnetism was observed coexisting with superconductivity depending on the lanthanide chosen for the magnetic sublattice.206?209 The stabilization of magnetic order in these compounds was explained by dipolar electromagnetic interactions as the magnetic transition temperatures were below 1 K. Akin to these compounds were the later discovered borocarbides RNi2B2C where the magnetic ion sublattice now exhibits significant R-R exchange interactions pushing the magnetic phase transition much higher than in the Chevrel phases.210?212 The discovery of superconductivity in the cuprate fam- ily offered a new avenue for exploration. In these compounds, magnetic ion sublattices isolated from the CuO planes show low ordering temperature similar to the previously mentioned systems which can coexist with superconductivity. However, interestingly in these compounds, Cu ions (S = 1/2) of the CuO planes exhibited antiferromagnetic insulator behavior, but this antiferromagnetic order could be suppressed to induce su- perconductivity upon doping.27,213?215 The iron pnictide superconducting systems exhibit similar phenomena to the cuprate systems whereby magnetic order can coexist with su- perconductivity through an isolated magnetic ion sublattice and superconductivity arises from suppressing magnetic order of the iron pnictide layers through doping.27,216?220 Un- like the related iron pnictide phases, FeSe exhibits no parent magnetic phase;42,51 recent work has shown a high temperature nematic phase that precedes the superconducting phase.221?223 Although no parent magnetically ordered phase exists, strong magnetic fluc- tuations have been observed in a wide range of temperatures in FeSe through neutron and NMR spectroscopy experiments.51?54 Thus, the introduction of a magnetic spacer layer 87 between FeSe may help to further reveal the role magnetic fluctuations play in supercon- ductivity in the Fe-chalcogenide systems. In addition to its interesting magnetism, experimental evidence has shown that (Li1?xFexOH)FeSe exhibits a Majorana Zero Mode (MZM) which plays a critical role in topological quantum computing applications.224,225 Therefore, (Li1?xFexOH)FeSe is a perfect system to study and understand the coexistence of physical phenomena for possi- ble applications in functional materials as well as quantum computing. Since all the aforementioned exotic phenomena emerge from the interactions be- tween the superconducting FeSe?? and the insulating (Li Fe OH)?+1?x x layers, one may suggest to modify the hydroxide layer to induce new properties. Unfortunately, traditional solid-state reactions and methods will be insufficient to modify the solid solutions as these phases are metastable. Thus, we have developed a two-stage ion-exchange scheme to obtain (Li1?x?yFexMnyOD)FeSe by converting Mn-doped K0.85Fe1.8?zMnzSe2 hydrother- mally. We manipulate the chemistry of the hydroxide layer in (Li1?xFexOD)FeSe through manganese doping to tune the magnetic properties without significantly altering the su- perconductivity in the system (Fig. 4.1). The addition of manganese to supplant Li and Fe in the tetrahedral hydroxide layer increases the effective spin and therefore the effective moment of layer to observe long range magnetic order. 88 4.2 Methods 4.2.1 Synthesis The synthesis of (Li1?x?yFexMnyOD)FeSe (M = Mn, Co, Ni, Cu, Zn) single crystals was perfomed via a two-step ion-exchange route similar to previous works.79,82,226 High purity metallic K (Alfa Aesar, 99.5%), M granules (Alfa Aesar, 99.98%), Fe granules (Alfa Aesar, 99.98%), and Se shots (Alfa Aesar, 99.999%) were used as raw materials. In order to incorporate transition metal doping, single crystals of K0.85Fe1.8?zMnzSe2 were synthesized through a self-flux route with elemental mixture of K:Fe:Mn:Se in two nominal ratios of 0.85:(1.62/1.44):(0.18/0.36):2 were mixed in an argon glovebox sealed under vacuum in a double quartz ampoule. The quartz tubes were slowly heated at 50 ?C/hr to 1050 ?C, held at 1050 ?C to ensure a congruent melt, slowly cooled down to 550 ?C at a rate of 5 ?C/hr, and ended with natural cooling to room temperature. This method routinely produced 3 x 3 mm2 plate-like single crystals. The hydrothermal ion- exchange were performed in 100 mL stainless steel autoclaves lined with Teflon cups. For each batch, select single crystals of K0.85Fe1.8?zMnzSe2 with average total mass of approximately 2 grams, 1.2 mmol of selenourea (Sigma Aldritch, 98%), 13 mmols of iron granules (Alfa Aesar, 99.98%), 1.2 mmol of tin granules (used to regulate pH), 0.31 moles LiOH (anhydrous, Alfa Aesar, 98%), and 40 mL of D2O (Cambridge Isotope, 99.9%) were loaded into the autoclave and purged under argon flow for 2 minutes before being tightly sealed. Each autoclave was heated to 120 ?C and held for 72 hours in a convection oven. Silver plate-like single crystals were recovered by washing away excess powders 89 with D2O, excess iron granules were easily recovered with a magnetic bar. Crystals were dried under vacuum overnight and stored in an Ar filled glovebox. 4.2.2 Magnetic and transport measurements All magnetic property measurements were carried out using a Quantum Design Magnetic Susceptibility Measurement System (MPMS-3) on powders and single crys- tals of (Li1?x?yFexMnyOD)FeSe samples. Zero-field-cooled (ZFC) and field-cooled (FC) measurements were taken from 1.8 to 300 K with various applied direct current (DC) magnetic fields. Isothermal magnetization measurements were taken from H = ? 7 T at numerous temperatures to probe the magnetic and superconducting state. Electrical transport measurements were performed on a 9 T Quantum Design Physical Property Measurement System (PPMS-9T) with temperatures from 1.8 to 300 K and fields up to 9 T using a four-probe technique with current applied across the ab-plane due to the lamel- lar nature of the single crystals. Heat transport measurements were performed on a 14 T Quantum Design Physical Property Measurement System (PPMS-14T) over the range of 1.8 to 60 K using the relaxation technique.143,179,180 4.2.3 X-ray diffraction measurements Laboratory powder x-ray diffraction (PXRD) was collected using a Bruker D8 X- ray diffractometer with Cu K? radiation (? = 1.5406 ?, step size = 0.020?, 2? = 5 - 90?) for phase identification. In order to aid in structural refinements associated with the (Li1?x?yFexMnyOD)FeSe system, as three different elements occupy the same crystallo- 90 graphic site, high resolution synchrotron X-ray diffraction was performed on powders of ground single crystals at Beamline 11-BM at the Advanced Photon Source at Argonne National Lab. Ground powders of single crystals were packed in 0.4 mm Kapton capil- lary tubes and sealed with epoxy. Diffraction data was collected between 0.5? and 46? with a step size of 0.0001? using a constant wavelength ? = 0.413964 ?(30 keV) at 100 K. Rietveld refinements were performed using the TOPAS software suite.126 4.2.4 Neutron diffraction measurements Neutron powder diffraction (NPD) data was collected at the NIST Center for Neu- tron Research (NCNR) BT-1 High Resolution Powder Diffractometer and Oak Ridge Na- tional Lab High Flux Isotope Reactor (HFIR) HB-2A.227 Powder samples of ground sin- gle crystals of (Li1?x?yFexMnyOD)FeSe were loaded into vanadium cans under helium exchange gas and loaded into a closed-cycle refrigerator (BT-1) or three-sample changer in 70mm Orange Cryostat (HB-2A). Low temperature diffraction data was collected at 2 K and 9 K and high temperatures at 50 K, for direct comparison in attempts to find mag- netic satellite reflections, with Cu(311) (? = 1.54 ?) at BT-1 and Ge(115) (? = 1.54 ?) at HB-2A. Rietveld refinements were performed using the TOPAS software suite.126 In order to search for weak magnetic reflection, high intensity but coarse resolution diffrac- tion measurements were performed on single crystals and ground single crystal powder on the BT-7 spectrometer at the NCNR using the position sensitive detector (PSD) with PG (002) (? = 2.359 ?) in a range of temperature from 2 - 50 K to search for magnetic transitions inside the superconducting regime. 91 (a) SG Smoothed BT-7 PSD Gaussian Fitting ? = 2.359 ? Raw data (3 K) 10% Mn-doped(LiOD)FeSe Magnetic (002) 1.25 1.30 1.35 1.40 1.45 1.50 Q (??1) (Li0: 875Fe0: 062Mn0: 062OD)FeSe 1.1 (b) 2 K 1.0 HB-2A 50 K 2 K - 50 K 0.9? = 1.54 ? 0.8 (002)Magnetic 0.7 0.6 0.5 (c) 0.4 0.3 1.2 1.3 1.4 1.5 1.6 0 5 10 15 20 25 30 35 40 45 50 Q (??1) T (K) Fig. 4.2: Powder neutron diffraction data collected at BT-7 and HB-2A. a) Powder neutron diffrac- tion data collected at 3 K at BT-7 using the PSD on a nominally 10% Mn doped Li- ODFeSe sample of the (002) reflection showing a clear well-resolved magnetic peak proximate to the (002) reflection. This data smoothed using a Savitzsky-Golay filter and subsequently fit using a two-Gaussian model to extract temperature dependence. b) Pow- der neutron diffraction data on (Li0.875(2)Fe0.062(3)Mn0.062(3)OD)FeSe at HB-2A showing the (002) reflection and magnetic peak at 2 K which is absent in the 50 K data. c) Tem- perature dependence on the integrated intensity of the two-Gaussian model on the data presented in a) showing a sharp decrease in the magnetic peak integrated intensity up to 10 K and temperature independence above 10 K. Uncertainties represent one standard deviation. 4.3 Evidence for long range magnetic order To date, all previous powder neutron diffraction data on (LiOD)FeSe samples made via the bottom-up synthesis and ion-exchange method do not exhibit observable long- 92 Intensity (a.u.) Intensity (a.u.) Normalized Intensity range magnetic order.48,80,82,124,228 Small angle neutron scattering did reveal a character- istic ferromagnetic scattering below 12 K with a clear vortex scattering peak observed at Q = 0.0077 ??1 under a 0.4 T applied field.81 Two more recent works on single crystals of the (LiOD)FeSe system via the ion-exchange method used inelastic neutron spectroscopy to investigate spin excitations below the superconducting temperature to reveal their im- portance in driving high critical temperatures in these compounds.226,229 We have performed neutron powder diffraction on Mn-doped (LiOD)FeSe to ob- serve long-range magnetic order. The results are shown in Figure 4.2 which span the use of two diffraction instruments and two different compositions of the Mn-doped (LiOD)FeSe system. To start, Figure 4.2b shows the low Q data accessible from the powder neutron diffraction data collected at HB-2A on Mn-doped (LiOD)FeSe and presented in Figure 4.3b/c in full. At Q = 1.41 ??1, d = 4.45 ?, a pronounced peak is observed at 2 K which is absent at 50 K; this peak is proximate to the nuclear (002) reflection but it was well re- solved. This is the first time a satellite reflection has been observed in neutron diffraction for the (LiOD)FeSe system indicating the existence of some form of long-range order. We attribute this to the larger magnetic moment of Mn2+ (S = 5/2) compared to Fe2+ (S = 2). As the magnetic peak appears distinct from the nuclear peaks, we can conclude the ordering in the system must be of the antiferromagnetic or ferrimagnetic type. Un- fortunately, we observe only one magnetic peak in the data, most likely due to the small magnetic moment of the system or the small magnetic form factor in this compound. The full magnetic structure cannot be solved currently. In order to further understand how this magnetic peak behaved as a function of temperature, additional powder neutron diffraction data was collected on ground single 93 crystals of nominally 10% Mn-doped (LiOD)FeSe samples at BT-7 using the PSD. For these measurements, the (002) reflection was identified and measured as a function of temperature with a ?0.2 ??1 window of collection using the PSD. Thus, the satellite re- flection proximate to the (002) reflection observed in HB-2A data could be recorded con- currently. The raw data collected using the PSD at 3 K is shown in Figure 4.2a displayed with a two-Gaussian model fit to the smoothed data using a third-order Savitzsky-Golay filter due to the coarse resolution of the raw data. The raw data and two-Gaussian fit show two well resolved peaks, one for the (002) reflection at Q = 1.37 ??1 and one for the magnetic peak at Q = 1.44 ??1, in close agreement to the HB-2A data. The slight discrepancy is likely due to crystallographic differences in the compounds. Figure 4.2c shows the temperature evolution of the normalized integrated intensities calculated by the two-Gaussian model corresponding to the (002) reflection and magnetic peak. It is observed that the (002) reflection is temperature independent while the magnetic peak shows a sharp decrease in integrated intensity above 10 K in close agreement to the or- dering temperature observed in magnetization and heat capacity data. We can conclude that the long-range magnetic ordering observed is intrinsic to the system with an ordering temperature around 9 K and has been observed for the first time by targeted design of the hydroxide layer in (LiOD)FeSe. Futhermore, in the Mn-doped (LiOD)FeSe we have observed a high superconduct- ing critical temperature paired with a relatively high magnetic transition temperature com- pared to rare-earth containing phases. This arises through similar means to the previously mentioned Chevrel, borocarbide, copper and iron-based systems whereby the magnetic sublattice and superconducting sublattice are isolated from another. However, we have 94 done so without the need for rare-earth ions. Interestingly, in this system the addition of transition metal doping in the hydroxide layer both induces magnetic order as well as charge dopes the FeSe layers significantly raising the critical temperature of the com- pound. Thus, transition metal doped (LiOH)FeSe offers a tremendous platform for explo- ration of the role magnetism plays in stabilizing high temperature superconducitivity in the iron chalcogenides and for potential use in next generation devices. 4.4 Hydrothermal synthesis and crystallographic results As shown in Fig.4.1, we first prepare precursors of K0.85Fe1.8?zMnzSe2 with 10% and 20% nominal Mn doping level using direct elemental reactions from high-temperature. Our XRD analysis of K0.85Fe1.8?zMnzSe2 show significant different lattice constants com- pared to K0.85Fe1.8Se2, indicating replacement of Fe in the FeSe layer, presented in Table 4.1 10% Nominal Doping 20% Nominal Doping Transition Metal a lattice parameter (?) c lattice parameter (?) a lattice parameter (?) c lattice parameter (?) Mn 3.8963 14.1434 3.8937 14.1513 Fe 3.8341 14.2360 3.8341 14.2360 Co 3.8814 14.1640 3.8892 14.1088 Ni 3.8817 14.0735 3.8790 14.1187 Cu 3.9001 14.0957 3.9091 14.1177 Zn 3.8939 14.0932 3.9061 14.0323 Tab. 4.1: Refined lattice parameters of K0.85Fe1.8?xMxSe2 M = (Mn, Co, Ni, Cu, Zn) starting ma- terials made via a high temperature self flux reactions from the corresponding elements. The 10% and 20% nominal doping come from the stoichiometric additons of each ele- ment to the growth. The reported lattice parameters for the Fe (un-doped) version come from a similarly targeted stoichiometric compound of K0.85Fe1.9Se2 via Shoemaker, D.P., et al Phys. Rev. B 86.18 (2012): 184511. Obtained small single crystals of these precursors are used to exchange K with LiOH hydrothermally as described in our earlier work82 For such conversion, it is crucial to include additional Fe powders and Sn metal plates in the autoclave to avoid formation 95 (a) * (b) (c) Fig. 4.3: a) High resolution synchrotron PXRD patterns for (Li1?x?yFexMnyOD)FeSe collected at room temperature, b) and c) show NPD data for (Li1?x?yFexMnyOD)FeSe at 2 K and 50 K respectively. Green tick marks represent the targeted tetragonal layered phase and orange tick marks represent impurity selenium. These are shown below the calculated, observed, and differences curves from Rietveld analysis. The asterisk represent a single observable impurity peak that does not match any observable peak in the NPD data. 96 of iron oxides and complete destruction of superconductivity. In this scheme, because the transition metal dopant, M, prefers covalent coordination with OH? group, they are driven from the FeSe layer of the K0.85Fe1.8?zMzSe2 to replace Li in the LiOH layer.48,230 This is because the hard Lewis acid, Mn2+, prefers coordination with the hard Lewis base of OH? as opposed to the soft base S2?; this prevents the formation of impurity manganese sulfides. The vacancies in the FeSe layer are then supplemented by Fe2+ from the solution. After the hydrothermal conversion, we obtain the products in both powder and single crystal forms. After synthesis, we needed to determine the precise composition and crystallo- graphic occupancy of all dopants. There has been multiple works to date on similar transition metal doping in the (LiOH)FeSe system, but none have been able to quan- tify the crystallographic location of the transition metal dopant.231?234 The location of the transition metal dopant is incredibly important in the realization of different physi- cal properties as superconductivity in the iron chalcogenide systems is very sensitive to doping.44,48,230 To identify the crystallographic location of the Mn doping, we use exten- sive x-ray and neutron diffraction. Since, X-ray form factors of Mn are very close to Fe, it is extremely difficult to obtain their site occupancy using regular X-ray diffraction whereas elemental analysis methods are not able to tell where Mn is located nor whether they are from impurities. Therefore, high-resolution synchrotron powder x-ray diffrac- tion (PXRD) has been paired with neutron powder diffraction (NPD) to fully understand the composition of these materials. Fig. 4.3 shows powder x-ray diffraction and powder neutron diffraction data for (Li1?x?yFexMnyOD)FeSe. High resolution synchrotron PXRD allows for high quality Rietveld refinements of crystallographic structural parameters, Fe 97 vacancies in the FeSe layer and total transition metal (M? = Fe + Mn) occupancies in the LiOH layer; however it does not allow for discrimination of transition metal dopants on the same crystallographic site. Thus, NPD complements synchrotron PXRD for high contrast between Fe and other transition metals occupying the same site. Figure 4.3a shows powder x-ray diffraction data and corresponding Rietveld fit for (Li1?x?yFexMnyOD)FeSe. Refinements with PXRD were limited to occupation of total transition metal (M?) doping in the hydroxide layer as well as total transition metal in the FeSe layer yielding a formula: (Li ?1?xMxOD)MySe. Rietveld refinements of the powder x-ray diffraction yielded a composition of (Li ?0.875(2)M0.125(2)OD)MSe with lattice param- eters a = 3.8008(1) ? and c = 9.2394(2) ? which is in close agreement with previous works.47,48,81,82 (Li0.875(2)Fe0.062(3)Mn0.062(3)OD)FeSe (298 K, PXRD), P4/nmm, Rwp = 8.506% a = 3.8008(1) ?,c = 9.23949(2) ? atom Site x y z Occ. Uiso (?2) Li/Fe1/Mn1 2b 0 0.5 0 0.875(2)/0.062(3)/0.062(3) 0.0246(2) Fe2 2a 0.5 0.5 0 0.993(4) 0.0252(2) Se1 2c 0 0.5 0.1603(1) 1 0.011(4) O1 2c 0.5 0 0.4249(4) 1 0.002(6) Fe-Se (?) Se-Fe-Se (?) Se-Fe-Se (?) Fe-Fe (?) 2.409(8) 112.21(6) 104.11(1) 2.687(8) (Li0.875(2)Fe0.062(3)Mn0.062(3)OD)FeSe (2 K, NPD), P4/nmm, Rwp = 5.380% a = 3.7887(1) ?,c = 9.1480(6) ? atom Site x y z Occ. U (?2iso ) Li/Fe1/Mn1 2b 0 0.5 0 0.875(2)/0.062(3)/0.062(3) 0.012(3) Fe2 2a 0.5 0.5 0 0.993(4) 0.0041(4) Se1 2c 0 0.5 0.1624(2) 1 0.0031(5) O1 2c 0.5 0 0.4271(4) 1 0.0147(9) D1 2c 0.5 0 0.3267(8) 1 0.004(1) Fe-Se (?) Se-Fe-Se (?) Se-Fe-Se (?) Fe-Fe (?) 2.407(5) 112.38(6) 103.78(6) 2.679(1) Tab. 4.2: Structural parameters for ground single crystals of (Li0.875(2)Fe0.062(3)Mn0.062(3)OD)FeSe. Structures are for room temperature PXRD data and 3 K NPD data. All relevant bond angles and distances from the refinements are given. Standard uncertainties given in parantheses indicate one standard deviation. 98 These refined structural and occupation values were then used as the starting model for the NPD. As mentioned, NPD was pivotal in determining the different transition metal dopants in the hydroxide layer and FeSe layers. Figure 4.3b,c show the NPD and corre- sponding Rietveld analysis for (Li0.875(2)M0.125(2)OD)MSe at 2 and 50 K respectively. Sub- sequent Rietveld analysis yielded a composition of (Li0.875(2)Fe0.062(3)Mn0.062(3)OD)FeSe, full results shown in Table 4.2. Interestingly, it was found that all the manganese dopant migrates to the hydroxide layer with no refineable amount of manganese in the FeSe layer. It is understood that dopants in the FeSe layer destroys superconductivity,48,82 so this reaction scheme allows for the superconductivity to be marginally altered while the magnetic layer can be manipulated. 4.5 Magnetic and transport properties We have measured magnetic susceptibility at various applied fields and isothermal magnetization at 2 K and 50 K on a single crystal of (Li0.875(2)Fe0.062(3)Mn0.062(3)OD)FeSe. Figure 4.4a/b show temperature dependent magnetic susceptibility, corrected for demag- netization factors of a two-dimensional plate,235 with applied field applied parallel to the crystallographic ab-plane and c-axis respectively. Interestingly, the superconducting on- set temperature is 25 K which is significantly lower than in the un-doped system. The highest observed shielding fraction after correction for demagnetization factors is ? 25% which may indicate non-bulk superconductivity in this system. Within the (LiOH)FeSe system, many factors that affect the observable superconducting temperature: lattice pa- rameters, transition metal concentration in the hydroxide and selenide layers, and struc- 99 (Li0. 875Fe0. 062Mn0. 062OD)FeSe 0.05 1.0 (a) H ab-plane (c) H ab-plane H= 30 Oe 0.00 0.5 2 K 0.06 50 K 300 Oe 1 T 0.05 0.04 0.0 0.02 0.2 0.10 0.5 0.0 0.00 0 20 40 0 20 40 0.2 1.0 0.5 0.0 0.5 (b) H c-axis H c-axisH= 30 Oe (d) 0.0 0.5 2 K 50 K 0.1 0.00 0.0 0.2 0.01 300 Oe 1 T 0.5 0.0 0.2 0 20 40 0 20 40 0.2 1.0 0.5 0.0 0.5 0 20 40 60 5 0 5 T (K) H (T) Fig. 4.4: Magnetic property measurements on an aligned single crystal of (Li0.875(2)Fe0.062(3)Mn0.062(3)OD)FeSe. a)/b) Magnetic susceptibility data at various applied fields, aligned parallel to the crystallographic ab-plane and c-axis respectively, showing a superconducting transition at 25 K but no second magnetic transition at lower temperature. c)/d) Isothermal magnetization at 2 K and 50 K, with the same orientation convention, showing magnetic hysteretic behavior superimposed on a superconducting signal at 2 K and paramagnetic behavior at 50 K. Note: (1 Oe = (1000/4?) A/m) tural homogeneity.48,82,203 Unlike previous reports,80,124 no magnetic transition is observed in the magnetic susceptibility even at higher applied fields. Interestingly, the only reports to show a tran- sition in magnetic susceptibility below the superconducting transition are on powder sam- ples via hydrothermal method.80,82,124,236 Single crystals samples from the ion-exchange hydrothermal method fail to show the secondary magnetic transition below the observed 100 4?? (dimensionless) M (?B) superconducting transition.82,203,237?240 We hypothesize that sample uniformity in doping and structure caused by the different reaction conditions is likely the cause of these prop- erty differences. In the bottom-up synthesis, FeSe and LiOH layers are formed in-situ contrary to the ion-exchange method which requires the removal of interlayer ions before replacing with LiOH extended solid which would lead to more structural and composi- tional variation. We observe strong anisostropy in the magnetic susceptibility. Fields applied parallel to the crystallographic ab-plane have a much stronger effect on suppressing the supercon- ducting transition than fields applied parallel to the c-axis as shown in the insets of Figure 4.4a/b. This is consistent with previous reports.79,203 and the observed transition is sup- pressed at relatively low fields, 300 Oe. The supression is due to the superconducting irreversibility temperature which is unique in these samples due to the very large vortex liquid region caused by extremely high anisotropy and two-dimensionality.238 We see an effect of underlying magnetic order is observable in isothermal magneti- zation measurements shown in Figure 4.4c/d at 2 K and 50 K. For both field orientations, magnetization as a function of fields shows a superconducting signal superimposed on a weak magnetic background. Mirroring the behavior of magnetic susceptibility, the Meiss- ner shielding is more clearly observable at low fields for the field applied parallel to the c-axis as compared to field applied parallel to the ab-plane. Again, this fast suppression of Meissner shielding could be due to true anisotropy, the filamentary nature of the su- perconductivity, or the effects of vortex dynamics in the sample. At higher temperatures, 50 K, the magnetic and superconducting signals are absent and paramagnetic behavior is observed. This paired with x-ray and neutron diffraction data shows that the magnetic 101 signal is intrinsic to the system. Electrical and thermal transport measurements shed more light on the superconduct- ing and magnetic order in this system. Temperature dependence of electrical resistivity on a single crystal of (Li0.875(2)Fe0.062(3)Mn0.062(3)OD)FeSe, shown in Figure 4.5a, shows a sharp superconducting transition at 41 K in zero field. With field applied parallel to the crystallographic c-axis, it is observed that the superconducting transition onset is not changed, but the transition width in greatly increased at higher applied fields. This tran- sition width broadening matches previous reports for this system. Specifically where this behavior has been justified by strong vortex flow as well as a wide temperature range for vortex liquid behavior due to anisotropy and two-dimensionality.79,203,238,240?243 The superconducting onset temperature measured by resistivity is much higher than in the same crystal measured in magnetic susceptibility. This phenomena was ob- served in previous works as well and can be attributed to the vortex dynamics in the system.203,239,240,243 The residual resistivity ratio ?(300K)/?(Tc) for the sample presented is ? 25 which is higher than in previous reports.203,240?242 The observed normal state be- havior is non-linear which has been demonstrated in previous works.203,240?242 and has been attributed to over-doping in the sulfide analogue.87 Due to the very high critical field in these samples,79, heat capacity measurements were employed to evaluate the effect magnetic ordering has within the superconducting regime. Heat capacity measurements on this system fail to show anomalous behavior around the superconducting transition.124,204,244 Temperature dependence of heat capacity at various applied fields shown in Figure 4.5b shows similar behavior. No anomaly is observed at the superconducting transition temperature, however; a magnetic transition 102 is observed at ? 9 K in zero applied field parallel to the crystallographic c-axis. In these previous reports,124,204,244 the magnetic transition was suppressed at high applied fields which was justified by antiferromagnetic or spin-glass ordering in the system. However, these measurements were done on samples made via the bottom-up hydrothermal syn- thesis, which behave differently than those produced via the ion-exchange method. To that end, the magnetic transition shown in Figure 4.5b shows slight temperature suppres- sion at 14 T, but is still clearly observable. We can conclude that the magnetic order in this system is of an antiferromagnetic or ferrimagnetic nature when combined with the observation of satellite reflection in neutron diffraction. To calculate the magnetic contribution to the heat capacity at low temperatures, the temperature dependent Cp/T data at zero applied field was fit to a third order polynomial above 20 K. This fitting was then extrapolated to low temperatures and subsequently sub- tracted from the raw data below 20 K. The remainder was then integrated to yield ?S as shown in Figure 4.5c as a function of temperature. The change in entropy sharply increases up to 12 K before leveling off at a value of 0.38(8) J/(mol?K), which is signif- icantly lower than the expectation for a free electron spin. The entropy change through a magnetic transition is defined as ?S = cRln(2J + 1), where c is the concentration of magnetic ions in the system and R is the ideal gas constant. Our compositional analysis of single crystals of (Li0.875(2)Fe0.062(3)Mn0.062(3)OD)FeSe showed a concentration of mag- netic ions in the hydroxide layer as c = 0.125, yielding an effective J = 0.22(6) for the system and effective moment ?eff = 1.05(2) ?B, spin-only. The expected spin values for Fe2+ and Mn2+ in tetrahedral coordination (quenched orbital angular momentum) are 2 and 5/2 respectively. These values are much larger than 103 the observed 0.22(6) from entropy measurements. Previous reports observed similar be- havior and concluded that this spin underestimation was due to spin-glassy character of the magnetism in the hydroxide layer.204 The robustness of the observed transition up to 14 T trends away from a spin-glass description for the magnetic behavior but spin glassi- ness cannot be ruled out.124,139,204 In-depth analysis at the atomic level through imaging or spectroscopy would be useful to understand how the dilute magnetic ions in the hydroxide layer order to from long range correlations. 4.6 Effect of other transition metal dopants We have been able to synthesize nominally 10 and 20% transition metal doped (LiOD)FeSe using the same two-step soft chemistry hydrothermal reaction scheme de- scribed above for the Mn-doped system. In this scheme single crystals of K0.85Fe1.8?zMzSe2 M = (Mn, Co, Ni, Cu, Zn) were converted to (Li1?x?yFexMyOD)Fe1?zMzSe where differ- ent transition metal dopants in the K0.85Fe1.8?zMzSe2 behaved differently when converting to the desired product. We have performed high resolution powder x-ray diffraction and neutron powder diffraction on 20% Co, Ni, Cu, and Zn samples in order to understand how the different transition metals behave in our hydrothermal reaction scheme. These results are presenting in Figures 4.6, 4.9, 4.7, 4.8. Full details of the crystallographic data is shown in Tables 4.3, 4.4, 4.5, 4.6. Based on our refinement results, it is surprising to find that only Mn and Co replace Li at the Li site as intended. For Ni (Fig. 4.6), about 2% still remains in the FeSe layer while for Cu and Zn, no doping is present in neither FeSe nor LiOH layer (Fig. 4.8, 4.7). For Cu and Zn, as a softer cations, they prefer co- 104 ordination to softer anions such as Se compared to O. Therefore, significant amount of starting K0.85Fe1.44Cu0.36Se2 and ZnSe impurity is present in the product for Cu and Zn doped samples respectively (Fig. 4.7 4.8). The case of Ni doping is a bit more difficult to explain - it is likely that the Ni in the starting materials converts to some stable sol- uble product in solution as opposed to exchanging for Li in the LiOH layer and is then removed when the product is recovered from the hydrothermal autoclave. (Li0.842(1)Fe0.135(4)Co0.02(3)OD)Fe0.950(2)Se (298 K, PXRD), P4/nmm, Rwp = 7.862% a = 3.7783(9) ?,c = 9.3065(2) ? atom Site x y z Occ. U 2iso (? ) Li/Fe1/Co1 2b 0 0.5 0 0.842(1)/0.135(4)/0.02(3) 0.0257(1) Fe2 2a 0.5 0.5 0 0.950(2) 0.0130(2) Se1 2c 0 0.5 0.1593(4) 1 0.014(5) O1 2c 0.5 0 0.4246(1) 1 0.007(5) Fe-Se (?) Se-Fe-Se (?) Se-Fe-Se (?) Fe-Fe (?) 2.401(7) 112.41(2) 103.73(7) 2.671(7) (Li0.842(1)Fe0.135(4)Co0.02(3)OD)Fe0.950(2)Se (2 K, NPD, P4/nmm, Rwp = 4.606% a = 3.7887(1) ?,c = 9.1480(6) ? atom Site x y z Occ. U 2iso (? ) Li/Fe1/Co1 2b 0 0.5 0 0.842(1)/0.135(4)/0.02(3) 0.012(9) Fe2 2a 0.5 0.5 0 0.950(2) 0.0051(1) Se1 2c 0 0.5 0.1566(4) 1 0.0019(3) O1 2c 0.5 0 0.4162(6) 1 0.0119(6) D1 2c 0.5 0 0.325(1) 1 0.003(1) Fe-Se (?) Se-Fe-Se (?) Se-Fe-Se (?) Fe-Fe (?) 2.372(8) 111.58(4) 105.32(3) 2.667(1) Tab. 4.3: Structural parameters for ground single crystals of (Li0.842(1)Fe0.135(4)Co0.02(3)OD)Fe0.950(2)Se. Structures are for room temperature PXRD data and 3 K NPD data. All relevant bond angles and distances from the refinements are given. Standard uncertainties given in parantheses indicate one standard deviation. As one may expect the location of the transition metal dopant, whether it remains in the FeSe layer or migrates to the LiOH layer, has a huge effect on the superconducting and magnetic properties. We have shown that Co doping behaves similarly to Mn dop- ing where all the Co ions migrate to the LiOH layer leaving the FeSe layer un-doped; although the ratio of Co replacement of Li compared to Fe is much smaller than in the 105 (Li0.822(1)Fe0.176(3)OD)Fe0.964(3)Ni0.020(1)Se (298 K, PXRD), P4/nmm, Rwp = 7.778% a = 3.7783(9) ?,c = 9.3065(2) ? atom Site x y z Occ. Uiso (?2) Li/Fe1 2b 0 0.5 0 0.822(1)/0.176(3) 0.0273(4) Fe2/Ni2 2a 0.5 0.5 0 0.964(3)/0.020(1) 0.0163(2) Se1 2c 0 0.5 0.1608(1) 1 0.011(2) O1 2c 0.5 0 0.4253(2) 1 0.002(4) Fe-Se (?) Se-Fe-Se (?) Se-Fe-Se (?) Fe-Fe (?) 2.412(6) 112.19(2) 104.15(1) 2.691(5) (Li0.8229(1)Fe0.176(3)OD)Fe0.964(3)Ni0.020(1)Se (2 K, NPD, P4/nmm, Rwp = 3.455% a = 3.7887(1) ?,c = 9.1480(6) ? atom Site x y z Occ. U 2iso (? ) Li/Fe1 2b 0 0.5 0 0.822(1)/0.176(3) 0.006(5) Fe2/Ni2 2a 0.5 0.5 0 0.964(3)/0.020(1) 0.0372(3) Se1 2c 0 0.5 0.1623(2) 1 0.0060(3) O1 2c 0.5 0 0.4242(4) 1 0.0118(7) D1 2c 0.5 0 0.3229(7) 1 0.003(1) Fe-Se (?) Se-Fe-Se (?) Se-Fe-Se (?) Fe-Fe (?) 2.409(1) 112.20(4) 104.13(8) 2.687(9) Tab. 4.4: Structural parameters for ground single crystals of Li0.822(1)Fe0.176(3)OD)Fe0.964(3)Ni0.020(1)Se. Structures are for room temperature PXRD data and 3 K NPD data. All relevant bond angles and distances from the refinements are given. Standard uncertainties given in parantheses indicate one standard deviation. Mn-doped case. The retention of an undoped FeSe layer has been shown in previous works to be integral in retaining the superconducting properties of FeSe layers.44 Inter- estingly though, Co doping does not result in observed long range magnetic order through neutron diffraction measurements. Magnetic susceptibility measurements and isothermal magnetization measurements presented in Figure 4.13 show a clear superconducting tran- sition at 30 K wtih a ferromagnetic signal superimposed on Meissner shielding signal at 2 K. The saturation value of the magnetic moment at 7 T applied field is approximately 0.48 ?B per magnetic ion in the LiOH layer which is 20% smaller than the similar Mn-doped compound; this is likely why long range magnetic order was not observed in neutron powder diffraction measurements on this system. The effect of dopant in the FeSe layer 106 (Li0.857(1)Fe0.14(4)OD)Fe0.992(3)Se (298 K, PXRD), P4/nmm, Rwp = 12.330% a = 3.8056(3) ?,c = 9.2424(1) ? atom Site x y z Occ. U 2iso (? ) Li/Fe1 2b 0 0.5 0 0.822(1)/0.14(4) 0.0247(3) Fe2 2a 0.5 0.5 0 0.992(3) 0.0183(5) Se1 2c 0 0.5 0.1604(1) 1 0.0169(1) O1 2c 0.5 0 0.4293(3) 1 0.0083(3) Fe-Se (?) Se-Fe-Se (?) Se-Fe-Se (?) Fe-Fe (?) 2.412(1) 112.20(1) 104.12(2) 2.690(9) (Li0.857(1)Fe0.14(4)OD)Fe0.992(3)Se (2 K, NPD, P4/nmm, Rwp = 6.222% a = 3.8375(4) ?,c = 9.040(1) ? atom Site x y z Occ. Uiso (?2) Li/Fe1 2b 0 0.5 0 0.822(1)/0.14(4) 0.0118(2) Fe2 2a 0.5 0.5 0 0.992(3) 0.0595(6) Se1 2c 0 0.5 0.1651(5) 1 0.0024(7) O1 2c 0.5 0 0.4313(8) 1 0.019(1) D1 2c 0.5 0 0.327(1) 1 0.046(2) Fe-Se (?) Se-Fe-Se (?) Se-Fe-Se (?) Fe-Fe (?) 2.431(4) 112.16(1) 104.21(6) 2.713(5) Tab. 4.5: Structural parameters for ground single crystals of (Li0.857(1)Fe0.14(4)OD)Fe0.992(3)Se. Structures are for room temperature PXRD data and 3 K NPD data. All relevant bond angles and distances from the refinements are given. Standard uncertainties given in parantheses indicate one standard deviation. and superconductivity is clearly observed in the nominally 20% Ni doped sample. For Ni doping, Ni does not replace Li in the LiOH layer and instead we observe 2% Ni substi- tution of Fe in the FeSe layer. In previous work, superconductivity in the Ni-doped FeSe system was suppressed by >5% Ni doping,158 and we observe similar results in magnetic susceptibility of the (Li0.8229(1)Fe0.176(3)OD)Fe0.964(3)Ni0.020(1)Se powder sample. Magnetic susceptibility and isothermal magnetization of the sample shows paramagnetic behavior with slight bifurication of ZFC-FC curves around 130 K which is likely due to slight amount of iron oxide impurity; superconductivity is completely suppressed. This result is somewhat expected since the FeSe layers are no longer pristine with 2% Ni substitution and 1.5% vacancies on the Fe site in the FeSe layers. 107 (Li0.847(2)Fe0.15(1)OD)Fe0.995(2)Se (298 K, PXRD), P4/nmm, Rwp = 12.330% a = 3.7837(3) ?,c = 9.281(1) ? atom Site x y z Occ. U 2iso (? ) Li/Fe1 2b 0 0.5 0 0.847(2)/0.15(1) 0.0155(6) Fe2 2a 0.5 0.5 0 0.995(2) 0.0266(1) Se1 2c 0 0.5 0.1591(1) 1 0.0153(2) O1 2c 0.5 0 0.4293(5) 1 0.0015(8) Fe-Se (?) Se-Fe-Se (?) Se-Fe-Se (?) Fe-Fe (?) 2.400(1) 112.25(2) 104.03(5) 2.675(5) (Li0.847(2)Fe0.15(1)OD)Fe0.995(2)Se (2 K, NPD, P4/nmm, Rwp = 7.750% a = 3.8375(4) ?,c = 9.040(1) ? atom Site x y z Occ. Uiso (?2) Li/Fe1 2b 0 0.5 0 0.847(2)/0.15(1) 0.024(8) Fe2 2a 0.5 0.5 0 0.995(2) 0.0032(5) Se1 2c 0 0.5 0.1608(4) 1 0.0041(3) O1 2c 0.5 0 0.4234(7) 1 0.0131(6) D1 2c 0.5 0 0.322(1) 1 0.051(1) Fe-Se (?) Se-Fe-Se (?) Se-Fe-Se (?) Fe-Fe (?) 2.399(1) 112.21(1) 104.12(2) 2.675(8) Tab. 4.6: Structural parameters for ground single crystals of (Li0.847(2)Fe0.15(1)OD)Fe0.995(2)Se. Structures are for room temperature PXRD data and 3 K NPD data. All relevant bond angles and distances from the refinements are given. Standard uncertainties given in parantheses indicate one standard deviation. The Cu and Zn-dopes samples behave very similarly. In both cases, we have ob- served from high resolution x-ray powder diffraction and neutron powder diffraction that Cu and Zn ions do not replace Li ions in the LiOH or Fe ions in the FeSe layers. Thus, the resulting product shows very similar behavior to previous work on the un-doped (LiOD)FeSe system where the superconducting criticial temperature is dependent on the tetragonality of the system.82As with other un-doped (LiOD)FeSe works, no long range magnetic order was observed in the neutron powder diffraction measurements on the Cu and Zn-doped samples. The vacancies caused by the removal of Cu and Zn ions from the starting material were filled by the additional Fe powders/granules added in the reac- tion scheme which cause the FeSe layers of the products in both cases to remain close to 108 pristine. These results show that transition metal doping of the (LiOD)FeSe system is lim- ited to the harder low valence transition metals (Mn, Co, and Ni). We have shown that Mn-doping can lead to the observation of long range magnetic order in the system by increasing the effective moment of the transition metal in the LiOH layers without largely effecting the superconductivity as the FeSe layers remain pristine. Co doping behaves similarly, whereby Co ions only go to the LiOH layer, but the level of Co substitution was lower than in the Mn case leading to no observable long range magnetic order. Ni doping was observed to be detrimental to superconductivity in the system as Ni replaces Fe in the FeSe layer destroying superconductivity. Thus, increased Mn and Co doping, beyond 20% as long as the starting material can be synthesized, is a viable method to stabilize single crystals of magnetic heterolayer superconductors. 4.7 Conclusions The coexistence of superconductivity and magnetism may be the key to realize the next generation of multi-functional materials. To that end, we have successfully syn- thesized a series of late transition metal doped (LiOH)FeSe single crystals through a hydrothermal ion-exchange reaction has been shown to house the co-existence of super- conductivity and long-range magnetic order in the case of the nominally 20% Mn-doped sample. We have fully characterized the 20% Mn-doped sample through powder x-ray and neutron diffraction yielding a stoichiometry of (Li0.875(2)Fe0.062(3)Mn0.062(3)OD)FeSe showing that the doped Mn migrates fully to the hydroxide layer. Magnetic measurements 109 show a superconducting transition at 25 K with shielding fraction ? 25% indicating the superconductivity in single crystal samples may be filamentary in nature with magneti- zation showing clear Meissner shielding and magnetic signal co-existing at low temper- atures exclusively. Transport measurements show a sharp 41 K transition in resistivity with a distinct non-linear normal state with no clear anomaly detected around the super- conducting transition in heat capacity. However, a clear magnetic peak was observed at ? 9K in heat capacity measurement which remains robust up to 14 T applied field parallel to the crystallographic c-axis. Powder neutron diffraction data reveals a satellite magnetic reflection at Q = 1.41 ? 1.44 ??1 indicating the long-range character of the magnetic order in this system observed for the first time. Additional neutron experiments would be required to elucidate the nature of the magnetic ordering in this system. This work shows that the (LiOH)FeSe system offers a platform for chemical manipulation to induce the coexistence of long range magnetic order and superconductivity for possible exploitation as multi-functional materials and for use in quantum computing. 110 2 30 (a) 1 20 0 0 T2 T 10 30 40 4 T6 T 0 8 T 0 100 200 300 0.4 (b) 0.3 0.20 0 T 7 T 0.2 0.1514 T 0 50 100 0 1000 2000 3000 T 2 (K2) 0.4 (c) 0 T 0.2 0.0 5 10 15 20 T (K) Fig. 4.5: Electronic and thermal transport measurements on a single crystal of (Li0.875(2)Fe0.062(3)Mn0.062(3)OD)FeSe. a) Temperature dependence of electrical re- sistivity with mutiple applied fields, with the applied field parallel to the crystallographic c-axis, showing a sharp superconducting transition at 41 K. At higher applied fields, the transition onset does not change, but the transition width is increased drastically. b) Temperature dependence of heat capacity at various applied fields showing no observable anomaly at the superconducting transition temperature, however; a magnetic transition is observed at ? 9 K that is weakly suppressed with high field (inset). c) The change in entropy calculated as described in main text. 111 ?S ( J ) Cp/T ( J )mol ?K mol ?K2 ?ab (m? ? cm) Fig. 4.6: top) High resolution synchrotron powder X-ray diffraction patterns for (Li1?x?yFexNiyOD)Fe1?zMnzSe collected at room temperature, (middle and bot- tom) show powder neutron diffraction data for (Li1?x?yFexNiyOD)Fe1?zMnzSe at 2 K and 50 K respectively. Green tick marks represent the targeted tetragonal layered phase and orange tick marks represent impurity selenium. These are shown below the calculated, observed, and differences curves from Rietveld analysis. Analysis yielded a composition of (Li0.8229(1)Fe0.176(3)OD)Fe0.964(3)Ni0.020(1)Se. 112 Fig. 4.7: a) High resolution synchrotron powder X-ray diffraction patterns for (Li1?x?yFexCuyOD)Fe1?zCuzSe collected at room temperature, b) and c) show powder neutron diffraction data for (Li1?x?yFexCuyOD)Fe1?zCuzSe at 2 K and 50 K respectively. Green tick marks represent the targeted tetragonal layered phase, orange tick marks represent impurity selenium, and purple tick marks represent impurity starting material in the x-ray data. These are shown below the calculated, observed, and differences curves from Rietveld analysis. Due to remaining starting material, the sample was reacted for 3 additional days before neutron analysis, where green tick marks represented the targeted phase and orange tick marks show impurity lithium carbonate. Analysis yielded a composition of (Li0.857(1)Fe0.14(4)OD)Fe0.992(3)Se. 113 Fig. 4.8: a) High resolution synchrotron powder X-ray diffraction patterns for (Li1?x?yFexZnyOD)Fe1?zZnzSe collected at room temperature, b) and c) show powder neutron diffraction data for (Li1?x?yFexZnyOD)Fe1?zZnzSe at 2 K and 50 K respectively. Green tick marks represent the targeted tetragonal layered phase, orange tick marks represent impurity selenium, and purple tick marks represent impurity zinc selenide. These are shown below the calculated, observed, and differences curves from Rietveld analysis. Analysis yielded a composition of (Li0.847(2)Fe0.15(1)OD)Fe0.995(2)Se. 114 Fig. 4.9: a) High resolution synchrotron powder X-ray diffraction patterns for (Li1?x?yFexCoyOD)Fe1?zCozSe collected at room temperature, b) and c) show powder neutron diffraction data for (Li1?x?yFexCoyOD)Fe1?zCozSe at 2 K and 50 K respectively. Green tick marks represent the targeted tetragonal layered phase and orange tick marks represent impurity selenium. These are shown below the calculated, observed, and differences curves from Rietveld analysis. Analysis yielded a composition of (Li0.842(1)Fe0.135(4)Co0.02(3)OD)Fe0.950(2)Se. 115 Fig. 4.10: (left) Temperature dependent magnetic susceptibility at 30 Oe and (right) Isother- mal magnetization at 2 K and 50 K measured on ground single crystals of (Li0.8229(1)Fe0.176(3)OD)Fe0.964(3)Ni0.020(1)Se displaying non-superconducting behavior, most likely due to Ni substitution on the Fe-site in the FeSe layers. Fig. 4.11: (left) Temperature dependent magnetic susceptibility at 30 Oe and (right) Isother- mal magnetization at 2 K and 50 K measured on ground single crystals of (Li0.857(1)Fe0.14(4)OD)Fe0.992(3)Se displaying superconducting behavior with a super- conducting transition at 32 K. 116 Fig. 4.12: (left) Temperature dependent magnetic susceptibility at 30 Oe and (right) Isother- mal magnetization at 2 K and 50 K measured on ground single crystals of (Li0.847(2)Fe0.15(1)OD)Fe0.995(2)Se displaying superconducting behavior with a super- conducting transition at 42 K. Fig. 4.13: (left) Temperature dependent magnetic susceptibility at 30 Oe and (right) Isother- mal magnetization at 2 K and 50 K measured on ground single crystals of (Li0.842(1)Fe0.135(4)Co0.02(3)OD)Fe0.950(2)Se displaying superconducting behavior with a superconducting transition at 30 K. 117 Chapter 5: Alkali metal-free hydrothermal synthesis of ethylenedi- amine intercalated iron chalcogenides The work described within this chapter was submitted to Inorganic Chemistry. Brandon Wilfong, Xiuquan Zhou, Huafei Zheng, Johnpierre Paglione, Rishvi Jayathialake, Daniel J. Campbell, Sz-Chian Liou, and Efrain Rodriguez were contributing authors of the manuscript. B.W., X.Z., and H.Z. prepared the samples, B.W. collected X-ray diffrac- tion data, performed MPMS measurements and collected resistivity data, B.W. and S.C.L performed TEM and ED experiments, D.J.C. collected SEM and EDS data and R.J. per- formed TGA analysis. 5.1 Introduction The discovery of superconductivity in the iron-based superconductors La(O 221?xFx)FeAs and FeSe42 spurred a surge of research activity. Much like the copper-oxygen layers which are key for superconductivity in the copper-based superconductors245, all iron- based superconductors contain stacked layers of iron pnictides or chalcogenides.27 In the simplest case, FeSe contains stacked tetrahedral layers of an square iron sublattice with each iron tetrahedrally coordinated to either pnictide or chalcogenide atoms.42,45 With FeSe and FeS displaying superconductivity at 8 K42 and 4 K58, respectively, 118 there has been substantial research on intercalation chemistry within the system to probe how intercalated species will affect the properties. Much of the work has focused on FeSe as opposed to FeS due to the metastability of FeS.59 To date the successful intercalates of the iron chalcogenide system consists of simple cations76?78 and/or partially charged hydroxide or amine layers.46,79?81,83?87 After intercalation the Tc of FeSe can be raised significantly to 43-46 K from 8 K. Although both hydroxide or amine intercalation can increase the Tc of FeSe, only LiOH intercalation is known to increase the T of FeS.87c In the case of simple cations, ionic interactions between the cationic layer and the iron chalcogenide layers help stabilize the structure. However, in the partially charged amine or hydroxide case non-ionic guest-host interactions such as hydrogen bonding would also help stabilize the layered structures. In most works regarding intercalation of amines, the co-intercalation of alkali metal atoms has been required since they to coor- dinate with the amine molecules.46,83?86 We demonstrate here that alkali metal cations are not required and intercalated strucutures may be stabilized by hydrogen bonding alone. Amine intercalations are usually achieved by forming solvated electrons using al- kali metal and liquid amines, which inevitably leads to co-intercalation of alkali metal ions. This makes determining the electron doping effect on the Tc of FeCh (Ch = S and Se) quite difficult due to the formation of multiple intercalated species. Two differ- ent approaches in synthesis have been pursued to understand the doping effect of these intercalated species and samples. One approach is to directly react FeCh with amines solvothermally, to form ethylenediamine(EDA)-intercalated FeSe and FeS.246?248 How- ever, the type of potential intercalates are limited using this method as the intercalates have to be in the form of liquid or exhibit low melting points. The other approach is to 119 hydro/solvothermally prepare intercalated FeCh phases through bottom-up synthesis. We pursue this approach here in order to find synthetic routes that maximize the number of different intercalates that can be inserted into FeCh superconductors without the require- ment of alkali metals. In the previously mentioned synthetic approaches, alkali metals are directly reacted with amines before adding the iron chalcogenides species which results in a differently chemically active solvated alkali metal ions249?251 as opposed to aqueous al- kali metal hydroxide ions used in our synthetic route. Thus, alkali metal - amine reaction scheme, solvated alkali metals co-intercalate with amines whereas in our reaction scheme, the aqueous alkali metal does not coordinate to the amine and co-intercalate between the iron chalcogenide layers. 5.2 Experimental Methods For a typical preparation of EDA-intercalated FeCh (Ch = S, Se), 4 mmol of Fe powder (Alfa Aesar, 99.5%), 10 mmol of thiourea (Sigma-Aldrich, 99%) or 5 mmol se- lenourea (Sigma-Aldrich, 98%), 7 mmol of KOH (Fisher, 85%) which equates to ? 0.5 M solution, 10 mL H2O, and 3 mL ethylenediamine (Sigma-Aldrich, 99%) sealed within a Teflon cup within a stainless steel autoclave at 120 - 160 ?C for 2 - 6 days. After the hydrothermal process, the contents were washed and centrifuged with de-ionized water several times until the supernatant was clear. The recovered black powders were collected, vacuum dried and stored in an Argon glovebox. Powder X-ray diffraction (XRD) data was collected usining a Bruker D8 X-ray diffractometer utilizing Cu K? radiation (? = 1.5418 ?, 2? = 5 - 70?, step size = 0.020?). 120 Pawley refinements were performed using the TOPAS software.126 Microscopic images were examined on a Hitachi SU-70 SEM field emission scanning electron microscope (SEM), and their elemental compositions were determined by energy dispersive X-ray spectroscopy (EDS) using a Bruker EDS detector. Electron diffraction patterns were ob- tained using a JEM 2100 LaB6 transmission electron microscope (TEM) at an acceler- ation voltage of 200 KeV. Thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC) were performed using Mettler Toledo TGA/DSC 2 under high-purity Ar. TGA/DSC samples were placed in alumina crucibles covered with alumina lids and heated from room temperature to 600 ?C at the rate of 10 ?C/min. Magnetic susceptibility and magnetization measurements were performed using a Quantum Design Magnetic Property Measurement System (MPMS3). Temperature de- pendence of magnetic susceptibility was performed at various applied fields and magneti- zation isotherms were measured up to fields of ? 7 T. Electrical resistivity measurements were performed on as-recovered and annealed samples of pressed pellets of ground pow- ders cold-pressed into a pellet with 2000 psi of uniaxial stress. Temperature dependence of electrical resistivity was performed using the four-probe method from 2 K - 300 K with applied currents of 0.1 mA and frequencies around 17 Hz. 5.3 Results and Discussion 5.3.1 Alkali metal free hydrothermal ethylenediamine intercalation All successful hydrothermal intercalations of EDA occurred at 120 ?C; higher tem- peratures led to increased FeSe/FeS and Fe3O4 formation with no intercalation. The re- 121 sults of varying the reaction time are presented in Figure 5.1 and 5.2. For the sulfide analogue, we find successful intercalation after two day reaction time; however,some tetragonal FeS remained. At four days, the remaining FeS is fully intercalated by EDA. After six days, no intercalation is observed and only poorly crystalline FeS and iron pow- der remain. For the selenide analogue, two day reactions reveal a possible intermediate intercalated phase. This new layered phase could be indexed with the structurally related KxFeySe2 (122-type), although its stoichiometry could not be determined. In previous works, this phase was shown to form in the sulfide case when using KOH as the base in the reaction.87 After a four day reaction, the 122-type structure had converted to EDA- intercalated phase and the formation of Fe3O4 is observed. After six days, there is no intercalated phase, all that remains is poorly crystalline FeSe, additional Fe3O4, and Fe powder. Previous work on amine intercalation in FeCh species focused on the co-intercalation of alkali metal ions which would stabilize the amine intercalate through alkali metal coor- dination to the amide groups.46 This method has been successful for a number of different amines and alkali metals,77,83,85,86,252,253 but the amount of electron doping level in these systems is difficult to solve. Amines can intercalate as both neutral molecules and nega- tively charged amides or imides.46 Therefore, simply determining the alkali metal content does not gave the electron doping level. However, using our hydrothermal method the for- mation of metal amides are prevented. Also, due to the co-intercalated alkali metal, all compounds made via the previous method are extremely air sensitive making extended characterization increasingly difficult. Thus, this work aims to use our hydrothermal method in order to remove the necessity for alkali metal coordination as the stabilizing 122 Ycalc EDAFeS Yobs - Ycalc Fe Yobs t = 4 days Ycalc EDAFeS Yobs - Ycalc Fe Yobs FeS t = 2 days 2? (deg) Ycalc FeS Yobs - Ycalc Fe Yobs t = 6 days 10 20 30 40 50 60 70 2? (deg) Fig. 5.1: Pawley refinements on ground powders of (C2H8N2)xFeS after varying hydrothermal reaction times. factor. Previously, we intercalated ammonia molecules along with protons hydrothermally into FeCh layers. We resolved the structure of intercalated species through the use of neu- tron powder diffraction. In this analysis, the location of the hydrogen atoms and ammonia molecules were determined and the N-H? ? ?Ch had clear implications for hydrogen bond- ing. However, this type of structural analysis is nearly impossible for ethylenediamine molecule NH2-C2H4-NH2 where all 12 atoms are in the highly disordered 16i site, for the I4/m space group.86 Secondly, previous work showed that the proton position is at the 2b 123 Relative Intensity (arb. units) Ycalc Fe Yobs - Ycalc Fe3O4 Yobs FeSe EDAFeSe t = 4 days Ycalc KFe2Se2 Yobs - Ycalc Fe Yobs FeSe t = 2 days 2? (deg) Ycalc FeSe Yobs - Ycalc Fe Yobs Fe3O4 t = 6 days 10 20 30 40 50 60 70 2? (deg) Fig. 5.2: Pawley refinements on ground powders of (C2H8N2)yFeSe after varying hydrothermal reaction times. site with about 50% of the occupancy, which allows for precise refinement from neutron diffraction and direct understanding of the doping effects within the system. Previous work246?248 has shown this synthesis can be done solvothermally, however; this severely limits the types of intercalate adducts that can be used as they must be liquid at or below 120 ?C, especially in the case of FeS due to its limited thermal stability. The results of the SEM, EDS and TGA measurements are presented in Figure 5.3 and 5.4, which help us understand the morphology, composition, and stability. Elemental analysis from EDS is in general agreement with previous reports,246,248 although direct 124 Relative Intensity (arb. units) stoichiometry cannot be matched due to remaining iron and iron-containing impurities, especially showing that our samples do not show alkali metals which are present in sam- ples grown by the alkali metal - amine method.254,255This confirms that our samples are alkali metal free as the aqueous alkali metals from the alkali metal hydroxides do not co-intercalate with the ethylenedimaine as solvated alkali metal ions do in the previously discussed synthetic method. This is likely due to the difference in chemical activity of solvated alkali metal ions as opposed to aqueous alkali metal ions. Elemental analysis for C/H/N cannot be performed with EDS due to their low effective charge and highly overlapping spectra. TGA analysis was performed under argon atmosphere to show that the hydrothermal intercalated phases are stable up to temperatures higher than previously reported for intercalated FeCh phases.246,248 (C2H8N2)xFeS showed weight loss, owning to ethylenediamine de-intercalation, at 225 ?C, while (C2H8N2)yFeSe did not show de- intercalation until 350 ?C. Although these de-intercalation temperatures are different than the reported weight loss temperatures for the ammonia intercalated phases256, we cannot definitely rule out that ammonia intercalated species do not exist in this compound due to the use of thiourea and selenourea as the chalcogenide sources. Weight loss before and after the de-intercalation give rough estimates of x = 0.19 and y= 0.16 ethylenediamine occupancy per unit cell for (C2H8N2)xFeS and (C2H8N2)yFeSe respectively; these values are both less than their solvothermal analogues which may mean the ethylenediamine co- ordination or intercalated structure may differ due to the different synthetic method.246,248 Surprisingly, this means the ethylenediamine intercalated FeS phase is stable up to tem- perature comparable to the metastable FeS binary.59. The selenide analogue is much more stable than the solvothermal synthesized version246 and matches the highest tem- 125 peratures at which alkali metal co-intercalated ethylenediamine samples were annealed.84 Both samples do not appear to be air sensitive after exposure for a week. The large ther- mal range of stability for these hydrothermally prepared iron chalcogenides gives a good basis for post-modifications to alter the physical properties. ?1 1400 7.0 1200 ?2 6.8 1000 ?3 800 6.6 600 C/N ?4 Fe: 55.56% ? 3.25% 400 6.4 Fe S: 44.43% ? 1.77% ?5 200 O 6.2 0 ?6 0 2 4 6 8 10 0 50 100 150 200 250 300 350 400 Temperature (?C) Energy (keV) Fig. 5.3: TGA and SEM/EDS analysis of (C2H8N2)xFeS ground powders. 3.15 ?1.5 ?2.0 1400 3.10 ?2.5 1200 3.05 ?3.0 1000 ?3.5 800 3.00 ?4.0 C/N600 2.95 ?4.5 400 Fe Fe: 55.19% ? 2.94%Se: 44.80% ? 5.96% ?5.0 2.90 200 O ?5.5 Fe 0 2.85 ?6.0 0 2 4 6 8 10 0 50 100 150 200 250 300 350 400 Temperature (?C) Energy (keV) Fig. 5.4: TGA and SEM/EDS analysis of (C2H8N2)yFeSe ground powders. 5.3.2 Intercalated ethylenediamine crystallography and symmetry Figure 5.5 shows XRD patterns for (C2H8N2)xFeS and (C2H8N2)yFeSe as recovered from the hydrothermal autoclave. Pawley fits were used to model the diffraction data since the intercalated structure is too complex to solve by Rietveld analysis with the laboratory XRD sources. However, neutron powder diffraction (NPD) could help solve the position 126 Weight (mg) Weight (mg) Heat Flow (mW) Heat Flow (mW) Counts (C/s) Counts (C/s) of the light elements of the organic clusters. Previously Jin et al. had utilized NPD to solve the structure of a related intercalated phase Na (C N H ) Fe Se .860.39(1) 2 2 8 0.77(1) 2.02(1) 2 . Therefore, we used a similar model with space group I4/m for our Pawley fits of the intercalated structure. Both patterns can be indexed well using this space group yielding lattice parameters close to previous reports.248 It should be noted that in both powder patterns only the (00l) reflections are resolved due to strong preferred orientation as well as possible stacking faults along the c-axis. For (C2H8N2)xFeS, Pawley fits gave lattice parameters of a = 3.700(5) ?, c = 20.67(1) ? which are in close agreement to alkali metal and alkali metal-free ethyelen- diamine intercalated FeS.77,248 In this case, the a-axis lattice parameter is slightly larger due to the in ? situ formation of the FeS layers as opposed to intercalating pre-formed tetragonal FeS powders. In in-situ formation of FeS layers may be influenced by the intercalated species through hydrogen bonding directionality and stability. Interestingly, the c-axis lattice parameter is in much closer agreement to the potassium co-intercalated ethylenediamine FeS rather than the solvothermal EDA treatment possibly due to the role of proton intercalation via the hydrothermal process. The only remaining impurity comes from excess Fe powder remaining from the in? situ growth of FeS layers. To avoid excess Fe, we attempted syntheses using pre-reacted tetragonal FeS at 120 ?C over 2 - 6 days, but were unsuccessful. Remaining Fe powder could be mostly removed from the recovered product by using a weak magnet. However, this recovery method significantly reduced product yield since the desired product was intergrown with unreacted Fe powders. Pawley fits for selenide analogue yielded lattice parameters of a = 3.813(2) ?, c = 21.64(3) ?. The previous solvothermal work indexed the EDA-intercalated FeSe with 127 (a) Ycalc EDA-FeS Yobs - Ycalc Fe Yobs (b) Ycalc Fe Yobs - Ycalc Fe3O4 Yobs FeSe EDA-FeSe 10 20 30 40 50 60 70 2? (deg.) Fig. 5.5: Pawley refinements with powder XRD for the structures of (C2H8N2)xFeS and (C2H8N2)yFeSe as recovered from hydrothermal synthesis. (a) Refinement of (C2H8N2)xFeS in the body-centered tetragonal structure (I4/m) at room temperature; (b) refinement of (C2H8N2)yFeSe with a body-centered tetragonal structure (I4/m) at room temperature. Lower tick marks correspond to the tetragonal phases. Additional tick marks denote impurities present with the sample. an orthorhombic cell, but we did not find the same result since we did not observe or- thorhombic splitting of the (013) or (015) peaks. The a-axis and c-axis lattice param- eters closely agree with previously reported for alkali metal co-intercalated EDA-FeSe and solvothermal EDA-FeSe.84,86,246,247,255 The powder XRD pattern for (C2H8N2)yFeSe is more complicated than the sulfide analogues due to the synthetic challenges raised by selenium. To start, additional impurities included tetragonal FeSe and iron oxide Fe3O4. 128 Relative Intensity (arb. units) Fig. 5.6: (a) Electron diffraction pattern of ethyelenediamine-intercalated, (b) model FeS layers rotated as observed in electron diffraction images, and (c) FeS, green circles indicating allowed reflections and rotated pairs for the I4/m unit cell with red circles indicating for- bidden reflections and blue showing possible reflections from intercated EDA molcules. We easily identify un-intercalated FeSe by its (001) reflection (d = 5.5258 ?), which is proximate to the (004) reflection of the intercalated phase (d = 5.41 ?). We hypothesize that remaining tetragonal FeSe is due to the thermodynamic stability of FeSe compared to that of metastable FeS. The appearance of Fe3O4 only in the selenide analogue is due to the electrochemical potential differences of S2? and Se2?. Attempts to remove Fe3O4 using reaction times less than 4 days lead to no EDA intercalation while both longer reac- tion and higher temperatures lead to no intercalation and/or more impurity. Interestingly, the Fe3O4 would not form in the product if the reaction was carried out in D2O instead of H2O despite reducing the crystallinity. Similar solvent effects were observed for the synthesis of LiOD-intercalated FeSe, where the use of D2O suppressed the formation of iron oxide impurities.82 Since it was not possible to resolve the structure from the powder XRD data, we have investigated their structures via electron diffraction (ED) using transmission elec- tron microscopy (TEM). Despite poor resolution for reflections other than (00l) on their XRD patterns, we observed well-resolved (hk0) reflections by ED. This indicates that the 129 (a) (220) (120) (210) (110) (1-10) (010) (300) (100) (200) (-100) (2-10) (0-10) (-110) (-1-10) (-2-10) Fig. 5.7: (a) Electron diffraction pattern of ethyelenediamine-intercalated FeSe; green indexing and circles indicate allowed reflections for the I4/m unit cell while red indicates forbidden reflections. seemingly poor crystallinity of the samples is a result of strong anisotropic broadening on the Bragg-Brentano geometry due to extremely short coherence lengths over the c-axis of the layers. The FeS analogue exhibits complicated ED patterns. As shown in Fig. 5.6, the diffraction spots highlighted by green circles could be assigned to three different square lattices with a = 3.69 ?, which is very close to our refined value by PXRD. Interestingly, the two weaker sets of spots are placed at the positions by rotating the stronger set by 26? and -26? respectively. This suggests that the extra two sets of reflections likely do not originate from two other randomly stacked nanocrystals. Indeed, we found another set of weak reflections (highlighted in blue in Fig. 5.6c) that do not show such twin reflections at ?26?. These extra diffraction spots (blue) cannot be indexed with the tetragonal lattice. In addition, considering they are much weaker compared to the square-lattice reflections, 130 these spots likely originate from the EDA molecules within the sheets. Since they do not rotate with the FeS sheets, it is likely that the EDA molecules do not arrange in a commensurate fashion with the square lattice of FeS and are vertically arranged between the FeS sheets. This configuration of the EDA molecules may be a result of stronger hydrogen bonding in the sulfide over the selenide of the type N-H? ? ?Ch. This hydrogen bonding may lead structure directing of subsequent FeS layers which would result in the observed c-axis disorder. More structural disorder in the intercalated FeS is revealed in Figure 5.8 in Supple- mental Information. These images are taken from different crystallites within the same synthetic batch. It is difficult to understand the driving force leading to varying degrees of c-axis disorder in this system, but it is likely caused by different degrees of hydrogen bonding or coordination with adjacent FeS sheets. This increased c-axis disorder and sub- sequent ED ??smudging? was observed in ethylenediamine intercalated SnS2 as well.257 Figure 5.8 also shows a TEM image of lattice fringes within the sample which display an average length of less than 10 nm indicating short coherence lengths along the c-axis due to stacking faults or possible twisting of the FeS layers due to intercalated EDA. Fig. 5.8: Electron diffraction pattern of ethyelenediamine-intercalated-FeS and TEM image of lat- tice fringes of EDA-FeS. 131 For the FeSe sample, its ED pattern agrees with the XRD data as the unit cell can be indexed to a tetragonal one with a = 3.83 ?(Fig. 5.7). However, we were able to observe a set a forbidden reflections at the condition (h + k = 2n + 1). This could be caused by disorder or a charge density wave, which is also seen for the ammonia- intercalated FeCh.256 More interestingly, their ED patterns are almost identical. However, unlike ammonia, whose atoms can be arranged into a nearly perfect tetrahedron, the EDA molecule is much more complex and does not exhibit 4-fold rotational symmetry. The fact that no extra reflection spot breaks the 4-fold symmetry suggests a completely disordered configuration for the EDA molecules. Therefore, all the C, H and N atoms in EDA- intercalated FeSe are likely to occupy the 16i Wyckoff position in a disordered manner. We cannot determine whether protons co-intercalate with EDA to occupy the edge-center (2b site) of the unit cell, but it is a reasonable assumption given the synthetic and structural similarities between both intercalated FeSe phases. 5.3.3 Intercalated ethylenediamine configuration and guest-host in- teractions To date, much of the work on the synthesis and characterization of ethylenedi- amine intercalated iron selenide has focused on the increase in critical temperature ob- served.84,86,246,255,258 In these intercalated iron selenide systems, increasing two dimen- sionality of the FeSe layers as well as charge doping due to intercalated species combine to produce the increase in superconducting critical temperature. Both of these factors are determined by the new structures which form as intercalated species exist between the 132 FeSe layers. Thus, fully understanding the novel structures which arise from molecu- lar intercalation between these inorganic layers is immensely important for future guided synthesis and to understand how the superconducting properties change in these systems due to the changes in structure. There exist two main forces which likely help stabilize the intercalation of ethylene- diamine in FeSe, electrostatic forces due to charge species formed via ethylenediamine coordination to metal ions and/or hydrogen bonding from the ethleyendiamine molecules to the chalcogenides in the FeSe layers. There has been extensive work on the character- ization of ammonia and ethylenediamine intercalation into various layered chalcogenide compounds. For ammonia intercalation of transition metal dichalcogenides, nuclear mag- netic resonance and diffraction work has shown that the formation of ammonium ions and/or the arrangement of intercalated ammonia molecules to maximize hydrogen bond- ing to nearby chalcogenide ions helped stabilize the structure of these compounds.259?269 Ethylenediamine intercalation is similar although there has been no reports thus far show- ing a similar homosolvation model to ammonia intercalation. Thus, ethylenediamine ex- hibits coordination to metal ions in solution to generate interlayer charged species to stabilize the structure, and/or the ethylenediamine molecules arrange to maximize amino- hydrogen bonding toward the chalcogenide layers. In the former case, dissolved metals in solution coordinate with the ethylenediamine molecules to form metal-ethylenediamine coordinated molecules.270?278 Harder Lewis acid metals preferentially coordinate with the ethylenediamine to form these coordinated species whereas softer cations typically form the chalcogenide layers or chains. When no metal ion coordination is present, free ethylenediamine molecules arrange between layers so that amino-hydrogen bonding to 133 chalcogenide or oxide ions stabilize the structure.257,279?283 The hydrogen bonding inter- action in these inorganic/organic hybrid materials stabilizes the structure, but also directs the inorganic constituents structure since hydrogen bonding is directional in nature. There have been three main works which discuss possible crystal structures for ethylenediamine intercalated iron selenide, shown in Figure 5.986,247,255 Two of these works consider akali-metal co-intercalation while the other uses an alkali-metal free solvother- mal method similar to the hydrothermal method considered in this work. All three have different solutions for how ethyelendiamine intercalates between FeSe layers determined through different methods, such as neutron powder diffraction and single crystal diffrac- tion. For the two with alkali metal co-intercalation, the coordination of the ethylene- diamine molecule to the alkali metal is different. The almost ubiquitous role ethylene- diamine plays in coordination chemistry is well catalogued.284,285 With regards to the first structure Figure 5.9 (left), the coordination of the ethylenediamine molecule to the sodium ion does not go through the nitrogen atoms but instead the carbon-carbon bond is proximate to the alkali metal. This is unlikely as the nitrogen atoms in ethylenedi- amine are strong Lewis bases for which coordination to the alkali metal should dominate. This structure was improved upon up the same authors in Figure 5.9 (center) where now monodentate or bridging ethylenediamine molecules arrange vertically between the FeSe layers and the FeSe layers are arranged in a primitive setting. Although this makes more sense from a coordination chemistry perspective, monodentate ethyelendiamine is very rare286?289 as the two nitrogen atoms predominately form a bidentate to the same coordi- nation site. Even amongst similar reaction schemes, the bidentate nature of ethylenedi- amine is well reported.247,276?278. The authors report that increasing Li content leads to the 134 transformation of the body centered stacking of the FeSe layers to the primitive setting. This may be due to the effects of Li-ethylenediamine coordination which leads to direc- tional hydrogen bonding which guides the FeSe layers into the primitive setting through amino-hydrogen interactions with the selenide ions of the FeSe layers.255 The stability of both of these intercalated compounds likely comes from electrostatic forces from the introduction of interlayer charged species; however, the arrangement of amino-hydrogens in the ethylenediamine molecules exhibit possible hydrogen bonding to selenide ions in the FeSe layers. In both of these cases, these compounds exhibit high superconducting critical temperatures, likely due to charge doping the FeSe layers, but without a concrete understanding of the interlayer species, determining the exact doping leading to this in- creased critical temperature is impossible. For the last case, Figure 5.9 (right), without alkali metal co-intercalation, the ethylene- diamine molecule is not coordinated to anything and is free between the FeSe layers. From this work, the reaction scheme requires Fe powder to be oxidized in solution to form the FeSe layers247, thus the lack of Fe ion centered ethylenediamine coordination complexes is questionable as the formation of Fe ion centered ethylenediamine complexes in similar hydrothermal and solvothermal syntheses has been documented.276?278 Since, no metal ion coordination exists electrostatic forces do not exist to stabilize the inter- calation, thus requiring hydrogen bonding. Previous work on ammonia intercalation of FeSe46 showed how hydrogens bonded to the nitrogen are directed toward the selenide ions leading to a bond distance of 2.75 ? consistent with weak hydrogen bonding ob- served in a wide range of works.290 The structure reported by Stahl et. al has a closest N-H ? ? ? Se distance of 3.63 ? which is outside the range for hydrogen bonding. Thus, 135 without the electrostatic forces due to metal ion-ethylenediamine coordination and no hydrogen bonding, it is hard to understand what stabilizes the structure in this case. In re- lation to the alkali metal intercalated phase, since no charged species exist in this system, there is no effective charge doping to the FeSe layers and may be one cause as to why superconductivity is not observed.247 The current work, without powder neutron diffraction or single crystal diffraction, does not allow for complete structural analysis. However, similar previous work on hy- drothermal ammonia intercalation has showed proton intercalation plays the role of elec- trostatic stabilization and hydrogen bonding. Hydrogen bonding from the intercalated ammonia atoms was also determined to be significant through neutron powder diffraction analysis. Thus, it is likely that the hydrothermal method for ethylenediamine interca- lation may exhibit proton co-intercalation similar to the ammonia case and significant hydrogen bonding from the hydrogens attached to the nitrogen directed at the chalco- genide ions. Current ED patterns for the hydrothermal ethylenediamine synthesis exhibit clear four fold symmetry due to the Fe sublattice in the iron chalcogenide layers. This rules out the formation of iron chalcogenide chains which has been observed in simi- lar synthetic works.247,276?278 Interesingly, in previous works using neat ethylenediamine only iron chalcogenide chains form,247,276?278 whereas when a hydroxide source is added, glyercol247 or water (in the current work), iron chalocogenide layers can be formed. This hints at a fundamental role hydroxide ions play in the synthesis of extended iron chalco- genide layers as opposed to chains. Further structural analysis of organic molecule in- tercalation in iron chalcogenides is needed to understand the role metal ion coordination plays in structural stabilization and charge doping to enhance superconductivity. In the 136 Model 1 Model 2 Model 3 Na0.39(C2N2H8)0.77Fe2Se2 Li0.31(C2N2H8)0.52Fe2Se2 (C2N2H8)0.3Fe0.85Se Fig. 5.9: Proposed structures for ethylenediamine intercalated FeSe from previous works.86,247,255 Left/center depict alkali-metal stabilized ethylenediamine intercalation and right shows solvothermal alkali metal free ethylenediamine intercalated FeSe. Occupany and disor- der for the ethylenediamine molecules have been simplfied for illustrative purposes. case where metal ion coordination does not occur, hydrogen bonding must stabilize the structure which requires in-depth structural analysis through neutron and/or single crystal diffraction. 5.3.4 Magnetic and transport properties of intercalated species Intercalation of various species between FeSe and FeS layers has shown to change the properties drastically, from enhancing superconductivity to inducing magnetic order. Previous work on alkali metal co-intercalated ethylenediamine showed an increase in critical temperature for FeSe from 8 K to 45 K84,86,258,291 while superconductivity was suppressed in FeS displaying paramagnetic behavior.254 For the case of intercalated FeSe species, the increase in critical temperature is understood to be a combination of effects due to an increase in adjacent FeSe layer distance as well as electron-doping from the 137 1e?3 1.5 1e?3 1.5 1.0 0.5 1.0 0.00 20 40 60 EDA-FeS H = 30 Oe 0.5 2.5 1e?3 EDA-FeSe H = 70 Oe 2.0 3 1e?3 1.5 2 1.0 1 0 0 20 40 0.5 0 100 200 300 T (K) Fig. 5.10: Temperature dependence of magnetic susceptibility of (C2H8N2)xFeS (top) and (C2H8N2)yFeSe (bottom) at various applied fields. The insets show low temperature behavior for each species. For (C2H8N2)xFeS a broad antiferromagnetic transition is shown at 55 K while for (C2H8N2)yFeSe a clear transition is shown at 40 K which is superconducting-like in appearance at 70 Oe applied field. intercalated species. Only one intercalated FeS species has been shown to increase the critical temperature87 while the change to paramagnetic behavior is likely due to over- doping caused by the intercalated species. Figure 5.10 shows the temperature dependence of magnetic susceptibility for (C2H8N2)xFeS and (C2H8N2)yFeSe at various applied fields. The magnetic susceptibility of (C2H8N2)xFeS is shown in Figure 5.10 (top). At 138 ? (emu/Oe/g) ? (emu/Oe/g) 10 EDA-FeS T = 2 K 5 0 ?5 ?10 Field (T) EDA-FeSe T = 2 K 5 0 ?5 ?10 ?5 0 5 H (T) Fig. 5.11: Isothermal magnetization of (C2H8N2)xFeS (top) and (C2H8N2)yFeSe (bottom). (C2H8N2)xFeS (top) shows weak hysteretic behavior at 2 K owning to possible canted antiferromagnetism or some remaining net moment in the sample down to base tem- perature. (C2H8N2)yFeSe (bottom) shows antiferromagnetic behavior due to the over- whelming signal from possible impurity phases; no Meissner shielding is observed in the magnetization. 30 Oe applied field, a broad antiferromagnetic transition is observed at 55 K indicative of low-dimensional antiferromagnetic coupling. The absence of superconductivity is in agreement to other EDA intercalated FeS compounds.248,254 The isothermal magnetiza- tion for (C2H8N2)xFeS shown in Figure 5.11 (top) shows very weak hysteretic behav- ior likely due to ferrimagnetism or an uncompensated moment. Figure 5.10 (bottom) 139 M (emu/g) M (emu/g) 1.0 EDA-FeS EDA-FeSe 0.8 0.6 0.4 0.2 0 100 200 300 T (K) Fig. 5.12: Temperature dependence of resistance of pressed pellets of ground powders of C2H8N2)yFeS and (C2H8N2)xFeSe. Both samples show semiconducting behavior with no features. for (C2H8N2)yFeSe shows a clear transition in magnetic susceptibility at 70 Oe applied field. The transition at 40 K is highly reminiscent of a superconducting transition. Sev- eral previous studies on alkali metal and solvothermal ethylenediamine intercalated sam- ples yielded superconducting samples with similar critical temperature.84,86,246,258 Unfor- tunately, the parasitic contribution from the possible impurities does not allow us to see the true diamagnetic contribution and shielding volume fraction of the superconducting transition. Figure 5.11 shows the isothermal magnetization, which is dominated by the contribution from the impurity that masks the Meissner shielding. Resistivity measurements were performed on pressed pellets of ground powders of (C2H8N2)yFeS and (C2H8N2)xFeSe using the standard four-probe method with silver paint as contacts. Figure 5.12 shows the temperature dependence of resistance of the samples. Both samples show semiconducting behavior with no transitions. Ethylenediamine inter- 140 Resistancenorm calated into FeS by solvothermal methods showed similar behavior to our compounds, and pressed pellets of other FeS and intercalated FeS samples show semiconducting be- havior even if the system is known to be metallic.59,87,248,292 Previous solvothermal work on EDA-intercalated FeSe did not report resistance measurements, although supercon- ducting transitions were observed in alkali metal co-intercalated FeSe samples that were metallic in the normal state.258 Resistance measurements on pressed pellets may be highly effected by grain boundaries, surface oxidation, and high contact resistances. Annealing pressed pellets of each sample at 120 and 200 ?C had no effect on the resistance measure- ments with both samples showing semiconducting behavior and no phase transitions. 5.4 Conclusion We have successfully synthesized EDA-intercalated FeS and FeSe from a novel hydrothermal synthetic route. Previous solvothermal methods were only able to syn- thesize the selenide analogue and were not suitable for the metastable sulfide. The sta- bilization of the (C2H8N2)yFeS phase demonstrates the robustness and universality of our basic hydrothermal methods. In addition to cationic species and metal hydroxides, we have extended topotactic intercalation chemistry to small molecules (NH3 and EDA) for FeCh using a generic facile hydrothermal route. The successful synthesis of neu- tral EDA-intercalated FeSe can help with precise determination of the electron doping level in comparison to alkali metal co-intercalated FeSe. The removal of the necessity of co-intercalating alkali metals to stabilize ethylenediamine intercalation in these iron chalcogenides points at the possible exploitation of hydrogen bonding as an important 141 factor for stabilizing and directing the structure and properties of novel organic-inorganic hybrid materials. To obtain high quality superconducting samples further optimization for the synthesis, such as D2O, EDA and chalcogenide source concentrations, higher tem- perature and/or the addition of different mineralizers and solvents, is still required. 142 Chapter 6: Conclusions and Future Work 6.1 Conclusions To summarize the entire dissertation, we have systematically targeted the class of materials known as tetrahedral transition metal chalcogenides (TTMCs) due to their struc- ture and proximity to superconductivity in the iron chalcogenide family of superconduc- tors. This class of materials offers a tremendous platform for exploration of interesting chemical and physical phenomena due to its tetrahedral coordination, square transition metal sublattice, and van der Waals layered structure. Through the use of this structure, we have explored topotactic synthetic routes through hydrothermal synthesis and other chimie douce techniques to synthesize metastable phases while retaining the important building block of the TTMC family. We demonstrated that metastable layered tetragonal CoSe single crystals could be synthesized by a room temperature topochemical deintercalation of a thermodynamically stable template precursor, KCo2Se2, in highly basic aqueous media. In opposition to su- perconductivity in FeSe, CoSe displays spin glass like behavior with a signature magnetic transition at 10 K. We propose this spin glass behavior arises from magnetic frustration as opposed to conventional means due to lack of vacancies and other factors. Magnetic frus- tration in this compound likely arises from competing magnetic interaction of the square 143 transition metal sublattice as well as interlayer coupling of adjacent square lattices. Arrott plot analysis shows the moments likely lie in the ab-plane which is similar to the starting material KCo2Se2 except with a suppressed transition temperature and effective moment. Transport experiments show metallic behavior with no discernible anomoloues behavior around the transition temperature in resistivity or specific heat measurements. When revisiting the iron chalcogenide system, we have moved past simple dein- tercalation and considered topochemical ion exchange through a hydrothermal synthetic route. Previous work on the (LiOH)FeSe system showed possible magnetism coexist- ing with superconductivity due to Fe ion substitution on the Li site in the LiOH layer of the compound. We have attempted to increase that magnetic signal by substituting other transition metal cations into the LiOH layer and in turn have developed a method for tar- geted doping of the LiOH layer without altering the FeSe layers. In particular, we have observed the formation of long range magnetic order in the 20% nominally doped Mn sample which is the first time long range order has been observed in this system. Neutron diffraction measurements show the onset of magnetic order is well within the supercon- ducting state without significantly altering the superconducting properties of the system. As mentioned, we have expanded the work beyond Mn with other transition metals from Mn - Zn showing different doping behavior and physical properties depending on their chemical behavior. In total, this work shows that (LiOH)FeSe system offers a platform for guided chemical manipulation to induce the coexistence of long range magnetic order and superconductivity. Beyond the intercalation of extended solids like LiOH, we have also expanded possible adducts to include organic molecules through a hydrothermal synthetic route. 144 Ethylenediamine intercalation of FeSe and FeS have been synthesized without the need for alkali metals which were required in previous works to stabilize the structure. The benefit of this is that charge doping is easier to understand as alkali metal co-intercalation often caused the formation of multiple charged species which makes the actual doping be- havior of the interlayer species difficult to truly understand. Beyond that, this hydrother- mal method expands the possible intercalants as long as they have some solubility in water at low temperature. The intercalation of organic molecules opens the window for hydro- gen bonding as a possible mechanism for guided structure directing for how adjacent TTMC layers are stacked which could offer novel functionalities. 6.2 Future Work The work described in this thesis only encapsulates a small piece of all the projects undertaken as a graduate student at University of Maryland. Some of the ongoing work offers a logical next step with regards to expanding the work described in this thesis and some new avenues of exploration are possible as well. One project which is currently ongoing but will require significant additional work is the expansion of the phase diagram as FeSe to changed to CoSe. Our work on the deintercalation method to form single crystals of CoSe offers a pathway for the stabilization of any concentration of Fe1?xCoxSe to be formed as long as the starting material KFe2?yCoySe2 is able to be synthesized. To date, current work has been able to synthesize a number of concentrations along the phase diagram but one outstanding problem is that the starting material exhibits phase separation as true doping does not occur and KFe2Se2 and KCo2Se2 form separately as products. 145 This means that the deintercalated product may exhibit both FeSe and CoSe in different concentrations as opposed to doping. To start, a systematic study of the KFe2?yCoySe2 starting material would need to be employed in order to understand how to synthesize the material as a solid solution. From there, the process should be rather straightforward but the deintercalation conditions will likely need to be tuned depending on the cobalt concentration as the pure iron phase requires hydrothermal conditions and the pure cobalt case only requires room temperature conditions. Beyond expanding the work on FeSe-CoSe, we would like to continue exploration of expanding the class of TTMC binaries to include other late transition metals. In par- ticular, NiSe should be able to be synthesized from a reductive deintercalation reaction of KNi2Se2. Previous work and preliminary work in our lab has yet to find the correct conditions required to stabilize this compound. This composition would be particularly interesting because like the iron case, the KNi2Se2 is superconducting which means the deintercalated binary could offer a platform for additional studies of superconductivity. Along with Ni, the synthesis of any related binary or ternary phase would be interest- ing as the functionality of the square transition metal sublattice in incredible rich in its physical properties. Although two of our works have been focused on the synthesis and properties of CoSe, additional works are still to be done with regards to magnetic order in the system. All magnetic and transport measurements have shown spin glass type behavior which has been corroborated by neutron diffraction. However, we do not understand why spin glass behavior develops in this system. To date, we explain this behavior as a function of in- teraction frustration as competing magnetic interactions on the square cobalt sublattice 146 and interlayer interactions compete to lead to no long range magnetic ordering. To that end, inelastic neutron spectroscopy should be employed to determine what spin fluctua- tions exist in this material at low temperatures which can be used to understand how these competing interactions lead to the suppression of long range order. The current work on transition metal doping the (LiOH)FeSe system was limited to a maximum of nominally 20% for an initial investigation. However, due to the observation of long range magnetic order in the 20% Mn doped (LiOH)FeSe, a continuation of that work at higher doping level is warranted. Like the FeSe-CoSe diagram case, this requires that the formation of the KFe2?yMySe2 (M = Mn, Co, Ni, Cu, Zn) can be stabilized in the desired structure with doping rather than phase separation. The only two transition metals of particular interest would be Mn and Co which were shown to selectively dope the LiOH layer. Not included in this work is work done on the Cr-doped system. Preliminary work on Cr-doped KFe2Se2 shown the formation of a new layered phase at mild Cr content but we have not had the opportunity to follow up on that work. The unknown layered phase remained after hydrothermal cation exchange of the known Cr-doped KFe2Se2 phase. The intercalation of organic molecules and extended hydroxide solids have opened an avenue for the exploration of the role hydrogen bonding can play on the stacking of adjacent layers. Hydrogen bonding is weak like van der Waals bonding but is not isotropic so the directionality of the hydrogen bond may be utilized to dictate the stacking of layers in this class of materials. In doing so, we may be able to show that rational design of heterolayer materials could be done through chemical means as opposed to the current paradigm of mechanical methods. Finally, one project currently under work is the development of in-situ hydrothem- 147 ral x-ray diffraction techniques and experiments at 17-BM through collaboration with Dr. Andrey Yakovenko and Wenqian Xu. Research in our group has shown that the hydrother- mal method is a tremendously powerful technique for the synthesis of iron chalcogenides and heterolayer materials. An example time-resolved x-ray diffraction powder pattern is shown in Figure 6.1 on the in-situ synthesis of (LiOH)FeS illustrating how the formation of products from starting materials can be shown in a time-resolved manner. Fig. 6.1: Rietveld refinement results for in-situ synthesis of (LiOH)FeS showing the phase fraction of starting Fe powder and (LiOH)FeS product as a function of time with accompanying lattice parameter evolution of the (LiOH)FeS phase. Thus, we have worked to develop a method in which x-ray diffraction patterns can be collected and analyzed in a high-throughput manner so that many different reactions and conditions can be scanned. We hope this work can shed light on how these metastable phases form and are stabilized in these reaction schemes. In addition, time resolved x- ray diffraction will allow kinetic analysis of the formation of these phases which is an under-studied aspect of hydrothermal synthesis. This work offers a light into the realm of hydrothermal synthesis of metastable materials and the associated kinetics which may reveal additional interesting avenues for research and exploitation. 148 Bibliography [1] Kamerlingh Onnes, H. Commun. Phys. Lab. Univ. Leiden, b 1911, 120. [2] Meissner, W.; Ochsenfeld, R. Naturwissenschaften 1933, 21, 787?788. [3] Bardeen, J.; Cooper, L. N.; Schrieffer, J. R. Physical review 1957, 108, 1175. [4] Tinkham, M. Introduction to superconductivity; Courier Corporation, 2004. [5] Cooper, L. N. Physical Review 1956, 104, 1189. [6] Pippard, A. B.; Bragg, W. L. Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences 1953, 216, 547?568. [7] London, F.; London, H.; Lindemann, F. A. Proceedings of the Royal Society of London. Series A - Mathematical and Physical Sciences 1935, 149, 71?88. [8] Maxwell, E. Physical Review 1950, 78, 477. [9] Fr?hlich, H. Proceedings of the Physical Society. Section A 1950, 63, 778. [10] McMillan, W. Physical Review 1968, 167, 331. [11] Ashcroft, N. W. Physical Review Letters 1968, 21, 1748. 149 [12] Proust, C.; Taillefer, L. Annual Review of Condensed Matter Physics 2019, 10, 409?429. [13] Tsuei, C.; Kirtley, J. Reviews of Modern Physics 2000, 72, 969. [14] Park, C.; Snyder, R. L. Journal of the American Ceramic Society 1995, 78, 3171? 3194. [15] Norman, M.; Pepin, C. Reports on Progress in Physics 2003, 66, 1547. [16] Wu, M. K.; Ashburn, J. R.; Torng, C. J.; Hor, P. H.; Meng, R. L.; Gao, L.; Huang, Z. J.; Wang, Y. Q.; Chu, C. W. Phys. Rev. Lett. 1987, 58, 908?910. [17] Maeda, H.; Tanaka, Y.; Fukutomi, M.; Asano, T. Japanese Journal of Applied Physics 1988, 27, L209. [18] Physical Properties of High Temperature Superconductors; John Wiley and Sons Ltd, 2015; Chapter 8, pp 151?164. [19] H?fner, S.; Hossain, M.; Damascelli, A.; Sawatzky, G. Reports on Progress in Physics 2008, 71, 062501. [20] Shen, Z.-X.; Spicer, W.; King, D.; Dessau, D.; Wells, B. Science 1995, 267, 343? 350. [21] Malozemoff, A. P.; Mannhart, J.; Scalapino, D. Phys. Today 2005, 58, 41?6. [22] Kamihara, Y.; Watanabe, T.; Hirano, M.; Hosono, H. J. Amer. Chem. Soc. 2008, 130, 3296?3297. 150 [23] Johnston, D. C. Adv. Phys. 2010, 59, 803?1061. [24] Ishida, K.; Nakai, Y.; Hosono, H. J. Phys. Soc. Jpn. 2009, 78, 062001?062001. [25] Wen, H.-H.; Li, S. Annu. Rev. Condens. Matter Phys. 2011, 2, 121?140. [26] Wen, J.; Xu, G.; Gu, G.; Tranquada, J.; Birgeneau, R. Rep. Prog. Phys 2011, 74, 124503. [27] Paglione, J.; Greene, R. L. Nature Physics 2010, 6, 645. [28] Mazin, I. I. Nature 2010, 464, 183. [29] Yang, W.; Sorini, A.; Chen, C.; Moritz, B.; Lee, W.-S.; Vernay, F.; Olalde- Velasco, P.; Denlinger, J.; Delley, B.; Chu, J.-H., et al. Physical Review B 2009, 80, 014508. [30] Lu, D.; Yi, M.; Mo, S.-K.; Erickson, A.; Analytis, J.; Chu, J.-H.; Singh, D.; Hus- sain, Z.; Geballe, T.; Fisher, I., et al. Nature 2008, 455, 81. [31] Lee, P. A.; Nagaosa, N.; Wen, X.-G. Reviews of modern physics 2006, 78, 17. [32] Anderson, P. W. science 1987, 235, 1196?1198. [33] Zhang, F.; Rice, T. Physical Review B 1988, 37, 3759. [34] Stewart, G. Reviews of Modern Physics 2011, 83, 1589. [35] Hirschfeld, P.; Korshunov, M.; Mazin, I. Reports on Progress in Physics 2011, 74, 124508. 151 [36] Cvetkovic, V.; Tesanovic, Z. EPL (Europhysics Letters) 2009, 85, 37002. [37] Yin, Z.; Haule, K.; Kotliar, G. Nature materials 2011, 10, 932. [38] Medici, L. d. arXiv preprint arXiv:1707.03282 2017, [39] Georges, A.; de?Medici, L.; Mravlje, J. 2013, [40] Fanfarillo, L.; Bascones, E. Physical Review B 2015, 92, 075136. [41] Lanata, N.; Strand, H. U.; Giovannetti, G.; Hellsing, B.; de?Medici, L.; Capone, M. Physical Review B 2013, 87, 045122. [42] Hsu, F.-C.; Luo, J.-Y.; Yeh, K.-W.; Chen, T.-K.; Huang, T.-W.; Wu, P. M.; Lee, Y.- C.; Huang, Y.-L.; Chu, Y.-Y.; Yan, D.-C.; Wu, M.-K. Proc. Natl. Acad. Sci. USA 2008, 105, 14262?14264. [43] Kotegawa, H.; Masaki, S.; Awai, Y.; Tou, H.; Mizuguchi, Y.; Takano, Y. J. Phys. Soc. Jpn. 2008, 77, 113703. [44] McQueen, T. M.; Huang, Q.; Ksenofontov, V.; Felser, C.; Xu, Q.; Zandbergen, H.; Hor, Y. S.; Allred, J.; Williams, A. J.; Qu, D., et al. Physical Review B 2009, 79, 014522. [45] Margadonna, S.; Takabayashi, Y.; McDonald, M. T.; Kasperkiewicz, K.; Mizuguchi, Y.; Takano, Y.; Fitch, A. N.; Suard, E.; Prassides, K. Chem. Comm. 2008, 5607?5609. 152 [46] Burrard-Lucas, M.; Free, D. G.; Sedlmaier, S. J.; Wright, J. D.; Cassidy, S. J.; Hara, Y.; Corkett, A. J.; Lancaster, T.; Baker, P. J.; Blundell, S. J., et al. Nature Materials 2013, 12, 15. [47] Chen, W.; Zeng, C.; Kaxiras, E.; Zhang, Z. Physical Review B 2016, 93, 064517. [48] Sun, H.; Woodruff, D. N.; Cassidy, S. J.; Allcroft, G. M.; Sedlmaier, S. J.; Thomp- son, A. L.; Bingham, P. A.; Forder, S. D.; Cartenet, S.; Mary, N., et al. Inorganic Chemistry 2015, 54, 1958?1964. [49] Margadonna, S.; Takabayashi, Y.; Ohishi, Y.; Mizuguchi, Y.; Takano, Y.; Ka- gayama, T.; Nakagawa, T.; Takata, M.; Prassides, K. Physical Review B 2009, 80, 064506. [50] He, S.; He, J.; Zhang, W.; Zhao, L.; Liu, D.; Liu, X.; Mou, D.; Ou, Y.-B.; Wang, Q.- Y.; Li, Z., et al. Nat. Mater. 2013, 12, 605. [51] B?hmer, A. E.; Arai, T.; Hardy, F.; Hattori, T.; Iye, T.; Wolf, T.; L?hneysen, H. v.; Ishida, K.; Meingast, C. Physical Review Letters 2015, 114, 027001. [52] Wang, Q.; Shen, Y.; Pan, B.; Zhang, X.; Ikeuchi, K.; Iida, K.; Christianson, A.; Walker, H.; Adroja, D.; Abdel-Hafiez, M., et al. Nature Communications 2016, 7, 12182. [53] Rahn, M. C.; Ewings, R. A.; Sedlmaier, S. J.; Clarke, S. J.; Boothroyd, A. T. Physical Review B 2015, 91, 180501(R). 153 [54] Wang, Q.; Shen, Y.; Pan, B.; Hao, Y.; Ma, M.; Zhou, F.; Steffens, P.; Schmalzl, K.; Forrest, T.; Abdel-Hafiez, M., et al. Nature Materials 2016, 15, 159. [55] Pomjakushina, E.; Conder, K.; Pomjakushin, V.; Bendele, M.; Khasanov, R. Phys. Rev. B 2009, 80, 024517. [56] Koz, C.; Schmidt, M.; Borrmann, H.; Burkhardt, U.; Robler, S.; Carrillo- Cabrera, W.; Schnelle, W.; Schwarz, U.; Grin, Y. Z. Anorg. Allg. Chem 2014, 640, 1600?1606. [57] Greenfield, J. T.; Kamali, S.; Lee, K.; Kovnir, K. Chem. Mater. 2015, 27, 588?596. [58] Lai, X.; Zhang, H.; Wang, Y.; Wang, X.; Zhang, X.; Lin, J.; Huang, F. J. Am. Chem. Soc. 2015, 137, 10148?10151. [59] Borg, C. K. H.; Zhou, X.; Eckberg, C.; Campbell, D. J.; Saha, S. R.; Paglione, J.; Rodriguez, E. E. Phys. Rev. B 2016, 93, 094522. [60] Evans Jr, H. T.; Milton, C.; Chao, E.; Adler, I.; Mead, C.; Ingram, B.; Berner, R. A. US Geological Survey Professional Paper 1964, 475, 64?69. [61] Zhou, X.; Wilfong, B.; Vivanco, H.; Paglione, J.; Brown, C. M.; Rodriguez, E. E. J. Am. Chem. Soc 2016, 138, 16432?16442, PMID: 27935693. [62] Schuster, W.; Mikler, H.; Komarek, K. L. Monatshefte f?r Chemie/Chemical Monthly 1979, 110, 1153?1170. [63] Zhou, X.; Rodriguez, E. E. Chem. Mater. 2017, 29, 5737?5752. 154 [64] Hoffmann, R.; Zheng, C. The Journal of Physical Chemistry 1985, 89, 4175?4181. [65] Zheng, C.; Hoffmann, R. Journal of Solid State Chemistry 1988, 72, 58 ? 71. [66] Hoffmann, R. Angewandte Chemie International Edition in English 1987, 26, 846? 878. [67] Burdett, J. K. Progress in Solid State Chemistry 1984, 15, 173 ? 255. [68] Papoian, G.; Hoffmann, R. Journal of the American Chemical Society 2001, 123, 6600?6608, PMID: 11439046. [69] Tremel, W.; Hoffmann, R. Journal of the American Chemical Society 1987, 109, 124?140. [70] Chandra, P.; Doucot, B. Phys. Rev. B 1988, 38, 9335?9338. [71] Shannon, N.; Momoi, T.; Sindzingre, P. Phys. Rev. Lett. 2006, 96, 027213. [72] Vannimenus, J.; Toulouse, G. Journal of Physics C: Solid State Physics 1977, 10, L537?L542. [73] Skoulatos, M.; Goff, J. P.; Geibel, C.; Kaul, E. E.; Nath, R.; Shannon, N.; Schmidt, B.; Murani, A. P.; Deen, P. P.; Enderle, M.; Wildes, A. R. EPL (Euro- physics Letters) 2009, 88, 57005. [74] Schmidt, B.; Thalmeier, P.; Shannon, N. Physical review B 2007, 76, 125113. [75] Wilfong, B.; Zhou, X.; Vivanco, H.; Campbell, D. J.; Wang, K.; Graf, D.; Paglione, J.; Rodriguez, E. Phys. Rev. B 2018, 97, 104408. 155 [76] Dagotto, E. Rev. Mod. Phys. 2013, 85, 849. [77] Guo, J.; Jin, S.; Wang, G.; Wang, S.; Zhu, K.; Zhou, T.; He, M.; Chen, X. Phys. Rev. B. 2010, 82, 180520. [78] Yan, X.-W.; Gao, M.; Lu, Z.-Y.; Xiang, T. Phys. Rev. B. 2011, 84, 054502. [79] Dong, X.; Jin, K.; Yuan, D.; Zhou, H.; Yuan, J.; Huang, Y.; Hua, W.; Sun, J.; Zheng, P.; Hu, W.; Mao, Y.; Ma, M.; Zhang, G.; Zhou, F.; Zhao, Z. Physical Review B 2015, 92, 064515. [80] Pachmayr, U.; Nitsche, F.; Luetkens, H.; Kamusella, S.; Br?ckner, F.; Sarkar, R.; Klauss, H.-H.; Johrendt, D. Angewandte Chemie International Edition 2015, 54, 293?297. [81] Lynn, J. W.; Zhou, X.; Borg, C. K. H.; Saha, S. R.; Paglione, J.; Rodriguez, E. E. Physical Review B 2015, 92, 060510(R). [82] Zhou, X.; Borg, C. K. H.; Lynn, J. W.; Saha, S. R.; Paglione, J.; Rodriguez, E. E. J. Mater. Chem. C 2016, 4, 3934?3941. [83] Hosono, S.; Noji, T.; Hatakeda, T.; Kawamata, T.; Kato, M.; Koike, Y. J. Phys. Soc. Jpn. 2014, 83, 113704. [84] Noji, T.; Hatakeda, T.; Hosono, S.; Kawamata, T.; Kato, M.; Koike, Y. Phys. C: Supercon. and App. 2014, 504, 8?11. [85] Krzton-Maziopa, A.; Pomjakushina, E.; Pomjakushin, V. Y.; Von Rohr, F.; Schilling, A.; Conder, K. J. Phys. Condens. Matter 2012, 24, 382202. 156 [86] Jin, S.; Fan, X.; Wu, X.; Sun, R.; Wu, H.; Huang, Q.; Shi, C.; Xi, X.; Li, Z.; Chen, X. Chem. Comm. 2017, 53, 9729?9732. [87] Zhou, X.; Eckberg, C.; Wilfong, B.; Liou, S.-C.; Vivanco, H. K.; Paglione, J.; Rodriguez, E. E. Chem. Sci. 2017, 8, 3781?3788. [88] Li, Z.; Pradeep, K. G.; Deng, Y.; Raabe, D.; Tasan, C. C. Nature 2016, 534, 227? 230. [89] Stein, A.; Keller, S. W.; Mallouk, T. E. Science 1993, 259, 1558?1564. [90] Gopalakrishnan, J. Chem. Mater. 1995, 7, 1265?1275. [91] Pervov, V.; Manokhina, E.; Dobrokhotova, Z. V.; Zotova, A.; Zavrazhnov, A. Y. Inorganic materials 2011, 47, 1407?1427. [92] Berner, R. A. The Journal of Geology 1964, 72, 293?306. [93] Lennie, A. R.; Redfern, S. A.; Champness, P. E.; Stoddart, C. P.; Schofield, P. F.; Vaughan, D. J. American Mineralogist 1997, 82, 302?309. [94] Lennie, A.; Redfern, S. A.; Schofield, P.; Vaughan, D. Mineralogical Magazine 1995, 59, 677?683. [95] Rouxel, J.; Tournoux, M. Solid State Ionics 1996, 84, 141?149. [96] Livage, J. New Journal of Chemistry 2001, 25, 1?1. [97] Corriu, R. J. New Journal of Chemistry 2001, 25, 2?2. 157 [98] R?dorff, W. Advances in inorganic chemistry and radiochemistry; Elsevier, 1959; Vol. 1; pp 223?266. [99] DiSalvo, F.; Safran, S.; Haddon, R.; Waszczak, J.; Fischer, J. Physical Review B 1979, 20, 4883. [100] Gamble, F.; Osiecki, J.; Cais, M.; Pisharody, R.; DiSalvo, F.; Geballe, T. Science 1971, 174, 493?497. [101] Whittingha, S. M. Intercalation chemistry; Elsevier, 2012. [102] Dresselhaus, M.; Dresselhaus, G. Advances in Physics 1981, 30, 139?326. [103] Whittingham, M. S. Science 1976, 192, 1126?1127. [104] Li, Y. J.; Whittingham, M. S. Solid State Ionics 1993, 63, 391?395. [105] Whittingham, M. S. Current Opinion in Solid State and Materials Science 1996, 1, 227?232. [106] Rajamathi, M.; Seshadri, R. Current Opinion in Solid State and Materials Science 2002, 6, 337?345. [107] Yaghi, O.; Li, H. Journal of the American Chemical Society 1995, 117, 10401? 10402. [108] Yang, S.; Zavalij, P. Y.; Whittingham, M. S. Electrochemistry Communications 2001, 3, 505?508. [109] Feng, S.; Xu, R. Accounts of chemical research 2001, 34, 239?247. 158 [110] Norton, D. L. Annual Review of Earth and Planetary Sciences 1984, 12, 155?177. [111] Tarasov, V.; Gebruk, A.; Mironov, A.; Moskalev, L. Chemical Geology 2005, 224, 5?39. [112] Martin, W.; Baross, J.; Kelley, D.; Russell, M. J. Nature Reviews Microbiology 2008, 6, 805?814. [113] Russell, M.; Hall, A.; Cairns-Smith, A.; Braterman, P. Nature 1988, 336, 117?117. [114] Byrappa, K. Springer Handbook of Crystal Growth; Springer, 2010; pp 599?653. [115] Byrappa, K.; Yoshimura, M. Handbook of hydrothermal technology; William An- drew, 2012. [116] Rabenau, A. Angewandte Chemie International Edition in English 1985, 24, 1026? 1040. [117] Rudolph, P. Handbook of crystal growth: Bulk crystal growth; Elsevier, 2014. [118] Channei, D.; Phanichphant, S.; Nakaruk, A.; Mofarah, S. S.; Koshy, P.; Sor- rell, C. C. Catalysts 2017, 7, 45. [119] Bonomi, F.; Werth, M. T.; Kurtz, D. M. Inorganic Chemistry 1985, 24, 4331?4335. [120] Zhou, X. Intercalation Chemistry of Tetrahedral Transition Metal Chalcogenides. Ph.D. thesis, 2018; Copyright - Database copyright ProQuest LLC; ProQuest does not claim copyright in the individual underlying works; Last updated - 2019-10-19. 159 [121] Zhou, X.; Wang, L.; Fan, X.; Wilfong, B.; Liou, S.-C.; Wang, Y.; Zheng, H.; Feng, Z.; Wang, C.; Rodriguez, E. E. Chemistry of Materials 2020, 32, 769?775. [122] Neilson, J. R.; McQueen, T. M. Journal of the American Chemical Society 2012, 134, 7750?7757. [123] Vivanco, H. K.; Rodriguez, E. E. J. Solid State Chem 2016, 242, Part 2, 3 ? 21, Solid State Chemistry of Energy-Related Materials. [124] Lu, X.; Wang, N.; Wu, H.; Wu, Y.; Zhao, D.; Zeng, X.; Luo, X.; Wu, T.; Bao, W.; Zhang, G., et al. Nature Materials 2015, 14, 325. [125] Shoemaker, D. P.; Chung, D. Y.; Claus, H.; Francisco, M. C.; Avci, S.; Llobet, A.; Kanatzidis, M. G. Physical Review B 2012, 86, 184511. [126] Coelho, A. A. Journal of Applied Crystallography 2018, 51, 210?218. [127] Cheary, R. W.; Coelho, A. Journal of Applied Crystallography 1992, 25, 109?121. [128] Pecharsky, V.; Zavalij, P. Fundamentals of powder diffraction and structural char- acterization of materials; Springer Science & Business Media, 2008. [129] Squires, G. L. Introduction to the theory of thermal neutron scattering; Courier Corporation, 1996. [130] Pynn, R. Neutron Scattering: A Primer; Los Alamos Neutron Science Center, 1990. [131] Spaldin, N. A. Magnetic materials: fundamentals and applications; Cambridge university press, 2010. 160 [132] Blundell, S. Magnetism in Condensed Matter; Magnetism in Condensed Matter, 2001. [133] MPMS 3 User?s Manual. Quantum Design: 11578 Sorrento Valley Rd., San Diego, CA 92121-1311 USA, 2016. [134] Kittel, C. Introduction to Solid State Physics; Wiley, 2005. [135] Kleiner, R.; Koelle, D.; Ludwig, F.; Clarke, J. Proceedings of the IEEE 2004, 92, 1534?1548. [136] PPMS DynaCool Options Manual. Quantum Design: 11578 Sorrento Valley Rd., San Diego, CA 92121-1311 USA, 2014. [137] Mydosh, J. A. Spin Glasses: An Experimental Introduction; Taylor and Francis, 1993. [138] Mydosh, J. J. Magn. Magn. Mater. 1996, 157, 606?610. [139] Binder, K.; Young, A. P. Rev. Mod. Phys 1986, 58, 801. [140] Physical Property Measurement System AC Measurement System (ACMS) Op- tion User?s Manual. Quantum Design: 11578 Sorrento Valley Rd., San Diego, CA 92121-1311 USA, 2003. [141] Physical Property Measurement System Resistivity Option User?s Manual. Quan- tum Design: 11578 Sorrento Valley Rd., San Diego, CA 92121-1311 USA, 1999. 161 [142] Physical Property Measurement System Heat Capacity Option User?s Manual. Quantum Design: 11578 Sorrento Valley Rd., San Diego, CA 92121-1311 USA, 2004. [143] Bachmann, R.; DiSalvo, F. J.; Geballe, T. H.; Greene, R. L.; Howard, R. E.; King, C. N.; Kirsch, H. C.; Lee, K. N.; Schwall, R. E.; Thomas, H.; Zubeck, R. B. Rev. Sci. Instrum. 1972, 43, 205?214. [144] Hohenberg, P.; Kohn, W. Phys. Rev. 1964, 136, B864?B871. [145] Kohn, W.; Sham, L. J. Phys. Rev. 1965, 137, A1697?A1705. [146] Kresse, G. Title. Ph.D. thesis, Technische Universi?t Wien, 1993. [147] Kresse, G.; Hafner, J. Phys. Rev. B 1993, 47, 558?561. [148] Kresse, G.; Furthm?ller, J. Comput. Mater. Sci. 1996, 6, 15 ? 50. [149] Kresse, G.; Furthm?ller, J. Phys. Rev. B 1996, 54, 11169?11186. [150] Bl?chl, P. E. Phys. Rev. B 1994, 50, 17953?17979. [151] Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett. 1996, 77, 3865?3868. [152] Monkhorst, H. J.; Pack, J. D. Phys. Rev. B 1976, 13, 5188?5192. [153] Martin, R. M.; Martin, R. M. Electronic structure: basic theory and practical meth- ods; Cambridge university press, 2004. [154] Koch, W.; Holthausen, M. C. A chemist?s guide to density functional theory; John Wiley & Sons, 2015. 162 [155] Williams, A. J.; McQueen, T. M.; Ksenofontov, V.; Felser, C.; Cava, R. J. J. Phys. Condens. Matter 2009, 21, 305701. [156] Wu, M.; Hsu, F.; Yeh, K.; Huang, T.; Luo, J.; Wang, M.; Chang, H.; Chen, T.; Rao, S.; Mok, B., et al. Physica C Supercond. 2009, 469, 340?349. [157] Urata, T.; Tanabe, Y.; Huynh, K. K.; Oguro, H.; Watanabe, K.; Tanigaki, K. arXiv preprint arXiv:1608.01044 2016, [158] Mizuguchi, Y.; Tomioka, F.; Tsuda, S.; Yamaguchi, T.; Takano, Y. J. Phys. Soc. Jpn. 2009, 78, 074712?074712. [159] Sines, I. T.; II, D. D. V.; Misra, R.; Popczun, E. J.; Schaak, R. E. J. Solid State Chem. 2012, 196, 17 ? 20. [160] Denholme, S.; Demura, S.; Okazaki, H.; Hara, H.; Deguchi, K.; Fujioka, M.; Ozaki, T.; Yamaguchi, T.; Takeya, H.; Takano, Y. Mater. Chem. Phys. 2014, 147, 50 ? 56. [161] Lei, H.; Abeykoon, M.; Bozin, E. S.; Petrovic, C. Phys. Rev. B 2011, 83, 180503. [162] Oledzka, M.; Ramanujachary, K.; Greenblatt, M. Mater. Res. Bull. 1998, 33, 855 ? 866. [163] Oledzka, M.; Lee, J.-G.; Ramanujachary, K.; Greenblatt, M. J. Solid State Chem. 1996, 127, 151 ? 160. [164] Huan, G.; Greenblatt, M.; Croft, M. Eur. J. Solid State Inorg. Chem. 1989, 26, 193?220. 163 [165] Bouroushian, M. Electrochemistry of metal chalcogenides; Springer, 2010; pp 57? 75. [166] Wen, H.-H.; Li, S. Annu. Rev. Condens. Matter Phys. 2011, 2, 121?140. [167] Rodriguez, E. E.; Stock, C.; Zajdel, P.; Krycka, K. L.; Majkrzak, C. F.; Zavalij, P.; Green, M. A. Phys. Rev. B 2011, 84, 064403. [168] Rodriguez, E. E.; Sokolov, D. A.; Stock, C.; Green, M. A.; Sobolev, O.; Rodriguez- Rivera, J. A.; Cao, H.; Daoud-Aladine, A. Phys. Rev. B 2013, 88, 165110. [169] Stock, C.; Rodriguez, E. E.; Green, M. A. Phys. Rev. B 2012, 85, 094507. [170] Zhang, X.; Oh, Y. S.; Liu, Y.; Yan, L.; Saha, S. R.; Butch, N. P.; Kirshenbaum, K.; Kim, K. H.; Paglione, J.; Greene, R. L.; Takeuchi, I. Phys. Rev. B 2010, 82, 020515. [171] Quirinale, D. G.; Anand, V. K.; Kim, M. G.; Pandey, A.; Huq, A.; Stephens, P. W.; Heitmann, T. W.; Kreyssig, A.; McQueeney, R. J.; Johnston, D. C.; Goldman, A. I. Phys. Rev. B 2013, 88, 174420. [172] Anand, V. K.; Dhaka, R. S.; Lee, Y.; Harmon, B. N.; Kaminski, A.; Johnston, D. C. Phys. Rev. B 2014, 89, 214409. [173] Jayasekara, W.; Lee, Y.; Pandey, A.; Tucker, G. S.; Sapkota, A.; Lamsal, J.; Calder, S.; Abernathy, D. L.; Niedziela, J. L.; Harmon, B. N.; Kreyssig, A.; Vaknin, D.; Johnston, D. C.; Goldman, A. I.; McQueeney, R. J. Phys. Rev. Lett. 2013, 111, 157001. 164 [174] Pandey, A.; Quirinale, D.; Jayasekara, W.; Sapkota, A.; Kim, M.; Dhaka, R.; Lee, Y.; Heitmann, T.; Stephens, P.; Ogloblichev, V., et al. Phys. Rev. B 2013, 88, 014526. [175] Anand, V. K.; Quirinale, D. G.; Lee, Y.; Harmon, B. N.; Furukawa, Y.; Ogloblichev, V. V.; Huq, A.; Abernathy, D. L.; Stephens, P. W.; McQueeney, R. J.; Kreyssig, A.; Goldman, A. I.; Johnston, D. C. Phys. Rev. B 2014, 90, 064517. [176] Mansart, J.; Le F?vre, P.; Bertran, F. m. c.; Forget, A.; Colson, D.; Brouet, V. Phys. Rev. B 2016, 94, 235147. [177] Kovnir, K.; Garlea, V. O.; Thompson, C. M.; Zhou, H.; Reiff, W. M.; Ozarowski, A.; Shatruk, M. Inorg. Chem. 2011, 50, 10274?10283. [178] Tan, X.; Fabbris, G.; Haskel, D.; Yaroslavtsev, A. A.; Cao, H.; Thompson, C. M.; Kovnir, K.; Menushenkov, A. P.; Chernikov, R. V.; Garlea, V. O.; Shatruk, M. J. Am. Chem. Soc. 2016, 138, 2724?2731. [179] Physical Property Measurement System Heat Capacity Option User?s Manual. Quantum Design: 11578 Sorrento Valley Rd., San Diego, CA 92121-1311 USA, 2004. [180] Hwang, J. S.; Lin, K. J.; Tien, C. Rev. Sci. Instrum. 1997, 68, 94?101. [181] Balents, L. Nature 2010, 464, 199?208. [182] Shtrikman, S.; Wohlfarth, E. Phys. Lett. A 1981, 85, 467?470. [183] Souletie, J.; Tholence, J. L. Phys. Rev. B 1985, 32, 516?519. 165 [184] Ogielski, A. T.; Morgenstern, I. Phys. Rev. Lett. 1985, 54, 928?931. [185] Ogielski, A. T. Phys. Rev. B 1985, 32, 7384?7398. [186] Yang, J.; Chen, B.; Wang, H.; Mao, Q.; Imai, M.; Yoshimura, K.; Fang, M. Phys. Rev. B 2013, 88, 064406. [187] Yelland, E.; Yates, S.; Taylor, O.; Griffiths, A.; Hayden, S.; Carrington, A. Phys. Rev. B 2005, 72, 184436. [188] Bhattacharyya, A.; Jain, D.; Ganesan, V.; Giri, S.; Majumdar, S. Phys. Rev. B 2011, 84, 184414. [189] Liu, Z.; Zhao, Y.; Li, Y.; Jia, L.; Cai, Y.; Zhou, S.; Xia, T.; B?chner, B.; Borisenko, S.; Wang, S. J. Phys. Condens. Matter 2015, 27, 295501. [190] Kaul, S. N. J. Phys.: Condens. Matter 1999, 11, 7597. [191] Moriya, T. J. Magn. Magn. Mater. 1979, 14, 1?46. [192] Tari, A. The Specific Heat of Matter at Low Temperatures; World Scientific, 2003. [193] Mohn, P. Magnetism in the Solid State: An Introduction; Springer Science & Busi- ness Media, 2006; Vol. 134. [194] Johnston, D. C. Phys. Rev. Lett. 2012, 109, 077201. [195] Yu, R.; Si, Q. Phys. Rev. Lett. 2015, 115, 116401. [196] Ma, F.; Ji, W.; Hu, J.; Lu, Z.-Y.; Xiang, T. Phys. Rev. Lett. 2009, 102, 177003. 166 [197] Wang, F.; Kivelson, S. A.; Lee, D.-H. Nature Phys. 2015, 11. [198] Glasbrenner, J.; Mazin, I.; Jeschke, H. O.; Hirschfeld, P.; Fernandes, R.; Valent?, R. Nature Phys. 2015, 11, 953?958. [199] Shannon, N.; Schmidt, B.; Penc, K.; Thalmeier, P. Eur. Phys. J. B 2004, 38, 599? 616. [200] Liu, Z.-H.; Richard, P.; Xu, N.; Xu, G.; Li, Y.; Fang, X.-C.; Jia, L.-L.; Chen, G.-F.; Wang, D.-M.; He, J.-B., et al. Phys. Rev. Lett. 2012, 109, 037003. [201] Qian, T.; Wang, X.-P.; Jin, W.-C.; Zhang, P.; Richard, P.; Xu, G.; Dai, X.; Fang, Z.; Guo, J.-G.; Chen, X.-L.; Ding, H. Phys. Rev. Lett. 2011, 106, 187001. [202] Wang, G.; Yi, X.; Shi, X. Physics Letters A 2015, 379, 2106?2109. [203] Zhang, G.-Y.; Chou, M.; Lin, C.-T. Crystals 2017, 7, 167. [204] Topping, C. V.; Kirschner, F. K. K.; Blundell, S. J.; Baker, P. J.; Woodruff, D. N.; Schild, F.; Sun, H.; Clarke, S. J. Physical Review B 2017, 95, 134419. [205] Wu, Y. P.; Zhao, D.; Lian, X. R.; Lu, X. F.; Wang, N. Z.; Luo, X. G.; Chen, X. H.; Wu, T. Physical Review B 2015, 91, 125107. [206] Lynn, J.; Moncton, D.; Thomlinson, W.; Shirane, G.; Shelton, R. Solid State Com- munications 1978, 26, 493?496. [207] Lynn, J.; Moncton, D.; Shirane, G.; Thomlinson, W.; Eckert, J.; Shelton, R. Journal of Applied Physics 1978, 49, 1389?1391. 167 [208] Moncton, D. E.; Shirane, G.; Thomlinson, W.; Ishikawa, M.; Fischer, O. Physical Review Letters 1978, 41, 1133. [209] Lynn, J.; Shelton, R.; Horng, H.-E.; Glinka, C. Physica B+ C 1983, 120, 224?226. [210] Lynn, J. W.; Skanthakumar, S.; Huang, Q.; Sinha, S. K.; Hossain, Z.; Gupta, L. C.; Nagarajan, R.; Godart, C. Phys. Rev. B 1997, 55, 6584?6598. [211] Gupta, L. Advances in Physics 2006, 55, 691?798. [212] Grigereit, T. E.; Lynn, J. W.; Huang, Q.; Santoro, A.; Cava, R. J.; Krajewski, J. J.; Peck, W. F. Phys. Rev. Lett. 1994, 73, 2756?2759. [213] Lynn, J. W.; Keimer, B.; Ulrich, C.; Bernhard, C.; Tallon, J. L. Phys. Rev. B 2000, 61, R14964?R14967. [214] Tokunaga, Y.; Kotegawa, H.; Ishida, K.; Kitaoka, Y.; Takagiwa, H.; Akimitsu, J. Physical Review Letters 2001, 86, 5767. [215] Lynn, J. W.; Clinton, T. W.; Li, W.-H.; Erwin, R. W.; Liu, J. Z.; Vandervoort, K.; Shelton, R. N. Phys. Rev. Lett. 1989, 63, 2606?2609. [216] Lynn, J. W.; Dai, P. Physica C: Superconductivity 2009, 469, 469?476. [217] Xiao, Y.; Su, Y.; Meven, M.; Mittal, R.; Kumar, C. M. N.; Chatterji, T.; Price, S.; Persson, J.; Kumar, N.; Dhar, S. K.; Thamizhavel, A.; Brueckel, T. Physical Review B 2009, 80, 174424. [218] Nandi, S.; Jin, W. T.; Xiao, Y.; Su, Y.; Price, S.; Shukla, D. K.; Strempfer, J.; Jeevan, H. S.; Gegenwart, P.; Br?ckel, T. Physical Review B 2014, 89, 014512. 168 [219] Bao, J.-K.; Willa, K.; Smylie, M. P.; Chen, H.; Welp, U.; Chung, D. Y.; Kanatzidis, M. G. Crystal Growth & Design 2018, 18, 3517?3523. [220] Stolyarov, V. S.; Veshchunov, I. S.; Grebenchuk, S. Y.; Baranov, D. S.; Golovchan- skiy, I. A.; Shishkin, A. G.; Zhou, N.; Shi, Z.; Xu, X.; Pyon, S., et al. Science Advances 2018, 4, eaat1061. [221] Watson, M. D.; Kim, T. K.; Rhodes, L. C.; Eschrig, M.; Hoesch, M.; Haghighi- rad, A. A.; Coldea, A. I. Physical Review B 2016, 94, 201107(R). [222] Watson, M. D.; Kim, T. K.; Haghighirad, A. A.; Davies, N. R.; McCollam, A.; Narayanan, A.; Blake, S. F.; Chen, Y. L.; Ghannadzadeh, S.; Schofield, A. J., et al. Physical Review B 2015, 91, 155106. [223] Chubukov, A. V.; Fernandes, R. M.; Schmalian, J. Physical Review B 2015, 91, 201105(R). [224] Qin, S.; Hu, L.; Wu, X.; Dai, X.; Fang, C.; Zhang, F.-c.; Hu, J. arXiv preprint arXiv:1901.03120 2019, [225] Liu, Q. et al. Physical Review X 2018, 8, 041056. [226] Ma, M.; Wang, L.; Bourges, P.; Sidis, Y.; Danilkin, S.; Li, Y. Physical Review B 2017, 95, 100504(R). [227] Calder, S. et al. Review of Scientific Instruments 2018, 89, 092701. [228] Davies, N. R.; Rahn, M. C.; Walker, H. C.; Ewings, R. A.; Woodruff, D. N.; Clarke, S. J.; Boothroyd, A. T. Physical Review B 2016, 94, 144503. 169 [229] Pan, B.; Shen, Y.; Hu, D.; Feng, Y.; Park, J.; Christianson, A.; Wang, Q.; Hao, Y.; Wo, H.; Yin, Z. Nature Communications 2017, 8, 123. [230] Woodruff, D. N.; Schild, F.; Topping, C. V.; Cassidy, S. J.; Blandy, J. N.; Blun- dell, S. J.; Thompson, A. L.; Clarke, S. J. Inorganic Chemistry 2016, 55, 9886? 9891. [231] Gu, Q.; Tang, Q.; Wan, S.; Du, Z.; Yang, X.; Yang, H.; Wang, Q.-H.; Lin, H.; Zhu, X.; Wen, H.-H. Physical Review B 2018, 98, 134503. [232] Du, Z.; Yang, X.; Altenfeld, D.; Gu, Q.; Yang, H.; Eremin, I.; Hirschfeld, P. J.; Mazin, I. I.; Lin, H.; Zhu, X., et al. Nature Physics 2018, 14, 134. [233] Zhou, H.; Ni, S.; Yuan, J.; Li, J.; Feng, Z.; Jiang, X.; Huang, Y.; Liu, S.; Mao, Y.; Zhou, F., et al. Chinese Physics B 2017, 26, 057402. [234] Mao, Y.; Li, Z.; Zhou, H.; Ma, M.; Chai, K.; Ni, S.; Liu, S.; Tian, J.; Huang, Y.; Yuan, J., et al. Chinese Physics B 2018, 27, 077405. [235] Prozorov, R.; Kogan, V. G. Physical Review Applied 2018, 10, 014030. [236] Urban, C.; Valmianski, I.; Pachmayr, U.; Basaran, A. C.; Johrendt, D.; Schuller, I. K. Physical Review B 2018, 97, 024516. [237] Khasanov, R.; Zhou, H.; Amato, A.; Guguchia, Z.; Morenzoni, E.; Dong, X.; Zhang, G.; Zhao, Z. Physical Review B 2016, 93, 224512. [238] Wang, C.; Yi, X.; Sun, X.; Tang, Q.; Qiu, Y.; Luo, Y.; Yu, B. Superconductor Science and Technology 2017, 30, 085004. 170 [239] Du, Z.; Yang, X.; Lin, H.; Fang, D.; Du, G.; Xing, J.; Yang, H.; Zhu, X.; Wen, H.- H. Nature Communications 2016, 7, 10565. [240] Wang, C.; Yi, X.; Qiu, Y.; Tang, Q.; Zhang, X.; Luo, Y.; Yu, B. Superconductor Science and Technology 2016, 29, 055003. [241] Lin, H.; Xing, J.; Zhu, X.; Yang, H.; Wen, H.-H. Science China Physics, Mechanics & Astronomy 2016, 59, 657404. [242] Wang, Z.; Yuan, J.; Wosnitza, J.; Zhou, H.; Huang, Y.; Jin, K.; Zhou, F.; Dong, X.; Zhao, Z. Journal of Physics: Condensed Matter 2016, 29, 025701. [243] Yi, X.; Wang, C.; Tang, Q.; Peng, T.; Qiu, Y.; Xu, J.; Sun, H.; Luo, Y.; Yu, B. Superconductor Science and Technology 2016, 29, 105015. [244] Lu, X. F.; Wang, N. Z.; Luo, X. G.; Zhang, G. H.; Gong, X. L.; Huang, F. Q.; Chen, X. H. Physical Review B 2014, 90, 214520. [245] Cava, R. Science 1990, 247, 656?662. [246] Gao, Z.; Zeng, S.; Zhu, B.; Li, B.; Hao, Q.; Hu, Y.; Wang, D.; Tang, K. Sci. China Mater. 2018, 61, 977?984. [247] Stahl, J.; Shlaen, E.; Singer, H.; Johrendt, D. Dalton Transactions 2018, 47, 3264? 3271. [248] Wu, D.; Guo, Z.; Liu, N.; Zhou, L.; Mao, Y.; Wan, L.; Sun, F.; Yuan, W. Inorg. Chem. Commun. 2018, 91, 72?76. 171 [249] Windwer, S.; Sundheim, B. R. The Journal of Physical Chemistry 1962, 66, 1254? 1258. [250] Zurek, E. Journal of the American Chemical Society 2011, 133, 4829?4839. [251] Cotton, F. A.; Wilkinson, G.; Murillo, C. A.; Bochmann, M.; Grimes, R. Advanced inorganic chemistry; Wiley New York, 1988; Vol. 5. [252] Hatakeda, T.; Noji, T.; Sato, K.; Kawamata, T.; Kato, M.; Koike, Y. J. Phys. Soc. Jpn. 2016, 85, 103702. [253] Xu, H.; Wang, X.; Gao, Z.; Huang, H.; Long, Y.; Lu, Y.; Tang, K. J. Alloys Compd. 2018, 735, 2053?2057. [254] Guo, Z.; Zhou, L.; Jin, S.; Han, B.; Sun, F.; Yuan, W. RSC Adv. 2017, 7, 17539? 17544. [255] Zhao, L.; Liang, A.; Yuan, D.; Hu, Y.; Liu, D.; Huang, J.; He, S.; Shen, B.; Xu, Y.; Liu, X., et al. Nature Communications 2016, 7, 10608. [256] Zhou, X.; Wilfong, B.; Liou, S.-C.; Hodovanets, H.; Brown, C. M.; Ro- driguez, E. E. Chem. Comm. 2018, [257] Toh, M.; Tan, K.; Wei, F.; Zhang, K.; Jiang, H.; Kloc, C. Journal of Solid State Chemistry 2013, 198, 224?230. [258] Hatakeda, T.; Noji, T.; Hosono, S.; Kawamata, T.; Kato, M.; Koike, Y. J. Phys.: Conf. Series. 2014; p 022032. 172 [259] Dines, M.; Levy, R. The Journal of Physical Chemistry 1975, 79, 1979?1982. [260] Brown, L. A.; Glaunsinger, W. S.; McKelvy, M. J. Journal of Solid State Chemistry 1999, 147, 38?44. [261] Silbernagel, B.; Dines, M.; Gamble, F.; Gebhard, L.; Whittingham, M. The Journal of Chemical Physics 1976, 65, 1906?1913. [262] Gamble, F.; Silbernagel, B. The Journal of Chemical Physics 1975, 63, 2544?2552. [263] Young Jr, V.; McKelvy, M.; Glaunsinger, W.; Von Dreele, R. Chemistry of Materi- als 1990, 2, 75?81. [264] McKelvy, M.; Glaunsinger, W. Solid State Ionics 1987, 25, 287?294. [265] Dunn, J. M.; Glaunsinger, W. S. Solid State Ionics 1988, 27, 285?294. [266] Bouwmeester, H.; Wiegers, G. Journal of Solid State Chemistry 1988, 76, 391? 397. [267] Bernard, L.; McKelvy, M.; Glaunsinger, W.; Colombet, P. Solid State Ionics 1985, 15, 301?310. [268] Young Jr, V.; McKelvy, M.; Glaunsinger, W.; Von Dreele, R. Solid state ionics 1988, 26, 47?54. [269] Sch?llhorn, R.; Zagefka, H.-D. Angewandte Chemie International Edition in En- glish 1977, 16, 199?200. [270] Vaqueiro, P. Inorganic chemistry 2006, 45, 4150?4156. 173 [271] Vaqueiro, P.; Darlow, D. P.; Powell, A. V.; Chippindale, A. M. Solid State Ionics 2004, 172, 601?605. [272] Feng, M.-L.; Ye, D.; Huang, X.-Y. Inorganic chemistry 2009, 48, 8060?8062. [273] Feng, M.-L.; Li, P.-X.; Du, K.-Z.; Huang, X.-Y. European Journal of Inorganic Chemistry 2011, 2011, 3881?3885. [274] Bensch, W.; N?ther, C.; Schur, M. Chemical Communications 1997, 1773?1774. [275] Schur, M.; Bensch, W. Zeitschrift f?r Naturforschung B 2002, 57, 1?7. [276] Wu, M.; Rhee, J.; Emge, T. J.; Yao, H.; Cheng, J.-H.; Thiagarajan, S.; Croft, M.; Yang, R.; Li, J. Chemical communications 2010, 46, 1649?1651. [277] Greenfield, J. T.; Pak, C.; Kamali, S.; Lee, K.; Kovnir, K. Chemical Communica- tions 2015, 51, 5355?5358. [278] Pak, C.; Kamali, S.; Pham, J.; Lee, K.; Greenfield, J. T.; Kovnir, K. Journal of the American Chemical Society 2013, 135, 19111?19114. [279] Li, W.; Xia, F.; Qu, J.; Li, P.; Chen, D.; Chen, Z.; Yu, Y.; Lu, Y.; Caruso, R. A.; Song, W. Nano Research 2014, 7, 903?916. [280] Gao, Q.; Chen, P.; Zhang, Y.; Tang, Y. Advanced Materials 2008, 20, 1837?1842. [281] Ogata, H.; Fujimori, H.; Miyajima, S.; Kobashi, K.; Chiba, T.; Taylor, R. E.; Endo, K. Journal of Physics and Chemistry of Solids 1997, 58, 701?710. 174 [282] Chen, Z.; Dilks, R. E.; Wang, R.-J.; Lu, J. Y.; Li, J. Chemistry of materials 1998, 10, 3184?3188. [283] Spetzler, V.; Rijnberk, H.; Naether, C.; Bensch, W. Zeitschrift f?r anorganische und allgemeine Chemie 2004, 630, 142?148. [284] Hathaway, B.; Wilkinson, G.; Gillard, R.; McCleverty, J. The synthesis, reactions, properties and applications of coordination compounds 1987, 5, 533?774. [285] Li, J.; Chen, Z.; Wang, R.-J.; Proserpio, D. M. Coordination Chemistry Reviews 1999, 190, 707?735. [286] Sharma, J.; Jiang, Z.; Bhutani, A.; Behera, P.; Shoemaker, D. P. CrystEngComm 2019, 21, 2718?2726. [287] Alexander, M. D.; Spillert, C. A. Inorganic Chemistry 1970, 9, 2344?2346. [288] Childers Jr, R. F.; Vander Zyl Jr, K.; House, D. A.; Hughes, R. G.; Garner, C. S. Inorganic Chemistry 1968, 7, 749?754. [289] Schwarzenbach, G.; Magyar, B. Helvetica Chimica Acta 1962, 45, 1454?1457. [290] Allen, F. H.; Bird, C. M.; Rowland, R. S.; Raithby, P. Acta Crystallographica Sec- tion B: Structural Science 1997, 53, 680?695. [291] Miao, X.; Terao, T.; Yang, X.; Nishiyama, S.; Miyazaki, T.; Goto, H.; Iwasa, Y.; Kubozono, Y. Phys. Rev. B. 2017, 96, 014502. 175 [292] Denholme, S.; Demura, S.; Okazaki, H.; Hara, H.; Deguchi, K.; Fujioka, M.; Ozaki, T.; Yamaguchi, T.; Takeya, H.; Takano, Y. Mater. Chem. Phys. 2014, 147, 50?56. 176