Research Interests

Our interests encompass a range of structural, electronic and magnetic problems in contemporary condensed matter science with emphasis on superconducting and magnetic materials.
The primary focus is on the study of strongly correlated electron systems, which typically display remarkable electronic properties that challenge existing theory for satisfactory explanations.
The delicate balance between localised and itinerant behaviour related to the strongly correlated nature of the systems lies at the heart of our current research. While lattice, charge, spin and orbital are usually independent degrees of freedom, in these materials they are strongly coupled so that the balance between the resulting phases is very subtle and sensitively susceptible to external stimuli.
Our principal strategy is to tune the interplay of the above parameters both chemically (changing the chemical composition) and physically (varying an external parameter like pressure, photo-irradiation and magnetic or electric field).

Molecular superconductors – metal fullerides

The structural and electronic properties of C60-based solids have been exhaustively studied in the last 30 years.
A3C60 (A = alkali metal) superconductors adopt face-centred cubic (fcc) structures and their superconducting TC increases monotonically with increasing interfullerene spacing, reaching a 33 K maximum for RbCs2C60; this physical picture had remained essentially unaltered since 1992 with the established fulleride chemistry chronically failing to deliver new materials. Trace supreconductivity (s/c fraction < 0.1%) at 40 K under pressure was reported in 1995 in multiphase samples with nominal composition Cs3C60. Despite numerous attempts by many groups worldwide, this remained unconfirmed and the structure and composition of the material responsible for superconductivity unidentified. Thus the possibility of enhancing fulleride superconductivity and understanding the structures and properties of these archetypal molecular superconductors close to the Mott-Hubbard metal-insulator (M-I) transition had remained elusive.

Our recent work much in collaboration with the group of Professor Matt Rosseinsky at Liverpool has now removed this bottleneck. We specifically targeted hyperexpanded high-symmetry metal fullerides with varying C60 orientations and packings. This approach first led to the discovery of bulk superconductivity under pressure at 38 K in bcc-structured A15 Cs3C60, the highest TC known for any molecular material [12]. This was the first example of a superconducting C603- fulleride with non-fcc sphere packing and this single non-fcc material had a higher TC than all the fcc A3C60 fullerides studied before. Moreover, the electronic ground state in competition with superconductivity not only contains magnetic moments localised on the C603- anions but is antiferromagnetically ordered with TN = 46 K. The pressure-induced antiferromagnetic insulator-superconductor transition maintains the threefold degeneracy of the active orbitals in both competing electronic states, and is thus a purely electronic transition to a superconducting state with a dependence of the transition temperature on anion packing density that is not explicable by BCS theory [8]. The simplicity of the structural and electronic properties of A15 Cs3C60 – high symmetry, absence of positional and valence disorder – firmly established it as a model system for all high-TC superconductors.

Such behaviour had not been observed before for any of the known superconducting A3C60 fcc fullerides because they are too far from the metal-insulator transition (too dense) to allow the role of electron correlations to emerge. Following isolation of fcc Cs3C60 – the most expanded member of the family – we went on to study the influence of lattice packing and symmetry in controlling the competition between magnetism and superconductivity in a high-TC superconductor. Unlike all other members of the fcc A3C60 family with a C603- charge, fcc Cs3C60 is insulating at ambient pressure. The frustrated geometry of the fulleride packing was shown to have a dramatic effect on the magnetic ground state – in contrast to its bcc-packed polymorph, the onset of antiferromagnetic order is suppressed by more than one order of magnitude [7]. This is accompanied by the emergence of the unconventional high-TC superconducting ground state (at 35 K upon pressurisation) at a much smaller critical bandwidth than in bcc packing. The two distinct lattice geometries (bcc and fcc) control the superconductivity via the proximity of each polymorph to the magnetic insulating state; TC scales as a universal function of the electronic bandwidth relative to the bandwidth at which the correlated electrons become localised and insulating independent of the precise lattice packing adopted.

Recently using infrared spectroscopy, we presented evidence for the dynamic Jahn‐Teller effect as the source of the dramatic change in electronic structure occurring during the transition from the metallic to the localized state in expanded fullerides [4]. The temperature dependence of the spectra in the insulating Cs3C60 phase can be explained by the gradual transformation from two temperature-dependent solid-state conformers to a single one, typical and unique for Jahn-Teller systems. Such results unequivocally established the relevance of the dynamic Jahn-Teller effect to overcoming Hund’s rule and forming a low-spin state, leading to a magnetic Mott-Jahn-Teller insulator.

Fe-based pnictide and chalcogenide superconductors

Superconductivity at the surprisingly high temperature of 55 K has been recently reported in fluorine-doped rare-earth iron oxyarsenides, REFeAsO1-xFx (RE= rare earth). The magnitude of TC and the apparent similarities with the high-TC cuprate superconductors; layered structural motifs of the conducting FeAs slabs and proximity to antiferromagnetic (AFM) and structural instabilities; have made these systems an intensely studied research field. We have studied in detail the properties of SmFeAsO1-xFx (0 ≦ x ≦ 0.20) in which superconductivity emerges near x~0.07 and TC increases monotonically with doping up to x~0.20. We found that orthorhombic symmetry survives through the metal-superconductor boundary well into the superconducting regime and the structural distortion is only suppressed at doping levels, x>0.15 when the superconducting phase becomes metrically tetragonal [11]. Remarkably this crystal symmetry crossover coincides with drastic anomalies in the resistivity and the Hall coefficient and a switch of the pressure coefficient of TC from positive to negative [13].

These discoveries have catalysed the search for superconducting compositions in related materials in which two-dimensional FeQ (Q = chalcogen) slabs are also present. This search has now led to the report that superconductivity at ~8 K occurs in the simple binary a-FeSe phase with the PbO structure for which we have found that the crystal structure in the superconducting state shows remarkable similarities to that of the REFeAsO1-xFx superconductors [14]. FeSe is a key member of this family of high-TC superconductors, as while it possesses the basic layered structural motif of edge-sharing distorted FeSe4 tetrahedra, it lacks interleaved ion spacers or charge-reservoir layers. We have found that application of hydrostatic pressure first rapidly increases TC which attains a broad maximum of 37 K at ~7 GPa (this is one of the highest Tc ever reported for a binary solid and is surpassed only by MgB2 at 39 K and by Cs3C60 at 38 K) before decreasing to 6 K upon further compression to ~14 GPa [10]. At the same time, the complementary diffraction data well within the superconducting state reveal an extremely soft solid with strong bonding anisotropy between inter- and intra-layer directions that transforms to the more densely packed b-polymorph above ~9 GPa. The non-monotonic TC(P) behaviour of FeSe coincides with a drastic anomaly and collapse in the pressure evolution of the interlayer spacing, providing compelling evidence for the key role of this structural feature in defining the electronic properties of these superconductors. A similar intimate link between crystal and electronic structures is also revealed in Fe1+dSe0.57Te0.43 in which there is an exact coincidence of the pressure at which TC is maximum with the onset of a structural phase transition [9]. Finally superconductivity in FeSe is enhanced at ambient pressure to 32 K by insertion of alkali metal ions in the interlayer spacing to afford KxFeySe2 compositions. The parent insulating material of the intercalated iron selenides has a stoichiometry close to K0.8+xFe1.6-x/2Se2 retaining an iron valency close to +2. Ordering of the tetrahedral site vacancies produces a five-fold expansion of the parent ThCr2Si2 unit cell in the ab plane which can accommodate 20% vacancies on a single site within the square FeSe layers [6].

Photoswitchable multifunctional molecular materials

We are employing X-ray light as a powerful tool to induce photoswitching between the ground and hidden excited states — which can be simultaneously structurally, electronically and magnetically characterised — in suitably selected model systems (e.g. Prussian blue analogues) in the bulk. Our work has led to the discovery of X-ray light illumination induced cooperative transitions which in a controllable way can lead to otherwise inaccessible metastable states with differing electronic/magnetic properties over a broad temperature range [16].
Recent work also examined the structural properties of the photosensitive nanocrystalline mixed valency binary oxide Ti3O5, which has recently emerged as a promising phase-change material that exhibits rapid photo-reversible optical and resistance changes at ambient temperature. Nanosizing led to a remarkable increase in the stability range of the large-volume high-temperature conducting phase[8]. Hydrostatic pressure is also being used as a complementary efficient stimulus to tune the magnetic interactions and allow trapping of novel inaccessible states[15].

Molecular superconductivity at the extreme

Superconductors have no electrical resistance and carry electricity without losing energy, so finding new superconductors with enhanced performance is of paramount importance for energy applications. Most superconductors have simple structures built from atoms, but superconductors made from molecules arranged in regular solid structures also exist. We have unveiled new molecular superconductors by resorting to chemistry (doping) and employing physical means (high pressure) [3]. The success of our synergic approach (synthesis/measurement/calculation) has led to the discovery of the unconventional strongly-correlated fulleride superconductors with the highest working temperature (at 38 K) for any molecular system that can also withstand record high critical magnetic fields exceeding 90 T [2]. The electronic structure of the molecular building block directly controls the formation of the state of matter (an anomalous metal), from which both the superconductor and the parent antiferromagnetic Mott insulator emerge, providing the link between molecule and property. This opens the way to designing-in features conferring robust high-temperature superconductivity to classes of superconductor that are not restricted by the elements of the Periodic Table.

π-electron molecular quantum magnetism

Molecular solids whose cooperative electronic properties are based purely on π-electrons from carbon atoms offer a fertile ground in the search for exotic states of matter. We integrate chemical synthesis with physical control of structure and function to discover new carbon-based molecular π-electron quantum magnets. We employ as building units members of the large family of electronically-active polyaromatic hydrocarbons and target control over the molecular oxidation state in order to yield open-shell radical anionic charge states. In particular, our research strategy has yielded new candidate quantum spin liquids [1] — a state of matter in which the spins never order and continue to fluctuate rapidly even at a temperature of absolute zero. Each individual spin points simultaneously along an infinite number of directions and is highly entangled with other spins. As such, quantum spin liquids can host many exotic phenomena of fundamental and technological interest with potential applications including data storage for quantum computers.

Anionogenic mixed valency in molecular solids

The electronic properties of materials arising from d- and f-shells have drawn the bulk of research attention in condensed matter science. Although unconventional p-electron-based electronic, magnetic and conducting properties remain somewhat rare to the present date, they are a fascinating topic that presents the added challenge of molecular structural and electronic complexity. Among the alkali oxides, the black molecular sesquioxides A4O6 (A = alkali metal) are of special interest. Their formula can be represented as (A+)4(O2−)2(O22−), comprising three molecular anions, one closed-shell nonmagnetic peroxide anion, O22− and two open-shell magnetic hyperoxide anions, O2−. They are therefore rare examples of binary ionic compounds with oxygen in two different oxidation states (mixed valence molecular oxides) with a complex interplay of spin, lattice, charge, and orbital degrees of freedom tunable by the application of external stimuli such as temperature and pressure.

The group makes extensive use of international synchrotron X-ray, neutron and muon facilities (ISIS and Diamond , U.K., Institut Laue Langevin (ILL) and European Synchrotron Radiation Facility (ESRF) , Grenoble, France, and SPring-8 , Japan).

Our group was coordinating a large EU-Japan collaborative project entitled “Light Element Molecular Superconductivity: an Interdisciplinary Approach (LEMSUPER) ” [Internet archive] involving a range of university partners across the European Union and from Japan.

Selected publications

  • [1] Y. Takabayashi, M. Menelaou, H. Tamura, et al.,
    π-electron S = 1/2 quantum spin-liquid state in an ionic polyaromatic hydrocarbon’,
    Nature Chem. 2017, 9, 635.
  • [2] Y. Kasahara, Y. Takeuchi, R. H. Zadik, et al.,
    ‘Upper critical field reaches 90 tesla near the Mott transition in fulleride superconductors’,
    Nature Commun. 2017, 8, 14467.
  • [3] R. H. Zadik, Y. Takabayashi, G. Klupp, et al.,
    ‘Optimized unconventional superconductivity in a molecular Jahn-Teller metal’,
    Science Adv. 2015, 1, e1500059.
  • [4] G. Klupp, P. Matus, K. Kamaras et al.,
    ‘Dynamic Jahn-Teller effect in the parent insulating state of the molecular superconductor Cs3C60,
    Nature Commun. 2012, 3, 912.
  • [5] R. Makiura, Y. Takabayashi, A. N. Fitch et al.,
    ‘Nanoscale effects on the stability of the l-Ti3O5 polymorph’,
    Chem. Asian J. 2011, 6, 1886.
  • [6] J. Bacsa, A. Y. Ganin, Y. Takabayashi et al.,
    ‘Cation vacancy order in the K0.8+xFe1.6-ySe2
    system: five-fold cell expansion accommodates 20% tetrahedral vacancies’
    Chem. Sci. 2011, 2, 1054.
  • [7] A. Y. Ganin, Y. Takabayashi, P. Jeglic et al.,
    ‘Polymorphism control of superconductivity and magnetism in Cs3C60 close to the Mott transition’,
    Nature 2010, 466, 221.
  • [8] Y. Takabayashi, A. Y. Ganin, P. Jeglic et al.,
    ‘The disorder-free non-BCS superconductor Cs3C60 emerges from an antiferromagnetic insulator parent state’,
    Science 2009, 323, 1585.
  • [9] N. C. Gresty, Y. Takabayashi, A. Y. Ganin et al.,
    ‘Structural phase transitions and superconductivity in Fe1+δSe0.57Te0.43 at ambient and elevated pressures’,
    J. Am. Chem. Soc. 2009, 131, 16944.
  • [10] S. Margadonna, Y. Takabayashi, Y. Ohishi et al.,
    ‘Pressure evolution of low-temperature crystal structure and bonding of the FeSe superconductor (TC = 37 K)’,
    Phys. Rev. B 2009, 80, 064506.
  • [11] S. Margadonna, Y. Takabayashi, M. T. McDonald et al.,
    ‘Crystal structure and phase transitions across the metal-superconductor boundary in the SmFeAsO1-xFx (0£x£0.20) family’,
    Phys. Rev. B 2009, 79, 014503.
  • [12] A. Y. Ganin, Y. Takabayashi, Y. Z. Khimyak et al.,
    ‘Bulk superconductivity at 38 K in a molecular system’,
    Nature Mater. 2008, 7, 367.
  • [13] Y. Takabayashi, M. T. McDonald, D. Papanikolaou et al.,
    ‘Doping dependence of the pressure response of TC in the SmO1-xFxFeAs superconductors’,
    J. Am. Chem. Soc. 2008, 130, 9242.
  • [14] S. Margadonna, Y. Takabayashi, M. T. McDonald et al.,
    ‘Crystal structure of the new FeSe1-x superconductor’,
    Chem. Commun. 2008, 5607.
  • [15] D. Papanikolaou, W. Kosaka, S. Margadonna et al.,
    ‘Piezomagnetic behavior of the spin crossover Prussian blue analogue CsFe[Cr(CN)6]’,
    J. Phys. Chem. C 2007, 111, 8086.
  • [16] D. Papanikolaou, S. Margadonna, W. Kosaka et al.,
    ‘X-ray illumination induced Fe(II) spin crossover in the Prussian blue analogue cesium iron hexacyanochromate’,
    J. Am. Chem. Soc. 2006, 128, 8358.