Session W41: Focus Session: Search for New Superconductors - Silicides, Nickelates and Cobaltates

A Possible Path from BCS through HTS to VHTS

Three years after celebrating the 50th anniversary of the BCS theory and the 20th anniversary of the discovery of high temperature superconductivity (HTS), it appears to be most fitting for us to contemplate the possibility of very high temperature superconductivity (VHTS). VHTS, preferably at room temperature, if achieved, could change the world both scientifically and technologically. Unfortunately, it has long been considered by some to belong to the domain of science fiction and to occur only ``at an astronomical distance and under an astronomical pressure.'' With the advent of liquid nitrogen superconductivity in 1987, the outlook has become much brighter. Currently, there appears to be no reason, either theoretical or experimental, why VHTS would be impossible, in spite of the 2006 prediction of the death of HTS by 2010-2015 through the so-called scientometric analysis of the publication record of the previous 20 years. The recent discovery of the new class of Fe-pnictide HTSs fuels more cautious optimism. Since its inception, BCS theory has provided the basic framework for the occurrence and understanding of superconductivity, but it has failed to show where and how to find superconductivity at a higher temperature. This may be attributed to the small energy scale of superconductivity in comparison with those of other excitations in the solids. After examining existing data, we believe that a holistic multidisciplinary enlightened empirical approach appears to be the most effective way to discover novel superconductors with higher transition temperatures. In this talk, I shall present several possible approaches toward VHTS that we are currently pursuing, after briefly summarizing what has happened in the long search for HTS and VHTS.   

Electronic structures of 122- and 111-type silicides and germanides in comparison to iron pnictide superconductors

Despite the high transition temperatures, iron pnictide superconductors have difficulty in practical use, due to high toxicity of pnictogens. Besides, the parent crystal structures of iron-based superconductors have been observed over wide range of constituent elements. In this study, we investigated the electronic structures of ThCr$_{2}$Si$_{2}$ (122)-type and CeFeSi (111)-type compounds containing Si or Ge instead of pnictogens. While superconducting $A$Fe$_{2}$As$_{2}$ ($A$: group 1-2 elements) has cylindrical Fermi surfaces with strong interband nesting, Fermi surfaces of isoelectronic compounds such as `$A$Co$_{2}$Ge$_{2}$' did not have these characteristics, mainly due to distortion of the `CoGe$_{4}$' tetrahedra. We investigated the relationship between the crystal geometry and the chemical species of 122- and 111-type compounds, and calculated their effects onto the electronic structures. We found that 111-type phases tend to have smaller distortion of tetrahedra than 122-type phases, although more electron-deficient metals must be chosen in order to reproduce the electronic structures of the superconducting phases. We will suggest some compositions that are expected to reproduce such electronic structures, and report about the synthesis of these compounds.   

NaAlSi: An Unusual Self-Doped Semimetal With Free Electrons and Covalent Holes

The ternary compound NaAlSi is a layered $sp$ conductor that superconducts at a relatively high $T_c$ of 7 K. Using first principles electronic structure calculations (the FPLO code), we find that NaAlSi is a self-doped semimetal with several unusual characteristics (in addition to the interesting fact that its structure is the same as those of the Fe-pnictide 111 coupounds). Na gives up its electron, allowing the Si-Al sublayer to form a distinct set of covalent valence bands, which are nearly filled. A strongly directional, Al-derived free electron band overlaps the valence bands, providing the electron carriers. The $k_z$ dispersion is small, but it occurs at the Fermi level, leading to unusual Fermi surfaces. Wannier functions will be provided to assist in the understanding of the bonding, and comparison with the CaAlSi compound, with one additional electron and a different structure (the MgB$_2$ structure) will be presented.   

Disorder-Sensitive Superconductivity and Bonding Network in the Iron-Silicide Superconductor Lu$_2$Fe$_3$Si$_5$

Iron silicide superconductor Lu$_2$Fe$_3$Si$_5$ exhibits relatively high $T_c$ = 6.0 K among Fe-based substances. Recent specific heat, penetration depth, and thermal conductivity measurements have provided evidences for the multigap superconductivity. We have studied non-magnetic and magnetic impurity effects on superconductivity in Lu$_2$Fe$_3$Si$_5$ by investigating $T_c$ variations in non-magnetic (Lu$_{1-x} $Sc$_x$)$_2$Fe$_3$Si$_5$, (Lu$_{1-x}$Y$_x$)$_2$Fe$_3$Si$_5$ and magnetic (Lu$_{1-x}$Dy$_x$)$_2$Fe$_3$Si$_5$. Small amount of non-magnetic impurities (Sc and Y) on the Lu-site rapidly depresses $T_c$ in accordance with the increase in the residual resistivity. Such a disorder-sensitive superconductivity strongly suggests the sign reversal of the superconducting order parameter. Lu$_2$Fe$_3$Si$_5$ has a complicated crystal structure compared to other multigap superconductors such as MgB$_2$ and iron pnictides. Thus it is important to map out the accurate bonding network in the crystal structure for the better understanding of the electronic structure. We have observed the charge density distribution of Lu$_2$Fe$_3$Si$_5$ by analyzing the synchrotron radiation powder diffraction data using the maximum entropy method/Rietveld method.   

$^{139}$La NMR in La$_{4}$Ni$_{3}$O$_{8}$: a possible analog to the cuprate high temperature superconductors

The Ni$^{1+}$/Ni$^{2+}$ states in the nickelates have identical electronic configurations as Cu$^{2+}$/Cu$^{3+}$ in the high temperature superconducting cuprates (3d$^{9}$/3d$^{8})$, and may exhibit similar properties. However, the Ni$^{1+}$ state is rare and cannot be easily stabilized. Recently, Martha Greenblatt and collaborators at Rutgers University have succeeded in growing a family of such compounds, Ln$_{n+1}$Ni$_{n}$O$_{2n+2}$ with a layered structure similar to the cuprates. The La$_{4}$Ni$_{3}$O$_{8}$ compound is particularly interesting as it undergoes an antiferromagnetic transition at T$_{N}$ = 100 K. We have done La NMR on powder samples to investigate the nature of this phase. Our spin lattice relaxation rate measurements clearly reveal a second order electronic phase transition similar to that observed in other antiferromagnets. Although we found clear signatures of changes to the spectra below T$_{N}$, we are unable to assign these changes to the presence of an internal field from the antiferromagnetic structure, or changes to the electric field gradient at the La site.   

Structure and electronic properties of the Ln$_{n+1}$Ni$_{n}$O$_{2n+2}$ layered nickelates

Structures of the recently discovered Ln$_{n+1}$Ni$_{n}$O$_{2n+2}$ (Ln = La, Nd) nickelates have infinite NiO$_{2}$ layers, similarly with superconducting cuprates. Moreover, due to unusually low Ni valance determined by their structures, Ni$^{+/2+}$ atoms posses the same 3d$^{9}$/3d$^{8}$ electronic configuration as Cu$^{2+/3+}$ in cuprates. An important question is whether mimicking the electronic and structural features of cuprates would also result in similar physical properties in nickelates? We have prepared several Ln$_{n+1}$Ni$_{n}$O$_{2n+2}$ phases, including the new ones, in a pure form, which make possible detailed characterization of their structures and properties for the first time. In particular, it was found that resistivity of Ln$_{n+1}$Ni$_{n}$O$_{2n+2}$ is decreasing with the increase of n, revealing a tendency to metallization with the change of doping in NiO$_{2}$ layers. Our recent neutron diffraction, resistivity, magnetic and NMR measurements clarifying the nature of the magnetic transition in La$_{4}$Ni$_{3}$O$_{8}$ is also discussed. 1. Poltavets V.V. \textit{Am. Chem. Soc.} \textbf{2006}, $128$, 9050. 2. Poltavets V.V. \textit{Inorg. Chem.} \textbf{2007}, $46$, 10887. 3. Poltavets V.V. Phys. Rev. Lett. \textbf{2009}, submitted.   

Ferromagnetism, superconductivity and magnetoresistance of Y$_{9}$Co$_{7}$

Crystal structures and physical properties of Y$_{9}$Co$_{7}$ and the related compound Y$_{8}$Co$_{5}$ will be discussed. In both materials, the crystallographic units consist of trigonal prisms formed by yttrium atoms centered around a cobalt atom. However, superconductivity and ferromagnetism occur only in Y$_{9}$Co$_{7}$. Magnetic susceptibility measurements (Arrot's method) on Y$_{9}$Co$_{7}$ give a Curie temperature T$_{Curie}$ = 4.25K and specific heat measurement provide evidence for bulk superconductivity with T$_{SC}$= 2.6K. The very large residual resistance ratio (RRR=30) confirm the excellent quality of the samples. Results of magnetoresistance up to 35T obtained in National High Magnetic Field Laboratory (NHMFL) will be discussed. The electrical resistivity measured at a temperature of 1.8K exhibits three different regions: for the field H$<$H$_{C}$ the material is superconducting, for H$_{C} \quad <$ H $<$ 1T resistivity rapidly increases and then for H $>$ 1T magnetic field causes excellent linear response of rho(H).   

Systematic band-filling dependence in the strongly-correlated triangular lattice

Strongly correlated triangular lattices are of interest because of their applicability to real systems such as superconducting organic charge-transfer solids (CTS) and the hydrated sodium cobaltate. Experimental work on nonhydrated Na$_x$CoO$_2$ have found (a) a charge-ordered semiconductor at $x=0.5$, (b) Curie-Weiss metal for $x>0.5$, and (c) paramgnetic metal for $x<0.5$. The strong $x$-dependence is reminescent of the band-filling dependence in CTS conductors \footnote{S. Mazumdar and A. N. Bloch, Phys. Rev. Lett. \textbf{50}, 207 (1983.} We have performed numerical calculations based on the extended Hubbard Hamiltonian to understand the $x$-dependence. We show that for {\it finite} Hubbard $U$, and moderate nearest neighbor interaction $V$, the normalized probability of double occupancy, which is a measure of the strength of the electron correlation, varies strongly as a function of the density of carriers. We are able to explain (i) the different behavior of $x<0.5$ and $x>0.5$, (ii) the absence of the $\sqrt{3} \times \sqrt{3}$ charge-ordering at $x=1/3$, and (iii) why $x=0.5$ is unique. Cobalt valence of 3.5 in the superconducting hydrated cobaltate is in agreement with our proposed mechanism of superconductivity in the CTS. \footnote {S. Mazumdar and R. T. Clay, Phys. Rev. B 77, 180515 (R), (2008).}   

New magnetic orderings in Na${_x}$CoO${_2}$

Na${_x}$CoO${_2}$ has a rich phase diagram and intriguing physical properties. Very recently, it has been found that there exist further magnetic orderings at low temperatures in additional to the known 22 K antiferromagnetic phase. In this paper, we report two distinct types of ordering occurring at 8 K and 15 K for $x$=0.834 respectively, revealed by the specific heat and magnetization measurements. Both orderings are of metamagnetism.   

Transition of Fermi surface topology in highly-doped NaxCoO2 studied by angle-resolved photoemission spectroscopy

We have performed an angle-resolved photoemission spectroscopy of NaxCoO2 to clarify the origin of the antiferromagnetic (AF) transition in the highly-doped region (x $>$ 0.75). We determined the electronic band structure along the out-of-plane direction by varying the photon energy, and revealed that a three dimensional electron pocket emerges at the $\Gamma $ point in the AF sample (x $>$ 0.75), while it is absent in the non-AF counterpart (x $<$ 0.75), showing the transition of the Fermi surface topology with the band filling. The present result suggests that the emergence of the electron pocket is closely related to the origin of the AF transition in highly-doped NaxCoO2.