II. New Families of Superconducting Materials

  • Superconductivity is a fascinating subject! It is not easy to master, neither theoretically, nor experimentally. And rationalized design of new superconducting systems is practically lacking in the literature. Based on our extensive studies on chemistry of vibronic coupling, we theoretically predicted a new family of superconducting materials. Our 40-pages review paper “Real and Hypothetical Intermediate-Valence Fluoride AgII/AgIII and AgII/AgI Systems as Potential Superconductors” appeared in ANGEW CHEM INT ED ENGL 40 (15): 2743-2781 2001 (see our paper here). Below we describe results of our investigations in brief.
  • AgII is an incredible chemical species. It is peculiar to fluorine systems, and disproportionates to AgI and AgIII in oxides. AgII is also a potent oxidizer, one of the most powerful known. For example, cationic AgII in anhydrous HF solutions oxidizes Xe0 to XeII. AgIII is common to fluoride systems, as well, in agreement with the rule that fluorine stabilizes high oxidation states of most elements. Cationic AgIII in anhydrous HF solutions is an even stronger oxidizing agent than AgII; it oxidizes [MF6] (M = Pt, Ru) to MF6. Solvated AgIII is an oxidizer of unsurpassed power, better than KrII, and (together with solvated NiIV) probably the best oxidizer available in inorganic chemistry. Dealing with AgIII solvated in aHF is living on the edge of chemistry; a tour to the regions penetrated by so few people!
  • AgII and AgIII have the same d-electron count as CuII (d9) and CuIII (d8), respectively. F and O2- are isoelectronic, closed-shell (s2p6) species. Both F and O2- are weak-field ligands. Led by these similarities, we examined analogies between the superconducting cuprates (CuII/III-O2- and CuII/I-O2- systems) and the formally mixed-valence AgII/III-F and AgII/I-F phases. For this purpose we performed the electronic structure computations for a number of structurally characterized binary and ternary fluorides of AgI, AgII, and AgIII, and compared the results with similar calculations for oxocuprate superconductors. Their computations reveal that states in the vicinity of the Fermi level (x2-y2 or z2) have usually strongly mixed Ag(d)/F(p) character and are Ag-F antibonding, thus providing the potential of efficient vibronic coupling (typical for d9 systems with substantially covalent bonds). This is the result not only of a coincidence in orbital energies; surprisingly the Ag–F bonding is substantially covalent in fluorides of AgII and AgIII. The DOSF for Ag/F materials and frequencies of the metal-ligand stretching modes have values close to those for copper oxides. The above features suggest that properly hole- or electron-doped AgII fluorides might be good BCS-type superconductors. We in addition analyzed a comproportionation / disproportionation equilibrium in the hole-doped AgII fluorides, and the possible appearance of holes in the F(p) band. It seems that there is a chance of generating an AgIII-F/AgII-F0 “ionic/covalent” curve crossing in the hole-doped AgII fluorides, significantly increasing critical superconducting temperature (TC).


  • Our early predictions on substantial covalency of the Ag-F bonds have been confirmed by experiment. Using the ultra-high resolution XPS spectra we have proved, that electronegativity of AgIII center is so large, that it can attract not only valence but also – to some extent – core electrons from F ! In the result, the mixing of the Ag(d) and F(p) states is very large, and Ag(3+) even introduces holes to the F(p) band. As the result of this atypical fight for electrons between Ag and F, leading to the “frustration” of the electron assignement to a particular atomic core, some higher fluorides of Ag, such as AgFBF4 or KAgF3, exhibit metallic conductivity and Pauli (temperature-independent) paramagnetism (read about it here). This “non-ionic” character of the Ag-F bonding in these compounds is unusual to whole inorganic chemistry, to put it midly. And it has been recently confirmed by advanced hybrid HF/DFT spin-polarized calculations for AgF2 (N. Harrison, unpublished results). Note, that KAgF3 is the first fluoride, for which it has been unequivocally proved that strongly bound valence electrons of AgII can be set free at quite low temperatures, and they freely  move through the array of  Ag and F atoms while using for this purpose the significantly compact, enertetically low lying orbitals of F (PHYS STAT SOL B 240 (3): R11-R14 2003, PHYS STAT SOL B, 242(1): R1-R3 2005). Thus fluorides (i.e. connections of chemical elements with the most electronegative element except for Kr), need no longer be ionic insulators! Just give a proper partner to fluorine (i.e. the one which is also eagerly fighting for electrons), and you will generate the frustration of the electron assignement to either of the two, resulting in … liberation of the charge carriers. Physicists should now smile and think of their Hubbard on-site U energy…
  • Now, prepare for good news. While at The University of Birmingham with Pete P. Edwards, we have discovered Meissner effect in the (probably self-doped or electron-doped) BeF2/AgF2 phases,  at the temperatures as high as 64 K (read here more about our discovery). Unofrtunately, despite the extensive effort and use of various techniques, we could not determine the chemical identity of the phase(s) responsible for the Meissner effect. This result confirms our earlier theoretical hypothesis that superconductivity may be induced in the higher fluorides of silver. We now try to squueze up some of these systems, in order to help superconductivity appear under ultra high pressures.
  • More recent study (Nature Mater. 5(7): 561-566 2006) has indicated that Cs2AgF4 is a 2D Heisenberg ferromagnet. We have proposed that it may be turned into antiferromagnet via use of high pressures, or alternatively AgF2 should be doped with compounds which contain covalently bound F (NATURE MATER 5(7): 513-514 2006).
  • d9 and s1 configurations, when joint with p5 in an avoided crossing scheme, provide electronic basis for particularly impresive high-temperature superconductivity.  Nowadays, the reasons for huge potential of s1, p5, and d9 species for superconductivity are rather obvious. Vibronic effects, governing to large extent superconductivity, are very strong in open–shell systems, with electrons delocalized in sigma (but not pi) bonding or antibonding levels. s1 and d9 species are usually very susceptible to the Jahn–Teller effect, and to vibronic coupling in general. But vibronic effects may be even further magnified if electronic levels of cations with s1 or d9 configuration are very close in energy scale to electronic levels of anions, accompanying cations in the extended nets (“covalency” discussed by Sleight). Such situation is provided e.g. by an oxide anion, O2–, for Bi and Cu at high–oxidation states (Bi4+/5+, Cu2+/3+). These observations generated our interest in systems based on a f13 electronic configuration. If we are success (chances for high-TC are rather moderate), all important inorganic families of SCs will meet under one roof (J MOL MODEL 11(4-5): 323-329 2005).
  • Admittedly, f orbitals are usually strongly contracted, spatially deeply hidden under other valence orbitals, and thus not available for bonding. However, there are cases when f orbitals participate to some extent in bonding to ligand; it happens when energetic and spatial proximity of ligand and f-block metal orbitals is provided. We will soon describe design of new superconductors based on systems with the f13 electronic configuration, using strongly reducing ligands to create extended nets with chosen lanthanide elements.
  • Superconducting hydrides and hydrogen storage are, as told above, related to each other by the equation: Xn+ + H–1 –> X(n–1)+ + H0. We now progress in a design of new superconducting hydride materials, based on unique electronic properties of four chosen inner- and outer-transition metals. Our target is to generate an array of radical-like hydride anions via chemical means, and not via applying external pressure. Results of these investigations halv recently been published (J MATER CHEM 16(12): 1154-1160 2006, J MOLEC MODEL, submitted 2006).

Ultra-high pressures allow for generating and tuning superconductivity in various classes of materials. Recently, we have predicted that compressed silane and germane may become metallic, or even superconducting (PHYS REV LETT 96(1): 017006 2006, J PHYS CHEM SOLIDS (SMEC 2005 Proceedings), in press 2006). Squeezing attempts now continue in several labs worldwide.