Skip to main content

Full text of "A common origin for the resistivity of Cd$_2$Re$_2$O$_7$, the cuprates, and Sr$_2$RuO$_4$?"

See other formats

A common origin for the resistivity of Cd 2 Re 2 07, the cuprates, and Sr 2 Ru04? 

George Kastrinakis 

Institute for Electronic Structure and Laser (IESL), Foundation for Research and Technology - Hellas (FORTH), 

P.O. Box 1527, Irakho, Crete 71110, Greece* 
(November 9, 2001) 

We propose an explanation for the temperature dependence of the resistivity of Cd2Re207, 
including the regime above the structural phase transition at T=200 °K. The mechanism involved 
relies on the existence of a strong van Hove singularity close to the Fermi surface, which is evidenced 
by relevant band structure calculations. The same mechanism has successfully described the T-linear 
resistivity of the cuprates and S^RuCU, and the one-particle scattering rate in the former materials, 
as corroborated by recent experiments. We describe a few predictions for Cd2Re207 and Cd20s207. 

Recent experiments have indicated an intriguing behavior of the resistivity of the pyrocbJore oxide Cd2Re2 07. 
Above the superconducting transition at T = 1 K found by Hanawa et alEl and Sakai et al.H, the resistivity scales 
quadratically with T up to 60 K. A. linear in T regime .follows up to T*=200 K, whereby a structural phase transition 
takes place, as shown by Jin et al.EI and Hanawa et al.Q Above T* the resistivity increases very slowly, approximately 
linearly with T. This latter behavior is also observed by Mandrus et al.EI in Cd 2 0s207, above the metal-insulator 
transition at 226 K. Both Cd2Re207 and Cd20s207 are described within the realm of the interacting Fermi liquidity 

We believe that the T dependence of the resistivity described above can be easily and coherently understood within 
the frame of our model, originally proposed tej-.tackle the issue of the T-linear resistivity observed for several cuprates 
in the optimally doped and overdoped regimeOu. 

Below we summarize the salient features of the microscopic model. The prerequisites for the application of the 
model are (a) the existence of a strong van Hove singularity close to the Fermi surface, andlb) interactions, giving rise 
to screening. Our derivation is valid Jjor any Baym-Kadanoff many-body approximationCm. We note that the model 
applies irrespective of dimensionality!] d > 2, as long as a substantial part of the carrier spectral weight is contained 
within the van Hove peak._What is more, there is convincing evidence from the band structure calculations of Singh 
et al.EI and Mandrus et alE, that a strong van Hove singularity lies close to the Fermi levelppf both aforementioned 
pyrochlores. (For the cuprates and Sr2Ru04, besides numerical calculations - e.g. see refill and therein - relevant 
experimental evidence can be found in ref.a). 

We have shown analytically that the one-particle scattering rate goes like x 2 , a;=max{T, energy}, for x — >0. The 
scattering rate becomes linear in energy if the latter exceeds a crossover value x a , or linear in T for T > x a /A. x 
is the difference between the chemical potential and the energy of the van Hove singularity. The result holds true 
everywhere in the Brillouin zone. This predictionjsias directly supported by the ARPES expts. of Valla et al.E3 Hence, 
calculating the resistivity via the Kubo fownulauEl yields a quadratic in T behavior for T < x D /4, and T-linear for 
T > x /4. The recent expts. of Ono et alO on Bi2Sr2- 2; La ;r Cu06+<5 support this T dependence. 

The resistivity of Cd2Re2C>7 is readily understood in the frame of the model, with x /4 w 60 K. We attribute the 
slow T dependence of the resistivity above T* to a drastic net reduction of the screened interaction. This can take 
place through the reduction of the bare coupling, i.e. reduction of the Hubbard U and/or reduction of the density of 
states at the Fermi level, or the susceptibility of the carriers, or through all of the above. We believe that the slowly 
increasing resistivity of Cd20s207 above the 226 K metal-insulator transition can be likewise understood. 

On the basis of the above, we predict that the one-particle scattering rate of Cd2Re207 and Cd20s207 should have 
a T 2 , T < £ /4,pdependence, followed by a T-linear one for T > x /4. The energy dependence of the rate follows a 
similar behaviorLHa. ARPES experiments on these compounds should shed light on the question. Moreover, taking 
Im x(q, w) = w/uJo + 0(uj 3 ), x being tngxarrier susceptibility, we predict uj to be a small energy scale, e.g. less than 
50 (fifty) meV, according to the modelEHa. Inelastic neutron scattering should yield relevant data. 

I acknowledge useful correspondence with Y. Ando and D.J. Singh. 

* e-mail : 

1 M. Hanawa et al., Phys. Rev. Lett. 87, 187001 (2001). 


2 H. Sakai et a l,, J 

3 R. Jin et al 

4 M. Hanawa, J. Yamaura, Y. Muraoka 

Phys. Conden s. Matter 13, L785 (2001). 

Z. Hiroi, |sond-mat/010905C 

' D. Mandrus et al., Phys. Rev. B 63, 195104 (2001) 
; D.J. Singh, P. Blaha, K. Scharz & J.O. Sofo, |cond-mat/0108226 

' G. Kastrinakis, Physica C 340, 119 (2000); v2 of |cond-mat/0005485 . 

8 G. Kastrinakis, Physica C 317-319, 497 (1999). 

9 D.H. Lu et al., Phys. Rev. Lett. 76, 4845 (1996). 

10 T. Valla et al., Science 285, 2110 (1999); T. Valla et al., Phys. Rev. Lett. 85, 828 (2000). 

11 S. Ono et al. Phys. Rev. Lett. 85, 638 (2000).