Probing the Coulomb interaction of the unconventional superconductor PuCoGa5

The discovery of superconductivity in the plutonium based intermetallic compound PuCoGa5 in 2002 came as a surprise with regard to the high transition temperature Tc=18.5 K for an actinide material [1]. This lead to a renewed interest in actinide research and launched an intensive search for new 5f electron-based superconductors resulting in the discovery of the heavy fermion superconductor NpPd5Al2 (Tc=4.9 K) in 2007 [2]. Beyond the specific field of actinide research, it turns out that, given its characteristic energy scales, the unconventional superconductor PuCoGa5 is playing the central role of a missing link between the canonical heavy fermion superconductors and the high-Tc cuprates [3]. The understanding of its physical properties will thus allow advances in the global understanding of unconventional superconductivity. The analogy to the isostructural heavy fermion superconductor compound CeCoIn5 (Tc=2.3 K) strongly suggests a magnetic mechanism for the electron pairing. Despite this, the study of the phonon spectrum is nonetheless of interest. This is partly due to tremendous recent progress in band structure calculations that allow an accurate computation of phonon spectrum of strongly-correlated electron systems [4]

Fig. 13: IXS phonon spectra of PuCoGa5 measured at T = 297 K. The full and open symbols correspond to modes measured at Q = (0, 4, 3-q) and Q = (0, 4, 1-q), respectively. The dashed lines indicate the dispersion of the corresponding TA and TO modes.


We determined the phonon dispersion curves of single crystalline PuCoGa5 samples along the [100], [110] and [001] directions by inelastic X-ray scattering (IXS) at beamline ID28 at room temperature. Representative spectra are shown in Figure 13 for transverse acoustic (TA) and transverse optic (TO) modes propagating along [001]. The IXS data are compared with a density functional theory (DFT) ab initio calculation using the generalised gradient approximation with finite U (GGA+U) method [4]. We conclude that the inclusion in the calculation of a finite on-site repulsion between f electrons, U, of approximately 3 eV is essential to describe quantitatively the lattice dynamics of PuCoGa5. This conclusion is primarily drawn from the sensitivity of the lowest energy TO modes to the Coulomb repulsion that undergo up to 30% change in energy between the calculations with U = 0 and U = 3 eV. This makes apparent the strong influence of the 5f charge distribution on the force constants. The measured and calculated phonon dispersion is shown in Figure 14; an overall good agreement with the calculation with U = 3 eV is obtained for all the investigated branches. In contrast, an inelastic neutron scattering study performed on the parent compound UCoGa5 indicated that the phonon spectrum of this later system is better described with U = 0 eV in agreement with its itinerant paramagnet ground state [5].

Fig. 14: Measured (circles) and calculated (lines) phonon dispersion relations of PuCoGa5.

Our study gives new evidence for localised degrees of freedom of f electrons in PuCoGa5 in agreement with photoemission results. We show that phonon spectroscopy is an alternative way of probing electronic properties of strongly-correlated electron systems. The localised versus itinerant nature of the f electrons, whose duality is at the strongest in actinides, is an essential information for any theory aiming to describe superconductivity in such strongly-correlated electron system.

 

References

[1] J.L. Sarrao et al., Nature 420, 297 (2002).
[2] D. Aoki et al., J. Phys. Soc. Japan 76, 063701 (2007).
[3] N.J. Curro et al., Nature 434, 622 (2005).
[4] P. Piekarz et al., Phys. Rev. B 72, 014521 (2005).
[5] N. Metoki et al., Physica B 378-380, 1003 (2006).

Principal publications and authors

S. Raymond (a), P. Piekarz (b), J.P. Sanchez (a), J. Serrano (c), M. Krisch (c), B. Detlefs (d), J. Rebizant (d), N. Metoki (e), K. Kaneko (e), P.T. Jochym (b), A.M. Oleś (b,f), and K. Parlinski (b), Phys. Rev. Lett. 96, 237003 (2006); and J. Alloys Comp. 444-445, 104 (2007).
(a) CEA-Grenoble, DRFMC / SPSMS, Grenoble (France)
(b) Institute of Nuclear Physics, Polish Academy of Sciences, Krakow (Poland)
(c) ESRF
(d) European Commission, Institute for Transuranium Elements, Karlsruhe (Germany)
(e) Japan Atomic Energy Agency, Tokai (Japan)
(f) Max Planck Institut für Festkörperforschung, Stuttgart (Germany)