Raman Spectroscopy of High-Temperature Superconductors


C. Kendziora
Materials Science and Technology Division

Background: Superconductors carry current without resistive losses. This makes them very attractive for virtually any application requiring electrical power generation, distribution, or use. As the Navy moves toward the electric warship, efficient electrical power for (among other things) ship propulsion and electronic weapons will be essential, creating additional need for superconducting technology. Most of the electrical power generated today is wasted because of resistive losses. This has already driven the implementation of superconducting technology in certain niche markets, such as high-field magnets and short-distance transmission. Obstacles remaining to further enabling superconductivity include the need for low temperatures (cryogenics) as well as materials problems associated with the high-temperature superconductors based on copper and oxygen (cuprates).

Superconductivity: The mechanism of superconductivity depends on a seemingly impossible phenomenon: the attraction of electrons (or holes-the electrons' positively charged equivalent) to each other in pairs. This attractive pairing occurs despite the large coulomb repulsion the carriers feel due to their charge. Such pairs carry the "supercurrent," a "frictionless" electronic motion. Rising temperature, which accelerates electrons and shakes the crystal lattice through which they flow, tends to separate, or "break" pairs. The critical temperature (Tc)-the temperature below which superconductivity sets in-is thus dependent on the pairing strength, and ultimately on the mechanism by which the pairs form. The cuprates have demonstrated by far the highest known Tc's as high as 164 K, still well below room temperature (296 K). However, despite having been discovered more than 13 years ago, several fundamental principles of cuprate superconductivity remain in dispute. Ultimately, to optimize this class of materials as well as to predict superconductors with even higher Tc , we must understand how they work and, specifically, why they work so much better than anything else.

The cuprates as a material class are generally insulators. However, when extra electrons are added (or removed, for the hole-type case) through chemical substitution, they become metallic and superconducting. Through extensive study on hole-doped cuprates, a consensus has been reached in the research community that the carriers that pair in hightemperature superconductors are not only very strongly coupled, but that they possess a certain angular momentum as well, making them "d-wave" superconductors. This is in contrast to the elemental and alloy "conventional" superconductors that have zero angular momentum ("s-wave") and may be crucial for the very high Tc they achieve. Until recently, electron-doped cuprates were thought to be more "conventional" than the hole-doped cuprates and to have s-wave pairing. Experimentally, the pairing state can be determined by measuring the strength of the pairing as a function of direction within the crystal. As shown in Fig. 7, s-wave pairing is typically nearly isotropic, with no directional dependence, while dwave pairing results in a characteristic anisotropy that includes "nodes," or directions where the strength is zero.

Fig7 Image




FIGURE 7
The superconducting pairing strength as a function of direction. The isotropic s-wave functional form (red) has no directional dependence. In contrast, the dx2- y2 functional form (blue) has directions of high and low pairing strength separated by 45 deg in 4-fold symmetry.

Raman Spectroscopy: Spectroscopy is the study of phenomena across some energy scale. The relevant energy scale for pairing of electrons in superconductors is 0 to 0.1 eV, which is convenient for Raman spectroscopy. In the Raman process, highenergy (generally 0.5 to 3 eV) photons interact benignly with a material and produce scattered photons that have been shifted in color (energy). The number of scattered photons collected as a function of energy shift is called the Raman spectrum, and this contains unique information about both the phonon (sharp spikes) and electronic (continuum) nature of the material.

The Raman spectrum of the electron-doped cuprate Nd1.85Ce0.15 CuO4+d crystal is plotted for two temperatures in Fig. 8. The incident laser energy (1.92 eV) is defined as zero, and only the downward "red" shift in energy is plotted. The red curves plot data taken at 28 K, in the normal state, while the blue curves show the Raman spectrum at 8 K in the superconducting state. As is clear in the top curve (and inset), a peak forms in the superconducting state, the energy of which (0.008 eV) is a measure of the strength of pairing. For a crystalline sample, the symmetry can be exploited using polarized photons to extract additional information. In this case, we have chosen a symmetry configuration in which no phonons are allowed and where the electronic signal is strong. Because the Raman interaction gives a weighted average of the different directions within the crystal, we cannot uniquely map out the pairing strength as a function of direction. However, we establish that the pairing is anisotropic and has nodes by the curvature of the spectrum at energies below the peak (inset). Specifically, it has the d-wave functional form, which predicts an intensity rise proportional to w3 at frequencies up to the peak.

Fig8 Image
FIGURE 8
The Raman spectrum of Nd1.85Ce0.15CuO4+d measured above and below the Tcof 21K. The incident laser energy of 1.92 eV is defined as zero, and only the downward 'red' shift is plotted. The inset expands the low-frequency scale for the 8 K data to accentuate the curvature.

Conclusions: The discovery of d-wave superconductivity in the electron-doped cuprates solved a puzzle that had troubled the high-Tc community for nearly 10 years: How can one class of materials exhibit such a high Tc via two different mechanisms? Adding the electron-doped cuprates to the class of dwave superconductors allows the research community to focus on possible mechanisms for d-wave pairing and why this leads to such high critical temperatures. Ultimately, the goal of this undertaking is to both raise Tc in the cuprate class of materials and to use what we have learned to predict and optimize new classes of materials with even higher critical temperatures.

Acknowledgments: Crystals were supplied by P. Fournier and R.L. Greene of the Center for Superconductivity Research and Department of Physics at the University of Maryland, College Park.

[Sponsored by ONR]




Sign Up for NRL News