We report a study on the heat transport of an S = 1 Haldane chain compound Ni(C_3H_{10}N_2)_2NO_2ClO_4 at low temperatures and in magnetic fields. The zero-field thermal conductivities show a remarkable anisotropy for the heat current along the spin-
chain direction (kappa_b) and the vertical direction (kappa_c), implying a magnetic contribution to the heat transport along the spin-chain direction. The magnetic-field-induced change of the spin spectrum has obviously opposite impacts on kappa_b and kappa_c. In particular, kappa_b(H) and kappa_c(H) curves show peak-like increases and dip-like decreases, respectively, at sim 9 T, which is the critical field that minimizes the spin gap. These results indicate a large magnetic thermal transport in this material.
To investigate the validity of the Wiedemann-Franz (WF) law in disordered but metallic cuprates, the low-temperature charge and heat transport properties are carefully studied for a series of impurity-substituted and carrier-overdoped La_{1.8}Sr_{0.2
}Cu_{1-z}M_zO_4 (M = Zn or Mg) single crystals. With moderate impurity substitution concentrations of z = 0.049 and 0.082 (M = Zn), the resistivity shows a clear metallic behavior at low temperature and the WF law is confirmed to be valid. With increasing impurity concentration to z = 0.13 (M = Zn) or 0.15 (M = Mg), the resistivity shows a low-T upturn but its temperature dependence indicates a finite conductivity in the T to 0 limit. In this weakly-localized metallic state that is intentionally achieved in the overdoped regime, a {it negative} departure from the WF law is found, which is opposite to the theoretical expectation.
The temperature and magnetic-field (H) dependences of thermal conductivity (kappa) of Bi_2Sr_2CaCu_2O_{8+delta} (Bi2212) are systematically measured for a broad doping range by using both pure Bi2212 single crystals with tuned oxygen contents and Bi_
2Sr_2Ca_{1-x}Dy_xCu_2O_{8+delta} (Dy-Bi2212) single crystals with different Dy contents x. In the underdoped samples, the quasiparticle (QP) peak below T_c is strongly suppressed, indicating strong QP scattering by impurities or oxygen defects, whereas the phonon conductivity is enhanced in moderately Dy-doped samples and a phonon peak at 10 K is observed for the first time in Bi2212 system, which means Dy^{3+} ions not only introduce the impurities or point defects but also stabilize the crystal lattice. The subkelvin data show that the QP heat conductivity gradually decreases upon lowering the hole doping level. The magnetic-field dependence of kappa at temperature above 5 K is mainly due to the QP scattering off vortices. While the underdoped pure Bi2212 show very weak field dependence of kappa, the Dy-doped samples present an additional dip-like term of kappa(H) at low field, which is discussed to be related to the phonon scattering by free spins of Dy^{3+} ions. For non-superconducting Dy-Bi2212 samples with x simeq 0.50, an interesting plateau feature shows up in the low-T kappa(H) isotherms with characteristic field at 1 -- 2 T, for which we discuss the possible revlevance of magnon excitations.
To study the effects of paramagnetic spins on phonons, both the in-plane and the c-axis heat transport of GdBaCo_{2}O_{5+x} (GBCO) single crystals are measured at low temperature down to 0.36 K and in magnetic field up to 16 T. It is found that the p
honon heat transport is very strongly affected by the magnetic field and nearly 5 times increase of the thermal conductivity in several Tesla field is observed at 0.36 K. It appears that phonons are resonantly scattered by paramagnetic spins in zero field and the application of magnetic field removes such strong scattering, but the detailed mechanism is to be elucidated.
Recently, it was proposed that phonons are specularly reflected below about 0.5 K in ordinary single-crystal samples of high-T_c cuprates, and that the low-temperature thermal conductivity should be analyzed by fitting the data up to 0.5 K using an a
rbitrary power law. Such an analysis yields a result different from that obtained from the conventional analysis, in which the fitting is usually restricted to a region below 0.15 K. Here we show that the proposed new analysis is most likely flawed, because the specular phonon reflection means that the phonon mean free path ell gets LONGER than the mean sample width, while the estimated ell is actually much SHORTER than the mean sample width above 0.15 K.
