Symposimu A-5: THERMAL CONDUCTIVITY, PHONONS, THERMAL TRANSPORT PHENOMENA II /Aug. 31
In an invited talk entitled “Low electronic thermal conductivity and violation of the Wiedemann-Franz law in correlated metallic vanadium oxides”, Kedar Hippalgaonkar (Agency for Science, Technology and Research, Singapore) reported about a study moving from the inconsistency between the measured value of VO2 thermal conductivity and its predicted electronic component based upon the Wiedemann-Franz law. Through independent measurements of the lattice thermal conductivity he showed how the total conductivity would be expected to be much larger than its experimental value. Doping with W recovered the expected electronic contribution, showing how the Wiedemann-Franz “constant” is here a linear function of W content. A sophisticated model explained this anomaly as due to the electronic correlation in the material, where electrons behave as a coherent/incoherent Fermi liquid.
Kunpeng Zhao – Cu2S has lower lattice k, Cu2Se has higher PF, what happens if we put them together, nanodomains with 20-8 0nms are formed and a modulated cubic phase is found, which is good for reducing lattice thermal conductivity. Phase transition from 300-700K giving large heat capacity below 700K… Electrical conductivity determines that the PF lies between the two materials. The carrier concentration is increased, so the seebeck drops and conductivity increases. Maximum value comes from those with Cu vacancies. Pisarenko Plot gives an effective mass of 4.2m0, but higher mobility, giving a higher weighted mobility. Low lattice conductivity (~0.3-0.4 W/m-K) with higher PF gives ZT as high as 2.3 at 1000K at 10^20-10^21 cm-3 carrier concentration. A question on stability was asked and the author showed data that the phase transition has hysteresis, but proving that there’s stability.
Lanling Zhao: Investigation of TE properties of Cu2Se(S) – Liquidlike materials should have lower specific heat at higher temperature. Increased electrical resistivity and Seebeck at higher temperatures and slightly reduced kL (~0.6 W/m-K). zT ~ 1.2 up to 1.8 (single crystals) comparable to polycrystalline bulk. Stability is not an issue for water-quenched CuSe samples. Copper deficient CuS shows increased electrical conductivity, similar kL, but smaller S. But due to larger sigma, zT is higher, but this has poor electrical and thermal stability. Next, the author talked about Te and I doped samples and found that Te is better for electrical conductivity. Highest ZT is for CuSe, and the Te/I doping does not help. Then, the author introduced Sulfur doped CuSe, for doping levels <0.16 where the structure remains monoclinic. For higher doping, the samples are composite multiphase. TE properties show lower electrical conductivity and increased Seebeck, but the CuSe ZT cannot be improved. Finally, the author introduced composites with CNTs, graphite and hard Carbon – graphite forms nanoclusters with a grain size ~30-60nms. Carbon doping gives higher sigma and comparable Seebeck and reduced kappa, obtaining higher ZT around 2.4 at 970K.
Terumasa Tadano – Peierls Boltzmann Theory (phonon gas model) deals with cubic anharmonic scattering. Full anharmonicity is considered in ab initio MD. The author proposed a new method to decrease computational cost, which is between BTE and ab-initio. Usually, DFT considers 2nd and 3rd order force constants, then uses BTE, neglecting 4th order anharmonicity. This breaks down when there’s an imaginary mode underlining that quartic anharmonicity is important, for eg. SnSe (TE materials) ‘soft’ materials causes frequency shift and introduce 4-phonon scattering. The author uses Self-Consistent Phonon Theory to obtain an effective Hamiltonian to obtain a temperature dependent phonon dispersion. Inelastic Nuetron Scattering (INS) experimentally corroborates this. Tested on cubic STO.T-dependent phonon dispersion but PbTe and SnSe show very strong temperature dependence due to more soft modes. Hence, this allows for self-corrected phonons to be added to a BTE calculation, and the key is to use 2nd and 4th order force constants. This calculation was also performed on clathrates (BGG), where a hardening of the rattling mode is observed. These rattling modes scatter acoustic phonons and hence the details are important resulting in an enhancement in lattice thermal conductivity due to quartic anharmonicity. Software is now available at ALAMODE at http://ttadano.github.io/alamode/