Significantly different normal state Lorenz numbers have been found in two independent direct measurements based on the Righi-Leduc effect, one about 6 times smaller and the other one about 2 times larger than the Sommerfeld value in single cuprate crystals of the same chemical composition. The controversy is resolved in the model where charge carriers are mobile lattice bipolarons and thermally activated nondegenerate polarons. The model numerically fits several longitudinal and transverse kinetic coefficients providing a unique explanation of a sharp maximum in the temperature dependence of the normal state Hall number in underdoped cuprates.
This is a pre-print. It is also available at: http://arxiv.org/abs/cond-mat/0601202. The definitive version: ALEXANDROV, 2006. Hall-Lorenz number paradox in cuprate superconductors. Physical Review B, 73(10), 100501(R), is available at: http://prb.aps.org/.