In this work, patterns in the behavior of different classes and types of thermoelectric materials are observed, and an alchemy that could help engineer a highly efficient thermoelectric is proposed. A method based on cross-correlation of Seebeck waveforms is also presented in order to capture physics of magnetic transition. The method is used to compute Curie temperature of LaCoO3 with an accuracy of 10K. In total, over 26 systems are analyzed, and 19 presented: Chalcogenides (PbSe, PbTe, Sb2Te3, Ag2Se), Skutterudites and Clathrates (CoSb3, SrFe4Sb12, Cd (CN)2, CdC, Ba8Ga16Si30*), Perovskites (SrTiO3, BaTiO3, LaCoO3, CaSiO3, Ce3InN*, YCoO3*), Half-Heuslers (ZrNiSn, NbFeSb, LiAlSi, CoSbTi, ScPtSb*, CaMgSi*), and an assorted class of thermoelectric materials (FeSi, FeSi2, ZnO, Ag QDSL*). Relaxation time is estimated from experimental conductance curve fits. A maximum upper bound of zT is evaluated for systems that have no experimental backing. In general, thermoelectric parameters (power factor, Seebeck coefficient and zT) are estimated for the aforementioned crystal structures. Strongly correlated systems are treated using LDAU and GGAU approximations. LDA/GGA/L(S)DA+U/GGA+U approach specific errors have also been highlighted. Densities of experimental results are estimated.

No description available.