Decomposition of spin susceptibility of tetrathiafulvalenium-tetracyanoquinodimethanide (TTF-TCNQ) and analogous compounds

The method of separating the spin susceptibility, $\chi$, of tetrathiafulvalenium-tetracyanoquinodimethanide (TTF-TCNQ) and analogous quasi-one-dimensional compounds into the contributions of the individual chains by means of EPR data is reexamined in the light of more recent information on the compounds. For TTF-TCNQ it is concluded that, if proper consideration is given to the experimental uncertainties, the method can be used to obtain reliable qualitative results. For hexamethylene-tetrathiafulvalenium tetracyanoquinodimethanide (HMTTF-TCNQ), however, it is shown that the decomposition is unreliable since analysis of the data for different directions of magnetic field yields quite different results for the individual stack susceptibilities. It is suggested that the reason the decomposition does not work in this case, among others, is that the interstack interaction is too large. Recognizing these difficulties with the decomposition removes a number of apparent paradoxes or anomalies, such as: (1) the sign for thermoelectric power being opposite to the sign of charge of the carrier responsible for conduction; (2) $\chi$ for TTF being 5 or more times $\chi$ for tetraselenafulvalenium; (3) $\chi$ for the HMTTF chain being independent of temperature over the range where for all other chains investigated in this family it is not.