Charge-density study of the nonlinear optical precursor DED-TCNQ at 20 K

A charge-density study of the nonlinear-optical (NLO) precursor {4-[bis(diethylamino)-methylium] phenyl}dicyanomethanide (DED-TCNQ), space group ${P2\mathrm{}}_1/c,$ $a=11.174\mathrm{}\left(2\right)\mathrm{}\AA ;$ $b=12.859\mathrm{}\left(2\right)\mathrm{}\AA ;$ $c=12.486\mathrm{}\left(2\right)\mathrm{}\AA ;$ $\beta =112.00\mathrm{}\left(1\right)\mathrm{}°,$ is presented. The results derive from a suitable combination of complementary 20 K x-ray and neutron diffraction data, the latter being important for locating the hydrogen atoms precisely. The compound is one in a series of TCNQ derivatives that exhibit varying degrees of quinoidal and zwitterionic character, these two electronic states being very close energetically. Bond-length-alternation type calculations show that the molecule at 20 K exists in a mixture of the two states, the zwitterionic ground state being dominant (63:37% zwitterionic: quinoidal). A topological analysis of the bonding density within the benzenoid ring provides for a more direct, alternative method to calculate this ratio which utilizes ellipticity values derived from the charge-density study. Results are identical thus corroborating the validity of the “strength-length” relationship implicitly assumed in bond-length-alternation type calculations. The ratio determined corresponds well to the electronic configuration needed to meet the requirements of the general rule for obtaining a maximum value of β (a measure of the NLO response on the molecular scale) as a function of bond-length alternation. The promise of this class of compounds for nonlinear optics also lies partly in their high molecular dipole moments and so the pseudoatomic charges derived from this study were used to evaluate the nature of the molecular charge transfer in detail and the solid-state dipolar vector moment μ. Such measurements of μ are otherwise difficult in the solid state. A value of $|\mu |=91\mathrm{}\times {10}^{-30}\mathrm{Cm}$ was deduced which compares with liquid and gas phase theoretical calculations of $\mu =66.71\mathrm{}\times {10}^{-30}\mathrm{Cm}$ and $\mu =33.36\mathrm{}\times {10}^{-30}\mathrm{Cm},$ respectively. This comparison, combined with an analysis of the sense of this vector, show that local crystal-field effects are highly influential in the solid state.