Absolute Infrared Intensities of the Fundamental Absorption Bands in Solid Cyanogen Halides

Absolute absorption intensities of the three infrared‐active fundamental vibrations of each of the cyanogen halides—CICN, BrCN, and ICN—have been measured for crystalline samples, using experimental techniques developed by Hollenberg and Dows. Comparison between integrated intensities obtained with “drawn in” backgrounds and those obtained after integrating the absorption coefficient, $\kappa$ , obtained from the measured transmittance by a Kramers–Kronig method, shows that the results agree within their estimated experimental uncertainty. The intensities of the stretching fundamentals in the solid phase differ drastically from those in the gas phase, e.g., ${\nu }_1$ for ClCN goes from an estimated 900 darks in the gas phase to 0 in the solid, ${\nu }_3$ increases by a factor of 2 from gas to solid for ClCN, and by a factor of about 4 for BrCN. However, the solid phase intensities of ${\nu }_2$ , the bending motion perpendicular to the molecular chains in the solid, are only slightly larger than the gas phase values, indicating that the major perturbation of the vibrational intensities in crystals of these molecules is parallel to the molecular chains. The dipole derivatives with respect to the change in CN and in CX bond lengths ${\mathrm{(\partial \mu \hspace{0.17em}/\hspace{0.17em}\partial r}}_i)$ have been computed for both gas and solid phases. A resonance structure argument leads to predictions for the signs expected for ${\mathrm{\partial \mu }}_g{\mathrm{\hspace{0.17em}/\hspace{0.17em}\partial r}}_{\mathrm{CX}}$ and ${\mathrm{\partial \mu }}_g{\mathrm{\hspace{0.17em}/\hspace{0.17em}\partial r}}_{\mathrm{CN}}$ , and for the changes expected as a result of intermolecular interactions in the solid phase. This qualitative resonance structure argument (which also successfully accounts for the nuclear quadrupole resonance results) is consistent with only one choice of signs for the ${\mathrm{\partial \mu }}_g{\mathrm{\hspace{0.17em}/\hspace{0.17em}\partial Q}}_i$ and ${\mathrm{\partial \mu }}_s{\mathrm{\hspace{0.17em}/\hspace{0.17em}\partial Q}}_i$ values. An alternative argument, independent of the model for the interaction, also leads to the same conclusion regarding the sign choice, increasing confidence that this choice is correct.