Knowledge Bank

The unexpectedly high column density of $H3+$ $observeda,b$ in the diffuse interstellar medium towards Cygnus OB2 12 presented quite an enigma for interstellar chemistry. Since $H3+$ has now been detected in many $sightlinesc$, it is clear that there is a general problem with the chemical models. The standard model of $H3+$ chemistry in diffuse clouds contains only three parameters: $\zeta$ (the $H2$ ionization rate), $ke$ (the $H3$ dissociative recombination rate constant), and $[e]/[H2]$ (the electron fraction) --- evidently, at least one of the assumed values for these parameters is in error by one to two orders of magnitude. There are only three options: (1) the electron fraction is lower than expected from complete conversion of $C\to C+,(2)$ the value of $ke$ for $J=1H3+$ (and cold electrons) is lower than that reported in the laboratory for rotationally hot $H3+$, and/or (3) the value of $\zeta$ is higher in diffuse clouds than in dense clouds due to the presence of low-energy cosmic rays. We report new observations of $H3+$ in the diffuse interstellar medium using CGS4 at UKIRT and IRCS at Subaru, which seem to rule out option (1). In particular, we have tentatively detected $H3+$ in the classical diffuse cloud towards $\zeta$ Per, toward which both $C+$ and $H2$ have been spectroscopically measured. It now seems that it is up to the dissociative recombination community to unambiguously determine the appropriate value of $ke$ through both experiment and theoretical calculation. With a definitive value of $ke$ in hand, observations of $H3+$ will then serve as a direct probe of the cosmic-ray ionization rate $\zeta$ in diffuse clouds.