Direct Determination of the HF/H2 Abundance Ratio in Interstellar Gas

TL;DR

This study reports the first direct measurement of the HF/H2 abundance ratio in interstellar gas using ro-vibrational transitions, revealing lower ratios than predicted by models, especially in dense regions.

Contribution

First direct determination of HF/H2 abundance ratio in interstellar gas through ro-vibrational transition observations, providing new insights into fluorine chemistry in different interstellar environments.

Findings

HF/H2 ratios are below predicted values in most observed regions.

Warm, dense gas near protostars shows different HF/H2 ratios.

Chemical models partially reproduce observed ratios but fail in dense, cold regions.

Abstract

We report the first detection of the v=1--0, R(0) ro-vibrational transition of HF at 2.499385 microns arising from interstellar gas. The line is seen in absorption toward 3 background sources---HD 154368, Elias 29, and AFGL 2136 IRS 1---all of which have reported H2 column densities determined from observations of H2. This allows for the first direct determination of the HF/H2 abundance ratio. We find values of N(HF)/N(H2)=1.15*10^-8 and 0.69*10^-8 for HD 154368 and Elias 29, respectively. The sight line toward AFGL 2136 IRS 1 also shows absorption from the v=1--0, R(1) transition of HF, indicating warm, dense (n_H > 10^9 cm^-3) gas, likely very close to the central protostar. Ascribing portions of the HF absorption to warm and cold gas, we find N(HF)/N(H2)=(1.7--2.9)*10^-8 and (0.33--0.58)*10^-8 for the two components, respectively. Except for the warm component toward AFGL 2136 IRS 1, all observed HF/H2 ratios are well below N(HF)/N(H2)=3.6*10^-8, the value predicted if nearly all gas-phase fluorine is in the form of HF. Models of fluorine chemistry that account for depletion onto grains are able to reproduce the results toward HD 154368, but not in the cold, dense gas toward AFGL 2136 IRS 1 and Elias 29. Most likely, some combination of simplifying assumptions made in the chemical models are responsible for these discrepancies.