JWST observations of cosmic-ray-excited H$_2$ in Barnard 68: spatial variations and constraints on cosmic-ray attenuation

TL;DR

This study uses JWST/NIRSpec to map spatial variations of cosmic-ray excited H$_2$ (CRXH$_2$) emission in Barnard 68, distinguishing it from UV-pumped H$_2$ and measuring how CR ionization attenuates with shielding depth. By modeling the cloud as a Bonnor-Ebert sphere and parameterizing CR attenuation as $\zeta_{H_2}(N_s)=\zeta_0/(1+N_s/N_0)^{\alpha}$, the authors fit the observed $I_{CRXH_2}/N(H_2)$ profile across 16 sightlines, deriving a reference ionization rate $\zeta_{ref}$ of about $1.4\times10^{-16}$ s$^{-1}$ at $N_{ref}=3\times10^{21}$ cm$^{-2}$ and documenting clear attenuation with shielding. The results show a significant spatial variation in CRXH$_2$ strength that cannot be explained by dust extinction alone, providing the most direct constraints to date on CR penetration into cold, dense gas and informing microphysical CR–H$_2$ interactions. CRXH$_2$ emission thus emerges as a powerful diagnostic of the local cosmic-ray environment in molecular clouds and star-forming regions.

Abstract

We present James Webb Space Telescope (JWST) NIRSpec observations of the starless dark cloud Barnard 68 that reveal the spatially-resolved signature of cosmic-ray excited molecular hydrogen (CRXH$_2$) emissions for the first time. Following up on our initial detection of CRXH$_2$ emissions from B68 (Bialy et al. 2025), we now exploit JWST's sensitivity and spatial multiplexing to map CRXH$_2$ rovibrational lines across 16 sight lines through the cloud. By disentangling the CRXH$_2$ and UV-pumped H$_2$ components, we isolate the para-H$_2$-dominated spectrum attributable to cosmic-ray excitation. We find that there are significant spatial variations in the ratio of the CRXH$_2$ line intensity to the line-of-sight H$_2$ column density; these cannot be accounted for by dust extinction alone and demonstrate a clear attenuation of the cosmic-ray flux with increasing shielding column. Modeling B68 as a Bonnor-Ebert sphere, we constrain both the unshielded cosmic-ray ionization rate, $ζ_{\rm H_2}$, and how it decreases with shielding column. At a reference depth of $N({\rm H}_2) = 3 \times 10^{21}$ cm$^{-2}$, we infer $ζ_{\rm H_2} \approx 1.4 \times 10^{-16}$ s$^{-1}$, a factor of $\approx 3$ higher than the average value derived from H$_3^+$ absorption studies. These results provide the most direct probe to date of cosmic-ray penetration into cold, dense gas, offering new constraints on both the microphysics of CR-H$_2$ interactions and the attenuation of low-energy cosmic rays in molecular clouds. Our findings establish CRXH$_2$ emission as a powerful new diagnostic of the cosmic-ray environment in interstellar space.

Figures (12)

• Figure 1: Arrangement of NIRSpec MOS shutters across Barnard 68. White lines show the locus of the 382 open shutters used to obtain spectra across the cloud. The 16 averaged slitlets ($\sim$11$"$ each) are marked with white dots. The background ("OFF") position, observed 30$'$ north of B68, used the same configuration. The extinction map is adapted from Alves et al. (2025, private comm.).
• Figure 2: Example H$_2$ rovibrational line profiles from B68 (black histograms) with Gaussian fits (red curves).
• Figure 3: Measured line widths (upper panel) and centroid velocities (lower panel) as a function of wavelength for all detected H$_2$ lines. Para-H$_2$ lines are shown in red, and ortho-H$_2$ lines in blue. The dashed curve in the upper panel is a quadratic fit to the unresolved instrumental profile. In the lower panel, horizontal lines show mean velocity centroids for ortho-, para-, and all H$_2$ lines; the scatter is consistent with calibration uncertainties.
• Figure 4: Separation of UV-excited and cosmic-ray--excited H$_2$ emission. Top: integrated line intensities at the OFF position, divided by predicted UV fluorescence fractions, from the 100 K models of Sternberg (1988). Ortho-H$_2$ and para-H$_2$ transitions are shown in blue and red. Bottom panel: residual B68 line strengths attributable to CR excitation.
• Figure 5: Spatial variation of CR-excited H$_2$ line intensities across B68. Shown are three of the strongest lines, plotted as a function of position along the slit (see Figure 1). Intensities are corrected for UV contamination using the OFF position, and are given in units of $10^{-7}\,\rm erg\,cm^{-2}\,s^{-1}\,sr^{-1}$.