I led JWST GTO-1190 as a part of the NIRCam Science team to characterize the initial mass function (IMF) in NGC2024, a nearby (~ 400 pc) young (0.5 Myr) star-forming region. With deep NIRCam imaging in eight filters across 0.7-5 microns, we identified 48 likely cluster members that are all sub-stellar. I created an automated point source detection technique that distinguishes point sources from extended, nebular emission, which most modern point source detection algorithms are unable to perform. Then, we determined cluster membership by comparing the observed photometry to that of a simulated Galactic field population assuming a local extinction estimate from archival Herschel data. From these observations, we were sensitive to objects of 0.5 Jupiter masses and 1 Jupiter mass through 30 magnitudes of visual extinction. From the likely cluster members, we found that a double power law function is strongly favored over any other model with an increasing mass function in linear space down to ~ 12 Jupiter masses, then a decreasing mass function down to our sensitivity limit of 0.5 Jupiter masses. This is the first identification of a turnover in the mass function in any stellar population to date. We find no likely cluster members below 3 Jupiter masses, although we are sensitive to lower masses. While a small sample, our results demonstrate that the mass function does not continue to increase down to smaller masses, but turns over toward the opacity limit of fragmentation (still an unknown quantity). We were awarded JWST Cycle 3 (GO 5409, PI: De Furio) time with NIRSpec/MSA to obtain 0.6-5.3 micron spectroscopy of the observed sources in NGC 2024. In the same program, we obtained NIRCam parallel imaging in the periphery of NGC 2024 to explore intra-cluster differences in the mass function, e.g. do high mass stars in the core of NGC 2024 inhibit the production of low mass objects down to the opacity limit of fragmentation. This work is in progress, and we anticipate the publication of the spectra in early 2026. See this link for the NASA press release, and this link for the ApJL publication.

Fig. 1. From De Furio et al. (2025). Color combined image of half of our NGC 2024 NIRCam data. Filters F115W and F140M are blue, F182M is green, F360M is orange, and F430M is red. Gaps in shortwave channel data are filled in using VISTA\VIRCAM data for reference. The F430M filter serves as the luminosity filter for the VISTA image, where the blue colors come from VISTA but the resolution comes from the F430M data. Shifting colors and balance adjustments are then made to fill in the gaps from the shortwave data. Credit: Alyssa Pagan, STScI.

Fig. 2. From De Furio et al. (2025). Completeness across the field of view of our NIRCam data in the F182M and F430M filters. Color bar shows the fractional area where we are sensitive to the specific color–magnitude combination. The purple background indicates no sensitivity, either due to saturation or being below our detection limit. Red line is a 1 Myr isochrone from the ATMO2020 chemical equilibrium models, cyan dashed lines are extinction vectors, and red circles are the likely cluster members. While we are sensitive to 0.5 Jupiter mass objects, we only identified one 3 Jupiter mass object as a likely cluster member.

Fig. 3. From De Furio et al. (2025). Best-fit mass function to the NIRCam data shown in black, 68% confidence interval and 95% confidence intervals of the low-mass power law are shaded in red and gray, respectively. Our results demonstrate that the mass function does not continue to increase down to the opacity limit, that there is a turnover in the mass function, and that it is either flat or decreasing down to 0.5 Jupiter masses.