Research

Galaxies form stars out of molecular gas, so measuring how much molecular gas a galaxy has is a precondition for understanding how it grows. That measurement is harder than it sounds. Most of our census of molecular gas comes from carbon monoxide, but CO is fragile: in diffuse, weakly shielded gas it is photodissociated away while the hydrogen molecules survive. The result is “CO-dark” molecular gas — real molecular material that simply does not show up in the surveys we use to find it.

My work is largely about finding that gas, weighing it, and figuring out what it means for star formation. The main tool is the ground-state 18 cm hyperfine lines of the hydroxyl radical (OH), which stays visible where CO does not.

CO-dark molecular gas in the Milky Way

Spectra toward the outer Galaxy comparing HI 21 cm, OH 18 cm from the GBT and a 20 m telescope, and CO 3 mm. OH emission appears across the Local Arm, inter-arm gas, Perseus Arm, and outer Galaxy, while CO is detected only in the Local Arm.
HI 21 cm, OH 18 cm (GBT and 20 m), and CO 3 mm spectra along one outer-Galaxy sightline. OH is detected in the Local Arm, the inter-arm gas, the Perseus Arm, and out to R ~ 10–12 kpc. CO shows up only in the Local Arm. From Busch et al. 2021.

My thesis work used the Green Bank Telescope to survey 18 cm OH emission in the outer Galaxy and the Perseus Arm, where the CO-dark fraction should be largest. Two results came out of it. First, OH emission is widespread in regions with no detectable CO, and the gas it traces is diffuse and cold rather than clumped into recognizable clouds (Busch et al. 2019). Second, this material is not confined to the thin molecular disk — there is a thick disk of dark molecular gas extending well off the Galactic plane, which conventional CO-based inventories miss entirely (Busch et al. 2021).

Model grid of CO(1-0) integrated intensity versus OH column density for a range of extinctions and densities, with observed sightlines overplotted.
CO(1–0) integrated intensity against OH column density, with model tracks for different total extinction AV and density nH. Sightlines with plenty of OH but little or no CO are where the CO-dark gas lies. From Busch et al. 2019.

Taken together, these say that the molecular reservoir of the Galaxy is both larger and more extended than the CO maps imply.

CryoPAF on Murriyang

Surveys of faint, extended emission are limited by how much sky you can see at once. A single-pixel receiver looks at one point on the sky; a phased array feed forms many beams simultaneously, turning the telescope into a proper survey instrument for exactly the diffuse signals I care about.

I am currently working with CryoPAF, the cryogenically-cooled phased array feed on Murriyang, the CSIRO Parkes 64 m telescope. A cooled PAF is the combination that matters here: the wide field of view makes large-area mapping tractable, and the cryogenic front end keeps the system temperature low enough that genuinely faint diffuse HI and OH stay within reach. The goal is to map the diffuse molecular interstellar medium over far more sky than pointed observations have allowed.

The first step is a 100-hour Murriyang program (P1385) mapping OH in a high-latitude cloud — a region where the CO-dark fraction should be high and the emission faint enough that the survey speed of a cooled PAF is what makes the measurement possible at all.

Animated channel maps of OH 1667 MHz emission across a high-latitude cloud from the P1385 CryoPAF program, with the mean spectrum below marking each channel's velocity.
Channel maps of OH 1667 MHz emission from the P1385 CryoPAF data toward the π cloud, a high-latitude cloud at b ≈ −70° about 300 pc away on the edge of the Local Bubble. The panel underneath shows the mean spectrum, with the current channel's velocity marked. CO emission is found only in the southeastern part of the cloud, so OH is tracing a significant amount of CO-dark gas. Busch et al., in prep.

Extragalactic OH and the Local Group

Dust mass surface density map of M31 with two GBT pointing positions marked by circles in the southern part of the disk.
Dust mass surface density in M31. The circles mark the two GBT pointings whose spectra are shown below. From Busch 2024.

If OH reveals hidden molecular gas in the Milky Way, it should do the same in other galaxies — but until recently nobody had detected thermal OH emission outside our own. Using the GBT, I realized I could carry out a pilot search in nearby galaxies for the same signal I had found in the Milky Way. I reported the first extragalactic detection of thermal 18 cm OH emission, toward Andromeda (M31), and found it traces an abundant reservoir of CO-faint molecular gas (Busch 2024).

This opens the technique up to the Local Group, where I also work on cold gas in low-metallicity dwarf galaxies through the Local Group L-band Survey (LGLBS) with the VLA — systems like NGC 6822, IC 10, and WLM, where low metallicity should make the CO-dark fraction larger still.

GBT spectra at two positions in M31. The left position shows HI, OH 1667, 1665, and 1720 MHz, and CO emission near -550 km/s; the right position shows HI but no clear OH.
GBT spectra toward the two M31 positions. Left, 00:39:32 +40:26:09: OH 1667 and 1665 MHz emission lines up in velocity with HI and CO. Right, 00:40:24 +40:28:11: HI is present but OH is not clearly detected. From Busch 2024.

The diffuse ISM in absorption

Emission is not the only way to find this gas. Absorption spectroscopy against bright background continuum sources gives a sharp, sensitive probe of individual diffuse clouds along a sightline. I contribute to work using HCO+ absorption to trace CO-dark gas in the diffuse ISM, and to the HyGAL program, which characterizes the Galactic interstellar medium through hydrides and other small molecules. These absorption measurements provide the physical conditions — temperature, density, abundance — that the emission surveys need in order to be converted into masses.

Observing programs

Most of this work depends on getting large amounts of time on single-dish and interferometric telescopes, since diffuse emission is faint and the surveys have to be wide. Since 2019 I have been awarded 2,803 hours as PI or co-I across the GBT, the VLA, ALMA, NOEMA, and Murriyang. A representative sample:

ProgramFacilityTime
Is the broad 18 cm OH emission “disk” in concordance with Galactic structure?GBT300 h
Deep OH absorption in the Outer Galaxy (co-I)VLA350 h
Probing the structure of the diffuse molecular ISM — mixed or clumped?GBT216 h
A GBT search for the diffuse molecular gas disk in absorptionGBT177 h
A survey for OH in the northern molecular ring of M31GBT176 h
A GBT survey of M33’s molecular gas in OHGBT168 h
Turbulent motions, mass content and distribution of dark H2 in the Outer GalaxyGBT135 h
Extending the GBT M31 molecular ring survey in OHGBT132 h
A follow-up survey for OH in the Milky Way’s nuclear windGBT121 h
Tracing dark molecular gas in the Eos molecular cloud with 18 cm OHGBT100 h
CryoPAF mapping of OH in a high-latitude cloudMurriyang100 h
An ultra-deep search for OH and CH in the diffuse ISM of the Magellanic systemMurriyang100 h
Detangling the 3D multi-phase structure of the Perseus cloud with 18 cm OHGBT92.5 h

Alongside these I contribute to multi-tracer NOEMA and VLA programs on the structure and chemistry of diffuse molecular gas near the HI-to-H2 transition.


Interested in working on any of this? See the students page, or just email me at mpbusch@nrao.edu.