🔭 astrophysics

The CRIMSON survey I: super-stellar SiO in the directly imaged companion TWA 5 B from high-resolution M-band spectroscopy

The CRIMSON survey presents the first high-resolution M-band detection of super-stellar gaseous SiO in the directly imaged companion TWA 5 B, revealing an atmosphere devoid of significant silicate clouds and consistent with a formation history involving core-accretion beyond the CO snowline or gravitational instability with substantial solid enrichment.

Original authors: Luke T. Parker, Jayne L. Birkby, Siddharth Gandhi, Vivien Parmentier, Vatsal Panwar, Matteo Brogi, Sophia R. Vaughan

Published 2026-06-03

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Luke T. Parker, Jayne L. Birkby, Siddharth Gandhi, Vivien Parmentier, Vatsal Panwar, Matteo Brogi, Sophia R. Vaughan

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: A Cosmic "Forensic" Investigation

Imagine the universe as a giant crime scene, and astronomers are the detectives trying to figure out how different "suspects" (planets and brown dwarfs) were born. Usually, detectives look at the "volatile" clues—things like water vapor and carbon monoxide (think of these as the steam and smoke rising from a fire). But in this paper, the team (the CRIMSON survey) decided to look for something much heavier and harder to find: Silicon.

Think of silicon as the "bricks and mortar" of the universe. While water and carbon are like the smoke that floats away easily, silicon is the heavy stone that stays put. By finding out how much silicon is in the atmosphere of a specific object, the team can figure out exactly what kind of "ingredients" were used to build it.

The Target: TWA 5 B

The object they studied is called TWA 5 B. It's not a normal planet orbiting a sun; it's a "super-Jupiter" (a giant gas ball) that is very young, very hot, and orbits a pair of stars far away.

The Temperature: It's about as hot as a blast furnace (2,400 Kelvin).
The Location: It's about 50 light-years away.
The Challenge: It's very dim compared to its host stars, making it like trying to see a firefly next to a spotlight.

The Tool: The "M-Band" Microscope

To see this faint object, the astronomers used a powerful telescope in Chile called the VLT, equipped with a special camera called CRIRES+.

The Problem: Usually, when looking at these objects, the Earth's atmosphere blocks the view, and the heat from the ground creates a lot of "static" noise. It's like trying to hear a whisper in a room full of screaming fans.
The Solution: They looked at a specific color of light called the "M-band" (infrared). In this specific color range, a molecule called Silicon Monoxide (SiO) sings a very loud song.
The Innovation: The team had to invent a new way to process the data. The telescope's detector was "stuttering" slightly, creating a fake pattern in the data (like a camera sensor that flickers). They wrote a custom computer program to smooth out this flicker, allowing them to hear the real signal.

The Discovery: A "Super-Stellar" Silicon Surprise

When they finally listened to the data, they found three things:

Water (H2O): Lots of it (as expected).
Carbon Monoxide (CO): A lot of it.
Silicon Monoxide (SiO): A huge amount of it.

This is the big news. They found silicon in the gas phase. Usually, in cooler objects, silicon gets trapped inside clouds (like dust or fog) and disappears from the gas. But because TWA 5 B is so hot, the silicon stayed as a gas.

The "Cloud" Clue:
Think of the atmosphere like a kitchen. If you have a pot of boiling water (the atmosphere) and you drop in flour (silicon), the flour usually sinks to the bottom or gets trapped in a cloud of steam.

The Finding: The team found that the "flour" (silicon) was still floating freely in the air.
The Conclusion: This means there are no thick clouds of magnesium-silicate dust in the upper atmosphere of TWA 5 B. If there were clouds, the silicon would have been hidden inside them, and the team wouldn't have seen it. The high amount of silicon gas proves the sky is clear.

The Formation Mystery: How was it built?

Now that they know how much silicon is there, they asked: "How did this object get so much silicon?"

They compared the amount of silicon in TWA 5 B to the amount in its host stars (the parents).

The Result: TWA 5 B has about 25 times more silicon than its parents.
The Analogy: Imagine a child who eats 25 times more broccoli than their parents. Where did all that broccoli come from?

This suggests TWA 5 B didn't just form from the same gas cloud as its parents. It must have "ate" a massive amount of solid, rocky material (silicate rocks) while it was growing.

Two Theories on how it grew:

The "Core Accretion" Theory: It started as a small rock that slowly sucked up gas and more rocks. However, the math doesn't quite work out; there wasn't enough "food" (rocky material) in that part of the solar system to build such a massive object this way.
The "Gravitational Instability" Theory: A chunk of the gas cloud collapsed on itself to form the object instantly (like a star forming). This theory fits the mass better, but to explain the extra silicon, this "baby" must have swallowed a huge pile of rocky debris after it was born.

The paper concludes that while we can't be 100% sure yet, the evidence points to the object forming far out in the system (beyond the "ice line" where water freezes) and then eating a lot of rocky dust to get so rich in silicon.

Why This Matters

This paper is a "proof of concept." It shows that we can now detect silicon gas in these distant, hot worlds.

Before: We could only guess what clouds were made of by looking at how the object looked in pictures.
Now: We can "taste" the atmosphere. If we see silicon gas, we know the clouds are gone or thin. If we don't see it, we know thick clouds are hiding it.

This gives astronomers a new, powerful tool to understand how giant planets and brown dwarfs are built, essentially allowing them to read the "receipt" of what ingredients were used to cook up these cosmic giants.

Technical Summary: The CRIMSON Survey I: Super-stellar SiO in the directly imaged companion TWA 5 B

Problem and Motivation
Understanding the formation history of giant planets requires linking their present-day atmospheric composition to the chemical partitioning of their protoplanetary discs. While volatile species (e.g., CO, H $_2$ O) have been extensively used to trace formation locations via C/O ratios, these measurements suffer from degeneracies due to complex thermal and chemical processing in discs. A complementary approach involves measuring refractory elements (e.g., Si, Fe), which remain in the solid phase throughout most of the disc. The ratio of refractory to volatile elements can constrain the accretion of solids versus gas. However, detecting refractory species in directly imaged planets is challenging. In ultra-hot Jupiters, refractory elements are often detected in atomic or ionized forms at visible wavelengths, but these species are limited to very high temperatures ( $T_{eq} \gtrsim 2000$ K). In cooler, directly imaged companions, refractory elements are expected to condense into clouds (e.g., magnesium-silicates), depleting the gas phase. To trace the initial refractory content and cloud properties, high-resolution spectroscopy must access refractory elements in molecular form. The M-band (3.5–5.2 $\mu$ m) offers a window to detect gaseous silicon monoxide (SiO), the dominant silicon carrier in hot atmospheres without silicate clouds, providing a direct probe of the atmosphere's refractory inventory and cloud condensation state.

Methodology
This study presents the first results of the CRIMSON survey, utilizing the CRIRES+ instrument on the VLT to perform high-resolution ( $R \sim 92,000$ ) M-band spectroscopy of the directly imaged companion TWA 5 B. TWA 5 B is a young ( $\sim 10$ Myr), hot ( $T_{eq} \approx 2400$ K) sub-stellar companion, ideal for detecting gas-phase SiO before significant cloud condensation occurs.

Observations: Two observing blocks were conducted on March 19, 2025, totaling 2.2 hours of exposure. The observations utilized a 0.2" slit and an ABBA nodding pattern to subtract the high thermal background inherent to M-band observations.
Data Reduction and Extraction: A critical methodological contribution is the development of a custom spectral extraction routine. The authors identified a periodic systematic in CRIRES+ 2D spectra caused by the discrete spatial sampling of the detector (0.056"/pixel) relative to the curved spectral trace. To mitigate this, they implemented an oversampling extraction routine (oversampling by a factor of 100) using spline interpolation to reconstruct the Point Spread Function (PSF) before extracting spectra at each spatial position. This reduced the systematic to $\le 10^{-3}$ of the flux level.
Post-Processing: The data were corrected for telluric absorption using molecfit and stellar contamination using Principal Component Analysis (PCA) applied in the wavelength domain. Crucially, the PCA eigenvectors were calculated excluding the spatial rows containing the companion signal to prevent the erosion of the companion's spectral features.
Analysis: Molecular detections were achieved via cross-correlation with atmospheric model templates. Atmospheric properties, including chemical abundances, dynamics, and temperature-pressure (TP) profiles, were retrieved using the HyDRA framework with a Bayesian nested sampling approach. The retrieval employed a "free chemistry" approach for H $_2$ O, CO, and SiO, and included a parametrized cloud deck model.

Key Results

Molecular Detections: The study reports the first conclusive detection of gaseous SiO in a directly imaged companion, with a signal-to-noise ratio (S/N) of 7.5. Strong detections of volatile species were also confirmed: CO (S/N = 9.1) and H $_2$ O (S/N = 18.8). The combined model was detected at S/N = 21.6.
Atmospheric Abundances: The retrieved volume mixing ratio (VMR) for SiO is $\log(\text{SiO}) = -3.56^{+0.42}_{-0.32}$ . This abundance is high, implying that the atmospheric silicon is contained almost entirely within the gas phase SiO, with negligible depletion into condensates.
Cloud Properties: The high SiO abundance, combined with a retrieved low cloud opacity ( $\log(\kappa) \approx -7.0$ ) and a poorly constrained cloud deck pressure pushed toward high pressures, indicates the absence of significant magnesium-silicate cloud condensation at or below the probed pressures ( $-2 \lesssim \log(P [\text{bar}]) \lesssim 0$ ). This is consistent with the object's temperature being above the condensation threshold for these clouds.
Elemental Ratios and Formation:
- Si/H: The atmosphere shows a super-stellar silicon enrichment, with $[\text{Si/H}]_\star = 1.41^{+0.42}_{-0.32}$ , representing a $\sim 25$ -fold enrichment relative to the host star.
- Volatile-to-Refractory Ratios: The C/O ratio is stellar ( $\sim 0.26$ ), while O/Si and C/Si are marginally sub-stellar.
- Formation Scenarios: These ratios are consistent with two formation pathways: core-accretion just within or beyond the CO snowline, or gravitational instability followed by substantial solid enrichment. However, the high mass of TWA 5 B ( $25 \pm 5 M_J$ ) and the wide orbit ( $\sim 86$ AU) make core-accretion difficult to reconcile with the estimated disc mass of the system. Gravitational instability followed by the accretion of $\sim 20 M_\oplus$ of silicate-rich solids post-collapse is presented as a viable alternative to explain the silicon enrichment.
Dynamics: The projected rotational velocity is measured at $v \sin(i) = 14.19^{+0.88}_{-0.79}$ km/s.

Significance and Claims
The paper claims that the detection of gaseous SiO in TWA 5 B provides a unique diagnostic for cloud properties in hot giant planets. Because SiO is the primary carrier of silicon in hot atmospheres and is depleted only by condensation, its abundance serves as a direct "litmus test" for the presence of magnesium-silicate clouds. The authors argue that this measurement confirms the absence of such clouds in TWA 5 B, allowing for a pristine measurement of the refractory content of the atmosphere.

Furthermore, the study demonstrates that high-resolution M-band spectroscopy can access refractory chemistry in molecular form, breaking degeneracies inherent in volatile-only studies. The authors posit that SiO abundance surveys across the population of directly imaged planets and brown dwarfs can determine the dominant cloud species forming (e.g., forsterite vs. enstatite) and their condensation temperatures, offering a ground-based alternative to space-based mid-infrared observations for characterizing cloud physics. Finally, the work highlights the necessity of custom extraction techniques for high-resolution slit spectrographs to overcome detector sampling systematics, a challenge relevant to future instruments like METIS on the ELT.