Observational constraints on CO sequestration and disk mass in the planet-hosting PDS 70 disk

arXiv:2611.14632 · astro-ph.EP, astro-ph.SR · Submitted 2026-11-10 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Astrophysics Radio. Generated commentary on the latest astrophysics papers.

Vera: We're looking at the paper: "Observational constraints on CO sequestration and disk mass in the planet-hosting PDS 70 disk".

Jocelyn: Here's the gist of the paper. Exoplanets accrete gas, dust, and ice from their natal environment, setting the chemical composition of their atmospheres and interiors The derived peak CO abundance is (4.8 + 4.3 − 2.7 − 17+8 −8) × 10−6 H2 assuming that the N2H+ emission is optically thin and is characterized by an excitation temperature of 21 K The disk mass of the PDS 70 disk is 0.011+0.011 − 0.039+0.067 − 0.018 M⊙ (12+12 − 6 − 4) MJup assuming that the N2H+ emission is optically thin and emits at 21 K This work presents the first detection of N2H+ and C17O in the PDS 70 disk, both of which were serendipitously observed by ALMA Discovery of Molecular Tracers The research focuses on quantifying the peak CO abundance and disk mass of the PDS 70 disk, which is noted as "the first disk with a direct confirmation of two accreting planets in its dust cavity" The detection of N2H+ and C17O in this disk was achieved using ALMA observations, specifically the N2H+ J = 1 − 0 transition at 93.173 GHz and the C17O J = 3 − 2 transition at 337.061 GHz The C17O emission shows "marginal statistical evidence for an azimuthal asymmetry trailing that previously seen in the continuum emission" These detections are crucial because they allow for constraints on the disk's chemical composition and formation location Constraining Disk Properties via Modeling The authors leverage a framework developed by Trapman et al.

Vera: (2025b) to constrain the disk mass and CO abundance using these line fluxes together with archival continuum and other CO isotopologue observations The comparison is made against a grid of thermochemical DALI models, specifically utilizing the transition disk model grid where models have a dust cavity but no gas cavity The results from this method are in "good agreement with independent disk mass measurements from rotation curves for the disks in the exoALMA large program" Sensitivity to Observational Choices The derived parameters are sensitive to which fluxes are included in the fitting, as demonstrated by a comparison between fits including and excluding the C17O line flux The fit including the C17O line flux prefers a more extended disk with Rc = 58+9 − 7 au and a disk mass of Mdisk = 0.011+0.011 − 0.005 M⊙, whereas the fit excluding the C17O emission favors a smaller disk with Rc = 46+7 − 6 au and disk mass that is ∼ 3.4× lower The peak CO abundance is more or less independent of whether the C17O line flux is included in the fitting because "the CO depletion is mainly set by the N2H+ flux" Systematic Uncertainties and Future Directions The large gas gap observed in the PDS 70 disk, which is not included in the adopted model grid, may introduce "substantial systematic uncertainties in the inferred disk parameters that are not captured by the formal posterior distributions" A sensitivity test showed that doubling the integrated C17O flux almost exactly matches the posteriors when C17O is not included in the fitting, highlighting that derived disk mass, size and gas-to-dust mass ratio are sensitive to a factor of 2 change in the C17O flux A more flexible model grid including a range of gas cavity sizes and depletion factors is therefore needed to quantify these systematic effects The CO sequestering in the outer disk gas probed by the observations suggests that "the PDS 70 planets may accrete the final portion of their atmospheres from this CO sequestered reservoir" The findings suggest that CO sequestration may need to be considered when interpreting the chemical composition of planetary atmospheres The system is a good candidate to explore a direct link between the atmospheric composition of the planets and their natal environment Including grain surface chemistry involving CO ice can reconcile observed fluxes with model predictions, suggesting that at least part of the sequestered CO is locked in the form of e.g., CO2 ice and icy complex organic molecules METIS and HARMONI on the ELT can further improve our understanding by characterizing exoplanetary atmospheres The PDS 70 disk mass is similar to that of Herbig disks and massive T Tauri disks The derived peak CO abundance is similar to what is found in other T Tauri disks and is approximately an order of magnitude lower than that in the ISM after accounting for CO freeze-out and isotope selective photodissociation of CO Final Summary In summary, the effective peak CO abundance relative to H2 in the PDS 70 disk layers probed by the observations is (4.8 + 4.3 − 2.7 − 17+8 −8) × 10−6 The disk mass of the PDS 70 disk is 0.011+0.011 − 0.039+0.067 − 0.018 M⊙ assuming that the N2H+ emission is optically thin and emits at 21 K The range of a factor of ∼ 3.4 in the derived disk mass is driven by the dependence on whether the C17O 3 − 2 line flux is included in the mcmc sampling procedure The range is due to the optical depth and excitation temperature of the N2H+ emission and to a lesser degree to whether the C17O flux is included in the fitting The peak CO abundance is independent of this The derived peak CO abundance is not necessarily the global CO/H2 ratio in the disk but rather the peak CO abundance in the emitting layer probed by the observations Key Findings 1.

Jocelyn: The effective peak CO abundance relative to H2 in the PDS 70 disk layers probed by the observations is (4.8 + 4.3 − 2.7 − 17+8 −8) × 10−6 2.

Vera: The disk mass of the PDS 70 disk is 0.011+0.011 − 0.039+0.067 − 0.018 M⊙ (12+12 − 6 − 4) MJup assuming that the N2H+ emission is optically thin and emits at 21 K 3.

Paper summary: Vera: So, wrapping up this discussion on the paper "Observational constraints on CO sequestration and disk mass in the planet-hosting PDS seventy disk," we've seen how ALMA data lets us measure two key things: the peak carbon monoxide abundance and the total mass of that disk.

Jocelyn: And what this means is that these measurements give us a real number for how much carbon monoxide is present in the layers we can actually see, which is important for understanding the chemistry of planets forming around these disks >.

Subrahmanyan: The authors are using a framework from Trapman et al 2025b to combine their line fluxes with other data, and they're getting results for disk mass that match what we get from rotation curves of other exoplanet disks >.

Vera: And the authors are really careful about the uncertainties, showing how changing which lines you include, like C17O, shifts your calculated disk size and mass significantly >. That dependence on observational choices is something we have to keep in mind.

Jocelyn: But they also point out that their main finding on CO abundance is pretty stable because it's mainly set by the N2H plus flux. That’s a specific chemical detail about how those molecules behave in the disk environment >.

Subrahmanyan: That stability suggests that when we look at planet atmospheres later, we have to factor in this sequestration effect, meaning the carbon isn't just freely floating around; it's locked up somewhere else in that disk structure >. It’s not just a free abundance; there's a process happening.

Vera: Exactly. And they flag some limitations too because they didn't include the large gas gap observed in PDS seventy when they were building their model grid, which could introduce some big uncertainties into the final numbers >. That’s a specific thing that stops them from getting a perfectly clean result.

Jocelyn: So, to put it simply, this paper uses ALMA data to measure these molecular tracers to get constraints on disk mass and carbon monoxide sequestration, which helps us model the environments where planets actually grow >. It connects the gas we see directly to how those atmospheres form later on.

Conclusion: Vera: So we've been talking about these ALMA measurements on the PDS seventy disk, and now we're looking at what this paper is actually titled: "Observational constraints on CO sequestration and disk mass in the planet-hosting PDS seventy disk."

Jocelyn: It’s a pretty descriptive title. It tells you exactly what they’re measuring: not just the mass, but how much carbon monoxide is sequestered, which means locked up somewhere else in the dust or gas.

Subrahmanyan: That sequestration idea is key because it directly links the chemistry of those planets to their birth environment around PDS seventy.

Vera: The authors are using this data to show that these measurements aren't just numbers; they constrain a specific chemical process happening in the disk.

Jocelyn: They found that even though the peak carbon monoxide abundance is stable, you still have to be careful because it’s just the peak in one layer, not the whole thing.

Subrahmanyan: That dependence on where you are means that when we model a planet's atmosphere later, we need to account for this sequestered reservoir.

Vera: They suggest that grain surface chemistry involving carbon monoxide ice might actually help bridge the gap between what they see and what the models predict.

Jocelyn: So, it’s about linking the molecules we observe right now to how those atmospheres will evolve later on.

Subrahmanyan: The implication is that PDS seventy is a great place to study that direct link between a planet's atmosphere and its natal disk structure.

Vera: And they point to future tools like METIS and HARMONI on the ELT for further characterization of those exoplanetary atmospheres.

Jocelyn: So, the next thing we look at is how those specific molecular tracers reveal this hidden carbon chemistry that might be setting up planet formation.

M. Leemker, S. Facchini, F. Zagaria, K. Doi, J. Bae, M. Benisty, H. B., Liu, C., Longarini, V., §ampinelli, L., §tapper, L. M., Ɠeague, R.

Dipartimento di Fisica of the University of Milano · Max Planck Institute for Astronomy · Department of Astronomy of the University of Florida · Department of Physics at National Sun Yat-Sen University · Center of Astronomy and Gravitation at National Taiwan Normal University · Institute of Astronomy at the University of Cambridge · Department of Earth, Atmospheric, and Planetary Sciences at Massachusetts Institute of Technology

astro-ph.EP, astro-ph.SR

Submitted: 2026-11-10

Updated: 2026-11-10

Comments: Accepted for publication in A&A

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 88/100

The gist: Exoplanets accrete gas, dust, and ice from their natal environment, setting the chemical composition of their atmospheres and interiors The derived peak CO abundance is (4.8 + 4.3 − 2.7 − 17+8…

Key concepts

CO Sequestration
This refers to the process where carbon monoxide (CO) is locked away in the outer layers of a disk, often forming ice or complex organic molecules. The study suggests these planets might accrete the final part of their atmospheres from this sequestered CO reservoir, which is important for understanding planetary chemistry.
N2H+ Emission
N2H+ is a molecule detected in the PDS 70 disk, observed via its J=1-0 transition at 93.173 GHz. Its emission characteristics are used to constrain the disk's chemical composition and formation location when assumed to be optically thin and excited at 21 K.
Disk Mass Constraints
The mass of the PDS 70 disk was estimated by fitting line fluxes with a framework developed by Trapman et al. This method compared results against thermochemical DALI models, yielding a disk mass consistent with independent measurements from rotation curves for similar disks.

Terminology

Summary

Exoplanets accrete gas, dust, and ice from their natal environment, setting the chemical composition of their atmospheres and interiors The derived peak CO abundance is (4. p. 0

Where the paper says it

5 Center of Astronomy and Gravitation, National Taiwan Normal University, Taipei 116, Taiwan 6 Institute of Astronomy, University of Cambridge, Madingley Rd, Cambridge CB30HA, UK 7 Department of Earth, Atmospheric, and Planetary Sciences, Massachusetts Institute of Technology, Cambridge, MA 02139, USA Received XXX; accepted XXX ABSTRACT Exoplanets accrete gas, dust, and ice from their natal environment, setting the chemical composition of their atmospheres and interiors. The abundance of the major volatiles in disks can be used to possibly link exoplanet atmospheres to their formation location in the disk.

8 + 4.3 − 2.7 − 17+8 −8) × 10−6 H2 assuming that the N2H+ emission is optically thin and is characterized by an excitation temperature of 21 K The disk mass of the PDS 70 disk is 0. p. 0
Where the paper says it

Astronomy & Astrophysics manuscript no. main ©ESO 2026 November 13, 2026 Observational constraints on CO sequestration and disk mass in the planet-hosting PDS 70 disk M.

p. 6
Where the paper says it

All in all, we derive a disk mass of 0. 018 M⊙ where the range is due to the C17O flux and assuming that the N2H + is optically thin and is characterized by an excitation temperature of 21 K.

011+0.011 − 0.039+0.067 − 0.018 M⊙ (12+12 − 6 − 4) MJup assuming that the N2H+ emission is optically thin and emits at 21 K This work presents the first detection of N2H+ and C17O in the PDS 70 disk, both of which were serendipitously observed by ALMA

Discovery of Molecular Tracers

The research focuses on quantifying the peak CO abundance and disk mass of the PDS 70 disk, which is noted as the first disk with a direct confirmation of two accreting planets in its dust cavity The detection of N2H+ and C17O in this disk was achieved using ALMA observations, specifically the N2H+ J = 1 − 0 transition at 93.173 GHz and the C17O J = 3 − 2 transition at 337. p. 1

Where the paper says it

The N2H + J = 1 − 0 transition at 93. 173 GHz is covered in the Band 3 spectral setup of ALMA program 2022.

061 GHz The C17O emission shows marginal statistical evidence for an azimuthal asymmetry trailing that previously seen in the continuum emission These detections are crucial because they allow for constraints on the disk's chemical composition and formation location

Constraining Disk Properties via Modeling

The authors leverage a framework developed by Trapman et al. (2025b) to constrain the disk mass and CO abundance using these line fluxes together with archival continuum and other CO isotopologue observations The comparison is made against a grid of thermochemical DALI models, specifically utilizing the transition disk model grid where models have a dust cavity but no gas cavity The results from this method are in good agreement with independent disk mass measurements from rotation curves for the disks in the exoALMA large program

Sensitivity to Observational Choices

The derived parameters are sensitive to which fluxes are included in the fitting, as demonstrated by a comparison between fits including and excluding the C17O line flux The fit including the C17O line flux prefers a more extended disk with Rc = 58+9 − 7 au and a disk mass of Mdisk = 0.011+0.011 − 0.005 M⊙, whereas the fit excluding the C17O emission favors a smaller disk with Rc = 46+7 − 6 au and disk mass that is ∼ 3. p. 4

Where the paper says it

005 M⊙, where the uncertainties are the 16th and 84th percentiles of the posterior distribution and do not include any systematic uncertainties. In contrast, the fit excluding the C 17O emission favors a smaller disk with Rc = 46+7 −6 au and disk mass that is ∼ 3.

4× lower The peak CO abundance is more or less independent of whether the C17O line flux is included in the fitting because the CO depletion is mainly set by the N2H+ flux

Systematic Uncertainties and Future Directions

The large gas gap observed in the PDS 70 disk, which is not included in the adopted model grid, may introduce substantial systematic uncertainties in the inferred disk parameters that are not captured by the formal posterior distributions A sensitivity test showed that doubling the integrated C17O flux almost exactly matches the posteriors when C17O is not included in the fitting, highlighting that derived disk mass, size and gas-to-dust mass ratio are sensitive to a factor of 2 change in the C17O flux A more flexible model grid including a range of gas cavity sizes and depletion factors is therefore needed to quantify these systematic effects The CO sequestering in the outer disk gas probed by the observations suggests that the PDS 70 planets may accrete the final portion of their atmospheres from this CO sequestered reservoir

Conclusion on Planetary Chemistry

The findings suggest that CO sequestration may need to be considered when interpreting the chemical composition of planetary atmospheres The system is a good candidate to explore a direct link between the atmospheric composition of the planets and their natal environment Including grain surface chemistry involving CO ice can reconcile observed fluxes with model predictions, suggesting that at least part of the sequestered CO is locked in the form of e.g., CO2 ice and icy complex organic molecules METIS and HARMONI on the ELT can further improve our understanding by characterizing exoplanetary atmospheres The PDS 70 disk mass is similar to that of Herbig disks and massive T Tauri disks The derived peak CO abundance is similar to what is found in other T Tauri disks and is approximately an order of magnitude lower than that in the ISM after accounting for CO freeze-out and isotope selective photodissociation of CO

Final Summary

In summary, the effective peak CO abundance relative to H2 in the PDS 70 disk layers probed by the observations is (4.8 + 4.3 − 2.7 − 17+8 −8) × 10−6 The disk mass of the PDS 70 disk is 0. p. 0

Where the paper says it

Astronomy & Astrophysics manuscript no. main ©ESO 2026 November 13, 2026 Observational constraints on CO sequestration and disk mass in the planet-hosting PDS 70 disk M.

011+0.011 − 0.039+0.067 − 0.018 M⊙ assuming that the N2H+ emission is optically thin and emits at 21 K The range of a factor of ∼ 3. p. 8
Where the paper says it

018 M⊙ (12+12 −6 − 41+70 −19 MJup) assuming that the N2H + emission is optically thin and emits at 21 K. The range of a factor of ∼ 3.

p. 8
Where the paper says it

018 M⊙ (12+12 −6 − 41+70 −19 MJup) assuming that the N2H + emission is optically thin and emits at 21 K. The range of a factor of ∼ 3.

4 in the derived disk mass is driven by the dependence on whether the C17O 3 − 2 line flux is included in the mcmc sampling procedure The range is due to the optical depth and excitation temperature of the N2H+ emission and to a lesser degree to whether the C17O flux is included in the fitting The peak CO abundance is independent of this The derived peak CO abundance is not necessarily the global CO/H2 ratio in the disk but rather the peak CO abundance in the emitting layer probed by the observations

Key Findings

  1. The effective peak CO abundance relative to H2 in the PDS 70 disk layers probed by the observations is (4. p. 8

    Where the paper says it

    However, observations of a second N2H + transition at sufficiently high spectral resolution to resolve the hyperfine components are needed to constrain the optical depth and excitation temperature of the N2H + emission, thereby further narrowing the currently allowed effective peak CO abundance of (4. 7 − 17+8 −8 ) × 10−6 relative to H2 in the layers probed by the PDS 70 disk observations and under the assumption of the adopted chemical network and ionization prescription.

    8 + 4.3 − 2.7 − 17+8 −8) × 10−6

  2. The disk mass of the PDS 70 disk is 0.011+0.011 − 0.039+0.067 − 0.018 M⊙ (12+12 − 6 − 4) MJup assuming that the N2H+ emission is optically thin and emits at 21 K

  3. The range of a factor of ∼ 3.4 in the derived disk mass is driven by the dependence on whether the C17O 3 − 2 line flux is included in the mcmc sampling procedure

  4. As CO is sequestered in the outer disk gas probed by the observations, the PDS 70 planets may accrete the final portion of their atmospheres from this CO sequestered reservoir

  5. The large gas gap observed in the PDS 70 disk is not included in the adopted model grid and may introduce substantial systematic uncertainties in the inferred disk parameters that are not captured by the formal posterior distributions

  6. We find tentative (∼ 3σ) evidence of an azimuthal asymmetry in the C17O emission that is trailing the asymmetry in the sub-mm continuum emission by ∼ 40 − 50◦

References

Anderson, D. E., Cleeves, L. p. 9

Where the paper says it

References Anderson, D. , Cleeves, L.

p. 9
Where the paper says it

References Anderson, D. , Cleeves, L.

p. 9
Where the paper says it

References Anderson, D. , Cleeves, L.

I., Blake, G. p. 9
Where the paper says it

, Cleeves, L. , Blake, G.

p. 9
Where the paper says it

, Cleeves, L. , Blake, G.

A., et al. 2022, ApJ, 927, 229

Andreu, A., Coutens, A. p. 9

Where the paper says it

2022, ApJ, 927, 229 Andreu, A. , Coutens, A.

, Cruz-Sáenz de Miera, F. p. 9
Where the paper says it

, Coutens, A. , Cruz-Sáenz de Miera, F.

, et al. 2023, A&A, 677, L17

Ansdell, M., Williams, J. p. 9

Where the paper says it

2023, A&A, 677, L17 Ansdell, M. , Williams, J.

P., van der Marel, N. p. 9
Where the paper says it

, Williams, J. , van der Marel, N.

, et al. 2016, ApJ, 828, 46

Baruteau, C. & Zhu, Z. 2016, MNRAS, 458, 3927

Benisty, M., Bae, J., Facchini, S. p. 0

Where the paper says it

2018; Bosman & Bergin 2021; Leemker et al. 2021; Facchini et al.

p. 0
Where the paper says it

2018; Bosman & Bergin 2021; Leemker et al. 2021; Facchini et al.

p. 0
Where the paper says it

2018; Bosman & Bergin 2021; Leemker et al. 2021; Facchini et al.

p. 0
Where the paper says it

2018; Bosman & Bergin 2021; Leemker et al. 2021; Facchini et al.

p. 0
Where the paper says it

2018; Bosman & Bergin 2021; Leemker et al. 2021; Facchini et al.

p. 0
Where the paper says it

2018; Bosman & Bergin 2021; Leemker et al. 2021; Facchini et al.

, et al. 2021, ApJ, 916, L2

Bergin, E. A., Booth, R. p. 0

Where the paper says it

2011, 2023). The bulk of the ice reservoir is likely inherited from the earliest phases of star and planet formation (Booth et al.

A., Colmenares, M. p. 9
Where the paper says it

, Booth, R. , Colmenares, M.

J., & Ilee, J. p. 9
Where the paper says it

, Colmenares, M. , & Ilee, J.

D. 2024, ApJ, 969, L21

Bergin, E. A., Cleeves, L. I., Gorti, U. p. 9

Where the paper says it

, Cleeves, L. , Gorti, U.

p. 9
Where the paper says it

, Cleeves, L. , Gorti, U.

, et al. 2013, Nature, 493, 644

Bergin, E. A., Du, F., Cleeves, L. I., et al. 2016, ApJ, 831, 101

Boehler, Y., Ménard, F. p. 9

Where the paper says it

2016, ApJ, 831, 101 Boehler, Y. , Ménard, F.

, Robert, C. p. 6
Where the paper says it

In addition to evidence for radial and azimuthal trapping of dust through continuum observations, vortices could also leave an imprint in the gas kinematics (e. , Robert et al.

M. T., et al. 2021, A&A, 650, A59

Booth, A. S., Walsh, C. p. 0

Where the paper says it

2017; Cridland et al. 2019; Walsh 2025).

p. 0
Where the paper says it

2017; Cridland et al. 2019; Walsh 2025).

, Terwisscha van Scheltinga, J. p. 9
Where the paper says it

, Walsh, C. , Terwisscha van Scheltinga, J.

, et al. 2021, Nature Astronomy, 5, 684

Bosman, A. D., Alarcón, F. p. 9

Where the paper says it

2021, Nature Astronomy, 5, 684 Bosman, A. , Alarcón, F.

, Bergin, E. p. 0
Where the paper says it

2015; Carr et al. 2018; Bosman & Bergin 2021; Leemker et al.

A., et al. 2021, ApJS, 257, 7

Bosman, A. D. & Bergin, E. p. 0

Where the paper says it

2015; Carr et al. 2018; Bosman & Bergin 2021; Leemker et al.

A. 2021, ApJ, 918, L10

Bosman, A. D., Walsh, C., & van Dishoeck, E. p. 9

Where the paper says it

, Walsh, C. , & van Dishoeck, E.

p. 9
Where the paper says it

, Walsh, C. , & van Dishoeck, E.

F. 2018, A&A, 618, A182

Bridle, A. H. & Schwab, F. p. 1

Where the paper says it

A summary of the observations is presented in Table A. Before combining the Execution Blocks (EBs) of the Band 3 data, a pseudocontinuum of all four spectral windows in each EB was created by flagging the data within 15 km s−1 of the bright lines and the N2H + 1 − 0 transition and averaging over a 30 MHz bandwidth to avoid bandwidth smearing (Bridle & Schwab 1989).

R. 1989

Bruderer, S., van Dishoeck, E. p. 9

Where the paper says it

, Walsh, C. , & van Dishoeck, E.

p. 9
Where the paper says it

, Walsh, C. , & van Dishoeck, E.

p. 9
Where the paper says it

, Walsh, C. , & van Dishoeck, E.

p. 9
Where the paper says it

, Walsh, C. , & van Dishoeck, E.

p. 9
Where the paper says it

, Walsh, C. , & van Dishoeck, E.

F., Doty, S. D., & Herczeg, G. p. 9
Where the paper says it

, van Dishoeck, E. , & Herczeg, G.

J. 2012, A&A, 541, A91

Carr, J. S., Najita, J. p. 9

Where the paper says it

2012, A&A, 541, A91 Carr, J. , Najita, J.

R., & Salyk, C. p. 9
Where the paper says it

, Najita, J. , & Salyk, C.

2018

Casassus, S. & Cárcamo, M. p. 2

Where the paper says it

The beam is indicated in the bottom left corner of each panel and a 20 au scalebar is shown in the bottom right corner of the last panel. are measured using the non-JvM corrected cubes (Casassus & Cárcamo 2022).

2022

Cazzoletti, P., van Dishoeck, E. F., Pinilla, P. p. 9

Where the paper says it

, van Dishoeck, E. , Pinilla, P.

, et al. 2018, A&A, 619, A161

Cazzoli, G., Cludi, L. p. 9

Where the paper says it

2018, A&A, 619, A161 Cazzoli, G. , Cludi, L.

, Buffa, G. p. 9
Where the paper says it

, Cludi, L. , Buffa, G.

, & Puzzarini, C. p. 9
Where the paper says it

, Buffa, G. , & Puzzarini, C.

2012

Cortes, P., Carpenter, J. p. 9

Where the paper says it

2012, ApJS, 203, 11 Cortes, P. , Carpenter, J.

, Kameno, S. p. 9
Where the paper says it

, Carpenter, J. , Kameno, S.

, et al. 2025

Cridland, A. J., Facchini, S., van Dishoeck, E. F., & Benisty, M. p. 2

Where the paper says it

Isella et al. Benisty et al.

2023

Cridland, A. J., van Dishoeck, E. F., Alessi, M. p. 9

Where the paper says it

, van Dishoeck, E. , Alessi, M.

, & Pudritz, R. p. 9
Where the paper says it

, Alessi, M. , & Pudritz, R.

E. 2019

Czekala, I., Loomis, R. p. 1

Where the paper says it

5 down to a 4σ threshold. The relative alignment of these EBs was determined following the procedure outlined in Loomis et al.

A., Teague, R. p. 1
Where the paper says it

The results from the recent ALMA large programs AGEPRO (Zhang et al. 2025) and exoALMA (Teague et al.

p. 1
Where the paper says it

The results from the recent ALMA large programs AGEPRO (Zhang et al. 2025) and exoALMA (Teague et al.

p. 1
Where the paper says it

The results from the recent ALMA large programs AGEPRO (Zhang et al. 2025) and exoALMA (Teague et al.

p. 1
Where the paper says it

The results from the recent ALMA large programs AGEPRO (Zhang et al. 2025) and exoALMA (Teague et al.

, et al. 2021

Deng, D., Gorti, U., Pascucci, I. p. 8

Where the paper says it

, CO2 ice and icy complex organic molecules (Ruaud et al. 2022; Pascucci et al.

p. 8
Where the paper says it

, CO2 ice and icy complex organic molecules (Ruaud et al. 2022; Pascucci et al.

, & Ruaud, M. p. 8
Where the paper says it

Including grain surface chemistry involving CO ice can reconcile the observed CO isotopologue fluxes with model predictions, suggesting that at least part of the sequestered CO is locked in the form of e. , CO2 ice and icy complex organic molecules (Ruaud et al.

2025

Dickman, R. L. 1978

Doi, K., Kataoka, A. p. 9

Where the paper says it

1978, ApJS, 37, 407 Doi, K. , Kataoka, A.

, Liu, H. B., et al. 2024

Endres, C. P., Schlemmer, S. p. 9

Where the paper says it

2024, ApJ, 974, L25 Endres, C. , Schlemmer, S.

, Schilke, P. p. 9
Where the paper says it

, Schlemmer, S. , Schilke, P.

, Stutzki, J. p. 9
Where the paper says it

, Schilke, P. , Stutzki, J.

, & Müller, H. p. 3
Where the paper says it

5 N2H + 3 − 2 279. Line constants are taken from the CDMS database (Müller et al.

S. P. 2016

Facchini, S., Teague, R., Bae, J., et al. 2021

Facchini, S., Testi, L. p. 9

Where the paper says it

2021, AJ, 162, 99 Facchini, S. , Testi, L.

, Humphreys, E. p. 9
Where the paper says it

, Testi, L. , Humphreys, E.

, et al. 2024

Fasano, D., Benisty, M. p. 2

Where the paper says it

Isella et al. Benisty et al.

p. 2
Where the paper says it

Isella et al. Benisty et al.

p. 2
Where the paper says it

Isella et al. Benisty et al.

p. 2
Where the paper says it

Isella et al. Benisty et al.

p. 2
Where the paper says it

Isella et al. Benisty et al.

, Curone, P. p. 9
Where the paper says it

, Benisty, M. , Curone, P.

p. 9
Where the paper says it

, Benisty, M. , Curone, P.

, et al. 2025

Foreman-Mackey, D., Hogg, D. W., Lang, D., & Goodman, J. p. 9

Where the paper says it

2025, A&A, 699, A373 Foreman-Mackey, D. , & Goodman, J.

2013

Francis, L., Johnstone, D. p. 9

Where the paper says it

2013, PASP, 125, 306 Francis, L. , Johnstone, D.

, Herczeg, G. p. 9
Where the paper says it

, van Dishoeck, E. , & Herczeg, G.

p. 9
Where the paper says it

, van Dishoeck, E. , & Herczeg, G.

, Hunter, T. p. 9
Where the paper says it

, Herczeg, G. , Hunter, T.

R., & Harsono, D. p. 0
Where the paper says it

2025; Slavicinska et al. These ices are oxygen rich due to the high desorption temperature of water that only desorbs in the innermost regions of disks (Harsono et al.

2020

Gaia Collaboration, Brown, A. G. A., Vallenari, A. p. 9

Where the paper says it

2020, AJ, 160, 270 Gaia Collaboration, Brown, A. , Vallenari, A.

, et al. 2018

Goldsmith, P. F. & Langer, W. p. 3

Where the paper says it

Similarly, the N2H + J = 1 − 0 line covered in the ALMA Band 3 data is not included in the DALI model grid that only includes the J = 3 − 2 and J = 4 − 3 transitions. Therefore, the line flux of the J = 1 − 0 transition is scaled to that of the J = 3 − 2 transition assuming local thermodynamical equilibrium, optically thin emission, and using the line constants from the CDMS database that does not separate between the hyperfine components (Goldsmith & Langer 1999; Müller et al.

D. 1999

Haffert, S. Y., Bohn, A. J., de Boer, J. p. 9

Where the paper says it

1999, ApJ, 517, 209 Haffert, S. , de Boer, J.

, et al. 2019

Harsono, D., Bruderer, S. p. 3

Where the paper says it

A&A proofs: manuscript no. main N2H + , and the continuum are compared to those in a large grid of thermochemical DALI models (Bruderer et al.

p. 3
Where the paper says it

A&A proofs: manuscript no. main N2H + , and the continuum are compared to those in a large grid of thermochemical DALI models (Bruderer et al.

, & van Dishoeck, E. F. 2015

Hsu, C.-C., Wang, J. J., Blake, G. A., et al. 2024

Isella, A., Benisty, M., Teague, R., et al. 2019

Jiang, S. D., Huang, J. p. 9

Where the paper says it

2019, ApJ, 879, L25 Jiang, S. , Huang, J.

, Czekala, I. p. 1
Where the paper says it

Similar to the continuum, the lines were cleaned down to a 1σ threshold. In addition, the line cubes are JvM-corrected to accurately measure the flux by accounting for the difference between the non-Gaussian shape of the beam in the dirty image and the Gaussian clean beam (Jorsater & van Moorsel 1995; Czekala et al.

, et al. 2026

Jorsater, S. & van Moorsel, G. p. 1

Where the paper says it

Similar to the continuum, the lines were cleaned down to a 1σ threshold. In addition, the line cubes are JvM-corrected to accurately measure the flux by accounting for the difference between the non-Gaussian shape of the beam in the dirty image and the Gaussian clean beam (Jorsater & van Moorsel 1995; Czekala et al.

A. 1995

Kama, M., Bruderer, S., van Dishoeck, E. F., et al. 2016

Keppler, M., Benisty, M., Müller, A. p. 3

Where the paper says it

5 N2H + 3 − 2 279. Line constants are taken from the CDMS database (Müller et al.

, et al. 2018

Keppler, M., Teague, R., Bae, J., et al. 2019

Krijt, S., Bosman, A. p. 0

Where the paper says it

2015; Carr et al. 2018; Bosman & Bergin 2021; Leemker et al.

D., Zhang, K. p. 0
Where the paper says it

2017; Miotello et al. 2017; Zhang et al.

, et al. 2020

Law, C. J., Benisty, M., Facchini, S., et al. 2024

Leemker, M., Facchini, S., Curone, P., et al. 2026

Leemker, M., Tobin, J. p. 0

Where the paper says it

2021; Andreu et al. 2023; Tobin et al.

J., Facchini, S., et al. 2025

Liu, H. B., Casassus, S. p. 2

Where the paper says it

The beam is indicated in the bottom left corner of each panel and a 20 au scalebar is shown in the bottom right corner of the last panel. are measured using the non-JvM corrected cubes (Casassus & Cárcamo 2022).

p. 2
Where the paper says it

The beam is indicated in the bottom left corner of each panel and a 20 au scalebar is shown in the bottom right corner of the last panel. are measured using the non-JvM corrected cubes (Casassus & Cárcamo 2022).

, Dong, R., et al. 2024

Liu, H. B., Doi, K., Casassus, S., et al. 2026

Long, F., Herczeg, G. J., Pascucci, I., et al. 2017

Long, Z. C., Akiyama, E. p. 9

Where the paper says it

2017, ApJ, 844, 99 Long, Z. , Akiyama, E.

, Sitko, M. p. 9
Where the paper says it

, Akiyama, E. , Sitko, M.

, et al. 2018

Longarini, C., Lodato, G. p. 9

Where the paper says it

2018, ApJ, 858, 112 Longarini, C. , Lodato, G.

, Rosotti, G. p. 9
Where the paper says it

, Lodato, G. , Rosotti, G.

, et al. 2025

Loomis, R. A., Facchini, S., Benisty, M., et al. 2025

Lovelace, R. V. E., Li, H., Colgate, S. p. 9

Where the paper says it

2025, ApJ, 984, L7 Lovelace, R. , Colgate, S.

A., & Nelson, A. p. 9
Where the paper says it

, Colgate, S. , & Nelson, A.

F. 1999

Ma, X., Yu, F., Dong, R., et al. 2026

Appendix Details

The analysis involved running the mcmc with 50 walkers for 3000 steps where the first 600 steps are used for the burn-in phase The integrated autocorrelation time corresponding to the unconstrained parameters (dust structure, vertical structure, and cosmic ray ionization rate) and to the stellar luminosity (characterized by an approximately Gaussian posterior) flattens off after 200-500 steps The resulting distribution of all 1000 posteriors is presented in Fig. F.1 where the blue histograms are identical to those in Fig. p. 6

Where the paper says it

5: Distribution of the posteriors for derived peak CO abundance, disk mass, Rc, and gas-to-dust mass ratio for 1000 mcmc runs with input fluxes (including the C17O flux) and R90%( 13CO) values sampled from the measured fluxes (grey). The orange histograms are identical to those in Fig.

p. 6
Where the paper says it

5: Distribution of the posteriors for derived peak CO abundance, disk mass, Rc, and gas-to-dust mass ratio for 1000 mcmc runs with input fluxes (including the C17O flux) and R90%( 13CO) values sampled from the measured fluxes (grey). The orange histograms are identical to those in Fig.

3 and show the result of a single mcmc fit using the measured values and their associated uncertainties The uncertainty on the derived parameters is not dominated by the uncertainty on the measured fluxes but rather by the underlying modelling grid

Final Results Summary

The resulting distribution of all 1000 posteriors is presented in Fig. F.1 where the blue histograms are identical to those in Fig. 3 and show the result of a single mcmc fit using the measured values and their associated uncertainties The derived disk mass is 0. p. 6

Where the paper says it

The orange histograms are identical to those in Fig. 3 and show the result of a single mcmc fit using the measured values and their associated uncertainties.

011+0.011 − 0.006 − 0.039+0.067 − 0.018 M⊙ where the range is due to the C17O flux and assuming that the N2H+ is optically thin and is characterized by an excitation temperature of 21 K The peak CO abundance in this case is (4. p. 6
Where the paper says it

018 M⊙ where the range is due to the C17O flux and assuming that the N2H + is optically thin and is characterized by an excitation temperature of 21 K. The peak CO abundance in this case is (4.

p. 6
Where the paper says it

018 M⊙ where the range is due to the C17O flux and assuming that the N2H + is optically thin and is characterized by an excitation temperature of 21 K. The peak CO abundance in this case is (4.

8 + 4.3 − 2.7 − 5.8 + 4.7 − 3.2)×10−6 with respect to H2 The derived peak CO abundance is not necessarily the global CO/H2 ratio in the disk but rather the peak CO abundance in the emitting layer probed by the observations

Key Data Points

The uncertainty on the measured line and continuum fluxes is not dominated by the uncertainty on the measured fluxes but rather by the underlying modelling grid The derived disk mass is 0.011+0.011 − 0.039+0.067 − 0.018 M⊙ where the range is due to the C17O flux and assuming that the N2H+ is optically thin and is characterized by an excitation temperature of 21 K The peak CO abundance in this case is (4.8 + 4.3 − 2.7 − 5.8 + 4.7 − 3.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improved AI can perform systematic uncertainty analysis by incorporating results from bootstrapping, which shows the uncertainty on the derived parameters is not dominated by the uncertainty on the measured fluxes but rather by the underlying modelling grid.

  2. Improved AI can better constrain disk mass and peak CO abundance by using a more flexible model grid that includes a range of gas cavity sizes and depletion factors to quantify systematic effects.

  3. Improved AI can perform parameter inference that accounts for degeneracies by including the 90 percent flux radius of the 13CO J = 2 - 1 emission in the fitting procedure, as this can help to break some of the degeneracies in the fitting.

  4. Improved AI can distinguish between different physical scenarios by analyzing how model parameters change based on whether specific lines are included in the fitting, such as how the fit including the C17O line flux prefers a more extended disk with Rc = 58+9−7 au and a disk mass of Mdisk = 0.011+0.011−0.005 M⊙.

  5. Improved AI can identify potential physical phenomena by looking for asymmetries in emission, such as the tentative (∼ 3σ) evidence of an azimuthal asymmetry in the C17O emission that is trailing the asymmetry in the sub-mm continuum emission by ∼ 40 − 50◦, which could hint at a vortex.

Abstract

Exoplanets accrete gas, dust, and ice from their natal environment, setting the chemical composition of their atmospheres and interiors. The abundance of the major volatiles in disks can be used to possibly link exoplanet atmospheres to their formation location in the disk. In this work, we aim to quantify the peak CO abundance and disk mass of the PDS 70 disk, the first disk with a direct confirmation of two accreting planets in its dust cavity. We present the first detection of N 2 H+ and C 17 O in this disk. The C 17 O emission shows marginal statistical evidence for an azimuthal asymmetry trailing that previously seen in the continuum emission. We then compare the disk integrated fluxes of N 2 H+, CO isotopologues, and the continuum to an existing grid of thermochemical models. We find that the derived disk mass drops by a factor of about 3.4 from 0.039+0.067-0.018 M to 0.011+0.011-0.006 M if the C 17 O line flux is included in the fitting procedure; the quoted intervals do not include systematic uncertainties. CO on the other hand is sequestered with an effective abundance of (4.8+4.3-2.7 - 17+8-8) times 10-6 relative to H 2 in the probed disk layers after accounting for CO freeze-out and isotope selective CO photodissociation for the adopted chemical network and ionization prescription. The range in peak CO abundances reflects the uncertainty in the N 2 H+ emission, assuming either optically thin emission at 21 K or optically thick emission. Therefore, the PDS 70 planets are possibly accreting the final portion of their atmospheres from the outer disk molecular layer where CO is sequestered. Thus, CO sequestering may need to be considered when interpreting the chemical composition of planetary atmospheres.

Sources

Related papers