Chemistry in externally FUV irradiated disks in the outskirts of the Orion Nebula

Most stars are born in stellar clusters and their protoplanetary disks, which are the birthplaces of planets, can therefore be affected by the radiation of nearby massive stars. However, little is known about the chemistry of externally irradiated disks, including whether or not their properties are similar to the so-far better-studied isolated disks. Motivated by this question, we present ALMA Band 6 observations of two irradiated Class II protoplanetary disks in the outskirts of the Orion Nebula Cluster (ONC) to explore the chemical composition of disks exposed to (external) FUV radiation fields: the 216-0939 disk and the binary system 253-1536A/B, which are exposed to radiation fields of $10^2-10^3$ times the average interstellar radiation field. We detect lines from CO isotopologues, HCN, H$_2$CO, and C$_2$H toward both protoplanetary disks. Based on the observed disk-integrated line fluxes and flux ratios, we do not find significant differences between isolated and irradiated disks. The observed differences seem to be more closely related to the different stellar masses than to the external radiation field. This suggests that these disks are far enough away from the massive Trapezium stars, that their chemistry is no longer affected by external FUV radiation. Additional observations towards lower-mass disks and disks closer to the massive Trapezium stars are required to elucidate the level of external radiation required to make an impact on the chemistry of planet formation in different kinds of disks.


INTRODUCTION
Protoplanetary disks of gas and dust around young stars are the birthplace of planets; hence, understanding the physical and chemical structures of disks is essential to determine the composition of the material that can be incorporated into planets and planetesimals.
During the last decades, the chemical structure and distribution of numerous molecules have been widely studied in protoplanetary disks (e.g.Dutrey et al. 2007;Pegues et al. 2021;Öberg et al. 2021).A variety of molecules have been observed and detected using the Atacama Large Millimeter/submillimeter Array (ALMA), the Submillimeter Array (SMA), the Northern Extended Millimeter Array (NOEMA), and the Institute for Radio Astronomy in the Millimeter Range (IRAM) 30m tele-scope; examples are CO (and isotopologues; e.g., Koerner et al. 1993;Dutrey et al. 1997;Thi et al. 2001;Booth et al. 2019), small organics (C 2 H, CS, CN, HCN, HNC, H 2 CO, HCO + , DCO + ; e.g., Dutrey et al. 1997;van Dishoeck et al. 2003;Qi et al. 2008;Öberg et al. 2011;Guzmán et al. 2015;Hily-Blant et al. 2017;Furuya et al. 2022), and even complex species (CH 3 OH, CH 3 CN, HC 3 N, c−C 3 H 2 , HCOOH, and CH 3 OCH 3 ; e.g., Chapillon et al. 2012;Qi et al. 2013;Öberg et al. 2015;Walsh et al. 2016;Favre et al. 2018;Brunken et al. 2022).These studies have advanced our understanding of the chemistry of planet formation; however, they have focused on disks around isolated stars in low-mass star-formation regions, whereas many stars form in clusters, with low-mass and high-mass stars forming in con-junction (Lada & Lada 2003;Krumholz et al. 2019).Thus, the chemistry of externally irradiated protoplanetary disks remains poorly constrained observationally.Following the nomenclature also used in Walsh et al. (2013), "isolated" refers to a disk irradiated by its central star only, whereas "irradiated" refers to a disk illuminated by the central star and the interstellar radiation field (ISRF), which includes at least one nearby massive star.
Most studies of extremely irradiated protoplanetary disks have been focused on the disk mass (e.g., Eisner et al. 2018;Mann et al. 2014;Boyden & Eisner 2020;Boyden & Eisner 2023), radius (e.g., Vicente & Alves 2005;Clarke 2007;Mann et al. 2014;Boyden & Eisner 2020;Boyden & Eisner 2023), evolution (e.g., Champion et al. 2017;Haworth et al. 2017), and lifetime (e.g., Adams et al. 2004;Haworth et al. 2021;Winter et al. 2019).Theoretical models and observational studies have predicted and discovered, respectively, that disks exposed to external radiation decrease in their mass and size when they are close to the massive stars, due to photoevaporation (e.g., Johnstone et al. 1998;Störzer & Hollenbach 1999;Matsuyama et al. 2003;Adams et al. 2004;Mann et al. 2014;Boyden & Eisner 2020;Concha-Ramírez et al. 2023;Boyden & Eisner 2023).This is expected to affect the ability of these disks to form planets.However, photoevaporation is expected to mainly remove gas from the outer disk.This is because dust can grow in size and the subsequent settling/drift lead to a depletion of dust in the surface and outer regions of the disk where external photoevaporation is mainly operating (Facchini et al. 2016).Therefore, rocky planets may still be able to form in the inner disk midplane (Adams et al. 2004;Concha-Ramírez et al. 2023).
We currently do not know the impact of an external radiation field on the chemistry of a disk.This is important to investigate because isolated disks are not the norm in the Galaxy.Instead, the vast majority of stars form within rich stellar clusters where disks are constantly affected by intense radiation fields from nearby massive stars, and exposed to extremeultraviolet (EUV) radiation and far-ultraviolet (FUV) radiation (Adams et al. 2004), that affects their evolution through time.Our Sun is thought to have been born in such a cluster, so the protosolar nebula would have been irradiated by its massive neighbors (Lada & Lada 2003;Adams 2010;Parker 2020).Hence, understanding the initial chemical conditions of the protosolar nebula requires an understanding of the chemistry in disks located in massive star-forming regions.
Observing molecular lines in disks that are close to massive stars is challenging, because of their physical size (R disk ∼ 10 2 − 10 3 au; Mann et al. 2014;Boyden & Eisner 2023) and the large distances to massive starformation regions, meaning that they typically span only ∼ 0. ′′ 2 − 1. ′′ 2 on the sky (at the distances of the nearest massive star forming regions).Additionally, many disks in clustered regions are still embedded in the parent molecular cloud, and it can be challenging to disentangle the disk molecular line emission from the cloud molecular line emission.
A few astrochemical disk models have investigated the effect of strong external FUV fields, characteristic of the environments close to O/B stars, on disk chemistry.These models predict that external FUV sources will significantly impact the thermal and chemical structure of the disk (Nguyen et al. 2002;Walsh et al. 2013Walsh et al. , 2014)).Therefore, the conclusions reached for isolated disks may not be directly transferred to disks born near massive stars.For example, the gas temperature is expected to be significantly higher in the outer disk midplane in externally irradiated disks, which will result in the release of molecules that would usually be frozen out onto dust grains (Walsh et al. 2013).However, these models remain speculative and their predictions are yet to be tested through comparisons with observations.
In this work, we present observations of two protoplanetary disks around pre-main sequence stars located in the outskirts of the Orion Nebula Cluster (ONC), which is the closest (∼ 414 pc) massive star-forming region (Menten et al. 2007;Rzaev et al. 2021).Our aim is to investigate the possible differences between the chemistry of isolated disks and of externally irradiated disks.By studying disks with similar initial conditions to the protosolar nebula, such as externally irradiated disks, we can learn about the formation conditions of planets in the Solar System.The ONC is an excellent laboratory to study these kinds of sources.This rich cluster has thousands of stars, but the radiation field is dominated by a single star: θ 1 Ori C (O'dell & Wen 1994;Smith et al. 2005;Ricci et al. 2008), a young (∼ 1 Myr; Hillenbrand 1997), massive (45 M ⊙ ; Kraus et al. 2009;Rzaev et al. 2021), O6-type star (O'Dell et al. 2017).Hubble Space Telescope (HST) observations revealed hundreds of disks in Orion that are externally irradiated, some of them surrounded by a cometary ionization frontthe so-called proplyds (O'dell et al. 1993;O'dell & Wen 1994;Ricci et al. 2008;Eisner et al. 2018).More recent ALMA surveys have shown that disks close to the stellar cluster (≲ 0.5 pc) can have a significant amount of their disk mass removed due to external photoevaporation, as was predicted by Johnstone et al. (1998) and Störzer & Hollenbach (1999), while disks that are farther away have disk masses that are relatively intact (Mann et al.
2014; Eisner et al. 2018;van Terwisga et al. 2019).However, the chemistry of disks in these less exposed regions could be affected because the FUV radiation, even if it is not as extreme as in the EUV regime, is still higher than the FUV field to which the isolated disks are exposed.Section 2 describes the ALMA observations, and provides information on the sources, including the disk and stellar properties, and the spectroscopic parameters of the molecular lines.Section 3 presents the results of the continuum emission and molecular lines.In Section 4 we compare the results of the externally irradiated disks with those for isolated disks.Finally, Section 5 presents our conclusions.

OBSERVATIONS
We have targeted two protoplanetary disk systems located in the Orion Nebula Cluster at a distance of 414 pc (e.g., Rzaev et al. 2021).The two systems, 216-0939 and 253-1536A/B, are located in the outskirts of the ONC (projected distance of > 0.9 pc from θ 1 Ori C), where the radiation field is estimated to be < 10 3 G 0 (Mookerjea et al. 2003), where G 0 is the average strength of the integrated ISRF; G 0 ≈ 1.6 × 10 −3 erg cm −2 s −1 (Draine 1978).The sources have been previously identified and classified with HST (e.g., Smith et al. 2005;Ricci et al. 2008Ricci et al. , 2011a)), and have been observed with ALMA as part of a survey that observed the continuum emission in 22 disks near the ONC (e.g., Mann et al. 2014).The three disks have estimated gas masses of the same order of magnitude, ranging between ∼ 30 − 80 M Jup .The two systems are exposed to different external radiation fields (by almost an order of magnitude), which allows us to investigate for the first time the effect of different external radiation fields on the chemistry.The stellar and disk propeties are summarized in Table 1.This section presents the sources, observational details, and data reduction process.
Source 216-0939 is a K5 star, determined spectroscopically by Hillenbrand (1997).However, the dynamical mass of 2.17 ± 0.07 M ⊙ estimated by Factor et al. (2017) is inconsistent with this spectral type.They propose that the star could instead be a tight equal-mass binary of two 1.1 M ⊙ stars.For the purposes of this study we consider both cases.The disk is located at a projected distance of 1.59 pc from θ 1 Ori C (O6 star; Hillenbrand 1997;Factor et al. 2017), and at 0.8 pc from ν Ori (B3V star; Terada & Tokunaga 2012).The disk around 216-0939 (J2000 R.A. = 05h35m21.57s;J2000 DEC.= −05d 09m 38.9s) has an inclination of 32 • , a position angle (PA) of 173 • , a systemic velocity (v LSR ) of 10.75 km/s (Factor et al. 2017), and a dust disk size of ∼ 290 au (Mann & Williams 2009).This source is one of the most massive protoplanetary disks in the ONC.The gas mass of the disk is estimated to be ∼ 46 M Jup (∼ 0.04 M ⊙ ; Ricci et al. 2011b;Mann et al. 2014;Mann & Williams 2009;Factor et al. 2017), computed from the 856µm dust continuum emission.Additionally, because of its distance to the massive stars, the external FUV radiation field is estimated to be G 0 ≈ 180, based on Herschel FIR observations (Pabst et al. 2021).Previous studies by Mann & Williams (2009, 2010) and Factor et al. (2017) reported the detection of HCO + (4 − 3), CO (3 − 2), and HCN (4 − 3) lines and a weak detection of CS (7 − 6) with an rms noise of 0.41 mJy beam −1 and a synthesized beam of 0. ′′ 57 × 0. ′′ 52, and provided the first view of the molecular content and structure of the disk.
Additionally, HST observations of the 216-0939 star revealed the presence of water ice absorption, concluding that it most likely originated from the surrounding disk (Terada & Tokunaga 2012).The water ice detection is associated with a large silhouette disk, about ∼ 1000 au in diameter and showed that there are regions of the disk that are sufficiently cold to host substantial ice reservoir (Terada & Tokunaga 2012;Terada et al. 2012).
Source 253-1536A/B is a wide binary system with dynamically estimated stellar masses of ∼ 3.5 M ⊙ and ≳ 0.2 M ⊙ / sin 2 i, for the A and B members, respectively (Williams et al. 2014).The lower-mass star, 253-1536B, has a spectral type M2, while a spectral type of F/G has been reported for 253-1536A by Ricci et al. (2011b).The disk of 253-1536A has an inclination of 65 • , and a position angle (PA) of 69.7 • (Williams et al. 2014).The inclination is unknown for 253-1536B, but the position angle has been estimated to be 136 • (Williams et al. 2014).The systemic velocity of the system (v LSR ) is 10.55 and 10.85 km/s for the A/B members (Williams et al. 2014).This system, an ONC proplyd (Smith et al. 2005), is located inside the M43 HII region and is associated with a bright proplyd ionization front (Williams et al. 2014) and a bipolar jet (Smith et al. 2005).This binary system is at a projected distance of 0.92 pc from θ 1 Ori (J2000 R.A. = 05h35m25.30s/05h35m25.23s;J2000 DEC.= −05d15m35.40s/− 05d15m35.69s),and its external FUV radiation field is estimated to be G 0 ∼ 500, based on Herschel FIR observations (Pabst et al. 2021).The gas masses of the disks are estimated to be ∼ 79 and 30 M Jup for A and B (∼ 0.08 and 0.03 M ⊙ ), respectively, computed from the 856µm dust continuum emission (Williams et al. 2014).253-1536A has an estimated disk radius of 0. ′′ 75 equivalent to ∼ 300 au, and is 1. ′′ 1 (440 au) away from 253-1536B (Williams et al. 2014).No evidence of a larger circumbinary disk around the system has been found (Smith et al. 2005;Mann & Williams 2009).
The three protoplanetary disks of this study are massive enough (≥ 30 M Jup ) to potentially form planets (Mann & Williams 2009), providing test cases for studying planet formation under extreme irradiation.

Observations details
The observations of the 216-0939 and 253-1536A/B protoplanetary disks were obtained with ALMA as part of the Cycle 5 project #2018.1.01190.S (PI: V. Guzmán).The ALMA Band 6 observations included two spectral settings, at 1.2 and 1.3 mm.The correlator setup was configured with narrow spectral windows targeting different molecular lines.The main targets of the observations were lines from species commonly observed in isolated disks, such as the CO isotopologues, HCN, small carbon chains, H 2 CO, and deuterated species.Table 2 summarizes the molecular line targets and their spectral properties.
The Band 6 observations were carried out in August 2018 with baseline lengths spanning between 41 and 3640 m approximately.The total on-source time was 82 and 84 minutes per spectral setting, for the 216-0939 and 253-1536 disks, respectively.The quasar J0423-0120 was observed to calibrate the frequency bandpass and amplitude, and the quasars J0529-051 and J0607-0834 were observed to calibrate phase temporal variations (see Table A1).
Additionally, we used archival ALMA Band 7 observations of the two protoplanetary disks obtained as part of the Cycle 0 project #2011.0.00028.S (PI: R. Mann), which include the HCO + (4 − 3) and HCN (4 − 3) lines.More details about these observations can be found in Mann et al. (2014).

Data reduction
The initial data calibration was performed by ALMA staff using standard procedures in CASA (Common Astronomy Software Applications) version 6.4 (McMullin et al. 2007).Additionally, to increase the signal-to-noise (SNR) of the observations, we further self-calibrated the data using the continuum.The first step was to create pseudo-continuum visibilities by flagging all channels that contained line emission.Then, we ran a total of seven phase-only calibration iterations and one amplitude calibration iteration.This procedure improved the SNR of the continuum emission only by a factor of ∼ 1.2 for 216-0939 and ∼ 1.8 for 253-1536A/B on average.Then, the self-calibration solutions were applied to each spectral window including the channels with line emission.The continuum was then subtracted from the visibilities using the uvcontsub routine to produce the self-calibrated visibilities of the different lines.
The continuum images were produced from the selfcalibrated visibilities by first extracting the continuum channels using the split routine and imaging the visibilities using the tclean routine with a Briggs robust weighting parameter of +0.5, which balances the SNR and spatial resolution.We used an elliptical cleaning mask created using CASA regions for each source (216-0939, 253-1536A, and 253-1536B).The continuum images are shown in Fig. 1, and the final rms and beam sizes of the continuum images are listed in Table 3.
For the molecular lines, we used a Briggs robust parameter value of +0.5 because the line emission was not bright enough to be imaged at higher angular resolution.We also used the uvtaper parameter in tclean to reduce the weight of the longest baselines in the uv−plane, in particular we used uvtaper=0.5arcsec.In this way, higher spatial frequencies are weighted down relative to lower spatial frequencies, increasing the sensitivity to larger-scale emission.This technique is used when there are poorly sampled areas in the uv−plane or to increase the SNR.We also explored the effects of uv cuts to minimize the cloud contamination, which affected some of Note-Molecular data extracted from the CDMS (Müller et al. 2001(Müller et al. , 2005;;Endres et al. 2016), JPL (Pickett et al. 1998), and LAMDA molecular (Schöier et al. 2005) catalogues, obtained through www.splatalogue.net(Remijan et al. 2007).
the lines.The same was investigated by Factor et al. (2017) for the HCO + and HCN lines (4-3).They excluded baselines shorter than 70 kλ and found some improvement in the quality of the observations.However, we did not find a significant improvement in the quality of our observations after removing the shortest baselines.This can be explained by the different baseline coverage, which spans from 41 to 3640 m in our dataset, while it spans from 21.2 to 384.2 m in the data analyzed by Factor et al. (2017).
To help the cleaning process, we created a Keplerian mask adapted to each source in the tclean process task, that selects regions with line emission in each channel.The Keplerian masks were generated with the publicly available Python code from Teague (2020), that computes the Keplerian motion of the disk given the mass of the central star, the disk geometry and the cube parameters, such as spectral resolution, line frequency, source position, and systemic velocity.For the binary system, we created a Keplerian mask for 253-1536A and 253-1536B separately1 and then added them to produce a total mask that was used in the cleaning process.The inclination and PA used to create the masks are listed in Table 1.However, the reported PA for 253-1536B disk did not capture well the Keplerian rotation of the disk, so after some visual inspection we instead used a more conservative mask with an inclination of 45 • and PA of −90 • .Appendix B shows the resulting Keplerian masks overlaid on the channel maps.The rms and beam sizes of the line cubes can be found in Table 3.
In order to better compare the dust continuum emission with the molecular line emission, we created a second version of the dust continuum images, using the imsmooth routine to degrade the angular resolution of the continuum observations.This task performs a Fourier-based convolution to smooth the image and increase the SNR.In particular, we used it to obtain dust continuum images with the same angular resolution as the images of the different molecular lines.

RESULTS
In this section, we present the continuum and molecular line detections and non-detections in the two externally irradiated systems.First, we describe the highresolution continuum emission of the disks.Then we present the results for the molecular line emission.In particular, we extract disk-integrated flux densities for the molecular lines, and disk-averaged column densities.Finally, we estimate the HCN excitation temperature for both disks, using archival Band 7 observations.

Dust continuum emission
The beam size of the high-angular resolution continuum images of our sources is ∼ 0. ′′ 13, resulting in a spatial resolution of ∼ 54 au (for a distance of 414 pc).These images are shown in the left panels of Fig. 1.The Note-Lines were imaged with a spectral resolution of 0.4 km/s.a Average rms estimated from the line free channels which correspond to the first 5 and last 5 channels in the image cube.
smoothed continuum images have an angular resolution of ∼ 0. ′′ 5 and are shown in the left panels in Figure 2.
The new high-angular resolution observations allow us to spatially resolve the dust emission in these disks.Using the Python package GoFish (Teague 2019a) to generate deprojected radial profiles (see right panels in Fig. 1), we estimate a disk size of 311.5 ± 14.5 au for the 216-0939 disk, defined as the radius containing 95% of the total flux.The observations also reveal a central cavity with an outer edge at 120 − 135 au.The cavity is also clearly seen in the cleaned image (see Figure 1).We note that this inner cavity is consistent with the scenario of the central star being a tight binary system, that would clear the inner disk from material due to tidal interactions with the disk, which was proposed by Factor et al. (2017).Additionally, the 216-0939 disk seems to be asymmetric or eccentric, with the southern side being more elongated and 23% brighter than the northern side.The origin of this eccentricity is unknown and should be investigated in the future.
We are able to spatially resolve the region between the two members of the binary system 253-1536A/B.We estimate an angular separation of 0. ′′ 3 between the two disk edges (equivalent to ∼ 124 au), and a disk size of 239.1 ± 14.5 au and 108.7 ± 14.5 au for 253-1536A and 253-1536B, respectively.At the current angular resolution, the continuum emission looks very symmet-rical for both disks, contrary to other binary systems where spirals have been observed due to the interaction between the disks (e.g., Kurtovic et al. 2018).However, substructure could appear with even higher angular resolution observations.

Molecular lines
In this section, we present the results for the observed molecular lines in both disks.First, we explain our criteria to determine whether a line is detected or not.We then describe the spatial distribution of the detected lines.

Line detections
A line was considered to be detected if emission within the Keplerian mask was ≥ 3σ (rms) in at least 3 velocity channels.Following this criteria, 12 CO (2 − 1), 13 CO (2 − 1), C 18 O (2 − 1), and HCN (3 − 2) are robustly detected in both sources.Indeed, > 3σ emission is detected in almost every channel in both sources, with 10σ emission detected in at least 3 channels.H 2 CO (3 − 2) and C 2 H (3 − 2) are also detected according to this criteria but at lower SNR.In particular, 5σ emission is detected in a couple of channels for H 2 CO (3 − 2), and C 2 H (3 − 2) is only detected at the 3σ level in five or six channels in both disks.The other observed lines did not show emission ≥ 3σ in any channel, and are therefore considered as non-detections.For examples of channel maps of detected and non-detected lines, see Appendix B.
To confirm the detection of H 2 CO and C 2 H, we used the matched-filter method described in Loomis et al. (2018).The advantage of this method is that it looks for a signal directly in the visibilities, and therefore removes the uncertainties of the cleaning process which is present in the images.This method requires a model image as a matched filter; we used the HCN (3 − 2) and C 18 O (2 − 1) cubes as filters, as these molecular lines were robustly detected and they are less affected by the cloud contamination.A molecular line is considered detected if a ≥ 3σ peak is found near the source velocity in the filter-response spectrum, using at least one of the filters 2 .Following this criteria, we confirm the detection of both H 2 CO and C 2 H (3 − 2) lines in both disks.We applied the same method to the lines that were not detected in the image plane, and confirm that DCN

Spatial distribution
2 We note that the filter response spectrum depends on the used filter because the method assumes that the two lines (weak data and filter) have the same spatial distribution.If this is not the case, it is possible to obtain a false negative (see   .The profile of the smoothed dust continuum emission is shown in black in all panels, in units of mJy/beam.Color-shaded regions show the 1σ scatter of the averaged emission at each radial bin.The black vertical line marks the edge of the high-resolution dust continuum emission on each system, defined as the radius within which the 95% of the total flux arises.In the case of the binary system, the radial profiles are obtained with 253-1536A as the center of the emission, and the dust edge shown corresponds to 253-1536A. We created velocity-integrated maps for each detected molecular line.Figure 2  To do this, we integrated the emission from the first channel that presented > 3σ line emission to the channel that was closest to the systemic velocity of the source (blue-shifted); the same was done to the redshifted part, integrating the channels from the systemic velocity to the last channel that presented > 3σ line emission.We find that the 12 CO line emission is more extended than the 13 CO and C 18 O line emission in both 216-0939 and 253-1536A/B systems.Moreover, the CO isotopologue emission is more extended compared to the dust continuum emission.The line emission from H 2 CO and C 2 H is weaker than the HCN and CO isotopologue line emission for both systems.In addition, we note that the 12 CO line emission is heavily contaminated by extended cloud emission and foreground absorption near the systemic velocity.To a lesser degree, the other CO isotopologues and HCN also suffer from cloud contamination (see the channel maps in Appendix B and the disk-integrated spectra in Appendix D).
In the velocity-integrated maps it is also possible to disentangle the Keplerian rotation of the smaller disk from the larger disk in 253-1536A/B, in particular for the CO isotopologues and HCN, where the emission of the minor companion can be observed in the red-and blue-shifted contours (see also the channel maps in Appendix B).The emission can also be tentatively disentangled for C 2 H. Figure E.1 shows the first moment map for the CO (2−1) line in 253-1536A/B, where a tentative deviation from Keplerian rotation can be seen towards the northern side of 251-1536A (black arrow).This deviation could be associated with the interaction between the two companions or be due to the cloud contamination.Unfortunately, the angular resolution of 0. ′′ 5 is not high enough to resolve the separation between the two disks and to disentangle if smaller spatial signals of dynamical interaction are present.
To further investigate the spatial distribution of the different molecular lines, we generated radial profiles using GoFish to deproject and azimuthally average the line emission in the zero moment maps, using the disk parameters.Figure 3 shows the radial profiles of the brighter (top panels) and weaker (bottom panels) lines.For both 216-0939 and 253-1536A/B disks, all molecular line emission is more extended than the dust continuum emission.Most of the lines show little substructure at the current angular resolution (the apparent gaps/rings in C 2 H (3 − 2) are probably related to the noise in the images).However, an interesting feature is that the central emission of the C 2 H line is flat (and not centrally peaked), suggesting the emission could be arise from a ring.C 2 H ringed emission has been observed in other isolated disks (e.g., Bergin et al. 2016;Cleeves et al. 2021;Pegues et al. 2021;Guzmán et al. 2021).

Disk integrated fluxes
Disk-integrated fluxes were estimated for the detected lines, and upper limits are reported for non-detections.
To calculate the integrated flux of each line, we first multiplied the image cubes by the same Keplerian masks used to clean the data, using the immath routine in CASA, and then summed over the line emission.This additional step removes some of the noise in the image and results in a better-integrated disk spectrum where the Keplerian rotation is shown.To estimate the uncertainty in the integrated flux for each line, we used the same method described above to extract the fluxes but now changing the disk center in 1000 random samples in regions without emission outside of the original Keplerian mask via bootstrapping.The uncertainty in the integrated flux was then computed as the standard deviation of the resulting distribution added in quadrature with a 10% systematic flux calibration uncertainty.The number of samples is chosen to be sufficiently large so that the estimated uncertainty does not vary significantly.The resulting integrated intensities of the detected lines are listed in Table 4.We also list upper limits for the non-detected lines.The CO and CO isotopologue fluxes have larger errors because of the cloud contamination.Indeed, the cloud is more pronounced in CO than in the other species.

Column density retrieval
To estimate the disk-averaged column densities (N T ) we consider that the gas obeys local thermal equilibrium (LTE), and assume a range of possible excitation temperatures (T ex ) based on typical gas temperatures observed in the outer regions of disks (Guzmán et al. 2021).The LTE assumption is reasonable because typical gas densities in disks are high compared to the critical densities of the observed HCN, H 2 CO, and C 2 H lines3 , assuming their emission arises mainly from layers that are close to the midplane (Law et al. 2021a).
Under LTE conditions, the energy levels are populated following Boltzmann's law, where N u corresponds to the upper-level column density, N tot is the total column density of the molecule, Q(T ex ) is the partition function, g u is the upper-level degeneracy, E u is the upper-level energy, and k B is the Boltzmann constant.If the line is optically thin, the upper-level column density can be written as where ν is the line frequency, W is the integrated line intensity, h is the Planck constant, c is the speed of light, and A ul is the Einstein coefficient for spontaneous emission.All frequencies, Einstein coefficients, and partition functions values were taken from the CDMS catalog (Müller et al. 2001(Müller et al. , 2005;;Endres et al. 2016), and can be found in Table 2.We assume the optically thin approximation and a disk-averaged excitation temperature that ranges from 20 − 50 K, which are typical temperatures found in the disk molecular layer (e.g., Walsh et al. 2010).The results are shown in Fig. 4. We note that for the 253-1536A/B system, the reported column densities include the contribution from both members A and B. We can see that the HCN, H 2 CO, and C 2 H column densities are not too sensitive to the excitation temperature assumptions in this range.Overall, we find N tot (HCN) ∼ (0.5 − 1.8) × 10 13 cm −2 , N tot (H 2 CO) ∼ (0.4 − 6.2) × 10 13 cm −2 , and N tot (C 2 H) ∼ (0.3−1.0)×10 14 cm −2 for both disks.These column densities have to be considered as  lower limits because we are assuming that the lines are optically thin, which might not be the case, in particular for HCN.
We find that the HCN column density is quite similar between the two disks, with differences of only a factor of ∼2.Moreover, the C 2 H column density is almost the same for the two systems.In contrast, the formaldehyde column density varies between the disks.In particular, the H 2 CO/HCN ratio is a factor ∼ 5 larger in 216-0939 compared to 253-1536A/B.These column densities are within the range of what has been found in previous studies of isolated disks (e.g., Pegues et al. 2020;Guzmán et al. 2021), which reported disk integrated column densities for HCN and C 2 H in the range of 10 12 − 10 15 cm −2 in disks around low-mass stars, and > 10 16 cm −2 in Herbig Ae/Be disks.

HCN excitation temperature
We estimated the excitation temperature of HCN using the ratio between the HCN (3 − 2) observations presented in this work and the HCN (4 − 3) line initially published in Mann et al. (2014) and later presented in more detail in Williams et al. (2014) and Factor et al. (2017).Figure 5 shows the HCN (3 − 2) and (4 − 3) moment zero maps for both sources.For consistency, we computed the disk-integrated flux of the HCN (4 − 3) line in the same manner as was done for the (3 − 2) line.Additionally, we smoothed the HCN (3 − 2) emission to match the angular resolution of the HCN (4 − 3) line (0. ′′ 59 × 0. ′′ 53).We used the spectroscopic parameters of HCN listed in Table 2, and assumed τ = 0 (the optically thin case).We infer an excitation temperature of 20.0 +4.1 −2.9 K and 22.5 +5.4 −3.6 K for 216-0939 and 253-1536A/B, respectively.The derived excitation temperature of ∼ 20 K is similar for both disks and also similar to what has been found for other disks (e.g., Bergner et al. 2019a;Guzmán et al. 2021), suggests that the HCN emission arises from a relatively cold layer close to the midplane.In order to investigate whether the excitation temperature increases in the outer disk, which could be expected for externally irradiated disks, and taking advantage of the angular resolution of the observations, we also derived the excitation temperature as a function of radius.We found that, at the current angular resolution, the excitation temperature is constant across both disks, with no visible increase in the outer disk.

DISCUSSION
In this section, we first compare the measured line fluxes of 216-0939 and 253-1536A/B with those found in disks around stars of different masses that are located in low-mass star forming regions, where the external radiation field is significantly lower compared to the disks in our sample.Then, we compare our findings with predictions from chemical models.

Comparison between irradiated and isolated disks
Figures 6 and 7 show the distance-normalized integrated fluxes and upper limits as a function of stellar mass, for the detected and non-detected lines, respectively.For 216-0939, we consider two possible stellar scenarios: a tight equal mass binary of two 1.1 M ⊙ stars and a single star of 2.17 M ⊙ , as discussed in Section 2. For 253-1536A/B, we only consider the mass of the primary A star (3.5 M ⊙ ), since most of the emission we detect comes from this source.For comparison, we also include fluxes reported for isolated disks around M4-M5 stars, T Tauri stars, and Herbig Ae/Be stars.The literature disk sample was compiled from Huang et al. ( 2017 2023).We note that the different molecular line fluxes were compiled from different studies, and in some cases the same line flux is reported in two or more studies.In those cases, we selected the most recent study.The total disk sample includes 5 M4-M5 stars, 6 Herbig Ae disks, 13 T Tauri disks, and the 2 externally irradiated systems from our sample.The stellar masses range from 0.12 to 3.5 M ⊙ , and the stellar ages range between 0.4 to > 10 Myr.The two disks in Orion have ages within this range (∼ 1−3 Myr; Williams et al. 2014;Factor et al. 2017).
Considering the combined sample of isolated and externally irradiated disks, we find a positive correlation between the fluxes and stellar mass, for all the lines pre-  Ae/Be (Huang et al. 2017;Bergner et al. 2019bBergner et al. ,a, 2020;;Pegues et al. 2020;Öberg et al. 2021;Law et al. 2021b;Guzmán et al. 2021;Pegues et al. 2023).The fluxes have been normalized to a common distance of 414 pc.The left red diamond corresponds to the 216-0939 disk, assuming the source is a binary system of two ∼ 1 M⊙ stars.A horizontal error bar is added to consider the possibility the system is a single massive star of 2.17 M⊙.The right red diamond corresponds to the 253-1536A disk.
sented here (see Fig. 6).This trend was previously reported for isolated disks only by Pegues et al. (2021) and Pegues et al. (2023).The estimated Spearman correlation coefficient of the combined disk sample is shown in the bottom right corner of each panel, which corresponds to the dispersion of the data and measures the correlation between the two variables.The closer to one the value of the Spearman correlation coefficient, the stronger the correlation.The strongest correlations are found for C 18 O, HCN and C 2 H.In principle, one could have expected to find brighter CO emission in the two irradiated disks compared to the isolated disks because they are expected to be warmer.However, we find that the 12 CO and 13 CO lines are weaker than expected according to the observed trend, with an emission flux similar to the Herbig disks of lower masses.This is probably related to the cloud contamination in the irradiated disks, which results in a drop of the CO emission in the central channels (see the channels maps in Appendix B).This is also consistent with the fact that the less contam-inated C 18 O emission shows a stronger correlation than the more abundant CO isotopologues.This suggests that there is no difference in the chemistry between the isolated sources and the two irradiated disks presented here, and that the differences between the line fluxes in the disks are mainly due to the stellar masses.
Regarding the non-detected molecular lines, the derived upper limits are consistent with the fluxes measured in other disks, where these lines have been detected (see Figure 7).
In order to remove the dependence of the fluxes on the stellar mass, in Fig. 8 we show the normalized line fluxes of each line against one another, in a similar manner to Bergner et al. (2019b), Bergner et al. (2019a), Pegues et al. (2020), Pegues et al. (2021), andPegues et al. (2023).Overall, we find that for the combined disk sample, every molecular line combination in Fig. 8 Same as Figure 6, but displaying the upper limits for a subset of the non-detected molecular lines.
H 2 CO pair shows one of the worst correlation, with a Spearman coefficient of 0.66.Indeed, the observed C 2 H flux in the irradiated disks is lower than expected based on the observed trend in isolated disks.The strongest correlation corresponds to HCN vs 13 CO with a Spearman coefficient (ρ = 0.91); followed by C 18 O vs 13 CO (ρ = 0.89), and C 2 H vs HCN, with a correlation coefficient of 0.82.
Finally, Figure 10 shows four different flux ratios (C 2 H/C 18 O, H 2 CO/C 18 O, HCN/C 18 O, and C 2 H/HCN) as a function of stellar mass.Considering the isolated disks only, Pegues et al. (2021) found no clear correlation between these flux ratios and stellar mass.However, with the addition of the externally irradiated disks 216-0939 and 253-1536A/B, we find that the C 2 H (3 − 2)/HCN (3 − 2) flux ratio shows a tentative trend, with the ratio decreasing with stellar mass (Spearman coefficient of −0.24).This result is unexpected since the C 2 H emission should be brighter in irradiated disks because the C 2 H formation is expected to be enhanced in the presence of FUV radiation.However, other factors, like carbon depletion (Bergin et al. 2016) and dust growth/settling, also play a role in the formation of carbon chains, and could explain the faint C 2 H emission in the two Orion disks.In addition, HCN is also known to be sensitive to photochemistry (Guzmán et al. 2015;Visser et al. 2018;Bergner et al. 2021;Pegues et al. 2021).Future observations towards more irradiated disks are needed to confirm the observed tentative trend.
A similar comparison of flux ratios is presented in Figure 9 but for a subset of the non-detected molecular lines (DCN and c−C 3 H 2 ).We find that the observed upper limits in the two irradiated disks are, in general, consistent with the values found in isolated disks.However, the ratios involving DCN (e.g., DCN/HCN and DCN/H 2 CO) seem to be slightly lower for the irradiated disks compared to the isolated ones.If this is confirmed, it would suggest that irradiated sources have less cold material than isolated disks, which is consistent with the expectation of irradiated disks being warmer.However, in that case we would also expect to see an enhancement of H 2 CO in the outer disk due to ice desorption in the warm gas, assumming that H 2 CO was inherited, and this is not observed with the current observations (see Fig. 3).Observations with better sensitivity are needed to confirm this result.
These results suggest that the chemistry of the two irradiated disk systems presented here is not too different from the chemistry of isolated sources, at least for the molecular lines considered here.Indeed, the two irradiated disks follow the trends observed for nearby isolated disks.However, we note that the results presented here may not be representative of the chemistry of all irradiated disks, because our sample consists of only two particularly massive irradiated disk systems; 216-0939 and 253-1536A/B are massive disks, located at 1.59 and 0.92 pc away from the radiation field source, respectively.Indeed, it is possible that some material in the outer disk has been stripped away by the external radiation field, and our observations are just tracing the part of the disk that has survived, and is no longer affected by the external radiation field.Another possibility is that the external radiation fields are lower than the ones assumed here, due to projection effects or to the extinction of UV photons by surrounding cluster material.
Molecular line surveys of smaller disks and disks that are closer to the massive stars are needed to draw more general conclusions about the chemistry of externally irradiated disks.Unfortunately, this is very challenging because lines are usually heavily contaminated by the emission from the molecular cloud.However, recent surveys have been able to detect CO and HCO + lines in disks close to the ONC (ranging between 0.03−0.15pc), thanks to very sensitive and high angular resolution ALMA observations (Boyden & Eisner 2020).Our sample consists of two systems around intermediate-mass stars (1 − 3.5 M ⊙ ); unfortunately, chemical models that include an external radiation field have not been developed for this type of star/disk systems yet.However, previous theoretical studies have found that the chemistry of T Tauri and Herbig disks are not too different (Agúndez et al. 2018).Therefore, we compare our results with models of disks around lower-mass T Tauri stars.We use the chemical models presented in Walsh et al. (2013) and Walsh et al. (2014), that were developed for a disk around a T Tauri star with a mass of 0.5 M ⊙ , a radius of 2 R ⊙ , and an effective temperature of 4000 K.In the models, the disk is ex-ternally irradiated by UV photons from the interstellar radiation field and a nearby massive O-type star at a distance of 0.1 pc.We note that the radiation field in the model goes up to 4 × 10 4 G 0 , which is higher than the radiation field illuminating the two disks presented here (< 10 3 G 0 ).The main results of these models are the following: Brighter molecular line emission.Some molecular lines may be brighter for the irradiated disks because of the higher disk temperatures.In particular, this should occur for CO (and their isotopologues), CN, and HCN.Our detections of 12 CO, 13 CO, and C 18 O are not consistent with this prediction, but this could be due . Same as Figure 10 for a subset of the non-detected molecular lines.
to the cloud contamination discussed in Section 3.2.In the case of HCN, the brighter line emission in irradiated disks compared to isolated disks seems to be related to the stellar mass and not to the external radiation field.
In addition, the measured HCN excitation temperature of ∼ 20 K is similar to what has been found in isolated disks, suggesting that the HCN emission arises from relatively cold gas with the caveat that the lines could be optically thick.A possible explanation is that we are observing the inner regions of the disk that are currently shielded and no longer affected by the external radiation.But more importantly, it is possible that the external radiation field is just not high enough to produce a significant difference in the line emission.COMs enhancement.Chemical models predict that complex organic molecules that are typically frozen out onto dust grains could be observed in the gasphase in warmer and externally irradiated disks because of their higher temperatures that sublimate these molecules into the gas-phase, in particular in the outer disk.For example, formaldehyde, as a precursor and tracer of COMs, is expected to be enhanced in the outer disk, where the temperature should be higher.However, the observed H 2 CO radial profiles in the two irradiated disks are centrally peaked, with no additional emission component in the outer disks that would be indicative of ice desorption (see Figure 3).However, we note that the SNR of the detected H 2 CO lines is low, so it is possible that the current observations are not sensitive enough to detect such a component in the outer disk.
Radiation field tracers.The HCN/HCO + and CN/HCN line ratios are expected to be larger in irradiated disks compared to isolated disks.HCN/HCO + ratios larger than one can be indicative of enhanced external radiation, because HCO + traces the cold, dense regions of the disk, which are smaller for irradiated disks (Walsh et al. 2013).Factor et al. (2017) measured a HCN/HCO + ratio of 0.58 ± 0.04 for the 216-0939 disk, which is consistent with the isolated models from Walsh et al. (2013).CN/HCN > 1 can also be indicative of enhanced radiation fields, similar to what is observed in photodissociation regions (PDRs), where the CN/HCN ratio is found to correlate with the FUV field (Fuente et al. 1993).This is because FUV photons enhance the abundance of CN in the outer disk, which is more exposed to external radiation (Guzmán et al. 2015;Bergner et al. 2021).Unfortunately, CN has not yet been observed in these irradiated disks.Future observations of CN lines are needed to verify this prediction.

SUMMARY
We have presented ALMA observations of the continuum and line emission of several molecules toward two externally FUV-irradiated protoplanetary disk systems around pre-main sequence stars in the outskirts of the Orion Nebula Cluster.In particular, we presented observations of the CO isotopologues, the small organic molecules HCN and H 2 CO, the carbon chains C 2 H and c−C 3 H 2 , and the deuterated species DCN and N 2 D + , all of which have been previously detected in isolated disks.The main conclusions are the following: • The high-angular resolution observations of the dust continuum emission in 216-0939 shows the presence of a large gap in the inner disk, that is well resolved for the first time in our data.Additionally, the dust emission is asymmetric or eccentric, with the southern side of the disk being 23% brighter than the northern side.We estimate the outer edge of this gap to be around 120−135 au, and a disk size of 311.5±14.5 au.The high-angular resolution observations of the dust continuum emission allow us to separate the two members in the binary system 253-1536A/B, measure their disk sizes (239.1 ± 14.5 au for A, and 108.7 ± 14.5 au for B) and the separation between their edges (∼ 124 au).In addition, we do not observe substantial substructure with the current observations, in neither 216-0939 nor 253-1536A/B.The cold tracers N 2 D + and DCN are expected to be less abundant in the warmer irradiated disks compared to isolated disks, since these molecules are formed more efficiently at low temperatures.However, the estimated upper limits for the diskintegrated fluxes are consistent with detections of these lines in isolated disks.
• In general, we do not observe significant differences between the chemistry of isolated and the two irradiated disks presented in this work, based on the observed disk-integrated fluxes and flux ratios for the molecular lines presented here.The differences between the 216-0939 or 253-1536A/B disks and typical T Tauri and Herbig Ae/Be disks found in low-mass star forming regions, seem to be more closely related to the different stellar masses than to the presence of an enhanced external radiation field.
• The observed disk integrated fluxes and line ratios in the two irradiated disks presented here are not consistent with chemical model predictions of externally irradiated disks presented by Walsh et al. (2013).However, the 216-0939 and 253-1536A/B disk are irradiated by a weaker FUV field (< 10 3 G 0 ) than the one included in the models (∼ 10 4 G 0 ).Our results, therefore, suggest that these disks are far enough away from the ONC so that their chemistry is no longer substantially af-fected, but disks located closer to the stellar cluster may experience stronger chemical effects.
The results presented in this work show that the chemical composition in these moderately irradiated systems is similar to that in isolated disks, which suggests that the assembly of planetary systems and their atmospheres will proceed in a similar manner to that expected in the better studied isolated systems.
Future observations of disks exposed to higher radiation fields are needed to better determine the differences between isolated and externally irradiated disks.In particular, observations of disks closer to the ONC are needed to investigate how the chemistry changes with distance from the ionizing source.In addition, observations with better sensitivity are needed to detect lines from cold molecular tracers, such as DCN, and determine whether they are indeed less abundant in the warmer irradiated disks compared to isolated disks.Finally, chemical models including more moderate radiation fields are needed to further investigate how the chemistry is affected by an external radiation field.et al. 2018).The first figure shows the difference in the response for a bright detected line (H 2 CO) compared to a weak non-detected line (DCN), using the HCN line as a filter in both cases.The second figure shows the difference in the response for C 2 H when using different filters.A clear peak is seen in the response when the C 18 O line is used as a filter, suggesting the C 2 H line is detected, and a less clear detection is seen when instead HCN is used as a filter.This suggests that the distribution of C 2 H is likely more similar to that of C 18 O than that of HCN..1 shows the disk-integrated spectra for the detected lines in 216-0939 and 253-1536A/B.These spectra are obtained by adding all the emission inside the Keplerian mask in each channel.We note that the disk-integrated spectra of the 253-1536A/B system includes emission from both binary members.The CO isotopologues and HCN emission show a double peaked profile, typical of Keplerian rotation of an inclined disk, in both sources.However, the CO isotopologue lines are heavily affected by absorption from the molecular cloud, in particular in the central channels.
For molecular lines such as H 2 CO and C 2 H, it is more difficult to distinguish the Keplerian rotation associated with the disk, because these lines are faint.

Figure 1 .
Figure 1.Left: Dust continuum emission images at high-angular resolution from the protoplanetary disks 216-0939 (top) and253-1536 (bottom).Right: Deprojected radial profiles for the continuum emission of each disk.For 216-0939, the radial profile is divided in north (blue) and south (orange) side emission.In particular, we estimated each radial profile including the emission from an aperture of 30 • in each side.For 253-1536, the radial profile of 253-1536A and 253-1536B are shown in blue and orange, respectively.Vertical lines represent the estimated disk size for each case, containing 95% of the total flux of each disk.For 216-0939, the black vertical line represents the disk size estimated from the full azimuthally averaged radial profile.
shows the resulting maps for 216-0939 (top) and 253-1536A/B (bottom).In this figure, the low-angular resolution images of the continuum emission are shown in the left panels for comparison.The top panels show the 12 CO (2 − 1), 13 CO (2 − 1), and C 18 O (2 − 1) line emission, and the bottom panels show the emission from HCN (3 − 2), H 2 CO (3 − 2) and C 2 H (3 − 2).The red and blue contours show the Keplerian rotation of the disks, created by integrating the line emission over the red-and blue-shifted parts of the line.

Figure 4 .
Figure 4. Disk-averaged column densities for HCN, H2CO and C2H, assuming optically thin emission.The colors represent different excitation temperatures (Tex) assumed in the calculation.
Figures 6 and 7 show the distance-normalized integrated fluxes and upper limits as a function of stellar mass, for the detected and non-detected lines, respectively.For 216-0939, we consider two possible stellar scenarios: a tight equal mass binary of two 1.1 M ⊙ stars and a single star of 2.17 M ⊙ , as discussed in Section 2. For 253-1536A/B, we only consider the mass of the primary A star (3.5 M ⊙ ), since most of the emission we detect comes from this source.For comparison, we also include fluxes reported for isolated disks around M4-M5 stars, T Tauri stars, and Herbig Ae/Be stars.The literature disk sample was compiled from Huang et al. (2017), Bergner et al. (2019b), Bergner et al. (2019a), Bergner et al. (2020), Pegues et al. (2020), Pegues et al. (2021), Law et al. (2021b), Guzmán et al. (2021), Öberg et al. (2021), and Pegues et al. (2023).We note that the different molecular line fluxes were compiled from different studies, and in some cases the same line flux is reported in two or more studies.In those cases, we selected the most recent study.The total disk sample includes 5 M4-M5 stars, 6 Herbig Ae disks, 13 T Tauri disks, and the 2 externally irradiated systems from our sample.The stellar masses range from 0.12 to 3.5 M ⊙ , and the stellar ages range between 0.4 to > 10 Myr.The two disks in Orion have ages within this range (∼ 1−3 Myr;Williams et al. 2014;Factor et al. 2017).Considering the combined sample of isolated and externally irradiated disks, we find a positive correlation between the fluxes and stellar mass, for all the lines pre-

Figure 6 .
Figure6.Molecular line fluxes for FUV irradiated disks in the Orion Nebula Cluster (this work) as a function of stellar mass, compared with isolated disks compiled from the literature, including M4-M5 stars(Pegues et al. 2021), T Tauri and Herbig Ae/Be(Huang et al. 2017;Bergner et al. 2019bBergner et al. ,a, 2020;;Pegues et al. 2020;Öberg et al. 2021;Law et al. 2021b;Guzmán et al. 2021;Pegues et al. 2023).The fluxes have been normalized to a common distance of 414 pc.The left red diamond corresponds to the 216-0939 disk, assuming the source is a binary system of two ∼ 1 M⊙ stars.A horizontal error bar is added to consider the possibility the system is a single massive star of 2.17 M⊙.The right red diamond corresponds to the 253-1536A disk.For CO (2 − 1), lower limits are shown since these lines are highly affected by cloud contamination.a C2H (3 − 2) includes J = 5/2 − 3/2 and J = 7/2 − 5/2 lines.
has a strong and positive correlation, as was previously found by Bergner et al. (2019a); Pegues et al. (2021); Pegues et al. (2023) for the isolated disks only.The C 2 H vs

Figure 8 .
Figure 8.The molecular line ratios for FUV irradiated disks in the Orion Nebula Cluster (this work) compared with isolated disks from the literature.The two disk systems from this work are shown as red diamonds.The dashed line shows a one-to-one correlation. 12CO is not included in this comparison because the derived fluxes are affected by cloud contamination.

Figure 10 .
Figure 10.The molecular line ratios for FUV irradiated disks in the Orion Nebula Cluster (this work) as a function of stellar mass compared with isolated disks from the literature.Note that this figure shows only a subset of the non detected lines.
Figure D.1 shows the disk-integrated spectra for the detected lines in 216-0939 and 253-1536A/B.These spectra are obtained by adding all the emission inside the Keplerian mask in each channel.We note that the disk-integrated spectra of the 253-1536A/B system includes emission from both binary members.The CO isotopologues and HCN emission show a double peaked profile, typical of Keplerian rotation of an inclined disk, in both sources.However, the CO isotopologue lines are heavily affected by absorption from the molecular cloud, in particular in the central channels.For molecular lines such as H 2 CO and C 2 H, it is more difficult to distinguish the Keplerian rotation associated with the disk, because these lines are faint.

Figure B. 1 .
Figure B.1.Channel maps for the CO (2 − 1), 13 CO (2 − and C 18 O (2 − 1) lines in 216-0939 (systemic velocity of 10.75 km/s).The Keplerian mask is shown in black.Red contours correspond to the 3σ level of the emission.The synthesized beam is shown in the bottom left panel.

Table 2 .
Spectroscopic parameters of targeted molecular lines.

Table 3 .
Molecular line and continuum image parameters/properties.

Table 4 .
Disk integrated fluxes.Note-The integrated fluxes are measured within the Keplerian masks.In the case of 253-1536, they include both A and B members. a Reported fluxes for non-detections correspond to 3σ upper limits where σ is the uncertainty estimated via bootstrapping.bThesame process was performed on the ALMA Band 7 data obtained from Cycle 0 project #2011.0.00028.s(Mannet al. CO and C 2 H, assuming optically thin emission, and a range of excitation temperatures, and found values in the range of 10 13 − 10 14 cm −2 , similar to what is observed in isolated disks.• Molecular lines such as c−C 3 H 2 (6 06 − 5 15 ), c−C 3 H 2 (6 25 − 5 14 ), c−C 3 H 2 (7 07 − 6 16 ), DCN (3 − 2) and N 2 D + (3 − 2) were not detected in neither the 216-0939 nor 253-1536A/B disks.

Table A1 .
Spectral settings.MATCHED FILTER RESULTS Figs.C.1 and C.2 show examples of the filter response spectra from the matched-filter technique (VISIBLE; Loomis C.