Linking ice and gas in the Coronet cluster in Corona Australis

Context. During the journey from the cloud to the disc, the chemical composition of the protostellar envelope material can be either preserved or processed to varying degrees depending on the surrounding physical environment. Aims. This works aims to constrain the interplay of solid (ice) and gaseous methanol (CH 3 OH) in the outer regions of protostellar envelopes located in the Coronet cluster in Corona Australis (CrA), and assess the importance of irradiation by the Herbig Ae/Be star R CrA. CH 3 OH is a prime test case as it predominantly forms as a consequence of the solid-gas interplay (hydrogenation of condensed CO molecules onto the grain surfaces) and it plays an important role in future complex molecular processing. Methods. We present 1.3 mm Submillimeter Array (SMA) and Atacama Pathfinder Experiment (APEX) observations towards the envelopes of four low-mass protostars in the Coronet cluster. Eighteen molecular transitions of seven species were identified. We calculated CH 3 OH gas-to-ice ratios in this strongly irradiated cluster and compared them with ratios determined towards protostars located in less irradiated regions such as Serpens SVS 4 in Serpens Main and the Barnard 35A cloud in the λ Orionis region. Results. The CH 3 OH gas-to-ice ratios in the Coronet cluster vary by one order of magnitude (from 1.2 × 10 − 4 to 3.1 × 10 − 3 ) which is similar to less irradiated regions as found in previous studies. We find that the CH 3 OH gas-to-ice ratios estimated in these three regions are remarkably similar despite the different UV radiation field intensities and formation histories. Conclusions. This result suggests that the overall CH 3 OH chemistry in the outer regions of low-mass envelopes is relatively independent of variations in the physical conditions and hence that it is set during the prestellar stage.


Introduction
Solar-type stars and planets form inside molecular clouds, when dense cores undergo gravitational collapse.While collapsing, the gas and dust constituting the clouds is assembled into infalling envelopes, streamers and circumstellar discs, which supply the fundamental ingredients for planet formation (see e.g., Pineda et al. 2023).Recent observations (e.g., Andrews et al. 2018;Harsono et al. 2018;Keppler et al. 2018;Segura-Cox et al. 2020) suggest that planets form earlier than previously thought (a few 10 5 years; Tychoniec et al. 2020) and in tandem with their host stars (Alves et al. 2020).In this context, it is still un-clear whether the composition of the molecular cloud material is preserved when becoming part of planets or, instead, entirely thermally processed losing the natal cloud's chemical fingerprint (see e.g., Jørgensen et al. 2020;van Dishoeck & Bergin 2020 for recent reviews).
To address this query it is necessary to study how the environment impacts the physical and chemical properties of embedded protostars and their discs.For instance, external irradiation (Winter et al. 2020) and cosmic ray ionization (Kuffmeier et al. 2020) might shape the chemical and physical evolution of forming low-mass stars by e.g., photoevaporating their discs (van Terwisga et al. 2020; Ha- worth et al. 2021 and references therein) or leading to less massive discs (Cazzoletti et al. 2019).In this paper we investigate the variations of the methanol gas and ice towards deeply embedded sources in Corona Australis, to study the effects of external irradiation on the gas-to-ice ratios.CH 3 OH is the most suitable molecule for the main aim of this work because it is abundantly detected in both solid and gas phases (Boogert et al. 2015;McGuire 2022).Most importantly, it is primarily formed in the solid-state (Watanabe & Kouchi 2002;Fuchs et al. 2009;Qasim et al. 2018;Simons et al. 2020;Santos et al. 2022) as its gasphase formation pathways are considerably less efficient (Roberts & Millar 2000;Garrod & Herbst 2006;Geppert et al. 2006).Finally, CH 3 OH is regarded as the gateway species for the formation of complex organic molecules both in the solid-state (e.g., Öberg et al. 2009b;Chuang et al. 2016;Fedoseev et al. 2017) and in the gas phase (e.g., Shannon et al. 2014;Balucani et al. 2015).
Corona Australis (CrA) is one of the nearest regions with ongoing star-formation, located at a distance of 149.4±0.4 pc as estimated from Gaia-DR2 measurements (Galli et al. 2020).The most recent census of the cloud counts 393 young stellar object (YSO) candidates (Esplin & Luhman 2022): these are relatively evolved, mainly Class II and III sources, and they are predominantly concentrated in the most extincted region, the "head" (Peterson et al. 2011;Alves et al. 2014).The correlation between the star formation activity and the head-tail structure of Corona Australis has been investigated by Dib & Henning (2019), who found that the spatial distribution of dense cores is a consequence of the physical conditions of the large scale environment present at the time the cloud assembled.The youngest population of YSOs (Class 0/I) is situated in the Coronet cluster (Taylor & Storey 1984;Forbrich et al. 2007;Forbrich & Preibisch 2007), also associated with the luminous Herbig Ae/Be star R CrA (Peterson et al. 2011).
Figure 1 displays a section of the Coronet cluster and a summary of the principal cluster members is provided in Table 1.The Herbig Ae/Be R CrA (spectral type B5−B8; Gray et al. 2006;Bibo et al. 1992) is the brightest star in this very young cluster that has an estimated age of 0.5−1 Myr (Sicilia-Aguilar et al. 2011).Due to the variable nature of R CrA, its stellar mass and luminosity are uncertain (Mesa et al. 2019).Recent Very Large Telescope (VLT)/SPHERE observations of R CrA resolved a companion at a separation of ∼0.156 ′′ −0.184 ′′ and complex extended jet-like structures around the star (Mesa et al. 2019).A second variable star, the T Tauri star T CrA is present in the Coronet cluster, approximately 30 ′′ to the south-east of R CrA (Herbig 1960;Taylor & Storey 1984).The region surrounding the two variable stars harbours seven identified Class 0/I YSOs (Table 1).
The Coronet cluster members have been at the center of active multi-wavelength research, mostly aimed at the characterization of the properties of YSOs at very early stages of stellar evolution.Some of the most studied objects are IRS7B, IRS7A, SMM1A, SMM1C and SMM2 (Nutter et al. 2005;Groppi et al. 2007;Miettinen et al. 2008;Chen & Arce 2010;Peterson et al. 2011).The spectral energy distributions (SEDs) of the cluster members have been investigated by Groppi et al. (2007), suggesting that SMM1C is a Class 0 YSO and IRS7B is a transitional Class 0/I object.IRS7A and SMM2 are likely Class I sources (Peterson et al. 2011) and SMM1A is classified as a pre-stellar core candidate (Nutter et al. 2005; Chen & Arce 2010).IRS1, IRS5N and IRS5A are suggested as Class I YSO based on submillimeter and infrared observations (Peterson et al. 2011).
Along with the identification and the evolutionary stage designation of the YSOs in the Coronet, their chemical evolution has been the subject of a large number of studies over the last decades.The line-rich spectra of IRS7B and IRS7A have been investigated with the Atacama Pathfinder EXperiment (APEX) single-dish telescope by Schöier et al. (2006), who reported different kinetic temperatures for formaldehyde (H 2 CO) and methanol (>30 K for H 2 CO and ≈20 K for CH 3 OH).Lindberg & Jørgensen (2012) suggested that the high temperatures in the region (>30 K) traced by the emission of H 2 CO, are caused by external irradiation from the Herbig Ae/Be star R CrA.With the purpose of analyzing the impact of external irradiation on the molecular inventory of low-mass protostars, systematic unbiased line surveys have been carried out firstly towards IRS7B with the Atacama Submillimeter Telescope Experiment (ASTE) by Watanabe et al. (2012), and extended to other Coronet cluster members using APEX by Lindberg et al. (2015).These systematic surveys confirmed that external irradiation can affect the chemical nature of protostars, particularly by enhancing the abundances of Photon-Dominated Regions (PDRs) tracers (such as CN, C 2 H, and c-C 3 H 2 ).High-resolution Atacama Large Millimeter/submillimeter Array (ALMA) observations found a low column density of CH 3 OH in the inner hot regions Article number, page 2 of 20 of the protostellar envelopes which could either be due to the suppressed CH 3 OH formation in ices at the higher temperatures in the region or due to the envelope structures on small scales being affected by the presence of a disc (Lindberg et al. 2014).The latter interpretation is supported by recent observations (Artur de la Villarmois et al. 2018;van Gelder et al. 2022) and models (Nazari et al. 2022b).
This paper explores the variations of CH 3 OH gas and ice abundances, investigating the effects of external irradiation on the gas-phase and solid-state chemistries of the youngest objects in the Coronet cluster.We present Submillimeter Array (SMA) observations towards five Coronet cluster members, thus complementing previous observations at 1.3 mm (Lindberg & Jørgensen 2012) by increasing the number of detected chemical species in this region.To probe the CH 3 OH large-scale emission, an APEX 150 ′′ ×150 ′′ map of the CH 3 OH J K = 5 K − 4 K Q-branch is also presented.The column densities of H 2 O and CH 3 OH ice are taken from Boogert et al. (2008).Furthermore, the dependence of gas-to-ice ratios on the physical conditions are addressed, by determining gas-to-ice ratios of CH 3 OH in this strongly irradiated cluster and comparing them with ratios obtained towards less irradiated star-forming regions (Perotti et al. 2020(Perotti et al. , 2021)).
The paper is laid out as follows.Section 2 describes both the SMA and APEX observational setups, and the data reduction strategy.In Section 3, we present our key observational results, while the variations on the ice and gas abundance are analysed in Section 4. Section 5 discusses the determined CH 3 OH gas-to-ice ratios and Section 6 summarises our conclusions and lists avenues for future studies.

Observations
The Coronet cluster was observed with the SMA (Ho et al. 2004) on February 25, 2020, April 28, 30, and May 19, 31, 2021 in the compact configuration, resulting in projected baselines between 17−72 m.The region was covered by one pointing centered on IRS7B with coordinates α J2000 = 19 h 01 m 56 s .40,δ J2000 = −36 • 57 ′ 27 ′′ .00 (Fig. 2).The SWARM correlator provided a total frequency coverage of 36 GHz, with the lower sideband covering frequencies from 210.3 to 228.3 GHz and upper sideband from 230.3 to 248.3 GHz.The spectral resolution was 0.6 MHz corresponding to 0.7 km s −1 .
Data calibration and imaging were performed with the CASA v. 5.7.2 package 1 (McMullin et al. 2007; Bean  et al. 2022).The complex gains were calibrated through observations of the quasars 1924-292 and 1957-387 and the bandpass through observations of the bright quasar 3C279.The overall flux calibration was carried out through observations of Vesta and Callisto.Imaging was done using the tclean algorithm and a Briggs weighting with a robust parameter of 0.5.The typical synthesised beam size of the SMA data set was 4 ′′ .9 × 2 ′′ .6 (400 − 700 AU).The position angle (PA) for CH 3 OH, our main molecule of interest, at 241.791 GHz was -8.2 • .The PA varies of ≈ 4.5 • across the covered frequencies.Notes. (a) The spectroscopic data are from Xu et al. (2008).They were taken from the Cologne Database for Molecular Spectroscopy (CDMS; Müller et al. 2001Müller et al. , 2005;;Endres et al. 2016) and the Jet Propulsion Laboratory catalog (Pickett et al. 1998). (b) Calculated using a collisional temperature of 30 K and collisional rates from Rabli & Flower (2010).The latter were taken from the Leiden Atomic and Molecular Database (LAMDA; Schöier et al. 2005).
Additional information on the CH 3 OH large-scale emission towards the Coronet cluster was provided by observations with APEX (Güsten et al. 2006) during the nights of October 25 − November 4, 2020.The pointing was the same as reported for the SMA observations.The achieved map size is 150 ′′ × 150 ′′ , fully covering the region observed by the SMA.The frequencies probed by the APEX observations are in the range 235.1 − 243 GHz, corresponding to the SMA upper side band observations, with a frequency resolution of 0.061 MHz (0.076 km s −1 ).The resulting beam-size of the observations is 27.4 ′′ for CH 3 OH.Data reduction was carried out with the GILDAS package CLASS2 .Subsequently, the reduced APEX and SMA datasets were combined in CASA v. 5.7.2 using the feathering technique, following the procedure presented in Appendix B.1. of Perotti et al. (2020).The combination of SMA and APEX short-spacing data resulted in mapping the outer regions of protostellar envelopes in the Coronet cluster on scales from ∼ 400 − 10 000 AU.

Results
Figure 2 displays the region targeted by the SMA and APEX observations.The positions of the SMA 1.3 mm continuum peaks coincides with the location of the young stellar objects in the Coronet cluster.The strongest emission is observed towards the two Class 0 sources SMM1C and IRS7B and the Class I objects SMM2 and IRS7A.Fainter emission north of the pre-stellar core candidate SMM1A (Chen & Arce 2010) is also seen.Apart from the continuum, line emission of seven species was identified (Table A.1).These results are presented and discussed in Appendix A (Figures A.1−A.3).In the remainder of this section we will uniquely focus on the observed methanol (CH 3 OH) emission, our molecule of interest.
Amongst the molecular species detected in the Coronet, methanol is the most suited for the main aim of this work: assessing the effect of external irradiation on the gas-phase and solid-state (ice) chemistries in the outer regions of protostellar envelopes.These are the regions  1 for their identification.
probed in the presented SMA and APEX observations, at radii r > 400 AU from the protostar.Here, gas-phase methanol is not thermally desorbed and it is a tracer of energetic input such as outflows releasing frozen CH 3 OH to the gas phase (e.g., Tychoniec et al. 2021).The latter is due to the fact that the most efficient methanol formation pathway in star-forming regions (A V > 9 mag) occurs on ice-coated dust grains (Watanabe & Kouchi 2002;Fuchs et al. 2009;Simons et al. 2020;Santos et al. 2022).Consequently, the observed non-thermal CH 3 OH emission in the outer regions of protostellar envelopes does not trace the bulk of the ices, and instead, it probes a fraction of CH 3 OH ice sputtered in outflows or photodesorbed (see Kristensen Article number, page 4 of 20 Figure 3 presents primary beam corrected integrated intensity (moment 0) maps of the CH 3 OH J = 5 0 − 4 0 A + line for the SMA, APEX, and the combination of interferometric and single-dish data (SMA+APEX).The combined moment 0 maps of the other CH 3 OH transitions belonging to the J = 5 K − 4 K Q-branch are displayed in Appendix A (Figure A.2). We specifically targeted this branch because its multiple transitions are conveniently observable in a narrow spectral range (0.2 GHz; Table 2) and they are associated to E u from 34.8 to 60.7 K, well below the typical CH 3 OH sublimation temperature as inferred from both observations and laboratory experiments (70 − 130 K; e.g., Kristensen et al. 2010;Penteado et al. 2017).The latter condition has to be satisfied in order to investigate methanol non-thermal desorption processes.Additionally, observations of CH 3 OH J = 5 K − 4 K Q-branch make it possible to directly compare our results with previous work on CH 3 OH gas-to-ice ratios in nearby star-forming regions using the same facilities, observational strategy, and spectral setups (Perotti et al. 2020(Perotti et al. , 2021)).
By looking at the SMA moment 0 map (Figure 3; panel a) it is clear that the interferometric observations do not sufficiently recover the spatially extended emission originating from the Coronet cluster.In contrast, they filter out ≈90% of the extended emission detected in the APEX data.Unlike for the other species detected in the SMA data, we here want to perform a quantitative analysis of the CH 3 OH emission therefore, we combine the SMA data with the APEX map to avoid underestimating the CH 3 OH column densities by up to one order of magnitude.
Similarly to the CH 3 OH J = 4 2 − 3 1 E transition (Fig. A.2; panel g), the CH 3 OH J = 5 0 − 4 0 A + emission is tracing the condensations associated to the prestellar core candidate SMM1A.The emission is extended in the APEX map (Fig. 3; panel b), and confined to two main regions, around SMM1A and to the east of SMM2.The peak intensity is observed also in this case in the vicinity of SMM1A.The same pattern is visible in the SMA + APEX moment 0 map (Fig. 3; panel c), but some CH 3 OH emission is also present at the SMM2, IRS7A, and IRS7B source positions although approximately a factor of 4 fainter compared to the peak position.
The CH 3 OH column densities towards the Coronet cluster members were determined from the integrated line intensities of the feathered SMA + APEX maps for all five CH 3 OH transitions belonging to the J = 5 K − 4 K branch ( For column density calculations, we followed two different treatments: a local thermodynamic equilibrium (LTE) analysis, assuming optically thin emission (Goldsmith & Langer 1999), and a non-LTE study employing the RADEX code (van der Tak et al. 2007).For both methods, a temperature of 30 K is assumed (Lindberg & Jørgensen 2012).Although the non-LTE analysis indicates that the CH 3 OH lines are not thermalised (see Figure B.3) for densities below 10 7 cm −3 , and four out of five transitions are optically thick (see Figure B.4), the estimated CH 3 OH column densities derived with both methods agree with each other (for typical envelope densities n H 2 between 10 5 and 10 6 cm −3 ; see Table B.2).More details about the LTE and non-LTE analysis are presented in Appendix B. Table 3 lists the resulting column densities calculated from the LTE analysis and using T rot = 30 K. The uncertainties were calculated from the spectral rms noise and 20% calibration uncertainty.

Linking methanol ice and gas
In this section, we analyse the gas-ice variations in the Coronet cluster.The H 2 O and CH 3 OH ice column densi-Article number, page 5 of 20 ties were determined by Boogert et al. (2008) from Spitzer IRS mid-infrared spectra as part of the Spitzer Legacy Program "From Molecular Cores to Planet-Forming Disks" (c2d).For the gas-phase counterpart, we made use of CH 3 OH gas column densities calculated in Sect. 3 and reported in Table 3.

H 2 column densities
Prior to searching for gas-ice correlations, the physical structure of the Coronet cluster is investigated by producing an H 2 column density map of the targeted region.This step is required as fractional abundances instead of absolute column densities need to be compared when studying gas-ice variations.This is due to the fact that gas and ice observations, by their very nature, are tracing different spatial scales, and therefore, a comparison between absolute values can lead to misinterpretations (e.g., Noble et al. 2017).At the same time it important to note that the H 2 column density calculated below provides the total H 2 column density -and hence it traces regions where methanol is in the gas phase as well in the ices.As a result, the CH 3 OH abundances presented in this work have to be interpreted as "average" abundances.
Figure 4 shows the H 2 column density map of the Coronet cluster obtained by making use of SCUBA submillimeter continuum maps at 850 µm from Nutter et al. (2005).A detailed description of the formalism adopted to obtain the H 2 column density from SCUBA maps is provided in Appendix C of Perotti et al. (2020).Briefly, in an optically thin thermal dust emission regime, the strength of the submillimeter radiation depends on the column density (N), the dust temperature (T) and the opacity (κ ν ) (Kauffmann et al. 2008).The inserted value for the opacity per unit dust+gas mass is 0.0182 cm 2 g −1 ("OH5 dust"; Ossenkopf & Henning 1994).The adopted value for the dust temperature is 30 K, based on previous observational studies of the Coronet and radiative transfer models by Lindberg & Jørgensen (2012).
The youngest cluster members are all situated in the densest areas of the targeted region (Fig. 4).The apex of the estimated H 2 column density lies to the south-east of R CrA, and in particular at the IRS7B, SMM1A, SMM1C and IRS7A positions (N H 2 > 1.45×10 23 cm −2 ).SMM2, IRS1, IRS5N and IRS5A are located in slightly less dense regions (0.5×10 23 cm −2 < N H 2 < 1.25×10 23 cm −2 ).The measured H 2 column densities and their uncertainties for individual sources are listed in Table 3.

Gas-ice variations
The analysis of gas-ice variations in the Coronet is addressed by comparing fractional abundances (X) of gas and ice species relative to H 2 (Table 3).In addition, this section also provides an overview of gas-ice variations in the Coronet with respect to two other nearby low-mass starforming clusters: Serpens SVS 4 and Orion B35A, which have been observed with the same facilities, angular resolution, and receiver settings adopted for observations of the Coronet to assure a meaningful comparison and reduce observational bias (see Perotti et al. 2020Perotti et al. , 2021)).Fig. 5 displays CH 3 OH gas abundances as function of H 2 O (a) and CH 3 OH ice (b).A correlation between CH 3 OH gas and H 2 O ice for the three star-forming regions is seen in panel a with the Coronet cluster showing gas-ice abundances as low as the Orion B35A cloud.The lower methanol gas abundances in the Coronet and in Orion B35A may be due to reduced ice mantle formation and enhanced photodissociation of methanol molecules upon desorption in these two regions with stronger UV field (see Section 5 for more details).The spread among the H 2 O abundances for the Coronet cluster and Orion B35A is limited, spanning less than a factor of 2.6 including the uncertainties.The same applies to the CH 3 OH gas abundances.In contrast, Serpens SVS 4 exhibits the highest values compared to the other two regions and the largest spread, with CH 3 OH gas abundances ranging up to ∼ 20 × 10 −8 .One outlier is identified, Serpens SVS 4−12, which is characterized by the lowest CH 3 OH gas and highest H 2 O ice abundances relative to the other Serpens SVS 4 protostars.This result is not surprising since this is the most extincted ob-ject of all targeted sources (A V ∼ 95 mag; Pontoppidan et al. 2004) with the lowest S/N (see Perotti et al. 2020 for a detailed discussion on Serpens SVS 4−12).Fig. 5 b compares CH 3 OH gas and ice abundances.In contrast to panel a, no clear trend is seen.However, when Serpens SVS 4−12 is excluded from the analysis it is possible to identify three different groups corresponding to the three regions: i) Orion B35A with the lowest measurements of both CH 3 OH in the gas and in the solid state; ii) Serpens SVS 4 with intermediate to high values of CH 3 OH ice and the highest gas abundances, and finally iii) the Coronet cluster, with some of the highest CH 3 OH ice but surprisingly low CH 3 OH gas abundances.The observed behaviour for the Coronet cluster is likely due to CH 3 OH destruction in the gas phase as a consequence of strong external irradiation (Lindberg & Jørgensen 2012).
Figure 6 shows H 2 O (a) and CH 3 OH (b) ice column densities (N) as function of cloud visual extinction (A V ) for lines of sights in the three nearby star-forming clusters.The cloud extinction estimates were taken from Pontoppidan et al. (2004) for Serpens SVS 4, from Perotti et al. (2021) for Orion B35A, and from Alves et al. (2014) for the Coronet cluster, then converted to visual extinction using the conversion factors (3.6 for A J , 5.55 for A H and 8.33 for A K ) from Weingartner & Draine (2001) for RV = 5.5 for the dense ISM.
The distribution of data points in panels a and b shows a positive trend for Serpens SVS 4 and Orion B35A, indicating that to higher visual extinctions correspond higher columns of H 2 O (a) and CH 3 OH (b).This observation is particularly valid for H 2 O but it also applies to CH 3 OH Article number, page 7 of 20 The gray lines represent linear fits to the detections.We note that the visual extinction values obtained for Corona Australis have to be taken with care as their determination is hampered by the lack of background stars towards the densest region of the Coronet cluster (see Section 4.2).The data for Serpens SVS 4 and Orion B35A were taken from Perotti et al. (2020) and Perotti et al. (2021), respectively.
when an evaluation is done per star-forming region.The opposite behaviour is found for the Coronet cluster members, for both ice species.This negative trend could simply reflect inaccurate extinction values due to the lack of background stars in the densest area of the Corona Australis complex, where the Coronet cluster members are located (see Section 2 of Alves et al. (2014) for more details).

Discussion
In this study, the CH 3 OH chemical behavior has been investigated with millimetric and infrared facilities.This provides direct observational constraints on the CH 3 OH gas-to-ice ratio in protostellar envelopes and on its dependency on the physical conditions of star-forming regions.Laboratory experiments predict that CH 3 OH ice in cold dark clouds is non-thermally desorbed with an efficiency spanning over approximately three orders of magnitude (10 −6 − 10 −3 molecules/photon; Öberg et al. 2009b;Bertin et al. 2016;Cruz-Diaz et al. 2016;Martín-Doménech et al. 2016).This range is expected to increase up to ∼10 −2 molecules/photon (Basalgète et al. 2021a,b), as X-rays emitted from the central YSO strongly impact the abundances of CH 3 OH in protostellar envelopes (Notsu et al. 2021).Generally, the majority of non-thermally desorbed CH 3 OH molecules fragment during the desorption and a fraction of those recombine (e.g., Bertin et al. 2016).
The CH 3 OH non-thermal desorption efficiency and by inference the gas-to-ice ratio is highly dependent on the ice structure and composition (pure CH 3 OH versus CH 3 OH mixed with CO molecules), as well as on the temperature, photon energy and flux (e.g., Öberg 2016;Carrascosa et al. 2023).To support laboratory experiments and computations, in this paper we provide observational measurements of the CH 3 OH gas-to-ice ratio in the outer regions of protostellar envelopes.
Our targeted regions (Serpens SVS 4, Orion B35A and Corona Australis Coronet) share a number of similarities: e.g., they are low-mass star forming regions, they are clustered and finally, they are affected by the presence of outflows and/or Herbig-Haro objects.However, they show distinct physical conditions for instance, the Serpens SVS 4 cluster is influenced by the presence of the Class 0 binary SMM4 (Pontoppidan et al. 2004), whereas B35A is affected by the nearby high-mass star λ Orionis (Reipurth & Friberg 2021) and the Coronet is strongly irradiated by the Herbig Ae/Be star R CrA (Lindberg & Jørgensen 2012).B35A is exposed to an interstellar radiation field χ ISRF of 34 (Wolfire et al. 1989) 3 enhanced by the neighbouring λ Orionis (Dolan & Mathieu 2002), whereas the radiation field in the Coronet cluster has been estimated to approximately χ ISRF ∼750 to account for the high fluxes at millimeter wavelengths (Lindberg & Jørgensen 2012).
In addition, the molecular clouds in which our targeted regions are located do not share a common formation history.The ongoing low-mass star formation in Serpens Main might have been triggered by cloud-cloud collisions (Duarte-Cabral et al. 2010) or compression from a shock wave from a supernova.However, there is no clear evidence that a nearby supernova ever occurred (Herczeg et al. 2019).In contrast to Serpens, the formation of the ring constituting the λ Orionis region is attributed to a supernova explosion that occurred roughly 1−6 Myr ago (Dolan & Mathieu 1999, 2002;Kounkel 2020).The low-mass star formation in B35A was likely generated by the presence of neighboring massive stars and their stellar winds (Barrado et al. 2018).Finally, star-formation in Corona Australis has been supposedly promoted by a high-velocity cloud impact onto the Galactic plane (Neuhäuser & Forbrich 2008) or by the expansion of the UpperCenLupus (UCL) superbubble (Mamajek et al. 2002).The fact that the three regions likely did not form in the same way implies different initial physical conditions for the production of CH 3 OH.
Figure 7 illustrates the distribution of the CH 3 OH gasto-ice ratios (N gas /N ice ) towards low-mass protostars located in the different molecular clouds described above.The average CH 3 OH gas-to-ice ratio (1.2×10 −4 ) determined from millimetric (single-dish) and infrared measurements by Öberg et al. (2009a) for four Class 0/I sources in Perseus, Taurus and Serpens is over-plotted.A de- Corona Australis Serpens Orion Fig. 7: CH 3 OH gas-to-ice ratios (N gas /N ice ) towards lowmass protostars in the Coronet cluster in Corona Australis (triangles).The square and the circle symbolize the average ratios for ten sources in Serpens SVS 4 and four sources in Orion B35A, respectively.The green shaded area corresponds to the interval of gas-to-ice ratios observed for the ten sources in the Serpens SVS 4 cluster.The dotted line indicates the average value for four sources from Öberg et al. (2009a).Upper and lower limits are marked as arrows.
tailed comparison between the gas-to-ice ratio determined by Öberg et al. (2009a) and the ratios calculated for the sources in the Serpens SVS 4 and the Orion B35A cloud is provided in Perotti et al. (2021).In this section we focus on the ratios obtained for Corona Australis and their comparison with the values determined for the two other star-forming regions.
The distribution of CH 3 OH gas-to-ice ratios towards the Coronet cluster covers approximately one order of magnitude.The uncertainty on the determination of the CH 3 OH ice column densities for HH100 IRS1 and IRS7A (Boogert et al. 2008) results in lower limits for their gasto-ice ratios, consequently we can not provide solid comparisons for these two sources.In contrast, the values obtained for IRS5A and IRS7B fall in the range of gas-to-ice ratios previously obtained for the protostars located in the Serpens SVS 4 cluster.
The CH 3 OH gas-to-ice ratios constrained from gas and ice observations of a sample of protostellar envelopes (Figure 7) do not point to one value but, instead, to a distribution (1.2×10 −4 to 3.1×10 −3 ) which validates previous observational studies (e.g., Öberg et al. 2009a;Perotti et al. 2020).This distribution agrees with a more complex scenario for the CH 3 OH desorption elucidated in the laboratory in which a large fraction of CH 3 OH molecules do not desorb intact, but instead fragment and eventually recombine during and after the desorption (e.g., Bertin et al. 2016) leading to CH 3 OH gas-to-ice ratios lower than 10 −3 .
A second interesting observation is that the CH 3 OH gas-to-ice ratios fall in a fairly narrow range; this suggests that the CH 3 OH chemistry at play in the cold outer regions of protostellar envelopes belonging to different low-mass star-forming regions is relatively independent of variations in the physical conditions.We note that a similar conclusion could not be tested for the innermost envelope regions where thermal desorption dominates and young disks are present; our observations are simply not sensitive to these regions.Our analysis implicitly indicates that the CH 3 OH chemistry in the outer envelope regions is set to a large degree already during the prestellar stage.A related result came to light when comparing abundances of complex organic molecules observed in distinct starforming environments (e.g., Galactic disc versus Galactic center, high-mass versus low-mass star-forming region; Jørgensen et al. 2020;Coletta et al. 2020;Yang et al. 2021;Nazari et al. 2022a).A larger sample is however necessary to prove this statement, and additional observations, laboratory experiments and modelling efforts are required to elucidate further the impact of the physical evolution of protostars on their chemical signatures and vice versa.

Conclusions
We present 1.3 mm SMA and APEX observations towards the Coronet cluster in Corona Australis, a unique astrochemical laboratory to study the effect of external irradiation on the chemical and physical evolution of young protostars.In addition, we determined CH 3 OH gas-phase abundances and compared those with CH 3 OH ice abundances from Boogert et al. (2008) to directly constrain the CH 3 OH gas-to-ice ratio in cold protostellar envelopes.Our key findings are: -We find a positive trend between H 2 O ice and CH 3 OH gas abundances in Serpens SVS 4, Coronet, and Orion B35A with the latter two regions showing the lowest gas abundances.This result is attributed to reduced ice mantle formation and substantial destruction of methanol gas molecules in these two regions characterized by stronger external UV field.-The distribution of CH 3 OH gas-to-ice ratios in the Coronet constrained from millimetric and infrared observations spans over one order of magnitude and reinforces previous observational constraints on the CH 3 OH gas-to-ice ratio.-Similarities are found between CH 3 OH gas-to-ice ratios determined in different low-mass star forming regions (Serpens SVS 4, Orion B35A and Coronet), characterized by distinct physical conditions and formation histories.This result suggests that the CH 3 OH chemistry -in the outer regions of low-mass envelopesis relatively independent of variations in the physical conditions and thus it is set to a large degree during the prestellar stage.
Multiple avenues for future studies can be pursued.Comparisons of gas and ice abundances of key species are currently limited by the low number of infrared surveys.Current infrared facilities such as the James Webb Space Telescope are now routinely probing interstellar ices towards background stars (McClure et al. 2023), pre-stellar cores, proto-stellar envelopes (Yang et al. 2022) and near edge-on discs (Sturm et al. 2023), thus significantly increasing the number of studied regions.Therefore, it will be feasible to search for gas-ice trends found in other regions, improve our constraints on the CH 3 OH gas-to-ice ratios, and ultimately determine ratios for a larger set of major Article number, page 9 of 20 interstellar molecules.This approach will provide important feedback on the interactions between ice and gas material during its journey from the molecular cloud to the disc and on the impact of the physical conditions on the physico-chemical evolution of protostars.In this context, the Corona Australis complex still has much to teach us, hosting one of the youngest population of protostars observed so far.Notes. (a) From the Cologne Database for Molecular Spectroscopy (CDMS; Müller et al. 2001Müller et al. , 2005;;Endres et al. 2016) and the Jet Propulsion Laboratory catalog (Pickett et al. 1998).The CH 3 OH transitions belonging to the CH 3 OH J K = 5 K − 4 K Q-branch are also detected in the APEX data.
Article number, page 13 of 20     3.Only transitions belonging to the J = 5 K − 4 K Q-branch have been considered.
Article number, page 17 of 20     Article number, page 20 of 20

Fig. 1 :
Fig. 1: Three-color image of the Coronet cluster (WISE 3.4 µm (blue), 4.6 µm (green), and 12 µm (red) bands; Wright et al. 2010) overlaid with SCUBA 850 µm density flux Nutter et al. (2005); contours are in decreasing steps starting at the peak flux (3.7 Jy beam −1 ) and subtracting 30% from the previous level.The white stars mark the positions of Class 0/I YSOs in the R CrA/Coronet region (Peterson et al. 2011; Nutter et al. 2005), whereas the orange stars indicate the Herbig Ae/Be star R CrA and the T Tauri star T CrA.The pre-stellar core candidate SMM1A is indicated with an orange circle and the orange square represents the Herbig Haro object HH100.

Fig. 2 :
Fig.2: SMA continuum at 1.3 mm (contours) overlaid with SCUBA 850 µm density flux fromNutter et al. (2005).The contours start at 5σ and continue in steps of 5σ (σ = 18 mJy beam −1 ).The empty circle indicates the size of the SMA primary beam, whereas the SMA synthesised beam is shown with a white ellipse in the bottom left corner.The empty rectangle shows the map covered by APEX observations.The stars mark the position of the objects located in the Coronet cluster; refer to Table1for their identification.

Fig. 3 :
Fig. 3: Primary beam corrected integrated intensity maps for CH 3 OH J = 5 0 − 4 0 A + transition (E u = 34.8K) at 241.791 GHz detected by the SMA (a), by the APEX telescope (b), and in the combined interferometric SMA and single-dish APEX data (c).All lines are integrated between 6 and 12.5 km s −1 .Contours start at 5σ (σ SMA = 0.16 Jy beam −1 km s −1 , σ APEX = 2 Jy beam −1 km s −1 , σ SMA+APEX = 0.17 Jy beam −1 km s −1 ) and follow in steps of 5σ.The circular field-of-view corresponds to the primary beam of the SMA observations.The synthesised beams are displayed in white in the bottom left corner of each panel.
Figure A.2 and Table B.1).The spectra are shown in Figure B.1.

Fig. 4 :Fig. 5 :
Fig. 4: H 2 column density map of the R CrA region calculated from SCUBA dust emission maps at 850 µm by Nutter et al. (2005).Refer to Fig. 1 for a guide to the symbols of the Coronet cluster objects.

Fig. 6 :
Fig. 6: Relationship between H 2 O (a), CH 3 OH (b) ice column densities (N) and visual extinction (A V ) for lines of sights in nearby star-forming clusters: the Coronet in Corona Australis (triangles), Serpens SVS 4 (squares), and Orion B35A (circles).Upper limits are marked as arrows.The gray lines represent linear fits to the detections.We note that the visual extinction values obtained for Corona Australis have to be taken with care as their determination is hampered by the lack of background stars towards the densest region of the Coronet cluster (see Section 4.2).The data for Serpens SVS 4 and Orion B35A were taken fromPerotti et al. (2020) andPerotti et al. (2021), respectively.
Fig. B.2: Rotational diagrams of CH 3 OH for the Coronet cluster members.The solid line shows the fixed slope for T rot = 30 K. Error bars are for 1σ uncertainties.The column densities are reported in Table3.Only transitions belonging to the J = 5 K − 4 K Q-branch have been considered.
Fig. B.3: RADEX model results for the five CH 3 OH transitions (Table A.1) using T kin = 30 K. The cyan line indicates T ex = 30 K.

Table 1 :
Overview of the Coronet cluster objects.

Table 2 :
Identified methanol transitions in the SMA and APEX data sets.

Table A .
1: Identified molecular transitions in the SMA data set.

Table B .
2: Comparison between total CH 3 OH gas column densities (N) calculated using the LTE and the non-LTE analyses towards the Coronet cluster members.
Table B.1.The vertical line is the CH 3 OH column density calculated assuming LTE conditions and optically thin emission reported in Table 3.