The high temperatures that prevail in high-mass star forming regions boost their gas-phase chemical complexity via thermal desorption of the complex organic molecules (COMs) trapped in dust-grain ice-mantles. These COMs are excellent diagnostic tools of the physical conditions and the structure of protostellar envelopes.

I will present the first insights of the SPARKS project, conducted with ALMA, which unveiled hundreds of compact objects, from low- to high-mass stars, of which about one quarter are surrounded by a dense and hot environment rich in COMs. The statistical study of this large sample of sources allows us to place meaningful constraints on the physical and chemical conditions required to form COMs in the interstellar medium, and to investigate molecular diversity towards star-forming regions.

Dense interstellar filaments and ridges are the preferred birthplace of stars. Here, we confront two rather different star forming filaments: The Musca filament which forms isolated low-mass stars and the DR21 ridge which forms massive star clusters.

Analyzing the Musca filament and its associated ambient cloud with APEX CO observations shows that the filament is continuously accreting mass from the ambient cloud. To understand the origin of these local dynamics, we studied the global kinematics of the Chamaeleon-Musca complex with NANTEN2 CO and GASS HI data. This points to a scenario where the star formation activity in this region is the result of large scale (~50 pc) converging HI flows. The bending of the magnetic field during the collision then explains the observed accretion dynamics around the Musca filament.

CO, [CI], and [CII] observations (IRAM 30m, SOFIA) towards the massive DR21 ridge indicate that gravity should dominate the DR21 cloud dynamics. However, the large scale kinematics also shows signatures, very similar to Musca, that point to the bending of the magnetic field associated with a collision. The formation of the DR21 ridge thus appears to be triggered by a cloud-cloud collision as well, but with different collision parameters than Musca. In contrast to Musca, gravity then takes over in the compressed DR21 cloud due to its high density.

During star formation the molecular gas undergoes significant chemical evolution leading to a molecular richness at the emergence of hot cores. The chemical formation pathways even for simpler molecules are debated. Based on the 870 micron ATLASGAL survey of the inner Galaxy, the SPARKS project discovered 6 massive clumps dominated by a single collapsing massive object down to 2000 au scales. This makes them relatively easy targets for single-dish observations to study the early warm-up phase chemistry leading to the appearance of hot cores. We performed a complete unbiased spectral survey covering the frequency range between 159GHz and 374GHz with the APEX telescope towards these sources. Towards the hot core precursor G328.25, our results show a molecular richness and we find more than 40 species (and 26 isotopologues) (Bouscasse et al. subm).  We could pin down their location within the envelope combining a detailed analysis of the line profiles and LTE modeling. Our results show a highly structured envelope, where we could distinguish between a warm compact inner region and a colder and more extended envelope. The sulfur chemistry shows not only high abundances compared to hot cores but also a large diversity of molecules and ions especially in the cold envelope. In addition, several deuterated molecules have been detected in the warm and cold gas of the envelope. We detect about 10 COMs in different parts of the envelope: nine originate from the warm envelope, and about half of them are also present in the cold gas phase. We also clearly identify COMs originating from accretion shocks imaged with ALMA towards this object. Based on the 26 isotopologues detected, we find isotopic ratios for carbon and sulfur at least two times smaller than expected. We explore to which extent the molecular composition can be used to put constrains on the chemical age of the source. A comparison to the molecular composition of the archetypical hot corino IRAS 16293 and a sample of radiatively heated hot cores suggests strong similarities for COMs between our object to the hot corino IRAS 16293. 

Turbulence plays a fundamental role in the evolution of molecular clouds as an effective energy transfert process from large scales to small scales. Energy is injected at scales larger than a kiloparsec by a variety of sources, and dissipated at small scales through viscous and ohmic processes. Our work focuses on this last part of the turbulence cascade where energy is dissipated into heat.

Although the viscous dissipation scale is ~10 au, the actual structures could be orders of magnitude smaller. Nevertheless, in the framework of the MIST project, we have started an observational study of a closeby molecular cloud (~110pc) located in the Chamaeleontis region observed with ACA in 12CO(1-0) with 13" spatial resolution. The main goal of this pathfinder study is to search for regions with high velocity gradients in which dissipative turbulence could be observed at a higher resolution with ALMA. We also complete this structural analysis by a statistical one. Using the pixel per pixel differences of the second moment order of the velocity field, we can detect high velocity gradient and unveil more filamentary structures of the clouds. To achieve our aims, we have developped a code written in Julia, using the  Principal Component Analysis, which allows to enhance the signal-to-noise ratio by significant factor.

The resulting observation datacube shows structures of large velocity gradient (up to 75km/s/pc). Our preliminary results also show multiple filamentary structures of about 20 mpc width and aspect ratios larger than ~10, the orientation of which being consistent with the projected magnetic fields. In this talk, we will present the present context of turbulence in diffuse molecular clouds, our methodology and preliminary results, and open with observational perspectives with ALMA and JWST.

CO2 is an abundant molecule in cometary and planetary atmospheres. Furthermore, it is also found in the interstellar medium. In those media, the local thermodynamic equilibrium conditions are rarely fullfilled. Thus, in order to correctly model the observations, there is a need for accurate collisional data for this molecule. This is the only way to derive CO2 abundance in such media.  

    Studying the CO2-He collision may be a good template for more complicated CO2 containing collisional systems. The rotational excitation of CO2 by helium was hence investigated. A new potential energy surface was calculated for this van der Waals complex using a Coupled Cluster method and a complete basis set extrapolation. The accuracy of the PES was then validated though comparison between theoretical and experimental spectroscopical data, such as bound state and pressure broadening. Revised collisional data, cross sections and rate coefficients, are provided in the 10 - 300K temperature range. Thanks to our new data, the abundance of CO2 may be re-evaluated in astrophysical media.

Using recursively the  dbscan (density-based spatial clustering of applications with noise)  algorithm to Spitzer detected YSOs coordinates and nearest neighbour statistics, we are able to retrieve significant spatial 2D multiscale structures in star forming regions that we connect either to primordial structures as issued from the fragmentation of natal molecular gas clump for the more condensed and small ones or to evolved dynamical structures for the largest ones. Besides that analysis and using Gaia data, we use the maximum likelihood estimation (MLE) approach to investigate at which conditions the depth of the nearest star forming regions (distance < 750pc) may be determined. We illustrate our results in two different SFRs:  Taurus & NGC2264. 

The complex filamentary patterns observed in the interstellar medium (ISM) are prime examples of highly non-Gaussian structures emerging from the non-linear interactions between a variety of physical processes at play (turbulence, gravity, magnetic fields, thermodynamics, …). Our understanding of how these structures form and evolve now largely relies on the statistical analysis of observations and their comparison with numerical simulations, in order to assess physical processes and their interactions, across a vast range of spatial scales.

The quantitive comparison of simulation results to sets of observational data therefore requires an adequate statistical description of non-Gaussian structures. I will present a set of novel tools that provide such a general ensemble of statistical descriptors, able to characterize the multi-scale geometry of images by encoding the couplings between oriented scales. I will present several applications of these "scattering transforms", from the description of interstellar emission maps in total intensity and polarization to their use in component separation.

Despite the increasing number of observations of several nitrogen hybrides molecules (NH, NH₂, NH₃,...) in the last decade, the nitrogen abundance in the interstellar medium (ISM) is still subject to debate. In particular, NH is involved in many reactions in a nitrogen chemical network. Accurately modelling the NH abundance in molecular clouds is then of high interest.

Density conditions in the ISM does not allow maintaining collisional processes large enough to maintain the local thermodynamic equilibrium (LTE) condition. Hence, the modelling of the observational spectra requires to take into account the competition between radiative and collisional processes. In the ISM, H₂ is the dominant collider but to the best of our knowledge, there are no collisional data for the NH-H₂ system. 

Calculations of fine structure resolved excitation cross sections and rate coefficients for the NH-H₂ collisional complex will be presented. The collisional data were computed from a new four dimensional potential energy surface, taking into account the orientation of H₂ over the inter molecular axis. Binding energy of the complex was calculated to check the validity of the PES through experimental data as a benchmark. Cross sections calculations are performed using the close-coupling approach for NH with both ortho-H₂ and para-H₂ partners. It appears a very different general behavior between the two partners. Calculations implying NH with ortho-H show a propensity rule in favor of $\Delta N = 1$ transitions whereas a propensity rules in favor of Delta N = 2 is seen for collisions with para-H₂. F-conserving transitions were found to be larger than F-changing transitions. We compare the new NH-para-H₂ rate coefficients with previous NH-He data, used as a proxy of NH-H₂ for astrophysical models. We find differences up to one order of magnitude. The new data are expected to improve abundances calculations for astrophysical applications.

The origin of stellar masses, arguably the most central question in star formation remain a major open issue in modern astrophysics \citep[see review by, e.g., ballesteros et al., 2020). The main goal of the ALMA-IMF Large Program (PIs Motte, Ginsburg, Louvet, \& Sanhueza) is to determine how the origin of the initial mass function (IMF) is in fact independent of cloud characteristics or not. Thanks to its unmatched angular resolution, sensitivity, image quality, and excellent frequency coverage, we used ALMA to survey 15 massive protoclusters, covering a wide variety of Galactic environments and evolutionary stages.
Our pilot study and preliminary ALMA-IMF results (Motte et al., 2018; Louvet et al., in prep) show that the mass distribution of cores (CMFs) in these typical yet extreme environments of the Milky Way present an excess of high-mass cores with respect to the canonical IMF (kroupa et al., 2013). A detailed study of two contiguous protoclusters which are at different evolutionary stages suggests that the CMF deviates from the canonical IMF form when and where a burst of star formation develops (Pouteau et al. in prep.). The CMF would thus strongly depend on environmental parameters and may lead to non-universal stellar IMFs and heterogenous, mass-segregated spatial distribution of stars. In addition, we compare the CMF of prestellar (i.e. starless) cores and protostellar (i.e. currently forming star by gravitational collapse) cores and prove that the relationship between the CMF and IMF should be revised because, in the case of  intermediate-to-high-mass star, the prestellar phase is evanescent if it exists at all (Nony et al. in prep.).

The ALMA-IMF Large Program therefore has the potential to transform our understanding of the IMF origin. It also provides the community with an unprecedented database with high legacy value for protocluster clouds (60 pc$^2$ at 2000 AU resolution), cores (about two thousand cores with 0.15-250 M$_{\odot}$ masses), many of which associated with outflows, filaments (hundreds of star-forming filaments), often tracing gas inflows, and hot cores (several tens of hot cores).

Dust plays a crucial role in numerous physical and chemical processes in the interstellar medium (ISM). Variations of physical conditions in the ISM (i.e. particle density and radiation field) trigger evolution of the dust properties (i.e optical properties, abundances, size distribution, composition) which strongly impact the gas. It is therefore important to understand how dust evolves with the local environment.

We study dust evolution, through its emission and scattering properties, in nearby photon- dominated regions (PDRs) where the physical conditions vary widely but can be spatially resolved. We focus on the Horsehead Nebula a well-known PDR, which has been extensively studied. This moderately excited PDR (radiation field intensity corresponding to G0 ~100) has a nearly edge on geometry well suited to follow the impact of photons on dust grains, while the density increases towards the interior of the cloud. Furthermore, observations from the Herschel Space Observatory together with those from the Spitzer Space Telescope and the Hubble Space Telescope are available and provide us with a wealth of spatial and spectral information of dust and gas emission from the mid-IR to the submillimeter spectral ranges. To model the dust emission and scattering across this PDR, we use the THEMIS dust model [1], included in the dust physics numerical tool DustEM [2]. A 3D continuum radiative transfer code, SOC [3], is then used to assess dust emission and scattering at different positions inside the PDR. Considering first dust populations from the diffuse interstellar medium [1] with fixed abundances across the cloud we show that the short and long wavelength dust emission cannot be simultaneously matched [4].

We propose the following scenario. In the Inner part of the Horsehead, the grains are aggregates with or without ice mantles and are grains typical of dense clouds environment [5]. The radiative feedback of the binary stars on the Horsehead edge photo-processes the coagulated grains into smaller grains, similar to those of the diffuse ISM. However, the typical time-scale during which grains are photo-processed must be shorter than the typical time-scale associated with the photo- fragmentation of aggregates into smaller grains. Hence, small grains containing aromatic cycles does not have time to form which explains why the abundance of these grains is lower than in the diffuse ISM. Also, the increase in the minimum size of these grains may suggest that the smallest of these grains are photo-destructed and/or they don’t have time to form from photo-fragmentation of larger grains. The slope steepening of the power-law size distribution of these grains may suggest that photo- fragmentation of aggregates leads to the creation of small grains instead of large grains [4].

As the gas physics and chemistry are tightly connected to these dust properties, we used a model for atomic and molecular gas in PDRs, the Meudon PDR code, using diffuse ISM-like dust and Horsehead-like dust to study the influence of nano-grain depletion on the gas physics and chemistry,. We focus on the impact on photoelectric heating and H2 formation and, therefore, on the H2 gas lines. We show that the nano-grain depletion in the outer part of the Horsehead has a strong influence on several gas tracers that will be prominent in JWST observations of irradiated clouds. [6]

Références: 

• [1] Jones, A. P., Fanciullo, L., Köhler, M., et al., A&A, 558, A62 (2013)
• [2] Compiègne, M., Verstraete, L., Jones, A., et al., A&A, 525, A103 (2011)
• [3] Juvela M., A&A, 622, A79 (2019)
• [4] Schirmer, T. et al, A&A, 639, A144 (2020)
• [5] Köhler, M., Ysard, N., & Jones, A. P. A&A, 579, A15 (2015)
• [6] Schirmer, T. et al, accepted in A&A, arXiv:2104.08204 (2021)

The spatial distribution of YSOs might give some clues on the fragmentation process of molecular clouds. This work aims to investigate at which spatial scale gas clumps fragment to form groups of YSOs. Using the getsf procedure (Men’shchikov+21), we extracted clumps catalogs at various scales (5~30kAU) in Herschel maps of NGC2264. In addition, we include the class 0/I YSOs (Rapson+14) to perform spatial connection with those clumps. We then design a graph theory-based approach using specific metrics to characterise the properties of fragmentation within the gas component, in which clumps and YSOs are the nodes of a directed graph and the links define any inclusive relation between nested clumps and YSOs. Spatial distribution of structures in NGC2264 revealed that hierarchical nested fragmentation is primary observed in the central hub of NGC2264, and is highly productive of YSOs, whereas monolithic collapse occurs in the sparser regions with far less efficient to produce YSOs. Furthermore, in the 5~30kAU scale range, hierarchical fragmentation of clumps can be described by a binary fractal fragmentation each time clump's size is reduced by a factor of two.