The interstellar medium, the vast space between stars, is not empty. There, we can find diffuse gas that gathers into large molecular clouds, several parsec wide. Inside, the interplay between gravity and turbulence creates over-dense regions called "dense cores". If the gravity can overcome the support provided by the thermal pressure of the gas, the core collapses onto itself. The gas becomes hotter and denser as the core contracts. At some point, the conditions are extreme enough to allow for hydrogen atoms to start fusioning into helium. This reaction is exothermal and its rate increases with density and temperature. When the energy released by those reactions provides enough thermal pressure to support against the gravitational contraction, the system is at equilibrium and a star is formed.
Dense cores rotate slowly, a rotation that is accelerated as the core contracts. Some of the gas becomes supported by the centrifugal force against gravity and forms a disk. This is called a "protoplanetary disk" because that is the place of planet formation. The angular momentum is a variable that quantifies the rotation of a body. It is conserved in a closed system. However, observations show that the angular momentum in stellar systems is much smaller (by a factor 1000) than the angular momentum of dense cores, meaning that most of it is lost during the star formation process. This phenomenon can be explained by magnetic fields. The field threading the gas slows its rotation by a process called "magnetic braking". The efficiency of the braking depends on how the gas and the magnetic field are coupled. The gas is indeed composed of both charged particles, that are sensitive to magnetic forces, and neutral particles, that are not. Quantifying the magnetic braking therefore requires the knowledge of the abundance of those particles and the way they interact between each other. This framework is called "non-ideal MHD" (MagnetoHydroDynamics). My first work (Marchand et al. 2016) was the development of a chemical code to calculate the magnetic resistivities, which are used in non-ideal MHD simulations to quantify the magnetic braking. I have also developed a numerical scheme to implement the Hall Effect, a process of non-ideal MHD, in numerical simulations (Marchand et al. 2018, 2019). I have used those tools to perform simulations of star formation to quantify the contribution of non-ideal MHD effects in the regulation of angular momentum (typically for the formation of the protoplanetary disk) (Marchand et al. 2020).
Dust grains are agglomerates of molecules much larger that gaseous chemical species (typically larger than 5 nm). Despite representing only 1% of the mass of the gas, they play a major role in most processes involved in star formation, notably by becoming the seed of planets. They can indeed hold one to several electric charges, making them critical to non-ideal MHD. Knowing their size and abundance has been the challenge of the last few years. Indeed, it is difficult to constraint the size-distribution of grains with observations. In addition, grains collide with each other and grow in size or fragment. Some of my work involved calculating the evolution of the size-distribution of grains and their electric charge, in a star formation context (Marchand et al. 2021, 2022, 2023).
In particular, Complex Organic Molecules (COMs, molecules with carbon and at least 6 atoms), are thought to be the precursor of pre-biotic molecules, and are abundantly observed in protostellar systems. It is therefore critical to determine the chemical composition of protostars to constrain the chemical environment of planet formation and the potential formation of life. I have notably lead an observation project to make the chemical inventory of the envelope of the L1551 protostar (Marchand et al. 2024). The chemistry of protostar can also be explored theoretically. I have developed a code (APE, Marchand et al. 2025) that provides the physical conditions of star formation for chemical simulations, to compute maps of chemical abundances in protostellar systems. The code is also an interface to radiative transfer softwares, so as to produce synthetic (=simulated) observations. Those are critical to help predict and interpret the results of real observations.