Interstellar cosmic dust is a tiny but influential fraction of the Galaxy. Two main families—silicon (silicon, oxygen, magnesium) and carbon dust (carbon, hydrogen)—are born separately, depending on the C/O ratio in the envelopes of old stars. Once in the interstellar medium, dust gets hit by supernova blast waves, accelerating grains to tens of km/s, and undergoes countless collisions. Yet observations stubbornly suggest the two types don’t mix: there is no reliable evidence for mixed grains. A puzzle emerges: what happens when two irreconcilable worlds collide?
The researchers recreated head-on nanograin impacts in a computer using molecular dynamics. An amorphous silicate grain of composition (Mg₂SiO₄)₈ and a hydrogenated amorphous carbon (HAC) grain C₉₀H₄₀, both about 1 nm in diameter, were hurled at each other at speeds from 0,8 to 11,3 km/s. To describe interatomic interactions with quantum-chemical accuracy, they applied a machine-learning potential MACE-MH‑1, pre-trained on millions of structures. A set of 110 independent simulations (10 random orientations per velocity) was evolved for 4,5 picoseconds with a 0,01 femtosecond timestep, tracking the formation of chemical bonds and fragments. In the microcanonical ensemble, law of conservation of energy holds, and the kinetic temperature of fragments was computed based on the equipartition theorem.
The collision picture splits into four regimes, like the settings of a mill’s grinding stones. At the lowest speeds (<1,5 km/s), grains bounce elastically—dust remains segregated. In the range 1,5–3,5 km/s, the sticking probability rises to 40%—aggregation begins. Between 3,5 and 7,5 km/s, sticking becomes almost guaranteed: grains merge, forming robust chemical bonds between the silicate and carbonaceous components. Only above 7,5 km/s does fragmentation kick in—the interstellar mill grinds matter into molecular crumbs. During shattering, the main product is carbon monoxide CO—up to 6,5 molecules per collision at 11,3 km/s. The carbonaceous grain spawns a cascade of hydrocarbons: from simple radicals CH, C₂H to long polyyne chains C₁₀H and even cyclic structures. Complex organic molecules (COMs) also appear—ketene, propenal, polyynic aldehydes, and their precursor HCO. The silicate part produces silicon oxides SiO and molecular silicates like MgSiO₃ and MgSiO₂. Overall, the products found in the simulations cover about 10% of known interstellar organic molecules.
This result, for the first time, links the mechanical processing of dust in supernova shocks to the observed richness of interstellar chemistry. Collisions between dissimilar nanograins act as a natural reactor, producing both simple tracer molecules (CO, SiO) and complex organics. Revising the fragmentation threshold (7,5 km/s instead of the canonical ~3 km/s) means that in some environments, dust can grow by sticking where destruction was previously expected. Moreover, during impact, silicate grains capture hydrogen, forming surface silanol groups, which enhances their catalytic activity.
Future simulations will extend the range of sizes (currently only nanometer-sized grains have been explored), compositions, and impact angles. This will refine the fragmentation speed for different dust families and reveal how often mixed aggregates and complex molecules form in real clouds. It will also be important to determine whether such molecules survive the harsh conditions of the interstellar medium and whether they can be incorporated into planets. In protoplanetary disk, where speeds rarely exceed 1 km/s, collisions may help dust grains stick together into planetesimals.
The results will affect models of chemical evolution of galaxies, the interpretation of observations of molecular clouds, and the understanding of the delivery of organics to protoplanetary disks.
Perform simulations for larger grains (up to tens of nanometers) and for non-head-on impacts, and assess the contribution of pure carbon grain collisions to the production of polycyclic aromatic hydrocarbons (PAHs).
The article touches on the problem of the origin of complex organic molecules in the interstellar medium and the role of dust in prebiotic chemistry, and also revises dust destruction thresholds, impacting models of galaxy evolution.
This section is addressed to the author of the paper and their colleagues. Neighbouring works were found by meaning-based search across our archive, not by word overlap. We do not review: we show the field around the work.
What the work does
The authors numerically simulated collisions of nanoscale silicate and carbonaceous dust grains using a machine-learned potential to describe interatomic interactions with quantum-chemical accuracy. Analyzing 110 simulations at speeds from 0.8 to 11.3 km/s, they identified four regimes—from elastic rebound to fragmentation—and observed the formation of mixed aggregates and a wide range of molecules, including CO, hydrocarbons, and complex organic compounds.
Where its strength is
This work is the first to directly link mechanical processing of dust in shock waves to the production of observed interstellar molecules, offering a concrete and quantitatively calculated pathway for synthesizing complex organics without surface chemistry.
Where it could go next
Worth trying
Why this may matter
If the proposed mechanism indeed contributes significantly to the interstellar organic budget, it could change our understanding of where and how quickly prebiotic molecules form in the Galaxy, and suggest new markers for astrochemical observations.
Nearby works in our archive
The closest work in the archive sits at 0.63 by meaning — the usual level for our corpus.
In the high-speed simulations, exotic pure carbon chains up to eight atoms long sometimes formed — molecules like l‑C₃ and C₅, previously found in the atmospheres of cool stars and in molecular clouds.