Skip to Main Content

Certain particles in the atmosphere have the unique ability to change the properties of clouds by causing water droplets to freeze at higher temperatures than they would on their own. With this ability, these so-called ice nucleating particles can greatly affect the evolution of clouds, precipitation, and climate. Previous research has pointed to the burning of biomass such as in wildfires as a major source of atmospheric particles, sometimes including these rare and elusive ice nucleating particles, but this relationship between combustion and the release of ice nucleants has not been understood.

A new study recently published by Carnegie Mellon’s Center for Atmospheric Particle Studies aims to answer these questions about ice nucleating particles and their relationship with biomass combustion. Led by Ryan Sullivan, associate professor of chemistry and mechanical engineering, a team of researchers conducted extensive experiments on the emissions from authentic biomass fuels. They found that minerals from biomass burning are an unrecognized and important source of ice nucleating particles that can explain much of the freezing activity observed in wildfire smoke.

The team, including Ph.D. students Leif Jahn, Michael Polen, Lydia Jahl, and Thomas Brubaker, first considered preliminary evidence they obtained that revealed the ice nucleation ability of particles emitted from biomass burning—specifically aerosol—became stronger over time. This went against previous experiments in the field, which found that chemical aging degrades the ice nucleation ability of most particle types, or does not alter it.


The researchers hypothesized that this strengthened ability came from chemical changes to the black carbon soot particles in the aerosol. Soot particles have been thought to be the ice nucleants released by fuel combustion and the particles surfaces become more oxidized as they age. As the soot particles oxidized in the atmosphere, perhaps they became more hydrophilic, increasing their ice nucleation ability, since ice nucleation involves water molecules forming on the surface into an ice crystal embryo. 

“We did many experiments, and our experiments indicated that our original hypothesis was not right because the fuels that produced the most soot typically had the weakest ice nucleation properties, or none that we could measure,” said Sullivan. “So it didn’t look like soot was the explanation.” This gave them a critical clue that something other than graphitic soot was responsible for the ice nucleation they were measuring.

You might also like...

Still trying to explain why the aerosol’s ice nucleating properties grew after aging, Sullivan became curious about the ash remaining in the pan where they burned the fuels during their experiments. Via x-ray diffraction off the atoms that compose the ash, they found that the ash that had the strongest ice nucleation properties also had the most crystalline material in it. When they examined the tiny submicron aerosol particles using electron and x-ray microscopy, they also saw minerals in the samples that were the best ice nucleants. This was a key finding since the presence of crystalline minerals is known to drive ice nucleation ability, but this had not been explored in both biomass-burning aerosol and the ash that is left behind.

After collecting authentic biomass fuel samples from various national wildlife refuges, they conducted more experiments to explore how changes in the original fuel relate to differences in the freezing ability of the smoke emissions. They were able to link the production of these new minerals from biomass burning to higher levels of mineral-forming elements measured in some of the original fuels. They were also able to conclusively rule out black carbon soot particles as the source of the ice nucleants.

The atmospheric chemistry community had not focused much on minerals produced in biomass-burning aerosol because they are assumed to be from pre-existing soil particles or dust that had landed on the tree or plant and was then resuspended into the atmosphere during wildfires. But Sullivan and his team found that these minerals are actually produced from the combustion itself. If the fuel contains elements like silicon, iron, aluminum, and calcium, when burned, mineral-containing particles are created. Tall grass fuels tend to produce more ice nucleating particles than trees because they naturally contain more of the mineral-forming elements in them.

Our findings are a totally different perspective for the atmospheric chemistry community regarding the source of minerals in biomass-burning smoke.

Ryan Sullivan, Associate professor, Mechanical Engineering, Chemistry

Sullivan sees this as an example of the scientific method at work. Their original hypothesis that soot was the answer was supported by preliminary data and other literature studies, but their experimental data said something quite different. So, they developed different experiments and analysis methods to continue their investigation. This has been a five-year project and the primary focus of Sullivan’s National Science Foundation (NSF) Career Award.

“Our findings are a totally different perspective for the atmospheric chemistry community regarding the source of minerals in biomass-burning smoke,” he said. “They have helped to address long-standing uncertainties regarding the questions of why some biomass fuels create ice nucleating particles when they combust and others do not, what are the sources of the particles, and how will the evolve as they move through the atmosphere.”