When we pull a sample of urban air, we find a dense cocktail of fine particulate matter ($PM_{2.5}$). A massive portion of this mass consists of secondary aerosols—particles that were not emitted directly from a tailpipe or smokestack as solids, but instead formed in the atmosphere from the chemical transformation of gases.
Yet, when we try to simulate this process using state-of-the-art chemical transport models (CTMs), a glaring problem emerges: the models consistently predict far less aerosol mass than we actually measure in the real world. As noted by Cynthia Whaley et al., organic aerosols exhibit some of the largest simulation biases among short-lived climate forcers. This is not a minor tuning issue; it is a fundamental "mass gap." For example, Jun Zheng et al. discovered that the direct gas-phase condensation of sulfuric acid accounts for a mere 10% to 25% of the actual sulfate mass measured in urban fine particles.
To understand why our models are blind to the remaining mass, we have to look at what traditional simulations leave out: the hidden ingredients (precursors), the true nature of particles as liquid-phase chemical reactors, and the complex environmental engines that drive them.
Unaccounted Precursors: The Hidden Ingredients
A chemical model can only build what its ingredient list allows. Historically, emission inventories for air quality models focused heavily on highly volatile organic compounds (VOCs) originating from traditional combustion sources.
However, modern urban atmospheres are filled with a different class of gases. The authors of An updated modeling framework to simulate Los Angeles air quality – Part 1 demonstrate that standard chemical libraries lack pathways for non-combustion pollutants, specifically volatile chemical products (VCPs) like cleaning agents and solvents, alongside intermediate volatility organic compounds (IVOCs).
Because IVOCs and VCPs sit in a chemical sweet spot—less volatile than traditional VOCs but volatile enough to remain gaseous—they oxidize rapidly into less-oxidized oxygenated organic aerosols (lo-OOA). As Jianlin Hu et al. point out, completely omitting these intermediate precursor compounds creates an immediate deficit in simulated organic mass compared to real-world chemical mass balance estimates.
Tiny Chemical Reactors: Heterogeneous and Aqueous Chemistry
Even when a model includes the right ingredients, the physical rules it uses to simulate gas-to-particle conversion are often oversimplified. Traditional frameworks rely heavily on gas-phase oxidation followed by thermodynamic equilibrium partitioning. This approach assumes that gases gently condense onto a particle surface based almost entirely on their vapor pressure.
In reality, atmospheric aerosols do not just sit there absorbing gas; they act as complex, liquid-phase chemical reactors. Jian Xue et al. highlight that omitting these aqueous-phase chemical mechanisms prevents models from capturing the full complexity of aerosol formation under high-humidity or cloudy conditions.
When a gas dissolves into the water film surrounding an existing particle, it undergoes rapid, irreversible reactions that lock it into a solid or liquid state. For instance, Jianlin Hu et al. note that biogenic precursors like isoprene require specific atmospheric conditions, such as high particle acidity, to trigger reactive surface uptake—a process completely missed by simple equilibrium models.
Traditional Model:
Gas-Phase Precursor ──(Oxidation)──> Low-Volatility Gas ──(Condensation)──> Particle Surface
Aqueous/Heterogeneous Reality:
Gas-Phase Precursor ──(Dissolution)──> Liquid Aerosol Water ──(Acid-Catalyzed Reaction)──> Irreversible Particle Mass
This mechanism gap cuts both ways. CTMs can also overpredict certain gas-phase species if they miss these condensed-phase sinks. Zhenhao Ling et al. showed that standard models overpredict small dicarbonyl molecules like glyoxal because they fail to account for "heterogeneous sinks"—surface reactions that actually account for roughly 21% of total secondary organic aerosol mass in polluted environments. When these heterogeneous pathways are missing, models fail to simulate the explosive growth of haze episodes, a limitation documented by Ziyue Chen.
Multi-Stage Oxidation and Environmental Constraints
Capturing this chemistry is exceptionally difficult because the transformation of a gas into an aerosol is rarely a single-step reaction. As Shuai Wang et al. explain, secondary organic aerosol formation involves highly complex, multi-stage oxidation processes. Tracking these evolving chemical generations introduces severe uncertainties into transport models.
These kinetic pathways become even more uncertain under extreme environmental conditions. In high-latitude winter environments, high humidity often overrides cold temperatures to drive secondary aerosol production. Yet, as Yuan Cheng et al. observe, current modeling frameworks lack the precise kinetic parameters required to simulate chemical reactions occurring within supercooled aerosol water.
Finally, the entire process relies on the atmosphere's "chemical engine"—the availability of oxidants like hydroxyl (OH) radicals to kickstart the oxidation cascade. In winter or early morning hours when solar radiation is weak, standard ozone photolysis cannot produce enough OH radicals to explain observed aerosol growth. Yongchun Liu et al. demonstrate that the omission of traffic-related nitrous acid (HONO) photolysis starves models of a critical early-morning radical source, leaving them unable to replicate the rapid, real-world accumulation of secondary particles.
Key Foundational Work
Measurements of gaseous H 2 SO 4 by AP-ID-CIMS during CAREBeijing 2008 Campaign (2011) — contributed 3 concept elements to this direction.
Impact of nitrous acid chemistry on air quality modeling results over the Pearl River Delta region (2011) — contributed 3 concept elements to this direction.
Global budget of methanol: Constraints from atmospheric observations (2005) — contributed 1 concept elements to this direction.