Hydrocarbon-like and oxygenated organic aerosols in Pittsburghinsights into sources and processes of organic aerosols

Two-thirds of the organic aerosol floating over Pittsburgh isn't from tailpipes or smokestacks — it's made in the atmosphere itself, assembled from chemistry that happens after the pollution leaves the source. State that flat, and it sounds almost philosophical. However, Zhang, Worsnop, Canagaratna, and Jimenez turned it into a measurement. Achieving this required inventing a way to sort particles by their chemistry at the speed they were forming. Here's why that's harder than it sounds. Organic compounds typically make up twenty to fifty percent of fine particle mass in urban air. They influence how particles grow, how much sunlight they scatter, how toxic they are, and how they seed clouds. But a complete molecular picture is essentially out of reach — targeting over one hundred individual compounds usually accounts for only ten to twenty percent of the organic mass, because the rest is a sprawling zoo of thousands of molecules with wildly different properties. Therefore, researchers are forced to work in bulk, characterizing the chemistry by class rather than by compound. The instrument Zhang and colleagues used was the Aerodyne Aerosol Mass Spectrometer, or AMS. It works by thermally vaporizing particles at around six hundred degrees Celsius, then blasting the vapor with seventy electron volt electrons to ionize it. This process produces a mass spectrum — a fingerprint of fragments sorted by their mass-to-charge ratio. The issue is that electron ionization is aggressive. It shatters molecules into pieces, and multiple different compounds can produce the same fragment. The AMS provides a rich, quantitative, minute-by-minute signal, but it's inherently a bulk measurement. You're reading a palimpsest, not a single text. The key breakthrough in this paper was an algorithm to decode it. Zhang and colleagues used two specific mass-to-charge peaks as chemical tracers. Mass-to-charge fifty-seven, dominated by a C4H9 ion, is a marker of hydrocarbon-like material — the type of fragment you'd expect from diesel exhaust or lubricating oil. Mass-to-charge forty-four, dominated by carbon dioxide ions, is a marker of oxygenated material — the sort you'd get from organic acids and other heavily processed compounds. Using these two peaks as anchors, the team ran an iterative multivariate regression to extract the full mass spectra and time series of two distinct aerosol types: hydrocarbon-like organic aerosol, or HOA, and oxygenated organic aerosol, or OOA. Together, these two components explained more than ninety-nine percent of the variance in the measured organic mass. The Pittsburgh Air Quality Study provided the raw data, which included three weeks of continuous measurements at a fixed urban site during September two thousand two. What does HOA look like in the record? It resembles a commute. The HOA time series shows a sharp peak during the morning rush hour, then drops off as the planetary boundary layer rises through the morning and dilutes the surface air. It correlates with carbon monoxide with an r-squared of 0.73, tracks nitrogen oxides at 0.82, and tracks elemental carbon at 0.72 — all classic markers of combustion. The mass spectrum itself is dominated by ion series characteristic of long-chain hydrocarbons, consistent with vehicle exhaust and lubricant aerosol. HOA accounts for thirty-four percent of organic aerosol mass averaged across the campaign, with an organic mass-to-organic carbon ratio — the OM:OC ratio — of 1.2. That low ratio indicates the material is minimally oxidized, close to a pure hydrocarbon. Critically, Zhang and colleagues demonstrate that HOA aging on a timescale of hours is negligible — the HOA you see at noon is essentially the same material that came out of tailpipes that morning, not something chemically transformed. Size confirms the story. About a third of HOA mass is in the ultrafine mode, which consists of particles smaller than approximately one hundred nanometres. That's exactly where fresh combustion particles exist before they've had time to coagulate or grow. HOA dominates the particle population at the smallest sizes the instrument can detect. OOA presents a different story entirely. It averaged 2.93 micrograms per cubic metre across the campaign, compared to 1.48 for HOA — so OOA is roughly twice as abundant and more than twice as variable. Its diurnal profile is flat. There is no rush-hour spike or morning peak. The levels in the boundary layer and in air aloft are similar, indicating this aerosol didn't form locally that morning — it arrived, carried in from the broader region. OOA correlates with sulfate at an r-squared of 0.74 and tracks ammonium; both species form in the atmosphere over regional scales and are transported in. Its OM:OC ratio is 2.2, nearly double that of HOA — heavily oxidized and oxygen-rich material. Less than five percent of OOA mass is ultrafine; instead, it concentrates in the accumulation mode, consisting of particles larger than about 250 nanometres that have been aloft long enough to grow. In the afternoon, when the boundary layer is fully mixed and local combustion sources are diluted, OOA accounts for more than eighty percent of organic mass. One finding from this work might seem puzzling at first. Ozone and OOA essentially do not correlate — the r-squared is near zero across the full dataset. Since ozone drives the photochemistry that creates secondary organic aerosol, you might expect them to track each other. Zhang and colleagues explain why they do not: a large regional background of OOA overwhelms any local production signal. Sulfate turns out to be a much better surrogate for OOA than ozone in this dataset, precisely because both are regional species that accumulate and transport together. The clearest evidence that OOA is genuinely secondary — made in the air, not emitted directly — came from a single remarkable day. On September twelfth, two thousand two, conditions were ripe for nucleation: low aerosol loading, strong sunlight, and favorable photochemistry. New particles began forming at 8:10 in the morning. Zhang and colleagues tracked what happened in four stages throughout the day. In the nucleation phase, particle number surged. Then, through two growth periods, those particles grew from the nucleation mode into the Aitken mode, which is the size range the AMS can actually detect. And what grew on those particles? OOA. Ultrafine OOA increased steadily through the growth periods while HOA declined. By late afternoon, the size distributions of OOA and sulfate had converged — both showing the same multimodal structure, both tracking the condensational sink that the team calculated from particle size and assumed condensate properties. The rate of OOA production during the photochemically active window was about 0.19 micrograms per cubic metre per hour, while ozone was increasing at about five parts per billion per hour — yielding roughly thirty-eight nanograms of OOA per cubic metre per part per billion of ozone. This is a local, real-time photochemical production rate, observed in real-time. What makes this event so powerful as evidence is the mass spectrum. The OOA spectrum measured during the September twelfth event is nearly identical to the OOA spectrum averaged across the entire three-week campaign. If the freshly condensed secondary material on that day looks chemically the same as the OOA present on every other day, the most natural conclusion is that the same process operates throughout — that the background OOA is also largely secondary. Zhang and colleagues cross-checked their results against the EC/OC tracer method, an independent technique that uses the ratio of organic carbon to elemental carbon to partition primary and secondary contributions. The HOA and OOA estimates from the AMS algorithm agreed reasonably well: an r-squared of 0.69 for the primary comparison, 0.52 for the secondary comparison, and an r-squared of 0.87 with a slope of 1.01 plus or minus 0.11 for total carbon against an independent carbon analyzer. That convergence from two very different methods adds considerable weight to the overall picture. The practical implication is sharp. If two-thirds of the organic aerosol over a major American city is secondary — formed regionally from gas-phase precursors — then cutting tailpipe emissions alone won't solve the problem. You have to address the volatile organic compounds and nitrogen oxides that feed the photochemical machinery producing OOA across entire regions. The OM:OC ratios — 1.2 for HOA, 2.2 for OOA, and a bulk average near 1.8 — are practically important as well: atmospheric models need these numbers to convert measured particle mass into carbon mass, and getting them wrong propagates errors into everything downstream. The chemical fingerprinting approach pioneered in this work became a standard framework for understanding urban aerosol. It turns out, the morning rush hour is only a third of the story. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.