Grantee Research Project Results

Final Report: Leveraging comprehensive organic oxidation experiments for the development of improved atmospheric chemical mechanisms

EPA Grant Number: R840005
Title: Leveraging comprehensive organic oxidation experiments for the development of improved atmospheric chemical mechanisms
Investigators:
Institution:
EPA Project Officer:
Project Period: August 1, 2020 through May 10, 2025
Project Amount: $799,667
RFA: Chemical Mechanisms to Address New Challenges in Air Quality Modeling (2019) RFA Text | Recipients Lists
Research Category: Watersheds , Endocrine Disruptors , Environmental Engineering , Air Quality and Air Toxics , Air

Objective:

The original goal of this work was the development of a systematic, general approach towards development/improvement of mechanisms (both explicit and reduced) for complex organic compounds, based on new laboratory datasets describing their oxidation chemistry, and in a way that conserves carbon and retains the organic species’ key chemical properties. Work centered on the detailed, systematic comparison of predictions from different mechanisms with results from comprehensive laboratory measurements of the evolving product distributions from a range of organic oxidation systems. These laboratory oxidation data was compared to predictions using key species, species classes, and chemical properties of the organics. Such results help identify possible mechanistic changes, as well as insights into key uncertainties in the experimental measurements themselves.

Summary/Accomplishments (Outputs/Outcomes):

Initial work involved detailed mechanism-measurement comparisons for a number of chemical systems (OH+trimethylbenzene, isoprene, n-butane, toluene, and α-pinene, all under high-NO conditions), using mechanisms generated from the GECKO mechanism generator. A number of comparison approaches were attempted, including an isomer-by-isomer comparison, in which species of the same formula were compared; a modified isomer-by-isomer approach which takes into account decomposition processes within the mass spectrometers; and a binned approach that instead compares carbon oxidation state (OS_C) and carbon number (n_C). This last approach was found to provide the clearest comparison, and is least susceptible to decomposition processes or calibration errors.

Based on this comparison a “similarity metric” was developed, providing a quantitative standard by which changes to the GECKO mechanism (namely the underlying SARs) can be assessed, as function of reaction time (atmospheric aging). Overall agreement (amount and phase of the organics) was good, but the agreement in speciation was poor. To investigate the reason for this, we carried out tests of how changing SARs describing key branch points or rates (the rates of alkoxy decomposition, the rates of ozonolysis reactions, the rates of photolysis reactions, or the yield of organic nitrates from RO₂+NO reactions) affected this agreement. We also tried removing “manually-inputted” reactions. Unfortunately these changes did not lead to substantial improvements in mechanism-measurement agreement across all systems studied: while some changes led to improved agreement for individual systems, it worsened agreements for others.

Nonetheless, we were able to identify a major source of discrepancy between chamber measurements and mechanistic predictions, related to peroxyacyl nitrates (PANs, of formula RC(O)OONO₂). GECKO predicts high levels of these species, but measurements indicated very low concentrations (<10 ppbC) of species with formulas corresponding to PANs. It is unclear whether this difference is because GECKO overpredicts such species (with formation that is too fast or loss that is too slow), or because our mass spectrometric instruments do not measure them efficiently, or some combination of the two. If the problem is with detection, then PANs likely decompose in the instruments to smaller species, since overall carbon balance is good. Regardless, this discrepancy likely explains the large differences in speciation, and resolving it would represent an important step forward in measuring and/or modeling PAN species.

To more systematically compare laboratory measurements and mechanistic predictions, we developed a new system for data integration and mechanism development called Chemical Assessment of Mechanism Evaluation and Optimization (CAMEO) framework. CAMEO is a generalizable, modular platform designed to support the development and evaluation of chemical mechanisms for complex organic compounds. It integrates laboratory observations, model simulations, and mechanistic details within a unified relational database that links experimental data, simulation outputs, and detailed reaction pathways and products. The framework includes advanced querying capabilities and interactive visualization tools, enabling efficient analysis and direct comparison between experimental results and model predictions. CAMEO thus provides a transparent and comprehensive platform for mechanism evaluation and continuous improvement.

The system is implemented as a modular Python package, utilizing SQLite for relational data management and custom Python scripts for data workflow. It combines (1) scalable ETL (Extract, Transform, Load) pipelines for automated data ingestion, (2) a well-designed database structure that ensures chemical data quality through built-in validation rules, and (3) interactive visualization tools built within dashboard for exploring and comparing experimental and modeling results. A key feature currently in development is a customized, structure-based speciation algorithm that maps molecular observations to lumped mechanism surrogates for targeted applications. This design aims to balance technical flexibility with reproducibility, making the framework adaptable across diverse research setups.

We used CAMEO to systematically examine the ability of EPA’s new Community Regional Atmospheric Chemistry Multiphase Mechanism (CRACCM2) to describe the observed chemistry of α-pinene + OH, across a range of RO₂ chemistries. This evaluation required accurately representing the experimental conditions that influence simulation results. This included not only time-dependent concentrations (pinene, NO_x, O₃) but also chamber-related parameters, such as dilution rate, light intensity, and major wall effects. Some of these (light intensity, H₂O₂ injections, and NOₓ offgassing) were not measured, so were estimated by conducting sensitivity simulations with varying values to obtain the closest matches. Such improved representation of chamber conditions led to substantially improved measurement-mechanism agreement. Overall the agreement for the various organic species was found to be reasonably good, across a range of RO₂reactivities, though some disagreements were seen for specific species (e.g., acetone) and species classes (peroxides, nitrates), highlighting potential future improvements to both measurements and mechanisms. We also began comparisons for a second chemical system, the OH-initiated oxidation of isoprene (C₅H₈), again across a range of RO₂ conditions.

Journal Articles:

No journal articles submitted with this report: View all 8 publications for this project

Supplemental Keywords:

Air quality, photooxidation, volatile chemical products, tropospheric ozone

Relevant Websites:

n/a

Progress and Final Reports:

Original Abstract

The perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.