What kinetic data do we need to improve indoor air models?

There are currently inadequacies in our representations of physical processes in indoor air chemistry models. We need to understand and parameterise these processes better if we want our model predictions to improve. This is important as models give us additional information in situations where measurements cannot be made and provide additional insight into chemical processing indoors. Having confidence in models also means that we can use them to explore the impacts of emerging pollutants or predict future IAQ problems. 

All chemistry models contain kinetic data and specifically for indoor air chemistry models, this includes rate coefficients and degradation pathways, as well as a description of how gases form particles indoors. If this information is absent or uncertain, it will impact the model predictions.

There are several good sources of information around indoor air chemistry models and what we still need to understand (e.g. our editorial, or Section 3 of the Summary Report that resulted from our WG1 activities). Section 2 of the Final report from WG2, also summarises what we can learn from outdoor air models that could potentially be useful for indoor environments. Morrison et al. (2017) discussed what a framework for indoor air chemistry modelling might look like, whilst Shiraiwa et al. (2019) reported on the development of a new modelling consortium for indoor environments called MOCCIE (MOdelling Consortium for the Chemistry of Indoor Environments): MOCCIE aims to connect models over a range of temporal (from seconds to days) and spatial (from molecular to room) scales. Recent studies have shown that large gains in our understanding are made when modellers are fully integrated into measurement campaigns (e.g. HOMEChem). Measured data can be used to evaluate models, whilst the improved models inform future measurements.

Many indoor air chemistry models use chemical mechanisms designed for outdoor air applications. Although chemical reactions indoors are often the same as outdoors, their rates of reaction may differ. For instance, photolysis is more important outdoors compared to indoors. However, the species that are important outdoors are not always relevant for indoors and vice versa. Terpenes are ubiquitous indoors because of their use in cleaning products and these species are also widely studied outdoors because they are emitted from vegetation. However, explicit degradation mechanisms are still only available for a handful of these species. 

For many common fragrance products, such as linalool, dihydromyrcenol, terpinene and geraniol, only the preliminary oxidation rate coefficients exist with little information about subsequent products or pathways. Similarly, detailed degradation schemes for squalene (C30H50) and many of its oxidation products (e.g. carbonyls and acids) are also absent beyond the preliminary oxidation steps. Even for explicit mechanisms such as the MCM with 20,000 reactions (Jenkin et al., 1997), relatively few of the rate coefficients have actually been measured experimentally. Most of them have been estimated using structure activity relationships and pathways can be uncertain.  Considerable uncertainty therefore exists with how these species are treated in indoor models, particularly after the first few oxidation steps.

Although there is a wealth of kinetic information that would be useful, based on our work, we suggest priortisation of the following:

• oxidation rate coefficients and degradation pathways for key fragrance compounds such as geraniol, camphene, dihydromyrcenol, linalool, citral, eucalyptol etc. 

• more chamber work investigating how air mixtures from different sources indoors age over time and for different conditions

• improved parameterisation for gas-to-particle partitioning, including a consideration of the role of highly oxygenated molecules (HOMs).

Modellers need to work closely with kineticists/chamber scientists to prioritise these suggested experimental activities. The new data should be used to update models, which should be evaluated against measured data of relevance (e.g. use of cleaning products that contain the new chemicals).

The rate coefficient measurements need to be made by experimentalists using flow tubes and GC-MS equipment in labs, whilst the degradation pathways, ageing studies and gas-to-particle investigations could be studied using chambers (see the WG2 report).