Novel approaches for quantifying the highly uncertain thermodynamics and kinetics of atmospheric gas-to-particle conversion

Lead Research Organisation: University of Manchester
Department Name: Earth Atmospheric and Env Sciences


Atmospheric aerosol particles, which can be both anthropogenic and biogenic in origin, remain a major uncertainty in the Earth system: they impact the climate by directly scattering and absorbing solar radiation (the direct effect), as well as regulating the properties of clouds (the indirect effect). On regional scales, aerosol particles are among the main pollutants deteriorating air quality, their impacts on both poorly quantified. Reducing these critical uncertainties requires accurate knowledge on the chemical composition of these particles, their concentrations and size as they are suspended in the atmosphere. Unfortunately, there are currently huge uncertainties in many fundamental parameters that are required to predict these evolving chemical and physical characteristics of aerosols. This inhibits us from ultimately understanding their true environmental impacts. A significant fraction of atmospheric aerosol particles are comprised of organic material (20-90% of particle mass). Unfortunately, this fraction could comprise thousands of, largely unidentified, compounds with a wide range of chemical properties. This, in essence, creates the uncertainties listed above. The specifics of these uncertainties are now discussed.

As aerosol particles reside in the atmosphere, condensation of low volatility organic compounds changes the amount and composition of condensed phase organic material, thus their climatic and health impacts. This condensation is highly dynamic and, presently, there are 3 fundamental restrictions in reconciling this behaviour from a single particle to wider scales:

1) It is common to regard aerosol particles as a simple liquid comprised of multiple components. However, it is becoming increasingly evident that atmospheric particles exist as viscous amorphous states, rather than simple liquid/solid mixtures. Partitioning between the gas and condensed phase is then kinetically limited in such amorphous states. Traditional aerosol models do not account for this. This adds significant uncertainty to predictions of gas/particle mass transfer as mixing timescales are ultimately governed by the diffusion coefficients of the aerosol constituents in the aerosol, which, on the other hand, are connected to the viscosity of the particulate matter. For typical aerosol sizes, the characteristic time for mixing could increase from a few milliseconds to hours or even days!

2) In addition to diffusivity and viscosity, the equilibrium vapour pressure of each aerosol constituent is largely determined by its pure component saturation vapour pressure, which depends on the molecular properties of the compound. Saturation vapour pressures of organic components are currently poorly known, particularly for the least volatile compounds. The uncertainty in this parameter is already known to introduce 4 orders of magnitude of uncertainty in predicted aerosol mass!

3) Finally, to assess the atmospheric importance of these phenomena, modelling approaches that treat the organic condensation/evaporation as a dynamic process and couple the gas phase transport to the condensed phase diffusion are urgently needed, although these remain almost non-existent.

Presently, there is a fundamental lack of data and modelling tools to resolve the importance of these topical issues. Whilst predictive techniques for viscosity, diffusivity and vapour pressure exist, they are developed for chemical engineering purposes and remain unevaluated for atmospheric science. In this proposal we aim to make new and novel measurements of the properties listed here aswell as evaluating/improving existing models. Developing a new kinetic model we will assess the sensitivity to these properties at the single particle level and compare with actual measurements of single particle growth using optical tweezer experiments. We will also develop simple parameterised schemes so that large scale models can assess the wider sensitivity to the climate.

Planned Impact

Although direct beneficiaries will be largely academic, results derived from this study will indirectly benefit policy driven end-users through through existing links between the PI and the UKCA Climate- Chemistry-Community-Aerosol model, a joint NCAS-Met Office programme funded by NCAS, GMR and DEFRA. Specifically the PI is tasked with developing reduced complexity thermodynamic models for use in large-scale schemes. These simplified schemes have to be informed with appropriate partitioning tools. This UKCA model is under joint development with the Met Office and so will be directly available to the most appropriate non- academic end-user in the UK. detailed discussion has already taken place about its inclusion in the non-UK ECHAM-HAM global model. Importantly, currently there exists no parameterization of the kinetic effects of diffusivity within the condensed phase in large-scale models. The PI will be able to directly assimilate these developments into the WRF-CHEM regional climate model developed at Manchester, the results made available to policy driven models through existings links described above.

It is crucial that tools are developed to ensure engagement with end users becomes sustainable beyond the duration of the project. To this end, a novel data portal will be constructed and hosted at The University of Manchester that provides access to continually updated peer reviewed data derived from this program and maintains consistent in-house data stewardship and continual support. Aimed at academic and non-academic end-users, this differs from services provided by current facilities such as BADC who compile data from field campaigns and operational model outputs for, for example, weather forecasts. In line with the new NERC data policy, all peer-reviewed data derived from this project will be made openly accessible with hosting of QA-ed and non QA-ed data clearly differentiated. Registered users will receive optional notification by RSS, email or SMS when a requested datum has been added (or its status changed, e.g. from non-QA to QA). All users will be able to download a catalogue of all data available. A frequently updated blog will announce new updates, and service notifications for registered users, regarding the progress of this project and data to be included in future revisions. We will also document new publications where appropriate. Using Research Computing Services (RCS) at Manchester, the data-portal will include a rigorous infrastructure: Hosting in a secure data centre, fully maintained back-up service, security and logging of transactions.

The incorporation of modeling tools improved in this project within an informatics suite developed under a separate NERC grant (NE/H002588/1), for which DOT is PI, can be tailored to provide a useful online teaching resource to engage end-users.


10 25 50
Description We have found that techniques used to measure aerosol phase state could force a given state and thus have huge implications for gas-particle mass transfer. This in itself cuased us to report the community on the dangers of choice of exerpimental method. We have also started work on using a new numerical method for modelling diffusion in complex systems.
Exploitation Route To better target avenues of research in an area that is gaining huge momentum and international focus.
Sectors Chemicals,Environment,Pharmaceuticals and Medical Biotechnology

Title UManSysProp 
Description UManSysProp is an online facility for calculating the properties of individual organic molecules, ensemble mixtures and aerosol particles. Built using open source chemical informatics, and currently hosted at the University of Manchester, the facilities are provided via: a browser-friendly web-interface. a programmer friendly JSON API that enables you to call our suite of tools from your own code. access to the source code GitHub respository for all predictive techniques provided on the site. 
Type Of Material Computer model/algorithm 
Year Produced 2015 
Provided To Others? Yes  
Impact The facility is too new for assessing any impact, through appropriate reference to our description paper. Nonetheless, we receive roughly 800 users per month on the site already and is being used as a focal point for a NERC international network grant starting this year [2016]. 
Title New online and open source facility for predicting the properties of organic molecules and mixtures. 
Description UManSysProp is an online facility for calculating the properties of individual organic molecules, ensemble mixtures and aerosol particles. Built using open source chemical informatics, and currently hosted at the University of Manchester, the facilities are provided here via: a browser-friendly web-interface. a programmer friendly JSON API that enables you to call our suite of tools from your own code. access to the source code behind all predictive techniques provided on the site. 
Type Of Technology Webtool/Application 
Year Produced 2016 
Impact There have been 1800 unique users within its first 9 months of deployment, including 4 references since the paper release.