GR-CLUE: A simple tool to assess the validity of nanoparticle growth rate deduced from size distribution data

GR-CLUE is a tool for assessing if observed growth of ~sub-10 nm aerosol particles can be interpreted with conventional growth rate (GR) analysis.

• GR calculations are generally not valid for the very smallest nanoparticles or molecular clusters, as their growth involves stochastic effects that require modeling of the cluster size distribution dynamics by a discrete, molecule-by-molecule model.
• GR-CLUE uses high-resolution nanoparticle size distribution data to assess the threshold size range above which GR becomes a valid concept.
• GR is expected to become valid when the metric δ²:δ decreases to very low values of << 1.
• The principles of the applied metric and GR are discussed in the study by Olenius et al. (DOI:10.1038/s41598-018-32610-z; see the citation below).

Note that the metric is indicative:

• The assessed threshold sizes are not expected to be exact.
• Instead, the approximative size range and its trends are meaningful: the threshold sizes are expected to increase with e.g. (1) increasing molecular size, (2) increasing nanoparticle evaporation, i.e. less efficient particle formation, and (3) decreasing condensable vapor concentration.

Usage

The input is given in two .m files: one that contains info on the settings on e.g. fitting size distribution functions, and one that gives the size distribution data by loading and possible pre-processing of files. Example files "input_settings_example.m" and "input_data_example.m" are included, and contain detailed information on the settings. GR-CLUE is called by

gr_clue('input_settings_example.m','input_data_example.m')

This generates visualization of the metric δ²:δ and of fits to the size distribution data that are used to determine the metric. Note that the fits need to be examined: if they do not seem reasonable and/or give bad goodness-of-fit assessment (marked in the figures, if the fits are very bad), the results should be discarded.

• The threshold sizes can be determined at different times, e.g. at the so-called appearance times that are often used in GR analysis, over the whole time range, or at a steady state.
• For demonstration purposes, the synthetic example data, generated by a discrete cluster dynamics model, has a very high size resolution. In practice the resolution is typically lower and data may also involve more noise. In such case, examining (averaged) steady-state distributions may be a reasonable approach for assessing the discrete regime, while keeping in mind that it can vary under time-dependent conditions.
• The tool doesn't include assessment of any growth rates, as sophisticated methods are available for this (e.g. TREND (DOI:10.5194/acp-18-1307-2018); BAYROSOL (DOI:10.5194/gmd-14-3715-2021); in the absence of coagulation and scavenging effects also the so-called appearance time method).

Prerequisites

• Matlab

Citation

If you use the codes provided in this repository in any study, please cite at least one of the following works:

• This repository (https://github.com/tolenius/GR-CLUE)
• Olenius et al.: Robust metric for quantifying the importance of stochastic effects on nanoparticle growth, Sci. Rep. 8, 14160 (2018), DOI:10.1038/s41598-018-32610-z
• Kontkanen et al.: What controls the observed size-dependency of the growth rates of sub-10 nm atmospheric particles? Environ. Sci.: Atmos. 2, 449-468 (2022), DOI:10.1039/D1EA00103E

License

This project is licensed under the terms of the GNU General Public License v3.0, as provided with this repository.

Authors

Code developer: Tinja Olenius (tinja.olenius@alumni.helsinki.fi)