Computational aerosol physics group (public)
What does our group do?
List of group members
Presentations and visitors
Computational aerosol physics group (restricted access)
Out of office
Quantum chemistry data (restricted access)
Work in progress
What does our group do?
In recent years, increasing interest has gathered around the properties and effects of atmospheric aerosols, i.e. solid or liquid particles suspended in atmospheric air.
Selected facts about aerosols:
- Aerosol particles are found in all parts of the Earth's atmosphere, both in polluted urban areas, and also naturally in remote areas, e.g. as a consequence of volcanic and biological activity.
- The size of aerosol particles range from small molecular clusters of less than 1 nanometer up to rain drops of more than 1 millimeter
- Aerosols can harm human health and reduce visibility%0a*An aerosols particle can both warm and cool the climate, their combined net effect is cooling
- Every cloud droplet has an aerosol particle as a seed, and especially the effect of aerosol%0a via cloud formation is perhaps the largest uncertainty in predicting climate change
- Up to 50-90%25 of the total particle number, and 30-50%25 of the cloud condensation nuclei, are produced by clustering from gas-phase molecule
Members of the computational aerosol physics group are working on deepening the theoretical understanding of the formation of the very smallest molecular clusters in the atmosphere. For this, we use a palette of different modeling approaches, each with strengths and weaknesses. Hereby, we are able to cover all interesting cluster sizes from only a few weakly bound molecules up to large, thermally stable clusters containing hundreds of molecules. These models cover a wide spectrum of different levels; from inexpensive, hard sphere collision theories and bulk liquid approximations to high level quantum mechanical treatment of each electron. Currently, our most active areas of research are
and are all briefly described in the following. We have projects suited for B.Sc. or M.Sc. theses, but also welcome new ideas. If You are interested in our research and would like to work with us, please feel free to contact Prof. Hanna Vehkamäki, or any other member or our group.Quantum Based Thermodynamics
Contact person: Kenty Ortega frame width=800px http://www.atm.helsinki.fi/images/stories/webfigure_v2.jpg|Figure: The homogeneous vapor-liquid nucleation of the water/ammonia/dimethylamine/sulphuric acid system. Experimental measurements and computational methods based on classical physics are not able to accurately describe the first steps of aerosol formation. On the sub-nanometer scale where these initial steps occur, there is no substitute for quantum mechanics. We use methods based on numerical solutions of Schrödinger's equation (with various approximations) to explore the structures and formation energetics of small molecular clusters. These quantum chemistry methods yield thermodynamic data which governs cluster evaporation rates. This data may be used directly in our model of cluster growth and dynamics, or may be parametrized for use combined with macroscopic classical nucleation theory. Some of our future quantum chemistry related projects are:
- Studying the role of oxidized organic compounds in the first steps of cluster formation.
Related recent publications from the group:
\[[http://www.atmos-chem-phys.net/12/225/2012/acp-12-225-2012.pdf|Ortega et al., Atmos. Chem. Phys., 12, 225-235, 2012]\]
\[[http://www.atmos-chem-phys-discuss.net/11/30853/2011/acpd-11-30853-2011.pdf|Kupiainen et al., Atmos. Chem. Phys. Discuss., 11, 30853-30875, 2011]\]
\[[http://www.atm.helsinki.fi/%257Ehvehkama/publica/Toivola_THEOCHEM2011.pdf|Toivola et al., Computational and Theoretical Chemistry, 967, 219-225, 2011.]\]
\[[http://www.atm.helsinki.fi/%257Ehvehkama/publica/Ruusuvuori_THEOCHEM2011.pdf|Ruusuvuori et al., Computational and Theoretical Chemistry, 966, 322-327, 2011.]\]
\[[http://www.atm.helsinki.fi/%257Ehvehkama/publica/kurten_acp_cims2011.pdf|Kurtén et al., Atmos. Chem. Phys., 11, 3007-3019, 2011.]\]
Back to the topAtmospheric Cluster Dynamics Code
Contact person: Kenty Ortega:flash http://www.atm.helsinki.fi/~ortega/QMvideo_001_1.swf width=600 height=400 From the quantum-chemistry based thermodynamic data we can estimate the relative stabilities of different molecular clusters in terms of their formation free energy. Using this information, we can then compute the rate at which clusters evaporate different molecules or break into smaller clusters. The evaporation rates can be used in Atmospheric Cluster Dynamics Code (ACDC), a kinetic code developed in our group. ACDC generates the birth-death equations governing the collision and evaporation processes within a set of clusters, and solves them numerically to find the time-dependent concentration of each cluster.We have lately been focusing on the formation of sulfuric acid – ammonia – amine clusters in conditions similar to various nucleation experiments and to atmospheric cluster formation. Ambient gas-phase base concentrations are often below the detection limit of even the most sophisticated instruments, but our simulations indicate that even such extremely low concentrations enhance sulfuric acid clustering significantly. Some of our future ACDC related projects are
- Parametrization of cluster formation by sulfuric acid and different bases for use in climate and air quality models
- Explaining what happens to clusters inside instruments while they are measured
- Explaining differences in the relation of particle formation rate versus sulfuric acid concentration in chamber experiments, flow tube experiments and field measurements
- Including (implicitly and/or explicitly) the effect of water inevitably attached to the clusters in atmospheric conditions and in most instruments
Related recent publications from the group:
\[[http://www.atmos-chem-phys.net/12/2345/2012/acp-12-2345-2012.pdf|McGrath et al., Atmos. Chem. Phys., 12, 2345-2355, 2012]\]
Back to the topChemistry of clusters
Contact person: Nicolai Bork frame width=500px%25 http://www.atm.helsinki.fi/~bork/Simuwiki/Fig_4_new.png|Transition states of the oxidation of SO'2' by O'3''-' as varying degree of hydration.The local environment inside or on the surface of any aerosol is very different from the free gas phase. Hence, a number of chemical reactions may be facilitated by the presence of aerosols, leading to a changing composition of the gas phase or the aerosol itself. Until now, experimental studies of aerosol induced chemistry has been limited to large aerosol particles from primary emissions, e.g. dust or sea salt. However, we are interested in secondary particles, i.e. particles originating from a purely gasseous phase. These are much smaller and hence, experimental studies are difficult and rare. However, the clusters are small enough that the reactions may be studied using ab initio methods. We calculate the energy barriers, separating reactants and products, also known as the activation energy. By combining this knowledge with statistical physics the rate of the chemical reaction may be calculated as function of temperature. Hereby, the importance of the reaction may readily be determined at varying temperature, altitude and atmospheric composition, and used for input in the ACDC code or for explaining experimental or field data. Related recent publications from the group:
\[[http://www.atmos-chem-phys.net/11/7133/2011/acp-11-7133-2011.pdf|Bork et al., Atmos. Chem. Phys., 11 (2011) 71337142]\]
\[[http://www.atmos-chem-phys-discuss.net/11/29647/2011/acpd-11-29647-2011.pdf|Bork et al., Atmos. Chem. Phys. Discus., 11 (2011) 29647-29679]\]
Back to the topMolecular Dynamics simulations
Contact person: Ville Loukonen frame height=400 width=500px%25http://www.atm.helsinki.fi/~bork/Simuwiki/movie5.gif-
Molecular dynamics simulation of a SO'2' molecule collision with a O'3''-'(H'2'O)'5' cluster. Rapidly, SO'3''-' is formed and the new O'2' molecule is ejected from the cluster due to the high reaction entropy.Although ab initio methods applied with statistical physics are excellent tools for predicting cluster formation and aging, the huge number of possible molecular configurations of the cluster is a possible weakness. However, instead of high precision calculations on a few important configurations, another approach is to let the system evolve in time by numerically solving the equations of motion. Thereby, a realistic movie of the actual dynamics of a cluster of molecules is obtained and the likelihood of a given event, e.g. evaporation of H'2'O, oxidation of SO'2', or a "sticky" molecule-cluster collision, may be evaluated. Another related line of study is kinetic simulation of cluster population dynamics, i.e. whether clusters grow by colliding with each other or single molecules, and break up into two daughter clusters or a daughter cluster and a monomer.
Back to the topClassical Nucleation Theory and related approaches
Contact person: Hanna Vehkamäk http://www.atm.helsinki.fi/~hvehkama/drops_trans.gif|Figure: An illustration of the formation free energy barrier to cluster growth. Although the above mentioned methods are all state-of-the-art, the severe computational scaling of any ab initio method effectively prevents these methods from being applied to clusters larger than a few nanometers, containing at most 10-15 molecules. Such small clusters are, however, often unstable and may quickly evaporate e.g. due to increase in temperature or changes in the chemical composition of the surrounding gas. Entirely different from the explicit molecular modeling of ab initio approaches, classical nucleation theory (CNT) treats molecular clusters as spherical droplets consisting of bulk liquid. CNT is thus best applied to larger clusters more similar to bulk liquids. We have also experience on more advanced classical theories taking some of the micro-structure of the clusters into account. Currently, work is being dedicated to the unification of ab initio methods and CNT enabling the development of a typical cluster to be followed, starting from its initial nucleation to the point where the cluster is large and thermally stable. Even thermodynamic theories like CNT are computationally too expensive for use in larger scale atmospheric models, and results have to be parameterized to reduce computational cost. Related recent publications from the group:
\[[http://www.atm.helsinki.fi/%257Ehvehkama/publica/Vehkamaki_minmax_JCP2012.pdf|Vehkamäki et al., J. Chem. Phys., 136, 094107, 2012.]\]
\[[http://www.atm.helsinki.fi/%257Ehvehkama/publica/Napari2010_JChemPhys_133_154503.pdf|Napari et al., J. Chem. Phys., 133, 154503, 2010.]\]
\[[http://www.atm.helsinki.fi/%257Ehvehkama/publica/Julin_JCP_2010_surf_tension.pdf|Julin et al.,J. Chem. Phys., 133, 044704, 2010.]\]
Back to the top
Navigate space pagetree
Our group uses computational and theoretical methods to understand cluster and particle formation for atmospherically relevant molecules. We use molecular dynamics, Monte Carlo simulations and cluster size distribution dynamics with molecular interactions taken either from quantum chemical models or thermodynamics. We are part of the VILMA (Virtual Laboratory for Molecular Level Atmospheric Transformations), Academy of Finland center of excellence.
in close collaboration with Theo Kurtén's Atmospheric Computational Chemistry group
Possible topics for BSc and MSc theses
Watch the rap music video about molecular clusters that won the Dance Your PhD 2021 competition and was created by PhD students of this group:
Back to the top