Homogeneous heteromolecular nucleation of natural atmospheric aerosols was first observed in the coastal environment more than 100 years ago. Aitken [1897] reported elevated condensation nucleus (CN) concentrations close to the shoreline on the west Scottish coast, where he observed particle concentrations to rise from a background level of ª300 cm-3 to ª150,000 cm-3 during clean marine air flow under sunny conditions. The cause of these new particles was thought to be photochemical breakdown of gases released from the shore biota, though the nucleating species were never identified. These 'nucleation' events were also observed at other coastal sites around western Europe, including the Mace Head Atmospheric Research Station in Ireland, where new particle events have been found to be related to the oceanic tidal cycle [O'Dowd et al., 1998].

The coastal environment is a strong and frequently active source of natural atmospheric aerosol particles and, consequently, serves as an excellent natural laboratory to study the conditions under which particle formation occurs. The PARFORCE consortium and program was set up to focus on the processes and conditions which promote and control homogeneous heteromolecular nucleation in the coastal boundary layer. These goals are being achieved through a long-term continuous aerosol monitoring program at Mace Head, along with two intensive field campaigns (September 1998 & June 1999) comprised of detailed atmospheric chemistry measurements. In parallel with the experimental program, aerosol nucleation and growth models are being developed for use in conjunction with the experimental data to identify the primary mechanisms involved.

Primary Objectives

To determine the environmental conditions and rates under which homogeneous heteromolecular nucleation occurs in the coastal boundary layer (i.e., pre-existing aerosol surface area, precursor gas concentration, and micro- and macro-scale meteorology).

To examine whether these nucleation events can be explained by binary or ternary heteromolecular nucleation via the following schemes: H2SO4-H2O or NH3-H2SO4-H2O or NH3-H2O-MSA/HCl/HNO3, or whether alternative nucleation schemes are more likely to explain the observed events (i.e., organic/halocarbon embryo formation followed by organic/halogen and/or sulfate growth).

Description of the Experiment

The Mace Head Research Station was chosen as the experiment site primarily due to the known high frequency of nucleation events. Additionally, its comprehensive facilities are suited to these studies: two main laboratories are situated less than 100 m from the tidal regions and are adjacent to 10 m and 20 m sampling towers. The primary sample duct istaken from a community aerosol sampler adjacent to the 10 m tower. A third laboratory is situated approximately 300 m from the shoreline.

Detailed meteorological data on atmospheric structure and turbulent fluxes at the coastal interface were acquired through use of Lidar measurements and micro meteorological flux packages. Ultrafine particle concentrations were measured using an array of condensation particle counters (CPC's), each measuring total particle concentration above a certain cut-off size (3, 5, 10 nm). This combination of measurements allows fast response source rates to be determined by examining differences between pairs of total particle counts. A pyramid array of particle counters was deployed around the station to elucidate the spatial scales over which these events occur. Within this array, the horizontal distances were of the order of 300 m apart while vertical gradient measurements were taken at 10 and 20 m. Size distributions of ultrafine and fine mode particles, using Differential Mobility Particle Sizer techniques, were also taken with 10 minute time resolution to examine the growth of this mode. Aerosol size distributions were measured through use of optical particle counters up to sizes of 300 mm to determine the existing aerosol condensation sink.

Continuous measurements were made of relevant inorganic precursor gases: SO2, NH3, HCl, OH, H2SO4 and MSA, the latter three compounds measured by chemical ionization mass spectrometry (CIMMS). Additional measurements were made of halocarbon species, DMS, DMDS, and VOCs (C6-C20). Seaweed stressing experiments were also carried out to examine how the emissions of various biogenic gases depend on environmental stressing.

Preliminary Results

A continuous measurement program of ultrafine particle concentrations and size distributions was initiated at Mace Head in February 1998. These long-term measurements illustrate the frequency of the nucleation events. In Figure 1, a contour plot of particle concentration (1-hour average) as a function of day-of-year and time-of-day is plotted for a 3-month period. Low tide occurrences are also highlighted. From the current database it is evident that these tidal particle events occur almost daily, all year round and do not appear to be seasonal in any way.

Figure 1. Contour plot of ultrafine particles as a function of day-of-year and time-of-day. The map scale is in units of particles cm-3. Occurrence of low tide is marked by '+'. It can be seen that during daylight hours, the occurrence of elevated particle concentrations follows closely the low tide cycle.

Aerosol Physical Characteristics

The number of new, ultrafine aerosols was measured by taking the difference in concentration between 3 and 10 nm sized particles. A typical nucleation event is illustrated in Figure 2a which shows particle concentrations from 0.6-1.1 x 106 cm-3 over a duration of a few hours. The peninsular location of the Mace Head station (see accompanying article) allows sampling in the tidal plume at distances ranging from ª100 m up to ª10 km from the tidal source region, depending on wind direction [O'Dowd et al., 1999]. For moderate wind speeds, these distances correspond to point measurements at ª30 seconds and ª30 minutes into the evolving coastal plume. Two characteristic differences are observed depending on distance from the plume source. For the measurements corresponding to 30 seconds into plume evolution, ultrafine (3-10 nm) particle concentration increases to about 10,000 cm-3, while no enhancement is seen at sizes larger than 10 nm. By comparison, 30 minutes into the plume evolution, concentrations at sizes larger than 3 nm reach levels >600,000 cm-3 with approximately 50,000 cm-3 of these having grown into sizes larger than 10 nm.

Aerosol Precursor Characteristics

A wide variety of possible aerosol precursors has been measured during both campaigns; however, no clear pattern linking any one species with aerosol formation and low tide occurrence has been observed. During the intensive field campaigns, no observable tidal signal was observed between NH3, DMS, SO2, VOCs (C2-C16), although some heavier organics were present at low tide; however, these remain to be identified. Gaseous H2SO4 exhibited a typical diurnal cycle, following OH concentrations, and was found to be independent of low tide conditions. Figure 2b illustrates the typical diurnal variation of H2SO4 with peak concentration occurring around midday. MSA is also shown in this figure and, although there appears to be some coherence between both acid concentrations on this day, it should be noted that, for the most part, MSA concentrations showed little correlation with H2SO4 concentrations. Preliminary interpretation suggests that there is a significant source of MSA in the absence of the OH radical. Typical peak marine air concentrations of OH, H2SO4 and MSA were of the order of 4-8 x 106 cm-3, 1.5-3 x 107 cm-3 and 1-2 x 10-7 cm-3, respectively.

Figure 2: (a) Ultrafine particle concentration and tidal height variation (from University of Sunderland). The units on the gray-scale bar are cm-3. (b) H2SO4 and MSA concentrations (from H. Berresheim, Deutscher Wetterdienst).

Aerosol Chemical Characteristics

The chemical composition of new particles must first be characterized in order to identify particular nucleating species. This task is quite difficult as it requires separating out sufficient new particle mass from existing particles, which otherwise would contaminate the chemical sample. One technique to elucidate ultrafine particle chemical characteristics involves examination of a particle's hygroscopic properties. Although indirect, this technique can reveal important aerosol properties in near-real time. The technique generally involves examining the hygroscopic growth of a particle of known size between two fixed relative humidities (typically 50-90%). The so-called growth factor over this range of humidities is well known for common aerosol species. The smallest size that could be conditioned for hygroscopic growth analysis was 8 nm, consequently this technique elucidates only the nature of the primary condensing vapor leading to growth of new particles into ultrafine sizes rather than the actual new particle composition.
For example, at 90% relative humidity, an 8 nm ammonium sulfate particle will exhibit an increase in diameter of 1.4 over its size at 40% humidity. Sulfuric acid particles of this size exhibit a slightly higher growth factor of 1.5 and less soluble organic aerosol will typically have a growth factor of the order of 1.1 or less. Thus, using this technique we were able to determine if the recently formed ultrafine particles were composed of sulfate mass or a less-soluble organic mass. Preliminary findings indicate the new particles are primarily composed of less soluble material, probably organic in nature, possessing a growth-factor of ª1.05 ­ 1.1. However, when very high concentrations of H2SO4 vapor coincided with the occurrence of ultrafine particles, the growth factor indicated a particle composition between that of organic mass and ammonium sulfate, suggesting that sometimes H2SO4 can also contribute to ultrafine aerosol growth.

Modeling of aerosol formation and growth

In the absence of direct chemical and physical characterization of new particles, a combination of modeling tools and experimental data must be used to elucidate the nucleation mechanisms and species involved. From the experimental results, we have measurements at two stages in the coastal plume development: ª30 seconds and ª30 minutes from the plume source. If, using an aerosol dynamics and condensation growth model, we can accurately predict the observed aerosol growth at these two stages of plume evolution, then we can determine the initial nucleation rates and condensation vapor source strength required to produce the observed particle concentrations. O'Dowd et al. [1999] concluded that, in order to explain a particle concentration of 10,000 cm-3 after 30 seconds into the plume, and of the order of 600,000 cm-3 after 30 minutes, with 50,000 cm-3 of these at sizes larger than 10 nm, a nucleation rate of the order of 107 cm-3 s-1 was required, followed by condensation from a continuous vapor concentration of 5 x 107 cm-3.

Revised nucleation theory [Kulmala et al., 1998] indicates that for binary nucleation of H2SO4-H2O, an acid concentration of >109 cm-3 would be required to sustain such nucleation rates. Thus, we can rule out binary nucleation as a feasible explanation. In the presence of NH3, the free energy barrier to nucleation is removed, thus suggesting that a ternary nucleation system involving NH3 would be more likely. Recent theoretical development of ternary nucleation models indicates that to achieve nucleation rates of 107 cm-3s-1, NH3 concentrations of the order of 20 ppt, and H2SO4 concentrations of 107 cm-3 are required [Korhonen et al., 1999]. Additionally, under higher NH3 concentrations of 100-200 ppt, a H2SO4 concentration of the order of 1-4 x 106 cm-3 is required to achieve the same nucleation rate. Given the range of NH3 (20-2500 ppt) and H2SO4 concentrations (105 to > 1.5 x 107 cm-3) encountered, this mechanism appears likely to explain the predicted nucleation rates. However, although the ternary H2SO4-H2O-NH3 system can explain nucleation rates of this magnitude, these concentrations of H2SO4 are insufficient to explain the growth of the new particles into detectable sizes of 3 nm. A fourth species, X, is required to provide the condensing material. X is thought to be an organic acid possessing a very low vapor pressure and is a product of biogenic emissions from the tidal region. Given the range of H2SO4 and NH3 concentrations encountered in this environment, ternary nucleation theory suggests that nucleation is happening more often than we can detect ultrafine particles. We only detect nucleation when there is a sufficient source of condensable material to grow these new particles to detectable sizes of larger than 3 nm [O'Dowd et al., 1999]. Although X remains to be identified, the hypothesis of an organic condensing species is consistent with the growth factor analysis of ultrafine particles.

The coastal environment appears to be a very strong and frequent source of natural particles and is likely to significantly influence the natural background aerosol population on a regional scale. However, there are still many unanswered questions concerning the chemical species involved in nucleation and growth of new natural particles in this environment. While there is a considerable amount of data analysis to be conducted in conjunction with numerical simulations before we can determine the primary chemical species leading to coastal ultrafine particle formation, it is clearly important that theoretical and technical developments are promoted to improve our ability to measure and predict the formation of new aerosol particles in this (and other) environments.

References

PARFORCE homepage: http://www.sunderland. ac.uk/~es0cmas/parforce

1. Aitken, J.A., On some nuclei of cloudy condensation, Trans. Roy. Soc. Edin., Vol XXXIX, 1897.
2. Korhonen, P., M. Kulmala, A. Laaksonen, Y. Viisanen, R. McGraw, J.H. Seinfeld, Ternary nucleation of H2SO4, NH3, and H2O in the atmosphere, J. Geophys. Res. (in press).
3. Kulmala M., A. Laaksonen and L. Pirjola, Parameterizations for sulfuric acid/water nucelation rates, J. Geophys. Res., 103, 8301-8308, 1998.
4. Mäkelä, J.M., P. Aalto, V. Jokinen, T. Pohja, A. Nissinen, S. Palmroth, T. Markkanen, K. Seitsonen, H. Lihavainen and M. Kulmala, Observations of ultrafine aerosol particle formation and growth in boreal forest, Geophys. Res. Lett., 24, 1219-1222, 1997.
5. Nilsson, E.D. and M. Kulmala, The potential for atmospheric mixing processes to enhance the binary nucleation rate, J. Geophys. Res., 103, 1381-1389, 1998.
6. O'Dowd, C.D., M. Geever, M.K. Hill, S.G. Jennings, M.H. Smith, New particle formation: Spatial scales and nucleation rates in the coastal environment, Geophys. Res. Lett., 25, 1661-1664, 1998.
7. O'Dowd, C.D., G. Mcfiggens, L. Pirjola, D. J. Creasey, C. Hoell, M.H. Smith, B. Allen, J.M.C. Plane, D. E. Heard, J. D. Lee, M. J. Pilling, and M. Kulmala, On the photochemical production of new particles in the coastal boundary layer, Geophys. Res. Lett., 26(12), 1707-1710,1999.

PARFORCE Consortium

Jyrki Makela, Markku Kulmala (University of Helsinki), Hans-Christen Hansson, Johan Strom (ITM, Stockholm University), Gerard Jennings, (NUI, Galway), Harald Berresheim, (DWD Meteorological Observatory Hohenpeissenberg), Gerrit de Leeuw, (TNO, The Netherlands), Roy Harrison, (University of Birmingham), Nick Hewitt, (Lancaster University), Yrjo Viisanen, (Finnish Meteorological Institute), Andrea Jackson, Alistair Lewis, (Leeds University), Spyros Rapsomanikis, (Demokritos University of Thrace).

Collaborators

Peter Liss, Lucy Carpenter, (University of East Anglia), Peter Simmonds, (Bristol University), Thorsten Hoffman, (ISAS, Dortmund), Keith Bigg, (not affiliated), Jochen Stutz, Ulrich Platt, (University of Heidelberg).