• Andronache, C., , T. Gronholm, , L. Laakso, , V. Phillips, , and A. Venäläinen, 2006: Scavenging of ultrafine particles by rainfall at a boreal site: Observations and model estimations. Atmos. Chem. Phys., 6, 47394754.

    • Search Google Scholar
    • Export Citation
  • Baron, P., , and K. Willeke, 2005: Aerosol Measurement: Principles, Techniques, and Applications. 2nd ed. Wiley-Interscience, 1131 pp.

  • Beard, K., 1974: Experimental and numerical collision efficiencies for submicron particles scavenged by small raindrops. J. Atmos. Sci., 31, 15951603.

    • Search Google Scholar
    • Export Citation
  • Beard, K., , and H. Pruppacher, 1971: A wind tunnel investigation of collection kernels for small water drops in air. Quart. J. Roy. Meteor. Soc., 97, 242248.

    • Search Google Scholar
    • Export Citation
  • Byrne, M., , and S. Jennings, 1993: Scavenging of sub-micrometre aerosol particles by water drops. Atmos. Environ., 27, 20992105.

  • Chate, D., 2005: Study of scavenging of submicron-sized aerosol particles by thunderstorm rain events. Atmos. Environ., 39, 66086619.

    • Search Google Scholar
    • Export Citation
  • Chate, D., , and A. Kamra, 1997: Collection efficiencies of large water drops collecting aerosol particles of various densities. Atmos. Environ., 31, 16311635.

    • Search Google Scholar
    • Export Citation
  • Croft, B., , U. Lohmann, , R. Martin, , P. Stier, , S. Wurzler, , J. Feichter, , R. Posselt, , and S. Ferrachat, 2009: Aerosol size-dependent below-cloud scavenging by rain and snow in the ECHAM5-HAM. Atmos. Chem. Phys., 9, 46534675.

    • Search Google Scholar
    • Export Citation
  • Denman, K., and Coauthors, 2007: Couplings between changes in the climate system and biogeochemistry. Climate Change 2007: The Physical Science Basis, S. Solomon et al., Eds., Cambridge University Press, 541–584.

    • Search Google Scholar
    • Export Citation
  • Feng, J., 2007: A 3-mode parameterization of below-cloud scavenging of aerosols for use in atmospheric dispersion models. Atmos. Environ., 41, 68086822.

    • Search Google Scholar
    • Export Citation
  • Greenfield, S., 1957: Rain scavenging of radioactive particulate matter from the atmosphere. J. Atmos. Sci., 14, 115125.

  • Grover, S., , H. Pruppacher, , and A. Hamielec, 1977: A numerical determination of the efficiency with which spherical aerosol particles collide with spherical water drops due to inertial impaction and phoretic and electrical forces. J. Atmos. Sci., 34, 16551663.

    • Search Google Scholar
    • Export Citation
  • Houk, R., , V. Fassel, , G. Flesch, , H. Svec, , A. Gray, , and C. Taylor, 1980: Inductively coupled argon plasma as an ion source for mass spectrometric determination of trace-elements. Anal. Chem., 52, 22832289.

    • Search Google Scholar
    • Export Citation
  • Jiang, Y., , A. Umemura, , and C. Law, 1992: An experimental investigation on the collision behaviour of hydrocarbon droplets. J. Fluid Mech., 234, 171190.

    • Search Google Scholar
    • Export Citation
  • Kerker, M., , and V. Hampl, 1974: Scavenging of aerosol particles by a falling water drop and calculation of washout coefficients. J. Atmos. Sci., 31, 13681376.

    • Search Google Scholar
    • Export Citation
  • Kim, D., , K. Lee, , and Y. Kim, 2006: Characterization of a particle trap impactor. J. Aerosol Sci., 37, 10161023.

  • Lai, K., , N. Dayan, , and M. Kerker, 1978: Scavenging of aerosol particles by a falling water drop. J. Atmos. Sci., 35, 674682.

  • Leong, K., , K. Beard, , and H. Ochs III, 1982: Laboratory measurements of particle capture by evaporating cloud drops. J. Atmos. Sci., 39, 11301140.

    • Search Google Scholar
    • Export Citation
  • Leonhard, P., , R. Pepelnik, , A. Prange, , N. Yamada, , and T. Yamada, 2002: Analysis of diluted sea-water at the ng L−1 level using an ICP-MS with an octopole reaction cell. J. Anal. At. Spectrom., 17, 189196.

    • Search Google Scholar
    • Export Citation
  • Martin, G., , D. Johnson, , and A. Spice, 1994: The measurement and parameterization of effective radius of droplets in warm stratocumulus clouds. J. Atmos. Sci., 51, 18231842.

    • Search Google Scholar
    • Export Citation
  • Montaser, A., Ed., 1998: Inductively Coupled Plasma Mass Spectrometry. Wiley-VCH, 964 pp.

  • Park, S., , C. Jung, , K. Jung, , B. Lee, , and K. Lee, 2005: Wet scrubbing of polydisperse aerosols by freely falling droplets. J. Aerosol Sci., 36, 14441458.

    • Search Google Scholar
    • Export Citation
  • Pranesha, T., , and A. Kamra, 1996: Scavenging of aerosol particles by large water drops. 1. Neutral case. J. Geophys. Res., 101 (D18), 23 37323 380.

    • Search Google Scholar
    • Export Citation
  • Pruppacher, H., , and J. Klett, 1997: Microphysics of Clouds and Precipitation. 2nd ed. Kluwer Academic, 954 pp.

  • Schumann, T., 1989: Large discrepancies between theoretical and field determined scavenging coefficients. J. Aerosol Sci., 20, 11591162.

    • Search Google Scholar
    • Export Citation
  • Slinn, W., , and S. Shen, 1970: Anisotropic Brownian diffusion and precipitation scavenging of submicron particles. J. Geophys. Res., 75, 22672270.

    • Search Google Scholar
    • Export Citation
  • Slinn, W., , and J. Hales, 1971: A reevaluation of the role of thermophoresis as a mechanism of in- and below-cloud scavenging. J. Atmos. Sci., 28, 14651471.

    • Search Google Scholar
    • Export Citation
  • Starr, J., , and B. Mason, 1966: The capture of airborne particles by water drops and simulated snow crystals. Quart. J. Roy. Meteor. Soc., 92, 490499.

    • Search Google Scholar
    • Export Citation
  • Tinsley, B., , R. Rohrbaugh, , and M. Hei, 2001: Electroscavenging in clouds with broad droplet size distributions and weak electrification. Atmos. Res., 59, 115135.

    • Search Google Scholar
    • Export Citation
  • Tinsley, B., , L. Zhou, , and A. Plemmons, 2006: Changes in scavenging of particles by droplets due to weak electrification in clouds. Atmos. Res., 79, 266295.

    • Search Google Scholar
    • Export Citation
  • Tripathi, S., , and R. Harrison, 2002: Enhancement of contact nucleation by scavenging of charged aerosol particles. Atmos. Res., 62, 5770.

    • Search Google Scholar
    • Export Citation
  • Ulmke, H., , M. Mietschke, , and K. Bauckhage, 2001: Piezoelectric single nozzle droplet generator for production of monodisperse droplets of variable diameter. Chem. Eng. Technol., 24, 6970.

    • Search Google Scholar
    • Export Citation
  • Vali, G., 1996: Ice nucleation—A review. Nucleation and Atmospheric Aerosols, M. Kulmala and P. E. Wagner, Eds., Elsevier, 271–279.

  • Vohl, O., , S. Mitra, , K. Diehl, , G. Huber, , S. Wurzler, , K. Kratz, , and H. Pruppacher, 2001: A wind tunnel study of turbulence effects on the scavenging of aerosol particles by water drops. J. Atmos. Sci., 58, 30643072.

    • Search Google Scholar
    • Export Citation
  • Wang, P., , and H. Pruppacher, 1977: An experimental determination of the efficiency with which aerosol particles are collected by water drops in subsaturated air. J. Atmos. Sci., 34, 16641669.

    • Search Google Scholar
    • Export Citation
  • Wang, P., , S. Grover, , and H. Pruppacher, 1978: On the effect of electric charges on the scavenging of aerosol particles by clouds and small raindrops. J. Atmos. Sci., 35, 17351743.

    • Search Google Scholar
    • Export Citation
  • Wang, X., , L. Zhang, , and M. Moran, 2010: Uncertainty assessment of current size-resolved parameterizations for below-cloud particle scavenging by rain. Atmos. Chem. Phys., 10, 56855705.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Overview of previous experimental studies carried out with different instruments to determine the collection efficiency of droplets with aerosol particles at RH < 100% (Starr and Mason 1966; Beard 1974; Wang and Pruppacher 1977; Lai et al. 1978; Leong et al. 1982; Byrne and Jennings 1993; Pranesha and Kamra 1996; Vohl et al. 2001). The dashed line separates the cloud drops from the drizzle and rain drops (Martin et al. 1994).

  • View in gallery

    (a) Experimental setup. The T marks the location of the thermocouples. (b) Cross section of the top of the chamber.

  • View in gallery

    Typical output image from the microscope camera in a real experiment. The droplet radius is 12.8 μm.

  • View in gallery

    Diagram of the impactor used in our experiments.

  • View in gallery

    Theoretical collection efficiency as a function of (a) droplet and particle size for a fixed RH of 90% and (b) RH and particle size for a fixed droplet radius of 12.8 μm.

  • View in gallery

    Experimentally determined collection efficiency as a function of particle radius and droplet radius. For r = 12.8 μm, eight particle radii were investigated, and for a = 0.25 μm, four droplet sizes were investigated. For r = 18.2 and 20.0 μm, the values of a are shifted by ±0.003 μm for better visibility.

  • View in gallery

    Experimentally determined collection efficiency and theoretical calculation, as a function of (a) particle radius and (b) droplet radius. The theoretical calculation assumes an aerosol density of 1.4 g cm−3, RH of 90%, and a coalescence efficiency of 1. Solid error bars represent the average uncertainty and the dashed error bars are the maximum and minimum range for E at a specific a and r.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 59 59 12
PDF Downloads 37 37 9

Experimental Study of Collection Efficiencies between Submicron Aerosols and Cloud Droplets

View More View Less
  • 1 Institute for Atmospheric and Climate Science, ETH, Zurich, Switzerland
  • 2 Laboratory for Inorganic Chemistry, D-CHAB, ETH, Zurich, Switzerland
  • 3 Department of Physics and Atmospheric Science, Dalhousie University, Halifax, Nova Scotia, Canada
  • 4 Institute for Atmospheric and Climate Science, ETH, Zurich, Switzerland
© Get Permissions
Full access

Abstract

Collection efficiency E experiments for aerosol particles scavenged by cloud droplets were carried out in the newly built Collision Ice Nucleation Chamber (CLINCH). Pure water droplets having radii between 12.8 and 20.0 μm were allowed to fall freely and to collide in a laminar flow with lithium metaborate particles having radii between 0.05 and 0.33 μm. At the bottom of the chamber, the droplets and the particles captured were collected using a cup impactor. The collected solution was analyzed for the scavenged aerosol mass by inductively coupled plasma mass spectrometry. Evaporation of droplets was taken into account since the relative humidity inside the chamber was below 100%, resulting in final theoretical droplet sizes between 4.2 and 17.6 μm. The resulting experimental measurements were compared with theoretical values to see their correlation. The authors found an experimental trend similar to theory, as well as the “Greenfield gap” at the particle radius of 0.24 μm (E = 0.038) for the smallest cloud droplet size investigated in this study. The experimental values of collection efficiency found herein agree with those from theory within one order of magnitude, similar to previous studies reported in the literature.

Corresponding author address: Luis Ladino, Institute for Atmospheric and Climate Science, ETH Zurich, Universitätstrasse 16, CHN O 17.2, CH-8092 Zurich, Switzerland. E-mail: luis.ladino@env.ethz.ch

Abstract

Collection efficiency E experiments for aerosol particles scavenged by cloud droplets were carried out in the newly built Collision Ice Nucleation Chamber (CLINCH). Pure water droplets having radii between 12.8 and 20.0 μm were allowed to fall freely and to collide in a laminar flow with lithium metaborate particles having radii between 0.05 and 0.33 μm. At the bottom of the chamber, the droplets and the particles captured were collected using a cup impactor. The collected solution was analyzed for the scavenged aerosol mass by inductively coupled plasma mass spectrometry. Evaporation of droplets was taken into account since the relative humidity inside the chamber was below 100%, resulting in final theoretical droplet sizes between 4.2 and 17.6 μm. The resulting experimental measurements were compared with theoretical values to see their correlation. The authors found an experimental trend similar to theory, as well as the “Greenfield gap” at the particle radius of 0.24 μm (E = 0.038) for the smallest cloud droplet size investigated in this study. The experimental values of collection efficiency found herein agree with those from theory within one order of magnitude, similar to previous studies reported in the literature.

Corresponding author address: Luis Ladino, Institute for Atmospheric and Climate Science, ETH Zurich, Universitätstrasse 16, CHN O 17.2, CH-8092 Zurich, Switzerland. E-mail: luis.ladino@env.ethz.ch

1. Introduction

In the atmosphere and within clouds, there are a large number of aerosol particles of different sizes, shapes, and composition. Aerosol particles can act as cloud condensation nuclei (CCN) and ice nuclei (IN) and form droplets and ice crystals, respectively (nucleation scavenging). In addition, particles can collide and coalesce with these hydrometeors (impaction scavenging), which is the focus of this study. The process of collision followed by coalescence is termed collection. These cloud-borne particles can be removed from the atmosphere by precipitation formation and subsequent wet deposition (Wang et al. 1978). Collisions at temperatures below 0°C can also initiate freezing of supercooled droplets by contact freezing (Vali 1996). This process can influence cloud lifetime and precipitation formation in mixed-phase clouds, and the global radiation budget.

Aerosol dynamics, and the ability of aerosols to be scavenged, are governed by various forces such as gravitational, electrical, Brownian diffusion, and phoretic forces (thermophoresis, diffusiophoresis, and photophoresis). Photophoresis was neglected because our chamber does not have any source of light, whereas thermophoresis and diffusiophoresis were not neglected because of the presence of a temperature gradient and a water molecule density gradient, respectively. Therefore, from here on “phoretic effects” refers exclusively to a combination of thermophoresis and diffusiophoresis. At our experimental conditions, thermophoresis and diffusiophoresis take place because our droplets evaporate because of a subsaturated condition with respect to water. When a droplet evaporates, its surface becomes colder than the environment. Therefore, because of the temperature gradient, the air molecules push the aerosol particles toward the droplet. In contrast, the water vapor gradient is directed away from evaporating droplets so that some aerosol particles are pushed away from the droplet surface. Theoretical studies and parameterizations of the aerosol–hydrometeor collision efficiency (CE) were conducted by Greenfield (1957) and Slinn and Hales (1971), among others. Andronache et al. (2006) and Croft et al. (2009) used some of the above-mentioned forces to calculate their influence on droplet–particle collisions. These studies show that Brownian motion is only important for small particles (a < 0.1 μm) and gravitational forces for large particles (a > 1 μm). Moreover, they also show that there is a collision efficiency minimum, referred to as the Greenfield gap, for particle radii in the range of 0.1–2 μm. The Greenfield gap takes place because in this particle range Brownian motion is not large anymore and gravitational settling is not yet important. The particle radius for the Greenfield gap varies depending on the collector size. Near the collection minimum, phoretic forces are relatively more important. Additionally, in the presence of electric charges, electroscavenging is the dominant force (Tinsley et al. 2001).

To validate previous theoretical calculations, experimental studies have been carried out. Among others, Beard (1974), Kerker and Hampl (1974), Wang and Pruppacher (1977), Lai et al. (1978), Leong et al. (1982), Pranesha and Kamra (1996), and Vohl et al. (2001) measured the collection efficiencies between rain drops and aerosol particles using different instruments and mathematical expressions. These studies varied the aerosol and/or the droplet sizes to examine the change in the collection efficiency. However, only the Wang et al. (1978) and Vohl et al. (2001) studies are readily comparable with each other because they used similar parameters and conditions during their experiments. Unfortunately, previous studies cannot be compared with our results given the different experimental conditions. Experimental validation of our theoretical knowledge related to collection efficiencies, particularly for droplet–aerosol collisions, is difficult and far from complete.

Figure 1 shows that all of the previous experimental studies were done for raindrops (drop radius > 50 μm). The collection of submicron aerosol particles by cloud droplets has not been studied in laboratory experiments to our knowledge. This contributes to the high uncertainty related to the quantification of the impact of anthropogenic aerosols on clouds in climate modeling studies (Denman et al. 2007).

Fig. 1.
Fig. 1.

Overview of previous experimental studies carried out with different instruments to determine the collection efficiency of droplets with aerosol particles at RH < 100% (Starr and Mason 1966; Beard 1974; Wang and Pruppacher 1977; Lai et al. 1978; Leong et al. 1982; Byrne and Jennings 1993; Pranesha and Kamra 1996; Vohl et al. 2001). The dashed line separates the cloud drops from the drizzle and rain drops (Martin et al. 1994).

Citation: Journal of the Atmospheric Sciences 68, 9; 10.1175/JAS-D-11-012.1

Given the importance of aerosol impaction scavenging within clouds on the climate system, and the lack of experimental data to date, our study focuses on particles scavenged by cloud droplets of radii 20 μm and smaller. With the aim of improving the understanding of the microphysical processes, and their atmospheric implications related to scavenging, we ran different experiments using pure water droplets (Milli-Q, 18.2 MΩ) with radii between 12.8 and 20.0 μm and aerosol particles [lithium metaborate (LiBO2)] with radii between 0.05 and 0.33 μm. We also present a comparison between our experimental results and a theoretical calculation of collection efficiencies.

2. Experimental setup

All experiments were run in the newly developed Collision Ice Nucleation Chamber (CLINCH). Figure 2 shows a detailed schematic with the major features of the chamber. The collision chamber consists of two vertical parallel plates of variable length (20–60 cm) where the aerosol particles can interact with the liquid droplets at constant temperature and humidity. Each section is 20 cm in length and 10 cm wide with a distance of 1 cm between the plates. The droplet generator from Bremen University placed above the chamber contains a piezo element, which produces liquid drops with a defined frequency (1000 s−1) and radii between 12.8 ± 0.6 μm and 20 ± 0.6 μm (Ulmke et al. 2001). The solution used in the generator is pure Milli-Q water with 1.0 mg L−1 of rubidium (Rb) in the form of nitrate to allow the determination of the volume and thus number of water droplets collected. The droplets generated have a spherical shape (Fig. 3) with high uniformity. The calculated terminal velocity of the droplets (used in the experimental and theoretical calculations) varies between 1.95 and 4.65 cm s−1 depending on the droplet size (r = 12.8–20.0 μm). The terminal velocity was reached after 1.3 cm (for the smallest droplets), that is, before the droplets enter the main chamber and collide with the aerosol particles. The total velocity (droplet terminal velocity plus velocity of the air mass carrying the aerosol particles) was 4.45 cm s−1 for the smallest droplets and 7.15 cm s−1 for the biggest one.

Fig. 2.
Fig. 2.

(a) Experimental setup. The T marks the location of the thermocouples. (b) Cross section of the top of the chamber.

Citation: Journal of the Atmospheric Sciences 68, 9; 10.1175/JAS-D-11-012.1

Fig. 3.
Fig. 3.

Typical output image from the microscope camera in a real experiment. The droplet radius is 12.8 μm.

Citation: Journal of the Atmospheric Sciences 68, 9; 10.1175/JAS-D-11-012.1

The aerosol inlet is located on the top of the instrument. The head of the chamber has been designed to avoid turbulence (Fig. 2b). Aerosol particles are generated by a custom-made atomizer and size selected using a Differential Mobility Analyzer (DMA; TSI 3081). Particles of LiBO2 (Sigma-Aldrich ≥ 98.0%) were produced and injected into the system for all experiments. LiBO2 was used as aerosol because it can easily be detected with inductively coupled plasma mass spectrometry (ICPMS) and also because of its high water solubility (the solubility is rather important to avoid any digestion before the analysis), low density (see below), high deliquescence relative humidity (DRH), and low toxicity. Since the aerosols generated with the atomizer are solution droplets, they were passed through a diffusion dryer (filled with silica gel) to produce solid salt particles before being size selected in the DMA. The relative humidity (RH) after the dryer was less than 6.0%. To reduce evaporation of the droplets as much as possible in the chamber, we passed the air mass with the aerosol particles over a bath of potassium sulfate (K2SO4; Sigma-Aldrich ≥ 99.0%) solution to obtain a constant relative humidity of around 89.0 ± 2.3%, which is below the DRH of LiBO2 (99.2%). The humidity and temperature of the air mass containing the aerosols were measured with a relative humidity/temperature sensor (DKRF400, Driesen+Kern GmbH) before the chamber inlet. The chamber temperature was measured by three thermocouples (Type K, NiCr-Ni), one located in the head of the chamber and one on each wall of the chamber. A microscope (Navitar Zoom 6000) camera (Moticam 2300) system (Ryf AG) is attached to the head of the chamber to measure the droplet size in situ through a window.

In the lowest section of the chamber the aerosol particles and liquid droplets were collected and separated by a custom-made cup impactor [modified from Kim et al. (2006)]. The separation is done according to density (see Fig. 4) and is the reason why the low density of LiBO2 is crucial for our study (ρp = 1.4 g cm−3). The particles that collided with the droplets and coalesced were collected in a plastic flask (previously cleaned and washed in an ultrasonic bath with Milli-Q water, HCl 1% and HNO3 1% solutions) at the end of the impactor, and then analyzed by ICPMS (see section 3a) at the Laboratory for Inorganic Chemistry at ETH Zurich. The aerosol mass collected in each sample was then used to calculate the collection efficiencies (product of collision efficiency and coalescence efficiency).

Fig. 4.
Fig. 4.

Diagram of the impactor used in our experiments.

Citation: Journal of the Atmospheric Sciences 68, 9; 10.1175/JAS-D-11-012.1

The air pressure inside the system was 960 hPa and the total flow was 1 L min−1. The flow was controlled by the condensation particle counter (CPC; TSI 3010) and its pump, which was used to measure the aerosol number concentration at the end of the system. The aerosol number concentrations measured at the inlet and outlet of the system have a negligible difference of less than 1%.

3. Data analysis

a. Chemical analysis

After the collection experiments, 100 μL of a solution containing 100 μg L−1 cesium (Cs) in 10% (volume-to-volume) nitric acid (HNO3) was added to all samples and the flasks were made up with ultrapure water (H2O; Millipore SA, France) to a total amount of 4 g of solution. The elemental concentration of lithium (Li) and Rb in each sample was determined using inductively coupled plasma mass spectrometry (Houk et al. 1980; Montaser 1998). This method allows the highly sensitive determination of the elements with limits of detection in the low nanogram-per-liter range. Depending on the experimental conditions, the concentrations of Li and Rb in the analyzed samples ranged between 0.66 and 5.5 μg L−1 (Li) and 0.006 and 5.5 μg L−1 (Rb). The added Cs was used as internal standard to compensate for potential nonspectral interferences and allow determination of the absolute amount of the elements collected by the impactor. The analysis was carried out using standard operating conditions in no-gas mode (Leonhard et al. 2002) since spectral interferences from molecular ions were considered negligible. Nonetheless, several samples were found to contain significant amounts of strontium (Sr), which interfered with 87Rb+ in the ICPMS and thus all Rb data were calculated based on the 85Rb+ ion signals. The operating conditions of the ICPMS are listed in Table 1.

Table 1.

Operating conditions of the ICPMS used.

Table 1.

Based on the added mass of the internal standard solution and the final mass of each solution, the concentration of Cs in all samples was determined and the actual value was used to calculate the respective concentrations of Li and Rb. Calibration of the instrument was carried out using aqueous standards freshly prepared before each analysis session. All samples, stock solutions, and calibration standards were prepared and diluted by weighing.

b. Experimental calculation

To calculate the experimental collection efficiencies we have to know the number of aerosol particles collected in one drop, and the number of aerosol particles in the droplet sweep-out volume. The following mathematical equation was used to determine the experimental collection efficiency:
eq1
The collision efficiency is determined by
e1
where mICPMS is the mass (μg) of LiBO2 collected in the sample; Nr is the droplet number (unitless) defined as
eq2
where fr is the frequency of droplets (s−1) and te is the experimental time (s); Ca the aerosol number concentration (cm−3); ma the mass (μg) of a LiBO2 particle (ma = ρpυa); r the droplet radius (μm), and a the particle radius (μm). Finally, L is the effective interaction length of the chamber (cm), defined as
eq3
where Vr and Va are the droplet terminal (settling) velocity and the velocity of the air carrying the particles, respectively (cm s−1), and l is the length of the chamber (cm). Refer to appendix A for a comprehensive list of variables and their definitions that are used in this paper.

For comparison of the experimental results with theory we assume a coalescence efficiency of 1 and use collection efficiency from now on.

c. Uncertainties

The biggest source of error in our calculations is the aerosol mass in the samples, which was determined by ICPMS with a relatively high reproducibility (uncertainty < 3%). However, particles smaller than 0.15 μm in radius have a very small mass and their samples can be contaminated easily. Each data point (particle–droplet size combination) was acquired at least 3 times for almost all cases, except in a few of them because of technical problems, in order to account for reproducibility. To assess the instrumental background levels, blank samples were taken for all experimental conditions, including samples without aerosol particles and droplets and also samples with no particles or no droplets.

We performed a correction for the small particles (a = 0.05, 0.1, 0.15 μm) because multiple charges of large particles in the DMA contribute significantly to these sizes. This was necessary to correct for the broad size distribution even when selecting a monodisperse sample of aerosol particles. We used a DMA to produce a monodisperse aerosol and thereafter passed the air mass through a Scanning Mobility Particle Sizer (SMPS) to obtain the size distribution for the multiple charge correction. According to the size distributions the weight of each peak was determined and the correct sizes were calculated (a = 0.055, 0.12, 0.17 μm). We did not have equipment to measure the droplet and particle charges. If the droplets have one or more charges, electroscavenging could play a role in our experiment, generating a source of uncertainty since electroscavenging was not included in the theoretical calculations.

The amount of droplets used in each experiment can be validated through the Rb concentrations determined by ICPMS. This should reduce the uncertainty that might potentially arise from incomplete droplet collection in the impactor.

The uncertainties in the collection efficiency for each particle–droplet size combination were calculated using the propagation of uncertainty. For that, all measurement uncertainties were taken into account in Eq. (1). The instrument precisions used to measure each single variable in Eq. (1) were taken into account.

4. Results and discussion

a. Theoretical calculations of collection efficiency

There have been a number of studies (Slinn and Shen 1970; Beard 1974; Wang and Pruppacher 1977; Grover et al. 1977; Wang et al. 1978; Schumann 1989; Tripathi and Harrison 2002; Park et al. 2005; Chate 2005; Tinsley et al. 2006; Andronache et al. 2006; Feng 2007; Croft et al. 2009; Wang et al. 2010) that presented theoretical calculations, and parameterizations to determine the collision efficiencies between liquid drops and aerosol particles. With the aim to make a comparison between our experimental data and theoretical values, we used the theoretical flux model from Wang et al. (1978) where Brownian motion, thermophoresis, and diffusiophoresis are included. Since our aerosol particles are less than 0.33 μm and our droplet radius is less than 20 μm, inertial impaction was not included in the theoretical calculations. The aerosols normally have a single charge after the neutralizer (from the DMA), but the electrical effect was not taken into account because Jiang et al. (1992) showed that the droplets are basically not charged (or have a maximum of one charge) when using a similar instrument and also because it was not possible to measure and control the electrical charges in both particles and droplets. Tinsley et al. (2001) found that the electric effect plays an important role even when a small amount of charge [1–5 electrons (e)] is present on the particles and/or the droplets (>50 e). The electrical effect is, however, more important at smaller particle sizes than the ones used in our study, as shown in Tinsley et al. (2001).

Figure 5 shows E from the theoretical flux model as a function of droplet size and relative humidity. Appendix B gives the detailed mathematical expressions and input quantities required for the theoretical flux model following Wang et al. (1978) and Pruppacher and Klett (1997). Brownian motion is the dominant force until a ≈ 0.01 μm for r = 12.8 μm, and thereafter we approach the Greenfield gap. Here thermophoresis is the dominant force. As shown in Fig. 5a, the collection efficiency decreases with increasing droplet size since the drop’s cross section increases [see Eq. (1)]. The relative humidity in our experimental setup is below 100%, and thermophoresis cannot be neglected. Thermophoresis has been neglected in some previous studies (e.g., Lai et al. 1978; Pranesha and Kamra 1996) despite the relative humidity being lower than 100%.

Fig. 5.
Fig. 5.

Theoretical collection efficiency as a function of (a) droplet and particle size for a fixed RH of 90% and (b) RH and particle size for a fixed droplet radius of 12.8 μm.

Citation: Journal of the Atmospheric Sciences 68, 9; 10.1175/JAS-D-11-012.1

The relative humidity within clouds can be below 100% especially at the cloud edges where entrainment can occur. At these conditions, the drops can evaporate, causing a temperature gradient and water vapor density gradient from the droplet surface to the environment so that thermophoresis and diffusiophoresis take place. Thermophoresis dominates over diffusiophoresis for particle radii less than about 2 μm. Figure 5b shows how E varies if RH changes. If RH is reduced from 90% to 50%, E is higher, especially in the Greenfield gap, where E increases by a factor of approximately 8 for cloud droplets with radius of 12.8 μm.

b. Experimental data

Figure 6 shows the 41 individual data points of collection efficiency obtained in our study for different droplet and aerosol particle sizes. At a = 0.21 μm the differences between the E values are higher than the other aerosol sizes studied. There is a possibility that some charge is retained on the droplets after their generation and/or that the aerosol particles have more than one charge from the DMA. This could account for the scatter. However, we cannot estimate the amount of charges in both particles and droplets because of technical limitations. Previous experiments by Beard (1974), Kerker and Hampl (1974), Wang and Pruppacher (1977), Lai et al. (1978), Leong et al. (1982), Pranesha and Kamra (1996), Chate and Kamra (1997), and Vohl et al. (2001) were conducted in two different ways. Either the droplet radius was fixed (case I) and the aerosol radii were varied or, vice versa (case II). Our experiments included both the case I and case II approaches. However, comparison of our data with previous studies for both cases cannot be done because our experimental conditions differ largely from the previous studies (e.g., different particle and/or droplet sizes and relative humidities).

Fig. 6.
Fig. 6.

Experimentally determined collection efficiency as a function of particle radius and droplet radius. For r = 12.8 μm, eight particle radii were investigated, and for a = 0.25 μm, four droplet sizes were investigated. For r = 18.2 and 20.0 μm, the values of a are shifted by ±0.003 μm for better visibility.

Citation: Journal of the Atmospheric Sciences 68, 9; 10.1175/JAS-D-11-012.1

Figure 7 shows our experimental collection efficiencies as calculated using Eq. (1). For our case I experiments, we fixed the cloud droplet size at 12.8 μm, and we scanned eight different particle radii in the range from 0.05 to 0.33 μm as shown in Fig. 7a. We found the largest collection efficiency for the smallest particles (a = 0.05 μm). For increasingly larger particle sizes, up to 0.24 μm, we found the collection efficiency decreased. This trend agrees with the theoretical calculations and is similar to the trends of previous experimental studies by Lai et al. (1978) and Leong et al. (1982). However, we did find a stronger decrease in E with increasing particle radius than predicted by the theoretical calculation. Varying and uncontrolled amounts of retained electric charge on our droplets could account for the discrepancy with theory. The differences relative to the theoretical calculation are small for particle radii between 0.15 and 0.19 μm. However, for the other investigated sizes the differences increase to a maximum difference of one order of magnitude for a = 0.24 μm, where we found the Greenfield gap. The value of E exceeds 1 for particles with radii between 0.05 and 0.01 μm. We expect this increased collection because of Brownian motion and phoretic forces in this size range. If the droplets collide with all particles within a swept-out volume, E would be 1. If particles at the sides and behind the swept-out volume are collected by the droplet, then E exceeds 1. Both phoretic and electric effects can contribute to this.

Fig. 7.
Fig. 7.

Experimentally determined collection efficiency and theoretical calculation, as a function of (a) particle radius and (b) droplet radius. The theoretical calculation assumes an aerosol density of 1.4 g cm−3, RH of 90%, and a coalescence efficiency of 1. Solid error bars represent the average uncertainty and the dashed error bars are the maximum and minimum range for E at a specific a and r.

Citation: Journal of the Atmospheric Sciences 68, 9; 10.1175/JAS-D-11-012.1

For the fixed aerosol radius experiments (case II) shown in Fig. 7b, we find that E decreases with increasing droplets size as predicted by the theory. The experimental values are smaller than the theoretical calculations by up to one order of magnitude. Wang et al. (2010) also found discrepancies between theory and observations of similar magnitude. They suggested that this can be attributed to one or more processes, which were not well represented in the theoretical calculations. For our experiments we have been very careful with controlling the sizes of the colliding partners and the relative humidity to match our theoretical calculations. However, the possibility of retained electric charge on the droplets is a likely contributor to the discrepancy between the experimental results and the theoretical calculations.

Table 2 summarizes the theoretical calculations and experimental values with their corresponding uncertainties calculated from Eq. (2):
e2
where ΔE and ΔCE are the total collection and collision efficiency errors respectively, as a function of mICPMS, Nr, Ca, ma, L, r, and a, and Δ represents the error of the corresponding variables.
Table 2.

Experimental Ee and theoretical ET collection efficiencies. All the experiments were run at RH = 88.3 ± 2.3% and T = 25°C. The error is the mean value of the single uncertainties of the experimental data [Error = (ΣΔE)/n] as discussed in the text. The experimental values are based on the mean values for each particle–droplet size combination.

Table 2.

The expected values from theory are also shown. The experimental values are the average for each particle and droplet size. The uncertainties are quite high for the small particle sizes. This is due to the system being sensitive to small changes in concentration, and controlling small particles is challenging. With the current experimental configuration the uncertainties can only marginally be improved.

5. Conclusions

The CLINCH was used to conduct a series of experiments to study the collection efficiency between cloud droplets and aerosol particles, taking important parameters such as relative humidity and aerosol concentration, which were missing in previous studies, into account. The relative humidity was measured and controlled to account for thermophoresis in the theoretical calculations.

The collection efficiencies for our experimental data agree within one order of magnitude with the theoretical calculations. For our experiments we carefully controlled the aerosol number concentration, the particle size, the droplet size (measured in situ), and the relative humidity. As Wang et al. (2010) suggested in their paper, certain physical processes maybe neglected and/or are difficult to represent in the theoretical calculations and/or are difficult to control in experimental studies. This contributes to the discrepancy between measurements and theory. Turbulence and electric charges are some of these factors.

We experimentally demonstrated the presence of the “Greenfield gap” for cloud droplets. In the case of a fixed collector droplet size, we confirmed experimentally that the collection efficiency decreases with increasing particle size up to a particle size of 0.24 μm and thereafter increases. In the case of a fixed aerosol size we demonstrated that the collection efficiency decreased with increasing droplet radius.

For the smallest particles (a = 0.05 and 0.1 μm) we found mean values of E > 1 (1.93 and 1.02 respectively). We had an increased number of samples for these two particle sizes and found a relatively high variability in the collection efficiency from 0.83 to 2.67 for a = 0.05 μm, and 0.97 to 1.14 for a = 0.1 μm. Moreover, the uncertainties for these two particle sizes are much higher than for the larger ones, which is due to the difficulty of performing experiments with small particle sizes.

Unfortunately, literature data are rare on E for such small particle and droplet sizes as used in our experiments. We therefore suggest that future studies should extend the range of droplet and particle sizes to cover the entire ranges of naturally occurring aerosol and droplet combinations for E.

Acknowledgments

The authors thank Hannes Wydler and Edwin Hausammann for their contribution in building the chamber and their unconditional help. This work was supported by the Swiss National Foundation Project 200021–107663/1.

APPENDIX A

List of Variables

Table A1 lists the variables and their definitions that are used throughout this paper.

Table A1.

Notation of the different variables used in the theoretical and experimental calculations.

Table A1.

APPENDIX B

Theoretical Expressions

The mathematical expressions that were used to determine the total collection efficiency ETot as a function of the droplet and particle radii are based on the flux model (Wang et al. 1978; Pruppacher and Klett 1997) as follows:
eb1
where
eqb1
is the collection kernel due to phoretic forces and Brownian motion (m3 s−1),
eqb2
is the phoretic force (kg m3 s−2),
eqb3
is the thermophoretic force (kg m3 s−2),
eqb4
with T in degrees Celsius, is the thermal conductivity of moist air (J m−1 s−1 °C−1) (Pruppacher and Klett 1997),
eqb5
is the diffusiophoretic force (kg m3 s−2),
eqb6
is the particle mobility (s kg−1),
eqb7
is diffusion coefficient of aerosol particles (m2 s−1),
eqb8
is the Cunningham slip correction factor (unitless),
eqb9
is the Knudsen number (unitless),
eqb10
is the mean free path (μm),
eqb11
is the air viscosity (kg m−1 s−1),
eqb12
is the droplet terminal velocity (cm s−1),
eqb13
is a unitless coefficient,
eqb14
is the Reynolds number (unitless) (Pruppacher and Klett 1997), where
eqb15
(unitless), with X = ln(CdRe2) (unitless),
eqb16
is referred to as either the Davies or Best number (unitless) (Pruppacher and Klett 1997),
eqb17
is the air velocity (carrying the aerosol particles) (cm s−1),
eqb18
is the mean ventilation coefficient for aerosol particles flux (unitless),
eqb19
is the mean ventilation coefficient for mass (unitless) [fp and fυ were taken from Beard and Pruppacher (1971); for simplicity we assume that fυ = fh based on Wang et al. (1978) and Pruppacher and Klett (1997), where fh is the mean ventilation coefficient for heat transfer (unitless)],
eqb20
is the Schmidt number of collected particles (unitless),
eqb21
is the Schmidt number for water vapor in air (unitless),
eqb22
is the water vapor density in the air (kg m−3) (Tinsley et al. 2006), and
eqb23
is the water vapor density in the droplet surface (kg m−3) (Tinsley et al. 2006).

Input quantities

We used the following input quantities:

  • KB = 1.381 × 10−23 kg m2 s−2 K−1 (Boltzmann constant),
  • Rυ = 461.5 J kg−1 K−1 (gas constant for water vapor),
  • Mw = 18.02 g mol−1 (molecular weight of water),
  • Ma = 28.96 g mol−1 (molecular weight of air) (Pruppacher and Klett 1997),
  • g = 9.807 m s−2 (acceleration due to gravity) (Feng 2007),
  • Dw = 0.225 × 10−4 m2 s−1 (diffusivity of water vapor),
  • ρa = 1.10 kg m−3 (air density),
  • kp = 4.19 × 10−1 J K−1 m−1 s−1 (thermal conductivity of the particles) (Wang et al. 1978),
  • (vapor pressure of water at temperature T),
  • ρw = 1000 kg m−3 (water density),
  • T = 25°C (298.15 K, air temperature),
  • ρp = 1400 kg m−3 (particle density),
  • p = 960 hPa (air pressure),
  • flow rate = 16.67 cm3 s−1, and
  • cross section = 10 cm2 (experimental conditions).

The input values for droplet temperature Tr and vapor pressure of water as a function of relative humidity (RH) are given in Table B1.

Table B1.

Input values for droplet temperature and vapor pressure of water as a function of RH (Baron and Willeke 2005).

Table B1.

REFERENCES

  • Andronache, C., , T. Gronholm, , L. Laakso, , V. Phillips, , and A. Venäläinen, 2006: Scavenging of ultrafine particles by rainfall at a boreal site: Observations and model estimations. Atmos. Chem. Phys., 6, 47394754.

    • Search Google Scholar
    • Export Citation
  • Baron, P., , and K. Willeke, 2005: Aerosol Measurement: Principles, Techniques, and Applications. 2nd ed. Wiley-Interscience, 1131 pp.

  • Beard, K., 1974: Experimental and numerical collision efficiencies for submicron particles scavenged by small raindrops. J. Atmos. Sci., 31, 15951603.

    • Search Google Scholar
    • Export Citation
  • Beard, K., , and H. Pruppacher, 1971: A wind tunnel investigation of collection kernels for small water drops in air. Quart. J. Roy. Meteor. Soc., 97, 242248.

    • Search Google Scholar
    • Export Citation
  • Byrne, M., , and S. Jennings, 1993: Scavenging of sub-micrometre aerosol particles by water drops. Atmos. Environ., 27, 20992105.

  • Chate, D., 2005: Study of scavenging of submicron-sized aerosol particles by thunderstorm rain events. Atmos. Environ., 39, 66086619.

    • Search Google Scholar
    • Export Citation
  • Chate, D., , and A. Kamra, 1997: Collection efficiencies of large water drops collecting aerosol particles of various densities. Atmos. Environ., 31, 16311635.

    • Search Google Scholar
    • Export Citation
  • Croft, B., , U. Lohmann, , R. Martin, , P. Stier, , S. Wurzler, , J. Feichter, , R. Posselt, , and S. Ferrachat, 2009: Aerosol size-dependent below-cloud scavenging by rain and snow in the ECHAM5-HAM. Atmos. Chem. Phys., 9, 46534675.

    • Search Google Scholar
    • Export Citation
  • Denman, K., and Coauthors, 2007: Couplings between changes in the climate system and biogeochemistry. Climate Change 2007: The Physical Science Basis, S. Solomon et al., Eds., Cambridge University Press, 541–584.

    • Search Google Scholar
    • Export Citation
  • Feng, J., 2007: A 3-mode parameterization of below-cloud scavenging of aerosols for use in atmospheric dispersion models. Atmos. Environ., 41, 68086822.

    • Search Google Scholar
    • Export Citation
  • Greenfield, S., 1957: Rain scavenging of radioactive particulate matter from the atmosphere. J. Atmos. Sci., 14, 115125.

  • Grover, S., , H. Pruppacher, , and A. Hamielec, 1977: A numerical determination of the efficiency with which spherical aerosol particles collide with spherical water drops due to inertial impaction and phoretic and electrical forces. J. Atmos. Sci., 34, 16551663.

    • Search Google Scholar
    • Export Citation
  • Houk, R., , V. Fassel, , G. Flesch, , H. Svec, , A. Gray, , and C. Taylor, 1980: Inductively coupled argon plasma as an ion source for mass spectrometric determination of trace-elements. Anal. Chem., 52, 22832289.

    • Search Google Scholar
    • Export Citation
  • Jiang, Y., , A. Umemura, , and C. Law, 1992: An experimental investigation on the collision behaviour of hydrocarbon droplets. J. Fluid Mech., 234, 171190.

    • Search Google Scholar
    • Export Citation
  • Kerker, M., , and V. Hampl, 1974: Scavenging of aerosol particles by a falling water drop and calculation of washout coefficients. J. Atmos. Sci., 31, 13681376.

    • Search Google Scholar
    • Export Citation
  • Kim, D., , K. Lee, , and Y. Kim, 2006: Characterization of a particle trap impactor. J. Aerosol Sci., 37, 10161023.

  • Lai, K., , N. Dayan, , and M. Kerker, 1978: Scavenging of aerosol particles by a falling water drop. J. Atmos. Sci., 35, 674682.

  • Leong, K., , K. Beard, , and H. Ochs III, 1982: Laboratory measurements of particle capture by evaporating cloud drops. J. Atmos. Sci., 39, 11301140.

    • Search Google Scholar
    • Export Citation
  • Leonhard, P., , R. Pepelnik, , A. Prange, , N. Yamada, , and T. Yamada, 2002: Analysis of diluted sea-water at the ng L−1 level using an ICP-MS with an octopole reaction cell. J. Anal. At. Spectrom., 17, 189196.

    • Search Google Scholar
    • Export Citation
  • Martin, G., , D. Johnson, , and A. Spice, 1994: The measurement and parameterization of effective radius of droplets in warm stratocumulus clouds. J. Atmos. Sci., 51, 18231842.

    • Search Google Scholar
    • Export Citation
  • Montaser, A., Ed., 1998: Inductively Coupled Plasma Mass Spectrometry. Wiley-VCH, 964 pp.

  • Park, S., , C. Jung, , K. Jung, , B. Lee, , and K. Lee, 2005: Wet scrubbing of polydisperse aerosols by freely falling droplets. J. Aerosol Sci., 36, 14441458.

    • Search Google Scholar
    • Export Citation
  • Pranesha, T., , and A. Kamra, 1996: Scavenging of aerosol particles by large water drops. 1. Neutral case. J. Geophys. Res., 101 (D18), 23 37323 380.

    • Search Google Scholar
    • Export Citation
  • Pruppacher, H., , and J. Klett, 1997: Microphysics of Clouds and Precipitation. 2nd ed. Kluwer Academic, 954 pp.

  • Schumann, T., 1989: Large discrepancies between theoretical and field determined scavenging coefficients. J. Aerosol Sci., 20, 11591162.

    • Search Google Scholar
    • Export Citation
  • Slinn, W., , and S. Shen, 1970: Anisotropic Brownian diffusion and precipitation scavenging of submicron particles. J. Geophys. Res., 75, 22672270.

    • Search Google Scholar
    • Export Citation
  • Slinn, W., , and J. Hales, 1971: A reevaluation of the role of thermophoresis as a mechanism of in- and below-cloud scavenging. J. Atmos. Sci., 28, 14651471.

    • Search Google Scholar
    • Export Citation
  • Starr, J., , and B. Mason, 1966: The capture of airborne particles by water drops and simulated snow crystals. Quart. J. Roy. Meteor. Soc., 92, 490499.

    • Search Google Scholar
    • Export Citation
  • Tinsley, B., , R. Rohrbaugh, , and M. Hei, 2001: Electroscavenging in clouds with broad droplet size distributions and weak electrification. Atmos. Res., 59, 115135.

    • Search Google Scholar
    • Export Citation
  • Tinsley, B., , L. Zhou, , and A. Plemmons, 2006: Changes in scavenging of particles by droplets due to weak electrification in clouds. Atmos. Res., 79, 266295.

    • Search Google Scholar
    • Export Citation
  • Tripathi, S., , and R. Harrison, 2002: Enhancement of contact nucleation by scavenging of charged aerosol particles. Atmos. Res., 62, 5770.

    • Search Google Scholar
    • Export Citation
  • Ulmke, H., , M. Mietschke, , and K. Bauckhage, 2001: Piezoelectric single nozzle droplet generator for production of monodisperse droplets of variable diameter. Chem. Eng. Technol., 24, 6970.

    • Search Google Scholar
    • Export Citation
  • Vali, G., 1996: Ice nucleation—A review. Nucleation and Atmospheric Aerosols, M. Kulmala and P. E. Wagner, Eds., Elsevier, 271–279.

  • Vohl, O., , S. Mitra, , K. Diehl, , G. Huber, , S. Wurzler, , K. Kratz, , and H. Pruppacher, 2001: A wind tunnel study of turbulence effects on the scavenging of aerosol particles by water drops. J. Atmos. Sci., 58, 30643072.

    • Search Google Scholar
    • Export Citation
  • Wang, P., , and H. Pruppacher, 1977: An experimental determination of the efficiency with which aerosol particles are collected by water drops in subsaturated air. J. Atmos. Sci., 34, 16641669.

    • Search Google Scholar
    • Export Citation
  • Wang, P., , S. Grover, , and H. Pruppacher, 1978: On the effect of electric charges on the scavenging of aerosol particles by clouds and small raindrops. J. Atmos. Sci., 35, 17351743.

    • Search Google Scholar
    • Export Citation
  • Wang, X., , L. Zhang, , and M. Moran, 2010: Uncertainty assessment of current size-resolved parameterizations for below-cloud particle scavenging by rain. Atmos. Chem. Phys., 10, 56855705.

    • Search Google Scholar
    • Export Citation
Save