Prototype Development and Laboratory Evaluation of an Aerosol to Hydrosol Sampler

Transcription

1 Aerosol and Air Quality Research, 15: , 215 Copyright Taiwan Association for Aerosol Research ISSN: print / online doi: 1.429/aaqr Prototype Development and Laboratory Evaluation of an Aerosol to Hydrosol Sampler Maimaitireyimu Wubulihairen, Sabrina Yanan Jiang, Zhi Ning * School of Energy and Environment, City University of Hong Kong, Kowloon, Hong Kong ABSTRACT Improved particle sampling techniques are needed to counter the disadvantages of traditional filter sampling and the limitations of those currently available. This study presents the development and evaluation of an Aerosol to Hydrosol Air Sampler (ATHAS). It consists principally of a steam generator, condenser and collector. ATHAS collects particles at the flow rate of 5 litres per minute through condensational growth followed by particle-containing droplets collection of combined impaction and centrifugal flow. Monodisperse polystyrene latex (PSL) particles ranging in size from.1 µm to 2 µm were used to test the performance of size dependent collection at the extreme condition of highly hydrophobic aerosols. Results show that the sampler has near 9 1% collection efficiency for supermicron PSL particles, and 65% and 5% for.5 and.1 µm particles, respectively. NaCl collection efficiency tests showed that the sampler has over 95% collection efficiency for soluble aerosols. The medium size and sampling flow make it suitable for field deployment with high potential for semi-continuous online chemical analysis. Keywords: Particulate matter; Condensational growth; Aerosol sampler. INTRODUCTION There have been many studies on fine particles and their effect on human health (Dockery and Pope, 1994; Stone et al., 1998; Nemmar et al., 1999; Pope, 2; Wilson et al., 22; Antonella and Joel, 29; Pope et al., 29). Effective control of adverse health effects requires knowledge about aerosol sources through analyzing their chemical composition (Grahame and Schlesinger, 27). Conventional filter sampling has been widely used for chemical analysis of particle components. Filter sampling typically requires hours to days of sampling time. However, variations in meteorological condition and traffic over short time periods can cause rapid changes in concentration and chemical composition (Ning et al., 27; Daher et al., 213). Moreover, long sampling times may lead to a loss of volatile components in collected particles (Chow, 1995). Thus, the measurements obtained through filter sampling cannot fully describe the changes in aerosol chemistry in the way possible with high time resolution sampling methods (Orsini et al., 23; Wang et al., 213). In addition, the high loading of filters may allow for chemical reactions to occur on the filter, resulting in loss or changes in composition. The process of manually * Corresponding author. Tel.: Tel: ; Fax: address: zhining@cityu.edu.hk extracting compounds from filters before chemical analysis is time consuming (Slanina et al., 21). The use of condensational growth for detection and measurement of fine particles has been known since the studies of Aitken in 19 th century (Aitken, 1888). It has led to the development of many devices for the efficient detection, measurement and collection of aerosol particles. In most condensation instruments, the aerosol is first passed through a vapor (water, butanol, or alcohol) filled saturator. The mixture of aerosol and vapor then passes through a condenser where the vapor condenses on particle surfaces and these grow to sizes that are easy to detect or collect (Agarwal and Sem, 198; Ahn and Liu, 199). A range of devices have been developed using this method (Khlystov et al., 1995; Ito et al., 1998; Hering and Stolzenburg, 25); in the method developed by the latter group, the particle laden flow is introduced to a wet walled growth tube, where its temperature is greater than that of the entrained air stream. The growth tube serves as a condenser in which the particle sizes are enlarged through condensation of liquid vapor. The Steam Jet Aerosol Collector (SJAC) (Khlystov et al., 1995) enlarges particles through condensation of water vapor on to particles. However, the sample collected includes large amount of condensed liquid formed on the inner wall of the device which dilutes of sample. Slanina et al. (21) described an enhanced steam jet aerosol collector which collects particles into liquid. Particles were enlarged with assistance of water vapor condensation and these are collected as droplets with a cyclone. However, the resultant sample is also diluted by

2 Wubulihairen et al., Aerosol and Air Quality Research, 15: , the condensed water. A widely used Particle-into-Liquid Sampler (PILS) has a sampling flow rate of 5 L/min (Weber et al., 21) and L/min (Orsini, 23) for applications of online bulk chemical components measurements. In some cases, PILS samples also need considerable off-line treatment before the actual analysis (Parshintsev et al., 21). Demokritou, Gupta et al. (22) developed a high volume (25 L/min) apparatus that enlarges particles through condensational growth. However, the system is large and not suitable for field deployment and sampling. Wang et al. (213) introduced a high volume aerosol into liquid collector. It samples at a flow rate of 2 L/min and uses a condensational growth component of versatile aerosol concentration enrichment system (VACES) (Kim et al., 21). It has two large slit nozzles and collects particles into liquid directly. Despite its high collection efficiency and preservation of particle chemical characteristics, the relatively large size may limit its portability for field deployment. A summary of the comparison of the various systems performance and main specifications is shown in Table 1. Here we describe an aerosol to hydrosol air sampler. It combines condensational particle growth with the collection of particle-containing droplets through integrated impaction and centrifugal motion at a designed medium sampling rate of 5 litres per minute. The sampler features an array of uniquely constructed steam generators, condensation water splitter and a novel swirling collector. The device is small in size and light in weight, making it easy to move and suitable to use in different field sampling applications. EXPERIMENTAL APPROACH Design of the Aerosol to Hydrosol Air Sampler (ATHAS) Fig. 1 shows the overall system diagram, which consists of three major components: (i) steam generator and mixing chamber, (ii) condenser and (iii) collector. The total height of the system is 7 cm. The steam generator forms the upper part of the system with four steam generating units placed in an angle towards the axis of system inlet. Inside each steam generating unit, is a PID (Proportional-Integral- Derivative) controlled electric heater tape wrapped stainless steel capillary tube. Different steam temperatures were evaluated, and by observation lower temperature setting was not able to generate enough steam resulting in water droplets directly coming out of capillary tube, while higher temperature created burst of steam that is not stable for aerosol mixing. Throughout the study, the heater has a set temperature of 14 C with optimum operation condition. A peristaltic pump transports deionized water at the flow rate of 1 milliliter per minute to one end of capillary tube. The steam passes through the heated tube and exits from the lower end of tube, where it mixes with particle laden air stream. Despite of the high injected steam temperature, the rapid cooling and mixing with incoming aerosol significantly lowered the mixed air temperature to 4 45 C thus suppressing the possible side effect of alternation of chemical and biological properties of particles. The middle part of the system consists of the condenser with a total Table 1. Summary of aerosol collectors with liquid interface in literature. ATHAS in this study Steam Jet Aerosol Collector High Volume Apparatus High-Volume Aerosol-into-Liquid Collector Cyclone with recirculating liquid film PILS Batch-type wetted wall cyclone PSL:.5 µm: 65% 1 µm: 89% 2 µm: 99% Ultrafine particles - 99% Particle loss in system is 1%.1 µm PSL: 9% 1 µm PSL: 99%.5 µm PSL: 7 3% 1 µm PSL: 7 9% (NH4)2SO4.1 µm, 8% 1 µm, 1% 1 µm psl - 51% 3 µm psl - 55% Collection efficiency 7 15 ml per hour depending on aerosol types Flow rate 4 L/min 5 L/min 8 1 L/min 2 L/min 25 L/min 22.5 L/min 5 L/min Liquid collection Liquid flow of.1 rate ml per minute N.A 4 6 ml per hour N.A 3.5 g injection water vapor per minute Continuous liquid flow to keep 12 ml of liquid Integrated system frame: cm Mixing reservoir is Saturator only Ф cm. System dimension not provided. Water tank is 1 L container Cyclone height: 55 mm Diameter: mm System dimension N.A Dimension Cyclone dimension 4 cm 3 Condenser is 2 L container of Ф cm Liquid tank size unknown This study Khlystov et al. (1995) Wang et al. (213) Demokritou et al. (22) Tolchinsky et al. (21) Reference King et al. (29) Weber et al. (21)

3 Wubulihairen et al., Aerosol and Air Quality Research, 15: , Inlet (i) (ii) (iii) Fig. 1. Schematic diagram of compact Aerosol to Hydrosol Air Sampler (ATHAS): (i) steam generator and mixing chamber, (ii) condenser, (iii) collector. height of 25 cm. It has four tubes, each 2.5 cm in diameter (ϕi.d) and 25 cm in length, surrounded by liquid ethylene glycol. The liquid is cooled by a temperature controlled circulating liquid chiller. The condenser is covered with 1 cm of insulation. The mixture of water vapor and particles exits mixing section and passes to the chilled section, where particles grow through condensing of supersaturated vapor on to the surface of the particles. To separate the wall condensation water with the particle containing droplets, a special splitter was designed as shown in the top part of Fig. 2, in which four stainless steel air flow guiding tubes (ϕo.d = 2 mm; ϕi.d = 16 mm) were aligned concentrically with the condensation tubes. Cold wall condensation liquid flows by gravity inside the wall and accumulates in the bottom of the splitter, where it is continuously drained through a drainage port by peristaltic pump. The air flow, instead, carries the particle containing droplets to the swirling collector in the lower part of system (Fig. 2). The collector has four skewed nozzles mounted at the top with 9 intervals. The exit of each nozzle is 2.5 mm in diameter. The choice of the sampling flows as well as the nozzle design parameters is made so that the predicted cut point of the impactor is 1.2 µm based on impactor design criteria: Stk5 D p 5 2 CV p 9 Dan (1) where Stk5 is the Stokes number, Dp5 is the cutoff diameter of particle size with 5% efficiency, C is the Cunningham slip correction factor, V is the average velocity at acceleration nozzle exit plane, ρp is the particle density, Dan is acceleration nozzle diameter, µ is the dynamic viscosity of air. The choice of the cut point diameter serves only as a guideline since the grown droplets may shrink with decreased effective size due to the lower pressure zone inside collector. A vacuum pump is connected to the collector and the peristaltic pump used for transporting deionized water to the steam generator also extracts condensation water from the drainage port. Particle-laden air stream passes the condenser and enters the collector through nozzles such that the injected particles impact on the cone shaped inner wall and follow a swirling flow as shown with dashed line in Fig. 2. Ultimately, the particle containing droplets accumulate while following the flow and form large liquid droplets that drip down into the 5 ml centrifuge tube and yield particle suspension liquid samples. An aluminum frame with wheels was built for the system which is able to accommodate all components. The frame dimension is cm with total weight of about 4 kg. PSL Experiment for Size-Dependent Collection Efficiency Monodispersive particles of.1 µm,.5 µm, 1 µm and 2 µm were used to determine the size-dependent collection efficiency of sampler. Fig. 3 provides a diagram of the experimental setup. A six jet collision nebulizer (BGI CN25, BGI Inc., Waltham MA) was used to generate monodisperse polystyrene latex (PSL) particles (Fluoro-max, Thermo Scienctific) suspended in distilled water (5 ml) at a flow rate of 5 L/min. The size distribution of generated particles was measured with a SMPS (38, TSI) to ensure they were monodispersed. Particles pass through a mixing chamber where they are neutralized by an array of ionizers (Staticmaster Ionizers 2U5, Amstat, Glenview, IL),

4 Wubulihairen et al., Aerosol and Air Quality Research, 15: , Fig. 2. Schematic diagram of swirling aerosol collector. Pump Fig. 3. Schematic diagram of the PSL experimental setup for size-dependent collection efficiency of the sampler. followed by dilution and drying with particle free air filtered with HEPA (High Efficiency Particulate Air) filter. After passing through a mixing chamber, the air stream was split into two parts, one of which was used for collector sampling at flow rate of 5 L/min and the other used for reference sampling at a flow rate of 3 L/min. A filter holder with a 37 mm Teflon filter was used for reference sample collection. The 5 L/min stream entered the sampler, which was operated as described earlier. While sampling, the particle containing droplets continuously swirled before dripping down into the centrifuge tube. After an hour of sampling, the particle enriched liquid sample was collected in the centrifuge tube and the volume was measured from tube scale. The liquid sample was then transferred to a 1 cm quartz cuvette for fluorescence measurement after recording the total sample

5 78 Wubulihairen et al., Aerosol and Air Quality Research, 15: , 215 volume. The Teflon filter was rinsed with ethyl acetate (6 ml). Both samples were tested for fluorescence intensity with a multi-mode microplate reader (SpectraMax M5e, Sunnyvale, CA). The collection efficiency (E c ) was calculated from the following equation: E c f f S REF VS /5 V /3 REF where, f S and V S are the fluorescence intensity and the volume of liquid sample collected in the sampler, while f REF and V REF are the fluorescence intensity and volume of rinsed solution from the reference filter measured by the fluorimeter. Particle Loss, Condensational Growth, NaCl Test During the PSL experiments that evaluated the sizedependent collection efficiency, the mass balance of sample collection in the system was also checked for system losses of PM. A filter holder with 37 mm Teflon filter was placed in the effluent flow section to check if particles escaped from the collector. The loss on various collector surfaces was assessed by washing the inner surface wall of the collector with ethyl acetate (6 ml), and rinsing the 4 nozzles with ethyl acetate (6 ml). Therefore, three additional samples were taken and their fluorescence intensity and volume were compared with that of the reference sample after the same steps for collection efficiency. The experiments were repeated three times for each test. Proper conditions for condensational growth were assessed by testing under three different cooler temperature conditions (a. 8 C to 4 C, b. C, c. 4 C to 8 C). Lower temperature was not attempted to prevent the liquid freezing inside the condenser tubes. The test aerosol temperature was maintained at around 22 C 26 C throughout the entire tests by room air conditioning. Prior to sampling, an Aerodynamic Particle Sizer (APS 3321, TSI) was connected to one of the nozzles of the collector to measure the condensational growth of PSL particles (.1 µm and.5 µm). Bypass flow was connected to make sure of the same total flow rate of 5 L/min through the system and keep the same operating conditions. A Condensation Particle Counter (CPC 37, TSI) was connected to the reference sampling port. Flow through nebulizer was controlled so that the particle concentration at the reference sampling port was around cm 3, resembling particle concentrations found in typical urban environments. Typically, a liquid recovery rate of 7 15 ml per hour was observed for the lab aerosol tests. Sodium containing aerosols were generated by the nebulizer containing NaCl solution to check the collection efficiency for water soluble aerosols. The experimental setup was the same as for the PSL size-dependent collection efficiency and particle loss protocols. Samples collected on teflon filters were extracted with deionized water (6 ml). The nozzles and inner surface wall were each washed with deionized water (6 ml). After sample treatments, all samples were analyzed by Inductively Coupled Plasma-Optical Emission Spectroscopy (ICP-OES) to determine sodium ion (2) concentration. The collection efficiency (E c ) was calculated by the following equation: E c C C S REF VS /5 V /3 REF In which, C S and V S are the sodium ion concentration and volume of liquid sample collected in our sampler. C REF and V REF are the sodium ion concentration and volume of reference filter extractions. Preliminary Field Experiment Upon the completion of laboratory performance evaluation, a preliminary field experiment was carried out with the same system settings as an initial attempt to establish the field operation protocol for future investigations. The field experiment was conducted at an urban ambient site in Kowloon Tong ( N, E), located about 1.5 km north of Mongkok, one of the busiest commercial districts in Hong Kong. Ambient air was sampled directly at 5 L/min without size cut and two short sampling campaigns were performed on 18 th and 25 th of March, 214. For each sampling day, the system was thoroughly cleaned using ethanol and water rinsing, followed by warming up to the optimum laboratory conditions for one hour and then a continuous 3-hour sampling was performed during 13: to 16:. The sample was collected directly in the 5mL standard centrifuge tube and each of the 3 hour sampling yielded 25 ml of liquid sample. After volume reading, the liquid sample was preserved in freezer prior to analysis by Inductively Coupled Plasma Mass Spectrometry (ICP MS). We would also like to point out that this set of data serves only to show possible application of the system in environmental studies. Further intensive investigation and robust protocols will be developed in the next phase of study. Sample Analysis Detailed sample analysis protocol for ICP-OES and ICP-MS follows Jiang et al. (213) and a brief summary is described here. For Na ion analysis by ICP-OES, the analysis was carried out using an Optima 21 DV system (Perkin Elmer, USA) in scanning mode, with ion lenses tuned for maximum sensitivity. The plasma flow was set at 15 L/min, auxiliary flow at.3 L/min, and nebulizer flow at.8 L/min. The RF power was set at 1 3 Watt and flow rate of pump at 1. ml/min. Analytical drifts were corrected by spiking with five analytes using 1. mg/l standard solution. For ICP-MS analysis, In order to improve the extraction efficiency, 16 N nitric acid (purity > 65%, Merck KGaA, Germany) was added into 25 ml collected samples to make final concentration of 16.25% acid solutions in 5 ml metal free centrifuge tubes. Acid assisted extraction was carried out in a multi-tube vortex mixer (Model X-25, VWR). After 2 hours of extraction, the extracts were filtered with.22 µm filter membranes and stored at 4 C before test. Inductively Coupled Plasma Mass Spectrometry (ICP-MS) was used to analyze (3)

6 Wubulihairen et al., Aerosol and Air Quality Research, 15: , concentrations of trace metals that includes Al, Cd, Cr, Co, Cu, Pb, Mn, Mo, Ni, V, Zn. Parameters were set at RF power 16; gas flow at.24 L/min; carrier flow at.93 L/min; sample pump rate at 1. rps and nebulizer pump rate at.2 rps. RESULTS AND DISCUSSION Condensational Growth Fig. 4 shows the condensational growth of.1 µm and.5 µm PSL particles in 3 different cooler temperature conditions a). 8 C to 4 C; b). C; c). 4 C to 8 C, representing an average temperature difference of 2, 26 and 32 C between incoming aerosol and condensation tube wall. The wall temperature was in the range of 5 1 C depending on the cooler temperature settings. SMPS and APS were used to confirm the monodispersity of PSL particles prior to the tests to make sure the particle geometric standard deviation (G.S.D.) < 1.7. As shown in Figs. 4(a) and 4(b), when the cooler temperature is over C or above, particle sizes after growth showed evident multiple modal distributions with several peaks and the dominant mode diameters were 1.7 and 2.8 µm for.1 µm PSL and 1.7, 2.8 and 3.8 µm for.5 µm PSL particles, respectively, indicating inadequate supersaturation condition for complete particle growth. With decreasing condensation a) b) c).1 μm 16 8 C to 4 C 8 C to 4 C μm C C.1 μm.5 μm C to 8 C 4 C to 8 C μm.5 μm Fig. 4. Condensational growth of.1 µm and.5 µm PSL particles in 3 different cooler temperature conditions: a). 8 C to 4 C; b). C; c). 4 C to 8 C. 6 4

7 782 Wubulihairen et al., Aerosol and Air Quality Research, 15: , 215 tube wall temperature to 4 C to 8 C, there is a clear trend of enhanced particle growth in which peaks with larger size start to emerge and combine with smaller peaks forming a mono-modal distribution. The modal diameter of the particle distribution after growth was 1.7 µm and 2.8 µm for PSL particles of.1 and.5 µm, respectively. These also represent the optimum condition for the condensational growth performance in the designed system configuration. The same condition was then fixed for future efficiency and particle loss tests. Larger PSL particles of 1 µm and 2 µm were also tested with the same temperature settings showing clear particle growth performance but the results were not shown here since the sampler collection efficiency was mainly affected by the sub-micron particles. Size Dependent Collection Efficiency and Particle Loss With the optimal condition set for particle growth performance, size dependent PSL particle collection efficiency was further investigated. Fig. 6 shows the collection efficiency (denoted as Sampler in the plot) of the sampler for monodispersive PSL particle sizes ranging from.1 µm to 2 µm. Each data point represents the average of repeated tests results of at least three times for each particle size. Error bar is the standard deviation of the collection efficiency. For particle size larger than 1 µm, the collection efficiency was close to 9 1%, while.5 µm and.1 µm PSL particles had efficiency of 65% and 5%, respectively. The size dependent efficiency curve fits well with the classic impaction theory and it was possible that smaller particles that were not effectively condensed with water vapor were not able to grow to desired droplet size and escaped from the collector following the outlet flow. It is also possible that high speed air jet exiting the nozzles followed the swirling flow and may stick to the inner surface of the collector if desired continuous rinsing by water vapor was not well maintained, resulting in lower collection efficiency. Therefore, further particle loss tests were used to identify the mechanisms of particle loss inside the system. Fig. 5 also shows the percentage of particle losses that occurred in different parts of system including nozzles, collector inner wall surface and the outlet air flow. Particles with each size were tested for at least three times and the error bars represent the standard deviation from the tests results. A clear increasing trend of sampler collection efficiency with particle size was demonstrated while the particle loss in the nozzles and on the collector wall surface was higher for smaller particle sizes. The outlet particles also showed a clear decreasing trend with increased particle size from 37% for.1 µm to 12% for 2 µm PSL particles. The results indicate that majority of the particle losses occur by escaping from the sampler, which may be due to either the limitation of particle growth for small sized hydrophobic particles or the evaporation of water vapor from the particle containing water droplets surface that shrinks their effective size due to the lower pressure zone inside sampler before they impact on the inner wall surface, both of which facilitate the escape of small size particles from the collector. The other important particle loss fractions occur as the deposition of particles on the inner wall surface and in the nozzles, combined of which contributes from 22% to 1% of total incoming aerosol concentrations for.1 µm and 2 µm particles, respectively. The mechanisms of these particle losses may be due to the impaction and interception while aerosol flow carries the particles and changes direction through the bend nozzle tubes and along the inner wall, however, both mechanisms favor larger particles in contrast to the small particle escaping. The observed trend of decreasing losses with increasing size may be due to the insoluble characteristics of the PSL particles that when particle-containing droplets deposit and stick on the surfaces after particle growth, larger particles are possibly washed out more easily than small sized ones by more condensational water on their surface they carry with droplets. This can be also demonstrated by the soluble NaCl tests showing very little loss on the nozzle and inner wall surface by washing out. Details will be discussed in following sections. It should be noted that the highly hydrophobic nature of the PSL particles represent the extreme condition for the sampler application for atmospheric aerosol sampling. Ambient aerosols are largely hydrophilic with mixture of water soluble chemical components, such as sulfate, nitrate, sodium and etc, which makes it more favorable for water based condensational growth instruments. Soluble NaCl Collection Efficiency The soluble NaCl aerosol size distribution, sampler collection efficiency and system loss test results are shown in Fig. 6. The collection efficiency test was repeated three times and error bar in Fig. 6(b) represents the standard deviation of the results. The incoming NaCl particle size distribution was measured using SMPS and the mode diameter was 3 4 nm (Data not shown). After condensational growth, the NaCl particle size was enlarged to mode diameter of 3.5 µm. Although the NaCl size is significantly smaller than the tested PSL particles, the sampler had an average collection efficiency of 98% under the experiment conditions, clearly demonstrating the importance of aerosol hygroscopicity on the condensational growth efficiency and resulting sampling performance. The negligible outlet aerosol flow (1%) from the test results also confirms the effective growth of the NaCl that greatly limits their escaping from the collector. On the other hand, particle losses on the inner wall surface and in the nozzles were much lower (both 3% on average) compared with the PSL test results. As discussed earlier, the driving mechanisms of these particle losses by impaction and interception favor larger size particles, while the collection efficiency is also a function of particle size. The results from the NaCl tests showed both high collection efficiency and low losses in the nozzles and on the wall surfaces, demonstrating the effectiveness of the designed swirling collector in collecting particle containing droplets and the advantage of water condensational growth that facilitates the washing out of the deposited aerosols for much enhanced collection efficiency. Preliminary Field Experiment The preliminary field experiment produced 3-hour timeintegrated metals concentration of ambient aerosols as

8 Wubulihairen et al., Aerosol and Air Quality Research, 15: , SAC Sampler OUT Nozzle NOZZLE Surface SURFACE Out Percent (%) Fig. 5. PSL particles collection efficiency (Sampler) and amount of lost particles in nozzles (Nozzle), collector inner wall (Surface) and outlet air flow (Out) a) Percent (%) b) Sampler SAC Nozzle Surface Out Fig. 6. a) NaCl particle size distribution after growth; b) NaCl collection efficiency (Sampler), particles loss in nozzles (Nozzle), collector inner wall (Surface) and outlet air flow (Out) shown in Table 2, together with a comparison with results from a recent study carried out in the same area as reference (Jiang et al., 213). A good recovery of the trace metals in ambient particles was demonstrated for the two sets of 3- hour samples from different days. Although the reference data were sampled at different time intervals, a consistent relative abundance of the metals was clear indicating the feasibility of the sampler for high time resolution sampling and metal analysis. It should be also noted that the preliminary investigation serves only to explore the protocol establishment for field studies and also experiment possible applications of the system in environmental studies. Robust experiment protocols will be developed in further investigation with intensive field evaluation in future studies. SUMMARY AND APPLICATIONS This study describes the prototype development and laboratory performance testing of an aerosol to hydrosol air sampler (ATHAS). It has a unique structure and a novel aerosol collection method by combined effects of impaction and centrifugal motion. Size dependent collection efficiency was demonstrated for highly hydrophobic PSL particles with about 5% for 1 nm and more than 9% for supermicron particles. Mass balance tests showed the particle losses occur mostly on the nozzle outlets and inner wall of the sampler due to impaction. However, near 1% collection efficiency was demonstrated for soluble NaCl particles even though the mode diameter was only < 4 nm. The sampler integrates a few novel features including a specially designed splitter to prevent sample dilution by wall condensation liquid and a swirling sample collector compatible with standard centrifuge tubes. The laboratory performance tests also showed very promising results with ideal collection efficiency for soluble aerosol and classic performance curve for insoluble aerosols. For ambient aerosols largely containing soluble components, the sampler at current configuration has high potential for its application in sample collection

9 784 Wubulihairen et al., Aerosol and Air Quality Research, 15: , 215 Table 2. Preliminary data of field sampling of ambient particle metals comparison. The method detection limit of the metals for ICP-MS is.1 µg/m 3. µg/m 3 Preliminary test ( ) Preliminary test ( ) Jiang et al. (213) Al Cd N.A.1.3 Cr Cu Mn Mo Ni V N.A.8.8 Zn or semi-continuous online chemical analysis. For aerosols of hydrophobic nature, such as diesel exhaust emissions, the sampler has its major limitations due to the performance of condensational growth. Based on the system loss tests, further improvement can be possibly made in future investigations to optimize the sampler design for sub-1 nm aerosol collection including enhanced condensation growth and collector surface treatment to improve the efficiency. We will also carry out further field investigations on real world ambient aerosol to develop operational protocols and demonstrate its applicability and performance for sample collections and feasibility for future online chemical analysis. Compared with other existing technologies, the developed sampler has a medium size and sampling flow, presenting an attractive alternative for applications that require easier field deployment and higher PM collection rate for trace level chemical constituents analysis. ACKNOWLEDGMENTS The work described in this paper was supported by the grants from City University of Hong Kong (Project No and ). The authors wish to thank Prof. Peter Brimblecombe for the kind help in editing and proofreading the manuscript. REFERENCES Agarwal, J.K. and Sem, G.J. (198). Continuous Flow, Single-particle-counting Condensation Nucleus Counter. J. Aerosol Sci. 11: Ahn, K.H. and Liu, B.Y.H. (199). Particle Activation and Droplet Growth Processes in Condensation Nucleus Counter I Theoretical Background. J. Aerosol Sci. 21: Aitken, J. (1888). On the Number of Dust Particles in the Atmosphere. Proc. R. Soc. Edinb. 16: 135. Antonella, Z. and Joel, S. (29). The Effect of Fine and Coarse Particulate Air Pollution on Mortality: A National Analysis. Environ. Health Perspect. 117: Chow, J.C. (1995). Measurement Methods to Determine Compliance with Ambient Air Quality Standards for Suspended Particles. J. Air Waste Manage. Assoc. 45: Daher, N., Hasheminassaba, S., Shafer, M.M., Schauer, J.J. and Sioutas, C. (213). Seasonal and Spatial Variability in Chemical Composition and Mass Closure of Ambient Ultrafine Particles in the Megacity of Los Angeles. Environ. Sci. Processes Impacts 15: Demokritou, P., Gupta, T. and Koutrakis, P. (22). A High Volume Apparatus for the Condensational Growth of Ultrafine Particles for Inhalation Toxicological Studies. Aerosol Sci. Technol. 36: Dockery, D.W. and Pope, C.A. (1994). Acute Respiratory Effects of Particulate Air Pollution. Annu. Rev. Publ. Health 15: Grahame, T.J. and Schlesinger, R.B. (27). Health Effects of Airborne Particulate Matter: Do We Know Enough to Consider Regulating Specific Particle Types or Sources? Inhalation Toxicol. 19: Hering, S.V. and Stolzenburg, M.R. (25). A Method for Particle Size Amplification by Water Condensation in a Laminar, Thermally Diffusive Flow. Aerosol Sci. Technol. 39: Hering, S.V., Stolzenburg, M.R., Quant, F.R., Oberreit, D.R. and Keady, P.B. (25). A Laminar-Flow, Water- Based Condensation Particle Counter (WCPC). Aerosol Sci. Technol. 39: Ito, K., Chasteen, C.C., Chung, H.K., Poruthoor, S.K. Genfa, Z. and Dasgupta, P.K. (1998). A Continuous Monitoring System for Strong Acidity in Aerosols. Anal. Chem. 7: Jiang, S.Y.N., Yang, F., Chan, K.L. and Ning, Z. (214). Water Solubility of Metals in Coarse PM and PM 2.5 in Typical Urban Environment in Hong Kong. Atmos. Pollut. Res. 5: Khlystov, A., Wyers, G.P. and Slanina, J. (1995). The Steamjet Aerosol Collector. Atmos. Environ. 29: Kim, S., Jaques, P.A., Chang, M., Froines, J.R. and Sioutas, C. (21). Versatile Aerosol Concentration Enrichment System (VACES) for Simultaneous in Vivo and in Vitro Evaluation of Toxic Effects of Ultrafine, Fine and Coarse Ambient Particles Part I: Development and Laboratory Characterization. J. Aerosol Sci. 32: King, M.D. Thien, B.F., Tiirikainen, J.S. and McFarland, A.R. (29). Collection Characteristics of a Batch-type Wetted Wall Bioaerosol Sampling Cyclone. Aerobiologia 25: Nemmar, A., Delaunois, A., Nemery, B., Dessy-Doize, C., Beckers, J.F. Sulon, J. and Gustin, P. (1999). Inflammatory

10 Wubulihairen et al., Aerosol and Air Quality Research, 15: , Effect of Intratracheal Instillation of Ultrafine Particles in the Rabbit: Role of C-fibers and Mast Cells. Toxicol. Appl. Pharmacol. 16: Ning, Z., Geller, M.D., Moore, K.F., Sheesley, R., Schauer, J.J. and Sioutas, C. (27). Daily Variation in Chemical Characteristics of Urban Ultrafine Aerosols and Inference of Their Sources. Environ. Sci. Technol. 41: Orsini, D.A., Ma, Y., Sullivan, A., Sierau, B., Baumann, K. and Weber, R.J. (23). Refinements to the Particleinto-liquid Sampler (PILS) for Ground and Airborne Measurements of Water Soluble Aerosol Composition. Atmos. Environ. 37: Parshintsev, J., Kivilompolo, M., Ruiz-Jimenez, J., Hartonen, K., Kulmala, M. and Riekkola, M.L. (21). Particleinto-liquid Sampler on-line Coupled with Solid-phase Extraction-liquid Chromatography-mass Spectrometry for the Determination of Organic Acids in Atmospheric Aerosols. J. Chromatogr. A 1217: Pope, C.A. (2). Review: Epidemiological Basis for Particulate Air Pollution Health Standards. Aerosol Sci. Technol. 32: Pope, C.A., Ezzati, M. and Dockery, D.W. (29). Fine- Particulate Air Pollution and Life Expectancy in the United States. New Engl. J. Med. 36: Slanina, J., ten Brink, H.M., Otjes, R.P., Even, A., Jongejan, P., Khlystov, A., Waijers-Ijpelaan, A., Hu, M. and Lu, Y. (21). The Continuous Analysis of Nitrate and Ammonium in Aerosols by the Steam Jet Aerosol Collector (SJAC): Extension and Validation of the Methodology. Atmos. Environ. 35: Stone, V., Shaw, J., Brown, D.M., MacNee, W., Faux, S.P. and Donaldson, K. (1998). The Role of Oxidative Stress in the Prolonged Inhibitory Effect of Ultrafine Carbon Black on Epithelial Cell Function. Toxicol. Vitro 12: Tolchinsky, A.D., Sigaev, V.I., Sigaev, G.I., Varfolomeev, A.N., Zvyagina, E.V., Brasel, T. and Cheng, Y.S. (21). Development of a Personal Bioaerosol Sampler Based on a Conical Cyclone with Recirculating Liquid Film. J. Occup. Environ. Hyg. 7: Wang, D., Pakbin, P., Saffari, A., Shafer, M.M., Schauer, J.J. and Sioutas, C. (213). Development and Evaluation of a High-Volume Aerosol-into-Liquid Collector for Fine and Ultrafine Particulate Matter. Aerosol Sci. Technol. 47: Weber, R.J., Orsini, D., Daun, Y., Lee, Y.N., Klotz, P.J. and Brechtel, F. (21). A Particle-into-Liquid Collector for Rapid Measurement of Aerosol Bulk Chemical Composition. Aerosol Sci. Technol. 35: Wilson, M.R., Lightbody, J.H., Donaldson, K., Sales, J. and Stone, V. (22). Interactions between Ultrafine Particles and Transition Metals in Vivo and in Vitro. Toxicol. Appl. Pharmacol. 184: Received for review, August 23, 214 Revised, November 3, 214 Accepted, December 22, 214