The Dew Particle Interception Abilities of Typical Plants in Northeast China Plant Leaves Capture Particles in Dew

e dew condensation frequency is high, and the dew amount is heavy in urban ecosystems. During the condensation process, particulate matter acts as a condensation core, playing an important role in purifying the air. At night, dew mainly condenses on plant leaf surfaces, the plant leaves settle the particles in the dew, and some of the particles are resuspended into the atmosphere in the process of dew evaporation after sunrise. is paper monitored the condensation and evaporation processes of dew on four common plants in Changchun city from June to September 2020. By analyzing the mass and size of particles on dierent leaves after dew condensation and evaporation, the ability of dierent plants to retain particles in dew was analyzed. e results showed that there was no signicant dierence in the TSP capture ability during dew condensation between Buxus sinica (Rehd. et Wils.) Cheng subsp. sinica var. parvifolia M. Cheng, Syringa oblata Lindl., Hemiptelea davidii (Hance) Planch., and Pinus tabuliformis Carrière, with a TSP content of 0.21± 0.06 μg/cm. Coarse particulate matter is the main type of deposit in the dew condensation stage. Particulate deposition varied according to species, leaf shape, and microstructure. e proportion of TSP remaining on leaves after dew evaporation from Pinus tabuliformis Carrière, Hemiptelea davidii (Hance) Planch., Buxus sinica (Rehd. et Wils.) Cheng subsp. sinica var. parvifolia M. Cheng, and Syringa oblata Lindl. tree was 89.7± 3.9%, 80.6± 3.6%, 75.9± 4.5%, and 71.4± 3.7%, respectively. e ability of the leaves to trap ne particles was signicantly higher than that for coarse particles (P< 0.05) after dew evaporation.e highest amount of particle captured by Syringa oblata Lindl. individual was 15.17 g/y during dew condensation, and the amount of remaining particles after dew evaporation was 10.83 g/y. is paper provides a theoretical basis for the selection of tree species for urban greening.


Introduction
Atmospheric particulate matter poses a threat to human health, especially to the human respiratory system and immune function [1]. e urban plants are considered to play the role of natural lter in the interception and removal of particulate matter [2]. In Beijing, 772 t of PM 10 was captured by leaves in one year [3]. e dust retaining capacity of leaves was 8012.89 t/y in Guangzhou [4]. e urban canopy of the Greater London Authority is currently estimated to remove between 852 and 2121 t of PM 10 annually, representing between 0.7% and 1.4% of PM 10 from the urban boundary layer [5]. Dew condensation is a common weather phenomenon, and it is also an e ective natural atmospheric puri cation process with the particles in the air serving as a condensation core [6,7]. Particles in the atmosphere, as condensation cores, experience the process of moisture absorption and homogeneous-heterogeneous phase transition, nally turning to dew. During this condensation stage, particles migrate and settle toward the surface continuously with the condensation of dew, and the dew becomes the sink of near-surface particles (PM 2.5 , PM 10 , etc.) at night [8,9].
is leads to the concentration of particulate matter in dew being signi cantly higher than that in rainwater [10,11], especially when the concentrations of total dissolved solids (TDS) are high [12]. e average TDS in the dew of Changchun (China) [10] and Delhi (India) [13] can reach 154 and 135 mg/L, respectively.
Due to the large-scale hardening of underlying surfaces and the heat island e ect in urban ecosystems, urban areas have become a "landscape unit" with frequent dew condensation, which not only occurs at a high frequency but also in large amount [8,14]. In Changchun, for example, the annual dew amount in 2017 was 35 mm, and the mass concentrations of TDS, PM 2.5 , and PM 10 in the dew were 175. 31, 24.39, and 49.65 mg/L, respectively. In normal weather, the removal rates of TSP, PM 10 , and PM 2.5 by dew were 28.2%, 25.9%, and 21.5%, respectively [15]. e annual removal of particulate matter by dew was 3.11-4.73 kg/ha. However, atmospheric particulate matter temporarily settles on the leaf surfaces of plants and returns to the air via resuspension [16,17]. Dew is effective at removing particulate matter from the air near surfaces in urban ecosystems. As an intermediate product forming and dissipating by day and night, the dew evaporates after sunrise, and the particulate matter condensed at night is transferred from the sink to the source during the evaporation process. e PM 2.5 concentration is highest in the morning but lower in the afternoon and evening in the urban forest of Beijing [18]. Generally, the concentrations of PM 10 and PM 2.5 near the surface are minimal from 22 : 00 to 3 : 00 and peak from 7 : 00 to 9 : 00 [18,19], coinciding with the dew evaporation period [20]. Dew condensation occurs between 20 : 00 and 5 : 00, and the particles can be reduced by wet deposition of dew [21,22]. e low concentration of particulate matter in the atmosphere at night is closely related to the condensation of dew. However, whether the peak concentration of particulate matter in the atmosphere is related to dew evaporation has not yet been determined.
Because of the hygroscopic behavior of aerosol particles [23,24], humidity can affect the particle shape and size [25,26]. e particle size in the atmosphere will change due to aerosol hygroscopicity or volatility whether in the process of dew condensation or evaporation, and it is very complicated and hard to measure.
ere are some advanced techniques such as ultraviolet aerodynamic particle sizer (UV-APS) fluorescence techniques or fluorescence in situ hybridization (FISH) to measure biological aerosol [27][28][29]. In this research, the particles deposited on dew condensation and evaporation nodes were observed. e sediment particles and the leaf microstructures were examined by scanning electron microscopy (SEM) in this study. In addition, studies on the interception of particulate matter in plant leaves focus on the interception of dry deposits over time [30]. Some studies identify the source of aerosols by analyzing the chemical composition and backward trajectories [31,32]. In this study, more attention was paid to the retention capacity of leaves. ere is a large and frequent amount of dew water on the leaves of plants in the urban ecosystem, and the concentration of particulate matter is high. erefore, it is urgent to discuss the amount of particulate matter on leaves in dew deposited at night and evaporated in the morning. e particles monitored in this study are the sediment particles intercepted by leaves. e purpose of this paper is to investigate the capacity of common plants to retain particles of different sizes during the dew condensation and evaporation processes and to discuss the main influencing factors of particles trapped by leaves in dew. e particles measured in this study were naturally captured by leaves, and it will improve the understanding of the ability of dew material captured by different leaves in the urban ecosystems.

Experimental Site.
Dew and leaf samples were collected at Jilin Jianzhu University, which is located in the southeastern part of Changchun in Jilin Province, China. e continental monsoon climate of the area produces average annual temperatures that vary from −15°C in January to 25°C in August. Precipitation ranges from 522-615 mm/a. e research area is located at the edge of the city, with no industrial and traffic pollution. e dew intensity in summer and autumn is relatively high. For example, in 2014 and 2015, the amount of dew in July, August, and September in the Changchun green area was 46.29 mm and 72.24 mm, respectively, accounting for 22.52% and 23.61% of the precipitation in the same period [15]. erefore, samples were collected during the peak dew period of summer and fall (June-September) in 2020.  Table 1. To avoid differences in particle interception under different pollution conditions, the selected tree species were all in relatively concentrated areas (see Figure 1).

Time of Dew Evaporation and Condensation Periods.
A tray (diameter of 30 cm) with turf covering both sides was placed on an electronic balance (range of 0.01 to 3000 g). e balance was placed in a waterproof box (with small holes to discharge rain water), and the box was placed on an observation shelf. e observation shelf was equipped with an arm that adjusts in height according to the growth of plants so that the tray and the plant canopy remained at the same height. e weight change of the turf can be continuously monitored, and it is the objective indicator as dew condensation or evaporation. e initial value was recorded half an hour after sunset; when the peak value was reached (approximately half an hour before sunrise), the weight gain value was taken as the dew condensation amount, and the corresponding period was considered the dew condensation period. When the value dropped to a constant weight (the value remained unchanged for half an hour), the dew was considered to have evaporated completely, and the corresponding period was defined as the dew evaporation period.

Particulate Matter Mass in Dew.
Four plant leaves were collected from four directions of the branches at random points in the tree canopy. irty to fifty leaves were collected, sealed to avoid vibration, and immediately brought back to the laboratory. Cotton balls moistened with alcohol and deionized water were used to thoroughly wash and wipe the surfaces of leaves, to remove dust from the leaves, and to reach an initial state of zero dust accumulation. e leaves were divided into two groups. e leaves in each group were evenly divided into two parts, with half adaxial surface up and half abaxial surface down. e area of leaves (cm 2 ) was recorded. e leaves were fixed on a polystyrene foam board. e polystyrene board was kept parallel to the surface and was placed at the height of the plant canopy at the beginning of dew condensation.
At the point node when dew condensation turns to evaporation, a piece of foam board was sealed and brought back to the laboratory. e 10.00, 2.50, and 0.22 μm membrane filters were dried (65°C) to constant weight (W jf , g). e surface of the leaves was washed using the ultrasonic cleaner with deionized water for 1 hour, and leaching water was filtered through the filters. e filters were weighed again after dried (65°C), and the weight of filters W jc (g) was recorded. e TSP, PM 2.5 , and PM 10 values of the dew were calculated based on the weight differences. After the dew evaporated completely, the other foam plate were sealed and brought back to the laboratory. e above steps for weighing leaves (W e , g) and determining particle residue weight (W je , g) were repeated. All particles were naturally intercepted by leaves, and it was passive sampling. PM 2.5 /PM 10 /TSP retention by individual plants during dew condensation and after evaporation was calculated as follows: where j is the type of PM 2.5 , PM 10 , or TSP; F j is the weight of particles during dew condensation for a single plant (mg/ plant); W jc is the weight of particulate matter in deionized water (washed off the leaves) when the dew condenses to the evaporation node (g); W jf is the initial weight of the filter paper (g); n is the number of collected leaves on one polystyrene foam; N is the number of leaves on one plant; 2 is the transfer coefficient; P j is the weight of particles after dew

Microstructural Analysis of Leaves.
e microstructures of the leaves was examined using scanning electron microscopy (SEM). e drop contact angle (DCA) was determined using a Canon EOS550D camera attached to a macrolens (MP-E 65 MM 1 : 2.8) [33]. A 6 μL droplet (for broadleaves) and a 2 μL droplet (for needles) of distilled water were placed on the leaf sample.
irty leaves were randomly selected with 3 replicates for each species. Total leaf area was recorded (S). e weighing bottle weight was recorded with a 0.0001 g analytical balance (W 0 ). Leaf blades were held with tweezers and sprayed with chloroform onto the front or back surfaces for 60 seconds. e extraction solution was then transferred into the weighing bottle.
e weighing bottle was weighed on a balance again when the chloroform was completely volatilized in a fume cupboard (W 1 ). e twice difference (W 1 −W 0 ) of the wax content was measured. e wax content per unit leaf area (W, g/m 2 ) was determined based on the following formula:

Particles Settled by Leaves in the Dew Condensation Stage.
As shown in Figure 2, the TSP contents settled by Pinus, Hemiptelea, Buxus, and Syringa leaves in the dew condensation stage were 0.21 ± 0.07, 0.21 ± 0.06, 0.22 ± 0.06, and 0.21 ± 0.06 μg/cm 2 , respectively. ere was no significant difference in TSPs per unit area among leaves of each tree.
Coarse particulate matter is the main type of deposit captured by leaves in the dew condensation stage. As shown in Figure 3, the four kinds of trees mainly removed coarse particles (>PM 10 ) in the dew deposition stage. e proportions of PM 2.5 , PM 2.5-10 , and >PM 10 were 9.2%-16.0%, 26.3%-32.2%, and 56.7%-62.4%%, respectively. As evergreen tree (Pinus), the interception of PM 2.5 proportion was significantly higher compared to other deciduous trees (Hemiptelea, Buxus, and Syringa) (P < 0.05), but there was no significant difference in the PM 2.5-10 and >PM 10 proportion (P > 0.05). It indicated that evergreen plants were better at trapping fine particles. is is similar to the results of other studies. More than 80% of the particles intercepted were PM 10 of several common green plants in Qingdao [30]. PM larger than 10 μm comprised 73% of PM deposited on leaves in Beijing, with the coarse and fine particle size fractions comprising 16% and 11% of the deposited PM, respectively [22].

Particles Settled by Leaves after Dew Evaporation.
After dew evaporation, the ability of the four kinds of trees to retain particulate matter was obviously different. e proportions of intercepted TSP settled by Pinus, Hemiptelea, Buxus, and Syringa leaves were 89.7 ± 3.9%, 80.6 ± 3.6%, 75.9 ± 4.5%, and 71.4 ± 3.7%, respectively (see Figure 4). e ability of Pinus (evergreen) to trap particles after dew evaporation was significantly higher compared to several other deciduous tree species. e ability of the four kinds of trees to trap fine particles on their leaves was significantly higher than that for coarse particles (P < 0.05). Taking Syringa as an example, the leaf interception ratios of PM 10 , PM 2.5-10 , and PM 2.5 after dew evaporation were 65.1 ± 3.9%, 83.7 ± 5.5%, and 88.5 ± 7.3%, respectively. After the dew evaporation process, some of the particles were still trapped on the surface of the leaves, and some of them left. e proportion of fine particles trapped in the leaves was higher. After dew evaporation, the particles trapped on the leaves were still mainly coarse particles, but the proportion of PM 2.5-10 increased to 30.7%-33.7%, and the proportion of PM 2.5 increased to 11.6%-17.6%. Both the dew condensation and the secondary suspension process were dominated by coarse particles.
ere was no significant difference in the mass of particulate matter captured by leaves of different tree species per unit area during dew condensation, but the deposited PM loads varied significantly among individual tree species due to the leaf area (see Table 1). Syringa showed the highest PM deposition, followed by Hemiptelea, Pinus, and Buxus. According to the calculation of 130 dew days in one year, the settlement of TSP by one Syringa, Hemiptelea, Pinus, and Buxus individual was 15.17 g/y, 12.87 g/y, 2.97 g/y, and 1.07 g/y, respectively, during dew condensation. e remaining TSP trapped by these leaves was 10.83 g/y, 10.35 g/y, 2.67 g/y, and 0.81 g/y, respectively, after dew evaporation. Evergreen or deciduous plants cannot differentiate the intercept of the TSP amount. As the leaves' area increases for each species, the TSP values should be slightly higher than the actual TSP considering the leaves inclination and mutual shadowing with respect to the vertical direction.

Factors Affecting Leaf Interception of Particles.
e relationship between precipitation and particles captured by leaves of the four plants during the dew deposition period was close. e mass of particles on leaves decreased after precipitation (see Figures 2 and 5). For example, the cumulative precipitation reached 53.3 mm from August 13 to 16, and on August 18, the deposition of dew particles in the four plants showed a significant declining trend. is may have been due to purification by precipitation, i.e., the dew captured few particles and resulted in a reduction in 4 Advances in Meteorology Advances in Meteorology interception [10]. With the increase in pollutants in the atmosphere after rain, samples were taken continuously on August 19 and 20, and the deposition of particles in dew gradually increased. As demonstrated in Table 2, there was no relationship between wind speed and particulate matter interception amount of four plants (P > 0.05). Some researchers found the correlation was positive or negative depending on the local wind speed values, and there was a threshold in the correlation between wind speed and particle concentration [34,35]. In this study, neither mean wind speed (0.83 m/s) nor 1 m/s wind speed was related to particulate interception (P > 0.05). e wind speed had no significant effect on the particulate matter during the dew evaporation period. is is because dew formation events always occur at wind speeds under 2 m/s, and high wind speed conditions are not conducive to dew formation. In addition, the dew evaporation period lasted 3 to 4 hours, and wind has little effect on particulate matter in a short time. e relative humidity during the dew evaporation period was positively correlated with the interception amount (P < 0.05) because the particulate matter was not easily dispersed with high air humidity. Air temperature or dew point temperature had no correlation with the interception amount (P > 0.05). e four kinds of trees showed little difference in terms of trapping particulate matter in the dew deposition stage, but there were significant differences in their ability to trap particles after the evaporation process (see Figure 4). e ability of plants to capture particle matter is related to multiple factors, including leaf shape and the leaf surface microstructure [36]. According to their leaf morphology, conifers have the strongest ability to trap particles in dew. Conifers exhibited higher PM 2.5 adsorption amounts than broadleaf trees [37]. e oil secreted by Pinus leaves has strong particle adsorption properties (see Figure 6(c)), which also makes it difficult for particles to fall off. e leaves of Hemiptelea have serrated edges (see Figure 7(c)), which can also make them more strongly adsorb particles in dew.    From the aspect of microstructure, the trichome density and length of Hemiptelea leaves was significantly higher compared to Syringa and Buxus leaves (see Figure 6). e trichome surface structures were helpful in enhancing the effective surface area for PM adsorption. Species with trichomes accumulated significantly more particulate matter than other species [38]. Particulate matter adsorption can be observed at the tip and base of the trichomes of Hemiptelea (see Figures 6(d), 6(g), and 6(j)). e difference between the removal efficiencies of T. usneoides with and without trichome structures was approximately 7% for PM 2.5 and 2% for PM 10 [39]. e effect of trichome structure is particularly notable for the adsorption of PM 2.5 . Trichome density has a significantly positive correlation with the maximum particle size (P < 0.05), and when the wind force is weak, stomatal density and trichome density have a significant effect on particulate matter resuspension [17]. erefore, Hemiptelea can capture more PM (see Figure 6(f )) than Syringa and Buxus.
As Figure 6 shows, many fine and coarse particles accumulated on the adaxial leaf surfaces, while some large particles attached to the abaxial leaf surfaces, and few particles became attached in the vicinity of the stomata. Grooves are the main parts of a blade that absorb PM 2.5 . Deep grooves can intercept more particles (see Figure 6(k)); additionally, deeper grooves make the release of PM less likely. Ridge ultrastructures at 1-2 μm scale are the most efficient at capturing PM. e stomatal densities of Syringa, Hemiptelea, and Buxus were 213.0 ± 20.5/mm 2 , 92.5 ± 7.3/ mm 2 , and 84.5 ± 18.6/mm 2 , respectively. Both stomata and trichomes are adjunct structures that are conducive to the adsorption of particles. Different shapes of leaves captured particles in dew with different ways. Evergreen tree with acicular leaves such as Pinus adhered to particulate matter by the secreted oil. While deciduous plants trapped to particulate matter around its unique microstructure, deep grooves or dense trichome such as the elliptical leave (Hemiptelea) used dense trichome to trap particulate matter (see Figure 6(f )) and the compound leave (Syringa) used deep grooves (see Figure 6(k)).

Conclusions
rough the monitoring of leaf particle deposition during the dew condensation and evaporation periods of four common urban plants, it was found that the amounts of TSPs settled by Pinus, Hemiptelea, Buxus, and Syringa trees in the dew condensation period were 0.21 ± 0.07, 0.21 ± 0.06, 0.22 ± 0.06, and 0.21 ± 0.06 μg/cm 2 , respectively.
ere was no significant difference in the TSP 8 Advances in Meteorology amounts in the dew condensation period by leaves per unit area among tree species (P > 0.05). Coarse particulate matter (>PM 10 ) accounted for 56.7%-62.4% of PM and was the main type of deposit in the dew condensation stage. According to the difference in leaf shape (conifer and broadleaf ), microstructure (stomata and trichomes), and hydrophobicity, the proportion of the TSP trapped by Pinus, Hemiptelea, Buxus, and Syringa was 89.7 ± 3.9%, 80.6 ± 3.6%, 75.9 ± 4.5%, and 71.4 ± 3.7%, respectively, after dew evaporation. e fine particle interception ability of leaves accounted for 88.5 ± 7.3% of the condensed fine particles for the four kinds of trees and was significantly higher than that for coarse particles (65.1 ± 3.9%) (P < 0.05) during the dew evaporation process. Temperature and wind speed had no significant effect on the secondary suspension of particles in the dew evaporation period (P > 0.05). More particles were trapped by leaves at higher relative humidity in the morning than at other times. Based on 130 dew days per year, the settlement of TSP by one Syringa, Hemiptelea, Pinus, and Buxus individual was 15.17 g/y, 12.87 g/y, 2.97 g/ y, and 1.07 g/y, respectively, during dew condensation, and the corresponding values were 10.83 g/y, 10.35 g/y, 2.67 g/y, and 0.81 g/y after dew evaporation. Identifying sources of particles with different particle sizes can help to understand the trajectory of particles and reveal the role of plants in trapping particles. So, the source analysis of particulate matter should be further discussed.
Data Availability e datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Conflicts of Interest
e authors declare that they have no conflicts of interest.