Home

Particle Analysis

Phase II—Results

Comparisons Between Locations

Particle Concentrations

The graph in fig.3 shows the particle concentrations in order of increasing particle concentration. Orange implies urban, blue is suburban, and green is rural. The patterned bars indicate that that leaf was a Black Locust rather than a Honeylocust. What is interesting about this graph is that, although it is set to show the concentrations in increasing order, it also shows the leaves in order of increasing proximity to heavily travelled roads or highways. This points out another factor; the amount of traffic in the area that the leaf is collected. Leaf II-4 was located approximately three meters from a road heavily trafficked by diesel-fueled trucks from a nearby quarry, as well as regular traffic. Leaf II-12 was located in a cemetery approximately 7.6 meters from a very quiet road. Therefore, although this doesn’t follow the hypothesis that concentrations would be highest in urban areas, it is understandable on the basis of proximity to roads and the amount of traffic on those roads.

Particle Sizes

Fig.4 shows the median particle diameters from the Phase II leaves in order from urban to rural areas. The diameters are shown to be highly variable. This is probably based on the number of factors that influence the size of the particles in the area, such as the amount of traffic, etc. However, the graph does show the smallest median diameter is on a rural leaf (leaf II-5) and the largest is on a suburban leaf (leaf II-3). Also to be noted is the fact that both leaves II-1 and II-2, which were collected from suburban locations only approximately ten meters apart, show median diameters that are lower than the other suburban locations. This would be expected due to their close proximity to each other.

Elemental Compostion

The graph shown in fig.5 shows the average percent elemental composition of selected elements in the particles on each leaf, in order from urban to rural. One can see from this graph that several of the elements analyzed accounted for nearly 100% of the elemental composition of particles on most leaves analyzed. One can see that the lowest percentage of Si is on leaf II-4, which also has the highest percentage of Ca. The lowest percentage of Ca is on leaf II-6, a rural leaf, which also contained a high percentage of Si. Leaf II-13, an urban leaf, contained the highest percentage of Fe. Leaves II-1 and II-2, from very close locations, showed very similar percentages, except for a slightly higher percentage of K in leaf II-2. There appears to be no trend from urban to rural, which is probably also caused by the high number of factors influencing this data.

Effects of Rainfall

Particle Concentrations

The particle concentrations on leaves collected from the same tree before and after rain changed dramatically. Leaf II-3 was collected from a suburban tree during a prolonged dry period (see fig.6). Leaf II-16 was collected from the same tree after the first rain following the dry period, and leaf II-17 was collected several days after that, following more light rain. The particle concentration decreased from approximately 2100 particles/mm² before the rain (leaf II-3) to approximately 154 particles/mm² immediately following the rain (see fig.7), then increased to approximately 568 particles/mm² on leaf II-17. This shows that there was a large decrease in concentration caused by the rainfall. It is difficult to tell whether the regaining of particles by leaf II-17 was caused by further deposition after the rain or whether it was caused by technical differences (detection thresholding, etc.).

Particle Sizes

Fig.8 shows the median particle diameter on leaves from the same tree before (II-3) and after rainfall (II-16, II-17). One can see that the median diameter decreased with rainfall. These findings are similar to the findings of Johnson et al, which showed that particle diameters on leaves decreased with rinsing. This led me to design and perform rinsing experiments (see section on rinsing).

Elemental Composition

The leaves collected after rainfall, II-16 and II-17, showed less Ca and more Si as a percent of the total than the sample from the same tree, II-3, collected before the rain (See fig.9). Fig.10 displays the absolute concentrations of elements in particulates on the leaf surfaces (calculated by multiplying average elemental percent by particles/mm²). This demonstrates more dramatically the reduction in all elements, but especially in Ca.

Effects of Rinsing

Particle Concentrations

The graph in fig.11 shows the particle concentrations for the artificially rinsed leaves. The two leaves collected before the rainfall, leaves II-3 and II-4, showed decreases in concentration with rinsing. However, the leaf collected after rainfall, leaf II-16, showed negligible change. Because leaf II-16 was already washed by the rainfall, the rinsing was not expected to change it any further.

Particle Sizes

The median particle diameters on the rinsed leaves are shown in the graph in fig.12. Very little change was seen in the leaves. The reason that this does not match the results from the naturally rinsed leaves is unknown at present. Further studies are necessary to determine this.

Elemental Composition

In leaf II-3, the percentage of Ca decreased with rinsing and the percentage Cl and Si , as well as K increased (see fig.13). In leaf II-4, there was a slight decrease in Ca, with several elements increasing slightly in terms of fractional percentage composition. Leaf II-16 showed a slight increase in Si as the only apparent change. The findings for leaf II-3 indicate that besides rinsing off soluble Ca-containing particles, there may be some precipitation of solubilized K and Cl. It should be noted that these percentages are not adjusted for changes in absolute concentration of particles.

Comparisons with Aerosol Filter Samples

Concentrations were not examined on filter samples because they represent a different particle collection method, and concentrations are therefore not comparable to those of leaf surface sampling.

Particle Sizes

In general, the median particle diameters on the aerosol filter samples were lower than those on the leaf samples. Since the particles were collected by airflow through the filter and not by sedimentation as those on the leaves, one would expect to see smaller particles on the filter surfaces – those that are too small to settle out onto the leaves. Also the smallest particles (see fig.14) were seen in the rural sample, s#9 (which corresponds to rural leaf samples I-8 and II-5). The urban sample, s#8, had the highest median diameter. This sample corresponds to urban leaf samples I-9 and II-15. Suburban sample ns#27 showed an intermediate median diameter. This sample location corresponds to samples I-7 and II-3, II-16, and II-17. These findings are consistent with the findings from the leaf samples, and as expected, the rural sample showed finer particulates than the others.

Elemental Compositions

The percentage elemental composition on the aerosol filter samples was quite different from that found on the leaf samples. Fig.15 shows that, on the filters, S was a major component of the aerosol. Na was also more prevalent, especially in the rural and suburban samples. The Ca percentage was high in both the urban and suburban samples, but was much lower in the rural sample. Substantial Fe was detected in the urban aerosol sample. Si, which was a major element on the leaf sample, was less prevalent on the filters. This type of display showing the elemental percentages averaged over all the particles analyzed does not reveal the associations between elements on individual particles, which data is one of the most valuable features of SEM-EDS of individual particles. An example of graphical analysis of the relationships between elements in individual particles is shown in fig.16.

The graph in fig.16 shows Ca vs. S percentage composition for each individual particle on the three filter samples. This clearly shows the basis for the percentage differences seen in the previous figure. For the rural sample, s#9, the S-rich particles contain very little Ca. In the urban sample (s#8), in contrast, there is a cluster of particles containing nearly equal Ca and S. The suburban sample, ns#27, shows some Ca=S (probable CaSO4), and also a large cluster of Ca-rich particles with lower S, plus a cluster of S-rich particles.