NOTES
Aerosol Deposition on Plant Leaves James B. Wedding,' Roger W. Carlson, James J. Stukel," and Fakhri A. Bazzaz University of Illinois at Urbana-Champaign, Urbana, Ill. 61801
An aerosol generator and wind tunnel fumigation system designed for use in studies of aerosol deposition on leaf surfaces is described. Gross deposition on rough pubescent leaves was 10 times greater than on smooth, waxy leaves. Results suggest that aerosol deposition, on a per unit area basis, for single horizontal streamlining leaves, is similar to that for arrays of leaves under similar flow conditions.
It is well-known that the primary source of lead in most communities is from the particulate emissions from automobiles. This lead is transported by the atmosphere from the highway and deposited on soil surfaces and/or directly on plant surfaces. The deposited lead may either enter plants through the roots in the soil (1-3) or through the leaf surfaces after aerosol deposition (3-5). This study will examine the deposition of lead aerosols on leaf surfaces. The deposition of P b on leaves is controlled by characteristics of the aerosol (size, chemical composition, cloud density), the leaf surface (roughness, pubescence, moisture), and the environment in which the plant lives (relative humidity, wind speed). Once a lead aerosol is deposited on a leaf surface it may gain entrance into the interior of the plant and be translocated throughout the plant to cause adverse effects on the plant's physiological processes. It is this possibility of physiological harm to the plant that forms the impetus for this study. By being able accurately to predict the extenr, of particle deposition on individual plant surfaces or arrays of leaves under various conditions, an estimation of the plant dosage of lead can be ascertained. This paper will address the questions concerning the correlation of the aerosol deposition rate for single and multiple leaf clusters. In order that the effects of surface roughness on aerosol leaf deposition might be bounded, two plant species representing extreme cases were examined. These were the tulip poplar and the sunflower leaf. Since the main objective of this investigation was to study the gross deposition of an aerosol on plant leaves in a well-defined flow field and not the resulting physiological response, it was proposed r,hat a uranine aerosol be used in place of a lead aerosol. The primary reason for this was the ease with which uranine particles can be analyzed as opposed to lead aerosols. To check the validity of this approach, an initial series of experiments was run to demonstrate the feasibility of substituting uranine partiPresent address, Engineering Research Center, Colorado State University, Fort Collins, Colo. 80521.
cles for lead aerosols in the deposition studies. A second series of experiments compared particle deposition on smooth, waxy leaves with that of rough, pubescent leaves while a third set of results compared deposition on single leaves with that found on groups or assemblages of leaves.
Experimental System A 0.929-m2 steel wind tunnel -7.315 meters in length was constructed as a transport system for the deposition studies (6). Air entering the tunnel passes through an absolute particle filter (99.99% capture efficiency for particles greater than a 0.3 p diameter) and converges into the 0.929 m2 test section. Monodisperse aerosol particles, produced by a vibrating orifice aerosol generator (7) are carried into the wind tunnel by a stream of dry, filtered air that is passed through a static eliminator (Po 210 source) to remove any residual charge on the particles. The flow in the wind tunnel is then passed through a stairmand disk (8) to increase mixing and homogeneity of particles in the air stream. Consequently, the turbulent scale of the air stream is greatly increased. To reduce the large-scale turbulence, the air was first passed through a flow straightening element (grid of 2.54-cm diameter holes) and then through a series of 0.158-cm mesh and 0.045-cm diameter wire grids spaced 2.54 cm apart (9). The turbulence scale is thus reduced to nearly isotropic conditions. This is conductive to obtaining a uniform cloud profile during aerosol experiments. Also, the effects of surface characteristics in aerosol deposition can be best determined in a flow regime of this nature where deposition induced by random large-scale fluid perturbations is held to a minimum. The air stream then enters a 1.219-meter long Plexiglas test section equipped with access ports a t the top and bottom. Air flow is induced by a blower on the exit end of the wind tunnel, controlled by a damper on the blower, measured by the pressure drop across a calibrated American Society of Mechanical Engineers' standard nozzle, and exhausted through a ventilating hood. The contamination of exhaust air is prevented by two absolute particle filters placed in the exit section of the wind tunnel. Uranine was analyzed fluorometrically by washing fumigated leaves (or filter paper from the isokinetic sampling probe) with deionized water and measuring the total mass deposited with a calibrated Turner fluorometer. Similar preparations of unfumigated leaves were used as blanks in these determinations. Lead was analyzed by the Heavy Metals Analytical Laboratory at the University of Illinois, Uraban-Champaign. The technique used for the lead analysis called for the leaves to be ashed a t 450"C, dissolved in hydrochloric acid, and the resulting solution to be analyzed using an atomic adsorption unit. Measurements of particle cloud concentration were Volume 9, Number 2, February 1975
151
-
-
0.5-
1
Sunflower
/
Sunflower Single Leaf(-)
*.,-’/
..f
y = 0.004+ 0.00264x
1
r = 0.835
taken with an isokinetic sampling probe at different locations in the test section (7). The cloud concentration of 3.02 pg m-3 was uniform throughout the testing area.
Plants Leaves of sunflower (Helianthusannuus L . ) and tulip poplar (Liriodendron tulipifera) were used in the gross deposition studies reported here. These species were chosen to represent plants with large differences in leaf surface characteristics so as to bound the correlation between surface leaf roughness and deposition. Leaves were either fumigated individually or in groups (leaf assemblages). Single leaves were supported at the petiole in a trailing position, parallel to the air flow in the wind tunnel. Leaf assemblages were constructed by placing the petioles of leaves into small openings of a “stem” such that the base of the leaf stems were immersed in water. In this manner, leaf number and leaf area could be varied over a wide range. The leaf area was determined by planimetering outline tracings of individual leaves with a Lasico Polar Planimeter. By rotating the “stem” during fumigation a reasonable approximation to a tree was modeled with wind striking the assemblage from all sides. This is not unlike the situation that an isolated tree might experience over time in an urban environment. All leaf experiments were conducted at a wind speed of 268 cm/sec-l, an aerosol cloud density of 3 pg m-3 (44.9 pg min-1 for the 0.929 m2 cross-sectional area of the fumingation chamber), and a particle size of 6.77 p f 0.02 p diameter. Fumigation periods varied in length from 10-35 min. The relative humidity and the air temperature varied between 55-60% and 24-28“C, respectively. Results and Discussion The first experiment examined the similarity of gross deposition on leaves between particles of P b and those of uranine dye. While it is known that uranine particles with a diameter of 6.77 p are aerodynamically equivalent to PbC12 particles with a diameter of 3.36 p , it was not clear that the deposition of these two aerosols would be the same under identical conditions. To compare the two aerosols, sunflower leaves were used as the deposition surface. The sunflower leaves were placed horizontally within the test section, such that during fumigation they assumed a stable position in the air stream with very little movement. The results of experiments shown in Figure 1 show that the aerosol collected by leaves fumigated with uranine was essentially identical to that of leaves fumigated with PbClz particles. This confirmed that fumigation with 6.77-p diameter particles could be successfully substituted for 3.36-p-diameter PbC12 particles in at least 152
Environmental Science & Technology
00
40
eo
Ix)
160
m
240
283
Leaf Area (ern')
Deposition of uranine dye particles on single leaves and leaf assemblages of tulip poplar compared with single leaves of sunflower Figure 2.
some aerosol deposition experiments. It should be noted, however, that no conclusions on adverse physiological effects should be drawn from the use of particles with aerodynamically equivalent diameters because the ease of entrance of particles into the leaf is dependent on particle size. The second experiment was designed to test the hypothesis that differences in leaf surface characteristics play a major role in influencing aerosol deposition and to bound the effects of the leaf surface characteristics. Rough pubescent leaves of the sunflower were compared to the glaborous leaves of tulip poplar. Single leaves were positioned horizontally in the test chamber and fumigated with uranine particles. It was believed that this orientation would best discriminate between differences in surface characteristics by the following reason. The particles are transported to the depositing surface primarily by turbulent diffusion and under these flow conditions, if a particle attaches to the surface, it has a high probability of remaining (6). Thus, any greater degree of capturing ability exhibited by one species over another would be due essentially to an increased deposition area caused by greater roughness or hairiness. This roughness profile appears most prominently to the particles when the leaf is parallel to the direction of particle trajectory. The deposition rate on the pubescent leaves of sunflower was nearly 10 times that of the nonpubescent leaves of tulip poplar. As all other conditions during the fumigation of these two species were the same, it is concluded that the difference in aerosol deposition is due only to differences in leaf surface characteristics. The data appear to conform to alinear relationship as indicated by the regression coefficients. In the final experiment, the rate of deposition for a group or assemblage of leaves was studied. The results indicate that the deposition rate per unit leaf area for the ,assemblages were essentially identical to the single trailing leaves-i.e., the rate was linearly related to the total leaf area as shown in Figure 2. The increase in leaf area was achieved by making the “tree” more and more dense or decreasing the effective flow area between the leaves. As deposition by impaction is negligible for this particle size and flow rate, one might expect the rate of deposition to decrease or level off somewhat as the porosity of the
tree approached t‘hat of a solid object. This was not the case. Also, the random motion of the leaves flapping violently would cause considerably different flow conditions to prevail than the case of the single trailing leaf with little motion. Even though the fluid dynamics are different, the functional form and value of the results are the same. The demonstration of the validity and reproducibility of our results gives promise that a dose-response curve can be quantitatively determined for plants existing in an environment where the ambient aerosol cloud concentration is known. The simplicity of the results suggests that mathematical models for plant deposition via the diffusion process are possible. This would be a first step toward formulation of a quantitative model to assess the potential deleterious effects of aerosols on crop yields.
Literature Cited (1) Rolfe, G. L., J . Enu. Qual., 2, 153 (1973).
(2) Baumhardt, G. R., Welch, L. F., ibid., 1,92 (1972). (3) Lagenverff, J. V., Soil Science, 111, 129 (1971). (4) Buchauer, M. J., Enuiron. Sci. Technol., 7,131 (1973). (5) Lagerwerff, J. V., Specht, A . W., ibid.,7,583 (1970). (6) Wedding, 3. B., Stukel, J . J., Int. J . Multi-Phase Flou, in press, (1974). (7) Wedding, J. B., MS thesis, University of Illinois, Urbana, Illinois, 1972; Wedding, J. B., Stukel, J. J . , Enuiron. Sei. Technol., 8,456 (1974). (8) Green, H. L., Lane, W. R., “Particle Clouds, Dusts, Smokes and Mists,” E. and F. N. Spon, Ltd., London, 1957. (9) Baines, W. D., Peterson, E. A . , Trans. ASME, p 467 (1951). Received for reuieu; June 7, 1974. Accepted Oct 18, 1974. Work supported in part by Grant N S F GI 31605 from the R A N N program of the National Science Foundation.
Detection of Acrolein in Engine Exhaust with Microwave Cavity Spectrometer of Stark Voltage Sweep Type Mitsutoshi Tanimoto* and Hiromichi Uehara Sagami Chemical Research Center, Nishi-Ohnurna 4-4-1, Sagarnihara, Kanagawa, 229, Japan
Acrolein in automobile engine exhaust is detected using a microwave cavity spectrometer of Stark voltage sweep type. The cavity operates in the TE1,0.20 mode at a fixed microwave frequency in the X-band region. The exhaust gas is collected through a glass tube packed with phosphorus pentoxide and trapped on a cold adsorbent in a dry ice-acetone cold bath. A sample of the engine exhaust is found tcl contain about 5 ppm of acrolein. Aldehydes are produced by incomplete combustion processes of various organic substances and appear to be important chemical species in photochemical air pollution. Formaldehyde i; the richest component of the aldehydes, and a specific analytical procedure has been developed ( I ) . Much attention has been paid also to acrolein, for it produces significant eye irritation. Acrolein has been determined colorimetrically using 4-hexylresorcinol as the color-producing reagent (1, 2). The colorimetric method suffers from interference from molecules having structural features in common with acrolein, such as crotonaldehyde and butadiene (2) and formaldehyde ( 3 ) .Microwave spectroscopy, however, is appropriate t o the detection of trace contaminants in complex mixtures because of its high sensitivity and high resolution. Recently a new cavity spectrometer has been constructed and applied to the detection of formaldehyde in automobile exhaust ( 4 , 5 ) . The present paper reports the detection of acrolein in engine exhaust by means of the cavity spectrometer with a help of a preconcentration technique.
Experimental As the details of the cavity spectrometer used in the present investigation was described elsewhere ( 4 ) , only a brief explanation is given here. A rectangular cavity (10 mm in height, 20 mm in width, and 500 mm in length) operates in the ‘I’El.o,zo mode at about 8.9 GHz. It is coupled to a microwave system through an iris diaphragm and is terminated by a movable short-circuiting plunger.
A copper plate is inserted as a Stark electrode a t the center of the cavity so as to cut the microwave electric field perpendicularly. The plate is mounted on narrow grooved strips made of Teflon. The Stark electrode is fed with a variable dc and 100 kHz sine-wave voltage. The microwave absorption is detected by a lock-in amplifier. The output of the amplifier is displayed on a recorder in the form of the derivative of the absorption spectrum. A number of samples were taken from the exhaust of an automobile engine connected with a dynamometer (GoPower DA-300). The samples were collected in a glass sample holder of about 300-ml capacity through a glass tube packed with phosphorus pentoxide. T o apply the low-temperature trapping technique (6), 1 gram of the adsorbent (diatomaceous earth firebrick) was placed in the sample holder, which was evacuated before the sampling. The exhaust gas was carefully introduced into the sample holder after the moisture which initially deposited in the sampling line upstream to the dehydrating reagent was revaporized by the heat of the exhaust gas. This procedure was necessary to avoid undesirable reactions of sample components with deposited water. The sample holder was then immersed in an acetone-dry ice cold bath with recurrent agitation. After half an hour, noncondensable gases such as Nz, 0 2 , and COZ were pumped out, and the adsorbent was heated up to about 200°C. The desorbed gas was introduced into the spectrometer cavity, and the pressure was reduced to 0.20 torr. The Stark dc voltage was varied, the microwave frequency being fixed a t 8910.5 MHz. The frequency was close to the zero-field rotational 000 8902.2 MHz, which was calculated transition 101 from the rotational constants derived by Cherniak and Costain ( 7 ) .
-
Results and Discussion The spectrum of acrolein in automobile engine exhaust (4000 rpm; torque, 3.5 kg m) is shown in Figure 1.Two other signals are recognized in the spectrum. One is the strong signal on the side of the higher Stark voltage and is due to formaldehyde contained in the exhaust gas as much as 20 Volume 9, Number 2 , February 1975
153
