Kane, J. C., La Mer, V. K., Linford, H . B., J . Ain. Cliein. Soc. 86, 3450 (1964a). Kane, J. C., La Mer, V. K., Linford, H. B.. J . Phys. Cliein. 68,2273 (1964b). La Mer, V. K., Discussions Furuduy Soc. 1966, No. 42, 248. Matijevik, E., Janauer, G. E., J. Colloid Interface Sci. 21, 197 (1966. Phoenix Instrument Co., Philadelphia, Pa., technical bulletin, 1968. Posselt, H. S., Reidies, A. H.. Weber, W. J., J . Ain. Wuter Works Assoc. 60, 48 (1968). Pressman, M., J . Am. Water Worlcs Assoc. 59, 169 (1 967). Roberts, R . B., Abelson, P. H., Cowie, D. B., Bolton, E. T., Britten, R. J., “Studies of the Biosynthesis of E . coli,” Carnegie Institution of Washington, Pub. 607 (1963). Rock, R . M.. Burbank. N. C.. J . Am. Wuter Works Assoc. 58. 676 (1966): Rubin, A. J., Hanna, G. P., J. ENVIRON. Scr. TECHNOL. 2, 358 (1968). Scott, T. A,, Melvin, E. H., A n d . Clieni. 25, 1656 (1953)
Smellie, R. H., La Mer. V. K., J . ColloidSci. 13, 589 (1958). Stumm, W., O’Melia, C . R., J . Ani. Water Works Assoc. 60, 514 (1968). Tenney, M . W., Echelberger, W. F., Schuessler, R . G . , Bretthauer, R. K., Division of Water, Air and Waste Chemistry, 155th Meeting, ACS, San Francisco, Calif., April 5 , 1968. Tennev. M. M., Stumm, W., J . Water Pollution Control Fed: 37, 1370 (1 965). Thimann, K. V . , “The Life of Bacteria,” 2nd ed., Macmillan, New York, 1963. Receiced,fi)r reciew FebrLiurjs 5 , 1968. Accepted Noceinber 29, 1968. Dicision of‘ Colloid und Surfiice Clzeinistry, 156th Meeting, ACS, Atlantic City, N . J . , September 1968. Work S L I P ported in part 1 3 ~fiinds ~ procided b~ the United States Depurtment of‘the Interior us uuthorized under the Water Resources Reseurcli Act of‘ 1964, Public Law 88-379, and administered by the Institute of’ Wuter Resources ut the Unicersity qf ‘Connecticut.
Interaction of Airborne Particles with Gases Benjamin M. Smith’ and Jack Wagman National Air Pollution Control Administration, Cincinnati, Ohio 45227
Birney R. Fish Oak Ridge National Laboratory, Oak Ridge, Tenn. 37830
A promising method for making sorption measurements in aerocolloidal systems is described. Initial studies are being conducted on the interaction of sulfur dioxide with metal oxide aerosols. The latter are generated by exploding metal wires in air, resulting in particles with diameters from about 0.01 to 0.1 micron. These are mixed with S3j-labeled SO2 at controlled conditions in an 820-liter cylindrical chamber Samples for analysis are withdrawn through diffusion tubes lined with leadperoxide-coated lead foil and followed by membrane filters. Diffusion tube analysis of several S02-aerosol mixtures indicated that this is a possible method of distinguishing between chemisorption and physical adsorption. In measurements that included an adsorption isotherm for SO2 on iron oxide aerosol, preferential chemisorption on iron oxide and aluminum oxide was observed at low SO2concentrations and was followed by multilayered physical adsorption at higher concentrations.
E
xperimental evidence suggests the possible influence of particulate matter upon the toxicity of irritant gases in respiratory systems. The first such evidence was presented by Dautrebande (1939), who suggested that adsorption of a gas on inert aerosol particles increased the amount of gas reaching the lungs. Dautrebande, Shaver, et a/. (1951) observed a marked increase in the irritation produced in human subjects exposed to sulfur dioxide (SO?), formaldehyde, and other gaseous pol-
‘Present address, Department of Physics, Gainesville College, Oakwood, Ga. 30566 558 Environmental Science & Technolog?
lutants when exposures occurred in the presence of various aerosols. Goetz (1958) has proposed a physicochemical explanation of the physiologic effects of combinations of gases and aerosols. Very briefly, the theory recognizes the degree of sorption of the gas by the aerosol particle, the rate of desorption of the sorbed gas from the particle, the degree of chemical reaction which may occur, and the toxicity of the new chemical compound if such reaction occurs. This theory is sufficiently general to be consistent with the findings of Dautrebande (1939), Amdur (1961), LaBelle, Long, et al. (1955), and others, and can explain both synergism and antagonism between gas and particle. A special type of synergism may occur as a result of differences in the aerodynamic behavior of gases and particles. For example, it can be demonstrated in principle that a gas such as SO, can penetrate more deeply into the lungs when it is adsorbed on particles of respirable size. This report describes a study of gas-particle interactions that is being carried out as a joint effort of the National Air Pollution Control Administration (NAPCA) and the Oak Ridge National Laboratory (ORNL). The importance of this study stems not only from its bearing on the question of synergism in the physiological effects of pollutants, but also from the significance of such interactions in photochemical and other chemical transformations in the atmosphere. In the type of experiments usually carried out to study adsorption at solid-gas interfaces, the solid phase is in the form of bulk powder, film, o r filament, and great care is taken to clean the surfaces by outgassing or heating before they are exposed to the gas phase. Such experiments bear little relation to real conditions in the atmosphere, where the solid phase is in aerosol form and in contact with a complex mixture of gases. Consequently, the emphasis in this work is on reactions of dilute gases with particles in the dispersed or aerosol state. The experimental procedures include generation of submicron-
sized aerosols by an exploding-wire technique and use of radio-labeled gases to achieve sufficient sensitivity in sorption measurements. The first systems under study consisted o f sulfur-35-labeled SO2 in combination with airborne particles of iron oxide (FesOa), aluminum oxide (Also3), lead oxides (PbO, also P b C 0 3 and Pb(OH)J, and platinum. Techniques and Instrumentation The exploding-wire technique developed at ORNL (Karioris and Fish, 1962) was selected for this study because of its convenience for generating reproducible solid aerosols in the particle size range from about 0.01 to 0.1 micron in diameter and in the concentrations required for these experiments. The generator consists of a 1 4 - ~ f . 20-kv. , capacitor and a highvoltage trigger unit. A 20-kv. variable power supply charges the capacitor, and the trigger is operated by 5-kv. and 300-volt power supplies. Impulse switches are used to activate a series of relays that control the sequence of events leading to explosion of a wire sample by discharge of the capacitor. Photographic and photometric observations o f the cooling rates o f aerosol clouds resulting from wire explosions show that cooling rate is a function of the composition and size of the wire and is about lo6to lo7 "K. per second. The extremely high quench rate results in retention of the spherical shape of the liquid droplets that cool to room temperature too quickly to assume the equilibrium crystal form. Thus, the aerosol particles are hard, smooth spheres of known composition and surface area. Crystallography and permeability studies have confirmed this smooth-spherical nature of the particles. Thus, total particle surface area can be estimated from electron microscope data. A gas handling system (Figures 1 and 2) was designed for the study to permit volume calibration and proportioning of SO, for sampling and for sorption studies. The system employs the vacuum freeze-out principle and allows combination of stable and radioactive SO1 in known proportions to facilitate SO1
Figur,e 2. Gas-handlingsystem for :$0 in safety cabinet
..-_ _.._ -.....
...- -,.....-.
.....
T h - "1st c~ om i n r i l l r l p o 51 v n m i sampling t, + r s r p - +~,-hninrlpQ llll_ calibration flask equipped with a mercury-column vacuum gage. Used in connection with this flask are volume-portioning tubes of several calibrated sizes to permit sampling of a specified volume of SO,. The vacuum during operation is approximately 0.1 micron of Hg pressure, as determined by ion and thermocouple gages. Ampules containing S3%beled SO, l__l
Figure 1. Schematic aiagrarn of gas-nanaang system 101 qusnfitatwe mixing ana volume poroomng 01 raaio lauelea and unlabeledSOs Volume 3, Number 6, June 1969 559
0
on char
h.r
lrllr" L
U
Up'cL'cLL
~LuU"U-6Larr
,u,rrrr
'all"
L L L L L L C I I C U L U 1115.
system. At high vacuum the ampule seal breaks, releasing the SO, to the system. The gas is then transferred to the calibration flask (- 10 minutes) by freeze-out in a well immersed in liquid nitrogen. The stopcock is closed, the gas allowed to reach temperature equilibrium, and the volume measured. Stable SO, is handled in essentially the same manner; labeled and unlabeled SO$ are eventually mixed and portioned. Portions for use in experiments are transferred to sample "U" tubes and the excess to a storage flask. Figures 3 and 4 illustrate the gas-aerosol interaction chamher. The chamber is cylindrical and contains a volume of 820 liters. The aerosol from a n exploded metal wire is transferred either directly to the interaction chamber or by way of a holding tank for conditioning and dilution to produce the desired particle concentration. Before the aerosol is introduced, the
a known trace ami nt of sulfur-35 is allowed to he chamber. The ren ning negative,pressure effects e experimental aero I. (drawn from the interaction diffusion tubes followed by
Figure 4. Interaction chamber and other equipment used m this study 560 Environmental Science & Technology
. . . .. ..
100
--1
-------
!
MlX,NG
\-,EO
MIN A P T E R
SO* WITHOUT AEROSOL A N D THEORETICAL
\\
\
'\
'\
-
4 -
---
t
t
-
I N I T I A L 50, CONCENTRATION 1 4 5 p I N I T I A L N U C L E I COUhiT 2 2 r 1 O 8 / P
01
0
1
1
1
2
4
6
B
1
'
1
,
IO
12
44
(6
01'
18
1
'
0
20
2
4
1
6
I 12
14
16
18
23
Figure 6. Decline of free SOzconcentration on passage through diffusion tubes after various periods of mixing with platinum aerosol
Figure 5. Decline of free SOi concentration during transit through diffusion tubes in presence and absence of Fe304 aerosol
sulfur-35 by means of an internal gas-flow proportional counter. Particle samples are assayed for sorbed SO? in a similar manner.
Figure 5 shows the results of measurements of SOndeposited on diffusion tubes in the absence and presence of FesOaaerosol. Samples were withdrawn for 5 minutes at 300 cm.3 per minute. The theoretical decay curve for SO2 diffusion was plotted from a n equation for deposition by diffusion in a cylindrical channel, derived from the equation by Gormley and Kennedy (1949), as follows:
I 10
DISTANCE ALONG T U B E I c m l
DISTANCE ALONG TUBE rem)
Results and Discussion
' 8
#/l
9 8 P 96
s 4
i
$
a
:\
INITIAL S o p COYCEYTRAT'ON* 6 2 y i / INITIAL NUCLEI COUNT : < 8 r108 /a 7 3 '10 R H AT 2 3 ' C
;1
c Y
-
i
'3
9 4 .I-
'll
0
till
92
. i
.~ 15
I FREE
w
SO2 V A P O R
J
0
z 3
901':
Y
6:
'TNUCLEl
0 z
COUNT
c
-a il z Y
3 0 z
Figure 7. Sorption of SO2 on Fe30i particles in air
where: N ,
=
No D Q
= = =
2 =
number of particles o r molecules deposited per unit length number entering channel diffusion coefficient volumetric flow rate of carrier gas distance from channel entrance
The measurements showed that SO2 in the absence of aerosol diffused to the tube walls at the predicted rate. However, in the presence of FesOa aerosol particles, the free SO2 declined at a slightly higher rate (samples removed at intervals up to 180 minutes after mixing). This same general trend was observed also in studies with A h 0 3 particles. The higher disappearance rate of SO? may be due to continued chemisorption on the aerosol particles during flow through the tube, in addition to diffusion to the tube walls.
,
k/
'4
a c w
.
-
I N I T I A L S O p CONCENTRATION i 1 0 2 p 1/1 I N I T I A L NUCLEI COUNT: I 8 x 1 0 8 / 1 4 6 % R H 4 T 23-C
J
I
--'5
0 L
40
eo
120 160 TIME I M I N I
zoo
zio
zoo
Figure 8. Sorption of SO2 on A1203 particles in air Volume 3, Number 6, June 1969 561
‘3
1
I
/C
//
/
Table I. Number of Layers of SOzAdsorbed on F e 3 0 4Aerosol as Function of SOI Concentration SO2 Concn., Coverage, pl./liter SO2 Monolayers 1 13 0 38 6 2 3 20 5 10 9 5 33 35.0 66 75 0
/
24” C. and 7 0 x R H . The observed relation between specific adsorption ( A ) and SO:, concentration (C) was: A
5 4
3 100
‘0 C , SO2 CONCENTSA-ICh
,502
sorption isotherm for FeaOI aerosol
A reverse effect was found in studies with platinum aerosols, as shown in Figure 6. Diffusion of SOz to the walls of the sampling tubes appeared to be delayed, even very shortly after the gas was mixed with the aerosol particles. The apparent delay increased with time after mixing. The bulge in these free SO, decay curves may be explained on the basis of the desorption of physically adsorbed S3j, either in the form of SO, or SOj, during transit of the aerosol through the diffusion tube. Data on the sorption of SOr by FeaOl particles are presented in Figure 7. At an initial SO, concentration of 6.2 pl. per liter, sorption by 1.8 X lo6particles per ~ m(initial . ~ count) reached about 3 %. The drop in aerosol concentration during the experiment is also shown. This occurred principally as a result of agglomeration and, to a lesser extent, by settling. Sorption on AI2O3particles was much greater, as shown in Figure 8. Peak adsorption was about 50%, with an initial SO? concentration of about 1 111. per liter. Studies with oxidized lead aerosols resulted in the complete removal of SOnfrom the free gaseous state almost immediately (about 5 minutes) after mixing. This was expected, because SO, reacts readily with lead oxides. A sorption isotherm for SOn on Fe3O4particles in the disperse phase was measured, as shown in Figure 9. This log-log plot shows the increase in specific adsorption with increasing SO, concentration over the range from about 1 to 66 p.p.m. at
562 Environmental Science & Technology
1-07 x 10-4
~ 1 . 3 6
These data, restated in terms of the extent of surface coverage in Table I, show that single monolayer coverage through chemisorption is completed at about 2 p.p.m. of SOt; coverage increases to a surprising equivalent of 75 monolayers at 66 p.p.m. of SO, through physical adsorption.
’CL IN A G
I p P/P1
Figure 9.
=
Conclusions
The procedure described herein has proved useful in measurements of sorption in aerocolloidal systems. Of particular interest is the possibility of distinguishing between physical adsorption and chemisorption by diffusion tube analysis of gas-aerosol mixtures. In preliminary measurements of Sa6labeled SO, with iron oxide and aluminum oxide aerosols at ambient conditions, preferential chemisorption was observed at low SO, concentrations and was followed by multilayered physical adsorption at higher concentrations. Literature Cited Amdur, M. O., “Inhaled Particles and Vapours,” C. N. Davies. Ed., pp. 281-94, Pergamon Press, New York, 1961. Dautrebande. L., “Bases Experimentales de la Protection Contre les Gaz de Combat,” J. Duculot, Gembloux, Belgium, 1939. Dautrebande, L., Shaver, J., Capps, R.. Arch. Internat. Pharmacodynamie 85, 17-48 (1951). Goetz, A,, “An Interpretation of the Synergistic Effect Based Upon Specific Surface Action of Airborne Aerosols,” Final Report U S . Public Health Service Research Contract SAph69557, Part B, 1958. Gormlev. P. G.. Kennedv. M.. Proc. Rov. Irish Acad., Sec. A 52, 1i3-9 (1949). Karioris, F. G., Fish, B. R., J. ColloidSci. 17, 155-61 (1962). LaBelle. C. W.. Long, J. E.. Christofano. E. E., Arch. Ind. Health 11, 297-304-(1955). _
I
I
Receiced for reciew October 21, 1968. Accepted January 16, 1969. Dicision of Colloid and Surface Chemistry, 156th Meeting, ACS, Atlantic City, N.J., Sept. 1968. The work for this manuscript was carried out at Oak Ridge National Laborator), which is operated for the US. Atomic Energy Cornmission bj, Union Carbide Corp., Nuclear Dicision.
