Effect of Selected Cation Concentration on Coagulation and Adhesion to Silica Surfaces of 6 MnO, Stephen R . Jenkins Civil and Architectural Engineering Department, University of Wyoming, Lararnie, Wyo. 82070
The effect of Ca2+, Mn2+, and N a + concentrations on the aggregation and deposition characteristics of 6-Mn02 were studied. Refiltration techniques were used to study particle aggregation, and a rotating disc apparatus was used to study deposition of the 6-MnO2 on a glass surface. Experimental results obtained on colloid stability of MnO2 and on the attachment of colloidal particles on glass surfaces have been explained with the help of quantitative models on the association (coordination) of cations with the oxide surfaces. In particular, Ca2+ and Mn2+ ions have been shown to specifically interact with Mn(1V) surface. Because of the specificity of this interaction, a stoichiometric relationship between the critical coagulation concentration of Ca2+ (or M n 2 + ) and the surface area concentration is observed under suitable conditions.
The most widely used method for removal of manganese in the water treatment field is oxidation of Mn2+ to a higher valent manganese oxide which can be removed by sedimentation and/or rapid sand filtration. The oxidation of Mn2+ will occur with air, oxygen, ozone, chlorine, or potassium permanganate under appropriate conditions of PH . Manganese oxides formed by oxidation of Mn2+, as well as those formed by reduction of M n 0 4 - , are often in colloidal suspension. Therefore, a coagulation step is necessary prior to sedimentation. Morgan and Stumm (1964), Posselt et al. (1968 a,b,c) and Morgan and Chen (1968) have investigated aggregation of higher valent manganese oxides. These authors have shown effects of Ca2+ and other cations on the stability of manganese oxide particles. They have shown that there exists a stoichiometric relationship between the amount of Ca2+ necessary to cause aggregation and the concentration of the manganese oxide. Further, large concentrations of Ca2+ did not cause restabilization. The present study was undertaken in order to help understand the relationship between the concentrations of various cations in solution and the stability of a MnO2 colloid, and the effect of the concentrations of the cations on removal of MnO2 particles by sand filters. The MnO2 colloids employed for the experiments reported here were prepared by dropwise addition of a 10-2M Mn2+ solution prepared from G. Frederick Smith Reagent grade Mn( C104)2 and standardized by the method proposed by Lingane and Karplus (1946), t o a solution containing a purified laboratory grade (Fisher Scientific Co.) NaMnO4 which had been standardized by the Fowler and Bright method (Skogg and West, 1963). Stoichiometric quantities of MnOr- and Mn2+ were added to yield a M n 0 2 concentration of 4 X lO-3M. Oxides were formed a t p H levels 6.5, 8.0, and 9.2. The p H was held constant by
addition of 1 0 - l M reagent grade NaOH with a Beckman automatic titrator. Purified N2 (C02 removed with an ascarite filter) was continuously bubbled through the solution to obviate oxidation of Mn2+ by air a t the higher pH’s. I t was necessary to add the Mn2+ very slowly, because if the concentration of Mn2+ in solution became too high, the incipiently formed MnO2 would be aggregated by the Mn2+. The oxide formed had an average oxidation grade of MnOl.98. Samples of the manganese oxides were filtered, airdried, and spindles made for X-ray analysis. The d-spacings measured from the X-ray photographs compared favorably with those reported by Bricker (1965) for 6-Mn02.
Cation Effect on Colloid Stability A refiltration procedure similar to that developed by La Mer et al. (1957) and modified by O’Melia and Stumm (1967a) was employed to evaluate cation effects on the stability of 6-MnO2 dispersions. Theoretical justification for the application of refiltration rate measurements to aggregation studies has been developed by La Mer et al. (1957). These authors have shown that the refiltration time is directly proportional to the square of the surface area of particles through which filtration occurs. With the onset of aggregation, the effective surface area decreases, resulting in a decrease in refiltration time. In the present study, aliquots of MnO2 were filtered under a constant pressure differential of 660 mm Hg through a Millipore membrane having a 47-mm diam and a specified pore size of 0.220 1. The filtrate was collected and refiltered under identical conditions. The time required to collect 10 and 30 ml of filtrate during the refiltration step was noted and the difference designated a t the refiltration time (RT). Refiltration time (RT) is dependent on the thickness of the filter cake, which is itself dependent on the concentration of colloid. Of primary interest in this study is the critical coagulation concentration (ccc) of the cation-Le., the minimum concentration of cation necessary to completely aggregate the MnO2 colloid and produce a percent nornialized refiltration time (%NRT) of zero. The refiltration times were normalized by subtracting the RT of a completely aggregated suspension from each measured RT and then comparing these normalized times to the normalized R T of a completely dispersed colloid of the same concentration. Normalized refiltration times are shown in Figure 1 for p H 8.0 and for various concentrations of MnO2 when Ca2+ is used as the destabilizing cation. Similar curves were obtained a t p H 6.5 and 9.2. By use of this data, the ccc may be plotted as a function of MnO2 concentration (Figure 2). At all p H levels, as the concentration of colloidal MnO2 increases, the ccc of Ca2+ required to obtain a %NRT equal to zero also increases (Figure 2). This relaVolume 7,Number 1, January 1973
43
l
o
o
compact part of the double layer. Such a reaction can be formulated as follows:
w
-Mn-OH
80
J
K
z ap
J
40
2
+ H+
-Mn-Oca+
(1)
where -Mn-OH represents the protonated surface sites and -Mn-Oca+ represents the surface sites occupied by a calcium ion. The equilibrium association can be expressed by a stability constant
.
!-eo
+ Ca2+
K1 = 4 x 10-6 = [-Mn-OCa+][H+]/ (2)
[-Mn-OH][CaZ+]
To account for aggregation of MnOz particles by CaZ+, three assumptions are made: First, sorbed Ca2+ reduces the net surface charge to zero-Le.,
[c dtotol-M x lo4 Figure 1. Percent normalized refiltration time ( % N R T ) vs. total concentration of calcium in solution for pH 8.0 and ionic strength ( I ) = 6 . 5 X 10-3M Ionic strength is approximately equal to the concentration of the NaN03 in the solutions
I
I
I
I
I
[-Mn-OCa+]AcG
Figure 2. Total critical coagulation concentration of Ca2+ vs. concentration of
6 !X
1- a) = [capacity] - [-Mn-Oca+
= [-Mn-OH]/(
1
pH80
[capacity] = [-Mn-0-1
+ [-Mn-OH]
I
I
I
2
3
4
I
5
M x IO3
tionship is linear with respect to the MnOz concentration for each pH. These curves show little effect of p H on the ccc of Ca2+ producing aggregation of a particular molar concentration of MnOz. I t is apparent from these curves, however, that a t very low concentrations of MnOz, the ccc of calcium needed for aggregation of the MnOz decreases with an increase in pH. This is in agreement with Morgan and Chen (1968) and Posselt et al. (1968a,b,c), who used MnOz concentrations below 0.5 X 10-’3M. At higher MnOz concentrations the ccc of Ca2+ appears to increase as the pH increases. Most important, in the pH range studied, there exists a stoichiometric relationship between the ccc of Ca2+ and the concentration on the MnOz colloid. This stoichiometric relation between the critical coagulation concentrations and the surface area concentration (in the case of Ca2+ as the coagulant) confirms a specificity of interaction in the coagulation reaction as discussed by Stumm and 0 ’Melia ( 1968). Similar experiments were made using N a + as the destabilizing cation. No stoichiometric proportion was observed between the ccc of Na+ and the surface concentration of MnOz, indicating coagulation occurs by compaction of the double layer without significant specific chemical interaction.
Quantitative Interpretation of Adsorption of Ca2+ to Account for Destabilization Stumm et al. (1970) have shown that the Ca2+ reacts with the manganese oxide surface by being sorbed into the 44
Environmental Science & Technology
(4)
+ [-Mn-Oca+]
[MnO&
]
Here cy is the fraction of deprotonated acid groups and [capacity] is the total number of surface sites available. For the case when H + dissociates with the surface groups or when OH-, N a + , or Ca2+ associate of the -Mn-OH, with these groups,
MnO2 I = 6.5 x 1 0 - 3 ~ (NaN03); pH as shown
(3)
-]AGG
where the subscript AGG refers to concentration of surface sites for complete destabilization of the MnOz. Second, the relationship between the protonated (-MnOH) and unprotonated (-Mn-0-) sites remains the same even though Ca2+ has occupied some of the surface sites-Le., [-Mn-O-]/a
n
= [-Mn-0
(5)
This capacity can be experimentally determined by alkalimetric titration curves as explained by Stumm et al. (1970). The method consists essentially of measuring the proton or hydroxide ion consumption by the solid phase, by comparing the titration curve of a suspension of the inorganic oxide with that of the medium alone. If the surface area of the oxide is known, the capacity can be calculated. Third, the capacity of the MnOz in the presence of Ca2+ is the same as in its absence. Combining Equations 2-4, the following equation results: [ C a Z i ] ~ ~=c a [ H + l / ( 1 - ~ ) K I
(6)
In this equation the concentration of Ca2+ in solution for complete MnOz destabilization ( [CaZf]AGG)is expressed as a function of [ H + ] only, since Kl is constant and cy and 1 - cy are functions of p H only. Thus the ccc of calcium ion in solution is independent of the concentration of the MnOz. When Equation 3 is substituted into Equation 4, [ - M n - o c a + ] ~ ~= ~ a[capacity]/(l - a )
(7)
The total amount of calcium required for complete aggregation of an MnOz colloid (ccc) is then equal to the [-Mn-OCa+]AGG + [Ca2+]AGG: [CBTIAGG= [ca2+]~GC + [-M~-OC~+]AGG
(8)
The above calculation has been made for various MnOz concentrations a t pH values of 8.0 and 9.2. The results [-Mn-Oca+]/[-Mn-0-1
= 1or
[-Mn-Oca+
] = [MnO-]
are compared with the experimentally determined ccc in
Figure 3. I t is apparent from Figure 3 that under the conditions stipulated, the ccc of calcium as calculated is always much larger than that determined experimentally. However, if it is assumed that some Na+ also adsorbs on the surface of the MnOz, and that some Mn2+ absorbs during the oxide formation, the critical coagulation concentration of CaZ+ will be reduced. Under these assumptions, at incipient aggregation the number of negative surface sites is not equal to number of -Mn-Oca+ sites. Instead, the concentration of -Mn-Oca+ sites will be sites: some fraction (X) of the concentration of -Mn-0[-Mn-OCaf]
= X[-Mn-0-1
(9)
It then follows from Equations 6-8 that [-M~-OC~+],AGG = Xa[capacity]/(l + ax)
(10)
[ C ~ ~ + ] A=GXGa [ H L ] / ( l - cu)K~
(11)
and
Using Equations 10 and 11, [CaT]AGG may be calculated for a given p H and [MnOz] and any assumed X. In Figure 3, the calculated values of [ C ~ Ta t] X~ equal ~ ~ to 0.35 are compared with the experimentally determined [CaTIAGG's. The experimental evidence thus confirms that calcium causes aggregation of MnOz colloids by penetrating into the compact layer a t the surface, reducing the charge (Stumm et al., 1970). Since the adsorption occurs by specific interaction, one might expect eventual charge reversal and restabilization. However, restabilization did not occur. Even though the sign of the surface charge might be changed by specific adsorption of added CaZ+, it is possible that restabilization was prevented by the accompanying double-layer compression resulting from the high NO3 - concentrations.
MnOz Particle Attachment t o Glass Surfaces O'Melia and Stumm (1967) have considered the mechanisms of particle removal in sand filtration as two separate and distinct steps: The transport of the suspended particles to the immediate vicinity of the solid-liquid interface presented by the filter, and the attachment of the particles to this surface. Particle transport, here being a convective or convective-diffusive process, is influenced by those parameters affecting fluid flow. Particle attachment, however, must be interpreted in part as a colloidchemical process and is influenced by both chemical and physical parameters (ionic strength, surface properties of the solid, temperature, etc.). Of most concern in the present study was the effect of solution variables on the attachment of the MnOz particle to a sand surface. Consequently, some method to assure reproducibility in the transport step was desired. To simulate uniformly accessible sand surfaces and to produce reproducible modes of particle transport, a rotating disc apparatus similar in design to that developed by Marshall and Kitchener (1966) was employed. The rotating disc containing a replaceable standard surface-e.g., glasswas submerged into a dispersion of MnOz colloids; for different solution conditions and different rotating speeds the quantity of MnOz that became attached to the surface within a given time was determined analytically. The plastic disc was turned from polymethacrylate plastic to fit the specially designed collet which connected directly to the drive shaft of the motor. All parts were carefully constructed and aligned to avoid vibration that would disturb the desired laminar flow of the solution across the face of the disc. The surfaces for deposition consisted of 2 2 ~ m mdiam
Figure 3. Comparison of ccc of calcium as determined experimentally with that calculated if a complex formation reaction is assumed I t is assumed that a complete aggregation of MnOz the -MnOcasites equal some fraction of Mn-0at I = 6.5 X 1 0 - 3 M (NaN03)
Corning circular microscope cover glasses attached to the disc surface with paraffin wax. These cover glasses were wiped with a lint-free tissue prior to being mounted on the disc. The disc was then placed in a distilled water solution and rotated a t a constant speed for 30 min. This washing of the glass surface allowed a degree of weathering to occur. Exposed glass surfaces leach borates and approach amorphous silica in nature (Hair, 1967). Because the surface of quartz is frequently amorphous in nature (Van Lier et al., 1960), it is felt that the glass disc surfaces simulate sand surfaces used in rapid sand filtration processes. Thus, glass has surface characteristics similar to sand and was used instead of quartz crystals because of the ease with which the cover glasses could be obtained and mounted to the disc. The disc was rotated a t a constant speed (200 rpm) in each suspension for 10 min and then in distilled water for 30 sec. This latter washing step was to prevent any suspended particles from remaining attached to the disc in a water drop clinging to the cover glass. Aliquots of cation (CaZ+, Mn2+, or N a f ) and predetermined amounts of NaOH for constant pH were added just before disc rotation began. After the cover glass was washed with distilled water, it was carefully removed and placed in an o-tolidine solution. After 10 min the absorbance of the o-tolidine solution was measured using the procedure of Morgan and Stumm (1965), and the amount of MnOz adhering to the cover glass was determined. This method provided very reproducible results even though application of the fluid dynamic flow characteristics to particle transport was not in accord with the theory proposed by Levich (1962) for transport of particles from the bulk solution to the deposition surface. Although the rotating disc provides a uniformly accessible surface for deposition, where the flux is independent of position, the tangential velocity of the fluid increases relative to the disc with increasing distance normal to the disc surface. Hydrodynamic forces caused by this velocity gradient act to shear the larger particles from the disc surface. After aggregation of a colloid has occurred, if one large particle is sheared from the glass surface, a sizable error (15%) in the measured amount of MnO deposited on the surface can be made. Data typical of all experiments made with MnZ+ and Ca2+ are presented in Figure 4. The amount of MnOz deposited on the cover glass remains incipiently small with Volume 7 , Number 1 , January 1973
45
6
6
I
1
I /
5 P
P 5
4
I
f
3
E
,o
-
2
I c
a u
I
Y
O
Figure 4. MnOz deposited/standard glass surface Vs. Ca concentration for various concentrations of M n 0 2 (shown)
Vertical arrows show the concentrations of cation necessary for visual aggregation. Simiiar data were obtained at pH 6.5 and 9.2
increasing concentrations of destabilizating cation, until a t some critical cation concentration the amount of deposited MnO2 increases with small subsequent increases of cation concentration. The onset of this increase in attachment is accompanied by the onset of visual particle agglomeration. The location of the peaks also coincides with particle agglomeration as shown by refiltration techniques explained above. Further increase of cation concentration causes the amount of MnOz adhering to the cover glass to decrease rapidly. If either calcium or manganese is added to a previous colloidal 6-MnO2, the amount of cation required to affect maximum adhesion of MnO2 is related to the concentration of MnOz (Figure 4). As shown in Figure 5 a t the pH levels indicated, there is an exact stoichiometry between the concentration of cation needed for maximum deposition and number concentration of MnOz particles. It may be inferred that the mechanisms responsible for particle attachment are similar to factors responsible for agglomeration. Ca2+ and Mn2+ associate with the MnOz and with the glass surface to reduce the charge of the diffuse parts of their double layers (Stumm et al., 1970) facilitating particle contact and an increase in particle deposition. After complete aggregation of the MnOz particles has occurred, the particles become so large that hydrodynamic forces tangent to the face of the disc shear the deposited particles from the disc surface. The amount of deposited MnOz thus decreases rapidly. If sodium is employed as the destabilizing cation instead of calcium or manganese, the concentration of Na+ necessary to produce maximum deposition of the MnOz is independent of the number concentration of MnOz particles. The sodium necessary to produce this maximum was always about 2.7 X 10-2M NaN03 regardless of the MnOz concentration. Thus, Na+ enhances MnOz Particle attachment to sand surfaces (as well as particle agglomeration) by double-layer compaction. Conclusions and S u m m a r y
The mechanisms of MnOz particle destabilization and MnOz particle adherence to glass (sand) surfaces appear to be similar in the presence of either Mnzf or Ca2+. In the comparison of Figures 3 and 5, the concentration of calcium required to cause maximum adherence of MnO2 on glass surfaces is the same as the concentration of calcium needed for complete aggregation of the MnO2, if the 46
Environmental Science & Technology
O
l [MnOJ
2
3
- M I IO3
4
Figure 5. Stoichiometric relationship between concentration of cation (Ca2+/Mn2+)necessary for maximum deposition of MnOn O n glass surface and concentration of Mn02 if / = 6.5 X
10-3~
MnOz concentration, pH, and ionic strength are the same. Both processes occur a t concentrations of Ca2+ which are well below those expected if double-layer compaction was necessary for agglomeration and attachment (Stumm and O’Melia, 1968). The concentrations of these two cations necessary to produce maximum deposition of MnOz particles and particle destabilization is proportional to the number concentration of particles (stoichiometry). The complex formation reaction postulated by Stumm et al., (1970) can be used to explain coagulation of MnOz particles with Caz+. Similar reasoning can explain MnOz particle attachment to glass. The cations Ca2+ and Mn2+ chemically associate with the MnOz surface, reducing the charge until van der Waals attractive forces overcome electrostatic repulsion between negative particles and negative glass surfaces, and attachment occurs. It is also possible that the cation attaches to the hydrous Si02 surface as well as to the hydrous MnOz surface, and may act as a bridge between negative particle and negative surface. Sodium concentrations a t maximum deposition were observed to be independent of the concentration of particles over the range investigated. Hence, compression of the diffuse part of the double layer can explain the destabilizing and deposition effects of N a + . Acknowledgment
The author thanks Werner Stumm for his guidance in this research.
cited Bricker, O., Arner. Mineral., 50,1296 (1985). Hair, M. L., “Infrared Spectroscopy in Surface Chemistry,” MarN.Y. 1967, P 288. cel Dekker, New La Mer, V. K., Smellie, R. W., Jr., P. Lee, J. Colloid Sci., 12, 566 (1957). Levich, V. G., “Physiochemical Hydrodynamics,” Prentice-Hall, 1962. Lingane, J . J., Karplus, R., Ind. Eng. Chern., 18,191 (1946). Marshall, J . K., Kitchener, J . A., J. Colloid Interface Sci., 22, 342 (1966). Morgan, J. J . , Chen, C., Div. of Water, Air and Waste Chemistry, 155th meeting, ACS, San Francisco, Calif., March-April 1968. Morgan, J. J., Stumm, W., J. Colloid Sci., 19,347 (1964). Morgan, J . J . , Stumm, W., J. Arner. Water Works Ass., 57, 107 (1965). O’Melia, C. R., Stumm, W., J. Colloid Interface Sci., 23, 437 (1967a). O’Melia, C. R., Stumm, W., J. Arner. Water Works Ass., 59, 1393 (1967b).
Posselt, H . S., Anderson, F. J., Weber, W. J., Jr., Enuiron. Sci. Technol., 2, 1087 (1968a). Posselt, H. S., Reidies, A. H., Weber, W. J., Jr., J . Amer. Water Works Ass., 60,48 (1968b). Posselt, H . S., Reidies, A. H., Weber W. J., Jr., ibid., 1968c, p 1366. Skogg, D. A,, West, D. M., “Fundamental of Analytical Chemistry,”Holt, Rinehart and Winston, N.Y., 437 (1963). Stumm, W., Huang, C. P., Jenkins, S. R., Croat. Chem. Acta, 42, 223 (1970)
Stumm, W., O’Melia, C . R., J . Amer. Water Works Ass., 60, 515 (1968). Van Lier, J. A,, deBruyn, P. L., Overbeck, J. T. G., J . Phys. Chem., 64,1675 (1960).
Received for reuieu, January 19, 1972. Accepted Nouember 1, 1972. This work u’as supported by the C‘S. Public Health Service Research Grant W P 00098. Presented a t the Division of Water, Air, and Waste Chemistry, I61st Meeting, A C S , Los Angeles, Calif March 1971.
‘
Persistence and Reactions of 4 ~ - ~ a c oicdAcid y ~ in Soils Edwin A, Woolson’ and Philip C. Kearney Agricultural Environment Quality Institute, Agricultural Research Center, Agricultural Research Service, U.S. Department of Agriculture, Beltsville. Md. 20705
Carbon-14-labeled cacodylic acid (hydroxydimethylarsine oxide) was prepared by reacting l4C-methyl iodide with methyl dichloroarsine. Concentrations of 1, 10, and 100 ppm of cacodylic acid were established in three soils of varying iron and aluminum content. At 2, 4, 8, 16, 24, and 32 weeks, soils were analyzed for 14C and total arsenic in the water-soluble (ws), calcium (Ca), iron (Fe), and aluminum (Al) fractions. Initially, cacodylic acid was distributed in the following fractions: ws >> A1 > Fe > Ca. After 32 weeks, the distribution was ws > A1 > Fe > Ca. In contrast, inorganic arsenate ( 5 + ) was largely present in the Fe and A1 fractions. Cacodylic acid persistence was a function of soil type and after 32 weeks the following amounts of 14C were recovered in each soil type by combustion: Christiana (23%), Hagerstown (%TO), Lakeland (62%). A decrease in both total 14C and total arsenic occurred in all soils with time. A pungent garlic odor was detected in soils receiving 100 ppm, suggesting the production of a volatile alkyl arsine. The loss of arsenic suggests that one route of cacodylic acid loss from aerobic and anaerobic soils is by alkyl arsine volatility. Degradation under aerobic conditions also occurred by cleavage of the C-As bond, presumably yielding COz and A s O ~ ~ - . This degradation is presumably due to microbiological action.
Cacodylic acid (CA), (hydroxydimethylarsine oxide) is a nonselective, postemergent, foliar contact herbicide. The other important organic arsenical herbicides are disodium methanearsonate (DSMA) and monosodium methanearsonate (MSMA). Several studies have been conducted with MSMA, including persistence and metabolism in soils (Dickens and Hiltbold, 1967; Von Endt et al., 1968), phytotoxicity to plants (Schweizer, 1967; Sachs and Michael, 1971), and metabolism in plants (Duble et al., 1968; Sckerl and Frans, 1969). Metabolism of CA in plants was examined by Sachs and Michael (1971). Inorganic arsenate (Jacobs et al., 1970b; Woolson et al., 1971a,b) has been characterized by different extractants and designated as water-soluble-, Fe-, Al-, and Ca-bound fractions. Arsenic in the fractions may not come entirely from the designated forms. Johnson and Hiltbold (1969) T o whom correspondence should be addressed.
extracted soils with nine successive extracting solutions to characterize arsenic residues from DSMA. Comparatively little is known about the persistence and metabolism of CA in soils. The object of this study was to examine the chemical distribution of cacodylic acid into water-soluble (ws), iron (Fe), aluminum (Al), and calcium (Ca) fractions in three soils, under aerobic and anaerobic conditions and to determine the persistence of CA in these soils. The mechanisms by which CA was transformed in soils were examined.
Methods and Materials 14C-Cacodylic Acid Synthesis a n d Purification. Methyldichloroarsine (CH3AsC12) was prepared by reacting an HCl solution of methylarsonic acid with KI and SOz. The purified methyldichloroarsine (bp = 132°C/760 mm) was added to 100 mg NaOH dissolved in several drops of water and contained in a 16 cm X 5 cm i.d. thick-walled tube sealed a t one end and restricted at the other. 14C-methyl iodide (0.25 mCi) was dissolved in 0.5 ml cold Et20 and transferred to the reaction vessel after 0.15 ml of CH3AsC12 was added. The tube was sealed and heated at 70-75°C for 24 hr. The CA was purified by tlc on cellulose plates developed in MeOH:NH40H (8:2). The CA had a specific activity of 2.7 mCi/mM. Soil Studies. Three soils (50 grams), Lakeland loamy sand, Hagerstown silty clay loam, and Christiana clay loam, were treated with 0.2 FCi 14C-CA and sufficient unlabeled CA to achieve final concentrations of 1, 10, and 100 ppm. The CA was added in MeOH and the soils thoroughly stirred after solvent evaporation. The soils were contained in 50-ml covered beakers. The soils in three replications were brought to 75% of field capacity and incubated at 25°C for 32 weeks. Grab samples (0.5 gram dry wt) were taken periodically and analyzed for the chemical distribution of CA into ws, Fe, Al, and Ca fractions (Petersen and Corey, 1966; Woolson et al., 1971a) by I4C-liquid scintillation analysis. The extracts from the 100-ppm CA treatment were analyzed for arsenic (As). The extracting solutions are described in Table 11. One-gram samples were taken periodically for total As analysis (Woolson et al., 1971b) and 0.1-gram samples were combusted for total 14C content. W02 was trapped in 2-methoxyethanolmonoethanol amine and diluted with PPO and POPOP dissolved in toluene. Volume 7, Number 1, January 1973
47