Vo-Dinh,T., Anal. Chem., 50,396 (1978). Lloyd, J.B.F., Nature, 231,64 (1971). Lloyd, J.B.F., J . Forensic Sci. SOC.,11,83, 153, 235 (1971). John, P., Souter, I., Anal. Chern., 48,520 (1976). Schulman. E. M.. Walling. C.. Science. 178.43 (1972). Wellons, S . L., Paynter, R.’A.; Winefordner;J. D., Spectrochim. Acta, 3,2133 (1974). ( 8 ) Vo-Dinh, T., LueYen, E., Winefordner, J. D., Anal. Chem., 48,
(2) (3) (4) (5) (6) (7)
1126 (1977). (11) Schmeltz, I., Hoffman, D., Wynder, E. L., “Trace Substances in Environmental Health. VIII”, D. D. Hemphill, Ed., p 281, Univ. of Missouri, Columbia, Mo., 1974. (12) Ho, C. H., Clark, B. R., Coal Technology Program Annual Interim Report for Fiscal Year Ending June 30, 7976, Oak Ridge National Laboratory Report ORNL-4208, p 95.
1186 (1976). (9) Vo-Dinh, T., LueYen, E., Winefordner, J. D., Talanta, 24, 146 (1977). (10) Vo-Dinh,T., Walden, G., Winefordner,J. D., Anal. Chem., 49,
Receiued for reuiew September 23, 1977. Accepted June 16, 1978. Research sponsored by the Energy Research and Development Administration under contract with Union Carbide Corp.
Chemical Modeling of Trace Metals in Fresh Waters: Role of Complexation and Adsorption Jasenka Vuceta’
and James J. Morgan
Department of Environmental Engineering Science, California Institute of Technology, Pasadena, Calif. 9 1125
Chemical interactions of trace metals with ligands and surfaces and competition among metals of natural waters are examined by chemical modeling. Investigations include the response of a body of oxic fresh water (pc = 12, pco2 = atm) containing major cations (Ca, Mg, K, Na), nine trace metals (Cu, Pb, Cd, Co, Ni, Zn, Hg, Mn, Fe), eight inorganic ligands (COB,SOa, C1, F, Br, “3, PO4, OH), and an adsorbent surface with the characteristics of SiOs(s). The distribution of trace metals is examined as a function of pH, type of additional adsorbing surfaces [Fe(OH)&), MnOz(s)], surface area, and the presence of selected organic ligands (EDTA, citrate, aspartic acid, histidine, cysteine). The speciation and distribution of metals depend upon availability of organic ligands and surfaces. Specific conclusions are reached for each metal. The modeling of the equilibrium distribution of trace metals is affected not only by the choice of metals, inorganic ligands, and surfaces, but also by the availability and choice of their respective stability constants. H
Limnological studies on the chemical behavior of trace metals are generally concerned with the distribution of trace metals in natural waters. Previous investigations into the distributions of trace metals in natural waters (1-5) indicated that in the case of chemical behavior of trace metals in aqueous systems, the most important environmental parameters are: the type of trace metal and its concentration, the types of the adsorbents and the’available surface area, pH, and types and concentrations of organic and inorganic ligands present. This paper examines the chemical interactions and the competitions between trace metals and different constituents of natural waters; biological and physical aspects are beyond the scope of this paper. The response of oxic fresh water ( p c = 12, pco2 = 10-3.5 atm), containing major cations (Ca, Mg, K, Na), nine trace metals (Pb, Cu, Ni, Zn, Cd, Co, Hg, Mn, Fe), eight inorganic ligands (CO3, SO4, C1, F, Br, “3, PO4,OH), and a surface with the adsorption characteristics of SiOZ(s), is investigated in terms of changes in pH, types of adsorbing surfaces, surface area, and the presence of organic ligands. Changes in the physical-chemical distribution of trace Present address, Bechtel Power Corp., 12400 East Imperial Highway, Norwalk, Calif. 90650. 1302
Environmental Science & Technology
metals are examined in terms of two chemical models: for inorganic systems, and for systems containing organic complexing agents. The inorganic systems consider the following three cases: distributions of metals as a function of pH in the presence of constant adsorbing surface area; distributions of metals as a function of different adsorbing surfaces; and adsorption of metals (at a constant pH) as a function of adsorbing surface area. The organic systems consider the effect of selected organic ligands (EDTA, citrate, aspartic acid, histidine, cysteine) on the speciation and distribution of metals. These organic models consider the following: distributions of metals as a function of concentration of the above-listed organic ligands a t constant pH of 7 and constant adsorbing surface area; distributions of metals as a function of adsorbing surface area under the conditions of constant pH of 7 and equimolar concentrations of the organic ligands; and adsorption of metals (at constant pH of 7) as a function of both surface area and organic ligand concentrations. The surface area of the adsorbent was estimated to be 1.55 X hectaredl (1hectare (ha)/L = lo4 m2/L) by assuming a representative value of 310 mg/L suspended solids, expressed as equivalent SiOs(s), which has a specific surface area of 5 m2/g. Si02 was chosen as a model adsorbent solid because its surface properties are reasonably well understood and adsorption energies are available for many metals. I t is, of course, only one of several important solid adsorbents known to be significant in river waters [Fe(OH),(s), MnOs(s), clays, etc.]. Although SiOZ(s) has been chosen as the principal adsorbent in these model calculations, Fe(OH)s(s)and MnOs(s) are also included in this paper. The values for the chemical free energies of the adsorption of metals were either taken from the results of James and Healy ( 6 ) ,MacNaughton (7), Vuceta ( 8 ) ,Vuceta and Morgan (9-11), or set to -6.5 kcal/mol if the experimentally determined values were not available. In our present state of knowledge, considerable uncertainty exists as to the correct values of the chemical free energies of adsorption. The results presented here can be improved in reliability as better adsorption free energy data become available. The uncertainties are, perhaps, on the order of 1-2 kcal/mol, a t the present time. The equilibrium computations were obtained using the REDEQL 2 program (12). REDEQL 2 is a general purpose computer program especially adapted to study acid-base,
0013-936X/78/0912-1302$01 .OO/O @ 1978 American Chemical Society
r0 L
Table 1. Inputs of Metals and Ligands Used for Modeling Oxic Fresh Waters (pt = 12, pcO2= atm) a constituent
-log molar total concn
Ca Mg K Na Fe Mn cu Cd Zn Ni Hg Pb co
F Br
3.4 3.77 4.23 3.56 5.0 5.5 6.0 6.0 7.0 6.5 9.0 9.0 7.0 3.9 3.65 5.5 6.62
"3
5.5
PO4
5.0
so4 ci
I
I
I
I
I
I
I
I
CuOH
-
7 8
3 9
u,
10 -
CuNH,
II
cuer
12
I
I
6.2
6.4
6.6
6.8
70
7.2
7.4
7.6
7.8
i
8.0
PH Figure 1. Speciation of Cu(li) in natural fresh waters as a function of pH (log'p2 c ~ ( O H ) ~ ( a = q ) -13.7) in presence of 1.55 X ha/L SiOp(s) I r
6
LU
I
I
2*
I
1
I
1
CuC(
7
These values are representative of those reported for natural waters by Livingstone ( 13)and Bowen ( 14).
I
I
e
8
9
precipitation-dissolution, redox, and adsorption processes. Adsorption equilibria were computed according to the James-Healy Model using the ADSORP subroutine. R e s u l t s and Discussion
Inorganic Models. Representative fresh water concentrations of metals and inorganic ligands (13, 14), which are used as the constant inputs for modeling fresh waters, are tabulated in Table I. The compilation does not, however, include chromium, nitrate, and iodine. They were omitted for the following reasons. The speciation and distribution of chromium in river waters are poorly known (15).I t is thought to be critically determined by oxidation-reduction processes in which the equilibrium for the Cr(II1)-Cr(V1) redox couple is not attained. Since the model computations predict the equilibrium distribution, chromium would be computed to be present exclusively as Cr(V1);Le., CrOZ-. Nitrate was not included in model computations because: it does not form any important aqueous complexes with trace metals, and the behavior of nitrate is dominated by its important role in biological cycles. Iodine is present in natural waters both as iodide (I-) and iodate (IO;). Analytically, it has been determined that iodide forms 20-30% of the total iodine present (16).The biological activity of iodine through photosynthesis, however, often brings about a disequilibrium. Distributions of Metals as a Function of p H . The pH range investigated in this section is 6.2-8.0, which corresponds to the ~ C range T of 4.7-3.22. The computed distribution of Cu(I1) depends upon the choice of the second hydrolysis constant [*&Cu(OH)z(aq)].A large stability constant (log*& = -13.7) (17-19) results in highly pronounced adsorption throughout most of the examined pH range (Figure 1).The hydrolysis products of Cu(I1) are present a t concentrations Cu-OH = 0.25 Cu-ADS. Cu-CO3 species are present a t much lower concentrations, and complexes with halides, sulfate, phosphate, and ammonia are of even smaller significance. A smaller stability constant (log*& = -17.3) (20) results in a shift in equilibrium away from the production of hydrolyzed species (Figure 2). At the given pH and alkalinity, there will
0
a
10 I1
12
6.2
6.4 6.6 6.0
7.0
7.2
7.4
7.6
7.0 0.0
PH Figure 2. Distribution of Cu(ll) in natural fresh waters as a function of pH (log'P2Cu(OH)2(aq) = -17.3) in presence of 1.55 X ha/L Si02(s)
6 7
-
0
U
a I9O
rrc
62
64
66
68
70
72
14
76
70
BO
on Figure 3. Speciation of Pb(ll) in natural fresh waters as a function of pH in presence of 1.55 X 10-4 ha/L Si02(s)
be a tendency toward complexation with carbonate. Precipitation of CuzCO3(OH)*(s)will occur around pH 7.2, and above p H 7.5 most of the Cu(I1) will be removed from solution as CU~CO~(OH)~(S). According to the model calculations, Pb(I1) is present mainly as the free metal ion below p H 7.1 and as Pb-COB species above that pH; hydroxo species and adsorption account equally for most of the remaining Pb. The combined total of Pb-OH and Pb-C03 never exceeds 15% (Figure 3). Volume 12, Number 12, November 1978
1303
ha/L SiOp(s)
Table II. Equilibrium Speciation a in Oxic River Water of pH 7 in Presence of 1.55 X
total metals
concn
3.4 3.77
3.41 3.78
4.23 3.56 5.0 5.5 6.0 6.0 7.0 6.5 9.0 7.0 7.5
4.23 3.56 17.77 9.91 6.93 6.16 7.01 6.52 17.06 7.29 7.52 7.00
1
Cab Mg
K Na Fe(lll)c Mn cu Cd e Zn Ni Hg Pb co H
-
4
free concn
co3
SO4
CI
F
3.0 7.27
3.9 3.92
3.65 3.65
5.5 5.50
6.41 6.77
5.41 5.67
...
7.24 6.97 18.14 11.90 8.92 8.15 9.00 8.51 18.95 8.88 9.31 8.88
i
...
9.82
.,
,
12.42 7.84 8.17 9.08 8.40 13.39 7.50 9.49 4.16
... ... ...
ligands Br
6.62 6.62
8.01 7.67
...
...
... ...
...
...
...
20.30 12.65 8.55 7.79 9.44 9.86 9.30 9.41 9.96
17.92 11.32 10.75 11.41 11.12 21.15
.., .,,
23.76 15.82 12.64 10.87 12.92 12.73 12.47 12.20 13.73
...
9.57
...
.,.
"3
Po4
SlOAs)
5.50 7.52
5.0 10.97
3.8 4.01
11.04 11.20
7.41 7.27
...
... ...
6.05 6.85 6.90 6.24 8.91 11.84 6.17 9.00 9.58 8.39 10.06 8.14 9.87
10.28 13.12 6.73 9.27 9.48 8.33 9.39 8.10 9.82
...
...
... ...
15.84 12.90 9.51 13.78 10.40 10.21
16.73 10.36 11.18 12.34 11.35 14.59
...
...
...
13.04 5.50
11.11 5.29
OH
... 6.93 8.72 8.09
... ...
Concentrationsare expressed as negative logarithms of mol/L. * pCa5(P04)aOH(s) = 5.8. pFe(OH),(s) = 5.0. pMn02(s)= 5.5. e pCdC03(s) = 6.56.
95 8
IO
t
1
9 c
IO
c1 0
I1 I2 I3 ~~
62
64
66
68
70
72
74
76
18
80
Figure 4. Speciation of Zn(ll) in natural fresh waters as a function of pH ha/L Si02 in presence of 1.55 X I
1
I
I
I
I
I
e 9
g,
IO
1 62
64
66
68
70
72
74
76
78
80
pn
Figure 5. Distribution of Cd(ll) in natural fresh waters as a function of pH in presence of 1.55 X ha/L Si02(s)
Complexes with sulfate, chloride, and bromide appear a t very small concentrations. For Zn(II), throughout the pH range 6.2-8.0, the predominant species is the free metal ion (Figure 4). At a pH less than 7, the second most dominant species are its sulfate complexes and a t pH > 7, its carbonate complexes. At pH 8, only 2.6% of Zn is sorbed on SiO&). Based on the comparison of the respective stability constants, similar results can be expected also for Co and Ni. 1304
62
64
66
68
70
72
14
76
76
80
pn
pn
EnvironmentalScience 8. Technology
Figure 6. Speciation of Hg(ll) in natural fresh waters as a function of pH ha/L SiOp(s) in presence of 1.55 X
The formation of CdCO:r(s) precipitate above pH 6.9 results in a sudden decrease in the concentration of Cd2+ (Figure 5). Consequently, adsorption and complexation are of much smaller importance. Hg( 11), under the model environmental conditions, is present mainly either in chloro complexes (below pH 7.1) or in hydroxo complexes (above pH 7.1). At pH 8.1, less than 15%of Hg(I1) is adsorbed on SiOz(s) (Figure 6). Distributions of M e t a l s at C o n s t a n t p H as a F u n c t i o n o f Different Adsorbing Surfaces. The equilibrium distributions of trace metals in oxic fresh waters a t pH 7 in the presence of 1.55 X hectaredl SiOn(s) are shown in Table 11. The numbers are the negative logarithms of the molar concentrations, and dashed lines indicate that the species was not considered in the computations. The following results are of particular interest. The free metal ion is the predominant species for the following nine metals: Ca, Mg, K, Na, Cd(II), Zn(II), Ni(II), Pb(II), and Co(I1). Cu(I1) is primarily adsorbed on Si02(s),while Hg(I1) is mainly hydrolyzed. The inorganic ligands are associated mainly with Ca, Mg, K, and Na. The complexes with other metals are less significant. Approximately 61.6% of the Si02(s)surface remains free; the rest is associated with Ca (14.5%),Cu(I1) (9.9"/0)Na , (9.4%),K (2.5%), and Mg (2.0%). Four solid phases are formed: MnOz(s), Fe(OH):j(s), CaS(PO&OH(s), and CdCOy(s). It has been proposed by Jenne (21) that hydrous oxides of Mn and Fe provide the principal control on the fixation of metals in soils
ha/L Si02(s),
Table 111. Equilibrium Speciation a of Metals in Oxidizing Fresh Water in Presence of 1.55 X 3X ha/L (OH)s(s), and 8 X ha/L Mn02(s) ligands
-
total metals
Cab Mg K Na Fe(lll) Fe(ll) Mn cu Cd e Zn Ni Hg Pb co H a
concn
free concn
1
i
3.4 3.77 4.23 3.56 5.0
3.41 3.78 4.23 3.56 17.77 17.00 9.91 6.93 6.16 7.01 6.52 17.07 7.29 7.52 7.00
... 5.5 6.0 6.0 7.0 6.5 9.0 7.0 7.5
...
-
Cog
SO4
CI
F
Br
NH3
3.0
PO4
SlO2(s)
Fe(OH)!j(s)
MnOe(s)
3.90
3.65
5.50
6.62
5.50
5.0
3.80
4.52
5.1
...
7.27
3.92
3.65
5.50
6.62
7.52
10.97
4.01
4.67
8.56
6.93
6.41 6.77
5.41 5.67 7.24 6.97 18.14 18.99 11.90 8.92 8.15 9.01 8.52 18.96 8.88 9.32 8.88
... ...
8.01 7.67
11.04 11.20
7.41 7.27
... ...
... ...
20.30 19.94 12.65 8.56 7.79 9.44 9.86 9.31 9.41 9.96
17.92
... ... ...
6.05 6.85 6.90 6.24 8.91
9.09 9.59 7.28 6.59 13.13
6.70 6.92 11.70 10.90 13.94
8.72 8.09
... ...
... ... ... ...
... 9.82
...
... 12.42 7.84 8.18 9.09 8.40 13.40 7.50 9.50 4.16
23.76
... ...
... 15.82 12.64 10.87 12.92 12.73 12.48 12.20 13.73
11.32 10.75 11.41 11.12 21.16
...
...
...
9.57
23.22 16.73 10.36 11.18 12.34 11.35 14.60
15.84 13.99 12.90 9.51 13.78 10.41 10.22
...
... ...
13.05 5.50
11.12 5.28
Concentrations are expressed as negative logarithms of rnol/L. pCa5(PO&0H(s) = 5.8.
OH
...
...
...
...
...
11.84 6.18 9.00 9.58 8.39 10.07 8.14 9.88
15.55 8.00 11.71 11.25 1 1.29 10.72 11.05 12.60
11.73 8.94 8.49 9.24 9.15 15.83 9.59 9.31
10.28 18.71 13.12 6.73 9.27 9.48 8.33 9.40 8.10 9.83
...
...
...
pFe(OH)3(s)= 5.0. pMn02(s) = 5.5. e pCdCO&) = 6.56.
Table IV. Compilation from Tables II and 111 adsorption in presence of three surfaces
SI02 Me
Fe(lll) Mn Cu Cd Zn Ni Hg Pb Co
moi/m2
7X 9.32 X 4.26 X lop7 6X 1 X 10-lO 2.6 x 10-9 5.49 X lo-’’ 4.6X 8.5 X lo-”
oh
Mer
0.01
