- Email: [email protected]
Lead(II) complexation by fulvic acid: how it differs from fulvic acid complexation of copper(II) and cadmium(II)
Geocbimica et Cosmochimica Acto Vol. 44, pp. 1381 to 1384 Q Pergamon Press Lid 1980.Printed in GreatBritain
NOTE
Lead(I1) complexation by fulvic acid: how it differs from fufvic acid complexation of loper and cadmium(II) ROBERT A. SAARand Department
JAMES H.
WEBER*
of Chemistry, Parsons Hall, University of New Hampshire, Durham, NH 03824, U.S.A.
(Received 4 December 1979; accepted in revised form 15 April 1980)
Abstract-Pb’+, like Cu ‘+, forms strong complexes with fulvic acids (Cd”-fulvate complexes are much weaker), but Pb-fulvate precipitates at a much lower mole ratio of metal ion to fulvic acid than either Cu-fulvate or Cd-fulvate does. Physical association of Pb2+ with Pb-fulvate solids as well as complexation by sites still available in the precipitates probably causes the increased removal of free Pb*+ from _ solution after precipitation beg&.
workers in our group (WEBER and WILSON, 1975; WILSON and WEBER, 1977a, b). One fulvic acid sample (SFA) was A BUILD-UP of heavy-metal ions in a soil profile or their derived from a podzol soil and has a number-average mol accumulation in a water body may harm plants and aniwt of 644. The other fulvic acid samule (WFA) was isolated mals. Knowing the types and strengths of complexes of from a freshwater river and has a iumber-average mol wt organic matter and heavy metal ions is important to our of 626. Metal-ion titrants were prepared from aqueous understanding of the movement of those ions in the envirFisher 1000 ppm atomic absorption standard solutions. onment (GAMBLE and SCHNITZER, 1973; REUTER and PER- The electrolyte for all experiments was 0.1 M KNO,, preDUE, 1977; MANTOURA et al., 1978; JACKX)N et al., 1978). pared in doubly deionized water from purified crystals. Experiments described in the literature show that Cu” The titrations were performed into a Princeton Applied and PbZC complexes with the same organic fraction have Research (PAR) 9301 water-jacketed ceil. A P.M. Tamsimilar conditional stability constants, and that Cd”son T9 circulating water bath maintained the experimental organic matter complexes have stability constants one to tem~rature at 25 & 0.2”C. An Orion model 94-82 lead two orders of ma~itude smaller (TAKAMATSJ and Yostelectrode, used in conjunction with a PAR model K77 IDA, 1978; BUFFLE et al.,1977; GUY and CHAKRABARTI, saturated calomel reference electrode, allowed us to 1976; CHEAM and GAMBLE 1974; STEVENSON,1976, 1977). measure free Pb2+. To ensure proper functioning of the We studied Pb’+, Cu2+ and Cd” complexes of fulvic Pb*+ electrode, we had to isolate the reference electrode acid (FA), which is the acid-soluble fraction of naturally from the solution with a KNOB bridge (thereby preventing occurring organic matter in soils and water. Although our chloride interference) and bubble nitrogen through the experiments affirm that Pb *+-FA and Cu2+-FA comsolution to prevent oxidation of the electrode surface. plexes have similar conditional stability constants (SAAR An Orion model 92-29 electrode measured free Cu”‘, and WEBER, 1980), a comparison of separate Pb*+-into-FA and an Orion model 92-48 electrode measured free Cd’+ and Cu*‘-into-FA titrations shows that fulvic acid can ion. Chloride ion does not interfere with Cu2+ or Cd’+ lower the free Pb*+ concentration more than it can lower electrode response, so we used a single-junction reference the free CL?+ concentration during a certain part of the electrode. For all titrations, we monitored pH by means of titration. a glass electrode. The electrodes were connected to two We designed ion-selective electrode experiments that Orion 7OlAmV/pH meters so that we could follow the could record the differences between Pb2+-FA and metal ion and hydrogen ion readings at the same time. A CuZf-FA complexation as the mole ratio of total metal magnetic stir bar and stirrer maintained the homogeneity ion to total FA (C&.&) increased during M’+-into-FA of the solution. Finally, a Perkin-Elmer model 204 fluortitrations. We compared these results with those from simiescence spectrophotometer, with excitation and emission lar titrations for which we recorded solution scattering rather wavelengths set at 400 nm, allowed us to measure solution than the concentration of free metal ion. Thus, we were scattering. able to show a connection between the onset of Pb*+-FA Procedures precipitation and the augmented removal of free Pb2+ ions from solution. We dissolved sufficient SFA or WFA powder in KNOs to produce solutions having either 5 x lo-’ or 2 x IO-&M FA concentration; the molar concentrations are based on EXPERIMENTAL the previously noted number-average molecular weights. The FA solutions were adjusted to pH 4.0, 4.5, 5.0 or 6.0, Materials and apparatus depending on the pH of the titration to follow. We then added aliquots of metal-ion titrant and waited for a steady Titrations included either of two types of fulvic acid, millivolt reading (l-3 min, depending on the type of elecwhich were extracted and characterized previously by trode and the concentration to be measured). The calibration curves, done after each titration in 0.1 M KNOJ, are not linear at low concentration; this is es* Author to whom reprint requests should be addressed. INTRODUCTION
1381
Note
1382
pecially true of the Pb2’ electrode, which has the poorest detection limit of the three metal-ion electrodes. Therefore, we computer-fitted polynomials to the calibrations and used these equations to calculate free metal-ion concentration for the just-completed titration containing fulvic acid. The uncertainty in the free metal-ion concentration is +2?* for the Cd’+ and Cuzt curves, based either on a standard deviation calculation for replicate titrations. or based on a IfrO.2mV uncertainty in the output of the ionselective electrode, estimated from replicate calibration curves. The data from the Pb’+ electrode are less reproducible, since the electrode has a repeatability of f 1mV (based on replicate calibrations), which translates into a f l&-15% uncertainty in free metal-ion concentration. This is also the relative standard deviation for three replicate titrations with Pb” and SFA at pH 5.0, with + 15?, at the low-concentration end and ilO’\,, at the highconcentration end. For the scattering experiments, we prepared several sets of &ml samples. Each set had a specific pH and FA type and concentration, but the members of a set had different C&r,,. We then measured the scattering intensity of each solution with the fluorescence spectrofluorometer. A sharp increase in scattering coincided with formation of centrifugable solids. RESULTS AND DISCUSSION The ion-selective electrode titrations show that, for early parts of a titration where Gw/CF is low, Cu2’-FA and Pb’ ‘-FA compiexation occur to nearly identical extents. Figure 1 shows three titrations, one each with Cd’+, Cur’ and Pb” as titrant. All were performed at pH 5.0 with 5 x 10-s M SFA. The SFA clearly responds differently to Cd’+ than to Cu’+ or Pb ” : At any value of C&r,, there is substantially more free Cd2+ ion than CL?+ or Pb’+ ion. If the data are used for stability-constant calculations, stability constants for Cd ‘+--FA will be substantially smaller than wit1 those for Cu2+-FA and Pb’+-FA. SFA appears to act similarly toward Cuzc and Pb’+ until CM/CsFA equals about 0.5. after which there is less free Pb*+ in solution at any C&C,,, than there is Cu2+ at the same value for Cc&s,,. Work in our group (BRESNAHAN et al., 1978: SAAR and WEBER, 1979) and elsewhere (BURGHet al., 1978) shows
fi 3
-zr
_~~1-.-.._._.._.. 00
05
.._.----. IO
15
20
25
30
CM Gc
Fig. I. Molar free metal ion (M’+) concentration vs the mole ratio of total metal ion to total fulvic acid (CM,/&,) for separate titrations with Pbzs, Cu’” and Cd” titrants. The soil-derived fulvic acid (SFA) Conc~Iltration was 5 x 10w5. the pH 5.0, the electrolyte 0.1 M KN03, and the temperature 25’C.
lOi
__
*
/
f-%‘frltront
n Cu’Ftlirant x
Cd7+tmant
Fig. 2. Results for the same titrations and conditions as in Fig. 1. The y-axis variable is the fraction of all metal ions in each titrant ahquot that becomes bound as it mixes with the fulvic acid solution. This is expressed as dG/d(C,/Cr,). where r = [C, - (M* “)1/C,,. and Mz+ is the molar concentration of free metal ion.
that FA contains a variety of sites with varying abilities to complex metal ions. We expect that metal ions added early in a titration will bind to the strongest complexing sites. Metal ions in successive M ‘+ aliquots will have access to an ever smaller selection of FA complexation sites, so we would expect FA to bind a diminishing fraction of the added metal ions as a titration continues. The variables included in Fig. 2 are designed to test this model. The y-axis variable is the fraction of metal ions in each metal-ion aliquot that becomes bound as the aliquot mixes with the fulvic acid solution. This fraction can be expressed as dP/d(C,;C;,)
The x-axis variabie is the mole ratio of total metal ion to total fulvic acid (Cn/CFA) and is a measure of the progress of a titration. The data included in Fig. 2 are the same as those in Fig. 1, except that the Pb’+ curve is the average of three replicate titrations; the raw data for one of these replicates appear as the Pb*’ line in Fig. 1. Within ex.. perimental error ( +2”<, for Cu” and Cd’ ’ )%the y-axis variable for the Cu*+ and Cd’+ titrations declines as the titration proceeds, until the complexing capacity of fulvic acid for the metal ion has been reached. For the range of f,/Cr, used here, the complexing capacity for Cd*+ is reached (hence the leveling off of the Cd’+ line), whereas it is not for Cu*‘. Thus, Cu* f and CdZ+ act according to the model of fulvic acid complexation we have described. The averaged PbZf results have a relative standard deviation of + 12% at C,,/Csr, below approximateIy 0.65 and +S< for C,,jCsr, above 0.65. Many Pb-into-FA titrat&s, with SFA and WFA a&several pH values, showed an increase in the y-axis variable during some part of the titration. Even for those titrations in which there was no rise in this variable, after a certain C,,/Cr,, the difference in Pb2+ and Cur + performance is statistically significant: the fraction of added Pb” that appeared to become bound in Pb2+-into-FA titrations dropped more slowly than did the fraction of added Cu*’ in Cut’-into-FA titrations. An explanation for this behavior appeared when we
1383
Note
x
00’
00
2 x lO-4 M FA
I
I 05
IO
15
20
25
cw'cFA
Fig 3. pH4.5 titrations of Pb2+ solutions into two different concentrations of soil-derived fulvic acid in 0.1 M KNOJ at 25°C. C,, is total molar lead-ion concentration, CrA is total molar fulvic acid concentration, and V where (Pb’+) is the molar free is CG, - (Pb2’)3/C,,, lead-ion concentration. compared two titrations with Pb2+ titrant: one with 2 x 10s4 M SFA and the other with 5 x 10m5M SFA in the titration cell. Figure 3 shows these two titrations. The halt in the decline of the y-axis variable, which signals a change in the complexation behavior of Pb’+-FA as compared to Cu “-FA or Cd*“-FA, occurs at lower Cr,/Cs,, for the more concentrated SFA solution than for the more dilute one. Precipitation of insoluble PbZf-FA complexes also occurs at lower C,,/Cr, for the more concentrated SFA solutions. Figure 4 shows, for the same conditions as in Fig. 3, that solution scatter rises abruptly when a certain is reached. These scattering thresholds correCP&SFA spond to nearly the same C,,/Cs,, value where the y-axis variable in Fig. 3 stops declining or even starts rising. The data, then, show that Pb2+ removal from solution increases when Pb2+-FA begins to precipitate, implying mechanisms of Pb’+ removal other than complexation. These could be adsorption on or entrapment within
developing aggregates (ONG et al., 1970). The occurrence of such aggregates and the increase in FA mol wt upon addition of metal ions are noted frequently in the literature (STEVENSON, 1976, 1977; SIPOS et al., 1978; RAMUNNIand PALMIERI,1975; JACKX~Nand SKIPPEN,1978). Even so, the Pb2+ binding curve begins to drop again as usable complexation sites become occupied and possibly as the amount of new Pb ” adsorption or entrapment declines. Insoluble complexes of FA and either Cu* + or Cd2 + can also form (MACCARTHY and O’CINNEIDE. 1974; WHITWORTH and PAGENKOPF,f979), but only when the metal ions are in large excess (Cfold excess for Cu2+ and 20-fold excess for Cd” in a 5 x IO-’ M solution of our SFA). The unusual curve shape we see for the Pb’+-SFA system in Figs 2 and 3 may occur for the Cu’+-FA and conceivably for the Cd ’ + -FA systems, but to see such effects for Cu2+ or CdZ+, we would have to add large quantities of metal ion. Data from the latest part of such a titration would have large errors because the FA complexing capacity would be far exceeded and [M”] would be only slightly smaller than CM.
CONCLUSIONS Lead ion, then, is unique among these three metal ions: Only Pb’*-FA begins precipitating before the FA complexing capacity for the metal ion has been reached. Our work with Cu *+--FA complexes shows there are two or more Cu2+ complexation sites per fulvic acid molecule. That is, during Cu 2”-into-FA titrations, V values rise above 2.0 before any Cu “‘---FA precipitation begins. If we assume that most of the complexation sites available to Cu2+ are available to Pb’+, it appears that PbFA solids settle out with complexation sites on the FA unfilled. At this point it is not clear whether formation of the precipitate blocks these unfilled sites. It would not be unreasonable to speculate that some of these sites are still available, thereby making Pb-FA solids unusually good at sequestering free metal ions-both through surface adsorption to the solid and through chemical chelation. Indeed, the fraction of PbZf ions in each ahquot that becomes bound remains significantly higher than does the fraction of Cu’+ ions that become bound for high C&r,. Our earlier work (SAAR and WEBER, 1979) showed an unusual property of Cd 2+-FA complexes. Their conditional stability constants increase as the FA concentration drops below about 1 x IO-“ M, in contrast to Cu’+-FA complexes, whose stability constants vary little with changing FA concentration. It appears that PbZ’ binds to FA much the way Cu2+ does at low C&r,: FA concentration does not influence complexation. But at higher Chl/CFA,a change in FA concentration may noticeably alter the degree to which Pb’+ is removed from solution. The extra removal of PbZ+ from solution mav be important in the study of Pb’+ movement in waters’that are polluted or rich in organic matter, and in soils to which sludges containing metal ions have been added. Acknowledgement-National Science Foundation grants OCE 77-08390 and OCE 79-10571 provided partial funding for this research.
*-*-it-
0' 00
03
06
09
12
I5
WC,
Fig. 4. Scattering of pH 4.5 Pb’+-fulvic acid solutions for two different concentrations of so&derived fulvic acid in 0.1 M KNO,. C&, is total molar lead-ion concentration, and C,, total molar fulvic acid concentration. O.C.A. 4419-I
REFERENCES BRESNAHANW. T., GRANT C. L. and WEBERJ. H. (1978) Stability constants for the complexation of copper(H) ions with water and soil fuivic acids measured by an ion selective electrode. Anat. Chem. SO, 167551679. BUFFLEJ., GRETER F.-L. and HAERDIW. (1977) Measurement of complexation properties of humic and f&vie
1384
Note
acids in natural waters with lead and copper ionselective electrodes. Anal. Chem. 49, 216-222. B~JRCH R. D., LANGFORD C. H. and GAMBLE II.S. (1978) Methods for the comparison of fuivic acid samples: the effects of origin and concentration on acidic properties, Can. J. C/rem. 56, 11961201. CHEAM V. and GAMBLE D. S. (1974) Metal-fulvic acid cheiation equilibrium in aqueous NaNO, solution. Hg(II), Cd(H), and Cu(II) fuivate complexes. Can. J. Soil Sci. 54, 413-417. GAMBLED. S. and SCHNITZERM. (1973) The chemistry of fuivic acid and its reactions with metai ions. In Trace Metals and Metal-Organic Interactions in Natural Wafers. (ed. P. C. Singer) Chap. 9. Ann Arbor. C&u R. D. and CHAKRABARTI C. L. (1976) Studies of metai-
organic interactions in model systems pertaining to natural waters. Can. /. Cfrefn. M, 26OO261 I. JACKSON K. S. and SKIPPEN G. B. (1978) Geochemicai dispersion of heavy metals via organic complexing: a laboratory study of copper. lead. zinc. and nickel behaviour at a simulated sediment-water boundary. J. Geochem. Explor. 10, 117.138. JACKSONK. S., JONAS~ONI. R. and SKIPPEN G. B. (1978) The nature of metals-sediment-water interactions in freshwater bodies. with emphasis on the role of organic matter. Earth-Sk Rw. 14, 97-146. MACCARTHYP. and O’CINNEIDES. (1974) Fuivic acid. 11. Interaction with metal ions. J. Soil Sri. 25, 429-437. MANTOURAR. F. C.. DICKERED A. and RILEYJ. P. (1978) The compiexation of metals with humic materials in natural waters. Est~ari~e Coasrui Mar. Sci. 6, 387408‘ ONG II. L., SWAMON V. E. and Brsy~s R. E. (1970) Natural organic acids as agents of chemical weathering. U.S. Geol. Surr:rr Pro/: Pap. 700C, C130-C137. RAMUNNIA. and PALMIER~F. (1975) Sui compiessi degli acidi umici di un suolo bruno con aicuni cationi bi e trivalenti. Agrochimica 19, 224-234.
REU-~E:RJ. H. and PEKUU~:E. M. (1977) Importance of heavy metal-organic matter jnteractions in natural waters. Geochim. Cosmochim. Acta 41, 325-334. SAARR. A. and WEBERJ. H. (1979) Compiexation of cadmium(H) with water- and soil-derived fuivic acids: effect of pH and fuivic acid concentration. Can J. Chrm. 57,
I x3-- 1268. SAAR R. A. and WEUERJ. H. (1980) Lead(H)-fuivic
acid complexes. Conditional stability constants, soiubiiity, and implications for lead(H) mobility. Enciron. Scr‘. Technol. 14, in press. Smos S.. SIPUSI?.. D~K~NY 1.. DE&RA., MEISEL.I. and LAK~‘ros B. (1978) Biopolymer-metal complex systems. Il. Physical properties of humic substances and their metal complexes. ncta &run. Acud. Sri. Hung. 27, 31-42. ST~:\IENSO~“~ F J. (1976) Stability constants of Cu”. Pb’“, and Cd” complexes with humic acids. Soil Sci. Sot. Ant. J. 40, 665 -672. STEVENS~FC F‘. J. (1977) Nature of divaient transition metal complexes of humic acids as revealed by a modified potentiometric titration method. Soil Sci. 123, 10-17. TAKAMATSI:T. and YOSHIDAT. (1978) Determination of stability constants of metal-humic acid complexes bv potentiometric titration and ion-selective electrodes. Soil Sci. 125, 377.-386. WEBERJ. H. and Wu.soh’ S. A. (1975) The isolation and characterization of fulvic acid and humic acid from river waler. Wafer Res. 9, 1079-1084. WHITWOKTHC. and PAC;ENKOPFG. K. (1979) Cadmium compiexation by coal humic acid. J. fnory. 1vucl. Cltem. 41, 317 -3’1. WIL.SONS. A and WE~EK .I. II. (1977a) A comp~ative study of number-average dissociation-corrected moiecular weights of fuivic acids isolated from water and soil. c/tc?n. Grol. 19. 2x5 29 3. WILSONS. A. and WER~XJ, H. (1977b) Electron spin resonance analysis of semiquinone free radicals of aquatic and soil fulvic and humic acids. Anal. Lett. 10, 75-84.
NOTE
Lead(I1) complexation by fulvic acid: how it differs from fufvic acid complexation of loper and cadmium(II) ROBERT A. SAARand Department
JAMES H.
WEBER*
of Chemistry, Parsons Hall, University of New Hampshire, Durham, NH 03824, U.S.A.
(Received 4 December 1979; accepted in revised form 15 April 1980)
Abstract-Pb’+, like Cu ‘+, forms strong complexes with fulvic acids (Cd”-fulvate complexes are much weaker), but Pb-fulvate precipitates at a much lower mole ratio of metal ion to fulvic acid than either Cu-fulvate or Cd-fulvate does. Physical association of Pb2+ with Pb-fulvate solids as well as complexation by sites still available in the precipitates probably causes the increased removal of free Pb*+ from _ solution after precipitation beg&.
workers in our group (WEBER and WILSON, 1975; WILSON and WEBER, 1977a, b). One fulvic acid sample (SFA) was A BUILD-UP of heavy-metal ions in a soil profile or their derived from a podzol soil and has a number-average mol accumulation in a water body may harm plants and aniwt of 644. The other fulvic acid samule (WFA) was isolated mals. Knowing the types and strengths of complexes of from a freshwater river and has a iumber-average mol wt organic matter and heavy metal ions is important to our of 626. Metal-ion titrants were prepared from aqueous understanding of the movement of those ions in the envirFisher 1000 ppm atomic absorption standard solutions. onment (GAMBLE and SCHNITZER, 1973; REUTER and PER- The electrolyte for all experiments was 0.1 M KNO,, preDUE, 1977; MANTOURA et al., 1978; JACKX)N et al., 1978). pared in doubly deionized water from purified crystals. Experiments described in the literature show that Cu” The titrations were performed into a Princeton Applied and PbZC complexes with the same organic fraction have Research (PAR) 9301 water-jacketed ceil. A P.M. Tamsimilar conditional stability constants, and that Cd”son T9 circulating water bath maintained the experimental organic matter complexes have stability constants one to tem~rature at 25 & 0.2”C. An Orion model 94-82 lead two orders of ma~itude smaller (TAKAMATSJ and Yostelectrode, used in conjunction with a PAR model K77 IDA, 1978; BUFFLE et al.,1977; GUY and CHAKRABARTI, saturated calomel reference electrode, allowed us to 1976; CHEAM and GAMBLE 1974; STEVENSON,1976, 1977). measure free Pb2+. To ensure proper functioning of the We studied Pb’+, Cu2+ and Cd” complexes of fulvic Pb*+ electrode, we had to isolate the reference electrode acid (FA), which is the acid-soluble fraction of naturally from the solution with a KNOB bridge (thereby preventing occurring organic matter in soils and water. Although our chloride interference) and bubble nitrogen through the experiments affirm that Pb *+-FA and Cu2+-FA comsolution to prevent oxidation of the electrode surface. plexes have similar conditional stability constants (SAAR An Orion model 92-29 electrode measured free Cu”‘, and WEBER, 1980), a comparison of separate Pb*+-into-FA and an Orion model 92-48 electrode measured free Cd’+ and Cu*‘-into-FA titrations shows that fulvic acid can ion. Chloride ion does not interfere with Cu2+ or Cd’+ lower the free Pb*+ concentration more than it can lower electrode response, so we used a single-junction reference the free CL?+ concentration during a certain part of the electrode. For all titrations, we monitored pH by means of titration. a glass electrode. The electrodes were connected to two We designed ion-selective electrode experiments that Orion 7OlAmV/pH meters so that we could follow the could record the differences between Pb2+-FA and metal ion and hydrogen ion readings at the same time. A CuZf-FA complexation as the mole ratio of total metal magnetic stir bar and stirrer maintained the homogeneity ion to total FA (C&.&) increased during M’+-into-FA of the solution. Finally, a Perkin-Elmer model 204 fluortitrations. We compared these results with those from simiescence spectrophotometer, with excitation and emission lar titrations for which we recorded solution scattering rather wavelengths set at 400 nm, allowed us to measure solution than the concentration of free metal ion. Thus, we were scattering. able to show a connection between the onset of Pb*+-FA Procedures precipitation and the augmented removal of free Pb2+ ions from solution. We dissolved sufficient SFA or WFA powder in KNOs to produce solutions having either 5 x lo-’ or 2 x IO-&M FA concentration; the molar concentrations are based on EXPERIMENTAL the previously noted number-average molecular weights. The FA solutions were adjusted to pH 4.0, 4.5, 5.0 or 6.0, Materials and apparatus depending on the pH of the titration to follow. We then added aliquots of metal-ion titrant and waited for a steady Titrations included either of two types of fulvic acid, millivolt reading (l-3 min, depending on the type of elecwhich were extracted and characterized previously by trode and the concentration to be measured). The calibration curves, done after each titration in 0.1 M KNOJ, are not linear at low concentration; this is es* Author to whom reprint requests should be addressed. INTRODUCTION
1381
Note
1382
pecially true of the Pb2’ electrode, which has the poorest detection limit of the three metal-ion electrodes. Therefore, we computer-fitted polynomials to the calibrations and used these equations to calculate free metal-ion concentration for the just-completed titration containing fulvic acid. The uncertainty in the free metal-ion concentration is +2?* for the Cd’+ and Cuzt curves, based either on a standard deviation calculation for replicate titrations. or based on a IfrO.2mV uncertainty in the output of the ionselective electrode, estimated from replicate calibration curves. The data from the Pb’+ electrode are less reproducible, since the electrode has a repeatability of f 1mV (based on replicate calibrations), which translates into a f l&-15% uncertainty in free metal-ion concentration. This is also the relative standard deviation for three replicate titrations with Pb” and SFA at pH 5.0, with + 15?, at the low-concentration end and ilO’\,, at the highconcentration end. For the scattering experiments, we prepared several sets of &ml samples. Each set had a specific pH and FA type and concentration, but the members of a set had different C&r,,. We then measured the scattering intensity of each solution with the fluorescence spectrofluorometer. A sharp increase in scattering coincided with formation of centrifugable solids. RESULTS AND DISCUSSION The ion-selective electrode titrations show that, for early parts of a titration where Gw/CF is low, Cu2’-FA and Pb’ ‘-FA compiexation occur to nearly identical extents. Figure 1 shows three titrations, one each with Cd’+, Cur’ and Pb” as titrant. All were performed at pH 5.0 with 5 x 10-s M SFA. The SFA clearly responds differently to Cd’+ than to Cu’+ or Pb ” : At any value of C&r,, there is substantially more free Cd2+ ion than CL?+ or Pb’+ ion. If the data are used for stability-constant calculations, stability constants for Cd ‘+--FA will be substantially smaller than wit1 those for Cu2+-FA and Pb’+-FA. SFA appears to act similarly toward Cuzc and Pb’+ until CM/CsFA equals about 0.5. after which there is less free Pb*+ in solution at any C&C,,, than there is Cu2+ at the same value for Cc&s,,. Work in our group (BRESNAHAN et al., 1978: SAAR and WEBER, 1979) and elsewhere (BURGHet al., 1978) shows
fi 3
-zr
_~~1-.-.._._.._.. 00
05
.._.----. IO
15
20
25
30
CM Gc
Fig. I. Molar free metal ion (M’+) concentration vs the mole ratio of total metal ion to total fulvic acid (CM,/&,) for separate titrations with Pbzs, Cu’” and Cd” titrants. The soil-derived fulvic acid (SFA) Conc~Iltration was 5 x 10w5. the pH 5.0, the electrolyte 0.1 M KN03, and the temperature 25’C.
lOi
__
*
/
f-%‘frltront
n Cu’Ftlirant x
Cd7+tmant
Fig. 2. Results for the same titrations and conditions as in Fig. 1. The y-axis variable is the fraction of all metal ions in each titrant ahquot that becomes bound as it mixes with the fulvic acid solution. This is expressed as dG/d(C,/Cr,). where r = [C, - (M* “)1/C,,. and Mz+ is the molar concentration of free metal ion.
that FA contains a variety of sites with varying abilities to complex metal ions. We expect that metal ions added early in a titration will bind to the strongest complexing sites. Metal ions in successive M ‘+ aliquots will have access to an ever smaller selection of FA complexation sites, so we would expect FA to bind a diminishing fraction of the added metal ions as a titration continues. The variables included in Fig. 2 are designed to test this model. The y-axis variable is the fraction of metal ions in each metal-ion aliquot that becomes bound as the aliquot mixes with the fulvic acid solution. This fraction can be expressed as dP/d(C,;C;,)
The x-axis variabie is the mole ratio of total metal ion to total fulvic acid (Cn/CFA) and is a measure of the progress of a titration. The data included in Fig. 2 are the same as those in Fig. 1, except that the Pb’+ curve is the average of three replicate titrations; the raw data for one of these replicates appear as the Pb*’ line in Fig. 1. Within ex.. perimental error ( +2”<, for Cu” and Cd’ ’ )%the y-axis variable for the Cu*+ and Cd’+ titrations declines as the titration proceeds, until the complexing capacity of fulvic acid for the metal ion has been reached. For the range of f,/Cr, used here, the complexing capacity for Cd*+ is reached (hence the leveling off of the Cd’+ line), whereas it is not for Cu*‘. Thus, Cu* f and CdZ+ act according to the model of fulvic acid complexation we have described. The averaged PbZf results have a relative standard deviation of + 12% at C,,/Csr, below approximateIy 0.65 and +S< for C,,jCsr, above 0.65. Many Pb-into-FA titrat&s, with SFA and WFA a&several pH values, showed an increase in the y-axis variable during some part of the titration. Even for those titrations in which there was no rise in this variable, after a certain C,,/Cr,, the difference in Pb2+ and Cur + performance is statistically significant: the fraction of added Pb” that appeared to become bound in Pb2+-into-FA titrations dropped more slowly than did the fraction of added Cu*’ in Cut’-into-FA titrations. An explanation for this behavior appeared when we
1383
Note
x
00’
00
2 x lO-4 M FA
I
I 05
IO
15
20
25
cw'cFA
Fig 3. pH4.5 titrations of Pb2+ solutions into two different concentrations of soil-derived fulvic acid in 0.1 M KNOJ at 25°C. C,, is total molar lead-ion concentration, CrA is total molar fulvic acid concentration, and V where (Pb’+) is the molar free is CG, - (Pb2’)3/C,,, lead-ion concentration. compared two titrations with Pb2+ titrant: one with 2 x 10s4 M SFA and the other with 5 x 10m5M SFA in the titration cell. Figure 3 shows these two titrations. The halt in the decline of the y-axis variable, which signals a change in the complexation behavior of Pb’+-FA as compared to Cu “-FA or Cd*“-FA, occurs at lower Cr,/Cs,, for the more concentrated SFA solution than for the more dilute one. Precipitation of insoluble PbZf-FA complexes also occurs at lower C,,/Cr, for the more concentrated SFA solutions. Figure 4 shows, for the same conditions as in Fig. 3, that solution scatter rises abruptly when a certain is reached. These scattering thresholds correCP&SFA spond to nearly the same C,,/Cs,, value where the y-axis variable in Fig. 3 stops declining or even starts rising. The data, then, show that Pb2+ removal from solution increases when Pb2+-FA begins to precipitate, implying mechanisms of Pb’+ removal other than complexation. These could be adsorption on or entrapment within
developing aggregates (ONG et al., 1970). The occurrence of such aggregates and the increase in FA mol wt upon addition of metal ions are noted frequently in the literature (STEVENSON, 1976, 1977; SIPOS et al., 1978; RAMUNNIand PALMIERI,1975; JACKX~Nand SKIPPEN,1978). Even so, the Pb2+ binding curve begins to drop again as usable complexation sites become occupied and possibly as the amount of new Pb ” adsorption or entrapment declines. Insoluble complexes of FA and either Cu* + or Cd2 + can also form (MACCARTHY and O’CINNEIDE. 1974; WHITWORTH and PAGENKOPF,f979), but only when the metal ions are in large excess (Cfold excess for Cu2+ and 20-fold excess for Cd” in a 5 x IO-’ M solution of our SFA). The unusual curve shape we see for the Pb’+-SFA system in Figs 2 and 3 may occur for the Cu’+-FA and conceivably for the Cd ’ + -FA systems, but to see such effects for Cu2+ or CdZ+, we would have to add large quantities of metal ion. Data from the latest part of such a titration would have large errors because the FA complexing capacity would be far exceeded and [M”] would be only slightly smaller than CM.
CONCLUSIONS Lead ion, then, is unique among these three metal ions: Only Pb’*-FA begins precipitating before the FA complexing capacity for the metal ion has been reached. Our work with Cu *+--FA complexes shows there are two or more Cu2+ complexation sites per fulvic acid molecule. That is, during Cu 2”-into-FA titrations, V values rise above 2.0 before any Cu “‘---FA precipitation begins. If we assume that most of the complexation sites available to Cu2+ are available to Pb’+, it appears that PbFA solids settle out with complexation sites on the FA unfilled. At this point it is not clear whether formation of the precipitate blocks these unfilled sites. It would not be unreasonable to speculate that some of these sites are still available, thereby making Pb-FA solids unusually good at sequestering free metal ions-both through surface adsorption to the solid and through chemical chelation. Indeed, the fraction of PbZf ions in each ahquot that becomes bound remains significantly higher than does the fraction of Cu’+ ions that become bound for high C&r,. Our earlier work (SAAR and WEBER, 1979) showed an unusual property of Cd 2+-FA complexes. Their conditional stability constants increase as the FA concentration drops below about 1 x IO-“ M, in contrast to Cu’+-FA complexes, whose stability constants vary little with changing FA concentration. It appears that PbZ’ binds to FA much the way Cu2+ does at low C&r,: FA concentration does not influence complexation. But at higher Chl/CFA,a change in FA concentration may noticeably alter the degree to which Pb’+ is removed from solution. The extra removal of PbZ+ from solution mav be important in the study of Pb’+ movement in waters’that are polluted or rich in organic matter, and in soils to which sludges containing metal ions have been added. Acknowledgement-National Science Foundation grants OCE 77-08390 and OCE 79-10571 provided partial funding for this research.
*-*-it-
0' 00
03
06
09
12
I5
WC,
Fig. 4. Scattering of pH 4.5 Pb’+-fulvic acid solutions for two different concentrations of so&derived fulvic acid in 0.1 M KNO,. C&, is total molar lead-ion concentration, and C,, total molar fulvic acid concentration. O.C.A. 4419-I
REFERENCES BRESNAHANW. T., GRANT C. L. and WEBERJ. H. (1978) Stability constants for the complexation of copper(H) ions with water and soil fuivic acids measured by an ion selective electrode. Anat. Chem. SO, 167551679. BUFFLEJ., GRETER F.-L. and HAERDIW. (1977) Measurement of complexation properties of humic and f&vie
1384
Note
acids in natural waters with lead and copper ionselective electrodes. Anal. Chem. 49, 216-222. B~JRCH R. D., LANGFORD C. H. and GAMBLE II.S. (1978) Methods for the comparison of fuivic acid samples: the effects of origin and concentration on acidic properties, Can. J. C/rem. 56, 11961201. CHEAM V. and GAMBLE D. S. (1974) Metal-fulvic acid cheiation equilibrium in aqueous NaNO, solution. Hg(II), Cd(H), and Cu(II) fuivate complexes. Can. J. Soil Sci. 54, 413-417. GAMBLED. S. and SCHNITZERM. (1973) The chemistry of fuivic acid and its reactions with metai ions. In Trace Metals and Metal-Organic Interactions in Natural Wafers. (ed. P. C. Singer) Chap. 9. Ann Arbor. C&u R. D. and CHAKRABARTI C. L. (1976) Studies of metai-
organic interactions in model systems pertaining to natural waters. Can. /. Cfrefn. M, 26OO261 I. JACKSON K. S. and SKIPPEN G. B. (1978) Geochemicai dispersion of heavy metals via organic complexing: a laboratory study of copper. lead. zinc. and nickel behaviour at a simulated sediment-water boundary. J. Geochem. Explor. 10, 117.138. JACKSONK. S., JONAS~ONI. R. and SKIPPEN G. B. (1978) The nature of metals-sediment-water interactions in freshwater bodies. with emphasis on the role of organic matter. Earth-Sk Rw. 14, 97-146. MACCARTHYP. and O’CINNEIDES. (1974) Fuivic acid. 11. Interaction with metal ions. J. Soil Sri. 25, 429-437. MANTOURAR. F. C.. DICKERED A. and RILEYJ. P. (1978) The compiexation of metals with humic materials in natural waters. Est~ari~e Coasrui Mar. Sci. 6, 387408‘ ONG II. L., SWAMON V. E. and Brsy~s R. E. (1970) Natural organic acids as agents of chemical weathering. U.S. Geol. Surr:rr Pro/: Pap. 700C, C130-C137. RAMUNNIA. and PALMIER~F. (1975) Sui compiessi degli acidi umici di un suolo bruno con aicuni cationi bi e trivalenti. Agrochimica 19, 224-234.
REU-~E:RJ. H. and PEKUU~:E. M. (1977) Importance of heavy metal-organic matter jnteractions in natural waters. Geochim. Cosmochim. Acta 41, 325-334. SAARR. A. and WEBERJ. H. (1979) Compiexation of cadmium(H) with water- and soil-derived fuivic acids: effect of pH and fuivic acid concentration. Can J. Chrm. 57,
I x3-- 1268. SAAR R. A. and WEUERJ. H. (1980) Lead(H)-fuivic
acid complexes. Conditional stability constants, soiubiiity, and implications for lead(H) mobility. Enciron. Scr‘. Technol. 14, in press. Smos S.. SIPUSI?.. D~K~NY 1.. DE&RA., MEISEL.I. and LAK~‘ros B. (1978) Biopolymer-metal complex systems. Il. Physical properties of humic substances and their metal complexes. ncta &run. Acud. Sri. Hung. 27, 31-42. ST~:\IENSO~“~ F J. (1976) Stability constants of Cu”. Pb’“, and Cd” complexes with humic acids. Soil Sci. Sot. Ant. J. 40, 665 -672. STEVENS~FC F‘. J. (1977) Nature of divaient transition metal complexes of humic acids as revealed by a modified potentiometric titration method. Soil Sci. 123, 10-17. TAKAMATSI:T. and YOSHIDAT. (1978) Determination of stability constants of metal-humic acid complexes bv potentiometric titration and ion-selective electrodes. Soil Sci. 125, 377.-386. WEBERJ. H. and Wu.soh’ S. A. (1975) The isolation and characterization of fulvic acid and humic acid from river waler. Wafer Res. 9, 1079-1084. WHITWOKTHC. and PAC;ENKOPFG. K. (1979) Cadmium compiexation by coal humic acid. J. fnory. 1vucl. Cltem. 41, 317 -3’1. WIL.SONS. A and WE~EK .I. II. (1977a) A comp~ative study of number-average dissociation-corrected moiecular weights of fuivic acids isolated from water and soil. c/tc?n. Grol. 19. 2x5 29 3. WILSONS. A. and WER~XJ, H. (1977b) Electron spin resonance analysis of semiquinone free radicals of aquatic and soil fulvic and humic acids. Anal. Lett. 10, 75-84.