United States Environmental Protection Agency Water Engineering Research Laboratory Cincinnati OH 45268 Research and Development EPA/600/S2-85/053 Sept. 1985 &ERA Project Summary Determination and Prediction of Chemical Forms of Trace Metals in Sewage Sludge and Sludge-Amended Soils L. J. Lund, G. Sposito, and A. L. Page A computer program. GEOCHEM, was developed to calculate trace metal equilibria in soil solutions affected by the application of sewage sludges. The structure of the program, data required. and typical applications of metal speci- ation in aqueous extracts from munici- pal sewage sludges and sludge- amended soils are described and discussed. Fulvic acids extracted from sewage sludge were found to contain 20 to 40 times more sulfur than fulvic acids extracted from soil organic matter. Sul- fonyls and surfhydryls therefore play a significant role in the complexation of metals in sludges and probably also in sludge-amended soils. The partitioning of metals in the organic and inorganic aqueous phases from sludge and sludge-amended soils differs among sludges. In general, Cu(ll) and Fe(lll) show high preference for organic com- plexation, and Cd, Ni, and Zn tend to form inorganic complexes preferen- tially. Following equilibrium of sludges from seven treatment plants with three soils for a period of 50 weeks at 1 /3 bar saturation, essentially all of the soluble Cu was organically complexed, and most of the soluble Cd, Ni, and Zn were present in solution as the free divalent ion. A procedure to fractionate sludges into various chemical forms was devel- oped. The procedure consisted of extraction with KNO3 (exchangeable), followed by distilled, deionized water (adsorbed). 0.5 M NaOH (organically bound), Na2 EDTA (carbonate), and HNO3 (sulfide-residual). The distri- bution of metals in various forms fol- lowing equilibration of soils with sludges is presented and discussed. In general, only very small percentages of the total quantity of metals in sludge- amended soils occur in soluble, exchan- geable, or adsorbed forms. This Project Summary was developed by EPA's Water Engineering Research Laboratory, Cincinnati. OH. to announce key findings of the research project that is fully documented in a separate report of the same title (see Project Report ordering information at back). Introduction The use of land for the disposal or recy- cling of municipal sewage sludges has increased during the past decade and is continuing to grow. Sewage wastes are receiving increased attention for use as phosphorus fertilizers, supplemental nit- rogen fertilizers, and soil conditioners. Evidence exists that trace metals in sludge-amended soils are toxic to a number of plant species, are accumu- lated by many types of vegetation, and may be a source of surface water and groundwater contamination. The basic chemistry of trace metals in sewage sludges and sludge-amended soils thus needs to be fully understood. The reac- tion rates, mechanisms, and products of trace metals following their application to soils in the form of sewage sludge are not well known. The soluble and insoluble forms in sludge-amended soils will be important in determining plant availabil- ------- ity, toxicity, and mobility of trace metals. This project investigated the chemical forms of five trace metals (Cd, Zn, Ni, Cu, and Pb) in municipal sewage sludges and the changes that occur in them following sludge application to soil as influenced by time, sludge source, soil characteristics, and solution composition. This informa- tion will help determine conditions under which sludges can be safely recycled on agricultural land, and it will aid in the selection of the most desirable soils for disposal of sludges. The data will also serve as a basis for determining potential degradation of soils and waters by sludge application to land. Sewage wastes vary greatly with the treatment plant, the time spent at any one treatment plant, and the treatment process. Because anaerobically digested sludges represent the highest percen- tage of those generated in the United States, this project emphasizes anaerobi- cally digested sludge. Because of varia- tions in the trace metal contents of sludge from different treatment facilities, sludges from a number of wastewater treatment plants were investigated. Description of GEOCHEM GEOCHEM is a computer program that enables the calculation of trace metal equilibria in soil solutions of widely vary- ing compositions. The program was developed and tested for use with sludge- amended soils. The method of calculation is based on chemical thermodynamics. A mole balance equation is set up for each component of a soil solution and thermo- dynamic equilibrium constants corrected for ionic strength are incorporated into the various terms of this equation accord- ing to the law of mass action. The solu- tion of the set of nonlinear algebraic equations that results from simultane- ously applying mole balance to all the components ultimately provides the con- centration of each dissolved, solid, or adsorbed species i n the soiI system under consideration. Applications of GEOCHEM Some typical applications of GEO- CHEM would include: (1) predicting the concentrations of inorganic and organic complexes of a metal cation in a soil solu- tion, (2) calculating the concentration of a particular chemical form of a nutrient element in a solution bathing plant roots so as to correlate that form with nutrient uptake, (3) predicting the fate of a pollu- tant metal added to a soil solution of known characteristics, and (4) estimat- ing the effect of changing pH, ionic strength, redox potential, water content. or concentration of some element on the solubility of a chosen chemical element in a soil solution. To illustrate the application of GEO- CHEM, Table 1 lists the analytical data for a saturation extract of an Altamont soil amended with sewage sludge and CdS04 . In this example, there were 9 metals (including H+ and 15 ligands (including OH") that could form 261 dif- ferent complexes according to the ther- modynamic data in the program. The output from a GEOCHEM subroutine for the data in Table 1 appears in Table 2. The percentage distribution figures for the metals and ligands constitute considera- ble worthwhile information, but what stands out is the strong complexation of Cu(ll) and Fe(lll) by the organic ligands and the relative lack of complexation of Cd, Mn(ll), and Zn. The major cations, Ca, Na, and K, are involved principally with the inorganic part of the system, whereas Cu(ll) and Fe(lll) are involved principally with the organic part. The remaining minor cations Cd, Mn(ll), and Zn, fall somewhere in between. Results Analytical Properties of the Fulvic Acid Fractions of Three Sludges An understanding of the mechanisms through which sewage sludges incorpo- rated in soil form water-soluble com- plexes with trace metals should provide a basis for predicting and controlling levels of these metals in the soil solution. In view of the existing literature on metal complexation by components of natural soil organic matter, it is reasonable to suggest that the fulvic acid fraction of sludge and of sludge-amended soil would be a major ligand for the complexation of trace metals. The fulvic acid fractions of three sludges were, therefore, extracted, purified, and then studied for their ana- lytical properties. The main difference between the fulvic acids extracted from sewage sludge (SS fulvic acids) and the fulvic acids extracted from soil, organic matter (SOM fulvic acids) is that the former contain 20 to 40 times more S than a typical fulvic acid extracted from natural soil organic matter (Table 3). Two of the SS fulvic acids also had somewhat less oxygen than a typical SOM fulvic acid. The principal conclusions to be drawn from these data are that SS fulvic acids can show significant variability in their chemical compositions, but that S- containing components should exert an important influence on their chemistry in any case. Ultraviolet and visible spectra of the SS fulvic acids suggested thatthey consist of highly aromatic and condensed struc- tures, somewhat comparable with those in SOM humic acids. The aromatic nature of the SS fulvic acids could be inferred indirectly from their X-ray diffraction patterns, which yielded a y -band cen- tered more nearly at 4 than 5A, thus implying a relatively small content of ali- phatic components. Partitioning of Metals after Sludge Incorporation into Soil At present, the chemical mechanisms by which potentially harmful trace metals in sewage sludges become soluble spe- cies after the incorporation of sludge into soil are not well defined. However, data obtained in this study show that the Table 1. Analytical Data for a Saturation Extract of an Altamont Soil Amended with Sewage Sludge and CdSOS Component pCf Component Ca K Na Fe(lll) Mn(ll) Cu(ll) Cd Zn CO so. Cl 2.07 3.70 3.00 4.75 4.70 5.72 5.85 5.13 2.70 2.70 2.28 P04 N03 Citrate^ Salicylate J Phthalate J Arginine $ Ornthine t Lysine \ Valine J Maleate J Benzylsulfonate | 4.00 2.77 4./4 4.27 3.97 4.49 4.36 4.36 4.36 3.97 4.27 » pH=6.30. f pC = - log [ ], where [ ] refers to a molar concentration. t Collectively considered organics in Table 2. ------- Table 2. Output from GEOCHEM Giving Primary Distribution of Metals and Ligands for Sludge-Amended Altamont Soil Free Metal or Ligand Ca K Na Fe(lll) Mn Cuflll Cd Zn C°3 so3. Cl P04 NO3 Organics Metal or Ligand 88.6 99.0 98.8 77.8 0.3 62.2 69.9 69.8 98.4 93.2 17.9 Bound With H 96.6 77.2 58.2 Ca 2.6 29.6 1.6 22.3 0.8 21.3 Na 0.5 0.1 2.2 Fe(lll) 0.5 0.02 Mn Cu(ll) Zn _ _ _ _ _ _ 0.1 _ 0.3 - - _ _ _ 0.3 0.01 C03 0.6 0.1 20.6 11.9 0.1 1.7 18.5 so. 6.9 0.6 0.9 6.1 4.9 6.9 Cl 1.0 0.2 17.2 0.1 P04 0.3 1.4 0.5 0.7 N03 0.6 0.3 0.1 1.4 1.1 Organics 0.19 79.3 2.2 99.2 12.4 3.7 Table 3. Ultimate Analyses* of SS Fulvic Acids and SOM Fulvic Acids Fulvic Acid Component C H N ot S Sludge A 30.3 ± f.7f 5.20 ±0.76 1.77 ±0.18 51.1 11. 56 ±0.95 SS Fulvic Acid Sludge B1 45.2 ± 3.0 7.31 ± 0.28 3.06 ± 0.06 40.0 4.46 ± 0.65 Sludge C 46.95 ± 0.84 7.21 ± 0.05 3.67 ± 0.14 35.7 8.44 ± 0.42 SOM Fulvic Acids 40-55 1.5 - 7.0 0.7 - 3.5 39-50 0.1 -3.6 * Percentage, water- and ash-free basis. f Mean deviation computed on data for three different samples of fulvic acid. t Calculated by difference: %O-100-% CC+W+/V+S/ organic constituents of sludges play a significant role in the speciation of metals such as Cd, Cu, Ni, Pb, and Zn. The degree to which soluble complexes between these metals and sludge- derived organic compounds actually form in a soil solution depends on the func- tional group character and the water sol- ubility of the organics, as well as on the number and stability of other possible trace metal compounds and surface phases. The association of five trace metals with the organic and inorganic fractions was studied by eluting sludge-FA from Sephadex G-10 columns. An elution diagram for a metal-SS ful- vic acid solution is shown in Figure 1 as an example of these studies. The relative concentrations of fulvic acid, Ni, and Cl vary with elution volume, and relation- ships between these concentrations can be used to infer something about metal complexes present. Results showed that the partition of metals among organic and inorganic forms differs among sludges. In general, Cu and Fe(lll) show high preference for organic complexa- tion, while Cd, Ni, and Zn tend to form inorganic complexes preferentially. These kinds of differences reflect clearly the fact that the complicated, hetero- geneous nature of sludge-derived fulvic acid makes for a detailed scheme of metal binding that will depend on the exact molecular composition of a given sludge. Incubation Studies to Evaluate Changes in Soils after Sludge Amendment A number of factors such as aeration and soil characteristics are known to affect the decomposition of organic materials and the fate of the constituents of the organic materials. For a fuller eva- luation of the changes that take place when sewage sludge is added to soils, a series of incubation experiments was carried out. In the first study, three soils were amended with Ontario sludge. Re- covery of added trace metals by HNO , DTPA (diethylenetriaminepentaacetate), and HOAc (acetic acid) was significantly affected by water content, sample prepa- ration, and soil. Application rate was gen- erally not a significant factor. This work demonstrated that for soil amended with Ontario sludge, metal extractability varied with time, water content, sample preparation, soil, and extractant. However, these results were only for one sludge. To determine the effects of these treatments more fully, additional studies were conducted using seven sludges representing a range of characteristics. The factors and levels used in this study were sludges (7), incubation water contents (1/3 bar and saturated), soils (Baywood, Delhi, and Calhi), sample preparation (wet and air-dried), times (7), and extractants (HNO3), DTPA, and sat- uration extracts). A constant rate of 100 T/ha was used, because in the previous study, rate did not affect the recovery of metals by HN03 or DTPA. The mean re- covery of the trace metals (Cd, Cu, Ni, and Zn) added to soils in sludges by extraction with 4/V HNO, was essentially complete (>90%). On the average, metal recovery by DTPA was 38%less than half that of HNO3. The recovery of metals by DTPA varied from 34% for Pb to 43% for Zn. Incubation time (1 to 52 weeks) had essentially no effect on the amounts of metals extracted from the sludge- amended soils by DTPA. Metal extracta- bility was significantly less for samples incubated at saturation than for samples incubated at 1/3 bar water content. Saturation extracts of sludge-amended soils incubated for 50 weeks were ana- lyzed for cationic and anionic composi- tion. These data were analyzed using GEOCHEM to predict the forms of the trace metals in the soil solutions. Data for Cd are used to illustrate the results (Table 4). Cadmium was found to be present in the extracts principally as free ionic forms, with smaller percentages present as organically complexed. For samples incubated at 1 /3 bar, the only real excep- tion to this result was for the LA liquid sludge, in which inorganically complexed ------- 120 150 200 Elution Volume, ml 250 300 Figure 1. Elution diagrams for Ni in metal-fulvic acid solutions Cd accounted for the greatest proportion of Cd in the extracts. Incubation at satu- ration resulted in a shift in form from free ionic to organically complexed. This result occurred in every case where Cd was detectable in the saturation extracts of the samples incubated at saturation. Fractionation Procedure A fractionation procedure was devel- oped and used on sludge-amended soils to estimate the proportion of selected trace metals in various forms. The proce- dure consisted of extraction with KN03 (exchangeable) followed by distilled- deionized water (adsorbed), 0.5 M NaOH (organically bound), Na, EDTA (carbon- ates), and HNO3 (sulfide residual). The use of distilled water to remove adsorbed metals was evaluated with freshly pre- pared Fe and Al hydrous oxide gels (Table 5). Close to 100% recoveries of sorbed metals were obtained for metals (Cd, Ni, and Zn) having the least affinity for adsorption by the hydrous oxides. The lowest recoveries (approximately 80%) were observed for Cu, which had the greatest affinity for both Fe and Al gels. Recovery of Pb from the Fe gel was also low. Recovery for each metal was virtu- ally independent of the amount of adsorbed metal present at the onset of the extraction. These data suggested that the use of three sequential deionized water washings forms a reliable basisfor the extraction of adsorbed metals from sludges and sludge-amended soils. The standard procedure for removal of fulvic and humic acid fractions of organic matter is by extraction with NaOH. This extractant was therefore used to evalu- ate the proportion of organic-bound metals. Data for the extraction of metals from pure carbonates and sulfides indi- cated that 0.5 M NaOH should be reason- ably specific to the extraction of organic-bound metals (Table 6). Only with Pb and Zn were significant quanti- ties of each metal carbonate brought into solution by 0.5 M NaOH. The use of a single 0.05 M NaEDTA extraction was generally specific to carbonate-bound metals. Only small proportions of the metals were extracted from the metal sulfides using this reagent. Essentially complete recovery of metals from the cor- responding metal sulfides was achieved only after extraction with 4/V HNO at 80°C (Table 6). Table 4. Speciation of Cd in Saturation Extracts of Sludge-Amended Soils After 50 Weeks of Incubation (percent of total) Incubation Water Content Soil Species 1/3 Bar Bay wood F* 1 O Delhi F 1 0 Calhi F 1 0 Saturation Baywood F 1 0 Delhi F 1 O Calhi F 1 O Rialto 58.6 22.0 19.1 75.2 20.6 4.2 75.7 19.9 4.1 -t 33.3 34.9 31.7 39.1 17.0 40.9 LA Compost 68.1 29.8 2.0 72.0 26.0 1.7 71.4 25.9 2.6 41.8 14.8 43.2 Colton 52.0 23.5 24.3 74.0 19.7 6.2 70.2 18.9 10.7 33.4 40.5 25.6 Sludge Ontario 70.9 26.0 3.1 71.0 22.2 6.6 68.4 21.0 10.3 24.8 6.2 69.0 30.3 30.3 39.2 Escondido 54.4 36.5 9.0 67.8 29.5 3.0 64.2 31.7 3.9 41.7 23.4 34.8 42.4 26.4 31.1 LA Liquid 37.6 58.0 4.3 42.9 54.9 2.0 41.2 52.5 5.3 24.8 60.0 14.9 31.9 43.1 25.0 26.3 58.0 15.6 Kokomo 73.3 23.1 3.5 79.7 15.9 4.2 72.2 23.4 4.4 61.1 29.3 9.6 52.7 31.8 15.4 * F = free ionic; I = inorganically complexed; O = organically complexed. f Indicates metal concentration below detection limit. ------- Table S. Water Extraction (3X) of Metals Absorbed by Fe and Al Hydrous Oxide Gels Metal FeGel AIGel Sorbed Extracted Percent Recovery Sorbed Extracted Percent Recovery (tig/0.05 g gel) (ug/0.05 g gel) Cd Cu Ni Pb Zn 21.2 40.0 50.0 99. 4 25.2 37.6 50.0 100.0 38.0 64.6 21.8 41.2 41.0 93.5 26.2 38.0 41.4 80.6 36.4 63.1 102 103 82 94 104 101 83 81 96 98 6.2 16.0 48.8 97.6 14.6 28.0 44.0 89.6 38.6 73.0 6.5 16.6 36.1 81.0 14.3 24.4 45.5 91.6 38.8 69.8 105 104 74 83 98 87 103 102 101 96 Table 6. Extractant Percent Recovery of Metals During NaOH. EDTA. and HNOa Extraction of 0.03 g of Pure Metal Carbonates and Sulfides Metal Extracted Cd Cu Ni Pb Zn 0.5 M NaOH 0.05 M EDTA 0.5 MNaOH 0.05 M EDTA 2 N HN03 4NHN03.80°C 0 99 0 4.3 16.3 98.7 1.1 100 1.3 3.3 0 47.5 Carbonate 0 1OO Sulfide 0 15 46.2 100 38.5 85.1 8.8 20 44.9 100 33.2 100 0 0 94.0 1OO amounts of sulfide-bound Cd relative to the amount of Cd in this form after incu- bation at saturation. Furthermore, for the Calhi and Delhi soils, the amounts of sulfide-bound Cd after 1/3-bar incuba- tion were usually lower than those observed for the whole sludges. In the case of the Kokomo sludge, which had a very high (81 %) proportion of Cd in the carbonate fraction, there was tendency for Cd to be redistributed into the organic-bound and adsorbed frac- tions during incubation. In fact, adsorbed Cd at detectable levels was observed only in the case of incubation with Kokomo sludge. No detectable, exchangeable Cd was observed after incubation of any of the sludge-amended soils, reflecting the absence of this form in the whole sludges. Where applied sludges resulted in Cd contents above 2 (ig/g of soil, a of soil, a significant proportion (0.1% to 2.2%) of the total Cd appeared in solution after incubation at 1/3 bar. The full report was submitted in fulfill- ment of Grant No. R-804516 by the Uni- versity of California under the sponsorship of the U.S. Environmental Protection Agency. Again, the data for Cd are used as an example of the results of applying this fractionation procedure to sludge- amended soils (Table 7). When the study used sludges with a low Cd content (Rialto, LA-composted, and Cotton), which resulted in Cd contents of less 2 /ug/g for the sludge-amended soils, Cd appeared almost exclusively in the organic and carbonate forms for each soil amended with these sludges. The ten- dency was for the distribution to be weighted in favor of the carbonate form (14 out of 18 cases) despite the greater proportion of Cd in the organic form in two out of three sludges. Though a small proportion of the total Cd (7.3% to 17.1%) was present in the Rialto, LA-composted, and Colton sludges in the sulfide form, only in one instance was Cd observed in this form after incubation of the sludges with the soils. In contrast, the absence of Cd in exchangeable and adsorbed fractions in the whole sludges was maintained after incubation. Only in one instance (after the incubation of Rialto sludge with the Baywood soil at 1/3 bar) was any sub- stantial quantity of Cd measured in the soil solution. Only minor variations occurred in the distribution of Cd for these sludges as a result of the different incubation conditions employed. All but one of the remaining sludges with higher Cd contents contained the major portion of the total Cd in the carbo- nate form after incubation with soil. In the majority of the sludge-amended soils, more than 60% of the Cd was in the car- bonate form after incubation at both 1/3 bar and saturation. In only one case did the proportion of Cd in this fraction fall below 50%. In the Ontario and Escondido sludges, the proportion of Cd in the car- bonate form increased after incubation with each soil at both moisture contents. A general increase also occurred in the amount of Cd in the sulfide fraction rela- tive to the whole sludge following incu- bation under saturated conditions. The increases in the carbonate and sulfide- bound fractions were made at the expense of organic-bound Cd. Incubation at 1 /3 bar resulted in appreciably lower ------- Table 7. Distribution of Cd in Sludges and Sludge-Amended Soils as Affected by Incubation Water Content (% of total) Sludge or Soil and Form of Cd Riatto LA-Composted Cotton Ontario Escondido LA-Liquid Kokomo Sludge: Solution Exchangeable Adsorbed Organic Carbonate Sulfide Baywood: Solution Exchangeable Adsorbed Organic Carbonate Sulfide Delhi: Solution Exchangeable Adsorbed Organic Carbonate Sulfide Calhi: Solution Exchangeable Adsorbed Organic Carbonate Sulfide * = None detected. f = Saturated, 1/3 = ND* ND ND NO NO ND ND 0 0 0 0 0 1.1 1.1 0000 0 66.0 32.0 65.5 36.6 32.8 7.0 26.4 50.9 34.5 51.6 53.3 81.0 7.6 st 0 0 0 44.6 55.4 0 0 0 0 53.3 46.7 0 0 0 0 54.3 45.7 0 1/3 bar saturation. 17.1 7.3 11.8 12.8 10.9 1/3 S 1/3 S 1/3 S 1/3 S 1/3 S 1/3 S 1/3 0.2 0 0 0 0 0 1.8 O 1.4 0 2.2 0 0.1 OOOOOOOOOOOOO 0 0 0 0 0 0 0 O 0 0 0 5.5 3.6 47.7 40.9 16.5 46.8 46.4 15.0 14.8 12.6 20.9 18.4 20.7 19.9 12.0 52.1 59.1 65.3 53.2 53.1 60.0 61.2 59.6 64.2 62.2 57.5 58.2 75.2 0 0 18.2 0 0 25.0 22.0 27.8 13.5 19.4 19.6 16.4 9.1 0 0 0 0 0 0 0.9 O 1.1 0 1.1 0 O.I OOOOOOOOOOOOO 0 0 0 0 0 0 0 O 0 0 0 6.4 3.9 43.3 39.6 41.8 44.6 43.8 26.5 43.1 23.3 34.5 22.3 56.9 23.2 23.7 56.5 60.4 55.6 55.6 56.2 60.1 51.4 54.7 59.6 60.1 42.0 60.3 70.7 00000 13.4 4.6 22.0 4.8 17.6 0 10.1 1.6 0 0 0 0 0 0 0.5 O 0.3 0 1.0 0 O.2 OOOOOOOOOOOOO 0 0 0 0 0 0 0 0 0 0 0 1.8 2.7 50.0 42.6 44.0 53.5 41.3 14.6 39.0 11.5 19.8 11.7 22.7 5.1 3.8 50.0 57.4 56.0 46.5 58.7 70.6 60.5 61.4 79.9 77.8 65.2 76.2 92.0 00000 14.8 0 27.1 0 10.5 12.1 16.9 1.3 L J. Lund, G. Sposito, and A. L. Page are with University of California, Riverside, CA 92521. J. A. Ryan is the EPA Project Officer (see below). The complete report, entitled "Determination and Prediction of Chemical Forms of Trace Metals in Sewage Sludge and Sludge-Amended Soils," (Order No. PB 85-197 1 19/AS; Cost $20.50, subject to change) will be available only from: National Technical Information Service 5285 Port Royal Road Springfield, VA 22161 Telephone: 703-487-4650 The EPA Project Officer can be contacted at: Water Engineering Research Laboratory U.S. Environmental Protection Agency Cincinnati, OH 45268 U. S. GOVERNMENT PRINTING OFFICE:1985/559-l 11/20681 ------- United States Environmental Protection Agency Center for Environmental Research Information Cincinnati OH 45268 BULK RATE POSTAGE & FEES PAI EPA PERMIT No. G-35 Official Business Penalty for Private Use $300 EPA/600/S2-85/053 0169J64 US SPA REGION V LIBRARY 230 S DEARSO^N ST. 60604 4 ------- |