Transport of Pb and Zn by carboxylate complexes in basinal ore fluids and related petroleum-field brines at 100°C: the influence of pH and oxygen fugacity

It is well established through field observations, experiments, and chemical models that oxidation (redox) state and pH exert a strong influence on the speciation of dissolved components and the solubility of minerals in hydrothermal fluids. log –pH diagrams were used to depict the influence of oxygen fugacity and pH on monocarboxylate- and dicarboxylate-transport of Pb and Zn in low-temperature (100°C) hydrothermal ore fluids that are related to diagenetic processes in deep sedimentary basins, and allow a first-order comparison of Pb and Zn transport among proposed model fluids for Mississippi Valley-type (MVT) and red-bed related base metal (RBRBM) deposits in terms of their approximate pH and conditions. To construct these diagrams, total Pb and Zn concentrations and Pb and Zn speciation were calculated as a function of log and pH for a composite ore-brine with concentrations of major elements, total sulfur, and total carbonate that approximate the composition of MVT and RBRBM model ore fluids and modern basinal brines. In addition to acetate and malonate complexation, complexes involving the ligands Cl-, HS-, H2S, and OH- were included in the model of calculated total metal concentration and metal speciation. Also, in the model, Zn and Pb are competing with the common-rock forming metals Ca, Mg, Na, Fe, and Al for the same ligands. Calculated total Pb concentration and calculated total Zn concentration are constrained by galena and sphalerite solubility, respectively. Isopleths, in log –pH space, of the concentration of Pb and concentration of Zn in carboxylate (acetate + malonate) complexes illustrate that the oxidized model fluids of T. H. Giordano (in Organic Acids in Geological Processes, ed. E. D. Pittman and M. D. Lewan, Springer-Verlag, New York, 1994, pp. 319–354) and G. M. Anderson (Econ. Geol., 1975, 70, 937–942) are capable of transporting sufficient amounts of Pb (up to 10 ppm) and Zn (up to 100 ppm) in the form of carboxylate complexes to form economic deposits of these metals. On the other hand, the reduced ore fluid models of D. A. Sverjensky (Econ. Geol., 1984, 79, 23–37) and T. H. Giordano and H. L. Barnes (Econ. Geol., 1981, 76, 2200–2211) can at best transport amounts of Pb and Zn, as carboxylate complexes, that are many orders of magnitude below the 1 to 10 ppm minimum required to form economic deposits. Lead and zinc speciation (mol% of total Pb or Zn) in the model ore fluid was calculated at specific log –pH conditions along the 100, 0.01, and 0.001 ppm total Pb and total Zn isopleths. Along the 100 ppm isopleth conditions are oxidized (∑SO4 >> ∑H2S) with Pb and Zn predominantly in the form of chloride complexes under acid to mildly alkaline conditions (pH from 3 to approximately 7.5), while hydroxide complexes dominate Pb and Zn speciation under more alkaline conditions. Sulfide complexes are insignificant under these oxidized conditions. For more reduced conditions along the 0.01 and 0.001 ppm isopleths chloride complexes dominate Pb and Zn speciation in the SO42- field and near the SO42--reduced sulfur boundary from pH = 4 to approximately 7.5, while hydroxide complexes dominate Pb and Zn speciation under alkaline conditions above pH = 7.5 in the SO42- field. In the most reduced fluids (∑H2S >> ∑SO4) along the 0.01 and 0.001 isopleths, sulfide complexes account for almost 100% of the Pb and Zn in the model fluid. Acetate (monocarboxylate) complexation is significant only under conditions of chloride and hydroxide complex dominance and its effect is maximized in the pH range 5 to 7, where it complexes 2 to 2.6% of the total Pb and 1 to 1.25% of the total Zn. Malonate (dicarboxylate) complexes are insignificant along all isopleths. The speciation results from this study show that deep formation waters characterized by temperatures near 100°C, high oxidation states and ∑H2S < 0.03 mg L-1 ( < 10-6), high chlorinities (~ 100000 mg L-1), and high but reasonable concentrations of carboxylate anions can mobilize up to 3% of the total Pb and up to 1.3% of the total Zn as carboxylate complexes. Furthermore, these percentages, under the most favorable conditions, correspond to approximately 1 to 100 ppm of these metals in solution; concentrations that are adequate to form economic deposits of these metals. However, the field evidence suggests that all of these optimum conditions for carboxylate complexation are rarely met at the same time. A comparison of the composite ore fluid compositions from this study and modern brine data shows that the ore brines, corresponding to log –pH conditions based on the Anderson (1975) and Giordano (1994) model fluids, are similar in many respects to modern, high trace-metal petroleum-field brines. The principal differences between modern high trace-metal brines and the composite ore fluids of Anderson (1975) and Giordano (1994) relate to their carboxylate anion content. The reported concentrations of monocarboxylate anions (∑monocbx) and dicarboxylate anions (Edicbx) in high trace-metal petroleum-field brines (< 1 to 300 mg L-1 and < 1 mg L-1, respectively) are significantly lower than the concentrations assumed in the modelled brines of this study (∑monocbx = 7 700 mg L-1 and ∑dicbx = 300 mg L-1). There are also major differences in the corresponding total chloride to carboxylate ratio (∑mCl/∑mcbx) and monocarboxylate to dicarboxylate ratio (∑mmonocbx/∑mdicbx). Modern high trace-metal brines have much higher ∑mCl/∑mcbx values and, therefore, the contribution of carboxylate complexes to the total Pb and Zn content in these modern brines is likely to be significantly less than the 1 to 3 percent for the composite ore fluids of Anderson (1975) and Giordano (1994). The composite ore-brine based on the Giordano and Barnes (1981) MVT ore fluid is comparable to the high salinity (> 170 000 mg L-1 TDS) subset of modern brines characterized by low trace-metal content and high total reduced sulfur (∑H2S). A comparison of the Sverjensky (1984) composite ore-brine with modern petroleum-field brines in terms of ∑H2S and Zn content, reveals that this ore fluid corresponds to a "border-type" brine, between modern high trace-metal brines and those with low trace-metal content and high ∑H2S. A brine of this type is characterized by values of ∑H2S, ∑Zn, and/or ∑Pb within or near the 1 to 10 mg L-1 range. Based on brine-composition data from numerous references cited in this paper, border-type brines do exist but are rare. The model results and field evidence presented in this study are consistent with other chemical simulation studies of carboxylate complexation in modern petroleum-field brines. Thus, it appears that carboxylate complexation plays a minor, if not insignificant, role as a transport mechanism for Pb and Zn in high salinity Na–Cl and Na–Ca–Cl basinal brines and related ore fluids.


Introduction
It is well established through field observations, experiments, and chemical models that oxidation (redox) state and pH exert a strong influence on the speciation of dissolved components and the solubility of minerals in hydrothermal fluids. These parameters are also known to act as major constraints on the stability of specific minerals and mineral assemblages in hydrothermal systems and to control sulfur and carbon isotope systematics in these systems. Of particular interest to this study is the influence of pH and redox state on the speciation of metals, ligands, and metal complexes. For example, in the model presented below, zinc monoacetate complexation, as represented by the reaction Zn 2z zCH 3 COO { ?Zn(CH 3 COO) z is directly influenced by pH through the control of acetate activity by protonation, CH 3  The oxygen fugacity indirectly affects reaction (1) because HS 2 in reaction (3) is constrained, assuming constant total dissolved sulfur, constant pH and equilibrium, by the SO 4 22 to HS 2 ratio of the fluid through the reaction, HS { z2O 2 g ð Þ?SO 2{ 4 zH z and corresponding mass action condition at equilibrium: In eqn. (5), the bracketed species represent the activity of the enclosed species and f O 2 is the fugacity of oxygen gas in equilibrium with the fluid. It has been suggested (cf. Shock, [1][2][3] Helgeson et al. 4 ) that dissolved acetate is coupled to dissolved carbonate in basinal fluids via metastable redox equilibria including the oxidation of acetate by molecular oxygen, CH 3 COO { z2O 2 g ð Þ?2HCO { 3 zH z This type of metastable link between carboxylate species and carbonate is not assumed in the model described below. In this model total acetate and total malonate concentrations are independent of total carbonate.
Oxygen fugacity (or activity) is the most commonly used redox variable for describing the oxidation state of hydrothermal systems. Its common use stems from the convenience of writing many hydrothermal reactions in terms of oxygen gas as a reactant or product and the generally close approach to (or actual attainment of) stable or metastable equilibrium under hydrothermal conditions. Consequently, log f O 2 -pH (or log a O 2 -pH) diagrams have become the most widely used approach to illustrate the combined effects of pH and redox state in the control of hydrothermal aqueous speciation and related phase equilibria (Barnes and Kullerud, 5 Barnes and Czamanske, 6 Ripley and Ohmoto, 7 Henley et al., 8 McPhail, 9 Wood 10 ). These diagrams can also be employed to graphically represent aqueous carbon and sulfur isotope systematics (Ohmoto, 11 Ohmoto and Rye, 12 Zhang 13 ) and organic matter composition (Helgeson et al., 4 Gize 14 ) under hydrothermal conditions. In addition to log f O 2 -pH diagrams, other graphical depictions of hydrothermal redox dependency are used, including: (a) plots in log (SSO 4 /SH 2 S)-pH and log (SSO 4 / SH 2 S)-T space (Ripley and Ohmoto,7,15 Ahmad et al., 16 Ohmoto and Goldhaber 17 ), (b) log f S 2 -T and log f S 2 -1/T diagrams (Ripley and Ohmoto, 15 Ohmoto et al., 18 Henley et al. 8 ), (c) log f O 2 -T, log f O 2 2 1/T, and log( f H 2 O /f H 2 )-T diagrams (Ripley and Ohmoto, 7 Ohmoto and Rye, 12 Eugster, 19 Helgeson et al., 4 Ohmoto and Goldhaber, 17 Giggenbach, 20 Cooke et al., 21 Wood and Samson 22 ), (d) plots in Eh (pe)-pH space (Rose,23,24 Wintsch, 25 Henley et al., 8 Pound et al., 26 Pourbaix and Pourbaix 27 ), (e) log f H 2 -T diagrams (Shock 3 ), and (f) the activity diagrams log a i -log a CO 2 (Helgeson et al. 4 ), log f O 2 2 log a H 2 S (Kettler et al. 28 ), log a SO 4 -log a H 2 S (Ohmoto et al. 18 ), and log f O 2 -log f S 2 (Barton et al., 29 Casadevall and Ohmoto, 30 Ohmoto et al., 18 Henley et al., 8 Eugster, 19 Sverjensky, 31,32 Ahmad et al. 16 ).
Over the past forty years these diagrams, particularly log f O 2 -pH diagrams, have been used to investigate transport and deposition mechanism of metals in hydrothermal ore fluids. In the majority of these studies, the principal focus is on inorganic complexation mechanisms that mobilize metals in solution and associated processes that cause complex destabilization and concomitant ore deposition. Although a large number of inorganic ligands are known to be complexing agents in hydrothermal ore fluids, inorganic transport of Hg, Pb, Zn, Cu, REE, and the precious metals appears to be dominated by the ligands Cl 2 , HS 2 , NH 3 , and OH 2 (Seward and Barnes 33 ). Thus, it is not surprising, that ore transport and deposition processes have been commonly evaluated using concentration contours of metal complexes involving these ligands in log f O 2 -pH space. This approach has been particularly fruitful in developing or evaluating genetic models for a wide variety of hydrothermal deposit-types, including mercury deposits (Wells and Ghiorso, 34 Barnes and Seward, 35 Fein and Williams-Jones 36 ), Mississippi Valley-type (MVT) deposits (Anderson,37,38 Giordano and Barnes, 39 Barnes, 40 Sverjensky 41 ), red-bed related base metal (RBRBM) deposits (Sverjensky 31,32 ), McArthur-type and Selwyn-type sedex deposits (Cooke et al. 21 ), porphyry copper deposits (Crerar and Barnes 42 ), gold and silver deposits (Barton et al., 29 Casadevall and Ohmoto, 30 Ahmad et al., 16 Gammons and Williams-Jones 43 ), as well as deposits of the more exotic metals including the platinum group elements (Mountain and Wood, 44 Wood et al., 45 Gammons et al. 46 ).
Although inorganic transport mechanisms adequately account for the genesis of many medium-and high-temperature hydrothermal deposits (e.g., porphyry Cu-Mo-W deposits; skarn deposits; and epithermal base metal, Hg, Au, and Ag deposits) it is less certain that such mechanisms can fully account for most lower-temperature hydrothermal deposits formed in the 50 to 200 uC temperature range, especially those deposits intimately associated with organic matter and formed by basinal diagenetic processes (e.g., MVT and RBRBM deposits). As a means of addressing metal transport in these lower-temperature hydrothermal ore systems, alternative complexing mechanisms have been tested including organic transport, involving transport of metals via organic ligands in the form of metal-organic complexes (cf. Giordano and Barnes, 39 Giordano, 47 Drummond and Palmer, 48 Manning, 49 Hennet et al., 50 Giordano and Kharaka, 51 Sicree and Barnes, 52 Giordano 53 ) and via pressure-sensitive species such as metalthiocarbonate complexes (Hennet et al. 54 ). Furthermore, many epithermal precious metal and mercury deposits contain organic matter with paragenetic relationships closely related to the timing of ore-formation. This observation has sparked speculation that a significant role for organic complexing may supplement or dominate apparently viable inorganic mechanisms of metal transport in these systems (cf. Fein and Williams-Jones, 36 Leventhal and Giordano 55 ).
Giordano 53 gives an overview of those investigations conducted during the past 25 years and designed to evaluate organic transport in hydrothermal ore systems. The majority of these investigations (experimental and theoretical) have focused on Pb and Zn transport in MVT and RBRBM deposits and deep sedimentary basin brines, generally thought to be ore fluids for these deposits. In all of these studies Pb and Zn concentrations and speciation are either measured or calculated for very restricted fluid compositions in terms of f O 2 , pH, and other solute activities. This approach is fine for evaluating the significance of complexing mechanisms under specific orefluid conditions. More useful, however, would be an evaluation based on metal concentration contours in log f O 2 -pH space. This is the method employed in those studies mentioned above in which log f O 2 -pH diagrams were used to develop transport and deposition models based on inorganic complexing. This approach allows a broad-based evaluation of a particular complexing mechanism over a wider range of possible f O 2 -pH conditions and the simultaneous evaluation of deposition processes as a function of f O 2 and pH. The author is not aware of any published studies that take this approach in evaluating organic-transport mechanisms of base or precious metals in low-to moderate-temperature hydrothermal ore fluids. However, organic transport of mercury has been evaluated in log f O 2 -pH space for hydrothermal ore-forming systems (Wells and Ghiorso, 34 Fein and Williams-Jones 36 ). These authors found that methyl mercury as well as acetate and oxalate complexes of mercury do not contribute significantly to Hg transport in the formation of hydrothermal mercury deposits.
The purpose of this paper is to present a preliminary evaluation of the influence of oxygen fugacity and pH on monocarboxylate-and dicarboxylate-transport of Pb and Zn in low-temperature (100 uC) hydrothermal ore fluids that are related to diagenetic processes in deep sedimentary basins. First a description of the chemical model used in this study will be presented. This is followed by a presentation of model results illustrating Pb and Zn solubility and speciation and a discussion of these results in terms of fluid parameters and the compositional characteristics of modern petroleum-field brines. Finally, some conclusions will be presented regarding (a) the similarities between modern petroleum-field brines and ore fluid models of MVT and RBRBM deposits and (b) the systematics of Zn and Pb transport by carboxylate complexation in these basinal brines.

Chemical model
Model ore fluids Over the past half century, geochemical studies of sedimentary basins, petroleum fields, and ore deposits related to these features have resulted in the general consensus that ore fluids responsible for a wide range of low-to medium-temperature (50 to 200 uC) hydrothermal deposits were in fact Na-Cl/Na-Ca-Cl basinal brines (Beales and Jackson, 56 Carpenter et al., 57 Hitchon, 58 Hanor,59,60,61,62 Anderson and McQueen, 63 Lyndon, 64 Sverjensky,31,41 Sangster, 65 Force et al., 66 Heijlen et al. 67 ). These fluids became enriched in metals as a result of diagenetic processes and migrated from hot, deep basin interiors to more peripheral and structurally active parts of the basin where metals were deposited in host rocks to form a variety of ore deposits ranging from purely epigenetic carbonate-hosted and siliciclastic-hosted deposits to those with a significant sedimentary exhalative component (Bjorlykke and Sangster, 68 Gustafson and Williams, 69 Sangster, 65 Cooke et al. 21 ). Within this wide range of deposit-types, two important categories of epigenetic ores are carbonate-hosed MVT deposits and RBRBM deposits, base metal ores associated with siliciclastic continental/marginal marine host rocks. Both of these ore-types have been extensively studied in terms of their geology and genesis (cf. Kisvarsanyi et al., 70 Sverjensky, 71 Haynes and Bloom, 72 Boyle et al., 73 Leach and Goldhaber, 74 Sicree and Barnes 52 ). Detailed ore-fluid models have also been proposed for each of these deposit-types and with respect to the ore fluids in these models, there is general agreement among researchers regarding temperature of transport and deposition, total sulfur and total carbonate concentration, and major element composition. For RBRBM deposits (Rose,23,24 Sverjensky,31,32 Branam and Ripley 75 ), there is also general agreement that the metal-transporting fluids were oxidized, very low in reduced sulfur (reduced sulfur is here defined as Sm H 2 S~mH 2 S 1 m HS 2 ), and slightly acidic. Parameters for a composite RBRBM ore fluid (Giordano 76 ) are shown in Table 1 (Anderson 37,38,77 ), similar to RBRBM ore fluids and many sulfide-poor petroleum-field brines.

Carboxylate ligands
It is well known that some petroleum-related brines from deep sedimentary basins contain measurable quantities (x to xxx mg L 21 ) of base metals (Pb, Zn, Cu) and that many petroleum-field brines contain measurable concentrations (x to xxxx mg L 21 ) of discrete, polar organic molecules (particularly carboxylic acid anions) capable of complexing Pb, Zn, and Cu, as well as the more common rock-forming metals mobilized in basinal diagenetic environments. The nature of Pb-Zn rich brines from sedimentary basins will be summarized in the Discussion section of this paper. Lundegard and Kharaka, 78 Giordano and Kharaka, 51 and Kharaka et al. 79 summarize the available data concerning the nature and distribution of discrete organic ligands in diagenetic environments. Based on a critical evaluation of the available analyses of dissolved organic species in petroleum-related brines, these authors conclude that the only clearly identified group of significant organic ligands in oil-and gas-field brines are short-chain aliphatic carboxylate anions. Furthermore, Kharaka et al. 79 give recommended maximum concentrations for the most abundant carboxylate anions found in these fluids ( Table 2). Acetate is the dominant monocarboxylate anion in brines with temperatures greater than 85 uC followed by propionate, butyrate, and valerate, while the dominant dicarboxylate species are succinate, glutarate, and malonate. Kharaka et al. 79 do not rule out the possibility of other important groups of organic ligands but the observational data are not available to support their importance in modern basinal brines and ancient ore fluids. In the model ore fluid evaluated in this paper, acetate with a concentration of 7 700 mg L 21 (0.13 m) represents the four dominant monocarboxylate ligands and malonate with a concentration of 300 mg L 21 (0.0029 m) is used to represent the three dominant dicarboxylate ligands ( Table 2). This approach was taken because there is a lack of experimental thermodynamic data for complexation equilibria involving the ligands propionate, butyrate, valerate, succinate, and glutarate. Based on experimental data at 25 uC (Smith et al. 80 ) and predicted constants at elevated temperatures (Shock and Koretsky, 81 Sverjensky et al., 82 Prapaipong et al. 83 ), it appears that malonate, succinate, and glutarate form metal complexes with similar stabilities and that metal acetate complexes are similar in terms of stability to corresponding complexes formed by propionate, butyrate, and valerate. Metal dicarboxylate complexes are generally one to two orders of magnitude stronger than metal monocarboxylate complexes because of the chelate effect. Thus, for a given total carboxylate ligand content, metal carboxylate complexation increases with increasing proportion of dicarboxylate ligand species, or decreasing concentration ratio of total monocarboxylate to total dicarboxylate.

Composite ore fluid
In this paper a composite ore fluid (column 6 in Table 1) is tested for Pb and Zn transport at 100 uC as a function of pH and log f O 2 . Concentrations of the major elements, total sulfur, and total carbonate of this hypothetical ore fluid only approximate the proposed concentrations given in columns 2 through 5 of Table 1 for the four model ore-fluids. Saturation  Nevertheless, the concentrations of the major elements, total carbonate, and total sulfur of this composite ore fluid are sufficiently close to those of the other model fluids, as well as many petroleum-field brines, to allow a first-order comparison of Pb and Zn transport among the proposed model fluids in terms of their approximate pH and f O 2 conditions. The primary purpose of this paper is to evaluate the influence of these two parameters on Zn and Pb transport by monocarboxylate ligands represented by acetate complexation and dicarboxylate ligands represented by malonate complexation. The stability constants used in this study to describe malonate and acetate protonation and malonate and acetate complexation of Pb and Zn are given in Table 3. Reliable 100 uC constants are available for the protonation equilibria and acetate complexation of Pb and Zn. However, the author is not aware of reliable 100 uC data for malonate complexation of Pb and Zn and, therefore, the available 25 uC data are used. Although thermodynamic data are available for Cu 1 -acetate complexes between 50 and 150 uC (Liu et al. 91 ), the lack of thermodynamic data for Cu 1dicarboxylate complexes (Martell and Smith,92,93 Smith and Martell, 94 Smith et al. 80 ) precluded the possibility of adequately modeling Cu complexation in this study. It should be noted that Sverjensky's1984 model MVT-fluid 41 has a temperature of 125 uC and, therefore, the log f O 2 and pH boundaries defined by Sverjensky 41 were used in this study to estimate correlative log f O 2 and pH boundaries at 100 uC ( Fig. 1 and Table 1). Total Pb and Zn concentrations and Pb and Zn speciation were calculated as a function of log f O 2 and pH for the composite ore-brine described in Table 1. In addition to acetate and malonate complexation, complexes involving the ligands Cl 2 , HS 2 , H 2 S, and OH 2 were included in the model of calculated total metal concentration and metal speciation. Also, in the model calculations, Zn and Pb are competing with the common-rock forming metals Ca, Mg, Na, Fe, and Al for the same ligands. Calculated total Pb concentration and calculated total Zn concentration are constrained by galena and sphalerite solubility, respectively. It was assumed that the brine is saturated with respect to quartz and K-feldspar at all log f O 2 -pH conditions and that total Fe concentration is controlled by either pyrite, hematite or magnetite solubility, depending on the given f O 2 -pH condition. The stability fields for pyrite, hematite, and magnetite are shown in Fig

Isopleths
It is generally assumed that an ore fluid must transport, at a minimum, between 1 and 10 ppm of a base metal to form an ore deposit of that metal (Barnes and Czamanske, 6 Barnes,40,95 Anderson, 38 Roedder, 96 Ohmoto, 97 Seward and Barnes 33 ). Examination of the Pb isopleths (representing galena solubility) in Fig. 2 shows that sufficient lead in the model ore fluid can only be transported under log f O 2 -pH conditions corresponding to low total sulfide (Sm SO 4 & Sm H 2 S ); for example, the relatively oxidized RBRBM ore fluid of Giordano 76 and the MVT ore fluid of Anderson. 37 The MVT ore fluids of Sverjensky 41 and Giordano and Barnes 39 are capable of  84 (2) Extrapolation based on 24 to 85 uC data from Giordano 85  transporting maximum amounts of lead that are 2 to 5 orders of magnitude below the 1 to 10 ppm minimum. It should be noted that Sverjensky's 1984 model ore fluid 41 was developed for a temperature of 125 uC and that the model fluid could barely transport 1ppm Pb at that temperature. The total Zn isopleths (representing sphalerite solubility) shown in Fig. 3 illustrate model results similar to those for Pb. The ore fluids of Sverjensky 41 and Giordano and Barnes 39 are not capable of transporting sufficient Zn at 100 uC. However, the oxidized RBRBM ore fluid of Giordano 76 and the Anderson MVT ore fluid can clearly mobilize 100s of ppm Zn, several orders of magnitude greater than the minimum required to form economic deposits of Zn.
In Fig. 4 and 5, respectively, isopleths of the total concentration of Pb and total concentration of Zn in carboxylate (acetate 1 malonate) complexes are plotted for the model ore fluid. The 1 to 10 ppm Pb isopleths shown in Fig. 4 illustrate that the oxidized model fluids of Giordano 76 and Anderson 37 are capable of transporting sufficient amounts of Pb in the form of carboxylate complexes to form economic deposits of lead. On the other hand, the reduced ore fluid models of Sverjensky 41 and Giordano and Barnes 39 can at best transport amounts of Pb, as carboxylate complexes, that are many orders of magnitude below the 1 to 10 ppm minimum. Similar to Pb, the 1 and 10 ppm Zn isopleths in Fig. 5 coincide with the log f O 2 -pH conditions of the RBRBM and Anderson MVT ore fluids and demonstrate that these brines can transport sufficient Zn, as carboxylate complexes at 100 uC, to form economic deposits of Zn. However, the more reduced Sverjensky 41 and Giordano and Barnes 39 ore fluids are incapable of transporting sufficient Zn, as carboxylate complexes, to form economically viable deposits of Zn. The isopleths in Fig. 4 and 5 curve upward on the left side of the diagram because protonation of acetate and malonate becomes significant at pH below 5, thereby reducing the activities of these ligands in solution and    activities are constant at a given pH. Therefore, the concentration of individual carboxylate complexes is also constant; thus producing the vertical 0.0001 ppm isopleths in Fig. 4 and 5.

Speciation
Lead speciation (mol% of total Pb) in the model ore fluid was calculated at specific log f O 2 -pH conditions along the 100, 0.01, and 0.001 ppm total Pb isopleths shown in Fig. 2. These speciation results are presented in Table 4 and Fig. 6, Table 5 and Fig. 7, and Table 6 Table 4) conditions are oxidized and lead is predominantly in the form of chloride complexes under acid to mildly alkaline conditions (pH from 3 to approximately 7.5), while hydroxide complexes dominate lead speciation under more alkaline conditions (Table 4 and Fig. 6). Sulfide complexes are insignificant under these oxidized conditions. Acetate (monocarboxylate) complexation accounts for up to 2.6% of the total Pb in the pH range 5 to 7, while malonate (dicarboxylate) complexes are insignificant along the isopleth. For the 0.01 ppm isopleth ( Fig. 2 and 7, Table 5), chloride complexes dominate lead speciation in the SO 4 22 field and near the SO 4 22reduced sulfur boundary from pH~4 to approximately 7.5, while hydroxide complexes dominate lead speciation under alkaline conditions above pH~7.5 in the SO 4 22 field. In the most reduced fluids along the 0.01 isopleth, lead sulfide complexes account for almost 100% of the Pb in the model fluid. Acetate (monocarboxylate) complexation is significant only under conditions of chloride and hydroxide complex dominance and its effect is maximized in the pH range 5 to 7, where it complexes 2 to 2.6% of the total Pb. Malonate (dicarboxylate) complexes are insignificant along the 0.01 isopleth. The lead speciation systematics along the 0.001 isopleth ( Fig. 2 and 8, Table 6) are similar to those for the 0.01   Table 4, along with corresponding % of total Pb values for various forms of Pb, and begin at the right end of the 100 ppm isopleth with data set #5 (8, 250.20), followed by #4 (7, 249.95), #3 (5, 249.90), #2 (4, 249.90), and #1 (3, 249.90).   isopleth. Chloride and hydroxide complexation dominate the speciation of Pb in the SO 4 22 field and near the SO 4 22 -reduced sulfur boundary, while sulfide complexes account for most of the Pb under reduced conditions below the SO 4 22 -reduced sulfur boundary. Acetate (monocarboxylate) complexes comprise a significant portion of the Pb (up to 3%) in oxidized, weakly acid to mildly alkaline (6 to 8 pH range) fluids, while malonate (dicarboxylate) complexes are insignificant under all conditions along the 0.001 isopleth.
The speciation systematics for Zn in the model fluid are, in general, similar to those for Pb. Zinc speciation (mol% of total Zn) in the model ore fluid was calculated at specific log f O 2 -pH conditions along the 100, 0.01, and 0.001 ppm total Zn isopleths shown in Fig. 3. These speciation results are presented in Table 7 and Fig. 9, Table 8 and Fig. 10, and Table 9 and Fig. 11 for the 100, 0.01, and 0.001 ppm isopleths, respectively. Along the 100 ppm isopleth ( Fig. 3 and 9, Table 7) conditions are oxidized and chloride complexation dominates the speciation of Zn in acid to mildly alkaline conditions (pH from 3 to approximately 7.5), while hydroxide complexation dominates zinc speciation under pHs ranging from approximately 7.5 into the alkaline region. Sulfide complexes are insignificant under these oxidized conditions. Carboxylate complexation is maximized in the 5 to 7 pH range (where chloride complexation dominates zinc speciation) and is responsible for approximately 1.25% of the total Zn under these conditions. Acetate (monocarboxylate) complexation accounts for approximately 1 to 1.25% of the total Zn, while malonate (dicarboxylate) complexes comprise less than 0.25% of the total Zn under those conditions most favorable for the formation of Zn dicarboxylate complexes. Zinc speciation along the 0.01 ppm ( Fig. 3 and 10, Table 8) and 0.001 ppm ( Fig. 3 and 11, Table 9) isopleths is similar. For the 0.01 isopleth, hydroxide complexation dominates zinc speciation under oxidized conditions above pHs of approximately 7.5. Zn chloride complexes are predominant in the SO 4 22 field and 4.3 6 10 28 1 a Speciation for the indicated complexes is given as mol% of total Pb. b The 0.001 ppm isopleth is illustrated in Fig. 2. c Saturation with respect to hematite (Hm), pyrite (Py), or magnetite (Mt).  Table 6, along with corresponding % of total Pb values for various forms of Pb, and begin at the right end of the 0.001 ppm isopleth with data set #7 (8, 252.70), followed by #6 (6, 252.40), #5 (6, 253.55), #4 (6.45, 254.00), #3 (7, 254.60), #2 (7.3, 255.00), and #1 (7.55, 257.00).  Fig. 3. c Saturation with respect to hematite (Hm), pyrite (Py), or magnetite (Mt). near the SO 4 22 -reduced sulfur boundary at pHs ranging from approximately 5 to 7.5, while Zn sulfide complexes account for approximately 87 to near 100% of the total Zn under reduced conditions in the pH range 4.8 to 6.5. Acetate (monocarboxylate) complexation is significant only in oxidized, weakly acid to mildly alkaline solutions and accounts for approximately 1.25 to 1.3% of the total Zn under these conditions. Malonate (dicarboxylate) complexation is maximized, but insignificant, under the same conditions. For the 0.001 isopleth, hydroxide complexation, again, dominates under alkaline conditions in the SO 4 22 field and chloride complexes are dominant under acid to mildly alkaline conditions in the SO 4 22 field and near the SO 4 22 -reduced sulfur boundary, while 90 to nearly 100% of the zinc is complexed by sulfide ligands under reduced conditions below the SO 4 22 -reduced sulfur boundary.

Percentage of total Pb and Zn in carboxylate complexes: key determinants
The percent of dissolved Zn or Pb complexed by carboxylate ligands in a basinal brine will depend on four key variables, assuming a systems remains invariant with respect to temperature and pressure, total sulfur and total carbonate of the brine, major cation ratios of the brine, and mineralogy of the aquifer. The four variables are (1) the total chloride molality to total carboxylate ligand molality (Sm Cl 2 /Sm cbx ), (2) the mix of specific carboxylate ligands, for example the monocarboxylate to dicarboxylate ratio (Sm monocbx /Sm dicbx ), (3) the concentration of total inorganic sulfide (Sm H 2 S~mH 2 S 1 m HS 2), and (4) the pH. Values for these four parameters are 3.2 6 10 24 6.9 6 10 25 1.4 6 10 26 1 a Speciation for the indicated complexes is given as mol% of total Zn. b The 0.01 ppm isopleth is illustrated in Fig. 3. c Saturation with respect to hematite (H), pyrite (Py), or magnetite (Mt).   Fig. 3. c Saturation with respect to hematite (H), pyrite (Py), or magnetite (Mt).  Table 9, along with corresponding % of total Zn values for various forms of Zn, and begin at the right end of the 0.001 ppm isopleth with data set #8 (8, 253.05, followed by #7 (7, 252.85), #6 (6, 252.80), #5 (6, 253.55), #4 (6.4, 254.00), #3 (7, 254.60), #2 (7.3, 255.00), and #1 (7.7, 256.00). summarized in Table 10 for modern petroleum-field brines and for the four-modelled ore fluids considered in this study. The Sm H 2 S of the system will of course be controlled by the log f O 2 -pH conditions of the brine, assuming sulfate-sulfide equilibrium among the sulfur-bearing species of the system.

Modern petroleum-field brines: composition and types
Excellent summaries of the origin of sedimentary basin brines and their chemical evolution are given by Carpenter, 115 Hanor, 59,60 Kharaka et al., 116 and Kharaka. 104 The compositional data gathered over the past 40 years from sedimentary basins worldwide provide unequivocal evidence that deep formation waters, in particular petroleum-field brines, vary widely in composition (White, 117 Billings et al., 99 Hitchon et al., 100 Collins, 102 Carpenter, 115 Hanor,59,60,61,62,111,112,118 Kharaka, 104 Kharaka et al., 106,116 Land et al., 107 McManus and Hanor, 119 Wilson and Long, 120 Kyle, 110 Land, 113 Palandri and Reed 114 ). This diversity is particularly evident for those variables that significantly influence carboxylate complexation of metals in these fluids. Of particular interest to this study are those modern formation waters classified as Na-Cl and Na-Ca-Cl brines, because of their compositional similarity to inclusion fluids from Mississippi Valley-type and red-bed related base metal deposits (Roedder, 96 Heijlen et al. 67 ). In Table 10, the model ore fluids considered in this paper are compared to modern Na-Cl and Na-Ca-Cl brines in terms of specific fluid parameters that influence the concentration of Pb and Zn in these waters. In the discussion that follows salinity is defined as total dissolved solids, TDS, in units of milligrams per liter and chlorinity is defined as total chloride in units of milligrams per liter. The salinity of a Na-Cl or Na-Ca-Cl brine  Table 1 and Fig. 1 through 5. b Rittenhouse et al., 98 Billings et al., 99 Hitchon et al., 100,101 Collins, 102 Carpenter et al., 57 Carpenter and Trout, 103 Kharaka, 104 Kharaka et al., 79,105,106 Land et al., 107 Saunders and Swann, 108 Moldovanyi and Walter, 109 Kyle, 110 Lundegard and Kharaka, 78 Hanor,59,60,61,62,111,112 Land,113 Polandri and Reed. 114 Table 2: Smonocbx~total dissolved monocbx in mg L 21 ; Sm monocbx~t otal molality of dissolved monocbx in mol kg 21 . j dicbx~sum of dicarbonxylate anions listed in Table 2: Sdicbx~total dissolved dicbx in mg L 21 ; Sm dicbx~t otal molality of dissolved dicbx in mol kg 21 . k For composite ore-brine; the metal concentration at indicated log f O 2 -pH conditions. Cluster~most common values observed for modern petroleum field brines. l BDL~below detection limit; less than about 0.01 mg L 21 .
can be estimated from its chlorinity by the relation salinity1 .7 chlorinity (Hanor 61 ). This relation is based on the Cl to NaCl mass ratio of the brine and, therefore, with decreasing Na/Ca values, this estimate becomes less accurate. Nevertheless, it provides a quick conversion between salinity and chlorinity for the brines considered in this study. Chloride is the dominant anion in formation waters with salinities greater than 20 000 mg L 21 , while a transition zone between 20 000 and 10 000 mg L 21 defines a TDS boundary below which the dominant anion is either bicarbonate, sulfate, or acetate (Hanor 61,112 ). Sodium is the dominant cation in basinal brines with salinities less than 300 000 mg L 21 . In general, the relative proportion of Na to K, Ca, or Mg decreases in these waters with increasing salinity and Ca is the dominant cation in basinal brines having salinities in excess of 330 000 mg L 21 TDS. Between 300 000 and 330 000 mg L 21 TDS, the dominant cation is either Na or Ca. Although salinities generally increase with depth in a sedimentary basin, there are many exceptions to this trend. Thus, the salinity of a formation water is not necessarily related to temperature or pressure.
True in situ pHs of deep subsurface brines are not will known (Hanor 61 ) and must be estimated from field pH measurements or calculated using a field pH and various assumptions regarding the mineral saturation state of the fluid and loss of volatiles (e.g., H 2 S and CO 2 ) during sampling (Kharaka et al., 105,116 Aggarwal et al., 121 Palandri and Reed 114 ). In general, the pH of formation waters decreases with increasing salinity from values of approximately 8 for salinities near 20 000 mg L 21 to values between 3 and 4 for halite-saturated brines with salinities near 330 000 mg L 21 (Hanor 59,61 ). The available field evidence suggests that most Na-Cl and Na-Ca-Cl brines have a pH in the 4-8 range.
The distribution of modern petroleum-field brines in terms of total reduced sulfur content (SH 2 S) is bimodal in character (Table 10). Low H 2 S brines have SH 2 S values less than 1.0 mg L 21 , however, most have SH 2 S concentrations less than 0.02 mg L 21 and often have values below the detection limit of the analytical instrument used (Kharaka et al., 106 Moldovanyi and Walter, 109 Hanor 111 ). High H 2 S brines have SH 2 S concentrations greater than 1.0 mg L 21 with the highest concentrations approaching 1000 mg L 21 (Wade et al., 122 Moldovanyi and Walter 109 ). The most commonly observed SH 2 S concentrations from high H 2 S formation waters appear to be in the 10 to 600 mg L 21 range. Total sulfate concentrations in modern petroleum-field brines (Table 10) rarely exceed 1000 mg L 21 and are more typically in the 0.1 to 250 mg L 21 range (Kharaka et al., 106 Hanor 111 ). Although there appears to be a rough inverse correlation between SH 2 S and SSO 4 , there is no statistical correlation between the concentration of SO 4 22 and chlorinity or salinity (Hanor 61,111 ).
Monocarboxylate anions (mainly acetate, propionate, butyrate, valerate) and dicarboxylate anions (mainly malonate, succinate, glutarate, maleate) are the dominant organic acid species in basinal brines. These species are commonly detected in formation waters associated with petroleum and obtained from reservoirs with temperatures ranging from 20 to 200 uC (Kharaka et al. 79 ). Concentrations of carboxylic acid anions in modern petroleum-field brines (Tables 2 and 10) are highly variable with respect to their major known controls, including age and temperature of reservoir rocks, type and amount of kerogen in source rocks, and the nature of metastable equilibria that characterized organic-matter interactions in the reservoir (Kharaka, 106 Shock, 1,123 Lundegard and Kharaka, 78 Giordano and Kharaka, 51 Franks et al., 124 Seewald 125 ). Although concentrations vary with temperature and reservoir depth for any given petroleum field, maximum values are generally present in Cenozoic-age reservoir rocks in the temperature range 80 to 120 uC (Hanor, 111 Kharaka et al., 79 Lundegard and Kharaka 78 ). Based on carboxylate anion concentration versus salinity relationships for petroleum-field brines worldwide, Hanor 111 has shown that there is a pronounced decrease in carboxylate anion concentration with increasing salinity. The highest concentrations of carboxylic acid anions are found in waters having salinities less than about 30 000 mg L 21 , while these species do not appear to be major constituents of petroleumfield brines having salinities that exceed 100 000 mg L 21 .
The total concentration of monocarboxylate anions (Smonocbx) in modern petroleum-field brines is generally less than 100 mg L 21 , however, many brines have concentrations in the 100 to 5000 mg L 21 range. The dominant monocarboxylate anion in most petroleum-field brines is acetate, followed in order of decreasing concentration by propionate, butyrate, and valerate ( Table 2). The highest concentration of acetate reported is 10 000 mg L 21 but values higher than 5000 mg L 21 are very rare (Lundegard and Kharaka, 78 Kharaka et al. 79 ). Typically acetate comprises 80 to 90% of the total monocarboxylate species (Giordano and Kharaka 51 ). The dominant dicarboxylate species in modern petroleum-field brines are malonate, succinate, and glutarate. Each of these species is not likely to have concentrations greater than about 100 mg L 21 ( Table 2), but controversially high values for oxalate (494 mg L 21 ) and malonate (2540 mg L 21 ) have been reported (MacGowan and Surdam 126 ). Reported total dicarboxylate anion concentrations (Sdicbx) range from 0 to 2640 mg L 21 but values higher than 300 mg L 21 are very rare.
Typically, brines high in dicarboxylate anions are also high in monocarboxylate anions; however, brines high in monocarboxylate species are not always enriched in dicarboxylate anions. The total dicarboxylate anion concentration is typically less than 5% of the total carboxylate concentration for modern petroleum-field brines. Thus, the ratio of total monocarboxylate anions to total dicarboxylate anions (Sm monocbx /Sm dicbx ) for modern basinal brines is nearly always greater than 20.

Modern petroleum-field brines: Pb and Zn content
The speciation results from this study show that deep formation waters characterized by temperatures near 100 uC, high oxidation states (i.e., low Sm H 2 S ), high chlorinities (y100 000 mg L 21 ), and high but reasonable concentrations of carboxylate anions can mobilize up to 3% of the total Pb and up to 1.3% of the total Zn as carboxylate complexes. Furthermore, these percentages, under the most favorable conditions, correspond to approximately 1 to 100 ppm of these metals in solution; concentrations that are adequate to form economic deposits of these metals. However, the field evidence suggests that all of these optimum conditions for carboxylate complexation are rarely met at the same time and that total concentrations of Pb and Zn are typically below 0.1 mg L 21 in most oil-field brines (Kharaka et al. 105,106 ). Nevertheless, as indicated in Table 10, formation waters high in trace metal content are common in a dozen or so localities worldwide (Carpenter et al., 57 Kharaka, 104 Kharaka et al., 105,106 Saunders and Swann, 108 Moldovanyi and Walter, 109 Kyle, 110 Land, 113 Hanor, 59,61,62 Hitchon et al. 101 ). These brines typically have concentrations of Pb and Zn in the 1 to 250 mg L 21 range, with maximum reported concentrations of 111 mg L 21 and 575 mg L 21 for Pb and Zn, respectively (Table 10).
To ascertain the factors influencing base metal concentrations in deep formation waters, the compositional data from a large number of modern petroleum-field brines, worldwide, have been evaluated by many investigators using a variety of correlation and chemical modelling methods (Sverjensky,31,41 Kharaka et al., 106,116,127 Lundegard and Kharaka, 128 Giordano and Kharaka, 51 Hanor,60,62,111,112 Hitchon et al., 101 Palandri and Reed 114 ). Based on these studies as well as others, it is now well established that three principal factors control Pb and Zn concentrations in modern basinal brines: (1) the availability of Pb and Zn in the basinal system, (2) the total reduced sulfide content (SH 2 S) of the brine, and (3) the chlorinity of the brine (SCl 2 ).
The field evidence for modern petroleum-field brines, as summarized in Table 10, suggests an inverse relationship between total sulfide content and the concentration of Pb and Zn. All high SH 2 S brines have low concentrations of these metals, while all brines with elevated Pb and Zn values have low concentrations of total sulfide. General details of this inverse relation are presented by Kharaka et al., 106 Moldovanyi and Walter, 109 and Hanor. 61 Hanor 61 also reports a pronounced covariation in Zn and Pb corresponding to a mass ratio of approximately 5 : l for brines with dissolved Zn concentrations greater than a few mg L 21 , while there is considerable scatter in Zn to Pb ratios for brines having Zn and Pb concentrations below 1 mg L 21 . These relationships (Zn 1 Pb vs. SH 2 S and Zn vs. Pb) support the conclusion that one control of Pb concentration and Zn concentration in high trace-metal brines is mineral buffering, possibly equilibrium with galena and sphalerite, respectively (Kharaka et al., 106 Hanor 61,62 ).
However, because most low SH 2 S petroleum-field brines also have low concentrations of trace metals (e.g., Pb and Zn v 1.0 mg L 21 each, see Table 10) and highly variable Zn to Pb ratios, the availability of Pb and Zn in the brines must also be a major control on the concentration these metals. Because galena and sphalerite are normally not present as components of basinal sediments, it seems more likely that Pb and Zn in most petroleum-field waters (as well as related ore fluids) were ultimately leached from other sources including metal oxides and hydroxides associated with red-bed sediments, organic matter, K-feldspar in sandstone, carbonate phases, or shales (Sverjensky,31,41 Kharaka et al., 106 Hanor 61 ). If such is the case, then it is not surprising that most low SH 2 S formation waters with low concentration of Pb and Zn are also characterized by variable Zn to Pb ratios.
Both field and modelling evidence demonstrate that chlorinity has a strong influence on the trace metal content of petroleum-field brines (Lundegard and Kharaka, 128 Hanor, 60,61,111 Kharaka et al. 79 ). The relationship of chlorinity versus Zn concentration for Gulf Coast saline waters (Hanor 61 ) shows unequivocally that petroleum-field brines which have Pb and Zn concentrations in excess of 1.0 mg L 21 nearly always have chlorinities in excess of approximately 100 000 mg L 21 (salinity~170 000 mg L 21 ). There is a SCl 2 transition zone between 60 000 mg L 21 (salinity~100 000 mg L 21 ) and 100 000 mg L 21 in which a few brines have Pb or Zn in excess of 1 mg L 21 . As mentioned by Hanor,61,111 these results strongly suggest a threshold chlorinity of roughly 100 000 mg L 21 below which formation waters are typically poor solvents of Pb and Zn and above which Pb and Zn, in concentrations capable of forming economic deposits of these metals, can be mobilized by formation waters. These field relationships are consistent with modelling results, including those of this study, which show that chloride complexing is the principal solubility mechanism for Pb and Zn in low SH 2 S petroleum-field brines and is probably the principal solubility mechanism for these metals in related ore fluids (cf., Sverjensky, 31, 41 Kharaka et al., 79,106 Lundegard and Kharaka, 128 Giordano and Kharaka, 51 Hanor 61,62 ).

Model ore fluids
A comparison of the composite ore fluid compositions from this study (Table 10 and related isopleths in Fig. 2 and 3) and the modern brine data in Table 10 shows that the ore brines, corresponding to log f O 2 -pH conditions based on the Anderson 37 and Giordano 76 model fluids, are similar in many respects to modern, high trace-metal petroleum-field brines. Both sets of fluids have the following characteristics in common: (a) chlorinities (SCl 2 ) near 100 000 mg L 21 or higher, (b) pH in the 4-6.5 range, (c) less than 0.1 mg L 21 total sulfide (SH 2 S), (d) total sulfate (SSO 4 ) in the 0.1 to 1000 mg L 21 range, and (e) Pb concentrations in the 1 to 100 mg L 21 range and Zn concentrations in the 1 to 1000 mg L 21 range. Note that the Pb and Zn isopleths transecting the more reduced log f O 2 -pH regions of the Anderson 37 and Giordano 76 ore fluids in Fig. 2 and 3, as well as the Pb content (65 mg L 21 ) and Zn content (420 mg L 21 ) for the corresponding composite fluids in Table 10, are close to the highest concentrations reported for these metals in petroleum-field brines (Pb, 111 mg L 21 ; Zn, 575 mg L 21 ). Also note that the 1 mg L 21 minimum for Pb and Zn in high trace-metal brines is consistent with the 1ppm Pb and 1 ppm Zn isopleths in Fig. 2 and 3, that are not far from the lower boundary of the Anderson 37 and Giordano 76 oreregions.
The modelled fluids in this study are in equilibrium with galena and sphalerite. Hence, the above similarities in Pb and Zn content between the composite ore fluids and modern high trace-metal brines is consistent with the above assertion that an important control of Pb and Zn content in high trace-metal brines is mineral buffering, most likely equilibrium with galena and sphalerite. However, as mentioned by other investigators (Sverjensky,31,41 Kharaka et al., 106 Hanor 61 ), it is likely that the original sources for much of the dissolved Pb and Zn in metalrich brines are phases other than galena and sphalerite; including organic matter, oxides, hydroxides, silicates, and carbonates. The principal differences between modern high trace-metal brines and the composite ore fluids of Anderson 37 and Giordano 76 relate to their carboxylate anion content ( Table 10). The reported concentrations of monocarboxylate anions (Smonocbx) and dicarboxylate anions (Sdicbx) in high trace-metal petroleum-field brines (v1 to 300 mg L 21 and v1 mg L 21 , respectively) are significantly lower than the concentrations assumed in the modelled brines of this study (Smonocbx~7 700 mg L 21 and Sdicbx~300 mg L 21 ). Also, the ratio of monocarboxylate acid anions to dicarboxylate ligand species (Smonocbx/Sdicbx) appears to be higher in modern high trace-metal brines compared to the composite ore fluid. These systematics in organic acid content suggest that carboxylate complexation of Pb and Zn in modern brines should be less than predicted by the composite brine modal in this study. There are also major differences in the corresponding total chloride to carboxylate ratio (Sm Cl /Sm cbx ). Modern high trace-metal brines have much higher Sm Cl /Sm cbx values and, therefore, the contribution carboxylate complexes to the total Pb and Zn content in these modern brines is likely to be significantly less than the 1 to 3 percent for the composite ore fluids of Anderson 37 and Giordano 76 (). This is consistant with the prediction of Giordano and Drummond 87 that zinc chloride complexes dominate over acetate complexes of Zn if Sm Cl /Sm acetate is greater than 10 at temperatures near 100 uC.
In contrast to these ore solutions, the composite ore-brine based on the Giordano and Barnes 39 MVT ore fluid (Table 10, Fig. 2 and 3) is comparable to the high salinity (w170 000 mg L 21 TDS) subset of modern brines characterized by low tracemetal content and high total reduced sulfur (SH 2 S) (Table 10, bottom half, right column). The Giordano and Barnes 39 composite fluid is saturated with respect to galena and sphalerite and Pb and Zn speciation is dominated by sulfide complexes (Tables 5, 6 , 8, 9). Thus, it is likely that concentrations of Pb and Zn in modern low trace-metal brines, with high sulfide content and slightly acidic to alkaline in pH, are also controlled by galena and sphalerite solubility and sulfide complexation of Pb 21 and Zn 21 . Overall concentrations of Pb and Zn are low because of the high sulfide content and, hence, the activities of Pb 21 and Zn 21 are suppressed because of mass action effects. It is likely that carboxylate complexation of Pb and Zn is insignificant in these fluids because of the presumably low concentration of carboxylate ligands and because Pb 21 and Zn 21 activities are low, thus suppressing the formation of carboxylate complexes, which are significantly weaker that sulfide complexes. Although high concentrations of carboxylate anions have been observed in certain high SH 2 S petroleum-field brines, these high concentrations are always from brines with salinities much lower than 170 000 mg L 21 TDS.
A comparison of the Sverjensky composite ore-brine 41 with modern petroleum-field brines in terms of SH 2 S and Zn content, revels that this ore fluid holds a border position between modern high trace-metal brines and those with low trace-metal content and high SH 2 S (Table 10). The composite Sverjensky fluid, as represented by the model results in Table 10 and Fig. 2 and 3, barely qualifies as a modern low trace-metal/high SH 2 S brine and it is not far removed from modern high trace-metal brines that have total Zn content and SH 2 S near the lower limit (1 mg L 21 ) and upper limit (0.1 mg L 21 ), respectively, for this brine category. As mentioned above, both of these modern brine-types are probably saturated, or close to saturation, with respect to galena and sphalerite. Also, these two modern brine-types generally coincide when SH 2 S, Zn, and/or Pb all have values within or near the 1 to 10 mg L 21 range. Thus, modern brines having these characteristics are referred to below as border-type brines.
Based on brine-composition data from numerous references cited in this paper, border-type brines do exist but are rare. Hanor 61 discusses the variation of total Zn with SH 2 S in Gulf Coast formation waters and suggests that brines which contain both dissolved Zn and dissolved total reduced sulfur in the 1 to 10 mg L 21 range are probable. He also implies that these brines are likely to be saturated with respect to galena and sphalerite. Hanor 61 also conducted computer simulations of sphalerite solubility, at 100 uC, in hypothetical basinal brines and reports that at chlorinites in excess of 150 000 mg L 21 (TDS~250 000 mg L 21 ) it may be possible for concentrations of Pb and Zn in excess of 1 mg L 21 to coexist in solution with comparable concentrations of dissolved inorganic sulfide. It is interesting to note that the original MVT ore fluid of Sverjensky 41 is initially saturated with respect to galena, but undersaturated with respect to sphalerite and contains 1.3 mg L 21 Zn, 2 mg L 21 Pb, and approximately 1 mg L 21 dissolved sulfide (SH 2 S). Thus, Sverjensky's initial MVT ore fluid also corresponds to a border-type brine, as defined above.
The principal compositional difference between the original Sverjensky MVT fluid 41 and the composite Sverjensky brine modelled in this study is in the total sulfur content. The original ore fluid has a total sulfur content of approximately 10 24.6 molal (1 mg L 21 as H 2 S), while total sulfur content in the composite fluid of this study is 10 22 molal (340 mg L 21 as H 2 S or 960 mg L 21 as SO 4 22 ). Thus, for the most oxidized and acidic conditions of the composite Sverjensky model, pH near 4.6 and log f O 2 near -51.8 (Table 10, Fig. 2 and 3), both ore fluid models, yield similar total Zn and SH 2 S values, close to those of a border-type brine. However, as log f O 2 -pH conditions deviate from this point in terms of log f O 2 , but still remain in the delimited ore-fluid region, the composite fluid yields values for SH 2 S and concentrations of Zn and Pb that are more in line with modern brines high in SH 2 S and low in trace-metals, while the original model fluid of Sverjensky, 41 remains a border-type brine because the total reduced sulfur content can not increase beyond 10 24.6 molal, or 1 mg L 21 SH 2 S.
Border-type brines and ore fluids with similar key characteristics, are probably at or near saturation with respect to galena and sphalerite and mobilization of Pb and Zn in these fluids is dominated by either chloride or sulfide complexation, depending on the log f O 2 -pH conditions, the chlorinity, and the total dissolved sulfide content. For the more oxidized conditions within the Sverjensky ore-fluid region 41 (Fig. 2 to 5, Tables 5 and 8), chloride complexation dominates Pb and Zn solubility, with only a few percent of the total Pb and Zn being transported by carboxylate complexes. Under more reduced conditions, sulfide complexation of Pb and Zn dominates transport of these metals and, once again, carboxylate complexes are insignificant.

Geochemical modelling of petroleum-field brines
The model results and field evidence presented above are consistent with chemical simulation studies of carboxylate complexation in modern petroleum-field brines. (cf., Kharaka et al., 79,105,106,116,127 Lundegard and Kharaka, 128 Giordano and Kharaka 51 ). The results from these studies indicated (1) that concentrations of Pb and Zn in carboxylate complexes form only a small percentage of the total Pb and Zn in metalrich petroleum-field brines, (2) chloride complexation dominates Pb and Zn speciation in these brines, and (3) significant amounts of Ca, Mg, Fe, and Al can be complexed by carboxylate ligands in petroleum-field brines. The influence of pH and oxygen fugacity on the solubilization of Ca, Mg, Al, and Fe by carboxylate complexes in fluids similar to those modelled in this study is currently being investigated by the author and results of this study will be the topic of a forthcoming paper.

Conclusions
The chemical simulation results from this study and the comparisons made between proposed ore fluids and modern brines lead to several important conclusions regarding the role of carboxylate complexation of Pb and Zn in Mississippi Valley-type (MVT) ore fluids, ore fluids responsible for red-bed related base metal(RBRBM) deposits, and modern petroleumfield brines.
(1) A comparison of dissolved Pb, Zn and sulfur content of modern petroleum-field brines with Pb and Zn solubility systematics in log f O 2 -pH space for the composite ore-brine in this study reveals good correlations between modern brinetypes and proposed ore fluid models for Mississippi Valley-type deposits and red-bed related base metal deposits.
(2) Based on analyses of formation waters worldwide and the modeling results of this study, modern petroleum-field brines and the model ore fluids can be divided into four main categories: (a) High trace-metal, low total sulfide brines which are similar to the 1975 model MVT fluid of Anderson 37 and the 1994 RBRBM ore fluid of Giordano. 76 Total sulfide is less than 0.1 mg L 21 , while total Pb and Zn have values between 1 and 100 mg L 21 and 1 and 1000 mg L 21 , respectively.
(b) Low trace-metal, high total sulfide brines which are similar to the 1981 MVT ore fluid of Giordano and Barnes. 39 Total sulfide is between 1 and 1000 mg L 21 , while total Pb and Zn are normally less than 0.1 mg L 21 .
(c) Low trace-metal, low total sulfide brines which may correlate to spent ore fluids or to basinal fluids not sufficiently enriched in Pb or Zn to be ore fluids of these metals.
(d) Border-type brines that in terms of reduced sulfur and trace metal content are on the border between high trace-metal brines and brines low in trace metals but high in sulfide. Brinetypes (a) and (b) above coincide when total sulfide, total Pb, and/ or total Zn all have values near or within the 1 to 10 mg L 21 range. Brines of this type are rare and generally correspond to the 1984 model ore fluid of Sverjensky. 41 (3) In modern basinal brines and related ore fluids with chlorinites near 100 000 mg L 21 , chloride complexes dominate the speciation of Pb and Zn under log f O 2 -pH conditions most favorable for the complexation of these metals by carboxylate ligands.
(4) The model results at 100 uC indicate that, for these basinal brines under the most favorable conditions, up to 3% of the total Pb and up to 1.3% of the total Zn can be mobilized by carboxylate complexes, principally those involving the acetate ligand.
(5) The percent of dissolved Zn and Pb complexed by carboxylate ligands in a basinal brine will depend primarily on four key variables, assuming a system remains invariant with respect to temperature, total sulfur content of the brine, and mineralogy of the aquifer. The four variables are (a) the ratio of total chloride molality to total carboxylate ligand molality, (b) the mix of specific carboxylate ligands, for example the molar ratio of monocarboxylate ligands to dicarboxylate ligands, (c) the concentration of total sulfide, which is controlled by the log f O 2 -pH conditions of the brine, assuming sulfate-sulfide equilibrium among the sulfur-bearing species of the system, and (d) pH.
(6) Optimum brine conditions, at 100 uC, for carboxylate complexation include (a) high oxidation state, reflecting a high total sulfate to total sulfide ratio and total sulfide concentrations less than 10 26 molal (v 0.03 mg L 21 as dissolved H 2 S), (b) high but reasonable concentrations of carboxylate anions; near 7000 mg L 21 for total monocarboxylate species and 300 mg L 21 for total dicarboxylate ligands, and (c) pH in the 4-8 range.
(7) At 100 uC, under optimum brine conditions for carboxylate complexation, approximately 1 to 100 mg L 21 and 0.l to 10 mg L 21 of Zn and Pb, respectively, can be mobilized as carboxylate complexes, principally those involving acetate.
(8) However, the field evidence based on analyses of petroleum-field brines worldwide, suggests that all of these optimum conditions are rarely, if ever, met at the same time, particularly with regard to the requisite carboxylate ligand content. Thus, it appears that carboxylate complexation plays a minor, if not insignificant, role as a transport mechanism for Pb and Zn in high salinity Na-Cl and Na-Ca-Cl basinal brines and related ore fluids.
(9) More refined simulations of base-metal solubility and speciation in basinal brines can be conducted by making the following modifications of the model used in this study: (a) the use of reliable 100 uC thermodynamic data for Pb 21 and Zn 21 malonate complexes, (b) the inclusion of Cu 1 complexation in the model, (c) better estimates of log f O 2 and pH conditions for the modelled ore fluids and modern petroleum-field brines, and (d) the use of mineral solubility constraints to control the activities of Na 1 , Ca 21 , and Mg 21 in the model.