Effect of fulvic acid on the sorption of Cu and Pb onto γ-Al2O3

(1)

Effect of fulvic acid on the sorption of Cu and Pb onto g-Al

2

O

3 Chung-Hsin Wu

a,

*, Cheng-Fang Lin

b

, Hwong-Wen Ma

b

, Ting-Qwet Hsi

b

a_{Department of Environmental Engineering and Health, Yuanpei Institute of Science and Technology, 306 Yuanpei Street,}

Hsinchu, Taiwan, ROC

b

Graduate Institute of Environmental Engineering, Nation Taiwan University, Taipei 106, Taiwan, ROC Received 19 October 2001; received in revised form 26 June 2002; accepted 2 August 2002

Abstract

This work investigated the adsorption of Cu and Pb at the surface of g-Al2O3in the presence of fulvic acid to address

the signiﬁcance of dissolved organic matters on metal partitioning. Fulvic acid, obtained from International Humic Substance Society, represented dissolved organic matter. Fulvic acid concentrations employed herein were 1, 5, and 10 mg C/L, which simulated the relevant environmental conditions. Ion selective electrodes were employed to ascertain free Cu and Pb measurements. The maximum adsorption of 10 mg C/L fulvic acid on g-Al2O3was 5 102mg C=mg

g-Al2O3: Fulvic acid promoted Cu adsorption in low pH conditions. The effects of fulvic acid on Pb adsorption were

similar to those of Cu. The conditional stability constants of sorbed fulvic acids with Cu and Pb were determined to be in the order of 4 to 6 ðlog KÞ: Cu and Pb species were modeled in heterogeneous systems using triple-layer model. Simulation results indicated that metal species are dominantly in complexation with fulvic acid, both in solution and at the g-Al2O3surface.

1. Introduction

Iron and/or aluminum oxides and hydroxides regulate the transport and concentration distribution of cations and anions in natural environments. The distribution and partitioning of reactive substances and other environmental pollutants are regulated primarily by reactions such as adsorption/desorption and precipita-tion, which occur at the interface between the aqueous solution and minerals [1]. In the 1970s, Stumm began to investigate the interfacial reactions of anions and cations onto the surface of synthetic oxides and developed the surface complexation model to describe the equilibrium conditions of interfacial complexation reactions qualita-tively.

Humic substances, a common component of soil and natural water bodies, are the primary metal-complexing chelates and, hence, perform a vital function in the environmental fate, bioavailability, toxicity, and mobi-lity of heavy metals in the biosphere [2–4] Liu and Gonzalez [5] and Jin et al. [6] demonstrated that a high complexation capacity of humic acid for metals and the strength of binding is in the sequence of Pb > Cu > Cd. Liu and Gonzalez [5] also showed that pH and ionic strength are the most important variables in controlling metal complexation with humic acid. Sebastien et al. [7] indicated that pH, soil organic matter, and metal concentrations control the solid-solution partitioning of metal. Neubauer et al. [8] also conﬁrmed that various parameters including, pH and clay and organic matter content govern mobility and toxicity of heavy metals in soil. Adsorption of natural organic matter (NOM) alters, modiﬁes and transforms the oxide surface physically and chemically. Adsorption renders negative charges to the oxide surface due to the anionic nature of *Corresponding author. Tel.: 2-2239-0986; fax:

+886-5-5334958.

E-mail address:chwu@pc.ymit.edu.tw (Chung-Hsin Wu).

(2)

NOM molecules. NOM performs a fundamental mod-iﬁcation in the distribution, transport and reactions of metals in aquifers. Therefore, the conditional stability constants of both dissolved and sorbed NOM with metal ions must be identiﬁed in order to model the adsorption behavior in such a system.

In an aqueous phase, metal ions form stable complexes with NOM and reduce the precipitation, therefore increasing the metal ion migration. Metal ions are able to associate with oxides either by binding directly to the oxide surface or by forming complexes with adsorbed NOM molecules. Alcacio et al. [9] explored that clay-organic complexes are common in soils and sediments, ternary complexes involving Cu(II) are likely to occur and to inﬂuence the binding and dissolution of copper in these nature systems. Airton and Wilson [10] demonstrated that the interactions with dissolved organic binding sites controlled copper species in raw sewage. Plavsic and Cosovic [11] displayed that the interaction of copper ions with alumina particles was enhanced in the presence of humic acid. Frimmel and Huber [12] indicated that NOM increased the dissolved fraction of Cu and Pb, and decreased that of Cd. Gagnon et al. [13] demonstrated that generally phenolic compounds enhanced the sorption of metallic cations on clay minerals. It becomes fairly important that under-standing the transport and distribution of metal ions relies heavily upon exploring the interfacial reactions between metal ions, organic matter, and oxides. In this study, fulvic acid was employed as a model dissolved organic matter (DOM) compound herein to explore the inﬂuences between metal ions (as Cu(II) and Pb(II)) and the interface of g-Al2O3=water: The adsorption of Cu

and Pb in varying fulvic acid concentrations was further simulated via the triple-layer model (TLM) to depict the adsorption speciation. The importance of oxide on metal retention and which parts of organics and oxides controlling the solubility of metal ions are addressed.

2. Materials and methods 2.1. Materials

Fulvic acid employed herein was obtained from the International Humic Substance Society, USA. For reference, the elemental analyses for Suwannee stream reference fulvic acid are: C, 53.3%; H, 4.24%; O, 41.29%; N, 0.69%; S, 0.59%; and P, 0.01%, and the molecular weight is 781 g=mole: The fulvic acid solu-tions were prepared by dissolving appropriate amounts of the fulvic acid powder in water (Milli-Q reagent water), which was then filtered through a 0:45 mm cellulose nitrate filter membrane (Whatman). The total organic carbon of filtrate was determined with a TOC analyzer (O.I. Corporation, Model 700).

Reagent grade chemicals CuðClO4Þ2 6H2O (Aldrich)

and PbðClO4Þ2(Orion Lead Standard 948206) were used

to prepare the stock Cu and Pb solutions, respectively. Free Cu and Pb concentrations (½Cu2þ_{and ½Pb}2þ_{) were}

determined via an ion selective electrode (Orion Model 94-29 Cupric electrode; Orion Model 94-82 Lead electrode) in conjunction with a reference electrode (Orion Model 9002, double junction reference elec-trode). The dissolved (sum of free-type and complexed-type) Cu and Pb concentrations were determined via analysis of the acidiﬁed ﬁltrates using inductively coupled plasma atomic emission spectrometry (ICP, Jobin Yvon 24). Reagent grade NaOH and HClO4were

used for system pH adjustment.

Based on the procedure described by Hohl and Stumm [14], the g-Al2O3 obtained from Aerosil Co.

(Japan) was ﬁrst prewashed with 0:1 M NaOH. Then, prior to its use, the g-Al2O3 was rinsed with deionized

water, dried, ground, and passed through a 200 mesh sieve. Puriﬁcation procedures removed impure sub-stances which might affect adsorption results. BET analysis of nitrogen gas adsorption was employed to ascertain the surface area of g-Al2O3(about 100 m2=g).

2.2. Sorption experiments

Fulvic acid (1, 5, and 10 mg C=L), Cu ð105MÞ; and Pb ð105_{MÞ were equilibrated respectively through}

g-Al2O3 suspension ð100 mg=LÞ with 0:01 M NaClO4

in single-solute equilibrium adsorption experiments. A miniscule amount of NaOH and HClO4 was added to

cover the pH range 4–6. Notably, all experiments were performed at 2570:51C for 24 h: At the end of the equilibrium period, the pH of each slurry suspension was immediately determined. The suspensions were passed through 0:45 mm ﬁlter membrane and TOC analyzer and ICP were employed to determine the fulvic acid, Cu, and Pb concentrations, respectively. The difference between initial and ﬁnal adsorbate concentra-tions provided the percentage of adsorbate adsorption.

Metal (Cu and Pb)-(fulvic acid)-g-Al2O3samples were

prepared and analyzed in the same manner as the cations (Cu and Pb)-g-Al2O3 and (fulvic acid)-g-Al2O3

systems were. The experimental results obtained from ICP and ion selective electrodes were the metal ions in total dissolved and free concentrations, respectively. The complexed metal concentration was the difference between total dissolved and free concentrations. 2.3. Model analysis

The TLM of Hayes and Leckie [15] was employed to simulate the equilibrium distribution of fulvic acid and metal ions at the g-Al2O3=water interface. Table 1

displays the TLM reactions and expressions of the intrinsic reaction constants normally used in a g-Al2O3

(3)

system [16]. Eqs. (1) and (2) (Table 1) deﬁne protonation of reacting surface sites, and Eqs. (3) and (4) describe the formation of complexes between the background electrolyte ions and the surface. TLM resolved the equilibrium and mass balance equations simultaneously. The surface area of the g-Al2O3; which was determined

by a BET adsorption isotherm, was equal to 100 m2_=g

and its site density was assumed to be 8 sites=nm2_[17].

Hohl and Stumm [14] illustrated the intrinsic equili-brium constants of Eqs. (1–4). Inner-and outer-plane capacitances were assumed to be 80 and 20 mF=cm2_;

respectively [18]. These parameters were then applied in the model analysis to simulate the experimental data for the fulvic acid þ Cu=g-Al2O3 and fulvic acid þ

Pb=g-Al2O3systems.

3. Results and discussion

3.1. Adsorption of fulvic acid=g-Al2O3systems

Fig. 1 presents the adsorption envelop of fulvic acid on 100 mg=L g-Al2O3 in the presence of three various

fulvic acid concentrations (1, 5, and 10 mg C/L). The pH dependent adsorption curve is similar to those of other anions. That is, with a system pH shift from alkaline to acidic conditions, the fulvic acid sorption increased. However, the maximum adsorption of fulvic acid was not present in the lowest pH value, which differs from other anions. Results in Fig. 1 indicate that the maximum adsorption was approximately at pH 5.0 that was comprised of 75% ðfulvic acid ¼ 1 mg C=LÞ; 62% ðfulvic acid ¼ 5 mg C=LÞ; and 51% ðfulvic acid ¼ 10 mg=LÞ; respectively. The experimental ﬁndings pre-sented herein are similar to those of previous studies [19,20]. Davis [19] and Schroth and Sposito [20] indicated that the maximum adsorption ratio of dissolved organic matters occurred between pH 6.3 and 4.5. These experimental results also indicated that the adsorption trend of dissolved organic matters decreased in both acidic and alkaline conditions,

probably because fulvic acids bear different functional groups which may exhibit quite distinct sorption characteristics.

The maximum sorption density of fulvic acid was 7:5 mg C/g g-Al2O3ðfulvic acid ¼ 1 mg C/L), 31 mg C/

g g-Al2O3 ðfulvic acid ¼ 5 mg C/L), and 51 mg C/g

g-Al2O3ðfulvic acid ¼ 10 mg C/L). This revealed that the

addition of a higher fulvic acid concentration resulted in the higher adsorption density within the equilibrium systems. Zhou et al. [21] demonstrated that mineral type, particle size and surface area affect humic substance adsorption. At a constant adsorbent concentration, the adsorption density varies with the amount of fulvic acid added. Gu et al. [22] presented that due to their heterogeneity and complexity, the adsorption mechan-isms of natural organic matter on mineral surface are not entirely understood. To model the adsorption of NOM on oxides, intrinsic study may not be easy and obtainable but the apparent description can be an alternative.

Table 1

The reactions and equilibrium expressions in the TLM

Reaction Intrinsic equilibrium expression/constant

SOHþ₂ ¼ SOH þ Hþ Kint a1 ¼ ½SOH½Hþ ½SOHþ 2 exp j0F RT ¼ 107:2 (1) SOH ¼ SO_{þ H}þ Kint a2 ¼ ½SO_½Hþ ½SOH exp j₀F RT ¼ 109:5 (2) SOH þ Naþ¼ SO_Naþ_{þ H}þ Kint Naþ¼ ½SO_Naþ_½Hþ ½SOH½Naþ exp ðj_b j₀ÞF RT ¼ 109:1 (3) SOH þ Hþþ ClO 4 ¼ SOH þ 2–ClO 4 _Kint ClOþ4 ¼ ½SOH þ 2 ClO 4 ½SOH½Hþ_½ClO 4 exp ðj0 jbÞF RT ¼ 108:2 (4)

Fig. 1. Adsorption of fulvic acids (1, 5, and 10 mg C/L) onto g-Al2O3as a function of pH.

(4)

3.2. Adsorption of Cu=g-Al2O3and Pb=g-Al2O3systems

Figs. 2a and b present the sorption envelope of 105_{M Cu and Pb on 100 mg=L g-Al}

2O3 during pH

4–6, respectively. The adsorption density of metal ions increased with a system pH shift from acidic to alkaline conditions. The removal may be attributed to the adsorption on oxide surface and not the formation of metal hydroxide precipitation, since the adsorption ratio of Cu and Pb failed to attain 100% and the experiment remained below pH 7 to avoid potential metal hydroxide precipitation. The sorption of Cu and Pb was negligible at pH less than 4.0; this is probably due to the competition for reaction sites between metal ions and Hþ as described by Airton and Wilson [10].

3.3. Adsorption of fulvic acid þ Cu=g-Al2O3and

fulvic acid þ Pb=g-Al2O3systems

Theoretically, soluble fulvic acids inﬂuence the adsorption of Cu and Pb in three manners. Firstly, adsorption of fulvic acids increases the negative charge on g-Al2O3 surface and, hence increases Cu and Pb

adsorption. Secondly, by competing with the adsorption sites for Cu and Pb, the presence of fulvic acid in a solution may decrease Cu and Pb adsorption. Lastly, the fulvic acid–metal complexes in solution also affect metal adsorption onto oxide. As proposed by Stumm [1], there are two possible structures for the adsorption of the metal and organic complex compounds on mineral surfaces. One is the S–Me–HA and the other is the S–HA–Me, where S represents the adsorption site on the solid surface and Me is the metal ion. Liu and Gonzalez [23] indicated that the most possible surface structure of bivalent metal ions between montmorillo-nite and humic acid should be S–Me–HA. The variable

effects of ligands on metal adsorption depend on pH, concentration of ligand, metal loading, formation constant of the complex, as well as the ionic strength of solution.

Fig. 3a illustrates the extent of Cu adsorption in the binary-solute system of fulvic acid and Cu with 100 mg=L g-Al2O3: The Cu adsorption ratio within 0,

1, 5, and 10 mg C=L fulvic acid, during pH 4–6, was 0– 55%, 0–65%, 10–80%, and 20–80%, respectively. Thus, fulvic acid seemed to promote Cu adsorption in concentrations greater than 5 mg C=L: However, the promotive ability of fulvic acid was signiﬁcant in lower pH conditions (such as pH near 4). The Pb adsorption ratio within 0, 1, 5, and 10 mg C=L fulvic acid, during pH 4–6, was 0–55%, 0–60%, 10–80%, and 20–80%, respectively (Fig. 3b). The effects of fulvic acid on Pb adsorption were similar to those of Cu expect that the presence of fulvic acid enhances Pb adsorption at the pH range studied. In a similar study conducted by Vermeer et al. [24], Cd sorption in Aldrich humic acid/hematite system increases as comparison to that on the single oxide system. Dalang et al. [25] revealed that adsorbed fulvic material on kaolinite produces an increase in the total quantity of adsorbed Cu. Generally speaking, observations of natural dissolved organic effects on metal adsorption seem to depend highly upon experi-mental conditions, particularly the relative concentra-tions of reacting components. In other words, at a ﬁxed pH condition, concentrations of fulvic acid may or may not inhibit the interactions between metal ions and oxide surfaces.

The Cu concentrations obtained from ICP and ion selective electrode in ﬁltrate were Cu in dissolved ([CuL] + [Cu(II)]) and free ð½Cu2þ_{Þ types. The complex}

concentration ([CuL]) was treated as the differences between dissolved and free concentrations. The Fig. 2. Adsorption of metal ions onto g-Al2O3as a function of pH; (a) Cu and (b) Pb.

(5)

Fig. 3. Adsorption of metal ions onto g-Al2O3in systems containing fulvic acids: (a) Cu and (b) Pb.

Fig. 4. Concentration distribution of dissolved and sorbed Cu species as a function of pH in a system containing fulvic acid: (a) 1 mg C/L, (b) 5 mg; C/L and (c) 10 mg C/L.

(6)

difference between total Cu(II) ð½Cu2þ þ ½CuL þ ½S L Cu þ ½SOCu þ ½SOCuOHÞ concentration and Cu(II) was the adsorbed Cu(II) ð½S L Cu þ ½SOCu þ ½SOCuOHÞ concentration. The adsorbed Cu concentration can be classiﬁed as the direct-adsorption of Cu(II) ð½SOCu þ ½SOCuOHÞ) and the ternary-complex of Cu(II) ([S–L–Cu]). The experimental results shown in Figs. 4a–c present the effects of 1, 5, and 10 mg C=L fulvic acid addition on the Cu species distribution under pH 4–6. Figs. 4a–c indicate that complexed Cu was 5–10%, 3–10%, and 10–20% and the free Cu2þ was 70–15%, 75–15%, and 60–10%, respectively. Furthermore, Figs. 4a–c also reveal that under pH > 6:0; pH > 5:3; and pH > 6:2 conditions, the amounts of complexed Cu(II) were larger than free Cu2þ; indicating that the importance of natural complexing ligands in controlling the fate and toxicity of metal species. Airton and Wilson [10] reported that the major

variable which controls the interactions between protons and naturally occurring organic matter is hydrogen ion activity. This thus causes a sensible inﬂuence in the chemical species of metals and toxicity in an aquatic body. Figs. 5a–c display the effects of 1, 5, and 10 mg C=L fulvic acid addition on the Pb species distribution under pH 4–6. The complexed Pb was 0–10%, 0–15%, and 10– 25% and adsorbed Pb was 0–60%, 0–85%, and 0–70%, respectively. Notably, with the addition of fulvic acid, the adsorbed-type Pb appear to increase with increased fulvic acid (up to 5 mg C=L) and then decreased at 10 mg C=L: This may be attributed to the higher fulvic acid concentration, which in turn caused more Pb complex formation. Subsequently, this then generated the higher Pb concentration in ternary-complex type ([S–L–Pb]) and enhanced the amount of adsorbed Pb under 1 and 5 mg C=L fulvic acid addition (Figs. 5a and b), but more complexed Pb was grabbed in solution at

Fig. 5. Concentration distribution of dissolved and sorbed Pb species as a function of pH in a system containing fulvic acid: (a) 1 mg C/L, (b) 5 mg; C/L and (c) 10 mg C/L.

(7)

even higher fulvic concentration as 10 mg=L: Davis [26] conﬁrmed that a higher dissolved organic compound (DOC) concentration enhanced the increase in Cu uptake at low pH. Conversely, on the surface of g-Al2O3with a 10 mg C=L fulvic acid addition, the high

concentration of Pb complex formation in a liquid phase decreased the generation of adsorbed Pb (Fig. 5c).

This study suggests that the sites of the g-Al2O3

surface for Pb ion adsorption may differ from those of fulvic acid adsorption as shown in Fig. 3b. This indicates that fulvic acid adsorption did not block the Pb complexing sites. Rather, adsorbed fulvic acid provided additional sites for Pb ﬁxation on the surface of g-Al2O3:

Adsorption of metal ions is likely to be promoted via complexation with sorbed fulvic acid as well as electrostatic attraction. Davis [26] explored the adsorp-tion of Cu(II) on alumina (50–1000 mg=L) suspended in 0.01 M NaCl solution as a function of pH (3.5–9) and organic ligand concentration (4.7–46:8 mg=L DOC) and indicated that the increase in Cu adsorption at low pH was due to Cu complexation by adsorbed organic matter. The magnitude of these effects is inﬂuenced by both the DOC and alumina concentrations. The experimental results of the present research implied that the organically complexed metal cations are maintained by the surface of g-Al2O3through ion bridging between

the negatively charged surface and fulvic acid, which further enriches metal cations on surface.

3.4. TLM simulation of fulvic acid þ Cu=g-Al2O3and

fulvic acid þ Pb=g-Al2O3systems

TLM was employed to simulate the interactions of fulvic acid and metal ions at the g-Al2O3=water

inter-face. Protonation of reacting surface sites, adsorption of background electrolyte ions, fulvic acid, metal ions, and fulvic acid-metal ions complexes on the surface of g-Al2O3; and complexation of fulvic acid and metal ions in

liquid phase were all modeled simultaneously within TLM analysis.

Hohl and Stumm [14] described the Pb=g-Al2O3

system as the exchange of surface proton with Pb as well as a subsequent formation of SOPbþ surface complex. The intrinsic coordination constant ðlog Kint

PbÞ

was calculated as 2:2: Chang et al. [27] illustrated that the reactions of the Cu=g-Al2O3system were the same as

the Pb=g-Al2O3 system and the intrinsic coordination

constant ðlog K_CuintÞ was determined as 0.09. The reactions and intrinsic conditional equilibrium constants of the Cu=g-Al2O3 and Pb=g-Al2O3 systems are presented as

follows: SOH þ Cu2þ¼ SOCuþþ Hþ; K_Cuint¼½SOCu þ_½Hþ ½SOH½Cu2þexp j₀F RT ; ð5Þ SOH þ Pb2þ¼ SOPbþþ Hþ_; K_Pbint¼½SOPb þ_½Hþ ½SOH½Pb2þexp j₀F RT : ð6Þ

Davis [19] investigated the complexation of S–L and metal ions as:

S2Lzþ Cu2þ¼ S2L2Cu2z; Kads Cu ¼ ½S2L2Cu2z ½S2Lz½Cu2þ; ð7Þ S2Lzþ Pb2þ_{¼ S2L2Pb}2z_; K_Pbads¼½S2L2Pb 2z ½S2Lz_½Pb2þ; ð8Þ

The fulvic acid reactions in these systems can be categorized into two portions. Firstly, the adsorbed fulvic acid complexed with Cu2þ _{and Pb}2þ _{on the}

surface of g-Al2O3 (Eqs. (7) and (8)). Secondly, fulvic

acid complexed with Cu2þ and Pb2þ in liquid phase (Eqs. (9) and (10)). Cu2þþ Lz_{¼ CuL}2z_; _Kcomp Cu ¼ ½CuL2z ½Cu2þ_½Lz ð9Þ Pb2þþ Lz¼ PbL2z; K_Pbcomp¼ ½PbL 2z ½Pb2þ½Lz: ð10Þ Davis [19] and Dalang et al. [25] illustrated that the characteristics of the complexation reaction within an adsorbed state is similar to those of the corresponding reaction within a solution. Houng [28] explored the conditional stability constants of Cu2þ and Pb2þ with fulvic acid at various pH levels at the ionic strength ¼ 0:01 M; fulvic acid ¼ 10 mg C=L; and 251C: The con-ditional stability constants determined are listed in Table 2 for TLM simulation in this study.

Table 2

The conditional stability constant of Cu and Pb with fulvic acid in aqueous phase

pH Log K_CuComp Log K_PbComp

3.80 3.46 2.66 4.00 3.58 2.78 4.20 3.69 2.91 4.40 3.81 3.03 4.60 3.92 3.15 4.80 4.04 3.28 5.00 4.16 3.40 5.20 4.27 3.53 5.40 4.39 3.65 5.60 4.50 3.78 5.80 4.62 3.90 6.00 4.74 4.03 6.20 4.85 4.15

(8)

The molecular weight of the fulvic acid offered by IHSS is 781 g=mole and its organic carbon content is 53%. The molar concentrations ðmMÞ of dissolved and sorbed fulvic acid in g-Al2O3 systems can be obtained

from the results of Fig. 1. Davis [19] veriﬁed that the adsorbed of Cu and Pb on the surface of oxide was dominant in [S–L–Cu] and [S–L–Pb], and that it could be neglected in [SOCu], [SOPb], [SOCuOH], and [SOPbOH]. Eqs. (7) and (8) determine the conditional stability constants of sorbed fulvic acid with Cu2þand Pb2þ: Within fulvic acid þ Cu=g-Al2O3 system, ½Cu2þ

and [S–L–Cu2z] were determined from Figs. 4a–c, thus, the conditional stability constant of sorbed fulvic acid with Cu2þ can be calculated (Table 3). The same procedure was used to calculate Kads

Pb: It is fairly

important to note that differences existed between K_Cuads in fulvic acid ¼ 1 mg C/L and fulvic acid ¼ 5 and 10 mg C=L: That is, fulvic acid in the former concen-tration was insufﬁcient to cover the entire surface of g-Al2O3; which provided reacting sites that adsorbed Cu.

Rebhn [29] also illustrated that clay minerals provided numerous reacting sites for adsorption at low organic content (o 0.5%). The Kads

Cu values in the fulvic acid ¼ 5

and 10 mg C=L concentrations were similar. This study postulated that all g-Al2O3reacting sites were covered by

fulvic acid in these concentrations. Therefore, the adsorbed fulvic acid of g-Al2O3 surface controlled the

adsorption behaviors. Davis [26] indicated that a sig-niﬁcant fraction of the alumina surface was covered by adsorbed organic matter. Cu(II) was partitioned primarily between the surface-bound organic matter and dissolved Cu-organic complexes in the aqueous phase. The variation in the adsorption density of organic matter had a signiﬁcant effect on Cu(II) complexation at the surface.

The parameters of Kint

Cu ð0:09Þ; KPbint ð2:2Þ; K comp

Cu and

K_Pbcomp; Kads

Cu and KPbads; as well as the surface reactions

were applied simultaneously in the TLM analysis to simulate the experimental data. Figs. 6a–d display the simulation results of fulvic acid þ Cu=g-Al2O3 and

fulvic acid þ Pb=g-Al2O3 systems at varying pH levels

and fulvic acid concentrations. Symbols and lines represent experimental data and TLM simulation results, respectively. As a whole, the simulation results of free metal ions were higher than those in the experimental data. Conversely, the complex metal ion concentrations, obtained from TLM simulation, were lower than the experimental data. The minor unfitness (lower estimate of free metal; higher estimate of complexed metal) between the simulations and experi-mental results are probably due to the biased estimates on the conditional stability constants. In addition, competition for metal between fulvic acid and oxide also plays a significant role for this. Vermeer et al. [24] has reported that if metal-humic affinity is greater than metal-oxide, the overall adsorption will be less than model prediction. The effects of surface potential between [S–L] and metal ions on the surface of g-Al2O3 may also explain the differences between the

TLM simulation results and the experimental data. Therefore, TLM simulated the complexation of [S–L] and metal ions as a reaction within a solution, but not on the g-Al2O3 surface, and resulted in the

aforemen-tioned differences. Modelling the metal binding beha-vior in a heterogeneous system (NOM/oxide) is different from that in simple solution metal-ligand reaction. The heterogeneity and various afﬁnity site of NOM render the metal sorption results difﬁcult to model.

4. Conclusions

This study investigated the interfacial reactions between fulvic acid and metal ions on the surface of g-Al2O3: Ion selective electrodes were employed for metal

ions measurements, which in turn established the equilibrium relations of fulvic acid and metal ions. Using conditional stability constants, this study illu-strated the equilibrium phenomena in heterogeneous systems at various pH levels. The experimental data were modeled by TLM to simulate both the complexa-tion and adsorpcomplexa-tion of metal ions and fulvic acid on g-Al2O3 surface. Generally, fulvic acid enhanced the

sorption of metal ions. In systems containing 5 and 10 mg C=L fulvic acids, sorption of Cu and Pb are enhanced at the pH range studied. However, low fulvic acid concentration might inhibit Cu sorption. Modeling concentration distributions of dissolved and sorbed metal speciations as a function of pH were consistent with the experimental observations. Simulation results indicated that metal species are dominantly in com-plexation with fulvic acid, both in solution and at the g-Al2O3 surface. Dissolved organic matters have a

Table 3

Conditional stability constant of Cu with sorbed fulvic acid pH FA 1 mg C/L FA 5 mg C/L FA 10 mg C/L

Log Kads Log Kads Log Kads

3.80 5.14 4.23 4.38 4.00 5.28 4.39 4.53 4.20 5.41 4.55 4.69 4.40 5.54 4.71 4.84 4.60 5.68 4.87 5.00 4.80 5.81 5.03 5.15 5.00 5.95 5.19 5.30 5.20 6.08 5.35 5.46 5.40 6.22 5.51 5.61 5.60 6.35 5.67 5.76 5.80 6.48 5.83 5.92 6.00 6.62 5.99 6.07 6.20 6.75 6.14 6.23

(9)

Fig. 6. Modeling concentration distribution of dissolved and sorbed Cu and Pb species as a function of pH in a system containing fulvic acid (a) Cu species as a function of pH in a system containing 5 mg C/L of fulvic acid, (b) Cu species as a function of pH in a system containing 10 mg C/L of fulvic acid, (c) Pb species as a function of pH in a system containing 5 mg C/L of fulvic acid and (d) Pb species as a function of pH in a system containing 10 mg C/L of fulvic acid.

(10)

profound role in controlling mobility, bioavailability and concentration distribution of metal ions in the aquatic environment.

Acknowledgements

This study was funded by the National Science Council, Taiwan, under contract NSC 89-2211-E-002-011.

References

[1] Stumm W. Chemistry of the solid–water interface. New York, NY, U.S.A: Wiley, 1992.

[2] Oden WI, Amy GL, Conklin M. Subsurface interactions of humic substances with Cu (II) in saturated media. Environ Sci Technol 1993;27:1045–51.

[3] Dumat C, Quiguampoix H, Staunton S. Adsorption of cesium by synthetic clay-organic matter complexes: effect of the nature of organic polymers. Environ Sci Technol 2000;34:2985–9.

[4] Pinheiro JP, Mota AM, Benedetti MF. Effect of aluminum competition on lead and cadmium binding to humic acids at various ionic strength. Environ Sci Technol 2000;34: 5137–43.

[5] Liu AG, Gonzalez RD. Modeling adsorption of copper (II), cadmium(II) and lead(II) on puriﬁed humic acid. Langmuir 2000;16:3902–9.

[6] Jin X, Bailey GW, Yu YS, Lynch AT. Kinetics of single and multiple metal ion sorption processes on humic substances. Soil Sci 1996;161:509–20.

[7] Sebastien S, William H, Herbert EA. Solid-solution partitioning of metals in contaminated soils: dependence on pH, total metal burden, and organic matter. Environ Sci Technol 2000;34:1125–31.

[8] Neubauer V, Nowack B, Furrer G, Schulin R. Heavy metal sorption on clay minerals affected by the side-rophore desferrioxamine B. Environ Sci Technol 2000;34: 2749–55.

[9] Alcacio TE, Hesterberg D, Chou JW, Martin JD, Beauchemin S, Sayers DE. Molecular scale characteristics of Cu(II) bonding in goethite-humate complexes. Geochim Cosmochim Acta 2001;65:1355–66.

[10] Airton K, Wilson FJ. Complexation and adsorption of copper in raw sewage. Water Res 2000;34:2061–8. [11] Plavsic M, Cosovic B. Voltammetric study of the role of

organic acids on the sorption of Cd and Cu ions by alumina particles. Colloid Surface A 1999;151:189–200. [12] Frimmel FH, Huber L. Inﬂuence of humic substances on

the aquatic adsorption of heavy metals on deﬁned mineral phases. Environ Int 1996;22:507–17.

[13] Gagnon C, Arnac M, Brindle JR. Sorption interactions between trace metals (Cd and Ni) and phenolic substances on suspended clay minerals. Water Res 1992;26:1067–72. [14] Hohl H, Stumm W. Interaction of Pb2þ with hydrous

g-Al2O3: J Colloid Interface Sci 1976;55:281–8.

[15] Hayes KF, Leckie JO. Modeling ionic strength effects on cation adsorption at hydrous oxide/solution interfaces. J Colloid Interface Sci 1987;115:564–72.

[16] Chang KS, Lin CF, Lee DY, Lo SL, Yasunaga T. Kinetic of Cr(III) adsorption/desorption at the g-Al2O3/water

interface by the pressure-jump technique. J Colloid Inter-face Sci 1994;165:169–76.

[17] Peri JB. Infrared and gravimetric study of the surface hydration of r-alumina. J Phys Chem 1965;69:211–9. [18] Hayes KF, Redden G, Ela W, Leckie JO. Surface

complexation models: an evaluation of model parameter estimation using FITEQL and oxide mineral titration data. J Colloid Interface Sci 1991;142:448–69.

[19] Davis JA. Adsorption of nature dissolved organic matter at the oxide/water interface. Geochim Cosmochim Acta 1982;46:2381–91.

[20] Schroth BK, Sposito G. Effect of landﬁll leachate organic acids on trace metal adsorption by kaolinite. Environ Sci Technol 1998;32:1404–8.

[21] Zhou JL, Rowland S, Mantoura RFC, Braven J. The formation of humic coatings on mineral particles under simulated estuarine conditions—a mechanistic study. Water Res 1994;28:571–9.

[22] Gu B, Schmitt J, Chen Z, Liang L, McCarty JF. Adsorption and desorption of natural organic matter on iron oxide: mechanisms and models. Environ Sci Technol 1994;28:38–46.

[23] Liu AG, Gonzalez RD. Adsorption/desorption in a system consisting of humic acid, heavy metals, and clay minerals. J Colloid Interface Sci 1999;218:225–32.

[24] Vermeer AWP, Mcculloch JK, Van Riemsdijk WH, Koopal LK. Metal ion adsorption to complexes of humic acid and metal oxides: deviation from the additivity rule. Environ Sci Technol 1999;33:3892–7.

[25] Dalang F, Bufﬂe J, Haerdl W. Study of the inﬂuence of fulvic substances on the adsorption of copper (II) ions at the kaolinite surface. Environ Sci Technol 1984;18:135–41. [26] Davis JA. Complexation of trace metals by adsorbed natural organic matter. Geochim Cosmochim Acta 1984;48:679–91.

[27] Chang KS, Lin CF, Lee DY, Lo SL. Mechanistic investigation of copper sorption onto g-Al2O3 using the

relaxation technique. Chemosphere 1993;27:1397–407. [28] Houng LM. Complexation of fulvic acid with copper ion.

MSc Thesis, Graduate Institute of Environmental En-gineering, National Taiwan University, Taipei, Taiwan, 1991.

[29] Rebhn M. Sorption of organics on clays and synthetic humic-clay complexes simulating aquifer process. Water Res 1992;26:79–84.