Comparison of zinc complexation properties of dissolved natural

organic matter from different surface waters

Tao Cheng *, Herbert E. Allen

Center for the Study of Metals in the Environment, Department of Civil and Environmental Engineering, University of Delaware, Newark, DE 19716, USA

Received 20 January 2005; received in revised form 23 July 2005; accepted 5 September 2005

Available online 9 December 2005

Abstract

The zinc binding characteristics of natural organic matter (NOM) from several representative surface waters were studied and compared. NOM

samples were concentrated by reverse osmosis. The samples were treated in the laboratory to remove trace metals. Square wave anodic stripping

voltammetry (SWASV) was used to study zinc complexing properties of those NOM samples at fixed pH, ionic strength, and dissolved organic

carbon (DOC) concentrations. Experimental data were compared to the predictions from the Windermere Humic Aqueous Model (WHAM)

Version VI. At the same pH, ionic strength, and temperature, the zinc titration curves for NOM samples from different surface water sources tested

in our study almost overlapped each other, indicating similarity in zinc binding properties of the NOM. A discrete two-site model gave good fits to

our experimental titration data. Non-linear fitting by FITEQL 4.0 shows that the conditional zinc binding constants at the same pH are similar for

NOM from different sources, indicating that zinc complexation characteristics of the NOM used in our study do not depend on their origin and one

set of binding parameters can be used to represent Zn-NOM complexation for NOM samples from those different surface water sources

representing geographically diverse locations. In addition, the total ligand concentrations (L1,T, L2,T, and LT) of all NOM show no observable

gradation with increasing pH (L1,TZ2.06G0.80 mmol/g carbon; L2,TZ0.12G0.04 mmol/g carbon; LTZ2.18G0.78 mmol/g carbon), while the

conditional binding constants of zinc by NOM ðlog KcZnLÞ show a linear increase with increasing pH ðlog Kc

1ðpHZ6:0ÞZ4:69G0:25;

log Kc1ðpHZ7:0ÞZ4:94G0:10; log Kc

1ðpHZ8:0ÞZ5:25G0:006; log Kc2ðpHZ6:0ÞZ6:29G0:13; log Kc

2ðpHZ7:0ÞZ6:55G0:08; log Kc2ðpH

Z8:0ÞZ6:86G0:023Þ with a slope of ca. 0.28, indicating the zinc-NOM complexes become more stable at higher pH. The WHAM VI predicted

free zinc ion activities at high zinc concentrations agree with our experimental results at pH 6.0, 7.0, and 8.0. However, the zinc binding of these

NOM samples is over estimated by WHAM VI at zinc concentrations below 10K6 M at pH 8.0.

q 2005 Elsevier Ltd. All rights reserved.

Keywords: Zinc complexation; Natural organic matter; Anodic stripping voltammetry

1. Introduction

Zinc (Zn) is a common element occurring naturally in the

environment and it is widely used by humans for domestic and

industrial purposes. Zinc is an essential element

and micronutrient required for normal growth by plants and

animals. At both high and low concentrations zinc can be

detrimental to organisms. In uncontaminated waters zinc

concentration is usually very low and can span a wide range

from 10K10 to 10K6 M (Stumm and Morgan, 1996). Both

human activity and natural processes have inevitably increased

the level of zinc concentrations in some natural water systems

and high concentrations of zinc that are toxic or even lethal to

0301-4797/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.

doi:10.1016/j.jenvman.2005.09.007

* Corresponding author. Tel.: C1 626 395 4385; fax: C1 626 395 2940.

E-mail address: tcheng@ce.udel.edu (T. Cheng).

organisms have been observed, which has caused great

environmental concern.

The speciation of zinc in natural waters is a critical factor to

consider when assessing the environmental impact of zinc. The

bioavailability, toxicity, transport and fate of zinc in the aquatic

environment, and water quality criteria have been recognized

as a function of water chemistry (Allen and Hansen, 1996).

Complexation of zinc by natural organic matter (NOM) has

important influence on the speciation of zinc in various natural

waters. The complexation of Zn by NOM in natural waters can

markedly lower the free Zn2C activity relative to total

dissolved Zn, leaving only a small fraction of the total zinc

as ‘free’ zinc, which is considered to be bioavailable, or toxic

(Allen and Hansen, 1996). Therefore, in order to understand

zinc toxicity in water bodies, we need to understand the

characteristics of zinc binding by NOM.

Modeling is an essential tool in quantifying the speciation of

metals in the environment. There are a number of models

Journal of Environmental Management 80 (2006) 222–229

www.elsevier.com/locate/jenvman

Table 1

Metal concentrations, dissolved organic carbon (DOC), dissolved inorganic

carbon (DIC), and percentage of fulvic and humic acids of NOM

Conc., (mg/L)

(with 10 mg

DOC/L)

Na K Ca Mg

SC-NOM 74.5 0.775 0.1071 0.045

NY-NOM 121.5 0.597 0.0813 0.017

GA-NOM 0.081 0.565 0.0372 !0.0018

Conc. (mg/L)

(with 10 mg

DOC/L)

Ni Cu Zn Cd Pb

SC-NOM !2.01 2.30 1.24 !0.68 !0.5

NY-NOM !2.01 3.62 0.43 !0.68 !0.5

GA-NOM !2.01 3.79 !0.10 !0.68 !0.5

DOC (mg/L) IC (mg/L) Percentage of HA and FA (%)

HA FA

SC-NOM 564.1 2.322 19 81

NY-NOM 331.8 1.006 10 90

GA-NOM 931.3 0.932 5 95

T. Cheng, H.E. Allen / Journal of Environmental Management 80 (2006) 222–229 223

available that predict metal complexation with inorganic and

organic compounds with defined chemical nature. NOM has

not been very well characterized due to its complicated nature,

although the main functional groups that bind to metals are

known. There exist a few models that model metal binding to

NOM (WHAM, NICA) (Tipping, 1994; 1998; Benedetti et al.,

1995). The Biotic Ligand Model (BLM), which uses WHAM

to compute organic speciation, was recently proposed to

predict acute toxicity of metals to aquatic organisms (Di Toro

et al., 2001; Santore et al., 2001). The key assumption of all the

above models is that metal complexation characteristics of

NOM do not depend on their origin (Tipping, 1994; 1998;

Benedetti et al., 1995; Di Toro et al., 2001; Santore et al., 2001;

Lu and Allen, 2002). Although NOM exhibit site specificity in

metal binding (Sarathy, 2002), there is evidence that copper

binding properties of NOM from surface water sources over

large spatial and temporal scales are similar (Lu and Allen,

2002). This is probably because the main mechanisms of metal

binding (metal complexation with carboxylic groups and

phenolic groups) in these NOM are similar. However, only

very limited literature is available on zinc binding by NOM. In

recent years, zinc binding by NOM has been studied by a range

of techniques including Donnan membrane (Oste et al., 2002),

cation ion-exchange (Fortin and Campbell, 1998), resin

equilibrium (Christensen and Christensen, 1999; 2000), and

voltammetry (Jansen et al., 1998; Xue and Sigg, 1994).

However, to our knowledge, studies that compare binding

characteristics of NOM from different surface water sources

have not been reported for Zn. This paper reports data from

experiments conducted on NOM samples from three surface

water sources to determine zinc complexing properties. The

data are analyzed to ascertain if all organic matters can be

generalized to behave in a similar fashion to complex zinc.

2. Materials and methods

All reagents used were analytical grade except the acids,

which were Optima grade. Unless otherwise mentioned, all

reagents were obtained from Fisher Scientific (Pittsburgh; PA;

USA) ‘Better’ buffers (Kandegedara and Rorabacher, 1999) of

MES (for pH 6.0), MOPS (for pH 7.0) and PIPBS (for pH 8.0)

were used in zinc titrations to keep the pH constant. The buffers

were added to samples to achieve a 0.01 M concentration of the

buffers. During the titrations, 0.1 M NaOH or 0.1 M HNO3 was

added as required to keep the pH change withinG0.1 pH unit.

Distilled de-ionized water was used in all experiments, for all

dilutions, and for blanks.

In order to test whether a single model of metal-NOM

complexation is adequate, or whether typically observed

variations in the characteristics of NOM samples from different

sources are sufficient to require site-specific chemical

characterizations or models, three sites were chosen as sources

of NOM in order to include geographically diverse locations,

ones that are likely to provide NOM samples that vary in

composition and chemical behavior. NOM was sampled from

the Big Moose Lake, a high elevation system in the Adirondack

Mountains of New York State, in May 2000; from the Edisto

River, a typical receiving water in South Carolina with a much

larger watershed and longer residence time, in March 2001;

and from the Suwannee River, Georgia, in June 1997.

Procedures used are described in detail in an earlier publication

(Ma et al., 2001). The source water was filtered through a

0.45 mm pore size filter and the samples were concentrated in

the field using a reverse osmosis (RO) unit (Model PROS/2S,

RealSoft, Norcross, GA) (Serkiz and Perdue, 1990). The

samples were stored in coolers with ice in the field and in a

refrigerator at 4 8C in the laboratory. The concentrated NOM

samples were passed through a HC-saturated cation-exchange

resin (Dowex 50WX8, Fluka Chemical Co., Milwaukee, WI)

column to remove both trace metals and major cations. To

avoid losing the humic acid (HA) fraction of the NOM on the

resin due to the strongly acidic condition, HA was separated in

advance by acidic precipitation (pHz1) and was later

recombined with the material that passed through the cation-

exchange column. The NOM samples thus treated were used in

all the subsequent experiments.

2.1. Characterization of NOM

We determined the concentration of the DOC (Dissolved

Organic Carbon) and the DIC (Dissolved Inorganic Carbon) of

the concentrated NOM samples using a Tekmar-Dohrman DC-

190 TOC analyzer. The metal concentration of the diluted

samples was analyzed using Inductively Coupled Plasma-

optical emission spectroscopy (ICP-OES) (Spectro Analytical

Instrument, Kleve, Germany). The DOC, DIC and the metal

concentrations of the NOM samples are reported in Table 1.

2.2. Zinc titrations

The NOM samples were also titrated against zinc using

square wave anodic stripping voltammetry (SWASV) over

a range of total zinc concentrations ranging from 10K7 to

T. Cheng, H.E. Allen / Journal of Environmental Management 80 (2006) 222–229224

10K4 M. Voltammetric measurements were performed using

an Analytical Instrument Systems AIS Model DLK-100A

electrochemical analyzer with an EG&G Princeton Applied

Research PARC 303A static mercury drop electrode in the

hanging mercury drop electrode (HMDE) mode. The mercury

drop size was ‘large’ with a surface area of 2.83 mm2. The

reference electrode was Ag/AgCl/3M KCl. For each measure-

ment, a new mercury drop was extruded and the sample

solution to be measured was purged with ultrapure (grade 5.0)

N2 for 4 min to eliminate the possible interference of O2. In

addition, the head space of the voltammetric cell was filled

with nitrogen after purging so that possible dissolution of CO2

from the atmosphere into the solution during the measurement

was minimized. SWASV mode was used to measure labile

(electro-active) Zn concentration. For each SWASV measure-

ment the deposition time was 30 s and deposition potential was

K1.500 V without stirring; an equilibration time of 15 s

followed; for the stripping step, the square wave mode was

used, the pulse height was 0.025 V and the potential scan began

from K1.500 V and ended at K0.800 V at a scan rate of

50 mV/s. SC, NY, and GA NOMwere titrated with Zn at 3 pHs

K6.0, 7.0, and 8.0. ‘Better’ buffers were used to control the pH

to the required value. These ‘better’ buffers do not complex

metal ions and thus do not interfere with the titrations

(Kandegedara and Rorabacher, 1999). For each titration,

10 mg/L of the dissolved organic carbon (DOC) was prepared

by dilution of the concentrated NOM. One M NaNO3 solution

was added to adjust the ionic strength to 0.02 M. Titrations

were conducted in a clean room at a constant temperature of

w22 8C. The reactions were allowed to stabilize after each

addition of zinc for at least 4 min. Our experiments on Zn-

NOM complexation kinetics show that the complexation

reactions between Zn and NOM reach equilibrium within

4 min under our experimental conditions (data not shown).

2.3. Computation of free zinc activities

The peak current measured by SWASV, Ip, which is

proportional to the labile fraction of metal in the voltammetric

measurement, is a weighted average of the diffusion of all

metal species (freeCcomplexed). For a fully labile system

(that is, a system in which all the relevant metal species (Zn and

ZnL in our system) are electroactive), the peak current Ip is

expressed as (van Leeuwen et al., 1989; De Jong and van

Leeuwen, 1987a,b,c):

Ip ZKpK1=2nFA �D1=2C�M;Tt

K1=2 (1)

where F is the Faraday constant, A is the electrode surface area,

n is the number of moles of electrons transferred per mole of

metal oxidized or reduced, C�M;T is the total soluble metal

concentration in the bulk solution, t is characteristics time,

which is constant in our experiments, and �D is the weighted

average of the diffusion coefficient of all metal species (freeCcomplexed), which is expressed as (van Leeuwen et al., 1989;

De Jong and van Leeuwen, 1987a,b,c):

�DZ½M��

C�M;T

DM C½ML��

C�M;T

DML (2)

where [M]* is the free metal ion concentration in bulk solution,

[ML]* is the complexed metal concentration in bulk solution,

DM and DML are the diffusion coefficients of the free and

complexed metal ion.

The mass balance of metal in bulk solution is,

½M�� C ½ML�� ZC�M;T (3)

Defining normalized current F as the ratio of peak current in

the presence of ligands to that of a ligand-free reference,

FZILp

IpZ

�D

DM

� �1=2

(4)

where F is the normalized current, ILp is the peak current in the

presence of ligands and Ip is that of the ligand-free reference.

Combining Eqs. (2)–(4), the free metal ion activity in bulk

solution is expressed as,

½M�� Z

ILpIp

� �2

KDML

DM

� �

1KDML

DM

� � C�M;T (5)

Eq. (5) is used to compute free zinc ion concentration for fully

labile Zn-NOM systems in our experiments. Normalized

current is obtained by comparing the peak current in the

presence of ligands and that of the ligand-free reference. Total

zinc concentration was determined by ICP. For a fully labile

system, the value of DML/DM can be estimated under

conditions when the ligand concentration is in large excess

of the total zinc concentration so that [M]*/[ML]* and the

weighted average �D tends to DML (Eq. (2)).

2.4. Models

To compute speciation of zinc in aquatic media that contain

NOM, a chemical equilibrium model is used. The simplest

metal complexation with homogeneous ligands can be

represented by:

Zn2CCLZZnL KcZnL Z

½ZnL�

½Zn2C�½L�(6)

where Zn represents the ‘simple’ (or more exactly, hydrated)

zinc ion, L is the ligand, ZnL is the complex formed between

Zn2C and L. The conditional zinc binding constant KcZnL is only

valid at constant pH and ionic strength. To simplify the

discussion, only the 1:1 ZnL complex is considered. While

other stoichiometry (other than 1:1) between Zn and ligand (L)

is possible, it is generally valid to assume the stoichiometry of

the complexation reaction between Zn and NOM is 1:1, since it

has been shown that the majority of the bond formed between

metal ion and NOM is monodentate and it has been shown by a

number of studies that this assumption is reasonable in

modeling metal ion and NOM complexation (Tipping, 1998;

Bugarin et al., 1994; Pinheiro et al., 1994).

Fig. 1. Estimation of DZnL/DZn value by titration of Zn with GA-NOM.

Titration of total Zn concentration of 7.33!10K7 M at pHZ7.0, IZ0.02 M,

TZ25 8C. The estimated DZnL/DZn value was 0.014.

T. Cheng, H.E. Allen / Journal of Environmental Management 80 (2006) 222–229 225

However, in order to represent the complexation reaction of

zinc with NOM, the heterogeneity property of NOM must been

taken into consideration. To this end, discrete multi-site models

are usually used. The simplest case of the discrete multi-site

model is the two-site model (van den Berg, 1984), represented

by:

Zn2CCLi ZZnLi KcZnLi

Z½ZnLi�

½Zn2C�½Li�(7)

Li;T ZLi CZnLi (8)

where KcZnLi

is the conditional stability constant, valid only at

constant pH and ionic strength, Li is free binding sites which

means the sites are not bound by zinc, Li,T is total binding sites.

iZ1, 2 is an index, representing two distinguishable ligands

present in NOM. Usually one is carboxyl group and the other

one is the phenolic group. The discrete two-site model (or

discrete multi-site model) has been successfully applied to

describe metal interaction with organic matter in a number of

conditions (van den Berg, 1984; Hering and Morel, 1988).

A fitting model, FITEQL 4.0 (Herbelin and Westall, 1999)

was used to calculate stability constants of ligand-zinc

complexes, and ligand concentrations. For the zinc titration

data, a 2-site model gives a good fit so it was adopted in our

modeling approach.

WHAM (Windermere Humic Aqueous Model) Version VI

(Tipping, 1998) was used to calculate zinc complexation with

NOM. The Zn binding constant of fulvic acid is log KMAZ1.6,

and the Zn binding constant of humic acid is log KMAZ1.5 in

WHAM Version VI. The free zinc data predicted by WHAM

were compared to experimental data to determine if WHAM

gave an accurate description of Zn-NOM complexation.

3. Results and discussion

DOC was determined on samples both before and after the

precipitation of the humic acid. All the organic matter that is

not humic acid is considered fulvic acid when calculating

speciation using WHAM. The DOC, IC (Inorganic Carbon)

and percentage of HA (Humic Acids) and FA (Fulvic Acids)

thus measured are reported in Table 1. The NOM from Edisto

River, SC, Big Moose Lake, NY and the Suwannee River, GA

were found to contain 81, 90 and 95% fulvic acid, respectively.

3.1. Zinc titrations

Concentrated SC, NY, and GA-NOM were added to a

solution with a fixed total Zn concentration of 7.37!10K7 M

buffered at pHZ7.0. For each addition of NOM, a voltam-

metric measurement was made and normalized current and

peak potential was plotted against DOC concentration. The

titration curve for GA-NOM is shown in Fig. 1. With the first

addition of NOM (DOCz20 mg/L), the normalized current

dropped from 1.0 to about 0.45, indicating a large fraction of

zinc was complexed by NOM and the diffusion coefficient of

the NOM complexed zinc was much lower than that of the free

zinc ion (Eq. (4)). When more NOM was added, however, the

decrease in the normalized current became more gradual and

the normalized current attained a limiting value at high DOC

concentrations. During the same titration the peak potential

tended to more negative values with addition of NOM

(At DOCZ0, the peak potential was K1.16 V, at DOCZ300 mg/L, the peak potential was K1.22 V, while at DOCZ600 mg/L, the peak potential was K1.26 V). This systematic

shift in peak potential indicates that the complex formed

between zinc and NOM (ZnL) is labile and the decrease in peak

current is due to a lower diffusion coefficient of the labile

complex (ZnL) compared to that of the free zinc ion (DZnL!DZn), not due to the presence of inert (non-labile) complexes

(Jansen et al., 1998; Cleven and Leeuwen, 1986; van Leeuwen

et al., 1989). The DZnL/DZn values which were determined for

the NOM samples using Eq. (4) are: SC-NOM, 0.04; NY-

NOM, 0.06; GA-NOM, 0.014. These values of DZnL/DZn are

close to the reported value of 0.05 (Jansen et al., 1998). In

addition, for the purpose of computing free Zn ion, Jansen et al.

(1998) showed that the influence of the value of DZnL/DZn on

the resulting free Zn ion is very small. This was also confirmed

by our calculation using Eq. (5).

Zinc was added to SC, NY, and GA-NOM samples having

10 mg/L DOC and the labile zinc ion was determined

following each change in total zinc. The titration curve of the

SC-NOM is shown in Fig. 2. For the same total zinc

concentration, a decrease in peak current, which is in

proportional to the labile zinc concentration (Eq. (1)), was

observed with increase in pH. It was demonstrated by

MINTEQCcalculation that under our experimental conditions

free zinc ion is the dominant inorganic zinc species and the

complexes formed between zinc and inorganic ligands (mainly

OHK, ClK, and NOK3 ) are negligible. It was also demonstrated

by MINTEQCthat no zinc solid formed under our experimen-

tal conditions. So the decrease in peak current was due to

formation of zinc-NOM complexes, not due to formation of

zinc solids or zinc inorganic complexes. As discussed

previously, the zinc-NOM complexes in our system were

labile; so the free zinc activities in our titration can be

Fig. 3. Zn titration curves of NOM samples from different sources

Fig. 2. Zn titration curves for the SC-NOM sample at pH 6.0, 7.0 and 8.0.

DOCZ10 mg/L; IZ0.02 M; TZ25 8C.

T. Cheng, H.E. Allen / Journal of Environmental Management 80 (2006) 222–229226

computed using Eq. (5). Comparison of zinc titrations

expressed as free zinc ion against total soluble zinc

concentrations for NY, SC and GA-NOM at pH 6.0, 7.0 and

8.0 are shown in Fig. 3. All the curves with the same pH and

ionic strength almost overlap each other. This indicates that

NOM from different surface water sources are similar in

complexation of zinc. To determine whether the three sets of

data are similar or different, we based our judgment on the

difference in log [Zn2C], not [Zn2C], considering the wide

range of the metal concentrations involved (several orders of

magnitude) and the complicated nature of the NOM. In

addition, in fitting the metal-NOM binding parameters in

WHAM, the relative error of the log of metal concentration, not

metal concentration, is used to estimate the goodness of fitting

(Tipping, 1998). In modeling metal complexation with NOM,

an error of a factor of 3–4 in the free metal ion activity is

acceptable (Christensen and Christensen, 2000).

at pH 6.0, 7.0, and 8.0. DOCZ10 mg/L; IZ0.02 M; TZ25 8C.

T. Cheng, H.E. Allen / Journal of Environmental Management 80 (2006) 222–229 227

Fig. 4 shows plots that compare the WHAM VI

predictions with our experimental data. For the NOM from

different sources, WHAM predictions account for the free

zinc very well (in terms of the relative error of log [Zn2C])

at high total Zn concentrations for all pH tested in our

titrations. However, WHAM under predicts the free zinc ion

activity at low total soluble Zn concentrations at pH 8.0. This

indicates that the strong binding sites of low concentration

for zinc complexation in NOM implied by WHAM VI may

not exist. Very limited literature is available on comparison

of experimentally measured Zn species in natural waters with

WHAM simulations. Christensen and Christensen (1999,

2000) reported that WHAM Version V tends to over estimate

Zn-NOM complexation. They suggested the default Zn-NOM

stability constant in WHAM Version V, which is the ‘best

average’ from a limited number of published data, is

overestimated. By using a Zn-NOM reaction constant

Fig. 4. Comparison of WHAM VI simulation and experimental Zn titration curves f

TZ25 8C. The symbols represent experimental measurement (open circles (B): pH 6

WHAM simulation.

of 1.7, instead of the default value of 1.3; they found good

agreements between their experimentally measured free Zn

activities and WHAM Version V prediction. It should be

noted that the Zn-organic matter binding constants reported

by Christensen and Christensen (2000) are for organic

matters that originate from leachate of solid waste disposal,

which presumably are different with respect to metal binding

compared to organic matters from surface water sources. In a

recent study, it was reported that the metal binding

characteristics of organic matters from the effluents of

municipal wastewater treatment plants were very different

from those of organic matters from surface water sources

(Sarathy, 2002). Metal binding sites other than carboxylic

and phenolic groups in those organic matters from leachate

and wastewater effluent might account for the different metal

binding properties compared to those of NOM from surface

water sources.

or SC, NY, and GA-NOM at pH 6.0, 7.0 and 8.0. DOCZ10 mg/L; IZ0.02 M;

.0; open squares (,): pH 7.0; open diamonds (,): pH 8.0). The lines represent

Table 2

Conditional stability constants and site densities of Zn-NOM complexation obtained by 2-site model fit with FITEQL 4.0 (SDZstandard deviation)

Sample pH log KcZnL;1

(mol/L)K1

log KcZnL;2

(mol/L)K1

L1,T(mmol/g C)

L2,T(mmol/g C)

LT(mmol/g C)

Percent of

Li,T/LT

1 2

SC-NOM 6.0 4.66 6.21 1.69 0.07 1.76 0.96 0.04

7.0 4.82 6.46 2.54 0.11 2.65 0.96 0.04

8.0 5.24 6.87 1.83 0.16 1.99 0.92 0.08

NY-NOM 6.0 4.95 6.44 1.33 0.10 1.43 0.93 0.07

7.0 4.99 6.59 2.04 0.13 2.17 0.94 0.06

8.0 5.25 6.87 1.84 0.16 2.00 0.92 0.08

GA-NOM 6.0 4.46 6.22 3.99 0.07 4.04 0.98 0.02

7.0 5.01 6.59 1.82 0.12 1.94 0.94 0.06

8.0 5.25 6.83 1.48 0.15 1.63 0.91 0.09

Avg 6.0 4.69 6.29

SD 0.25 0.13

Relative SD, % 5.3 2.0

Avg 7.0 4.94 6.55

SD 0.10 0.08

Relative SD, % 2.1 1.1

Avg 8.0 5.25 6.86

SD 0.006 0.023

Relative SD, % 0.1 0.3

Avg. 2.06 0.12 2.18 0.94 0.06

SD 0.80 0.04 0.78 0.023 0.023

Relative SD, % 39 32 36 2.4 38

T. Cheng, H.E. Allen / Journal of Environmental Management 80 (2006) 222–229228

3.2. Zinc ligand binding constants

Table 2 lists the conditional binding constant ðlog KcZnL;iÞ

values and the concentrations of the ligands for SC, NY, and

GA-NOM obtained by FITEQL 4.0 using a 2-site model. Since

temperature and ionic strengthwere held constant, these binding

constants varied onlywith pH.Our fitting results showed that the

conditional binding constant ðlog KcZnL;iÞ values increased with

pH, indicating that the zinc-ligand complexes become more

stable at higher pH. This is what we would expect since the

decrease in competing protons at higher pH results in higher

stability for the complexes. It was also observed that the

total ligand concentrations (L1,T, L2,T, and LT) showed

Fig. 5. Conditional stability constants obtained by the 2-site model for NOM

samples versus pH. Averaged log KcZnL;i of SC, NY, and GA-NOM obtained

from the 2-site model was plotted against pH.

no gradation with pH. Plots of log of conditional stability

constants ðKcZnL;iÞ versus pH (Fig. 5) for NOM samples illustrate

that log KcZnL;i is linearly pH dependent with slopes close to

0.283 (equations shown in Fig. 5). The slopes (0.280 and 0.285)

of our linear regression of the log of conditional stability

constants and pH are close to the slope (0.276) reported by

Christensen and Christensen (2000). They reported a linear

relationship between the log of conditional stability constants

and pH as:

log KcZnL Z 0:276pHC2:581 (9)

for Zn-NOM complexation by two leachate dissolved

organic carbon samples with similar ionic strength and pH

range (IZ0.056 and 0.023 M, the pH range is 5.0–8.0).

4. Conclusions

At the same pH, ionic strength, and temperature, the zinc

titration curves for NOM samples from different surface water

sources tested in our study almost overlap each other,

indicating that zinc complexation characteristics of the surface

water NOM used in our study do not depend on their origin.

This is probably because the main mechanisms of metal

binding (metal complexation with carboxylic groups and

phenolic groups) in these NOM are similar. These observations

added strength to the assumption that one set of binding

constants and ligand concentrations can be used to represent

Zn-NOM complexation for NOM from dissimilar surface

water sources. Clearly, however, more data on zinc binding by

NOM from other sources are required to determine whether the

zinc complexation characteristics of NOM depend on their

origin, or whether the NOM studied here happen to have

similar Zn binding affinity. Titrations of the type presented here

T. Cheng, H.E. Allen / Journal of Environmental Management 80 (2006) 222–229 229

on a wider range of natural water NOM would be useful in

investigating this issue further. Comparison of titration curves

with those predicted by WHAM VI shows a good fit in the case

of all NOM at high zinc concentrations for pH 6.0, 7.0 and 8.0,

and a poor fit in the case of low zinc concentrations at pH 8.0.

Zinc binding by NOM may be over estimated in the current

version of WHAM, especially at low zinc concentrations and

high pH.

Acknowledgements

The support from the International Lead Zinc Research

Organization for this research is gratefully acknowledged.

References

Allen, H.E., Hansen, D.L., 1996. The importance of trace metal speciation to

water quality criteria. Water Environ. Res. 68, 42–54.

Benedetti, M.F., Milne, C.J., Kinniburgh, D.G., van Riemsdijk, W.H., Koopal,

L.K., 1995. Metal ion binding to humic substances: application of the non-

ideal competitive adsorption model. Environ. Sci. Technol. 29, 446–457.

Bugarin, M.G., Mota, A.M., Pinheiro, J.P., Goncalves, M.L.S., 1994. Influence

of metal concentration at electrode surface in different pulse anodic

stripping voltammetry in the presence of humic matter. Anal. Chim. Acta

294, 271–281.

Christensen, J.B., Christensen, T.H., 1999. Complexation of Cd, Ni, and Zn by

DOC in polluted groundwater: a comparison of approaches using resin

exchange, aquifer material sorption, and computer speciation model

(WHAM and MINTEQA2). Environ. Sci. Technol. 33, 3857–3863.

Christensen, J.B., Christensen, T.H., 2000. The effect of pH on the

complexation of Cd, Ni, and Zn by dissolved organic carbon from

leachate-polluted groundwater. Water Res. 34, 3743–3754.

Cleven, R.F.M.J., Leeuwen, H.P., 1986. Electrochemical analysis of the heavy

metal/humic acid interaction. Int. J. Environ. Anal. Chem. 27, 11–28.

De Jong, H.G., van Leeuwen, H.P., 1987a. Voltammetry of metal complex

systems with different diffusion coefficients of the species involved. Part I.

Analytical approaches to the limiting current for the general case including

association/dissociation kinetics. J. Electroanal. Chem. 234, 1–16.

De Jong, H.G., van Leeuwen, H.P., 1987b. Voltammetry of metal complex

systems with different diffusion coefficients of the species involved. Part II.

Behaviour of the limiting current and its dependence on association/disso-

ciation kinetics and lability. J. Electroanal. Chem. 234, 17–29.

De Jong, H.G., van Leeuwen, H.P., 1987c. Voltammetry of metal complex

systems with different diffusion coefficients of the species involved. Part III.

The current–potential relation for the general case including association/

dissociation kinetics. J. Electroanal. Chem. 235, 1–10.

Di Toro, D.M., Allen, H.E., Bergman, H.L., Meyer, J.S., Paquin, P.R., Santore,

R.C., 2001. Biotic ligand model of the acute toxicity of metals. I. Technical

basis. Environ. Toxicol. Chem. 20, 2383–2396.

Fortin, C., Campbell, P.G.C., 1998. An ion-exchange technique for free-metal

ion measurements (Cd2C, Zn2C): applications to complex aqueous media.

Int. J. Environ. Anal. Chem. 72, 173–194.

Herbelin, A.L., Westall, J.C., 1999. FITEQL. A Computer Program for

Determination of Chemical Equilibrium Constants from Experimental

Data, Version 4.0, Report 99-01. Department of Chemistry, Oregon State

University, Corvallis, OR.

Hering, J.G., Morel, F.M.M., 1988. Humic acid complexation of calcium and

copper. Environ. Sci. Technol. 22, 1234–1237.

Jansen, R.A.G., van Leeuwen, H.P., Cleven, R.F.M.J., van den Hoop,

M.A.G.T., 1998. Speciation and lability of zinc (II) in river waters.

Environ. Sci. Technol. 32, 3882–3886.

Kandegedara, A., Rorabacher, D.B., 1999. Noncomplexing tertiary amines as

‘better’ buffers covering the range of pH 3–11. Temperature dependence of

their acid dissociation constants. Anal. Chem. 71, 3140–3144.

Lu, Y., Allen, H.E., 2002. Characterization of copper complexation with

natural dissolved organic matter (DOM)— link to acidic moieties of DOM

and competition by Ca and Mg. Water Res. 36, 5083–5101.

Ma, H., Allen, H.E., Yin, Y., 2001. Characterization of isolated fractions of

dissolved organic matter from natural waters and a wastewater effluent.

Water Res. 35, 985–996.

Oste, L.A., Temminghoff, E.J.M., Lexmond, T.M., van Riemsdijk, W.H., 2002.

Measuring and modeling zinc and cadmium binding by humic acid. Anal.

Chem. 74, 856–862.

Pinheiro, J.P., Mota, A.M., Goncalves, M.L.S., 1994. Complexation study of

humic acids with cadmium(II) and lead(II). Anal. Chim. Acta. 284,

525–537.

Santore, R., Di Toro, D.M., Paquin, P.R., Allen, H.E., Meyer, J.S., 2001. A

biotic ligand model of the acute toxicity of metals. 2. Application to acute

copper toxicity in freshwater fish and daphnia. Environ. Toxicol. Chem. 20,

2397–2402.

Sarathy V., 2002. Comparison of Copper Complexation Properties of

Dissolved Organic Matter from Surface Waters and Wastewater Effluents.

Master Thesis. University of Delaware, Newark, DE, USA.

Serkiz, S.M., Perdue, E.M., 1990. Isolation of dissolved organic matter from

the Suwannee River using reverse osmosis. Water Res. 24, 911–916.

Stumm, W., Morgan, J.J., 1996. Aquatic Chemistry: Chemical Equilibria and

Rates in Natural Waters, third ed. Wiley, New York.

Tipping, E., 1994. WHAM—a chemical equilibriummodel and computer code

for waters, sediments, and soils incorporating a discrete site electrostatic

model of ion-binding by humic substances. Comput. Geosci. 20, 973–1023.

Tipping, E., 1998. Humic ion-binding model VI: an improved description of the

interactions of protons and metal ions with humic substances. Aquat.

Geochem. 4, 3–48.

van den Berg, C.M.G., 1984. Determination of the zinc complexing capacity in

seawater by cathodic stripping voltammetry of zinc–APDC complex ions.

Mar. Chem. 16, 122–130.

van Leeuwen, H.P., Cleven, R., Buffle, J., 1989. Voltammetric techniques for

complexation measurements in natural aquatic media: role of the size of

macromolecular ligands and dissociation kinetics of complexes. Pure Appl.

Chem. 61, 255–274.

Xue, H.B., Sigg, L., 1994. Zinc speciation in lake waters and its determination

by ligand-exchange with EDTA and differential-pulse anodic-stripping

voltammetry. Anal. Chim. Acta 284, 505–515.