Effects of kaolinite colloids on Cd2 + transport through saturated sand under varying ionic strength conditions: Column experiments and modeling approaches

    Effects of kaolinite colloids on Cd 2 + transport through saturated sand under varying ionic strength conditions: Column experiments and modeling approaches Rakkreat Wikiniyadhanee, Srilert Chotpantarat, Say Kee Ong PII: DOI: Reference:

S0169-7722(15)30020-6 doi: 10.1016/j.jconhyd.2015.08.008 CONHYD 3156

To appear in:

Journal of Contaminant Hydrology

Received date: Revised date: Accepted date:

31 May 2014 19 August 2015 24 August 2015

Please cite this article as: Wikiniyadhanee, Rakkreat, Chotpantarat, Srilert, Ong, Say Kee, Effects of kaolinite colloids on Cd2 + transport through saturated sand under varying ionic strength conditions: Column experiments and modeling approaches, Journal of Contaminant Hydrology (2015), doi: 10.1016/j.jconhyd.2015.08.008

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Effects of kaolinite colloids on Cd2+ transport through saturated sand under varying ionic strength conditions: Column experiments and modeling approaches

IP

T

Rakkreat Wikiniyadhanee1,2, Srilert Chotpantarat2,3,*, Say Kee Ong4

1

International Postgraduate Programs in Environmental Management, Graduate School,

SC R

Chulalongkorn University, Bangkok, Thailand 2

Center of Excellence on Hazardous Substance Management (HSM), Chulalongkorn University, Bangkok, Thailand

3

NU

Department of Geology, Faculty of Science, Chulalongkorn University, Phayathai Road, Pathumwan, Bangkok 10330, Thailand 4

MA

Department of Civil, Construction and Environmental Engineering,

D

Iowa State University, Ames, IA, USA

AC

CE P

TE

*The corresponding author. Tel: +66-22185442; Fax: +66-22185464; E-mail address: [email protected]

1

ACCEPTED MANUSCRIPT

2

Abstract Column experiments were performed under various ionic strengths (0.0 – 0.9 mM) using 10 mg L-1 of Cd2+ without kaolinite colloids and 10 mg L-1 Cd2+ mixed with 100 mg L-1 kaolinite

IP

T

colloids. The nonequilibrium two-site model (TSM) described the behavior of both Cd2+ transport and Cd2+ co-transported with kaolinite colloids better than the equilibrium model (CDeq) (R2 = 0.978 -

SC R

0.996). The results showed that an increase in ionic strength negatively impacted the retardation factors (R) of both Cd2+ and Cd2+ mixed with kaolinite colloids. The presence of kaolinite colloids

NU

increased the retardation factors of Cd2+ from 7.23 to 7.89, 6.76 to 6.61 and 3.79 to 6.99 for ionic strengths of 0.225, 0.45 and 0.9 mM, respectively. On the other hand, the presence of kaolinite

MA

colloids decreased the retardation factor of Cd2+ from 8.13 to 7.83 for ionic strength of 0.0 mM. The fraction of instantaneous sorption sites (f) parameters, kinetic constant for sorption sites (α) and

D

Freundlich constant (Kf) were estimated from HYDRUS-1D of TSM for Cd2+ transport. The fraction of instantaneous sorption sites was found to increase for an increase in ionic strength. Kf values of

TE

Cd2+ transport without kaolinite colloids for 0.0, 0.225 and 0.45 mM were found to be higher than

CE P

those of Cd2+ transport with kaolinite colloids, except for ionic strength of 0.9 mM. Hence, the presence of kaolinite colloids probably retarded the mobility of Cd2+ in porous media for higher ionic strengths. Furthermore, retardation factors and Kf values of both Cd2+ transport and Cd2+ co-transport

AC

were shown to decrease when ionic strength increased. Interestingly, according to TSM, the fraction of instantaneous sorption sites tends to increase for an increase in ionic strength, which imply that the mechanism of Cd2+ sorption onto quartz sand can be better described using equilibrium sorption rather than nonequilibrium sorption for an increase in ionic strength.

Keywords: Cd2+ transport, colloids, Hydrus-1D, kaolinite, ionic strength

ACCEPTED MANUSCRIPT

3

1. Introduction Due to their sorption affinity for pollutants, colloidal particles in groundwater from

T

soils and waste sites have attracted considerable interest in the fate and transport of pollutants

IP

and the potential of increased health risks to humans (Hu et al., 2008). Colloidal particles

SC R

have surface charges and/or organic components and are generally smaller than inter-granular pores or fractures in rock; and therefore, can be transported over long distance via groundwater flow (Sen et al., 2004). Migration of pollutants can be potentially enhanced

NU

when the pollutants are adsorbed onto the colloids and transported in subsurface environment

MA

(McCarthy and McKay, 2004). Depending on the nature of the colloids, cations, anions and nonpolar and polar organic compound can be adsorbed and transported (Kersting et al., 1999;

D

Sen et al., 2002). Cationic forms of metals have a tendency to sorb onto soil colloids such as

TE

clay minerals which can act as a carrier for metals (McGechan and Lewis, 2002; Bradl, 2004; Sen et al., 2004; Usman et al., 2005).

CE P

There are some research on the effects of colloids on heavy metals transport. For example, Sun et al. (2010) examined the effects of kaolinite colloids on the transport of lead

AC

(Pb) in saturated porous media by using column experiments. They found the transport of kaolinite was faster than that of Pb and the delayed Pb transport indicates that Pb had interacted with the porous medium in the column. When they compared transport of Pb to the of kaolinite-Pb complex, they found the latter complex displayed a faster breakthrough than Pb itself. In addition, they speculated that sorbed Pb from the media surface can be scavenged by the mobile kaolinite which would further enhance the mobility of Pb in the presence of kaolinite. Similarly, Saiers and Hornberger (1996) found that kaolinite colloids enhanced cesium (Cs) migration in a packed sand column. Noell et al. (1996) found that amorphous silica colloids increased Cs transport through glass bead columns. In an aquifer, colloidal particles are commonly attached to the porous medium during normal water flow and ionic

ACCEPTED MANUSCRIPT

4

strength conditions. Many researchers have studied various factors affecting the transport of organic and inorganic colloids such as pH (Ryan and Elimelech, 1996; Grolimund et al.,

T

1998), ionic strength (Ryan and Elimelech, 1996; Grolimund et al., 1998; Saiers and

IP

Hornberger, 1999; Walshe et al., 2010) and ionic composition (Israelachvili,1992; Elimelech

SC R

et al., 1995). Ionic strength and ionic composition are known to have an influence on the transport behavior of colloids associated with contaminants. Cheng and Saiers (2010) reported that the binding capacity of sediment-colloids for

Cs decreased with increasing

137

Cs released from the columns packed

NU

ionic strength, leading to a decrease in the mass of

137

MA

with Hanford coarse sand. Cadmium (Cd), which naturally presents in ores together with copper (Cu), zinc (Zn) and lead (Pb), is a toxic heavy metal and it can come from natural

D

sources, such as wildfires and volcanic activities, and anthropogenic sources, such as metal

TE

refining and smelting (Filiplč, 2012). Cadmium has been categorized as a potential carcinogen in humans by the U.S. national Toxicology program and International Agency for

CE P

Research on Cancer (IARC, 1993; NTP, 2000; IARC, 2012). Some evidence indicates that genomic instability can be induced by exposure to Cd (Filiplč, 2012). Additionally, Cd is a

AC

cumulative toxin because of its long biological half-life about 10-30 years in a human body (Nordberg et al., 2007). Unfortunately, there have not been effective therapies for treating Cd intoxication (Nordberg et al., 2007; Candan et al., 2009). Previous studies about the Cd2+ transport have focused on solute transport without colloidal particles and investigated effects of pH and pore-water velocity. Pang et al. (2002) investigated the effects of pore-water velocity on chemical nonequilibrium during transport of Cd2+ through alluvial gravel columns and found that pore-water velocity was positively correlated with the partitioning coefficient, forward rate and backward rate, but was inversely correlated with the retardation factor, mass transfer coefficient and ratio of forward and backward rate. In addition, Moradi et al. (2005) applied HYDRUS-1D and MACRO to

ACCEPTED MANUSCRIPT

5

describe Cd2+ transport below the root zone of a multilayer field soil in an arid region. They found that the sorption of Cd2+ could be described using Freundlich isotherm and Cd2+ moved

T

to depths of 60 and 70 cm from soil surface for enriched sewage sludge by 38 and 80 mg

IP

Cadmium per a kilogram of sewage and the results showed that equilibrium CDE model

SC R

generated far better simulation but the sorption constants of Freundlich isotherm were lower than the measured values. Recently, Zhi-Ming et al. (2012) studied about Cd2+ transport in neutral and alkaline soil columns at various depths. The results revealed that the equilibrium

NU

CDE model was sufficient for modeling Cd transport and the dispersion coefficient were

MA

between 0.78 and 10.70 cm2 h-1 retardation factors were between 25.4 and 54.7. However, there are many studies on the influence of colloids on the transport of Cd2+ in groundwater but not many on the fate and transport of natural colloids (Zhou et al., 2011), and the effects

TE

D

of ionic strength on co-transport of colloidal particles and Cd2+. Many computer programs, for example HYDRUS-1D (Šimunek et al., 2008) and

CE P

CXTFIT (Toride et al., 1999), have been used to model the water and solute transports. However, CXTFIT is frequently applied to describe breakthrough curves of tracer in steady

AC

state one dimensional flow (Toride et al., 1999). HYDRUS-1D was chosen in this study because it is a freeware and its window interface is similar to CXTFIT making this program easy to use and HYDRUS-1D has many options to estimate parameters. Moreover, HYDRUS-1D has been developed and used in many studies to carry out new modules (Chotpantarat et al., 2012). Mathematical models were used to analyze the experimental data to obtain numerical parameters which may be used to describe the behavior of metal transport and metal co-transport with colloidal particles. Consequently, the estimation of Cd2+ transport in a geological aquifer can be more accurate and reliable by applying mathematical models. However, few research, especially mathematical models, are available on effects of kaolinite colloids under different ionic strengths on the Cd2+ transport in subsurface environment.

ACCEPTED MANUSCRIPT

6

As mentioned above, therefore, in this study, we investigated the influence of ionic strength and kaolinite colloids on the migration of Cd2+ in columns packed with quartz sand.

T

The objectives of this study were to (1) evaluate the effects of ionic strength on the co-

IP

transport of Cd2+ with kaolinite colloids in saturated sand columns and (2) investigate the

SC R

effects of mobile kaolinite colloids on the Cd2+ transport in saturated sand columns and to model the effects of ionic strength and kaolinite colloids on Cd2+ sorption and transport in

NU

saturated sand columns by HYDRUS-1D.

2. Materials and Methods Quartz sand

MA

2.1.

Ottawa sand (quartz sand) purchased from Fisher Chemical (England) had a grain size

D

of 0.6 – 0.8 mm in diameter. The sand was cleaned according to the procedure described by

TE

Zhou et al. (2011) to remove metal oxides and adsorbed particles from the sand surface. The

CE P

cleaned sand was then analyzed for its mineral compositions using X-ray diffraction (XRD) (AXS D8, Bruker, Germany).

AC

2.2. Kaolinite

Kaolinite used in this experiment was purchased from Fisher Chemical (England). Kaolinite was analyzed for its mineral compositions by using XRD (AXS D8, Bruker, Germany), and its chemical compositions in term of the compounds of oxide by using X-ray fluorescence (XRF) (AXS S4, Bruker, Germany).

2.3. Column transport experiments An acrylic column with an internal diameter of 2.5 cm and 10.0 cm in length with stainless-steel end caps on both ends were used for the column experiments. The column was packed with Ottawa sand with an approximate effective porosity and bulk density of 0.32 and

ACCEPTED MANUSCRIPT

7

1.51 g cm-1, respectively. Each column test was packed separately and reproduced column-tocolumn by consistently packing with similar bulk density.

T

The seepage velocity of the influent was approximately set at 1.56 ± 0.07 m day-1.

IP

The effluent was collected using a fraction collector (Frac-920, England). To achieve

SC R

uniformity, the sand was added to the column and the column gently shaked to improve compaction. The column was slowly saturated by introducing ultrapure water (18.2 MΩ) at the bottom end of the column. The porosities of packed sand columns varied between 0.30

NU

and 0.35. Bromide was used as a tracer in column transport experiments to estimate the

MA

longitudinal dispersivity (λ) of the packed sand columns. To determine the effects of kaolinite colloids under different ionic strengths on Cd2+ transport. Cd-bearing kaolinite suspensions for different ionic strengths were prepared as

TE

D

follows. Firstly, 0.25 g kaolinite clay and a 500 mL of ultrapure water (18.2 MΩ) were added to the 500 mL volumetric flask. Then the suspension was mixed and placed in an ultrasonic

CE P

bath for 30 min. After that, the suspension was left to stands for 24 hrs. After 24 hrs, the upper half of the suspension was siphoned into a new container. The concentration of

AC

kaolinite colloids in the siphoned suspension was determined before diluting the solution to 250 mg L-1. Secondly, 200 mL of 250 mg L-1 of kaolinite colloid suspension and 5 mL of 1000 mg L-1 Cd2+ solution were added to the 500 mL volumetric flask and ultrasonicated for 1 min. The ionic strength of the suspensions was then adjusted. The final volume was 500 mL, and the concentration of Cd2+ and kaolinite colloids were 10 mg L-1 and 100 mg L-1, respectively. Cd concentration in nature is normally not as high as that used in this experiment but Cd concentration in highly contaminated sites can be in the same ranges as that used in the experiment. Since sorption of Cd by kaolinite may reduce the final dissloved concentration, a high initial concentration of 10 mg L-1 was used. Other researchers such as Pang et al. (1999) have used cadmium concentration of about 10 mg L-1 in batch tests and

ACCEPTED MANUSCRIPT

8

column experiments which was similar with our study. For kaolinite concentration, although the concentration of kaolinite colloid in nature has not been reported, colloid concentrations

T

found in natural subsurface systems varied from 35 to 100 ppm (Jie et al., 2003). However,

IP

concentrations of colloids higher than 100 mg L-1, have been reported in some sites

SC R

(Gschwend, 1990; Ronen et al., 1992; Vilks et al., 1993). Thus, kaolinite concentrations in this study varied from extremely low to high concentration. A kaolinite concentration of 100 mg L-1 was used to mimic areas with high colloid concentration, which may provide results

NU

that can clearly explain such effects. Other researchers (Akbour et al. (2002); Saiers and

MA

Hornberger (1996); Zhu et al. (2012)) have used concentration of kaolinite colloid higher than 10 mg L-1 in their colloid transport experiments. The ionic strength of the Cd-bearing kaolinite suspensions were 0, 0.225, 0.45, and 0.9 mM of CaCl2. Generally, the ionic

TE

D

strengths of groundwater are between 0.1 – 10 mM (Mark and Bruce, 1995). Moreover, zetapotentials of the Cd-bearing kaolinite suspensions were measured using Zetasizer nano (ZS,

CE P

Malvern, England).

2.4. Experimental design

AC

2.4.1. Cd2+ transport experiments Prior to the use of the packed sand column, the column was equilibrated by flushing several pore volumes (PVs) of ultrapure water and at least 5 PVs of a solution with a fixed pH value of 5.60 ± 0.02 to standardize the chemical conditions and establish steady state flow. After that, 10 mg L-1 of Cd2+ solution with the same ionic strength of the background solution was pumped into the column from the bottom with a piston pump (FMI lab QG6, US) at a seepage velocity of 1.56 ± 0.07 m day-1 for 13 PVs. Several PVs of the background solution with the same pH and ionic strength were then applied into the column to flush and make sure that the Cd2+ was eluded from the column. The effluent was collected in 15 mL tubes at regular time interval using a fractional collector.

ACCEPTED MANUSCRIPT

9

2.4.2. Co-transport experiments The preparation and flushing of the packed sand column for the co-transport experiments was similar to that of the Cd2+ transport experiment. After preparation of the

T

column Cd-bearing kaolinite suspension solution, adjusted for a given ionic strength, was

IP

pumped into the column from the bottom using a piston pump (FMI lab QG6, US) at a

SC R

constant seepage velocity of 1.56 ± 0.07 m day-1 for about 13 PVs. After completing this step, several PVs of kaolinite colloid-free background solution with the same pH and ionic strength as the Cd-bearing kaolinite suspension solution was introduced into the column to flush and make sure that kaolinite colloids was eluded from the column. The effluent was

NU

collected in 15 mL tubes at regular time interval using a fractional collector. The columns were flushed with background solution and the solution was measured after the experiments

D

2.4.3. Tracer experiments

MA

were completed to ensure that there was no colloids in the column.

TE

Bromide was introduced to the column as a conservative tracer. The experimental

CE P

procedures were the same as Cd2+ transport and co-transport experiments and an electrical conductivity meter (sensION5, HACH, US) was used to determine the concentrations of bromide. A calibration curve was established by diluting a 1 M of NaBr. Electrical

AC

conductivity versus NaBr concentration was linear in the range of 0 – 1 M with a coefficient of determination of R2 of 0.999. Three replications of tracer experiments were performed.

2.4.4. Experimental runs The experiment conducted to investigate the effects of ionic strength on Cd2+ transport and co-transport is given in Table 1 along with the column properties and the seepage velocities.

[Insert Table 1] 2.5. Analytical procedures

ACCEPTED MANUSCRIPT

10

2.5.1. Kaolinite colloids and Cd2+ Concentration The concentration of kaolinite colloid in the samples was measured using a UV/vis

T

spectrophotometer (GENESYS 10S, Thermo Scientific, England) at wavelength of 350 nm.

SC R

mg L-1 kaolinite suspension to different concentrations.

IP

(Sun et al., 2010; Zhu et al., 2012). A calibration curve was established by diluting the 250

Preliminary experiments were conducted to compare the digestion method to analyze Cd concentration by digesting the kaolinite with concentrated HNO3 in the presence of heat

NU

(microwave) at 110 ºC (Zhu et al., 2012) and a modified digestion method without heat. The

MA

concentrations of Cd for both methods using the same samples, were found to be not significantly different (data not shown). Since the total number of samples for co-transport

D

experiments were about 400 and the microwave digestion method required lots of time, the

TE

modified method was carried out for kaolinite digestion. Five mL of 14.5 M HNO3 was added to 5 mL of all effluent samples to digest kaolinite colloids to determine the Cd2+

CE P

concentration. The Cd2+ concentration in each sample was measured using an atomic

AC

adsorption spectrophotometer (AAS) (AAnalyst 800, PerkinElmer, Germany).

2.5.2. CXTFIT model The CXTFIT 2.0 (Toride et al., 1999; Chotpantarat et al., 2011) program shows a number of analytical solutions for one-dimentaional transport model relied on the convectiondispersion equation (CDE). The equilibrium CDE equation may be represent as

Where, C (mg L-1) is the concentration of solute in liquid phase; t (day) is the duration time; DL (cm2 day-1) is the longitudinal dispersion coefficient; vx (cm day-1) is the average linear ground water velocity; ρ (g cm-3) is the bulk density of the aquifer; θ is the porosity; C*

ACCEPTED MANUSCRIPT

11

(mg g-1) is the amount of solute sorbed onto the solid; in addition, rxn is the subscript indicating a chemical or biological reaction of the solute.

T

2.5.3. HYDRUS-1D model

IP

The HYDRUS-1D model (Šimunek et al., 2008; Chotpantarat et al., 2011, 2012) can

SC R

be used to simulate the transport of heavy metal in soils. The equilibrium convectiondispersion (CDeq) transport model (Eq. (1)), and the chemical non-equilibrium model or the

NU

two-site model (TSM) using Freundlich isotherm model which was one of the most simple and adjusted description for heavy metal adsorption in formal mathematical terms (Mallmann

MA

et al., 2012) as showed (Eq. (2a) – (2d))





s k   k  sek  s k  k t

CE P



TE

S e  fK f C1 / n

S k  1  f K f C1 / n

(2a)

D

C s e s k   C  vxC  C     DL    t t t x  x  x  t  rxn

(2b) (2c) (2d)

AC

Where f is the fraction of exchange sites (-), αk (day-1) is a first-order kinetic rate coefficient. Eq 2 (a) explains solute transport in the column experiment. Sorption se (mg g-1 soil) on one fraction of the sites (Type 1 site), assumed to be instantaneous sorption while sk (mg g-1 soil) is the solid phase concentration on the remaining sites (Type 2 site), considered to be a first-order kinetic process, and ϕk represents a sink-source term that accounts for various zero- and first-order or other reactions at the kinetic sorption sites (mg L-1 day-1) in Eq 2 (b). Eq 2 (c) is a mass balance of the kinetic sorption sites while Eq 2 (d) represents the sorbed concentration of the kinetic sites when equilibrium would be reached with the solute concentration. In our experiments, the total Cd concentration (the sum of dissolved Cd + digested Cd) along with the Cd sorbed onto the kaolinite colloid were measured at each pore

ACCEPTED MANUSCRIPT

12

volume. HYDRUS-1D was used to model the total Cd in the effluent from the column experiments. The objective of the experiment was to determine the impact of kaolinite colloid

T

on the overall transport of Cd. As such, the models used was curve-fitted with the total Cd

IP

and therefore can be treated as a single phase in the advection-dispersion transport model and

SC R

the chemical non-equilibrium model. Zhu et al. (2012) similarly used the advectiondispersion transport equation to investigate the transport and interaction of Hg and kaolinite

MA

2.5.4. Retardation factor estimation

NU

in saturated sand media.

The values of retardation factor (R) were determined by using the relationship between retardation factor and sorption parameters as shown in Eqs 3 and 4. The retardation

CE P

TE

D

factors for equilibrium model were estimated using

AC

Whereas retardation factors for nonequilibrium model can be expressed as

Where Ceq is the equilibrium liquid-phase concentration.

2.5.5. Parameter estimation The bromide breakthrough curves (BTCs) were used to estimate the hydrodynamic dispersion coefficient (D) of the soil using the nonlinear least-squares parameters optimization method in CXTFIT program. The equilibrium convection-dispersion (CDeq) with retardation factor of 1 and the two-region non-equilibrium were applied to investigate the system as well. The hydrodynamic dispersion coefficient determined from bromide BTCs was used to determine the soil dispersivity. The average dispersivity (λavg) of bromide BTCs

ACCEPTED MANUSCRIPT

13

was used for the convection-dispersion model of HYDRUS-1D to estimate sorption parameters of the heavy metal under varying IS conditions. Moreover, TSM was used to

T

estimate the sorption parameters (Kf and n) and non-equilibrium parameters (f and α) for

IP

transport of the heavy metal. Coefficient of determination of R2 was used to determine the

SC R

suitableness of the curve fitting.

3. Results and Discussion

NU

The results of XRD study (data not shown) confirmed that the mineral composition of

MA

sand, purchased from Fisher Chemical, was pure quartz mineral (SiO2), suggesting that the groups located on the surface of quartz sand were Si-O-, which are generally negatively

D

charged. Wang et al. (2011) have studied the effects of ionic strength on facilitated transport

TE

of Cu with hydroxyapatite nanoparticles (nHAP) through quartz sand and they reported that the zeta-potentials of quartz sand were from -79.6 mV to -26.0 mV. Consequently,

CE P

electrostatic repulsion occurred between Cd-bearing kaolinite colloids and quartz. In addition, the XRF data for kaolinite (data not shown) showed that the major compounds of oxide in

AC

kaolinite were Al2O3 and SiO2, which were similar to findings of Shahmohammadi-Kalalagh et al. (2011). This imply that Cd2+ can be sorbed onto kaolinite and sand through ionic exchange reaction since the surface charges of both quartz sand and kaolinite were negative. However, increasing the ionic strength can enhance deposition of colloids due to a reduction of repulsive surface interaction energies between colloids and quartz sand, as well as colloids and colloids. Although the concentration of Cd sorbed onto kaolinite colloid was decreased for increasing ionic strength, colloid retention was promoted and hence the Cd2+ retention in the column. The study of Gu and Evans (2008) supported this view. They have studied the surface complexation of Cd2+ adsorption onto kaolinite under different pH and ionic strengths. They reported that under low pH range (≤6) Cd2+ were bound through ionic

ACCEPTED MANUSCRIPT

14

exchange reaction and formed outer-sphere complex on basal surface sites (siloxane groups). The possible sorption mechanism for Cd2+ onto beach sand may be due to the negative charge

T

in the structure of calcite, quartz and aragonite in the form of oxides and carbonates which

SC R

Cd2+ and negatively charged sand (Taqvi et al., 2007).

IP

may cause affinity for Cd2+ by the electrostatic attractive force between positively charged

3.1. Bromide tracer test

NU

The model fits for equilibrium and the two-region non-equilibrium models of CXTFIT are showed in Fig 1. The estimated parameters are listed in Table 2. The two-region

MA

non-equilibrium model for bromide BTCs demonstrated the partitioning coefficient (β) value of 1 and the mass transfer coefficient (ω) value of 100, suggesting that the equilibrium model

D

should be applied for analyzing the bromide BTCs (Pang et al., 2002). Therefore, the average

AC

[Insert Table 2]

CE P

[Insert Fig. 1]

TE

dispersivity estimated from CXTFIT program was 1.07 cm.

3.2. Effects of ionic strength on Cd2+ transport through saturated sand columns The BTCs of Cd2+ through the saturated sand columns are presented in Fig 2. The maximum relative concentration of Cd2+ (Ci/C0) in the effluent for every ionic strength was 1. However, the PV to reach maximum Ci/C0 was different for each ionic strength condition. For example, for ionic strengths of 0, 0.225, 0.45 and 0.9 mM, the maximum Ci/C0 were obtained at 10.5 PVs, 6.0 PVs, 5.5 PVs and 4 PVs, respectively. The BTC of Cd2+ transport for ionic strength of 0.9 mM was somewhat similar to that of tracer bromide, indicating that Cd2+ sorption onto media was low. The results of column transport experiments were consistent with those of batch sorption experiments (Data not shown) that the amount of Cd2+

ACCEPTED MANUSCRIPT

15

sorbed onto sand surface was relatively low (~ 0.018 mg g-1 of sand) and decreased as increasing in the ionic strength (Wikiniyadhanee, 2012). The BTCs of Cd2+ transport for

T

every ionic strength depicted a sharp rising front with a tailing of declining limb and no

IP

tailing phenomenon of Cd2+ breakthrough curves was found under high ionic strengths,

SC R

however, pronounced tailing phenomenon represented under low ionic strengths (Fig. 2). The same pattern of breakthrough curves was observed in Cs transport through silica sand (Flury et al., 2004). The sorption capacities and retardation factors of Cd2+ transport through the

NU

saturated sand columns were estimated as shown in Table 3 and Table 4. The retardation

MA

factors and sorption capacities of Cd2+ were found to decline from 8.13 to 3.79 (53.32 % decrease) and 3.0 x 10-3 to 7.0 x 10-4 mg g-1 (76.67 % decrease), respectively, for an increase in ionic strength. The results suggested that an increase in ionic strength can promote the

TE

D

transport of Cd2+ through saturated sand media. This may be due to a decrease in Cd2+ sorption onto the quartz sand. Srivastava et al. (2005) used the hydrolysis constant from Baes

CE P

and Mesmer (1976) to determine the aqueous species of Cd and they reported that Cd was predominantly present as Cd2+ for pH lower than 7. Moreover, Basualto et al. (2006) studied

AC

the extraction of Cd from aqueous solutions and reported that Cd2+ predominates along the acid pH range at low concentrations of chloride ion (1 mM). However, if the chloride ion concentration is high, the complex species CdCl42- is formed. The highest concentration of chloride ion in these co-transport experiments was 0.6 mM; therefore, the effects of chloride ion on Cd adsorption may ignored.

[Insert Fig. 2]

[Insert Table 3]

[Insert Table 4]

ACCEPTED MANUSCRIPT

16

3.3. Effects of kaolinite colloids on Cd2+ transport To determine the effects of kaolinite colloids on Cd2+ transport, the transport experiments of

T

Cd2+ were compared to those of Cd2+ co-transport with kaolinite colloids at the same ionic

IP

strength. In this study, we aimed to investigate the effects of ionic strength; therefore, the

SC R

range of ionic strength varied from extremely low to high (from 0 mM to 3.6 mM) but we cannot observe the different effects of ionic strength on transport of Cd-bearing kaolinite colloid for ionic strength higher than 0.9 mM. So, the ionic strength of the co-transport

NU

experiments in this manuscript showed only ionic strength in the range of 0 - 0.9 mM. The

MA

recoveries for total Cd2+ were 89.3% and 93.8% for ionic strengths of 0.0 and 0.225 mM, which were lower than those of Cd2+ co-transport with kaolinite colloids at the same ionic strengths (92.3% and 93.9%, respectively). In contrast, the recoveries for total Cd2+ for ionic

TE

D

strengths of 0.45 and 0.9 mM were 95.0% and 97.3%, which were higher than those of Cd2+ co-transport with kaolinite colloids at the same ionic strengths (94.2% and 94.1%,

CE P

respectively) (Table 3). As shown in Table 3, for ionic strength of 0 mM, 91.63% of dissolved Cd2+ eluted from the column with 8.37% of the Cd associated with the kaolinite

AC

colloids. For an increase in ionic strength, the percentages of dissolved Cd2+ increased (see Table 3; column nos. 6-8); consequently, the percentages of kaolinite-facilitated Cd2+ decreased. It can be concluded that more than 90% of Cd2+ transport was in the soluble phase. Batch sorption experiments were conducted to investigate the effects of various ionic strength on amount of Cd2+ sorbed onto kaolinite and sand (data not shown). The results indicated that the sorption capacity of Cd2+ onto kaolinite and sand decreased for an increase in the ionic strength (Wikiniyadhanee, 2012). The effects of ionic strength on Cd2+ sorption can be attributed to competition between Cd2+ ions and cations in the solution for vacant sites (Jiang et al., 2010). Moreover, the increase of ionic strength can neutralize the surface negative charge of kaolinite colloids and quartz sand, leading to a decrease in the sorption of Cd2+

ACCEPTED MANUSCRIPT

17

(Coles and Yong, 2002). Our column experiments presented in the manuscript were in agreement with our batch sorption experiments.

T

The BTCs for Cd2+ through the saturated sand columns with kaolinite colloids are also

IP

presented in Fig. 2. Figure 2e shows that the breakthrough time of Cd2+ co-injected with

SC R

kaolinite colloids was less than that of Cd2+ transport without kaolinite colloids for an ionic strength of 0 mM. On the other hand, the breakthrough time of Cd2+ decreased when the ionic strength increased. For example, the breakthrough time of Cd2+ was almost equal to the

NU

breakthrough time of Cd2+ with kaolinite colloids for an ionic strength of 0.225 mM (Fig. 2b

MA

and 2f) and 0.45 mM (Fig. 2c and 2g). Figure 2h shows the breakthrough time of Cd2+ coinjected with kaolinite colloids was much greater than the breakthrough time of Cd2+

D

transport without kaolinite colloids (Fig. 2d). The results indicated that kaolinite colloids

TE

obviously affected Cd2+ transport in the saturated sand columns and their effects depended on the concentration of ionic strength. Figures 2e, f and g, moreover, showed that values of Ci/C0

CE P

of the plateau portion of BTCs for Cd2+ transport were greater than for Cd2+ co-transported with kaolinite colloids, indicating that Cd2+ sorption onto quartz sand appears to be altered

colloids.

AC

from the equilibrium sorption to the nonequilibrium sorption as presenting of kaolinite

Data in Table 4 shows both retardation factors for Cd2+ transport and Cd2+ transport with kaolinite colloids decreased when ionic strength of the solution increased. The study of Zhu et al. (2012) showed that kaolinite colloids reduced the Hg’s mobility and from the breakthrough curve of Hg in the co-transport of 100 mg L-1 Hg and 100 mg L-1 kaolinite colloids and the maximum Ci/C0 of Hg was found to occur at 6.6 PVs, which was in later than occurrence of the maximum for 100 mg L-1 Hg transport at 1.2 PVs. In contrast, the study of Sun et al. (2010) showed that Pb co-transport with kaolinite in quartz sand columns was much faster than Pb alone. In addition, the Kf values estimated from HYDRUS-1D of TSM of

ACCEPTED MANUSCRIPT

18

Cd2+ transport without kaolinite colloids for ionic strengths of 0.0, 0.225 and 0.45 mM were higher than that of Cd2+ transport with kaolinite colloids, however, for ionic strength of 0.9

T

mM the Kf value of Cd2+ transport without kaolinite colloids was lower that of Cd2+ co-

IP

transport. Kf values of both Cd2+ transport and Cd2+ co-transport were decreased when ionic

SC R

strength increased.

Results of the transport experiments for Cd2+ without kaolinite colloids and Cd2+ with kaolinite colloids showed that for higher ionic strength conditions, kaolinite colloids tended

NU

to decrease the mobility of Cd2+ mobility and vice versa. Moreover, the zeta-potentials of

MA

Cd-bearing kaolinite suspensions were found to be less negative when ionic strength increased (Table 3), indicating that kaolinite colloids can retards Cd2+ transport through

D

packed sand columns. A possible reason is that increasing in ionic strength can promote

TE

retention of Cd-bearing kaolinite colloids in the sand media. This is due to a decrease in repulsive forces between kaolinite colloids and sand as a result of compression of the diffuse

CE P

double layer by divalent ions (Akbour et al., 2002). The results of this study showed the similar effects as in the results of the study of co-transport of hydroxyapatite nanoparticle

AC

(nHAP) and Cu in saturated sand column (Wang et al., 2011). In other studies, Magal et al. (2011) showed the impact of hyper-saline solutions on colloid transport in porous media. They reported that colloid transport was decreased for an increase in ionic strength. However, they may be released into the aqueous phase and be transported through the porous medium when the electrostatic repulsion between particles is increased due to lowering of the ionic strength of the aqueous medium (Roy and Dzombak, 1996). In addition, Grolimund et al. (1998) found that lowering the pH or increasing in ionic strength of bulk solutions resulted in the deposition of colloids due to a reduction of repulsive surface interaction energies between the colloids and the porous media, and colloids and colloids. Ryan and Elimelech (1996) reported that the deposition of colloids reduced when the ionic strength decreased or

ACCEPTED MANUSCRIPT

19

pH is increased. This was due to an increase in electrostatic repulsion and increase in the double layer thickness. Israelachvili (1992) and Elimelech et al. (1995) reported that

T

monovalent ions have a lower effect on colloidal transport in the medium than multivalent

IP

ions. Walshe et al. (2010) found that increasing in the ionic strength of a solution resulted in

SC R

reduced peak concentrations of both kaolinite and kaolinite-facilitated MS2 coliphage from columns packed with gravel aquifer media. In addition, an increase in the particle size of kaolinite colloids due to an increase in ionic strength can enhance retention of Cd2+ in the

NU

column due to size exclusion of kaolinite colloids. The concentrations of colloid

MA

breakthrough depend on the particle size of the colloids when migrating through watersaturated porous media (Santos and Bedrikovetsky, 2006). Mitropoulou et al. (2013) have

D

studied the effect of ionic strength and sand grain size on colloid transport in unsaturated

CE P

straining/filtering effects.

TE

columns and found that the larger colloids were more retained than smaller colloids due to

3.4. Modeling of Cd2+ transport by HYDRUS-1D

AC

The CDeq transport model and the chemical non-equilibrium model or the TSM were applied to simulate the transport behavior of Cd2+ in saturated sand columns. For the simulations, the water dispersion coefficient was estimated based on bromide tracer tests. The model revealed the TSM of both Cd2+ and Cd2+ co-transport with kaolinite colloids described the behavior of Cd2+ transport more suitable than the equilibrium model (CDeq), except Cd2+ transport under the ionic strength of 0.9 mM (R2 = 0.978 - 0.996). Chotpantarat et al. (2012) also showed that the two-site sorption model fitted breakthrough curves of heavy metals better than the equilibrium model. Table 4 represents the parameters estimated by HYDRUS1D and showed that the Freundlich constants (Kf) decreased as increasing in ionic strength for Cd2+ and Cd2+ with kaolinite colloids, which was consistent with decreased retardation

ACCEPTED MANUSCRIPT

20

factors. On the other hand, 1/n values increased as increasing in ionic strength, indicating that the preferential sorption of Cd2+ for lower ionic strength was greater than for higher ionic

T

strength. For TSM parameters, the fraction of instantaneous site (f) of Cd2+ under different

IP

ionic strengths was about 60%-90% for Cd2+ transport and about 60%–85% for Cd2+ with

SC R

kaolinite colloids. Furthermore, f parameter increased when ionic strength increased, suggesting that the fraction of instantaneous sorption site (f) of Cd2+ tend to be increased as increasing in ionic strength due to ionic competition. The high ionic strength in the systems

NU

the higher is the fraction of occupied instantaneous sorption sites. In other words, the higher

MA

fraction of instantaneous sorption site would imply that mechanism of Cd2+ sorption onto quartz sand appears to be changed from the nonequilibrium sorption to the equilibrium

D

sorption as seen in Fig.2. Furthermore, the shape of curves derived from TSM closely

TE

coincided with those from CDeq, particularly in the decreasing limbs, which do not show the tailing phenomenon (Figs. 2d and h). As the kinetic constant for sorption sites (α) decreases,

CE P

the α parameter, describing the kinetic sorption rate, becomes less important. The lower α value means the sorption reaction will reach the equilibrium condition faster, resulting in the

AC

higher metal concentration in the soil solution and the faster transport. However, during the desorption phase, the lower value of α parameter produces a faster rate of desorption from the soil sorption sites to soil solution and, consequently, the faster transport of the Cd2+. The lower sorption rate might be related to the stronger competition of Ca2+ with Cd2+ for the nonequilibrium sorption sites. Furthermore, the results revealed that mass recovery of Cd2+ in column experiments tend to be increase at high ionic strength, both column for Cd2+ (from 89.3 to 97.3 %) and Cd2+ with kaolinite colloids (from 92.3 to 94.1%), indicating that sorption mechanism significantly appear to be more reversible under high ionic strength. This is in agreement with the results of models as mentioned before.

ACCEPTED MANUSCRIPT

21

4. Conclusions Experimental results of this study showed that an increase in the ionic strength of the

T

aqueous phase can facilitate the transport of Cd2+ through sand porous media. This is to

IP

decrease in the sorption capacity of Cd2+ onto quartz sand from the competition of cations for

SC R

sorption sites on the media and complexation of Cd2+ with the anions in the aqueous phase. Results of the study also showed that kaolinite colloids can enhance the mobility of Cd 2+ in a saturated media column under condition of low ionic strength. The results showed that

NU

kaolinite colloids in the aqueous phase can inhibit the mobility of Cd2+ for higher ionic

MA

strength aqueous phase. This is due to the retention of kaolinite colloids in the porous media, which consequently retain the Cd2+ sorbed onto the kaolinite colloids. In the real world, under

D

normal groundwater flow, colloidal transport would be minimal but when the section of

TE

aquifer is distributed or there is a large flow, colloids can be distributed and move resulting in an increase transport of pollutants. Furthermore, the findings of this study revealed that the

CE P

chemical non-equilibrium TSM described the results more suitable than the CDeq model of both experiments and kaolinite colloids under different ionic strengths significantly affected

AC

on Cd2+ transport for risk assessment of Cd2+ mobility in subsurface environment. The understanding of Cd2+ co-transport with colloidal particles under different ionic strengths is necessary for predicting the transport of Cd2+ in the groundwater system because failure to estimate the fate and transport of heavy metals might result in wrong prediction of arrival time and concentration of them. Thus, the mathematical models, especially nonequilibrium model (TSM), should be considered to assess the transport of Cd2+ in the real groundwater system in which colloidal particles present.

ACCEPTED MANUSCRIPT

22

5. Acknowledgements The authors would like to thank the Graduate School of Chulalongkorn University,

T

the Ratchadaphiseksomphot Endowment Fund 2014 of Chulalongkorn University (CU-57-

IP

061-CC) for providing financial support. We are grateful for the thorough reviews by

SC R

Prof.Dr.Stefan B.Haderlein, Editor-in-Chief of Journal of Contaminant Hydrology, and anonymous reviewers. Their valuable comments significantly improved the earlier draft to

AC

CE P

TE

D

MA

NU

this article.

ACCEPTED MANUSCRIPT

23

References Akbour, R. A., Douch, J., Hamdani, M., and Schmitz. P., 2002. Transport of kaolinite

T

colloids through quartz sand: influence of humic acid, Ca2+ and trace metals. Journal of

IP

Colloid and Interface Science 253(1), 1-8.

SC R

Baes, C. F. and Mesmer, R. E. Hydrolysis of cations. Wiley, New York, 1976 Basualto, C., Poblete, M., Marchese, J. Ochoa, A., Acosta, A. and Sapag, J., 2006. Extraction

NU

of cadmium from aqueous solutions by emultion liquid membranes using a stirred transfer cell contactor. Journal of Brazilian Chemical Society 17(4), 1347-1354.

MA

Bradl, H. B., 2004. Adsorption of heavy metal ions on soils and soil constituents. Journal of Colloid and Interface Science 277, 1-18.

D

Candan, N., Tϋzmen, N., Andac, M., Andac, C. A., Say, R. and Denizli, A., 2009. Cadmium

TE

removal out of human plasma using ion-imprinted beads in a magnetic column. Materials

CE P

Science and Engineering C29, 144-152. Cheng, T., and Saiers, J.E., 2010. Colloid-facilitated transport of cesium in vadose-zone sediments: the importance of flow transients. Environmental Science and Technology

AC

44(19), 7443 - 7449.

Chotpantarat, S., Ong, S. K., Sutthirat, C. and osathaphan, K., 2012. Competitive modeling of sorption and transport of Pb2+, Ni2+, Mn2+ amd Zn2+ under binary and multi-metal systems in lateritic soil columns. Geoderma 189-192, 278-287. Chotpantarat, S., Ong, S. K., Sutthirat, C. and osathaphan, K., 2011. Effect of pH on transport of Pb2+, Mn2+, Zn2+ and Ni2+ through lateritic soil: Column experiments and transport modeling. Journal of Environmental Sciences 23(4), 640-648. Coles, C. A., and Yong, R. N. 2002. Aspects of kaolinite characterization and retention of Pb and Cd. Applied Clay Science 22, 39-45.

ACCEPTED MANUSCRIPT

24

Elimelech, M., Jia, X. Gregory, J., and Williams, R., 1998. Particle deposition and aggregation: Measurement, Modeling and Simulation. Butterworth-Heinemann Ltd.,

T

Oxford, pp. 440.

IP

Filiplč, M., 2012. Mechanism of cadmium induced genomic instability. Mutation Research

SC R

733, 69-77.

Flury, M., Czigány S., Chen, G. and Harsh, J. B., 2004. Cesium migration in saturated silica sand and Hanford sediments as impacted by ionic strength. Journal of Contaminant

NU

Hydrology 71, 111-126.

MA

Grolimund, D., Elimelech, M., Borkovec, M., Barmettler, K., Kretzschmar, R., and Sticher, H., 1998. Transport of in situ mobilized colloidal particles in packed soil column.

D

Environmental Science and Technology 32(22), 3562-3568.

TE

Gschwend, P. M., Backhus, D. A., MacFarlane, J. K. and Page, A. L. 1990. Mobilization of colloids in groundwater due to infiltration of water at a coal ash disposal site, Journal

CE P

of contaminant hydrology 6, 307-320. Gu, X., and Evans, J. L., 2008. Surface complexation modeling of Cd(II), Cu(II), Ni(II),

276.

AC

Pb(II) and Zn(II) adsorption onto kaolinite. Geochimica et Cosmochimica Acta 72, 267-

Hu, S., Chen, X., Shi, J., Chen, Y., and Lin, Q., 2008. Particle-facilitated lead and arsenic transport in abandoned mine sites soil influenced by simulated acid rain. Chemosphere 71, 2091-2097. IARC, International Agency for Research on Cancer, 1993. Beryllium, Cadmium, Mercury and Exposures in the Glass Manufacturing Industry. IARC Monographs on the Evaluation of Carcinogenic Risks to humans. Lyon, France, pp. 119-238.

ACCEPTED MANUSCRIPT

25

IARC, International Agency for Research on Cancer, 2012. A review of human carcinogens: arsenic, metals, fibres and dusts. IARC Monographs on the Evaluation of Carcinogenic

T

Risks to humans, 100c. Lyon, France, pp. 121-141.

IP

Israelachvili, J. N., 1992. Intermolecular and surface forces. Academic, New York.

SC R

Jiang, M., Jin X., Lu, X. and Chen, Z. 2010. Adsorption of Pb(II), Cd(II), Ni(II) and Cu(II) onto natural kaolinite clay. Desalination 252, 33-39.

Jie, Z., Markus, F., and Yan, J. Colloid-facilitated Cs transport through water-saturated

NU

Hanford sediment and Ottawa Sand. Environmental Science and Technology 37,

MA

4905-4911.

Kersting, A. B., Efurd, D. W., Finnegan, D. L., Rokop, D.J., Smith, D. K., and Thompson, J.

D

L., 1999. Migration of plutonium in ground water at the Nevada Test site. Nature 397, 56-

TE

59.

Magal, E., Weisbrod, N., Yechieli, Y., Walker, S. L., and Yakirevich, A., 2011. Colloid

3521-3532.

CE P

transport in porous media: Impact of hyper-saline solutions. Water Research 45(11),

AC

Mallmann, J. K. F., dos Santos, R. D., Cambier, P., Labanowski, J., Lami, I., Santanna, A. M., Tessier, D., and van Oort, F., 2012. Using a two site-reactive model for simulating one century changes of Zn and Pb concentration profiles in soils affected by metallurgical fallout. Environmental Pollution 162, 294-302. Moradi, A., Abbaspour, K. C. and Afyuni, M., 2005. Modelling field-scale cadmium transport below the root zone of a sewage sludge amended soil in an arid region in central Iran. Journal of Contaminant Hydrology 79, 187-206. Mark, J.G., and Bruce, E.L. Influence of different chemical treatments on transport of Alcalifenes paradoxus in porous media. Applied and environmental microbiology, 61: 1750-1756.

ACCEPTED MANUSCRIPT

26

McCarthy, J. F., and McKay, L. D., 2004. Colloid transport in the subsurface: past, present, and future challenges. Vadose Zone Journal 3(2), 326-337.

T

McGechan, M. B., and Lewis, D. R., 2002. Transport of particulate and colloid-sorbed

IP

contaminants through soil, Part 1: general principles. Biosystems Engineering 83, 255-

SC R

273.

Mitropoulou, P. N., Vasiliki, I. S. and Constantinos V. C., 2013. Transport of colloids in unsaturated packed columns: Role of ionic strength and sand grain size. Engineering

NU

Journal. 232, 237-248.

Noell, A. L., Thompson, J. L., Corapcioglu, M. Y., and Triay, I. R., 1996. The role of silica

MA

colloids on facilitated cesium transport through glass bead columns and modeling. Journal of Contaminant Hydrology 31(1-2), 23-56.

D

Nordberg, G. F., Nogawa, K., Nordberg, M. and Friberg, L. T., 2000. Cadmium. In: Aitio, A.,

TE

Anderson, I., Fowler, B. A., Friberg, L. T., Nordberg, G. F., Nordberg, M., Part, P., NTP,

CE P

National Toxicology Program, Tenth Report on Carcinogens, Department of health and human services, III-42-III-44, Triangle Park, NC.

AC

Pang, L., Close, M., Schneider, D., and Stanton, G., 2002. Effect of pore-water velocity on chemical nonequilibrium transport of Cd, Zn, and Pb in alluvial gravel columns. Journal of Contaminant Hydrology 57, 241-258. Ronen, D., Magaritz, M., Weber, U., Amiel, A. J. and Klein, E. 1992. Characterization of suspended particles collected in groundwater under natural gradient flow conditions. Water resource research 28, 1279-1291. Roy, B. S., and Dzombak, A. D., 1996. Colloid release and transport processes in natural and model porous media. Colloids and Surfaces A: Physiochemical and Engineering Aspects 107, 245-262.

ACCEPTED MANUSCRIPT

27

Ryan, J. N., and Elimelech, M., 1996. Colloid mobilization and transport in groundwater. Colloids and surfaces A: Physicochemical and engineering aspects 107, 1-56. 137

Cs through quartz sand: experimental

T

Saiers, J.E., Hornberger, G.M., 1996. Migration of

IP

results and modeling approaches. Journal of Contaminant Hydrology 22, 255-270.

SC R

Saiers, J.E., and Hornberger , G.E. 1999. The influence of ionic strength on the facilitated transport of cesium by kaolinite colloids. Water Resources Research. 35 (6), 1713-

NU

1727.

Santos, A. and Bedrikovetsky, P., 2006. A stochastic model for particle suspension flow in

MA

porous media. Transport in porous media 62, 23-53.

Sen, T. K., Nalwaya, N., and Khilar, K. C., 2002. Colloid-associated contaminant transport in

D

porous media: 1. Experiment studies AIChE Journal 48(10), 2366-2374.

TE

Sen, T. K., Shanbhag, S., and Khilar, K. C., 2004. Subsurface colloids in groundwater contamination: a mathematical model. Colloids and Surfaces A: Physicochemical and

CE P

Engineering Aspects 232(1), 29-38. Shahmohammadi-Kalalagh, Sh., Babazadeh, H., Nazemi, A. H., and Manshouri, M., 2011.

AC

Isotherm and kinetic studies on adsorption of Pb, Zn and Cu by kaolinite. Caspian Journal of Environmental Science 9(2), 243-255. Šimunek, J., Šejna, M., and van Genuchten, M. T., 2008. The HYDRUS-1D software package for simulating the one-dimentional movement of water, heat, and multiple solutes in variably saturated media. Version 4. Beta 1. ICGWMC-TPS-70, International Ground Water Modeling Center, Colorado School of Mines, Golden, Colorado. Srivastava, P., Singh, B. and Angove, M., 2005. Competitive adsorption behaviour of heavy metals onto kaolinite. Journal of Colloid and interface Science 290, 28-38.

ACCEPTED MANUSCRIPT

28

Sun, H., Gao, B., Tian, Y., Yin, X., Yu, C., Wang, Y., and Ma, L. Q., 2010. Kaolinite and lead in saturated porous media: Facilitated and impeded transport. Journal of

T

Environmental Engineering 136 (11), 1305-1308.

IP

Taqvi, S. I. H., Hasany, S. M. and Bhanger M. I., 2007. Sorption profile of Cd(II) ions onto

SC R

beach sand from aqueous solutions. Journal of Hazardous Materials 141, 37-44. Toride, N.,Leij, F. J., van Genuchten, M. T., 1999. The CXTFIT code for estimating transport parameters from laboratory or field tracer experiments version 2.1. In: Research Report

NU

Vol. 137 of US Salinity Laboratory. Agricultural Research Service, US Department of

MA

Agriculture, Riverside, CA, USA.

Usman, A., Kuzyakov, Y., and Stahr, K., 2005. Effect of clay minerals on immobilization of

D

heavy metals and microbial activity in a sewage sludge-contaminanted soil. Journal of

TE

Soil and Sediments 5, 245-252.

Vilks, P., Cramer, J. J., Bachinski, D. B., Doern, D. C. and Miller, H. G. 1993. Studies of

CE P

colloids and suspended particles, Cigar Lake uranium deposit, Saskatchewan, Canada. Applied geochemistry 8, 605-616.

AC

Walshe, G. E., Pang, L., Flury, M., Close, M. E., and Flintoft, M., 2010. Effects of pH, ionic strength, dissolved organic matter, and flow rate on the co-transport of MS2 bacteriophages with kaolinite in gravel aquifer media. Water Research 44(4), 1255-1269. Wang, D., Paradelo, M., Bradford., S. A., Peijnenburg, W. J., Chu, L., anf Zhou, D., 2011. Facilitated transport of Cu with hydroxyapatite nanoparticles in saturated sand: effects of solution ionic strength and composition. Water Research 45(18), 5909-5915. Wikiniyadhanee, R. Impacts of ionic strength on facilitated transport of Cd by kaolinite colloids in saturated sand column, Master degree Thesis, Graduate school Chulalongkorn University, Thailand, 2012.

ACCEPTED MANUSCRIPT

29

Zhi-Ming, Q., Shao-Yuan, F. and Helmers, M. J., 2012. Modeling cadmium transport in neutral and alkaline soil columns at various depths. Pedosphere 22(3), 273-282.

T

Zhou, D., Wang, D., Cang, L., Hao, X., Chu, L., 2011. Transport and re-entrainment of soil

IP

colloids in saturated packed column: effects of pH and ionic strength. Journal of Soil and

SC R

Sediments 11(3), 491-503.

Zhu, Y., Ma, L. Q., Gao, B., Bonzongo, J. C., Harris, W., and Gu, B., 2012. Transport and interaction of kaolinite and mercury in saturated sand media. Journal of hazardous

AC

CE P

TE

D

MA

NU

materials 213-214, 93-99.

ACCEPTED MANUSCRIPT

30

Figure captions Fig 1 – Measured bromide BTCs and CXTFIT-simulated bromide BTC: column 1(a); column

T

2(b); column 3(c)

IP

Fig 2. – Breakthrough curves for Cd2+ without kaolinite colloids for ionic strength (a) 0 mM,

SC R

(b) 0.225 mM, (c) 0.45 mM and (d) 0.9 mM and breakthrough curves for Cd2+ with kaolinite

AC

CE P

TE

D

MA

NU

colloids for ionic stregth (e) 0 mM, (f) 0.225 mM, (g) 0.45 mM and (h) 0.9 mM.

ACCEPTED MANUSCRIPT b

SC R

IP

T

a

31

MA

NU

c

AC

CE P

TE

D

Fig 1 – Measured bromide BTCs and CXTFIT-simulated bromide BTC: column 1(a); column 2(b); column 3(c)

ACCEPTED MANUSCRIPT e

SC R

IP

T

a

32

f

D

MA

NU

b

g

d

AC

CE P

TE

c

h

Fig 2. – Breakthrough curves for Cd2+ without kaolinite colloids for ionic strength (a) 0 mM, (b) 0.225 mM, (c) 0.45 mM and (d) 0.9 mM and breakthrough curves for Cd 2+ with kaolinite colloids for ionic stregth (e) 0 mM, (f) 0.225 mM, (g) 0.45 mM and (h) 0.9 mM.

ACCEPTED MANUSCRIPT

33

Table 1 – Cd2+ transport and co-transport experiments for various ionic strengths

MA D TE CE P

Porosity

T

(cm3 cm-3) 0.34 0.32 0.30 0.35 0.31 0.32 0.31 0.31

IP

Bulk density (g cm-3) 1.52 1.56 1.52 1.52 1.57 1.56 1.57 1.56

NU

(mg L-1) 10.00 Cd2+ 10.00 Cd2+ 10.00 Cd2+ 10.00 Cd2+ 2+ 10.00 Cd + 100.00 kaolinite colloids 10.00 Cd2++ 100.00 kaolinite colloids 10.00 Cd2++ 100.00 kaolinite colloids 10.00 Cd2++ 100.00 kaolinite colloids

Ionic Strength (mM) 0 0.225 0.45 0.9 0 0.225 0.45 0.9

SC R

Conditions

AC

Column Experiment (No.) 1 2 3 4 5 6 7 8

Seepage Velocity (m day-1) 1.4±0.01 1.5±0.05 1.6±0.05 1.4±0.02 1.6±0.03 1.6±0.01 1.6±0.02 1.6±0.02

pH

5.61±0.01 5.60±0.02 5.60±0.02 5.60±0.01 5.59±0.02 5.60±0.02 5.60±0.01 5.61±0.02

ACCEPTED MANUSCRIPT

34

Table 2 – Equilibrium convection-dispersion (CDeq) and two-region model parameter values estimated from bromide BTCs Two-region model

Column Experiment (No.) 1

Seepage velocity (cm day-1) 160.80

D (cm day-1)

λ (cm)

R

171.36

1.06

2

153.84

163.20

3

162.00

174.96

β

ω

R2

0.9999

100

0.992

1.08

0.9999

100

0.991

1.10

0.9999

100

0.990

D (cm day-1)

λ (cm)

0.992

175.68

1.06

0.991

167.04

1.08

0.990

168.32

1.09

IP

2

SC R

2

2

T

CDeq model

D: the longitudinal dispersion coefficient; λ: dispersivity; R2: coefficient of determination; β: the partitioning

AC

CE P

TE

D

MA

NU

coefficient between the equilibrium and non-equilibrium phases; ω: the mass transfer coefficient.

ACCEPTED MANUSCRIPT Table 3 – Recovery and sorption capacities of Cd2+ transport through saturated sand columns with and without kaolinite colloids Total Cd2+ Sorption Size ф Zeta (mg) capacity potential (μm) (mg/g) (mV) ф Dissolved Kaolinite2+ 2+ Cd F-Cd 1 1.86 (100.0) 3.0x10-3 2 1.89 (100.0) 1.7x10-3 3 1.74 (100.0) 1.0x10-3 4 1.98 (100.0) 7.0x10-4 5 1.64 (91.6) 0.15 (8.4) 1.9x10-3 0.923±0.024 -26.3±2.0 6 1.81 (96.3) 0.07 (3.7) 1.7x10-3 1.848±0.180 -24.4±0.7 7 1.74 (96.7) 0.06 (3.3) 1.6x10-3 2.494±0.032 -22.4±1.1 8 1.72 (97.7) 0.04 (2.3) 1.4 x10-3 2.691±0.084 -20.7±0.7 The percentages are represented in parenthesis ф Surface charge and size of Cd-bearing kaolinite colloids were measured before performing experiments

AC

CE P

TE

D

MA

NU

SC R

IP

T

Column Experiment (No.)

35

ACCEPTED MANUSCRIPT

36

Table 4 – Estimated parameters for Cd2+ BTCs from equilibrium convection-dispersion and nonequilibrium approaches (Two-site model) generated by HYDRUS-1D

λ* (cm)

1

Ionic strength (mM) 0.000

1.07

4.22

0.38

0.963

7.96

6.55

2

0.225

1.07

3.09

0.48

0.987

6.98

4.43

3

0.45

1.07

2.59

0.55

0.995

6.61

3.14

4

0.9

1.07

1.07

0.67

0.996

3.81

5

0.000

1.07

3.01

0.62

0.990

8.30

6

0.225

1.07

3.10

0.61

0.989

8.40

7

0.45

1.07

2.55

0.59

0.993

6.89

8

0.9

1.07

2.62

0.58

0.995

6.96

0.66

0.0046

0.978

8.13

0.42

0.80

0.0039

0.990

7.23

0.57

0.82

0.0014

0.994

6.76

1.20

0.66

0.90

0.0010

0.996

3.79

4.40

0.62

0.64

0.0016

0.994

7.89

4.12

0.67

0.63

0.0019

0.996

7.82

3.42

0.50

0.82

0.0022

0.995

6.61

3.25

0.55

0.85

0.0023

0.996

6.98

SC R

0.38

CE P

TE

D

MA

* Estimated from three bromide BTCs using CXTFIT program ** Estimated from Eq. 3 and Eq. 4 for equilibrium model and nonequilibrium model, respectively

AC

R**

IP

KF

NU

Column (No.)

Nonequilibrium model Fitting (TSM) 1/n f α R2 -1 (day )

T

Equilibrium model Fitting (CDeq) KF 1/n R2 R**

ACCEPTED MANUSCRIPT HIGHLIGHT  We investigated the effect of kaolinite and IS on Cd2+ transport in sand aquifer  The presence of kaolinite decreases the retardation factors of Cd2+ at low IS  The two-site model well describes Cd2+ co-transported with kaolinite

T

 Kaolinite colloids retard the mobility of Cd2+ in porous media for higher IS

AC

CE P

TE

D

MA

NU

SC R

IP

 Sorption of Cd2+ (with kaolinite) appears to be more equilibrium as increasing in IS

37