Literature review

w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 1 1 5 0 – 1 1 5 8

A v a i l a b l e a t w w w . s c i e n c e d i r e c t . c o m

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / w a t r e s

Sorption of perfluorooctane sulfonate and perfluorooctanoate on activated carbons and resin: Kinetic and isotherm study

Qiang Yua,b,c, Ruiqi Zhanga, Shubo Denga,b,c,*, Jun Huanga,b,c, Gang Yua,b,c

aDepartment of Environmental Science and Engineering, Tsinghua University, Beijing 100084, PR China bPOPs Research Center, Tsinghua University, Beijing 100084, PR China cState Key Joint Laboratory of Environment Simulation and Pollution Control, Tsinghua University, Beijing 100084, PR China

a r t i c l e i n f o

Article history:

Received 5 August 2008

Received in revised form

30 November 2008

Accepted 1 December 2008

Published online 13 December 2008

Keywords:

PFOS

PFOA

Activated carbon

Anion-exchange resin

Sorption kinetics

Sorption isotherm

* Corresponding author. Department of En Tel.: þ86 10 6279 2165; fax: þ86 10 6279 4006

E-mail address: [email protected] 0043-1354/$ – see front matter ª 2008 Elsevi doi:10.1016/j.watres.2008.12.001

a b s t r a c t

Perfluorooctane sulfonate (PFOS) and perfluorooctanoate (PFOA) have increasingly attrac-

ted global concerns in recent years due to their global distribution, persistence, strong

bioaccumulation and potential toxicity. The feasibility of using powder activated carbon

(PAC), granular activated carbon (GAC) and anion-exchange resin (AI400) to remove PFOS

and PFOA from water was investigated with regard to their sorption kinetics and

isotherms. Sorption kinetic results show that the adsorbent size influenced greatly the

sorption velocity, and both the GAC and AI400 required over 168 h to achieve the equilib-

rium, much longer than 4 h for the PAC. Two kinetic models were adopted to describe the

experimental data, and the pseudo-second-order model well described the sorption of

PFOS and PFOA on the three adsorbents. The sorption isotherms show that the GAC had

the lowest sorption capacity both for PFOS and PFOA among the three adsorbents, while

the PAC and AI400 possessed the highest sorption capacity of 1.04 mmol g�1 for PFOS and

2.92 mmol g�1 for PFOA according to the Langmuir fitting. Based on the sorption behaviors

and the characteristics of the adsorbents and adsorbates, ion exchange and electrostatic

interaction as well as hydrophobic interaction were deduced to be involved in the sorption,

and some hemi-micelles and micelles possibly formed in the intraparticle pores.

ª 2008 Elsevier Ltd. All rights reserved.

1. Introduction potentially toxic, PFOS and PFOA have increasingly attracted

Perfluorinated compounds (PFCs) have been widely used in

industrial and commercial applications for about 50 years as

surfactants, emulsifiers, fire retardants, polymer additives

and etc. (Key et al., 1997; Fujii et al., 2007). The PFCs most

commonly used and found in the environment are per-

fluorooctane sulfonate (PFOS) and perfluorooctanoate (PFOA),

which may be directly discharged from pollution sources or

generated by the microbial degradation of other per-

fluorinated compounds. Since they are found to be globally

distributed, environmentally persistent, bioaccumulative and

vironmental Science an . .cn (S. Deng).

er Ltd. All rights reserved

global concerns in recent years and also have been proposed

as the candidates of persistent organic pollutants (POPs)

(Giesy and Kannan, 2002; Loos et al., 2008).

Different from other typical POPs, PFOS and PFOA have

high water solubility, and thus can exist and easily transport

in water environments. So far, they have been detected in

wastewater, surface water, groundwater and even tap water

throughout the world (Fujii et al., 2007). Industrial wastewater

has been implicated as a point source for PFOS and PFOA as

well as their precursors entering into natural waters (Hansen

et al., 2002; Prevedouros et al., 2006). Many researchers found

d Engineering, Tsinghua University, Beijing 100084, PR China.

.

w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 1 1 5 0 – 1 1 5 8 1151

the higher concentrations of PFOA and PFOS in the rivers near

the PFCs-related factories in Germany (Loos et al., 2008) and

America (Hansen et al., 2002; Sinclair et al., 2006). Besides, the

PFOS concentrations at mg L�1 level were detected in the river

near a Canadian airport due to the accidental release of fire-

fighting foam (Moody et al., 2002). Tang et al. (2006) reported

that the concentration of PFOS in the original wastewater

generated from photolithographic processes of semi-

conductor manufacture was up to 1650 mg L�1, which would

cause serious pollution if being discharged into the environ-

ment. Therefore, the development of effective techniques to

remove PFOS and PFOA from industrial wastewater becomes

crucial.

Some conventional techniques including biological degra-

dation, oxidation and reduction are difficult to destruct PFOS

and PFOA in ambient environments due to their stable prop-

erties (Key et al., 1998; Schroder and Meesters, 2005). The

recent studies show that some special techniques such as

ultrasonic irradiation under argon atmosphere (Moriwaki

et al., 2005), zerovalent iron in subcritical water (Hori et al.,

2006), ultraviolet irradiation (Yamamoto et al., 2007) and

vitamin B12/Ti-citrate reduction in anoxic environment

(Ochoa-Herrera et al., 2008) may decompose PFOS or PFOA in

solution, but the specific conditions and high energy

consumption are required. Additionally, the commercial

reverse osmosis and nanofiltration membranes were also

used to separate PFOS from wastewater efficiently (Tang et al.,

2006, 2007), and our previous study found that the chitosan-

based molecularly imprinted adsorbents were effective for the

selective removal of PFOS from water (Yu et al., 2008).

It has been demonstrated in many cases that sorption is an

effective and economical method to remove many pollutants

from wastewater, but only few papers about PFOS or PFOA

removal using some commercial adsorbents were published

(Lampert et al., 2007; Ochoa-Herrera and Sierra-Alvarez, 2008).

Some researchers reported that the perfluorinated surfactants

could easily penetrate the granular activated carbon beds in

German waterworks and thus it was doubtful whether acti-

vated carbon was effective for PFOS and PFOA removal

(Schaefer, 2006). As no detailed investigation was conducted,

it is unclear whether the conventional adsorbents are effec-

tive for PFOS and PFOA removal from water. Therefore, it is

necessary to know which adsorbent should be used to effec-

tively remove them once water is polluted by PFOS or PFOA.

The objectives of this study are to investigate the sorption

behaviors of PFOS and PFOA on the commercial adsorbents

including activated carbons and resin, and evaluate their

feasibility for PFOS and PFOA removal from water. The

Table 1 – Physicochemical properties of PFOS and PFOA.

Compound Mol. formula Mol. weight

PFOS C8F17SO3K 538

PFOA C7F15 COONa 436

a Calculated by SPARC, http://ibmlc2.chem.uga.edu/sparc.

b Brooke et al. (2004).

c USEPA (2002).

sorption kinetics at different solution pH and sorption

isotherms for PFOS and PFOA using the granular activated

carbon, powder activated carbon and anionic resin were

studied in detail. The possible interactions between the

adsorbents and adsorbates were also discussed.

2. Materials and methods

2.1. Materials

Perfluorooctane sulfonate (PFOS, potassium salt) and per-

fluorooctanoate (PFOA, sodium salt) were purchased from

Tokyo Kasei Kogyo (Japan), and their properties are summa-

rized in Table 1. The Amberlite IRA 400 resin (AI400) and coal-

based activated carbon were obtained from Sinopharm

Chemical Regent Co., Ltd. (China) and Pengcheng Activated

Carbon Co., Ltd. (China), respectively. HPLC-grade methanol

was purchased from Fisher Chemical (USA). Other chemicals

were of reagent grade.

2.2. Adsorbent pretreatment

Prior to the use in the sorption experiment, the resin was first

washed in deionized water to remove dirt and then dried at

50 �C until constant weight. Similarly, the coal-based acti-

vated carbons were first rinsed with deionized water for

several times and then washed in 80 �C deionized water for 2 h

to remove the impurities. After being dried in an oven at

105 �C for 48 h, they were crushed by a mortar and screened.

The activated carbons in the size range of 0.9–1.0 mm were

used as the granular activated carbon (GAC), while the powder

activated carbon (PAC) in this paper represented the particles

below 0.1 mm.

2.3. Characterization of activated carbons

The specific surface areas of activated carbons were deter-

mined by nitrogen adsorption at 77 K using a surface area

analyzer (ASAP 2010, Micromeritics, USA). The determination

of the point of zero charge (pHpzc) for the activated carbons

was carried out as follows (Faria et al., 2004): 50 mL of 0.01 M

NaCl solution was placed into each conical flask, and the

solution pH was adjusted from 2 to 12 with 0.01 M HCl or NaOH

solution. Thereafter, 0.15 g of activated carbon was added into

each flask, and the flasks were sealed and shaken at 25 �C for

48 h. Finally, the equilibrium solution pH was measured. The

Mol. volume (cm3 mol�1)

Water solubility (mg L�1)

pKa

257a 570b �3.27b

226a 3400c 2.5c

Table 2 – Characteristics of the three adsorbents used in this study.

Adsorbent SBET a(m2 g�1) S1

b (m2 g�1) S2 c (m2 g�1) pHpzc Size (mm)

PAC 812 466 346 7.5 <0.1

GAC 712 313 399 7.5 0.9–1.0

Adsorbent Exchange capacity Functional group Matrix Size (mm)

AI400d 3.0–3.5 eq kg�1 –NþR3 (Cl �) Polystyrene divinylbenzene 0.3–1.2

a BET surface area.

b Micropore area.

c Mesopore and macropore area.

d Properties given by the manufacturer.

w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 1 1 5 0 – 1 1 5 81152

pH point where pHinitial ¼ pHfinal was taken as the pHpzc of the activated carbon. The obtained results are shown in Table 2.

2.4. Sorption experiments

Batch sorption experiments were carried out at 150 rpm in an

orbital shaker with 0.01 g of adsorbents (0.005 g for the sorp-

tion kinetics of PFOA on the AI400) in the 250-mL flasks con-

taining 100 mL PFOS or PFOA solution and 2 mM NaH2PO4 as

pH buffer. The solution pH was adjusted using 0.1 M NaOH or

HCl solution and the pH values were determined by an HQ40d

Digital Multi-Parameter Meter from Hach (USA). In the sorp-

tion kinetic experiments, 50 mg L�1 of PFOS or PFOA solution

at initial pH 3 or 7 was used and the sorption experiments

were conducted at 25 �C. After the sorption experiments, the

final solution pH was determined. The sorption isotherm

experiments were conducted at the initial PFOS or PFOA

concentrations ranging from 20 to 250 mg L�1 at 25 �C for 168 h

(12 h for the PAC), and the solution pH was continuously

adjusted to 5 and kept constant during the sorption. As a small

volume of NaOH or HCl solution (less than 0.5 mL) was added,

the effect of ionic strength on the sorption PFOS or PFOA was

negligible. All experiments were conducted twice and the

average value was adopted.

2.5. PFOS and PFOA determination

After the sorption experiments, the mixture was filtrated by

a filter with a 0.22 mm nylon membrane. The control experi-

ments indicated that the adsorption of PFOS or PFOA on the

membrane was negligible due to their high concentrations in

solution. The concentrations of PFOS and PFOA were deter-

mined using an LC-10ADvp HPLC with a CDD-6A conductivity

detector from Shimadzu (Japan). The TC-C18 column

(4.6 � 250 mm) from Agilent Technologies (USA) was adopted and the mixture of methanol/0.02 M NaH2PO4 (70/30 for PFOS,

65/35 for PFOA, v/v) was used as the mobile phase at

1.5 mL min�1 flow rate. The sample volume injected was 20 mL.

In this study, the detection limits for PFOS and PFOA are about

1 and 0.7 mg L�1, respectively. The sorption amount was

calculated according to the difference of PFOS or PFOA

concentrations before and after sorption.

3. Results and discussion

3.1. Sorption kinetics

Fig. 1 shows the sorption kinetics of PFOS and PFOA on the

three adsorbents including the GAC, PAC and AI400. Although

all the sorption processes are time dependent, their kinetic

profiles are quite different. It can be seen that both PFOS and

PFOA displayed the very slow sorption kinetics on the GAC

and AI400, and the sorption equilibrium was achieved after at

least 168 h, which may be the reason for the fast breakthrough

of perfluorinated surfactants when the activated carbon filter

was used in German waterworks (Schaefer, 2006). As the GAC

and AI400 are granular porous adsorbents with large propor-

tion of micropores and PFOS and PFOA molecules are about

1 nm in length (Erkoc and Erkoc, 2001; Johnson et al., 2007), it

took long time for the adsorbates to diffuse into the intra-

particle pores. In contrast, only about 4 h was required to

reach the sorption equilibrium for the PAC, suggesting that

the sizes of activated carbon influence the sorption velocity

significantly. The smaller particles have larger external

surface area and more functional groups are available for

PFOS or PFOA sorption, causing the faster adsorption on the

PAC than that on the GAC.

To further understand the sorption kinetics, the pseudo-

second-order model was selected to fit the kinetic data, which

assumes that the sorption rate is controlled by chemical

sorption and the sorption capacity is proportional to the

number of active sites on the sorbent (Ho and McKay, 1999).

This model has been successfully used in many adsorption

processes over the whole time range (Wu et al., 2001; Chiou

and Li, 2003), which can be expressed as follows (Ho and

McKay, 1998).

t qt ¼

1 k2q2e

þ t

qe ¼

1 y0 þ

t qe

(1)

where qe and qt are the amount of PFOS or PFOA adsorbed on

the adsorbents at equilibrium and time t (mmol g�1); k2 is the

sorption rate constant (g mmol�1 h�1); v0 represents the initial

sorption rate (mmol g�1 h�1).

As shown in Fig. 1 and Table 3, the pseudo-second-order

model fitted all the sorption data well according to the rela-

tively high correlation coefficients (r2 > 0.93), indicating that

the chemical interactions were possibly involved in the

0 40 80 120 160 200 0.0

0.1

0.2

0.3

0.4

0.5 GAC

PFOS pH=3 PFOS pH=7 PFOA pH=3 PFOA pH=7

q t ( m

m o

l g

- 1 )

t (h)

0 2 4 6 8 10 12 0.0

0.2

0.4

0.6

0.8

1.0

PFOS pH=3 PFOS pH=7 PFOA pH=3 PFOA pH=7

q t ( m

m o

l g

- 1 )

t (h)

PAC

0 40 80 120 160 200 0

1

2

3

PFOS pH=3 PFOS pH=7 PFOA pH=3 PFOA pH=7

AI400

q t ( m

m o

l g

- 1 )

t (h)

a

b

c

Fig. 1 – Sorption kinetics of PFOS and PFOA on the (a) GAC,

(b) PAC and (c) AI400 and modeling using the pseudo-

second-order equation.

w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 1 1 5 0 – 1 1 5 8 1153

sorption processes. In consideration of the anion property of

the sorbates due to their low pKa (�3.27 for PFOS, 2.5 for PFOA, shown in Table 1) and positive surface charge of the activated

carbons (pHpzc ¼ 7.5) in the pH range studied, the electrostatic interaction may take place between the sorbates and the

sorbents. It should be pointed out that the pKa value for PFOA

is hotly debated, and these reported values range from �0.5 to 3.8 (Burns et al., 2008). The value of 2.5 recognized by the

USEPA (2002) was adopted in this study. Higgins and Luthy

(2006) also found that this interaction played an important

role in the sorption of PFOA and PFOS on the sediments. As for

the sorption of PFOA and PFOS on the AI400, ion exchange was

expected to dominate the sorption, but the sorption amount of

PFOA was much higher than that of PFOS on the resin, indi-

cating that the sorption was complex and other interactions

may also be involved in the sorption process. The reasons will

be discussed in the final paragraph. In addition, it can be seen

in Table 3 that the y0 value for the PAC is much higher than

that for the GAC and AI400, indicating the fast sorption of

PFOS and PFOA on the PAC.

Generally, the adsorption process on a porous adsorbent

can be divided into three stages. The first stage is called

external diffusion, in which the adsorbates move from the

bulk solution to the external surface of the adsorbent; the

second stage is the intraparticle diffusion, and the adsorbates

diffuse further within the adsorbent to the adsorption sites; in

the last stage, the adsorbates are adsorbed at the active sites

on the adsorbent, which is a fast step and usually can be

negligible (Chingombe et al., 2006). As the pseudo-second-

order model cannot give a definite mechanism in the sorption

process, the intraparticle diffusion model was adopted to fit

the sorption kinetics, which can be expressed as (Boyd et al.,

1947; Chiou and Li, 2003; Yang and Al-Duri, 2005)

qt ¼ 6qe r

� D p

�1 2

t 1 2 ¼ kdt

1 2 (2)

where kd (mmol g �1 h�0.5) is the intraparticle diffusion rate

constant, related to the intraparticle diffusivity; r (mm) is the

particle radius; qe (mmol g �1) is the equilibrium sorption

amount; D (mm2 h�1) is the intraparticle diffusivity. The

intraparticle diffusion model assumes that the external

diffusion is negligible and intraparticle diffusion is the only

rate-controlling step, which is usually true for the well-mixed

solution (Yang and Al-Duri, 2005). The good linear relationship

should be obtained in the plot of qt vs. t 0.5 and the line should

also pass through the origin if the intraparticle diffusion is the

rate-controlling step.

Fig. 2 shows the modeling result of PFOS and PFOA sorption

on the three adsorbents using the intraparticle diffusion

model. The good linear plots between the qt and t 0.5 passing

through the origin can be seen for the GAC and AI400 at the

beginning, implying that the sorption kinetics of PFOS and

PFOA on the granular porous adsorbents in the initial stage

followed an intraparticle diffusion-controlled adsorption

(except the sorption of PFOA on the GAC at pH 3). This stage

lasted 72–144 h for PFOS or PFOA sorption at different solution

pH. The kinetic plots for the GAC and AI400 in Fig. 2a and c

exhibited the two-stage linearity, and the latter is final equi-

librium stage where the intraparticle diffusion slowed down.

However, this model failed to fit the sorption kinetics of PFOS

and PFOA on the PAC since the plots not only displayed a bad

linearity but also had significant positive intercepts (Fig. 2b).

This result clearly demonstrates that the intraparticle diffu-

sion was not the rate-controlling step in the sorption of PFOS

or PFOA on the PAC. The adsorbates may easily diffuse into

the inner pores due to the small PAC size, and the external and

Table 3 – Kinetic parameters of the pseudo-second-order model for PFOS and PFOA sorption on the three adsorbents.

Adsorbent Adsorbate Final pH Pseudo-second-order parameter

qe (mmol g �1) y0 (mmol g

�1 h�1) k2 (g mmol �1 h�1) r2

GAC PFOS 3.08 0.51 0.02 0.06 0.989

7.20 0.41 0.01 0.07 0.988

PFOA 3.10 0.38 0.05 0.37 0.943

7.28 0.30 0.01 0.07 0.993

PAC PFOS 3.00 0.69 2.33 4.89 0.984

7.18 0.60 1.96 5.45 0.997

PFOA 3.01 0.79 3.22 5.16 0.990

7.20 0.42 2.45 13.9 0.981

AI400 PFOS 3.10 0.36 0.02 0.16 0.930

7.10 0.34 0.01 0.12 0.940

PFOA 3.02 3.39 0.11 0.01 0.993

7.09 1.81 0.07 0.02 0.992

w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 1 1 5 0 – 1 1 5 81154

intraparticle diffusion may be comparable or even the

external diffusion becomes the rate-limited step.

It can be seen in Fig. 1 that solution pH significantly

affected the sorption of PFOS and PFOA on the three adsor-

bents. As mentioned before, the electrostatic interaction

should be involved in the sorption of anionic PFOS and PFOA

on the activated carbons, which were justified by the

enhanced sorption of PFOS and PFOA on the GAC and PAC

when solution pH decreased from 7 to 3. According to the

obtained parameters in Table 3, the initial sorption rate

increased greatly at lower pH as the parameter y0 for the GAC

and PAC at pH 3 was much higher than that at pH 7.

Solution pH not only influences the properties of the

adsorbent surface, but also affects the adsorbate speciation in

solution. In Fig. 1, it also can be found that solution pH had

more obvious effect on the PFOA sorption onto the three

adsorbents than PFOS, which is probably related to their

speciation at different pH. As solution pH 3 is very close to the

pKa of PFOA, some PFOA may exist in the form of neutral

molecules, but all PFOS molecules are still in anionic form

because of the negative pKa (�3.27). Therefore, the enhanced sorption amount of PFOS at pH 3 can be attributed to the

increased protonated groups on the GAC and PAC, while PFOS

sorption on the AI400 changed little because no protonated

groups were present on the resin. Under the same conditions,

the enhanced sorption amount of PFOA on the adsorbents at

pH 3 should be less than that of PFOS due to the decreased

anionic PFOA, but the much higher enhanced sorption

amount of PFOA on the three adsorbents at pH 3 was found in

Fig. 1, indicating that other interactions must participate in

the sorption at pH 3 and strengthen the sorption of PFOA. In

consideration of the hydrophobic perfluorinated chain of the

PFOA and the hydrophobicity of the activated carbons and

polymer backbone of resin, the hydrophobic interaction

should be also involved in the sorption process, especially this

interaction becomes more obvious at pH 3 since more

hydrophobic neutral PFOA exist in solution. Of course, the

hydrophobic interaction is also expected to occur between

PFOS and the adsorbents since PFOS has the similar per-

fluorinated chain as PFOA. At pH 3, the anionic PFOA species

in solution account for about 76%, higher than the removal

percent of PFOA in all experiments, and thus the anionic PFOA

species are enough to exchange with the anions on the resin

in the sorption process. At the same time, the neutral PFOA

species (about 24%) may adsorb on the adsorbent via the

hydrophobic interaction. Therefore, it’s reasonable that the

PFOA removal by the AI400 at pH 3 is much higher than that at

pH 7 (shown in Fig. 1c), while the removal of PFOS by the resin

hardly changed when solution pH decreased from 7 to 3 due to

the constant anionic PFOS. It was reported that the hydro-

phobic interaction also played an important role in the sorp-

tion of PFOS or PFOA on the sediments, sand and clay (Higgins

and Luthy, 2006; Johnson et al., 2007).

The critical micelle concentration (CMC) for fluorinated

surfactants in aqueous solution is mainly dependent on the

fluorocarbon chain length and the counterion, while other

factors such as temperature, pressure and electrolytes have

little effect on the CMC values. For PFOS or PFOA, the CMC

values are very different due to the different counterions. As

sodium ion is the dominant counterion in solution in our

study, the CMC values for PFOS and PFOA should be about

4573 and 15696 mg L�1, respectively (Kissa, 2001). Because the

PFOS or PFOA concentrations used in this study are far below

their CMC values (less than the CMC by a factor of 100 in most

cases), almost no micelles can form in solution. However, it is

possible to form some hemi-micelles on the adsorbent surface

when the PFOS or PFOA concentrations are in the range of

0.01–0.001 of the CMC (Johnson et al., 2007). Moreover, the

hemi-micelles and even micelles may also form in the inner

pores of the adsorbents after a large number of PFOS or PFOA

molecules adsorb on the porous adsorbents, where the PFOS

or PFOA concentrations are likely much higher than that in

solution. Based on the above discussion, the possible

adsorption models and sorbate–sorbent interactions are

proposed in Fig. 3.

3.2. Sorption isotherms

Sorption isotherm is critical to evaluate the sorption capacity

of adsorbents as well as understand the sorbate–sorbent

interactions. Fig. 4 shows the sorption amount of PFOS and

PFOA on the GAC, PAC and AI400 at different equilibrium

0 4 8 12 0.0

0.1

0.2

0.3

0.4

0.5

PFOS pH=3 PFOS pH=7

PFOA pH=3 PFOA pH=7 Linear fitting

q t ( m

m o

l g

- 1 )

t 1/2

GAC

0 1 2 3 4 0.0

0.2

0.4

0.6

0.8

PFOS pH=3 PFOS pH=7 PFOA pH=3 PFOA pH=7 Linear fitting

t 1/2

q t ( m

m o

l g

- 1 )

PAC

0 4 8 12 0

1

2

3

t 1/2

PFOS pH=3 PFOS pH=7

PFOA pH=3 PFOA pH=7 Linear fitting

q t ( m

m o

l g

- 1 )

AI400

a

b

c

Fig. 2 – Intraparticle diffusion model for the sorption of

PFOS and PFOA on the (a) GAC, (b) PAC and (c) AI400.

micelle

hemi-micelle

B

A

B

B

Activated carbon

N+R3

N+R3

Cl-

micelle

C

N+R3

Cl-N+R3

hemi-micelle

B

N+R3

Anion-exchange resin

a

b

Fig. 3 – Schematic diagram of the sorption of PFOS and

PFOA on the activated carbon (a) and resin (b) via some

possible sorbate–sorbent interactions: A. electrostatic

interaction; B. hydrophobic interaction; C. Ion exchange.

w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 1 1 5 0 – 1 1 5 8 1155

concentrations. Two commonly used models, the Langmuir

and Freundlich equations (Genc-Fuhrman et al., 2004) were

adopted to describe the experimental data, which can be

expressed respectively as

qe ¼ bqmCe 1 þ bCe

(3)

qe ¼ KC 1 n e (4)

where qe is the equilibrium sorption amount (mmol g �1), Ce

represents the equilibrium concentration (mmol L�1) of PFOS

or PFOA in solution, qm is the maximum sorption capacity

(mmol g�1), b is the sorption equilibrium constant (L mmol�1),

K is a constant representing the sorption capacity (mmol1�1/n

L1/n g�1), and n is a constant depicting the sorption intensity.

As shown in Fig. 4 and Table 4, the sorption isotherms of

PFOS on the GAC can be fitted better by the Langmuir model

than the Freundlich model, while the isotherms on the PAC

were described better by the Freundlich model. Both the

Langmuir and Freundlich models can well describe the sorp-

tion isotherms on the AI400. Obviously, it cannot be concluded

that which model described all sorption isotherms better

according to the obtained correlation coefficients. As the

Langmuir equation is derived from the assumption of mono-

layer coverage, the good fitting results of the sorption of PFOS

and PFOA on the adsorbents hint that the possible monolayer

sorption occurred. It is reasonable that the monolayer

0.0 0.1 0.2 0.3 0.4 0.5 0.0

0.1

0.2

0.3

0.4

Langmuir Freundlich

PFOS

PFOA

q e ( m

m o

l g

- 1 )

GAC

0.0 0.1 0.2 0.3 0.4 0.0

0.5

1.0

Langmuir Freundlich

PFOS

PFOA

q e ( m

m o

l g

- 1 )

C e (mmol L

-1 )

C e (mmol L

-1 )

PAC

0.0 0.1 0.2 0.3 0

1

2

3

PFOS PFOA Langmuir Freundlichq

e ( m

m o

l g

- 1 )

C e

(mmol L -1

)

AI400

a

b

c

Fig. 4 – Sorption isotherms of PFOS and PFOA on the GAC,

PAC and AI400 at 25 8C and modeling using the Langmuir

and Freundlich equations. The error bars denote the

standard deviation at each equilibrium concentration

point.

w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 1 1 5 0 – 1 1 5 81156

sorption of PFOS or PFOA on the AI400 occurred since ion

exchange dominated the sorption, and the sorption amount

reached the maximum values when the available sites were

saturated with the PFOS or PFOA molecules. Due to the

hydrophobic perfluorinated chain of PFOA and PFOS, the

multilayer sorption also probably occurred at higher equilib-

rium concentration. The Freundlich model is an empirical

isotherm model usually used in heterogeneous surface energy

systems. According to the results of Freundlich fitting, all the

isotherms are nonlinear as the values of n�1, an indicator of

nonlinearity, are in the range from 0.05 to 0.28 (n�1 ¼ 1 for a linear isotherm). Nonlinearity may result from many causes

such as the sorption site heterogeneity and sorbate–sorbate

interactions (Cheung et al., 2001). In this study, the sorbate–

sorbate interactions such as electrostatic repulsion may be

mainly responsible for the nonlinearity since the high

concentrations of PFOS and PFOA were used and the electro-

static repulsion became significant in the process of intra-

particle diffusion.

It can be found in Fig. 4 that the sorption amount of PFOS

and PFOA on the three adsorbents at the same equilibrium

concentration followed the order of GAC < AI400 < PAC and

GAC < PAC < AI400, respectively. This result is also in agree-

ment with the order of qm and K, which are also the indicators

of the sorption capacity of the adsorbents.

Although the GAC and PAC were obtained from the same

commercial activated carbons, their sorption capacities were

quite different from each other. The differences can be

partially attributed to their different physical properties such

as surface area and pore distribution, shown in Table 2. The

specific surface area of the PAC is a little higher than that of

the GAC, which may explain the higher sorption capacity on

the PAC to some extent, but the slightly higher surface area

cannot lead to about double sorption capacity of the PAC for

PFOS. Due to the big size of PFOS and PFOA molecules, some of

micropores (less than 2 nm) available for N2 adsorption may

be unaccessible for PFOS or PFOA molecules. Therefore, the

available micropores for the sorption of PFOS and PFOA on the

adsorbents are more important than the determined ones.

Because of the smaller size of the PAC than the GAC, more

micropores and surface functional groups on the PAC would

be easily exposed to the sorbates and some active sites origi-

nally unapproachable for the sorbates in the GAC probably

became available ones on the PAC, leading to the higher

sorption capacity. Especially, as shown in Fig. 3a, the micro-

pores on the surface may get blocked by the adsorbed PFOS or

PFOA and even the formed micelles or hemi-micelles, making

the intraparticle pores impassable for the coming adsorbates.

This effect will become more obvious on the GAC than that on

the PAC (Ahn et al., 2007).

It is surprise to find that the sorption capacity of PFOA on

the AI400 is over 5 times as large as that of PFOS, while the

activated carbons prefer the sorption of PFOS to PFOA. If ion

exchange is the dominant mechanism in the sorption process

and all the exchange sites are available for the adsorbates, the

sorption amount of PFOS and PFOA on the AI400 at pH 5

should be comparable. The significant difference in sorption

capacity of PFOS and PFOA on the AI400 may be attributed to

their different CMC and molecular size. As the PFOS molecular

volume is bigger than PFOA (shown in Table 1) and the AI400 is

a gel type micropore resin, the PFOS molecules are more

difficult to diffuse into the AI400. Moreover, since the PFOS

CMC is much lower than that of PFOA, it is easier for PFOS to

form the hemi-micelles or micelles on the surface or inside

the pores of the resin. Due to the bigger size of PFOS, the

micelles or hemi-micelles formed by PFOS should have bigger

sizes than that by PFOA, and easily block the pores and

Table 4 – Calculated equilibrium constants using the Langmuir and Freundlich equations for PFOS and PFOA sorption on the three adsorbents.

Adsorbent Adsorbate Langmuir constants Freundlich constants

qm (mmol g �1) b (L mmol�1) r2 K (mmol(1�1/n) L1/n g�1) n�1 r2

GAC PFOS 0.37 39 0.964 0.43 0.18 0.869

PFOA 0.39 18 0.965 0.47 0.28 0.974

PAC PFOS 1.04 55 0.835 1.27 0.18 0.960

PFOA 0.67 59 0.911 0.83 0.20 0.950

AI400 PFOS 0.42 69 0.944 0.52 0.17 0.930

PFOA 2.92 69 0.967 3.35 0.13 0.987

w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 1 1 5 0 – 1 1 5 8 1157

prevent the intraparticle diffusion of other PFOS molecules,

leading to the much lower sorption amount of PFOS on the

AI400. As perfluorinated chain is hydrophobic and oleophobic,

the adsorbed PFOA molecules may adsorb other PFOA mole-

cules and even form hemi-micelles or micelles at high

concentrations, shown in Fig. 3b. The higher sorption capacity

of PFOS on the GAC and PAC may be related to the more

hydrophobic PFOS. As PFOS has longer perfluorinated chain

than PFOA, the adsorbed PFOS molecules prefer to adsorb

other PFOS molecules to form the hemi-micelles or micelles.

In fact, the sorption mechanisms of PFOS and PFOA on the

adsorbents are complex, and these possible mechanisms need

to be verified in the near future.

4. Conclusions

The sorption kinetics and isotherms of PFOS and PFOA on the

activated carbons and anion-exchange resin were investi-

gated and the PAC was found to be the best adsorbent for PFOS

in terms of sorption kinetics and sorption capacity, while the

AI400 had the highest sorption capacity for PFOA. The sorp-

tion kinetic results reveal that the sorption of PFOS and PFOA

on the granular porous adsorbents including GAC and AI400

was very slow, and the sorption equilibrium was achieved

after at least 168 h, while the sorption equilibrium time was

only about 4 h using the powder activated carbon. Obviously,

the adsorbent size significantly affected the sorption velocity

of PFOS and PFOA. The pseudo-second-order model can well

describe the sorption kinetics of PFOS and PFOA on the three

adsorbents, and the intraparticle diffusion model can well fit

their sorption on the GAC and AI400 in the initial stage. The

sorption isotherms show that the PAC and AI400 were the best

adsorbents for PFOS and PFOA, respectively, and their

maximum sorption capacities were 1.04 and 2.92 mmol g�1

according to the Langmuir model. Additionally, their sorption

capacity for PFOS and PFOA increased with decreasing solu-

tion pH. Besides the electrostatic interaction and ion

exchange, the hydrophobic interaction was also possibly

involved in the sorption. PFOS and PFOA may form the hemi-

micelles or micelles in the adsorbent pores, which signifi-

cantly affected the sorption kinetic and sorption capacity. The

appropriate PAC is the promising adsorbent for PFOS and

PFOA removal from water.

Acknowledgments

We thank the National Nature Science Foundation of China

(project no. 50608045 and 50778095), special fund of State

Key Joint Laboratory of Environment Simulation and Pollu-

tion (project no. 08Z04ESPCT), and National Outstanding

Youth Foundation of China (50625823) for financial support,

and the Program for New Century Excellent Talents in

University is also appreciated. Additionally, the analytical

work was supported by the Laboratory Found of Tsinghua

University.

r e f e r e n c e s

Ahn, C.K., Kim, Y.M., Woo, S.H., Park, J.M., 2007. Selective adsorption of phenanthrene dissolved in surfactant solution using activated carbon. Chemosphere 69, 1681–1688.

Boyd, G.E., Adamson, A.W., Myers Jr., L.S., 1947. The exchange adsorption of ions from aqueous solutions by organic zeolites. II. Kinetics. J. Am. Chem. Soc. 69, 2836–2848.

Brooke, D., Footitt, A., Nwaogu, T.A., 2004. Environmental Risk Evaluation Report: Perfluorooctane Sulfonate (PFOS). UK Environment Agency.

Burns, D.C., Ellis, D.A., Li, H.X., McMurdo, C.J., Webster, E., 2008. Experimental pKa determination for perfluorooctanoic acid (PFOA) and the potential impact of pKa concentration dependence on laboratory-measured partitioning phenomena and environmental modeling. Environ. Sci. Technol. 42, 9283–9288.

Cheung, C.W., Porter, J.F., McKay, G., 2001. Sorption kinetic analysis for the removal of cadmium ions from effluents using bone char. Water Res. 35, 605–612.

Chingombe, P., Saha, B., Wakeman, R.J., 2006. Sorption of atrazine on conventional and surface modified activated carbons. J. Colloid Interface Sci. 302, 408–416.

Chiou, M.S., Li, H.Y., 2003. Adsorption behavior of reactive dye in aqueous solution on chemical cross-linked chitosan beads. Chemosphere 50, 1095–1105.

Erkoc, S., Erkoc, F., 2001. Structural and electronic properties of PFOS and LiPFOS. J. Mol. Struct. (Theochem) 549, 289–293.

Faria, P.C.C., Orfao, J.J.M., Pereira, M.F.R., 2004. Adsorption of anionic and cationic dyes on activated carbons with different surface chemistries. Water Res. 38, 2043–2052.

Fujii, S., Polprasert, C., Tanaka, S., Lien, N.P.H., Qiu, Y., 2007. New POPs in the water environment: distribution, bioaccumulation and treatment of perfluorinated compounds – a review paper. J. Water Supply Res. Technol. – Aqua 56, 313–326.

w a t e r r e s e a r c h 4 3 ( 2 0 0 9 ) 1 1 5 0 – 1 1 5 81158

Genc-Fuhrman, H., Tjell, J.C., McConchie, D., 2004. Adsorption of arsenic from water using activated neutralized red mud. Environ. Sci. Technol. 38, 2428–2434.

Giesy, J.P., Kannan, K., 2002. Perfluorochemical surfactants in the environment. Environ. Sci. Technol. 36, 146A–152A.

Hansen, K.J., Johnson, H.O., Eldridge, J.S., Butenhoff, J.L., Dick, L.A., 2002. Quantitative characterization of trace levels of PFOS and PFOA in the Tennessee River. Environ. Sci. Technol. 36, 1681– 1685.

Higgins, C.P., Luthy, R.G., 2006. Sorption of perfluorinated surfactants on sediments. Environ. Sci. Technol. 40, 7251–7256.

Ho, Y.S., McKay, G., 1998. A comparison of chemisorption kinetic models applied to pollutant removal on various sorbents. Process Saf. Environ. Protect. 76, 332–340.

Ho, Y.S., McKay, G., 1999. Pseudo-second order model for sorption processes. Process Biochem. 34, 451–465.

Hori, H., Nagaoka, Y., Yamamoto, A., Sano, T., Yamashita, N., Taniyasu, S., Kutsuna, S., Osaka, I., Arakawa, R., 2006. Efficient decomposition of environmentally persistent perfluorooctane sulfonate and related fluorochemicals using zerovalent iron in subcritical water. Environ. Sci. Technol. 40, 1049–1054.

Johnson, R.L., Anschutz, A.J., Smolen, J.M., Simcik, M.F., Penn, R.L., 2007. The adsorption of perfluorooctane sulfonate onto sand, clay, and iron oxide surfaces. J. Chem. Eng. Data 52, 1165–1170.

Key, B.D., Howell, R.D., Criddle, C.S., 1997. Fluorinated organics in the biosphere. Environ. Sci. Technol. 31, 2445–2454.

Key, B.D., Howell, R.D., Criddle, C.S., 1998. Defluorination of organofluorine sulfur compounds by Pseudomonas sp. strain D2. Environ. Sci. Technol. 32, 2283–2287.

Kissa, E., 2001. Fluorinated Surfactants and Repellents, second ed. Marcel Dekker, New York.

Lampert, D.J., Frisch, M.A., Speitel Jr., G.E., 2007. Removal of perfluorooctanoic acid and perfluorooctane sulfonate from wastewater by ion exchange. Pract. Periodical Haz., Toxic, and Radioactive Waste Mgmt. 11, 60–68.

Loos, R., Locoro, G., Huber, T., Wollgast, J., Christoph, E.H., De Jager, A., Gawlik, B.M., Hanke, G., Umlauf, G., Zaldivar, J.M., 2008. Analysis of perfluorooctanoate (PFOA) and other perfluorinated compounds (PFCs) in the River Po watershed in N-Italy. Chemosphere 71, 306–313.

Moody, C.A., Martin, J.W., Kwan, W.C., Muir, D.C.G., Mabury, S.C., 2002. Monitoring perfluorinated surfactants in biota and surface water samples following an accidental release of fire- fighting foam into Etohicoke Creek. Environ. Sci. Technol. 36, 545–551.

Moriwaki, H., Takagi, Y., Tanaka, M., Tsuruho, K., Okitsu, K., Maeda, Y., 2005. Sonochemical decomposition of

perfluorooctane sulfonate and perfluorooctanoic acid. Environ. Sci. Technol. 39, 3388–3392.

Ochoa-Herrera, V., Sierra-Alvarez, R., 2008. Removal of perfluorinated surfactants by sorption onto granular activated carbon, zeolite and sludge. Chemosphere 72, 1588–1593.

Ochoa-Herrera, V., Sierra-Alvarez, R., Somogyi, A., Jacobsen, N.E., Wysocki, V.H., Field, J.A., 2008. Reductive defluorination of perfluorooctane sulfonate. Environ. Sci. Technol. 42, 3260–3264.

Prevedouros, K., Cousins, I., Buck, R.C., Korzeniowski, S.H., 2006. Sources, fate and transport of perfluorocarboxyates. Environ. Sci. Technol. 40, 32–44.

Schaefer, A., 2006. Perfluorinated surfactants contaminate German waters. Environ. Sci. Technol. 40, 7108–7109.

Schroder, H.F., Meesters, R.J.W., 2005. Stability of fluorinated surfactants in advanced oxidation processes – a follow up of degradation products using flow injection-mass spectrometry, liquid chromatography–mass spectrometry and liquid chromatography–multiple stage mass spectrometry. J. Chromatogr. A 1082, 110–119.

Sinclair, E., Mayack, D.T., Roblee, K., Yamashita, N., Kannan, K., 2006. Occurrence of perfluoroalkyl surfactants in water, fish, and birds from New York State. Arch. Environ. Contam. Toxicol. 50 (3), 398–410.

Tang, C.Y.Y., Fu, Q.S., Robertson, A.P., Criddle, C.S., Leckie, J.O., 2006. Use of reverse osmosis membranes to remove perfluorooctane sulfonate (PFOS) from semiconductor wastewater. Environ. Sci. Technol. 40, 7343–7349.

Tang, C.Y.Y., Fu, Q.S., Criddle, C.S., Leckie, J.O., 2007. Effect of flux (transmembrane pressure) and membrane properties on fouling and rejection of reverse osmosis and nanofiltration membranes treating perfluorooctane sulfonate containing wastewater. Environ. Sci. Technol. 41, 2008–2014.

USEPA, 2002. Revised Draft Hazard Assessment of Perfluorooctanoic Acid and its Salts Available from: http:// www.ewg.org/files/EPA_PFOA_110402.pdf.

Wu, F.C., Tseng, R.L., Juang, R.S., 2001. Kinetic modeling of liquid- phase adsorption of reactive dyes and metal ions on chitosan. Water Res. 35, 613–618.

Yamamoto, T., Noma, Y., Sakai, S.I., Shibata, Y., 2007. Photodegradation of perfluorooctane sulfonate by UV irradiation in water and alkaline 2-propanol. Environ. Sci. Technol. 41, 5660–5665.

Yang, X., Al-Duri, B., 2005. Kinetic modeling of liquid-phase adsorption of reactive dyes on activated carbon. J. Colloid Interface Sci. 287, 25–34.

Yu, Q., Deng, S.B., Yu, G., 2008. Selective removal of perfluorooctane sulfonate from aqueous solution using chitosan-based molecularly imprinted polymer adsorbents. Water Res. 42, 3089–3097.