OCCURRENCE AND FATE OF PERFLUOROOCTANE SULFONATE (PFOS) AND PERFLUOROOCTANOIC … · 2018. 1....

OCCURRENCE AND FATE OF

PERFLUOROOCTANE SULFONATE (PFOS) AND

PERFLUOROOCTANOIC ACID (PFOA) IN WATER

AND WASTEWATER AND THEIR REMOVAL

USING A HYBRID PAC-MBR SYSTEM

YU JING

NATIONAL UNIVERSITY OF SINGAPORE

2010

brought to you by COREView metadata, citation and similar papers at core.ac.uk

provided by ScholarBank@NUS

https://core.ac.uk/display/48637644?utm_source=pdf&utm_medium=banner&utm_campaign=pdf-decoration-v1

OCCURRENCE AND FATE OF

PERFLUOROOCTANE SULFONATE (PFOS) AND

PERFLUOROOCTANOIC ACID (PFOA) IN WATER

AND WASTEWATER AND THEIR REMOVAL

USING A HYBRID PAC-MBR SYSTEM

YU JING

(M.Eng., Southeast University)

A THESIS SUBMITTED

FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

DEPARTMENT OF CIVIL ENGINEERING

NATIONAL UNIVERSITY OF SINGAPORE

2010

i

ACKNOWLEDGEMENT

I would like to take this opportunity to acknowledge and thank all those who

have helped me along the way.

First and foremost, I would like to express my utmost appreciation from

bottom of my heart to my academic supervisor, Associate Professor Jiangyong

HU, for sticking with me through these many years. Without her professional

guidance, constructive advice and constant encouragement, this work could

not have been completed.

Acknowledgements are made to all former and current staff at Environmental

Laboratory in the Centre for Water Research for their kind support and

cooperation in various ways. Heartfelt thanks are also given to my former

Final Year Project students who have made contributions to this work in

whatever aspects.

At last but not least, I would like to thank my family for their vital support and

consistent encouragement. Their love is always the power source for me to

struggle all difficulties in my life.

The research scholarship provided by the National University of Singapore

throughout the whole period of candidature is greatly appreciated.

ii

TABLE OF CONTENTS

ACKNOWLEDGEMENT…………………………………………...……….i

TABLE OF CONTENTS…………………………………………...………..ii

SUMMARY…………………………………………………..…………….viii

NOMENCLATURE…………………………………………………...……xii

LIST OF TABLES…………………………………………………...……...xv

LIST OF FIGURES…………………………………………..…………...xvii

CHAPTER 1 INTRODUCTION……………………………………………1

1.1 Background………………………………………………………….1

1.2 Objective and Scope of Study………………………………………4

1.3 Outline of Thesis…………………………………………………….5

CHAPTER 2 LITERATURE REVIEW....................………………………7

2.1 Introduction…………………………………………………………7

2.1.1 Physico-chemical properties of PFOS and PFOA…………7

2.1.2 Persistence, bioaccumulation and toxicity of PFOS and

PFOA...............................................................................................10

2.1.2.1 Persistence………………………………………………...10

2.1.2.2 Bioaccumulation………………………………………….11

2.1.2.3 Toxicity……………………………………………………11

2.1.3 Preliminary regulations for PFOS and PFOA……………12

2.2 Analytical method for PFCs ............................................................12

2.2.1 Introduction of LC/MS/MS analysis for PFCs…………...12

2.2.2 LC/MS/MS method for water and wastewater…………...13

2.2.3 LC/MS/MS method for sludge and sediment……………..14

iii

2.2.4 Limitation of Electrospray Ionization (ESI)……………...15

2.2.5 Matrix interference………………………………………...16

2.2.6 Post extraction clean-up method for analysis of

environmental matrices………………………………………...18

2.3 Occurrence of PFOS/PFOA in the environment...........................19

2.3.1 Occurrence in the surface water…………………………..19

2.3.2 Occurrence in the drinking water…………………………24

2.3.3 Occurrence in the seawater………………………………..25

2.3.4 Occurrence in the sludge and sediment…………………...26

2.4 Fate and behavior in the sewage treatment plants .......................27

2.4.1 Occurrence in the wastewater……………………………..27

2.4.2 Mass flow and mass change………………………………..28

2.5 Removal Technologies .....................................................................29

2.5.1 Advanced oxidation process……………………………….29

2.5.2 RO/NF membrane………………………………………….31

2.5.3 Adsorption…………………………………………………..32

2.5.3.1 Activated carbon adsorption…………………………….32

2.5.3.2 Adsorption onto sediment and sludge…………………..33

2.5.4 Membrane biological reactor (MBR)……………………..34

2.5.4.1 Introduction………………………………………………34

2.5.4.2 Configuration and application…………………………..35

2.5.4.3 Technology benefits and problems……………………...36

2.5.4.4 Hybrid PAC-MBR system……………………………….38

2.6 Research statement ..........................................................................40

CHAPTER 3 MATERIALS AND METHODS ...........................................48

iv

3.1 Chemicals, materials and reagents .................................................48

3.1.1 Chemicals and reagents……………………………………48

3.1.2 Materials…………………………………………………….48

3.2 Water sample collection and preparation......................................50

3.2.1 Water sample collection……………………………………50

3.2.2 Water sample preparation (Basic SPE extraction)………52

3.3 Experiment on development of clean-up method for wastewater

and sludge sample ..................................................................................53

3.3.1 Silica cartridge clean-up procedure……………………….53

3.3.2 Application of clean-up method to sludge samples………53

3.3.3 Evaluation of matrix effect and recoveries……………….54

3.4 Wastewater and sludge sample collection and preparation .........55

3.4.1 Wastewater and sludge sample collection………………...55

3.4.2 Wastewater and sludge sample preparation……………...57

3.5 PAC-MBR experimental setup and operation ..............................58

3.5.1 MBR configuration…………………………………………58

3.5.2 Synthetic wastewater and operational conditions………..59

3.5.3 PFCs mass balance calculation……………………………62

3.5.4 Membrane resistance calculation…………………………62

3.6 Adsorption study on PAC and activated sludge ...........................63

3.6.1 Preparation of EfOM………………………………………63

3.6.2 EfOM characterization…………………………………….64

3.6.3 Equilibrium adsorption experiments……………………...64

3.6.4 Adsorption kinetics experiments…………………………..65

3.6.5 Mathematical modeling……………………………………66

v

3.7 Analysis method ...............................................................................67

3.7.1 COD, DOC and UV254 analysis…………………………..67

3.7.2 Carbohydrate and protein analysis……………………….67

3.7.3 MLSS and MLVSS…………………………………………67

3.7.4 EPS and SOUR analysis……………………………………67

3.8 LC/MS/MS analysis .........................................................................68

3.8.1 Optimization of LC/MS/MS detection method…………...68

3.8.2 Method validation and quantification…………………….70

3.9 Fractionation process.......................................................................72

3.10 Quality assurance and control ......................................................74

3.11 Statistical analysis ..........................................................................75

CHAPTER 4 OCCURRENCE of PFOS AND PFOA IN WATER AND

WASTEWATER ............................................................................................76

4.1 Introduction ......................................................................................76

4.2 Results and discussions ....................................................................78

4.2.1 PFOS/PFOA Concentration in surface water…………….78

4.2.2 PFOS/PFOA Concentration in wastewater………………84

4.2.3 PFOS/PFOA Concentration in coastal water…………….87

4.2.4 Seasonal variations of PFOS and PFOA………………….89

4.2.5 Correlations between PFOS and PFOA…………………..90

4.3 Summary ...........................................................................................93

CHAPTER 5 DEVELOPMENT OF POST EXTRACTION CLEAN-UP

METHOD FOR PFOS/PFOA DETERMINATION IN WASTEWATER

AND SLUDGE SAMPLES............................................................................96

5.1 Introduction ......................................................................................96

vi

5.2 Results and discussions ....................................................................99

5.2.1 Effect of clean-up procedures on matrix effect…………..99

5.2.2 Effect of internal standards on matrix effect……………101

5.2.3 Detection of PFOS and PFOA in water and sludge

samples…………………………………………………..………103

5.3 Summary .........................................................................................104

CHAPTER 6 BEHAVIOR OF PFOS AND PFOA IN SEWAGE

TREATMENT PLANTS .............................................................................106

6.1 Introduction ....................................................................................106

6.2 Results and Discussion ...................................................................108

6.2.1 PFOS/PFOA in wastewater ………………………………109

6.2.2 Seasonal variation of PFOS and PFOA………………….112

6.2.3 Mass flow in aqueous sample during treatment………...114

6.2.4 PFOS/PFOA in sludge…………………………………….118

6.3 Summary .........................................................................................121

CHAPTER 7 PFOS/PFOA REMOVAL BY HYBRID PAC-MBR

PROCESS .....................................................................................................124

7.1 Introduction ....................................................................................124

7.2 Results and Discussion ...................................................................128

7.2.1 Adsorption study on PAC and activated sludge………...128

7.2.1.1 Characterization of EfOM……………………………...128

7.2.1.2 PFOS and PFOA adsorption onto PAC……………….129

7.2.1.3 Effect of EfOM on the PFOS and PFOA adsorption onto

PAC………………………………………………………………132

7.2.1.4 PFOS and PFOA adsorption onto activated sludge…..134

vii

7.2.2 Performance of MBR and PAC-MBR systems at different

SRT ………………………………………………………………137

7.2.2.1 Overall performance of MBR and PAC-MBR system..137

7.2.2.2 SMP and DOM fraction characteristics……………….140

7.2.3 Removal of PFOS and PFOA in PAC-MBR and MBR

system……………………………………………………………142

7.2.3.1 Removal by adsorption onto activated sludge………...142

7.2.3.2 Removal by adsorption onto PAC……………………..143

7.2.3.3 Mass balance…………………………………………….145

7.2.3.4 Effect of SRT on PFOS and PFOA removal…………..150

7.2.3.5 Effect of PAC dosage on PFOS and PFOA removal….152

7.2.4 Membrane fouling………………………………………...153

7.2.4.1 Variations of TMP………………………………………153

7.2.4.2 Effect of PAC on TMP………………………………….154

7.3 Summary .........................................................................................155

CHAPTER 8 CONCLUSIONS ...................................................................160

8.1 Conclusions .....................................................................................160

8.2 Recommendations ..........................................................................168

REFERENCE ...............................................................................................170

APPENDIX: PUBLICAIONS .....................................................................193

viii

Summary

Perfluorinated compounds (PFCs), particularly perfluorooctane sulfonate

(PFOS, C8F17SO3-) and perfluorooctanoic acid (PFOA, C7F15COO-

), have

emerged as a new class of environmentally persistent pollutants, which have

been widely used in different applications. PFOS and PFOA, regarded as the

terminal breakdown end-products of PFCs, have been detected in a wide array

of environmental matrices including biota, water, air, sediment and sludge.

The primary objective of this thesis is to contribute towards establishment of

fundamental understanding of fate and behavior of PFCs in the aquatic

environment and sewage treatment plants (STPs) as well as development of a

hybrid PAC-MBR process to effectively remove these two trace organic

compounds. More than one hundred water samples from reservoirs,

rivers/canals, coastal waters and treated effluents of wastewater treatment

plants (WWTPs) were collected and analyzed to characterize the spatial

distribution and seasonal variation of PFOS and PFOA in the aquatic and

oceanic environment of Singapore. Coastal waters had lower concentrations of

PFOS and PFOA as compared to surface waters and wastewaters, while

highest concentration of PFOS and PFOA were observed in treated effluents

of two WWTPs. Our results suggest that coastal waters in the western area of

Singapore were more heavily contaminated compared to those in the middle

and eastern areas. Between dry and wet season, significant seasonal difference

(p=0.025) was observed in surface waters for PFOS only, while no discernable

ix

seasonal differences were found for both PFOS and PFOA in coastal waters

and wastewaters.

An efficient sample clean-up method was developed in this study to

significantly remove co-eluting matrix components by applying the SPE

extracts onto a silica cartridge. Internal standardization was used to further

compensate for the matrix effect, which was also proven to improve the signal

reproducibility. The clean-up method described in this study was applied to

different water samples (surface water and wastewater) and sludge samples to

evaluate the efficiency of silica clean-up and the influence of sample origin on

the matrix effect. Results showed that the method was robust and could be

applied to analyze PFOS and PFOA in different environmental matrices. In

water and sludge samples, matrix effect and recovery efficiency were in the

range of 91.8%-98.3% and 81.3%-98.0%, respectively, indicating that clean-

up method can effectively remove co-eluting matrix components in various

environmental matrices.

The behavior of PFOS and PFOA in the biological units of various full-scale

municipal sewage treatment plants was also investigated. Mass flow of PFOS

increased significantly (mean 62.2%) in conventional activated sludge process

(CAS) of plant B, while it remained consistent after the secondary treatment

in plant A. Mass flow of PFOA increased 82.9% (mean) in CAS of plants A

and B and 62.3% (mean) in membrane biological reactor (MBR), while it

remained unchanged after the treatment of liquid treatment module (LTM). In

terms of behavior pattern of PFOS and PFOA, our results suggest that there

x

was no significant difference between conventional activated sludge process

and membrane biological reactor operated at comparable sludge retention time

(SRT). However, mass flow of these two compounds remained consistent

after treatment of activated sludge process operating at short SRT. Seasonal

variations of PFOS in concentrations of raw sewage were found in plant A,

while PFOA did not have significant seasonal variation in both plants A and B.

The adsorption of PFOS and PFOA onto powdered activated carbon (PAC)

was investigated in the presence and absence of effluent organic matter

(EfOM) at low concentration range (0.1-500 µg/L). Adsorption of PFOS and

PFOA onto PAC fitted the Freundlich model well (r2>98%) and adsorption

capacity of PFOS (KF=17.55) and PFOA (KF=10.03) in the absence of EfOM

was more than one order of magnitude higher than that in the presence of

EfOM, indicating EfOM greatly reduce the adsorption capacity of PAC.

Moreover, the EfOM fraction of <1 k, which had greater effect on the

adsorption than that of >30 k fraction, was the major contributor to the

adsorption competition. Additionally, the estimated partition coefficient Kd

was 729 and 154 L/kg for PFOS and PFOA, respectively, suggesting PFOS

and PFOA, especially PFOA, have a low tendency to partition onto the

activated sludge.

The overall performance and removal efficiencies of PFCs were also

investigated in PAC-MBRs which were operated with different PAC dosages

and SRTs. On the one hand, the effect of PAC dosage on the removal of PFCs

in PAC-MBR was studied at the SRT of 30 d. Removal efficiency of PFCs

xi

increased with the increase of PAC dosage from 30 to 100 mg/L, suggesting

adsorption on PAC was the efficient and predominant process in the removal

of PFCs in activated sludge system. On the other hand, the effect of SRT on

removal of PFCs in PAC-MBR was studied. Removal efficiencies of PFCs

were >90% for PFOS and >84% for PFOA at different SRT studied,

suggesting that adsorption onto PAC could be dominant and removal

efficiencies may be not significantly affected by different operational SRTs.

With the increase of SRT, PFCs concentration on PAC decreased significantly,

indicating significant effect of SRT on the PAC adsorption capacity in PAC-

MBR due to different PAC concentrations at different SRTs.

Keywords: PFOS, PFOA, Aquatic environment, Sewage treatment

plant (STP), Hybrid powdered activated carbon-membrane biological

reactor (PAC-MBR), Fate and behavior

xii

NOMENCLATURE

AHS Acid Humic Substance

AMWD Apparent Molecular Weight Distribution

BAFs Bioaccumulation Factors

CAS Conventional Activated Sludge Treatment

COD Chemical Oxygen Demand

DOC Dissolved Organic Carbon

DOMs Dissolved Organic Matters

EDCs Endocrine Disruptors Compounds

EfOMs Effluent Organic Matters

EPS Extra Cellular Polymeric Substance

ESI Electrospray Ionization

GC Gas Chromatography

GAC Granular Activated Carbon

HDPE High Density Polyethylene

HPLC High-performance Liquid Chromatography

HRT Hydraulic Retention Time

IDL Instrumental Detection Limit

J Permeate Flux

Kd Partition Coefficient

Koc Organic Carbon Partition Coefficient

Kow Octanol-water Partition Coefficient

LOD Limit Of Detection

xiii

LOQ Limit Of Quantification

MBR Membrane Bioreactor

ME Matrix Effect

MF Microfiltraiton

MLSS Mixed Liquor Suspended Solids

MRM Multiple Reaction Monitoring

MS Mass Spectrometry

NMR Nuclear Magnetic Resonance

NOMs Natural Organic Matters

NPS Nonpoint Source Pollution

PAC Powdered Activated Carbon

pKa Acid Dissociation Constant

PFASs Perfluoroalkyl Sulfonates

PFCAs Perfluoroalkyl Carboxylates

PFCs Perfluorinated Compounds

PFDoA Perfluorododecanoic Acid

PFEES Pperfluoro (2-ethoxyethane) Sulfonic Acid

PFO Perfluorooctanoate

PFOS Perfluorooctane Sulfonate

PFOA Prfluorooctanoic Acid

POPs Persistent Organic Pollutants

RE Recovery Efficiency

Rm Intrinsic Membrane Resistance

Rr Reversible Resistance

Ri Irreversible Resistance

xiv

SMPs Soluble Microbial Products

SOUR Specific Oxygen Uptake Rate

SPE Solid Phase Extraction

SRT Sludge Retention Time

STPs Sewage Treatment Plants

TMP Transmembrane Pressure

UF Ultrafiltraton

VP Vapor Pressure

VSS Volatile Suspended Solids

WWTPs Wastewater Treatment Plants

WAS Waste Activated Sludge

xv

LIST OF TABLES

Table 2.1 Definition of acronyms and structures of PFCAs and PFSAs….8

Table 2.2 Physico-chemical properties of PFOS and PFOA…………......10

Table 2.3 Review of PFOS and PFOA concentrations in surface water

(ng/L), drinking water (ng/L), wastewater (ng/L) and sludge (ng/g)….....21

Table 2.4 Sorption and desorption coefficients of PFOS from various

matrices...........................................................................................................33

Table 3.1 Characteristics of powdered activated carbon ...........………...49

Table 3.2 characteristics of the MF hollow fiber membrane………….....50

Table 3.3 Wastewater treatment plants characteristics………………......52

Table 3.4 Composition and concentration of synthetic wastewater……..60

Table 3.5 PAC added at the startup of PAC-MBR system…………….....61

Table 3.6 MRM-transitions, compound-dependent parameters of the

analytes............................................................................................................69

Table 3.7 IDL, LOD and LOQ of PFOS and PFOA…………………..….72

Table 5.1 Influence of sample clean-up and internal standardization on

ME% and RE% (n=5)………………………………………………….…101

Table 5.2 Influence of sample origin on ME% and RE% with internal

standard (n=3)……………………………………..……………..………..103

Table 5.3 ME% and RE% with internal standard by application of clean-

up method on sludge samples (n=3)………………….………………..….104

Table 6.1 Mass flow (mg/d) of PFCs in influent, effluent and solid waste in

CAS1, CAS2, LTM and MBR……………………...……………………..116

xvi

Table 6.2 Calculated partition coefficient Kd in primary sludge and

activated sludge……………………………………………………...……..121

Table 7.1 Characteristics of EfOM solution obtained from the lab scale

MBR (n=5)…………………………………………………………………128

Table 7.2 Langmuir isotherm constants and Freundlich isotherm

constants for the adsorption of PFCs onto PAC at 25 oC…………….....131

Table 7.3 Freundlich iostherm parameters for the adsorption of PFCs on

PAC in EfOM fractions……………………………….…………………..134

Table 7.4 Linear isotherm parameters for PFCs onto activated sludge..135

Table 7.5 Measured PFCs concentrations in activated sludge of MBR at

different SRT……………………………..………………………………..137

Table 7.6 Estimated mass flows of PFCs in activated sludge of WAS in

PAC-MBR operated at different SRTs…………………………..………150

Table 7.7 Effect of SRT on the PFCs removal in PAC-MBR system with

PAC dosage of 100 mg/L (based on mass balance)…………………..…..152

Table 7.8 Effect of PAC dosage on the PFCs removal in PAC-MBR

system (based on mass balance)…………………..………………………153

Table 7.9 Resistances of membrane for the MBR and PAC-MBR

systems……………………………………………………………………...155

xvii

LIST OF FIGURES

Figure 1.1 Research scope and content…………………………..…………5

Figure 2.1 Structures of PFOS and PFOA……………………………...…..9

Figure 2.2 Average mass flow (mg/d) for PFOS and PFOA in sewage

treatment plant...............................................................................................29

Figure 2.3 Configuration of MBR systems: (a) Side-stream MBR, (b)

Suctioned- filtration submerged MBR, and (c) Gravitational-filtration

submerged MBR…………………………………………………………….35

Figure 3.1 Sampling locations for reservoir waters, river/canal waters,

effluents of WWTPs, coastal waters and location for outfalls of

WWTPs…………………………………………………………..………….51

Figure 3.2 Flow scheme of the sewage treatment plants A and B……......56

Figure 3.3 Schematic diagram of lab-scale PAC-MBR system………......59

Figure 3.4 Mass balances of PFCs in PAC-MBR or MBR system……….62

Figure 3.5 LC/MS/MS chromatograms of PFOS, PFOA and internal

standards PFEES and PFDoA……………………………………………...70

Figure 3.6 Procedure for fractionation of DOM……………………..…...74

Figure 4.1 Concentrations of PFOS and PFOA in surface waters,

wastewaters and coastal waters from western area of Singapore collected

by: 1. Oct 2006; 2. Dec 2006; 3. Mar 2007; 4. Sep 2007; 5. Dec 2007…...80


wastewaters and coastal waters from middle area of Singapore collected

by: 1. Oct 2006; 2. Dec 2006; 3. Mar 2007; 4. Sep 2007; 5. Dec 2007…...81

xviii


wastewaters and coastal waters from eastern area of Singapore collected

by: 1. Oct 2006; 2. Dec 2006; 3. Mar 2007; 4. Sep 2007; 5. Dec 2007…....82

Figure 4.4 Total PFOS and PFOA concentrations in surface waters

summed up by 5 sampling campaigns.……………………………………83

Figure 4.5 Total PFOS and PFOA concentrations in wastewaters summed

up by 5 sampling campaigns……………………………………………….87

Figure 4.6 Total PFOS and PFOA concentrations in coastal waters

summed up by 5 sampling campaigns……………………………………..88

Figure 4.7 Correlations of PFOS and PFOA between coastal water C4

and wastewater W2........................................................................................89

Figure 4.8 Correlations between PFOS and PFOA concentrations in

surface waters.………………………………………………………………92


coastal waters…………….…….……………………………………………92


wastewaters……………………..…………………………………………...92

Figure 4.11 Correlations between PFOS and PFOA concentrations in the

effluents of (a) W2 and W4; (b) W5………………………...…………...93

Figure 5.1 LC-MS-MS chromatograms of PFOS and PFOA in the raw

sewage extracted by (a) HLB SPE and (b) HLB+silica…………..…….100

Figure 6.1 PFOS concentrations in wastewater of STP A (CAS1, LTM

and MBR) and STP B (CAS2)…………………………………………….109

Figure 6.2 PFOA concentrations in wastewater of STP A (CAS1, LTM

and MBR) and STP B (CAS2)…………….…………………...………….111

xix

Figure 6.3 Seasonal variations in influent concentrations of (a) PFOS and

(b) PFOA in STPA. 1: Oct 06 (CAS1), 2: Mar 07 (CAS1), 3: Sep 07

(CAS1), 4: Mar 07 (MBR), 5: Sep 07 (MBR), 6: Oct 06 (LTM), 7: Mar 07

(LTM), 8: Dec 06 (CAS1), 9: Dec 06 (LTM), 10: Dec 07 (CAS1), 11: Dec

07 (MBR)......................................................................................................113

Figure 6.4 Seasonal variations in influent concentrations of (a) PFOS and

(b) PFOA in STP B……………………………….………………………..114

Figure 6.5 Change of mass flow after primary treatment in (a) STP A and

(b) STP B……………….…………………………………………………..118

Figure 6.6 PFOS concentrations in sludge samples from STP A and STP

B………………………………………………………………………….....119

Figure 6.7 PFOA concentrations in sludge samples from STP A and STP

B…………………………………………………………………………….120

Figure 7.1 Adsorption isotherms of PFCs onto the PAC in the absence

and presence of EfOM: (a) PFOS; (b) PFOA. Experimental data fit to

Freundlich model (solid line)…………………...…………………………131

Figure 7.2 Adsorption of PFOS and PFOA onto PAC as a function of

contact time: (a) in the presence of EfOM; (b) in the Milli-Q water…...132

Figure 7.3 Log-log plot of PFCs adsorption isotherms in the presence and

absence of EfOM fractions: (a) PFOS and (b) PFOA…………………...133

Figure 7.4 Adsorption isotherms of PFCs onto the activated sludge…...136

Figure 7.5 COD removal in MBR and PAC-MBR systems with different

SRTs……………………………………………….………………..………138

Figure 7.6 DOC of supernatant and effluent in MBR and PAC-MBR

system with different SRTs………………………………………..………138

xx

Figure 7.7 MLSS/MLVSS in MBR and PAC-MBR systems with different

SRTs………………………..………………………………………………139

Figure 7.8 SOUR in MBR and PAC-MBR systems with different

SRTs………………………………………………………………………...140

Figure 7.9 AMWD of SMP in the supernatant of (a) MBR and (b) PAC-

MBR systems at different SRTs…………….…………………………….141

Figure 7.10 Hydrophobicity of DOM in the supernatant of (a) MBR and

(b) PAC-MBR systems at different SRTs……………….………..……...141

Figure 7.11 PFCs removal in MBR with different SRTs…………..…...143

Figure 7.12 PFCs removal in PAC-MBR system operated with different

PAC dosages…………………………………………………………….…144

Figure 7.13 PFCs removal in PAC-MBR system with PAC dosage of 100

mg/L at different SRTs………………………….……..……….…………145

Figure 7.14 Distribution of removed PFCs flow in MBR operated at

different SRT: (a) PFOS; (b) PFOA……………………..………..……..147

Figure 7.15 Estimated distributions of removed PFCs mass flow in waste

of PAC-MBR at SRT 30 d with different PAC dosage: (a) PFOS; (b)

PFOA………………………………………………………….……..……..148

Figure 7.16 Estimated distributions of removed PFCs mass flow in waste

of PAC-MBR operated at different SRTs: (a) PFOS; (b)

PFOA…………………………………………………………….…………149

Figure 7.17 Effect of SRT on the PFCs adsorption onto activated sludge in

MBR: (a) PFOS; (b) PFOA……………………………………….……...150

Figure 7.17 Long-term TMP profile for the MBR and PAC-MBR systems

at different SRTs……………………….…………………………………154

Chapter 1-Introduction

1

CHAPTER 1 INTRODUCTION

Perfluorinated compounds (PFCs) have been manufactured for over 50 years

and, due to their unique properties of repelling both water and oil, they have

been used as surfactants and surface protectors in carpets, leather, paper, food

containers, fabric, and upholstery and as performance chemicals in products

such as fire-fighting foams, floor polishes, and shampoos. Widespread use of

PFCs has led to ubiquitous occurrence of these chemicals in the environment

particularly Perfluorooctane sulfonate (PFOS, C8F17SO3-) and

perfluorooctanoic acid (PFOA, C7F15COO-), which are the final breakdown

products of PFCs. PFOS and PFOA are also well known for their application

in production of Teflon and other stain resistant materials.

1.1 Background

The occurrence of PFOS and PFOA have been reported in human blood,

biological tissues, water, air, sludge, sediment, and soil since 1968 that they

were first detected with nuclear magnetic resonance (NMR) spectroscopy

(Taves, 1968; Giesy et al., 2001; Taniyasu et al., 2003; Saito et al., 2003;

Martin et al., 2003; Higgins et al., 2005; Sinclair et al., 2006). Currently, high-

performance liquid chromatography (HPLC) with triple quadrupole mass

spectrometry in electrospray negative mode is the most promising and

extensively applied method for analyzing PFCs in various environmental and

biological matrices (Giesy et al., 2001; Taniyasu et al., 2003; Kannan et al.,


2

2004; Moody et al., 2001; Hansen et al., 2001; Martin et al., 2004b). Analysis

was accomplished by direct injection (Schultz et al., 2006) or preconcentration

on solid phase extraction (SPE) cartridges, followed by LC/MS/MS analysis

(Giesy et al., 2001; Tomy et al., 2004; Becker et al., 2008; Boulanger et al.,

2005; Sinclair et al., 2006).

It was reported that PFOS and PFOA were detected in surface waters (Hansen

et al., 2002; Boulanger et al., 2004; Loos et al., 2008), wastewaters (Boulanger

et al., 2005; Becker et al., 2008; Yu et al., 2009), drinking waters (Harada et

al.,2003), ground waters (Schultz et al., 2004) and coastal waters (So et al.,

2004; Saito et al., 2003; Yamashita et al.,2005) all over the world. As their

ubiquitous presence in the environment, PFOS and PFOA arouse great

concerns due to their impact on animal and human. They are known to cause

acute and subchronic toxicity effects in laboratory studies (Haughom et al.,

1992; Seacat et al., 2003). One main concern is their persistence and

bioaccumulativity on live tissue. PFOS and PFOA are readily absorbed by

mammals following oral and inhalation exposure. Once absorbed in the body,

they distribute mainly in the serum and the liver (Kudo et al., 2003; OECD,

2002; US EPA, 2003). However, there is no evidence of any metabolic

degradation of PFOS and PFOA (Kissa, 2001; Schultz et al., 2003).

Furthermore, both chemicals are poorly excreted in both urine and feces.

Biological half-life of PFOA in plasma of a few days for mice and rats and

approximately 4.4 years for humans are reported (Kudo et al., 2003). Half-life

of PFOS varies from 7.5 days in rats to 8.7 years in humans, estimated from

retired 3M production workers (3M, 1999; OECD, 2002; Thibodeaux et al.,


3

2003). Another main concern regarding their adverse effect on animals is

endocrine disruption. A well-known case, for example, is that some male

fishes that are exposed to these pollutants may undergo feminization. These

compounds can bind with the natural estrogen receptors (ER) in the organism

body, and consequently interfere with the normal binding of hormones

generated by the body with ER. So far, more and more evidences of

malfunction of organisms are considered to be related to estrogenic

compounds although direct evidences and clear mechanism of estrogenic

effect still need to be revealed (OECD, 2002; US EPA, 2003).

Due to the toxic and adverse estrogenic effects, investigations on the fate of

PFOS and PFOA have been extensively carried out. For example, the pathway

and distribution in aquatic environment such as river, lake and seawater have

been researched. The main contamination source resulting in their occurrence

in the environment could be the sewage treatment plants (STPs), which

receive industrial and domestic wastewater discharges and usually consist of

conventional activated sludge treatment (CAS). Even though the precursors

could be degraded and produce PFOS and PFOA in the atmosphere, STPs are

identified as the major contamination source, through which PFOS and PFOA

enter into the aquatic environment. These compounds are discharged into the

environment with increased mass flow as they are resistant to CAS. For

example, STPs played an important role in the release of these compounds

into the local environment in some cities in U.S.A, Europe and Japan

(Boulanger et al., 2005; Hansen et al., 2006; Moody et al., 2005). Also, it was


4

observed that mass flow of PFOS and PFOA increased after treatment of CAS

(Schultz et al., 2006).

As the STPs can not effectively remove PFOS and PFOA, these compounds

enter into the environment and occur in the drinking water at trace

concentrations. For example, Harada et al. (2003) observed that the mean

levels ranged from 0.1 to 40.0 ng/L for PFOA and from <0.1 to 12.0 ng/L for

PFOS in treated drinking water in Japan. Although the adverse effect under

such concentration is not clear till now, it is certain that long-term exposure

will cause unexpected adverse effect since they are persistent and easily

accumulated in biological tissue. Therefore, research on the removal

technologies is important and urgent. Currently, various biological and

physico-chemical treatment processes including adsorption, biological

treatment, advanced oxidation and membrane separation have also been

studied to remove these compounds. However, these processes cannot remove

these pollutants both technologically and cost-effectively. The removal of

these compounds is still a challenge, especially for the full-scale wastewater

treatment. Thus, new advanced processes and removal mechanism have to be

developed and studied to remove these PFCs compounds effectively at low

cost for wastewater treatment.

1.2 Objective and Scope of Study

The primary research objective is to contribute towards establishment of

understanding of fate and behavior of PFOS and PFOA in environment and

full-scale activated sludge treatment system as well as development of proper


5

removal technology for wastewater treatment. Figure 1.1 shows the detailed

research scope and content. The specific objectives are listed as follows:

і) Characterize the spatial distribution and seasonal variation of PFOS and

PFOA in the aquatic and oceanic environment of Singapore.

іі) Develop a novel post extraction clean-up method for the determination

of PFOS and PFOA in environmental matrices, such as wastewater and

sludge.

ііі) Investigate fate and behavior of PFOS and PFOA in full-scale activated

sludge treatment system.

іv) Study the adsorption of PFOS and PFOA onto powdered activated

carbon (PAC) and activated sludge as well as removal of PFOS and PFOA

by hybrid PAC-MBR process.

PFCs Concentrations (liquid phase)

i) Occurrence and fate in water

Rivers/Canals

Reservoirs/Lakes

Effluents from STPs

ii) Development of clean-up method

Wastewater Sample

(liquid phase)

Sludge Sample (solid phase)

Coastal Water

Solid Phase Extraction (SPE)

Post Extraction Clean-up method

Reduce Matrix Effect

Solid Phase Extraction (SPE) LC/MS/MS

iv) Removal in hybrid PAC-MBR process

liquid phase

iii) Behavior in STPs

Seasonal Variation

Mass Change

Effect of SRT

Partition Coefficient

CAS MBR LTM

solid phaseKinetic Study

Adsorption Experiment

Equili-brium Study

PAC&Sludge

PAC-MBR

MBR

Effect of PAC dosage

Effect of SRT

Adsorption by PAC and sludge

Overall Performance

Mass Balance

Figure 1.1 Research scope and content.

1.3 Outline of Thesis


6

This thesis provides an overview of the spatial and seasonal distribution of

PFOS and PFOA in the waters of Singapore, develops a novel post-extraction

clean-up method for the determination of these two compounds in

environmental matrices, investigates the effect of SRT on the behavior of

these two compounds in the activated sludge process and explores removal

strategy of hybrid PAC-MBR process. The background information and

literature review, which shows the necessity and importance of the study, are

presented in Chapter 1. Chapter 2 reviews current available literature on PFCs,

including their basic properties, analytical method, occurrence in the

environment, fate and behavior in STPs and removal technologies. Chapter 3

describes the detailed materials and methods used in this study. Spatial and

seasonal distribution of PFOS and PFOA in different water matrices in

Singapore are presented in Chapter 4. Chapter 5 discusses the development of

post-extraction clean-up for wastewater and sludge sample as well as its effect

on eliminating matrix interference in complicated environmental samples.

Chapter 6 compares the behavior of PFOS and PFOA in full-scale

conventional activated sludge processes and membrane biological reactor, as

well as in an activated sludge process operated with a short SRT. Chapter 7

explores overall removal performance and factors affecting PAC adsorption

capacity in hybrid PAC-MBR process. Conclusion from this study and

recommendations for improvements and future study directions are presented

in Chapter 8.

Chapter 2-Literature Review

7

CHAPTER 2 LITERATURE REVIEW

2.1 Introduction

Perfluorinated compounds (PFCs) include perfluoroalkyl carboxylates (PFCAs)

and sulfonates (PFASs) with variable chain-lengths usually between about 6

and 15 carbon atoms. In addition, they contain precursors, which may break

down to PFASs or PFCAs of different chain lengths. The final breakdown

products are the sulfonates and carboxylates like PFOS and PFOA. In

perfluorinated organic compounds or perfluorochemicals all hydrogen atoms

of the corresponding hydrocarbon compound are substituted for fluorine atoms.

The polar carbon-fluorine bond is the most stable bond in organic chemistry.

Therefore, PFCs are thermally and chemically more stable than the analogue

hydrocarbons. One important group of PFCs is the group of perfluorinated

surfactants. They consist of a hydrophilic end group, i.e., sulfonate or

carboxylate end group, and a hydrophobic perfluorinated carbon chain (Table

2.1). Perfluorinated alkylsulfonates and carboxylates occur in numerous

consumer products as active ingredients, impurities or as degradation products

of derivatives, e.g. in oil, water and stain repellents for paper, leather and

textiles or in fire fighting foams. They may be emitted to the aquatic

environment during production and application and also after waste disposal.

Among all PFCs, the most important key compounds are PFOS and PFOA.

2.1.1 Physico-chemical properties of PFOS and PFOA


8

Structures of PFOS and PFOA are shown by Figure 2.1. The reported pKa

values of PFOA is 2-3 (Gilliland et al., 1992), indicating PFOA are present in

the environment. At pH 7, only 3-6 in 100,000 molecules are PFOA, with the

remaining being perfluorooctanoate (PFO). Physico-chemical properties of

PFOS and PFOA are summarized in Table 2.2. The pKa for PFOS has not

been measured but is expected to be negligible. A calculated pKa of -3.27 for

PFOS indicates that PFOS will be present in the environment completely in

the ionized form (OECD, 2002).

Table 2.1 Definition of acronyms and structures of PFCAs and PFSAs.


9

Figure 2.1 Structures of PFOS and PFOA.

The vapor pressure (VP) of 3.31x10-4 Pa has been measured for the potassium

salt of PFOS, using the spinning rotor method (OECD, 2002). Vapor pressures

of PFOA and perfluorononanoic, -decanoic, -undecanoic, and –dodecanoic

acids have been measured at the temperature range of 59.25-190.80 oC (Kaiser

et al., 2005). Extrapolation of the Antoine equation to 25 oC for PFOA results

in an estimated VP of 4.2 Pa (Kissa, 2001; US EPA, 2003). The solubility of

PFOS in water is reported to be 519 mg/L at 20±0.5 oC, and 680 mg/L at 24-

25 oC (3M, 2003). The sharp increase of solubility with temperature is

qualitatively consistent with the reported Krafft point of PFOS. The Krafft

temperature is the limit at which compounds cease to be singly dispersed and

begin to form micelles. Above the Krafft point, the solubility increases

abruptly on account of the formation of micelles. The solubility of PFOA in

water has not been published, although it is expected to be less soluble than

PFOS. The aqueous solubility of PFOA could be determined in a concentrated

acid solution. The octanol-water partition coefficient (Kow) is often used to

estimate other properties such as bioconcentration factors and sorption

coefficients. The surface active properties of PFCs make a direct

determination of the Kow impossible. For example, PFO/PFOA is expected to

form multiple layers in octanol/water making determination of Kow extremely


10

difficult (US EPA, 2003). In a preliminary study reported by 3M an

inseparable emulsion was formed. No measurements of the Henry’s law

constant (H) have been made for PFOS or PFOA. H is usually given by the

ratio of vapor pressure and water solubility. H for PFOS is expected to be very

low and H for PFOA is expected to be relatively high. 3M (2003) reported H

of 3.19x10-4 Pa·m3/mole for PFOS by calculated as the ratio of vapor pressure

and water solubility.

Table 2.2 Physico-chemical properties of PFOS and PFOA.

Property PFOS PFOA Mocular weight 500a 414

Vapor pressure (Pa) 3.31 x 10-4 1.3 x 104 Kow N.A N.A

Henry's law constant (Pa·m3/mole) 3.19 x 10-4 1.52 x 103 Water solubility (g/L) 0.519 3.4

pKa -3.27b 2.5 Note: a. potassium salt; b. calculated

2.1.2 Persistence, bioaccumulation and toxicity of PFOS and PFOA

2.1.2.1 Persistence

PFCs are stable to acids, bases, oxidants, and reductants and are generally not

believed to undergo metabolic or other degradation in the environment

(Schultz et al., 2003; Kiss 2001). Hatfield (2001) reported that aqueous

photolytic degradation of PFOA showed rather long half-life times in natural

environment. PFOS also showed its resistance to advanced oxidation

processes including ozone, ozone/UV, ozone/H2O2 and Fenton reagent due to

very strong and stable carbon-fluorine bond (Hori et al., 2006; Moriwaki et al.,

2005). Biological half-life of PFOA in plasma of a few days for mice and rats

and approximately 4.4 years for humans were reported (Kudo et al., 2003).


11

Half-life of PFOS varied from 7.5 days in rats through 200 days in

Cynomolgus monkeys to 8.7 years in humans, estimated from retired 3M

production workers (3M, 1999; OECD, 2002; Thibodeaux et al., 2003).

2.1.2.2 Bioaccumulation

Bioaccumulation factors (BAFs) represent accumulation potentials of organics

from environment to organisms. BAFs are calculated by dividing the average

concentrations in organism by the concentrations in water environment as

partition coefficient between octane and water phases for PFOS and PFOA are

not measurable (OECD, 2002; US EPA, 2002). Preliminary study showed

dietary BAFs of PFOS were 2796 in bluegill sunfish and 720 in carp (OECD,

2002). BAFs of PFOA were about 2 in fathead minnow and 3~8 in carp (US

EPA, 2002), which are much lower than PFOS.

2.1.2.3 Toxicity

PFCs are known to cause acute and subchronic toxicity effects in laboratory

studies. PFOA can cause peroxisome proliferation and affect mitochondrial,

microsomal, and cytosolic enzymes and proteins involved in lipid metabolism

(Kudo et al., 2003; Lau et al., 2003; Lau et al., 2004). Also PFOA reportedly

exerts other toxic effects, including accumulation of triglycerides in liver and

reduction of thyroid hormone in circulation (US EPA, 2003). PFOA produces

hepatomegaly, focal hepatocyte necrosis, hypolipidemia, alteration of hepatic

lipid metabolism, peroxisome proliferation, induction of the cytochrome P450

superfamily, and uncoupling of oxidative phosphorylation in laboratory-

exposed animals (Case et al., 2001). Exposure of rats and rabbits to PFOS and


12

n-EtFOSA results in reduced body weight gain, feed consumption, litter size,

and fetal weight at doses >5 mg/kg∙d. There is lot of information on toxicity

and toxico-kinetics of perfluorinated chemicals in the literature.

2.1.3 Preliminary regulations for PFOS and PFOA

PFOS and PFOA were recently nominated as candidates for POPs by the

Stockholm Convention in May 2009. Exposure criteria of PFCs for human

health were still in debating and there was no agreement yet. Minnesota

Department of Health recommended 0.3 μg/L for PFOS and 0.5 μg/L for

PFOA in drinking water as the safe level for human health in 2007 (MDH,

2007). However, North California Division of Water Quality proposed 2 μg/L

of PFOA to be interim maximum allowable concentration (NC DWQ, 2006).

Rather high screening levels of PFOA was established by West Virginia of

USA (WV DEP, 2002), which were 150 μg/L for water environment and 1360

μg/L for aquatic life. On January 15, 2009 U.S. Environmental Protection

Agency (US EPA) set a "provisional health advisory" of 0.4 ppb for PFOA

and 0.2 ppb for PFOS as safe level in drinking water (US EPA, 2009).

However, the advisory is not meant to protect the public from long term

exposure but might protect individuals for a couple of years.

The European Parliament approved a new EU directive (2006/122/EU) on

restrictions of marketing and use of PFOS and PFOS-related substances,

which came into effect on June 27, 2008. The provisions imply a prohibition

to use PFOS and substances that could degrade to PFOS in chemical products


13

and articles. Fire-fighting foams that have been placed on the market before 27

December 2006 can be used until 27 June 2011.

2.2 Analytical method for PFCs

2.2.1 Introduction of LC/MS/MS analysis for PFCs

More than three decades ago, Taves and co-workers first postulated that

perfluoroalkyl substances were widespread environmental contaminants

(Taves, 1968; Martin et al., 2004a). They used arduous, yet elegant, methods

to extract, clean up, and detect organic fluorine in human serum with nuclear

magnetic resonance (NMR) spectroscopy. These first studies revealed

compounds that resembled perfluorooctanoic acid (PFOA), but the inherent

ambiguity of the detection system prevented definitive identification. In

addition, the low concentration, lack of authentic standards, and unusual

physical and chemical properties of perfluoroalkyl chemicals made it difficult

to confirm their identity by traditional techniques, such as gas

chromatography/mass spectrometry (GC/MS).

Perfluorinated surfactants can be determined using derivatization techniques

coupled with gas chromatography followed by electron capture detection and

mass spectrometric detection (Jahnke, et al., 2006; Shoeib, et al., 2006). Since

PFOS has low volatility and its derivatives are unstable (Hekster, et al., 2002),

gas chromatography is not applicable for the determination of PFOS. It

implies liquid chromatography, which separates the analyte from other

molecules in the mixture based on differential partitioning between the mobile

and stationary phases, could be the suitable method to analyze PFCs. Ohya et


14

al. (1998) applied high-performance liquid chromatography (HPLC) and

fluorescence detection to measure perfluorocarboxylic acid concentrations in

biological samples.

2.2.2 LC/MS/MS analytical method for water and wastewater

High-performance liquid chromatography (HPLC) with triple quadrupole

mass spectrometry in electrospray negative mode is the most promising and


biological matrices, including water, wastewater, sludge and sediment samples

(Giesy et al., 2001 ; Kannan et al., 2002; Tomy et al., 2004; Martin, et al.,

2004a; Higgins et al., 2005; Hansen et al., 2001; Tseng et al., 2006). Up to

date, internal standard is generally used for quantitation of perfluorinated

compounds in water and wastewater since internal standard compensates

matrix suppression. Sixteen short- and long-chain perfluorinated compounds

were quantified by internal standards in water sample (Taniyasu et al., 2005).

Seven perfluorinated compounds were detected at ppt level in seawater by

internal standard quantitation using LC/MS/MS (Yamashita et al., 2004). Six

precursors and PFOS were detected in lake water by internal standard

quantitation (Boulanger et al., 2005). In municipal wastewater, quantitative

determination of perfluorinated compounds were successfully conducted by

two internal standards (Higgins et al., 2005; Tseng et al., 2006).

External standard quantitation was applicable to detect surface water

(Boulanger et al., 2004), but not suitable for wastewater because matrix

interference caused low recovery. It was observed the matrix interference on


15

PFOA and PFOS analysis caused low recovery (<35%) in influent of one STP

in Iowa (Boulanger et al., 2005).

2.2.3 LC/MS/MS analytical method for sludge and sediment

Quantitative determination of perfluorinated compounds in sludge and

sediment was achieved by three internal standards using HPLC with triple

quadrupole mass spectrometry in electrospray negative mode by internal

standard quantification (Higgins et al., 2005). Internal standard (surrogate

standard) was recommended and it compensated the loss due to matrix

interference. External standard was not available to quantify PFOS and PFOA

in sludge and sediment due to matrix suppression.

2.2.4 Limitation of Electrospray Ionization (ESI)

Electrospray ionization (ESI) is a method used to generate gaseous ionized

molecules from a liquid solution. This is done by creating a fine spray of

highly charged droplets in the presence of a strong electric field. The sample

solution is sprayed from a region of a strong electric field at the tip of a metal

nozzle maintained at approximately 4000 V. The highly charged droplets are

then electrostatically attracted to the mass spectrometer inlet. Either dry gas,

heat or both are applied to the droplets before they enter the vacuum of the

mass spectrometer, thus causing the solvent to evaporate from the surface. As

the droplet decreases in size, the electric field density on its surface increases.

The mutual repulsion between like charges on this surface becomes so great

that it exceeds the forces of surface tension, and ions begin to leave the droplet


16

through what is known as a “Taylor cone”. The ions are directed into an

orifice through electrostatic lenses leading to the mass analyzer.

ESI is especially useful in producing ions from macromolecules because it

overcomes the propensity of these molecules to fragment when ionized. It is

currently indispensable for identifying and quantifying perfluorinated acids;

however, this method has some inherent limitations such as low salt tolerance,

low tolerance for mixtures and difficulty in cleaning overly contaminated

instrument due to high sensitivity for certain compounds. In particular, co-

eluting matrix components can either suppress or enhance ionization, which

must be controlled to achieve maximum accuracy. For example, Benijts et al.

(2004) observed a decrease of 66% and an increase of 72% in MS/MS

response for 4-t-Octylphenol and estriol, respectively. In addition, several

studies have shown that matrix effects resulting from co-eluting residual

matrix components enhanced or suppressed electrospray ionization of

perfluorinated analytes, leading to considerable inaccuracy (Boulanger et al.,

2005; Higgins et al., 2005).

2.2.5 Matrix interference

Matrix interference resulting from co-eluting residual matrix components

affects the ionization efficiency of target analytes and can lead to erroneous

results. It was reported that recoveries of STP influent are only 34% (PFOS)

and 16% (PFOA), while effluent was 74% (PFOS) and 80% (PFOA)

(Boulanger et al., 2005). This low recovery of influent is due to matrix

suppression of analyte signals, which is confirmed by standard addition to the


17

final extracts of influent. It was also observed the matrix interference on

PFOA and PFOS analysis caused low recovery (<35%) in influent of one STP

in Iowa (Schultz et al., 2006a).

Matrix-matched standards are one possible control measure but become

impractical when an appropriate “clean” matrix cannot be found. Standard

addition quantitation, which involves spiking successive known quantities of a

standard into the sample and reanalyzing, is common in atomic absorption

spectroscopy and an acceptable technique to use when matrix effects are

unavoidable. Successive spiking has already been proven necessary for

perfluorinated acid quantitation by direct-injection MS analysis. Unfortunately,

standard addition quantitation can place further demands on instrument and

sample preparation time but should be used for accuracy when spike/recovery

experiments indicate a problem. Therefore, sample clean-up is desired to

eliminate matrix interference in complicated environmental and biological

samples (van Leeuwen et al., 2006; Szostek et al., 2004; Simcik et al., 2005;

van de Steene et al., 2006).

In order to rule out the matrix interference, internal standard (Isotopically

labeled chemical) is an effective tool. An important prerequisite, however, is

that analyte and internal standard have very similar characteristics, and

identical, or at least very close, retention times. Both compounds should be

affected by the co-eluted matrix to the same extent. In this respect isotopically

labeled internal standards offer the best solution. However some researchers

are still using external standard to quantify perfluoroalkyl substances by


18

external calibration since the use of stable isotopes is generally very costly,

and commercial availability is often limited. PFOS and PFOA were detected at

ng/L level in lake water by external standard quantification (Boulanger et al.,

2004). For the determination of PFCs in complex environment samples,

external standard quantification is not applicable due to matrix interference.

2.2.6 Post extraction clean-up method for analysis of environmental

matrices

Analysis of complex environmental matrices such as sediment, sludge and

wastewater by electrospray LC/MS/MS can be significantly hampered by

ionization effects induced by co-eluting components present in the sample

extracts. Several studies have shown that matrix effects resulting from co-

eluting residual matrix components enhance or suppress electrospray

ionization of perfluorinated analytes, leading to considerable inaccuracy

(Boulanger et al., 2005; Sinclair et al., 2006; Higgins et al., 2005). Therefore,

it is very important to eliminate matrix effects when the LC/MS/MS method is

used to quantitatively determine the concentration of perfluorinated

compounds.

Post-extraction clean-up is desired to eliminate matrix interference in

complicated environmental and biological samples (Martin et al., 2004a, van

Leeuwen et al., 2006, Szostek et al., 2004, Simcik et al., 2005). Powley et al.

applied Envi-carb (graphitized carbon) and glacial acetic acid to purify the

crude extracts of biological matrices (blood, serum, live and plant tissue).

Szostek et al. (2004) used silica column to clean up fish tissues by eluting the


19

lipids with dichloromethane, while the target compounds (PFCAs and PFSAs)

were eluted with acetone. For surface water samples, fluorous silica column

chromatography was used to clean up the SPE extracts and remove the

interfering compounds prior to LC/MS detection (Simcik et al., 2005).

Although the effect of these post-extraction clean-ups was assessed by the

improved recoveries for PFCs, matrix effect issue has not been sufficiently

studied and addressed. The assessment of matrix effect during development

and validation of LC/MS/MS method is necessary to ensure the precision,

selectivity, and sensitivity would not be compromised (Matuszewski et al.,

2003).

2.3 Occurrence of PFOS/PFOA in the environment

2.3.1 Occurrence in the surface water

PFOS and PFOA concentrations in surface waters are summarized in Table

2.3. Surface water in developed countries and industrialized areas were usually

highly polluted by PFCs, such as U.S.A (Hansen, et al., 2002; Takino, et al.,

2003), Japan (Saito, et al., 2003; 2004), Germany (Skutlarek, et al., 2006) and

coastal areas of China (So, et al., 2004). It was reported that concentrations of

PFOS and PFOA in the Great Lakes ranged from 21-70 and 27-50 ng/L,

respectively (Boulanger et al., 2004). Also, PFOS was detected in all of the

surface seawater samples collected from Tokyo Bay, at concentrations ranging

from 8 to 59 ng/L (mean of 26 ng/L) (Taniyasu et al, 2003). Several studies

reported on the occurrence of PFCAs and PFASs in surface waters in the USA,

Canada, Japan, Hong Kong, South China and Korea, both in freshwater and in

seawater. Elevated concentrations of PFOS (114±19 ng/L) and PFOA


20

(394±128 ng/L) were detected downstream of the receiving water of the 3M

fluorochemical manufacturing facility at Decatur, USA (3M, 1999). Upstream,

the concentration of PFOS was 32±11 ng/L and there were no measurable

PFOA levels (<25 ng/L) (Hansen et al., 2002). A comprehensive study on the

occurrence of PFOS and PFOA at 78 sampling sites in Japanese rivers and

creeks demonstrated the widespread occurrence of these compounds. In

different districts geometric means between 0.97 and 21.2 ng/L were evaluated

for PFOA and between 0.89 and 5.7 ng/L for PFOS. Individual concentrations

comprised a range from 0.10 to 456 ng/L for PFOA and from 0.24 to 37.3

ng/L for PFOS. Systematic surveys revealed two highly contaminated sites, a

public-water-disposal site for PFOA and an airport for PFOS (Taniyasu et al.,

2003). Measurements in German rivers, predominantly located in the Rhine

River catchment area, demonstrated that PFCAs and PFOS also occurred in

comparable levels to those found in USA, Canada and Japan (Skutlarek et al.,

2006).

Table 2.3 Review of PFOS and PFOA concentrations in surface water (ng/L), drinking water (ng/L), wastewater (ng/L) and sludge (ng/g). Environmental

Matrix Internal Standard

LOQ Concentration Recovery Location Reference

PFOS PFOA PFOS PFOA

Surface water

0.7 13 21-70 27-50 56-176% USA Boulanger, et al., 2004 PFDoA 17 9 n.d-995,000 n.d-11,300 68-93% Canada Moody, et al., 2002

10 25 27-144 25-598 83-112% USA Hansen, et al., 2002 0.8 8 1.8-16.1 n.d-21.6 USA Sinclair et al., 2004 0.1 n.a 0.3-157 n.a 75-105% Japan Saito et al., 2003 0.1 0.7-157 n.a n.a Japan Harada et al., 2003 0.1 0.1 0.2-67,000 0.6-526 92-106% Japan Saito et al., 2004

13C-PFOA n.a n.a 0.8-1,090 10-173 70-130% USA Sinclair et al., 2006 0.05 0.05 3.4-14.5 2.4-12 69-83% Germany Weremiuk et al., 2006 0.005 0.03 n.d-99 0.85-260 94-105% China So et al., 2007 2 2 n.d-5,900 n.d-33,900 98-100% Germany Skutlarek et al., 2006

13C-PFOA 0.1 0.1 n.d-44.6 n.d-297.5 95-106% China Jin et al., 2009 13C-PFOA 4 1.1 n.d-35 n.d-19 92-106% Australia Clara et al., 2009 13C-PFOA 1.2 1.8 29-82 3.6-10.9 90-101% Switzerland Huset et al., 2008 13C-PFOA, 13C-PFOS

7.2-8.4 2.0-2.8 Italy Loos et al., 2007

PFDoA 0.5 2 4-79 113-181 78-81% Taiwan Tseng et al., 2006 Drinking

water 13C-PFOA 0.1 0.1 n.d-14.8 n.d-45.9 95-106% China Jin et al., 2009

2 2 n.d-22 n.d-519 98-100% Germany Skutlarek et al., 2006 0.1 0.3-59 n.a n.a Japan Harada et al., 2003

A C

hapter 2-Literature Review

21

Environmental Matrix

Internal Standard

LOQ Concentration Recovery Location

Reference

PFOS PFOA PFOS PFOA

Seawater

0.0008 0.0052

0.338-57.7 1.8-192

80-160%

Tokyo Bay

Yamashita et al., 2005

0.04-0.075 0.137-1.06 Offshore (Japan)

0.070-2.6 0.673-5.45 Coastal area of Hong Kong

0.023-9.68 0.243-15.3 Coastal area of China

0.039-2.53 0.239-1.135 Coastal area of Korea n.d-0.109 0.088-0.51 Sulu Sea (surface) n.d-0.024 0.076-0.117 Sulu Sea (deep)

0.008-0.113 0.16-0.42 South China Sea 0.054-0.078 0.136-0.142 Western Pacific Ocean 0.0011-0.02 0.015-0.062 Pacific Ocean

8.6-36 0.015-0.036 North Atlantic Ocean 0.037-0.073 0.1-0.439 Mid Atlantic Ocean

0.005 0.02 0.02-12 0.24-16

84-190% Hongkong and South

China Ocean So et al., 2004 0.035-755 0.22-345 South Korea

0.1 n.a 0.2-25.2 n.a 75-105% Japan Saito et al., 2003 PFDoA 0.5 2 60 270 78-81% Taiwan Tseng et al., 2006

A C


22

Environmental Matrix Internal Standard

LOQ Concentration Recovery Location

Reference PFOS PFOA PFOS PFOA

Wastewater

- - 48-454 41-674 - USA (2 STPs) 3M, 1999 - - 41-5290 67-2420 - USA (4 STPs) 3M, 1999 26 22 72-92% USA (1 STP) Boulanger et al., 2005

13C-PFOA, PFEES 0.5 0.5 1.1-400 1.7-65 88-97% USA (10 STPs) Schultz et al., 2006 13C-PFOA, PFBS 2.5 2.5 3-68 58-1,050 70-130% USA (6 STPs) Sinclair et al., 2006

13C-PFOA, PFBS - - 1.8-149 1-334 90% USA (2 STPs) Loganathan et al., 2007

n.a 0.2 0.07 4.1 5.5 85.5-91.2% China (1 STP) Zhao et al., 2007

PFDoA 0.5 2 21-79 36-170 78-81% Taiwan (2P) Tseng et al., 2006 n.a 0.5 0.5 3.4-67 49.1-548.4 40-70% Japan (2 STPs) Nozoe et al., 2008

13C-PFOA - - 14-336 14-41 76-109% Japan (1 STP) Murakami et al., 2009 Sludge 13C-PFOA, PFEES 0.9 1 14.4-2,610 n.d-13.3 71-87% USA (12 STPs) Higgins et al., 2005

13C-PFOA, PFEES 0.7-2.2 0.7-2.2 3.8-160 n.d-12 >70% USA (3 STPs) Schultz et al., 2006

13C-PFOA, PFBS 10 10 n.d-65 18-241 n.a USA (2 STPs) Sinclair et al., 2006

13C-PFOA, PFBS 2.5 2.5 n.d-990 7-219 37-89% USA (2 STPs) Loganathan et al., 2007

23

A C



24

2.3.2 Occurrence in the drinking water

Some data are available on drinking water contamination with PFCAs and

PFASs from the USA and Japan (Harada et al., 2003; Skutlarek, et al., 2006).

The USA studies were conducted to obtain data about the presence of

fluorochemicals in drinking waters in the vicinity of fluoropolymer production

plants and where secondary manufacturers use these chemicals (3M, 2001). In

internal studies of tap water in 1984 in the vicinity of this works, DuPont

detected PFOA at concentrations of 1.5 μg/L in a store tap in Lubeck, at

concentrations of 1.0 and 1.2 μg/L in a store tap in Washington, and at

concentrations of 0.8 and 0.6 μg/L in Little Hocking (US EPA, 2002).

Recently, PFOA and PFOS contamination was reported in private

groundwater wells in Lake Elmo, Minnesota and in some of the Oakdale,

Minnesota municipal wells (Schultz et al., 2004). These contaminations

originated from several landfills, where PFCs were disposed by the 3M

Company decades ago. PFOA and PFOS were also detected in treated

drinking water and tap water in Columbus, Georgia, where several secondary

manufacturers were located, which produced non-wovens, household additives,

apparel, carpet, and home textiles. The PFOS concentrations ranged from 53

to 63 ng/L, and the PFOA concentrations ranged from 25 to 29 ng/L (3M,

2001). In a Japanese study, PFOA and PFOS had also been found in tap water

samples. The mean levels ranged from 0.12 ng/L to 40.0 ng/L for PFOA and

from <0.1 to 12.0 ng/L for PFOS (Harada et al., 2003). Occurrence of high

concentrations of PFCs in tap water indicated poor performances of current

water treatment processes to remove PFCs from surface water (Saito et al.,

2003; 2004; Skutlarek, et al., 2006).


25

2.3.3 Occurrence in the seawater

Concentrations of PFOS and PFOA in open ocean water samples from the

Pacific and Atlantic Oceans, and from several coastal seawaters from Asian

countries, were shown in Table 2.3. PFOS and PFOA were found in 80% of

the surface seawater samples analyzed (Yamashita et al., 2005). The

similarities between PFCs composition in coastal and open ocean waters were

found in some regions, which suggests that tidal and/or water current

movements play a major role in the transport of these compounds from coastal

locations; therefore, information on oceanic currents appeared necessary to

explain the transport of PFCs from coastal waters to the open ocean.

Relatively high concentrations of PFOS and PFOA were detected in Tokyo

Bay waters. PFOA was the predominant fluorochemical detected, which was

in the range of 1800 to 192,000 pg/L, followed by PFOS (338–57,700 pg/L)

(Taniyasu et al., 2003). Concentrations of PFOS and PFOA in offshore waters

of the Pacific Ocean were approximately three orders of magnitude lower than

those in Tokyo Bay. Concentrations of all of the target fluorochemicals in

offshore waters were in the pg/L range (Yamashita et al., 2004; Yamashita et

al., 2005). In the offshore waters of Japan, PFOA was also the predominant

fluorochemical investigated, which was similar to what was observed for

coastal waters. All target PFCs in open-ocean water samples collected in the

mid-Atlantic Ocean were at pg/L levels (Table 2.3). PFOA and PFOS

concentrations were comparable to those in offshore waters collected in the

South China Sea and the Sulu Sea. These concentrations were one order of

magnitude lower than those found in offshore waters, and four orders of


26

magnitude lower than the concentrations measured in Tokyo Bay waters (So et

al., 2004). It seems that these are the background values for remote marine

waters far from local sources.

2.3.4 Occurrence in the sludge and sediment

The binding of PFCs to sediment and sewage sludge is in general strong

and stable, which means a high potential for accumulation herein. In the so-

called multi-city study of the 3M Company, PFOS and PFOA

concentration ranges in sludge samples were: 58.9-2,980 ng/g dw (dry

weight) for PFOS and 0.297-173 ng/g dw for PFOA (3M, 2001).

Concentrations of PFOS and PFOA in sludge were in the range of 26-65

ng/g dw and 69-241 ng/g dw, respectively, in combined sludge of a STP in

New York State (Sinclair et al., 2006), 31-55 ng/g dw and <6-8.2 ng/g dw,

respectively, in activated sludge of a STP in U.S.A (Boulanger et al., 2005).

Higgins et al. (2005) reported PFOS and PFOA concentrations were in the

range of 14.4 – 2610.0 and n.d - 29.4 ng/g dw in the sludge samples of 8

STPs, respectively. Loganathan et al. (2007) observed that concentrations

of PFOS and PFOA in sludge were in the range of <2.5-77 ng/g dw and

7.0-130 ng/g dw in a STP of Kentucky, respectively.

In general sediment samples had lower levels of PFOS and PFOA than

those in sludge. Concentration of PFOS and PFOA in sediment was in the

range of n.d-3.07 ng/g dw and n.d-0.625 ng/g dw in 17 sites located in San

Francisco Bay area (Higgins et al., 2005), 0.09-0.14 ng/g dw and 0.84-1.1

ng/g dw in samples collected in Tojin river estuary, Japan (Taniyasu et al.,

2003).


27

2.4 Fate and behavior in the sewage treatment plants

2.4.1 Occurrence in the wastewater

Schultz et al. (2006a) reported that PFOS and PFOA were ubiquitous in the

influent and effluent of ten STPs in U.S.A. In the effluents of those ten STPs,

PFOS concentrations were in the range of 1.1-130 ng/L, while PFOA

concentrations varied from 2.5-97 ng/L. It was reported that concentrations of

PFOA in effluents of the six WWTPs ranged from 58 to 1,050 ng/L, while a

much lower PFOS concentrations (3-68 ng/L) were observed in effluents of

these WWTPs (Sinclair et al., 2006). Loganathan et al. (2007) observed that

PFCs concentrations ranged from 1.8 to 22 ng/L for PFOS and from 1.0 ng/L

to 227 ng/L for PFOA in a STP of U.S.A, while higher PFOS (7.0-149 ng/L)

and PFOA (22-334 ng/L) concentrations were detected in another STP of

U.S.A. In the so-called multi-city study, elevated PFCs concentrations were

found in publicly owned treatment works effluent in the range of 48-4,980

ng/L for PFOS and 42-2,280 ng/L for PFOA (3M, 2001). The concentrations

were highly variable and differed much between the treated effluent of the so-

called supply-chain cities and the control cities.

2.4.2 Mass flow and mass change

Few articles published are available on the behavior of PFCs in STPs due to

the difficulty in determination of their concentrations in sludge and wastewater

samples. Past studies are still not enough to draw general and reliable

conclusions on PFC behavior in STPs. In most studies, only the influent and

effluent were analyzed to estimate performance of overall process. A survey in

ten US STPs showed no obvious removal of PFOS in ten plants except one


28

STP with influent as high as 400 μg/L. PFOA in effluent of seven STPs was

increased by 10~100% of influent which contained 16~49 μg/L PFOS (Schultz

et al., 2006a). Surveys of STPs in Iowa of USA also showed no removal of

PFOS and PFOA, as well as other PFCs (Boulanger et al., 2005; Sinclair and

Kannan, 2006). Studies in a STP of Japan obtained similar results of poor or

even negative removal of PFOA and PFOS by activated sludge process

(Nozoe et al., 2006). These results implied that activated sludge process might

be ineffective to remove PFOS or PFOA, and certain amount of PFCs was

discharged from WWTPs to environment. PFCs precursors like telomer

alcohols, sulfonamides or esters were suspected to degrade to PFASs and

PFCAs during activated sludge process. Furthermore, Sinclair and Kannan

(2006) observed that mass flow of PFOS and PFOA in aqueous phase

increased significantly after secondary treatment in a sewage treatment plant

with industrial influence, while no increase in mass flow of PFOA was found

in another sewage treatment plant with no industrial influence.

Up to now, the first and only one study which estimated performances of

individual facilities in activated sludge process revealed the interesting vision

of PFC behavior inside STP (Schultz, et al., 2006b), as shown in Figure 2.2.

Mass flows of PFOA were nearly unchanged as a result of wastewater

treatment, which indicates that conventional wastewater treatment is not

effective for removal of this compound. A net increase in the mass flows for

PFOS occurred from trickling filtration and activated sludge treatment was

observed since the more highly substituted perfluorooctyl surfactants had been

biodegraded. The precursor compounds formed an additional source of PFOA


29

and PFOS in the STP effluents. It seems that mass flow of PFOS or PFOA

either increased or remained consistent, indicating conventional activated

sludge process can not effectively remove these compounds.

2.5 Removal Technologies

PFOS and PFOA are not only metabolically but also photochemically inert,

resisting both biotic and abiotic degradation. PFOS and PFOA were

considered stable and persist in environment without natural degradations

(OECD, 2002; US EPA, 2002). Also, Schröder et al. (2003) observed that

PFOS was not degradable by activated sludge. Furthermore, past study on fate

and behavior of these pollutants in STPs implied that they can not be

effectively removed by biological treatment process. Thus, various physico-

chemical processes have been studied to remove PFOS and PFOA.

2.5.1 Advanced oxidation process

Figure 2.2 Average mass flow (mg/d) for PFOS and PFOA in sewage treatment

plant. PC-primary clarifier,TF -trickling filter, AS-activated sludge, SC-secondary clarifier, FC-final chlorination/dechlorination, TH-thickener, AD-anaerobic digester,

RAS-recycled activated sludge.


30

Advanced oxidation processes (AOPs) involve the generation of hydroxyl

radicals in sufficient quantity to effect water purification. These common

processes include O3/H2O2, O3/UV, UV/ H2O2. UV/TiO2 process and Fenton

reagent are also effective to specific wastewater (Gottschalk, et al., 2000).

AOPs such as O3, O3/UV, O3/H2O2 and Fenton reagents were unable to

decompose PFOS in normal state, but able to degrade PFOS precursors and

partially fluorinated polymers effectively (Schröder and Meesters, 2005).

Some oxidation processes can decompose some PFCs completely in critical

conditions or coupled with catalysts. It was observed that PFOS was

completely oxidized by subcritical water oxidation, with catalyst of zerovalent

metals like iron. PFOS molecules were observed to be strongly adsorbed on

Fe3O4 sediments and further decomposed to carbon dioxide and fluorine ions

due to oxidation by molecular oxygen in subcritic water (Hori et al., 2006).

The author, however, did not consider that PFOA was able to be decomposed.

It could be due to that catalyst iron could not excite the oxidation of PFOA

and no cleavage of C-F bond occurred.

Other oxidation process has been studied to decompose environmental

contaminants, including PFCs. One of the most effective oxidation processes,

sonochemical reaction was applied to degrade PFOS and PFOA. Under

ultrasonic irradiation (20, 3 W/cm2), PFOS molecules were firstly transformed

to PFOA by releasing the sulfonate group, and the product of PFOA was

consequently degraded to short-chained PFCAs (Moriwaki, et al., 2005).

Unfortunately only partial PFOS (28%) and PFOA (63%) were decomposed


31

although high energy was consumed. It seems that partial decomposition could

be resulted from unevenly and unfully irradiation in the reaction. Higher

energy and longer exposure irradiation may be needed to completely

decompose compounds.

2.5.2 RO/NF membrane

Membrane technology is one of the most promising technologies in water

reclamation and reuse. Reverse osmosis (RO) and nanofiltration (NF) are used

extensively in water and wastewater treatment. Both NF and RO are pressure

driven membrane processes, where an applied transmembrane pressure forces

water through the ‘pores’ and contaminants are retained due to charge and size

interactions. NF distinguishes itself from RO in that it only retains multivalent

ions, which makes it a very economic alternative where the retention of

monovalent salts is not required. RO and NF membranes are effective in

removing most organic (Kiso et al., 2001; Kimura et al., 2003; Schafer et al.,

2003) and inorganic compounds from water solutions. The main motivation to

use those processes in water and wastewater treatment is the removal of

micropollutants such as PPCPs. Because of the difficulty in effectively

removing trace organic compounds with low molecular weight from

wastewaters by conventional treatment process, membrane technology has

been investigated and applied to improve their removal. Nearly complete

retention of those micropollutants by RO and NF has been reported by many

researchers (Kimura et al., 2003; Schafer et al., 2003; Childress et al., 1996;

Kiso et al., 2001; Hu et al., 2007). Both size exclusion and adsorptive effects

appears to be essential in maintaining high retention of those micropollutants


32

on a variety of NF and RO membranes over a range of solution conditions.

Tang et al. (2006) reported that RO membrane rejected 99% or more of the

PFOS with feed concentration ranging from 0.5 to 1500 ppm. Although the

author did not mention PFOA rejection by RO, it could be predicted that

PFOA would be significantly removed based on size exclusion mechanism.

Consequently, membrane separation appears to be an effective technology for

removal of PFCs from wastewater. However, RO or NF filtration is rarely

used in wastewater treatment because of high cost.

2.5.3 Adsorption

2.5.3.1 Activated carbon adsorption

Compared with AOPs, adsorption is a more common and widely used method

for removing organic contaminants from wastewater steams. In water and

wastewater treatment the most often used adsorbent is granular activated

carbon (GAC) and powdered activated carbon (PAC). Activated carbon

adsorption is one of the most promising methods to remove PFCs in aqueous

stream due to the effectiveness and low cost. Ochoa-Herrera et al. (2008)

reported that PFOS could be effectively removed by granular activated carbon

(GAC) and Freundlich isotherm was applicable at high and low equilibrium

concentrations. In contrast, Yu et al. (2009) studied the feasibility of using

powder activated carbon (PAC), granular activated carbon (GAC) and anion-

exchange resin (AI400) to remove PFOS and PFOA from water. It was

observed that adsorption isotherms of PFOS and PFOA fitted Langmuir

isotherms better than Freundlich isotherm. Qiu et al. (2006) also reported that

GAC was able to effectively remove PFOS and PFOA. In 4 hours, 93% PFOS


33

and 99% PFOA in pure water at ppb level were adsorbed onto GAC.

Compared with AOPs, GAC is able to remove both PFOS and PFOA in

normal state effectively. Although PAC has similar adsorption capacity to

GAC, it has not yet found broader application in wastewater treatment as it is

not easily removed from the treated effluent.

2.5.3.2 Adsorption onto sediment and sludge

Sorption of the potassium salt of PFOS to three types of soil, sediment and

sludge from a domestic wastewater treatment plant has been measured using a

method based on OECD 106 (3M, 2003). Adsorption occurred rapidly in all

cases, and the concentrations remained fairly constant after 16 hours.

Desorption was also investigated which took place rapidly, and after 8 hours

the concentration in water did not vary significantly. Values for the sorption

and desorption coefficients were calculated and presented in Table 2.4.

Table 2.4 Sorption and desorption coefficients of PFOS from various matrices.

Matrix type Kd (L/kg) Kdes (L/kg) Mean Clay soil 18.3 47.1 32.7

Clay loam soil 9.72 15.8 12.8 Sandy loam soil 35.3 34.9 35.1 River sediment 7.42 10 8.7

STP sludge - - 1028 Note: mean values are mean of sorption and desorption coefficients. For sludge, value is the mean of the Freundlich coefficients for sorption and desorption, as direct values are only reported as limit values.

The occurrence of PFCs in sludge from STPs indicates an adsorption of these

compounds to the activated sludge during the treatment process. It was

reported that the measured log Koc value for PFOS and PFOA are 2.57 and

2.06, which all are in the range of 2.57-3.1 [log(L/kg)] for PFOS (3M, 2002)


34

and 1.9-2.17 [log(L/kg)] for PFOA (Dupont, 2003). In activated sludge

treatment process, 100-400 gSS/m3 of sludge is usually produced. Therefore

removal by sorption onto sludge is generally relevant (>10%) only for

compounds with a Kd>300 L/kg. According to the reported data, Kd ranged

from 371 to 1,258 L/kg for PFOS and 79 to 148 L/kg for PFOA, indicating

<35% PFOS and <6% PFOA were adsorbed onto sludge. Therefore, it can be

expected that PFOS and PFOA did not adsorb significantly onto sludge and

sorption was not an important removal process in conventional wastewater

treatment system, which were proven by Figure 2.2.

2.5.4 Membrane biological reactor (MBR)

2.5.4.1 Introduction

Since research on membrane bioreactor (MBR) technology began over 30

years ago, several generations of MBR systems have evolved (Gander et al.,

2000). Up to this date, MBR systems have mostly been used to treat industrial

wastewater, domestic wastewater and specific municipal wastewater, where a

small footprint, water reuse, or stringent discharge standards were required. It

is expected, however, that MBR systems will increase in capacity and broaden

in application area due to future, more stringent regulations and water reuse

initiatives.

In the early 1990s, MBR installations were mostly constructed in external

configuration, in which case the membrane modules are outside the bioreactor

and biomass is re-circulated through a filtration loop. This limited wider

application in treatment of municipal wastewater in North America because of


35

high power consumption. After the mid 1990s, with the development of

submerged MBR system, MBR applications in municipal wastewater extended

widely. In the past 10 years, MBR technology has been of increased interest

both for municipal and industrial wastewater treatment in North America.

2.5.4.2 Configuration and application

MBR systems are characterized by two configurations: submerged (immersed

or integrated) MBRs and external (recirculated or side-stream) MBRs. Due to

the absence of a high-flow recirculation pump, submerged MBRs consume

much lower power than external MBRs. This was the primary driver for

propelling submerged MBRs into the purview of large-scale wastewater

treatment plants in a few dozens of countries around the world. External

MBRs were considered to be more suitable for wastewater streams

characterized by high temperature, high organic strength, extreme pH, high

toxicity and low filterability. In the case of an external MBR system, the

membrane device is independent of the bioreactor. Feedwater enters the

Figure 2.3 Configuration of MBR systems: (a) Side-stream MBR, (b) Suctioned- filtration submerged MBR, and (c) Gravitational-filtration

submerged MBR.


36

bioreactor where organic matters are biodegraded by biomass. The mixed

liquor in the bioreactor is then pumped around a recirculation loop containing

a membrane unit where permeate is discharged and the retentate is returned

back to the bioreactor. The transmembrane pressure (TMP) and crossflow

velocity of the membrane device are both generated from a pump (Hillis, 2000;

Kim et al., 2001).

2.5.4.3 Technology benefits and problems

The technical benefits of MBR include high quality effluent, small footprint,

short start-up time and low operating and maintenance manpower requirement.

Of these, the prime ones are the excellent effluent quality, easy management,

high biomass concentration, and less sludge production (Xing et al., 2000;

Fleischer et al., 2005). MBR systems can provide high-quality effluents which

are free of solids and bacteria and can be directly reused for municipal

watering, toilet flushing, and car washing (Huang et al., 2001; Xing et al.,

2001). Since suspended solids are completely retained by membranes in MBR

systems, quality of effluent would no more be affected by the settling problem

caused by poor flocculation of microorganisms or proliferation of filamentous

bacteria (Bai and Leow, 2002). Consequently, it is much easier to operate and

maintain MBR systems as compared to conventional activated sludge systems.

The elimination of secondary settlement stage allows the use of high activated

sludge concentration in a small volume tank. For example, some authors have

investigated MBR system with MLSS ranging between 10,000 and 23,000

mg/L (Dijk and Roncken, 1997; Churchouse et al., 1998). Bouhabila et al.

(1998) studied critical fluxes for the operation of the MBR with MLSS


37

concentration of up to 15,000 mg/L. High biomass concentration in the reactor

enabled MBR to produce high quality effluent at short hydraulic retention time

(Gunder, 2001). Furthermore, MBR systems can be operated at low organic

loading rates with the combination of high biomass concentrations and the

complete retention of biosolids. These characteristics promote the

development of slow growth bacteria, such as nitrifiers, and result in lower

sludge production as compared with conventional aerobic treatment processes

(Chang et al., 2002).

Despite the many advantages of MBR systems, it has been shown that

membrane fouling is the most serious problem affecting system performance

(Visvanathan et al., 2000; Le-Clech et al., 2003; Kim et al., 2001). It is

reported that the nature and extent of fouling are strongly influenced by three

factors: characteristics of mixed liquor, operating conditions, and membrane

properties (Chang and Lee, 1998; Shimizu et al., 1996; Bouhabila et al., 2001;

Ng et al., 2005). It has been shown that membrane fouling is the most serious

problem affecting system performance in some recent reviews covering

membrane applications to bioreactors (Visvanathan et al., 2000; Kim et al.,

2001). Though numerous investigations of membrane fouling have been

published, the diverse range of operating conditions and feedwater matrices

employed, and the limited information reported in most studies on the mixed

liquor composition, have made it difficult to establish any generic behavior

with respect to membrane fouling in MBR systems (Chang et al., 2002).

However, it is evident that the nature and extent of fouling are strongly

influenced by characteristics of mixed liquor, operating conditions, and


38

membrane properties (Chang and Lee, 1998; Chang et al., 1999; Bouhabila et

al., 2001).

2.5.4.4 Hybrid PAC-MBR system

Membrane fouling in MBR results from the interaction between membrane

material and components in the activated sludge mixture. The latter includes

substrate components, cells, cell debris, and microbial metabolites such as

extracellular polymeric substances (EPS). Accordingly, the floc structure,

particle size distribution and EPS contents of activated sludge can all

contribute to membrane fouling. To prevent or mitigate membrane fouling in

MBR, various techniques have been adopted such as low-flux operation, high

shear slug flow aeration in a submerged configuration, periodical air or

permeate backflushing and intermittent suction operation. In recent years, the

addition of PAC to a MBR (referred to as hybrid PAC-MBR in this study) has

been applied for wastewater treatment. A few studies of hybrid PAC-MBR

process have been reported and results showed that the addition of PAC

improved the performance of MBR system (Munz et al., 2007; Ng et al., 2006;

Satyawali et al., 2009). On the one hand, some studies observed that

membrane flux was enhanced since PAC decreased the compressibility of

sludge flocs and increased the porosity of cake layer by acting as supporting

medium (Kim et al., 1998; Aquino et al., 2006). Li et al. (2005) further

identified that PAC addition significantly decreased membrane total resistance

by 44% for long term operation of submersible membrane bioreactor, which

resulted in extension of operation time by 1.8 times as compared to normal

MBR system. On the other hand, a few studies found that adding PAC into


39

MBR could not only increase porosity of cake layer but also reduce the

accumulation of foulants on the membrane surface and change the

composition and permeability of the cake layer (Kim et al., 1998; Ng et al.,

2006). Ng et al. (2006) even pointed out that the primary role of the PAC was

to provide adsorptive removal of foulants rather than providing supporting

medium. Other benefits of PAC addition include increase in the removal of

organics, reduction in the impact of organic shocking loadings and increase in

the resistance to toxic substances (Aktas et al., 2007; Lesage et al., 2007).

Therefore, it is evident that hybrid PAC-MBR system shows better

performance than normal MBR system in terms of effluent quality, stability

and fouling rate due to PAC effects on the foulants, sludge flocs and

membrane filtration.

Hybrid PAC-MBR could be an effective technique to remove micropollutants

in wastewater since the bioreactor combines three individual process

operations, namely physical adsorption, biological degradation and membrane

filtration in a single unit. A few studies have been conducted to investigate the

removal mechanism of micropollutants in hybrid PAC-MBR. Dosoretz et al.

(2004) reported that an almost complete removal of phenanthrene was

observed in hybrid PAC-MBR due to the simultaneous adsorption and

biodegradation. Baumgarten et al. (2007) also found that combination of MBR

and PAC could effectively remove some micropollutants, such as antibiotics.

However, little information is available on the removal of PFCs in hybrid

PAC-MBR till now.


40

2.6 Research statement

PFOS and PFOA, regarded as the terminal breakdown end-products of PFCs,

have been detected in the air, surface waters, wastewaters, drinking waters,

groundwaters, coastal waters, sediments as well as various biological tissues

all over the world. Their ubiquious presence in the environment could be due

to the worldwide use of PFCs and the high mobility of their precursors in the

air. A few studies have been conducted to identify the contamination source of

PFCs in the environment. Some researchers observed that effluents form the

STPs are the most important PFCs sources for the aquatic ecosystems (Sinclair

and Kannan, 2006; Loganathan et al., 2007). Zushi et al. (2008), however,

reported that loads of PFCs in rain runoff were about 2-11 folds greater than

those in STP effluents that were discharged into a river. It indicates that

nonpoint source of PFCs could be the most important source for the river

studied. In addition, Yamashita et al. (2004) reported that application of PFC-

containing products could also be an important source of aquatic environment.

It seems that effluents from STPs, nonpoint source from rain runoff and

application of PFC-containing products might be important sources and

determine the PFCs concentration levels in the aquatic environment. However,

these studies failed to prove that there are no other significant PFCs sources

such as atmospheric deposition or precipitation for the aquatic environment.

Kallenborn et al. (2004) and Scott et al. (2006) both reported relatively high

PFOA concentrations in the rainwater samples from Europe and North

America, which could be important PFCs sources. Therefore, further research

is needed to identify possible contamination sources and transportion

pathways of PFCs in environment. Furthermore, seasonal variations in the


41

PFCs concentraions were investigated. So et al. (2004) observed PFCs

concentrations in the winter were higher than in the summer in coastal waters

of China. In wastewater of STPs, Loganathan et al. (2007) found that mass

flow of PFCs were higher in winter than in summer. The authors suggest that

there were less rain in winter than in summer, which resulted in dilution effect

on the coastal waters or wastewaters. However, limited data is available on the

comparison of PFCs concentrations between dry season and wet season in the

aquatic environment. Singapore is an island country and also a true city-state

with a tropical rainforest climate and no distinctive seasons. Especially its

climate is characterized by uniform temperature, pressure and abundant

rainfall in wet monsoon season (November and December). In a such an

unique island city, it could be an ideal place to identify seasonal variations of

PFCs concentrations between dry seasons and wet seasons by excluding other

factors, such as temperature and atmospheric pressure variation. To the best of

our knowledge, this study is the first study to identify the seasonal varitions of

PFCs in aquatic environment between dry and wet seasons.

Solid phase extraction (SPE) followed by high performance liquid

chromatography coupled with tandem mass spectrometry (HPLC/MS/MS) are

widely applied to quantitatively identify PFOS and PFOA. However, analysis

of complex environmental matrices such as sediment, sludge and wastewater

by electrospray LC/MS/MS can be significantly hampered by ionization

effects induced by co-eluting components present in the sample extracts.

Several studies have shown that matrix effects resulting from co-eluting

residual matrix components enhance or suppress electrospray ionization of


42

perfluorinated analytes, leading to considerable inaccuracy (Boulanger et al.,

2005; Sinclair et al., 2006; Higgins et al., 2005). Therefore, post-extraction

clean-up is desired to eliminate matrix interference in complicated

environmental and biological samples (van Leeuwen et al., 2006; Szostek et a.,

2004; Simcik et al., 2005; van de Steene et al., 2006). A few studies applied

different methods to remove the interfering compounds prior to LC/MS

detection. For examples, Powley et al. (2005) applied Envi-carb (graphitized

carbon) and glacial acetic acid to purify the crude extracts of biological

matrices (blood, serum, live and plant tissue). Szostek et al. (2004) used silica

column to clean up fish tissues by eluting the lipids with dichloromethane. For

surface water samples, fluorous silica column chromatography was used to

clean up the SPE extracts and remove the interfering compounds prior to

LC/MS detection (Simcik et al., 2005). However, the above post-extraction

methods may not be applied to wastewater and sludge samples collected form

STPs, in which stronger matrix effect was observed in comparison with

surface water (Boulanger et al., 2005; Sinclair et al., 2006). Thus, it is

necessary to develop a novel post-extraction clean-up for different

environmental matrices, including wastewater and sludge samples. In addition,

limited data is available on the quantitive estimation of matrix effect and effect

of post-extraction clean-up on different environmental matrices. In order to

ensure the precision, selectivity, and sensitivity of extraction method, there is

also a need to quantitively investigate matrix effect during development and

validation of LC/MS/MS method.


43

Due to widespread usage of PFCs in industrial and commercial applications，

various contamination levels were reported in the influent and effluent of

municipal STPs in Iowa City (Boulanger et al., 2005), in Kentucky and

Georgia (Loganathan et al., 2007), in 10 national wide municipal STPs in

U.S.A (Schultz et al., 2006a) and in the effluent of 6 U.S.A cities (Sinclair and

Kannan, 2006). It is evident that the discharge of municipal wastewater

effluent is one of the major routes for introducing PFOS and PFOA that are

used in domestic, commercial and industrial settings into aquatic environment.

A few researchers studied the fate and behavior of PFCs in STPs. Sinclair and

Kannan (2006) observed that mass flow of PFOS and PFOA in aqueous phase

increased significantly after secondary treatment in a STP with industrial

influence, while no increase in mass flow of PFOA was found in another STP

with no industrial influence. Furthermore, Schultz et al. (2006b) identified the

fate and behavior of these two compounds in both aqueous phase and solid

phase (sludge) during each step of municipal wastewater treatment plant. It

was observed that mass flow of PFOS or PFOA either increased or remained

consistent, indicating conventional activated sludge process can not effectively

remove these compounds. Unfortunately, these investigations were conducted

at different STPs with different influents. Different influent of STP would

significantly affect the behavior pattern of PFOS or PFOA since their

precursors in the influent could biodegraded to PFOS or PFOA in the activated

treatment processes. Therefore, it is desired to investigate behavior of PFCs in

various activated sludge treatment processes which receive the same raw

sewage. In addition, sludge retention time (SRT) could also be an important

factor affecting the fate of PFOS and PFOA in STP. Clara et al. (2005) found


44

that the degradation of the micropollutants, such as endocrine disrupting

compounds and pharmaceuticals, was dependent on the SRT in the activated

sludge process since the SRT determines the enrichment of the microorganism

that is able to degrade the micropollutants. Therefore, behavior pattern of

PFCs may be different in the conventional activated sludge process operated

with different SRT. However, no data is available about the effect of SRT on

the behavior pattern of PFOS and PFOA in the activated sludge process. It is

desired to study the fate and behavior of PFOS and PFOA in full-scale STP

comprising of different activated sludge treatment processes with different

SRT, which treat the same raw sewage.

Although there is no maximum allowable concentration of PFCs in the

discharge of STPs, PFOS and PFOA, candidates for persistent organic

pollutants (POPs), are reported to have adverse effect on the human health.

Since PFOS and PFOA can not be effectively removed by conventional STPs

and drinking water treatment plants, it is urgent to develop a new technology

to remove these compounds effectively at low cost for the wastewater

treatment. The hybrid PAC-MBR technology integrates adsorption and

biodegradation of organic matters with membrane filtration in one unit, which

has been proved to be a simple and highly efficient way to remove compounds

in wastewater. In particular, PAC addition increases the removal of organic

matters with low molecular weight by adsorption; it also serves as a

supporting medium for attached bacterial growth (Kim et al., 1998). Even

though MBR may not be able to significantly remove PFOS and PFOA due to

similar biodegradation and adsorption behavior in activated sludge system,


45

combination of MBR and PAC technologies could effectively remove these

compounds while adsorption onto PAC occurs. It was reported that PFCs were

effectively removed by adsorption onto the activated carbon at high and low

equilibrium concentrations (Ochoa-Herrera et al., 2008; Qiu et al., 2006).

However, these studies were conducted in the buffer solution without the

presence of dissolved organic matters (DOMs). In STPs, effluent from

biological wastewater treatment contains complex and heterogeneous soluble

organic matters, which are so called effluent organic matters (EfOM). The

composition of EfOM is a combination of those of natural organic matter

(NOM), soluble microbial products (SMPs), and trace harmful chemicals. It

was observed that PAC adsorption capacity would be reduced dramatically

when EfOM was present during activated carbon treatment of wastewater

containing micropollutants (Newcombe et al., 2002; Matsui et al., 2003). The

direct competition for the adsorption sites was found to be the most likely

competition between EfOM and target micropollutants (Newcomber et al.

2002; Kilduff et al. 1998; Matsui et al., 2003). However, limited data is

available on the effect of EfOM on the PFCs adsorption to the activated

carbon. Therefore, study on the EfOMs effect on PFCs adsorption is needed

for the better understanding of competitive effects caused by the presence of

EfOM. In addition, it is essential to study the adsorption capacity and kinetics

of PFCs onto PAC to understand the removal mechanism in hybrid PAC-MBR

process.

It is evident that operation parameters can affect PFCs removal in the hybrid

PAC-MBR system. On the one hand, SRT, a commonly used parameter for


46

biological process design and operation, could be an important factor affecting

the removal of PFOS and PFOA. It was reported that SMPs was the dominant

DOMs in the supernatant and effluent of MBR (Lee et al., 2003; Barker et al.

1999). At different SRT, composition of SMPs could be different, which may

affect the PAC adsorption. For example, Liang et al. (2007) observed that

SMPs in MBR was significantly reduced as SRT was increased, indicating

reduced adsorption competition from DOMs. Thus, PAC adsorption capacity

may be significantly affected by characteristics of SMPs, which can be

influenced by SRT of MBR. However, no study is available on the adsorption

capacity of PAC in MBR operated at different SRT. Therefore, it is necessary

to study the effect of SRT on PFCs adsorption on the PAC. On the other hand,

it is generally accepted that PAC addition in the MBR can enhance membrane

flux and decrease fouling rate. Most studies focused on the effect of PAC on

membrane filtration and fouling. However, little data is available on the effect

of PAC dosage on micropollutant removal. It is necessary to explore the

optimum PAC dosage in order to achieve the desired PFCs’ removal in hybrid

PAC-MBR process.

In summary, this study aims to identify possible contamination sources and

transportion pathways of PFCs and seasonal variation of PFOS and PFOA in

the aquatic and oceanic environment; to develop a novel post extraction clean-

up method for the determination of PFOS and PFOA in environmental

matrices; to study the fate and behavior of PFOS and PFOA in full-scale STP

comprising of different activated sludge treatment processes with different

SRT; to study the EfOMs effect on PFCs adsorption for the better


47

understanding of competitive effects at the presence of EfOM; and to study the

mechanism of PFCs removal by hybrid PAC-MBR process running with

different SRT and PAC dosage.

Chapter 3-Materials and Methods

48

CHAPTER 3 MATERIALS AND METHODS

3.1 Chemicals, materials and reagents

3.1.1 Chemicals and reagents

Standards of perfluorooctane sulfonate potassium salt (PFOS, ≥98%),

perfluorooctanoate acid (PFOA, 96%), methanol (99.8%) and ammonium

acetate (97%) were purchased from Sigma-Adrich (Singapore). Internal

standard perfluoro (2-ethoxyethane) sulfonic acid (PFEES, 97%) and

perfluorododecanoic acid (PFDoA, 95%) was purchased from Oakwood

Research Chemicals (West Columbia, USA) and Sigma-Adrich (Singapore),

respectively. Oasis HLB (500mg, 6 cc) and Sep-Pak plus silica (1g) solid

phase extraction (SPE) cartridges were from Waters (Milford, USA). Nylon

syringe filter (0.2 μm) was from Millipore (USA).Stock solutions were

prepared in methanol at a concentration of 1 mg/mL. From these stock

solutions working solutions were prepared by diluting with 70:30 (v/v)

methanol/aqueous ammonium hydroxide (0.01%) solution. Stock solutions

and working solutions were stored at -20 o

C.

3.1.2 Materials

YM (Millipore, USA) series UF membranes with nominal molecular weight

cutoffs of 1, 10 and 30 kDa were used in this study. High density polyethylene

(HDPE) bottles were used for all adsorption experiments. Supelite™ XAD-8

resin (SUPELCO, U.S.A.), AG MP-50 cation exchange resin (Bio Rad,


49

U.S.A.) and Amberlite IRA-96 anion exchange resin (Rohm and Haas, France)

were used for DOM fractionation.

PAC obtained from Sigma-Adrich (Singapore) was used as adsorbents for this

study. The characteristics of PAC used are given in Table 3.1. Prior to

adsorption experiments, PAC was soaked in Mill-Q water for 24 h to release

potentially adsorbed organics, dried in the oven at 103 oC for 48 h and then

stored in a desiccator. PAC and activated sludge were used as adsorbents for

the adsorption experiment. The activated sludge was collected from the

laboratory-scale MBR immediately before adsorption experiment and sodium

azide (100 mg/L), a respiratory inhibitor, was added to prevent microbial

metabolism. Table 3.2 shows the characteristics of the MF hollow fiber

membrane.

Table 3.1 Characteristics of powdered activated carbon (PAC).

Specification Description/value

Type Wood based

Ash content ≤ 5%

Particle size (nominal) ≤40 μm (75%)

BET Surface area 763.8 m2/g

Mean pore diameter 12.724 Å

Pore volume 0.486 cm3/g


50

Table 3.2 characteristics of the MF hollow fiber membrane.

Properties of Membrane Unit Description/value

Material - PVDF

Type - Hydrophobic

Total surface area m2 0.2

Permeability at 20oC L·m-2·h-1·bar-1 50

Mean pore diameter µm 0.2

MWCO kDa 300 kDa

Internal diameter mm 0.46

External diameter mm 0.95

Flux L/m2 ·h 10

3.2 Water sample collection and preparation

3.2.1 Water sample collection

Water samples from reservoirs, rivers, canals and effluents of STPs and

coastal waters around the island were collected in October 2006, December

2006, March 2007, September 2007 and December 2007. Due to the tropical

climate in Singapore, November and December are the wet monsoon season,

while other months are relatively dry seasons in Singapore (Ooi and Chia,

1974). Monthly rainfall totals are 86, 788, 169, 197 and 413 mm for October

2006, December 2006, March 2007, September 2007 and December 2007,

respectively (National Environment Agency, 2008). For the sampling in

December, it was performed during the second week while it was raining.

During sampling, outside temperature ranged from 21 to 32 oC.


51

Figure 3.1 Sampling locations for reservoir waters, river/canal waters, effluents of WWTPs, coastal waters and location for outfalls of WWTPs.

Western area of Singapore is a heavily industrial influenced area, while

eastern and middle areas are predominantly commercial and light industrial

areas. Some samples were collected from urban regions, while others were

from industrial influenced districts (sampling sites illustrated in Figure 3.1).

Grab wastewater samples were collected from the secondary effluent of

conventional activated sludge processes (CAS) in the six WWTPs in

Singapore. Characteristics of WWTPs are listed in Table 3.3. According to the

composition of the influent, WWTPs are categorized to domestic WWTPs and

industrial WWTPs. Industrial WWTPs include W2 and W5, while the rest are

domestic WWTPs. For all WWTPs, raw sewage is collected by separate sewer

system which receives no stormwater runoff. Outfalls of W1 and W2, OW1

and OW2, are 0.4 km and 1 km away from the shore, by which their effluents

are partially discharged to the sea. Samples were collected and stored in high


52

density polyethylene bottles. Sample bottles were kept on ice and brought

back to the laboratory within 6 h of collection.

Table 3.3 Wastewater treatment plants characteristics.

WWTP Treatment process

Flow (m3/d) Wastewater treated

W1 CAS, MBR 232000 90% domestic/commercial, 10% industrial



W4 CAS,

MBR, LTM 361000 95% domestic/commercial, 5% industrial

W5 CAS 205000 40% domestic, 60% industrial

W6 CAS 282000 95% domestic/commercial, 5-10% industrial

CAS: conventional activated sludge system; MBR: membrane biological reactor; LTM: liquid treatment module (an activated sludge process operated

with short sludge retention time).

3.2.2 Water sample preparation (Basic SPE extraction)

Immediately after the collection, water samples were filtrated by GF/B glass

filter (Whatman, USA) and stored at -20 oC until extraction. In most cases,

samples were extracted within 24 h after collection. Extraction procedure used

was similar to that described previously (So et al., 2004) with minor

modifications. Briefly, analytes were extracted using 500-mg hydrophilic-

lipophilic balance (500mg, 6 cc HLB) cartridges, which were sequentially

preconditioned with 5 mL of methanol and 10 mL of Milli-Q water. 1 L of

filtrated water sample was loaded on cartridges at a fixed flowrate of 10

mL/min, after which the cartridges were rinsed with 5 mL of 40% methanol in

Milli-Q water. Then the cartridges were completely dried by air on the

manifold using the vacuum pump. Finally, the cartridges were eluted with 2×2


53

mL methanol into a polypropylene tube. The resulting extract was reduced to

1 mL under a gentle stream of pure nitrogen gas. Final extracts were filtrated

by 0.2 μm nylon syringe filters to remove fine particles and then transferred to

polypropylene vial for analysis.

3.3 Experiment on development of clean-up method for wastewater and

sludge samples

3.3.1 Silica cartridge clean-up procedure

An additional solid phase cartridge, Sep-Pak plus silica (1g) cartridge, was

applied to further remove the interfering matrix in the extracts from basic HLB

SPE procedure. After eluting the HLB cartridge with 2×2 mL methanol,

extracts were diluted with 6 mL of dichloromethane. The silica cartridge was

preconditioned with 5 mL of dichloromethane/methanol (60:40, v/v). Then the

diluted extracts were loaded onto the silica cartridge at 10 mL/min and

collected into a polypropylene tube. The extract was evaporated to dryness

under a gentle nitrogen stream and reconstituted in 0.5 mL of

methanol/aqueous ammonium hydroxide (0.01%) solution (70:30, v/v).

Particles that appeared in the final extracts were removed by 0.2 µm nylon

syringe filter.

3.3.2 Application of clean-up method to sludge samples

Sludge samples were extracted according to an established method (Higgins et

al., 2005) and further cleaned by silica clean-up method. 100 mg of freeze

dried sludge was transferred to a 50 mL polypropylene vial and washed by 7.5

mL of 1% acetic acid solution. Each vial was vortexed, placed in the 60 oC


54

sonication bath, and sonicated for 15 min. After sonication, the vials were

centrifuged at 3000 rpm for 10 min, and the acetic acid solution was decanted

into a second 50 mL polypropylene vial. An aliquot of the methanol/acetic

acid extraction solvent mixture (1.7 mL) was then added to the original vial

and the vial was again vortexed and sonicated for 15 min at 60 o

The extracts of both wastewater and sludge sample were filtrated by 0.2 μm

nylon syringe filters to remove fine particles and stored at -20

C before

centrifuging and decanting the extract. This process of acetic acid washing

followed by methanol/acetic acid extraction was repeated 3 times, and a final

7.5 mL of acetic acid wash was performed. For each sludge sample, all washes

and extracts were combined. The resulting extracts of 35.1 mL were loaded

onto Oasis HLB cartridge (500 mg) at 1-2 mL/min, which was preconditioned

by 5 mL of methanol and 10 mL of 1% acetic acid. Then the cartridge was

eluted by 2×2 mL methanol after it was rinsed by 10 mL of Mill-Q water.

Instead of reducing matrix interference by dilution, the silica clean-up method

was used to remove the matrix effect. After silica cartridge clean-up, the

eluent was concentrated under nitrogen gas to complete dryness. The final

extracts were kept to 1 mL by 70:30 (v/v) methanol/aqueous ammonium

hydroxide (0.01%) solution.

o

C until analysis.

Immediately prior to LC/MS/MS analysis, the 0.5 mL aliquot of extract was

transferred to an autosampler vial for analysis.

3.3.3 Evaluation of matrix effect and recoveries

Matrix effect was evaluated in correspondence to the strategy applied by

Matuszewski et al. (2003) with some modification. As analytes occurred


55

ubiquitously in the aquatic environment and were observed at the level of tens

ng/L in the collected raw sewage (data not shown), the principle of standard

addition was employed to modify the applied strategy. MS/MS peak areas of

known amount of working standards is defined as A, while those of raw

sewage extract as B. MS/MS peak areas of raw sewage extracts spiked with

the same amount of analytes after and before SPE extraction is defined as C

and D, respectively. The matrix effect (ME) is calculated by comparing

MS/MS area for known amount of analytes spiked after extraction of raw

sewage (C-B) with those of the same amount of working standards (A). The

comparison of MS/MS area for known amount of analytes spiked before

extraction of raw sewage (D-B) with those of same amount of working

standards (A) is defined as recovery efficiency (RE). ME and RE are

calculated as followed:

ME%=(C-B)/A×100 (eq. 3.1)

RE%=(D-B)/A×100 (eq. 3.2)

The absence of absolute matrix effect is denoted by a ME% value of 100%,

which implies that the response of standards is same as that of extracts. There

is matrix suppression if ME% is <100%, while ME% of >100% indicates

matrix enhancement. In case of calculations associated with internal standards,

area ratios (area of analyte/area of internal standard) were applied instead.

3.4 Wastewater and sludge sample collection and preparation

3.4.1 Wastewater and sludge sample collection

Samples were collected from two STPs, plant A and B in Singapore. Plant A,

the largest STP in Singapore, treats 361 MLD (million litre per day), which


56

consists of 95% domestic wastewater and 5% industrial and commercial

wastewater. The plant comprises of conventional activated sludge process

lines (CAS1) in parallel with liquid treatment module (LTM) and MBR

(Figure 3.2). CAS1 is operated with a SRT of ~15 days and hydraulic retention

time (HRT) of 8 h, while LTM with SRT of ~3.5 d and HRT of 6 h,

respectively. For the MBR, SRT and HRT are ~20 d and 6 h. Plant B, which

only has conventional activated sludge process (CAS2), treats 205 MLD,

among which 60% is industrial wastewater and 40% is domestic wastewater. It

is run with a SRT of ~12 d and HRT of 10 h. Both CAS1 and CAS2 have

similar treatment process, mainly consisting of primary clarifier (PC), aeration

tank (AT) and secondary clarifier (SC). The produced sludge together with the

solids obtained from the primary clarifier is disposed by thickening centrifuges,

followed by anaerobic digester and dewatering centrifuges.

Screen Grit Removal

Primary Clarifier

Aeration Tank

Secondary Clarifier

Thickening Centrifuge

Anaerobic Digester

Dewatering Centrifuge

Raw Sewage

Inorganic Solids

Primary Sludge Secondary

SludgeReturned Activated Sludge

Effluent

Dewatered Sludge

Primary Clarifier

Liquid Treatment Module

Seconday Clarifier

Primary Clarifier MBR

Effluent

Effluent

CAS 1 (CAS2 in Plant B)

LTM

MBR

267 MLD (205 MLD)*

71 MLD

23 MLD

Grab sample

Returned Liquid

Returned Activated Sludge

Influent Primary effluent

Aeration effluent

Secondary effluent

Thickened sludge

Digestered sludge

LTM-in LTM-pc LTM-eff

MBR-pc MBR-eff

MBR-sup

* denote the flowrate of CAS2 (plant B)

Figure 3.2 Flow scheme of the sewage treatment plants A (CAS1, LTM and MBR) and B (CAS2).


57

Sampling campaigns were conducted in the October 2006, December 2006,

March 2007, September 2007 and December 2007 in plants A and B.

December is the wet season, which has much more precipitations than other

months, and other months are dry seasons in Singapore. For the sampling in

December, it was performed after it had been raining for a week and when it

was raining. For the sampling in other than months, there was no significant

rainfall in the period. During the sampling campaigns, outside temperature

ranged from 21 to 32 o

C. Grab aqueous samples were taken in plants A and B

and sampling points were shown in Figure 3.2. Grab samples of primary,

activated, secondary, returned activated, thickened and anaerobically digested

sludges were also collected.

3.4.2 Wastewater and sludge sample preparation

Wastewater and sludge samples were extracted according to the developed

method (Silica clean-up method) described in section 3.3.1 and 3.3.2. The

extracts of both wastewater and sludge samples were filtered by 0.2 μm nylon

syringe filter to remove fine particles and then stored at -20 o

C until analysis.

Immediately prior to LC/MS/MS analysis, the 0.5 mL aliquot of extracts was

transferred to an autosampler vial and 50 μL of 50 μg/L aqueous internal

standard mixture containing perfluoro (2-ethoxyethane) sulfonic acid (PFEES)

and perfluorododecanoic acid (PFDoA) was added.


58

3.5 PAC-MBR experimental setup and operation

3.5.1 MBR and PAC-MBR configuration

Experiments were performed in four identical lab-scale submerged MBR and

PAC-MBR systems as illustrated in Figure 3.3. Each MBR consisted of

regular tank with an operating volume of 16 L and a microfiltration (MF)

membrane module submerged in the tank. The membrane module was made

of polyvinylidene fluoride (PVDF) hollow fibre membrane with a pore size of

0.2 µm and filtration area of 0.4 m2, which was mounted between two baffle

plates located above an air diffuser in the MBR. Two baffle plates were

mounted above the air diffuser to optimize the contact between air bubbles and

the membrane surface. Compressed air (36 L/h) was supplied through the air

diffuser to provide good mixing of the activated sludge and cross flow action

for effective scouring of the membrane surface. The membrane flux was kept

constant at 10 L/m2·h and followed a suction cycle of 8 min on and 2 min off.

Two water level sensors were installed at the high and low water level

respectively to maintain a constant water level in the bioreactor. Both the

bioreactor and the storage tank were initially filled with the synthetic

wastewater. The storage tank with an effective volume of 200 L was reloaded

everyday with the fresh wastewater to ensure the continuous supply to the

MBR over the entire experimental period. To minimize the variation of

wastewater characteristics, the storage tank was thoroughly cleaned every two

days to reduce the growth of microorganisms.


59

Magnetic stirrer

Synthetic wastewater storage tank

Feeding pump

PAC dosing pump

PAC tank

Suction pump

Pressure guage

Flow meter

Air

MF membrane module

Air diffuser

Baffles

Figure 3.3 Schematic diagram of lab-scale PAC-MBR system. (For MBR system, no PAC dosage system)

3.5.2 Synthetic wastewater and operational conditions

The composition of the synthetic wastewater used in this study is listed in

Table 3.4. The carbon source was mainly from sodium acetate which is simple

and readily biodegradable. The influent COD concentration was 600±20 mg/L

with the ratio of COD: N: P maintained at 100: 10: 1.

PAC dosage system


60

Table 3.4 Composition and concentration of synthetic wastewater.

Components Molecular weight (Da) Concentration (mg/L)

CH3COONa 82 768.75

(NH4)2SO4 132.1 284

KH2PO4 136.1 26

CaCl2 ·2H2O 147 0.368

MgSO4 ·7H2O 246.5 5.07

MnCl2·4H2O 197.9 0.275

ZnSO4·7H2O 287.5 0.44

FeCl3 162.2 1.45

CuSO4·5H2O 249.7 0.391

CoCl2·6H2O 237.9 0.42

Na2MoO4·2H2O 242 1.26

Yeast extract -- 30

Seed sludge was obtained from the aeration tank of a local pilot MBR system

for municipal wastewater treatment. After transferring into the lab-scale MBR,

the sludge was allowed to acclimate to the synthetic wastewater for 35 d.

During the startup period, the MBR was operated at the same condition as that

used in the experimental period except no sludge wastage. The experiments

were performed in three phases according to the change of SRT in the order of

30, 16 and 5 d. Before transferring to a new phase, a period of at least two

times of the new SRT was provided for MBR stabilization. In each phase, a

steady state of four weeks was maintained, during which measurements were

evenly conducted for parameters of interest. PAC was dosed into the MBRs


61

with the dosage 30, 80 and 100 mg/L. Table 3.5 shows the PAC dose added to

the MBR system.

Table 3.5 PAC added at the startup of PAC-MBR system.

SRT (d) PAC

dosage (mg/L)

PAC amount added (g)

PAC calculated concentration in PAC-MBR (g/L)

30

30 72 4.5

80 115.2 7.2

100 144 9

16 100 76.8 4.8

5 100 24 1.5

The hydraulic retention time (HRT) of 8 hours and DO concentration of

around 5 mg/L were maintained during the entire experimental period of 515 d.

The MBRs was operated under ambient temperature (28 ± 2 °C) and the pH

was controlled within a range of 6.8-7.5. Fouling development, indicated by

the increase in suction pressure, was monitored by pressure gauges. Membrane

cleaning was carried out in about 47-132 d when the suction pressure

increased beyond 26 kPa. Typically, the interval between two membrane

cleanings became shorter as SRT decreased indicating membrane fouling was

more serious at short SRTs. The membrane module was taken out of the MBR.

It was rigorously rinsed with tap water to remove the attached cake layer

followed by backwashing with 0.05% sodium hypochlorite solution for 2 h to

further remove the foulants adsorbed within membrane pores. The membrane

module was thoroughly cleaned again with tap water before it was mounted

back in the MBR.


62

3.5.3 PFCs mass balance calculation

The mass balance in MBR or PAC-MBR was shown in Figure 3.4. In the

PAC-MBR system, WAS includes waste activated sludge and PAC.

PAC-MBR (MBR)

Q0, C0 Qe, Ce

(Q0-Qe), Ce

Cs

WAS

Aqueous phase

Solid phase

Figure 3.4 Mass balances of PFCs in PAC-MBR or MBR system. 1. Q0 and Qe: flow rate of influent and effluent; 2. C0 and Ce: PFCs concentration in influent and effluent; 3. Cs: PFCs concentration in wasted solids; 4. WAS:

waste activated sludge.

3.5.4 Membrane resistance calculation

The transmembrane pressure (TMP) increased with the increase of operation

time while flux was maintained constant. The resistance-in-series model was

applied to evaluate the fouling characteristics. The permeate flux of a

membrane is governed by the basic membrane filtration equation as follows:

tR

PJµ∆

= (eq. 3.3)

Where J is the permeate flux, ΔP is the transmembrane pressure (TMP), µ is

the permeate viscosity, Rt is the total membrane resistance. The total

membrane resistance, typically, includes three parts, i.e.,

irmt RRRR ++= (eq. 3.4)

where Rm is the intrinsic membrane resistance, Rr is the resistance due to

reversible fouling caused by the cake layer deposited over the membrane


63

surface, and Ri is the resistance due to irreversible fouling caused by solute

adsorption into the membrane pores.

At the end of the experiment, the fouled membrane module was rigorously

rinsed three times with DI water. After physical cleaning, the TMP of

membrane (ΔP’) was measured by filtration of pure water. Based on the

experimental data, the values of Rm, Rr, and Ri can be determined as follows.

JP

Rm µ0∆

= (eq. 3.5)

mi RJPR −

∆=µ

' (eq. 3.6)

imf

r RRJP

R −−∆

=µ

(eq. 3.7)

where ΔP0 is the TMP measured by filtrating pure water with virgin

membrane, ΔP’ is the TMP measured by filtrating pure water with fouled-

membrane after physical cleaning, and ΔP f is the final TMP at the end of

experiment.

3.6 Adsorption study on PAC and activated sludge

3.6.1 Preparation of EfOM

EfOM solution was collected from the mixed liquor of the laboratory-scale

MBR. Immediately, the mixed liquor was centrifuged at 3000 rpm for 10 min

followed by filtration by GF/B glass filter (0.45 µm, Whatman, U.S.A) and

stored at 4 oC until adsorption experiments.


64

3.6.2 EfOM characterization

The fractionation method used in this study was basically based on the

procedure developed by Barker et al., (1999a) with minor modification. The

apparent molecular weight distribution (AMWD) of the EfOM was

determined using ultrafiltraton (UF) membrane in a stirred and pressurized cell

(Model 8200, Amicon, USA), operated in dead end mode. The filtrate

permeating through each YM membrane was collected and DOC

concentration was measured. Nitrogen gas regulated at 30 psi pressure was

used as a driving force for filtration. Gentle turbulence was created at the

membrane surface using a magnetic stirrer to minimize the build-up of a dense

macromolecular layer at the membrane surface. The percentage of organic

matters for each fraction was calculated in terms of DOC based on the mass

balance. The <1 kDa and >30 kDa fractions were obtained and used for the

adsorption experiments. Sodium chloride solution (0.013M), which is of same

ionic strength as MBR effluent, was added to the >30 kDa fraction to achieve

a same DOC concentration as that of <1 kDa fraction.

3.6.3 Equilibrium adsorption experiments

PAC equilibrium adsorption experiments were conducted in duplicate in

EfOM free solution (Mill-Q water), EfOM raw solution and EfOM fractions

(<1 kDa and >300 kDa fractions). PFOS or PFOA stock was spiked to the

solutions and initial single adsorbate concentration ranged from 0.1 to 500

µg/L. Different amount of PAC was added at the appropriate dosage and 1

mM phosphate buffer (0.5mM Na2HPO4 and 0.5 mM NaH2PO4) was spiked to

maintain pH at 7.2.


65

Sludge equilibrium adsorption experiments were conducted in duplicated with

activated sludge at the concentration of 2,000-5,000 mg/L. 1mM phosphate

buffer was spiked to maintain pH at 7.2. PFOS or PFOA stock was spiked to

the sludge solution at the concentration of 50-400 µg/L. Sodium azide (100

mg/L), a respiratory inhibitor, was added to prevent microbial metabolism.

All equilibrium adsorption batch experiments were carried out in an incubator

shaker (CMR, USA) at 25 oC with shaking speed of 120 rpm. Bottles was

sealed and agitated on the shaker for 5 d to reach adsorption equilibrium. Then

PAC particles or sludge were separated by GF/B glass filter (0.45 µm,

Whatman, USA) for the analysis of remaining PFCs concentration in liquid

phase.

3.6.4 Adsorption kinetics experiments

Batch kinetics experiment was conducted in duplicate to determine the

kinetics parameters that describe the rate of removal of the target

perfluorinated compounds by PAC. The initial PFOS or PFOA concentration

for kinetics experiments was 100 µg/L and 1mM phosphate buffer (0.5mM

Na2HPO4 and 0.5 mM NaH2PO4) was spiked to maintain pH at 7.2. PFOS or

PFOA stock was added to 1 L of EfOM solution stirred in a 1-L HDPE bottle

with magnetic stirrer. After 20 min mixing, PAC was added and samples were

collected at predetermined intervals over 6 h. Samples were then filtrated

through GF/B glass filter (0.45 µm, Whatman, USA) to remove PAC.


66

3.6.5 Mathematical modeling

The most frequently used two isotherm models, Langmuir and Freundlich

equations were applied to fit the experimental data to determine the adsorption

capacity of PAC and sludge. These equations describe the non-linear

equilibrium between adsorbed organic compounds on the solid surface and

organic compounds in solution at a constant temperature. The Langmuir

equation which is valid for monolayer adsorption onto a surface with a finite

number of identical sites is given by

CebCebaCs⋅+⋅⋅

=1

(eq. 3.8)

where Cs is the concentration of the solute in the solid phase, Ce is the

equilibrium concentration of the solute in solution; a and b are Langmuir

constants related to maximum adsorption capacity (monolayer capacity) and

bonding energy of adsorption, respectively. The Langmuir equation is used for

homogeneous surfaces. The Freundlich equation assumes neither

homogeneous site energies nor limited levels of adsorption. The Freundlich

equation is defined by

nF CeKCs /1⋅= (eq. 3.9)

where KF and n are the Freundlich constants in relation to adsorption capacity

and adsorption intensity, respectively. The KF

However, the relationship between equilibrium concentrations of organic

compounds in liquid and solid phase could be linear and defined by simple

partition coefficients. For n=1, the partition between the two phase is

independent of the concentration and isotherms becomes linear Freundlich

value corresponds to the

adsorption capacity (ug adsorbate/mg carbon) at an equilibrium concentration

of 1.0 µg/L.


67

euqation. In this case, the experimental data are fitted to linear adsorption

isotherm defined by

CeKCs d ⋅= (eq. 3.10)

where Kd

is the partition coefficient.

3.7 Analysis method

3.7.1 COD and DOC analysis

Chemical oxygen demand (COD) was determined in accordance with

Standard Methods (APHA-AWWA-WEF, 1998). Dissolved organic carbon

(DOC) was measured by 1010 Total Organic Carbon Analyzer (O.I.Analytical,

USA).

3.7.2 Carbohydrate and protein analysis

Carbohydrate and protein were determined according the method of Dubois et

al. (1956) and Lowry et al., (1951), respectively. The phenol-sulfuric acid

method (Dubois et al., 1956) was used to measure the content of carbohydrate

in DOM with glucose as the standard reference, whereas the modified Lowry

method (Lowry et al., 1951; Hartree, 1972) was used for protein determination

with bovine serum albumin (BSA) as the standard reference.

3.7.3 MLSS and MLVSS

Sludge concentration was measured as mixed liquor suspended solids (MLSS)

and volatile suspended solids (VSS) in accordance with Standard Methods

(APHA-AWWA-WEF, 1998).


68

3.7.4 EPS and SOUR analysis

EPS content in biomass was extracted and determined using the established

procedure (Frølund et al. 1996; Ng et al., 2005). First, 200 mL biomass sample

was centrifuged at 2000 g at room temperature for 15 min and the supernatant

decanted. The centrifuged biomass was resuspended back to 200 mL with a

fresh phosphate buffer (526 mg/L NaCl, 74.56 mg/L KCl, 760.2 mg/L Na3PO4

and 552 mg/L NaH2PO4). Then cation exchange resin (DOWEX Marathon C,

Fluka Cat No. 91973F) was added to resuspended sample which was

transferred to a closed container, at 90 g/gVSS. The mixture was stirred at 600

rpm for 2 h in an ice water bath and then centrifuged at 12,000 g for 30 min to

remove the resin and the microorganisms. The supernatant was then analyzed

for carbohydrates and proteins using the analyzing method above (section

3.7.2). Specific oxygen uptake rate (SOUR) was measured in accordance with

the Standard Methods (APHA-AWWA-WEF, 1998).

3.8 LC/MS/MS analysis

3.8.1 Optimization of LC/MS/MS method

High performance liquid chromatograph, composed of a HP100 liquid

chromatograph (Aligent Technologies, U.S.A) interfaced with a triple

quadrupole MS/MS (Applied Biosystems, U.S.A) was applied to detect

samples in the electrospray negative ionization mode. Separation of

compounds was performed on a 150×2.1 mm (5 μm) Zobax Extend C18

column (Aligent Technologies, U.S.A). A 10-μL aliquot of the sample extract

was injected into a guard column (XDB-C8, 2.1 mm i.d.×12.5mm, 5μm;

Agilent Technologies, U.S.A) connected sequentially to Zorbax Extend C18


69

column with 2 mM ammonium acetate aqueous solution (solvent A) and

methanol (solvent B) as mobile phases, starting at 3% of solvent B. The flow

rate was set at 300 uL/min. The gradient was held until 0.50 min, increased to

95% B until 6.00 min, held until 8.50 min, reverted to original conditions at

8.51 min and was held 3% B until 12.00 min. The column temperature was

kept constant at 30 o

C.

Table 3.6 MRM-transitions, compound-dependent parameters of the analytes.

Compound MRM-

transitions (amu)

Decluster Potential

(V)

Focusing Potential

(V)

Entrance Potential

(V)

Collision Energy (V)

Collision Cell Exit Potential

(V) PFOS 499 99 -31 -310 -3.5 -12 -10

PFOA 413 369 -51 -350 -9.5 -60 -6

PFEES 314.5 135 -41 -70 -9 -28 -6

PFDoA 613 569 -26 -150 -3 -16 -18

MS/MS was operated under multiple reaction monitoring (MRM) mode. The

mass spectrometer was operated using the TurboIonsprayTM (TIS) source in

the negative mode. The ionization source-specific parameters were: curtain

gas (CUR), 30 psi; collision gas (CAD), 6 psi; ionspray voltage, -4500V;

temperature of the turbo heater gas, 450 o

C; nebuliser gas (GS1), 40 psi.; turbo

gas (GS2), 90 psi. Nitrogen was used as the curtain gas, nebuliser gas and

turbo gas. Other analyte-dependent parameters were optimised for each

compound (Table 3.6). Methanol was run between the water samples to

prevent carryover effect.


70

3.8.2 Method validation and quantification

The analytical characteristics of the method, such as linear response range,

reproducibility and limits of quantification, were investigated to evaluate the

efficiency of the method and the possibility of the method application to

various water samples. Retention Time was 6.98 min (PFOS) and 6.80 min

(PFOA) in Figure 3.5, which showed a chromatogram of PFOS and PFOS in

spiked Mill-Q water at 1 ng/L.

Figure 3.5 LC/MS/MS chromatograms of PFOS, PFOA and internal standards PFEES and PFDoA.

Seven calibration curve points bracketing the concentrations in samples were

prepared routinely, to check for linearity. Quantification was based on the

response of the external standards that bracketed the concentrations found in

samples. The curve covered a range equivalent to the concentration of the

analytes in 1000 mL water sample after the extract was concentrated to 1.0 ml

(approx. 1000-fold concentration). A calibration curve containing 0.25, 0.5, 1,

XIC of -MRM (4 pairs): 412.4/368.5 amu from Sample 8 (6) of DataSET1.wiff (Turbo Spray) Max. 1320.0 cps.

1 2 3 4 5 6 7 8 9 10 11 12Time, min

0.0

5.0e4

Inte

nsity

, cps


1 2 3 4 5 6 7 8 9 10 11 12Time, min

0

1000

Inte

nsity

, cps

9.52

2.21 9.69 12.777.780.57 4.131.77 2.51 10.295.35 10.537.915.55 11.940.68 4.43 6.02 6.60 11.302.62 8.701.360.37 9.083.803.19 4.76 6.75 7.14

XIC of -MRM (4 pairs): 498.2/79.9 amu from Sample 8 (6) of DataSET1.wiff (Turbo Spray) Max. 5.5e4 cps.

1 2 3 4 5 6 7 8 9 10 11 12Time, min

0.0

5.0e4

Inte

nsity

, cps

9.68

XIC of -MRM (4 pairs): 314.5/134.9 amu from Sample 8 (6) of DataSET1.wiff (Turbo Spray) Max. 1.7e4 cps.

1 2 3 4 5 6 7 8 9 10 11 12Time, min

0.0

1.0e4

1.7e4

Inte

nsity

, cps

8.93


1 2 3 4 5 6 7 8 9 10 11 12Time, min

0

20002760

Inte

nsi

ty, c

ps

10.19

4.91 10.650.13 6.97 12.078.053.100.83 2.49 5.613.59 6.62 7.34 11.821.89 12.713.711.25 4.52 6.40 9.42 9.578.377.46 9.190.74

PFOS

PFOA

PFOS

PFEES

PFDoA


71

5, 10, 50, 100, and 250 μg/L standard was used. The correlation coefficients

(r2

) exceeded 0.995. The instrumental detection limits (IDL) were 1-5 pg, as

summarized in Table 3.7. The limit of detection (LOD), defined as the

concentration that yielded an S/N ratio of higher than or equal to 3, and the

limit of quantification (LOQ), defined as the concentration that yielded a S/N

ratio of higher than or equal to 10, were determined by the SPE extraction of

spiked Mill-Q water samples.

Procedural recovery was evaluated by spiking mixture of external standards

(100 ng/mL, 100 µL) to Milli-Q water. Procedural recoveries for PFOS and

PFOA were in the range of 95-103% (mean: 98.4%, n=3) and 90-98% (mean:

93.8%, n=3), respectively. During the analysis of samples, procedural and

instrumental blanks were analyzed. They were below the detection limit,

indicating no contamination occur in sampling and analysis. Sample extracts

with concentrations exceeding the range of calibration curve were

appropriately diluted by methanol and reinjected again. Furthermore, spiked

additions were applied to identify the matrix suppression on the ion signals for

each batch of samples based on the standard additoin method. They were

prepared by spiking mixture of external standards (100 ng/mL, 100 µL) into

the SPE extracts. Mean recoveries of spiked additions for PFOS and PFOA

were 84-91% and 70-84%, respectively. Sufficient recoveries achieved for

procedural blanks and spiked additions proved the reliability and efficiency of

analysis method.


72

Table 3.7 IDL, LOD and LOQ of PFOS and PFOA.

Analyte IDL (pg)

Water and Wastewater Sludge Recovery

LOD (ng/L)

LOQ (ng/L)

LOD (ng/g)

LOQ (ng/g) Water

Waste-water Sludge

PFOS 1 0.1 0.25 1 2 88% 91% 84% PFOA 5 0.5 1.25 5 8 84% 82% 70%

3.9 Fractionation process

The fractionation method used in this study was basically based on the

procedure developed by Leenheer (1981) and Thurman (1985) except that the

anion exchange resin Duolite A-7 was substituted by Amberlite IRA-96, since

this type of resin was also suggested for fractionation process by Chang et al.

(2002) and it was readily available. Resins used (XAD-8, AG MP-50and IRA-

96 exchange resin) were pre-purified using the Soxhlet extraction method

described by Leenheer (1981).

Prior to the fractionation process, the columns (i.d = 25mm x 100mm),

endpieces and the accompanying frits for uniform water distribution were

washed with HCl acid (~0.3M) to remove trace carbon. The service flow rate

used for XAD8 resin was about 15 BV/h; while the service flow rates used for

ion exchange resins were about 30 BV/h. After removal of suspended solids

by MF (microfiltration) and adjustment of pH to 7 by HCl, 100-300 L

(according to resin capacity) of water sample was introduced and passed

through three types of resins (Figure 3.6). The compounds adsorbed by the

first XAD-8 resin column were eluted using 100 ml 0.1N HCl, defined as

Hypho-B. The filtrate was acidified to pH 2 with 2M HCl and then re-

introduced into another XAD-8 column. The organic matters adsorbed by the


73

second XAD-8 resin column were eluted using 100 mL of 0.1N NaOH as a

brownish solution defined as AHS (acid humic substance, also called Hypho-

A) containing HA (humic acids) and FA (fulvic acids). Then the second XAD-

8 resin column was dried at 60 °C and the residual matters were washed out by

methanol (50 mL) to get the Hypho-N. A vacuum concentration instrument

(BÜCHI Rotavapor R-124, Switzerland) combined with high purity nitrogen

gas was used to concentrate this solution. The Rotavapor was operated under a

vacuum pressure around 900 mbar and at a temperature of 62 °C with a

rotation speed of 50 rpm, and the whole process lasted for 20-30 min. The

portion that passed through the second XAD-8 resin column, which contained

only hydrophilic solutes, was pumped through the AG-MP-50 cation-

exchange resin column. Hyphi-B retained on this cantion-exchange resin, was

eluted by 100 ml of 2M HCl. The filtrate was pumped through the IRA-96

anion-exchange resin column and the Hyphi-A absorbed on this resin was

eluted with 100 ml of 1M NaOH. The final effluent, which passed through

three types of resins, was defined as Hyphi-N.


74

Figure 3.6 Procedure for fractionation of DOM.

3.10 Quality assurance and control

Because of the presence of the fluoropolymer in some laborotary equiments,

precautions were taken to minimize the possible contamination during the

analysis (Yamashita et al., 2004). For example, teflon bottles, Teflon-lined

caps, and any suspected fluoropolymer materials were not utilized throughout

the analysis. In order to ensure the quality of the sampling, Milli-Q water was

used as field blanks to evaluate the possible contamination during the

transportation for each batch of samples. For each field blank, PFOS and

Methanol

Hyphi-N

pH =2

XAD-8 resin

0.1N NaOH

AG-MP-50 cation resin

IRA-96 anion resin

Hypho-N 2 N HCl

AHS

1N NaOH

Hyphi-B

Hyphi-A

MBR supernatant

MF filter

XAD-8 resin

0.1 N HCl

Hypho-B

(Hypho-A)


75

PFOA were below the detection limit, indicating that no discernable

contamination occurred during sampling.

Spiked additions were applied to identify the matrix suppression on the ion

signals for each batch of samples based on the standard addition method. They

were prepared by spiking mixtures of external standards (100 ng/mL, 100 µL)

into the SPE extracts of the effluents obtained from W4. Recoveries of spiked

additions for PFOS and PFOA were in the range of 80-93% (mean: 87.8%,

n=5) and 78-90% (mean: 83.9%, n=5), respectively. Sufficient recoveries

achieved for spiked additions demonstrated the reliability and efficiency of the

analysis method.

3.11 Statistical analysis

Statistical software Minitab (Minitab Inc, USA) was used to calculate the

correlation between PFOS and PFOA as well as the correlations of

concentrations between dry season and wet season. Statistical significance was

accepted at p<0.05.

Chapter 4-Occurrence of PFOS and PFOA in Water and Wastewater

76

CHAPTER 4 OCCURRENCE OF PFOS AND

PFOA IN WATER AND WASTEWATER

4.1 Introduction

PFOS and PFOA are ubiquitous in the environment because of their high

persistence, resulting from their exceptionally thermal and chemical stability.

Surveys have been conducted to monitor the extent of PFOS and PFOA

contamination in surface waters (Hansen et al., 2002; Boulanger et al., 2004;

Loos et al., 2008), wastewaters (Boulanger et al., 2005; Becker et al., 2008),

drinking waters (Harada et al.,2003), groundwaters (Schultz et al., 2004) and

coastal waters (So et al., 2004; Saito et al., 2003; Yamashita et al.,2005). The

pathways of PFCs to aquatic environment could include (a) discharge of

effluents from STPs, (b) direct discharge of wastewater from manufacture and

use of PFCs to the aquatic environment, (c) rain runoff moving PFCs

pollutants on ground (such as oil, fire-fighting foam) to the aquatic

environment, and (d) atmospheric transport of PFCs and subsequent

atmospheric loading of PFCs to surface waters (Prevedouros et al., 2006;

Zushi et al., 2008). A few studies have been conducted to identify the

contamination source of PFCs in the environment. Some researchers observed

that effluents from the STPs are the most important PFCs sources for the

aquatic ecosystems (Sinclair and Kannan , 2006; Loganathan et al., 2007).

Zushi et al. (2008), however, reported that loads of PFCs in rain runoff were

about 2-11 folds greater than those in STP effluents that were discharged into

a river, indicating that nonpoint source of PFCs could be the most important


77

source for the river studied. In addition, Yamashita et al. (2004) reported that

application of PFC-containing products could also be an important source of

contamination for aquatic environment. It seems that effluents from STPs,

nonpoint source from rain runoff and application of PFC-containing products

might be important sources and determine the PFCs concentration levels in the

aquatic environment. However, there could be other significant PFCs sources

such as atmospheric deposition or precipitation for the aquatic environment.

Therefore, further research is needed to identify possible contamination

sources and transportion pathways of PFCs in aquatic environment.

Furthermore, seasonal variations in the PFCs concentraions were investigated.

So et al (2004) observed that PFCs concentrations in the winter were higher

than those in the summer in coastal waters of China. In wastewater of STPs,

Loganathan et al. (2007) found that mass flow of PFCs were higher in winter

than in summer. The authors suggest that there were less rain in winter than in

summer, which resluted in dilution effect on the coastal waters or wastewaters

in summer. However, no data is available on the comparison of PFCs

concentrations between dry season and wet season in the aquatic environment.

Singapore is an island coutry and also a true city-state with a tropical

rainforest climate and no distinctive seasons. Especially its climate is

characterized by uniform temperature, pressure and abundant rainfall in wet

monsoon season (November and December). In a such an unique isoland city,

it could be an ideal place to identify seasonal variations of PFCs

concentrations between dry seasons and wet seasons by excluding other

factors, such as temperature and atmospheric pressure variation. To the best of


78

our knowledge, this study is the first study to identify the seasonal variations

of PFCs in aquatic environment between dry season and wet season.

In order to investigate the distribution of PFOS and PFOA in different water

matrices in Singapore, 138 water samples were collected from reservoirs,

rivers/canals, wastewater treatment plants and coastal waters around the island

over a year. The purpose of this study was to determine the magnitude and

extent of PFCs’ contamination and to provide an overview of the spatial

distribution of PFOS and PFOA in the waters of Singapore. Moreover, surface

water samples in the industrial districts and wastewater from all six WWTPs

in Singapore as well as coastal water samples were collected and analyzed in

an attempt to locate possible contamination sources within the island. In

addition, seasonal variations between dry season and wet season were studied.

The results of this study would identify the sources and transport pathways of

PFCs in the aquatic and oceanic environment of Singapore.

4.2 Results and discussions

4.2.1 PFOS/PFOA concentration in surface water

Concentrations of PFOS and PFOA in surface waters, wastewaters and coastal

waters from five batches of sampling campaigns are summarized in Figures

4.1-4.3, which show the spatial distribution of those two compounds in

western, middle and eastern areas of Singapore. Overall, PFOS and PFOA

concentrations in all samples were in the range of 2.2-532.1 ng/L and 2.4-

1,057.1 ng/L, respectively.


79

The concentrations of PFOS and PFOA in surface waters ranged from 2.2-

87.3 ng/L and from 5.7-91.5 ng/L, respectively. This is comparable to but

slightly higher than those observed in the Great Lakes (USA) (PFOS: 21-70

ng/L, PFOA: 27-50 ng/L) (Boulanger et al., 2004). Comparable PFOS

concentration range was also observed in Guangzhou (0.9-99 ng/L) (So et al.,

2007), one of most industrialized areas in China. The highest concentration of

PFOS (87.3 ng/L) in surface waters was detected at S5, eastern area subjected

to light industrial influence. This indicated potential PFCs contamination

sources nearby. In comparison to other studies, however, the highest PFOS

concentration was approximately half of that reported in Tama river in Japan

(157 ng/L) (Saito et al., 2003) and in downstream of discharge of 3M

fluorochemical manufacturing facility (144 ng/L) (Hansen et al., 2002). The

concentration of PFOS detected at S5 was also about 7 times lower than the

highest concentration (651 ng/L) measured in Lake Shihwa (South Korea),

which is heavily influenced by the industrial effluent from the Shihwa

industrial district (Rostkowski et al., 2004).

Compared to this study, lower PFOA concentration range was observed in

Guangzhou area (0.85-13 ng/L) (So et. al., 2007) and Pearl River Delta (0.24-

16 ng/L) (So et al., 2004), both of which are heavily associated with industrial

and urban activities. In particular, the highest PFOA concentration (91.5 ng/L)

in surface water was observed at S7, which was collected downstream of a

canal that flows along the edge of an airport. It suggests that the airport may

be a potential PFCs contamination source. In contrast, the PFOA level in this

study was approximately 2 and 3 times lower than those observed in Tokyo


80

Figure 4.1 Concentrations of PFOS and PFOA in surface waters, wastewaters and coastal waters from western area of Singapore collected by: 1. Oct 2006; 2.

Dec 2006; 3. Mar 2007; 4. Sep 2007; 5. Dec 2007; n.a: not available.

Bay in Japan (Yamashita et al., 2004) and Yangtze Rive in China (So et al.,

2007).

PFOS

PFOA

n.a

n.a

n.a

n.a

n.a


81

Figure 4.2 Concentrations of PFOS and PFOA in surface waters, wastewaters and coastal waters from middle area of Singapore collected by: 1. Oct 2006; 2.


The total PFCs (i.e., PFOS and PFOA) concentrations from 5 sampling

campaigns for all surface waters are summarized in Figure 4.4. It can be seen

that S9, located at the western area, had the highest total PFCs concentration,

which suggests the presence of potential PFCs contamination source in the

surrounding area. However, S7 had the next highest total PFCs concentration

among all surface waters even though its location is in the eastern area (urban

region). This may be due to the leakage of perfluorinated surfactants, such as

PFOS

PFOA

n.a

n.a

n.a

n.a

n.a n.a n.a


82

Figure 4.3 Concentrations of PFOS and PFOA in surface waters, wastewaters and coastal waters from eastern area of Singapore collected by: 1. Oct 2006; 2.


aqueous fire-fighting foams, gasoline, oil and lubricants (Moody et al., 2000),

from the adjacent airport. Similarly, the highest total PFCs concentration in

reservoir waters was detected in R8, which is the downstream of the S9.

Furthermore, R2, R3 and R4 which are in the nature reserve area (middle area)

had lower concentrations compared to other reservoirs which are in either

industrial or commercial influenced areas. In contrast, the higher

concentrations were observed in R7, R8 and R9 which are in the industrial

area (western area). It suggests that factories, such as petrochemical, paints,

coatings and surfactants manufacturing plants in the western region, may be

the potential PFCs sources, thus causing this area to be the most contaminated

area in Singapore. It is, however, worthy to point out that, even though the

PFOS

PFOA

n.a

n.a n.a

n.a


83

R1 R2 R9R3 R4 R5 R6 R7 R8 S1 S2 S3 S4 S8S6 S7S5 S9 S10S11S12

51.7%

64.5%61.4%

54.6%

56.6%

61.9%

69.2%

61.1%

49.9%

63.4%60.8%

61.2%

55.3%

29.5%

64.3%

55.4%

56.8%

59.9%

74.6%67.7%

69.6%

0.0

50.0

100.0

150.0

200.0

250.0

300.0

350.0

400.0

450.0To

toal

PFO

S a

nd P

FOA

con

cent

ratio

n (n

g/L) PFOA

PFOS

Figure 4.4 Total PFOS and PFOA concentrations in surface waters summed up

by 5 sampling campaigns. The value in pencentage above the each column indicates PFOA pencentage in total PFOS and PFOA concentration.

highest concentrations of PFCs in reservoirs were observed in R8 and R9,

their PFOS and PFOA concentrations were a few order of magnitude lower

than the lifetime drinking water health advisory for PFOS (1,000 ng/L)

(Hansen et al., 2002) and PFOA (150,000 ng/L) (Psoulsen et al., 2005).

Overall, the concentrations of PFOS and PFOA in reservoirs (mean: 17.6 ng/L

for PFOS, 24.9 ng/L for PFOA) were comparable to those of river water

samples (mean: 20.2 ng/L for PFOS, 28.4 ng/L for PFOA). Reservoir waters,

however, had relatively lower variation in concentrations for both PFOS and

PFOA in comparison with river/canal waters.

PFOA percentage in the total PFCs concentration varied from 29.5%-74.6% in

surface waters, indicating different contamination sources of perfluorinated


84

compounds (Figure 4.4). PFOA contribution to the PFCs, which were greater

than 50% in all surface water samples except for R9 and S5, indicated that

PFOA was predominant in most of surface waters. Predominance of PFOA

over other PFCs was also observed in the water samples collected along the

Yangtze River, the largest river in China (So et al., 2007). The lowest PFOA

percentage observed in S5 suggests there may be significant PFOS

contamination source upstream or in the surrounding area, which greatly

increase PFOS contribution to the PFCs. In addition, even though location of

R9 is geographically close to R8, the lower PFOA percentage at R9 suggests

different contamination sources occur in its catchment area or its tributaries.

4.2.2 PFOS/PFOA concentration in wastewater

PFOS concentrations were observed in the range of 5.8-532.1 ng/L in the

effluents of WWTPs. This is much higher than those of 10 municipal WWTPs

(1.1-130 ng/L) in USA (Schultz et al., 2006). The highest PFOS concentration

(532.1 ng/L) was detected in the effluent of W1, one of the domestic WWTPs.

This is comparable to those measured in a WWTP in Cleveland with no

known fluorochemical sources (417-454 ng/L) (3M, 2001). With the exception

of W1, PFOS concentrations in effluents of domestic WWTPs, in the range of

5.8-35.3 ng/L, were lower than those of 10 municipal WWTPs in USA

(Schultz et al., 2006). The higher PFOS concentration in W1 as compared to

other domestic WWTPs suggested the occurrence of significant PFOS

contamination sources (such as airport etc.) in the served area. High PFOS

concentration (461.7 ng/L) was also observed in the effluent of W5, which


85

receives 60% industrial wastewater with known fluorochemical sources such

as petrochemicals industry.

PFOA concentration in the effluents of WWTPs was observed in the range of

7.9-1,057.1 ng/L. PFOA concentrations in industrial WWTPs (22.6-1,057.1

ng/L) were much higher than those from domestic WWTPs (7.9-157.3 ng/L).

In the effluents of domestic WWTPs, PFOA concentrations were higher than

those of 10 WWTPs in USA (PFOA: 2.5-97 ng/L) (Schultz et al., 2006) and

comparable to 2 WWTPs in Kentucky and Georgia (PFOA: 6.7-183 ng/L). In

addition, the highest PFOA concentration (1,057.1 ng/L) was also observed in

the effluent of W5. This is significantly higher than those from Cleveland

(665-674 ng/L) but about two times lower than those from Decatur (2,140-

2,420 ng/L) with influence from fluorochemical manufacturing facilities (3M,

2001). The industries in the service area of W5 include petrochemical

intermediates, petroleum refining, paints, coatings and surfactants

manufacturing. Our results suggest that discharges from these fluorochemical

related plants may contain a large amount of PFOS and PFOA or their

precursors, thus leading to high PFCs concentrations in the effluent of

WWTPs. It is evident that discharge of effluents from WWTPs is an important

pathway through which PFCs enter the environment.

Figure 4.5 shows that W5 had the highest total PFCs mass load which was

calculated by multiplying WWTP’s daily flow rate with its total PFCs

concentration. Even though daily flow rate of W5 was comparable to that of

other WWTPs, its total PFCs’ mass load was 2.3-7.0 times of that of other


86

WWTP. It indicates that industrial influent significantly affect the PFCs’ mass

load which was discharged into the aquatic environment. However, much

higher total PFCs mass load occurred in W1, one of the domestic WWTPs,

than that of W2 (industrial WWTP). Compared to W5, there is fewer known

fluorochemical source in the served area of W2. It suggests that the

composition of the industrial influent entering W2 may be different and not

closely related to PFCs’ contamination. Moreover, PFOA was predominant

over PFOS in the WWTPs’ effluents except for W1. The lowest PFOA

percentage observed (22.5%) in the effluent of W1 indicates that there could

be potential contamination sources of PFOS in its service area. It is known that

PFOS is used widely in multiple photolithographic chemicals, such as

photoacid generators (PAGs) and anti-reflective coatings (ARCs) in the

semiconductor industry. Due to its health concerns, PFOS has been phased out

in the European Union semiconductor industry since 2008. However, there is

no restriction on the applications of PFOS in local semiconductor

manufacturing. Many semiconductor manufacturers are located in the service

area of W1, which would lead to high PFOS concentrations and predominance

of PFOS in the effluents of W1.


87

W6W5W4W3W2W1

22.5%

65% 83.3% 81.1%

57.3%

70.2%

050

100150200250300350400450500550600650

Toto

al P

FCs

mas

s lo

ad (g

/d) Total PFOA load

Total PFOS load

Figure 4.5 Total PFOS and PFOA mass load in the effluent of WWTP

summed up by 5 sampling campaigns. The value in pencentage above the each column indicates PFOA pencentage in total PFOS and PFOA mass load.

4.2.3 PFOS/PFOA concentration in coastal water

PFOS and PFOA were detected in all coastal waters and were in the range of

1.9-8.9 ng/L and 2.4-17.8 ng/L, respectively. This is relatively higher than

those detected in Hongkong coastal waters (PFOS: 0.09-3.1 ng/L, PFOA: 0.7-

5.5) (So et al., 2004), but much lower than those detected in Tokyo Bay

(PFOS: 0.3-57.7 ng/L, PFOA: 1.8-192.0 ng/L) (Yamashita et al., 2005). The

highest PFOS and PFOA concentration as well as total PFCs concentration in

coastal waters were detected at C4 (Figure 4.6). C4 is near the causeway

connecting the Singapore Island and the Malay Peninsula across the Johor

Strait. It suggests that Johor Straits is more heavily contaminated than the

southern and eastern coastal waters. Industries in the northwestern area may be

the significant contamination sources for Johor Straits since W2 discharges its

effluent nearby. The next highest total concentration occurred in C2, which

was collected from the confluence of rivers S2, S3 and S4 flowing through the

commercial areas. PFOS and PFOA concentrations in S2, S3 and S4 were in


88

the range of 4.8-28.3 ng/L and 7.4-36.6 ng/L respectively, and were expected

to be the major contributors for high level of PFOS and PFOA to C2. The

lowest total PFCs concentration was observed at the location of C1, suggesting

that eastern coastal waters is cleaner than other coastal waters in terms of

PFCs contamination. It was observed that PFOA was predominant over PFOS

in all coastal waters, which was indicated by the PFOA percentage. Such an

interesting observation was also reported by other studies (So et al., 2004;

Yamashita et al., 2005). The highest PFOA percentage (69.2%) was observed

at the site of C2, suggesting that commercial activity may lead to high PFOA

composition in waters.

C1 C2 C3 C4

60.6%

66.3%

69.2%

57.3%

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

90.0

100.0

Toto

al P

FOS

and

PFO

A c

once

ntra

tion

(ng/

L) PFOAPFOS

Figure 4.6 Total PFOS and PFOA concentrations in coastal waters summed up

by 5 sampling campaigns. The value in pencentage above the each column indicates PFOA pencentage in total PFOS and PFOA concentration.

As the OW1 is close to the location of C4, the correlation between coastal

water C4 and wastewater W2 was investigated (Figure 4.7). For both PFOS

and PFOA, water samples at C4 could be significantly correlated to effluent of

W2. Furthermore, PFOA percentage in C4 (60.6%) was in agreement with that


89

of W2 (65%). This suggests that discharge of WWTPs may be the major

source of perfluorinated compounds in the environment.

R2 = 0.8764

R2 = 0.5289

0.0

5.0

10.0

15.0

20.0

0.0 50.0 100.0 150.0PFCs concentration in W2 (ng/L)

PFC

s con

cent

ratio

n in

C4

(ng/

L)

PFOS

PFOA

Figure 4.7 Correlations of PFOS and PFOA between

coastal water C4 and wastewater W2.

4.2.4 Seasonal variations in concentration of PFOS/PFOA

For surface waters, significant seasonal difference (p=0.025) was observed for

PFOS between dry and wet seasons, while no significant difference (p=0.616)

was observed for PFOA. It seems that PFOS concentrations noticeably

decreased during wet season, while there was no discernable decrease in

PFOA concentrations in surface waters during wet season. The presence of

PFCs in rainfall indicates rainfall significantly affect their concentrations in

surface water (Prevedouros et al., 2006). PFOS have been observed at a low

concentration (0.59 ng/L) in the precipitation, while significant higher PFOA

concentrations were reported in rainwater. Kallenborn et al. (2004) reported

that PFOA was the predominant PFCs measured in rainwater samples from

Sweden and Finland with the greatest concentrations (11 ng/L and 17 ng/L,

respectively). Scott et al. (2006) also reported relatively high PFOA

concentrations (<0.1-89 ng/L) in the rainwater samples from U.S.A and


90

Canada. Based on the limited data, it seems that PFOS concentration in

rainwater is lower than that of PFOA, which leads to their different seasonal

variations in surface water. Furthermore, runoff could also be the potential

PFCs sources during rainy weather. Rainfall may pick up and carry away

PFCs pollutants (such as oil, fire-fighting foam) when it moves over and

through the ground, which leads to the occurrence of nonpoint source

pollution (NPS) of PFCs. Zushi et al. (2008) also observed that some PFCs

concentrations in a river did not decrease with the increase of river flow rate

during the rainy weather due to the nonpoint source pollution of PFCs. The

predominance of PFOA over PFOS in most of surface waters indicated that

PFOS is used less widely than PFOA in the island. Therefore, runoff contains

lower amount of PFOS, which dilutes the surface waters. Unlike the surface

waters, no discernable seasonal differences were found for both PFOS and

PFOA in coastal waters and wastewaters. NPS of PFCs occurring during wet

season may contribute to the indiscernible variations in PFCs concentrations

from coastal waters and wastewaters. However, it is possible that industrial

activities lead to high concentration variations in the wastewaters, which

override the seasonal differences between dry seasons and wet seasons. Since

wastewaters are discharged into the coastal water, the same effect could

subsequently apply to coastal waters. Similarly, no large-magnitude seasonal

variations in concentrations of PFCs were found among spring, summer, fall

and winter seasons in two municipal sewage treatment plants (Loganathan et

al., 2007).

4.2.5 Correlations between PFOS and PFOA


91

Correlations of PFOS and PFOA in the surface waters, coastal waters and

effluents of WWTPs were examined. It was found that PFOS and PFOA in the

coastal waters could be significantly correlated (R2=0.568, p=0.001), while

weak positive correlations were observed in surface waters (R2=0.197,

p=0.001) and wastewater (R2=0.282, p=0.003), as shown in Figures 4.8, 4.9

and 4.10. This suggests that the possibility of a common contamination source

for these two compounds in coastal waters may be higher than those of surface

waters and wastewaters. Such correlation between those two PFCs could be

attributed to the production and application of related products as well as their

subsequent release into the environment. Similarly, So et al. (2007) observed

strong positive correlations between PFOS and PFOA in coastal waters

collected from Hongkong and South China. In addition, correlations between

PFOS and PFOA in effluents of individual STPs were investigated. It was

found that concentrations of PFOS were significantly correlated to PFOA in

effluent of W4 (R2=0.739, p=0.005) and W5 (R2=0.629, p=0.002), while

medium positive correlations were observed in the effluent of W2 (R2=0.389,

p=0.001) (Figure 4.11). However, PFOS and PFOA concentrations were

weekly correlated in the effluent of W1 (R2=0.151, p=0.001), W3 (R2=0.053,

p=0.001) and W6 (R2=0.025, p=0.001). It seems that the correlations between

PFOS and PFOA are not determined by influent composition of STPs since

significant and weak correlations were observed in both domestic and

industrial STPs. Similarly, Loganathan et al. (2007) found that PFOS were

significantly correlated to PFOA in the wastewaters of one WWTP (R2=0.772),

while weak positive correlations were observed in those of another WWTP

studied (R2=0.084).


92

y = 0.466 x + 18.077R2 = 0.197

0102030405060708090

100

0 20 40 60 80 100PFOS concentration (ng/L)

PFO

A co

nent

ratio

n (n

g/L)

Figure 4.8 Correlations between PFOS and PFOA concentrations

in surface waters.

y = 1.434 x + 1.479R2 = 0.568

0

4

8

12

16

20


PFO

A c

once

ntra

tion

(ng/

L)


in coastal waters.

y = 0.739 x + 40.357R2 = 0.282

0

200

400

600

800

1000

1200

0 100 200 300 400 500 600PFOS concentration (ng/L)

PFO

A c

once

ntra

tion

(ng/

L)


in wastewaters.


93

y = 1.5589x + 11.092R2 = 0.389

y = 12.744x - 99.97R2 = 0.7386

0

50

100

150

200

0 20 40 60 80PFOS concentration (ng/L)

PFO

A c

once

ntra

tion

(ng/

L)

W2

W4

(a) (b)

Figure 4.11 Correlations between PFOS and PFOA concentrations in the effluents of (a) W2 and W4; (b) W5.

4.3 Summary

PFOS and PFOA were detectable in all 138 water samples from reservoirs,

rivers/canals, coastal waters and treated effluents from WWTPs around the

island. Ranges of PFOS concentrations in surface waters, wastewaters and

coastal waters were 2.2-87.3 ng/L, 5.8-532.1 ng/L and 1.9-8.9 ng/L,

respectively, while those of PFOA concentrations were 5.7-91.5 ng/L, 7.9-

1057.1 ng/L, 2.4-17.8 ng/L, respectively. Overall, coastal waters had lower

concentrations of PFOS and PFOA, compared with surface waters and

wastewaters.

In surface waters, the highest total concentrations of PFOS and PFOA were

observed in the western area because of the high levels of industrial activities

in that area. This region was noted to be the most highly contaminated by

PFCs. The next highest total concentration was observed at the location that is

adjacent to the airport, indicating that leakages of perfluorinated surfactants,

such as aqueous fire-fighting foams, gasoline, oil and lubricants (Moody et al.,

2000) from the airport may be potential contamination sources.

y = 2.1864x - 201.09R2 = 0.6286

0

200

400

600

800

1000

1200


PFO

A c

once

ntra

tion

(ng/

L)


94

In wastewaters, the highest total PFCs mass load and PFOA concentration

(1057.1 ng/L) were observed in W5, suggesting discharges of fluorochemical

related factories in the service area of W5 may contain a large amount of

PFOS and PFOA, thus resulting in high concentrations in the WWTPs

effluents. The highest PFOS concentration (532.1 ng/L) was detected in the

effluent of W1 treating mainly domestic and commercial wastewater. This

indicates the presence of potential PFOS contamination sources in its service

area. Compared with surface waters and coastal waters, much higher PFCs

concentrations in wastewaters indicate that discharge of effluents of WWTPs

is an important pathway by which PFCs enter the environment.

In coastal water, the high PFOS and PFOA concentrations at C4 suggest that

Johor Straits is more heavily contaminated than the southern and eastern

coastal waters. The high levels of industrial activities in the western area may

be the significant contamination sources for Johor Straits. In addition, for both

PFOS and PFOA water samples at C4 were significantly correlated with

effluents of W2.

Between dry and wet seasons, significant seasonal variation was observed in

surface waters for PFOS only, while no discernable seasonal differences were

found for both PFOS and PFOA in coastal waters and wastewaters. In addition,

PFOS and PFOA were significantly correlated in the coastal waters, while

weak positive correlations were observed in surface waters and wastewaters. It

suggests that the possibility of a common contamination source for these two


95

compounds in coastal waters is higher than those of surface waters and

wastewaters.

Chapter 5-Development of Post Extraction Clean-up Method For PFOS/PFOA Determination in Wastewater and Sludge Samples

96

CHAPTER 5 DEVELOPMENT OF POST

EXTRACTION CLEAN-UP METHOD FOR

PFOS/PFOA DETERMINATION IN

WASTEWATER AND SLUDGE SAMPLES

5.1 Introduction

High-performance liquid chromatography (HPLC) with triple quadrupole

mass spectrometry in electrospray negative mode is the most promising and


biological matrices (Giesy et al., 2001; Tomy et al., 2004; Becker et al., 2008;

Boulanger et al., 2005 ; Sinclair et al., 2006; Higgins et al., 2005 ; Taniyasu et

al., 2003; Kannan et al., 2004; Moody et al., 2001; Hansen et al., 2001;

Martin et al., 2004b). Analysis can be accomplished by direct injection

(Schultz et al., 2006) or preconcentration on solid phase extraction (SPE)

cartridges, followed by LC/MS/MS analysis (Giesy et al., 2001; Tomy et al.,

2004; Becker et al., 2008; Boulanger et al., 2005 ; Sinclair et al., 2006).

However, analysis of complex environmental matrices such as sediment,

sludge and wastewater by electrospray LC/MS/MS can be significantly

hampered by ionization effects induced by co-eluting components present in

the sample extracts. Several studies have shown that matrix effects resulting

from co-eluting residual matrix components enhance or suppress electrospray

ionization of perfluorinated analytes, leading to considerable inaccuracy

(Boulanger et al., 2005; Sinclair et al., 2006; Higgins et al., 2005). Therefore,


97

it is very important to eliminate matrix effects when the LC/MS/MS method is

used to quantitatively determinate the concentration of perfluorinated

compounds.

Structural analogues of analytes as internal standards are effective techniques,

which show similar behavior in the samples, to compensate for the matrix

effects. Standard addition quantitation, which involves spiking successive

known quantities of a standard into the sample and reanalyzing, is an

acceptable technique to use when matrix effects are unavoidable.

Unfortunately, standard addition quantitation places further demands on

instrument and sample preparation time. Structural analogues of analytes as

internal standards are valuable alternatives, which show similar behavior in

the samples and would compensate for the matrix effects (Petrovic et al., 2005;

Benijts et al., 2004; Matuszewski et al., 2006). An important prerequisite is

that analyte and internal standard have very similar characteristics, and

identical, or at least very close, retention times. Both compounds would

therefore be affected by the co-eluted matrix to the same extent. Structural

analogues of perfluoroalkyl such as PFDoA and PFEES have been used as

internal standards to determine perfluorinated compounds in water and

biological tissue samples with acceptable recovery (Higgins et al., 2005;

Moody et al., 2001; Benijts et al., 2004; Tseng et a., 2006). However,

isotopically labeled internal standard such as 1, 2-13C PFOA is either

expensive or limited by its commercial availability (Martin et al., 2004a).

Although structural analogues of analytes can serve as internal standards, the

potential for ionization enhancement or suppression remains high in complex


98

environmental and biological samples. On the other hand, standard addition

quantitation, which involves spiking successive known quantities of a standard

into the sample and reanalyzing, is a valuable alternative to use when matrix

effects are unavoidable. Without using any internal standard for compensation,

standard addition quantitation could be the most accurate analysis method for

the determination of PFCs in the water (Weremiuk et al., 2006; Furdui et al.,

2008). Unfortunately, standard addition quantitation places further demands

on instrument and sample preparation time. Therefore, post-extraction clean-

up is desired to eliminate matrix interference in complicated environmental

and biological samples (van Leeuwen et al., 2006; Szostek et a., 2004; Simcik

et al., 2005; van de Steene et al., 2006). Powley et al. (2005) applied Envi-carb

(graphitized carbon) and glacial acetic acid to purify the crude extracts of

biological matrices (blood, serum, live and plant tissue). Szostek et al. (2004)

used silica column to clean up fish tissues by eluting the lipids with

dichloromethane, while the target compounds (PFCAs and PFSAs) were

eluted with acetone. For surface water samples, fluorous silica column

chromatography was used to clean up the SPE extracts and remove the

interfering compounds prior to LC/MS detection (Simcik et al., 2005).

However, the above post-extraction methods may not be applicable to

wastewater and sludge samples collected form STPs, in which stronger matrix

effect was observed in comparison with surface water (Boulanger et al., 2005;

Sinclair et al., 2006). Furthermore, the developed method of this study is more

time efficient in comparison with other post-extraction clean-up method

(Simcik et al., 2005) as SPE extract was dried by gentle nitrogen stream only

once during the sample preparation. In addition, although the effect of these


99

post-extraction clean-ups was assessed by the improved recoveries for PFCs,

matrix effect issue has not been sufficiently studied and addressed. The

assessment of matrix effect during development and validation of LC/MS/MS

method is necessary to ensure the precision, selectivity, and sensitivity would

not be compromised. However, limited data is available on the quantitive

estimation of matrix effect and effect of post-extraction clean-up on different

environmental matrices.

The objective of this study was to develop a new post-extraction clean-up

method for the determination of PFOS and PFOA in environmental matrices.

The influence of different environmental matrices on the electrospray

ionization efficiency was assessed by comparing MS response of post-

extraction spiked sample and that of the standard. In addition, the developed

clean-up method was applied to sludge samples to further remove interfering

components after solid phase extraction.

5.2 Results and discussions

5.2.1 Effect of clean-up procedures on matrix effect

Figure 5.1 shows LC-MS-MS chromatograms of PFOS and PFOA in the raw

sewage extracted using the procedure of HLB SPE and HLB+silica. After

silica cartridge clean-up, the intensity of MS increased significantly by 57.6%

for PFOA and 60.4% for PFOS, respectively. It indicates that significant

enhancement in the MS responses of PFOS and PFOA could be due to the

efficient removal of co-eluting interfering compounds by silica cartridge

clean-up. The matrix effect and the extent of ionization suppression for the


100

basic HLB SPE procedure and HLB+silica procedure were evaluated by

comparing MS peak areas of the analyte standards and standards spiked after

extraction of raw sewages (Table 5.1). It can be seen that ME% for both PFOS

and PFOA were below 50%, indicating solid phase extraction alone was

insufficient to remove matrix components. During HLB SPE procedure, 40%

methanol in Milli-Q was applied to wash the cartridge and remove the matrix

components. However, this simple washing cannot effectively remove the

interfering compounds. Without further clean-up, the co-eluting matrix

constituents would lead to strong suppression of electrospray ionization and

result in large inaccuracies, which were observed by other studies on

perfluorinated compounds in environmental matrices (Boulanger et al., 2005;

van Leeuwen et al., 2006; Berger et al., 2005). Moreover, recoveries (RE%:

<50%) were significantly affected by the matrix effect in water samples due to

the ionization suppression even though this HLB SPE procedure can achieve

more than 90% recoveries (98.4% for PFOS and 93.8% for PFOA) for PFCs

spiked Mill-Q water (data not shown).

(a) (b)

Figure 5.1 LC-MS-MS chromatograms of PFOS and PFOA in the raw sewage extracted by (a) HLB SPE and (b) HLB+silica.

1 2 3 4 5 6 7 8 9 10 11Time min

0

500

1000

15009.60

9.865.672.54 11.00.06 5.133.79 6.936.31 8.557.34 8.09 8.724.040.48 10.301.781.46 3.542.21

1 2 3 4 5 6 7 8 9 10 110

500

1000

1500

2000

25009.96

9.89

9.75

8.58 8.952.03 8.010.17 2.71 7.12 10.362.920.46 1.65 11.16.505.60 5.780.88 3.97 4.85 5.483.63 4.28

1 2 3 4 5 6 7 8 9 10 1

1 2 3 4 5 6 7 8 9 10 1Time, min

10

1000

2000

3000

4000

9.95

9.758.63 9.19 10.598.11 8.467.796.930.50 4.862.961.64 5.334.641.76 4.201.27 3.582.36 6.36

10

500

1000

1500

2000

2490 9.61

9.41 15.90 6.571.35 5.332.92 8.664.77 10.434.02 4.622.56 8.127.31 10.606.69 7.392.301.720.76 3.09

Max.2750 cps.

Max.2490 cps.

Max.4410 cps.

PFOA

PFOS

PFOA

PFOS

9.60 min 9.61 min

Max.1580 cps.

9.95 min i

9.96 min iIn

tens

ity(c

ps.)

Time (min) Time (min)


101

In order to further clean up the extracts, silica cartridge was applied to reduce

the co-eluting interfering compounds. As can be seen, ME% (>70%) and RE%

(>67%) increased significantly for both PFOS and PFOA. It suggests that the

partition of analytes (PFOS and PFOA) and interfering compounds are

different between dichloromethane/methanol mixture and silica column.

Consequently, substantial amount of interfering compounds would be retained

by silica cartridge, while PFOS and PFOA would be eluted by mixture of

dichloromethane/methanol (Benijts et al., 2004; Powley et al., 2005). After

silica cartridge clean-up, the coefficient of variation (CV) decreased more than

44% (PFOS) and 34% (PFOA), indicating precision of the analysis increased

due to the reduced matrix effect. However, matrix suppression still existed

(ME%<90%) even though the additional silica cartridge clean-up had been

applied. Therefore, internal standardization was used to compensate for the

remaining matrix suppression.

Table 5.1 Influence of sample clean-up and internal standardization on ME% and RE% (n=5).

Analyte HLB SPE HLB+silica HLB+silica (Internal Standard)

ME% (CV)

RE% (CV)

ME% (CV)

RE% (CV)

ME% (CV)

RE% (CV)

PFOS 48.1

(8.4)

47.1

(9.1)

78.3

(4.1)

74.6

(4.7)

96.3

(2.9)

97.0

(3.2)

PFOA 37.8

(9.5)

36.3

(10.3)

71.2

(6.2)

67.1

(6.6)

93.2

(2.6)

90.3

(4.3)

5.2.2 Effect of internal standards on matrix effects

Internal standards have been shown to be an effective tool to compensate for

the matrix effect (Petrovic et al., 2005; Benijts et al., 2004; Matuszewski et al.,


102

2006). Ideally, isotopically labeled perfluoroalkyl internal standards are

prefered for negating ionization effects because they will have the same

retention times as their natural analogues. Moreover, both analytes and

internal standards are affected by the co-eluting matrix to the same extent.

Therefore, isotopically labeled perfluoroalkyl internal standards offer the best

solution. However, the use of stable isotopes of perfluoroalkyl is cost

prohibitive and commercial availability is often limited. For example,

perfluorooctanoic acid (1, 2-13C) is much expensive, while the required stable

radioactive perfluorinated acid standards (e.g., 14C PFOS) for toxicological

and environmental fate studies are not available. In many cases, structural

analogues such as PFDoA and PFEES which show similar behavior in the

source to compensate for matrix effect, have been used as internal standards to

determine the concentration of perfluorinated compounds in various

environmental samples with acceptable recovery (Higgins et al., 2005; Moody

et al., 2001; Benijts et al., 2004; Tseng et a., 2006). In this study, PFDoA and

PFEES served as internal standards for PFOA and PFOS, respectively. As can

be seen, they enhanced the ME% from 78.3% to 96.3% for PFOS and from

71.2% to 93.2% for PFOA (Table 5.1). The coefficient of variation (CV) was

greatly decreased by applying internal standardization with the value below

5%. In addition, a higher recovery (>90%) was achieved compared to that of

around 70% without internal standardization. The compensation effect of

internal standards was also observed in nine pharmaceuticals analysis, by

which ME% was increased by 20-78%. Also, RE% of those nine

pharmaceuticals were brought close to 100% (van de Steene et al., 2006). It

indicates similar behaviors of internal standards during the analysis


103

compensate for the matrix effect, resulting in minimized matrix effect and

maximized recovery.

5.2.3 Detection of PFOS and PFOA in water and sludge samples

Various types of water samples, including reservoir water, river water, treated

effluent and influent of WWTP B were collected and analyzed to evaluate the

matrix effect on the LC-ESI-MS/MS method (Table 5.2). As can be seen,

ME% was close to 100% for all water samples, indicating no noticeable

matrix effect was observed. Although influent of WWTP B (raw sewage)

showed a slightly lower ME% (more suppression) in comparison with other

three types of water samples, sample origin had limited impact on ME%. It

suggests that HLB together with silica cartridge method can effectively

remove most of the interfering matters which cause matrix suppression on the

detection of PFOS and PFOA by LC/MS/MS. In contrast, without a clean-up

procedure Boulanger et al. (2005) observed significant matrix suppression in

the influent of a wastewater treatment plant, while no signal suppression or

enhancement was observed in surface water and treated effluent. This

confirms that raw sewage has a very strong matrix effect and additional clean-

up procedure is necessary for the detection of PFOS and PFOA by LC/MS/MS.

Table 5.2 Influence of sample origin on ME% and RE% with internal standard (n=3).

Analyte Reservoir Water River Water Treated Effluent Influent of WWTP B

ME% (CV)

RE% (CV)

ME% (CV)

RE% (CV)

ME% (CV)

RE% (CV)

ME% (CV)

RE% (CV)

PFOS 96.2

(2.5)

97.0

(2.8)

98.3

(2.3)

98.0

(3.4)

95.8

(2.3)

95.7

(3.5)

92.8

(3.3)

93.2

(3.1)

PFOA 95.8

(2.6)

92.7

(2.9)

97.5

(2.2)

95.8

(2.8)

95.1

(2.3)

92.8

(2.4)

91.9

(3.6)

89.2

(4.8)


104

The developed silica cartridge clean-up method was applied to various sludge

samples collected from wastewater treatment plant. It was observed that ME%

was slightly lower than those of surface water samples, indicating stronger

matrix effects occurred in sludge samples (Table 5.3). The recovery was in the

range of 85.4-96.6% for PFOS and 81.3-83.8% for PFOA, which were higher

than those reported by Higgins et al. (2005). It suggests that the co-extraction

of lipids and other interfering matters in the extracts of various sludge samples

can be effectively removed by silica cartridge. Simcik et al. (2005) also found

that fluorous silica column could remove the interfering compounds to clean

up the SPE extracts. Moreover, this clean-up method can greatly improve the

detection limit (up to 10 times) as the extracts are concentrated instead of

being diluted after basic solid phase extraction.

Table 5.3 ME% and RE% with internal standard by application of clean-up method on sludge samples (n=3).

Analyte Digested Sludge Activated Sludge Primary Sludge ME% (CV)

RE% (CV)

ME% (CV)

RE% (CV)

ME% (CV)

RE% (CV)

PFOS 92.2

(3.3)

85.4

(4.8)

93.1

(2.8)

96.6

(3.3)

95.4

(2.9)

88.1

(2.1)

PFOA 91.8

(3.0)

81.3

(2.9)

92.5

(3.3)

83.8

(6.2)

93.8

(3.0)

83.6

(4.8)

5.3 Summary

High performance liquid chromatography (HPLC) with tandem mass

spectrometry (MS/MS) is considered the method of choice for the quantitative

determination of perfluorinated compounds in environmental matrices.


105

However, co-eluting matrix components may reduce or enhance the ion

intensity of the analytes and affect the reproducibility and accuracy of the

LC/MS/MS analyses. This study evaluated matrix effect on PFOS and PFOA

in raw sewage by comparing MS responses of standards and those of the same

known amount of analytes in post-extraction spiked samples. Strong matrix

suppression (ME%<49% and RE%<48% for raw sewage) confirmed that

further extracts clean-up after basic solid phase extraction was necessary. A

silica cartridge clean-up method was successfully developed to remove

remaining co-eluting interfering compounds in raw sewage, by which ME%

and RE% were increased to >71% and >67%, respectively. The application of

internal standards further compensated for matrix effect and brought the ME%

and RE% close to 100%, indicating minimal matrix effect was achieved

without significant loss of analytes. Moreover, internal standards improved

reproducibility by significantly decreasing coefficient of variation. The

developed LC/MS/MS detection method was applied to different water

samples and sludge samples. For sludge samples, the recovery was in the

range of 85.4-96.6% for PFOS and 81.3-83.8% for PFOA, respectively.

Results showed that this silica cartridge clean-up method can effectively

remove co-eluting matrix components in various environmental matrices.

Chapter 6-Behavior of PFOS and PFOA in Sewage Treatment Plants

106

CHAPTER 6 BEHAVIOR OF PFOS AND

PFOA IN SEWAGE TREATMENT PLANTS

6.1 Introduction

Effluent from wastewater treatment plants is one of the important routes for

the introduction of certain organic contaminants into aquatic ecosystems.

Several studies have described effluents of sewage treatment plants (STPs) as

an important source for metals, chlorinated organic compounds to aquatic

environments (Irvine et al., 1993; Loganathan et al., 1997). Perfluorinated

compounds (PFCs), a class of emerging environmental pollutants, have been

widely used for the last 50 years in industrial and commercial applications,

such as coatings, shampoos, electroplating, fire-fighting foams, stain repellants

for furniture and carpets. Widespread usage of these compounds has been

attributed to their contamination in wastewaters. Occurrence of two PFCs,

perfluorooctane sulfonate (PFOS, C8F17SO3-) and perfluorooctanoic acid

(PFOA, C7F15COO-) has been reported in STPs (Kissa, 2001). Various

contamination levels were observed in the influent and effluent of municipal

STPs in Iowa City (Boulanger et al., 2005), in Kentucky and Georgia,

respectively (Loganathan et al., 2007), in 10 national wide municipal STPs in

U.S.A (Schultz et al., 2006a) and in the effluent of 6 U.S.A cities (Sinclair and

Kannan, 2006). The discharge of municipal wastewater effluent is therefore

one of the major routes for introducing PFOS and PFOA that are used in

domestic, commercial and industrial settings into aquatic environment.

Recently, there has been increasing concern about the fate and behavior of


107

PFOS and PFOA in sewage treatment plant due to their biotic and abiotic

persistence and chronic toxicity (Prevedouros et al., 2006; Hof et al., 2004;

3M, 2003). Sinclair and Kannan (2006) observed that mass flow of PFOS and

PFOA in aqueous phase increased significantly after secondary treatment in a

sewage treatment plant with industrial influence, while no increase in mass

flow of PFOA was found in another sewage treatment plant with no industrial

influence. Furthermore, Schultz et al. (2006b) identified the fate and behavior

of these two compounds in both aqueous phase and solid phase (sludge)

during each step of municipal wastewater treatment plant. It was observed that

mass flow of PFOS or PFOA either increased or remained consistent,

indicating conventional activated sludge process can not effectively remove

these compounds. However, these investigations were conducted at different

STPs with different influents. Different influent of STP would significantly

affect the behavior pattern of PFOS or PFOA since their precursors in the

influent could be biodegraded to PFOS or PFOA in the activated sludge

treatment processes. Therefore, it is desirable to investigate behavior of PFCs

in various activated sludge treatment processes which receive the same raw

sewage.

Even though PFOS or PFOA is not biodegradable, their precursors could be

biodegraded and thus affect the mass flow of PFOS or PFOA in STP. For

example, it was found that secondary treatment by activated sludge

significantly increased the mass flows of PFOS and PFOA, probably resulting

from biodegradation of precursor compounds such as fluorotelomer alcohols

(Sinclair and Kannan, 2006). Currently, conflicting information is obtained


108

about the behavior patterns of PFCs in STPs. Some studies reported increase

in mass flow of PFCs in biological processes, while other studies observed

unchanged PFCs’ mass flow in the activated sludge treatment. Except for

effect of influent of STP, sludge retention time (SRT), a commonly used

parameter for sewage treatment plant design and operation, could also be an

important factor affecting the behavior of PFOS and PFOA in STPs. Clara et

al. (2005) found that the degradation of the micropollutants, such as endocrine

disrupting compounds and pharmaceuticals, was dependent on the SRT in the

activated sludge process since the SRT determines the enrichment of the

microorganism that is able to degrade the micropollutants. Therefore, behavior

pattern of perfluorinated compounds may be different in the conventional

activated sludge process operated with different SRT. However, no data is

available about the effect of SRT on the behavior pattern of PFOS and PFOA

in the activated sludge process.

The objective of this part of study was to compare the behavior of PFOS and

PFOA in full-scale conventional activated sludge processes and membrane

biological reactor, as well as in an activated sludge process operated with a

short SRT. This is the first study to investigate the effect of SRT on the

behavior of these two compounds in the activated sludge process. Furthermore,

seasonal variation in the concentrations of PFOS and PFOA in sewage

treatment plants was studied. In order to achieve these, aqueous and solid

samples were taken from each treatment unit of STPs, A and B.

6.2 Results and Discussion


109

6.2.1 PFOS/PFOA in wastewater

Sampling strategy can affect the concentrations of the analytes measured in

sewage treatment plants. The 24-h composite sample is appropriate to

represent average concentrations over 24 h in the wastewater streams of STPs.

Grab samples collected in this study, which could have been collected at high

or low flow period, may increase the variation in concentrations. Therefore,

care must be taken when concentrations of PFOS and PFOA measured were

compared.

1 2 3 4

16151312 14109 11

5 6 7 8PEAT

SE

Inf 0.0

100.0

200.0

300.0

400.0

500.0

600.0

PFO

S co

ncen

tratio

ns in

was

tew

ater

(ng/

L)

1.Dec 06-CAS1 2.Oct 06-CAS1 3.Mar 07-CAS1 4.Sep 07-CAS15.Dec 07-CAS1 6.Oct 06-LTM 7.Dec 06-LTM 8.Mar 07-LTM9.Mar 07-MBR 10.Sep 07-MBR 11.Dec 07-MBR 12.Oct 06-CAS213.Dec 06-CAS2 14.Mar 07-CAS2 15.Sep 07-CAS2 16.Dec 07-CAS2

Figure 6.1 PFOS concentrations in wastewater of STP A (CAS1, LTM and MBR) and STP B (CAS2). Inf: influent; AT: aeration tank; PE: primary clarifier effluent; SE: secondary clarifier effluent.

The measured PFOS and PFOA concentrations in wastewater samples from

STPs A and B are shown in Figures 6.1 and 6.2. PFOS and PFOA were


110

detected in all wastewater samples collected from STP A and B. PFOS was

observed at 5.3 - 29.8 ng/L in STP A, which are comparable to those measured

in the effluent of 4 STPs receiving domestic and commercial sewage (Sinclair

et al., 2006). However, much higher concentration of PFOS (48.1 - 560.9 ng/L)

was detected in STP B receiving 60% industrial wastewater. These measured

PFOS concentrations are also much higher than those in the influents and

effluents of 10 STPs mainly receiving domestic and commercial sewage in

USA (Schultz et al., 2006a), but much lower than those in the effluents of

Decatur which receives influent from fluorochemical manufacture or industry.

Nevertheless, the comparable concentration was observed in a wastewater

treatment plant in Cleveland (3M, 2001), which has no known fluorochemical

sources. It suggests that industrial sewage can contain a large amount of PFOS

in comparison with domestic and commercial sewage even though there is no

known source of fluorochemical exposure.


111

16151312 14109 115 6 7 8

1 2 3 4PEAT

SE

Inf 0.0

200.0

400.0

600.0

800.0

1000.0

1200.0

PFO

A c

once

ntra

tions

in w

aste

wat

er (n

g/L)

1.Dec 06-CAS1 2.Oct 06-CAS1 3.Mar 07-CAS1 4.Sep 07-CAS15.Dec 07-CAS1 6.Oct 06-LTM 7.Dec 06-LTM 8.Mar 07-LTM9.Mar 07-MBR 10.Sep 07-MBR 11.Dec 07-MBR 12.Oct 06-CAS213.Dec 06-CAS2 14.Mar 07-CAS2 15.Sep 07-CAS2 16.Dec 07-CAS2

Figure 6.2 PFOA concentrations in wastewater of STP A (CAS1, LTM and MBR) and STP B (CAS2). Inf: influent; AT: aeration tank; PE: primary clarifier effluent; SE: secondary clarifier effluent.

PFOA was the predominant contaminant in STP A, which was measured at

11.2 - 138.7 ng/L. Slight lower and comparable PFOA concentration was

reported in the influents and effluents of 10 STPs in USA (Schultz et al.,

2006a). However, Sinclair et al. (2006) observed much higher concentration in

4 STPs receiving domestic and commercial sewage. This suggests that

commercial sewage could be a significant source of PFOA, which includes a

wide range of sources (hospitals, shopping malls and so on) and provides more

variable amount of PFOA. In addition, the predominance of PFOA over other

perfluorinated compounds was observed in other STPs (Sinclair et al., 2006;


112

Loganathan et al., 2007). In STP B, PFOA concentration was detected in the

range of 31.8 – 1,057.1 ng/L, which is much higher than those of STP A and

those in a sewage treatment plant with similarly 60% industrial influent

(Sinclair et al., 2006). It suggests the effect of industrial influent on PFOA

concentration is dependent on the composition of the sewage that enters the

sewage treatment plants. Moreover, in this study, higher variation in

concentrations of both PFOS and PFOA was observed in STP B than those in

STP A, indicating industrial influent can result in high concentration variation.

6.2.2 Seasonal variation

In STP A, PFOS concentration in influent of dry season showed statistically

significant difference from the wet season (ｐ=0.003), while PFOA had no

such significant difference (ｐ=0.157) (Figure 6.3). It seems that PFOS

concentrations noticeably decreased during wet season, while there was no

discernable decrease in PFOA concentrations in surface waters. The presence

of PFCs in rainfall indicates rainfall significantly affect their concentrations in

surface water. PFOS have been observed at a low concentration (0.59 ng/L) in

the precipitation, while significant higher PFOA concentrations were reported

in rainwater. Kallenborn et al. (2004) reported that PFOA was the

predominant PFCs measured in rainwater samples from Sweden Finland with

the greatest concentrations (11 ng/L and 17 ng/L, respectively). Scott et al.

(2006) also reported relatively high PFOA concentrations (<0.1-89 ng/L) in

the rainwater samples from U.S.A and Canada. Based on the limited data, it

seems that PFOS concentration in rainwater is lower than that of PFOA, which

leads to their different seasonal variations in surface water. Furthermore,


113

0

5

10

15

20

25

30

35

1 2 3 4 5 6 7 8 9 10 11

PFOS Concentration (ng/L)

(a) (b)

Figure 6.3 Seasonal variations in influent concentrations of (a) PFOS and (b) PFOA in STPA. 1: Oct 06 (CAS1), 2: Mar 07 (CAS1), 3: Sep 07 (CAS1), 4:

Mar 07 (MBR), 5: Sep 07 (MBR), 6: Oct 06 (LTM), 7: Mar 07 (LTM), 8: Dec 06 (CAS1), 9: Dec 06 (LTM), 10: Dec 07 (CAS1), 11: Dec 07 (MBR)

runoff could also be the potential PFCs sources during rainy weather. Rainfall

may pick up and carry away PFCs pollutants (such as oil, fire-fighting foam)

when it moves over and through the ground, which leads to the occurrence of

nonpoint source pollution (NPS) of PFCs. Zushi et al. (2008) observed that

some PFCs concentrations in a river did not decrease with the increase of river

flow rate during the rainy weather possibly due to the NPS of PFCs. As a

result, the decreased PFOS concentrations in surface water may result in their

decrease in wastewater correspondently after surface water is treated by water

treatment plants and then subsequently utilized by various consumers.

However, in STP B no significant difference between dry season and wet

season for both PFOS (ｐ=0.520) and PFOA (ｐ=0.274) was observed despite

slightly lower concentration was observed in wet season (Figure 6.4). It is

likely that high concentration of industrial influent override the effect of

dilution by rainstorm. In comparison, no large-magnitude seasonal variation in

concentrations of perfluorinated compounds was found among spring, summer,

0

20

40

60

80

100

120

1 2 3 4 5 6 7 8 9 10 11

PFOA Concentration (ng/L)

Dry season Wet

season Wet season

Dry season


114

fall and winter seasons in two municipal sewage treatment plants (Loganathan

et al., 2007).

Dec07

Dec06

Sep07

Mar07

Oct06

0

50100

150200

250

300350

400450

500

PFOS

Con

cent

rati

on (

ng/L

)

(a) (b)

Figure 6.4 Seasonal variations in influent concentrations of (a) PFOS and (b) PFOA in STP B.

6.2.3 Mass flow in aqueous sample during treatment

The average mass flow was calculated by multiplying average PFOS and

PFOA concentrations in aqueous and solid phase by the daily average flow of

each treatment unit (Table 6.1). Total solid waste is the daily mass of PFOS or

PFOA associated with primary sludge and waste activated sludge. Related

information on the wastewater and solid stream was obtained from individual

STPs. It is worthy to note that sampling strategy can affect the concentrations

of the analytes measured in sewage treatment plants. Specially, grab sample,

which could have been collected at high or low flow period, may increase the

variation in concentration. As the concentration was based on grab samples, it

would result in additional variation in mass flow besides the error of

measurement. Therefore, only change of more than 30% in mass flow would

be taken into consideration in this study (Loganathan et al., 2007).

Dec07

Dec06

Sep07

Mar07

Oct06

0

100

200

300

400

500

600

700

800

900

PFOA

Con

cent

rati

on (

ng/L

)

Wet season

Dry season Wet

season

Dry season


115

CAS1, MBR and LTM, which are different treatment processes, receive same

raw sewage, while CAS2 receives different raw sewage. No significant change

in mass flow of PFOS (-24.5%-16.0%) was observed in CAS1, MBR and

LTM. It is known that PFOS or PFOA can not be biodegraded by activated

sludge process (Lange, 2002). A reduction in mass flow of PFOS or PFOA is

neither expected nor observed (Schultz et al., 2006b; Sinclair et al., 2006)

since biodegradation of precursors such as fluorotelomer alcohols (FTOHs),

perfluoroalkyl phosphates (PAPS), or fluorotelomer sulfonates (FTSs) during

activated sludge treatment process are likely sources of increase of PFOS and

PFOA. Specially, it is known that 2-(N-ethyl-perfluorooctanesulfonamido)

ethanol (N-EtFOSE alcohol) and 2-(N-ethyl perfluorooctane sulfonamido)

acetic acid (N-EtFOSAA) can be biotransformed to PFOS during activated

sludge treatment (Boulanger et al., 2005; Lange, 2000 and 2002). Our result

suggests that either raw sewage of STP A did not introduce the precursors of

PFOS or no significant biotransformation occurred during these processes.

However, significant increase in mass flow of PFOS (mean 94.6%) was

observed in CAS2, indicating biodegradation of precursors may occur during

the secondary treatment. As CAS1 is running with the similar operational

parameters (e.g. SRT and HRT) compared to CAS2, the results suggest that no

precursors of PFOS be likely contained in the raw sewage of STP A.

Table 6.1 Mass flow (mg/d) of PFCs in influent, effluent and solid waste in CAS1, CAS2, LTM and MBR.

Site Date

PFOS PFOA

influent (aqueous)

influent (particu-

late) Effluent

Solid waste (total)

Mass change

Mass change

(%)

influent (aqueous)

influent (particu-

late)

Effluent (aqueous)

Solid waste (total)

Mass change

Mass change

(%)

CAS 1 Oct 06 3722 487 3372 1014 -177 -4.2% 4355 ND 6496 292 -2433 -55.9% Dec 06 3782 348 3540 534 57 1.4% 6782 ND 7565 403 -1186 -17.5% Mar 07 3396 207 3009 607 -13 -0.4% 6584 ND 9465 501 -3382 -51.4%

MBR Mar 07 321 16 261 104 -29 -8.6% 451 ND 739 58 -345 -76.6%

LTM Oct 06 1076 130 1125 376 -295 -24.5% 1457 ND 1221 164 72 4.9% Dec 06 1002 100 729 240 133 12.1% 1950 ND 2346 165 -561 -28.8% Mar 07 1361 165 1022 260 244 16.0% 1378 ND 1544 180 -346 -25.1%

CAS 2 Oct 06 39834 3862 66318 7870 -30492 -69.8% 11583 375 15867 800 -4709 -39.4% Dec 06 20550 3395 34894 6200 -17150 -71.6% 11607 646 28543 1351 -17641 -144.0% Mar 07 16072 2214 32308 12028 -26050 -142.5% 39360 1784 47191 1213 -7260 -17.6%

ND: not detectable; Mass change=Influent (aqueous)+Influent (particulate)-Effluent-Solid waste (total); Mass change (%)=Mass change/[Influent (aqueous)+Influent (particulate)]

116

A C

hapter 6-Behavior of PFO

S and PFOA

in Sewage Treatm

ent Plants


117

Mean mass flow of PFOA increased by 41.6% (17.5%-55.9%) and 76.6% in

CAS1 and MBR, respectively, while PFOA mass flow remained unchanged

after the treatment of LTM with a SRT of 3.5 d. During activated sludge

treatment some precursors, especially 8:2 FTOH, have been shown to

biotransform into PFOA (Dinglasan et al., 2004; Wang et al., 2005). This

suggests that no noticeable biodegradation of PFOA precursors can occur in

LTM though their presence in the raw sewage has been demonstrated by mass

increase in CAS1 and MBR. Similarly, Clara et al. (2005) found that no

biodegradation of micropollutants, such as endocrine disruptors compounds

(EDCs) or pharmaceuticals could occur when the activated sludge treatment

system (CAS or MBR) was operated with a SRT, which was lower than a

critical SRT (e.g. approx. 10 days for estrogens, 17b-estradiole, estrone and

bisphenol-A). Only at a higher SRT which is more than the critical one, the

microorganisms that biodegrade certain micropollutants are able to be

detained and enriched in the system. It seems that the SRT of LTM is lower

than the critical one, resulting in no biodegradation of precursors. Furthermore,

mass flow of PFOA and PFOS increased 17.6-144.0% and 69.8-142.5% after

the secondary treatment of CAS2, respectively. It suggests that the precursors

of PFOS and PFOA be biodegraded at a SRT of ~12 days, which may be

higher than the critical SRT. Our results confirm that change in mass flow of

PFOS and PFOA may be determined by both the presence of precursors and

operating SRT of the activated sludge system. SRT could thus be an important

operational parameter that affects the behavior pattern of PFOS and PFOA.


118

PFOS and PFOA mass change after the treatment of primary clarifier in STP

A and B are shown in Figure 6.5. As can be seen, mass flow change was in the

range of -27.3%-6.7% for PFOS and -35.7%-12.5% for PFOA, respectively. It

suggests that there is no discernable mass change after the treatment of

primary clarifier for both PFOS and PFOA. In addition, their mass flow in the

inflow and outflow of primary clarifier were equivalent at the 95% CI

(confidential interval). It seems primary clarifier has no noticeable effect on

the mass flow of PFOS and PFOA. Similarly, Schultz et al (2006b) observed

that only 10% (PFOS) and 0.1% (PFOA) reduction in mass flow occurred due

to their sorption onto primary sludge.

Dec07

Dec06

Sep07

Mar07

Oct06

-40%

-30%

-20%

-10%

0%

10%

20%

30%

Change of mass flow (%)

PFOSPFOA

(a) (b) Figuure 6.5 Change of mass flow after primary treatment in (a) STP A and (b) STP B

6.2.4 PFOS/PFOA in sludge

PFOS and PFOA were detected in all sludge samples except for one sample

from STP A which was below LOQ of PFOA (Figure 6.6 and 6.7). PFOS was

observed at 13.1 - 46.0 ng/g dw in STP A, while 3.2 - 53.6 fold higher

concentration (145.1 - 702.2 ng/g dw) was observed in STP B. Similarly,

while higher PFOA concentration (18.0 - 69.0 ng/g dw) in STP B was

detected.lower PFOA concentration (<5.0 - 44.2 ng/g dw) was observed in

Dec07

Dec06

Sep07

Mar07

Oct06

-30%

-20%

-10%

0%

10%

20%

30%

Change of mass flow (%) PFOS

PFOA


119

6 7 8 9 10

1 2 3 4 5

ASPS

SS

Inf

DS

0

100

200

300

400

500

600

700

800

PFO

S co

ncen

traio

ns in

slud

ge (n

g/g

dw)

1.Dec 06-CAS1 2.Oct 06-CAS1 3.Mar 07-CAS1 4.Oct 06-LTM (AS)5.Dec 06-LTM (AS) 6.Mar 07-LTM (AS) 7.Mar 07-MBR (AS) 8.Oct 06-CAS29. Dec 06-CAS2 10. Mar 07-CAS2

Figure 6.6 PFOS concentrations in sludge samples from STP A and STP B. Inf: influent particulate; PS: primary sludge; AS: activated sludge; SS: secondary clarifier sludge; DS: digester sludge.

STP A. In comparison, Higgins et al. (2005) reported PFOS and PFOA

concentrations were in the range of 14.4 – 2,610.0 and n.d - 29.4 ng/g dw in

the sludge samples of 8 STPs, respectively. In addition, our results suggest

that high concentration in wastewater lead to high concentration in sludge

which is due to the partition between aqueous and solid phases. Nevertheless,

due to its higher partition coefficient in comparison with PFOA (Higgins et al.,

2006), PFOS was dominant in sludge samples of both STPs A and B even

though PFOA was dominant in aqueous samples of STP A.


120

In terms of PFOS or PFOA concentrations, there was no noticeable difference

among activated sludge samples collected from LTM, MBR and CAS1. It

seems their concentrations in sludge are more relevant to the aqueous

concentration than sludge characteristics, which are affected by the SRT (Liao

et al., 2001; Ng et al., 2005).

6 7 8 9 10

1 2 3 4 5

ASPS

SS

Inf

DS

0

10

20

30

40

50

60

70

PFO

A c

once

ntra

tions

in sl

udge

(ng/

g dw

) 1.Dec 06-CAS1 2.Oct 06-CAS1 3.Mar 07-CAS1 4.Oct 06-LTM (AS)5.Dec 06-LTM (AS) 6.Mar 07-LTM (AS) 7.Mar 07-MBR (AS) 8.Oct 06-CAS29. Dec 06-CAS2 10. Mar 07-CAS2

Figure 6.7 PFOA concentrations in sludge samples from STP A and STP B. Inf: influent particulate; PS: primary sludge; AS: activated sludge; SS: secondary clarifier sludge; DS: digester sludge.

The partition coefficient Kd for primary sludge and activated are estimated

based on the data obtained by dividing PFCs concentrations in primary sludge

or secondary sludge by their aqueous concentration in primary effluent or

secondary effluent (Table 6.2). Kd value of PFOS was in the range of 894-

2,237 L/kg (primary sludge) and 720-2,324 L/kg (activated sludge), while

significant lower Kd value of PFOA was observed at 188-597 L/kg (primary


121

sludge) and 201-513 L/kg (activated sludge), respectively. The mean Kd value

of PFOS was more than 3 times higher than that of PFOA, indicating much

more amount of PFOS adsorbed onto sludge as compared to PFOA. High

variations in Kd value for PFOS and PFOA may be caused by different

retention time of aqueous and solid streams in primary and secondary

clarifiers. In addition, it seems that there is no significant difference between

Kd values in primary sludge and activated sludge. Based on the data on

organic carbon ƒoc

Table 6.2 Calculated partition coefficient Kd

, calculated activated sludge log Koc values (partition

coefficient for the compound onto a hypothetical pure organic carbon) were

2.98-3.49 for PFOS and 2.43-2.83 for PFOA, respectively. In contrast, lower

organic carbon-normalized log Koc values were reported by 3M (2000) (2.57-

3.1 for PFOS) and DuPont (2003) (1.9-2.17 for PFOA) determined on

sediments.

in primary sludge and activated sludge.

Compound Sludge type Kd (L/kg ) Range Mean (±SD)

PFOS Primary sludge 894-2237 1408 (±481) Activated sludge 720-2324 1645 (±511)

PFOA Primary sludge 188-597 405 (±149) Activated sludge 201-513 368 (±106)

6.3 Summary

PFOS and PFOA were detected in all aqueous samples collected from STP A

and STP B, ranging from 5.3 - 560.9 ng/L and 11.2 - 1057.1 ng/L, respectively.

In sludge of STPs A and B, PFOS and PFOA concentrations were in the range

of 13.1 – 702.2 ng/g dw and <5.0 - 69.0 ng/g dw, respectively. Due to

industrial influence, PFOS and PFOA were observed at higher concentration

in aqueous and sludge samples in STP B than that of STP A.


122

Significant increase in mass flow of PFOS (mean 94.6%) was observed in

CAS2, while it remained consistent after secondary treatment in CAS1. This is

likely due to no occurrence of PFOS precursors in the raw sewage of STP A.

Mean mass flow of PFOA increased 41.6% in CAS1, 67.0% in CAS2 and

76.6% in MBR, while it remained unchanged after the treatment of LTM.

Different behavior pattern of these two compounds were found in LTM, an

activated sludge process operated at a relatively short SRT. The findings

suggest that change in mass flow of PFOS and PFOA in secondary sludge

treatment may be determined by the absence/presence of precursors and

operating SRT of the activated sludge system.

Compared with STP A, higher concentrations of PFOS and PFOA were

detected in STP B receiving 60% industrial wastewater. It suggests that

industrial sewage contain a large amount of PFOS and PFOA in comparison

with domestic sewage even though there was no known source of

fluorochemical exposure. Furthermore, industrial influent caused little

seasonal variation in concentrations of PFOS and PFOA. Between dry and wet

seasons, seasonal variation of PFOS was observed in STP A, while PFOA had

no significant difference in both STP A and STP B. PFOS concentration in

rainwater observed by other studies was lower than that of PFOA, which could

lead to their different seasonal variations in surface water. It is also likely that

NPS of PFCs occurred in wet season, which would contribute to consistent

PFOA concentrations in surface waters and subsequently resulted in


123

indiscernible variation in PFOA concentrations in the wastewaters between

dry and wet seasons.

Chapter 7-PFOS/PFOA Removal by Hybrid PAC-MBR Process

124

CHAPTER 7 PFOS/PFOA REMOVAL BY

HYBRID PAC-MBR PROCESS

7.1 Introduction

The discharge of municipal wastewater effluent is one of the major routes for

introducing PFOS and PFOA that are used in domestic, commercial and

industrial settings into aquatic environment. They were detected in the influent

and effluent of municipal WWTPs in Iowa City (Boulanger et al., 2005), in 10

national wide municipal WWTPs in U.S.A (Schultz et al., 2006a) and in the

effluent of 6 U.S.A cities (Sinclair et al., 2006). High PFCs concentrations

were observed in the effluent of fluorochemical manufacture or related

industries (3M, 2001). For an example, Tang et al. (2006) reported PFOS

concentration of 1650 mg/L in the effluent of semiconductor manufacturing.

They were considered stable and persistent in environment without natural

degradations (Prevedouros et al., 2006; 3M, 2003). Also, Lange (2002)

observed that PFOS and PFOA were not degradable by activated sludge.

Studies on fate and behavior of these pollutants in WWTPs implied that they

can not be effectively removed by biological treatment process (Sinclair et al.,

2006; Schultz et al., 2006b).

Although there is no maximum allowable concentration of PFCs in the

discharge of STPs, PFOS and PFOA, candidates for persistent organic

pollutants (POPs), are reported to have adverse effect on the human health.

Since PFOS and PFOA can not be effectively removed by conventional STPs


125

and drinking water treatment plants, it is urgent to develop a new technology

to remove these compounds effectively at low cost for the wastewater

treatment. Various physico-chemical treatment processes including adsorption

(Ochoa-Herrera et al., 2008), sonochemical treatment (Moriwaki et al., 2005),

reduction with zero-valent iron in subcritical water (Hori et al., 2006) and

membrane filtration (Tang et al., 2006) have been studied to remove these

compounds. Activated carbon adsorption is one of the most promising

methods to remove PFCs in aqueous stream due to the effectiveness and low

cost. It was reported that PFCs were effectively removed by adsorption onto

the activated carbon at high and low equilibrium concentrations (Ochoa-

Herrera et al., 2008; Qiu et al., 2006). Ochoa-Herrera et al. (2008) reported

that PFOS could be effectively removed by granular activated carbon (GAC)

and Freundlich isotherm was applicable at high and low equilibrium

concentrations. In contrast, Yu et al. (2009) studied the feasibility of using

powder activated carbon (PAC), granular activated carbon (GAC) and anion-

exchange resin (AI400) to remove PFOS and PFOA from water. It was

observed that adsorption isotherms of PFOS and PFOA fitted Langmuir

isotherms better than Freundlich isotherm. Qiu et al. (2006) also reported that

GAC was able to effectively remove PFOS and PFOA. In 4 hours 93% PFOS

and 99% PFOA in pure water at ppb level were adsorbed onto GAC. Based on

the information available, it seems that PFCs compounds can be effectively

removed by adsorption onto the activated carbon in water solution without the

presence of dissolved organic matters (DOMs).


126

The hybrid PAC-MBR technology integrates adsorption and biodegradation of

organic matter with membrane filtration in one unit, which has been proved to

be a simple and highly efficient way to remove compounds in wastewater. In

particular, PAC addition increases the removal of organic matters with low

molecular weight by adsorption; it also serves as a supporting medium for

attached bacterial growth (Kim et al., 1998). Even though MBR may not be

able to significantly remove PFOS and PFOA due to similar biodegradation

and adsorption behavior in activated sludge system, combination of MBR and

PAC technologies could effectively remove these compounds while adsorption

onto PAC occurs. However, there is no data available on the removal of PFCs

in the hybrid PAC-MBR process till now.

Effluent from biological wastewater treatment contains complex and

heterogeneous soluble organic matter, which is so called effluent organic

matter (EfOM). EfOM is highly heterogengeneous, containing molecular of

various molecular weight ranging from the simple compounds such as acetic

acid to very complex polymers. The composition of EfOM is a combination of

those of natural organic matter (NOM), soluble microbial products (SMPs),

and trace harmful chemicals. Most of the NOM originates from drinking water,

which is one of major components in wastewater, while SMPs come from

biological treatment with the wastewater treatment plant (WWTP) and non-

biodegradable organic matter (Shon et al., 2006). The SMP are organic

compounds that are biologically derived from substrate metabolism during

biomass growth (utilization associated products, UAP) and that are released

from cell lysis during biomass decay (biomass associated products, BAP)


127

(Grady et al., 1999; Barker et al., 1999). SMP has been found to constitute the

majority of soluble organic matter in wastewater effluent from biological

treatment system (Barker et al., 1999). It is known that natural organic matter

(NOM) adversely affected the adsorption of micropollutants onto, such as

pesticides, onto the activated carbon (Newcombe et al., 2002, Quinlivan et al.,

2005; Matsui et al., 2003). When background NOM is present during activated

carbon treatment of water containing micropollutants, a competition will occur

between the target compound and the compounds composing NOM. As a

consequence the adsorption of micropollutant will usually be reduced,

sometimes dramatically (Newcombe et al., 2002; Matsui et al., 2003). The

direct competition for the adsorption sites was found to be the most likely

competition between EfOM and target micropollutants (Newcomber et al.

2002; Kilduff et al. 1998; Matsui et al., 2003). However, limited data is

available on the effect of EfOM on the PFCs adsorption to the activated

carbon.

The objective of this chapter was to investigate the effect of EfOM on the

adsorption of micropollutants PFOS and PFOA onto the powdered activated

carbon. The EfOM was characterized and fractionated to study the adsorption

competition between PFCs and EfOM. Our results would contribute to better

understanding of competitive effects caused by presence of EfOM. In addition,

the performance and removal efficiencies of PFCs were investigated in a

hybrid PAC-MBR process which operated with different PAC dosage and

SRTs. The effect of SRT and PAC dosage was also studied for the better


128

understanding of removal mechanism of PFCs in the hybrid PAC-MBR

process.

7.2 Results and Discussion

7.2.1 Adsorption study on PAC and activated sludge

7.2.1.1 Characterization of EfOM

Four fractions of nominal molecular weights were obtained by ultrafiltration

of EfOM solution (Table 7.1). It can be seen that supernatant of MBR had a

broad spectrum of molecular weight. The fractions of smallest molecular

weight (<1 kDa) accounted for 27.8%, while the fraction of largest molecular

weight (>30 kDa) was the largest fraction, accounting for 31.2%. Other

fractions, 1-10 kDa and 10-30 kDa accounted for 29.9% and 12.9%,

respectively. Ultrafiltration is a size exclusion method of fractionation, some

factors, such as molecular structure and charge as well as solution chemistry

(pH and ionic strength) strongly affected the actual molecular weight of the

fractions (Kuchler et al., 1994). Pelekani et al. (1999) observed that

ultrafiltration of fractionation overestimated the actual molecular weight

distributions of NOM based on the membrane nominal molecular weight

cutoff values.

Table 7.1 Characteristics of EfOM solution obtained from the lab scale MBR (n=5).

Fraction <1 kDa 1–10 kDa

10-30 kDa

>30 kDa EfOM

% EfOM 27.8% 29.9% 12.9% 31.5% 100% SD 4.3 3.3 1.7 4.8 -

SD: standard deviation.

7.2.1.2 PFOS and PFOA adsorption onto PAC


129

The effects of EfOM on the PFCs adsorption onto the activated carbon were

investigated by conducting single solute isotherm experiments for PFOS and

PFOA in the absence and presence of EfOM using PAC. Figure 7.1 shows the

adsorption isotherms of PFOS and PFOA in EfOM free and EfOM raw. It can

be seen that adsorption capacity in EfOM free (Mill-Q water) was much

higher than that of EfOM raw for both PFOS and PFOA, suggesting EfOM

significantly decreased the adsorption capacity of PFCs onto PAC. The

isotherm experiment data were fitted to Langmuir and Freundlich models and

constants determined were listed in Table 7.2. Adsorption of PFCs to PAC

fitted the Freundlich model better (r2>98%) than Langmuir model (r2<90%) in

the absence and presence of EfOM within the studied concentration range

(0.1-500 µg/L), indicating PFCs were adsorbed to the heterogeneous sites with

different affinities for the solutes. Moreover, the adsorption capacity of PAC

tended to increase as the equilibrium concentration increased, which suggests

the possibility of more than just one monomolecular layer of coverage. The

application of Freundlich model is also extensively used to describe the

adsorption of organic solutes, such as Polychlorinated biphenyls (PCBs) and

polyacromatic hydrocarbons (PAHs) onto the activated carbon (Ahn et al.,

2005; Newcombe et al., 2002, Matsui et al., 2003).

The adsorption isotherms of PFOS and PFOA obtained in this study showed a

consistency within the concentration range of 0.1-500 µg/L, while Ochoa-

Herrera et al. (2008) observed significant difference in adsorption capacity of

activated carbon between high concentration range (15-150 mg/L) and low

concentration range (50-500 µg/L). Due to the different adsorbents used and


130

experiment conditions, the results may not be comparable. In addition, the

PFOS adsorption capacity was higher than that of PFOA, which is in

agreement with the data of other study (Ochoa-Herrera et al, 2008). This

observation is expected since PFOA has higher solubility and degree of

attraction by water molecules could be higher, which tends to prevent species

being bound by the carbon surface (Cooney, 1998).

According to Freundlich model, adsorption capacity of PFOS (KF=17.55) and

PFOA (KF=10.03) onto the PAC in pure water is more than one order of

magnitude higher than that of EfOM raw (KF=0.66 for PFOS, KF=0.20 for

PFOA), indicating the presence of EfOM greatly reduced the adsorption

capacity of PAC. Similarly, much lower adsorption capacity (e.g. Freundlich

constant KF) was observed for the simultaneous adsorption of organic

compounds and NOM onto the activated carbon in comparison with those

adsorption isotherms in pure water (Newcombe et al., 2002, Quinlivan et al.,

2005; Matsui et al., 2003). Moreover, it seems that the presence of EfOM

resulted in a linear Freundlich isotherm (n≈1) for both PFOS and PFOA.

Linear isotherm is the simplest expression of equilibrium adsorption, which is

valid for dissolved species that is present at concentrations less than one-half

of its solubility (Schwarzenbach et al., 2003). The effect of adsorption of

effluent organic matter on the efficiency of activated carbon for the removal of

PFCs is significant and would require high carbon dosage to effectively

remove PFCs in wastewater and attain desired water quality.


131

0

50

100

150

200

250

300

0 100 200 300 400Ce (µg/L)

Cs (

µg/m

g) `

EfOM freeEfOM

(a) (b)

Figure 7.1 Adsorption isotherms of PFCs onto the PAC in the absence and presence of EfOM: (a) PFOS; (b) PFOA. Experimental data fit to Freundlich

model (solid line).

Table 7.2 Langmuir isotherm constants and Freundlich isotherm constants for the adsorption of PFCs onto PAC at 25 oC.

Adsorbate Solution

Langmuir isotherm Freundlich isotherm

a (µg/mg) b (L/µg) r2 KF

[(µg/mg) (L/µg)1/n]

1/n r2

PFOS EfOM

free

232.6 0.0545 0.881 17.5469 0.479 0.988

PFOA 200.0 0.0322 0.813 10.03 0.5369 0.981

PFOS EfOM

raw

232.6 0.003 0.345 0.6593 0.9321 0.98

PFOA 0.27 2.359 0.883 0.2043 1.1083 0.984

The adsorption kinetics of PFCs onto PAC was investigated in the presence

and absence of EfOM with initial concentration of 100 µg/L (Figure 7.2).

Equilibrium was observed at contact time of 72 h for both PFOS and PFOA,

while less contact time (4 h) was needed for the adsorption of PFOS or PFOA

onto PAC in Mill-Q water to reach steady state. It was observed that majority

of PFCs were adsorbed in 4 h, which suggests a rapid initial adsorption rate

for both PFOS and PFOA. Furthermore, it seems contact time of 8 h is

sufficient for the most of the PFCs that could be removed by PAC and be

0

50

100

150

200

250

0 100 200 300 400Ce (µg/L)

Cs (

µg/m

g) `

EfOM freeEfOM


132

adsorbed onto the carbon surface from the EfOM solution even though

equilibrium has not reached.

00.10.20.30.40.50.60.70.80.9

1

0 8 16 24 32 40 48 56 64 72Time (hour)

C/C

o

PFOSPFOA

(a) (b)

Figure 7.2 Adsorption of PFOS and PFOA onto PAC as a function of contact time: (a) in the presence of EfOM; (b) in the Milli-Q water.

7.2.1.3 Effect of EfOM on the PFOS and PFOA adsorption onto PAC

Figure 7.3 shows partial adsorption isotherms of PFCs onto PAC in the

presence of 3 type of EfOM fractions as well as absence of EfOM. It can be

seen that the adsorption capacity for PFCs onto PAC was in the following

order: EfOM free>30 k fraction>1 k fraction>EfOM raw. The adsorption

capacity of <1 k fraction was close to that of EfOM raw, especially for PFOA,

while >30 k fraction was close to EfOM free. The fraction of <1 k has greater

effect on the PFCs adsorption than >30 k fraction, indicating direct site

competition between PFCs and the <1 k fraction. It seems that larger

molecular size fraction, which absorb mainly in the larger pores, may not

compete directly for these adsorption sites. As smaller molecules, such as < 1

k fraction, diffuse faster than larger molecules (>30 k EfOM), larger

molecules may be still diffusing through pore structures after the PFCs have

been adsorbed, thereby causing no hindrance to PFCs adsorption. However,

larger molecular weight fraction, >30 k fraction, significantly reduced the

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 1 2 3 4 5 6 7 8

Time (hour)

C/Co

PFOS

PFOA


133

PFCs adsorption capacity compared to EfOM free. This competition effect

could be caused by the presence of the low molecular weight compounds in

the >30 k EfOM fraction. Since UF characterization is influenced by a variety

of factors, including membrane pore size distribution, solution ionic strength,

as well as molecule size and shape (Logan and Jiang, 1990), Newcomber et al.

(2002) found that a certain amount of compounds with small molecular weight

(<500 Da) appeared in the fraction of high molecular weight (>30000 Da).

The presence of the low molecular weight compounds in the >30 k EfOM

fraction could cause decrease in PFCs adsorption by competition effect.

Therefore, the low molecular weight compounds, which have similar

molecular size of PFCs, are the major contributors to the competition. The

direct site competition between target micropollutants and low molecular

weight compounds of similar molecular size has been observed to be the

dominant mechanism by which NOM significantly reduced micropollutants

adsorption capacity onto activated carbon (Newcomb et al., 2002; Kilduff et

al., 1998; Matsui et al., 2003).

0

0.5

1

1.5

2

2.5

3

-1 0 1 2 3log Ce (µg/L)

log

Cs (

µg/m

g) <1K>30KEfOM rawEfOM free

(a) (b)

Figure 7.3 Log-log plot of PFCs adsorption isotherms in the presence and absence of EfOM fractions: (a) PFOS and (b) PFOA.

0

0.5

1

1.5

2

2.5

-1 0 1 2 3

log Ce (µg/L)

log Cs (µg/mg) <1K

>30KEfOM rawEfOM free


134

Freundlich iostherm parameters for the adsorption of PFCs on PAC in EfOM

fractions are shown in Table 7.3. The Freundlich constant 1/n for PFCs

increased with the decreasing molecular weight of the EfOM. It suggests that

EfOM occupied the high energy adsorption sites, which resulted in a decrease

in site heterogeneity. Therefore, the decease in adsorption of PFCs is due to

the decrease in suitable adsorption sites. Moreover, Freundlich constants (1/n

and KF) of >30 k fraction is much more closer to those of EfOM free than the

other samples with background of EfOM, indicating much less competition for

adsorption sites in the >30 k fraction than in EfOM raw or <1 k fraction. The

small molecular weight compounds may be present in the >30 k fraction and

cause the difference in adsorption between >30 k fraction and EfOM free.

Table 7.3 Freundlich iostherm parameters for the adsorption of PFCs on PAC in EfOM fractions.

Compund Parameter EfOM

raw <1 k >30 k EfOM free

PFOS

KF

[(µg/mg)(L/µg)1/n] 0.6593

1.052

2 7.4029 17.5469

1/n 0.9321 1.006 0.746 0.479

PFOA

KF

[(µg/mg)(L/µg)1/n] 0.2043

0.508

5 3.3045 10.03

1/n 1.1083 0.906

9 0.7292 0.5369

7.2.1.4 PFOS and PFOA adsorption onto activated sludge

Biosorption of PFOS and PFOA onto activated sludge were studied and

adsorption isotherms were shown in Figure 7.4. It can be seen that

experimental data fitted linear isotherms well (r2>0.9) for both PFOS and


135

PFOA, indicating that the partitions were independent of the concentrations.

According to the linear isotherm, partition coefficient Kd was 729 L/kg for

PFOS and 154 L/kg for PFOA, respectively (Table 7.4). In comparison,

relatively higher Kd values (720–2,324 L/kg for PFOS, 201–513 L/kg for

PFOA) were observed for the activated sludge in the WWTPs (Yu et al., 2009).

Based on the data on organic carbon ƒoc, calculated activated sludge log Koc

values (partition coefficient for the compound onto a hypothetical pure organic

carbon) were 2.86 for PFOS and 2.19 for PFOA, respectively. The log Koc

value for PFOS of this study is within the range measured by 3M Co. (log Koc

=2.57-3.1), while lower log Koc value (2.57) was reported by Higgins et al.

(2006). For PFOA, log Koc value of this study is slightly higher than those (log

Koc =1.9-2.17) observed by DuPont (2003) and that (log Koc =2.06) reported

by Higgins et al. (2006). In addition, no discernable difference in PFCs

biosorption onto activated sludge of different SRT was observed (Table 7.5),

indicating no effect of SRT on the PFCs biosorption. Some studies reported

the effect of SRT on the sludge characteristics (e.g surface charge, contact

angle) (Liao et al., 2000; Ng et al., 2005; Masse et al., 2006), which could

affect sludge biosorption capacity of organic matters. However, the effect of

SRT on the biosorption of micropollutants was not observed in study.

Table 7.4 Linear isotherm parameters for PFCs onto activated sludge.

Adsorbate Linear isotherm

Kd (L/kg) r2

PFOS 729 0.93

PFOA 154 0.90


136

The observed partition coefficients Kd for PFOS and PFOA are several of

orders of magnitude lower than those of bioaccumulative organic compounds

such as polychlorinated biphenyls and organochlorine pesticides (Katsoyiannis

et al., 2005). It suggests that PFCs have a lower tendency to partition onto the

sludge and sorption onto the activated sludge has no significant effect on the

removal of PFOS and PFOA in activated sludge treatment. Schultz et al.

(2006b) found that about less than 5% PFOS and PFOA were adsorbed onto

the activated sludge in the aeration tank of wastewater treatment plant with

conventional activated sludge treatment system. In addition, the adsorption

capacity of PFOS was more than 3 times higher than that of PFOA, suggesting

more PFOS could be adsorbed onto the activated sludge in wastewater

treatment process as compared to PFOA.

0

0.02

0.04

0.06

0.08

0.1

0 100 200 300Ce (µg/L)

Cs (

µg/m

g) `

PFOSPFOA

Figure 7.4 Adsorption isotherms of PFCs onto the activated sludge.

Experimental data fit to linear isotherm (solid line).


137

Table 7.5 Measured PFCs concentrations in activated sludge of MBR at different SRT.

SRT (d) 5 16 30

Compound PFOS PFOA PFOS PFOA PFOS PFOA

PFCs concentration

in sludge (µg/g) 106±12 22±2 116±13 27±2 111±10 23±3

7.2.2 Performance of MBR and PAC-MBR systems at different SRT

7.2.2.1 Overall performance of MBR and PAC-MBR system

The overall performance of the MBR in terms of COD and DOC in the

supernatant and effluent at different SRTs is summarized in Figures 7.5 and

7.6. The COD removal efficiencies were excellent and stable with an average

of over 95% at all investigated SRTs. Our results are generally consistent with

those reported in the literature (Huang et al., 2001; Lee et al., 2003). Also, it

can be seen that more than 30% DOC was rejected by membrane for both

MBR and PAC-MBR, indicating membrane separation play an important role

in maintaining satisfactory organic removal of MBR/PAC-MBR systems. In

addition, organic removal efficiencies of PAC-MBR at all studied SRTs were

a little higher than those of MBR. It suggests that PAC adsorption of organic

matters improved the overall performance in comparison with MBR.


138

30 day16 day5 day60%

65%

70%

75%

80%

85%

90%

95%

100%

SRT (d)

CO

D re

mov

al (%

) `

MBRPAC-MBR

Figure 7.5 COD removal in MBR and PAC-MBR systems with different SRTs.

5 d 16 d 30 d

0.6

2.6

4.6

6.6

8.6

10.6

12.6

14.6

16.6

18.6

SRT (d)

DO

C (m

g/L) Supernatant (MBR)

Effluent (MBR)Supernatant (PAC-MBR)Effluent (PAC-MBR)

Figure 7.6 DOC of supernatant and effluent in MBR and

PAC-MBR systems with different SRTs. Figure 7.7 shows sludge concentrations in terms of MLSS and MLVSS in the

MBR and PAC-MBR system at different SRTs. As can be seen, average

MLSS concentration decreased accordingly with the decrease of SRT.

However, the ratios of VSS/SS were almost independent of SRT with an

average value over 0.95, indicating no considerable accumulation of inorganic

matter in the MBR system since synthetic wastewater was used as feed rather


139

than real wastewater. Furthermore, it was noted that the metabolic activity of

sludge, characterized by SOUR, slightly decreased as SRT lengthened (Figure

7.8). It could be attributed to the increase of inert biomass (i.e., metabolic

products mainly from endogenous respiration) at long SRTs and possibly to

the potential inhibition effect of soluble microbial products as observed by

Huang et al (2000). At different SRT, the MLVSS of PAC-MBR was found to

be slightly lower than that of MBR, while MLSS of PAC-MBR was

significantly higher than that of MBR. The increase in MLSS of PAC-MBR

could be due to the addition of a certain amount of PAC to the reactor, which

is confirmed by the comparable MLVSS between MBR and PAC-MBR.

Furthermore, the SOUR of PAC-MBR was close to that of MBR at different

SRTs, suggesting no discernable difference in metabolic activity of sludge was

observed between these two systems.

30 d16 d5 d02468

101214161820

SRT (d)

ML

SS/M

LV

SS (g

/L) `

MLSS(MBR)MLVSS(MBR)MLSS (PAC-MBR)MLVSS (PAC-MBR)

Figure 7.7 MLSS/MLVSS in MBR and PAC-MBR

systems with different SRTs.


140

16 d 30 d5 d0

2

4

6

8

10

12

14

SRT (day)

SOUR (mgO2/gVSS h)

SOUR (MBR)

SOUR (PAC-MBR)

Figure 7.8 SOUR in MBR and PAC-MBR systems with different SRTs.

7.2.2.2 SMP and DOM fraction characteristics

Figure 7.9 shows the apparent molecular weight distributions (AMWD) of

DOM in the MBR and PAC-MBR at different SRTs. It can be seen that DOM

in the MBR systems had a broad spectrum of molecular weight. The majority

of DOM, accounting for around 53%, had molecular weight of less than 10

kDa, whereas the components with molecule weights between 10kDa and 30

kDa formed the smallest fraction, constituting 6.1-7.3% of DOM. The fraction

with molecule weights > 30 kDa account for 29-42% of DOM In addition, it

was noted that >30 kDa fraction increased with the increase of SRT, even

though the concentrations of DOM were significantly different. The results are

consistent with those reported in conventional biological treatment systems

where the AMWD of DOM have been found to be greatly affected by SRT

with high molecular weight components becoming more evident at long SRTs

(Barker and Stuckey, 1999).


141

>30K10K-30K1K-10K<1K0%

10%

20%

30%

40%

50%

Molecular weight (Da)

Perc

enta

ge (%

) SRT 5SRT 16SRT 30

(a) (b)

Figure 7.9 AMWD of SMP in the supernatant of (a) MBR and (b) PAC-MBR systems at different SRTs.

The DOM fractionations are shown in Figure 7.10. It can be seen that

hydrophilic HiA were the most abundant fraction of DOM, though their

proportion significantly increased in the MBR or decreased in the PAC-MBR

with the increase of SRTs. AHS accounted for the second largest fraction in

MBR and PAC-MBR systems, probably consisting of humic and fulvic acids.

In addition, it was noted that the proportion of AHS in total DOM gradually

decreased as SRT was lengthened, suggesting that DOM generated at long

SRTs tend to be more hydrophilic. As shown in Figure 7.10, HiB components

constituted the smallest fraction of in the MBR. In addition, proportions of

HoN and HoB were relatively stable and independent of SRT.

HiNHiBHiAHoNHoBAHS0%

10%

20%

30%

40%

50%

60%

DOM fraction

Perc

enta

ge (%

) SRT 5SRT 16SRT 30

Figure 7.10 Hydrophobicity of DOM in the supernatant of (a) MBR and

(b) PAC-MBR systems at different SRTs.

>30K10K-30K1K-10K<1K0%

10%

20%

30%

40%

50%

60%

Molecular weight (Da)

Perc

enta

ge (%

) SRT 5SRT 16SRT 30

HiNHiBHiAHoNHoBAHS0%

10%20%30%40%50%60%70%80%

DOM fraction

Perc

enta

ge (%

) SRT 5SRT 16SRT 30


142

7.2.3 Removal of PFOS and PFOA in PAC-MBR and MBR

7.2.3.1 Removal by adsorption onto activated sludge

Figure 7.11 shows the removal efficiency of PFCs in the MBR system

operated at different SRTs. The highest removal efficiency for both PFOS and

PFOA was observed in MBR with shortest SRT (5 d), while MBR with

longest SRT had lowest removal efficiency. Removal efficiencies of these two

compounds seem to decrease with the increase of SRT, implying no

improvement of biodegradation for these PFCs compounds at longer SRT. It

was reported that some micropollutants, such as endocrine disruptors

compounds (EDCs) or pharmaceuticals could be biodegraded when the

activated sludge treatment system (e.g MBR) was operated with longer SRT

(Clara et al. 2005a; Clara et al, 2005b). Some studies reported increase in

biodegradation of toxic or recalcitrant organic compounds at longer SRT due

to the acclimation and enrichment of certain microorganism (Kimura et al,

2007). However, this study confirmed that these two PFCs compounds can not

be biodegraded in activated sludge system. Furthermore, removal efficiencies

were in the range of 6-14.8% for PFOS and 1.4-3.8% for PFOA at the studied

SRT. As PFOS and PFOA can not be biodegraded, these two compounds can

only be removed by adsorption onto activated sludge or membrane. Filtration

experiment showed that removal efficiency for PFCs was negligible by

membrane (data not shown), indicating MF membrane can not significantly

remove PFCs. It suggests that adsorption onto the sludge would be the major

mechanism for PFCs removal in MBR system. However, low removal

efficiency of PFCs in MBR indicates PFCs can not be efficiently removed by


143

activated sludge system, which is also confirmed by some studies on fate and

behavior of PFCs in WWTPs (Sinclair et al., 2006; Schultz et al., 2006b; Yu et

al., 2009).

30 day16 day5 day0%

5%

10%

15%

20%

SRT (day)

PFC

s re

mov

al in

MB

R (%

)

PFOSPFOA

Figure 7.11 PFCs removal in MBR with different SRTs.

7.2.3.2 Removal by adsorption onto PAC

In PAC-MBR system, PFOS and PFOA could be effectively removed at

appropriate PAC dosage. Figure 7.12 shows the PFCs removal efficiency in

the PAC-MBR system operated at SRT of 30 d with PAC dosage varied from

30 to 100 mg/L. With the increase of PAC dosage, the removal efficiency

increased from 77.4% to 94.8% for PFOS and 67.7% to 90.6% for PFOA. In

contrast, negligible removal efficiencies for these two compounds were

observed in MBR with the same SRT (30 d), which suggest that adsorption of

PFCs onto PAC could play an important role in their removal in the PAC-

MBR system, instead of biosorption onto the activated sludge. Furthermore,

more PFCs were removed by the PAC-MBR at PAC dosage of 100 mg/L in

comparison with that of 30 mg/L, indicating the removal efficiency of PFCs

depend on the PAC dosage.

PFC

s rem

oval

in M

BR

(%)


144

100 mg/L80 mg/L30 mg/L0%

10%20%30%40%50%60%70%80%90%

100%

PAC dosage

PFCs

rem

oval

inPA

C-M

BR (%

)

PFOSPFOA

Figure 7.12 PFCs removal in PAC-MBR system

operated with different PAC dosages

PFCs removal in PAC-MBR system with PAC dosage 100 mg/L was studied

at different SRTs. It can be seen that the removal efficiencies were >90% for

PFOS and >84% for PFOA at different SRT (Figure 7.13). It suggests that

adsorption onto PAC was dominant and removal efficiencies may be not

significantly affected by different operational SRTs. Compared to those of

SRT at 16 d and 30 d, removal efficiencies at SRT of 5 d were slightly lower.

It seems PAC concentration in the reactor would affect the PFCs’ removal

efficiency as the there was the lowest PAC concentration in the reactor at SRT

of 5 d.


145

30 day16 day5 day0%10%20%30%40%50%60%70%80%90%

100%

SRT (day)

PFC

s rem

oval

in P

AC

-MB

R (%

)

PFOSPFOA

Figure 7.13 PFCs removal in PAC-MBR system with

PAC dosage of 100 mg/L at different SRTs.

7.2.3.3 Mass balance

The mass balance of PFOS and PFOA in MBR system was established by

measuring PFCs concentration in aqueous and solid phases of inflow and

outflow. Mass flows of removed PFCs in MBR operated at different SRT are

shown in Figure 7.14. It can be seen that mass flow of PFOS or PFOA in

WAS accounted for more than 82.5% of its total removed amount. PFOS and

PFOA are not biodegraded in the activated sludge process due to their

exceptionally thermal and chemical stability. Since SPE extraction and other

analysis errors would lead to experimental errors, distribution of removed

PFCs mass flow suggests adsorption onto activated sludge could be the only

mechanism that removed PFCs in activated sludge system. In addition, more

PFCs were removed at shorter SRT since mass flow of PFCs in both liquid

and solid phases increased with the decrease of SRT. It seems that more

activated sludge (including solid and liquid phases) wasted out of the reactor

at shorter SRT result in more removed PFCs. Furthermore, mass flow of PFOS


146

or PFOA in the solid phase of WAS decreased with the increase of SRT. The

amount of PFCs in solid phase of WAS in MBR was determined by its

concentration on the sludge surface and mass flow of sludge in WAS. Since no

discernable effect of SRT on the PFCs adsorption on the sludge was found in

this study, decrease in WAS mass flow led to less sludge mass flow

discharged from the MBR with the increase of SRT, which could result in the

reduction of adsorbed PFCs mass flow in WAS. For PFOS, majority of

removed PFOS was adsorbed onto sludge and discharged with WAS at

different SRTs. In contrast, majority of removed PFOA was discharged from

the MBR system in the aqueous phase of WAS at SRT of 5 and 16 d,

indicating different behavior of PFOA in MBR at short SRT in comparison

with PFOS. Based on this study (section 7.2.1.4), adsorption capacity of PFOS

(Kd: 729 L/kg) was more than 3 times higher than that of PFOA (Kd: 154

L/kg). As can be seen, mass flow of PFOS on sludge of WAS was more than

3.5 times of that of PFOA at the same SRT. It suggests that more PFOS was

adsorbed onto the activated sludge, which could result in different behavior in

comparison with PFOA. In addition, mass flow of PFOA in the solid phase of

WAS at SRT of 30 d was more than that in the aqueous phase of WAS since

higher MLSS (avg 7.8 g/L) was observed at SRT of 30 d in comparison with

SRT of 5 and 16 d (3.5 g/L and 5.7 g/L, respectively).


147

545 184 100

1193 666459

686 μg/d943 μg/d1966 μg/d

0%

20%

40%

60%

80%

100%

5 16 30

SRT (d)

Dis

tribu

tion

of re

mov

ed P

FOS

mas

s in

MB

R (%

) errorsludgeaqueous

(a) (b)

Figure 7.14 Distribution of removed PFCs flow in MBR operated at different SRT: (a) PFOS; (b) PFOA. The value on the top of column represents the total mass flow (μg/d) removed in the MBR system; the value in columns indicates

the mass flow of PFCs (μg/d) in aqueous and solid phases.

As the PFCs concentrations in PAC surface can not be measured, their mass

balances in the PAC-MBR system were established by calculations.

Distributions of removed PFCs mass flow in the PAC-MBR at SRT of 30 d

with different PAC dosages were estimated and shown in Figure 7.15. With

the increase of PAC dosage, more PFOS or PFOA was removed by adsorption

on the PAC and activated sludge. However, mass flow in the solid phase of

WAS only increased by 22% for PFOS and 33% for PFOA even though PAC

dosage increased from 30 to 100 mg/L. Based on the PAC mass balance, PAC

concentrations were 2.7, 7.2 and 9.0 g/L in MBR. It seems adsorption capacity

of PAC decreased significantly as PAC concentration in MBR increased

greatly. Furthermore, it can be seen that more than 98% of removed PFCs was

in the solid phase (including activated sludge and PAC) of WAS. Compared to

MBR with the same SRT, most of the PFCs in the solid phase of WAS seems

to be adsorbed onto the PAC instead of activated sludge. For example, 459

mg/d of PFOS and 120 mg/d of PFOA were removed by adsorption onto the

activated sludge of the MBR, while mass flows in solid phase of WAS of the

616195 105

247

155120

240 μg/d426 μg/d980 μg/d

0%

20%

40%

60%

80%

100%

5 16 30

SRT (d)

Dis

tribu

tion

of re

mov

ed P

FOA

mas

s in

MB

R (%

) differencesludgeaqueous


148

PAC-MBR with PAC dosage of 30 mg/L were 7,430 mg/d for PFOS and

6,499 mg/d for PFOA at same SRT (30 d). It suggests adsorption on PAC was

an efficient and predominant process in the removal of PFCs in activated

sludge system. PAC adsorption would be much more effective than

biosorption for the removal of PFCs in the wastewater treatment even though

its adsorption capacity was significantly reduced by EfOM.

100 mg/L80 mg/L30 mg/L

9106 µg/d8608 µg/d7455 µg/d

0%

20%

40%

60%

80%

100%

PAC dosage (mg/L)

Dis

tribu

tion

of re

mov

ed P

FOS

mas

s in

PAC

-MB

R (%

)

solid WAS(PAC+sludge)aqueousWAS

(a) (b)

Figure 7.15 Estimated distributions of removed PFCs mass flow in waste of PAC-MBR at SRT 30 d with different PAC dosage: (a) PFOS; (b) PFOA. The

value on the top of column represents the total mass flow removed in the PAC-MBR system.

Figure 7.16 shows the estimated distributions of removed PFCs mass in PAC-

MBR operated at different SRT with a PAC dosage of 100 mg/L. The total

removal mass flow of PFOS or PFOA was comparable at different SRT,

indicating insignificant effect of SRT on the PFCs removal with the presence

of PAC. It seems that the effect of SRT on the PFCs’ removal could be

overridden by the effect of PAC adsorption. Furthermore, even though PAC-

MBR was operated at different SRT, mass flow of PFCs in the solid phase was

more than 98% of the total removed PFCs mass flow. Compared to the MBR

with the same SRT, most of PFCs in the solid phase of WAS seemed to be

adsorbed onto the PAC instead of activated sludge. For example, 1,193 mg/d

6534 µg/d 8100 µg/d 8706 µg/d

100 mg/L80 mg/L30 mg/L0%

20%

40%

60%

80%

100%

PAC dosage (mg/L)

Dis

tribu

tion

of re

mov

ed P

FOA

mas

s in

PAC

-MB

R (%

)

solid WAS(PAC+sludge)aqueousWAS


149

of PFOS was removed by adsorption onto the activated sludge of MBR, while

mass flow of PFOS in solid phase of WAS was 8,650 mg/d at same SRT (5 d)

of PAC-MBR. According to this study (section 7.2), PAC adsorption was

much more than biosorption. With presence of PAC, most of the PFCs is

expected to adsorb onto the PAC in the PAC-MBR. It was estimated about

171 mg/d of PFOS, instead of 1193 mg/d, was adsorbed onto the activated

sludge in WAS based on the partition coefficient of PFOS (Table 7.6). Table

7.6 shows estimated mass flows of PFCs in activated sludge of WAS in the

PAC-MBR operated at different SRTs. Biosorption accounted for <2% of total

removed PFCs amount at different SRT, indicating PFCs removal due to

biosorption was negligible in the PAC-MBR. It also confirmed that adsorption

on PAC is the predominant process in the removal of PFCs in activated sludge

system at appropriate PAC dosage, which would not be significantly affected

by SRT.

5 d 16 d 30 d

8713 µg/d 9139 µg/d 9106 µg/d

0%

20%

40%

60%

80%

100%

SRT (d)

Dis

tribu

tion

of re

mov

ed P

FOS

mas

s in

PAC

-MB

R (%

) solid WAS(PAC+sludge)aqueous WAS

(a) (b)

Figure 7.16 Estimated distributions of removed PFCs mass flow in waste of PAC-MBR operated at different SRTs: (a) PFOS; (b) PFOA. The value on the

top of column represents the total mass flow removed in the MBR system.

8703 µg/d8735 µg/d8820µg/d

5 d 16 d 30 d0%

20%

40%

60%

80%

100%

SRT (d)

Dis

tribu

tion

of re

mov

ed P

FOA

mas

s in

PAC

-MB

R (%

) solid WAS(PAC+sludge)aqueous WAS


150

Table 7.6 Estimated mass flows of PFCs in activated sludge of WAS in PAC-MBR operated at different SRTs.

SRT (day) 5 16 30 PFOS PFOA PFOS PFOA PFOS PFOA

PFCs concentration in sludge (µg/g)

14.43 4.74 7.14 2.83 7.58 2.91

PFCs mass flow in sludge (µg/d) 171.36 56.31 39.72 15.75 30.53 11.72

Total removed PFCs (µg/d) 8712.96 8220.16 9139.4 8735.2 9106.35 8702.88

Percentage in WAS (%) 1.98% 0.69% 0.44% 0.18% 0.34% 0.13%

7.2.3.4 Effect of SRT on PFOS and PFOA removal

Figure 7.17 indicates that PFCs concentration in sludge were slightly different,

varying from 106 to 116 µg/g (PFOS) and 22 to 27 µg/g (PFOA). Furthermore,

calculated PFCs concentrations in sludge were estimated by dividing mass

flow of PFCs in solid phase of WAS by the amount of activated sludge

discharged from MBR. Calculated PFCs concentrations on the sludge surface

were consistent with the measured values. It seems that sludge adsorption

capacity was consistent at different SRTs, indicating SRT had no significant

effect on the PFCs adsorption onto activated sludge.

5 d 16 d 30 d0

20

40

60

80

100

120

140

160

SRT (day)

PFO

S co

ncen

tratio

n on

slud

ge su

rface

(u

g/g)

measuredcalculated

(a) (b)

Figure 7.17 Effect of SRT on the PFCs adsorption onto activated sludge in MBR: (a) PFOS; (b) PFOA.

30 d16 d5 d0

10

20

30

40

SRT (day)

PFO

A c

once

ntra

tion

on sl

udge

surfa

ce

(ug/

g)

measuredcalculated


151

The effect of SRT on the adsorption of PFCs onto PAC in the PAC-MBR

system is shown in Table 7.7. As the mass flow of PFCs in aqueous phase of

the WAS were negligible, the normalization of PAC adsorption was calculated

by dividing total removed PFCs mass flow by the mass flow of PAC in the

WAS. Expected PAC adsorption capacity was predicted by the partition

coefficient Kd of this study (see section 7.2.1.4). It can be seen that PFCs

concentrations on PAC at SRT of 5 d were 5 times more than those at SRT of

30 d. With the increase of SRT, PFCs concentration on PAC decreased

significantly, indicating significant effect of SRT on the PAC adsorption

capacity in the PAC-MBR due to different PAC concentrations at different

SRTs. Furthermore, PAC adsorption capacity was not fully utilized at different

SRT in comparison with expected adsorption capacity when PAC was dosed

at 100 mg/L in the MBR. With the increase of SRT, utilized PAC adsorption

capacity decreased from 54.1% to 17.3% (PFOS) and 65.5% to 19.8% (PFOA).

It seems that PAC adsorption capacity could decrease significantly with the

increase of SRT. Therefore, PAC could have highest adsorption capacity in

the PAC-MBR at shortest SRT, which suggests fouling of PAC may

deteriorate and result in significant reduction in its adsorption capacity (Lee et

al., 2005; Ng et al., 2006). In addition, PFOA concentrations on PAC at

different SRT were comparable to those of PFOS even though PAC adsorption

capacity of PFOS was higher than that of PFOA with the presence of EfOM. It

may be due to the high PAC dosage added in the system (100 mg/L), which

overrided the difference in their adsorption capacity.


152

Table 7.7 Effect of SRT on the PFCs removal in PAC-MBR system with PAC dosage of 100 mg/L (based on mass balance).

SRT (day) 5 16 30 PFOS PFOA PFOS PFOA PFOS PFOA

Inflow mass flow (µg/d) 9600 9600 9600 9600 9600 9600

Outflow mass flow (µg/d) 887.0 1379.8 460.6 864.8 493.6 897.1

Total removed mass (µg/d) 8713 8220 9139 8735 9106 8703

Mass flow in aqueous WAS (µg/d)

63.36 98.56 9.80 18.40 5.55 10.08

Mass flow in solid WAS (PAC+sludge) (µg/d)

8649.60 8121.60 9129.60 8716.80 9100.80 8692.80

PFCs concentrations on PAC (µg/g)

5766.40 5414.40 1902.00 1816.00 1011.20 965.87

Expected PAC adsorption capacity (µg/g)

10666.84 8265.42 5538.17 4738.42 5853.55 4877.63

Utilized PAC adsorption capacity (%)

54.1% 65.5% 34.3% 38.3% 17.3% 19.8%

7.2.3.5 Effect of PAC dosage on PFOS and PFOA removal

The effect of PAC dosage on the adsorption of PFCs in PAC-MBR system is

shown in Table 7.8. As PAC dosage was increased from 30 to 100 mg/L,

PFCs concentrations on PAC decreased from 2,750 µg/g to 1,011 µg/g,

indicating significant effect of PAC dosage on PAC adsorption capacity for

PFCs in the PAC-MBR. According to the PAC adsorption study, PFCs

adsorption on PAC fitted Freundlich isotherms with the presence of EfOM,

which predicted that PAC would have lower adsorption capacity at higher

PAC dosage. Furthermore, utilized PAC adsorption capacity varied from

11.9% to 17.3% (PFOS) and 13.1% to 19.8% (PFOA) even though PAC

dosage tripled. The comparable utilized PAC capacity at different PAC


153

dosages indicates that fouling effect on the PAC could be similar at the same

SRT. In addition, PFOA concentrations on PAC at different PAC dosages

were slightly lower than those of PFOS even though PAC adsorption capacity

of PFOS was much higher than that of PFOA. It is possible that fouling effect

on the PAC could significantly reduce the difference in PFCs adsorption onto

PAC.

Table 7.8 Effect of PAC dosage on the PFCs removal in PAC-MBR system (based on mass balance).

PAC dosage (mg/L)

30 80 100 PFOS PFOA PFOS PFOA PFOS PFOA

Inflow mass flow (µg/d) 9600 9600 9600 9600 9600 9600

Outflow mass flow (µg/d) 2145.5 3066.3 992.1 1499.9 493.7 897.1

Total removed mass flow (µg/d)

7454.5 6533.7 8607.9 8100.1 9106.3 8702.9

Mass flow in aqueous WAS (µg/d)

23.96 34.24 11.08 16.75 5.51 10.02

Mass flow in solid WAS (PAC+sludge) (µg/d)

7430.55 6499.42 8596.87 8083.31 9100.83 8692.86

PFCs concentrations on PAC (µg/g)

2750.34 2405.69 1194.01 1122.68 1011.20 965.87

Expected PAC adsorption capacity (µg/g)

23021.46 18394.22 11218.12 8497.52 5853.55 4877.63

Utilized PAC adsorption capacity (%)

11.9% 13.1% 10.6% 13.2% 17.3% 19.8%

7.2.4 Membrane fouling

7.2.4.1 Variations of TMP

Figure 7.17 shows the long-term TMP profile for MBR and PAC-MBR system

at different SRT. For MBR system, noticeable membrane fouling was first

observed on day 38, 57 and 61 in the MBR operated at SRT of 5, 16 and 30


154

day, respectively. Subsequently, the TMP increased rapidly until day 47, 74

and 79 when membrane was removed for chemical cleaning. For PAC-MBR

system, noticeable fouling was detected after 67 d of operation for SRT of 5 d,

which was 1.76 time longer than the MBR system without PAC addition. It

seems that PAC addition would decrease the TMP of PAC-MBR at the same

operation condition as that of MBR, thus allowing the PAC-MBR system to

operate for a longer time to reach maximum total membrane resistance caused

by cake layer formation and solute adsorption on the membrane.

0

1

2

3

4

5

6

7

8

9

10

11

0 20 40 60 80 100 120 140days

Nor

mal

ized

TM

P ΔP/

ΔP0

SRT 5d (MBR)

SRT 16d (MBRC)

SRT 30d (MBR)

SRT 5d (PAC-MBR)

SRT 16d (PAC-MBR)

SRT 30d (PAC-MBR)

Figure 7.17 Long-term TMP profile for the MBR and

PAC-MBR systems at different SRTs.

7.2.4.2 Effect of PAC on TMP

Resistances of membrane for MBR and PAC-MBR system are summarized in

Table 7.9. It can be seen that total resistance (Rt) and intrinsic resistance (Rm)

are nearly same for MBR and PAC-MBR system. Compared to MBR system,

reversible resistance (Rr) decreased form 2.11 to 1.95 (1012·m-1) in PAC-

MBR system, while irreversible resistance (Ri) increased from 0.18 to 0.33

(1012 ·m-1). It suggests that PAC addition could reduce the cake layer


155

resistance and increase the percentage of Ri in the total membrane resistance.

Thus it allows the PAC-MBR system to operate for a longer time to reach

maximum irreversible fouling resistance caused by solute adsorption on the

membrane. Li et al. (2005) confirmed that Rr of PAC-MBR was 17.9% lower

than that of MBR, while Ri was 25.5% higher than that of MBR. In addition,

Rt of PAC-MBR was significantly lower than that of MBR when they were

operated at the same condition before membrane fouling occurred, which was

found to be due to the significant reduction in reversible resistance (Rr). It

suggests that PAC play an important role in reducing cake resistance and

changing an overall particle size distribution to a greater size range. (Li et al.,

2005; Munz et al., 2007).

Table 7.9 Resistances of membrane for the MBR and PAC-MBR systems

Resistances MBR (1012·m-1) PAC-MBR

(1012·m-1)

Rm 0.31 0.32

Rr 2.11 1.95

Ri 0.18 0.33

Rt 2.6 2.6

7.3 Summary

The simultaneous adsorption of EfOM and PFOS or PFOA onto PAC was

investigated in this study. The presence of EfOM significantly decreased the

adsorption capacity of PFCs onto PAC in comparison with that in the absence

of EfOM. Adsorption of PFCs to PAC fitted the Freundlich model well

(r2>98%) in the absence and presence of EfOM within the studied

concentration range (0.1-500 µg/L). According to Freundlich model,


156

adsorption capacity of PFOS (KF=17.55) and PFOA (KF=10.03) onto the PAC

in pure water was more than one order of magnitude higher than that of EfOM

solution (KF=0.66 for PFOS, KF=0.20 for PFOA), indicating that presence of

EfOM greatly reduce the adsorption capacity of PAC. The adsorption kinetics

of PFCs was investigated in the presence of EfOM with initial concentration

of 100 µg/L. A rapid initial adsorption rate was observed in 4 h for both PFOS

and PFOA. It seems the contact time of 8 h was sufficient for the PFCs to be

adsorbed onto the carbon surface from the EfOM solution even though

equilibrium had not been reached.

EfOM solution was characterized by ultrafiltration and four EfOM fractions

were obtained to investigate their effects on the PFCs adsorption. The

adsorption capacity for PFCs onto PAC was in the following order: EfOM

free> 30 k fraction>1 k fraction>EfOM solution. It seems that larger

molecular size fraction, which was absorbed mainly in the larger pores, may

not compete directly for these adsorption sites. However, the smaller

molecular weight compounds, which had the similar molecular size of PFCs,

were the major contributors to the competition. The direct site competition

between target PFCs and low molecular weight compounds of similar

molecular size seems to be the dominant mechanism by which EfOM

significantly reduced PFCs adsorption capacity onto activated carbon.

Adsorption of PFCs to activated sludge fitted linear isotherms (r2>0.9) within

concentration range of 50-400 µg/L, which indicated that the partitions were

independent of the concentrations for both PFOS and PFOA. Based on our


157

data, the estimated partition coefficient Kd was 729 L/kg for PFOS and 154

L/kg for PFOA, respectively. It suggests that PFOS and PFOA, especially

PFOA, have a low tendency to partition onto the sludge, indicating sorption

onto the activated sludge has insignificant effect on the removal of PFOS and

PFOA in activated sludge treatment process.

Removal efficiencies of PFCs in MBR were investigated at different SRT,

which were in the range of 6-14.8% for PFOS and 1.4-3.8% for PFOA. PFCs

low removal efficiency (<15%) in MBR indicates PFCs can not be efficiently

removed by activated sludge system. Distribution of removed PFCs mass flow

suggests that adsorption onto activated sludge could be the only mechanism

that removed PFCs in activated sludge system. More PFCs was removed at

shorter SRT since mass flow of PFCs in both liquid and solid phases increased

with the decrease of SRT. Furthermore, PFCs mass flow in the solid phase of

WAS decreased with the increase of SRT. It is possibly attributed to the

decrease in sludge mass flow discharged from the MBR, which could result in

the reduction of adsorbed PFCs mass flow in WAS. In addition, PFCs

concentrations in sludge were slightly different at different SRT, varying from

106 to 116 µg/g (PFOS) and 22 to 27 µg/g (PFOA). It seems that sludge

adsorption capacity was consistent at different SRTs, indicating SRT had no

significant effect on the PFCs adsorption onto activated sludge.

The overall performance and removal efficiencies of PFCs were investigated

on PAC-MBRs which operated with different PAC dosages and SRTs. The

effect of PAC dosage on the removal of PFCs in PAC-MBR was studied at the


158

SRT 30 d. Removal efficiency increased from 77.4% to 94.8% for PFOS and

67.7% to 90.6% for PFOA with the increase of PAC dosage from 30 to 100

mg/L. Based on the established mass balance, it suggests that adsorption on

PAC was the efficient and predominant process in the removal of PFCs in

activated sludge system. PAC adsorption would be much more effective than

biosorption for the removal of PFCs in the wastewater treatment even though

its adsorption capacity was significantly reduced by EfOM. As PAC dosage

increased from 30 mg/L to 100 mg/L, PFCs concentrations on PAC decreased

from 2,750 to 1,011 µg/g, indicating the significant effect of PAC dosage on

PAC adsorption capacity for PFCs in PAC-MBR. However, utilized PAC

adsorption capacity was relatively consistent in the range of 11.9% to 17.3%

(PFOS) and 13.1% to 19.8% (PFOA) even though PAC dosage was

significantly increased. The comparable utilized PAC capacity at different

PAC dosage indicates that fouling effect on the PAC could be similar at the

same SRT.

The effect of SRT on removal of PFCs in PAC-MBR was further investigated.

Removal efficiencies were >90% for PFOS and >84% for PFOA at different

SRT, suggesting that adsorption onto PAC could be dominant and removal

efficiencies may be not significantly affected by different operational SRT.



MBR due to different PAC concentrations at different SRTs. In addition,

utilized PAC adsorption capacity decreased from 54.1% to 17.3% (PFOS) and

65.5% to 19.8% (PFOA) when SRT was increased from 5 to 30 d. It seems


159

that PAC adsorption capacity could decrease significantly with the increase of

SRT, which was possibly due to deteriorating fouling of PAC.

TMP profile for MBR and PAC-MBR system at different SRT observed in

this study suggests that PAC could significantly extend the operation time of

PAC-MBR. PAC addition could reduce the cake layer resistance and increase

the percentage of irreversible fouling resistance in the total membrane

resistance, thus allowing the PAC-MBR system to operate for a longer time to

reach the maximum irreversible fouling resistance caused by solute adsorption

on the membrane.

Chapter 8-Conclusions

160

CHAPTER 8 CONCLUSIONS

8.1 Conclusions

This study investigated the occurrence and fate of PFOS and PFOA in water

and wastewater as well as explored removal strategy of hybrid PAC-MBR

process. For the first time, it provided data on spatial and seasonal occurrence

and distribution of PFOS and PFOA in Singapore water environment,

including rivers, reservoirs and lakes and sea water around the island. PFOS

and PFOA were detected in all collected samples in the range of 1.9~532.1

ng/L (PFOS) and 2.4~1,057.1 ng/L (PFOA). Seawater had lower concentration

of PFOS and PFOA, compared with surface waters and treated effluents. In

surface waters, the highest total concentrations of PFOS and PFOA were

observed in the western area because of the high levels of industrial activities

in that area. This region was noted to be the most highly contaminated by

PFCs. In wastewaters, the highest total PFCs mass load and PFOA

concentration (1,057.1 ng/L) were observed in W5, suggesting discharges of

fluorochemical related factories in the service area of W5 may contain a large

amount of PFOS and PFOA, thus resulting in high concentrations in the

WWTPs effluents. The highest PFOS concentration (532.1 ng/L) was detected

in the effluent of W1 treating mainly domestic and commercial wastewater.

This indicates the presence of potential PFOS contamination sources in its

service area. Compared with surface waters and coastal waters, much higher

PFCs concentrations in wastewaters indicate that discharge of effluents of

WWTPs is an important pathway by which PFCs enter the environment. In


161

coastal water, the high PFOS and PFOA concentrations at C4 suggest that

Johor Straits is more heavily contaminated than the southern and eastern

coastal waters. The high levels of industrial activities in the western area may

be the significant contamination sources for Johor Straits. Furthermore,

significant seasonal variation between dry seasons and wet seasons was

observed in surface waters for PFOS only, while no discernable seasonal

differences were found for both PFOS and PFOA in coastal waters and

wastewaters. In addition, PFOS and PFOA were significantly correlated in the

coastal waters, while weak positive correlations were observed in surface

waters and wastewaters. It suggests that the possibility of a common

contamination source for these two compounds in coastal waters is higher than

those of surface waters and wastewaters.

An efficient sample clean-up method was developed in this study to

significantly remove co-eluting matrix components by applying the SPE

extracts onto a silica cartridge after dilution with dichloromethane. Matrix

effect on PFOS and PFOA were evaluated by comparing MS responses of

standards and those of the same known amount of analytes in post-extraction

spiked samples. It was found that that ME% for both PFOS and PFOA were

below 50%, indicating SPE alone was insufficient to remove matrix

components. Also, recoveries (RE%: <50%) were significantly affected by the

matrix effect due to the ionization suppression even though this HLB SPE

procedure can achieve more than 90% recoveries (98.4% for PFOS and 93.8%

for PFOA) for PFCs spiked Mill-Q water. Therefore, silica cartridge was

applied to reduce the co-eluting interfering compounds and ME% (>70%) and


162

RE% (>67%) were increased significantly for both PFOS and PFOA. It

suggests that substantial amount of interfering compounds were retained by

silica cartridge, while PFOS and PFOA were eluted by mixture of

dichloromethane/methanol (60:40, v/v). After silica cartridge clean-up, the

coefficient of variation (CV) decreased more than 44% (PFOS) and 34%

(PFOA) for ME% and RE%, indicating precision of the analysis increased due

to the reduced matrix effect. The application of internal standards further

compensated for matrix effect and brought the ME% and RE% close to 100%,

indicating minimal matrix effect was achieved without significant loss of

analytes. CV was greatly decreased by applying internal standardization with

its value below 5%. In addition, a higher recovery (>90%) was achieved

compared to that of around 70% without internal standardization. The

developed LC-MS-MS detection method was applied to different water and

sludge samples. Results showed that this silica cartridge clean-up method can

effectively remove co-eluting matrix components in various environmental

matrices with ME% >95% for water samples and >90% for sludge samples.

The behavior of PFOS and PFOA in the biological units of various full-scale

municipal wastewater treatment plants was studied. Samples of influent,

primary effluent, aeration tank effluent, final effluent and grab samples of

primary, activated, secondary and anaerobically digested sludge were

collected by 5 sampling events over one year. The two sewage treatment

plants (STPs) selected for this study included plant A receiving 95% domestic

wastewater and plant B receiving 60% industrial wastewater and 40%

domestic wastewater. PFOS and PFOA were detected in all aqueous samples


163

collected from STP A and B, ranging from 5.3 - 560.9 ng/L and 11.2 – 1,057.1

ng/L, respectively. In sludge of STPs A and B, PFOS and PFOA

concentrations were in the range of 13.1 – 702.2 ng/g dw and <5.0 - 69.0 ng/g

dw, respectively. It is noted that PFOS and PFOA were observed at higher

concentration in aqueous and sludge samples in STP B than those of STP A,

indicating that industrial sewage contain a larger amount of PFCs in

comparison with domestic sewage. Significant increase in mass flow of PFOS

(mean 94.6%) was observed in CAS2, while it remained consistent after

secondary treatment in CAS1. This is likely due to no occurrence of PFOS

precursors in the raw sewage. Mean mass flow of PFOA increased 41.6% in

CAS1, 67.0% in CAS2 and 76.6% in MBR, while it remained unchanged after

the treatment of LTM. Different behavior pattern of these two compounds

were found in LTM, an activated sludge process operated at a relatively short

SRT. The findings suggest that change in mass flow of PFOS and PFOA in

secondary sludge treatment may be determined by the presence of precursors

and operating SRT of the activated sludge system. Furthermore, between dry

and wet seasons, seasonal variation of PFOS was observed in STP A, while

PFOA had no significant difference in both STP A and STP B. PFOS

concentration in rainwater observed by other studies was lower than that of

PFOA, which could lead to their different seasonal variations in surface water.

It is also likely that NPS of PFCs occurred in wet season, which would

contribute to consistent PFOA concentrations in surface waters and

subsequently resulted in indiscernible variation in PFOA concentrations in the

wastewaters between dry and wet seasons.


164

The adsorption of PFOS and PFOA onto powdered activated carbon (PAC)

was investigated in the presence and absence of EfOM at low concentration

range (0.1-500 µg/L). Adsorption of PFOS and PFOA to PAC fitted the

Freundlich model well (r2>98%) and adsorption capacity of PFOS (KF=17.55)

and PFOA (KF=10.03) in the absence of EfOM was more than one order of

magnitude higher than those in the presence of EfOM (KF=0.66 for PFOS,

KF=0.20 for PFOA), indicating EfOM greatly reduced the adsorption capacity

of PAC. The adsorption kinetics of PFCs was investigated in the presence of

EfOM with initial concentration of 100 µg/L. A rapid initial adsorption rate

was observed in 4 h for both PFOS and PFOA. It seems the contact time of 8 h

was sufficient for the PFCs to be adsorbed onto the carbon surface from the

EfOM solution even though equilibrium had not been reached. Moreover,

EfOM was characterized by ultrafiltration and fractions of nominal molecular

weights were obtained to investigate their effect on the PFOS and PFOA

adsorption. The fraction of <1 k had greater effect on the adsorption than >30

k fraction, indicating the similar molecular size of target compounds, were the

major contributors to the adsorption competition. The direct site competition

between target PFCs and low molecular weight compounds of similar

molecular size seems to be the dominant mechanism by which EfOM

significantly reduced PFCs adsorption capacity onto activated carbon.

Additionally, biosorption of PFOS and PFOA to the activated sludge fitted the

Linear isotherm (r2>0.9) within concentration range of 50-400 µg/L. Based on

our data, the estimated partition coefficient Kd was 729 L/kg for PFOS and

154 L/kg for PFOA, suggesting PFOS and PFOA, especially PFOA, have a

low tendency to partition onto sludge.


165

This study explored overall removal performance and factors affecting PFCs’

adsorption onto activated sludge and PAC in MBR and hybrid PAC-MBR

processes. Laboratory-scale MBR and PAC-MBR were operated in parallel at

SRT of 5, 16, and 30 days for treatment of readily biodegradable synthetic

wastewater. Removal efficiencies of PFCs in MBR were in the range of 6-

14.8% for PFOS and 1.4-3.8% for PFOA at different SRT studied. PFCs low

removal efficiency (<15%) in MBR indicates PFCs can not be efficiently

removed by activated sludge system. Distribution of removed PFCs mass flow

suggests adsorption onto activated sludge could be the only mechanism that

removed PFCs in activated sludge system. More PFCs was removed at shorter

SRT since mass flow of PFCs in both liquid and solid phases increased with

the decrease of SRT. In addition, PFCs concentrations in sludge were slightly

different at different SRT, varying from 106 to 116 µg/g (PFOS) and 22 to 27

µg/g (PFOA). It seems that sludge adsorption capacity was consistent at

different SRTs, indicating SRT had no significant effect on the PFCs

adsorption onto activated sludge.

The overall performance and removal efficiencies of PFCs were also

investigated in PAC-MBRs which operated with different PAC dosage and

SRTs. On the one hand, the effect of PAC dosage on the removal of PFCs in

PAC-MBR was studied at the SRT of 30 d. Removal efficiency increased

from 77.4 to 94.8% for PFOS and 67.7 to 90.6% for PFOA with the increase

of PAC dosage from 30 to 100 mg/L. Based on the established mass balance,

it suggests adsorption on PAC was the efficient and predominant process in


166

the removal of PFCs in activated sludge system. PAC adsorption would be

much more effective than biosorption for the removal of PFCs in the

wastewater treatment even though its adsorption capacity was significantly

reduced by EfOM. As PAC dosage was increased from 30 to 100 mg/L, PFCs

concentrations on PAC decreased from 2,750 to 1,011 µg/g, indicating

significant effect of PAC dosage on PAC adsorption capacity for PFCs in

PAC-MBR. However, utilized PAC adsorption capacity was relatively

consistent in the range of 11.9 to 17.3% (PFOS) and 13.1 to 19.8% (PFOA)

even though PAC dosage significantly increased. The comparable utilized

PAC capacity at different PAC dosage indicates that biofouling effect on the

PAC could be similar at the same SRT. On the other hand, the effect of SRT

on removal of PFCs in PAC-MBR was studied. Removal efficiencies of PFCs

were >90% for PFOS and >84% for PFOA at different SRT studied,

suggesting that adsorption onto PAC could be dominant and removal

efficiencies may be not significantly affected by different operational SRT.



MBR due to different PAC concentrations at different SRTs. In addition,

utilized PAC adsorption capacity decreased from 54.1 to 17.3% (PFOS) and

65.5 to 19.8% (PFOA) when SRT was increased from 5 to 30 d. It seems that

PAC adsorption capacity could decrease significantly with the increase of SRT,

which was possibly due to deteriorating fouling of PAC.

TMP profiles for MBR and PAC-MBR system at different SRTs observed in

this study suggests that PAC could significantly extend the operation time of


167

PAC-MBR. PAC addition could reduce the cake layer resistance and increase

the percentage of irreversible fouling resistance in the total membrane

resistance. Overall, PAC addition would decrease the TMP of PAC-MBR at

the same operation condition as that of MBR, thus allowing the PAC-MBR

system to operate for a longer time to reach the maximum total membrane

resistance caused by cake layer formation and solute adsorption on the

membrane.

Contributions of this study would provide a better understanding of occurrence

and fate of PFOS and PFOA in aquatic environment as well as behavior in

sewage treatment plants. The seasonal varitions of PFCs in aquatic

environment were explored between dry and wet seasons in an ideal island,

where other climate factors were excluded from this study. Moreover, the

developed post extraction cleanup method should contribute to higher

accuracy for detection of wastewater and sludge samples. Also, it should be

noted that the effect of SRT on the PFCs mass change would deepen the

understanding of their behavior patterns in STPs. To our best of knowledge, it

is the first study to examine the effect of SRT on the PFCs’ behavior in

activated sludge treatment process. On the other hand, the study on

simultaneous adsorption of EfOM and PFOS or PFOA onto PAC investigated

in this study would provide valuable new insights into the characteristics of

PAC adsorption in the MBR and consequently further advance our knowledge

on the removal of PFOS and PFOA in the hybrid PAC-MBR process as well

as activated sludge process.


168

8.2 Recommendations

The transport pathways of PFCs to aquatic environment could include

discharge of effluents from STPs, direct discharge of wastewater from

manufacture and use of PFCs, rain runoff and atmospheric transport of PFCs

and subsequent atmospheric loading of PFCs to surface waters. It is of note

that research with respect to fate of PFCs in the aquatic environment is far

from complete and much work is needed to fully understand this important

issue. PFCs concentrations in the air, drinking water, ground water, rainwater

and rain runoff should be investigated to identify possible contamination

sources and transport pathways of PFCs in environment.

The developed silica cartridge clean-up can effectively remove interfering

components and significantly improve the accuracy of the LC/MS/MS

analyses. In this study the developed method is limited to the analysis of PFOS

and PFOA. As other PFCs compounds have similar physico-chemical

properties, the developed clean-up method could be applied to analysis of

other PFCs compounds. It would significantly enlarge the contribution of

developed clean-up method. Therefore, further research should be conducted

to explore the possibility of extending the developed clean-up method to other

PFCs compounds analysis in environmental matrices.

In order to achieve better understanding of behavior of studied compounds in

sewage treatment plants, this study is restricted to investigate compounds

PFOS and PFOA only and there is no intention to identify their precursors and

product compounds. Future research should identify the related precursors


169

and explore the mass transfer between aqueous and solid phases along the

treatment processes. Degradation of precursors can lead to occurrence and

variation of studied compounds in concentration and mass flow in water and

wastewater. Relationship between precursors and studied compounds should

be studied to identify the contribution in mass increases from their precursors.

Furthermore, mass balance in STPs in this study was not attempted because

only grab samples, instead of composite samples, were collected from STPs.

Composite samples should be collected and analyzed to establish the mass

balance of studied compounds in the whole STP. Then behavior of studied

compounds may be completely and accurately understood by identifying mass

flow increase due to degradation of their precursors and mass flow decrease

due to sludge adsorption.

It was found that PAC adsorption could be the removal mechanism of PFCs

for the hybrid PAC-MBR process. Site competition was suggested to be the

adsorption mechanism of PFCs onto the PAC in the presence of EfOM. It

should be noted that other adsorption mechanism such as pore blockage could

override site competition and be the dominant mechanism. In order to exclude

the pore blockage mechanism, a few adsorbents with different pore size

distributions should be studied to investigate the effect of pore size

distribution on the adsorption in the presence of EfOM. Moreover, it would

provide the knowledge on the characteristics of optimal adsorbent for removal

of PFCs in the hybrid PAC-MBR process.

Reference

170

REFERENCE

3M (1999) Fluorochemical Use, Distribution and Release Overview. US EPA

Administrative Record AR226-0550. U.S. Environmental Protection Agency,

Washington, DC.

3M (2000) Soil Adsorption/Desorption Study of Potassium Perfluorooctane

Sulfonate (PFOS). US EPA Docket AR 226-1030a 030. U.S. Environmental

Protection Agency, Washington, DC.

3M (2001) Environmental Monitoring-Multi-City Study (Water, Sludge,

Sediment, POTW Effluent and Landfill Leachate Samples). Docket AR-226-

1030a; Office of Pollution and Prevention and Toxic Substances, U.S.

Environmental Protection Agency, Washington, DC.

3M (2003) Health and environmental assessment of perfluorooctane sulfonic

acid and its salts. US EPA docket no. AR-226-1486. US Environmental

Protection Agency, Washington, DC.

Ahn, S., Werner, D., Karapanagioti, H.K., Zare, R.N., Luthy, R.G. (2005)

Phenanthrene and pyrene sorption and intraparticle diffusion in

polyoxymethylene, coke, and activated carbon. Environ. Sci. Technol. 39,

6516-6526.

Aktas, O., Cecen, F. (2007) Bioregeneration of activated carbon: A review. Int.

Biodeter. & Biodegr. 59, 257-272.

APHA-AWWA-WEF. Standard Methods for Examination of Water and

Wastewater, 20th ed; APHA-AWWA-WEF, Washington, DC, 1998.

Reference

171

Aquino, S.F., Hu, A.Y., Akram, A., Stuckey, D.C. (2006) Characterization of

dissolved compounds in submerged anaerobic membrane bioreactors

(SAMBRs). J. Chem. Technol. Biotechnol. 81, 1894-1904.

Austin, M.E., Kasturi, S.B., Barber, M., Kannan, K., MohanKumar, P.S. and

MohanKumar, S.M.J. (2003) Neuroendocrine effects of perfluorooctane

sulfonate in rats. Environ. Health Persp. 111, 1485-1489.

Bai, R. and Leow, H.F. (2002) Microfiltration of activated sludge wastewater-

the effect of system operation parameters. Sep. Purif. Technol. 29, 189-198.

Barker, D.J., Mannucchi, G.A., Salvi, S.M.L. and Stuckey, D.C. (1999a)

Characterization of soluble residual chemical oxygen demand (COD) in

anaerobic wastewater treatment effluents. Water Res. 33, 2499-2510.

Barker, D.J., Stuckey, D.C. (1999b) A review of soluble microbial products

(SMP) in wastewater treatment systems. Water Res. 33, 3063-3082.

Baumgarten, S., Schroeder, H., Charwath C., Lange M., Beier S., Pinnekamp,

J. (2007) Evaluation of advanced treatment technologies for the elimination of

pharmaceutical compounds. Water Sci. Technol. 56, 1-8.

Becker, A.M., Gerstmann, S. and Frank, H. (2008) Perfluorooctane surfactants

in waste waters, the major source of river pollution. Chemosphere 72, 115-121.

Benijts, T., Dams, R., Lambert, W. and De Leenheer, A. (2004) Countering

matrix effects in environmental liquid chromatography-electrospray ionization

tandem mass spectrometry water analysis for endocrine disrupting chemicals.

J. Chromatogr. A 1029, 153-159.

Berger, U. and Haukas, M. (2005) Validation of a screening method based on

liquid chromatography coupled to high-resolution mass spectrometry for

http://apps.isiknowledge.com.libproxy1.nus.edu.sg/OneClickSearch.do?product=UA&search_mode=OneClickSearch&db_id=&SID=N2nGMl57c8jmHn87fDg&field=AU&value=Beier%20S&ut=000251531600001&pos=5�

Reference

172

analysis of perfluoroalkylated substances in biota. J. Chromatogr. A 1081,

210-217.

Boudreau, T.M., Sibley, P.K., Mabury, S.A., Muir, D.C.G. and Solomon, K.R.

(2003) Laboratory evaluation of the toxicity of perfluorooctane sulfonate

(PFOS) on Selenastrum capricornutum, Chlorella vulgaris, Lemna gibba,

Daphnia magna and Daphnia pulicaria. Arch. Environ. Contam. Toxicol. 44,

307-313.

Bouhabila, E.H., Aїm, R.B. and Buisson, H. (2001) Fouling characterisation in

membrane bioreactors. Sep. Purif. Technol. 22-3, 123-132.

Boulanger, B., Vargo, J., Schnoor, J.L., Hornbuckle, K.C. (2004) Detection of

perfluorooctane surfactants in Great Lakes water. Environ. Sci. Technol. 38,

4064-4070.

Boulanger, B., Vargo, J.D., Schnoor, J.L. and Hornbuckle, K.C. (2005)

Evaluation of perfluorooctane surfactants in a wastewater treatment system

and in a commercial surface protection product. Environ. Sci. Technol. 39,

5524-5530.

Bouhabila, E. H., Aim, R. B., Buisson, H. (2001) Fouling characteristics in

membrane bioreactors. Sep. Purif. Technol. 22, 123-132.

Case, M.T., York, R.G., Christian, M.S. (2001) Rat and rabbit oral

developmental toxicology studies with two perfluorinated compounds. Int. J.

Toxicol. 20, 101-109.

Chang, I.S., Lee, C.H. (1998) Membrane filtration characteristics in

membrane-coupled activated sludge systems-The effect of physiological states

of activated sludge on membrane fouling. Desalination 120, 221-223.

Reference

173

Chang, I.S., Lee, C.H. and Ahn, K.H. (1999) Membrane filtration

characteristics in membrane coupled activated sludge system: The effect of

floc structure of activated sludge on membrane fouling. Sep. Sci. Technol. 34,

1743-1758.

Chang, I.S., Clech, P.L., Jefferson, B. and Judd, S. (2002) Membrane fouling

in membrane bioreactors for wastewater treatment. J. Environ. Eng.-ASCE

128, 1018-1029.

Childress, A.E., Elimelech, M. (1996) Effect of solution chemistry on the

surface charge of polymeric reverse osmosis and nanofiltration membranes. J.

Membr. Sci. 119, 253.

Churchouse, S.J. (1998) Kubota Maintains Popular Trend. Water and Environ.

International 26-27.

Clara, M., Kreuzinger, N., Strenn, B., Gans, O. and Kroiss, H. (2005a) The

solids retention time - a suitable design parameter to evaluate the capacity of

wastewater treatment plants to remove micropollutants. Water Res. 39, 97-106.

Clara, M., Strenn, B., Gans, O., Martinez, E., Kreuzinger, N.,and Kroiss, H.

(2005b) Removal of selected pharmaceuticals, fragrances and endocrine

disrupting compounds in a membrane bioreactor and conventional wastewater

treatment plants. Water Res. 39, 4797-4807.

Cooney, D.O. (1998) Adsorption design for wastewater treatment. Lewis

Publishers, NJ.

Deng, H., Peng, P., Huang, W., Song, J. (2004) Distribution and loadings of

polycyclic aromatic hydrocarbons in the Xijiang River in Guangdong, South

China. Chemosphere 64, 1401-1411.

Reference

174

Dinglasan, M.J.A., Ye, Y., Edwards, E.A. and Mabury, S.A. (2004)

Fluorotelomer alcohol biodegradation yields poly- and perfluorinated acids.

Environ. Sci. Technol. 38, 2857-2864.

Dosoretz C.G., Boddeker K.W. (2004) Removal of trace organics from water

using a pumped bed-membrane bioreactor with powdered activated carbon. J.

Membr. Sci. 239, 81-90.

DuPont (2003) Adsorption/Desorption of Ammonium Perfluorooctanoateto

Soil (OECD 106); E.I. du Pont de Nemours and Company: Wilmington, DE,

US EPA, Docket OPPT-2003-0012-040. U.S. Environmental Protection

Agency, Washington, DC.

Fleischer, E.J., Broderick, T.A., Daigger, G.T., Fonseca, A.D, Holbrook, R.D.,

Murthy, S.N. (2005) Evaluation of membrane bioreactor process capabilities

to meet stringent effluent nutrient discharge requirements. Water Environ. Res.

77, 162-178.

Gander, M., Jefferson, B., Judd, S. (2000) Aerobic MBRs for domestic

wastewater treatment: a review with cost considerations. Sep. Purif. Technol. 8,

119-130.

Giesy, J.P. and Kannan, K. (2001) Global distribution of perfluorooctane

sulfonate in wildlife. Environ. Sci. Technol. 35, 1339-1342.

Giesy, J.P. (2006) Perfluorinated compounds in streams of the Shihwa

industrial zone and lake Shihwa, South Korea. Environ. Toxicol. Chem. 25,

2374-2380.

Gilliland, F. (1992) Fluorochemicals and human health: Studies in an

occupational cohort. Doctoral thesis, Division of Environmental and

Occupational Health, University of Minnesota. US EPA, AR226-0473.

Reference

175

Grady Jr., C.P.L., Daigger, G.T., Lim, H.C. (1999) Biological Wastewater

Treatment, second ed. Marcel Dekker, New York.

Gunder, B. (2001) The membrane-coupled activated sludge process in

municipal wastewater treatment. Technomic, Lancaster, PA, USA.

Hansen, K.J., Clemen, L.A., Ellefson, M.E. and Johnson, H.O. (2001)

Compound-specific, quantitative characterization of organic: Fluorochemicals

in biological matrices. Environ. Sci. Technol.35, 766-770.

Hansen, K.J., Johnson, H.O., Eldridge, J.S., Butenho., 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.

Harada, K., Saito, N., Sasaki, K., Inoue, K., Koizumi, A. (2003)

Perfluorooctane sulfonate contamination of drinking water in the Tama River,

Japan: estimated effects on resident serum levels. Bull. Environ. Contam.

Toxicol. 17, 31-36.

Hatfield, T.L. (2001) Screening studies on the aqueous photolytic degradation

of perfluorooctanoic acid (PFOA), 3M Environmental Laboratory, USEPA

docket: EPA-HQ-OPPT-2002-0051-0023. U.S. Environmental Protection

Agency, Washington, DC.

Haughom, B., Spydevold, O. (1992) The mechanism underlying the

hypolipemic effect of perfluorooctanoic acid (PFOA), perfluorooctane

sulphonic acid (PFOSA) and clofibric acid. Biochim. Biophys. Acta. 1128, 65-

72.

Hekster, F., Pijnenburg, J.R.W, Laane, P.M. and Voogt, P.d. (2002)

Perfluoroalkylated substances: Aquatic environmental assessment, ministry of

Reference

176

transport, public works and water management, 99 pp, Report RIKZ 2002.043,

Hague, Netherlands.

Higgins, C.P., Field, J.A., Criddle, C.S., Luthy, R.G. (2005) Quantitative

determination of perfluorochemicals in sediments and domestic sludge.


Higgins, C.P. and Luthy, R.G. (2006) Sorption of perfluorinated surfactants on

sediments. Environ. Sci. Technol.40(23), 7251-7256.

Hof, P. T., Scheirs, J., van de Vijver, K., van Dongen, W., Esmans, E. L.,

Blust, R., de Coen, W. (2004) Biochemical Effect Evaluation of

Perfluorooctane Sulfonic Acid-Contaminated Wood Mice (Apodemus

sylvaticus). Environ. Health Perspect. 112, 681-686.

Houde, M., Martin, J.W., Letcher, R.J., Solomon, K.R., Muir, D.C.G. (2006)

Biological monitoring of polyfluoroalkyl substances: a review. Environ. Sci.

Technol. 40, 3463-3473.

Hu, J.Y., Jin, X, Ong, S.L. (2007) Rejection of estrone by nanofiltration:

Influence of solution chemistry. J. Membr. Sci. 302, 188-196.

Huang, X., Liu, R. and Qian, Y. (2000) Behaviour of soluble microbial

products in a membrane bioreactor. Process Biochem. 36, 401-406.

Huang, X., Gui, P. and Qian, Y. (2001) Effect of sludge retention time on

microbial behaviour in a submerged membrane bioreactor. Process Biochem.

36, 1001-1006.

Huset, C.A., Chiaia, A.C., Barofsky, D.F., Jonkers, N., Kohler, H.E., Ort, C.,

Giger, D.W., Field, J.A. (2008) Occurrence and mass flows of fluorochemicals

in the Glatt Valley watershed, Switzerland. Environ. Sci. Technol. 42, 6369-

6377.

Reference

177

Irvine, K.N., Loganathan, B.G., Pratt, E.J. and Sikka, H.C. (1993) Calibration

of PCSWMM to estimate metals, PCBs, and HCB in CSOs from an Industrial

sewershed. In. New Techniques for Modelling the Management of Stormwater

Quality Impacts. (Ed. W. James). Lewis Publ. pp 215-242.

Iversen V, Mehrez R, Horng RY. (2009) Fouling mitigation through

flocculants and adsorbents addition in membrane bioreactors: Comparing lab

and pilot studies. J. Membr. Sci. 345, 21-30.

Jahnke, A., L. Ahrens, R. Ebinghaus and C. Temme (2007). Urban versus

Remote Air Concentrations of Fluorotelomer Alcohols and Other

Polyfluorinated Alkyl Substances in Germany. Environ. Sci.Technol. 41, 745-

752.

Jin, YH, Liu, W, Sato, I, Nakayama SF, Sasaki K, Saito N, Tsuda S . (2008)

PFOS and PFOA in environmental and tap water in China. Chemosphere 77,

605-611.

Kallenborn, R., Berger, U., Järnberg, U. (2004) Perfluorinated Alkylated

Substances (PFAS) in the Nordic Environment, Norwegian Institute for Air

Research, Oslo, Norway, 107 pp.

Kannan, K., Newsted, J.L., Halbrook, R.S., Giesy, J.P. (2002) Perfluorooctane

sulfonate and Related Fluorinated Hydrocarbons in Mink and River Otters

from the United States. Environ. Sci. Technol. 36, 2566-2571.

Kannan, K., Corsolini, S., Falandysz, J., Fillmann, G., Kumar, K.S.,

Loganathan, B.G., Mohd, M.A., Olivero, J., Van Wouwe, N., Yang, J.H. and

Aldous, K.M. (2004) Perfluorooctanesulfonate and related fluorochemicals in

human blood from several countries. Environ. Sci. Technol. 38, 4489-4495.

Kaiser, M.A., Larsen, B.S., Kao, C.-P.C., Buck, R.C. (2005) Vapor Pressures

Reference

178

of Perfluoro -Octanoic, -Nonanoic, -Decanoic, -Undecanoic, and -Dodecanoic

acids. J. Chem. Eng. Data published online Aug. 26,

http://dx.doi.org/10.1021/je050070r.

Katsoyiannis, A. and Samara, C. (2005) Persistent organic pollutants (POPs)

in the conventional activated sludge treatment process: fate and mass balance,

Environ. Res. 97, 245-257.

Key, B.D., Howell, R.D. and Criddle, C.S. (1997) Fluorinated organics in the

biosphere. Environ. Sci. Technol. 31, 2445-2454.

Kilduff, J.E., Karanfil, T. and Weber, Jr. W.J. (1998) TCE adsorption by GAC

preloaded with humic substances, J. Am. Water Works Assoc. 90, 76-89.

Kim, J.S., Lee, C.H. and Chang, H.D. (1998) Comparison of ultrafiltration

characteristics between activated sludge and BAC sludge. Water Res. 32,

3443-3451.

Kim, J.S., Lee, C.H. and Chang, I.S. (2001) Effect of pump shear on the

performance of a crossflow membrane bioreactor. Water Res. 35, 2137-2144.

Kimura, K., Amy, G., Drewes, J. E., Heberer, T., Kim, T. U., Watanabe, Y.

(2003) Rejection of organic micropollutants (disinfection by-products,

endocrine disrupting compounds, and pharmaceutically active compounds) by

NF/RO membranes. J. Membr. Sci. 227, 113.

Kimura, K., Hara, H., Watanabe, Y. (2007) Elimination of selected acidic

pharmaceuticals from municipal wastewater by an activated sludge system and

membrane bioreactors. Environ. Sci. Technol. 41, 3708-3714.

Kiso, Y., Sugiura, Y., Kitao, T., Nishimura, K. (2001) Effects of

hydrophobicity and molecular size on rejection of aromatic pesticides with

nanofiltration membranes. J. Membr. Sci. 192, 1-9.

Reference

179

Kissa, E. (2001) Fluorinated surfactants and repellents, second ed. Marcel

Dekker, New York.

Kuchler, I.L. and Miekeley, N. (1994) Ultrafiltration of humic compounds

through low molecular mass cut-off level membranes. Sci. Total Environ. 154,

23-28.

Kudo, N. and Y. Kawashima (2003). Toxicity and toxicokinetics of

perfluorooctanoic acid in humans and animals. J. Toxicol. Sci. 2, 49-57.

Lange, C.C. (2000) 3M Environmental Laboratory. The aerobic

biodegradation of N-EtFOSE alcohol by the microbial activity present in

municipal wastewater treatment sludge; Report CA058, Docket AR-226-

1030a078; U.S. Environmental Protection Agency: Washington, DC.

Lange, C.C. (2002) 3M Environmental Laboratory. Biodegradation screen

study for telomer type alcohols; Docket AR226-1149; U.S. Environmental

Protection Agency: Washington, DC.

Lau, C., Thibodeaux, J.R., Hanson, R.G., Rogers, J.M., Grey, B.E., Stanton,

M.E., Butenhoff, J.L. and Stevenson L.A. (2003) Exposure to perfluorooctane

sulfonate during pregnancy in rat and mouse. II: Postnatal evaluation, Toxicol.

Sci. 74, 382-392.

Lau, C., Butenhoff, J.L. and Rogers, J.M. (2004) The developmental toxicity

of perfluoroalkyl acids and their derivatives. Toxicol. Appl. Pharmacol. 198,

231-241.

Le-Clech, P., Jefferson B., Chang, I.S., Judd, S.J. (2003) Critical flux

determination by the flux-step method in a submerged membrane bioreactor. J.

Membr. Sci. 227, 81-93.

Lee, K.M., Lim, P.E. (2005) Bioregeneration of powdered activated carbon in

Reference

180

the treatment of alkyl-substituted phenolic compounds in simultaneous

adsorption and biodegradation process. Chemosphere 58, 407-416.

Lesage, N., Sperandio, M., Cabassud, C. (2007) Study of a hybrid process:

adsorption on activated carbon/membrane bioreactor for the treatment of an

industrial wastewater. Chem. Eng. Process 47, 303-307.

Li, Y.Z., He, Y.L., Liu, Y.H., Yang, S.C., Zhang, G.J. (2005) Comparison of

the filtration characteristics between biological powdered activated carbon

sludge and activated sludge in submerged membrane bioreactors. Desalination

174, 305-314.

Liao, B.Q., Allen, D.G., Droppo, I.G., Leppard, G.G. and Liss, S.N. (2001)

Surface properties of sludge and their role in bioflocculation and settleability.

Water Res. 35, 339-350.

Loewen, M., Halldorson, T., Wang, F., Tomy, G. (2005) Fluorotelomer

Carboxylic Acids and PFOS in Rainwater from an Urban Center in Canada.


Löfstrand, K., Jörundsdóttir, H., Tomy, G., Svavarsson, J., Pál Weihe, Torgeir

Nygård, Åke Bergman. (2008) Spatial trends of polyfluorinated compounds in

guillemot (Uria aalge) eggs from North-Western Europe. Chemosphere 72,

1475-1480

Logan, B.E. and Jiang, Q. (1990) Molecular size distributions of dissolved

organic matter. J. Environ. Eng. Div. ASCE 116, 1046-1062.

Loganathan, B.G., Irvine, K.N., Kannan, K. Pragatheeswaran, V. and Sajwan,

K.S. (1997) Distribution of selected PCB congeners in the Babcock Street

Sewer District: A multimedia approach to identify PCB sources in combined

Reference

181

sewer overflows (CSOs) discharging to the Buffalo River, New York. Arch.

Environ. Contam. Toxicol. 33, 130-140.

Loganathan, B.G., Sajwan, K.S., Sinclair, E., Kumar, K.S. and Kannan, K.

(2007) Perfluoroalkyl sulfonates and perfluorocarboxylates in two wastewater

treatment facilities in Kentucky and Georgia. Water Res. 41, 4611-4620.

Loos, R., Wollgast, J., Huber, T., Hanke, G. (2007) Polar herbicides,

pharmaceutical products, perfluorooctanesulfonate (PFOS),

perfluorooctanoate (PFOA), and nonylphenol and its carboxylates and

ethoxylates in surface and tap waters around Lake Maggiore in Northern Italy.

Anal. Bioanal. Chem. 387, 1469-1478.

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.

Martin, JW, Mabury, SA, Solomon, KR and Muir, DCG (2003a). Dietary

accumulation of perfluorinated acids in juvenile rainbow trout (Oncorhynchus

mykiss). Environ. Toxicol. Chem. 22, 189-195.

Martin, JW, Mabury, SA, Solomon, KR and Muir, DCG (2003b).

Bioconcentration and tissue distribution of perfluorinated acids in rainbow

trout (Oncorhynchus mykiss). Environ. Toxicol. Chem. 22, 196-204.

Martin, J.W., Kannan, K., Berger, U., De Voogt, P., Field, J., Franklin, J.,

Giesy, J.P., Harner, T., Muir, D.C.G., Scott, B., Kaiser, M., Jarnberg, U.,

Jones, K.C., Mabury, S.A., Schroeder, H., Simcik, M., Sottani, C., Van Bavel,

B., Karrman, A., Lindstrom, G. and Van Leeuwen, S. (2004a) Analytical

http://apps.isiknowledge.com.libproxy1.nus.edu.sg/OneClickSearch.do?product=UA&search_mode=OneClickSearch&doc=1&db_id=&SID=Y1cd6pb6n1@258PnJoo&field=AU&value=Hanke%20G&ut=000254020900013&pos=8�

http://apps.isiknowledge.com.libproxy1.nus.edu.sg/OneClickSearch.do?product=UA&search_mode=OneClickSearch&doc=1&db_id=&SID=Y1cd6pb6n1@258PnJoo&field=AU&value=Umlauf%20G&ut=000254020900013&pos=9�

Reference

182

challenges hamper perfluoroalky research. Environ. Sci. Technol. 38, 248A-

255A.

Martin, J.W., Whittle, D.M., Muir, D.C.G. and Mabury, S.A. (2004b)

Perfluoroalkyl contaminants in a food web from lake Ontario. Environ. Sci.

Technol. 38, 5379-5385.

Masse, A., Sperandio, M., Cabassud, C. (2006) Comparison of sludge

characteristics and performance of a submerged membrane bioreactor and an

activated sludge process at high solids retention time. Water Res. 40, 2405-

2415.

Matsui, Y., Fukuda, Y., Inoue, T., Matsushita, T. (2003) Effect of natural

organic matter on powdered activated carbon adsorption of trace contaminants:

characteristics and mechanism of competitive adsorption. Water Res. 37,

4413-4424.

Matuszewski, B.K. (2006) Standard line slopes as a measure of a relative

matrix effect in quantitative HPLC-MS bioanalysis. J. Chromatogr. B 830,

293-300.

Matuszewski, B.K., Constanzer, M.L. and Chavez-Eng, C.M. (2003)

Strategies for the assessment of matrix effect in quantitative bioanalytical

methods based on HPLC-MS/MS. Anal. Chem. 75, 3019-3030.

MDH (2007) Health Based Values for Perfluorooctanoic Acid (PFOA).

http://www.health.state.mn.us/divs/eh/hazardous/topics/pfcs/pfoamemo0307.p

df. U.S. Environmental Protection Agency, Washington, DC.

Moody, C.A., Field, J.A. (2000) Perfluorinated surfactants and the

environmental implications of their use in fire-fighting foams. Environ. Sci.

Technol. 34, 3864-3870.

Reference

183

Moody, C.A., Kwan, W.C., Martin, J.W., Muir, D.C.G. and Mabury, S.A.

(2001) Determination of perfluorinated surfactants in surface water samples

by two independent analytical techniques: Liquid chromatography/tandem

mass spectrometry and F-19 NMR. Anal. Chem. 73, 2200-2206.

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.

Munz, G., Gori, R., Mori, G., Lubello, C. (2007) Powdered activated carbon

and membrane bioreactors (MBR-PAC) for tannery wastewater treatment:

long term effect on biological and filtration process

performances. Desalination 207, 349-360.

Murakami, M., Shinohara, H., Takada, H. (2009) Evaluation of wastewater

and street runoff as sources of perfluorinated surfactants (PFSs). Chem. 74,

487-493.

National Environmental Agency (2008) Daily rainfall at 5 standard stations for

the period, January 2006 to December 2007. Ministry of the Environment and

Water Resources, Singapore.

Newcombe, G., Morrison, J., Hepplewhite, C., Knappe, D.R.U. (2002)

Simultaneous adsorption of MIB and NOM onto activated carbon: II.

Competitive effects. Carbon 40, 2147-2156.

North California Division of Water Quality (2006), Recommended interim

maximum allowable concentration for perfluorooctanoic acid. U.S.


Reference

184

Ng, C. A. , Sun, D. and Fane, A. G. (2006) Operation of Membrane Bioreactor

with Powdered Activated Carbon addition. Separation Sci. and Tech. 41,

1447-1466.

Ng, H.Y., Hermanowicz, S.W. (2005) Membrane bioreactor operation at short

solids retention times: Performance and biomass characteristics. Water Res.,

39, 981-992.

Nozoe, M., S. Fujii, S. Tanaka, H. Tanaka and N. Yamashita (2006).

Investigation of PFOS and PFOA in a wastewater treatment plant. Environ.

Eng. Res. 43, 105-111.

Ochoa-Herrera,V., Sierra-Alvarez, R. (2008) Removal of perfluorinated

surfactants by sorption onto granular activated carbon, zeolite and sludge.

Chemosphere 72, 1588-1593.

OECD (2002). Hazard Assessment of Perfluorooctane Sulfonate (PFOS) and

its Salts.

ENV/JM/RD(2002)17/FINAL. Organisation for Economic Co-operation and

Development, Paris, France.

Ohya, T., N. Kudo, E. Suzuki and Y. Kawashima (1998). Determination of

perfluorinated carboxylic acids in biological samples by high-performance

liquid chromatography. J. Chromatogr. B 720, 1-7.

Ooi, J.B. and Chia, L.S. (1974) The climate of west Malaysia and Singapore.

Oxford University Press.

Pelekani, C., Newcombe, G., Snoeyink, V., Hepplewhite, C., Assemi, S.,

Beckett, R. (1999) Characterization of natural organic matter using high

performance size exclusion chromatography. Environ Sci Technol. 33, 2807-

22813.

Reference

185

Poulsen, P.B., Jensen, A.A., Wallström, E. (2005) More environmentally

friendly alternatives to PFOS-compounds and PFOA. Environmental

Protection Agency, Danish Ministry of the Environment.

Petrovic, M., Hernando, M.D., Diaz-Cruz, M.S. and Barcelo, D. (2005) Liquid

chromatography-tandem mass spectrometry for the analysis of pharmaceutical

residues in environmental samples: a review. J. Chromatogr. A 1067, 1-14.

Powley, C.R., George, S.W., Ryan, T.W. and Buck, R.C. (2005) Matrix effect-

free analytical methods for determination of perfluorinated carboxylic acids in

environmental matrixes. Anal. Chem. 77, 6353-6358.

Prevedouros, K., Cousins, I.T., Buck, R.C., Korzeniowski, S.H. (2006)

Sources, fate and transport of perfluorocarboxylates. Environ. Sci. Technol. 40,

32-44.

Qiu, Y., Fujii, J., Tanaka, S., Lien, N.P.H., Koizumi, A., (2005). Removal

characteristics of perfluorooctane sulfonate and perfluorooctane acid by

granular activated carbon adsorption. Proc. of Joint 14th KU-KAIST-NTU-

NUS Sympo. in Env. Eng., 63-72.

Quinlivan, P.A, Li, L., Knappe, D.R.U. (2005) Effects of activated carbon

characteristics on the simultaneous adsorption of aqueous organic

micropollutants and natural organic matter. Water Res. 39, 1663-1673

Saito, N., Sasaki, K., Nakatome, K., Harada, K., Yoshinaga, T., Koizumi, A.

(2003) Perfluorooctane sulfonate concentrations in surface water in Japan.

Arch. Environ. Contam. Toxicol. 45, 149-158.

Satyawali, Y. and Balakrishnan, M. (2009) Effect of PAC addition on sludge

properties in an MBR treating high strength wastewater. Water Res. 43, 1577-

1588.

Reference

186

Schafer, A. I., Nghiem, L. D., Waite, T. D. (2003) Removal of the natural

hormone estrone from aqueous solutions using nanofiltration and reverse

osmosis. Environ. Sci. Technol. 37, 182.

Schultz, M.M., Barofsky D.F. and Field J.A. (2003). Fluorinated alkyl

surfactants. Environ. Eng. Sci. 20, 487-501.

Schultz, M.M., Barofsky, D.F., Field, J.A. (2004) Quantitative determination

of fluorotelomer sulfonates in groundwater by LC MS/MS. Environ. Sci.

Technol. 38, 1828-1835.

Schultz, M.M., Barofsky, D.F. and Field, J.A. (2006a) Quantitative

determination of fluorinated alkyl substances by large-volume-injection liquid

chromatography tandem mass spectrometry - Characterization of municipal

wastewaters. Environ. Sci. Technol. 40, 289-295.

Schultz, M.M., Higgins, C.P., Huset, C.A., Luthy, R.G., Barofsky, D.F. and

Field, J.A. (2006b) Fluorochemical mass flows in a municipal wastewater

treatment facility. Environ. Sci. Technol. 40, 7350-7357.

Schwarzenbach, R.P., Gschwend, P.M., Imboden, D.M. (2003) Environmental

Organic Chemistry, 2nd ed.; Wiley: Hoboken, NJ.

Scott, B.F., Spencer, C., Mabury, S.A., Muir, D.C.G. (2006) Poly and

perfluorinated carboxylates in North American precipitation. Environ. Sci.

Technol. 40, 7167-7174.

Seacat, A.M., Thomford, P.J., Hansen, K.J. Clemen, L.A., Eldridge, S.R.,

Elcombe, C.R., and Butenhoff J.L. (2003) Sub-chronic dietary toxicity of

potassium perfluorooctanesulfonate in rats. Toxicology 183, 117-131.

Reference

187

Shimizu, Y., Okuno, Y.I., Uryu, K., Ohtsubo S., Watanabe, A. (1996)

Filtration characteristics of hollow fiber microfiltration membranes used in

membrane bioreactor for domestic wastewater. Water Res. 30, 2385-2392.

Shon, H.K., Vigneswaran, S., Snyder, S.A. (2006) Effluent organic matter

(EfOM) in wastewater: Constituents, effects, and treatment. Crit. Rev. Env. Sci.

Tec. 36, 327-374.

Skutlarek, D., M. Exner and H. Färber (2006). Perfluorinated surfactants in

surface and drinking waters, Environ. Sci. Pollut. Res. Int. 13, 299-307.

Shoeib, M., T. Harner and P. Vlahos (2006). Perfluorinated Chemicals in the

Arctic Atmosphere. Environ. Sci. Technol. 40, 7577-7583.

Simcik, M.F. and Dorweiler, K.J. (2005) Ratio of perfluorochemical

concentrations as a tracer of atmospheric deposition to surface waters. Environ.

Sci. Technol. 39, 8678-8683.

Sinclair, E. and Kannan, K. (2006a) Mass loading and fate of perfluoroalkyl

surfactants in wastewater treatment plants. Environ. Sci. Technol. 40, 1408-

1414.

Sinclair, E., Mayack, D.T., Roblee, K., Yamashita, N., Kannan, K. (2006b)

Occurrence of perfluoroalkyl surfactants in water, fish, and birds from New

York State. Arch. Environ. Contam. Toxicol. 50, 398-410.

So, M.K., Taniyasu, S., Yamashita, N., Giesy, J.P., Zheng, J., Fang, Z., Im,

S.H. and Lam, P.K.S. (2004) Perfluorinated compounds in coastal waters of

Hong Kong, South China, and Korea. Environ. Sci. Technol. 38, 4056-4063.

So, M. K., Miyake, Y., Yeung, W. Y, Ho, Y. M, Taniyasu, S., Rostkowski, P.,

Yamashita, N, Zhou, B. S., Shi, X. J., Wang, J. X., Giesy, J. P, Yu, H., Lam, P.

Reference

188

K. S. (2007) Perfluorinated compounds in the Pearl River and Yangtze River

of China. Chemosphere 68, 2085-2095.

Stock, N.L., Lau, F.K., Ellis, D.A., Martin, J.W., Muir, D.C.G., Mabury, S.A.

(2004) Polyfluorinated telomer alcohols and sulfonamides in the north

American troposphere. Environ. Sci. Technol. 38, 991-996.

Szostek, B. and Prickett, K.B. (2004) Determination of 8:2 fluorotelomer

alcohol in animal plasma and tissues by gas chromatography-mass

spectrometry. J. Chromatogr. B 813, 313-321.

Tang, C.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, 7342-7349.

Taniyasu, S., Kannan, K., Horii, Y., Hanari, N. and Yamashita, N. (2003) A

survey of perfluorooctane sulfonate and related perfluorinated organic

compounds in water, fish, birds, and humans from Japan. Environ. Sci.

Technol. 37, 2634-2639.

Taniyasu, S., Kannan, K., So, M.K., Gulkowska, A., Sinclair, E., Okazawa, T.

and Yamashita, N. (2005) Analysis of fluorotelomer alcohols, fluorotelorner

acids, and short- and long-chain perfluorinated acids in water and biota. J.

Chromatogr. A 1093, 89-97.

Taves, D. (1968) Evidence that there are two forms of fluoride in human

serum, Nature 217, 1050-1051.

Thibodeaux, J.R., Hanson, R.G., Rogers, J.M., Grey, B.E., Barbee, B.D.,

Richards, J.H., Butenhoff, J.L, Stevenson, L.A. and Lau, C. (2003). Exposure

to perfluorooctane sulfonate during pregnancy in rat andmouse. I: Maternal

and prenatal evaluations. Toxicol. Sci. 74, 369-381.

Reference

189

Tomy, G.T., Budakowski, W., Halldorson, T., Helm, P.A., Stern, G.A.,

Friesen, K., Pepper, K., Tittlemier, S.A. and Fisk, A.T. (2004) Fluorinated

organic compounds in an eastern Arctic marine food web. Environ. Sci.

Technol. 38, 6475-6481.

Tseng, C.L., Liu, L.L., Chen, C.M. and Ding, W.H. (2006) Analysis of

perfluorooctanesulfonate and related fluorochemicals in water and biological

tissue samples by liquid chromatography-ion trap mass spectrometry. J.

Chromatogr. A 1105, 119-126.

US EPA (2002) Draft hazard assessment of perfluorooctanoic acid and its salts,

February 20, 2002. U.S. Environmental Protection Agency, Washington, DC.

US EPA (2003) Preliminary risk assessment of the developmental toxicity

associated with exposure to perfluorooctanoic acid and its salts. Office of

Pollution Prevention and Toxics, Risk Assessment Division. U.S.


US EPA (2005) Preliminary Risk Assessment of the Developmental Toxicity

Associated with Exposure to Perfluorooctanoic Acid and its Salts. Office of

Pollution Prevention and Toxics, Risk Assessment Division, U.S.


US EPA (2009) Provisional Health Advisories for Perfluorooctanoic Acid

(PFOA) and Perfluorooctane Sulfonate(PFOS).

http://www.epa.gov/waterscience/criteria/drinking

/pha-PFOA_PFOS.pdf. U.S. Environmental Protection Agency, Washington,

DC.

van De Steene, J.C., Mortier, K.A. and Lambert, W.E. (2006) Tackling matrix

effects during development of a liquid chromatographic-electrospray

Reference

190

ionisation tandem mass spectrometric analysis of nine basic pharmaceuticals

in aqueous environmental samples. J. Chromatogr. A 1123, 71-81.

van Leeuwen, S.P.J., Karrman, A., Van Bavel, B., De Boer, J. and Lindstrom,

G. (2006) Struggle for quality in determination of perfluorinated contaminants

in environmental and human samples. Environ. Sci. Technol. 40, 7854-7860.

van Leeuwen, S.P.J. and de Boer, J. (2007) Extraction and clean-up strategies

for the analysis of poly- and perfluoroalkyl substances in environmental and

human matrices. J. Chromatogr. A 1153, 172-185.

Vigneswaran, S., Kwon, D.Y., Ngo, H.H., Hu, J.Y. (2000) Improvement of

microfiltration performance in water treatment: is critical flux a viable solution?

Water Sci. Technol. 41, 309-315.

Visvanathan, C., Aim, R. B. and Parameshwaran, K. (2000) Membrane

separation

bioreactors for wastewater treatment. Crit. Rev. Environ. Sci. Technol. 30, 1-

48.

Wang, N., Szostek, B., Folsom, P.W., Sulecki, L.M., Capka, V., Buck, R.C.,

Berti, W.R. and Gannon, J.T. (2005) Aerobic biotransformation of C-14-

labeled 8-2 telomer B alcohol by activated sludge from a domestic sewage

treatment plant. Environ. Sci. Technol. 39, 531-538.

Weremiuk, A.M., Gerstmann, S., Frank, H. (2006) Quantitative determination

of perfluorinated surfactants in water by LC-ESI-MS/MS. J. Sep. Sci. 29,

2251-2255.

West Virginia Department of Environmental Protection (2002). Final

ammonium perfluorooctanoate (C8) Assessment of Toxicity Team (CATT)

report. U.S. Environmental Protection Agency, Washington, DC.

Reference

191

Xing, C.H., Tardieu, E., Qian, Y. and Wen, X.H. (2000) Ultrafiltration

membrane bioreactor for urban wastewater reclamation. J. Membr. Sci. 177,

73-82.

Xing, C.H., Wen, X.H., Qian, Y. and Tardieu, E. (2001) Microfiltration-

membrane coupled bioreactor for urban wastewater reclamation. Desalination

141, 63-73.

Yamashita, N., K. Kannan, S. Taniyasu, Y. Horii, T. Okazawa, G. Petrick and

T. Gamo (2004). Analysis of perfluorinated acids at parts-per-quadrillion

levels in seawater using liquid chromatography-tandem mass spectrometry,

Environ. Sci. Technol., 38, 5522-5528.

Yamashita, N., Kannan, K., Taniyasu, S., Horii, Y., Petrick, G., Gamo, T.

(2005) A global survey of perfluorinated acids in oceans. Mar. Pollut. Bull. 51,

658-668.

Yu, J., Hu, J.Y., Tanaka, S., Fujii, S. (2009a) Perfluorooctane sulfonate (PFOS)

and perfluorooctanoic acid (PFOA) in sewage treatment plants. Water Res. 43,

2399-2408.

Yu Q, Zhang R, Deng S, Huang J, Yu G. (2009b) Sorption of perfluorooctane

sulfonate and perfluorooctanoate on activated carbons and resin: kinetic and

isotherm study. Water Res. 43, 1150-1158.

Zhao, X., Li, J., Shi, Y., Cai, Y., Mou S. and Jiang G. (2007). Determination

of perfluorinated compounds in wastewater and river water samples by mixed

hemimicelle-based solid-phase extraction before liquid chromatography-

electrospray tandem mass spectrometry detection. J. Chromatogr. A 1154, 52-

59.

Reference

192

Zushi, Y., Takeda, T., Masunaga, S. (2008) Existence of nonpoint source of

erfluorinated compounds and their loads in the Tsurumi River basin, Japan.

Chemosphere 71, 1566-1573.

http://apps.isiknowledge.com.libproxy1.nus.edu.sg/OneClickSearch.do?product=UA&search_mode=OneClickSearch&doc=1&db_id=&SID=Y1cd6pb6n1@258PnJoo&field=AU&value=Zushi%20Y&ut=000255384300019&pos=1�

Appendix: Publications

193

APPENDIX: PUBLICATIONS

Journal Paper:

1. Yu, J., Hu, J.Y., Tanaka, S., Fujii, S. (2009) Perfluorooctane sulfonate

(PFOS) and perfluorooctanoic acid (PFOA) in sewage treatment plants. Water

Res. 43, 2399-2408.

2. Hu, J.Y. and Yu, J. (2010) Development and Validation of a LC-MS-MS

method for the Determination of perfluorinated compounds in environmental

matrices. Chromatographia. 72, 411-416.

3. Hu, J.Y., Yu, J., Tanaka, S., Fujii, S. (2011) Perfluorooctane sulfonate

(PFOS) and perfluorooctanoic acid (PFOA) in water environment of

Singapore. Water Air Soil Pollut. 216，179-191.

4. Yu, J. and Hu, J.Y. (2010) Adsorption of perfluorinated compounds onto

activated carbon and activated sludge. (submitted to J Environ. Eng.)

Conference Paper:

1. Yu, J. and Hu, J.Y. (2007) Occurrence of pharmaceuticals in treated sewage

of local WRPs. The 16th Joint KAIST-KYOTO-NTU-NUS Symposium on

Environmental Engineering, Taiwan, pp.206-215.

2. Hu, J.Y. and Yu, J. (2007) Occurrence of perfluorooctane sulfonate (PFOS)

and perfluorooctanoic acid (PFOA) in water and wastewater of Singapore. The

16th Joint KAIST-KYOTO-NTU-NUS Symposium on Environmental

Engineering, Taiwan, pp.129-139.

Date post:	28-Apr-2021
Category:	Documents
Upload:	others
View:	3 times
Download:	0 times

OCCURRENCE AND FATE OF PERFLUOROOCTANE SULFONATE (PFOS) AND PERFLUOROOCTANOIC … · 2018. 1....

Documents