STUDY OF N-NITROSODIMETHYLAMINE (NDMA)

FORMATION POTENTIAL AND ITS REMOVAL IN

TREATED EFFLUENT BY UV/H2O2 PROCESS

WANG QI

B.Eng. (Hons.), NUS

A THESIS SUBMITTED

FOR THE DEGREE OF MASTER OF ENGINEERING

DEPARTMENT OF CIVIL AND ENVIRONMENTAL

ENGINEERING

NATIONAL UNIVERSITY OF SINGAPORE

2013

DECLARATION

I hereby declare that this thesis is my original work and it has been written by me in its

entirety. I have duly acknowledged all the sources of information which have been used

in the thesis.

This thesis has also not been submitted for any degree in any university previously.

___________________________

Wang Qi

22 March 2013

i

ACKNOWLEDGEMENTS

I would like to express my sincere thanks and gratitude to Associate Professor HU

Jiangyong for her continuous guidance, supervision and encouragement throughout the

master degree study. Her timely and insightful advice led me to this completion of the

dissertation.

I would as well thank research fellow CHU Xiaona, LIM Fang Yee for their help and

advice throughout the research; to Ms TAN Xiaolan, Ms LEE Leng Leng and Mr S.G.

Chandrasegaran for their assistance and guidance on experiment set-up, sample analysis

and chemical purchase; and to fellow research students for their companionship and

assistance.

Finally I would like to thank my parents for their great support throughout my study and

my closest friends for their inspiration and encouragement.

ii

TABLE OF CONTENTS

SUMMARY ....................................................................................................................... iv

LIST OF TABLES ............................................................................................................. vi

LIST OF FINGURES ....................................................................................................... vii

LIST OF PLATES .............................................................................................................. x

NOMENCLATURES ........................................................................................................ xi

CHAPTER 1 INTRODUCTION ........................................................................................ 1

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

1.2 Project objectives and scope of study ....................................................................... 3

1.3 Outline of the thesis .................................................................................................. 4

CHAPTER 2 LITERATURE REVIEW ............................................................................. 5

2.1 NDMA as emerging DBPs ........................................................................................ 5

2.1.1 Background ........................................................................................................ 5

2.1.2 NDMA formation ............................................................................................... 7

2.1.3 NDMA occurrence in water environment .......................................................... 9

2.1.4 NDMA precursors ............................................................................................ 11

2.1.5 Current research on NDMA removal methods ................................................ 12

2.2 Dissolved organic matter (DOM) ........................................................................... 13

2.2.1 DOM characterization ...................................................................................... 13

2.2.2 Correlation between NDMAFP and DOM ...................................................... 18

2.2.3 DOM removal .................................................................................................. 19

2.3 UV/H2O2 AOP process ........................................................................................... 20

2.3.1 AOP background .............................................................................................. 20

2.3.2 UV/H2O2 mechanism ....................................................................................... 22

2.3.3 Key affecting factors ........................................................................................ 22

2.3.4 NDMA precursor removal by UV/H2O2 .......................................................... 28

CHAPTER 3 MATERIALS AND METHODS................................................................ 29

3.1 Introduction ............................................................................................................. 29

3.2 Experiment design part 1: NDMA formation potential and precursors study ........ 29

3.2.1 Experiment design ........................................................................................... 29

3.2.2 Experiment conditions ..................................................................................... 30

3.3 Experiment design part 2: UV/H2O2 effect on DOM and NDMAFP removal ....... 32

3.3.1 System set-up ................................................................................................... 32

iii

3.3.2 Wastewater characteristics ............................................................................... 35

3.3.3 Experiment design ........................................................................................... 35

3.4 Measurement and analysis method ......................................................................... 36

3.4.1 NDMA detection and analysis ......................................................................... 36

3.4.2 Free chlorine and total chlorine ....................................................................... 38

3.4.3 TOC and TN .................................................................................................... 39

3.4.4 Nitrate, Nitrite and Ammonium ....................................................................... 40

3.4.5 LC-OCD-OND ................................................................................................. 40

3.4.6 DON ................................................................................................................. 41

CHAPTER 4 RESULTS AND DISCUSSIONS ............................................................... 42

4.1 NDMAFP study results ........................................................................................... 42

4.1.1 NDMA formation kinetics and formation potential ......................................... 42

4.1.2 Organic precursor analysis ............................................................................... 45

4.1.3 Relationship between organic precursors and NDMAFP ................................ 49

4.2 Results of UV/H2O2 system I .................................................................................. 53

4.2.1 DOM removal .................................................................................................. 53

4.2.2 NDMAFP removal ........................................................................................... 59

4.3 Results of UV/H2O2 system II................................................................................. 60

4.3.1 DOM removal .................................................................................................. 60

4.3.2 NDMAFP removal ........................................................................................... 69

CHAPTER 5 CONCLUSIONS AND RECOMMENDATIONS ..................................... 77

5.1 Conclusions ............................................................................................................. 77

5.2 Recommendations ................................................................................................... 78

5.2.1 Improvement on NDMA detection .................................................................. 78

5.2.2 Improvement on UV/H2O2 process .................................................................. 79

5.2.3 Study on wastewater characteristics ................................................................ 79

BIBLIOGRAPHY ............................................................................................................. 80

APPENDICES .................................................................................................................. 84

Appendix 1 List of toxic DBPs ..................................................................................... 84

Appendix 2 Table of potential NDMA precursors ........................................................ 88

Appendix 3 Membrane bio-reactor specification .......................................................... 93

Appendix 4 Activated sludge process specification ..................................................... 94

iv

SUMMARY

NDMA formation from treated wastewater is one of the concerns in water reuse. NDMA

is a carcinogenic compound and can be formed at a level significantly higher than the

level in drinking water quality guidelines. In order to better control NDMA in water, it is

important to study NDMA formation potential (NDMAFP) and NDMAFP removal

technology. This study focused on NDMAFP in wastewater treated effluent and

NDMAFP removal by UV/H2O2 technology.

Experiments were carried out to measure NDMAFP and characterize NDMA organic

precursors. This aimed to investigate relationship between organic precursors and

NDMAFP. Followed by this, UV/H2O2 system was designed and tested on the removal

effect on both organic precursors and NDMAFP. Two types of wastewater treated

effluents were selected to represent effluents with different organic compositions:

activated sludge process (ASP) effluent and membrane bio-reactor (MBR) effluent.

Experiment results showed similar NDMA formation kinetics between ASP and MBR

effluents, with rapid increase in first two days followed by a plateau. NDMAFP was

between 240 to 400ng/L. Results indicated no correlation (R2 = 0.16) between dissolved

organic nitrogen (DON) and NDMAFP. Weak correlation (R2

= 0.45) was however

observed between DON in molecules with molecular weight smaller than 500Da and

NDMAFP. In batch experiment, UV/H2O2 system consisted of low pressure ultra-violet

(LPUV) with intensity of 2mW/cm2 and H2O2 dosage at the magnitude of 100ppm.

Results showed efficient dissolved organic carbon (DOC) removal, with 70% removal in

one hour. DON removal was less efficient, with 30% removal. 80% of NDMAFP in ASP

v

effluent was removed within one hour. This showed potential feasibility of applying

UV/H2O2 to remove NDMAFP in wastewater treated effluent. However, no NDMAFP

removal was discovered in MBR effluent. This indicated that NDMAFP removal was

water specific. During UV/H2O2 treatment of ASP effluent, generation of intermediate

NDMA precursors was observed. No correlation between DOC (R2 = 0.02) or DON (R

2 =

0.002) and NDMAFP was discovered. This brought the attention that DOC or DON

cannot be used as an indicator for NDMAFP during UV/H2O2 treatment. At the same

time, sufficient oxidation should be provided to lower intermediate NDMA precursors

and to achieve NDMAFP removal.

vi

LIST OF TABLES

Table 2.1 Physico-Chemical properties of NDMA ............................................................. 6

Table 2.2 Guideline level of NDMA ................................................................................ 10

Table 2.3 DON compounds in municipal wastewater effluent ......................................... 16

Table 2.4 Comparison of H2O2/O3, O3/UV, and H2O2/UV AOP processes ...................... 21

Table 2.5 Main difference between LP and MP mercury UV lamps ................................ 23

Table 3.1 Water quality parameters for wastewater treated effluents ............................... 31

Table 3.2 pH and UV transmittance for ASP and MBR effluents .................................... 35

Table 4.1 DOC and DON composition in each fraction in ASP and MBR effluents ....... 50

Table 4.2 Comparison of DOC and DON composition in ASP and MBR effluents ........ 66

vii

LIST OF FINGURES

Figure 2.1 Chemical structure of NDMA ........................................................................... 6

Figure 2.2 Typical NDMA concentrations in various water bodies ................................... 9

Figure 2.3 Classification of DOM based on DOC fractionation ....................................... 14

Figure 2.4 Chromatogram of surface water (River Pfinz, Karlsruhe, Germany) with

responses for organic carbon detection (OCD), UV-detection at 254 nm (UVD) and

organic nitrogen detection (OND). A: biopolymers, B: humic substances, C: building

block, D: low molecular-weight acids, E: low molecular-weight neutrals (LMW), F:

nitrate and G: ammonium. ................................................................................................ 17

Figure 2.5 Emission spectrum of a MP-lamp ( ) and the absorptions of a commonly

pretreated natural water ( ) and a 10mg/L H2O2 solution (according to the

Lambert-Beer law) ( ). Composition of natural water: DOC content: 3.6 mg/L C;

[NO3-]: 11.2 mg/L ............................................................................................................. 24

Figure 2.6 Emission spectrum of a LP-lamp ( ) and the absorptions of a commonly

pretreated natural water ( ) and a 10mg/L H2O2 solution (according to the

Lambert-Beer law) ( ). Composition of natural water: DOC content: 3.6 mg/L C;

[NO3-]: 11.2 mg/L ............................................................................................................. 25

Figure 3.1 Schematic chart of experiment design part 1 ................................................... 30

Figure 3.2 Schematic diagram of UV/H2O2 system I ....................................................... 33

Figure 3.3 Schematic diagram of UV/H2O2 system II ...................................................... 33

Figure 4.1 NDMA formation vs reaction time for selected wastewater samples ............. 43

Figure 4.2 Comparison of NDMAFP of ASP and MBR effluents ................................... 44

Figure 4.3 NDMA formation study of secondary effluent in Australia (Farré et al., 2010)

.......................................................................................................................................... 45

Figure 4.4 OCD chromatographs of selected wastewater samples ................................... 47

Figure 4.5 OND chromatographs of selected wastewater samples ................................... 48

Figure 4.6 NDMAFP and normalized NDMAFP in different wastewater samples.......... 49

Figure 4.7 Correlation between NDMAFP and organic precursors in secondary effluents

.......................................................................................................................................... 51

Figure 4.8 Correlation between NDMAFP and DON associated with four fractions in

secondary effluents ........................................................................................................... 52

Figure 4.9 DOC degradation profiles of unfiltered ASP effluent under UV/H2O2 system I

with different H2O2 dosages .............................................................................................. 54

viii

Figure 4.10 Comparison of DOC degradation profiles of unfiltered and filtered ASP

effluents under UV/H2O2 system I ([H2O2] = 200 ppm) ................................................... 54

Figure 4.11 DOC degradation profiles of filtered ASP effluent over the extended reaction

time under UV/H2O2 system I ([H2O2] = 200 ppm) .......................................................... 55

Figure 4.12 OCD chromatography of filtered ASP effluent treated by UV/H2O2 system I

([H2O2] = 200 ppm) .......................................................................................................... 56

Figure 4.13 DON degradation profiles of filtered ASP effluent under UV/H2O2 system I

([H2O2] = 200 ppm) .......................................................................................................... 57

Figure 4.14 OND chromatography of filtered ASP effluent under UV/H2O2 system I

([H2O2] = 200 ppm) .......................................................................................................... 58

Figure 4.15 Effect of UV/H2O2 system I on NDMAFP in ASP effluent ([H2O2] = 200

ppm) .................................................................................................................................. 59

Figure 4.16 DOC degradation profiles of ASP effluent under UV/H2O2 system II .......... 62

Figure 4.17 DOC degradation profiles in four DOM fractions in ASP effluent under

UV/H2O2 system II ([H2O2] = 100 ppm) ........................................................................... 63

Figure 4.18 DON degradation profiles of ASP effluent under UV/H2O2 system II ......... 64

Figure 4.19 DON degradation profiles in four DOM fractions in ASP effluent under

UV/H2O2 system II ( [H2O2] = 100 ppm) .......................................................................... 65

Figure 4.20 Comparison of DOC and DON removal in ASP and MBR effluents ........... 67

Figure 4.21 DOC and DON degradation profiles for four DOM fractions in MBR effluent

under UV/H2O2 system II ................................................................................................. 68

Figure 4.22 NDMAFP of ASP effluent under UV/H2O2 system II .................................. 69

Figure 4.23 NDMAFP per DON of ASP effluent under UV/H2O2 system II ................... 70

Figure 4.24 Ratio of NDMAFP and initial NDMAFP of ASP effluent under UV/H2O2

system II ............................................................................................................................ 70

Figure 4.25 DOC degradation profiles in four DOM fractions in ASP effluent under

UV/H2O2 system II ([H2O2] = 50 ppm) ............................................................................. 72

Figure 4.26 DON degradation profiles in four DOM fractions in ASP effluent under

UV/H2O2 system II ( [H2O2] = 50 ppm) ............................................................................ 72

Figure 4.27 Comparison of DOC in four DOM fractions in ASP effluent with different

H2O2 dosages .................................................................................................................... 73

Figure 4.28 Comparison of DON in four DOM fractions in ASP effluent with different

H2O2 dosages .................................................................................................................... 74

Figure 4.29 NDMAFP of MBR effluent under UV/H2O2 system II ([H2O2] = 100 ppm) 76

ix

Figure 4.30 NDMAFP of MBR effluent under UV/H2O2 system II with different H2O2

dosages over 60 minutes ................................................................................................... 76

x

LIST OF PLATES

Plate 3.1 UV/H2O2 system I set-up ................................................................................... 34

Plate 3.2 UV/H2O2 system II set-up .................................................................................. 34

Plate 3.3 Experiment set up for SPE procedure ................................................................ 37

xi

NOMENCLATURES

AOP Advanced Oxidation Process

BB Building Block

BP Bio-polymers

DOC Dissolved Organic Carbon

DOM Dissolved Organic Matter

DON Dissolved Organic Nitrogen

GC Gas Chromatography

HS Humic Substance

LC Liquid Chromatography

LMW Low Molecular Weight

LPUV Low Pressure Ultraviolet

MPUV Medium Pressure Ultraviolet

MS Mass Spectrometry

NDMA N-nitrosodimethylamine

NDMAFP NDMA Formation Potential

SPE Solid Phase Extraction

1

CHAPTER 1 INTRODUCTIONRODUCTION

1.1 Background

Clean drinking water supply is a challenging issue for the society. Currently, water

shortage and pollution problem is affecting human health and development. With

population increase and urbanization, the problem is becoming severer. One of the

strategies to face the problem is to look for alternative water supply from water reuse and

reclamation (Fujioka et al., 2012). However, due to complex composition of wastewater,

it is important to ensure reclaimed water quality. In current practice, treated water is

discharged into natural environment before consumption to reduce contamination levels

(Water Reuse: Potential for Expanding the Nation's Water Supply through Reuse of

Municipal Wastewater (2012), 2012). Future approach is leading to directly restore

treated wastewater to portable quality (Shannon et al., 2008). Both current and future

situations require stringent regulation and reliable engineering system to ensure drinking

water quality.

NDMA is one of the emerging concerning compounds affecting drinking water quality.

This carcinogenic compound poses health threat to human at very low concentration:

according to USEPA, 0.7ng/L NDMA in drinking water yields 10-6

lifetime cancer risk.

First incident of NDMA in drinking water supply dated in 1998 in California. In this case,

high level of NDMA was detected in places where treated wastewater was reused

indirectly for potable purpose. It was later discovered that NDMA was produced from

chlorination of wastewater. To response, California Department of Public Health (CDPH)

established interim action level of 10ng/L (CDPH, 2011). Moreover, Ontario Ministry of

the Environment (MOE) in Canada set Interim Maximum Acceptable Concentration for

2

NDMA as 9ng/L (MOE, 2000). The World Health Organization (WHO) established a

guideline value of 100ng/L (WHO, 2006).

With stringent regulation on NDMA, it is important to study and therefore control

NDMA formation in treated wastewater. The challenges associated include

characterization of NDMA precursors and evaluation of NDMA removal technology.

Characterization of NDMA precursors is important for treatment process selection.

Dissolved organic matters (DOM) are the main precursors. NDMA pathway studies

discovered dimethylamine (DMA) as most effective organic nitrogen precursor (Mitch et

al., 2003b). Organic nitrogen compounds with DMA and DMA functional group in

wastewater also contribute to NDMA precursors. However, DMA is not the only source

(Mitch et al., 2003). Other precursors reported include dithiocarbamate and nitrogen-

containing cationic polyelectrolytes (Weissmahr & Sedlak, 2000). However, studies

reported that majority of NDMA precursors in treated effluent are unknown compounds

other than DMA (Mitch & Sedlak, 2004; Sedlak & Kavanaugh, 2005). Therefore it is

important to characterize wastewater derived DOM and to better understand NDMA

precursors.

NDMA removal consists of the removal of NDMA and NDMA precursor. Many studies

have addressed the issue of NDMA removal. Ultraviolet (UV) photolysis is one

established method, where UV light cleaves the N-N bond and breaks NDMA into nitrite

and DMA (U.S.EPA, 2008). In comparison, fewer studies have investigated treatment

process for NDMA precursor removal in wastewater treated effluent. Microfiltration

demonstrates moderate removal efficiency for NDMA precursors. Reverse osmosis can

reduce NDMA precursors by one order. Result is not available for UV effect on NDMA

3

precursor removal (Deeb et al., 2010). Preliminary research result has shown UV/H2O2

effect on NDMA precursor removal (Chen et al., 2010). However the performance may

be overestimated for wastewater treated effluent, due to experiments being performed on

surface water combined with wastewater treated effluent. The preliminary research only

chose one UV/H2O2 condition, different UV/H2O2 conditions and various treated

wastewater should be further studied.

1.2 Project objectives and scope of study

The main objective of this project is to study NDMA formation potential in treated

wastewater and to evaluate UV/H2O2 technology on NDMA precursor removal. The

study includes NDMA formation potential (NDMAFP) study, organic precursor

characterization and UV/H2O2 process evaluation.

The scope of the study is:

To establish NDMA detection method and to measure NDMAFP in treated

wastewater from two treatment processes

To characterize organic precursors and to study the relationship between organic

precursors and NDMAFP

To establish UV/H2O2 technology and to assess its removal effect on organic

precursors and NDMAFP in treated wastewater

4

1.3 Outline of the thesis

This thesis presents the study of NDMA formation potential and its removal in treated

wastewater. Chapter one and two illustrate concerns of NDMA in wastewater treatment,

characterization of organic precursors, and background of UV/H2O2 technology. Chapter

three presents methodology of the study and experiment design for both NDMAFP study

and UV/H2O2 technology evaluation. Chapter four presents experiment results and its

interpretation including NDMAFP for selected treated wastewater, correlation between

organic precursors and NDMAFP, UV/H2O2 effect on organic precursors and NDMAFP

removal. Chapter five concludes this study and presents recommendations for future

study on NDMAFP removal.

5

CHAPTER 2 LITERATURE REVIEW

2.1 NDMA as emerging DBPs

2.1.1 Background

Disinfection by-products (DBPs) are the compounds generated during disinfection

process in water treatment, when organic and inorganic matter in water reacts with

disinfectant. Depending on disinfection process, DBPs include chlorinated and non-

chlorinated DBPs. Many DBPs are toxic and have adverse effect on human health. A list

of regulated and unregulated toxic DBPs (Richardson et al., 2007) is included in

appendix one.

In the forty years history of DBPs, trihalomethane (THM) and haloacetic acid (HAA)

were first discovered. They were regulated firstly in the United States. Water treatment

processes have then changed. Source water is controlled. Coagulation is added before

disinfection and THM and HAA levels are controlled. Other alternatives are applying

chloramines or use other oxidants as primary disinfectants.

In the past decade, new problems emerge. This includes: nitrification, formation of other

DBPs, nitrosamines and mobilization of lead. The most distinguish study in DBP recently

is the discovery of emerging DBPs. NDMA is one emerging DBP resulted from

chlorination and chloramination. Compared with DBPs discovered forty years ago, these

emerging DBPs in water are of much lower concentration (2 to 3 magnitudes lower), but

of much higher danger to human health (2 to 3 magnitudes higher). As mentioned in

6

chapter one, for NDMA, concentration associated with 10-6

cancer rate is 0.7ng/l;

however, the 10-6

cancer rate associated with bromodichloromethane, which is the most

dangerous of early discovered THMs, is 600ng/l. Given the high severity, NDMA has

received attention in both scientific community and water authority.

NDMA is a semi-volatile organic chemical. Molecular formula is (CH3)2N-NO and its

molecular mass is 74.08g/mol. It is a hydrophilic, polar compound. It was previously

used in production of liquid rocket fuel, antioxidants and softeners for copolymers. Its

chemical structure (Plumlee et al., 2008) is shown in Fig 2.1. The physical and chemical

properties are presented in Table 2.1 (Farré et al., 2010).

Figure 2.1 Chemical structure of NDMA

Table 2.1 Physico-Chemical properties of NDMA

Property Value

CAS Number 62-75-9

Boiling point (oC) 151-154

Molecular weight (g/mol) 78.08

Hydrophobicity soil –Log KoC 1.079

Hydrophobicity water –Log Kow -0.57

Vapour pressure (Pa) 1,080 (25 oC)

Henry’s law constant (Pa m3/mol) 3.34 (25

oC)

Density (kg/L) 1.006 (20 oC)

Solubility in water (g/L) Miscible 100 (19 oC)

7

NDMA has been found in different media: air, water and soil. It is completely miscible in

water and does not adsorb to particles. This makes decontamination by adsorption

difficult. Other properties of NDMA include its high mobility in soil and probable

leaching into ground water. Under sunlight, NDMA breaks down quickly in air.

2.1.2 NDMA formation

Two possible formation pathways have been discovered: nitrosation and unsymmetrical

dimethylhydrazine (UDMH) oxidation (Mitch et al., 2003b). The former was observed in

NDMA formation in food production. The later was demonstrated to be responsible for

NDMA formation during water chlorination.

During nitrosation, nitrosyl cation is firstly formed from acidification of nitrite; then, it

reacts with dimethylamine to form NDMA. This is illustrated in the two reaction

equations below:

( )

( ) ( ) ( )

This reaction takes place most rapidly at pH 3.4. Under water and wastewater treatment

conditions, this reaction is too slow for the NDMA formation to occur.

Recent studies showed that the formation could take place when UDMH acted as an

intermediate (Mitch & Sedlak, 2002). Two steps mechanism was further developed, and

it corresponds with the phenomena of NDMA formation from monochloramine. This

8

mechanism is illustrated in the equations below. Firstly, monochloramine is formed from

ammonia and hypochlorite. Followed by this, monochloramine reacts with organic

nitrogen containing precursor to form UDMH. The product is further oxidized into

NDMA.

Step 1

( )

( ) ( ) ( )

Step 2

( ) ( ) ( )

NDMA formation is generally slow. It was reported that the application of

monochloramine facilitated NDMA formation (Mitch et al., 2003b). Furthermore, it was

observed that hypochlorite alone could form NDMA by reacting with secondary amine.

This reaction rate is one order magnitude smaller than that of monochloramine (Mitch &

Sedlak, 2002).

In water and wastewater treatment, major factors that affect NDMA formation include

the following: pH, chloramine concentration, contact time, Cl/N ratio and treatment

process. Studies demonstrated that pH affected NDMA formation during chloramination

of DMA; the maximum reaction rate occurred between pH 6 and 8 (Mitch et al., 2003b).

With fixed DMA concentration, NDMA formation increases with the increase of

9

chloramine concentration. However, the formation reaches a plateau in the end. NDMA

formation increases with the increase of contact time and with the increase of Cl/N molar

ratio of chloramine. In terms of treatment process, dosing of chlorine or chloramines as

disinfectant is the direct source of NDMA formation. Amine containing resin or polymers

increase NDMA precursors and result in increase of NDMA formation.

2.1.3 NDMA occurrence in water environment

Representative values of NDMA detected in various water sources are shown in the

Figure 2.2. Normally, surface water without impact of industrial waste or wastewater

discharges has NDMA level less than 10 ng/L. However, secondary wastewater effluent,

as shown, has higher NDMA concentration: 100 to 1000 ng/L. It was observed that even

after advanced treatment, for instance microfiltration, reverse osmosis and UV

disinfection, chlorinated wastewater still contained NDMA concentration between 10 to

100 ng/L (Sedlak & Kavanaugh, 2005).

Figure 2.2 Typical NDMA concentrations in various water bodies

10

Surveys on NDMA in drinking water system have been carried out. In US, California

Department of Health Services examined 32 water treatment plants. The study showed

some raw waters contained NDMA. In Ontario, 179 water treatment plants were surveyed

for NDMA presence. The study confirmed the results from survey in California. It also

discovered that in chloraminated system, there was increase of NDMA concentration in

distribution system (Charrois et al., 2007). In Japan, national wide survey examined

NDMA in raw and finished waters. Concentration of NDMA in raw waters was detected

to be less than 10ng/L (Asami et al., 2009).

In respond to the occurrence of NDMA in drinking water, as mentioned in Chapter one,

regulatory authorities have established water quality guidelines and standards for NDMA.

It is summarized in Table 2.2 (Fujioka et al., 2012).

Table 2.2 Guideline level of NDMA

US EPA,

IRIS 10-6

risk level

(ng/L)

CDPH

10-6

risk

level

(ng/L)

CDPH

notification

level (ng/L)

Ontario MOE

interim

action level

(ng/L)

WHO

guideline

value

(ng/L)

ADWG

guidelin

e value

(ng/L)

AGWR

guideli

ne

value

(ng/L)

0.7 3 10 9 100 100 10

IRIS: Integrated Risk Information System

CDPH: California Department of Public Health

MOE: Ministry of the Environment

WHO: World Health Organization

ADWG: Australian Drinking Water Guidelines

AGWR: Australian Guidelines for Water Recycling

At present, not many water utilities are able to monitor NDMA concentration in their

product water. In the United States, screening focused on unregulated contaminants of

concern for future monitoring. Results from 1196 public water supplies showed NDMA

as the most frequently detected contaminant with a maximum concentration of 630ng/L

11

(USEPA, 2010). NDMA is very likely to be regulated in the coming years under the US

Safe Drinking Water Act (Roberson, 2010). This indicates the need to detect NDMA in

water supplies and to control NDMA precursors in recycled water.

2.1.4 NDMA precursors

Existence of NDMA precursors in wastewater has been documented. To measure the

quantity of NDMA precursors in water sample, the concept of NDMAFP was introduced.

NDMAFP refers to the maximum amount of NDMA formed. Mitch et al. (2003)

developed a method to measure NDMAFP by dosing excessive monochloramine into

water sample and leaving the reaction for 10 days period. In secondary wastewater

effluent in United States, Mitch et al. (2003) reported NDMAFP to be between 660 and

2000 ng/L. For secondary wastewater effluent in South East Queensland in Australia,

Farré et al. (2010) reported NDMAFP varied from 300 to 1000 ng/L. Sedlak and

Kavanaugh (2005) reported NDMAFP in domestic wastewater in California of ranging

from 25 to 55 μg/L (25,000 – 55,000 ng/L). They reported NDMAFP in an industrial

wastewater was 82.5 μg/L (82,500 ng/L).

As mentioned in chapter one, DMA is the most effective organic nitrogen precursors.

Organic compounds containing amine functional groups including DMA and tertiary

amines can be converted into NDMA during chloramination process. Mitch et al. (2003b)

measured DMA in primary wastewater effluent ranging between 20 to 80 μg/L. DMA

has been detected in human urine and feces as well as animal feces. Amines can also be

formed from microbiological degradation of amino acids. Other sources of DMA and

nitrogen containing organics precursors include resins used in water and wastewater

12

treatment. For example, 50 ng/l NDMA was detected after extraction of distilled water by

strong-base dimethyl-ethanol containing anion-exchange resins (Najm & Trussell, 2001).

Another source comes from DMA functional group containing industrial products; such

as fungicides, pesticides, and herbicides (Mitch et al., 2003a). Furthermore, amine from

polymers, for example polyDADMAC cationic polymers, is proved to be NDMA

precursors. A table of potential NDMA precursors in secondary effluent (Farré et al.,

2010) is listed in appendix two.

2.1.5 Current research on NDMA removal methods

Ultraviolet (UV) radiation in the wavelength of 225 to 250 nm is the most common

method for removing NDMA in drinking water. This treatment has been used in water

treatment plant both in Ontario and Water Factory 21 in California. The technologies

include: low-pressure UV lamps emitting monochromatic light at 254 nm, medium-

pressure UV lamps emitting polychromatic light, and pulsed UV systems. The

controversy of the technology is that it does not destroy NDMA precursor completely;

therefore, there is possibility of reforming NDMA afterward. Studies showed that aerobic

and anaerobic biodegradation was possible to treat in situ NDMA contaminated water.

However, little was proved for NDMA removal by biodegradation under field conditions

(Mitch et al., 2003a).

NDMA precursors can be removed by biological treatment in secondary treatment and

microfiltration as well as reverse osmosis in advanced treatment. An average of 60% of

NDMA precursors in domestic wastewater can be removed by secondary biological

treatment. During advanced treatment, Mitch et al. (2003b) reported that MF reduced

13

particle-associated NDMA precursors from activated sludge effluent. RO removed not

only colloidal NDMA precursors but charged, dissolved precursors as well. Deeb et al.

(2006) reported NDMA and NDMA precursors removal during advanced treatment:

microfiltration (MF), reverse osmosis (RO) and UV treatment in three water facilities.

Experiment results indicated that MF removed a fraction of NDMA precursors ranging

from 12% to 95%. RO removed well NDMA precursors. Farre et al. (2011) studied

NDMA precursors removal in two advanced water treatment plant (AWTP) in South East

Queensland Australia. Similar to the result of previous study, RO filtration removed more

than 98.5% NDMA precursors. However, contradicting with the previous study, no

NDMA precursor removal was observed during either MF or ultrafiltration (UF).

Researchers advised AWTP to be careful during disinfection prior to RO process to

prevent excessive formation of NDMA.

2.2 Dissolved organic matter (DOM)

2.2.1 DOM characterization

Dissolved organic matter (DOM) in treated wastewater effluent is concerning in

wastewater reuse. DOM can alter other pollutants’ behavior through redox reaction and

can be converted to DBPs (Wei et al., 2008). DOM is a complex mixture of organic

compounds. They are of aromatic and aliphatic hydrocarbon structures attached with

functional groups (Leenheer et al., 2007), such as polysaccharides, protein, lipids, humic

substances, hydrophilic acids, carboxylic acids, amino acids and hydrocarbons.

Carbon is the main element in DOM. One of DOM characterization methods is to

measure dissolved organic matter (DOC) content. This is the organic compounds with

14

diameter smaller than 0.45μm. This can be measured using TOC analyzer by subtracting

inorganic carbon from total carbon. With fractionation technique, organic carbon

concentration can also be measured for each fraction. Figure 2.3 shows the general

categorization of DOM based on DOC fractionation (Leenheer & Croué, 2003). DOM

fractions can be fractionated differently according to study objectives.

Figure 2.3 Classification of DOM based on DOC fractionation

Dissolved organic nitrogen (DON) is another important parameter. Available methods for

DON quantification has been summarized as following (Pehlivanoglu-Mantas & Sedlak,

2006).

The Kjeldahl-N method has been applied for over a century. In this method, the

DON in the N (-III) oxidation state is converted to ammonia, which is distilled

15

and measured by titration, colorimetry, or ion-selective electrode. The presence

of high concentration of nitrate (> 10 mg N/L) interferes with measurement of

organic nitrogen.

Persulfate digestion converts all forms of inorganic and organic N into nitrate

followed by detection by colorimetry, ion-selective electrode, or ion

chromatography. To measure DON, this method requires subtraction of IN from

TN. The precision can be low for the sample with high IN compared to DON.

Oxidation of DON to nitrate with UV light. It should be noted that UV oxidation

does not always oxidize certain forms of organic nitrogen. The measuring result

is usually smaller than that of the persulfate method.

High-temperature combustion (HTC) converts organic nitrogen compounds into

NO, which then is detected in the gas phase by chemiluminescence (Badr et al.,

2003). Commercial instruments have been manufactured based on this technique.

HTC also estimate DON from the difference between TN and IN. This results in

the same loss of precision in samples that contains high IN.

To gain insight into the structure of DON, fractionation by affinity for hydrophobic resins

is used. The standard fractionation technique used by the International Humic Substances

Society employs a XAD-8 resin in conjunction with acid and base precipitation steps to

isolate humic substances. This can also be used to quantify humic-associated DON. The

molecular weight distribution of the DON can be estimated by ultrafiltration or by gel-

permeation chromatography (GPC), which is known as high-performance size-exclusion

chromatography (HPSEC). Table 2.3 lists DON compounds in municipal wastewater

16

effluent (E. Pehlivanoglu-Mantas & Sedlak, 2008). However, it should be noted that there

is lack of identification of wastewater derived DON: around 70% of it cannot be

identified with current methods.

Table 2.3 DON compounds in municipal wastewater effluent

DON compound Concentration (μMN)

Dissolved organic nitrogen 75 - 150

Dissolved free amino acids (DFAA) 0.04 - 2

Dissolved combined amino acids (DCAA) 1 - 19

Dimethylamine (DMA) 0.3 - 5.1

Ethylenediaminetetraacetic acid (EDTA) 0.1 - 0.5

Nitrilotriacetic acid (NTA) 0.1 - 0.5

Caffeine 0.1 - 0.2

N-containing pharmaceuticals <0.001

Liquid chromatography-Organic carbon detector-Organic nitrogen detector (LC-OCD-

OND) instrument is able to provide information on molecular weight distribution of both

DOC and DON. LC-OCD-OND separates DOM into major fractions. It estimates the

quantity of organic carbon in each fraction and the quantity of organic nitrogen in

biopolymer and humic substances fractions.

Size-exclusion chromatography fractionates DOM based on size exclusion, ionic and

hydrophobic interaction; UV-reactor then converts organic carbon (OC) to carbonic acid

and organically bound nitrogen to nitrate, these products are further detected by OCD and

OND.

17

Major fractions are illustrated by the figure below. Biopolymers (BP) contain molecules

of molecular weight (MW) larger than 10 kDa. Approximate MW for humic substance

(HS) is 1000 Da. Building blocks (BB) fraction consists of breakdown products of HS

which have MW ranging from 300 to 500 Da. Low molecular weight (LMW) fraction

includes molecules smaller than 350 Da. LC-OCD-OND method is proven to be robust

and sensitive. Its sensitivity is sufficient to measure water sample with low NOM directly

(Huber et al., 2011).

Figure 2.4 Chromatogram of surface water (River Pfinz, Karlsruhe, Germany) with

responses for organic carbon detection (OCD), UV-detection at 254 nm (UVD) and

organic nitrogen detection (OND). A: biopolymers, B: humic substances, C: building

block, D: low molecular-weight acids, E: low molecular-weight neutrals (LMW), F:

nitrate and G: ammonium.

LC-OCD-OND has been applied as an advanced analysis method for organic

characterization in surface water, seawater and treated wastewater (Huber et al., 2011;

Zheng et al., 2008; Penru et al., 2011). It has been applied for the study of membrane

foulants characterization. Zheng et al. (2010) applied LC-OCD-OND to measure

biopolymers concentration and to identify protein-like substances in organic foulants in

18

secondary effluent treated by ultrafiltration. Biopolymer concentration was calculated by

the LC-OCD-OND software in mg C/L. N/C ratio was used to evaluate protein-like

organics. In the study of membrane biofouling in forward osmosis membrane reactor

(FOMBR), Zhang et al (2012) applied LC-OCD-OND to examine organic molecular

weight distribution in the supernatant; the researchers also evaluated percentage of humic

substance in total DOC using results of OC in each fraction. In the study of

characterization of Mediterranean coastal seawater (Penru et al., 2011), LC-OCD-OND

was used to analyze organics composition. Results showed humic substance and LMW

were the main composition. LC-OCD-OND analysis was also applied to study the

organics removal effect of pretreatment for seawater desalination (Naidu et al., 2007).

Removal effect in each organic fractions based on OC content were evaluated.

2.2.2 Correlation between NDMAFP and DOM

Studies have shown that DOM is responsible for halogenated DBPs generation during

chlorination process (Gang et al., 2003; Richardson, 2003). However little is known

regarding NDMA formation mechanisms and kinetics from DOM. Luo et al. (2005)

fractionated organic matters in drinking water and discovered that hydrophilic fraction

contributed most to NDMAFP.

Studies have investigated correlation between NDMAFP and DOC and DON. Farré et al.

(2010) studied NDMAFP in treated wastewater and observed correlation between

NDMAFP and DOC and DON with R2 = 0.4. In natural water, Gerecke and Sedlak (2003)

observed correlation between NDMAFP and DOC with R2 = 0.41. Pehlivanoglu-Mantas

and Sedlak (2008) studied wastewater in different stage of treatment. They reported that

most NDMA precursors (75%) existed in the size fraction of less than 1,000 Da. DON

19

consisted of hydrophilic low-molecular weight (less than 1,000 Da) compounds contain

most of NDMA precursors.

2.2.3 DOM removal

To achieve NDMAFP removal, it is of interest to remove DOM. Coagulation, ion

exchange and activated carbon absorption are the established methods. DOC is

commonly used as an indicator for DOM removal, and it can be effectively removed.

Recently, study tested GAC for DOC removal in biologically treated wastewater. GAC

and GAC bio-adsorption were compared, and the later had better performance. With

sufficient reaction time (up to 6 hours), GAC bio-adsorption achieved 84% DOC removal

(Xing et al., 2008). Another study reported GAC in pilot-scale column for DOC removal

in membrane bio-reactor (MBR) wastewater achieved 80 to 90% (Gur-Reznik et al.,

2008).

Conventional water treatment processes, coagulation and filtration in particular are

normally ineffective to remove DON. Experiments results showed treated wastewater had

low SUVA value; DON showed poor adsorption to alum or carbon calcium (Chen et al,

2011). With alum dosage of 8 mg/L per mg/L DOC, less than 25% DON removals were

achieved. Filtration resulted in 18% DON removal which also showed limited efficiency

(Mitch, 2009). Therefore focus of DON removal should be given in wastewater treatment

plant and more advanced processes are needed (Bond et al., 2011). The following section

discusses the possibility of applying UV/H2O2 process to remove DOM and NDMA

precursor.

20

2.3 UV/H2O2 AOP process

2.3.1 AOP background

AOP refers to the process where organic compounds are oxidized through reaction with

hydroxyl radical which can be produced from different reactions, such as H2O2/O3,

O3/UV, H2O2/UV, TiO2/UV, Fenton’s reaction. Hydroxyl radical has the second highest

oxidation potential (2.70 eV) among other strong oxidants: fluorine, ozone and chlorine.

AOPs are destructive and non selective. Some AOPs have disinfectant capabilities.

However, there are potential productions of oxidation by-products, bromated formation.

Scavenging effects from other compounds affect AOP’s effectiveness as well.

Currently, H2O2/O3, O3/UV, H2O2/UV are established technologies. Emerging ones

include high energy electron beam irradiation, cavitation (Sonication & Hydrodynamic),

TiO2-catalyzed UV oxidation, and Fenton’s reaction. Table 2.4 summarizes the

comparison among three established technologies (Kommineni et al., 2000).

21

Table 2.4 Comparison of H2O2/O3, O3/UV, and H2O2/UV AOP processes

AOP

technology

Description Advantages Disadvantages

H2O2/O3 OH∙ generated from

adding H2O2 and O3

to water reacts with

organics

Disinfectant

More effective

than H2O2 or O3

alone

Established

technology for

remediation

applications

Potential for bromate

formation

May require

treatment of excess

H2O2 due to potential

for microbial growth

May require O3 off-

gas treatment

O3/UV OH∙ is generated

when UV is applied

to ozonated water.

Three mechanisms

for organics removal

are: direct photolysis,

oxidation by

molecular ozone and

OH∙ reaction.

Disinfectant

More effective

than O3 or UV

alone

More efficient at

generating OH∙

than H2O2/UV

process for equal

oxidant

concentrations

Energy and cost

intensive process

Potential for bromate

formation

Turbidity can

interfere with UV

light penetration

Ozone diffusion can

result in mass transfer

limitation

May require O3 off-

gas treatment

Interfering

compounds can

absorb UV light

Potential increase in

THM and HAA9

formation when

combined with pre-

and/or post-

chlorination

H2O2/UV OH∙ is generated

when UV is applied

to water dosed with

H2O2. Two

mechanisms for

organics removal are:

direct photolysis, and

OH∙ reaction.

No potential for

bromate formation

MPUV and

pulsed-UV

irradiation can

serve as

disinfectant

Full scale drinking

water treatment

system exists

Not limited by

mass transfer

relative to O3

process

Turbidity can

interfere with UV

light penetration

Less

stoichiometrically

efficient at generating

OH∙ than O3/UV

process

Interfering

compounds can

absorb UV light

Potential increase in

THM and HAA9

formation when

combined with pre-

and/or post-

chlorination

22

2.3.2 UV/H2O2 mechanism

UV/H2O2 has been applied in drinking water treatment since 1990s ("New concepts of

UV/H2O2 oxidation," 2011). In this process, two mechanisms contribute to the

degradation of target compound: direct photolysis by UV and oxidation by hydroxyl

radical.

Photolysis takes place when chemical absorbs UV irradiation and is broken down into

smaller molecules. The range of UV wavelength has to include the one which can be

absorbed by chemicals. Hydroxyl radical is generated when H2O2 is catalyzed by UV.

This reaction is illustrated in the equation below.

(2.6)

Free hydroxyl radial (OH∙) is of one electron deficiency. It is hence very unstable and

reacts non-selectively with chemicals that contact with it. It is shown that free hydroxyl

radial can completely oxidize organic contaminants into carbon dioxide, water and salt.

2.3.3 Key affecting factors

Principle affecting factors in UV/H2O2 process include UV lamp technology, H2O2

dosage, and water quality parameters.

2.3.3.1 UV lamp technology

Conventionally, mercury lamps are used to produce UV irradiation. Mercury atoms are

brought to excited state under electrical charge, and irradiate photons when they return to

ground state. The wavelength of the irradiation depends on the gas pressure in the lamp.

("New concepts of UV/H2O2 oxidation," 2011).

23

Two types of UV lamps that are available for UV/H2O2 process: low pressure ultra-violet

lamps (LPUV) and medium pressure ultra ultra-violet lamps (MPUV). The differences

between these two lamps are summarized in the Table 2.5 ("New concepts of UV/H2O2

oxidation," 2011). LPUV emits UV light at the wavelength of 254nm. MPUV emits UV

light at the wavelength ranging from 200nm to 400nm. LPUV has higher UV efficiency

than MPUV. UV efficiency is defined as ratio of the power of light emitted in the range

of 200 to 300nm to the lamp input power. LPUV has an UV efficiency of around 30

percent; while MPUV's UV efficiency is of around 15 percent. However, MPUV allows a

higher energy input than LPUV. Therefore, higher intensity of emitting light can be

achieved.

Table 2.5 Main difference between LP and MP mercury UV lamps

LP mercury lamp MP mercury lamp

Low power: <1 kiloWatt High power: ≤30 kiloWatt

Emission: 253.7 nm Emission: 200-400 nm

High energy efficiency: ~ 30% Low energy efficiency: ~15%

Relatively long lifespan: ~9,000 hours Relatively short lifespan: 4,000 – 6,000 hours

In terms of free hydroxyl generation, LPUV has better hydroxyl generation efficiency.

The efficiency depends on the wavelength. Figures 2.5 and 2.6 ("New concepts of

UV/H2O2 oxidation," 2011) indicate UV absorbance in pretreated nature water and in

10mg/l H2O2 solution for MPUV and LPUV. As it is shown in the two figures below,

between 200 to 300nm, H2O2 extinction decreases with increasing of wavelength.

However, UV absorbance by water matrix is significantly higher than the absorbance by

H2O2 in the wavelength ranging from 200 to 230nm. As a result, UV in this range is

easily absorbed by water matrix before reacting with H2O2 (Li et al., 2008). Moreover,

UV light with wavelength above 260nm has very low H2O2 extinction coefficient. This

24

indicates that UV light above 260nm contributes little to hydroxyl radical production.

Consequently the following argument is formed: when UV wavelength is about 254nm,

due to low absorbance by water matrix, the production of hydroxyl radical may be

efficient. Since LPUV emits light at 254nm and MPUV emits light in a wider range

where the emission is not efficient for hydroxyl radical generation, LPUV is more

efficient in hydroxyl radical production.

Figure 2.5 Emission spectrum of a MP-lamp ( ) and the absorptions of a commonly

pretreated natural water ( ) and a 10mg/L H2O2 solution (according to the

Lambert-Beer law) ( ). Composition of natural water: DOC content: 3.6 mg/L C;

[NO3-]: 11.2 mg/L

25

Figure 2.6 Emission spectrum of a LP-lamp ( ) and the absorptions of a commonly

pretreated natural water ( ) and a 10mg/L H2O2 solution (according to the

Lambert-Beer law) ( ). Composition of natural water: DOC content: 3.6 mg/L C;

[NO3-]: 11.2 mg/L

2.3.3.2 H2O2 dosage

Low concentration of 10 mg/L is normally chosen for AOP design, since additional H2O2

has to be removed (Li et al., 2008). Excess H2O2 can be removed by granular activated

carbon (GAC). Higher concentration is used when 10 mg/L is not able to achieve

targeting removal efficiency.

2.3.3.3 Water quality parameters

AOP performance is determined by the second order OH∙ rate constant. In pretreated

wastewater, the removal effect of target compound is affected by the reactivity of target

compound with hydroxyl radical. At the same time, following factors need to be taken

26

into consideration: presence of carbonate species, natural organic matter (NOM), reduced

metal ions, and pH.

Carbonate species, in particular carbonate and bicarbonate react with OH∙ and hence have

scavenging effect on OH∙. The reactions are shown in the equations below (Li et al.,

2008).

106L/mole∙s (2.7)

108 L/mole∙s (2.8)

Although CO32-

and HCO3- have OH∙ rate constant one order smaller than that of many

organics, the concentration of carbonate species is usually several orders of magnitude

larger than that of the target organics. This makes the reduction of alkalinity important

for improving UV/H2O2 efficiency. The scavenge effect of carbonate on target compound

is measured by the factor:

(2.9)

Where = target organics reaction rate with OH∙ divided by total reaction rate of OH∙

with R and carbonate species

k = second order OH∙ rate constant, L/mole/s

C = concentration, mole/L

Between carbonate and bicarbonate, the former is more damaging. This is because OH∙

rate constant for carbonate is about 46 times of that of bicarbonate. Given that lower pH

value will result in lower carbonate concentration, in terms of reducing alkalinity, lower

pH is favorable. The pKa value for equilibrium between carbonate and bicarbonate is

10.4; the pKa value for equilibrium between HCO3- and H2CO3 is 4.3. At pH value of 8.3

or lower, no significant carbonate is presented (Li et al., 2008).

27

Similar to the scavenging effect of carbonate species, NOM in water matrix react with

OH∙ and have scavenging effect on OH∙. On the other hand, NOM absorbs UV light

before UV reaches H2O2. This lessens the amount of UV light available for OH∙

generation.

Study (Westerhoff et al., 1999) on 17 water sources provided the OH∙ rate constant of

NOM, which is about 3 108 to 4.5 10

8 L/mole NOM carbon∙s. This shows most of

NOM has higher OH∙ rate constant compared to carbonate and bicarbonate. At similar

concentration level, scavenge effect of carbonate species is less significant compared to

that of NOM.

Reduced metal ions, in particular iron and manganese react with OH∙ and have

scavenging effect. In addition, Fe (II) and Fe (III) absorb UV lights as well, which results

in reduced UV/H2O2 efficiency.

In general, UV/H2O2 process prefers higher pH value. pH affects UV/H2O2 efficiency in

the following aspects. It determines composition of carbonate species in water sample. It

varies the charge of target organics. For certain organics, ionic form is associated with a

higher rate constant, up to one or two orders higher. Moreover, pH affects HO2-

concentration. HO2- absorbs UV light with higher efficiency compared to H2O2. The pKa

value of H2O2 is 11.6. Therefore, at higher pH, more HO2- absorb UV with higher

efficiency, and more efficient the UV/H2O2 process will be (Li et al., 2008).

28

2.3.4 NDMA precursor removal by UV/H2O2

With laboratory synthetic water (surface water combined with biologically treated

wastewater), UV/H2O2 process achieved NDMAFP removal of 50% (Chen et al., 2010).

However, with extended reaction time, there was no further NDMAFP and DON

removal. This indicates the possible need of higher H2O2 dosage to degrade more

NDMAFP. Within the complex organic composition in wastewater matrix, organics with

higher oxidation demand reduces photo oxidation efficiency. This is demonstrated by

comparing removal efficiency in humic acid (HA) solution and humic acid solution

spiked with wastewater. DOC and DBPFP removal is more significant in HA solutions

(Chen et al., 2010).

More recent study applied MPUV/H2O2 and LPUV/H2O2 to selected nitrogenous organic

compounds. While high DOC reduction (up to 90%) were achieved, removal of DON

was less significant, except for histamine and caffeine. The maximum DON removal after

two treatments achieved only 30% and 20%, respectively (Chen et al., 2011). During

MPUV/H2O2 treatment on DMA, an increase of NDMAFP was observed at the beginning

of the reaction followed by reduction.

There is limited research on UV/H2O2 removal efficiency of NDMAFP in treated

wastewater, in terms of organic precursor and NDMAFP removal kinetics. It is of interest

to conduct bench scale experiment on UV/H2O2 process, to understand removal kinetics

and efficiencies, to assess the feasibility of applying UV/H2O2 to remove NDMAFP in

treated wastewater.

29

CHAPTER 3 MATERIALS AND METHODS

3.1 Introduction

Detection of trace level (ng/L) NDMA is the first challenge in the study. Before starting

batch experiment, detection method for NDMA was established in the lab. Based on this,

selected wastewater samples were collected and batch experiments were designed to

study NDMAFP in wastewater samples. At the same time, characterization of organic

precursors was conducted to gain more insights into relationships between organic

precursors and NDMAFP.

After the experiment on NDMAFP, following study was conducted on organic precursors

and NDMAFP removal by UV/H2O2. As mentioned in chapter two, the performance of

UV/H2O2 system depends on many factors, such as UV lamp technology, H2O2 dosage

and water parameters. Two configurations of UV/H2O2 systems were tested before

deciding the final system. A LPUV/H2O2 batch system was then used for assessing its

performance.

3.2 Experiment design part 1: NDMA formation potential and

precursors study

3.2.1 Experiment design

Wastewater samples were collected and dosed with monochloramine solution. Over 10

day period, reactions were stopped at 0, 0.5, 1, 2, 5 and 10 day by adding sodium

thiosulfate to quench chloramine residual. After this, NDMA concentration was measured.

30

On parallel, characterization of organic precursors in collected wastewater samples was

conducted: TOC, TN, IC, and LC-OCD-OND analysis were conducted. This part of

experiment is illustrated in the schematic diagram below.

Figure 3.1 Schematic chart of experiment design part 1

3.2.2 Experiment conditions

3.2.2.1 Materials

All chemicals used in this study were purchased from Sigma-Aldrich unless specifically

indicated. In this part of experiment, materials used were: NDMA and NDMA-d6

standard solution, EPA521 cartridge, methanol, dichloromethane, ammonium acetate,

ammonium chloride, sodium hydroxide, sodium thiosulfate, monosodium phosphate and

disodium phosphate.

3.2.2.2 Sample collection and characteristics

Three types of wastewater samples were collected from local wastewater treatment plant:

primary effluent, secondary effluent from activated sludge process (ASP), secondary

effluent from membrane bioreactor (MBR) process. All water samples were collected in

10 L or 20 L jerry can. They were transported by lab vehicle, stored in cold room with

Wastewater

sample collection

Chloramination NDMA

measurement

Organic precursor analysis: TOC, TN,

IC, LC-OCD-OND

31

water sample information (type, date, owner) clearly labeled. Water quality parameters

are presented in Table 3.1.

Table 3.1 Water quality parameters for wastewater treated effluents

Parameters Primary effluent ASP effluent MBR effluent

NH4 + (mg N/L) 32.1±0.85 0.25±0.05 0.13±0.04

NO3 - (mg N/L) < DL 8.59±0.40 7.07±0.30

NO2- (mg N/L) < DL < DL < DL

TOC, mg/L 54.01±0.41 7.06±0.62 4.85±0.38

TC, mg/L 82.68±0.46 12.91±0.42 14.94±0.28

IC, mg/L 28.67±0.04 5.85±0.24 10.08±0.09

TN, mg/L 39.11±1.04 10.93±0.46 7.38±0.12

DON*, mg/L 1.345±0.07 0.45±0.06 0.26±0.01

DL: detection limit

*: DON measured in LC-OCD-OND

3.2.2.3 NDMAFP study

NDMAFP study followed the study by Mitch et al. (2003a). All glassware used in this

experiment were washed with decon 90 water, rinsed with DI water and baked at 550 oC

for at least 2 h prior to use. Reactions were conducted in room temperature in 1-L amber

glass bottles or glass bottles shielded from light. Reactions were controlled at pH=6.8 by

adding 10mM phosphate buffer. For each reaction, 900mL water sample was dosed with

100mL 20mM monochloramine stock. The monochloramine stock solution was prepared

fresh daily from ammonium chloride and sodium hydroxide solution. Hypochlorite to

ammonia ratio followed 1:1.2. Excessive dose of monochloramine ensured that majority

32

of the NDMA precursors were converted into NDMA. Total chlorine concentration was

measured at the end of each stage, and 200mM sodium thiosulfate solution was applied to

quench the reaction. 900mL DI water was used as control blank. It followed the reaction

procedure to measure the possible effects from reagents. After reaction, 1L sample

underwent NDMA detection and analysis as described previously.

3.3 Experiment design part 2: UV/H2O2 effect on DOM and NDMAFP

removal

3.3.1 System set-up

Two batch reaction systems were tested (Figs 3.2 and 3.3). In system I, two 15 watt

LPUV lamps (Max Cure UV Technik, Singapore) were placed on top of two 1L beakers

with magnetic stirrers at the bottom. UV lamps produced UV intensity of 1.3mW/cm2 at

254nm. In system II, one LPUV lamp (Trojan UV Max Model B lamp) was placed in the

center of 1L beaker with magnetic stirrer at the bottom. UV lamp produced UV intensity

of 2mW/cm2 at 254nm.

In both systems, lamps were heated for 10 minutes before each batch experiment to

ensure steady UV output. One liter of water sample was added into each beaker for

reaction. The magnetic stirrers provided vigorous mixing.

33

Figure 3.2 Schematic diagram of UV/H2O2 system I

Figure 3.3 Schematic diagram of UV/H2O2 system II

Beaker 1

Beaker 2

LPUV lamp1

LPUV lamp2

Magnetic Stirrer Magnetic Stirrer

To power

Beaker

Magnetic Stirrer

To power

LP

UV

lamp

34

Plate 3.1 UV/H2O2 system I set-up

Plate 3.2 UV/H2O2 system II set-up

35

3.3.2 Wastewater characteristics

Both ASP and MBR effluents were collected on site in polyethylene containers and

stored at 4oC in cold room. Effluents were filtered through 0.45 μm filter to remove

particles before reaction in photoreactor. pH and UV transmittance values for both

effluents after filtration are shown in Table 3.2. Reagents used include the ones in

experiment part 1, hydrogen peroxide and catalase.

Table 3.2 pH and UV transmittance for ASP and MBR effluents

Parameters ASP effluent MBR effluent

pH 7.35±0.50 7.96±0.06

UVT, % 73.27±1.1 76.33±0.64

3.3.3 Experiment design

ASP effluent was used first to test on the performance of system one. A dosage of

200ppm H2O2 was spiked into water samples. Reaction time was 120 minutes at first and

then extended up to three hours.

Both ASP effluent and MBR effluent were tested in system two. A dosage of 100ppm

H2O2 was spiked into water samples. Reaction time was 60 minutes. Different H2O2

dosages were chosen to examine the effect of H2O2 dosage.

During each batch experiment, 1L water sample was dosed with H2O2 and treated by UV

irradiation. 100mL sample was collected at specific time for organic precursor analysis.

H2O2 were stopped immediately by adding catalase. The H2O2 in the remaining 900mL

36

sample was stopped by adding 100mL 20mM monochloramine stock. This sample was

then used for NDMAFP detection. Both NDMAFP detection and organic precursor

analysis followed the procedures in experiment part 1. Precursor removal profile and

efficiency were evaluated by DOC, DON removal and DOC, DON removal in each

organic fraction defined in LC-OCD-OND results.

3.4 Measurement and analysis method

3.4.1 NDMA detection and analysis

3.4.1.1 SPE

SPE procedure concentrates NDMA before analyzing. EPA 521 cartridge (coconut

charcoals) was chosen as the solid phase. SPE manifold was used. Solvents used were

dichloromethane and methanol which were purchased from Sigma-Aldrich.

Before each experiment, 100ppt deuterated NDMA (NDMA-d6) was added as an internal

standard to water sample. As mentioned in chapter two, this is to monitor extraction

efficiency. Plastic tubes connecting sample to cartridge were cleaned with DI water.

Cartridges were mounted on the SPE manifold. The vacuum port was joined to a

volumetric flask connecting to a pump. During SPE, solvents and water sample were

pulled through the cartridge by vacuum. Figure 3.1 shows experiment set up for SPE

procedure. The SPE procedure was: EPA 521 cartridge was pre-cleaned by 3 ml

dichloromethane, followed by 3ml methanol. Air dry was followed. Cartridge was

conditioned with 3 ml methanol and 15 ml DI water, the SPE bed was not allowed to dry

during the condition step.Water sample was loaded at flow rate of 3-5 ml/min. Air dry

37

with full vacuum for 60 min or until cartridge was completely dry. SPE cartridge was

filled with one tube full of dichloromethane and left for 5 mins to soak the sorbents

before elution. Elution was performed with 15 ml dichloromethane at drop wise flow

rate. The extract was concentrated to 1 ml under a gentle N2. The sample was filtered

with 0.45 μm filter and transferred to 1.8 ml crimp-top vial for LC/MS/MS analysis.

Plate 3.3 Experiment set up for SPE procedure

3.4.1.2 LC-MS/MS analysis

Samples after SPE were analyzed by Shimadzu LCMS-8030 triple quadrupole mass

spectrometer. A liquid chromatography (LC-30AD) with a Shimadzu SIL-30AC auto

sampler was connected to a triple quadrupole mass spectrometer. Analyst software

LabSolutions was used for data acquisition and analysis. A C18 capillary column (2.1 x

150 mm i.d., 5μm) was applied for separation. Mobile phases consisted of solvent A (100%

methanol) and solvent B (2 mM ammonium acetate).

The solvent gradient was programmed to be of 50% of solvent B for 3.5 min, increasing

solvent B from 50% to 100% over 1 min, and returning back to 50% of solvent B over

38

0.01min, followed by a 1.5 min re-equilibration prior to the next sample injection. The

flow rate for mobile phase was 0.3 μL/min. The sample injection was 20 μL.

The mass spectrometer was operated in multiple reaction-monitoring transition mode.

Each transition was in positive-ion mode. MS/MS parameters were selected after the

injection of NDMA, NDMA-d6 standards in continuous-flow mode. Selected ion pairs

for NDMA and NDMA-d6 were: NDMA 75.10 > 43.10, NDMA-d6 81.20 > 46.10. The

mass spectrometer was operated at unit mass resolution, dwell time per ion pair was at

150 msec.

Prior to each batch run, NDMA and NDMA-d6 standard solutions of 10, 20, 50, 100, 200

μg/L were analyzed to establish calibration curve. Methanol was used as blank solution to

check for carryover in the column. Recovery of NDMA was indicated by the recovery of

NDMA-d6. NDMA concentration in water sample was calculated based on recovery as

shown in the equations 3.1 and 3.2.

( )

( )

3.4.2 Free chlorine and total chlorine

During NDMAFP study, chlorine was dosed into water sample. It was necessary to

measure chlorine concentration. Both free chorine and total chlorine were measured using

the DPD (N, N-diethyl-p-phenylenediamine) method. In this study, a Hach DR5000

39

spectrophotometer, DPD Free Chlorine Reagent Powder Pillows and DPD Total Chlorine

Reagent Powder Pillows were used.

Experiment procedure is described as the following. For free chlorine, chlorine test

program was chosen from DR5000. Blank was prepared by filling a sample cell with 10

mL of sample solution. The blank was wiped and inserted into the cell holder. The

instrument was zeroed. Sample was prepared by filling a second cell with 10 mL of

sample solution and adding DPD Free Chlorine Powder Pillow. The sample cell was

swirled for 20 seconds to mix. At the presence of chlorine, a pink color would develop.

Within one minute, the prepared sample was inserted into the cell holder. Results were

read in mg/L Cl2.

For total chlorine, chlorine test program was chosen from DR5000. Sample was prepared

by filling a sample cell with 10 mL of sample solution and adding DPD Total Chlorine

Powder. The sample cell was swirled for 20 seconds to mix. The instrument was started

to time three minutes. Within three minutes, a blank was prepared by filling a second

sample cell with 10 mL of sample solution. The blank sample cell was wiped and inserted

into the cell holder. The instrument was zeroed. At three minutes, the prepared sample

was wiped and inserted into the cell holder. The results were read in mg/L Cl2.

3.4.3 TOC and TN

Total organic carbon (TOC) and total nitrogen (TN) were measured by Shimadzu TOC

analyzer with total nitrogen measuring unit. Calibration range for TOC and TN were both

from 0 to 20 ppm. For wastewater sample with high TOC and TN content, dilution was

40

performed before the measurement. DI waters were added before and after each batch of

measurements for quality control.

3.4.4 Nitrate, Nitrite and Ammonium

Inorganic ions such as NO3- and NO2- were measured by Dionex ion chromatography.

Calibration ranges for NO3- and NO2

- were both from 0.5 to 40 ppm. DI water was added

before and after each batch of measurement for quality control. Ammonium was

measured using Hach Nitrogen-Ammonia Reagent Set.

3.4.5 LC-OCD-OND

10 ml of sample was loaded into sampling vial, covered by aluminum foil, and loaded

into auto sampler for testing. DI water was placed before and after each batch of

measurement. Injection volume was 1000 μl; running time was 130 minutes for each

sample. Mobile phases applied were phosphate buffer prepared by mixing 1.5g/l

Na2HPO4 and 2.5g/l KH2PO4. Acidification solution was prepared by mixing 4 ml o-

phosphoric acid and 0.5 g potassium peroxodisulfate in 1 L of DI water. OCD and UVD

calibration was established on potassium hydrogen phthalate. OND calibration was

established using potassium nitrate.

Data collection and processing were carried out by customized software (ChromCALC,

DOC-LABOR, Karlsruhe, Germany). OC in four fractions and ON content in biopolymer

and humic substances fractions can be estimated by the software. For ON in building

block and low molecular weight neutrals, manual integration was carried out. Integration

was done both to estimate the nitrogen content in LMW and the total organic nitrogen

41

detected. Organic nitrogen in building block was then calculated by subtracting nitrogen

content in other fractions out of total nitrogen content:

ON in Building Block = Total ON– ON in Biopolymers – ON in HS – ON in LMW (3.4)

3.4.6 DON

DON was measured initially by subtracting inorganic nitrogen (IN) from total dissolved

nitrogen (TDN). However, after monitoring both IN and TDN, it was observed that

selected samples had high IN level. The ratio of IN to TDN was higher than 0.7. As

mentioned in chapter two, the precision of this method is low when IN is high compared

to DON. In another study (Chen et al., 2011), it was shown that at IN/TDN higher than

0.6, IN measurements variance could be greater than DON level. Therefore, TDN and IN

measurement in this study was used for indicating water parameter, instead of calculating

DON.

The measurement of DON was approached by the oxidation method in LC-OCD-OND.

As mentioned in chapter two, organic nitrogen in the fraction passed through the column

is oxidized by UV and measured by OND. Manual integration can calculate organic

nitrogen in four fractions. However, during LC-OCD-OND measurement, a small portion

of organic matter remained on the column due to strong retention caused by hydrophobic

interaction (Huber et al., 2011). Therefore, DON measured by LC-OCD-OND

represented only the DON in hydrophilic portion. Although total DON could be

calculated by subtracting IN from TN (both measured in LC-OCD-OND), the high

percentage of IN lowered as well the precision of total DON in this case.

42

CHAPTER 4 RESULTS AND DISCUSSIONS

This study focused on NDMAFP in terms of NDMA formation kinetics, characteristics of

NDMA organic precursors, relationship between organic precursors and NDMAFP, and

NDMAFP removal. Results from experiment part one include NDMA formation and

organic characteristics in selected wastewater samples. Results from experiment part two

include UV/H2O2 effect on organic precursors and NDMAFP removal in two secondary

wastewater samples.

4.1 NDMAFP study results

4.1.1 NDMA formation kinetics and formation potential

Experiment results (Fig 4.1) show the similarity in formation kinetics for two types of

secondary effluents. NDMA concentration increased rapidly during the first two days of

reaction. It reached a plateau afterward. In the first graph, NDMA concentration in ASP

effluent increased from zero (not detectable) at the beginning to 302 ng/L at day two. The

concentration remained at the same level after two days. Formation profiles of ASP and

MBR effluent corresponded with previous study (Mitch et al., 2003a) where NDMA

concentration in secondary effluent plateaued after two days.

However, for primary effluent, although a rapid increase was observed as well within

first two days, concentration did not reach a plateau afterward. Figure 4.1 shows that

there was still a trend of increase after 10 days. This may be due to the high organic

precursor level in primary effluent. Primary effluent contained 54 ppm TOC, much

higher than that of secondary effluent (around 6 ppm). Within 10 day reaction period,

43

organic precursors in primary effluent had not been completely converted to NDMA. In

order to measure NDMAFP of primary effluent or wastewater sample with high organic

content, additional monochloramine dosage and/or extended reaction time may be

needed.

Figure 4.1 NDMA formation vs reaction time for selected wastewater samples

Primary effluent had also high level of NDMA formed within 10 days: 3764 ng/L. Two

types of secondary effluent formed similar level of NDMA which corresponded to

previous studies (Wilczak et al., 2003; Sedlak & Kavanaugh, 2005; Farré et al., 2010).

Figure 4.2 shows NDMA formed from two effluents ranging from 240 to 400 ng/L; the

0

50

100

150

200

250

300

350

0 50 100 150 200 250 300

ND

MA

, ng/

L

reaction time, h

ASP effluent0

50

100

150

200

250

300

350

0 50 100 150 200 250 300

ND

MA

, ng/

L

reaction time, h

MBR effluent

0

500

1000

1500

2000

2500

3000

3500

4000

4500

0 100 200 300

ND

MA

, ng/

L

reaction time, h

Primary effluent

44

error bars represent standard deviations of the measurements. Figure 4.3 shows NDMA

formed in secondary effluents from seven different wastewater treatment plants (Farré et

al., 2010). Concentration of NDMA formed varied from 300 to 1000 ng/L. These two

results corresponded with each other. However, research in the United States reported

NDMA formed from secondary effluents ranging from 660 to 2000 ng/L (Wilczak et al.,

2003; Sedlak & Kavanaugh, 2005). This higher level of NDMA formed might be due to

the different degrees of industrial/commercial contribution. In Sedlak and Kavanaugh’s

study (2005), higher levels of NDMA formed were identified in industrial area.

Figure 4.2 Comparison of NDMAFP of ASP and MBR effluents

13

/8/2

01

2

21

/8/2

01

2

1/1

0/2

01

2

5/4

/20

12

5/4

/20

12

19

/11

/20

12

3/1

2/2

01

2

0

50

100

150

200

250

300

350

400

450

ASP MBR

ND

MA

FP,

ng/

L

45

Figure 4.3 NDMA formation study of secondary effluent in Australia (Farré et al., 2010)

4.1.2 Organic precursor analysis

Table 3.1 indicates that primary effluent had DOC and DON level much higher than

secondary effluent. Primary effluent’s DOC was nine times of that of secondary

effluents; its DON was three to six times of that of secondary effluents. DOC and DON

values of two secondary effluents were of the similar level.

Figs 4.4 and 4.5 show organic precursors molecular weight distributions in selected

wastewater samples. Due to high DOC concentration in primary effluent, the sample was

diluted 10 times for LC-OCD-OND analysis. All three wastewaters contented organics in

all fractions: biopolymer (BP), humic substance (HS), building block (BB), low

molecular weight acid and neutral (LMW). However, the organics composition of the

wastewaters differed from each other.

46

Primary effluent had relatively higher percentage of OC and ON (26% and 44%) in

biopolymer (BP) fraction compared to secondary effluents. Between two secondary

effluents, MBR effluent had little OC and ON (2% and 0%) in BP fraction, due to large

molecules removal by ultrafiltration membrane. This corresponded with the result in

Zhang et al.’s study (2012) that BP accounted for very small portion of DOC (4.4%) in

MBR effluent. Except the difference in BP fraction, MBR and ASP effluents had similar

molecular weight distributions in other fractions.

47

Figure 4.4 OCD chromatographs of selected wastewater samples

Bio

poly

mers

Hum

ics

Build

ing B

locks

LM

W A

cid

s a

nd H

S

LMW Neutrals

Prim Eff (1/10)

ASP Eff

MBR Eff

0

1

2

3

4

5

6

7

8

9

10

11

12

0 20 40 60 80 100

rel. S

ign

al R

esp

on

se

Retention Time in Minutes

Similar MW distribution

26%

16%

2%

High %

of BP

48

Figure 4.5 OND chromatographs of selected wastewater samples

Bio

poly

mers

Hum

ics

Build

ing B

locks

LM

W A

cid

s a

nd H

S

BY

PA

SS

Prim Eff (1/10)

ASP Eff

MBR Eff

0

1

2

3

4

5

6

7

8

9

10

11

12

0 20 40 60 80 100

rel. S

ign

al R

esp

on

se

Retention Time in Minutes

44%

Simialar MW

distribution

28%

0%

High %

of BP

49

4.1.3 Relationship between organic precursors and NDMAFP

Since wastewater sample contains different levels of organic precursors, in order to

access NDMAFP by unit precursor, NDMAFP was normalized to DOC and DON. Figure

4.6 indicates that before normalization, NDMA formed in primary effluent was 10 times

of that of secondary effluent. After normalization, NDMAFP/DOC and NDMAFP/DON

for primary effluent and secondary effluent were relatively comparable. However, it

should be noted that, in primary effluent, not all NDMA precursors had been converted to

NDMA, therefore ultimate NDMAFP would be higher. Due to this reason, primary

effluent result was not used for the following correlation study; only NDMAFP of

secondary effluents were analyzed.

Figure 4.6 NDMAFP and normalized NDMAFP in different wastewater samples

0

500

1000

1500

2000

2500

3000

3500

4000

4500

Prim Eff ASP Eff MBR Eff

ND

MA

FP,

ng/

L

0

10

20

30

40

50

60

70

80

90

Prim Eff ASP Eff MBR Eff

ND

MA

FP/D

OC

, p

pb

/pp

m

0

500

1000

1500

2000

2500

3000

3500

Prim Eff ASP Eff MBR Eff

ND

MA

FP/D

ON

, p

pb

/pp

m

50

Figure 4.6 also shows that ASP effluent had smaller NDMAFP/DOC and

NDMAFP/DON than that of MBR effluent. This may be due to different organic

compositions. ASP effluent had smaller percentage of DOC in BB and LMW, and

smaller percentage of DON in BB (Table 4.2). This indicates that less percentage of small

molecular weight molecules may lead to less NDMAFP. In other words, small molecular

weight precursors may contribute more to NDMAFP. This finding was further supported

in the correlation analysis in the following paragraphs.

Table 4.1 DOC and DON composition in each fraction in ASP and MBR effluents

DOC DON

BP HS BB LMW BP HS BB LMW

ASP Eff 16% 38% 23% 23% 28% 40% 20% 12%

MBR Eff 2% 26% 43% 29% 0% 17% 73% 10%

Figure 4.7 indicates weak correlations between DOC and DON and NDMAFP.

Correlation between DON and NDMAFP was relatively better than that of DOC, with R2

= 0.16. In a similar study, Farré et al. (2010) observed correlation between DOC and

DON and NDMAFP with R2

= 0.4, where seven types of secondary effluent were

collected. This suggests that to obtain a better understanding of correlation between water

parameter and NDMAFP in future study, various types of secondary effluent should be

collected.

51

Figure 4.7 Correlation between NDMAFP and organic precursors in secondary effluents

Figure 4.8 shows that among DON in each fraction, DON in BB and LMW fraction

correlated better to NDMAFP, with R2

= 0.42 and 0.47, respectively. BB fraction is

consisted of molecules with MW within 300 – 500 Da. LMW fraction is consisted of

molecules with MW smaller than 350 Da. Therefore, these correlation results suggest that

in order to control NDMAFP, more attention should be given to DON control in small

molecules with MW smaller than 500 Da.

In previous study, other researchers examined NDMAFP correlation with fractions of

organics. They fractionated organic precursors using resins. Results showed that

y = 6.6773x + 261.35 R² = 0.0905

0

100

200

300

400

500

0 5 10 15

ND

MA

FP,

ng/

l

DOC, ppm

y = 193.61x + 240.87 R² = 0.1639

0

100

200

300

400

500

0 0.2 0.4 0.6

ND

MA

FP,

ng/

l

DON, ppm

52

NDMAFP was more correlated to precursors in the size fraction less than 1,000 Da

(Pehlivanoglu-Mantas & Sedlak, 2008). Results in this study corresponded with the

previous finding. Moreover, LC-OCD-OND analysis in this study provides another

approach to examine DOC and DON distribution.

Figure 4.8 Correlation between NDMAFP and DON associated with four fractions in

secondary effluents

y = 0.186x + 287.88 R² = 0.1031

0

50

100

150

200

250

300

350

400

450

0 100 200 300

ND

MA

FP,

ng/

l

DON in BP, ppb

y = -0.1151x + 133.75 R² = 0.0394

0

20

40

60

80

100

120

140

160

0 200 400 600

ND

MA

FP,

ng/

l

DON in HS, ppb

y = 0.3313x - 3.8795 R² = 0.4186

0

20

40

60

80

100

120

140

160

180

0 200 400 600

ND

MA

FP,

ng/

l

DON in BB, ppb

y = 0.0774x + 18.69 R² = 0.4735

0

10

20

30

40

50

60

0 200 400 600

ND

MA

FP,

ng/

l

DON in LMW, ppb

53

4.2 Results of UV/H2O2 system I

As discussed earlier, since DOM correlates to NDMAFP, it is important to study DOM

removal before assessing removal effect on NDMAFP. Experiments on system one tested

on optimal H2O2 dosage and reaction time on DOM removal, as well as UV/H2O2 effect

on NDMAFP.

4.2.1 DOM removal

4.2.1.1 DOC removal

Fig 4.9 shows that for unfiltered ASP effluent, addition of 200 ppm H2O2 dosage

achieved most removal. Maximum DOC removal was 24% at 120 minutes. DOC

concentration increased slightly or did not change in first 60 minutes. This is due to the

breakdown of large organic particles and the influence of poor UV transmittance. When

the particles were broken down and became dissolved, they contributed to DOC

concentration initially. Therefore, although there might be DOC removal, it was not

reflected in the measured DOC level. UVT of unfiltered sample was 56%. Poor UV

transmittance hindered UV radiation’s effect on H2O2, which leads to less amount of OH

radical generation. Therefore removal efficiency was not significant.

54

Figure 4.9 DOC degradation profiles of unfiltered ASP effluent under UV/H2O2 system I

with different H2O2 dosages

Fig 4.10 shows that after filtration, removal efficiency was improved. Maximum DOC

removal achieved 31% compared with 25% in the unfiltered sample. Linear relationship

was also observed between reaction time and DOC residual. The improvement in

performance comes from the following reasons: there was no DOC increase at the

beginning, since organic particles were filtered; higher UVT transmittance (76% in

filtered sample compared with 56% in unfiltered sample) improved UV radiation effect

on H2O2, therefore more OH radical was generated and better removal efficiency was

achieved.

Figure 4.10 Comparison of DOC degradation profiles of unfiltered and filtered ASP

effluents under UV/H2O2 system I ([H2O2] = 200 ppm)

0

0.2

0.4

0.6

0.8

1

1.2

0 30 60 90 120 150

DO

C, C

/Co

time, min

20 ppm

50 ppm

200 ppm

300 ppm

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0 30 60 90 120 150

DO

C, C

/Co

time, min

unfiltered

filtered

55

When the reaction time was extended to three hours, DOC removal was further

improved, with 71% final DOC removal (Fig 4.11). Linear correlation was also observed

over the extended reaction time. From LC-OCD-OND analysis results in Figure 4.12, it

can be seen that organic carbon containing in large molecular weight fractions was

degraded first. From 0 to 60 minutes, OC in biopolymer (BP) fraction decreased

significantly. This was followed by decreasing of OC in humic substance (HS) fraction.

Figure 4.11 DOC degradation profiles of filtered ASP effluent over the extended reaction

time under UV/H2O2 system I ([H2O2] = 200 ppm)

0

0.2

0.4

0.6

0.8

1

1.2

0 60 120 180 240 300 360

DO

C, C

/Co

time, min

56

Figure 4.12 OCD chromatography of filtered ASP effluent treated by UV/H2O2 system I

([H2O2] = 200 ppm)

Bio

poly

mers

Hum

ics

Build

ing B

locks

LM

W A

cid

s a

nd H

S

LMW Neutrals

-12

-11

-10

-9

-8

-7

-6

-5

-4

-3

-2

-1

0

1

2

3

4

5

6

7

8

9

10

11

12

0 20 40 60 80 100

rel. S

ign

al R

esp

on

se

Retention Time in Minutes

0 min

30 min

60 min

120 min

180 min

300 min

57

4.2.1.2 DON removal

DON removal is less significant than that of DOC. Figure 4.13 shows that DON

increased in the first 30 minutes and then deceased slowly. In the study of Chen et al.

(2010), UV/H2O2 was applied to 5% to 20% treated wastewater combined with surface

water, similar DON removal trend was observed.

Figure 4.13 DON degradation profiles of filtered ASP effluent under UV/H2O2 system I

([H2O2] = 200 ppm)

OND analysis in Figure 4.14 shows significant removal of DON in BP and HS fraction,

which is similar to that in OCD. However, DON in LMW fraction increased obviously.

This may be the reason of less DON removal in wastewater sample.

Both effects on DOC and DON demonstrate UV/H2O2’s effect on DOM removal in ASP

effluent. More DOC removal was achieved compared to DON. This agrees with results in

previous study (Chen et al, 2011). Both DOC and DON in BP and HS fractions were

efficiently removed. However, DON in LMW fraction increased significantly. DOM

removal in each fraction will be discussed in more details in section 4.3.

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0 60 120 180 240 300 360

DO

N, C

/Co

time, min

58

Figure 4.14 OND chromatography of filtered ASP effluent under UV/H2O2 system I

([H2O2] = 200 ppm)

Bio

poly

mers

Hum

ics

Build

ing B

locks

LM

W A

cid

s a

nd H

S

LMW Neutrals

0 min

30 min

60 min

120min

180 min

300 min

-12

-11

-10

-9

-8

-7

-6

-5

-4

-3

-2

-1

0

1

2

3

4

5

6

7

8

9

10

11

12

0 20 40 60 80 100

rel. S

ign

al R

esp

on

se

Retention Time in Minutes

59

4.2.2 NDMAFP removal

Figure 4.15 shows that NDMAFP decreased within first 60 minutes to 50% and started to

increase afterward. NDMAFP at 300 minutes reached 2.9 times of the initial value. This

indicates that prolonged UV/H2O2 reaction time had adverse effect on NDMAFP

removal. UV/H2O2 treatment may generate intermediate NDMA precursors. Previous

study demonstrated the generation of intermediate NDMA precursors, when UV/H2O2

was applied to diltiazem. Increase of NDMAFP was observed in first 30 minutes (Chen et

al., 2011). Figure 4.15 also shows that there is no direct correlation between DOC and

DON removal with NDMAFP removal.

Figure 4.15 Effect of UV/H2O2 system I on NDMAFP in ASP effluent ([H2O2] = 200

ppm)

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0

200

400

600

800

1000

1200

0 30 60 120 180 300

DO

C, D

ON

, C/C

o

ND

MA

FP,

ng/

l

time , min

NDMAFP

DOC

DON

60

4.3 Results of UV/H2O2 system II

System II was improved based on system I. UV intensity was increased. A stronger

LPUV lamp with UV intensity of 2 mW/cm2 replaced the previous one with UV intensity

of 1.3 mW/cm2. System configuration was also improved. Instead of having UV lamp

above water sample, in system II, UV lamp was submerged in water sample. This

ensured better contact between UV lamp and water sample; emitted UV radiation were

absorbed by the water sample more efficiently.

Similar to system I, DOC removal was linearly related to reaction time. However, DOC

removal rate increased significantly. Similar DOC removal within 300 minutes reaction

time in system I was achieved in 60 minutes in system II. The significant improvement

on DOC removal is due to stronger UV radiation emitted into water sample. This leads to

more OH radical generation, which means stronger oxidation capacity. Consequently,

DOC removal became faster.

4.3.1 DOM removal

4.3.1.1 Removal kinetics

Within 60 minutes, DOC decreased continuously with constant degradation rates (Fig

4.16). For 100 ppm and 50 ppm H2O2 dosage, degradation rates were 0.08 ppm DOC/min

and 0.07 ppm DOC/min respectively. Removal efficiencies of DOC were 65% and 55%.

Similar results were observed in a previous study, when LPUV/H2O2 system with UV

intensity of 40 mW/cm2 and H2O2 dosage of 11.2 ppm was applied to selected organic

compounds. Continuous decreasing trend and constant oxidation rate for DOC were

61

observed during 60 minutes reaction time. For histamine and caffeine, degradation rates

were 0.05 ppm DOC/min and 0.04 ppm DOC/min, respectively. Removal efficiencies of

DOC were 70% and 50%, respectively (Chen et al., 2011).

Figure 4.17 shows that DOC removal kinetics in each fraction was different from that of

total DOC. No linear relation was observed. Removal kinetics also differed among the

fractions. BP contains molecules with high MW (20,000 Da to 100,000Da). Initial DOC

concentration in BP was 894.1 ppb. DOC in BP decreased continuously and achieved

92% removal. DOC degraded fast from 0 to 20 minutes with an average degradation rate

of 33.4 ppb DOC/min; the degradation slowed down afterwards with an average

degradation rate of 3.8 ppb DOC/min. BP fraction contains mainly polysaccharide,

proteins and amino sugars, with polysaccharide as dominating component (Huber et al.,

2011). Polysaccharide contains aliphatic chains which are easily oxidized (Chen et al.,

2011). This contributes to the significant removal of BP.

HS contains molecules of MW around 1000 Da. HS was dominant in ASP effluent, with

initial DOC concentration of 2636.9 ppb. DOC in HS degraded continuously and total

removal was 81%. Average degradation rate was 35.2 ppb DOC/min. In a previous study

(Wang et al., 2010) when 108.8 ppm H2O2 was dosed in UV system treating humic

substance ([DOC]o = 5 ppm), degradation rate was 71.9 ppb DOC/min. However, HS was

dosed in DI water. The higher degradation rate may be due to the less complicated water

matrix. The HS solution was less complicated than ASP effluent, therefore, UV

transmittance was better and hydroxyl radical was more efficient.

62

BB comprises molecules of MW from 300 to 500 Da. Its initial DOC concentration was

1438.9 ppb. Low degree of removal (9%) was observed in this fraction with an average

DOC degradation rate of 2.1 ppb DOC/min. This may be due to the generation of

intermediate compounds in BB fraction from oxidation of BP.

LMW compounds (acids and neutrals) are of MW less than 350 Da. Its initial DOC

concentration was 1596.0 ppb. Significant removal (64%) was achieved. Concentration

increased in first 5 minutes and decreased with an average degradation rate of 22.7 ppb

DOC/min.

Figure 4.16 DOC degradation profiles of ASP effluent under UV/H2O2 system II

0

0.2

0.4

0.6

0.8

1

1.2

0 10 20 30 40 50 60 70 80

DO

C, C

/Co

time, min

100 ppm

50 ppm

63

Figure 4.17 DOC degradation profiles in four DOM fractions in ASP effluent under

UV/H2O2 system II ([H2O2] = 100 ppm)

DON degradation was less efficient than that of DOC. 30% removal was achieved (Fig

4.18). DON increased in the first five minutes and then decreased. No linear relation was

observed between DON removal and reaction time. In wastewater, approximately 0.6% -

13% DOM is in the form of combined amino acid, 14% - 25% is from protein (Dignac et

al., 2000; Manka & Rebhun, 1982). Hydroxyl radicals can break down these substances,

result in decreasing of DON. However, for complex organic structures, in particular when

nitrogen atoms are located in the centre of molecules, more oxidation capacity is required

(Chen et al., 2011).

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0 10 20 30 40 50 60 70 80

DO

C, C

/Co

time, min

BP

HS

BB

LMW

64

Figure 4.18 DON degradation profiles of ASP effluent under UV/H2O2 system II

DON degradation in each fraction (Fig 4.19) shows most removal was in BP, HS and BB

(63%, 74% and 76% respectively). However, there was a significant increase of DON in

LMW fraction. This is similar to the observation in system I.

Initial DON concentration in BP was 192.7 ppb, which was dominant among four

fractions. DON in BP degraded fast from 0 to 20 minutes with an average degradation

rate of 7.4 ppb DON/min; the degradation slowed down afterwards with an average

degradation rate of 0.09 ppb DON/min. Initial DON concentration in HS was 161.1 ppb,

which was the second highest among four fractions. Average DON degradation rate was

2.2 ppb DON/min. BB fraction contained initial DON of 73.2 ppb. Its average

degradation rate was 0.7 ppb DON/min. Initial DON concentration in LMW was the

lowest (41.1 ppb), however, it increased with an average rate of 2.8 ppb DON/min. DON

degradation kinetics were distinct from that of DOC. This indicates that removals of

DOC and DON in ASP effluent were not directly related under UV/H2O2 treatment.

0

0.2

0.4

0.6

0.8

1

1.2

0 10 20 30 40 50 60 70 80

DO

N, C

/Co

time, min

100 ppm

50 ppm

65

The increase of DON in LMW was observed in a previous study when UV/H2O2 was

applied to synthetic wastewater containing melanoidins (Dwyer & Lant, 2008). The

increase was significant, from 29% of initial nitrogen to 46%. It was reported that DON

associated with small molecular weight molecules (<1,000Da) cannot be chemically

oxidized easily, unlike the associated DOC (Dwyer & Lant, 2008). Therefore, the

increase of DON in LMW may be because of the accumulation of DON in smaller

molecules broken down from larger molecules.

Figure 4.19 DON degradation profiles in four DOM fractions in ASP effluent under

UV/H2O2 system II ( [H2O2] = 100 ppm)

0.000

0.500

1.000

1.500

2.000

2.500

3.000

3.500

4.000

4.500

0 10 20 30 40 50 60 70

DO

N, C

/Co

time, min

BP

HS

BB

LMW

66

4.3.1.2 Comparison of ASP and MBR effluents

ASP effluent and MBR effluent were selected to test the effect of organics composition

on DOM removal. MBR effluent had lower DOC and DON concentration (Table 3.1). It

also had lower DOC and DON in both BP and HS fractions (Table 4.3 and 4.4).

Table 4.2 Comparison of DOC and DON composition in ASP and MBR effluents

DOC, ppb DON, ppb

% DOC % DON

BP HS BB LMW BP HS BB LMW

ASP 894.06 2636.88 1438.93 1596.03 192.66 161.12 73.23 41.11

14% 40% 22% 24% 41% 34% 16% 9%

MBR 258.32 1950.81 1337.38 1508.72 16.60 114.73 90.17 43.20

5% 39% 26% 30% 6% 43% 34% 16%

Figure 4.20 shows similar DOC removal kinetics for both ASP and MBR effluents.

However, less DON was removed in MBR effluent. Significant increase of DON was

observed in MBR effluent within 5 minutes.

67

Figure 4.20 Comparison of DOC and DON removal in ASP and MBR effluents

DOC and DON increased significantly in BP fraction at 5 minutes of reaction (Fig 4.21).

These increases were not expected in UV/H2O2 process. With repeated experiments, the

same trend was observed. Within 5 minutes, DOC increased from 258 ppb to 316 ppb

and DON increased from 17 ppb to 129 ppb. After 5 minutes, DOC degraded with

average rates of 15.5 ppb DOC/min from 5 to 20 minutes and 1 ppb DOC/min from 20 to

60 minutes. DON dropped from 129 ppb to 40 ppb from 5 to 10 minutes with a rate of

17.8 ppb DON/min and stayed relatively constant afterwards. Except the difference in BP

fraction, DOC and DON removal trends in other three fractions were similar between

MBR and ASP effluents (Figs 4.17, 4.19, 4.21).

0

0.2

0.4

0.6

0.8

1

1.2

0 10 20 30 40 50 60 70

DO

C, C

/Co

time, min

ASP

MBR

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

0 10 20 30 40 50 60 70

DO

N, C

/Co

time, min

ASP

MBR

68

The observation indicates the formation of large molecules in MBR effluent at the initial

stage of UV/H2O2 treatment. This may be caused by UV irradiation under which small

organic molecules can be polymerized to form large polymers. For example, in UV

polymerization technology, monomer is formed into polymer under UV irradiation (Hu et

al., 2009).

Figure 4.21 DOC and DON degradation profiles for four DOM fractions in MBR effluent

under UV/H2O2 system II

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0 10 20 30 40 50 60 70

DO

C, C

/Co

time, min

BPHSBBLMW

0

1

2

3

4

5

6

7

8

0 10 20 30 40 50 60 70

DO

N, C

/Co

time, min

BPHSBBLMW

69

4.3.2 NDMAFP removal

4.3.2.1 Removal kinetics

Figure 4.22 shows the result of NDMAFP removal. For both 100 ppm and 50 ppm H2O2

dosages, humps were observed in NDMAFP profiles. NDMAFP decreased at the

beginning, then increased significantly and finally decreased. With 100 ppm H2O2

dosage, NDMAFP reached highest value of 404 ng/L at 20 minutes; 80% NDMAFP

removal was achieved after 60 minute treatment. With 50 ppm H2O2 dosage, NDMAFP

reached a higher value of 739 ng/L at a later time, 40 minutes. No NDMAFP removal

was achieved within 60 minutes. Peak NDMAFP for 50ppm dosage was 1.89 times of the

initial NDMAFP.

Figure 4.22 NDMAFP of ASP effluent under UV/H2O2 system II

0

100

200

300

400

500

600

700

800

900

0 5 10 20 30 40 60

ND

MA

FP,

ng/

l

time, min

50ppm

100ppm

70

Figure 4.23 NDMAFP per DON of ASP effluent under UV/H2O2 system II

NDMA per DON profiles (Fig 4.23) were similar to that of NDMAFP. 1 mg/L DON

yielded highest NDMAFP of 2979 ng/L (2.979 μg/L) at 40 minutes with 50 ppm H2O2

dosage. This is smaller than the value tested with diltiazem under UV/H2O2 (Chen et al.,

2011), where 1 mg/L DON yielded 59 μg/L NDMA. H2O2 dosage of 50 ppm yielded

more NDMA compared with 100 ppm dosage. This is consistent with the finding that

inadequate quantity of oxidant yielded maximum of NDMA (Chen et al., 2011).

Figure 4.24 Ratio of NDMAFP and initial NDMAFP of ASP effluent under UV/H2O2

system II

0

500

1000

1500

2000

2500

3000

3500

0 5 10 20 30 40 60

ND

MA

FP/D

ON

(n

g m

g-1

)

time, min

50ppm

100ppm

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

2

0 20 40 60 80

ND

MA

FP/i

nit

ial N

DM

AFP

time, min

100ppm

50ppm

71

Figure 4.24 shows NDMAFP decreased slightly at 5 minutes. With H2O2 dosage of 100

ppm, NDMAFP increased at a constant rate of 12.9 ng/L/min between 5 and 20 minutes.

For 50 ppm dosage, NDMAFP increased at a constant rate of 10.4 ng/L/min between 5

and 40 minutes. These NDMAFP increases infer that at the beginning of UV/H2O2

reaction, some organics were degraded to form intermediate NDMA precursors; as a

result the amount of NDMA precursors increased. However, with longer reaction time,

total amount of NDMA precursors could be decreased. NDMAFP removal profiles may

be H2O2 dosage dependent. In this study, higher H2O2 dosage resulted in higher increase

rate in shorter period and larger degree of removal. In a similar study when NDMAFP of

diltiazem was tested under UV/ H2O2 treatment, comparison was done between single

H2O2 dosage of 220 ppm and initial H2O2 dosage of 108 ppm with 88 ppm addition every

10 minutes. Results from both conditions showed increases of NDMAFP followed by

decreases; however, the later showed a lower increase rate in longer period and larger

degree of removal (Chen et al., 2011). To further understand NDMAFP removal profiles’

dependency on H2O2 dosage, NDMAFP removal under various H2O2 dosages and dosing

methods should be studied.

The similarity in NDMAFP removal profiles of 50 and 100 ppm dosages may be because

of the same organics removal kinetics. As shown previously, both 100ppm and 50ppm

systems had similar DOC and DON removal trends (Figs 4.16 and 4.18). Furthermore,

both had similar DOC and DON removal trends in each fraction of organics (Figs 4.17,

4.19, 4.25 and 4.26).

72

Figure 4.25 DOC degradation profiles in four DOM fractions in ASP effluent under

UV/H2O2 system II ([H2O2] = 50 ppm)

Figure 4.26 DON degradation profiles in four DOM fractions in ASP effluent under

UV/H2O2 system II ( [H2O2] = 50 ppm)

Better NDMAFP removal efficiency with 100ppm dosage may be due to faster DOM

removal. With 100 ppm H2O2, DOC removal was faster, and more DOC was removed in

each fraction (Fig 4.27). In terms of DON, although total DON removal was similar for

0.000

0.200

0.400

0.600

0.800

1.000

1.200

1.400

1.600

1.800

0 20 40 60 80

DO

C, C

/Co

time, min

BP

HS

BB

LMW

0.000

0.500

1.000

1.500

2.000

2.500

3.000

3.500

4.000

0 20 40 60 80

DO

N, C

/Co

time, min

BP

HS

BB

LMW

73

both H2O2 dosages, 100 ppm system had faster DON increase in LMW (Fig 4.28). This

indicates faster oxidation effect on DON with higher H2O2 dosage.

Figure 4.27 Comparison of DOC in four DOM fractions in ASP effluent with different

H2O2 dosages

0

0.2

0.4

0.6

0.8

1

1.2

0 5 10 20 30 40 60

DO

C in

BP

, C/C

o

time, min

[H2O2]=100ppm

[H2O2]=50ppm

0

0.2

0.4

0.6

0.8

1

1.2

0 5 10 20 30 40 60

DO

C in

HS,

C/C

o

time, min

[H2O2]=100ppm

[H2O2]=50ppm

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

0 5 10 20 30 40 60

DO

C in

BB

, C/C

o

time, min

[H2O2]=100ppm[H2O2]=50ppm

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0 5 10 20 30 40 60

DO

C in

LM

W, C

/Co

time, min

[H2O2]=100ppm

[H2O2]=50ppm

74

Figure 4.28 Comparison of DON in four DOM fractions in ASP effluent with different

H2O2 dosages

Under UV/H2O2 treatment, although DOM was decreasing, there was generation of

intermediate NDMA precursors. Depending on the treatment condition, NDMAFP could

increase up to three times of initial NDMAFP (in system I). In other study, NDMAFP

increased up to 2.3 times with UV/H2O2 treatment on diltiazem (Chen et al., 2011). With

stronger oxidation effect, intermediate NDMA precursors could be further degraded.

These results indicate that during UV/H2O2 treatment, it is bias to use DOM

concentration as an indicator for NDMAFP removal. Alternatively, one potential suitable

parameter to indicate for NDMAFP removal would be DON concentration in LMW,

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0 5 10 20 30 40 60

DO

N in

BP

, C/C

o

time, min

[H2O2]=100ppm[H2O2]=50ppm

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0 5 10 20 30 40 60

DO

N in

HS,

C/C

o

time, min

[H2O2]=100ppm[H2O2]=50ppm

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0 5 10 20 30 40 60

DO

N in

BB

, C/C

o

time, min

[H2O2]=100ppm

[H2O2]=50ppm

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

0 5 10 20 30 40 60

DO

N in

LM

W, C

/Co

time, min

[H2O2]=100ppm[H2O2]=50ppm

75

since the increase of DON in LMW indicates the degree of nitrogen associated organic

oxidation. Caution should be taken to remove intermediate NDMA precursors by dosing

sufficient oxidant (100 ppm in this study) and providing adequate reaction time (60

minutes in this study).

4.3.2.2 Comparison of ASP and MBR effluents

Within 60 minutes, there was no hump presented in NDMAFP profile for MBR effluent

(Fig 4.29). NDMAFP reached a plateau after 20 minutes. Further comparison on H2O2

dosage showed no NDMAFP removal at 60 minutes with 75, 100, 125 and 150ppm

dosage (Fig 4.30). This may be explained by the formation of large nitrogen rich

molecules at the initial stage of UV/H2O2 treatment. This formation may increase the

need of oxidation capacity to degrade NDMA precursors. It was reported that continuous

H2O2 addition could achieve more DON removal (Chen et al., 2011). As discussed in

section 4.1, NDMAFP correlates with DON fractions; therefore the method of continuous

H2O2 dosing may be used to improve NDMAFP removal for MBR effluent. The

difference between NDMAFP removal kinetics in ASP and MBR effluents indicates that

UV/H2O2 treatment for NDMAFP is water specific. Further research is needed to

evaluate UV/H2O2 performance on different wastewater samples with different organics

composition.

76

Figure 4.29 NDMAFP of MBR effluent under UV/H2O2 system II ([H2O2] = 100 ppm)

Figure 4.30 NDMAFP of MBR effluent under UV/H2O2 system II with different H2O2

dosages over 60 minutes

0

50

100

150

200

250

300

350

400

0 5 10 20 30 40 60

ND

MA

FP,

ng/

l

time, min

0

50

100

150

200

250

300

350

75 100 125 150

ND

MA

FP,

ng/

l

H2O2, ppm

initial NDMAFP

NDMAFP

77

CHAPTER 5 CONCLUSIONS AND RECOMMENDATIONS

5.1 Conclusions

This study examined NDMAFP and its removal in treated wastewater collected from

ASP and MBR treatment systems. Similar NDMA formation kinetics was observed for

both effluents. NDMAFP resulted from chloramination was in the range of 240 to 400

ng/L. Within experiment scope, no correlation (R2

= 0.16) was observed between

NDMAFP and DON. Further correlation analyses were carried out between NDMAFP

and each fraction of DON defined by LC-OCD-OND. Results showed weak correlation

(R2=0.45) between NDMAFP and DON in molecules with molecular weight smaller than

500 Da. Primary effluent results were not included in the analysis, since extended

reaction time and/or additional chloramine dosage is needed to measure its NDMAFP.

UV/H2O2 treatment was proved to be able to remove DOC efficiently (up to 70% for both

effluents) and remove DON with less efficiency (up to 30% or 20%) in ASP and MBR

effluents. UV/H2O2 removed NDMAFP in ASP effluent up to 80%. This indicates the

potential usage of UV/H2O2 as a mean to remove DOM and NDMAFP in treated effluent.

However, under the experimental condition, no NDMAFP removal was achieved in MBR

effluent. This shows using UV/H2O2 process to remove NDMAFP may be water specific.

DOM removal in each fraction exhibited distinct kinetics. During the UV/H2O2 treatment

on ASP effluent, DOC was degraded in all four fractions (Biopolymers, Humic

Substance, Building Block and Low Molecular Weight compounds); DON was decreased

in first three fractions but increased significantly in low molecular weight (LMW)

78

fraction. This demonstrates the difficulty in treating DON in LMW which is the challenge

in DON removal and also in N-DBPs control. For MBR effluent, similar removal kinetics

was observed expect the unusual increases of DOC and DON in biopolymers fraction at

the beginning. This indicates the potential formation of large molecules in MBR effluent

at initial stage of UV/H2O2 treatment. This may cause difficulty in the treatment for MBR

effluent.

Results also indicated the generation of intermediate NDMA precursors during the

treatment, where high yield of NDMA was generated from insufficient oxidant dosing.

No linear correlation was observed between DOM and NDMAFP. These bring attentions

for the UV/H2O2 treatment aiming to remove NDMAFP: using DOM concentration as an

indicator for NDMAFP could be biased; sufficient oxidant should be applied to lower

NDMA yield during the treatment and to achieve NDMAFP removal. In this study, 100

ppm H2O2 dosage with 60 minutes treatment time was proved to be suitable for ASP

effluent.

5.2 Recommendations

5.2.1 Improvement on NDMA detection

NDMA detection method can be improved to reduce sample volume and shorten analysis

time. In this study, 1L of water sample was needed to go through SPE which took

approximately 8h. To improve, solid-phase micro-extraction (SPME) could be adopted.

The method of SPME coupled with GC and chemical ionization MS/MS requires only

79

less than 5ml water sample and the analysis needs only 1.5h (Hung et al., 2010). This will

be able to fasten NDMA analysis. Therefore more water samples can be examined.

On the other hand, NDMA analysis sensitivity can be further improved. During LC-

MS/MS analysis, the fragmentation patterns from electron impact ionization mass

spectrum were not favorable. Molecular ion at 75m/z may be a reliable quantitation ion;

however the confirming ions at 42 and 43m/z were hardly unique. To overcome this,

positive chemical ionization (PCI) can be applied. At the same time, utilizing large

volume injection method could help lower the detection limit.

5.2.2 Improvement on UV/H2O2 process

It was shown that by increasing UV intensity, performance has improved in system two.

Further improvement should be done to increase UV intensity by choosing UV lamp with

stronger UV output or improving configuration of the system to ensure more UV

radiation emitted to wastewater sample.

5.2.3 Study on wastewater characteristics

This study mainly focused on two wastewater characteristics, DOC and DON. Other

wastewater characteristics could be analyzed; for example: SUVA which indicates

unsaturated bond or aromaticity; chloride concentration and alkalinity which might

inhibit oxidation process (Liao et al., 2001). In this study, two types of treated wastewater

were collected. For future study, more types of treated wastewater from different

treatment processes and different treatment plants should be tested. These would provide

a wider variety of organic composition.

80

BIBLIOGRAPHY

Asami, M., Oya, M., & Kosaka, K. (2009). A nationwide survey of NDMA in raw and

drinking water in Japan, Sci. Total Environ. 207(2009) 3540-3545.

Badr, El-Sayed A, Achterberg, Eric P, Tappin, Alan D, Hill, Steve J, & Braungardt,

Charlotte B. (2003). Determination of dissolved organic nitrogen in natural

waters using high-temperature catalytic oxidation. TrAC Trends in Analytical

Chemistry, 22(11), 819-827.

Bond, T., Huang, J., Templeton, M. R., & Graham, N. (2011). Occurrence and control of

nitrogenous disinfection by-products in drinking water - A review. Water

Research, 45(15), 4341-4354. doi: 10.1016/j.watres.2011.05.034

CDPH. (2011). California Department of Public Health: NDMA and Other Nitrosamines

- Drinking Water Issues. Retrieved 1/22, 2013, from

http://www.cdph.ca.gov/certlic/drinkingwater/pages/NDMA.aspx

Charrois, Jeffrey WA, Arend, Markus W, Froese, Kenneth L, & Hrudey, Steve E. (2004).

Detecting N-nitrosamines in drinking water at nanogram per liter levels using

ammonia positive chemical ionization. Environmental science & technology,

38(18), 4835-4841.

Charrois, Jeffrey WA, Boyd, J.M., Froese, S.E., & Hrudey, Steve E. (2007) Occurence of

Nnitrosoamines in Alberta public drinking-water distribution systems.

J.Environ. Eng. Sc, 6(2007) 103–114.

Chen, B., Kim, Y., & Westerhoff, P. (2011). Occurrence and treatment of wastewater-

derived organic nitrogen. Water Research, 45(15), 4641-4650. doi:

10.1016/j.watres.2011.06.018

Chen, H. W., Chen, C. Y., & Wang, G. S. (2010). Influence of ultraviolet light coupled

with hydrogen peroxide treatment on organic nitrogen and carbon precursors and

disinfection by-product formation. Environmental Engineering Science, 27(5),

423-430. doi: 10.1089/ees.2009.0431

Chen, H. W., Chen, C. Y., & Wang, G. S. (2011). Performance evaluation of the

UV/H2O2 process on selected nitrogenous organic compounds: Reductions of

organic contents vs. corresponding C-, N-DBPs formations. Chemosphere, 85(4),

591-597. doi: 10.1016/j.chemosphere.2011.06.090

Deeb, R., Hawley, E., Sedlak, D., Loveland, J., & Kavanaugh, M. (2006). Removal and

destruction of NDMA and precursors during wastewater treatment. Proceedings

of the Water Environment Federation, 2006(12), 1468-1477.

Dignac, M.F., Ginestet, P., Rybacki, D., Bruchet, A., Urbain, V., & Scribe, P. (2000).

Fate of wastewater organic pollution during activated sludge treatment: nature of

residual organic matter. Water Research, 34(17), 4185-4194.

Dwyer, J., & Lant, P. (2008). Biodegradability of DOC and DON for UV/H2O2 pre-

treated melanoidin based wastewater. Biochemical Engineering Journal, 42(1),

47-54. doi: 10.1016/j.bej.2008.05.016

Farré, M.J., Döderer, K., Keller, J., & Gernjak, W. (2010). N-nitrosodimethylamine

(NDMA) in Purified Recycled Water.

http://www.urbanwateralliance.org.au/publications/UWSRA-tr34.pdf

Farré, M.J., Keller, J., Holling, N., Poussade, Y., & Gernjak, W. (2011). Occurrence of

N-nitrosodimethylamine precursors in wastewater treatment plant effluent and

their fate during ultrafiltration-reverse osmosis membrane treatment. Water

Science and Technology, 63(4), 605.

Fujioka, T., Khan, S. J., Poussade, Y., Drewes, J. E., & Nghiem, L. D. (2012) N-

nitrosamine removal by reverse osmosis for indirect potable water reuse–A

81

critical review based on observations from laboratory, pilot and full scale

studies. Separation and Purification Technology.

Gang, D., Clevenger, T. E., & Banerji, S. K. (2003). Relationship of chlorine decay and

THMs formation to NOM size. Journal of hazardous materials, 96(1), 1-12.

Gerecke, A.C., & Sedlak, D.L. (2003). Precursors of N-nitrosodimethylamine (NDMA)

in natural waters. Environ.Sci. Technol, 37, 1331–1336.

Gur-Reznik, S., Katz, I., & Dosoretz, C. G. (2008). Removal of dissolved organic matter

by granular-activated carbon adsorption as a pretreatment to reverse osmosis of

membrane bioreactor effluents. Water research, 42(6), 1595-1605.

Hu, X., Zhang, J., & Yang, W. (2009). Preparation of transparent polystyrene nano-

latexes by an UV-induced routine emulsion polymerization. Polymer,50(1), 141-

147.

Huber, S.A., Balz, A., Abert, M., & Pronk, W. (2011). Characterisation of aquatic humic

and non-humic matter with size-exclusion chromatography–organic carbon

detection–organic nitrogen detection (LC-OCD-OND). Water research, 45(2),

879-885.

Huber, S. A., Balz, A., & Abert, M. (2011). New method for urea analysis in surface and

tap waters with LC-OCD-OND (liquid chromatography-organic carbon

detection―organic nitrogen detection). Journal of water supply: research and

technology. AQUA 60(3)): 159-166.

Hung, H. W., Lin, T. F., Chiu, C. H., Chang, Y. C., & Hsieh, T. Y. (2010). Trace

Analysis of N-Nitrosamines in Water Using Solid-Phase Microextraction

Coupled with Gas Chromatograph–Tandem Mass Spectrometry. Water, Air, &

Soil Pollution, 213(1), 459-469.

Kommineni, S. , Zoeckler, J., Stocking, A.J., Liang, S., Flores, A.E., & Kavanaugh, M.C.

(2000). Treatment Technologies for Removal of Methyl Tertiary Butyl Ether

(MTBE) from Drinking Water. Center for Groundwater Restoration and

Protection, National Water Research Institute, 2000

Liao, C. H., Kang, S. F., & Wu, F. A. (2001). Hydroxyl radical scavenging role of

chloride and bicarbonate ions in the H2O2/UV process. Chemosphere, 44(5),

1193-1200.

Leenheer, J. A., & Croué, J.P. (2003). Characterizing aquatic dissolved organic matter.

Environmental Science & Technology, 37(1), 18-26.

Leenheer, J.A., Dotson, A., & Westerhoff, P. (2007). Dissolved organic nitrogen

fractionation. Annals of Environmental Science, 1(1), 6.

Li, K., Hokanson, D. R., Crittenden, J. C., Trussell, R. R., & Minakata, D. (2008).

Evaluating UV/H2O2 processes for methyl tert-butyl ether and tertiary butyl

alcohol removal: Effect of pretreatment options and light sources. Water

Research, 42(20), 5045-5053.

Luo, X, Clevenger, T.E., & Deng, B. (2005). Role of NOM in the formation of N-

nitrosodimethylamine (NDMA) in surface waters. Paper presented at the Papers,

229th ACS National Meeting, San Diego, CA, United States.

Manka, J., & Rebhun, M. (1982). Organic groups and molecular weight distribution in

tertiary effluents and renovated waters. Water Research, 16(4), 399-403.

Membrane Bio-Reactor Demonstration Plant at Ulu Pandan Water Reclamation Plant

(2012)

Mitch, W. A. (2009). Occurrence and formation of nitrogenous disinfection by-products.

Water Research Foundation.

Mitch, W.A., Gerecke, A.C., & Sedlak, D.L. (2003). A N-Nitrosodimethylamine

(NDMA) precursor analysis for chlorination of water and wastewater. Water

Research, 37(15), 3733-3741.

82

Mitch, W.A., & Sedlak, D.L. (2002). Formation of N-nitrosodimethylamine (NDMA)

from dimethylamine during chlorination. Environmental Science & Technology,

36(4), 588-595.

Mitch, W.A., & Sedlak, D.L. (2004). Characterization and fate of N-

nitrosodimethylamine precursors in municipal wastewater treatment plants.

Environmental science & technology, 38(5), 1445-1454.

Mitch, W.A., Sharp, J.O., Trussell, R.R., Valentine, R.L., Alvarez-Cohen, L., & Sedlak,

D.L. (2003). N-nitrosodimethylamine (NDMA) as a drinking water contaminant:

a review. Environmental Engineering Science, 20(5), 389-404.

MOE. (2000). Ontario Ministry of the Environment and Energy. Regulation Made Under

the Ontario Water Resources Act: Drinking Water Protection—Larger Water

Works.

Najm, I., & Trussell, R.R. (2001). NDMA formation in water and wastewater. J AWWA,

92–99.

Naidu, G., Jeong, S., Vigneswaran, S., & Rice, S. A. (2013). Microbial activity in

biofilter used as a pretreatment for seawater desalination. Desalination, 309, 254-

260.

New concepts of UV/H2O2 oxidation. (2011). Retrieved Jan 10 from

http://www.waterrf.org/PublicReportLibrary/4040.pdf

Pehlivanoglu-Mantas, E., & Sedlak, D.L.(2008). Measurement of dissolved organic

nitrogen forms in wastewater effluents: Concentrations, size distribution and

NDMA formation potential. Water Research 42, 3890-3898.

Pehlivanoglu-Mantas, E., & Sedlak, D.L.(2006). Wastewater-derived dissolved organic

nitrogen: analytical methods, characterization, and effects—a review. Critical

Reviews in Environmental Science and Technology, 36(3), 261-285.

Penru, Y., Simon, F. X., Guastalli, A. R., Esplugas, S., Llorrens, J., & Baig, S. (2011,

July). Characterization of natural organic matter from Mediterranean coastal

seawater. In Fourth IWA Specialty Conference on Natural Organic Matter: From

Source to Tap and Beyond.

Plumlee, M.H., López-Mesas, M., Heidlberger, A., Ishida, K.P., & Reinhard, M. (2008).

N-nitrosodimethylamine (NDMA) removal by reverse osmosis and UV treatment

and analysis via LC–MS/MS. Water research, 42(1), 347-355.

Richardson, S.D. (2003). Disinfection by-products and other emerging contaminants in

drinking water. TrAC Trends in Analytical Chemistry, 22(10), 666-684.

Richardson, S. D., Plewa, M. J., Wagner, E. D., Schoeny, R., & DeMarini, D. M. (2007).

Occurrence, genotoxicity, and carcinogenicity of regulated and emerging

disinfection by-products in drinking water: a review and roadmap for research.

Mutation Research/Reviews in Mutation Research, 636(1), 178-242.

Roberson, J.A. . (2010). How USEPA Is looking at regulating groups. Water Works

Assoc. 102, 16-18.

Sedlak, D.L., & Kavanaugh, M. (2005). Removal and destruction of NDMA and NDMA

precursors during wastewater treatment.

Shannon, M. A., Bohn, P. W., Elimelech, M., Georgiadis, J. G., Mariñas, B. J., & Mayes,

A. M. (2008). Science and technology for water purification in the coming

decades. Nature, 452(7185), 301-310.

Treatment process and basic design principles. (n.d.), Retrieved March 9, 2013, from

Sustainable sanitation and water management website,

http://www.sswm.info/sites/default/files/toolbox/ENDRESS%2BHAUSER%202

002%20Activated%20Sludge_medium.jpg

U.S.EPA. (2008). Emerging Contaminant - N-Nitrosodimethylamine (NDMA) Fact

Sheet.

83

USEPA. (1996). Test Methods for Evaluating Solid Wastes: Physical/Chemical Methods.

Publication 955-001-00000-1.

USEPA. (2010). Second cycle of the unregulated contaminant monitoring regulation

(UCMR2) data summary.

Wang, G. S., Liao, C. H., Chen, H. W., & Yang, H. C. (2006). Characteristics of natural

organic matter degradation in water by UV/H2O2 treatment. Environmental

technology, 27(3), 277-287.

Water Reuse: Potential for Expanding the Nation's Water Supply through Reuse of

Municipal Wastewater (2012). Washington, D.C.: The National Academies

Press.

Wei, L. L., Zhao, Q. L., Xue, S., & Jia, T. (2008). Removal and transformation of

dissolved organic matter in secondary effluent during granular activated carbon

treatment. Journal of Zhejiang University: Science A, 9(7), 994-1003. doi:

10.1631/jzus.A071508

Weissmahr, K.W., & Sedlak, D.L. (2000). Effect of metal complexation on the

degradation of dithiocarbamate fungicides. Environ. Toxicol. Chem., 19, 820–

826.

Westerhoff, P., Aiken, G., Amy, G., & Debroux, J. (1999). Relationships between the

structure of natural organic matter and its reactivity towards molecular ozone and

hydroxyl radicals. Water Research, 33(10), 2265-2276. doi: 10.1016/S0043-

1354(98)00447-3

WHO. (2006). Guidelines for drinking-water quality, third edition. Retrieved from,

http://www.who.int/water_sanitation_health/dwq/gdwq3rev/en/

Xing, W., Ngo, H. H., Kim, S. H., Guo, W. S., & Hagare, P. (2008). Adsorption and

bioadsorption of granular activated carbon (GAC) for dissolved organic carbon

(DOC) removal in wastewater. Bioresource Technology, 99(18), 8674-8678. doi:

10.1016/j.biortech.2008.04.012

Zhang, J., Loong, W. L. C., Chou, S., Tang, C., Wang, R., & Fane, A. G. (2012).

Membrane biofouling and scaling in forward osmosis membrane

bioreactor. Journal of Membrane Science. 403-404, 8-14.

Zhao, Y. Y., Boyd, J., Hrudey, S. E., & Li, X. F. (2006). Characterization of new

nitrosamines in drinking water using liquid chromatography tandem mass

spectrometry. Environmental science & technology, 40(24), 7636-7641.

Zheng, X., Ernst, M., & Jekel, M. (2010). Pilot-scale investigation on the removal of

organic foulants in secondary effluent by slow sand filtration prior to

ultrafiltration. Water research, 44(10), 3203-3213.

Zheng, X., Mehrez, R., Jekel, M., & Ernst, M. (2009). Effect of slow sand filtration of

treated wastewater as pre-treatment to UF. Desalination, 249(2), 591-595.

84

APPENDICES

Appendix 1 List of toxic DBPs

(Richardson et al, 2007)

85

86

87

88

Appendix 2 Table of potential NDMA precursors

(Source: data from REAXY database)

89

90

91

92

93

Appendix 3 Membrane bio-reactor specification

(Source: “Membrane Bio-Reactor Demonstration Plant”, 2012)

94

Appendix 4 Activated sludge process specification

(Source: “Treatment process and basic design principles”, 2013)

ASP