laboratory column study for evaluating a multimedia permeable reactive barrier for the remediation...

This article was downloaded by: [Monash University Library]On: 04 December 2014, At: 10:15Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Environmental TechnologyPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/tent20

Laboratory column study for evaluating a multi-mediapermeable reactive barrier for the remediation ofammonium contaminated groundwaterXiangke Konga, Erping Bib, Fei Liub, Guoxin Huangc & Jianfei Maa

a Institute of Hydrogeology & Environmental Geology, CAGS, Shijiazhuang, Chinab School of Water Resources and Environment, China University of Geosciences (Beijing),Beijing, Chinac Beijing Academy of Food Sciences, China Meat Research Center, Beijing, ChinaAccepted author version posted online: 27 Nov 2014.

To cite this article: Xiangke Kong, Erping Bi, Fei Liu, Guoxin Huang & Jianfei Ma (2014): Laboratory column study forevaluating a multi-media permeable reactive barrier for the remediation of ammonium contaminated groundwater,Environmental Technology, DOI: 10.1080/09593330.2014.992482

To link to this article: http://dx.doi.org/10.1080/09593330.2014.992482

Disclaimer: This is a version of an unedited manuscript that has been accepted for publication. As a serviceto authors and researchers we are providing this version of the accepted manuscript (AM). Copyediting,typesetting, and review of the resulting proof will be undertaken on this manuscript before final publication ofthe Version of Record (VoR). During production and pre-press, errors may be discovered which could affect thecontent, and all legal disclaimers that apply to the journal relate to this version also.

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

http://www.tandfonline.com/loi/tent20

http://www.tandfonline.com/action/showCitFormats?doi=10.1080/09593330.2014.992482

http://dx.doi.org/10.1080/09593330.2014.992482

http://www.tandfonline.com/page/terms-and-conditions

http://www.tandfonline.com/page/terms-and-conditions

XK. Kong et al. multi-media PRB for remediating ammonium contaminated groundwater

1

Publisher: Taylor & Francis 1

Journal: Environmental Technology 2

DOI: 10.1080/09593330.2014.992482 3

4

5

Laboratory column study for evaluating a multi-media permeable 6

reactive barrier for the remediation of ammonium contaminated 7

groundwater 8

Xiangke Konga, Erping Bib*, Fei Liub, Guoxin Huangc, Jianfei Maa 9

aInstitute of Hydrogeology & Environmental Geology, CAGS, Shijiazhuang, China; 10

bSchool of Water Resources and Environment, China University of Geosciences 11

(Beijing), Beijing, China; cChina Meat Research Center, Beijing Academy of Food 12

Sciences, Beijing, China 13

Abstract: In order to remediate the ammonium contaminated groundwater, an 14

innovative multi-media permeable reactive barrier (M-PRB) was proposed, which 15

consisted of sequential columns combining oxygen releasing compound (ORC), 16

zeolite, spongy iron and pine bark in the laboratory-scale. Results showed that both 17

ammonium and nitrate could be reduced to levels below the regulatory discharge 18

limits through ion exchange and microbial degradation (nitrification and 19

denitrification) in different compartments of the M-PRB system. The concentration of 20

dissolved oxygen (DO) increased from 2 mg/L to above 20 mg/L after the simulated 21

groundwater flew through the oxygen releasing column packed with ORC, 22

demonstrating that ORC could supply sufficient oxygen for subsequent microbial 23

nitrification. Ammonium was efficiently removed from about 10 mg N/L to below 0.5 24

mg N/L in the aerobic reaction column which was filled with biological zeolite. After 25

54 operating days, above 70% ammonium could be removed by microbial nitrification 26

* Corresponding author. E-mail: [email protected]

Accep

ted

Man

uscr

ipt

Dow

nloa

ded

by [

Mon

ash

Uni

vers

ity L

ibra

ry]

at 1

0:15

04

Dec

embe

r 20

14

2

in the aerobic reaction column, indicating that the combination use of ion exchange 27

and nitrification by biological zeolite could ensure the ammonium removal efficiency 28

high and sustainable. To avoid the second pollution of nitrate that produced by the 29

former nitrification, spongy iron and pine bark were used to remove oxygen and 30

supply organic carbon for heterotrophic denitrification in the oxygen removal column 31

and anaerobic reaction column separately. The concentration of nitrate decreased from 32

14 mg N/L to below 5 mg N/L through the spongy iron-based chemical reduction and 33

microbial denitrification. 34

Keywords: ammonium; groundwater; multi-media permeable reactive barrier; ion 35

exchange; microbial degradation 36

37

1. Introduction 38

Nowadays, the rapid industrial and agricultural development brings massive 39

contaminants to groundwater, in which ammonium is the most commonly 40

encountered nitrogenous compound [1-2]. Groundwater aquifers, especially those 41

locating adjacent to surface water supplies are easily contaminated by ammonium due 42

to the recharge of contaminated rivers and lakes [3], which has seriously threaten the 43

public water supplies and deteriorated groundwater quality [4]. Thus, available 44

methods for the efficient treatment of ammonium contaminated groundwater are 45

receiving increased attention [5]. 46

47

Permeable reactive barrier (PRB) has been demonstrated to be a promising in-situ 48

remediation technology, which can offer a cost-effective way to control large-scale 49

contaminated groundwater plumes [6-7]. It is installed at the downgradient of 50

groundwater recharge path, and the contaminants could be removed passively by the 51

Accep

ted

Man

uscr

ipt

Dow

nloa

ded

by [

Mon

ash

Uni

vers

ity L

ibra

ry]

at 1

0:15

04

Dec

embe

r 20

14


3

reactive materials filled in PRB [8]. 52

53

Commonly, ion exchange has been used in various types of PRB for the removal of 54

ammonium, and the utilization of natural zeolite is considered to be a competitive and 55

effective treatment method due to its low cost and relative simplicity of application 56

and operation [9]. But it has to face the bottleneck that zeolite will finally reach its 57

maximum adsorption ability [10] and even desorb ammonium when the concentration 58

in solution is lower than that in zeolite [11]. The sequential reactive barrier combining 59

biological degradation and ion exchange processes is thought to be a prospective 60

technology [12,13]. However, the growing of nitrifying bacteria in aquifers is a hard 61

process in groundwater environment [14] in that the concentration of DO is usually 62

below 3 mg/L [15]. 63

64

Nooten et al. [16] firstly described a laboratory-scale multifunctional permeable 65

reactive barrier with clinoptilolite for the removal of ammonium from landfill 66

leachate, which has thrown a new light on sustainable ammonium removal. However, 67

different from landfill leachate, groundwater commonly faces a relative low 68

concentration of ammonium contamination (a few to dozens of mg N/L), which 69

indicates that the indigenous bacteria in aquifer may play an important role in the 70

ammonium depletion by biological nitrification process in an aerobic environment 71

[17]. Besides high adsorption capacity for ammonium [18], zeolite also has a high 72

specific surface area and an abundant porous structure which are helpful for the 73

nitrifying bacteria growing on. The growth of microorganisms on the surface of the 74

zeolite immerged in ammonium aqueous solution or wastewater has been confirmed 75

in previous studies [19,20]. Furthermore, biofilm covered on zeolite does not affect 76

Accep

ted

Man

uscr

ipt

Dow

nloa

ded

by [

Mon

ash

Uni

vers

ity L

ibra

ry]

at 1

0:15

04

Dec

embe

r 20

14

4

the ion exchange (for the smaller particle size 1-3.2 mm), and can realize the recovery 77

of ion exchange capacity of zeolite [20]. Thereby, successfully creating an aerobic 78

environment for PRB filled with zeolite is of great importance to ensure sustainable 79

use for in situ groundwater remediation. 80

81

In this study, a set of multi-media columns in series was designed to simulate a 82

multi-media permeable reactive barrier (M-PRB) for ammonium removal, which 83

described a strategy that combined adsorption and microbial degradation (nitrification 84

and denitrification). ORC was filled in the oxygen releasing column to release oxygen 85

for the following aerobic nitrification. Biological zeolite was used to ensure a robust 86

ammonium removal efficiency by ion exchange and nitrification process in the 87

aerobic reaction column. To realize the ammonium removal and avoid the second 88

pollution of nitrate that produced by the former nitrification, the oxygen removal 89

column was filled with spongy iron to deplete excess oxygen and create an anaerobic 90

environment for denitrification. Pine bark was filled in the anaerobic reaction column, 91

which could dissolve slowly in water to supply organic carbon for the growth of 92

heterotrophic denitrification bacteria. 93

94

The objectives of this study were to evaluate the performance of designed M-PRB, 95

investigate the regulating effects of DO and pH by ORC and spongy iron, analyse the 96

competing adsorption on zeolite by co-existed cations, and explore the ammonium 97

and nitrate removal mechanisms in different columns. 98

99

2. Materials and Method 100

2.1 Materials 101

Accep

ted

Man

uscr

ipt

Dow

nloa

ded

by [

Mon

ash

Uni

vers

ity L

ibra

ry]

at 1

0:15

04

Dec

embe

r 20

14


5

The zeolite (0.83-1.70 mm) used was supplied by Zhejiang Shenshi Mining Co. Ltd. 102

Its chemical compositions were as follows (in mass): SiO2 69.58%, Al2O3 12.20%, 103

Fe2O3 0.87%, K2O 1.13%, MgO 0.13%, CaO 2.59%, Na2O 2.59%, and others 10.91%. 104

The cation exchangeable capacity (CEC) was 130-180 meq/g. ORC was made to 105

small bead (1.0-2.0 cm) by blending cement (27.0 wt.%), calcium peroxide (21.0 106

wt.%), sand (16.0 wt.%), bentonite (10.0 wt.%) and water (26.0 wt.%) together as 107

reported in our previous study [21], and it could effectively release oxygen in water 108

and avoid the occurrence of extremely high pH. Spongy iron (0.83-1.70 mm) was 109

obtained by Kai bi yuan Co. Ltd, which was typically composed of Fe (60.6 wt.%). 110

Pine bark (0.83-1.70 mm) was purchased from a nursery store in Beijing. The nature 111

sand (0.83-1.70 mm) was from a river channel. Unless otherwise indicated, all 112

chemicals used were analytical reagent grade as received. 113

114

The simulated groundwater was prepared by adding pre-calculated NH4Cl into the 115

groundwater which had a concentration of ammonium (10±2 mg N/L), nitrite (＜0.1 116

mg N/L), and nitrate (4±1 mg N/L). The groundwater used was from the well on the 117

campus of China university of Geosciences (Beijing). The soil used for culturing 118

biological inoculums was obtained from the garden of the school (0.1-0.3 m depth 119

from surface), which was passed through 80-100 mesh sieves to remove the plant 120

roots and other impurities. 121

122

2.2 Enrichment culture to prepare bacterial inoculums 123

The enrichment culture of bacterial inoculums was conducted using a 5 L brown glass 124

bottle to which was added 300 g soil and 2.5 L groundwater. The culture procedure of 125

bacterial inoculums was as follows: (1) Add 74.31 mg NH4Cl to the brown glass 126

Accep

ted

Man

uscr

ipt

Dow

nloa

ded

by [

Mon

ash

Uni

vers

ity L

ibra

ry]

at 1

0:15

04

Dec

embe

r 20

14

6

bottle, making the ammonium concentration of 10 mg N/L. Intermittent aeration 127

device (aeration for 22 h every 24 h) was used to ensure the DO of 8-10 mg/L. (2) 128

Detect the ammonium, nitrite and nitrate concentrations of the solution at an interval 129

of 24 h. Upon depletion of ammonium (<0.5 mg/L), NH4Cl was added again to ensure 130

the ammonium concentration of about 10 mg N/L. (3) Repeat this culture method 131

until the inoculums reached to a steady degradation half-life (1 day) of 10 mg N/L 132

ammonium. The enrichment culture was conducted at room temperature (20±5 ℃). 133

134

2.3 Columns setup 135

An overview of the columns setup was given in Figure1. Four Polymethyl 136

methacrylate columns (height, 50 cm; inside diameter, 3 cm) were set up to simulate 137

the M-PRB for ammonium removal. The outlet of each column was connected to a 138

capped 15 mL vial, and the column effluent was pumped from this vial into another 139

column. The first oxygen releasing column was packed with optimized ORC. Zeolite 140

was filled in the second aerobic reaction column. The third oxygen removal column 141

was filled with spongy iron. Pine bark was filled in the last anaerobic reaction column. 142

ORC and Spongy iron were all mixed packing with granular sand to ensure the 143

reactivity and permeability of the M-PRB. The parameters of each column were 144

listed in Table 1. 145

146

The simulated groundwater was kept in a 10 L hermetic bottle connected with a 147

nitrogen gas bag to maintain the concentration stability, and it was pumped in an 148

upward flow through the columns with constant flow rate. The whole column system 149

was operated in the dark environment at room temperature (20±5 ℃). 150

151

Accep

ted

Man

uscr

ipt

Dow

nloa

ded

by [

Mon

ash

Uni

vers

ity L

ibra

ry]

at 1

0:15

04

Dec

embe

r 20

14


7

2.4 Operation of the columns 152

To visually observe the process of ammonium transformation by nitrifying bacteria 153

grown on zeolite in the aerobic reaction column, the bacterial inoculums was added 154

into the aerobic reaction column to promote the formation of biofilm on zeolite. The 155

culture procedures of biological zeolite were as follows: (1) NH4Cl was added to give 156

the bacterial inoculums of 10 mg N/L ammonium, and then the inoculums was 157

introduced into the aerobic reaction column immediately to make the nature zeolite 158

saturated; (2) Discharge the solution of the column when the ammonium was 159

completely transformed to nitrate in the effluent, and add new bacterial inoculums to 160

soak the zeolite again. Repeat this step for a total of four times; (3) After the culture 161

period, all the columns were connected together to start the experiment. To enable 162

bacteria to accumulate and form biofilm on the zeolite, the flow rate of the column 163

was set for 105 mL/d within the first 33 days of operation, then the flow rate was 164

increased to 263 mL/d to investigate the removal performance of M-PRB columns 165

under high groundwater velocity. Correspondingly, the hydraulic retention time in 166

each column was about 2.5 days (0-33 days) and 1 day (33-100 days) respectively. 167

168

2.5 Analytical methods 169

Water samples were taken from the inlets and outlets of the caped vials at appropriate 170

intervals and analyzed immediately. Ammonium, nitrite and nitrate were analyzed by 171

a HP-8453 UV–Vis spectrophotometer at λmax 420 nm, 540 nm, 220 nm and 275 nm 172

respectively. Other cations (K+, Na+, Ca2+, Mg2+) were determined by ion 173

chromatography (Dionex-120, USA). DO was determined by a portable dissolved 174

oxygen analyzer (Hach HQ30d, USA), and pH was measured by a pH meter 175

(Sartorius PB-10, Germany). The instruments were calibrated before measuring the 176

Accep

ted

Man

uscr

ipt

Dow

nloa

ded

by [

Mon

ash

Uni

vers

ity L

ibra

ry]

at 1

0:15

04

Dec

embe

r 20

14

8

samples. 177

178

3. Results and discussion 179

3.1 The variation of DO and pH in different columns 180

The changes of DO and pH in M-PRB columns are presented in Figure 2A and 2B. 181

During the complete operation period, after the solution flew through the oxygen 182

releasing column, the effluent DO concentrations reached from 3 mg/L to above 20 183

mg/L, and the pH only had an increase in the first 20 days and then decreased to 184

8±0.5. When the high DO solution flew through the oxygen removal column, the DO 185

concentrations decreased from 7±1 mg/L to below 4 mg/L. The pH increased 186

drastically and the maximum value reached to 9.36 in the oxygen removal column, 187

but it slowly decreased to below 8.3 after 54 days. 188

189

The measured DO concentrations in the aerobic reaction column reached to above 20 190

mg/L, because the solubility of DO in water could increase or even exceed its 191

saturated solubility with the increase of pressure in one-dimensional column system 192

[22]. The high proportion of cement and bentonite in ORC had a good package effect 193

on CaO2, which could prevent high pH of water caused by the fast reaction of CaO2. 194

In addition, the nature groundwater also had a good pH buffering capacity, especially 195

its HCO3-/CO3

2- buffering system. After the solution flew through the oxygen 196

releasing column, the HCO3- concentration decreased from 148.6 mg/L to 69.0 mg/L, 197

which means about 1.32 mmol/L OH- could be neutralized. Additionally, the 198

microbial nitrification could also decrease the pH of the system as it is an acid 199

production process. The effluent pH value of the aerobic reaction column was below 200

7.8, which demonstrated that the slight pH increase caused by ORC did not have a 201

Accep

ted

Man

uscr

ipt

Dow

nloa

ded

by [

Mon

ash

Uni

vers

ity L

ibra

ry]

at 1

0:15

04

Dec

embe

r 20

14


9

significant effect on the following nitrification in the aerobic reaction column. 202

203

Spongy iron was used as an oxygen scavenger in the oxygen removal column, which 204

had a more porous internal structure and larger specific surface area compared with 205

other iron types [23]. DO concentrations were averagely decreased to 3 mg/L in the 206

oxygen removal column, and therefore the following anaerobic reaction column was 207

almost in a suboxic to anoxic environment. Furthermore, the pH value of the 208

anaerobic reaction column was not very high and decreased from 9.0 to below 8.3 209

with the iron oxidation, which was favorable for the microbial denitrification [24]. It 210

was supposed that the mass proportion (1:7) of spongy iron and sand filled in the 211

oxygen removal column was helpful to reduce the pH arise caused by spongy iron. 212

213

3.2 Ammonium removal by adsorption and biological nitrification 214

An overview of the ammonium and nitrate removal in different columns is shown in 215

Figure 3A and 3B. During the complete operation period, the influent groundwater 216

had an ammonium concentration ranging from 9.5 mg N/L to 13 mg N/L. After the 217

groundwater flew through the M-PRB columns, the ammonium could be efficiently 218

removed and its concentration was decreased to below 0.5 mg N/L. 219

220

Within the first 33 days of operation, ammonium was completely removed but there 221

was little nitrate formed in the aerobic reaction column, indicating that the main 222

removal mechanism was the adsorption of ammonium onto the biological zeolite. The 223

nitrification was weak in this period, since the nitrifying bacteria were adapting 224

themselves to the environment. Thereafter, the nitrate concentration gradually 225

increased from 5 mg N/L to above 7.5 mg N/L in the aerobic reaction column and 226

Accep

ted

Man

uscr

ipt

Dow

nloa

ded

by [

Mon

ash

Uni

vers

ity L

ibra

ry]

at 1

0:15

04

Dec

embe

r 20

14

10

above 25% of the ammonium was oxidized to nitrate, demonstrating that the 227

nitrifying bacteria began to refresh their activities and appeared accumulation in the 228

aerobic reaction column. After 45 days of operation, the nitrate in the aerobic reaction 229

column effluent increased quickly, and its concentration kept at a level above 13 mg 230

N/L from day 60 to 100, suggesting that the biological zeolite in the column had 231

realized the accumulation of microbial community. The increased flow rate had an 232

effect on the nitrification using the biological zeolite under the experimental 233

conditions. At the low flow rate (105 mL/d), ammonium was adsorbed quickly by the 234

biological zeolite (the maximum adsorption capacity 9.68 mg/g) filled in the upstream 235

of the column, and thus little nitrification occurred in the downstream of the column 236

due to a lack of ammonium in solution. After the flow rate was increased to 263 mL/d, 237

the ammonium adsorbed by biological zeolite in the upstream decreased and the 238

soluble ammonium migrated farther along the column. This was favorable for the 239

biological zeolite in the downstream contacting with the soluble ammonium, causing 240

more production of nitrate due to the sufficient substrate in the aerobic reaction 241

column. 242

The concentration of ammonium slightly increased in the oxygen removal column 243

owing to the reaction of spongy iron with nitrate. Iron has a high reduction potential 244

and can react with nitrate to realize partial removal of nitrate by chemical reduction, 245

which has been confirmed by most of the reviewed works [25-26]. Rodríguez-Maroto 246

et al. found ammonium was the main end product of nitrate reduction by zero-valent 247

iron in experiment [27]. The mechanism is expressed in equation (1). 248

4Fe+NO3-+7H2O=4Fe2++NH4

++10OH- (1) 249

The remaining ammonium in the oxygen removal column (＜3 mg/L) could be further 250

removed in the anaerobic reaction column as demonstrated by the column effluent 251

Accep

ted

Man

uscr

ipt

Dow

nloa

ded

by [

Mon

ash

Uni

vers

ity L

ibra

ry]

at 1

0:15

04

Dec

embe

r 20

14


11

concentration of below 0.5 mg N/L. It was analyzed that ammonium could be 252

adsorbed by the nature river sand packed in the column which contained a variety of 253

negatively charged minerals [28]. In addition, ammonium was partly removed as a 254

nitrogen source for microbial metabolism [29]. 255

256

3.3 Nitrate removal by chemical reduction and biological denitrification 257

The concentration change of nitrate in different columns is shown in Figure 3B. The 258

nitrate concentration in the aerobic reaction column increased from 4.5 mg N/L to 259

above 13 mg N/L after 54 days due to the existing nitrification process, which was 260

higher than the Chinese drinking water standard of 10 mg N/L [30]. But after the 261

solution flew through the anaerobic reaction column, the nitrate concentration was 262

decreased to below 5 mg N/L, suggesting that more than 90% nitrate formed by the 263

transformation of ammonium in the nitrification compartment could be removed. 264

265

Within the first 45 days of operation, the concentrations of nitrate reduced and 266

ammonium formed were nearly the same in the oxygen removal column (Figure3A, 267

3B). It was supposed that the removal of nitrate completely relied on chemical 268

reduction by spongy iron, which was explained in section 3.2. There was little 269

decrease in nitrate concentration in the anaerobic reaction column, because the 270

indigenous denitrifying bacteria needed to adapt themselves to the environment. With 271

the experimental operation, slow decomposition of pine bark in water could provide 272

organic carbon for the growth of heterotrophic denitrifying bacteria in the anaerobic 273

reaction column. After 54 days, the denitrification rate gradually increased and the 274

nitrate was decreased from 9 mg N/L to 5 mg N/L in the anaerobic reaction column. It 275

was worth attention that the nitrate decreased is much higher than ammonium 276

Accep

ted

Man

uscr

ipt

Dow

nloa

ded

by [

Mon

ash

Uni

vers

ity L

ibra

ry]

at 1

0:15

04

Dec

embe

r 20

14

12

increased in the effluent of the oxygen removal column, indicating that the 277

denitrification also occurred in the oxygen removal column. These phenomena 278

suggested that the indigenous denitrifying bacteria growing on the river sand had 279

refreshed their activities. During the entire operation, above 25% nitrate could be 280

removed by chemical reduction due to the presence of spongy iron in the oxygen 281

removal column, and the remaining nitrate was further removed by denitrificaiton. 282

The nitrate removal efficiency of denitrificaiton wasn’t high in the experiment in that 283

no denitrifying bacterial suspension was inoculated in the anaerobic reaction column. 284

285

3.4 Contributions of ion exchange and nitrification to ammonium removal 286

The removal of ammonium in the aerobic reaction column is presented in Figure 4. 287

During the first 45 days of operation, the ammonium was mainly removed by ion 288

exchange onto the zeolite. After 54 days of operation, with the adaption of nitrifying 289

bacteria, the nitrification process became the primary mechanism to remove 290

ammonium. Within this period, above 70% ammonium was removed by nitrification. 291

Zeolite utilized had a high ammonium sorption capacity (9.68 mg/g), therefore it 292

could ensure a robust ammonium removal efficiency in the aerobic reaction column. 293

Furthermore, the biofilm covered on zeolite did not affect ion exchange [18]. Thus, 294

with the sustainable supply of oxygen by ORC, zeolite was able to be biologically 295

regenerated due to further removal of ammonium by microbial nitrification. Thereby, 296

it is beneficial for ammonium removal efficiency keeping high and sustainable in the 297

aerobic reaction column. 298

299

3.5 Selective exchange of metal cations by zeolite 300

Zeolite consists of alluminosilicate frameworks with exchangeable cations (typically 301

Accep

ted

Man

uscr

ipt

Dow

nloa

ded

by [

Mon

ash

Uni

vers

ity L

ibra

ry]

at 1

0:15

04

Dec

embe

r 20

14


13

K+, Na+, Ca2+, Mg2+), which indicates that these co-existed cations have strong 302

competition effects for exchange sites on the zeolite [9]. 303

304

As shown in Figure 5, the concentrations of Mg2+ and K+ decreased clearly in effluent 305

of the aerobic reaction column, no substantial Ca2+ increase or decrease was observed, 306

and Na+ increased markedly during the operation period. Zeolite showed a high 307

preference for NH4+, Mg2+ and K+ over Na+ and Ca2+, which suggested that the NH4

+, 308

Mg2+ and K+ were primarily exchanged with Na+ adsorbed in the zeolite crystal 309

structure. Compared to other cations, Na+ has a smaller ionic radius and a higher 310

chemical activity, indicating that it is easier to overcome the internal diffusion 311

resistance in zeolite and be exchanged with other cations in solution. NH4+ and K+ are 312

adsorbed quickly because zeolite has a high adsorption affinity with them [31]. Unlike 313

the previous results [10,32], zeolite showed a high adsorption affinity to Mg2+. Taking 314

into account that the pH in the aerobic reaction column was less than 9 after 20 days 315

(Figure 2B), there was no Mg(OH)2 formed in the column (Mg(OH)2, Ksp=5.1×10-12, 316

25℃). In addition, the driving force from the influent concentration of cations would 317

promote the ion exchange reaction between the solid-liquid phases [33]. Therefore, 318

Mg2+ was primarily adsorbed by zeolite. 319

320

During the operation period, the total concentration of cations in the effluent of the 321

aerobic reaction column had an obvious decrease (>0.7 mmol/L), indicating that the 322

nitrification occurred in the aerobic reaction column, and partial ammonium was 323

removed by nitrification instead of ion exchange. 324

325

3.6 Performance evaluation of M-PRB columns 326

Accep

ted

Man

uscr

ipt

Dow

nloa

ded

by [

Mon

ash

Uni

vers

ity L

ibra

ry]

at 1

0:15

04

Dec

embe

r 20

14

14

As shown in Figure 6, at the operating condition of the influent ammonium 327

concentration of about 10 mg N/L, the sequential M-PRB columns had the capacity of 328

reducing ammonium and its nitrification products (nitrite & nitrate) to levels below 329

the regulatory discharge limits, and a total nitrogen content could be decreased from 330

15 mg N/L to below 6 mg N/L. Due to the sufficient supply of oxygen by ORC, the 331

ammonium could be further removed by microbial nitrification in the aerobic reaction 332

column, ensuring that the column had a robust ammonium removal efficiency. In 333

addition, the nitrate formed in the first nitrification compartment could be reduced by 334

the spongy iron-based chemical reduction and biological denitrification in the oxygen 335

removal column and anaerobic reaction column, avoiding the second pollution of 336

nitrate. 337

338

Although the proposed M-PRB columns showed a good removal performance in the 339

laboratory-scale experiment, further studies are necessarily required to evaluate the 340

performance of its field application. Firstly, the real groundwater temperature of 341

12-17℃ in most sites is slightly lower than the room temperature, so the performance 342

of the system at a lower temperature need to be investigated. Secondly, the materials 343

utilized in the field-scale M-PRB could be further optimized by adjusting operating 344

parameters such as the material filling proportions. For example, ORC filled in the 345

oxygen releasing column was mixed with sand to avoid surplus gas accumulation 346

affecting the column permeability. But in field-scale, the released oxygen is easily 347

diffused and depleted in aquifer, and the proportion of ORC in M-PRB need increase 348

to ensure a sufficient oxygen supply. Spongy iron was an excellent oxygen 349

consumption material, but an excess dosage in M-PRB could result in a drastic 350

increase in pH and form ammonium by nitrate reduction. Although the spongy iron 351

Accep

ted

Man

uscr

ipt

Dow

nloa

ded

by [

Mon

ash

Uni

vers

ity L

ibra

ry]

at 1

0:15

04

Dec

embe

r 20

14


15

was mixed with sand at a mass ratio of 1: 7 in this study, a small amount of 352

ammonium formation (less than 3 mg N/L) was still observed. Fortunately, the 353

ammonium produced by the reduction of spongy iron was kept on acceptable level 354

and could be further removed in the anaerobic reaction column. In addition, it is easy 355

to create an anaerobic reaction environment by increasing the distance between the 356

aerobic and anaerobic reaction compartments of M-PRB system. Then, it would be 357

helpful to reduce the dosage of spongy iron utilized in the oxygen removal 358

compartment. 359

360

4. Conclusion 361

ORC could supply sufficient dissolved oxygen for microbial nitrification in anaerobic 362

groundwater. The combination use of ion exchange and microbial nitrification by 363

biological zeolite could ensure the ammonium removal efficiency (> 98%) in the 364

aerobic reaction column. A nitrate concentration of 9 mg N/L produced by microbial 365

nitrification in the aerobic reaction column was removed through the spongy 366

iron-based chemical reduction and microbial denitrification in the following oxygen 367

removal column and anaerobic reaction column, avoiding the nitrate pollution. 368

369

The driving force from influent concentration of cations can promote the ion 370

exchange between the zeolite-liquid phases. Na+ is easier to overcome the internal 371

diffusion resistance in zeolite and be exchanged with other cations in solution. Mg2+ 372

and K+ would compete with ammonium for exchange sites on zeolite. Ca2+ has no 373

substantial influence for ammonium exchange. 374

375

The proposed M-PRB columns were useful for removing ammonium and its 376

Accep

ted

Man

uscr

ipt

Dow

nloa

ded

by [

Mon

ash

Uni

vers

ity L

ibra

ry]

at 1

0:15

04

Dec

embe

r 20

14

16

nitrification products (nitrite & nitrate) to levels below the regulatory discharge limits. 377

Results from this study will be useful in designing a permeable reactive barrier system 378

for field remediation of ammonium contaminated groundwater. 379

380

Acknowledgements 381

This work was supported by the <National Program of Control and Treatment of 382

Water Pollution> under Grant<2009ZX07424-002>; and <the Fundamental Research 383

Funds of the Institute of Hydrogeology and Environmental Geology> under 384

Grant<SK201304>. 385

386

References 387

[1] Joanna F, Kazimierz G. Evaluation of zeolite-sand mixtures as reactive materials protecting 388

groundwater at waste disposal sites. J Environ Sci. 2013; 25: 1764-1772. 389

[2] Šiljeg M, Foglar L, Kukučka M. The ground water ammonium sorption onto Croatian and 390

Serbian clinoptilolite. J Hazard Mater. 2010; 178: 572-577. 391

[3] Ning Z, Zhang CY, Zhang S. Study of groundwater contamination of ammonium and its 392

nitrogen isotope. South to north water transfers and water sci & technol. 2011; 9:129-132. 393

(in chinese) 394

[4] Ma J, Shi Y, Chen X, Wang HX. Nitrogen load from rural non-point source pollution in suburb 395

area of Shenyang: a case study of Damintun town. Advanced Mater Res. 2013; 610: 396

3277–3281. 397

[5] Wang YQ, Liu SJ, Xu Z, Han TW, Chuan S, Zhu T. Ammonia removal from leachate solution 398

using natural Chinese clinoptilolite. J Hazard Mater. 2006; 136: 735-740. 399

[6] Phillips DH, Nooten TV, Bastiaens L, Russell MI, Dickson K, Plant S, Ahad JME, Newton T, 400

Elliot T, Kalin RM. Ten Year Performance Evaluation of a Field-Scale Zero-Valent Iron 401

Permeable Reactive Barrier Installed to Remediate Trichloroethene Contaminated 402

Groundwater. Environ Sci & Technol. 2010; 44: 3861-3869. 403

[7] Vignola R, Bagatin R, De Folly D’Auris A, Massara EP, Ghisletti D, Millini R, Sisto R. 404

Zeolites in a permeable reactive barrier (PRB): One year of field experience in a refinery 405

groundwater—Part 1: The performances. Chem Eng J. 2011; 178: 204-209. 406

[8] Andrew DH, Avery HD. Long-Term Performance of Zero-Valent Iron Permeable Reactive 407

Barriers: A Critical Review. Environ Eng Sci. 2007; 24: 401-423. 408

[9] Huang HM, Xiao XM, Yan B, Yang LP. Ammonium removal from aqueous solutions by using 409

natural Chinese (Chende) zeolite as adsorbent. J Hazard Mater. 2010; 175: 247-252. 410

[10] Lin L, Lei ZF, Wang L, Liu X, Zhang Y, Wan CL, Lee DJ, Tay JH. Adsorption mechanisms of 411

high-levels of ammonium onto natural and NaCl-modified zeolites. Separation and 412

Accep

ted

Man

uscr

ipt

Dow

nloa

ded

by [

Mon

ash

Uni

vers

ity L

ibra

ry]

at 1

0:15

04

Dec

embe

r 20

14


17

Purification Technology. Sep Purif Technol. 2013; 103: 15-20. 413

[11] Yang K, Zhang X, Chao C, Zhang B, Liu J. In-situ preparation of NaA zeolite/chitosan porous 414

hybrid beads for removal of ammonium from aqueous solution. Carbohyd Polym. 2014; 415

107: 103-109. 416

[12] Chavez-Crooker P, Obreque-Contreras J. Bioremediation of aquaculture wastes. Curr. Opin. 417

Biotechnol. 2010; 21: 313-317. 418

[13] Zhang Y, Bi EP. Effect of dissolved organic matter on ammonium sorption kinetics and 419

equilibrium to Chinese clinoptilolite. Environ Technol. 2012; 33: 2395-2407. 420

[14] Buss SR., Herbert AW, Morgan P, Thornton SF, Smith JWN. A review of ammonium 421

attenuation in soil and groundwater. Q J Eng Geol Hydrogeol. 2004; 37: 347-359. 422

[15] Yeh CH, Lin CW, Wu CH. A permeable reactive barrier for the bioremediation of 423

BTEX-contaminated groundwater: Microbial community distribution and removal 424

efficiencies. J Hazard Mater. 2010; 178: 74-80. 425

[16] Van NT, Diels L, Bastiaens L. Design of a Multifunctional Permeable Reactive Barrier for the 426

Treatment of Landfill Leachate Contamination: Laboratory Column Evaluation. Environ 427

Sci Technol. 2008; 42: 8890-8895. 428

[17] Hedstrom A. Ion exchange of ammonium in zeolites: A literature review, J. Environ. 429

Eng-Asce. 2001; 127: 673-681. 430

[18] Krishnani KK, Zhang Y, Xiong L, Yan YS, Boopathy R, Mulchandani A. Bactericidal and 431

ammonia removal activity of silver ion-exchanged zeolite. Bioresource Technol. 2012; 432

117: 86-91. 433

[19] Van NT, Diels L, Bastiaens L. Microbially mediated clinoptilolite regeneration in a 434

multifunctional permeable reactive barrier used to remove ammonium from landfill 435

leachate contamination: Laboratory column evaluation. Environ Sci Technol. 2010; 44: 436

3486-3492. 437

[20] Wen DH, Ho Y-S, Tang XY. Comparative sorption kinetic studies of ammonium onto zeolite. 438

J Hazard Mater. 2006; 133: 252-256. 439

[21] Kong XK, Zhang Y, Bi EP. Optimization of oxygen-releasing materials and pH regulation in 440

groundwater remediation system. Chinese J Environ Eng, 2012; 6: 2935-2940. (in 441

chinese) 442

[22] Liu SJ, Jiang B, Huang GQ, Li XG. Laboratory column study for remediation of 443

MTBE-contaminated groundwater using a biological two-layer permeable barrier. Water 444

Res. 2006; 40: 3401-3408. 445

[23] Huang GX, Fallowfield H, Guan HD, Liu F. Remediation of Nitrate-Nitrogen Contaminated 446

Groundwater by a Heterotrophic-Autotrophic Denitrification Approach in an Aerobic 447

Environment. Water Air Soil Pollut. 2012; 223:4029–4038. 448

[24] Rivett MO, Buss SR, Morgan P, Smith JWN, Bemment CD. Nitrate attenuation in 449

groundwater: A review ofbiogeochemical controlling processes. Water Res. 2008; 42: 450

4215-4232. 451

[25] Westerhoff P, James J. Nitrate removal in zero-valent iron packed columns. Water Res, 2003; 452

37: 1818-1830. 453

[26] Mishra D, Farrell J. Understanding nitrate reactions with zerovalent iron using tafel analysis 454

and electrochemical impedance spectroscopy. Environ Sci & Technol. 2005; 39: 645–650. 455

[27] Rodríguez-Maroto JM, García-Herruzo F, García-Rubio A, Gómez-Lahoz C, Vereda-Alonso 456

Accep

ted

Man

uscr

ipt

Dow

nloa

ded

by [

Mon

ash

Uni

vers

ity L

ibra

ry]

at 1

0:15

04

Dec

embe

r 20

14

18

C. Kinetics of the chemical reduction of nitrate by zero-valent iron. Chemosphere. 2009; 457

74: 804-809. 458

[28] Jellali S, Diamantopoulos E, Kallali H, Bennaceur S, Anane M, Jedidi N. Dynamic sorption 459

of ammonium by sandy soil in fixed bed columns: Evaluation of equilibrium and 460

non-equilibrium transport processes. J Environ Manage. 2010; 91:897-905. 461

[29] Barber WP, Stuckey DC. Nitrogen removal in a modified anaerobic baffled reactor (ABR): 1, 462

Denitrification. Water Res. 2000; 34: 2413-2422. 463

[30] GB5749-2006. Standards for drinking water quality. Beijing:Ministry of Environmental 464

Protection of the People’s Republic of China; 2006. 465

[31] Guo XJ, Zeng L, Li XM, Park H-S. Ammonium and potassium removal for anaerobically 466

digested wastewater using natural clinoptilolite followed by membrane pretreatment. J 467

Hazard Mater. 2008; 151: 125-133. 468

[32] Ames LL. The cation sieve properties of clinoptilolite. Amer Mineral. 1960; 45: 689-700. 469

[33] Wang Y, Kmiya Y, Okuhara T. Removal of low concentration ammonia in water by 470

ion-exchange using Namordenite. Water Research. 2007; 41: 269-276. 471

472

Table 1. The parameters of each column 473

Column mark Packing Mass of dry material (g)

porosity

Oxygen releasing column ORC/sand

mixed ORC(475) sand(1100)

0.32

Aerobic reaction column zeolite 1075 0.51

Oxygen removal column spongy iron/sand

mixed spongy iron(172)

sand(1376) 0.37

Anaerobic reaction columnpine bark/sand

layered pine bark(70) sand(1343)

0.43

474

475

476

Accep

ted

Man

uscr

ipt

Dow

nloa

ded

by [

Mon

ash

Uni

vers

ity L

ibra

ry]

at 1

0:15

04

Dec

embe

r 20

14


19

477

Accep

ted

Man

uscr

ipt

Dow

nloa

ded

by [

Mon

ash

Uni

vers

ity L

ibra

ry]

at 1

0:15

04

Dec

embe

r 20

14