Microbial photoelectrotrophic denitrification as a sustainable and

Subscriber access provided by LAURENTIAN UNIV

Article

Microbial photoelectrotrophic denitrification as a sustainable and efficient way for reducing nitrate to nitrogen Hao-Yi Cheng, Xia-Di Tian, Chuanhao Li, Shusen Wang, Shi-Gang Su, HongCheng Wang, Bo Zhang, Hafiz Muhammad Adeel Sharif, and Aijie Wang Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.7b02557 • Publication Date (Web): 13 Oct 2017 Downloaded from http://pubs.acs.org on October 13, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 31

Environmental Science & Technology

1

Microbial photoelectrotrophic denitrification as a sustainable

2

and efficient way for reducing nitrate to nitrogen

3 4

Hao-Yi Cheng,a Xia-Di Tian,ab , Chuan-Hao Li,c Shu-Sen Wang, ab Shi-Gang Su, ab

5

Hong-Cheng Wang,ab Bo Zhang,a Hafiz Muhammad Adeel Sharif,ab Ai-Jie Wang,*ab

6

a

7

Eco-Environmental Sciences, Chinese Academy of Sciences, No. 18 Shuangqing Road,

8

Haidian District, Beijing 100085, China

9

b

Key

Laboratory

of

Environmental

Biotechnology,

Research

Center

for

University of Chinese Academy of Sciences, No. 19 Yuquan Road, Shijingshan

10

District, Beijing 100049, China

11

c

12

Campus, No. 135 Waihuan Road, Daxuecheng District, Guangzhou 510006, China

13

* Corresponding Author: Tel & Fax: +86-10-62915515; Email: [email protected].

School of Environmental Science and Engineering, Sun Yat-Sen University, East

14 15 16 17 18 19 20 21 22 23 1

ACS Paragon Plus Environment


24

Abstract：：

25

Biological removal of nitrate, a highly-concerning contaminant, is limited when the

26

aqueous environment lacks bioavailable electron donors. In this study, we

27

demonstrated, for the first time, that bacteria can directly use the electrons originated

28

from photoelectrochemical process to carry out the denitrification. In such

29

photoelectrotrophic denitrification (PEDeN) systems (denitrification biocathode

30

coupling with TiO2 photoanode), nitrogen removal was verified solely relying on the

31

illumination dosing without consuming additional chemical reductant or electric

32

power. Under the UV illumination (30 mW·cm-2, wavelength at 380±20 nm), nitrate

33

reduction in PEDeN apparently followed the first order kinetics with the constant of

34

0.13±0.023 h-1. Nitrate was found almost completely converted to nitrogen gas at the

35

end of batch test. Compared to the electrotrophic denitrification systems that driven

36

by organics (OEDeN, biocathode /acetate consuming bioanode) or electricity (EEDeN,

37

biocathode/abiotic anode), denitrification rate in PEDeN equaled to that in OEDeN

38

with COD/N ratio at 9.0 or that in EEDeN with applied voltage at 2.0 V. This study

39

provides a sustainable technical approach for eliminating nitrate from water. PEDeN

40

as a novel microbial metabolism may shed further light onto the role of sunlight

41

played in the nitrogen cycling in certain semi-conductive and conductive minerals

42

enriched aqueous environment.

43 44

Key words: photoelectrotrophic; photoelectrochemical; denitrification; biocathode;

45

selective. 2


Page 2 of 31

Page 3 of 31


46

Introduction

47

Nitrate is one of the most concerning contaminants due to its ubiquity and adverse

48

effects on human health and ecosystem.1 Autotrophic denitrification, the biological

49

process of transforming nitrate to the innocuous nitrogen gas with inorganic electron

50

donors, has received great attention in past decades.

51

denitrification does not rely on the presence of organic matter, it is particularly

52

attractive during the treatment of drinking water and the secondary effluent of

53

wastewater treatment plants, in which the organic form of electron donors is typically

54

lacking.4, 5 Compared to the heterotrophic denitrification by adding organic electron

55

donors, autotrophic denitrification is suggested to be more cost-effective and it avoids

56

the risk of secondary organic pollution.2

2, 3

Because the autotrophic

57

In autotrophic denitrification, hydrogen is most frequently used electron donor due

58

to its innocuous nature.6 However, the direct use of hydrogen gas may lead to safety

59

issues and suffer from a low H2 utilization efficiency due to its low water solubility.

60

Therefore, denitrification using electrodes, which generates hydrogen through

61

hydrogen evolution reactions in-situ, was developed as an alternative approach.2, 7

62

After that, bacteria was found to directly receive electrons from electrode for

63

denitrification without hydrogen as an electron mediator (here defined as the

64

electrotrophic denitrification, EDeN).8 In this process, EDeN gains the benefit of less

65

energy loss because of its occurrence at more positive potential compared to the

66

hydrogen evolution.8 In the case of the denitrification driven by electrode (as the

67

cathode), another electrode has to be coupled as the anode. Oxygen evolution is the 3



68

main anodic reaction in most cases with the theoretical occurrence potential at 0.82 V

69

(versus standard hydrogen electrode, SHE, at pH 7). Because this potential is more

70

positive than that of nitrate reduction (i.e. ENO3-/NO2- is 0.43 V vs SHE at pH 7),

71

electric energy input is required. Alternatively, microbial catalytic anode system can

72

be employed without the power supply, where biodegradable organics are oxidized by

73

anode respiration bacteria at a more negative potential (e.g. the typical Eacetate/HCO3- is

74

-0.28 V vs SHE at pH 7) compared to the nitrate reduction.9, 10

75

Solar energy is considered to be a clean and sustainable energy source. Conversion

76

of solar energy to chemical energy is realized with the help of a semiconductor, which

77

respectively provides the reducing and oxidizing power through the photoexcited

78

electrons and holes under illumination.11 To date, TiO2 is the most intensively studied

79

semiconductor since it is inexpensive and easily synthesized.12 Photocatalytic nitrate

80

reduction has been reported in TiO2 systems,13, 14 but the selectivity of N2 formation is

81

low and the over reduction of nitrate to ammonium ions is particularly undesired.

82

Although the

83

TiO2/Ag, TiO2/Ni-Cu, etc.,13, 15 the cost of the catalysts would inevitably increase,

84

while the formation of ammonium is still not negligible.15 On the contrary, microbial

85

denitrification usually has high selectivity.

86

formation was reported to be at very low level or even not detected.8, 16 Considering

87

that the conduction band of TiO2 (-0.5 V vs SHE at pH 7) lied at an energy level that

88

is noticeably higher than that of nitrate reduction, the photo-excited electrons should

89

have energy high enough to drive EDeN without additional energy input, such as

selectivity is improved by doping other metals, such as TiO2/Pt,

Regarding the EDeN, the ammonium

4


Page 4 of 31

Page 5 of 31


90

power supply as in the previous study.17 Therefore, we hypothesize that combining the

91

TiO2 semiconductor system with the EDeN will gain a high selectivity for reducing

92

nitrate to N2 and the occurrence of this process can solely rely on the illuminating

93

dose. In such a novel photoelectrotrophic denitrification (PEDeN) process, the

94

generation of photoelectrons is independent of the presence of organics, which brings

95

about the PEDeN particularly promising in the denitrification under organics limited

96

conditions. To the best of our knowledge, the proposed PEDeN here has not been

97

verified yet.

98

Herein, the PEDeN was developed by constructing a photoelectrochemical cell,

99

where a denitrification biocathode was connected to a TiO2 photoanode without

100

voltage supply (Figure 1a). The objectives of this study are i) to demonstrate the

101

PEDeN is capable of efficiently reducing nitrate to nitrogen without additional

102

reductant and energy input; ii) to understand the performance of PEDeN by

103

comparing with the already established electricity driven EDeN system (EEDeN, also

104

known as electrochemical bioreactor of denitrification, Figure 1b) and organics driven

105

EDeN system (OEDeN, also known as denitrification microbial fuel cell, Figure 1c).

106 107

Materials and Methods

108

Reactor construction. The reactor was designed as three chambers which were

109

separated by cation exchange membranes (Ultrex CMI-7000,

110

International, U.S.). The cathode chamber was arranged in the middle, while the two

111

side chambers were set as photoanode chamber and bioanode chamber, respectively 5


Membranes


112

(Figure 1). The side plate of the photoanode chamber was equipped with a quartz

113

glass window (6 cm in diameter) that allowed the light passage. The TiO2/Ti plate

114

(surface area =16 cm2), graphite granules (3~4 mm in diameter) and carbon brush

115

were employed as the electrode materials in each chamber as the photoanode,

116

biocathode and bioanode, respectively. The working volumes of these three chambers

117

were 230 mL, 100 mL and 180 mL respectively. The biocathode was connected to the

118

photoanode in the PEDeN test while to the bioanode in the OEDeN test. A 10 Ω

119

resistance was inserted into the electric circuit in series to allow the sampling of

120

current. In case of the EEDeN test, the TiO2/Ti plate was replaced by a piece of

121

carbon paper with the same projected area. The biocathode at this moment was

122

connected to the carbon paper electrode through a DC power. The operation manners

123

of this three-chamber reactor as PEDeN, OEDeN and EEDeN were schematically

124

illustrated in Figure 1. Each chamber was equipped with a Ag/AgCl reference

125

electrode (saturated KCl, 197 mV vs standard hydrogen electrode, SHE) to measure

126

the potentials of the corresponding electrodes.

127

Preparation of TiO2/Ti photoanode and denitrification biocathode. The TiO2/Ti

128

photoanode was prepared by a modified sol-gel and spin-coating method as described

129

previously by Dong et al.18 Briefly, 38 mL solution A (76% C2H5OH, 9% H2O and 15%

130

CH3COOH) was added dropwise to 62 mL solution B (54% Ti(OC4H9)4 and 46%

131

C2H5OH). Nitric acid was then added in the mixture for adjusting the pH to 1~2. This

132

solution was further kept at room temperature for 24 h for the sol-gel aging.

133

Subsequently, the prepared sol-gel was spin-coated in three layers on the Ti sheet and 6


Page 6 of 31

Page 7 of 31


134

calcined for 2 h at 500 °C in a furnace (KSL-1200X, MTI KJ Group, Hefei, China).

135

The denitrification biocathode was acclimated in the above-mentioned three-chamber

136

reactor by coupling a bioanode which

137

bioelectrochemical system using acetate as the electron donor. The active sludge

138

obtained from local domestic wastewater treatment plant was used as the inoculum for

139

the biocathode. The modified M9 nutrient medium (pH=7) amended by 10 mL·L-1

140

Wolfe’s trace elements and KNO3 (NO3-N=35 mg·L-1) was employed as the catholyte,

141

while the same solution but replacing the KNO3 by 500 mg·L-1 sodium acetate as the

142

anolyte. Both of these solutions were prior deoxygenated using argon gas (purity >

143

99.9%) and then were continuously fed into the chambers with a flowrate of 120

144

mL·d-1. The cathode and the anode were connected through a 10 Ω resistance. Once

145

the current generation was observed and kept stable, the acclimation of denitrification

146

biocathode was considered to be successful.

147

Operation. For the PEDeN, OEDeN and EEDeN tests, the reactors were operated as

148

batch-fed mode. In the PEDeN test, the denitrification biocathode was coupled with

149

the TiO2/Ti photoanode. The cathode chamber was fed with the same solution as in

150

the acclimation stage except the inoculum, while the anode chamber was just fed with

151

the modified M9 nutrient medium. A xenon lamp (PLS-SXE300UV, PerfectLight,

152

Beijing, China) was used as the light source and equipped with a 380-nm (±20 nm)

153

bandpass filter (DT380, PerfectLight, Beijing, China). The intensity of the

154

illumination was adjusted to be 30 mW·cm-2, 45 mW·cm-2 and 60 mW·cm-2

155

respectively. In order to understand the effect of the illumination on the denitrification,

was pre-cultured in a separate

7



156

the xenon lamp was switched on/off to provide a light/dark condition. In parallel, the

157

reactors equipped with the abiotic cathode were set to explore the role of bacteria in

158

denitrification. For EEDeN and OEDeN tests, the experiments were carried out under

159

dark conditions with the modified M9 nutrient medium and that amended by sodium

160

acetate (0, 68, 135, 270, 405, 540 mg·L-1) as the anolyte, respectively. In PEDeN and

161

EEDeN tests, the bioanode chamber was just filled with phosphate buffered solution

162

to exclude the possibility of acetate diffusing to the biocathode chamber. All of the

163

above-mentioned experiments were conducted with replicate reactors in a temperature

164

controlled laboratory (~28°C). For each condition, the reactors were operated at least

165

3 runs.

166

Analysis and Calculation. The TiO2/Ti photoanode and denitrification biocathode

167

were electrochemically characterized through linear scan voltammetry (LSV) and

168

cyclic voltammetry (CV), respectively, using an electrochemical working station

169

(model-660D, CHI Instruments Inc. Shanghai, China). The LSVs for photoanodes

170

were performed from the potential of -0.5 V to 1.2 V (versus SHE) with a scan rate of

171

10 mV·s-1. The CVs for denitrification biocathode were carried out with a scan rate of

172

0.5 mV·s-1 in the potential window between -0.4 V and 0.8 V (versus SHE). The

173

morphology of the TiO2/Ti photoanode and cathode biofilm were examined by a field

174

emission scanning electron microscope (FESEM, SU8020 Hitachi, Ltd, Japan). The

175

crystallographic structure of the TiO2 thin film was characterized by an X-ray

176

diffraction (PANalytical PW 3050 Philips X’pert Pro). The electrode potential and the

177

current flowing through the sampling resistance were recorded every 10 minutes by a 8


Page 8 of 31

Page 9 of 31


178

data acquisition system (Model 2700, Keithley Instru. Inc., U.S.). The NO3- and NO2-

179

were measured using an ion chromatograph (883 Basic IC plus. Metrohm,

180

Switzerland) equipped with a Metrosep A Supp 5-250 column （ Metrohm,

181

Switzerland）. The concentration of N2O in the headspace of cathode chamber (~10

182

mL) was determined using a gas chromatographic mass spectrometry as described in

183

Cheng et al.19 The N2O concentration in the liquid phase was calculated according to

184

that in the gas phase using Henry’s law. The reported aqueous N2O concentration was

185

modified by adding those N2O in the gas phase. The NH4+ ions were detected

186

according to the Nessler reagent colorimetry method. The sodium acetate was

187

measured by a high performance liquid chromatography (HPLC, e2695, Waters Co.,

188

U.S.) equipped with a Aminex HPX-87 column (Bio-Rad Laboratories, Inc, U.S.) at a

189

wavelength of 210 nm. The current efficiency for denitrification at cathode was

190

calculated as the ratio of the theoretical numbers of electron transfer basing on the

191

transformation of nitrogen compounds to the observed numbers of electron transfer in

192

the batch-fed experiments:

193

current efficiency =

) ]×× [ ×(

!"#

(1)

194

' # where $%& (mM) is the initial nitrate nitrogen concentration. $%& (mM) % %

195

and $%# &% (mM) are the concentrations of nitrite nitrogen and nitrous oxide

196

nitrogen, respectively at t (s) time. V (L) is the working volume of the cathode

197

chamber. I (mA) is the recorded current. F is the Faraday’s constant (96485 C·mol-1).

198

5 is the electron transfer number of nitrate to nitrogen gas. 2 is the electron transfer

199

number of nitrate to nitrite. 4 is the electron transfer number of nitrate to nitrous 9



200

oxide.

201

Results

202

Characterization of TiO2 photoanode and denitrification biocathode. As shown in

203

Figure 2a, the sol-gel formed a film of TiO2 on Ti sheets. This TiO2 film has visible

204

cracks on it, which was consistent with the previous studies and would likely provide

205

a large reaction surface.20,

206

crystallographic structure of the TiO2 is indexed as the anatase phase. Current

207

generation as a function of electrode potential is shown as in Figure 2c. Negligible

208

photocurrent was observed under dark condition, while photocurrent produced when

209

the electrode potential was positively scanned to -0.25 V vs SHE and finally reached

210

4.5 mA at 1.2 V vs SHE (scan rate of 10 mV·s-1). Since the performance of

211

light-to-current conversion could be varied among individual prepared photoanode

212

(data not shown), only those photoanodes with similar photoelectrochemical behavior

213

were used for the following experiments.

21

According to the XRD analysis (Figure 2b), the

214

In terms of the acclimation of denitrification biocathode (Figure 2e), the current

215

generation presents a lag time of two days and then rapidly increases from day 2 to

216

day 8. Subsequently, the current kept stable for 4 days at around -3.5 mA. These

217

results suggest the successful development of denitrification biocathode. The

218

denitrification biocathode was then electrochemically characterized by the CV test

219

with the scan rate at 0.5 mV·s-1. As shown in Figure 2f, an obvious reduction peak is

220

observed in biocathode CV with the presence of nitrate, which has the onset potential

221

and peak potential at around 0.15 V and 0 V (versus SHE), respectively. As the 10


Page 10 of 31

Page 11 of 31


222

control, CV of abiotic cathode was also performed, which showed negligible faradaic

223

current, clearly indicating the peak related to nitrate reduction in the biocathode CV is

224

a result of microbial catalysis. At the end of the denitrification biocathode acclimation,

225

the graphite granules were characterized by SEM, which showed the coverage of the

226

biofilm (Figure 2d).

227

Denitrification driven by light. Nitrate concentration as well as the photocurrent was

228

monitored in the ON-OFF intermittent illumination test. As shown in Figure 3a, the

229

nitrate reduction of PEDeN system is coordinated with the current generation and

230

only observed under illumination. This result clearly reveals the nitrate reduction is

231

capable of being driven by light. Control test by coupling the photoanode with an

232

abiotic cathode was carried out under the same ON-OFF intermittent illumination

233

condition, in which the dosing of illumination did not arouse current generation. As a

234

result, the change of nitrate concentration at the cathode chamber was negligible.

235

These observations confirm that the nitrate reduction in PEDeN system depends upon

236

both of the illumination dosing at photoanode and the microbial activity at cathode.

237

Upon illumination, the reduction of nitrate followed the pseudo first-order reaction

238

with a kinetic constant of 0.13±0.023 h-1 (n=4, fitted according to Figure 3b). During

239

the nitrate reduction, the accumulations of nitrite and nitrous oxide were observed,

240

which is consistent with the previous works.22, 23 These denitrification intermediates

241

depleted after approaching to a maximum concentration. In all batch tests, ammonium

242

was always below the detection limit, which is 0.02 mg·L-1 (Nessler reagent

243

colorimetry method), indicating the nitrate was selectively reduced to nitrogen. As 11



244

shown in Figure 3c, initially photocurrent keeps at about 1.5 mA and then declines

245

after 12 h. With the exhaustion of oxidized nitrogen, the photocurrent declined to zero.

246

The current efficiency at cathode was calculated and its value for denitrification was

247

97.02±1.4% (n=4) at the end of the batch tests. Cathode potential during the

248

experiments fell into the range between 0.05 V and -0.03 V (versus SHE), which is

249

more positive than the theoretical potential of hydrogen evolution and consequently

250

confirms the denitrification does not rely on the hydrogen as the electron mediator. In

251

addition to hydrogen evolution, other side reactions, such as the CO2 reduction to

252

methane or organics, can be also excluded because they were thermodynamically

253

unfavorable (i.e. the standard potentials for CO2 reduction to methane and organics at

254

pH 7 are about -0.24 V vs SHE and -0.28 V vs SHE, respectively, which are more

255

negative to the cathode working potential of PEDeN). Therefore, the small numbers

256

of electron loss here is likely due to the consumption by microbial growth.

257

As mentioned above, the photocurrent kept stable but not declined with the

258

depletion of oxidized nitrogen compounds during the initial 12 hours. This suggests

259

the rate-limiting factor of denitrification in the current set of PEDeN was not come

260

from the cathode side but more likely due to the low electrons providing capability of

261

the photoanode. In order to get insight to this issue, PEDeN was further tested under

262

higher illumination intensities (45 mW·cm-2 and 60 mW·cm-2). As shown in Figure S1,

263

both of the photocurrent and denitrification rate enhanced with increasing the

264

illumination intensity. When the illumination density got to 60 mW·cm-2, the

265

denitrification rate constant reached 0.15±0.02 h-1 (n=2), which was 36.36% higher 12


Page 12 of 31

Page 13 of 31


266

than that observed under 30 mW·cm-2 illumination. Compared to the PEDeN operated

267

at lower illumination intensity, no substantial difference of the current efficiency was

268

observed at higher illumination intensity (96.86±1.8%, n=4, 60 mW·cm-2). Under

269

high illumination intensity condition, the variation of photocurrent followed the

270

identical trends as that of total nitrogen, indicating the denitrification rate-limiting

271

factor switched to the cathode side (i.e. controlled by the available oxidized nitrogen

272

compounds serving as electron acceptors).

273

PEDeN verus OEDeN and EEDeN. As the reactors employed in this study have a

274

three-chamber configuration, the denitrification performance in PEDeN, OEDeN and

275

EEDeN can be compared using the same denitrification biocathode. In case of

276

OEDeN, the denitrification biocathode was coupled by an acetate consuming

277

bioanode. As shown in Figure 4a, the performance of OEDeN is sensitive to the

278

COD/N (g COD/g NO3--N) ratio. The total nitrogen (TN) removal efficiency of

279

OEDeN can be comparable to that of PEDeN only at COD/N ratio higher than 9.0.

280

The number of electrons transferred, in terms of coulombs, to biocathode are plotted

281

in Figure 4a, which also shows a positive correlation to the COD/N ratio, indicating

282

the inferior TN removal efficiency of OEDeN at low COD/N ratio was due to the

283

lacking of electron donors. The theoretical COD/N ratio for completely converting

284

nitrate to nitrogen gas should be 2.86 in OEDeN based on the stoichiometry. However,

285

the electron equivalents could be lost for the microbial assimilation or competitive

286

methanogenesis at the bioanode.24 Thus, the observed COD/N can be always higher

287

than the theoretical value. The COD/N ratio obtained in our study falls in the same 13



288

Page 14 of 31

order of magnitude as that reported by others (4.5-16 g COD to g NO3--N).10, 25

289

In the EEDeN test, the TiO2/Ti photoanode was replaced by a piece of carbon paper

290

with the same projected area. Without the applied voltage, the current generation was

291

negligible (Figure S2b). This is because the potential of electrochemical oxygen

292

evolution is more positive than that of the nitrate reduction. Although carbon dioxide

293

formation at a carbon based electrode is thermodynamically possible at more negative

294

potential (

Microbial photoelectrotrophic denitrification as a sustainable and

Recommend Documents