Removal of fluoride and hydrated silica from underground water by electrocoagulation in a flow channel reactor

Removal of fluoride and hydrated silica from underground water by electrocoagulation in a flow channel reactor

Journal Pre-proof Removal of fluoride and hydrated silica from underground water by electrocoagulation in a flow channel reactor Locksley F. Castañeda...

2MB Sizes 1 Downloads 68 Views

Journal Pre-proof Removal of fluoride and hydrated silica from underground water by electrocoagulation in a flow channel reactor Locksley F. Castañeda, Oscar Coreño, José L. Nava, Gilberto Carreño PII:

S0045-6535(19)32657-8

DOI:

https://doi.org/10.1016/j.chemosphere.2019.125417

Reference:

CHEM 125417

To appear in:

ECSN

Received Date: 2 August 2019 Revised Date:

9 November 2019

Accepted Date: 18 November 2019

Please cite this article as: Castañeda, L.F., Coreño, O., Nava, José.L., Carreño, G., Removal of fluoride and hydrated silica from underground water by electrocoagulation in a flow channel reactor, Chemosphere (2019), doi: https://doi.org/10.1016/j.chemosphere.2019.125417. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier Ltd.

Ms. Ref. No.: CHEM64794R1

Removal of fluoride and hydrated silica from underground water by

1

electrocoagulation in a flow channel reactor

2 3

Locksley F. Castañedaa, Oscar Coreñob, José L. Navaa,*, Gilberto Carreñoa

4 5

a

Departamento de Ingeniería Geomática e Hidráulica, Universidad de Guanajuato, Av.

6

Juárez 77, Centro, 36000, Guanajuato, Guanajuato, Mexico. E-mail: [email protected];

7

[email protected]; [email protected]

8 9

b

Departamento de Ingeniería Civil, Universidad de Guanajuato, Av. Juárez 77, Centro, 36000, Guanajuato, Guanajuato, Mexico. E-mail: [email protected]

10 11 12 13 14 15 16 17 18 19 20 21 22

*Corresponding

author: [email protected]

23

Tel: + 52-473-1020100 ext. 2289; fax: + 52-473-1020100 ext. 2209

24

1

Ms. Ref. No.: CHEM64794R1 25

Abstract

26

This paper concerns simultaneous removal of fluoride and hydrated silica from

27

groundwater (4.08 mg L-1 fluoride, 90 mg L-1 hydrated silica, 50 mg L-1 sulfate, 0.23 mg L-

28

1

29

up-flow EC reactor, with a six-cell stack in a serpentine array, opened at the top of the cell

30

to favor gas release. Aluminum plates were used as sacrificial electrodes. The effect of

31

current density (4 ≤ j ≤ 7 mA cm-2) and mean linear flow rate (1.2 ≤ u ≤ 4.8 cm s-1), applied

32

to the EC reactor, on the elimination of fluoride and hydrated silica was analyzed. The

33

removal of fluoride followed the WHO guideline (< 1.5 mg L-1), while the hydrated silica

34

was abated at 7 mA cm-2 and 1.2 cm s-1, with energy consumption of 2.48 kWh m-3 and an

35

overall operational cost of 0.441 USD m-3. Spectroscopic analyses of the flocs by XRD,

36

XRF-EDS, SEM-EDS, and FTIR indicated that hydrated silica reacted with the coagulant

37

forming aluminosilicates, and fluoride replaced a hydroxide from aluminum aggregates,

38

while sulfates and phosphates were removed by adsorption process onto the flocs. The

39

well-engineered EC reactor allowed the simultaneous removal of fluoride and hydrated

40

silica.

phosphate, pH 7.38 and 450 µS cm-1conductivity) by electrocoagulation (EC), using an

41 42

Keywords: Hydrated silica removal; Electrocoagulation; Aluminum electrodes; Fluoride

43

removal; Groundwater.

44 45 46 47 48

2

Ms. Ref. No.: CHEM64794R1 49 50

1. Introduction

51

Groundwater is an important source of water supply for human consumption, which in arid

52

and semiarid regions worldwide is contaminated by inorganic salts, metalloids, and metals,

53

among other pollutants. Mexico is not the exception and in different areas, it is common to

54

find groundwater contaminated mainly with fluoride and hydrated silica, whose

55

concentrations range between 1-9.5 mg L-1, and 50-132 mg L-1, respectively (Guzmán et

56

al., 2016; Rosales et al., 2018; Sandoval et al., 2014). Other ions such as phosphates and

57

sulfates are also found in groundwater. The contamination of groundwater by fluorides and

58

hydrated silica occurs mainly from the dissolution of minerals in contact with water

59

(Battula et al., 2014; Behbahani et al., 2011; Emamjomeh et al., 2011; Zhu et al., 2007).

60 61

Both Mexican standard and international regulations, such as the World Health

62

Organization (WHO), indicate that the maximum allowable fluoride concentration in water

63

for human consumption must be <1.5 mg L-1. While at low concentrations fluoride is

64

beneficial in helping to combat tooth decay (Emamjomeh et al., 2009a; Essadki et al.,

65

2009), at concentrations > 4 mg L-1 it can cause severe harm to human health, such as

66

thyroid disorder, neurological damage, mottled teeth, skeletal and dental fluorosis,

67

osteoporosis and diseases in the kidneys, lungs, liver, muscles, and nerves, among others

68

(Camargo, 2003; Emamjomeh et al., 2011; Gosh et al., 2008; Hu et al., 2003; Maleki et al.,

69

2015).

70 71

Regarding hydrated silica, there is currently no official standard that establishes the

72

maximum permissible concentrations in water for human consumption. However, it is 3

Ms. Ref. No.: CHEM64794R1 73

known that prolonged exposure to these crystals can cause damage to human health, mainly

74

in the lungs, generating diseases such as tuberculosis, silicosis, bronchitis, and cancer

75

(Merget et al., 2002; Rosales et al., 2018; Sakar and Paul, 2016; Sariñana-Ruiz et al., 2017).

76

In addition to the above, hydrated silica can also seriously affect pipelines and certain unit

77

operations of industries that use this kind of water in their processes, permeating the walls

78

of equipment and causing failures (Gelover et al., 2012; Rosales et al., 2018).

79 80

One of the most used processes for fluoride removal from water is the precipitation-

81

flocculation by aluminum and calcium salts (Singh et al., 2016). This process produces too

82

much sludge because the counter-ion of the above-mentioned salts consumes coagulant

83

(Singh et al., 2013). Other processes such as adsorption (Gosh et al., 2008; Zhao et al.,

84

2011), chemical precipitation (He and Cao, 1996), and membrane process (Hu and

85

Dickson, 2006; Tor, 2007) have been used to remove fluorides from water as well.

86 87 88

In terms of electrochemical processes, electrocoagulation (EC) is a very efficient process

89

that destabilizes anions and dispersed fine particles from contaminated water by electrolysis

90

(Emamjomeh et al., 2009b). The EC technology allows the possibility of automating the

91

process, while the installation of the treatment is compact, does not need the addition of

92

chemicals, and the generation of sludge is minimal (Hu et al. 2007; Essadki et al., 2009;

93

Emamjomeh et al., 2009b), It can even be used together with other electrochemical

94

processes (Medel et al., 2019; Tirado et al., 2018).

95 96

Some reported works indicate the use of aluminum electrodes as sacrificial anodes in the

97

EC process to remove fluoride from contaminated water (Sandoval et al. 2014), using 4

Ms. Ref. No.: CHEM64794R1 98

parallel plate electrodes fitted in continuous flow reactors. These flow cell reactors, at lab-

99

scale, can be easily scalable at pilot plants (Castañeda and Nava, 2019).

100 101

The EC process involves the generation of coagulants in situ by electrodissolution of

102

aluminum sacrificial anodes, Eq. 1, while in the bulk of the solution, at neutral pH, the

103

formation of aluminum salts takes place, Eqs. 2, 3. At the cathode, the evolution of

104

hydrogen gas bubbles occurs, Eq. 4.

105 106

Al() → Al + 3e

(1)

107

  + 3  →  ( )() + 3 

(2)

108

2   + 3  →   + 6 

(3)

109

3H O + 3e → 1.5H + 3OH

(4)

110 111

According to the literature, the removal of fluoride occurs via co-precipitation of fluoro-

112

aluminum complexes and a chemical substitution reaction between a fluoride ion and a

113

hydroxide from aluminum flocs (Hu et al., 2003; Mohamad et al., 2011). Meanwhile, the

114

removal of hydrated silica occurs by the formation of aluminosilicates (Guzman et al.,

115

2016; Rosales et al., 2018). Other coexisting ions such as phosphates and sulfates have

116

been partially removed by adsorption processes on aluminum aggregates (Rosales et al.,

117

2018; Sandoval et al., 2019).

118 119

A common problem with aluminum electrodes is the anodic passivation. The use of flow

120

cells with plate electrodes favors the transport of Al3+ ions from the electrode surface to the

5

Ms. Ref. No.: CHEM64794R1 121

bulk of the solution, diminishing the passivation of the anode (Rosales et al., 2018;

122

Sandoval et al., 2019).

123

On the other hand, hydrated silica (72 mg L-1) has been efficiently removed from

124

groundwater by EC using several aluminum plate electrodes fitted in a stack of a flow EC

125

reactor opened to the atmosphere (Rosales et al., 2018). It is worth mentioning that this

126

paper shows for the first time the efficient abatement of hydrated silica by EC, highlighting

127

that aluminum reacts with silica species to yield aluminosilicates. The removal of pollutants

128

from groundwater containing silica by EC using Fe as sacrificial anodes has already been

129

tested, but unfortunately, the silica removal has been very deficient (Wan et al., 2011).

130 131

The novelty of this paper consists in the simultaneous removal of fluoride and hydrated

132

silica from real groundwater, using a filter-press reactor with a novel design, in which the

133

horizontally located aluminum plate electrodes make up a six-cell stack, while at the top,

134

the cell is opened to the atmosphere allowing the fast release of hydrogen bubbles produced

135

at the cathodes. It examines the influence of the coagulant dosage (in terms of current

136

density employed) and the retention time (dictated by the mean linear flow velocity) on the

137

efficiency of the simultaneous removal of fluoride and hydrated silica. Spectroscopic

138

analyses such as XRD, SEM-EDS, XRF-EDS, and FTIR are performed to elucidate the

139

mechanism of elimination of the pollutants contained in groundwater.

140 141

2. Materials and methods

142

2.1. Deep well water

143

The real groundwater sample was obtained from the plateau region of Guanajuato in

144

Mexico (4.08 mg L-1 fluoride, 90 mg L-1 hydrated silica, 50 mg L-1 sulfate, 0.23 mg L-1 6

Ms. Ref. No.: CHEM64794R1 145

phosphate, 263 mg L−1 alkalinity, 50 mg L−1 hardness, pH 7.38 and 450 µS cm-1

146

conductivity), which exceeds the WHO guideline for fluoride; a high concentration of

147

hydrated silica was also found.

148 149

2.2 Electrocoagulation reactor

150

The sketch of the EC reactor and its components is shown in Fig. 1. The EC flow reactor is

151

composed of a stack of aluminum plate electrodes horizontally fitted so that the electrolyte

152

flows in the form of a serpentine; the cell is opened at the top to facilitate the fast release of

153

hydrogen bubbles formed on the cathode during the EC process.

154

155 156

Fig. 1. (a) Sketch of the reactor, (b) bottom plate, (c) channel separator, (d) aluminum

157

electrode, and (e) electrolyte collector at the exit.

158

7

Ms. Ref. No.: CHEM64794R1 159

The reactor in a serpentine array consist of a six-cell stack containing 8 empty channels

160

with 3 cm width, 8 cm length and 0.46 cm thickness, and 7 parallel aluminum plates as

161

electrodes (3 cm × 8 cm × 0.46 cm, width, length and thickness, in contact with electrolyte,

162

respectively), out of which four are used as cathodes and three as anodes. The electrolyte

163

inlet is located at the bottom of the cell, having a diameter of 1.27 cm. The top of the

164

reactor was designed, after several CFD simulation trials (not shown herein), to allow the

165

fast release of the gas generated in the cell, and therefore the cell was opened to the

166

atmosphere. Moreover, at the top of the reactor, there is a window of 3.4 cm length and 1.5

167

cm height, followed by a liquid collector of 10 cm in length to transport the electrolyte

168

towards the exit. More details on the cell can be consulted elsewhere (Castañeda and Nava,

169

2019). The dimensions of the EC reactor are shown in Table SM-1.

170 171

Fig. SM-1 shows a schematic diagram of the hydraulic and electric system coupled with the

172

EC reactor, which contains a 15 L-capacity reservoir for the groundwater sample, a

173

centrifugal pump (1/125 HP, Iwaki, MD-10L), a valve and a flowmeter (0.1-1 L min-1,

174

White Industries), all this joined to each other by a 0.5-inch diameter PVC pipe. A B&K

175

Precision 1090 power source was used to supply the current during the EC trials, which

176

directly records the cell potential.

177 178

2.3 Methodology

179

The EC tests were carried out in the system shown in Fig. SM-1. Current densities (j) of 4,

180

5, 6 and 7 mA cm-2 and mean linear flow velocities (u) of 1.2, 2.4, 3.6 and 4.8 cm s-1 were

181

implemented for EC tests, which matched volumetric flow rates of 0.1, 0.2, 0.3 and 0.4 L

182

min-1, and retention times () of 55.9, 27.8, 18.5 and 13.9 s, respectively. The Faraday´s 8

Ms. Ref. No.: CHEM64794R1 183

law was used to calculate the theoretical value of the aluminum used as coagulant,

184

CAl(III)(N):

185

186

() =

" ∙ % ∙ &' ( ∙ ) ∙ * ∙ +

(1 × 10. )

(5)

187

188

where () and j are given in mg L 2 and A cm  , respectively, the molecular weight

189

of aluminum is Mw = 26.98 g mol 2 , 5 is the channel length (8 cm), the Faraday constant is

190

6= 96,485 C mol 2 , 8 = 3 is the number of electrons, 9 is the interelectrode gap (0.46

191

cm), and the factor 1×106 permits to obtain ()

192

electrocoagulation tests, 1 mg L-1 of hypochlorite (typical concentration used for

193

disinfection purposes) was added to groundwater to avoid passivation (Guzmán et al., 2016;

194

Rosales et al., 2018; Sandoval et al., 2014). It is worth to mention that the EC tests in the

195

reactor were stabilized for a period of 10 minutes to achieve the steady state. The EC tests

196

were performed in triplicate, obtaining similar results.

in mg L 2 . Before the

197 198

Once the electrolyte left the electrocoagulation reactor, it went to the jar test device, where

199

the coagulant produced inside the EC cell was slowly mixed (30 rpm) during 15 minutes,

200

so that the aggregate could grow; then, the flocs were left to rest for 60 minutes until the

201

aggregates settled. Afterward, the clarified solution, free of flocs, was analyzed to quantify

202

the residual concentration of fluoride, hydrated silica, and coexisting ions. The

203

experimental aluminum dose, () , formed in the EC trials was determined after the

204

redissolution of the flocs, using sulfuric acid to attain a pH = 2. Spectroscopy analyses were 9

Ms. Ref. No.: CHEM64794R1 205

carried out on the dry flocs. Before electrolysis, the electrodes were polished with 600

206

grade carbon emery paper, and then rinsed with plenty of water. The results were the

207

average of three EC tests.

208 209

2.4 Analytical procedure

210

2.4.1. Groundwater analysis

211

A HI 83200 multiparameter bench photometer, from Hanna instruments, was the equipment

212

used to measure hydrated silica, phosphate and sulfate. The silica analysis was carried out

213

by heteropoly blue method using the kit HI 93705. Phosphate was determined by amino

214

acid method using the HI 93706 kit and sulfate was determined by precipitation with

215

barium chloride crystals (light absorbance method) using the kit HI. The detection limit of

216

hydrated silica, phosphate, and sulfate was 0.2 mg L-1. The concentration of fluoride was

217

measured by a fluoride ion selective electrode (27502-19, Cole Palmer) with a detection

218

limit of 0.02 mg L−1. A Perkin Elmer AAnalyst™ 200 atomic absorption spectrometer, with

219

a detection limit of 0.1 mg L-1 (309.27 nm wavelength), was used to determine the

220

concentration of aluminum. Conductivity and pH measurements were carried using a

221

waterproof instrument from Hanna, model HI 991300. Analytical grade reagents were used.

222

The results were the average of three analyses.

223 224

2.4.2. Flocs characterization

225

The scanning electron microscopy (SEM) analysis was carried out using a JEOL JSM-6010

226

PLUS/LA device. The energy dispersive analysis of X-rays (EDS) was performed using a

227

JEOL detector incorporated in the SEM microscope. X-ray diffraction (XRD) analyzes

10

Ms. Ref. No.: CHEM64794R1 228

were made on a diffractometer Rigaku Ultima IV, with nickel filter and Cu K:2 radiation.

229

The elemental compositions of the flocs were determined by energy dispersive X-ray

230

fluorescence (XRF), using a Rigaku Nex CG X-ray fluorescence spectrometer, equipped

231

with an X-ray tube with Pd anode. The Fourier transform infrared spectroscopy (FTIR)

232

examination in the flocs was carried out in a Perkin Elmer Spectrum GX FTIR

233

Spectrometer, using an EasiDiff diffuse reflectance accessory.

234

2.5. Energy consumption and costs of EC

235

The energy consumption (ABCD), cost of aluminum dose ($F() ), and overall cost of EC

236

($OC) were calculated by Eqs. (6), (7) and (8), respectively:

237

238

ABCD =

GHIJJ ∙ K (..) ∙* ∙L ∙ +

(6)

239

240

where the units of ABCD , cell potential (ABM ), and I are kWh m-3, V, and C s 2 ,

241

respectively. N is the channel width (3 cm), S is the electrode spacing (0.46 cm), and the

242

factor 3.6 is used to obtain Econs in kWh m-3.

243

$() = O() P(2.008 USD Kg 2 )(0.001)

(7)

244

245

The aluminum price, in Mexico, is 2.008 USD kg 2 and 0.001 is a conversion factor to

246

obtain $() in USD m  .

247

$ = $() + :ABCD + :EXYZX + βMY\]M

(8) 11

Ms. Ref. No.: CHEM64794R1 248

249

$OC is expressed in units of USD m-3, α is the cost of the electricity in central Mexico

250

(0.0976 USD (kWh)-1), Epump is in units of kWh m-3, and β is the sludge confinement cost

251

in Mexico (0.035 USD Kg 2 ).

252

253

3. Results and discussion

254

3.1 Removal of fluoride and hydrated silica by EC

255

Both the theoretical, () , and experimental, () , aluminum dosages and the residual

256

fluoride concentration, ^_ , are shown in Fig. 2. Results were obtained at different mean

257

linear flow velocities (1.2 < u < 4.8 cm s-1) and current densities of 4, 5, 6 and 7 mA cm-2.

258

The residual fluoride concentration after all the EC trials meets the WHO guideline (^_ <

259

1.5 mg L-1), evidencing a decrease with current density owing to the increase in the

260

experimental aluminum dosage (coagulant). A modest decrease in ^_ was obtained as a

261

function of mean linear flow velocity. The experimental aluminum dosage is greater than

262

that theoretically obtained by Faraday's law, Eq. 5, which is attributed to the chemical

263

oxidation of aluminum plates with the hypochlorite (1 mg L-1) present in the groundwater

264

sample.

265

12

Ms. Ref. No.: CHEM64794R1

266 267

Fig. 2. Effect of the mean linear flow velocity on the remaining fluoride concentration after

268

EC trials at different current densities: (a) 4 mA cm-2, (b) 5 mA cm-2, (c) 6 mA cm-2 and (d)

269

7 mA cm-2.

270 271

The residual concentrations of hydrated silica, sulfate, and phosphate are shown in Figs.

272

3(a), (b) and (c), respectively. This Fig. shows that the residual concentration of hydrated

273

silica (Chs) increases with flow rate due to the decrease in the coagulant dosage, and Chs

274

decreases with current density owing to the massive reaction between aluminum and silica

275

to yield aluminosilicates (Rosales et al., 2018). The best removal of hydrated silica was

276

obtained at 7 mA cm-2 and at 1.2 cm s-1 giving Chs = 6.2 mg L-1, equivalent to the removal

277

of 96%.

278 279 280

13

Ms. Ref. No.: CHEM64794R1

281 282

Fig. 3. Effect of the mean linear flow rate on the removal of: (a) hydrated silica, (b) sulfate

283

and (c) phosphate concentrations, after the same EC trials shown in Fig. 2.

284 285

Figure 3 (b) shows the residual concentration of sulfate, *`ab_ , that remains almost

286

constant as a function of flow rate but decreases with current density, which is attributable

287

to the coagulant dosage. The best removal of sulfate, *`ab_ = 25 mg L-1 (50% removal), was

288

obtained at 7 mA cm-2 and 1.2 cm s-1. The residual phosphate concentration, c`ad_ , Figure 14

Ms. Ref. No.: CHEM64794R1 289

3 (c), increases with u, owing to the decrease in the aluminum dosage, but decreases with j

290

(coagulant dosage). Once again, the best removal was obtained at 7 mA cm-2 and 1.2 cm s-1,

291

where the phosphate concentration was completely removed. These results agree with what

292

is reported in the literature, a well-known fact is that sulfate and phosphate ions are

293

adsorbed in the active sites of aluminum flocs (Guzmán et al., 2016). It is worth to mention

294

that the pH remains almost constant during the EC trials at a value of around 8.7. This

295

small variation is attributed to the substitution reaction between fluoride and hydroxide

296

from flocs during fluoride removal (Guzmán et al., 2016).

297 298

3.2 Flocs characterization

299 300

Figs. SM-2 (a)-(b) show SEM micrographs of the flocs obtained from EC test at 7 mA cm-2

301

and 1.2 cm s-1 at different scales. SM-2 (a) shows aggregates with sizes from less than 10

302

µm up to around 500 µm; whereas the aggregates from SM-2 (b) consist of particles with

303

sizes below 100 nm. SEM-EDS and XRF-EDS analyses were used to define the chemical

304

composition of aluminum flocs, and the results are shown in Table 1. According to the high

305

percentage of silicon obtained by SEM-EDS and XRF-EDS analyses, these revealed the

306

generation of aluminosilicate complexes (Guzmán et al., 2016; Rosales and Nava, 2018).

307

However, it was not possible to detect fluorine. The element compositions from both

308

analyses were similar, with small differences attributed to different sizes of sampled areas,

309

since the area sampled for XRF-EDS was 8.02 cm2 , compared with five areas of around

310

0.16 mm2 for SEM-EDS.

311 312

15

Ms. Ref. No.: CHEM64794R1 313

Table 1. Flocs composition obtained by XRF-EDS and SEM-EDS from EC at 7 mA cm-2

314

and 1.2 cm s-1. wt. %

Al

Si

S

Cl

K

Ca

O

Na

SEM-EDS

29.02

8.79

0.15

0.68

ND

4.49

54.92

2.01

XRF-EDS

16.45

4.90

0.27

0.45

0.22

1.87

73.24

2.60

315 316

Fig. 4 (a) shows a typical XRD pattern of flocs obtained by EC at 7 mA cm-2 and 1.2 cm s-

317

1

318

calcite

319

(Al1.52Ca0.52Na0.48O8Si2.48), oligoclase (Na0.723Ca0.277)(Al1.277Si2.723)O8, lisetite (Ca0.98Na

320

(Al3.96Si4.04)O16),

321

((Na0.98Ca0.02)(Al0.02Si2.97)O8). These aluminosilicate phases adsorb arsenates, sulfates, and

322

phosphates (Guzmán et al., 2016; Rosales and Nava, 2018). The carbonates (alkalinity)

323

contained in the groundwater precipitate as calcite after the EC tests.

. The broad peaks are produced by the superposition of peaks that could correspond to (CaCO3),

bytownite

labradorite

(Ca0.43Na0.07(Al0.92Si1.0O)4),

(Ca0.325Na0.16

(Al0.81Si1.19)O4,

anorthite

and

1.97

albite

324 325

FTIR spectrum obtained for wave numbers between 4000 and 400 cm-1 is shown in Fig. 4

326

(b), after EC tests at 7 mA cm-2 and u = 1.2 cm s-1. The peaks were identified using the

327

reference (Socrates, 2004). The peaks match with the chemical bonds O-H, Na-F, Al-O, Al-

328

O-Si, Si-O and Al-F, which agree with those reported in the literature for EC tests with

329

aluminum as a sacrificial anode (Drouiche et al., 2009; Ghosch et al., 2008; Guzmán et al.,

330

2016). The peaks placed on 599 cm-1 related to the Al-F bounding prove the chemical

331

substitution reaction between fluoride and hydroxide from aluminum aggregates (Sandoval

332

eta al., 2014). The Al-O-Si bond confirms the reaction between aluminum coagulant and

333

silica to yield aluminosilicates. It is worth mentioning that sulfate and phosphate bounds 16

Ms. Ref. No.: CHEM64794R1 334

were not detected in the FTIR spectra, possibly because those were encapsulated inside the

335

flocs; however, it is well known that both anions are removed by adsorption onto aluminum

336

aggregates (Guzmán et al., 2016; Thakur and Modal, 2017).

337

338 339

Fig. 4. Characteristic (a) XRD and (b) FTIR spectra of the flocs from EC at 7 mA cm-2 and

340

1.2 cm s-1.

341 342

3.3 Energy consumption and operational EC costs

343

According to the results condensed in Table 2, Ecell decreases with u for all the j tested,

344

which is related to the fast removal of the coagulant from the electrode to the bulk solution, 17

Ms. Ref. No.: CHEM64794R1 345

decreasing the resistance on the electrode (preventing passivation), as well as the ohmic

346

drop in the interelectrode space, to provide the fast release of H2 bubbles from the solution

347

to the atmosphere. It is noteworthy that the Msludge decreases with u but increases with j, as

348

expected; however, the sludge generated by EC is minor and varies between 0.037 < Msludge

349

< 0.425 kg m-3. The residual concentration of fluoride adheres to the WHO guideline (1.5

350

mg L-1 < CF) for all the EC trials.

351 352

The best removal of hydrated silica was obtained at 7 mA cm-2 and of 1.2 cm s-1, giving a

353

Chs = 6.2 mg L-1, with the operational cost of EC of $OC = 0.441 USD m-3. Finally, it is

354

important to mention that the operational cost of EC reported here may vary in other

355

countries, due to the fluctuation of prices of the electricity and sludge confinement.

18

Ms. Ref. No.: CHEM64794R1

Table 2. Remaining fluoride and hydrated silica concentrations after EC tests, as well as experimental aluminum dose, cost of aluminum dose, Ecell, Econs, Msludge, and overall cost of EC. Initial composition of groundwater: 4.08 mg L−1 fluoride, 90 mg L−1 hydrated silica, 0.23 mg L−1 phosphate, 50 mg L−1 sulfate, 263 mg L−1 alkalinity, 50 mg L−1 hardness, conductivity 450 µS cm−1 and pH 7.4. CAl(III) Chs Econs Epump Msludge j u $ OC Ecell ^_ τ -2 -1 -3 -1 -1 -1 -3 -3 -3 (s) (mg L ) (mg L ) (V) (kWh m ) (kWh m ) (kg m ) (USD m ) (mA cm ) (cm s ) (mg L ) 55.9 1.2 0.995 27.4 7.8 0.754 0.5 0.164 0.208 43.2 27.8 2.4 1.213 64 7.3 0.353 0.25 0.093 0.116 29.3 4 18.5 3.6 16.8 1.345 67.5 7.05 0.227 0.16 0.037 0.070 13.9 4.8 14.6 1.370 72.5 6.99 0.169 0.125 0.075 0.058 55.9 1.2 54.3 0.942 17 8.8 1.063 0.5 0.316 0.264 27.8 2.4 36.1 0.966 57 8.6 0.519 0.25 0.080 0.144 5 18.5 21.4 3.6 1.080 64 8.1 0.326 0.16 0.124 0.091 13.9 4.8 1.360 71 8.5 0.257 0.125 0.695 0.095 18.2 55.9 1.2 58 0.847 13.6 11 1.594 0.5 0.313 0.322 27.8 2.4 0.871 51 10.9 0.790 0.25 0.089 0.180 41.1 6 18.5 3.6 26.1 0.970 60.5 9.5 0.459 0.16 0.095 0.112 13.9 4.8 1.052 69 9.4 0.341 0.125 0.061 0.087 21.4 55.9 1.2* 0.764 6.2 14.7 2.486 0.5 0.425 0.441 73.2 27.8 2.4 43.3 0.804 21 14.2 1.200 0.25 0.247 0.230 7 18.5 3.6 26.5 0.823 57 12.6 0.710 0.16 0.069 0.136 13.9 4.8 20.4 0.858 65 10.3 0.435 0.125 0.050 0.094 -1 -1 -1 -1 *Other residual concentrations are: 0 mg L phosphate, 25 mg L sulfate, 178.8 mg L alkalinity, 24 mg L hardness, 416 µS cm-1 conductivity and pH 8.7.

19

Ms. Ref. No.: CHEM64794R1 337

4. Conclusions

338

The residual concentration of fluoride after the EC treatment adhered to the WHO guideline

339

(< 1.5 mg L-1), while the hydrated silica was completely removed at 7 mA cm-2 and 1.2 cm

340

s-1, with energy consumption and overall operational cost of 2.48 kWh m-3 and 0.441 USD

341

m-3, respectively. XRD, XRF-EDS, SEM-EDS and FTIR analyses on the flocs confirmed

342

that during the EC process, the aluminum reacted with silica forming aluminosilicates.

343

Meanwhile, fluoride substituted a hydroxide from aluminum flocs, and the sulfates and

344

phosphates were removed by adsorption process onto aluminum aggregates.

345 346

The well-engineered EC reactor permitted to apply high current densities to generate high

347

aluminum dosages (14-73 mg L-1) that reacted with hydrated silica allowing its complete

348

removal. Therefore, the EC process has great potential to be applied in the industry,

349

particularly for the preconditioning of water containing silica and affordable treatment

350

costs.

351 352

Acknowledgments

353

The authors thank to SICES (project No. IJ-19-78), CONACYT (project No. 759) and the

354

University of Guanajuato (projects No. 102/2019, 150/2019) for financial support. Authors

355

acknowledge Dr. Raul Miranda and Daniela Moncada from LICAMM-UG Laboratory for

356

spectroscopy analysis.

357 358

References

20

Ms. Ref. No.: CHEM64794R1 359

Battula, S.K., Cheukuri, J., Raman, N.V.V.S., Himabindu, V., Bhagawan, D., 2014.

360

Effective removal of fluoride from ground water using electro-coagulation. Int. J. Eng. Res.

361

Appl. 4, 439-445.

362 363

Behbahani, M., Alavi, M.R., Arami, M., 2011. Techno-economical evaluation of fluoride

364

removal

365

methodology. Desalin. 271, 209-218.

by

electrocoagulation

process:

Optimization

through

response

surface

366 367

Camargo J.A., 2003. Fluoride toxicity to aquatic organisms: a review. Chemosphere 50,

368

251–264.

369 370

Castañeda, L.F., Nava, J.L., 2019. Simulations of single-phase flow in an up-flow

371

electrochemical reactor with parallel plate electrodes in a serpentine array. J. Elecroanal.

372

Chem. 832, 31-39.

373 374

Drouiche, N., Aoudja, S., Hecinia, M., Ghaffour, N., Lounici, H., Mameri, N., 2009. Study

375

on the treatment of photovoltaic wastewater using electrocoagulation: fluoride removal

376

with aluminum electrodes–characteristics of products. J. Hazard. Mater. 169, 65–69.

377 378

Emamjomeh, M.M., Sivakumar, M., 2009a.

Fluoride removal by a continuous flow

379

electrocoagulation reactor. J. Environ Manage. 90, 1204–1212.

380 381

Emamjomeh, M.M., Sivakumar, M., 2009b.

Review of pollutants removed by

382

electrocoagulation and electrocoagulation/flotation processes. J. Environ Manage. 90,

383

1663–1679.

384 385

Emamjomeh, M.M., Sivakumar, M., Varyani, A.S., 2011. Analysis and the understanding

386

of fluoride removal mechanisms by an electrocoagulation/flotation (ECF) process. Desalin.

387

275, 102-106.

388

Essadki, A.H., Gourich, B., Vial, Ch., Delmas, H., Bennajah, M., 2009. Defluoridation of

389

drinking water by electrocoagulation/electroflotation in a stirred tank reactor with a 21

Ms. Ref. No.: CHEM64794R1 390

comparative performance to an external-loop airlift reactor. J. Hazard. Mater. 168, 1325-

391

1333.

392 393

Gelover, S.L., Pérez, S., Martín, A., Villegas, I.E., 2012. Electrogeneration of aluminium to

394

remove silica in water. Water Sci. Technol. 65, 434-439.

395 396

Ghosh, D., Medhi, C.R., Purkait, M.K., 2008. Treatment of fluoride containing drinking

397

water by electrocoagulation using monopolar and bipolar electrode connections.

398

Chemosphere 73, 1393–1400.

399 400

Guzmán, A., Nava, J.L., Coreño, O., Rodríguez, I., Gutiérrez, S., 2016. Arsenic and

401

fluoride removal from groundwater by electrocoagulation using a continuous filter-press

402

reactor. Chemosphere 144, 2113–2120.

403 404

He, G.L., Cao, S.R., 1996. Assessment of fluoride removal from drinking water by calcium

405

phosphate systems. Fluoride 29, 212–216.

406 407

Hu C.Y., Lo S.L., Kuan W.H., 2003. Effects of co-existing anions on fluoride removal in

408

electrocoagulation (EC) process using aluminum electrodes. Water Res. 37, 4513–4523.

409 410

Hu, Ch.Y., Lo, S.L., Kuan, W.H., 2007. Simulation the kinetics of fluoride removal by

411

electrocoagulation (EC) process using aluminum electrodes. J Hazard Mater. 145, 180-185.

412 413

Hu, K., Dickson, J.M. 2006. Nanofiltration membrane performance on fluoride removal

414

from water. J. Membr. Sci. 279, 529–538.

415 416

Maleki, A., Mahvi, A.H., Daraei, H., Rezaei, R., Meihami, N., Mohammadi, K., Zandi S.,

417

2015.

418

electrocoagulation/electroflotation. Fluoride. 48, 23-47.

Influence

of

selected

anions

on

fluoride

removal

in

419

22

Ms. Ref. No.: CHEM64794R1 420

Medel A., Ramírez J.A., Cárdenas J., Sirés I., Meas Y., 2019. Evaluating the

421

electrochemical and photoelectrochemical production of hydroxyl radical during

422

electrocoagulation process. Sep. Purif. Technol. 208, 59-67.

423 424

Merget, R., Bauer, T., Küpper, H.U., Philippou, S., Bauer, H.D., Breitstad, R., Bruening,

425

T., 2002. Health hazards due to the inhalation of amorphous silica. Arch. Toxicol. 75, 625-

426

634.

427 428

Mohammad, M., Muttucumaru, S., 2011. Analysis and the understanding of fluoride

429

removal mechanisms by an electrocoagulation/flotation (ECF) process. Desalin. 275 102–

430

106.

431 432

Rosales, M., Coreño, O., Nava, J.L., 2018. Removal of hydrated silica, fluoride and arsenic

433

from groundwater by electrocoagulation using a continuous reactor with a twelve-cell

434

stack. Chemosphere 211, 149-155.

435 436

Sakar, A., Paul, B., 2016. The global menace of arsenic and its conventional remediation –

437

A critical review. Chemosphere 158, 37-49.

438 439

Sandoval, M.A, Fuentes, R., Nava, J.L., Rodríguez, I., 2014. Fluoride removal from

440

drinking water by electrocoagulation in a continuous filter press reactor coupled to a

441

flocculator and clarifier. Sep. Purif. Technol. 134, 163–170.

442 443

Sandoval, M.A, Fuentes, R., Nava, J.L., Coreño, O., Li, Y.M., Hernández, J.H., 2019.

444

Simultaneous removal of fluoride and arsenic from groundwater by electrocoagulation

445

using a filter-press flow reactor with a three-cell stack. Sep. Purif. Technol. 208, 208-216.

446 447

Sariñana-Ruiz, Y.A., Vazquez-Arenas, J., Sosa-Rodríguez, F.S., Labastida, I., Armienta

448

M.A., Armienta, A., Aragón-Piña, A., Escobedo-Bretado, M.A., González-Valdez, L.S.,

449

Ponce-Peña, P., Ramírez-Aldaba, H., Lara R.H., 2017. Assessment of arsenic and fluorine

23

Ms. Ref. No.: CHEM64794R1 450

in surface soil to determine environmental and health risk factors in the Comarca Lagunera,

451

Mexico. Chemosphere 178, 391-401.

452 453

Singh, J., Singh, P., Singh, A., 2016. Fluoride ions vs removal technologies: a study. Arab.

454

J. Chem. 9, 815–824.

455 456

Singh, K., Lataye, D.H., Wasewar, K.L., Yoo, Ch.K., 2013. Removal of fluoride from

457

aqueous solution: status and techniques. Desalin. Water Treat. 51, 3233–3247.

458 459

Socrates, G., 2004. Infrared and Raman Characteristic Group Frequencies: Tables and

460

Charts, third ed. Wiley, United Kindom.

461 462

Thakur, L.S., Modal, P., 2017. Simultaneous arsenic and fluoride removal from synthetic

463

and real groundwater by electrocoagulation process: Parametric and cost evaluation, J.

464

Environ. Manag. 190, 102-112.

465 466

Tirado L., Gökkuş Ö., Brillas E. Sirés I., 2018. Treatment of cheese whey wastewater by

467

combined electrochemical Processes, J. Appl. Electrochem. 48, 1307-1319.

468 469

Tor, A., 2007. Removal of fluoride from water using anion-exchange membrane under

470

Donnan dialysis condition. J. Hazard. Mater. 141, 814–818.

471 472

Wan, W., Pepping, T.J., Banerji, T., Chaudhari, S., Giammar, D.E., 2011. Effects of water

473

chemistry on arsenic removal from drinking water by electrocoagulation. Water Res. 45,

474

384-392.

475 476

Zhao, X., Zhang, B., Liu, H., Qu, J., 2011. Simultaneous removal of arsenite and fluoride

477

via an integrated elctro-oxidation and electrocoagulation process. Chemosphere 83, 726-

478

729.

479

24

Ms. Ref. No.: CHEM64794R1 480

Zhu, J., Zhao, H., Ni, J., 2007. Fluoride distribution in electrocoagulation defluoridation

481

process. Sep. Purif. Technol. 56, 184-191.

25

Highlights • Abatement of fluoride and silica from underground water by electrocoagulation. • Flow channel cell with aluminum electrodes open to the atmosphere to favor H2 exit. • Significant aluminum dosages were produced at current densities > 6 mA cm-2. • Hydrated silica reacted with the coagulant forming aluminosilicates as flocs. •

The removal of fluoride followed the WHO recommendation, while silica was abated.

Authors contribution section All authors participated in the preparation of the paper contributing ideas to carry out experiments, discussions, and assisted in the final writing of the manuscript.

Declaration of Interest Statement

The authors declare that there is no conflict of interest regarding the publication of this paper.