JES-00152; No of Pages 10 J O U RN A L OF E N V I RO N ME N TA L S CI EN CE S X X (2 0 1 4 ) XX X–XXX

Available online at www.sciencedirect.com

ScienceDirect www.journals.elsevier.com/journal-of-environmental-sciences

4Q3

Qinghai Hu, Zhongjin Xiao, Xinmei Xiong, Gongming Zhou, Xiaohong Guan⁎

5 6

O

R O

2

F

3

Predicting heavy metals' adsorption edges and adsorption isotherms on MnO2 with the parameters determined from Langmuir kinetics

1

State Key Laboratory of Pollution Control and Resources Reuse, College of Environmental Science and Engineering, Tongji University, Shanghai 200092, China. Email: [email protected]

P

7

AR TIC LE I NFO

ABSTR ACT

16 11

Article history:

Although surface complexation models have been widely used to describe the adsorption of

12 17

Received 26 February 2014

heavy metals, few studies have verified the feasibility of modeling the adsorption kinetics,

13 18

Revised 7 May 2014

edge, and isotherm data with one pH-independent parameter. A close inspection of the

14 19

Accepted 11 May 2014

derivation process of Langmuir isotherm revealed that the equilibrium constant derived from

E

D

10 9

15 20

the Langmuir kinetic model, KS-kinetic, is theoretically equivalent to the adsorption constant

T

21 34 22

in Langmuir isotherm, KS-Langmuir. The modified Langmuir kinetic model (MLK model) and modified Langmuir isotherm model (MLI model) incorporating pH factor were developed. The

35 23 36 24

Heavy metals

MLK model was employed to simulate the adsorption kinetics of Cu(II), Co(II), Cd(II), Zn(II) and

Adsorption edge

Ni(II) on MnO2 at pH 3.2 or 3.3 to get the values of KS-kinetic. The adsorption edges of heavy

37 25 38 26

Adsorption isotherm

metals could be modeled with the modified metal partitioning model (MMP model), and the

Adsorption constants

values of KS-Langmuir were obtained. The values of KS-kinetic and KS-Langmuir are very close

E

C

Keywords:

39 27

R

to each other, validating that the constants obtained by these two methods are basically the

28

same. The MMP model with KS-kinetic constants could predict the adsorption edges of heavy

29

R

metals on MnO2 very well at different adsorbent/adsorbate concentrations. Moreover, the

30

adsorption isotherms of heavy metals on MnO2 at various pH levels could be predicted

31 33 42 41 40 43

N C O

reasonably well by the MLI model with the KS-kinetic constants.

32

© 2014 The Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences.

Introduction

46 47 48 49 50 51 52 53 54 55 56

The behavior, transport, and ultimate fate of heavy metals in aquatic environment are often controlled by the sorption reactions on the reactive surfaces of soil minerals (Wen et al., 1998; Serrano et al., 2009). Moreover, adsorption is found to be a promising technique to remove heavy metal ions from aqueous solutions because of its low operational and maintenance costs and high efficiency, especially for the heavy metal ions at low concentrations (Shi et al., 2005). Different empirical approaches, in particular the measurements of adsorption kinetics, adsorption isotherms and adsorption edges, have been commonly used in the studies of the adsorption behavior of heavy metals on various adsorbents including minerals. However, it

U

45 44

Published by Elsevier B.V.

is troublesome to get these data under various practical conditions. Thus, accurate description of the sorption process with mathematic modeling is critical for the assessment of environmental risk of heavy metals and the development of sound adsorption technologies for removing heavy metals. Therefore, sorption modeling has received considerable attention. Many studies have been performed on surface complexation models (CCM, DLM, TLM, CD-MUSIC) to describe the adsorption of heavy metals on pure mineral surfaces or soils and the results are quite promising (Wen et al., 1998; Ponthieu et al., 2006; Weerasooriya et al., 2007). These above-mentioned models incorporate the surface electrostatic effects and thus contain many empirical parameters, which makes the modeling process very complex and limits their

⁎ Corresponding author. E-mail: [email protected] (X. Guan).

http://dx.doi.org/10.1016/j.jes.2014.05.036 1001-0742/© 2014 The Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences. Published by Elsevier B.V.

Please cite this article as: Hu, Q., et al., Predicting heavy metals' adsorption edges and adsorption isotherms on MnO2 with the parameters determined from Langmuir kinetics..., J. Environ. Sci. (2014), http://dx.doi.org/10.1016/j.jes.2014.05.036

57 58 59 60 61 62 63 64 Q4 65 66 67 68 69

2

139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171

1. Materials and methods

173 172

1.1. Materials

174

All chemicals were of analytical grade and were purchased from Shanghai Reagent Station (China). The stock solutions containing target heavy metals were prepared by dissolving their corresponding chloride salts into 18.2 MΩ Milli-Q water and acidified with HCl to pH 2.0.

175

1.2. Batch equilibrium titration

180

Surface site densities and acidity constants of MnO2 were determined through a batch acidimetric-alkalimetric titration method. Unless otherwise specified, 0.01 mol/L NaCl was used as a background electrolyte in all experiments. The titration procedure consisted of: (1) adding appropriate amount of MnO2 to double distilled deionized water and shake for 12 hr to disperse the aggregates and hydrate adsorption sites prior to each experiment; (2) distributing 100 mL of the suspension to a series of 125-mL conical flasks; (3) adding different amounts of acid (HCl) and base (NaOH) to the 125-mL flasks to adjust pH to a range of 2.0–10.0; a control flask without acid or base addition was included in each batch; (4) capping all flasks and then shake at 220 oscillation/min using an SHZ-C shaker for 24 hr at 25 ± 1°C; (5) recording the final pH and plot the acid/base addition volume as a function of pH to obtain an overall titration curve; (6) filtering the suspensions and titrate the filtrate (50 mL) back to the final pH of the control; and (7)

181

E

D

P

R O

O

F

The consistency between the Langmuir kinetic rate constant (equilibrium constant) and the Langmuir isotherm parameter (adsorption constant) can be expected since the Langmuir isotherm expression is derived based on the Langmuir kinetics (Yiacoumi and Tien, 1995a, 1995b). Moreover, Wang et al. (2004) incorporated the variation of surface site density with pH in the Langmuir isotherm and then developed a modified metal partitioning model (MMP model) to simulate and predict the adsorption edges of heavy metals on fly ash very well with only one parameter, KS (adsorption constant). Therefore, it is feasible to model the adsorption kinetics, edge, and isotherm data with one pH-independent parameter, adsorption constant or equilibrium constant, by developing proper adsorption models. However, to the best of our knowledge, there was no researcher taking advantage of the consistency between the equilibrium constant and adsorption constant to predict the adsorption edges and isotherms based on the equilibrium constant derived from the adsorption kinetics. Manganese oxides are highly reactive minerals in sedimentary environments, where they influence the fate and transport of potentially toxic trace elements and essential micronutrients (Frierdich and Catalano, 2012). Moreover, manganese dioxide (MnO2) exhibits higher affinity and larger adsorption capacity for heavy metals than other metal oxides such as Al2O3, Fe2O3, Fe3O4, and TiO2 (Tamura et al., 1996, 1997). Therefore, MnO2 was employed as the adsorbent in this study. The objectives of this article were (1) to incorporate the pH effects into the Langmuir kinetic model and employ this modified Langmuir kinetic model (MLK model) to simulate and calculate the adsorption kinetics of heavy metals on MnO2; (2) to calculate the adsorption edges and adsorption isotherms of heavy metals on MnO2 with the equilibrium constants obtained from the MLK model; and (3) to further verify this practice by comparing the simulation results and parameters with the MMP model developed by Wang et al. (2004, 2006).

N

C

O

R

R

E

C

T

practical application (Guan et al., 2009). Venema et al. (1996) discussed the scope and limitations of such models and commented that, none of them can simultaneously describe all available data for a particular extended data set although most of these models can satisfactorily quantify the adsorption phenomenon under the specific conditions used in the experiments. Background electrolyte within the environmentally relevant ionic strength range of 0.005–0.04 mol/L (in lakes, rivers, groundwaters but not in sea) has negligible impact on the adsorption of metals over the pH range studied, as illustrated in Appendix A Fig. S1, suggesting that metal ions are specifically adsorbed on the MnO2 (Christophi and Axe, 2000). The negligible influences of ionic strength on heavy metal adsorption on various minerals had also been reported (Christophi and Axe, 2000; Tamura and Furuichi, 1997). Mahmudov and Huang (2011) reported that the change of specific chemical energy constituted the major portion of the change of total free energy of adsorption whereas electrostatic energy appeared to play a much less significant role on the adsorption process. It was argued that simpler hypotheses are more likely to be a correct priori to make sharper, more testable predictions (Jeffery and Berger, 1992). McBride (1997) pointed out that the adsorption behavior of mineral surfaces can usually be understood in terms of simple mass action principles and coordination chemistry without involving electrostatic effects. Thus, the adsorption models ignoring the electrostatic effects may be preferred for practical applications. Serrano et al. (2009) developed a non-electrostatic equilibrium model with both surface complexation and ion exchange reactions to describe the sorption of heavy metals on four soils over a wide range of pH and metal concentration. Pagnanelli and his colleagues presented two approaches to model the effect of pH on heavy metal adsorption on biomaterials (Esposito et al., 2002; Pagnanelli et al., 2003). One introduced qmax as a pH function into the Langmuir isotherm to obtain the heavy metal biosorption capacities at different pH and the other was based on the non-competitive mechanism between heavy metals and H+ protons (Esposito et al., 2002; Pagnanelli et al., 2003). To describe the binding of heavy metals to humic substances containing heterogeneous ligands, Benedetti et al. (1995) developed a Non-Ideal Competitive Adsorption model based on a Langmuir– Freundlich type affinity distribution function. Wang et al. (2004, 2006) developed a modified metal partitioning model from the Langmuir isotherm which could successfully quantify the adsorption edges of heavy metals on fly ash. Rudzinski and Plazinski (2010) also presented an equilibrium sorption isotherm equation incorporating pH effects, which could describe the pH-dependent adsorption of heavy metals. However, the above-mentioned studies did not relate the parameters of adsorption kinetics with those of adsorption edges and isotherms. The kinetics of adsorption is one of the most important factors in adsorption system design, since the adsorbate residence time and the reactor dimensions are controlled by the adsorption/ desorption kinetics. Various kinetic models, such as Statistical Rate Theory approach, pseudo-first order, pseudo-second order, and the Elovich equations, had been proposed to simulate the adsorption kinetics (Plazinski et al., 2009). Their theoretical derivations existing in literature could not be interpreted as general ones but rather as those related to a specific system or to a limited range of operating conditions (Plazinski et al., 2009). The common practice is to determine the adsorption rate constants by simulating the adsorption kinetics of heavy metals on minerals obtained under various conditions with a specific adsorption kinetic model. Thus, the adsorption rate constants calculated with these models are specific and not inter-related. For instance, the rate constants obtained by these models are pH-specific. One reason is that none of these models incorporates pH effect, which strongly affects the adsorption processes (Wang et al., 2004). Plazinski and his colleague modified the statistical rate approach, which could describe the pH-dependent kinetics of metal ion adsorption with one rate constant, but there were two uncertain parameters (α or β) ranging from 0 to 1 (Plazinski and Rudzinski, 2009; Plazinski, 2010).

U

70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138

J O U RN A L OF E N V I RO N ME N TA L S CIE N CE S X X (2 0 1 4 ) XXX –XXX

Please cite this article as: Hu, Q., et al., Predicting heavy metals' adsorption edges and adsorption isotherms on MnO2 with the parameters determined from Langmuir kinetics..., J. Environ. Sci. (2014), http://dx.doi.org/10.1016/j.jes.2014.05.036

176 177 178 179

182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197

3

J O U RN A L OF E N V I RO N ME N TA L S CI EN CE S X X (2 0 1 4 ) XX X–XXX

203

ΔV SS ¼

1.3. Batch adsorption experiments

205

218

The batch adsorption experiments were carried out to investigate the adsorption kinetics, adsorption edges and adsorption isotherms of heavy metals on MnO2. The first two steps of adsorption experiments were the same as those of the titration experiments. The pH of the solute solutions and the MnO2 suspension were adjusted to the desired level with NaOH and/or HCl solution before the adsorption experiments. The adsorption reaction was initiated by dosing pre-determined amount of stock solution of heavy metals to the MnO2 suspension. Then the flasks were sealed and shaken at 220 oscillation/min using an SHZ-C shaker at 25 ± 1°C. To investigate the adsorption kinetics, the flasks were taken out from the shaker at different time intervals. However, the adsorption reaction lasted for 24 hr to determine the adsorption edges and isotherms.

219

1.4. Chemical analyses

220 Q6

236

The morphology of MnO2 was observed with a scanning electron microscope (Philips XL-30 ESEM). The particle size was determined with laser particle analyzer (Better Size 2000). The X-ray diffraction pattern of MnO2 was collected on a Bruker D & Advance X-ray powder diffractometer (D8 Advance X). The specific surface area was determined by the Brunauer–Emmett– Teller (BET) (Micromeritics ASAP 2020 Surface Area and Porosity Analyzer) method. ZetaSizer Nano ZS 90 (Malvern Instruments, Worcestershire, UK) was used to determine the zeta potential of MnO2. Our preliminary study confirmed that the same results were obtained using either 0.22 or 0.45 μm membrane filter made of cellulose acetate. Thus, at the end of each adsorption test, the suspensions were immediately filtered through a 0.45 μm membrane filter and the filtrates were collected, acidified, and analyzed using the ICP-AES (Agilent 720ES) to determine the concentration of heavy metals. A pHS-3C pH electrode, calibrated with buffers at pH 4.00, 6.86, and 9.18, was used to measure pH.

237

1.5. Data analyses

238

SigmaPlot, a software that can perform nonlinear regression, was employed to fit the titration curves of MnO2 to determine its surface density and acidity as well as to fit the metal adsorption data to determine the adsorption constants of metals.

213 214 215 Q5 216 217

221 222 223 Q7 224 225 226 227 228 229 230 231 232 233 234 Q8 235 Q9

239 240 241 242

244 243

2. Model development

245

2.1. Titration model

246

The determination of surface site density and acidity is a key step for adsorption modeling (Guan et al., 2009; Su et al., 2008; Wang et al., 2004, 2008). The following Eq. (1) was employed to

248

1 1  þ −  H þ K Hi Hþ 0 þ K Hi

ΔV SS ¼ ΔV overall −ΔV water

250 251 252 253

ð1Þ 255 254 256 257 258 259 260 261 262 263 264 265 266 267

ð2Þ

where, ΔVoverall (mL) is the total volume of acid or base added during titration and ΔVwater for a certain pH is determined based on the ΔVwater–pH curve, and the ΔVwater–pH curve is determined by a titration experiment using a water sample that has the same ionic strength as that in MnO2 suspensions.

269 268 270 271 272 273

2.2. Incorporation of pH effect into the Langmuir kinetic model 274 The adsorption reaction on metal oxide surface, postulating that there is one type of surface site and two surface reactions controlling the metal adsorption, can be elucidated by Eqs. (3)–(4): SOH↔SO− þ Hþ

KH

275 276 277 278

ð3Þ −

where, SOH and SO are the protonated and deprotonated 279 280 MnO2 sites, respectively; KH is the acidity constant. 281 ka

SO− þ M2þ ⇌ ¼ SOMþ kb

KS

ð4Þ

where, M2+ is the dissolved free divalent heavy metal ion; SOM+ is metal-adsorbent complex, KS (L/mol) is the equilibrium constant; ka is the adsorption rate constant, and kd is the desorption rate constant. Moreover, it is assumed that one free-surface site (SO−) binds one metal ion (Wang et al., 2004). According to Langmuir kinetic model, the adsorption kinetics of heavy metals can be expressed as: −

247

V 0 STi K Hi C

where, ΔVSS (mL) is the net volume of stock acid/base (negative value for acid) solution consumed by surface sites; V0 (mL) is the total volume of MnO2 suspension; STi (mol/g) is the total concentration of surface site species i; KHi is the acidity constant of species i; C (mol/L) is the concentration of the acid/base stock solution; and [H+]0 (mol/L) is the hydrogen ion concentration of the control unit. Note the total surface site concentration STi = Γi × SS., here Γi is the surface site density for species i (mol/g) and SS (g/L) is the solid concentration. The net volume of stock acid/base solution consumed by surface sites under a certain pH condition, ΔVSS, an be calculated using Eq. (2):

T

212

C

211

E

210

R

209

R

208

N C O

207

U

206

i ¼ 1−n

249

)

D

204

X

(

F

202

O

201

simulate the net titration data for determining the surface site concentration and the corresponding acidity constant, based on the assumption that multiple types of monoprotic surface sites were present on the particle surface (Wang et al., 2004):

R O

200

developing the net titration curve by subtracting the acid/base consumed by the supernatant from the overall titration curve under the same pH conditions. Carbon dioxide was not purged since the impact arose from dissolved CO2 was counteracted by subtracting the acid/base consumed in back titration from the corresponding forward titration.

P

199

E

198

h i d M2þ dt

h i   ¼ ka M2þ ½SO− −kd SOMþ

283 282 284 285 286 287 288 289

ð5Þ

where, [] (mol/L) stands for the aqueous concentration, and t 290 291 Q10 (sec) is time. 292 At equilibrium, 293 h i d M2þ =dt ¼ 0;

K S ¼ ka =kd

ð6Þ

which is referred to KS-kinetic in the following part.

Please cite this article as: Hu, Q., et al., Predicting heavy metals' adsorption edges and adsorption isotherms on MnO2 with the parameters determined from Langmuir kinetics..., J. Environ. Sci. (2014), http://dx.doi.org/10.1016/j.jes.2014.05.036

295 294

4

J O U RN A L OF E N V I RO N ME N TA L S CIE N CE S X X (2 0 1 4 ) XXX –XXX

296

The following mass balance equations are considered:

metal adsorption, the dissolved metal concentration can be 341 expressed as: 342

297

Surface site balance:

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi   r 2 4M ffi 1 MT T MT 1 − þ 1− þ 1 þ − þ h i ST αH ST K S αH ST K S ST ST 2þ ¼ M 2 ST

h i   MT ¼ M2þ þ SOMþ

ð8Þ

where, ST (mol/L) and MT (mol/L) are the total surface site concentration and the total metal concentration, respectively. It is postulated that the concentration of soluble metal hydroxide species is negligible (Tamura and Furuichi, 1997; Pagnanelli et al., 2003).

304 305 306 307 308

R ¼ 1−

From the definition of KH, one can get: ½SOH ¼

310 309 311

 þ − H ½SO  : KH

ð9Þ

 þ − h i H ½SO   ST ¼ ½SO  þ þ MT − M2þ K H h i 1 ¼ ½SO−  þ MT − M2þ αH −

316 317 318

dt

Eq. (11) can be further modified to be: h i d M2þ dt

320 319 321

ð11Þ

h i2 h i þ αH ka M2þ þ fαH ka ðST −MT Þ þ kd g M2þ ¼ kd M T : ð12Þ

336

2.3. Adsorption edge and isotherm

337

A mathematic model describing the adsorption edge of metal ions on fly ash was developed by Wang et al. (2004). Assuming that only free metal ions are present in the system and that only one type of acid site is responsible for the

326 327 328 329 330 331 332 333 334

338 339 340

C

325

N

324

U

323

O

335

Eq. (12) indicates that the concentration of residual metal in the solution, [M2+], is a function of t, pH, total surface site concentration, and initial metal ion concentration. It is inferred as MLK model that incorporates the effect of pH (included in the αH term). The values of ka and kd could be obtained by simulating the kinetics data of heavy metal adsorption and the details of the procedure are present in Appendix A Supplementary data. The derivation process of MLK model revealed that it was only applicable at constant pH. Since buffer may influence heavy metal adsorption, no buffer was used in the adsorption tests. Thus, the adsorption kinetics experiments should be performed at low pH, which could maintain almost constant without buffer during the adsorption experiments, to obtain the values of KS-kinetic with MLK model.

322

:

h

346 347 348 349

351 350 352 353 354 355 356 357 358 359 360

ð15Þ

361 362 363

If the surface concentration (mol/g) is defined as,   SOMþ Mass adsorbent

ð16Þ

and Γmax ¼

345

i

αH K s ST M2þ   h i: SOMþ ¼ 1 þ αH K S M2þ

Γ¼

344 343

ð14Þ

The MMP model can be obtained by incorporating Eqs. (13) in (14), which can be used to describe metal partitioning under different metal loadings and different pH conditions. It was easy to derive the MMP model for a system that contains multiple surface sites following the practice of Wang et al. (2004). In addition, the modified Langmuir isotherm model (MLI model) that incorporates the pH effect in metal adsorption isotherm can be expressed as (Wang et al., 2004):

T

 h i h i h i ¼ ka αH ST þ M2þ −M T Þ M2þ −kd ðMT − M2þ :

C

−

h i d M2þ

E

315

where, αH ¼ K þK HHþ , the ratio of deprotonated sites to the total  H ½ surface sites that are not occupied by heavy metals. Then inserting Eqs. (8) and ((10) into Eq. (5), one gets:

R

314

ð10Þ

R

313 312

MT

P

Inserting Eqs. (8) and (9) into Eq. (7), one gets:

h i M2þ

R O

303

where, KS (L/mol) is the adsorption constant of heavy metal ions in Langmuir adsorption isotherm, which is referred to KS-Langmuir in the following part. Then, the metal partitioning or the ratio of metal concentration in the solid phase to the total metal concentration in the system (R) can be expressed as:

D

302 301

ð13Þ

F

Metal balance:

E

298 299 300

ð7Þ

O

  ST ¼ ½SO−  þ ½SOH þ SOMþ

365 364

ST : Mass adsorbent

ð17Þ 367 366 368

Thus, the MLI equation can be re-organized to be: h i αH K S Γ max M2þ h i : Γ¼ 1 þ αH K S M2þ

ð18Þ

370 369

3. Results and discussion

372 371

3.1. Characterization of MnO2

373

The MnO2 particles were aggregated with many small particles (Fig. 1a), resulting in a rough surface and a porous structure. Its particle size is mainly in micro-size range with the average size of 20 μm, as shown in Fig. 1b. The BET specific surface area of MnO2 was determined to be 49.1 m2/g. The XRD pattern, as illustrated in Fig. 1c, revealed that MnO2 employed in this study was a γ-MnO2 (Adelkhani, 2012). As demonstrated in Fig. 1d, the point of zero charge (pHpzc) of MnO2 was 3.1, which fell within the range of point of zero charge values reported by Zhang et al. (2008). The concentration of soluble Mn(IV) was less than 0.1 mg/L at pH ≥ 4.2, as

374

Please cite this article as: Hu, Q., et al., Predicting heavy metals' adsorption edges and adsorption isotherms on MnO2 with the parameters determined from Langmuir kinetics..., J. Environ. Sci. (2014), http://dx.doi.org/10.1016/j.jes.2014.05.036

375 376 377 378 379 380 381 382 383 384

5

J O U RN A L OF E N V I RO N ME N TA L S CI EN CE S X X (2 0 1 4 ) XX X–XXX

391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409

F

390

O

Fig. 2a shows the net titration results for MnO2 at different concentrations. The site densities and acidity constants of MnO2 were determined by fitting the titration curves using Eq. (1). The best curve fitting results were based on the two-site assumption, as indicated by smooth curves. Thus, it was hypothesized that the MnO2 surface contains two types of surface sites, denoted as sites S1 and S2. The densities (Γ) of acid sites S1 and S2 are 1.6 × 10− 4 ± 0.1 × 10−4 and 1.3 × 10− 4 ± 0.1 × 10−4 mol/g, respectively and their acidity constants (pKH) are 3.5 ± 0.1 and 7.0 ± 0.2, respectively. Since the values of pKH for these two sites are all greater than pHpzc, the deprotonated surface sites are negatively charged. As the surface is positively charged at pH below 3.1, there must be another surface site S0 which is positively charged before deprotonation. However, site S0 was not observed since our titration was limited to pH above 2.0 due to the considerable dissolution of MnO2 at pH ≤ 2.0. Fig. 2b demonstrates the speciation of S1 and S2 at various pH levels. Then the deprotonated form of surface sites S1 and S2, S1O− and S2O−, which are negatively charged and can act as Lewis base, should contribute to the adsorption of heavy metal ions

R O

389

C

a

b

E

0

0.1

1.0

10.0

100.0

Size (μm)

d

10 (221) Zeta Potential (mV)

(121)

Intensity (a.u)

U

(021)

5

1

(110)

0

-10

-20

-30 20

40

60

2 Theta (degrees) Q1

80

413 414 415 416 417 418 419 420 421 422 423 424

426

2

2

412

Fig. 3 shows the adsorption kinetics of Cu(II), Co(II), Cd(II), Zn(II), and Ni(II) on MnO2. For all metals, the adsorption was fast and majority of metals were removed within the first 120 min of contact with the adsorbent. However, the adsorption experiments lasted for 24 hr to ensure that the adsorption equilibrium could be achieved. Moreover, Fig. 3 reveals that the adsorption of heavy metals increased significantly with the increase of pH, which may be ascribed to the increase in the amount of deprotonated surface site (S2) as illustrated

3

R c γ-MnO

411

425

4

R

N C O

500 nm

410

3.3. Modeling the adsorption kinetics of metal ions

P

3.2. Surface site acidity and speciation

D

388

386

(Lewis acids). Compared to S1O−, S2O− was expected to have much greater tendency to coordinate with heavy metals since the pKH of S2 was much larger than that of S1. Simulating the adsorption edges of heavy metals with the MMP model for two types of surface sites, when both S1O− and S2O− were considered as adsorption sites, revealed that the adsorption constants of heavy metals on surface site S1 were much smaller than those on surface site S2 or even negative, as listed in Appendix A Table S1. Moreover, the simulation results with MMP model for single type of surface sites (S2O−) were almost identical to those with MMP model for two types of surface sites. Therefore, S2O− should be the adsorption sites accounting for the adsorption of heavy metals and only one type of surface sites (S2O−) was considered in the following modeling process.

T

387

shown in Appendix A Fig. S2. However, the dissolution of MnO2 was not negligible at very low pH and the concentration of Mn(IV) was as high as 11.8 mg/L at pH = 2.0.

E

385

0

2

4

6

8

10

12

pH

Fig. 1 – (a) SEM image of MnO2; (b) Particle size distribution of MnO2; (c) XRD patterns of γ-MnO2 employed in this study; (d) zeta potential of MnO2 as a function of pH using 0.01 mol/L NaCl as the background electrolyte, the concentration of MnO2 was 0.1 g/L. Please cite this article as: Hu, Q., et al., Predicting heavy metals' adsorption edges and adsorption isotherms on MnO2 with the parameters determined from Langmuir kinetics..., J. Environ. Sci. (2014), http://dx.doi.org/10.1016/j.jes.2014.05.036

427 428 429 430 431 432 433 434

6

J O U RN A L OF E N V I RO N ME N TA L S CIE N CE S X X (2 0 1 4 ) XXX –XXX

1.6 Surface site density (×10-4 mol/g)

a

pH

8

6 Modeling 5 g/L 10 g/L 50 g/L

4

2 -0.15

-0.10

-0.05 0.00 0.05 0.10 0.1 mol/L base added (mL L/g)

b

1.4 1.2 S1OH

1.0

S2OH

0.8 0.6 S1 O -

0.4

S2O-

0.2 0.0 2

0.15

4

6 pH

8

10

F

10

446 447 448 449 450 451 452

3.4. Predicting the metal adsorption edges

466

E

D

P

453

T

445

C

444

E

443

100

Cu(II), pH = 4.5 ± 0.3

80

Cu(II), pH = 3.2

60

80

40

50

30

25

15

0

0 100 80 60

0

04080120

40

20

04080120 Co(II), pH = 4.5 ± 0.3

0

Cd(II), pH = 5.1 ± 0.3

Co(II), pH = 3.3

Cd(II), pH = 3.3

04080120

0 40

200 400 600 800 1000 1200 1400 Time (min)

20 20 10 0

40

0 04080120

20

Zn(ll), pH = 5.1 ± 0.3 Zn(ll), pH = 3.3

0 0

200 400 600 800 1000 1200 1400 0 Time (min)

454 455 456 457 458 459 460 461 462 463 464 465

The adsorption edges of Cu(II), Co(II), Cd(II), Zn(II), and Ni(II) ions 467 on MnO2 were determined in the pH range of 2.0–7.0 and shown 468

R

442

demonstrates that the calculations (dotted lines) coincide with the experimental data reasonably well, indicating that the adsorption kinetics of heavy metal on MnO2 at various pH could be described with a single set of parameters. However, this model could not predict the adsorption kinetics of heavy metal at higher pH so well as it did at lower pH, which should be ascribed to the variation of pH (±0.3) when the adsorption kinetics was performed in the pH range 4.5–5.1. In addition, this MLK model can be employed to predict the adsorption kinetics carried out at different concentrations of metals and adsorbent, since they are also incorporated in this model. Appendix A Fig. S4 shows the Ni(II) adsorption kinetics on MnO2 in the first 120 min of adsorption at different pH levels.

R

441

O

440

C

439

N

438

U

437

in Fig. 2b. The adsorption kinetics of heavy metals on MnO2 fitted very well to the pseudo second order kinetic model with a high correlation coefficient (R2 > 0.99), as illustrated in Appendix A Fig. S3, indicating that the adsorption of heavy metals on MnO2 was reaction-controlled rather than mass transfer-controlled (Özer et al., 2004). The MLK model was used to simulate the kinetics of heavy metal adsorption on MnO2 in 120 min at pH 3.2 or 3.3, which was very low and kept almost constant during the adsorption tests, to obtain the adsorption rate constant (ka), the desorption rate constant (kd) and then the corresponding equilibrium constant (KS-kinetic). The determined ka, kd, and KS values were summarized in Table 1. As shown in Fig. 3, the MLK model could simulate the kinetics of heavy metal adsorption on MnO2 at pH 3.2 or 3.3 quite well. Because the MLK model incorporates the effect of pH, the MLK model could be applied to predict the adsorption kinetics of heavy metals at other pH levels with the parameters (ka and kd) listed in Table 1. Fig. 3

Removal efficiency (%)

436

Removal efficiency (%)

435

R O

O

Fig. 2 – (a) Net titration results for MnO2 of three concentrations (ΔVSS/SS as a function of pH); (b) surface speciation of MnO2 at different pH levels. S1, S2: surface sites.

04080120 Ni(ll), pH = 5.1 ± 0.3 Ni(ll), pH = 3.3

200 400 600 800 1000 1200 1400 Time (min)

Fig. 3 – Adsorption kinetics of heavy metals. Experimental condition: adsorbent dose, 5 g/L; initial metal concentrations (mmol/L), Cu(II) 0.104, Co(II) 0.126, Cd(II) 0.113, Zn(II) 0.125, Ni(II) 0.110. Symbols are experimental data; solid lines are simulation results with MLK model; dashed lines are adsorption kinetics predicted with the parameters got from MLK model. The insets show the first 120 min of the kinetics data at low pH for each metal ion. Please cite this article as: Hu, Q., et al., Predicting heavy metals' adsorption edges and adsorption isotherms on MnO2 with the parameters determined from Langmuir kinetics..., J. Environ. Sci. (2014), http://dx.doi.org/10.1016/j.jes.2014.05.036

7

J O U RN A L OF E N V I RO N ME N TA L S CI EN CE S X X (2 0 1 4 ) XX X–XXX

482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514

t1:1 Q2 t1:2 t1:3

t1:4

3.6. Linear correlation

535

The divalent heavy metal ions also form hydroxo complexes in homogeneous solutions (hydrolysis) with different reactivities, which can be evaluated by the reported stability constants of the hydroxo complexes. The adsorption constants of heavy metals on MnO2 were compared with the reactivities of these heavy metals to form hydroxo complexes, as illustrated in Appendix A Fig. S5a. There is a good correlation between equilibrium constants and the stability of MOH+ (Stumm and Morgan, 1996). Similar phenomenon was observed for heavy metal adsorption on both MnO2 and Fe2O3 (Tamura and Furuichi, 1997). The similar tendency of heavy metal ions for adsorption and hydrolysis indicates similarities in the two reactions, which can be regarded as the interaction between Lewis acid and Lewis base. The deprotonated surface hydroxyl sites and hydroxide ions are Lewis base and they interact with heavy metal ions, which are Lewis acid, via their oxygen atoms. Thus, the heavy metal ions with higher reactivity to form hydroxo complexes have higher adsorption affinities for MnO2. Consequently, the adsorption constants of other heavy metal ions on MnO2 can be estimated based on the correlation shown in Appendix Fig. S5a. Moreover, the correlation between the equilibrium constants

536

F

481

O

480

518

R O

479

Eq. (18) is a MLI model which incorporates the pH factor, included in the αH term, in the Langmuir isotherm. According to this MLI model, the variation of the adsorption capacity of one adsorbent with pH is not due to the change of adsorption constant with pH but to the change of effective adsorption site density with pH. Thus the MLI model can be applied to simulate or predict the adsorption isotherms over a wide pH range with only one adsorption constant. The adsorption isotherms of Ni(II) on MnO2 at pH 3.3–6.0 were predicted by the MLI model with KS-kinetic as the adsorption constant, as demonstrated by the surface plot in Fig. 6. The predicted adsorption isotherms matched the experimental results reasonably well over the entire concentration range. This prediction approach is also applicable to other divalent heavy metal ions and to predict the adsorption edges and isotherms of these metals over a wide range of pH and broad metal loading conditions.

P

478

517

D

477

3.5. Predicting the adsorption isotherms

E

476

and are present in Fig. 5. There is a quite good agreement 515 between the experimental data and the model predictions. 516

T

475

C

474

E

473

R

472

R

471

in Fig. 4. The adsorption of heavy metals on MnO2 increased with increasing pH. As pH increased, the net surface charge of MnO2 became more negative, resulting in an increase in cation adsorption and the “S” shaped adsorption edge was observed. These adsorption edge results are consistent with what other researchers have reported (Christophi and Axe, 2000). In neutral and low pH ranges, adsorption of the free metal ion is the only mechanism for metal adsorption because the formation of metal-hydroxide complexes is not significant (Wang et al., 2004). Tamura and Furuichi (1997) confirmed that if metal adsorption onto metal oxides took into account hydroxo complexes (MOH+), it made this adsorption path difficult to accept since the adsorption affinity of those species would be too high to be believable. Moreover, Tamura and Furuichi (1997) suggested that the adsorption of divalent heavy metal ion be through the interaction between the positive charge of the heavy metal ions and the negative charge of deprotonated hydroxyl sites –O−. Thus, only the interaction of free metal ions with the deprotonated form of surface site S2, i.e., S2O−, was considered in this study. The MMP model developed by Wang et al. (2004) (Eq. (14)) was employed to simulate the adsorption edges and the results are presented in Fig. 4. The KS-Langmuir values of different heavy metals, listed in Table 1, revealed that the adsorption affinities of heavy metal ions toward MnO2 surface followed the order of Cu(II) > Co(II) > Cd(II) ≈ Zn(II) > Ni(II). Different sequences of affinity of heavy metals for various adsorbents had been reported (Christophi and Axe, 2000), while Cu(II) always had greater affinity than the other four heavy metals investigated in this study. As discussed above, the equilibrium constants (KS-kinetic) got from the adsorption kinetics of heavy metals are theoretically identical to the adsorption constants in Langmuir isotherm. Therefore, this MMP model was employed to predict the adsorption edges of heavy metal ions on MnO2 with the equilibrium constants (KS-kinetic) determined from Eq. (5) (Fig. 5). With the equilibrium constants (KS-kinetic) determined from MLK model, MMP model could predict the adsorption edges superbly. The MMP model can also be used to predict the adsorption edges of heavy metals at different heavy metal concentrations and sorbent concentrations. Thus, the adsorption edges of Ni(II) on MnO2 (1.0, 5.0 g/L) over a wide Ni(II) concentration range were predicted by the MMP model with KS-kinetic, as shown by the surface plots in Fig. 5. The adsorption edges of Ni(II) on MnO2 at different initial Ni(II) concentrations (0.058, 0.110, 0.225, 0.550 mmol/L) were determined experimentally

N C O

470

U

469

Table 1 – Comparison of the equilibrium constants determined from the MLK model (logKS-kinetic) with those from the MMP (logKS-Langmuir) (confidence level: α < 0.05). Element

t1:5 t1:6 t1:7 t1:8 t1:9 t1:10

Cu(II) Co(II) Cd(II) Zn(II) Ni(II)

t1:11

a

Eq. (12) ka (× 104)

kd (×10− 3)

3.21 2.55 1.42 1.13 0.64

1.82 6.68 7.26 3.76 5.16

± ± ± ± ±

0.96 0.52 0.45 0.18 0.21

± ± ± ± ±

0.71 1.50 2.48 0.65 1.88

Eq. (6)

Eq. (14)

R

logKS-kinetic ((mol/L)−1)

logKS-Langmuir ((mol/L)−1)

0.937 0.980 0.951 0.974 0.929

7.25 6.58 6.29 6.48 6.09

7.35 6.54 6.39 6.34 6.17

± ± ± ± ±

0.04 0.05 0.03 0.06 0.02

R

logK1 a

0.995 0.992 0.997 0.980 0.999

6.00 4.80 3.92 5.04 4.14

Stability constant of MOH+.

Please cite this article as: Hu, Q., et al., Predicting heavy metals' adsorption edges and adsorption isotherms on MnO2 with the parameters determined from Langmuir kinetics..., J. Environ. Sci. (2014), http://dx.doi.org/10.1016/j.jes.2014.05.036

519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534

537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556

8

J O U RN A L OF E N V I RO N ME N TA L S CIE N CE S X X (2 0 1 4 ) XXX –XXX

Removal efficiency (%)

100 Cu(II)

Co(II)

Zn(II)

Ni(II)

Cd(II)

80 60 40 20

2

3

4 5 6 Equilibrium pH

7

80 60

F

Modeling 40

Predicting

20

O

Removal efficiency (%)

0 100

2

3

4 5 6 Equilibrium pH

7

2

3

4 5 Equilibrium pH

6

7

R O

0

P

Fig. 4 – Adsorption edges of heavy metals. Experimental conditions: adsorbent dose, 5 g/L; initial metal concentrations (mmol/L): Cu(II) 0.104, Co(II) 0.126, Cd(II) 0.113, Zn(II) 0.125, Ni(II) 0.110. Solid lines are simulation results with MMP model; dashed lines are predicted with the corresponding adsorption constants got from the MLK model.

561 562

4. Conclusions

565

We succeeded to use only one pH-independent metal adsorption constant (KS-kinetic), determined from the adsorption kinetics of heavy metals, for the description of heavy metals' adsorption edges and isotherms on MnO2 over a wide range of

578 577

D

Acknowledgment

R

R

C

100

60 40

20

0 8

7 Equ 6 5 ilib rium pH 4

MnO2, 5.0 g/L

100

N

80

0 ) 0.12 ol/L 0.04 (mm 0.06 i(II) 0.08 of N 0.10 ation 3 0.12 entr c con ial t i In

570 571 572 573 574 575 576

This work was supported by the National Natural Science 579 Foundation (No. 21277095), and the Tongji University Open 580

O

MnO2, 1.0 g/L

U

568

ency (%) Removal effici

567

E

564 563

566

569

E

560

pH, metal loading conditions and adsorbent concentrations. Many factors neglected in this study (surface heterogeneity, effect of coexisting ions, mass transfer etc.) may modify the mechanism of adsorption of both protons and metal ions, and thus further theoretical studies considering those weak points seem necessary. However, the simple model approach proposed in this study can be treated as a starting point for revealing the relationship among adsorption kinetics, edge, and isotherm.

T

559

determined from the MLK model (KS-kinetic) and the adsorption constants obtained from the MMP model (KS-Langmuir) is demonstrated in Appendix A Fig. S5b. There is a sound linear relationship between KS-kinetic and KS-Langmuir and their values are so close to each other, confirming that the constants obtained by these two methods are basically the same.

ency (%) Removal effici

558

C

557

80 60 40 20 0 8

7 Equ 6 ilib 5 rium

pH

4

0 ) 0.1 ol/L 0.2 ) (mm 0.3 i(II 0.4 n of N 0.5 ratio 3 0.6 cent on lc tia i n I

Fig. 5 – Adsorption edges of Ni(II) determined at different adsorbate/adsorbent concentrations. Solid lines are the prediction results at different initial metal concentrations with the corresponding adsorption constants got from the MLK model. The surfaces, plotted by the MMP model with the equilibrium constant (KS-kinetic), describe the Ni(II) adsorption edges on MnO2 at different metal/adsorbent concentrations. Please cite this article as: Hu, Q., et al., Predicting heavy metals' adsorption edges and adsorption isotherms on MnO2 with the parameters determined from Langmuir kinetics..., J. Environ. Sci. (2014), http://dx.doi.org/10.1016/j.jes.2014.05.036

9

J O U RN A L OF E N V I RO N ME N TA L S CI EN CE S X X (2 0 1 4 ) XX X–XXX

(mmol/L.g) Ni(II) adsorbed

0.14 0.12 0.10 0.08 0.06 0.04 0.02

14 12 l/L) mo 8 10 II) (m i( 6 of N 4 3.5 on i t 2 a r 3.0 0 ent onc al c i t i In

F

4.5 pH 4.0

O

5.0

R O

0 6.0 5.5

584 583

Appendix A. Supplementary data

585 586

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.jes.2014.05.036.

58 7

REFERENCES

588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611

Adelkhani, H., 2012. The effect of acidity of electrolyte on the porosity and the nanostructure morphology of electrolytic manganese dioxide. Appl. Surf. Sci. 258 (17), 6232–6238. Benedetti, M.F., Milne, C.J., Kinniburgh, D.G., Vanriemsdijk, W.H., Koopal, L.K., 1995. Metal ion binding to humic substances: application of the nonideal competitive adsorption model. Environ. Sci. Technol. 29 (2), 446–457. Christophi, C.A., Axe, L., 2000. Competition of Cd, Cu, and Pb adsorption on goethite. J. Environ. Eng.-ASCE 126, 66–74. Esposito, A., Pagnanelli, F., Veglio, F., 2002. pH-related equilibria models for biosorption in single metal systems. Chem. Eng. Sci. 57 (3), 307–313. Frierdich, A.J., Catalano, J.G., 2012. Distribution and speciation of trace elements in iron and manganese oxide cave deposits. Geochim. Cosmochim. Acta 91, 240–253. Guan, X.H., Su, T.Z., Wang, J.M., 2009. Quantifying effects of pH and surface loading on arsenic adsorption on NanoActive alumina using a speciation-based model. J. Hazard. Mater. 166 (1), 39–45. Jeffery, W., Berger, J., 1992. Ockham's Razor and Bayesian analysis. Am. Sci. 64–72. Mahmudov, R., Huang, C.P., 2011. Selective adsorption of oxyanions on activated carbon exemplified by Filtrasorb 400 (F400). Sep. Purif. Technol. 77 (3), 294–300.

D

E

Funding for Materials Characterization (No. 2013080). The authors also appreciate the constructive suggestions from Dr. Li Wang.

U

N C O

R

R

E

C

T

582

P

Fig. 6 – Experimental data and predictive curves for the adsorption isotherms of Ni(II) at different pH (3.3 ± 0.1 and 4.3 ± 0.1). Experimental conditions: adsorbent dose, 5 g/L; initial Ni(II) concentration 0.110 mmol/L. Solid lines are predicted results with adsorption constants got from the MLK model. The surface, plotted by the MLI model with the equilibrium constant (KS-kinetic), describes the Ni(II) adsorption isotherms on MnO2 at different pH.

581

McBride, M.B., 1997. A critique of diffuse double layer models applied to colloid and surface chemistry. Clay Clay Miner. 45 (4), 598–608. Özer, A., Özer, D., Ekiz, H.I., 2004. The equilibrium and kinetic modelling of the biosorption of copper(II) ions on Cladophora crispata. Adsorption 10 (4), 317–326. Pagnanelli, F., Esposito, A., Toro, L., Veglio, F., 2003. Metal speciation and pH effect on Pb, Cu, Zn and Cd biosorption onto Sphaerotilus natans: Langmuir-type empirical model. Water Res. 37 (3), 627–633. Plazinski, W., 2010. Statistical rate theory approach to description of the pH-dependent kinetics of metal ion adsorption. J. Phys. Chem. C 114 (21), 9952–9954. Plazinski, W., Rudzinski, W., 2009. Modeling the effect of pH on kinetics of heavy metal ion biosorption. A theoretical approach based on the statistical rate theory. Langmuir 25 (1), 298–304. Plazinski, W., Rudzinski, W., Plazinska, A., 2009. Theoretical models of sorption kinetics including a surface reaction mechanism: a review. Adv. Colloid Interf. Sci. 152 (1–2), 2–13. Ponthieu, M., Juillot, F., Hiemstra, T., van Riemsdijk, W.H., Benedetti, M.F., 2006. Metal ion binding to iron oxides. Geochim. Cosmochim. Acta 70 (11), 2679–2698. Rudzinski, W., Plazinski, W., 2010. How does mechanism of biosorption determine the differences between the initial and equilibrium adsorption states? Adsorption 16 (4–5), 351–357. Serrano, S., O'Day, P.A., Vlassopoulos, D., Garcia-Gonzalez, M.T., Garrido, F., 2009. A surface complexation and ion exchange model of Pb and Cd competitive sorption on natural soils. Geochim. Cosmochim. Acta 73 (3), 543–558. Shi, Z.Q., Di Toro, D.M., Allen, H.E., Ponizovsky, A.A., 2005. Modeling kinetics of Cu and Zn release from soils. Environ. Sci. Technol. 39 (12), 4562–4568. Stumm, W., Morgan, J.J., 1996. Aquatic Chemistry. John Wiley & Sons, Inc., New York. Su, T.Z., Guan, X.H., Gu, G.W., Wang, J.M., 2008. Adsorption characteristics of As(V), Se(IV), and V(V) onto activated alumina: effects of pH, surface loading, and ionic strength. J. Colloid Interface Sci. 326 (2), 347–353. Tamura, H., Furuichi, R., 1997. Adsorption affinity of divalent heavy metal ions for metal oxides evaluated by modeling with the Frumkin isotherm. J. Colloid Interface Sci. 195 (1), 241–249. Tamura, H., Katayama, N., Furuichi, R., 1996. Modeling of ion-exchange reactions on metal oxides with the Frumkin isotherm .1. Acid–base and charge characteristics of MnO2, TiO2, Fe3O4, and Al2O3 surfaces and adsorption affinity of alkali metal ions. Environ. Sci. Technol. 30 (4), 1198–1204. Tamura, H., Katayama, N., Furuichi, R., 1997. The Co2+ adsorption properties of Al2O3, Fe2O3, Fe2O3, TiO2, and MnO2 evaluated by modeling with the Frumkin isotherm. J. Colloid. Interf. Sci. 195 (1), 192–202. Venema, P., Hiemstra, T., van Riemsdijk, W.H., 1996. Comparison of different site binding models for cation sorption: description of pH dependency, salt dependency, and cation–proton exchange. J. Colloid Interface Sci. 181 (1), 45–59. Wang, J.M., Teng, X.J., Wang, H., Ban, H., 2004. Characterizing the metal adsorption capability of a class F coal fly ash. Environ. Sci. Technol. 38 (24), 6710–6715. Wang, J.M., Ban, H., Teng, X.J., Wang, H., Ladwig, K., 2006. Impacts of pH and ammonia on the leaching of Cu(II) and Cd(II) from coal fly ash. Chemosphere 64 (11), 1892–1898. Wang, J.M., Wang, T., Burken, J.G., Chusuel, C.C., Ban, H., Ladwig, K., et al., 2008. Adsorption of arsenic(V) onto fly ash: a speciation-based approach. Chemosphere 72 (3), 381–388. Weerasooriya, R., Tobschall, H.J., Bandara, A., 2007. Modeling interactions of Hg(II) and bauxitic soils. Chemosphere 69 (10), 1525–1532. Wen, X.H., Du, Q., Tang, H.X., 1998. Surface complexation model for the heavy metal adsorption on natural sediment. Environ. Sci. Technol. 32 (7), 870–875.

Please cite this article as: Hu, Q., et al., Predicting heavy metals' adsorption edges and adsorption isotherms on MnO2 with the parameters determined from Langmuir kinetics..., J. Environ. Sci. (2014), http://dx.doi.org/10.1016/j.jes.2014.05.036

612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680

10 681 682 683 684 685 686

J O U RN A L OF E N V I RO N ME N TA L S CIE N CE S X X (2 0 1 4 ) XXX –XXX

Yiacoumi, S., Tien, C., 1995a. Modeling adsorption of metal-Ions from aqueous-solutions. 1. Reaction-controlled cases. J. Colloid Interface Sci. 175 (2), 333–346. Yiacoumi, S., Tien, C., 1995b. Modeling adsorption of metal ions from aqueous solutions. 2. Transport controlled cases. J. Colloid Interface Sci. 175 (2), 347–357.

Zhang, L.Z., Ma, J., Yu, M., 2008. The microtopography of manganese dioxide formed in situ and its adsorptive properties for organic micropollutants. Solid State Sci. 10 (2), 148–153.

687 688 689 690 691

U

N

C

O

R

R

E

C

T

E

D

P

R O

O

F

692

Please cite this article as: Hu, Q., et al., Predicting heavy metals' adsorption edges and adsorption isotherms on MnO2 with the parameters determined from Langmuir kinetics..., J. Environ. Sci. (2014), http://dx.doi.org/10.1016/j.jes.2014.05.036