Ads (1)

Journal of Molecular Liquids 207 (2015) 9098

Contents lists available at ScienceDirect

Journal of Molecular Liquids

j ourna l homepage: www.e lsev ie r .com/ locate /mol l iqA comparative study for the removal of methylene blue dye by N and Smodified TiO2 adsorbentsShila Jafari a,, Feiping Zhao a,b, Dongbo Zhao b, Manu Lahtinen c, Amit Bhatnagar d, Mika Sillanp a

a Lappeenranta University of Technology, LUT Chemistry, Laboratory of Green Chemistry, Sammonkatu 12, FI-50130 Mikkeli, Finlandb Colleges of Chemistry & Chemical Engineering, Hunan Normal University, Chinac University of Jyvskyl, Department of Chemistry, Laboratories of Inorganic and Analytical Chemistry, P.O. Box 35, FI-40014 Jyvskyl, Finlandd Department of Environmental Science, University of Eastern Finland, P.O. Box 1627, FI-70211 Kuopio, Finland Corresponding author.E-mail addresses: [email protected], shila.sanaz.jafari@

http://dx.doi.org/10.1016/j.molliq.2015.03.0260167-7322/ 2015 Elsevier B.V. All rights reserved.a b s t r a c ta r t i c l e i n f oArticle history:Received 9 November 2014Received in revised form 26 January 2015Accepted 14 March 2015Available online 17 March 2015

Keywords:Methylene blueTiO2AdsorbentKineticsIsothermspHSuccessful removal of methylene blue (MB) dye from aqueous solutions using nitrogen and sulfur modifiedTiO2(P25) nanoparticles has been demonstrated in this study. The modified adsorbents were characterizedusing various analytical methods, such as X-ray diffraction (XRD), scanning electron microscopy (SEM) andenergy-dispersive X-ray spectroscopy (EDS). The adsorption potential of S-TiO2, N-TiO2 and TiO2(P25) type ad-sorbents was tested for the removal of MB dye. The kinetic studies indicated that the adsorption of MB dyefollowed the pseudo-first ordermodel, while desorption processes followed the secondordermodel. The adsorp-tion capacity of the adsorbent proved to be increasing as a function of initial pH of the solution. The maximumadsorption capacities were found to be 350.66, 410.12 and 282.84mg/g of S-TiO2, N-TiO2 and TiO2(P25), respec-tively. It can be concluded thatmodification of TiO2(P25) nanoparticles with N and S leads to a higher adsorptiveremoval of MB.

2015 Elsevier B.V. All rights reserved.1. Introduction

Dyes are used as coloring substances in plethora of industries, suchas various textile industry applications, food, paper, carpets, rubbers,plastics and cosmetics [13]. The discharge of colored wastewaterfrom these industrial plants into natural streams has caused many sig-nificant problems such as increasing the toxicity and chemical oxygendemand (COD) of the effluent, and also reducing light penetration,which has a derogatory effect on photosynthetic phenomena. Fromthe aesthetic and health point of view, the presence of dyes (carcino-genic compounds in particular) in surface and underground waters isneither not safe, pleasant, nor welcomed [4]. Methylene blue (MB), acationic dye, is one of the dyes that is used extensively for dying cotton,wool and silk. It is currently estimated that 3040% of the used dyesoriginating from industrial sources are released into waste waters [5].These dyes are chemically and photolytically stable and the complex ar-omatic structures of these substances may hinder their natural biodeg-radation processes. Therefore, color removal from these waste watershas been attracted much attention [6].

Recently, research efforts have focused on the processes dealingwith removal of different types of dyes from wastewaters by physicaland chemical methods. These methods include ozonation, membranegmail.com (S. Jafari).separation, electrochemical treatment, ultrasonic techniques, photo-catalysis and adsorption [79]. All these processes have their meritsand disadvantages and among them, the adsorption process is preferredas an environmentally friendly and cost effective technique [5]. Addi-tionally, a literature survey shows that the selection of adsorbent playsan important role in determining its economic feasibility [1013].

For this purpose, the search for efficient and low cost adsorbents isongoing. Although there are a wide variety of adsorbents, the majorityof studies focus on themost common adsorbents, such as chitosan, acti-vated carbon,fly ash and sepiolite [5,1421]. Despite the availability andlow-cost of titanium dioxide (TiO2), its use has largely been overlookedas dye adsorbent [2224]. TiO2 is widely used as a photocatalyst and asan ideal adsorbent for the degradation of various organic pollutants.TiO2 is a very promising adsorbent due to the high surface reactivity, ad-sorption capacity and low-cost. Furthermore, as the pH of zero pointcharge (pHpzc) of TiO2 is ca. 6.06.8 [25,26]. It is a suitable adsorbentfor the adsorption of charged groups due to its favorable electrostatic at-traction mechanism.

Most of the studies concerning TiO2 as an adsorbent, concentratedon adsorption capacities, mechanisms, process parameters to mentiona few [9,2729]. Moreover, the adsorption ability of modified TiO2 wasalso investigated. Janus et al. reported that nitrogen modified TiO2showed higher adsorption capacity for the removal of Reactive Red198 and Direct Green 99 as compared to an unmodified TiO2 [30]. Liet al. [22,31] described dark adsorption and photodegradation experi-ments of Orange II dye on TiO2 supported porous adsorbents at different

http://crossmark.crossref.org/dialog/?doi=10.1016/j.molliq.2015.03.026&domain=pdfhttp://dx.doi.org/10.1016/j.molliq.2015.03.026mailto:[email protected]:[email protected]://dx.doi.org/10.1016/j.molliq.2015.03.026http://www.sciencedirect.com/science/journal/01677322www.elsevier.com/locate/molliq

91S. Jafari et al. / Journal of Molecular Liquids 207 (2015) 9098pH values. All supported catalysts exhibited greater adsorption thanthe unsupported TiO2. Anionic S was considered as a less favorable dop-ant for this purpose, because of the large anionic size difference be-tween S and O [32]. However, recent studies have found that S canalso be an effective dopant [32,33]. Ohno et al. have shown thatphoto-absorptivity of solgel prepared S-doped TiO2 is stronger thanthat of N-doped TiO2 in the region of visible light [33]. Adsorption is avital stage in photocatalytic degradation of organic pollutants on TiO2surface [34].

In our previouswork, adsorptive removal ofmethyl violet dye byun-modified TiO2 andmodified TiO2, was investigated [23,24]. In continua-tion of our previous work, the present study reports the performanceand capability of unmodified TiO2 as well as modified TiO2 (N and Sforms) for the adsorptive removal of MB dye from aqueous solution.TiO2 was used as an adsorbent in this study, as adsorption process isone of the most widely used and applicable methods for water treat-ment. Furthermore, it is a low cost, easy handling and environmentalfriendly process. As compared to the photocatalytic method, adsorptionprocess does not require to use UV lamp which has some limitation ofapplication from the European Union [35,36]. As pure anatase-type tita-nium dioxide (TiO2) particles have the relatively low adsorption capac-ity, so they can bemodified or doped to improve their adsorption ability.Therefore, surface modification of TiO2 by sulfur and nitrogen was doneto prepare S-TiO2 and N-TiO2 and their adsorption capacity was com-pared with that of pure TiO2 nanoparticles, in order to determine theirapplicability for the removal of MB dye as a model compound. Herein,N and S modification of TiO2 adsorbents have been introduced thatcan be applied as an efficient and competing adsorbents for those ofother TiO2 types that have already been reported as potential adsorbentmaterials.2. Experimental

2.1. Materials and methods

The materials for the preparation of N-doped TiO2 and S-doped TiO2were purchased from Sigma and Aldrich companies. TiO2 (P25, 20 nm,Degussa) was used as a precursor for the doped TiO2 and adsorbent.MB dye was supplied by Merck Co. and was used as an adsorbate with-out further purification. To prepare the stock solution, accuratelyweighed quantities of MB were dissolved in distilled water in order toobtain other desired concentrations of solutions through successivedilutions.

The X-ray powder diffraction (XRD) data was acquired using aPANanalytical X'Pert PRO alpha 1 diffractometer in BraggBrentanogeometry and a Johansson monochromator to produce Cu K1 radi-ation (1.5406 ; 45 kV, 30 mA). A gently hand-ground powder sam-ple was placed on silicon made zero-background holder usingpetrolatum jelly as an adhesive. The diffraction intensities were re-corded from a spinning sample by an X'Celerator detector using con-tinuous scanning mode in a 2-range of 10100 with a step size of0.017 and counting times of 120 s per step. Data processing wasmade by the program X'Pert High Score Plus v. 2.2 d and search-match phase identification routines were made with the same programusing an ICCD-PDF4+ powder diffraction database as the data retrievalsource [37].

An S-4800 Ultra-High Resolution Scanning Electron Microscope(SEM) with a resolution of 1 nm was used to determine the surfacecomposition of the samples. It was equipped with energy dispersiveX-ray analysis (EDS) for determining chemical composition.

The charge distribution of MB was determined with the help of thenatural bond orbital (NBO) method using Gaussian 09 program [38].The density functional theory (DFT)was employed for optimizing struc-ture at B3LYP/6-31* level [39]. No symmetry constraints were imposedduring the optimization of MB.2.2. Adsorbent synthesis

10 g/L of TiO2 was placed in an ultrasonic bath (Finn sonic m08) for30 min and was mixed with aqueous solution of 10 mL ethanol, 10 mLammonia and 2 mL nitric acid as origins for N-dopant organic nitrogencompound and stirred at room temperature for 12 h. The mixture wasdried at 80 C for 36 h. N-TiO2 powder was obtained after calcinationat 400 C for 4 h in air with a heating rate of 3 C/min [40].

S-doped TiO2 was prepared in the same way but using 64 g thioureaas the origin for S-dopant and 250 mL methanol.

2.3. Equilibrium experiments

All batch adsorption experiments were conducted in brown bottlescovered with aluminum foil (to avoid any reaction by light) providingdark conditions and at an ambient temperature. The equilibrium ad-sorption experiments were carried out to gauge the efficiency of TiO2in removing MB from aqueous solution. Batch adsorption tests wereperformed to study the adsorptive removal of MB from aqueous solu-tion by TiO2, where 5 mL of MB solution of known initial concentration(110 mg/L) (C0) and a known amount of the adsorbent (0.01 g) weretaken in the 10 mL test tubes. The solution was then placed in an IKAKS 4000 control mixer and was shaken at 200 rpm for 24 h to ensurethat adsorption process has reached equilibrium under dark conditions.The experiments were conducted in triplicates. Next, the suspensionwas separated from aqueous solution using a 0.45 m acetate mem-brane syringe filter andwas centrifuged for doped-TiO2 and unmodifiedTiO2 nanoparticles. The equilibrium concentration (Ce) of MB wasdetermined using an UV/Vis spectrometer at 664 nm (Lambda 45PerkinElmer Instruments). The following equationwas used to calculatethe amount of adsorbed dye per unitmass of adsorbents (mg/g) at equi-librium (qe):

qe C0Ce

mV 1

where in C0 and Ce are the initial and equilibrium concentrations of MB,m is the adsorbent mass (g) and V is the volume (L) of the solution.

2.4. Kinetic experiments

For kinetic study, 0.01 g of adsorbentwasmixedwith 3mL ofMB so-lution with a constant concentration (6 mg/L) under dark conditions atan ambient temperature. The variation of MB concentration over timewas measured by a UV/Visible spectrophotometer. The adsorbed dyeamounts were calculated from Eq. (1).

2.5. Effect of solution pH

Twomethodswere used to determine the influence of pH onMB re-moval efficiency. First, different initial pH values (211) were tested.The pH values were adjusted by adding 0.1 mol/L HNO3 and 0.1 mol/LNaOH solutions and measured using a pH meter (inoLab pH 730). Fur-ther experiments were carried out by adjusting pH at two constant pHvalues whichwere prepared in ammonia buffer (pH 10) and phosphatebuffer (pH 2). The initial MB concentration was 5 mg/L and the experi-ments were carried out at a constant temperature using a shaker. Agita-tion at 200 rpm was performed for 24 h. After 24 h, the samples werecentrifuged and filtered and their absorbance (A) was measured witha UV/Vis spectrophotometer. The removal percentage was calculatedfor each sample by the following equation:

A% A0A A0

100 2

where A0 is the initial absorbance of the samples.

92 S. Jafari et al. / Journal of Molecular Liquids 207 (2015) 9098Secondly, the prepared solutionswere used for kinetic studies as ex-plained in the kinetic experiments section above. The removal percent-age was calculated for each sample by Eq. (2).

3. Results and discussion

Fig. 1 shows that there are three possible mesomers of MB. The den-sity functional theory (DFT) calculations as illustrated in Fig. 2 revealedthat the S atom carries a positive charge (0.573e), while the central Natom of MB (N2 mesomer) has a negative charge (0.674e) andbranched N atoms (N1 mesomer) more negative (0.721e). This sug-gests that S mesomer more closely resembles the structure of MB. Thisis consistent with the assumption which has been reported [38].

3.1. XRD measurement

X-ray diffraction patterns for the examined S- and N-doped TiO2samples are shown in Fig. 3. According to the XRD data, the untreatedTiO2 is a mixture of anatase and rutile polymorphs (Fig. 3), with an ap-proximate weight fraction ratio of 80/20% in favor of the anatase form[37]. No other phases could be identified in the XRD patterns. Whenthe TiO2 was doped with N and S, no significant changes were observedin the case of N-doped TiO2 (the weight fraction ratio of TiO2 poly-morphs remained unchanged), whereas noticeable changes were visi-ble in S-doped TiO2 (Fig. 3).

Interestingly, the crystallinity of both TiO2 polymorphism S-dopedTiO2 clearly decreased along with the emergence of two new verybroad scattering humps on the 2-ranges of 1620 and 2232(Fig. 4). Generally, broad scattering humps originate from amorphouscontent, which has clearly increased in S-doped sample in contrast tothe untreated TiO2. This may originate from several phenomena for in-stance, some fraction of anatase and rutile phases may have experi-enced loss of crystallinity during the doping procedure, the dopant orits derivatives may be in amorphous form. The average crystallinity ofthe titania phases was evaluated using the Scherrer method [41]. Forthe anatase and rutile phases observed in the untreated TiO2 sample,crystal lattice direction dependent crystal sizes of 2530 nm and 5567 nm were determined, respectively (Table 1). For N-doped TiO2,slightly larger sizes were determined for the anatase phase, whereassimilar or slightly smaller sizes were found for the rutile phase. In thecase of S-doped TiO2 sample size-evaluation is less accurate as the crys-tallinity of the samplewas noticeably lower and someof the peaks over-lapped with the amorphous humps, clearly affecting to the peak profileanalysis. As a result, crystal size of the anatase phase seemed to remainunchanged, while the crystal size of rutile phase decreased (1724 nm)as a result of the doping process.

3.2. SEM and EDAX analysis

The morphology of TiO2(P25) and the prepared doped TiO2 (N-TiO2and S-TiO2) measured by SEM is given in Fig. 5. It is clear that the rawmaterial P25 consists of granular crystals with an average diameter ofabout 25 nm as can be seen in Fig. 5a. Also, N-TiO2 and S-TiO2 (Fig. 5band c) exhibit spherical particles and present porous structures whichare somewhat larger than those of seen on untreated TiO2, therebyFig. 1. Three possiblesuggesting that due to some aggregation processes occurring duringthe synthesis process, the larger particles are seen in latter case.

The N-TiO2 and S-TiO2 particles were about 50 and 30 nm in size, re-spectively. Table 2 shows the results of EDX measurements and atomiccomposition of the TiO2(P25), N-TiO2 and S-TiO2 samples. Ti, O, N and Sare the only detected elements in the samples. Obviously, N and S wereassigned to the doping species.

3.3. Equilibrium studies

Equilibrium data are important in the basic design of adsorption sys-tems and are critical in optimizing the use of adsorbent. They describethe relationship between the amount of adsorbate and the concentra-tion of dissolved adsorbate in the liquid at equilibrium [17]. Because ofthe marked nonlinear behavior of all the isotherms, at least a two-parameter model must be used to fit the experimental equilibriumdata. The Langmuir isotherm has been commonly used for many adsor-bate/adsorbent systems for both liquid and gas phase adsorption withsatisfactory results [42]. It can be written as:

qe qmKLCe1 KCe

3

where qe is the amount of adsorbate per unit mass of adsorbent at equi-librium, qm is the maximum adsorption capacity to form a completemonolayer coverage on the surface bound at high equilibrium of adsor-bate concentration Ce (ppm), and KL is a model parameter accounting,somehow, for the degree of affinity between the adsorbate andadsorbent.

The second model used in the present work was the Freundlich iso-therm (Eq. (4)). The Freundlich equation can be expressed as:

qe K FCe1=n 4

where KF and n are constants. The Freundlich isotherm was first con-ceived as an empirical model, although it can be derived from the as-sumption that the surface is composed of patches following anexponential decay energy distribution with Langmuir-type isothermbehavior on each patch. Yet the equation does not account for Henry'slaw behavior at low surface coverage and for the saturation of theadsorbed phase.

This last limitation was overcome using the Sips isotherm (Eq. (5),which is a modified Freundlich equation including the asymptotic satu-ration effect. Yet even this does not adequately describe the Henry's lawregion at a low concentration. The Sips isotherm is:

qqm

KCe 1=n

1 KCe 1=n: 5

Because of the resemblance of the above equation to the Langmuirisotherm, the Sips model is also known as the LangmuirFreundlichisotherm.

The adsorption equilibrium data of MB onto the TiO2(P25), N-TiO2and S-TiO2 are shown in Fig. 6. Comparison of the adsorption dataonto these adsorbents shows that N-TiO2 has a higher adsorption capac-ity (410.12 mg/g) than both S-TiO2 (350.66 mg/g) and untreated TiO2mesomers of MB.

Fig. 2.NBO charge (a) and potential distribution (b) on atoms ofMB calculated by using Gaussian 09 program at B3LYP/6-31* level. Blue ball: N; yellow ball: S; dark gray ball: C; light grayball: H. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

12 16 20 24 28 32 36 40 44 48

Inte

nsity

(co

unts

)

2 ()

TiO2

N-doped

Anatase Rutile

**S-doped

Fig. 3. XRD patterns of fresh TiO2(P25) and S- and N-doped TiO2.

12 16 20 24 28 32 36 40 44 48 52

TiO2

S-dopedInte

nsity

(co

unts

)

2 ()

Fig. 4. Comparison of XRD patterns of untreated TiO2(P25) and S-doped TiO2.

93S. Jafari et al. / Journal of Molecular Liquids 207 (2015) 9098

Table 1Crystal sizes of TiO2 polymorphs on untreated, N- and S-doped TiO2.

Peak pos. (2) Miller indices Crystallite size (nm) TiO2 form

TiO2(P25)25.33 101 29 Anatase27.46 110 52 Rutile36.11 101 56 Rutile37.83 004 25 Anatase41.26 111 67 Rutile48.06 200 22 Anatase

N-TiO225.33 101 36 Anatase27.46 110 46 Rutile36.11 101 58 Rutile37.83 004 25 Anatase41.26 111 55 Rutile48.06 200 24 Anatase

S-TiO225.33 101 26 Anatase27.46 110 17 Rutile36.11 101 22 Rutile37.83 004 34 Anatase41.26 111 24 Rutile48.06 200 24 Anatase

Table 2EDX measurement of the adsorbents.

Adsorbent Ti (A%) O (A%) N (A%) S (A%)

TiO2(P25) 52.61 47.39N-TiO2 40.14 48.73 11.12S-TiO2 46.29 35.37 18.34

94 S. Jafari et al. / Journal of Molecular Liquids 207 (2015) 9098(282.84 mg/g). The reason for higher potential for N-TiO2 might bedue to its higher surface area (53.88 m2/g) as compared to S-TiO2(12.23 m2/g) and TiO2(P25) (49.27 m2/g). Furthermore, pore size andpore volume were also higher in the case of N-TiO2. The amount ofadspecies, qe, increases with an increase in dye concentration, Ce. Fig. 6clearly indicates that the amount of adsorbed MB on N-TiO2 is greateron every studied concentration. It should be noted that dopant such asN is incorporated as anion and replaced by oxygen in the lattice of TiO2.

The experimental data was fitted into three above stated equilibri-um isothermmodels. Table 3 summarizes the obtained correlation coef-ficients, r2 values, and constants of the mentioned isotherm models foradsorption of MB onto TiO2 nanoparticles, N-TiO2 and S-TiO2. The Lang-muirFreundlich isotherm model provided the highest r2 values formodified TiO2 and the Langmuir isotherm model showed the highestr2 values for unmodified TiO2. The equilibrium adsorption data showedFig. 5. SEM images of a) TiO2(P2that the order of the adsorbed amount of MB per unit mass of adsor-bents at any concentration, decreased in following order: N-TiO2 N S-TiO2 N TiO2(P25). N-TiO2 has the lowest LangmuirFreundlich constantvalue (KLF, Table 3). These results can be explained by N- and S-dopingeffects. It can be observed that N and S doping significantly increase theadsorption capacity.

A photocatalytic process increases the efficiency of N-doping evenfurther [32]. Also, in this study, adsorption efficiency is increased byN-doping which might be explained due to increase in surface area ofN-doped TiO2. The higher adsorption capacity observed with thenitrogen-modified TiO2 was due to the size of agglomeration particlesof N-TiO2 which are smaller than S-TiO2 and TiO2(P25) as is presentedin Fig. 5. The smaller particle leads to a higher surface area which pro-vides more sorption sites for adsorption process. By comparing the re-sults obtained in this study with previously reported works (Table 4)on adsorption capacities of various adsorbents and modified TiO2 forMB removal suggests that better results are achieved in the presentstudy.3.4. Kinetic studies

The adsorption kinetics of MB onto TiO2 nanoparticles and N-TiO2and S-TiO2 were also studied. A plot for the amount of adspecies (qt)versus time is shown in Fig. 7. The results show that for N-TiO2, adsorp-tion is initially very rapid and reached to maximum value after about2 h, when the adspecies starts to desorb. Adsorption of TiO2(P25) isslower and reached to a maximum value within 6.5 h, followed by de-sorption. S-TiO2 reached the maximum value within 10 h but interest-ingly no desorption was observed for up to 24 h. It is suggested thatthese observations are due to the significant changes in crystallinity of5), b) N-TiO2, and c) S-TiO2.

Fig. 6. Adsorption of MB on TiO2(P25), N-TiO2 and S-TiO2.

Table 4Comparison of maximum adsorption capacities of dye adsorption on S-TiO2, N-TiO2, andTiO2(P25) with other adsorbents.

Adsorbent Adsorption capacity (mg/g) Reference

S-TiO2 350.66 This workN-TiO2 410.12 This workTiO2_(P25) 282.84 This workCarbon nanotubes 35 [54]Fly ash 13.42 [55]PDA microspheres 90.7 [45]Titanate nanotube 133.33 [56]

95S. Jafari et al. / Journal of Molecular Liquids 207 (2015) 9098the S-TiO2 phase, whereas such changes were not observed for N-TiO2or TiO2 nanoparticles.

For modeling of adsorption kinetic data, themost widely used equa-tions, i.e. pseudo-first order and pseudo-second order models, wereused [43,44]. The linear forms of these models are as follows:

lnqeqtqe

k1t 6

tqt

1k2q

2e tqe

7

where qe is the equilibriumvalue of qt, k1 and k2 are the rate constants ofpseudo-first order and pseudo-second order models, respectively. Theresults of fitting by the mentioned equations are presented in Table 5.As shown in Fig. 7 for N-TiO2 and TiO2(P25), the desorption processtakes place after a matter of hours, indicating that adsorption of MBonto their surface is reversible.

In other words, the majority of adspecies is not thermodynamicallystable and after time laps, the adspecies leave the surface to reach themost stable condition. This result is in agreement with earlier studies[23,24,28,45,46]. For modeling desorption kinetic data, following equa-tions were used:

Zero order:

qt q0kdt 8

First order:

qt q0ekdt 9

Second order:

qt q0q0t

t 1kdq0

10Table 3Obtained parameters of different adsorption isotherms for N-TiO2.

TiO2(P25) S-TiO2 N-TiO2

Langmuir KL (L/mg) 0.50 0.58 0.21362qm (mg/g) 300.44 338.01 400.10r2 0.9748 0.9085 0.9811

Freundlich KF 107.25 207.78 125.17n 2.31 5.59 2.32r2 0.9214 0.9310 0.9462

LangmuirFreundlich KLF (L/mg) 0.52 0.59 0.23qm (mg/g) 282.84 350.66 410.12n 0.60 2.25 0.9382r2 0.9612 0.9390 0.9814where q0 is the amount of adspecies per unit mass of adsorbent at thestart of the desorption process and kd is the desorption rate constant.

The results of fitting the experimental data are listed in Table 6.Based on the obtained correlation coefficient values, r2, the zero ordermodel describes that the desorption ofMB by N-TiO2 and TiO2 nanopar-ticles is better than the other kinetic models.

3.5. Effect of pH

The surface charge of TiO2 is highly dependent on the solution pHbecause its amphoteric nature affects adsorption capacity [34,45,47].The pH at which the zeta potential equals zero is called the isoelectricpoint (IEP) and it can be used to qualitatively assess the adsorbent sur-face charge. The reported empirical isoelectric points (abbreviated aspHiep) for TiO2 and TiO2(P25) are 28.9 and 6.26.9, respectively [45,4850]. For an initial dye concentration (5.00 mg/L, 25 C), the effectof initial pH on the absorbance percentage (A%) of MB onto TiO2(P25)and prepared doped TiO2 samples was studied after 24 h equilibration.Our findings in Fig. 8 indicate that there is an increase in the adsorbedamount of MB with increase in initial pH.

It is suggested that titanium surface is positively charged due to theprotonation of the surface hydroxyl group at low pH, while it becomesnegatively charged at higher pH values. MB has a cationic character, sothe obtained higher uptake values at higher pH are due to electrostaticattraction between the positive charged cationic dye and the negativelycharged TiO2 (pH pHiep) according to Eq. (12). Thus the presence ofTiIV-O could be responsible for the dye adsorption onto TiO2(P25) andthe modified TiO2 surface. It is clear that electrostatic attraction leadsto the observed adsorption [34,45,49].

TiIVOHHTiIVOH2 pHbpHiep 11

TiIVOH OHTiIVO H2O pHNpHiep 12

Therefore, for the cationic dye MB, more electrostatic attraction be-tween TiO2(P25), doped TiO2 and MB is available with increasedsolution pH. Therefore, TiO2(P25), S-TiO2 and N-TiO2 have a highFig. 7. Adsorption of MB onto TiO2 nanoparticles and doped TiO2 as a function of time.

Table 5The obtained constants of adsorption kinetic models at 30 C.

Adsorbent Pseudo second order Pseudo first order

qe(exp) (g/gads) qe (g/gads) k2 (gads/mgmin) r2 qe (mg/gads) k1 (1/min) r2

S-TiO2 116.5390 116.7068 0.0060 0.9714 115.6256 0.0053 0.9610N-TiO2 84.7500 85.1707 0.0057 0.9125 84.4758 0.0048 0.8691TiO2_(P25) 116.4288 116.1134 0.0030 0.9949 118.6100 0.0045 0.9942

Table 6The obtained constants of desorption kinetic models.

Adsorbent Zero order First order Second order

kd (mg/gadsmin) r2 kd (1/min) r2 kd (gads/mgmin) r2

TiO2(P25) q0 (g/gads) = 116.43 0.1153 0.9827 1.14 103 0.9437 1.10 105 0.9252N-TiO2 q0 (g/gads) = 84.76 0.0337 0.9815 5.00 104 0.8637 7.00 106 0.8218

96 S. Jafari et al. / Journal of Molecular Liquids 207 (2015) 9098adsorption capacity forMB in a broadpH range. It is noteworthy that theadsorption of MB increases slowly at pH 24 but eventually increaseswith increasing pH. The strong pH-dependent adsorption suggeststhat MB adsorption is attributed to inner-sphere surface complexationrather than ion exchange or outer-sphere surface complexation [34].Inner-sphere complexes are solution complexes that closely associatewith the charged mineral surface (TiO2). Therefore, TiO2 can adsorbMB cationic ions via formation of inner-sphere complexes through theTiIV-O group which is pH dependent and particularly influenced bypH because at pH values 4, most TiIV-OHs are protonated. Hence,there is a lower adsorption capacity at lower pH values, the adsorptioncapacity increases with increasing pH values.

FTIR spectra of TiO2 adsorbents before and after MB adsorptionwerecompared to explore the adsorption mechanism. Further (Fig. 9), com-pared with the FTIR spectra of S-TiO2, the FTIR spectra of S-TiO2(Fig. 9a) after MB adsorption showed the appearance of two newpeaks at around 881 and 1076 cm1 associated with the bending vibra-tions of the CH groups of the heterocycle of MB [51]. The peaks at1151 cm1 associated with groups of TiOS strengthened and shiftedto 1147 cm1 afterMB adsorption, indicating that sulfur of S-TiO2mightparticipate in adsorption [32]. Meanwhile, the broad band at around3110 cm1 was bathochromically shifted to 2985 cm1 after MB ad-sorption, indicating that the OH on the surface of TiO2 was also in-volved in the adsorption [52]. The bathochromic shift phenomenonFig. 8. Effect of initial pH on the adsorption ofMB onN-TiO2, S-TiO2 and TiO2 nanoparticles(initial dye concentration: 5 mg/L, time: 24 h).from 3394 cm1 to 3297 cm1 was also observed on the spectrum ofN-TiO2 (Fig. 9 b). In the case of N-TiO after MB adsorption, four newpeaks at 1338, 1388, 1602, and 1645 cm1 could reflected to thepeaks at 1332, 1388, 1593, and 1656 cm1 of MB. This confirmed thesuccessful loading of MB on the surface of N-TiO2. Furthermore, theFig. 9. FTIR spectra of a) S-TiO2 before and after adsorption on MB and b) N-TiO2 beforeand after adsorption on MB.

Fig. 10. Absorbance percentage of MB onto S-TiO2, N-TiO2 and TiO2(P25) at various pHlevels.

Fig. 11. Adsorption of MB onto TiO2(P25), S-TiO2 and N-TiO2 as a function of time at pH=10.

Fig. 12. The plot of k1 as a function of type adsorbent at different pH values, filled squares,filled circles and open triangles represent TiO2(P25), N-TiO2 and S-TiO2, respectively.

97S. Jafari et al. / Journal of Molecular Liquids 207 (2015) 9098peak at 1633 cm1 associated with bending vibration of hydroxylgroups which is indicated that N-TiO2 had more surface adsorbedwater and hydroxyl group [53]. All these observations confirmedinner-sphere complex mechanism.

Interestingly, prepared S-TiO2 demonstrates a different pattern ofcompression by two other adsorbents in the acidic and basic pH valuerange (Fig. 10). As can be seen, removal of MB by S-TiO2 is highest atpH 3 and 10. It is therefore, suggested that the distinct trend of S-TiO2might be dependent on a different phase composition that is in agree-ment with the XRD analysis, which proved a different phase composi-tion for S-TiO2. The amount of adspecies (qt) versus time at pH = 10was observed as presented in Fig. 11. The results show that the amountof adspecies increases with increasing pH. For N-TiO2 and TiO2(P25),which are in the same phase ratio, increasing pH from neutral to alka-line leads to maximum adsorption within 250 min, but no desorptionwas observed for up to 24 h, while in the neutral pH range, adsorptionwas reversible and desorption processes were observed for both N-TiO2 and TiO2(P25).Table 7The obtained constants of adsorption kinetic models at 30 C and pH= 10.

Adsorbent Pseudo-second order Pseudo-first order

qe(g/gads)

k2(gads/mgmin)

r2 qe(mg/gads)

k1(1/min)

r2

S-TiO2 1.3457 0.0015 0.9908 633.1522 0.0019 0.9958N-TiO2 1.1121 0.0038 0.9314 241.6916 0.0033 0.9900TiO2(P25) 2.0661 0.0048 0.9871 350.9451 0.0046 0.9907Adsorption kinetic data were modeled with pseudo-first order andpseudo-second order equations (Eqs. (6) and (7)). The results of fittingwith these equations are presented in Table 7. Based on the obtainedcorrelation coefficient and the equilibrium value of qt values, it is con-cluded that the pseudo-first order model describes the adsorption ki-netics of MB onto S-TiO2, N-TiO2 and TiO2(P25). It is interesting thatthe adsorption rate of MB onto TiO2(P25) increases at pH = 10, whileat neutral pH, N-TiO2 demonstrates the highest adsorption rate. Fig. 12shows the plot of k1 as a function of adsorbent type at various pH levels.Increase in solution pH led to decreased k1 values and the adsorptionrate changed at pH = 10 (TiO2(P25) N N-TiO2 N S-TiO2). It is proposedthat MB adsorption is thermodynamically stable on N-TiO2 andTiO2(P25) surfaces at pH = 10. This may be due to electrostatic attrac-tion between the positively charged MB which is a cationic dye, andthe negatively charged TiO2(P25) (pH pHiep).4. Conclusions

The results of this study indicate that TiO2 nanoparticles modifiedwithN- and S- are efficient adsorbents for the removal ofMB fromaque-ous solution. The modification of TiO2 nanoparticles with N and S in-creases both adsorption rate and capacity, thereby suggesting that dueto its higher adsorption capacity, it can be used in treating dye pollutedwastewater. The equilibriumobeys the Sips (LangmuirFreundlich) iso-therm. Thus it can be concluded that both S-TiO2 and N-TiO2 provide aheterogeneous surface for adsorption of MB. Increasing pH from 2 to11 increases their adsorption capacity. At pH = 10, the adsorption ofMB onto N-TiO2 stabilizes, while in the neutral pH range, it is notstable and starts to desorb after sometime. The result of the presentstudy suggested that S-mesomer is a possible structure ofMB. The Lang-muir model was found to fit reasonably well for TiO2-(P25) while aftermodification of TiO2 byN- and S-, data was found to fit well in the Lang-muirFreundlich model. It has been observed in our previous work andin this study also that TiO2 has a good potential to be used as an adsor-bent for the removal of cationic dyes from aqueous solution. And also,the kinetic analysis of dye adsorption onto TiO2 nanoparticles indicatesthat adsorption is reversible when anatase phase is present and TiO2 ismodified by nitrogen.Acknowledgment

The authors would like to thanks the financial support ofLappeenranta University of Technology.

98 S. Jafari et al. / Journal of Molecular Liquids 207 (2015) 9098References

[1] S. Rengaraj, S. Venkataraj, C.-w. Tai, Y. Kim, E. Repo, M. Sillanp, Self-assembledmesoporous hierarchical-like In2S3 hollow microspheres composed of nanofibersand nanosheets and their photocatalytic activity, Langmuir 27 (2011) 55345541.

[2] M. Muruganandham, R. Amutha, M.S.M.A.Wahed, B. Ahmmad, Y. Kuroda, R.P.S. Suri,J.J. Wu, M.E.T. Sillanp, Controlled fabrication of -GaOOH and -Ga2O3 self-assembly and its superior photocatalytic activity, J. Phys. Chem. C 116 (2011) 4453.

[3] M. Muruganandham, R. Amutha, G.-J. Lee, S.-H. Hsieh, J.J. Wu, M. Sillanp, Facilefabrication of tunable Bi2O3 self-assembly and its visible light photocatalytic activity,J. Phys. Chem. C 116 (2012) 1290612915.

[4] Y. Bulut, H. Aydn, A kinetics and thermodynamics study of methylene blue adsorp-tion on wheat shells, Desalination 194 (2006) 259267.

[5] B. Ismail, S.T. Hussain, S. Akram, Adsorption of methylene blue onto spinel magne-sium aluminate nanoparticles: adsorption isotherms, kinetic and thermodynamicstudies, Chem. Eng. J. 219 (2013) 395402.

[6] D.R. Baker, P.V. Kamat, Photosensitization of TiO2 nanostructures with CdS quantumdots: particulate versus tubular support architectures, Adv. Funct. Mater. 19 (2009)805811.

[7] M. Zhou, H. Srkk, M. Sillanp, A comparative experimental study on methyl or-ange degradation by electrochemical oxidation on BDD and MMO electrodes, Sep.Purif. Technol. 78 (2011) 290297.

[8] M.A. Rauf, S.S. Ashraf, Fundamental principles and application of heterogeneousphotocatalytic degradation of dyes in solution, Chem. Eng. J. 151 (2009) 1018.

[9] M. Behnajady, S. Yavari, N. Modirshahla, Investigation on adsorption capacity ofTiO2-P25 nanoparticles in the removal of a mono-azo dye from aqueous solution:a comprehensive isotherm analysis, Chem. Ind. Chem. Eng. Q. 20 (2014) 97107.

[10] S. Elemen, E.P. Akakoca Kumbasar, S. Yapar, Modeling the adsorption of textile dyeon organoclay using an artificial neural network, Dyes Pigments 95 (2012) 102111.

[11] E.A. El-Sharkawy, A.Y. Soliman, K.M. Al-Amer, Comparative study for the removal ofmethylene blue via adsorption and photocatalytic degradation, J. Colloid InterfaceSci. 310 (2007) 498508.

[12] A. Bhatnagar, W. Hogland, M. Marques, M. Sillanp, An overview of the modifica-tion methods of activated carbon for its water treatment applications, Chem. Eng.J. 219 (2013) 499511.

[13] A. Bhatnagar, M. Sillanp, Utilization of agro-industrial and municipal waste mate-rials as potential adsorbents for water treatmenta review, Chem. Eng. J. 157(2010) 277296.

[14] S. Hashemian, M. Salimi, Nano composite a potential low cost adsorbent for removalof cyanine acid, Chem. Eng. J. 188 (2012) 5763.

[15] A. Bhatnagar, M. Sillanp, Applications of chitin- and chitosan-derivatives for thedetoxification of water and wastewater a short review, Adv. Colloid Interf. Sci.152 (2009) 2638.

[16] E. Kumar, A. Bhatnagar, U. Kumar, M. Sillanp, Defluoridation from aqueous solu-tions by nano-alumina: characterization and sorption studies, J. Hazard. Mater.186 (2011) 10421049.

[17] S. Hokkanen, E. Repo, M. Sillanp, Removal of heavy metals from aqueous solutionsby succinic anhydride modified mercerized nanocellulose, Chem. Eng. J. 223 (2013)4047.

[18] E. Iakovleva, M. Sillanpaa, The use of low-cost adsorbents for wastewater purifica-tion in mining industries, Environ. Sci. Pollut. Res. Int. (2013)http://dx.doi.org/10.1007/s11356-013-1546-8.

[19] T.A. Kurniawan, M.E.T. Sillanp, M. Sillanp, Nanoadsorbents for remediation ofaquatic environment: local and practical solutions for global water pollution prob-lems, Crit. Rev. Environ. Sci. Technol. 42 (2011) 12331295.

[20] A. Bhatnagar, F. Kaczala, W. Hogland, M. Marques, M. Sillanp, Valorization of solidwaste products from olive oil industry as potential adsorbents for water pollutioncontrol a review, Environ. Sci. Pollut. Res. 21 (2014) 268298.

[21] E. Kumar, A. Bhatnagar, W. Hogland, M. Marques, M. Sillanp, Interaction of inor-ganic anions with iron-mineral adsorbents in aqueous media a review, Adv.Colloid Interf. Sci. 203 (2014) 1121.

[22] M. Janus, E. Kusiak, J. Choina, J. Ziebro, A.W. Morawski, Enhanced adsorption of twoazo dyes produced by carbon modification of TiO2, Desalination 249 (2009)359363.

[23] S. Jafari, S. Azizian, B. Jaleh, Enhancement of methyl violet removal by modificationof TiO2 nanoparticles with AgI, J. Ind. Eng. Chem. 18 (2012) 21242128.

[24] S. Jafari, S. Azizian, B. Jaleh, Adsorption kinetics of methyl violet onto TiO2 nanopar-ticles with different phases, Colloids Surf. A Physicochem. Eng. Asp. 384 (2011)618623.

[25] K.K. Ioannis, A.A. Triantafyllos, TiO2-assisted photocatalytic degradation of azo dyesin aqueous solution: kinetic and mechanistic investigations: a review, Appl. Catal. BEnviron. 49 (2004) 114.

[26] X. Tan, Q. Fan, X. Wang, B. Grambow, Eu(III) sorption to TiO2 (anatase and rutile):batch, XPS, and EXAFS studies, Environ. Sci. Technol. 43 (2009) 31153121.

[27] L. Yonggang, L. Daji, Experimental study of nanometer TiO2 for Use as an adsorbentfor SO2 removal, Dev. Chem. Eng. Miner. Process. 10 (3/4) (2002) 443457.[28] L. Zhang, Y. Zhu, H. Li, N. Liu, X. Liu, X. Guo, Kinetic and thermodynamic studies ofadsorption of gallium(III) on nano-TiO2, Rare Metals 29 (2010) 1621.

[29] S. Pimpaporn, W. Sumpun, Amorphous titanium dioxide as an adsorbent for dyepolluted water and its recyclability, J. Sol-Gel Sci. Technol. 71 (2014) 8695.

[30] M. Janus, J. Choina, E. Kusiak, A.W. Morawski, Study of nitrogen-modified titaniumdioxide as an adsorbent for azo dyes, Adsorpt. Sci. Technol. 26 (2008) 501513.

[31] G. Li a, X.S. Zhao, Madhumita B. Ray, Advanced oxidation of orange II using TiO2 sup-ported on porous adsorbents:, The role of pH, H2O2 and O3, Sep. Purif. Technol. 55(2007) 9197.

[32] S.-H. Nam, T.K. Kim, J.-H. Boo, Physical property and photo-catalytic activity of sulfurdoped TiO2 catalysts responding to visible light, Catal. Today 185 (2012) 259262.

[33] T. Ohno, T. Mitsui, M. Matsumur, Photocatalytic activity of S-doped TiO2photocatalyst under visible light, Chem. Lett. 32 (2003) 364365.

[34] V. Belessi, B. Lambropoulou, I. Konstantinou, R. Zboril, J. Tucek, D. Jancik, T. Albanis,D. Petridis, Structure and photocatalytic performance of magnetically separable tita-nia photocatalysts for the degradation of propachlor, Appl. Catal. B Environ. 87(2009) 181189.

[35] H. Moseley, Thematic Network for UltravioletMeasurements, The Photobiology UnitUniversity of Dundee, 2000. 120.

[36] I.W. Group, Exposure to Artificial UV Radiation and Skin Cancer, International Agen-cy for Research on Cancer, Lyon, 2006.

[37] C. Boulevard, International Centre for Diffraction Data, ICDD-PDF4+, Release2011,12 Campus Boulevard, Newton Square, in: Pennsylvania (Ed.), USA, 2011.

[38] A. Mills, D. Hazafy, J. Parkinson, T. Tuttle, M.G. Hutchings, Effect of alkali on methy-lene blue (C.I. Basic Blue 9) and other thiazine dyes, Dyes Pigments 88 (2011)149155.

[39] F. Zhao, E. Repo, D. Yin, M.E.T. Sillanp, Adsorption of Cd(II) and Pb(II) by a novelEGTA-modified chitosan material: kinetics and isotherms, J. Colloid Interface Sci.409 (2013) 174182.

[40] X. Cheng, X. Yu, Z. Xing, J. Wan, Enhanced photocatalytic activity of nitrogen dopedTiO2 anatase nano-particle under simulated sunlight irradiation, Energy Procedia 16(Part A) (2012) 598605.

[41] A. Patterson, The scherrer formula for X-ray particle size determination, Phys. Rev.56 (1939) 978982.

[42] L.F. Bautista, J. Pinilla, J. Aracil, M. Martnez, Adsorption isotherms of aspartame oncommercial and chemically modified divinylbenzenestyrene resins at differenttemperatures, J. Chem. Eng. Data 47 (2002) 620627.

[43] K. Sven, Vetenskapsakad, About the theory of so-called adsorption of soluble sub-stances, Handl. 4 (1898) 139.

[44] Y.S. Ho, G. McKay, Pseudo-second-order model for sorption processes, ProcessBiochem. 34 (1999) 451465.

[45] F. Jianwei, C. Zhonghui, W. Minghuan, L. Shujun, Z. Jinghui, Z. Jianan, H. Runping, X.Qun, Adsorption of methylene blue by a high-efficiency adsorbent (polydopaminemicrospheres): kinetics, isotherm, thermodynamics and mechanism analysis,Chem. Eng. J. 259 (2015) 5361.

[46] B. Delmon, J.T. Yates, Studies in Surface Science and Catalysis, Elsevier Publications,2001. 271275.

[47] K. Bourikas, M. Stylidi, D.I. Kondarides, X.E. Verykios, Adsorption of acid orange 7 onthe surface of titanium dioxide, Langmuir 21 (2005) 92229230.

[48] Z.E. Allouni, M.R. Cimpan, P.J. Hl, T. Skodvin, N.R. Gjerdet, Agglomeration andsedimentation of TiO2 nanoparticles in cell culture medium, Colloids Surf. B:Biointerfaces 68 (2009) 8387.

[49] P.K. Dutta, A.K. Ray, V.K. Sharma, F.J. Millero, Adsorption of arsenate and arsenite ontitanium dioxide suspensions, J. Colloid Interface Sci. 278 (2004) 270275.

[50] P. Fernndez-Ibez, J. Blanco, S. Malato, F.J. d.l. Nieves, Application of the colloidalstability of TiO2 particles for recovery and reuse in solar photocatalysis, Water Res.37 (2003) 31803188.

[51] O.V. Ovchinnikov, S.V. Chernykh, M.S. Smirnov, D.V. Alpatova, R.P. Vorob'eva, A.N.Latyshev, A.B. Evlev, A.N. Utekhin, A.N. Lukin, Analysis of interaction the organicdye methylene blue and the surface of AgCl(I) microcrystals, J. Appl. Spectrosc. 74(2007) 809816.

[52] F. Zhao, E. Repo, M. Sillanp, Y. Meng, D. Yin,W.Z. Tang, Green synthesis of magnet-ic EDTA- and/or DTPA-cross-linked chitosan adsorbents for highly efficient removalof metals, Ind. Eng. Chem. Res. (2014)http://dx.doi.org/10.1021/ie503874x.

[53] X. Cheng, X. Yu, Z. Xing, Characterization and mechanism analysis of N doped TiO2with visible light response and its enhanced visible activity, Appl. Surf. Sci. 258(2012) 32443248.

[54] Y. Yunjin, X. Feifei, C. Ming, X. Zhongxiao, Z. Zhiwen, Adsorption behavior of meth-ylene blue on carbon nanotubes, Bioresour. Technol. 101 (2010) 30403046.

[55] S. Wang, Y. Boyjoo, A. Choueib, A comparative study of dye removal using fly ashtreated by different methods, Chemosphere 60 (2005) 14011407.

[56] X. Lin, Y. Ye, M. Jiaxing, S. Weiling, Z. Chaoying, W. Dapeng, C. Qing, N. Jinren, Ad-sorption behavior of methylene blue onto titanate nanotubes, Chem. Eng. J. 156(2010) 313320.

http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0005http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0005http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0005http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0010http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0010http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0010http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0015http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0015http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0015http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0015http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0015http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0020http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0020http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0025http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0025http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0025http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0030http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0030http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0030http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0030http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0035http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0035http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0035http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0040http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0040http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0045http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0045http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0045http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0045http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0050http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0050http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0055http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0055http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0055http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0060http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0060http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0060http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0065http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0065http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0065http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0070http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0070http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0075http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0075http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0075http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0080http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0080http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0080http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0085http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0085http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0085http://dx.doi.org/10.1007/s11356-013-1546-8http://dx.doi.org/10.1007/s11356-013-1546-8http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0095http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0095http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0095http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0100http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0100http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0100http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0105http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0105http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0105http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0110http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0110http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0110http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0110http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0115http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0115http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0115http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0120http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0120http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0120http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0120http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0125http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0125http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0125http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0125http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0130http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0130http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0130http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0135http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0135http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0135http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0135http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0140http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0140http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0140http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0145http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0145http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0150http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0150http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0155http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0155http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0155http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0155http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0155http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0155http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0155http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0160http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0160http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0160http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0165http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0165http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0170http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0170http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0170http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0170http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0175http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0175http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0180http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0180http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0185http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0185http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0185http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0190http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0190http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0190http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0195http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0195http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0195http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0195http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0200http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0200http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0205http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0205http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0205http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0210http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0210http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0215http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0215http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0220http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0220http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0220http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0220http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0225http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0225http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0230http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0230http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0230http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0235http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0235http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0235http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0235http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0240http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0240http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0245http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0245http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0245http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0245http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0250http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0250http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0250http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0250http://dx.doi.org/10.1021/ie503874xhttp://refhub.elsevier.com/S0167-7322(15)00170-1/rf0260http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0260http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0260http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0265http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0265http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0270http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0270http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0275http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0275http://refhub.elsevier.com/S0167-7322(15)00170-1/rf0275

A comparative study for the removal of methylene blue dye by N and S modified TiO2 adsorbents1. Introduction2. Experimental2.1. Materials and methods2.2. Adsorbent synthesis2.3. Equilibrium experiments2.4. Kinetic experiments2.5. Effect of solution pH

3. Results and discussion3.1. XRD measurement3.2. SEM and EDAX analysis3.3. Equilibrium studies3.4. Kinetic studies3.5. Effect of pH

4. ConclusionsAcknowledgmentReferences