Journal of Environmental Chemical Engineering 2 (2014) 1495–1505

Investigating role of sulphur specific carbon adsorbents in deepdesulphurization

Sandip V. Patil a, Laxmi Gayatri Sorokhaibam a, Vinay M. Bhandari a,*, D.J. Killedar b,Vivek V. Ranade a

aChemical Engineering & Process Development Division, CSIR-National Chemical Laboratory, Pune 411 008, Indiab Shri G.S. Institute of Technology and Science, Indore 452003, India

A R T I C L E I N F O

Article history:Received 10 April 2014Accepted 9 July 2014

Keywords:ThiopheneDesulphurizationAdsorptionFuelPollution

A B S T R A C T

Adsorptive desulphurization of model fuel (thiophene in isooctane) was studied by using sulphur specificcarbon based adsorbents, SHIRASAGI GH2x 4/6 and SRCx 4/6, in an attempt to obtain insight into theadsorptive behaviour of sulphur moiety on the modified surfaces. Both adsorption equilibria andadsorption kinetics have been reported. Characterization of the modified adsorbents was carried outusing scanning electron microscopy, energy dispersive X-ray spectroscopy, X-ray diffraction (XRD) andsurface area analyser. Adsorption experiments along with characterization results highlight potential forhigher sulphur adsorption on the modified adsorbents. Substantially higher capacities for adsorption ofthiophene were obtained, �5.5 and �20 mg g�1 of sulphur for SHIRASAGI GH2x 4/6 and SRCx 4/6respectively. Experimental investigations reveal possible influence of Al/Si ratio, 12.46 and 9.35 in GH2xand SRCx respectively that can have implication in adsorption behaviour. It was found that surfacemodification plays important role in enhancing sulphur removal capacity as compared to conventionalcarbon adsorbents. A plausible mechanism to account for surface interactions and role of surfacemodification has been proposed. The higher mesopore volume along with higher oxygen content isbelieved to influence preferential adsorption of thiophenic sulphur by SRCx. The work also indicatedspecific role of Al and Si content in the carbon matrix. The utility of these adsorbents and higher sulphurremoval capacity has also been confirmed using synthetic mixture of refractory sulphur compounds suchas benzothiophene, dibenzothiophene along with thiophene.

ã 2014 Elsevier Ltd. All rights reserved.

Contents lists available at ScienceDirect

Journal of Environmental Chemical Engineering

journal homepage: www.else vie r .com/ locat e/ jece

Introduction

The ever increasing stringent environmental regulationsrequire sulphur content in transportation fuels to be reduceddrastically. Whereas the earlier environmental norms allowedsulphur content in diesel and gasoline to be 500 and 300 ppmrespectively, the existing norms require this to be brought downnearly 10 fold i.e. 15 and 30 ppm respectively [1–3]. Gasoline, anattractive option for application in fuel cells, also has limitationsdue to stringent requirement of sulphur content well below 1 ppmto avoid poisoning of catalyst. In such cases, adsorptive desulphu-rization can be one of the most promising processes for deepdesulphurization to meet the challenge. Understanding surfaceinteractions and surface modifications can substantially aiddevelopment in this regard.

* Corresponding author: Tel.: +91 020 25902171; fax: +91 2025893041.E-mail addresses: laxmigayatri1@gmail.com (L.G. Sorokhaibam),

vm.bhandari@ncl.res.in (V.M. Bhandari).

http://dx.doi.org/10.1016/j.jece.2014.07.0092213-3437/ã 2014 Elsevier Ltd. All rights reserved.

The removal of sulphur compounds during petroleum refiningoperations to obtain different fuel fractions is not straightforwardand becomes increasingly more and more difficult with lowering ofsulphur concentrations. The three important fuel fractions namelygasoline, diesel and jet fuel contain variety of sulphur compounds.Generally, gasoline contains mercaptans, sulfides, disulfides,thiophene, alkylated derivatives of thiophene and benzothiophenewhile the sulphur compounds in the diesel comprise mainly ofalkylated benzothiophenes, dibenzothiophene and its alkylatedderivatives. Conventionally, hydrodesulphurization (HDS) [4] isused to bring down sulphur concentration in liquid fuels. However,with existing stringent norms for drastically lower sulphurconcentrations, HDS processes have severe limitations in bringingdown sulphur concentrations to desired levels both technically andeconomically. Hydrodesulphurization process requires suitablecatalyst for removal of sulphur compounds and operates at hightemperatures of the order of 450 �C and high pressures of the orderof 20–40 atm. The removal of these sulphur compounds, includingrefractory sulphur compounds by HDS process to meet the desiredlevels would demand more than 3 fold increase in the catalyst

Nomenclature

aL Langmuir theoretical adsorption capacity, mg g�1

AC activated carbonADS adsorptive desulphurizationbL Langmuir adsorption constant, L mg�1

BDS biodesulphurizationBT benzothiopheneCe liquid phase concentration at equilibrium, mg L�1

C0 liquid phase initial concentration, mg g�1

Cs Solid phase concentration of sulphur at equilibrium,mg L�1

DBT dibenzothiopheneGH2x SHIRASAGI GH2x 4/6 adsorbentHDS hydrodesulphurizationKF Freundlich adsorption coefficient,

ðmg�g�1ÞðL�mg�1Þ1=nF

nF Freundlich constantqe solid phase concentration at equilibrium, mg g�1

qe,cal calculated equilibrium adsorption capacityqe,exp experimental equilibrium adsorption capacityRL Langmuir dimensionless constant separation factorSRCx SHIRASAGI SRCx 4/6 adsorbentT thiophene

1496 S.V. Patil et al. / Journal of Environmental Chemical Engineering 2 (2014) 1495–1505

volume/reactor size and as a consequence, rise in cost of operationof this high temperature and high pressure process. HDS methodmay also lead to decrease in fuel efficiency due to decrease inoctane number [5] of the fuel resulting from saturation of aromaticor olefinic groups in meeting the new stringent specifications. It istherefore instructive to evaluate alternate solutions to thisproblem and possible alternatives include adsorptive desulphu-rization [6], biodesulphurization [7–9] and oxidative desulphuri-zation [10–12].

The need for practically sulphur free fuels like gasoline neededfor fuel cell application has necessitated the scope not just fordevelopment of newer methods but for more techno-economicallyfeasible options like adsorption process that operate at ambienttemperature and pressure conditions. Deep desulphurizationconsiderations have provided impetus for research in differenttechnological platforms, and adsorptive desulphurization seems tobe a promising one. Adsorptive separations also have limitations insulphur removal, mainly due to low capacity for sulphur removal,difficulty in removing refractory sulphur compounds and poorunderstanding of surface interactions that play crucial role insulphur removal. This in turn requires improved understanding ofadsorption behaviour of various sulphur compounds on differentadsorbent surfaces, modification of surfaces, extent and efficacy ofnewer materials/modifications in deep desulphurization.

Adsorptive desulphurization (ADS) can remove sulphur com-pounds through physisorption, chemisorption or p complexation.Adsorption process using porous forms of activated carbons [13]and modified adsorbents can be an excellent technique that mayintegrate with the existing HDS process to meet the threshold limitof sulphur content in diesel and other fuels. The advantages ofconventional adsorption process and issues with other processesmake it imperative to study production of ultra-low sulphur fuelsusing adsorption and newer adsorbents. Apart from this, ADSmethod does not impart impurity and fuel composition remainsessentially unchanged. The typical adsorption process is expectedto offer selective removal of sulphur compounds under ambientconditions with ease of operation/process control and economi-cally achieve removal of sulphur compounds from the

transportation fuels with greater efficiency. The ease of regenera-tion with minimum requirement of chemical or energy is crucialfor satisfactory implementation of this technology. A number ofsorbents starting from simple activated carbons [14] to p-com-plexation adsorbents have been reported in the literature withvarying degree of success, though unsatisfactory in most cases.Sorbents impregnated with transition metals like Ni [15], Fe [16],Cu [17], Zn, Mn [18], Pd, Vn [19] and Ce have also been reported. Ionexchanged zeolites have been found to be promising adsorbentswith sulphur removal capacity of 42 mg g�1 using model fuel ofbenzothiophene, dibenzothiophene and 4,6-dimethyl dibenzo-thiophene [2]. Hernández-Maldonado and Yang [20] showed thatCu-Y and Ag-Y zeolites have good capacity for thiophene sulphurremoval from benzene and n-octane mixtures. Researchers haveworked with solid adsorbents like activated carbon, Ag-loadednanofibrous membranes [21], zeolites, polymeric adsorbents etc.for several separation/purification applications including industri-al applications due to their high surface area and good adsorptioncapacity. In this regard, though a large number of adsorbents withdifferent matrix, porous structure and adsorbent capacity areavailable for commercial use, the selection for any particularapplication is not an easy process and not straightforward. It iswhere the role of surface functional groups has to be investigatedwherein certain functional groups affect the interaction withspecific or polar adsorbate. Sulphur is known to be slightly polar innature and this aspect of sulphur may be exploited for successfuldesulphurization of transportation fuels like diesel. This is alsoexpected to provide insight into effective regeneration strategythat is difficult in most cases as of today. The selection of thepresent adsorbents, GH2x and SRCx for desulphurization applica-tion was envisaged in the light of polarity interactions for obtainingimprovement in desulphurization performance.

Experimental

Materials, reagents and methods

Thiophene (99.9% purity, Loba Chemicals), isooctane (99.9%purity, Loba Chemicals), benzothiophene (95%, Fluka) and diben-zothiophene (98%, Sigma–Aldrich) were used without furtherpurification. Modified carbon based adsorbents SHIRASAGI GH2x4/6 and SHIRASAGI SRCx 4/6 procured from Japan EnviroChemicalsLtd. were used for liquid phase adsorption of thiophene from iso-octane and also for synthetic fuel mixture. Pelletized Noritactivated carbon procured from Aldrich was also used in theexperimental investigation. The samples were analysed forsulphur removal using total sulphur analyser, TN-TS 3000(Thermoelectron Corporation, Netherlands) and also with GasChromatograph (Agilent GC 7980) equipped with CPSil 5CB forsulphur as column (30 m � 320 � 4 mm) in conjunction with flamephotometric detector. Helium was used as a carrier gas with flowrate of 1.25 mL min�1 and split ratio of 10:1. The injectortemperature employed was 250 �C with injection volume of 1 mLand total analysis time of 40 min. The oven temperature wasramped at 5 �C min�1 from 40 to 100 �C and 40 �C min�1 to 230 �C.Reproducibility of the experimental results was checked and wasfound satisfactory.

Adsorption studies

Equilibrium adsorption studies were carried out at 30 �C usingstandard procedures. Known volume of solution – predeterminedconcentration was equilibrated with known weight of theadsorbent in the range of 0.05–1 g per 10 mL of model fuel (initialconcentration of 442.5 mg L�1 of total sulphur using thiophene assulphur containing component in iso-octane). Equilibration time of

S.V. Patil et al. / Journal of Environmental Chemical Engineering 2 (2014) 1495–1505 1497

minimum 8 h to 24 h was used with Spectralab HM8T orbitalshaker. The adsorption studies using GH2x and SRCx were furtherinvestigated in model fuel comprising of mixture of thiophene,benzothiophene and dibenzothiophene where total sulphurcontribution from each component was 50, 100 and 150 mg L�1

respectively under similar set of experimental conditions. Kineticstudies were carried out using model fuel of known concentration,adding known weight of adsorbent, stirring speed of 120 rpm andwithdrawing samples at regular intervals of time.

Adsorbent characterization

Textural characterizations of the carbonaceous adsorbents wereconducted by scanning electron microscope, SEM (Leo-Leica,Stereoscan 440, Cambridge, UK) attached with Energy dispersiveX-ray spectroscopy, EDX (Bruker, Quanrax-200, Berlin, Germany),Pan Analytical XRD in the scan range of 2u between 2–90� for GH2xand 2–60� for SRCx in continuous mode using Cu Ka source andQuantachrome Autosorb Automated Gas Sorption system for N2

adsorption and desorption isotherm. Specific surface area wascalculated by applying Brunaer–Emmett–Teller (BET) methodwhile BJH (Barrett–Joyner–Halenda) equation was applied forpore size/pore distribution and Dubunin–Radushkevich methodfor micropore volume determination. Surface functional groups onthe modified carbon adsorbents were determined by Cary 600 FT-IR (Agilent) spectrophotometer having 4 cm�1 resolution using KBrpellet method in the range 400–4000 cm�1.

Results and discussion

Selection of adsorbents and model fuel

In the present study, adsorption behaviour of two diverseforms of chemically modified activated carbons, namely SHIR-ASAGI GH2x 4/6 and SHIRASAGI SRCx 4/6 was studied by varyingadsorbent loading in the range of 0.5–5% for thiophene in iso-octane as model liquid fuel. These two modified carbonaceousadsorbents are known to exhibit affinity towards sulphurcompounds. SHIRASAGI GH2x 4/6 (hereafter referred as GH2x)is pelletized form of activated carbon reportedly prepared underhigh temperature steam and impregnated with specialitychemicals to provide characteristics functions for application inremoval of hydrogen sulfide, sulphur compound and fatty acidsfrom air. Similarly, SHIRASAGI SRCx 4/6 (hereafter referred asSRCx), also reportedly prepared in identical manner withcharacteristic high adsorption performance for sulphur com-pounds like sulfides and mercaptans. However, the ability of thesemodified adsorbents for removal of thiophene, a commonorganosulphur compound and higher thiophenic derivatives intransportation fuel has not been reported so far. These adsorbentsare expected to offer selectivity for adsorption of sulphurcompounds compared to more abundant aromatic compoundsthat are present in real transportation fuels with order ofmagnitude higher concentrations. Thus, the role of surfacemodification of carbon based adsorbents can be evaluated moreeffectively with such adsorbents and using simple, yet represen-tative, compound such as thiophene as model compound.

Table 1Physical and chemical properties of modified adsorbents.

Adsorbent Appearance Sp.

(g m

GH2x 4/6 Black cylindrical pellet 2.0�SRCx 4/6 Black granule 1.6�

The crucial factor in thiophene adsorption is difference inpolarity between thiophene and other hydrocarbon componentspresent in liquid fuels. Model fuel of thiophene in isooctane wasselected as presence of alkylated groups which are known electrondonors may provide steric hindrance thereby blocking access ofthiophenic sulphur to surface active groups on the adsorbentsurfaces [22]. Thus, selection of thiophene facilitates understand-ing the actual capacity and nature of interaction in absence ofinterfering groups. Further, to validate the applicability of thepresent adsorbents for refractory compounds, adsorption usingsynthetic fuel mixture consisting of thiophene (T), benzothiophene(BT) and dibenzothiophene (DBT) was also explored. The propor-tion of the mixture 50:100:150 of T:BT:DBT was selected in view ofhigher initial concentration of the more refractory fraction in thefuel composition. Prior to adsorption studies, the specificadsorbents were dried at 200 �C for 12 h.

It is expected that such investigation can provide insight intothe selective removal of sulphur compounds from liquid fuels. Thepresent study also comprises of various characterization studies,comparative and parametric adsorption performance of GH2x andSRCx. The adsorption behaviour of commercial activated carbon,Norit [10] was also studied to give a comparative overview of theperformance of modified adsorbents over commercial activatedcarbon. In an effort to understand the adsorption behaviour, factorslike temperature effect and diffusion were considered in evaluat-ing the adsorption data. The present investigation is expected toprovide conceptual understanding of studied adsorbent–adsorbatemodification and surface interaction with sulphur. This study isalso likely to provide insights into developing better and neweradsorbents with abilities for capture of specific sulphur moieties.

Characterization of adsorbents

The different physico-chemical properties of GH2x and SRCxare given in Table 1. Both GH2x and SRCx have superior surface areaand pore volume as summarized in Table 2. It is seen from Table 2that SRCx has a narrower pore volume of 0.710 cm3 g�1 comparedto 0.807 cm3 g�1 of GH2x. The observed higher sulphur removalcapacity of SRCx in single solute thiophene model fuel may also beattributed to narrower pore size distribution. The nitrogenadsorption isotherms at 77 K (Fig. 1a, b) indicated high adsorptionat low relative pressures below 0.3 in both GH2x and SRCx showingcontribution from the micropores towards adsorption process.GH2x has very high BET surface area of 1329 m2g�1 while SRCx hasa relatively low surface area of 1098 m2g�1 which may be due tothe partial reduction in the porous surface during modification.This is complemented by the fact that the micropore volume is alsoslightly reduced in SRCx (Table 2). The N2 adsorption–desorptionisotherm of GH2x and SRCx depicts typical type I isotherm withsteep initial region showing possible adsorption in the micropores.The correlation between high surface areas to that of higheradsorption capacity was not seen in case of adsorption of modelfuel (thiophene in isooctane) by GH2x. The specific surface areaand micropore volume of SRCx is lesser than GH2x which may bedue to lesser extent of impregnation/modifying chemical concen-tration as stated earlier. The effect is visible in the nature ofadsorption capacity of the two adsorbents being used. It may be

GravityL�1)

Bulk density(g mL�1)

Ignition point

2.2 0.40�0.45 490 �C1.8 0.15�0.19 300 �C

Table 2Surface area and porosity of the modified adsorbents.

Adsorbent BETsurface area(m2g�1)

Total pore volume(cm3g�1)

Micropore volume(cm3g�1)

Microporesurface area(m2g�1)

Averagepore diameter(nm)

Adsorption energy(kJ mol�1)

GH2x 4/6 1329 0.807 0.677 1901 2.43 14.50SRCx 4/6 1098 0.710 0.552 1551 2.58 14.98

Fig. 1. Nitrogen adsorption–desorption isotherm plots; a) SHIRASAGI GH2x 4/6 andb) SHIRASAGI SRCx 4/6.

1498 S.V. Patil et al. / Journal of Environmental Chemical Engineering 2 (2014) 1495–1505

inferred that certain micropore sites were blocked to greaterextent in SRCx and selective modification was ensured leading tohigher adsorption by SRCx. The nature of the adsorbents underinvestigation is mainly meso-microporous type. However, there is

Fig. 2. BJH pore size distribution curve; a) SHIRASAGI GH2x 4/6 and b) SHIRASAGISRCx 4/6.

greater concentration of micropores with radii below 2 nm whichis visible in BJH analysis of the pore size distribution in Fig. 2(a) and(b).

The type of the treatment or modification affected the surfacetexture of the adsorbent as depicted by the SEM images (Fig. 3a, b)of GH2x and SRCx showing heterogeneous nature. The surface ofSRCx as revealed by SEM is a combination of smooth and irregularportion with cracks on the surface. EDX spectra of GH2x and SRCxare clubbed along with their respective SEM images (Fig. 3). Asseen from Table 3, the percent composition of oxygen and Si isslightly more in case of SRCx attributing some role in the higheradsorption capacity as observed in adsorption equilibrium data ofmodel fuel – thiophene adsorption.

The surface functional groups of the carbon based adsorbentswere investigated by FTIR. The FTIR spectra of GH2x and SRCx aregiven in Fig. 4 depicting presence of certain functional groups suchas C��O stretching and ��OH groups. The broad band at3740.38 cm�1 is due the presence of isolated hydroxyl groups inSRCx where as the one at 1187.35 cm�1 is attributed to C��O stretchof amines, while the medium peak at 1498.62 cm�1 is due to CQCstretching. A common IR absorption band at 2365.43 cm�1 isobserved in both GH2x and SRCx which is peculiar of CO2 in air[23]. The usual OH stretching band at 3416.74 cm�1 is also found inGH2x but it is associated with two more closed peaks which maybe assigned due to the co-existence of isolated hydroxyl groups aswell as associated OH groups. IR absorption band at1176.01 cm�1 inGH2x is assigned to C��O bond while methyl groups areresponsible for IR absorption at 1487.28 cm�1. The IR peak dueto the vibration of Si–Si near or on the surface of SiO2 is moreprominent in GH2x as seen from weak peak at 621.50 cm�1[24,25].The presence of these functional groups shows the effect of surfacemodification which is believed to have contribution in theadsorption process.

The wide angle XRD patterns (Fig. 5) of GH2x and SRCx showedtwo broad bands around 2u = 24 and 43� indicating amorphousstructure of the carbonaceous adsorbent matrix [26]. The absenceof other prominent phases is indicated by the absence of sharppeaks corresponding to oxides/minerals phases of Si and Al, whichare present as trace amounts as indicated in EDX data. It alsoimplies that Al and Si present in the modified adsorbent are welldispersed in the carbon matrix. Diffraction peaks belonging to SiO2

at 20.8 and 26.7 (JCPDs 46-1045) are diffuse in the broad peaks ofactivated carbon. The lower content and more dispersed nature ofSi and Al is responsible for the absence of prominent peaksassociated with these metals. It is also observed that the intensityof the peak is slightly lesser in SRCx than in GH2x. This can beattributed to the relatively lower content of Al in SRCx than inGH2x. This is in agreement with SEM image of SRCx showingmixture of smooth surface and aggregate particles which is lesserin SRCx than in GH2x.

Equilibrium adsorption experiments

The study of adsorption isotherm is essential in designing ofadsorption systems and defines maximum capacity for theremoval. The results of equilibrium studies on the two different

Fig. 3. SEM-EDX image of a) GH2x and b) SRCx.

S.V. Patil et al. / Journal of Environmental Chemical Engineering 2 (2014) 1495–1505 1499

adsorbents are given in Fig. 6. The adsorbents have high capacity ofsorption for sulphur compound and SRCx exhibits much highercapacity than GH2x. The high capacity for sorption in theseadsorbents and the substantial variation in the capacity for the twoadsorbents or such adsorbents needs to be explained andinterpreted theoretically. The extent of adsorption and equilibriumrelations can be represented by using two most commonly used

Table 3EDX report for modified adsorbents.

Elements GH2x 4/6 SRCx 4/6

% C 70.26 68.32% Al 3.74 3.18% O 25.68 28.16% Si 0.30 0.34

adsorption isotherms – Langmuir and Freundlich adsorptionisotherm. The linear regression method was used for determiningthe best fit adsorption isotherm and isotherm parameters usingcoefficient of determination, R2-isotherm giving R2 value closest tounity representing best fit isotherm. Adsorption isotherms (Fig. 6)indicate potential for still higher capacity realizations forthiophene removal by GH2x and SRCx. The linearized mathemati-cal expression of Langmuir [27] and Freundlich adsorption [28]isotherms are given below:

Langmuir adsorption isothermCe

qe¼ 1

aLbLþ Ce

aL(1)

where Ce is the equilibrium sulphur concentration in mg L�1, aL ismonolayer adsorption capacity in mg g�1 and bL (L g�1) is the

Fig. 4. FTIR spectra of a) GH2x and b) SRCx.

Fig. 5. X-ray diffractogram of a) GH2x and b) SRCx.

1500 S.V. Patil et al. / Journal of Environmental Chemical Engineering 2 (2014) 1495–1505

Langmuir adsorption constant. The Langmuir constant, aL corre-sponds to the theoretical maximum adsorption capacity.

The favourable nature of adsorption can be indicated fromdimensionless factor RL of Langmuir adsorption which is defined as

RL ¼ 1ð1 þ bLC0Þ (2)

Favourable adsorption is indicated if RL is between 1 and 0 [29],unfavourable if greater than 1, while a value of 1 and 0 representsunfavourable and irreversible isotherms, respectively.

For adsorption using GH2x, the RL value was evaluated to 0.355showing favourable adsorption.

Freundlich adsorption isotherm

log qe ¼ log KF þ 1nFlog Ce (3)

where KFðmg:g�1ÞðL:mg�1Þ1=nF is the Freundlich adsorption coeffi-cient and nF being the Freundlich constant. The Freundlichisotherm takes into account the heterogeneity factor. 1/nF is afunction of the strength of adsorption and it indicates the affinitybetween the adsorbate and the adsorbent.

In the present study, interpretation of the experimental datawas done using Langmuir and Freundlich adsorption equations andthe corresponding constants along with the coefficient ofdetermination values are given in Table 4. SRCx exhibited excellentsulphur removal capacity of �20 mg g�1 of the adsorbent whileGH2x gave a maximum capacity of �5.5 mg g�1 of the adsorbentused. The key factor affecting the adsorption behaviour may be thedifference in ratio of Al and Si impregnation over activated carbon.The presence of Si group in the matrix of activated modified carbonhas its own significance as there have been reports of thiopheneadsorbed on SiO2 [30]. The characterization data of EDX showedhigher Al/Si ratio of 12.46 in GH2x and lower Al/Si ratio of 9.35 inSRCx. It can be inferred that there is role of higher Si contentcontributing to higher adsorption by SRCx. Other factors includeslightly smaller pore size of GH2x that may be a governing factor inlower adsorption capacity of GH2x compared to SRCx. This isconsistent with the fact that adsorption is dependent on a numberof factors, mainly the porosity and surface functional groups incase of selective/specific adsorbents. The applicability or the fittingof the isotherm is compared by judging coefficient of determina-tion, R2 value. The better fit of sulphur removal by SRCx towardsFreundlich isotherm implied that cooperative adsorption ispromoted by the modified carbon. On the other hand, Langmuirisotherm exhibited a better fit for adsorption on GH2x asillustrated by R2 value in Table 4. A deeper analysis showed thatboth Langmuir and Freundlich isotherm showed near applicabilityfor adsorption by GH2x (R2 value of 0.834 and 0.817 respectively),where SRCx showed very poor applicability for Langmuir type ofadsorption model (Table 4). It indicates that there is distinctdifference in the mechanism of adsorption of SRCx and GH2x. Thedifference in these carbon based adsorbents can be explained onthe basis of the surface composition and characteristic analysis ofthe adsorbent material.

A comparison of adsorption isotherm of the modified adsorb-ents with commercial Norit A.C. having BET surface area of905.9 m2g�1 and total pore volume of 0.699 cm3 g�1 is given inFig. 6 which indicates substantially lower adsorption by Norit AC(�2 mg�1) as compared to the modified carbons, GH2x and SRCx. It,therefore, substantiates greater potential of these specificallytailored adsorbents towards removal of sulphur. Norit AC is mainlycomposed of C and small fractions of Au, O and Al followed bytraces of Ca, Na and Mg in descending order. Due to the lowercapacity of commercial activated carbons towards removal of

Fig. 6. Experimental adsorption isotherm for thiophene removal a) GH2x, b) SRCxand c) Norit AC at 30 �C.

S.V. Patil et al. / Journal of Environmental Chemical Engineering 2 (2014) 1495–1505 1501

sulphur, there is limited information on adsorption capacity ofactivated carbon in single solute adsorption of thiophene in iso-octane as model fuel. Antonio Chica [31] reported the adsorptioncapacity of 1.7 and 2.8 thiophene/Al on zeolites like on H-ZSM5 andH-Y zeolites. Deep desulphurization on zeolites as reported by Kinget al. [32] indicated adsorption capacity of H-ZSM5 as 0.13thiophene/Al. It is suggested that with specific modification, thecapacity of commercial activated carbon may also be enhanced.Though for most carbon based adsorbents, the capacity for sulphurremoval is not high, some recent results on modified carbons showhigher capacities of the order of 20–40 mg S g�1 of adsorbentthrough chemical modification such as oxidation and probably due

Table 4Adsorption isotherm constants.

Adsorbent Langmuir constants Freundlich constants

aLa bL

b R2 KFc 1/nFd R2

GH2x 4/6 8.432 4.278 � 10�3 0.834 0.273 0.5 0.817SRCx 4/6 7.097 1.696 � 10�3 0.503 1.746 � 10�4 1.88 0.763

a Langmuir theoretical adsorption capacity, mg g�1.b Langmuir adsorption constant, L mg�1.c Freundlich adsorption coefficient, ðmg�g�1ÞðL�mg�1Þ1=nF .d Freundlich constant.

to exceptionally high surface area such as BY-1 A.C. [33],commercial granulated A.C. [34], commercial coconut A.C. [35]and polystyrene based activated carbon spheres [36].

Effect of temperature and adsorption thermodynamics

The effect of temperature on thiophene adsorption by GH2x andSRCx was studied using three different temperatures viz., 303, 313and 323 K. In each case the adsorbent loading was varied from 0.1to 1.0 g in 20 mL volume of the model fuel solution havingconcentration of 450 mg L�1. It was observed that lower tempera-ture was favourable for removal of sulphur using GH2x, while SRCxindicated favourable adsorption at slightly higher temperature inthe range of temperature studied. This fact is also supported by thethermodynamic studies on the two systems as given by van’t Hoff’splot (Fig. 7a, b) where positive slope indicated exothermicadsorption and negative slope indicated endothermic adsorptionfor GH2x and SRCx respectively. The contrasting thermal behaviourfor the two adsorbents indicated structural changes and differencein nature of interaction for the two adsorbents. The endothermicnature may be due to the low activation energy for adsorption bySRCx. It may be noted that adsorption may be either exothermic orendothermic depending on the system under consideration. Thenature of adsorption i.e. adsorption thermodynamics also dependon the activation temperature, raw material used etc. Additionally,for physisorption, weak van der Waals forces are involved and isfavourable at low temperature or exothermic which is the case ofGH2x whereas for higher temperature adsorption, endothermicbehavior is predominant as observed in the case of SRCx. Thethermodynamic parameters viz., DH�, DS�, DG� were determinedfrom Fig. 7 and are given in Table 5. They are estimated using therelations below:

ln KC ¼ DH�

R�DS�

RT(4)

where Kc, is the ratio of solid phase concentration at equilibrium tothat in the solution in mg L�1 while the values of DH� and DS� aredetermined from Fig. 7 from the intercept and slope respectively.As indicated earlier, the values of enthalpy change, DH�, werenegative for adsorption on GH2x and positive for SRCx, suggestingexothermic and endothermic nature respectively. The Gibb’s freeenergy change, DG� was determined using Eq. (5).

DG� ¼ DH� � TDS� (5)

Fig. 7. Van’t Hoff plot for sulphur removal a) GH2x and b) SRCx.

Table 5Thermodynamic parameters for thiophene adsorption.

Adsorbent DH�

(kJ mol�1)DS�

(J mol�1 K�1)DG�

(kJ mol�1)R2

303 K 313 K 323 K

GH2x �169.07 �522.65 �10.71 �5.48 �0.25 0.971SRCx 55.38 193.79 �3.33 �5.27 �7.21 0.999

1502 S.V. Patil et al. / Journal of Environmental Chemical Engineering 2 (2014) 1495–1505

As seen from Table 5, the values of DG� are negative at all thethree different temperatures for adsorption with GH2x as well asSRCx, thereby indicating spontaneous and thermodynamicallyfavourable adsorption. It is also implied from these values that theadsorption process is more of physisorption in nature as themagnitude of DG� is well below 20 kJ mol�1 in both the cases ofadsorption.

The positive value of entropy change in case of SRCx indicatesthe high affinity of the sulphur moieties towards SRCx. The positiveentropy change is also attributed to the increasing randomness onthe adsorbent–adsorbate interface due to the rapid adsorption asseen from the nature of adsorption equilibrium. It must bementioned that entropy change is calculated for sulphur moleculeswhile getting displaced from the bulk to the adsorbent surface byusing Eq. (5). And for spontaneous endothermic systems havingnegative DG and DH, the DS is bound to give positive valueaccording to the above equation.

Kinetics of adsorption

The kinetics data (Fig. 8) giving change of concentration withtime were used for analysing the kinetics of the adsorption process.Lageren’s pseudo-first order [37,38] and second order kineticmodels [37,38] were investigated using the contact time data andtheir linearized mathematical expressions are respectively givenby Eqs. (6) and (7)

log ðqe � qtÞ ¼ logqe �k1

2:303t (6)

and

tqt

¼ 1k2q2e

þ tqe

(7)

Fig. 8. Thiophene uptake profile at 31 �C, adsorbent dosage = 44 g L�1 of GH2x andSRCx.

where qt is the amount adsorbed at the time t in mg g�1, k1 is 1storder rate constants in min�1 and k2 is the second order rateconstant in g mg�1min�1.

The values of these kinetic constants k1 and k2 can bedetermined from the linear plots of log(qe–qt) versus t forpseudo-1st order, t/qt versus t for second order rate equationand are given in Table 5. The maximum removal of sulphur wasnearly 27.49% with adsorbent dosage of 44 g L�1 of GH2x. With theincrease in surface coverage after some initial period, the rate ofremoval slowed down and remained almost asymptotic with timeafter 210 min.

To understand the mechanism of adsorption process, theIntraparticle diffusion model [39] was applied to the kinetic data;linearized mathematical expression being given by Eq. (8)

qt ¼ kp�t0:5 þ c (8)

where kp is the rate constant of intraparticle diffusion step inmg g�1min1/2 whose value is determined from the slope of the plotbetween qt and t0.5. The intercept c refers to the boundary layereffect whose value is proportional to boundary layer thicknesswhere thicker is the boundary layer, slower is the rate oftransportation. The study of adsorption diffusion models isnecessary to determine whether adsorption kinetics is controlledby film diffusion or intraparticle diffusion, the slowest step beingthe rate limiting step. Also, the fact that pseudo-first order andsecond order kinetic models cannot explain the diffusion mecha-nism between the solid adsorbent and adsorbate necessitates thestudy of diffusion models. Adsorption on liquid–solid interface isoften controlled by various steps like bulk diffusion, film diffusionand intraparticle diffusion. In physisorption, bulk diffusion being avery rapid process, it is not considered as rate determining step. Itmay also be noted that determining the rate controlling step inadsorption requires order of magnitude analysis for various rateprocesses to decide if any particular step can be rate controlling ormore than one steps are needed to be accounted for in describingthe kinetics.

In case of GH2x, the R2 value of pseudo-second order model isvery high (0.994) and moreover the qe value calculated from thismodel is close to experimental value. Hence it may be concludedthat pseudo-second order is the best fit model to explain theobserved kinetics. The regression of qt versus t0.5 does not showgood linearity as elucidated by R2 value from Table 6 and Fig. 9; nordoes it pass through the origin. Thus it can be stated thatIntraparticle diffusion is not the sole rate determining step. Itindicates some degree of boundary layer control in thiopheneadsorption by GH2x. The Weber–Morris plot (Fig. 9) shows theuptake – square root of time profile of SRCx. From Table 6, it is also

Table 6Kinetic parameters for sulphur removal by GH2x and SRCx (adsorbent dose = 44g L�1 at 31 �C).

Pseudo first-order model qe,exp.(mg g�1)

k1(min�1)

qe,cal.(mg g�1)

R2

GH2x 4/6 5.475 0.020 2.205 0.883SRCx 4/6 20.174 0.246 18.633 0.897

Pseudo second-order model qe,exp.(mg g�1)

k2(g mg�1min�1)

qe,cal.(mg g�1)

R2

GH2x 4/6 5.475 0.012 3.067 0.994SRCx 4/6 20.174 0.086 11.507 0.937

Intraparticle diffusion model kp(mg g min0.5)

c(mg g�1)

R2

GH2x 4/6 0.155 0.596 0.843SRCx 4/6 0.480 �0.028 0.970

Fig. 9. Weber–Morris plot for intraparticle diffusion kinetics of thiophene ontoGH2x and SRCx, C0 = 441.38 mg L�1.

S.V. Patil et al. / Journal of Environmental Chemical Engineering 2 (2014) 1495–1505 1503

evident that SRCx showed relatively higher rate constant thanGH2x proving greater selectivity for thiophene/S.

Mechanism of sulphur removal on carbon adsorbent

Understanding the mechanism of interaction in case of differentadsorbents with the organosulphur compounds is important andcan provide insight into adsorbent material, material modificationand consequently leading to tailoring of new materials. There islimited amount of research on mechanism of adsorptive desul-phurization. Thus, an effort to explain the mechanism will alsohelp in industrial application of adsorption technology. Two mainfactors responsible in explaining the surface interaction are poresize/pore structure and presence of doped metal ions/oxides –

extent to which they are able to enhance certain active sitesresponsible for adsorption on the activated carbon surface.

Role of pore size/pore structure and surface chemistry in adsorption

Porosity and surface chemistry of the carbon surfaces are twomain features affecting removal of sulphur, especially for the

Fig. 10. A typical chromatogram

refractory sulphur compounds. Table 2 indicates that the higheraverage pore diameter in SRCx can have a crucial role in adsorption.A good correlation can be drawn between the average porediameter and adsorption of sulphur moieties in thiophene onmodified carbon surfaces. The higher adsorption capacity andhigher average pore diameter in SRCx implied that wider poresfacilitated adsorption of sulphur moieties. As stated earlier, it isevident from Fig. 2 that the surface of GH2x and SRCx are mainlycomposed of micro- and mesopores with majority of the poreslying in the micropore region. It can be deduced from Table 2 thatthe percentage of pore contribution from mesopores is 16.11% and22.25% for GH2x and SRCx respectively. This higher volume ofmesopores (2–50 nm) facilitated transport of thiophene leading tohigher adsorption capacity on SRCx. The data indicate preferentialadsorption of thiophenic sulphur on mesoporous sites. Also, it isseen that the adsorption energy is higher for SRCx carbons withlarger pore diameter. The higher adsorption energy is also linked tothe interaction of adsorbate molecules on surface functionalgroups.

In spite of the relatively lower surface area of SRCx, the highcapacity of sulphur removal is quite significant. This can becorrelated with the total average pore size that is higher in SRCx.The pore size has prominent effect in adsorption process. Thelower adsorption energy of 14.50 and 14.98 kJ mol�1 as calculatedfrom D–R method of micropore analysis suggest physical adsorp-tion. The slightly higher % of basic groups containing oxygen(Table 3), 28.16% in SRCx as against 25.68% in GH2x is alsoresponsible for the stronger physical adsorption of SRCx. Orienta-tion of the active pore sites with respect to the sulphur compoundin the model fuel is also an important factor in overall adsorptionuptake.

Degree of impregnation of Al/Si

The sulphur removal efficiency on activated carbon surface isenhanced by the presence of basic groups like oxygen containingfunctional groups as confirmed by FTIR; Al and Si groups whichattribute to certain degree. There can be blocking of pore volume toa certain extent in the process but the overall capacity of removalcan be enhanced. Al and Si content within the carbon matrix couldbe in the form of miniaturized zeolite like aluminosilicate structurerepresenting zeolite form within the carbon matrix. With thispresumption, it can be expected that enhancement of sulphuradsorption is probably due to interaction of sulphur moieties with

of synthetic fuel mixture.

Table 7Adsorption data of component fractions in synthetic fuel mixture.

Adsorbent Component sulphur fraction removal capacity (mg g�1)

Thiophene Benzothiophene Dibenzothiophene

GH2x 3.22 7.86 13.74SRCx 3.27 8.89 13.37

1504 S.V. Patil et al. / Journal of Environmental Chemical Engineering 2 (2014) 1495–1505

ions in the zeolite like structure similar to ion-exchanged zeolites.Thus the adsorption of sulphur compound on such modifiedcarbon materials may involve more than one mechanism forsulphur removal. Therefore understanding of sulphur removalmechanism on such modified carbon/material is important toelucidate role of surface functionalities in enhancing the sulphurremoval capacity. This will also help in tailor made adsorbents forremoval of specific pollutants. In view of low capacities observedfor activated carbon materials, the high capacity for sulphurremoval in the case of modified carbons is not commensurate withpore modifications alone and indicates specific role of Al and Si inmodification. It is therefore postulated that the surface texture ofGH2x and SRCx is composed of metal oxides especially of Al and Siwhich are believed to be well dispersed on the carbon matrix. Themicropore volume was slightly reduced in SRCx as these oxides arefurther well dispersed on activated carbon support.

The understanding of surface modification and interaction ofsulphur moiety can be crucial for the effective regeneration,thermal or chemical, of the adsorbents. This aspect requires furtherdetailed investigations.

Evaluating sulphur removal from synthetic fuel

It would be instructive to substantiate high sulphur removalcapacity observed for thiophene with that using refractory sulphurcompounds as well. For this reason a synthetic model fuel wasprepared using two refractory sulphur compounds benzothio-phene and dibenzothiophene along with thiophene. A proportionof 50 ppm:100 ppm:150 ppm was used for thiophene, benzothio-phene and dibenzothiophene respectively. A typical chromato-gram of the three components is shown in Fig. 10. Batch studies onGH2x and SRCx adsorbents using this model fuel have indicatedpreference for refractory sulphur compounds such as benzothio-phene and dibenzothiophene. The normalized % distribution of thethree sulphur compounds on these adsorbents is shown in Fig. 11.The results on this model fuel are commensurate with the findingson sulphur removal for thiophene of this work. Table 7 gives theamount of individual sulphur removed with the respectiveadsorbents in the synthetic fuel mixture. It is evident from thecomparison of adsorption data of the individual fractions with thatof Fig. 10 that individual capacity remains nearly the same in boththe adsorbents GH2x and SRCx in synthetic fuel composition of50:100:150 of T, BT and DBT respectively. The favourableadsorption of more refractory sulphur compounds, DBT and BThas also been observed in our earlier findings [40]. As seen fromTable 7, lower adsorption of thiophene relative to that of DBT andBT, shows dependence on initial sulphur concentration. This is inagreement with the findings of Zhao et al. [41] where higher initialthiophene concentration indicated higher removal due to hydro-phobic interaction (p–p) between the thiophenic molecules. Thishowever requires further detailed investigation.

Fig. 11. Distribution of sulphur content on adsorbent surfaces in synthetic fuelmixture of T:BT:DBT (50:100:150) mg L�1.

Conclusions

Thiopheneadsorptiononcarbonadsorbents,GH2x andSRCx, thatrepresent modified materials showed adsorption uptake of nearly5.5 and 20 mg g�1 of sulphur respectively. It was found that for GH2xadsorption equilibrium can be explained using Langmuir modelwhile in the case of SRCx, Freundlich isotherm is more appropriate.The kinetics of adsorption indicate that adsorption process is notsolely controlled by intraparticle diffusion model. A number offactors like pore size/pore size distribution, pore volume and surfacechemistry are required to be accounted for describing the adsorptioninteraction mechanism. Functional groups including C��O, N��Hand ��OH group were observed on the modified carbon surfaces andcan play crucial role in the sulphur removal mechanism. Also, thedegree of Al/Si ratio seems to affect the adsorption behaviour. Theenhancement of sulphur adsorption is probably due to interaction ofsulphur moieties with ions in the miniaturized zeolite likealuminosilicate structure similar to ion-exchanged zeolites. Theequilibrium adsorption of thiophene on SHIRASAGI GH2x 4/6 andSRCx 4/6 at different temperatures indicates exothermic andendothermic nature of adsorption respectively. The study onsynthetic mixture containing refractory sulphur compounds likebenzothiophene and dibenzothiophene, confirmed high sulphurremoval capacity. The understanding of the adsorption mechanismand insights from this study will help in design and development ofnew adsorbents with potential for selective adsorption of refractorysulphur compound from liquid fuels.

References

[1] R.T. Yang, A.J. Hernández-Maldonado, F.H. Yang, Desulfurization of transpor-tation fuels with zeolites under ambient conditions, Science 301 (2003) 79–81,doi:http://dx.doi.org/10.1126/science.1085088. 12843388.

[2] V.M. Bhandari, C.H. Hyun Ko, J. Geun Park, S.-S. Han, S.-H. Cho, J.-N. Kim,Desulfurization of diesel using ion-exchanged zeolites, Chem. Eng. Sci. 61(2006) 2599–2608, doi:http://dx.doi.org/10.1016/j.ces.2005.11.015.

[3] K.A. Cychosz, A.G. Wong-Foy, A.J. Matzger, Liquid phase adsorption bymicroporous coordination polymers: removal of organosulfur compounds, J.Am. Chem. Soc. 130 (2008) 6938–6939, doi:http://dx.doi.org/10.1021/ja802121u. 18465854.

[4] Y. Huang, Z. Zhou, Y. Qi, X. Li, Z. Cheng, W. Yuan, Hierarchically macro-/mesoporous structured Co–Mo–Ni/g-al2O3 catalyst for the hydrodesulfuriza-tion of thiophene, Chem. Eng. J. 172 (2011) 444–451, doi:http://dx.doi.org/10.1016/j.cej.2011.06.006.

[5] C. Fredrick, Sulfur reduction; what are the options? Hydrocarbon Proc. 81(2002) 45–50.

[6] W.H. Tian, L.B. Sun, X.L. Song, X.Q. Liu, Y. Yin, G.S. He, Adsorptivedesulfurization by copper species within confined space, Langmuir: ACS J.Surf. Colloids 26 (2010) 17398–17404, doi:http://dx.doi.org/10.1021/la101856d. 20882950.

[7] M. Soleimani, A. Bassi, A. Margaritis, Biodesulfurization of refractory organicsulfur compounds in fossil fuels, Biotechnol. Adv. 25 (2007) 570–596, doi:http://dx.doi.org/10.1016/j.biotechadv.2007.07.003. 17716849.

[8] B.R. Folsom, D.R. Schieche, P.M. DiGrazia, J. Werner, S. Palmer, Microbialdesulfurization of alkylated dibenzothiophenes from a hydrodesulfurizedmiddle distillate by Rhodococcus erythropolis I-19, Appl. Environ. Microbiol.65 (1999) 4967–4972. 10543810.

[9] C. Song, X. Ma, New design approaches to ultra-clean diesel fuels by deepdesulfurization and deep dearomatization, Appl. Catal. B: Environ. 41 (2003)207–238, doi:http://dx.doi.org/10.1016/S0926-3373(02)00212-6.

[10] L.C. Caero, E. Hernández, F. Pedraza, F. Murrieta, Oxidative desulfurization ofsynthetic diesel using supported catalysts: part I. Study of the operationconditions with a vanadium oxide based catalyst, Catal. Today 107–108 (2005)564–569.

S.V. Patil et al. / Journal of Environmental Chemical Engineering 2 (2014) 1495–1505 1505

[11] Z.E.A. Abdalla, B. Li, Preparation of MCM-41 supported (Bu4N)4H3(PW11O39)catalyst and its performance in oxidative desulfurization, Chem. Eng. J. 200–202 (2012) 113–121.

[12] Y. Shiraishi, K. Tachibana, T. Hirai, I. Komasawa, Desulfurization anddenitrogenation process for light oils based on chemical oxidation followedby liquid–liquid extraction, Ind. Eng. Chem. Res. 41 (2002) 4362–4375, doi:http://dx.doi.org/10.1021/ie010618x.

[13] J. Bu, G. Loh, C.G. Gwie, S. Dewiyanti, M. Tasrif, A. Borgna, Desulfurization ofdiesel fuels by selective adsorption on activated carbons: competitiveadsorption of polycyclic aromatic sulfur heterocycles and polycyclic aromatichydrocarbons, Chem. Eng. J. 166 (2011) 207–217, doi:http://dx.doi.org/10.1016/j.cej.2010.10.063.

[14] C.O. Ania, J.B. Parra, A. Arenillas, F. Rubiera, T.J. Bandosz, J.J. Pis, On themechanism of reactive adsorption of dibenzothiophene on organic wastederived carbons, Appl. Surf. Sci. 253 (2007) 5899–5903, doi:http://dx.doi.org/10.1016/j.apsusc.2006.12.065.

[15] J.X. Guoa, J. Lianga, Y.H. Chua, M.C. Suna, H.Q. Yina, J.J. Li, Desulfurizationactivity of nickel supported on acid-treated activated carbons, Appl. Catal. PartA : Gen. 421–422 (2012) 142–147.

[16] X. Gao, S.J. Liu, Y. Zhang, Z.Y. Luo, K.F. Cen, Physicochemical properties of metal-doped activated carbons and relationship with their performance in theremoval of SO2 and NO, J. Hazard. Mater. 188 (2011) 58–66, doi:http://dx.doi.org/10.1016/j.jhazmat.2011.01.065. 21324585.

[17] C.O. Ania, T.J. Bandosz, Metal-loaded polystyrene-based activated carbons asdibenzothiophene removal media via reactive adsorption, Carbon 44 (2006)2404–2412, doi:http://dx.doi.org/10.1016/j.carbon.2006.05.016.

[18] J.J. Bian, S. Zhang, J.W. Zhang, X. Min, C.H. Li, Supported manganese dioxidecatalyst for seawater flue gas desulfurization application, Chem. Eng. J. 189(2012) 57–61.

[19] J.R. Ma, Z.Y. Liu, Q.Y. Liu, S.J. Guo, Z.G. Huang, Y. Xiao, SO2 and NO removal fromflue gas over V2O5/AC at lower temperatures-role of V2O5 on SO2 removal,Fuel Process. Technol. 89 (378) (2008) 242–248.

[20] A.J. Hernández-Maldonado, R.T. Yang, Desulfurization of commercial liquidfuels by selective adsorption via p-complexation with Cu(I)-Y zeolite, Ind.Eng. Chem. Res. 42 (2003) 3103–3110, doi:http://dx.doi.org/10.1021/ie0301132.

[21] C.-J. Li, Y.-J. Li, J.-N. Wang, L. Zhao, J. Cheng, Ag+-loaded polystyrenenanofibrous membranes preparation and their adsorption properties forthiophene, Chem. Eng. J. 222 (2013) 419–425, doi:http://dx.doi.org/10.1016/j.cej.2012.09.107.

[22] A.F. Brady, Global refining margins look poor in short term, buoyant later nextdecade, Oil Gas J. 97 (1999) 75–80.

[23] Y.C. Ning, Interpretation of Organic Spectra, first edition, John Wiley & Sons,Singapore, 2011.

[24] T.F. Young, C.P. Chen, J.F. Liou, Y.L. Yang, T.C. Chang, Study of Si–Si vibrationalstates of the near surface region of porous silicon, J. Porous Mater. 7 (2000)339–343, doi:http://dx.doi.org/10.1023/A:1009622601723.

[25] Q. Yang, D. Wang, Y. Guo, K. Ding, J. Zhang, The effect of annealing on thephotoluminescence of Si/SiOx, Integ. Ferroelect. 135 (2012) 89–93, doi:http://dx.doi.org/10.1080/10584587.2012.685410.

[26] X. Gao, S. Liu, Y. Zhang, Z. Luo, K. Cen, Physicochemical properties of metal-doped activated carbons and relationship with their performance in the

removal of SO2 and NO, J. Hazard. Mater. 188 (2011) 58–66, doi:http://dx.doi.org/10.1016/j.jhazmat.2011.01.065. 21324585.

[27] I. Langmuir, The constitution and fundamental properties of solids and liquids.Part I. Solids, J. Am. Chem. Soc. 38 (1916) 2221–2295, doi:http://dx.doi.org/10.1021/ja02268a002.

[28] H. Freundlich, über die adsorption in losungen, Z. Physiol. Chem. 57 (1906)385–470.

[29] G. McKay, H.S. Blair, J.R. Gardner, Adsorption of dyes on chitin. I. Equilibriumstudies, J. Appl. Polymer Sci. 27 (1982) 3043–3057, doi:http://dx.doi.org/10.1002/app.1982.070270827.

[30] C.H. Rochester, R.J. Terrell, Infrared study of the adsorption of sulphurcompounds on silica and silica-supported nickel, J.Chem. Soc. Faraday Trans.1:Phys. Chem. Condensed Phases 73 (1977) 596–608, doi:http://dx.doi.org/10.1039/f19777300596.

[31] A. Chica, K. Strohmaier, E. Iglesia, Adsorption, desorption, and conversion ofthiophene on H-ZSM5, Langmuir: ACS J. Surf. Colloids 20 (2004) 10982–10991,doi:http://dx.doi.org/10.1021/la048320.

[32] D.L. King, C. Faz, T. Flynn, Desulfurization of Gasoline Feedstocks forApplication in Fuel Reforming, Society of Automotive Engineers, Detroit,MI, 2000.

[33] W. Dai, Y.P. Zhou, S.N. Li, W. Li, W. Su, Y. Sun, L. Zhou, Thiophene capture withcomplex adsorbent SBA-15/Cu(I), Ind. Eng. Chem. Res. 45 (2006) 7892–7896,doi:http://dx.doi.org/10.1021/ie060428s.

[34] Z. Jiang, Y. Liu, X. Sun, F. Tian, F. Sun, C. Liang, W. You, C. Han, C. Li, Activatedcarbons chemically modified by concentrated H2SO4 for adsorption of thepollutants from wastewater and dibenzothiophene from fuel oils, Langmuir 19(2003) 731–736, doi:http://dx.doi.org/10.1021/la020670d.

[35] W. Dai, Y. Zhou, S. Wang, W. Su, Y. Sun, L. Zhou, Desulfurization oftransportation fuels targeting at removal of thiophene/benzothiophene, FuelProcess. Technol. 89 (2008) 749–755, doi:http://dx.doi.org/10.1016/j.fuproc.2008.01.002.

[36] Q. Wang, X. Liang, R. Zhang, C. Liu, X. Liu, W. Qiao, L. Zhan, Preparation ofpolystyrene-based activated carbon spheres and their adsorption of dibenzo-thiophene, New Carbon Mater. 24 (2009) 55–60, doi:http://dx.doi.org/10.1016/S1872-5805(08)60036-0.

[37] Y.S. Ho, G. McKay, Kinetic models for the sorption of dye from aqueous solutionby wood, Process Safety Environ. Protection 76 (1998) 183–191, doi:http://dx.doi.org/10.1205/095758298529326.

[38] S. Azizian, Kinetic models of sorption: a theoretical analysis, J. Colloid Interf.Sci. 276 (2004) 47–52, doi:http://dx.doi.org/10.1016/j.jcis.2004.03.048.15219428.

[39] W.J. Weber, J.C. Morris, Advances in water pollution research: removal ofbiologically resistant pollutants from waste water by adsorption, Proceedingsof International Conference on Water Pollution Symposium, Pergamon Press,Oxford, 1962, pp. 231–266.

[40] L.G. Sorokhaibam, V.M. Bhandari, V.V. Ranade, New adsorbents and materialmodifications – an approach for ultraclean fuels, Proceedings of AIChE AnnualMeeting, Hilton San Francisco, CA, 2013).

[41] Q. Zhao, Z. Chang, X. Lei, X. Sun, Adsorption behavior of thiophene fromaqueous solution on carbonate- and dodecylsulfate-intercalated ZnAl layereddouble hydroxides, Ind. Eng. Chem. Res. 50 (2011) 10253–10258, doi:http://dx.doi.org/10.1021/ie201021k.