using activated carbon based on agricultural waste (date stones) A kinetic, equilibrium and thermodynamic study of l -phenylalanineadsorption Technology Journal of Applied Researchand

(1)

Availableonlineatwww.sciencedirect.com

Journal of Applied Research and Technology

www.jart.ccadet.unam.mx JournalofAppliedResearchandTechnology14(2016)354–366

Review

A kinetic, equilibrium and thermodynamic study of l-phenylalanine adsorption using activated carbon based on agricultural waste (date stones)

Badreddine Belhamdi

^a,∗

, Zoulikha Merzougui

^a

, Mohamed Trari

^b

, Abdelhamid Addoun

^a

aLaboratoryofPhysicalandChemicalStudyofMaterialsandApplicationsintheEnvironment,FacultyofChemistry(USTHB),BP32-16111EL-Alia,Algeria

bLaboratoryofStorageandValorizationofRenewableEnergies,FacultyofChemistry(USTHB),BP32-16111EL-Alia,Algeria

Abstract

The mainpurposeof this workis to produce lowcostactivated carbonsfrom date stoneswastesforthe adsorption ofl-phenylalanine.

The activated carbons were prepared by chemicalactivation with KOH(ACK) and ZnCl2 (ACZ) and characterizedby scanning electron microscopy,N₂ adsorption–desorptionisothermsandFT-IRspectroscopy.BothTheactivatedcarbonsACKandACZhavehighspecificsur- face areasandlargepore volumes,favorableforthe adsorption. Batch experimentswere conducted to determinethe adsorptioncapacities.

AStrongdependenceofthe adsorptioncapacityonpHwasobserved,thecapacitydecreaseswithincreasing pHuptooptimalvalueof5.7.

Theadsorptionfollowsapseudo-secondorder kineticmodel.Additionally,the equilibriumadsorptiondata werewellfittedto theLangmuir isotherm, and themaximumadsorption capacitiesof l-phenylalanineonto ACKand ACZ were188.3 and133.3mgg⁻¹ atpH 5.7,respectively.Thethermodynamicstudyrevealedthattheadsorptionofl-phenylalanineontoactivatedcarbonswasexothermicinnature.Theproposed adsorptionmechanismstakeintoaccountthehydrophobicandelectrostaticinteractionswhichplayedthecriticalrolesinthel-phenylalanine adsorption.

Keywords: Activatedcarbon;l-phenylalanine;Sizedistribution;Sorbentsurface;Adsorptionisotherm;Thermodynamic

1. Introduction

Adsorptionofaminoacidsontosolidsurfaceshasreceived muchattentionbecause ofitsscientificimportanceandappli- cations in the separation and purification processes (Han &

Yun, 2007; Hong &Bruening, 2006; Kostova &Bart, 2007;

O’Connoretal.,2006;Sánchez-Hernández,Bernal,delNozal,

&Toribio,2016).Aminoacidsare biomoleculesof greatrel- evancethat arewidelyused inmanyindustries such asfood, cosmetic,medicine,biochemistryandothers(Bourke&Kohn, 2003; Hartmann, 2005; Infante etal., 2004; Oshima, Saisho, Ohe, Baba, & Ohto,2009; Palit &Moulik, 2001). They are non-toxic andare used as buildingblocks for the production of pharmaceutical andagrochemicalcompounds.In addition,

∗Correspondingauthor.

E-mailaddress:[email protected](B.Belhamdi).

PeerReviewundertheresponsibilityofUniversidadNacionalAutónomade México.

they are interesting moleculesas adsorbates because of their molecularsizeandzwitterionicnature(O’Connoretal.,2006).

Similar tomany aminoacids, l-phenylalanineisessential for animalsandthehumanbody.Itisextensivelyusedasingredi- ent infood or feedadditive, ininfusionfluids,neutraceutical andpharmaceutical(Pimentel,Alves,Costa,Fernandes,etal., 2014;Pimentel,Alves,Costa,Torres,etal.,2014;Zhou,Liao, Wang,Du,&Chen,2010).Generally,theaminoacidshavebeen studied by adsorption on well-ordered surfacesof solids. On theotherhand,mostof thecurrentmethodsemployedforthe removalofl-phenylalaninefromproteinhydrolysatesarebased on theadsorption onactivated carbon,polymericresins,zeo- lites andion exchangers (Lopes, Delvivo, &Silvestre, 2005;

Outinen et al., 1996; Shimamuraet al., 2002).These studies giveinformationforpracticalresearchesonthepurificationand separation of amino acids. Over the last years, several studies have been reported for the adsorption of amino acids on poroussolids(Casadoetal.,2012;ElShafei,2002;ElShafei

&Moussa,2001;Ghosh,Badruddoza,Uddin,&Hidajat,2011;

http://dx.doi.org/10.1016/j.jart.2016.08.004

1665-6423/©2016UniversidadNacionalAutónomadeMéxico,CentrodeCienciasAplicadasyDesarrolloTecnológico.Thisisanopenaccessarticleunderthe CCBY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/4.0/).

(2)

Goscianska,Olejnik,&Pietrzak,2013b,2013c;Jiao,Fu,Shuai,

&Chen,2012;Longetal.,2009;Mei,Min,&Lü,2009;Palit

&Moulik,2001;Silvério, DosReis, Tronto,&Valim, 2008;

Titus,Kalkar,&Gaikar,2003;Wu,Zhao,Nie,&Jiang,2009).

However,theiradsorption capacityisstill lowbecauseof the smallporevolumeorwideporesoftheseadsorbents,whichare consequentlyinappropriatetothemolecularsizeofaminoacids.

Moreover,themajordrawbackoftheadsorptionprocessisthe highcostfortheproductionandregenerationofadsorbents.Such inconvenientresultedingrowingresearchoninexpensiveadsor- bents(Alves,Franca,&Oliveira,2013a,2013b;Clark,Alves, Franca, & Oliveira, 2012; Goscianska, Nowicki, & Pietrzak, 2014;He,Lin,Long,Liang,&Chen,2015;Sebben&Pendleton, 2015).

In this study, porous activated carbon-based materials obtainedfromdatestones(seeds) arepotentialadsorbentsfor l-phenylalanineaminoacid.Thisismainlyduetotheirphysi- calandchemicalcharacteristicssuchashighlydevelopedporous structure,goodthermalstability,lowcostandmoreaccessibility.

Datestonesareamongthemostcommonagriculturalbyprod- uctsavailableinpalmsgrowingintheMediterraneancountries likeAlgeria,whichisoneofthelargestproducersintheworld.

Algeriaproducesmorethan400differentvarietiesofdateswith anannualproductionof about 400,000tons(Chandrasekaran

&Bahkali,2013).Date stonesconstitute roughly 10%of the dateweightandthislignocellulosic-basedagriculturalwasteis agoodprecursorforpreparingactivatedcarbonbecauseofits excellentnatural structure andlow ashcontent (Bouchenafa- Saib,Grange,Verhasselt,Addoun,&Dubois,2005;Merzougui

&Addoun,2008).Asitiswellknown,twomethodsarecom- monlyusedforthepreparationofactivatedcarbon:physicaland chemicalactivations.Comparedwiththephysicalprocess,the chemicalactivationpresentssomeadvantageslike lowactiva- tiontemperature,shortactivationtime,highsurfacearea,well developedmicroporosityofactivatedcarbon,simpleoperation andlowenergyconsumption(Deng,Yang,Tao,&Dai,2009;

Pereiraetal.,2014).Therefore,thedatestonescanbeactivated withchemicalagentssuchasKOH,ZnCl2,H3PO4,K2CO3and NaOH,toobtainactivatedcarbonswithwell-developedtextural characteristics.To thebestof ourknowledge,theuseof activatedcarbons forthe l-phenylalaninerecoveryfromaqueous solutionsbyadsorbentsbasedondatestonesarenotavailable intheopenliterature.Thus,theprincipalobjectiveofthiswork wastoprepareporousactivatedcarbonswithhighsurfaceareas fromdatestonesbychemicalactivationwithKOHandZnCl2. The activated carbons proved to be good candidates for the adsorptionofl-phenylalanineinanaqueousmedium.

2. Materialsandmethods 2.1. Materials

The date stones used in this study were from Algerian origin. The following reagents were used: l-phenylalanine standard(>98%,Fluka,France),potassiumhydroxide(>98%, Sigma Aldrich, USA), zinc chloride (>98%, Sigma Aldrich, USA),KH2PO4(>99%,Fluka,France),K2HPO4(>99%,Fluka,

France), NaHCO3 (>99%, Sigma Aldrich, USA), Na2CO3

(>99%,SigmaAldrich,USA),HCl(37%,SigmaAldrich,USA), KCl(>98%,Fluka,France),NaCl(>99%,SigmaAldrich,USA), NaOH (>99%, Sigma Aldrich, USA). Ultrapure water was obtainedfrommilli-Qsystem(Millipore,France).

2.2. Preparationoftheactivatedcarbons

The activated carbons wereprepared from datestones. At first, the stones were thoroughly washedwith distilled water anddriedinanairovenat120^◦C;suchprotocolwaseffective tofacilitate crushingandgrinding. Afraction particlesizeof between0.5and1mmwasusedforthepreparationofactivated carbonsbyimpregnationwithZnCl2andKOH.Theprecursor wasimpregnatedwithachemicalactivatingagentinasolidform.

Theimpregnatedprecursorwascarbonizedinahorizontaltubu- larfurnaceundernitrogenflowwithaheatingrateof5^◦Cmin⁻¹, toallowfreeevolutionofvolatiles,uptotheholdtemperaturefor 1h.TheresultingactivatedcarbonwasimmersedinHClsolu- tion(0.1molL⁻¹)underrefluxebullition(3h)inordertoextract the compoundformed andreagentexcess.Then, the solution wasfilteredandtheblacksolidwaswashedwithhotdistilled wateruntil thetestwithAgNO3becamenegative. Theadsor- bentwasdriedat120^◦C,andkeptintightlyclosedbottlesuntil use.The activated carbons were namedACZ (1g ZnCl2:1g datestones,activatedat600^◦C),ACK(9mmolKOH:1gdate stones,activatedat800^◦C).

2.3. Characterization

Thespecificsurfaceareaandporestructureoftheactivated carbons werecharacterizedbynitrogenadsorption-desorption isotherms at −196^◦C using the ASAP 2010 Micromeritics equipment.Alltheactivatedcarbonswereoutgassedat150^◦C overnight. The specific surface area was calculated by the Brunauer–Emmett–Teller(BET)equation(Brunauer,Emmett,

&Teller,1938).Theexternalsurfacearea,microporeareaand micropore volumewere calculatedbythe t-plot method.The totalporevolumewasevaluatedfromtheliquidvolumeofN2

atahighrelativepressurenearunity0.99(Guo&Lua,2000).

Themesoporevolumewascalculatedbysubtractingthemicro- porevolumefromthetotalvolume.Theporesizedistribution (PSD)wasdeterminedusingthedensityfunctionaltheory(DFT) model.Themorphologyofactivatedcarbonswasvisualizedby scanning electron microscopy (SEM) using a Philips XL 30 equippedwith anenergy dispersivespectrometer (EDS). The Fourier transform infrared (FT-IR) spectroscopy was used to determine thefunctionalgroups of the activatedcarbons; the spectra were recorded over the range (400–4000cm⁻¹) on a Perkin-ElmerspectrumtwospectrometerusingKBrpellets.

2.4. DeterminationofzeropointchargepHPZC

Thedeterminationofthepointofzerocharge(pHPZC)was conducted toinvestigatehow the surfacechargeof ACKand ACZ adsorbents dependson pH. pHPZC of the activated car- bonswasdeterminedusingtheproceduredescribedelsewhere

(3)

Fig.1.SEMimagesofACZ(A)andACK(B)beforeadsorption.

(Prahas,Kartika,Indraswati,&Ismadji,2008):0.01MofNaCl waspreparedandtheinitialpHwasadjustedbetween2and12 usingHClorNaOHsolution(0.1M).50mLofNaClsolution wasplacedinErlenmeyer flakeswith0.1g ofadsorbent.The flaskswerekeptunderagitation(150rpm,48h),andthefinal pHofthesolutionwasmeasured.Theintersectionpointofthe curvespHfinalvs.pHinitialandthebisectorwastakenaspHPZC.

2.5. Adsorptionexperiments

Thebatchadsorptionexperimentswereperformedin100mL Erlenmeyerflaskscontainingamassofadsorbentmixedwitha knownvolumeofl-phenylalaninesolution(200mgL⁻¹)under agitation (150rpm).The effect of the contact timewas studied to determinethe time requiredfor equilibriumat natural pH at 20^◦C. For the temperature effect, 50mg of activated carbonwasaddedto50mLofl-phenylalaninesolutionswith concentrations rangingfrom50 to1000mgL⁻¹,prepared by dissolvingappropriateamountsofl-phenylalanineinultrapure water(18.2Mcm),theflasksweremaintainedunderconstant agitation at various temperatures (20, 25, 35 and 40^◦C) for 300min.TheeffectofpHontheadsorptionwasperformedby mixing50mgofactivatedcarboninto50mLofl-phenylalanine solutions inthe pH range (2.0–9.4), under constantagitation at 20^◦C for 300min. The pH value of the l-phenylalanine solutionswaschangedbyusingdifferentbuffersolutions(pH 5.7–7.2potassiumphosphatebuffer:pH2.0HCl–KClbuffer:pH 9.4bicarbonatebuffer).Theconcentrationsofl-phenylalanine remaining in the supernatant solutions were filtered using a hydrophilic syringe filter with a pore size of 0.45 um. The adsorbedamountwasdeterminedusingaUV–visspectropho- tometerat257nm.(Theadsorptioncapacity)Theequilibrium adsorption capacity per unit mass of activated carbons qe

(mgg⁻¹)andtheremovalpercentageofthel-phenylalanineη (%)werecalculatedfromthefollowingequation:

qe= (C0−C_e)V

W (1)

η=(C0−C_e)

C0 ×100 (2)

whereC0andCearetheinitialandequilibriuml-phenylalanine aminoacidconcentrationsintheliquidphase(mgL⁻¹),W(g) theweightofadsorbentandV(L)thevolumeofsolution.

2.6. Desorptionstudy

Desorption ofl-phenylalaninewasinvestigatedinorderto exploretheregeneration andrecycling abilityofthetwo activated carbons.Forthis,50mgofACK andACZ weremixed with50mLofl-phenylalaninesolutionatasaturatedconcentra- tion,andstirredat150rpmatoptimumadsorptiontemperature (20^◦C)for300min.Theamountofadsorbedaminoacidwas determinedbythesameequationusedintheadsorptionexper- iments (see Section 2.5). Thereafter, ACK and ACZ were washed with ultrapure water until the residual concentration of l-phenylalanine becomes negligible. The loaded activated carbonswerethenallowedtobeincontactwith50mLoftwo eluentsolutions(MNaOHandMHCl,10⁻²M) for300min.

Thedesorbedcarbonswereagainsubjectedtothenextbatchin ordertocheckdesorptionandreusabilityofACKandACZ.The amountof desorbedaminoacidwascalculatedfromthe concentration of desorbedl-phenylalanine in liquid phase using UV–vis spectrophotometerat257nm.Thepercentageofdes- orbedl-phenylalaninefromtheactivatedcarbonswascalculated accordingtothefollowingequation:

Desorption(%)=

Massdesorbed Massadsorbed×100

(3)

3. Resultsanddiscussion

3.1. Characterizationofactivatedcarbons

3.1.1. Scanningelectronmicroscopy(SEM)

TheSEMmicrographsshowtheeffectsofZnCl2andKOH onthesurfaceporestructuresoftheactivatedcarbons(Fig.1).

Theexternalmorphologyshowsmoreorlesshomogeneouscav- itiesonthesurfacesofACKandACZ.Thesecavitiesresulted fromthe evaporationof chemicalagentsduringtheactivation process,leavingspacepreviouslyoccupiedbyKOHandZnCl2. Figure 1suggeststhatthe largepores onthesurfaceare con- nectedtoawholenetworkofsmallerporesinsidetheactivated

(4)

Table1

TexturalcharacteristicsofACKandACZ.

Adsorbent Surfacearea(m²g⁻¹) Porevolume(cm³g⁻¹) DFTporesize(nm)

SBET Sext Smic VT Vmes Vmic

ACK 1209 276 933 0.550 0.180 0.370 1.48

ACZ 1235 525 710 0.630 0.341 0.289 1.59

0.0 0.2 0.4 0.6 0.8 1.0

100 200 300 400 500

Adsorption isotherm ACZ

Adsorption isotherm ACK Desorption isotherm ACK Desorption isotherm ACZ Volume adsorbed (cm3 g-1)

Relative presure (p/p₀)

Fig.2.Nitrogenadsorption–desorptionisothermsofACKandACZsamplesat

−196^◦C.

carbon.Therefore,the ACKandACZhavegreatpotential as goodadsorbentsforl-phenylalanine.

3.1.2. Nitrogensorption

The N2 adsorption–desorption isotherms of activated carbons ACK and ACZ are given in Figure 2. The isotherms aretypeI,accordingtothe IUPACclassification,assigned to microporous materials, and they present at a very low rela- tivepressure(P/P0<0.2)asignificantincreaseofN2adsorption correspondingtothemicroporesfilling(Fig.2).However,the amountofadsorbednitrogenisreducedathigherpressures,sug- gesting the developmentof bothmicro andmeso-porosity in ACKandACZ.Thepresenceofhysteresisloopsindicatesthat somemesoporositystartstobedevelopedbycapillaryconden- sation.Thetexturalparametersofactivatedcarbonsdetermined fromnitrogenadsorption-desorptionaregatheredinTable1.It canbeconcluded that the activation of datestones byZnCl2

and KOH leads to active coals of a well-developed surface areaandahigh porevolume(Table1).Amongthe activated carbons,ACZexhibitsthe highestsurfacearea(1235m²g⁻¹) andporevolume(0.63cm³g⁻¹)whileACKhasasmallersur- facearea (1209m²g⁻¹) andpore volume(0.55cm³g⁻¹). As shownin Table 1,the activated carbon prepared by KOH is essentially microporous, with 77% of its surface area. Simi- lartrends havebeen found for the influence of the chemical activating agent on the development of the surface area and porevolumeofactivatedcarbonsobtainedthroughZnCl2and KOHactivationofotherlignocellulosicmaterials(Angin,2014;

Bagheri&Abedi,2009;Foo&Hameed,2011;Sre´nscek-Nazzal, Kami´nska,Michalkiewicz, &Koren,2013; Yorgun, Vural,&

0 10 20 30 40 50 60 70 80 90 100 110

0.000 0.004 0.008 0.012 0.016 0.020

ACZ ACK

Incremental pore volume (cm3 g–1)

Pore diameter (nm)

Fig.3.DFTporesizedistributionforACKandACZsamples.

Demiral,2009;Zhu,Wang,Peng,Yang,&Yan,2014).Thepore sizedistributionisanimportantpropertyintheadsorptionmech- anismbecausetheadsorptionofmoleculesofdifferentsizesand shapesisdirectlyrelatedtotheporesizeofadsorbents.Accord- ing tothe classification adopted by IUPAC, adsorbent pores areclassifiedasmicropores(<2nm),mesopores(2–50nm)and macropores(>50nm).Figure3showstheporesizedistribution fortheactivatedcarbons,whichclearlyindicatesthatthepore diameterisinthemicroporerange.Itisimportanttomention thatal-phenylalaninemoleculeisrelativelysmallwithasize of 0.7nm×0.5nm×0.5nm(Alvesetal.,2013b;Longetal., 2009).Therefore,suchmicropores withasizeof (<2nm) are accessibleforl-phenylalaninemolecules.

3.1.3. Infraredspectroscopy(FT-IR)

TheFT-IRisanimportanttechniquetoqualitativelydeter- minatethe characteristicfunctionalgroupsof the adsorbents.

Thespectraofporousactivatedcarbons(Fig.4)showmultiple functionswhichcanalsobeobservedinothercarbonsactivated byKOHandZnCl2(Huang,Ma,&Zhao,2015;Lua&Yang, 2005;Saka,2012).TheFT-IRanalysisindicatesthatACKand ACZ exhibitasimilarshape andthesamefunctionalgroups.

The broad band in the range (3000–3500cm⁻¹) is ascribed to the O–H stretchingmode of hydroxyl groups withhydro- genbendingof adsorbedwater.Bands(2900–2950cm⁻¹) are assignedtoasymmetricandsymmetricstretchingvibrationsof aliphatic bond–CH,–CH2and–CH3whilethebandsaround 1580cm⁻¹maybeduetothepresenceof aromaticC Cring stretchingvibration.Thebandat1400cm⁻¹isassociatedwith –COO–asymmetricvibrationofcarboxylicgroupswhilethatat 1384cm⁻¹isduetostretchingvibrationof–CH3group.Finally, thevibrationbandcenteredat1115cm⁻¹isattributedtoC–O

(5)

4000 3500 3000 2500 2000 1500 1000 500 ACK

Wavenumber (cm^–1) ACZ

Transmittance, %

Fig.4.TheFTIRspectraofACKandACZsamplesbeforeadsorptioninthe range4000–500cm⁻¹.

0 60 120 180 240 300 360

0 20 40 60 80 100 120 140

ACK ACZ qt (mg g–1)

t (min)

Fig.5.Effectofcontacttimeonthel-phenylalanineadsorptionontoACKand ACZ(l-phenylalanineconcentration=200mgL⁻¹,adsorbentdose=1gL⁻¹, agitationspeed=150rpm,contacttime=300min,temperature=20^◦C,natural pH).

stretchingvibrations,as inalcohols,phenols, acids, ethersor esters.Thepresenceofthefunctionalgroupssuchascarboxyl andhydroxylarepotentialadsorptionsitesforl-phenylalanine aminoacid.

3.2. Adsorptionstudies

3.2.1. Effectofcontacttime

Thecontact timeisafundamentalparameterinany trans- ferphenomenasuchasadsorption.Theequilibriumadsorption capacityofl-phenylalanineonactivatedcarbonwasinvestigated todeterminethetimerequiredtoreachtheequilibriumbetween adsorbents(50mg)andl-phenylalaninesolution,(200mgL⁻¹) (Fig. 5); it can be observed that the adsorption capacities of activatedcarbonsgraduallyincreasewiththecontacttimeand doesnotstopuntilanequilibriumstateisreached(180min).No obvious variationinthe amount of adsorbed l-phenylalanine was observed; the adsorbed mass at equilibrium reflects the maximumadsorption capacityof the activated carbons under the operating conditions; the equilibrium adsorption capacity ofl-phenylalanineonACK(114mgg⁻¹)ishigher thanACZ

0 100 200 300 400 500 600 700 800

0 20 40 60 80 100 120

ACZ 20 °C ACZ 25 °C ACZ 35 °C ACZ 40 °C

ACK 20 °C ACK 25 °C ACK 35 °C ACK 40 °C C_e (mg L^–1)

C_e (mg L^–1)

0 100 200 300 400 500 600 700

0 40 80 120 160 200 qe (mg g–1)qe (mg g–1)

Fig.6.Effectoftemperatureonl-phenylalanineadsorptionontoACKandACZ (adsorbent dose=1gL⁻¹,agitationspeed=150rpm,contacttime=300min, naturalpH).

(68mgg⁻¹) (Fig. 5). Comparing the results obtained in this work (porous activated carbons) with mesoporous materials, such as the SBA-3 mesoporous silica tested elsewhere (Goscianska,Olejnik,&Pietrzak,2013a),itcanbeconcluded that the mesoporous materials need amuch longer period to reachtheequilibrium;inthiscasetheporesizedistributionof ACKandACZ,centeredat1.48and1.59nm,respectively,are beneficialfortheadsorptionbecausel-phenylalaninemolecules haveeasyaccesstothepores.

3.2.2. Temperatureeffect

The temperature has a direct influence on the adsorption of amino acids.Figure 6shows thetemperature effectonthe adsorptionofl-phenylalaninebyactivatedcarbons.Thesame behaviorisobservedforACKandACZandtheisothermsplot- tedatvarioustemperaturesshowthattheequilibriumadsorption capacitydecreaseswithincreasingtemperaturefrom20to40^◦C, indicatingthattheadsorptionofl-phenylalanineisofexother- micnature.Theseresultsshowthatthedecreaseofadsorption atahightemperaturecanbe ascribedtothe greatertendency of l-phenylalanine moleculesto form hydrophobic bonds in anaqueousmedium,thushinderingtheirhydrophobicinterac- tionswiththeadsorbentsurface(El Shafei&Moussa,2001).

(6)

0 100 200 300 400 500 600 700 800 0

20 40 60 80 100 120

0 100 200 300 400 500 600 700 800

0 20 40 60 80 100 120 140 160 180 200

ACK pH 2.0 ACK pH 5.7 ACK pH 7.2 ACK pH 9.4 ACZ pH 2.0 ACZ pH 5.7 ACZ pH 7.2 ACZ pH 9.4

C_e (mg L^–1)

C_e (mg L^–1) qe (mg g–1)qe (mg g–1)

Fig. 7.Effect of pH on l-phenylalanine adsorption onto ACK and ACZ (adsorbentdose=1gL⁻¹,agitationspeed=150rpm,contacttime=300min, temperature=20^◦C).

Ontheotherhand,thedecreasedadsorption atequilibriumis duetodecreasedsurface activity athigher temperatures.The bestuptakeofl-phenylalaninewasobtainedat20^◦Cwhichis selectedasanoptimaladsorptiontemperatureandwillbeused forfurtherexperiments.

3.2.3. EffectofpH

pHisknowntobeacrucialparameterthataffectstheadsorp- tionbehavioratwater–solidinterfaces.Itseffectontheamino acidadsorptiononACKandACZatdifferentbuffersolutions rangingfrom2.0to9.4at20^◦CisillustratedinFigure7.All equilibriumisothermsareoftheL-type(Langmuirisotherms) andtheamountofadsorbedl-phenylalanineincreaseswithrais- ingtheinitialconcentration.Atalowconcentration,theamino acidisrandomlydepositedontheadsorbent,andthefastuptake canbeattributedtoalargenumberofemptysitesonthesurface (Goscianskaetal.,2014).Incontrast,athighconcentrations,the nonpolargroupsofaminoacidsareclosetoeachotheruntilthey touchinsidetheirvanderWaalsradii,leadingtoadensepacking ofmoleculesontheactivesitesoftheadsorbent.Theremovalof l-phenylalaninefromanaqueoussolutionisstronglydepend- entonpH(Fig.7)andthiscanbeexplainedbypHpzc.pHPZC

(Fig.8)isfoundtobe6.18(ACZ)and6.80(ACK).Thesurface

0 2 4 6 8 10 12 14

ACK pH _PZC=6.80 ACK pH _PZC=6.18 Bisector

Final pH

Initial pH

Fig.8.DeterminationofthepHofzeropointcharge(PZC)ofACZandACK.

oftheactivatedcarbonispositivelychargedbelowpHpzcand negatively charged above pHpzc. The maximum adsorption capacityofbothactivatedcarbonswasobtainedatpH5.7,which isclosetotheisoelectricpoint(PI=5.48)ofl-phenylalanine.

The latter is known as a Zwitterion containing both amine and carboxylic groups, near to the isoelectric point and presentsbothnegativeandpositivecharges(Jiaoetal.,2012).

Therefore, the Coulombrepulsive interactionbetween the l- phenylalaninemoleculesisalmostnegligible.Furthermore,the strong hydrophobic interactions amino acid/adsorbents, and intra-molecularinteractionbetweenaminoacidmoleculesare responsibleof closepacking ofl-phenylalanineinthemicro- poresofadsorbents,leadingtothehighestadsorptioncapacity atthispH.AtlowpH(<2),thesurfacechargeoftheactivated carbonisnegativeandthecationic formof theamino acidis not favorable for the adsorption because of the electrostatic repulsion;thisexplainsthedecreaseintheadsorptionefficiency.

Figure7alsoshowsthattheadsorbedamountofl-phenylalanine decreases withincreasing pHfrom 5.7to9.4andthiscanbe explainedbythestrongelectrostaticrepulsionadsorbent/amino acid molecules, negatively charged. This effect is similar to thatreportedpreviously(Alvesetal.,2013b;Goscianskaetal., 2013c);as consequence, the adsorptionof l-phenylalanine is inhibitedaboveandbelowpH5.7.Basedontheexperimental results,pH5.7wasselectedasanoptimumvalue.

3.2.4. Adsorptionkineticstudy

Several kinetic models were proposed to understand the behavior ofadsorbents andtostudy themechanismscontrol- ling the adsorption. In this study, the experimental data of l-phenylalanine adsorptionareexamined usingapseudo-first andpseudo-secondorderkineticmodel.Thepseudofirst-order kineticmodelisexpressedinitslinearformbythefollowing equation(Heetal.,2015):

ln(qe−qt)=lnqe−k1t (4)

(7)

Table2

Kineticparametersforl-phenylalanineadsorptiononACKandACZ.

Model Parameters Adsorbent

ACK ACZ

qe,exp(mgg⁻¹) 114.36 68.03

η(%) 56.89 33.85

Pseudo-firstorder qe,cal(mgg⁻¹) 62.49 45.01

k1(min⁻¹) 0.0194 0.0205

R² 0.982 0.995

q(%) 30.63 23.00

Pseudo-secondorder qe,cal(mgg⁻¹) 119.05 72.46 k2(gmg⁻¹min⁻¹) 0.0007 0.0009

R² 0.999 0.999

q(%) 2.65 4.00

Whilethepseudosecond-orderkineticmodelisbasedonthe assumptionofachemisorptionoftheadsorbateontheadsorbent (Heetal.,2015):

t q_t = 1

k2q²_e + t

q_e (5)

whereqtandqe(mgg⁻¹)aretheamountofaminoacidadsorbed attimet(min)andatequilibrium,respectively,k1(min⁻¹)and k2(gmg⁻¹min⁻¹)aretherateconstantofthepseudofirstorder andpseudosecondorderadsorptionmodels.

The best fit was validated on the base of the correlation coefficient(R²),the difference betweenthe experimentaland theoretical adsorptioncapacities andthe normalized standard deviationq(%)(SenGupta&Bhattacharyya,2011):

q(%)=

[(q_e,exp−q_e,cal)/q_e,exp]²

n−1 100 (6)

wherenisthenumberof datapoints,qe,exp andqe,cal arethe experimental and calculated equilibrium adsorption capacity values(mgg⁻¹),respectively.

Theaimofthiskineticstudywastofindtheappropriatemodel thatbetterdescribestheexperimentaldataandthatdetermines thekineticparametersofthemasstransferofl-phenylalanine.

ThedatawereexaminedbyusingEqs.(4)–(6).Thecalculated kineticparametersforl-phenylalanineadsorptiononactivated carbonsaregatheredinTable2.Theadsorptioncapacitiescal- culated from the pseudo second-order model are very close to the experimental ones as evidenced from the low values

q(%) andhighcorrelationcoefficient(R²>0.99).Although the R² values for the pseudo first-order model are all above 0.98(Table2),theeminent variances(the relativeerrorof l- phenylalanine ontoACK and ACZ are 45.35% and 33.84%, respectively)betweentheexperimentalandcalculatedadsorp- tioncapacitiesreflectthepoorfittingof thepseudofirst-order model.Hence,l-phenylalanineadsorptionkineticsonbothACZ andACKiswelldescribedbythepseudosecond-orderkinetics.

3.2.5. Adsorptionisotherms

Theadsorptionisothermisgenerallyappliedtoanalyzethe experimentaldata atequilibrium. The isothermswere further investigatedbyperformingbatchadsorptionexperimentsover

0 100 200 300 400 500 600 700 800

0 1 2 3 4 5 6

0 2 6 6 8 10 12

ACZ pH 2.0 ACZ pH 5.7 ACZ pH 7.2 ACZ pH 9.4 ACK pH 2.0 ACK pH 5.7 ACK pH 7.2 ACK pH 9.4

C_e (mg L^–1)

C_e (mg L^–1) Ce/qeCe/qe

Fig.9.Langmuirisothermforadsorptionofl-phenylalanineonACKandACZ atdifferentvaluesofpH(adsorbentdose=1gL⁻¹,agitationspeed=150rpm, contacttime=300min,temperature=20^◦C).

thepHrange(2–9.4)attheoptimaltemperatureof20^◦C.The curveswerefittedbythemostusedmodelsnamelytheLang- muirandFreundlichones.TheLangmuirmodelisbasedonthe assumptionthatthemaximumadsorptioncorrespondstoasatu- ratedmonolayerofadsorbatemoleculesonhomogeneoussites ofadsorbentwithaconstantenergy,andnointeractionbetween adsorbedspecies.Thelinearformisexpressedasfollows(Yang, Yu,&Chen,2015):

C_e

q_e = 1

qmaxK_L+ C_e qmax

(7)

whereqmaxandqearetheequilibriumandmaximumadsorption capacities (mgg⁻¹),respectively; KL, the Langmuir constant relatedtotheaffinityofthebindingsites(Lmg⁻¹);andCe,the equilibrium concentration of adsorbate inan aqueous phases (mgL⁻¹).The valuesof qmax andKL arecalculatedfromthe slopes (1/qmax)andintercept(1/qmaxKL)ofthelinearplotsof (Ce/qe)vs.Ce(Fig.9).

The maincharacteristics of the Langmuirisotherm canbe expressed by adimensionlessseparation factor,RL, whichis

(8)

definedbythefollowingformula(Liu,Zheng,Wang,Jiang,&

Li,2010):

R_L= 1 1+KLC0

(8) TheRLvalueindicatesthepossibilityoftheadsorptionpro- cessbeing favorable (0<RL<1), unfavorable (R1>1), linear (RL=1),orirreversible(RL=0).

TheFreundlichisothermmodel isbasedonheterogeneous adsorbentsurface(Yangetal.,2015):

lnqe=lnKF +1

nlnCe (9)

Thelinearplotoflnqevs.lnCe(Fig.10)enablesthedetermi- nationoftheFreundlichconstantsKFandnfromtheintercept andslope,respectively.

Theisothermparametersofl-phenylalanineare calculated from the Langmuir andFreundlich models (Table 3), all the R²valuesoftheLangmuirmodelaregreaterthan0.99.These values are much higher than thoseof the Freundlich model, whateverthepH,suggestingtheapplicabilityoftheLangmuir modelwhichrevealsamonolayercoverageofl-phenylalanine onhomogeneoussitesfor bothadsorbents.TheRL valuesare between0and1,indicatingthatthel-phenylalanineadsorption onACKandACZisfavorableundertheoperatingconditions.

AsthepHincreasesfrom2.0to9.4,qmaxshowsasignificant decrease,reflectingthattheadsorptionismorefavorableatpH 5.7, which is close to the isoelectric point (PI=5.48). ACK exhibitsamaximummonolayeradsorptionofl-phenylalanine (188.3mgg⁻¹) compared to ACZ (133.3mgg⁻¹) due to its largestmicroporoussurfacearea.Astheporesize ofACKis 1.48nm,mostporesincludingmicroporesareeasilyaccessible tol-phenylalaninewithasize of 0.7×0.5×0.5nm³.Conse- quently,themicroporoussurfaceareaplaysanimportantrolefor determiningtheadsorptioncapacityofthesmallbiomoleculel- phenylalanine.ThemaximumadsorptioncapacitiesofACKand ACZforl-phenylalaninearecomparedtothosevaluesreported

2 3 4 5 6 7

3.5 4.0 4.5 5.0 5.5

ACK pH 2.0 ACK pH 5.7 ACK pH 7.2 ACK pH 9.4

ACZ pH 2.0 ACZ pH 5.7 ACZ pH 7.2 ACZ pH 9.4

lnc_e

3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0

3.0 3.5 4.0 4.5 5.0

InqeInqe

Fig. 10.Freundlich isotherm for adsorption of l-phenylalanine on ACK and ACZ at different values of pH (adsorbent dose=1gL⁻¹, agitation speed=150rpm,contacttime=300min,temperature=20^◦C).

Table3

Isothermparametersforl-phenylalanineadsorptiononACKandACZ.

Model Adsorbent Parameters pH

2.0 5.7 7.2 9.4

Langmuir ACK qe,exp(mgg⁻¹) 144.07 176.01 154.68 122.66

qmax(mgg⁻¹) 151.97 188.32 170.06 137.74

kL(Lmg⁻¹) 0.0246 0.0262 0.0181 0.0146

RL 0.0483 0.0455 0.0646 0.0789

R² 0.999 0.998 0.998 0.993

ACZ qe,exp(mgg⁻¹) 84.01 107.45 96.94 75.65

qmax(mgg⁻¹) 99.40 133.33 119.33 91.99

kL(Lmg⁻¹) 0.0076 0.0063 0.0065 0.0073

RL 0.1412 0.1655 0.1613 0.1462

R² 0.988 0.993 0.993 0.997

Freundlich ACK kF 37.738 43.613 27.222 26.016

n 4.667 4.364 3.520 3.986

R² 0.963 0.952 0.911 0.921

ACZ kF 8.874 7.927 7.667 6.191

n 2.860 2.443 2.516 2.529

R² 0.978 0.985 0.980 0.944

(9)

Table4

Comparisonofthemaximumadsorptioncapacities(Qm)ofvariousadsorbents forl-phenylalanine.

Adsorbent Qm(mgg⁻¹) Reference

Polymericadsorbent 115.6 (GrzegorczykandCarta,1996) Polymericresins 65.9–100.8 (Díezetal.,1998)

Carbonatedcalcium phosphates

44.0 (Bihietal.,2002)

NAZSM-5zeolite 41.3 (Titusetal.,2003) Commercialactivated

carbon

100.0 (Garnieretal.,2007)

Organic-inorganic hybridmembranes

1.2 (Wuetal.,2009)

Sphericalcarbon aerogels

66.1 (Longetal.,2009)

Macroporousresins 12.8–84.0 (Meietal.,2009) Activateddefective

coffeebeans

69.5 (Clarketal.,2012)

Calcined CuZnAl-CO3

layereddouble hydroxides

46.4 (Jiaoetal.,2012)

Activatedcorncobs 109.2 (Alvesetal.,2013a,2013b) Mesoporousmaterials

CSBA-15,CSBA-16

andCKIT-6

0.27–0.30 (Goscianskaetal.,2013b)

Mesoporoussilica 36.0–69.0 (Goscianskaetal.,2013a) Mesoporouscarbon

CMK-3

273.0 (Goscianskaetal.,2014)

Multi-walledcarbon nanotubesCNTs

233.0 (Goscianskaetal.,2014)

Activatedcarbon ACK

188.3 Thisstudy

ActivatedcarbonACZ 133.3 Thisstudy

intheliteratureforotheradsorbents(Table4).In comparison withvariousadsorbents,ouractivatedcarbonsACKandACZ havehighadsorptioncapacitiesandcanbeconsideredaseffec- tiveadsorbentsfortherecoveryofl-phenylalaninefromaqueous solutions.

3.2.6. Adsorptionthermodynamics

A thermodynamic study was performed for the determi- nation of the free energy change (G^◦), entropy (S^◦) and enthalpy (H^◦). A previous study of the temperature effect onl-phenylalanineadsorptionoverACZandACKenabledus todeterminethe thermodynamic parametersat20–40^◦C and 800mgL⁻¹.Theywerecalculatedfromthefollowingequations (Bouguettoucha, Reffas, Chebli, Mekhalif,& Amrane, 2016;

Milonjic,2007):

K_C= qe

Ce (10)

ΔG⁰=−RTln(ρK_C) (11)

ln(ρK_C)=ΔS⁰

R −ΔH⁰

RT (12)

whereKCistheequilibriumconstant(Lg⁻¹),Ttheabsolutetem- perature(K),Rtheuniversalgasconstant(8.314Jmol⁻¹K⁻¹) andρ the densityof water (gL⁻¹).H^◦ andS^◦ are calcu- latedfromtheslopeandinterceptof theplotsof ln(ρKC)vs.

4.6 4.8 5.0 5.2 5.4 5.6 5.8 6.0

3.40 3.45 3.35 3.30 3.25 3.20 ACK ACZ

ln(ρKC)

1000/T(K^–1) 3.15

Fig. 11.Regressions of Van’t Hoff for thermodynamic parameters of l- phenylalanineadsorptiononACKandACZ.

1/T,respectively(Fig.11andTable5).Thenegativeenthalpies (H^◦)indicatetheexothermicnatureoftheadsorption,inagree- ment withthetemperature effectstudy(see Section3.2.2).It hasbeenreportedthatH^◦isintherange(2.1–20.9kJmol⁻¹), indicating a physisorption (Liu, 2009). H^◦ (−6.947 and

−7.153kJmol⁻¹ for ACK and ACZ, respectively) showed a physisorptionofl-phenylalanine,withweakinteractionswhile the negative free enthalpy (G^◦) indicates the spontaneous natureofl-phenylalanineuptakeoverthestudiedtemperature range.The variationof G^◦for physisorptionisintherange (0–20.9kJmol⁻¹), whereas this energy ranges from 80 to 200kJmol⁻¹ for a chemisorption (Liu, 2009). In our case,

G^◦ (Table 5) is characteristic of a physical adsorption.

The entropy (S^◦) is used to describe the randomness at the solid–solutioninterface during the recoveryprocess. The positive values of S^◦ demonstrate an increase in randomness during the adsorption of l-phenylalanine on ACK and ACZ.

3.3. Proposedmechanismofadsorption

To further understand the adsorption behavior and select a desorption approach, the adsorption mechanism of l- phenylalanine amino acid was discussed. The adsorption mechanismsoccurredmainlybecauseofthehydrogenbonding formation,hydrophobicandelectrostaticinteractionsofamino acid molecules withthe activated carbons surface.The main activesitesforbindingofl-phenylalaninebytheactivatedcar- bonsarethehydroxylandcarboxylgroupsonthesurfaceofACK andACZ,whichreactwithpolarmoleculesandvariousfunc- tionalgroups.Thesurfaceofporousactivatedcarboncaninclude electricallychargedgroups(ACZ/ACKsurface–OH2+:below pHpzc)and(ACZ/ACKsurface–O⁻:abovepHpzc),electrically neutral groups (ACZ/ACK surface –OH: near pHpzc). l- phenylalanineaminoacidhasdissociationconstants(pK1=1.83 and pK2=9.13) and isoelectric point (PI=5.48) (Jiao et al., 2012).Themoleculeispositivelycharged(⁺NH3–R–COOH)for pH<PI andnegativelycharged(NH2–R–COO⁻) forpH>PI,

(10)

Table5

Thermodynamicparametersforl-phenylalanineadsorptiononACZandACK.

Adsorbent H^◦(kJmol⁻¹) S^◦(kJmol⁻¹K⁻¹) G^◦(kJmol⁻¹)

293K 298K 303K 313K

ACK −6.947 0.019 −13.820 −13.930 −14.184 −14.280

ACZ −7.153 0.017 −12.300 −12.418 −12.604 −12.646

Electrostatic repulsion force

pH above 5.7 and pH_PZC pH=5.7 ≈PI pH below 5.7 and pH_PZC

Hydrogen bending Electrostatic

repulsion force

π−π Interaction Activated carbon surface

NH₂

H₂N

H₃N+

H₂O+

O

OH OH

O

O–

Fig.12.Proposedadsorptionmechanismofl-phenylalanineontoACZandACKadsorbents.

and behaves in an aqueous medium as a dipolar zwitterion (⁺NH3–R–COO⁻) at pH–PI. In acid solution, the presence of H3O⁺ ionsinthe surface of ACZ andACK causes repulsion of protonated amino groups withthe surface functional groups, and thus lower the adsorption efficiency. In a basic solution,OH⁻ ionspresentontheadsorbent surfacecompete withanioniccarboxylicgroupsforl-phenylalaninemolecules (repulsioneffect)andinhibittheadsorption.Thehighestuptake ofl-phenylalanine atpH5.7indicates adominant hydrophobic interaction with ␲–␲ type between the phenyl rings of aminoacidmoleculesandgrapheneringsoftheactivatedcar- bonssurface(Doulia,Rigas,&Gimouhopoulos,2001;Rajesh, Majumder,Mizuseki,&Kawazoe,2009).Electrostaticattrac- tion between anionic carboxylic groups of l-phenylalanine molecules and OH⁻ ions in the surface of activated car- bonsalsoaccountsfor theincreasedadsorption.Additionally, suchstrong bindings betweenl-phenylalanine andtheadsor- bentsurface can be explained by the formation of hydrogen binding between oxygenated groups at the activated car- bonssurfaceandaminogroupsofl-phenylalanine.According to these results, the adsorption mechanism is proposed in Figure12.

3.4. Desorptionbehaviorofl-phenylalaninefromactivated carbons

Regeneration and reuse of adsorbents for further cycles is important from the economic perspective. Desorption of l-phenylalanine from the activated carbons was evaluated using two different eluents: NaOH and HCl (0.01M). The highest desorption was achieved in the NaOH solution with almost 95.7% for ACZ and 88.8% for CAK against 21 and 6.5% in the HCl solution. This may be due tothe enhance- ment of the number of negatively charged sites at high pH whichincreasestheelectrostaticrepulsion, whichliberatesl- phenylalanine from ACKand ACZ. To check the adsorption efficiency,thedesorbedACKandACZweredriedovernightand subjectedtoanewadsorption/desorptioncycle.Duringthesec- ondcycle,theadsorptioncapacitiesobtainedwere29.4(ACK) and 23.5mgg⁻¹ (ACZ). A significant decay in the adsorp- tioncapacityofbothactivatedcarbonswasobservedandmay beattributed tothe depletionof activesitesofthe adsorbents being occupied by the amino acid. With the increase of the repeatedcycle,therateofdesorptionwasalsogreatlydecreased (Fig.13).