Availableonlineatwww.sciencedirect.com
Journal of Applied Research and Technology
www.jart.ccadet.unam.mx JournalofAppliedResearchandTechnology15(2017)27–35
Original
Sorption of cyanide from aqueous medium by coffee husk: Response surface methodology
Mebrahtom Gebresemati
a, Nigus Gabbiye
b, Omprakash Sahu
b,∗aDepartmentofChemicalEngineering,EiT-M,MekelleUniversity,Ethiopia
bSchoolofChemicalandFoodTechnology,BiT,BahirDarUniversity,Ethiopia Received29March2016;accepted15November2016
Availableonline16February2017
Abstract
Cyanideisaverytoxiccompoundthatisreleasedintotheenvironmentthroughindustrialeffluents;therefore,itneedstoberemovedprior towastewaterdischarge.Adsorptionisoneofthemostwidelyusedtreatmentmethodsforremovingcontaminantsfromwaterandwastewater.
Inpresentresearchworkresponsesurfacemethodologywassuccessfullyemployedforoptimizationand analysisof theadsorptionprocess.
Batchmodeexperimentswereperformedtodeterminetheadsorptionequilibriumofcyanideinaqueoussolution.Atoptimumconditioninitial concentration(10mg/l),contacttime(1h),adsorbentdose(1g)andpH(8),90.6%ofcyanideadsorptionwasobtained.Thefactorialdesignand RSManalysisdemonstratedthattheexperimentalandthepredictedvaluesfromthemodelswereinagreement.
©2017UniversidadNacionalAutónomadeMéxico,CentrodeCienciasAplicadasyDesarrolloTecnológico.Thisisanopenaccessarticleunder theCCBY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Keywords:Adsorption;Biomass;Pollutant;Wastewater;Toxic
1. Introduction
Organicpollutantshaveadverseeffectonhumanhealthand environment.Cyanideisamongoneofthem,andlistedastoxic pollutants. It is a singly-charged anion containing unimolar amountsofcarbonandnitrogenatomstriply-boundedtogether (Giraldo&Moreno-Piraján,2010).Cyanideisanatural com- poundthatisproducedbylivingorganismsincludingbacteria, fungi,algaeandplantsaspartofadefencemechanismagainst predation (Mekuto, Jackson, & Ntwampe, 2013). However, thesenaturalsourcesofcyanideareinsignificantwhencompared tocyanideproductionbyanthropogenicactivities.Mostlymin- ing,mineralprocessing, electroplating andplasticsindustries havesignificantlycontributiontocyanidecontainingwastewater intheenvironment(USEPA,1994).Uncomplexedcyanide,also calledfreecyanide,isthemosttoxicformofcyanideinwaste- waterbecauseit isreleasedasgaseoushydrogen(Osobamiro,
∗Correspondingauthor.
E-mailaddress:[email protected](O.Sahu).
PeerReviewundertheresponsibilityofUniversidadNacionalAutónomade México.
2012).The effects are so momentoustocause nervedamage andthyroidglandsmalfunctioningandestablishedtoxicitylevel (ASTM,2006).Asanenvironmentallyprotectivemeasure,the Environmental Protection Agency (EPA)has placed rigorous limitations(0.1mg/l) onthepermittedcyanide concentrations levelsinindustrialwastewatereffluentstreams(Abbas,Abbas,
&Ibrahim,2014).
To treat the cyanide contained industrial effluent gener- allychemicaloxidationtechniqueshasbeenused(Botz,2001;
Dash,Gaur,&Balomajumder,2009;Parga,Shukla,&Carrillo- Pedroza,2003).Someothertreatmenttechnologieshavebeen also introduced to treat the cyanide-containing wastewaters (Abbasetal.,2014;Hanela,Durana,&Jacobob,2015;Palmer, Breton,Nunno,&Sulivan,1988).However,thesemethodsare expensiveandhazardouschemicalsareusedasthereagentspro- cesses involving sorptionhave received greater attention and mostwidelyacceptedmethodfortoxiccontaminatedwastewa- ter(Agarwal,Balomajumder,&Thakur,2013;Singh,Agarwal,
&Balomajumder,2016).Theadsorptionshowsremarkableper- formance,including lowinvestmentcost,simplicityof design andoperation,insensitivitytotoxicantsandeffectivenesswith lowconcentrationofadsorbent(Mohammad,2013).Inliterature
http://dx.doi.org/10.1016/j.jart.2016.11.002
1665-6423/©2017UniversidadNacionalAutónomadeMéxico,CentrodeCienciasAplicadasyDesarrolloTecnológico.Thisisanopenaccessarticleunderthe CCBY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/4.0/).
some agricultural waste and by product has been used as absorbenttotreattheindustrialeffluent(Anastopoulos&Kyzas, 2014; Prasad & Santhi, 2012; Singh et al., 2016). The use of biosorbents shows high removal due to primary building blockofplantcellwalls,lignocellulosecontainscellulose,hemi- cellulose, lignin including little amount of protein, vitamins, lipids, combined with ash (Jorgensen, Kristensen, & Felby, 2007). Since, cyanide is a toxic compound well-known as a metabolicinhibitor,cyanide-containingeffluentscannotbedis- charged without being subjected to treatment in response to increasinghealthandenvironmentalawareness.
Themain aimof thisresearchworkis toanalyzethe effi- ciencyof coffee husk(adsorbent) for the removalof cyanide fromsyntheticaqueoussolutioninbatchreactors.Theadsorp- tionisusedforlowconcentrationsofcyanide.Theoptimization of initial concentration, contact time, mass loading and pH were deliberated via response surface method experimental design.Preparedbioadsorbenthasbeencharacterizedbyscan- ningelectronmicroscopy(SEM)andFouriertransformedinfra red(FTIR).
2. Materialandmethods 2.1. Chemicals
All laboratory grade chemicals sodium cyanide, sodium hydroxide, hydrochloricacidwere usedwithoutfurthertreat- ment and supplied by Himedia Laboratories Pvt., Mumbai, India.
2.2. Sample
To avoid interference with other elements in wastewater, theexperimentsinthisstudywerecarriedoutusingsimulated syntheticaqueoussolutionof differentcyanideionconcentra- tions.Topreparestocksolutionknowamountweightofsodium cyanidedissolvedin1Lofmilliporewater(Q-H2O,Millipore Corp.withresistivityof18.2MX-cm).Thecyanideionconcen- trationsweremeasuredusingUV-spectrophotometer.
2.3. Preparationofadsorbent
Thecoffeehuskwascollectedfromprocessingunit.Itwas clean by washing with distilled water. Dilute sulfuric acid, 2% (v/v)anddiluted sodiumhydroxide, 2% (v/v)were used to pretreat 20g of ground coffee husk biomass. Treatments were performed intriplicate at an autoclaved at 121◦C with
Table1
Characteristicofcoffeehusk.
S.no Characteristics Values
1 Specificgravity 0.71
2 Bulkdensity(kg/m3) 395
3 Porosity(%) 79
4 Meanporeradius(A◦) 5.4
5 Surfacearea(m2/g) 0.39
6 Moisturecontent(%) 17.1
7 BETsurfacearea(m2/g) 910
15psi (103.4kPa) for 90min. Sampleswere neutralizedwith hotdeionizedwateranddriedat60◦Covernighttocompletely removemoisture,thencooledtoroomtemperatureandstoredin polybagsuntilfurtheruse.Theadsorbentwasusedinitsoriginal piecesize.Thechemicalpropertyofcoffeehuskismentionin Table1.
2.4. Experimentaldesign
Atotalof30experimentshavebeenemployedinthiswork to evaluate the individual and interactive effects of the four main independent parameters onthe cyanide adsorption effi- ciency (Montgomery, 2001). Percentage adsorption has been takenasaresponse(Y)ofthesystem,whilefourprocessparam- eters,namely,initialconcentration5–40mg/l;pH:2–10;contact time1–3handadsorbentdose0.5–5ghasbeentakenasinput parameters. Forstatistical calculations,thelevels for thefour mainvariablesX1(IC),X2(t),X3(g),X4(pH) werecodedasxi
accordingtothefollowingrelationship:
xi= Xi−X0
δX (1)
whereX0isvalueofXiatthecenterpointandδXpresentsthe stepchange.Thevariablesandlevelsof thedesignmodelare giveninTable2.TheresultsoftheY(response)ofadsorption were measured according todesign matrix listed in Table 3.
Fromexperimentalobservations,itwasassumedthatthehigher orderinteractionsweresmallrelativetotheloworder.
Experimentalprocedurebatchexperimentsforoptimization ofprocessparameterswerecarriedoutin250mLroundbottom flaskswithworkingvolumeof100mLat120rpminanincubator cumorbitalshaker.Ameasureamountofadsorbatewasaddedin measureinitialconcentratedofcyanidesolution.Theexperiment carriedoutinroomtemperature(21◦C).ThepHofsolutionwas readjustedwith1NNaOHorHClduringoperation.Theamount
Table2
Factorsandlevelsoftheexperimentaldesignforadsorption.
Factors Level(−2α) Level(−α) Level(0) Level(−α) Level(+2α)
Inletconcentration(mg/l) 5 10 20 30 40
pH 2 4 7 8 10
Contacttime(H) 1 1.5 2 2.5 3
Adsorbentdose(g) 0.5 1 3 4 5
Table3
Thedifferentcombinationofthefactorsfortheexperimentaldesign.
Runs X1(initialconcentration,mg/l) X2(contacttime,h) X3(adsorbentdose,g) X4(pH) Y(adsorption%)
1 0 0 0 0 81
2 0 0 0 −2 98
3 1 −1 1 1 91
4 −1 −1 1 −1 51
5 0 0 0 0 81
6 −1 1 −1 −1 98
7 −1 1 1 1 98
8 0 0 0 0 98
9 −2 0 0 0 98
10 0 0 0 0 81
11 0 0 0 0 81
12 1 1 1 −1 96
13 −1 −1 1 1 77
14 2 0 0 0 81
15 0 0 0 2 17
16 −1 −1 −1 1 95
17 1 1 −1 1 57
18 0 0 0 0 81
19 1 −1 −1 1 38
20 1 −1 1 −1 69
21 0 0 −2 0 69
22 −1 1 −1 1 87
23 1 1 −1 −1 96
24 −1 −1 −1 −1 92
25 0 −2 0 0 81
26 1 1 1 1 38
27 −1 1 1 −1 98
28 1 −1 −1 −1 55
29 0 2 0 0 96
30 0 0 2 0 69
ofcyanideadsorbedperunitmassoftheadsorbentwasevaluated bythefollowingmassbalanceequation:
q= V(C0−Ct)
M (2)
whereC0istheinitialpollutantconcentration(mg/l)andCtis thepollutantconcentration(mg/l)attimet,Visthevolumeof thesolution(L)andMisthemassoftheadsorbentused(g).
Theadsorption efficiency was calculatedby the following equation:
Removal (%)=(C0−Ct)×100 C0
(3) whereC0isinitialconcentrationofcyanideinsolutionandCt
isfinialconcentrationintime(t).
2.5. Analysis
Fouriertransforminfrared(FTIR)spectroscopicanalysiswas performed(FTIR-2000,PerkinElmer)todeterminefunctional groupinadsorbent.The spectraweremeasured from4000to 500cm−1. The surface area, total pore volume, and average porediameterofthesamplesweredeterminedfromtheadsorp- tionisothermsof nitrogenat77Kusing AutosorbI,supplied byQuantachromeCorporation,USA.Thesurfacemorphology of the sample was examined using scanning electron micro- scope(ModelVPFESEMSupra35VP).Proximateanalysiswas
carried outusing thermo gravimetricanalyzer(TGA)(Model PerkinElmerTGA7,USA).
3. Resultanddiscussion
3.1. Modelfittingandstatisticalanalysis
Thecentralcompositedesign(CCD)wasemployedtofita second-orderpolynomialmodelandtoobtainanexperimental error.TheCCDwasappliedwithfourdesignfactorsandfive levels,tostudytheaffectofinitialconcentration,contacttime, adsorbentdosageandpHonadsorptionefficiencywithcoffee husk.Theresultsof theY(response)ofadsorptionweremea- sured accordingtodesignmatrixandthemeasured responses are listed in Table 3. Linear, interactive, quadraticandcubic modelswerefittedtotheexperimentaldatatoobtaintheregres- sion equations. Two different tests namely sequential model sumof squaresandmodel summarystatistics wereemployed to decide about the adequacy of various models torepresent adsorptionwithcoffeehusk.Resultsofthesetestsaregivenin Tables4and5,forpercentageadsorptionremovalrespectively.
Cubicmodelwasfoundtobealiased.Forquadraticandlinear models,pvaluewaslowerthan0.02,andbothofthesecouldbe usedforfurtherstudyaspersequentialmodelsumofsquares test. As per model summary statistics, the quadratic model wasfoundtohavemaximumadjustedR20.73,andpredicted
Table4
Sequentialmodelsumofsquares.
Source Sumofsquares Df Meansquare Fvalue p-Value
Prob>F
Remark
Meanvstotal 160,747.2 1 160,747.2
Linearvsmean 5773 4 1443.25 3.72056 0.0165 Suggested
2FIvslinear 2502.75 6 417.125 1.101504 0.3974
Quadraticvs2FI 1354.05 4 338.5125 0.869318 0.5049
CubicvsQuadratic 3772.5 8 471.5625 1.595812 0.2758 Aliased
Residual 2068.5 7 295.5
Total 176,218 30 5873.933
Table5
Modelsummarystatistics.
Source Std.Dev. AdjustedR2 PredictedR2 PRESS Remark
Linear 0.004591 0.737311 0.7285453 0.580842 0.0012024 Suggested
2FI 0.004937 0.783492 0.6773902 −0.3752962 0.0021785
Quadratic 0.005352 0.62245 0.27007 −0.7979303 0.0023669
Cubic 0.002031 0.766297 0.446086 −2.7872128 0.009409 Aliased
R2 0.59 values excluding cubic model which was aliased.
Therefore,quadraticmodelwaschosenforfurtheranalysis.To determinewhetherornotthequadraticmodelissignificant,it iscrucialtoperformanalysisofvariance(ANOVA)mentionin Table6.Theprobability(p-values)valuesareusedasadevice tocheckthesignificanceofeachcoefficient,whichalsoshows theinteractionstrengthofeachparameter(smallerthep-values biggersignificanceofthecorrespondingcoefficient)(Mohana, Shrivastava,Divecha,&Madamwar,2008).
Inadditiontoanalyzingtheindependentvariables’effects, thisexperimentalmethodologyalsogeneratesamathematical model.Thegraphicalviewpointofthemathematicalmodelhas ledtotheterm RSM.The relationshipbetweentheresponses andtheinputsisgiveninthefollowingequation:
Y =f(X1,X2,X3,X4,...,Xn)±ε (4)
whereYistheresponse;fistheunknownfunctionofresponse;
X1,X2,X3,...,Xnaretheinputvariables,whichcanaffectthe response;n isthenumberof theindependentvariables,andε isthestatisticalerrorthatrepresentsothersourcesofvariability notaccountedforbyf.
Afterselectionofthedesign,themodelequationisdefined and coefficientsof the model equation are predicted. Design summaryfor adsorption withfivelevels andfourfactors;the designmodeloftheexperimentsisquadraticpolynomialandthe centerpointiszerousingdesignexpert6.0.8software.Amanual regressionmethodwasusedtofitthesecondorderpolynomial givenbyEq.(5),respectivelytotheexperimentaldataandto identify therelevantmodelterms.Thefinalequation obtained intermsofcodedfactorsisgivenbelow:
percentageofadsorption
(Y)=73.2−9.1766666667X1+6.333333333X2
−0.916666667X3−10.75X4 (5)
Thestatisticalsignificanceoftheratioofmeansquarevari- ation due to regression and mean square residual error was testedusingANOVA(Myers,Montgomery,&Anderson-Cook, 2016).TheModelF-valueof3.72impliesthemodelissignif- icant. There is onlya1.65% chance that a “ModelF-value”
thislargecouldoccur duetonoise. Valuesof“Prob>F”less than0.05indicatemodeltermsaresignificant.Inthiscaseinitial
Table6
Analysisofvariance.
Source Sumofsquares Df Meansquare Fvalue p-Value
Prob>F
Remark
Model 5773 4 1443.25 3.7206 0.0165 Significant
X1 2016.667 1 2016.667 5.1988 0.0314
X2 962.6667 1 962.6667 2.481663 0.1278
X3 20.16667 1 20.16667 0.052 0.8215
X4 9697.8 1 2773.5 7.1498 0.0130
Residual 7966.3 25 387.912
Lackoffit 1731.5 20 398.315 1.1502 0.04812 Notsignificant
Pureerror 15,470.8 5 346.3
Cortotal 9697.8 29
concentrationandpHaresignificantmodelterms.Valuesgreater than 0.10 indicate the model terms are not significant; this includescontacttimeandadsorbentdose.The“F-value”1.15 of lack of fit implies that model is not significant relative to thepureerror.Thereisa4.81%chancethata“F-value”could occurduetonoise.Non-significantlackoffitisgoodbecausethe modelhastofit.Insummary,theANOVAanalysisdemonstrated thatthequadraticregressionmodelcouldeffectivelypredictthe adsorptionprocessofphenolicwastewater.
3.2. Effectoffactorsontheresponse
The percentage adsorption of cyanide increased with an increaseincontacttimeandtheequilibriumtimevariedfordif- ferentconcentrations, whichpresented inFigure 1. The time required toattain equilibriumfor 10 and 20mg/l of cyanide concentrationwas2h.Themaximumpercentageadsorptionare 99%for5mg/l,10mg/l;98%for20mg/l,96%for30mg/land 89%for40mg/l.Thisisfactthatatthebeginningnoofvacant siteare more withincrease intimeand concentration vacant site were disappeared. The same trend also observed when coconuthuskwasusedat30mg/lcyanideinitialconcentration
100
80
60
40
20
0
0 0.5 1 1.5 2
Time (hr)
5mg/l 10mg/l 30mg/l 20mg/l 40mg/l
Percentage of adsorption
2.5 3 3.5
Fig.1.Effectofinitialconcentrationofcyanidewithcontacttime.
(Singhetal.,2016),granularactivatedcarbonat100mg/lini- tial concentration (Agarwalet al.,2013) andeucalyptusbark for zincat70ppminitialconcentration(Afroze,Sen, &Ang, 2016).
FromFigure2(a)itwasobservedthatastheconcentration increased from5 to40mg/l,the rate ofadsorption decreased
50 60 70 80 90 100
1 1
0 –1
–1 –1 –1
0 1
1
Initial concentration Contact time
1 1
0 –1
–1 –1
–1 0 1 1
Initial concentration pH
Adsorption, %
50 60 70 80 90 100
Adsorption, %
1 1
0 –1
–1 –1
–1 0 1
1
Initial concentration Adsorption dose
50 60 70 80 90 100
Adsorption, %
a b
c
Fig.2.(a)Effectofinitialconcentrationofcyanidewithcontacttime(3D).(b)EffectofinitialconcentrationofcyanidewithpH(3D).(c)Effectofinitialconcentration ofcyanidewithdose(3D).
atlow level from93 to47%.In this waycyanide concentra- tionshowedaninverserelationwithadsorption,i.e.maximum adsorptionoccursatminimumconcentrationofcyanide.Itmay beduetogreaternumberofionsinthesolutioncausesmorenum- berofcollisionsandthusleadingtodesorptionofionsfromthe bindingsitesofadsorbentparticles.Whereaswithlimitednum- berofionstherearelimitednumberofcollisionbetweentheions thatiswhylowconcentrationshowedhigherratesofadsorption (Vedula,Dalal,&Majumder, 2013).TheinitialpHofsample playsignificantroleonremovalefficiencyofcyanidesolution, it was adjusted between pH 2 and 10 shownin Figure 2(b).
MaximumadsorptionofcyanideionobservedathigherpHof sample.Itattributestheinteractionsofionswithadsorbedfunc- tionalgroupsontheadsorbentsurfacesthoseareweakerthan interactionswith the surface hydroxylsof the sorbent. Atan optimumpHof10,over99%removalof100mg/lcyanidewas obtained for (adsorbent) pistachio hull wastesdose of 1.5g/l after a60min contacttime(Moussavi &Khosravi,2010).In literatureauthor foundthat the adsorptionof sodiumcyanide wasincreasefrom45.2%to61.7%atpH4–11and100mg/lof initialconcentrationincommercialactivatedcarbon.Thismay be duetopHof the solution,influencethe adsorbent surface chargeand the ionization degreewith speciationof different pollutants(Dash,Dash,&Balomajumdar,2014).AtalowpH thede-protonationoncoffeehusk surfaceprovidesfunctional groups,forchemisorptionsonitssurfacethatcanundergoion exchange type of interaction with cyanide ions, due to that lowerpercentageofadsorptionvaluesobserved.Solutesinteract with sorbent surfaces because they haveacquired an electri- cal surface charge (Stavropoulos, Skodras, & Papadimitriou, 2015).
Effectofcoffeehuskdosageonadsorptionwasinvestigated bychangingthesorbentdosefrom0.5to5gwithdifferentini- tialconcentration,showninFigure2(c).Theremovalefficiency ofcyanideincreasedwhensorbentdoseincreasedfrom1to3g forinitialconcentrationof 5,10and20mg/land4g dosefor initialconcentrationofcyanide30and40mg/l.Atalowsorbent dose,e.g.0.5g,theavailablesorptionsiteswerequiteinsuffi- cientcomparedwiththelargeamountofcyanideionsinsolution, resultinginlowremovalefficiency (Gupta,Balomajumder,&
Agrawal,2012).Atahighersorbentdose,e.g.5g,thesorption sitesweresufficientandafurtherincreaseinsorbentdosedid notleadtoasignificantincreaseinthecyanideremoval from aqueoussolution.Thismaybeduetoaccumulationofbiosorb- ent particle resulting in decrease in surface area existing for biosorption (Kilic,Apaydin-Varol,& Pütün,2011).Achieve- mentofhighcyanideremovalpercentagewitharelativelylow adsorbentdoseindicatesthehighaffinityandsuitabilityofcof- feehuskforremovalofcyanidefromwastewater.Theincreasing adsorptionefficiencywithincreasingcoffeehuskdose canbe attributedtothe increaseinsurfaceareaandbyextensionthe greaternumberofexchangeable sitesavailablefor interaction withcyanideions(Moussavi&Khosravi,2010).Similarlyby usingcoffeehusk(97%adsorption),coconutshellandcommer- cialactivated(99%adsorption)atinitialcyanideconcentration of260mg/l,doseof10g/l,andapHof11.0wasobserved(Halet etal.,2015).
1
b 1000
a
0.800 0.600 0.400
0.200 0.000
1 0
–1
–1 –1 –1
0 1
1
Initial concentration Contact time
1 1
0 –1
–1–1 –1
0 1
1
Initial concentration Contact time
50 60 70 80 90 100
Adsorption, %Desirability
Fig.3. Graphicrepresentationofthe(a)desirability3Dplot(b)optimized percentageadsorptionforsawdust.
3.3. Optimization
Inthepresentstudy,fornumericaloptimization,theoptimum responseresultwas90.6%,percentageadsorption.Theoptimum processing conditions using numericaloptimization were the codedlevels(−1,−1,−1,−1)orinitialconcentration(10mg/l), contacttime(1h),adsorbentdose(1g)andpH(8)andisshown inFigure3(a).Theresponsepredictiondesirabilityisfoundtobe verygood(0.976)ascomparedtotheultimatedesirabilityof1.
Thedesirabilityisplottedusing3DplotasshowninFigure3(b).
3.4. Adsorptionisothermmodel
An isothermequilibriummodel representstherelationship betweenthecyanideadsorbedontocoffeehusk(adsorbent)at givenexperimentalconditions.The equationused foradsorp- tion isotherm indicatesthe diffusionofadsorbate fromliquid phase to the solidphase at equilibriumcondition (Al-Asheh, Banat, Al-Omari,& Duvnjak,2000). Thusthe correlation of equilibrium data is required for interpretation. Different iso- therm model namely the Langmuir, Freundlich, and Temkin, were appliedtofit theexperimentalisotherm dataof cyanide adsorption on activatedcarbons has beenproposed (Adhoum
&Monser,2002;Dashetal.,2014).Amongtheentiremodel
6
4
2
0
1 2 3
Ce
R2=0.975
Ce/qe
4 5
Fig.4.ThelinearLangmuiradsorptionisothermforcyanidewithcoffeehusk.
Langmuiriswidelystudiedforadsorptionandusedforpresent study. The non-linear form of the Langmuir isotherm model is represented by the subsequent equation (Langmuir, 1918) mentioninthefollowingequation:
Ce qe = 1
Q0b + Ce Q0
(6) where ‘Ce’ is the equilibrium concentration and ‘qe’ is the amountofadsorbateadsorbedpergramofadsorbentatequilib- rium;‘Q0’and‘b’areLangmuirconstantsrelatedtothesorption capacityandintensityrespectively. The experimentalequilib- riumdataforcyanideadsorptiononpreparedactivatedcarbons at21◦C,calculatedfromEq.(6).ByplottingCe/qeversusCe, bandQ0canbe determinedwhen astraightline isobtained.
ThevalueofQ0andbvalueswerecalculatedfromtheslopeand interceptofthestraightline,respectively.Theequilibriumofthe adsorbentwasfittedverywelltotheLangmuirisothermshown inFigure4.TheLangmuiradsorptionisothermofcyanideon coffeehuskgavethecorrelationcoefficientwas0.975.Theval- ueswerefoundtofollowtheLangmuirisothermtosomeextent, dependingonthepHmostly(Foo&Hameed,2010).Thefea- sibilityoftheLangmuirmodelwasfurtherassessedusingthe equilibriumdimensionlessparameterseparationfactor(RL).
RL= 1 1+bC0
(7) ThevaluesofRL definethefavorabilityofadsorptionpro- cess(Hall,Eagleton,Acrivos,&Vermeulen,1966).IftheRL=1 indicated linear, RL>1 indicates unfavorable, RL=0<RL<1 indicates favorable and RL=0 indicates irreversible. In this experimentallthevalue ofRL wasfoundtobeless thanone forinitialconcentration5mg/lto40mg/lofcyanide.
3.5. Characteristicsoftheadsorbent
The percentage composition of coffee husk is mention in Table 7 and shown inFigure 5. The peak shows 35.98% of cellulose 17.8% hemicellulose, 25.4% lignin, ash 2.4% and otherelement.Duetothisselfdevelopedbiochemicalproperty ofcoffeehusk;cyanideaccumulatesfromwastewaterthrough metabolicallymediatedpathwaysofuptake.FromTable7itcan alsoseethathydrogenisavailableinrelativelylargepercentage
Table7
Physicochemicalcharacteristicofcoffeehusk.
S.no Components Weightpercentage
1 Totalcarbonate 51.7
2 Totalnitrogen 1.1
3 Cellulose 35.98
4 Hemicellulose 17.8
5 Fixedcarbon(%) 16.2
6 Carbon(%) 45.33
7 Hydrogen(%) 6.21
8 Nitrogen(%) 0.63
9 Oxygen(%) 43.7
10 Lignin 25.4
11 Ash 2.4
12 Volatilematter(%) 82.7
2500
2000
1500
1000
500
00
5 10 15
Hemicellulose
Lignin Cellulose
20 2-theta (Degree)
Intensity (Counts)
25 30 35
Fig.5.X-raydiffractionstudyofcoffeehusk.
in comparison to nitrogen compounds indicates that carbon- hydrogengroupsmightbeavailableforadsorptionofcyanide.
Therelativelylowpercentageofnitrogenshowsthatveryless percentage of protein might be present in the husks. This isadvantageous overprotein-richadsorbentssinceportentous materialsarelikelytoputrefyundermoistconditions.
Scanningelectronmicroscopy(SEM)isextensivelyusedto studythesurfacemorphologyandcharacteristicsofthebiosorb- ent. In thispresentstudy, the surfacemorphologies of coffee husk before and after adsorbing cyanide were associated by SEManalysis. The smoothmorphology andporousstructure ofCHshowninFigure6(a)and(b)makeitsuitablebiosorbent asitincreasesthebiosorptioncapacity.TheSEMmicrographs Figure6(a)indicatestheoccurrenceofnumerousporesonthe biosorbentsurface,poressizemaybeabout1–2mofdiame- tersandaunevenstructureontosurfacewithahugesurfacearea.
FromFigure6(b),itcouldevaluatethatsmallparticleadhereto theCHsurface.Thismaybeduetotheoccurrenceofcyanide ontoCHsurface.Thisisassumedtobeeffectivebiosorptionof phenolandcyanideontocoffeehusksurface.
Biomasshasbecomeapromisingalternativesourceofmate- rialsforindustrialapplications(Xu,Yu,Tesso,Dowell,&Wang, 2013). The FTIR spectrum of the coffee husk show charac- teristics bandof lignocellulosicmaterials,showninFigure 7.
AstrongbroadO Hbandin∼3350cm−1 (lignin);abandin
300 µm WD=7.5 mm
300 µm WD=7.5 mm
EHT=10.0kV Signal A=SE1 Max=50% Time 12:45:00 EHT=10.0kV
Max=50% Time 12:45:00 Signal A=SE1
a b
Fig.6.SEMofcoffeehusk(a)beforeadsorptionand(b)afteradsorptionprocess.
100.0
80.0
60.0
40.0
20.0
4000 3500
3750 2650 1100
1750 950
3000 2500 2000
Before used After used
Wavenumber (cm–1)
Transmittance, %
1500 1000 500
Fig.7.Fouriertransforminfraredstudyofcoffeehusk.
∼2650cm−1(lignin),revealedtoaC H;abandin∼1750cm−1 ketone group (hemicellulose), revealed to a C O; a band in
∼1582cm−1 revealed to a C C stretching vibrating in aro- matics (cellulose, hemicellulose) and a band in ∼1100 to 950cm−1revealed to aC O of cellulose,hemicelluloses and lignin.TheseregionsalsoindicatethepresenceofC Nstretch- ingvibrating.Thesebandsindicatethepossibleinvolvementsof thesefunctionalgroupsonthesurfaceofcoffeehuskincyanide adsorption process (Sathishkumar, Arulkumar, & Palvannan, 2006).Coffeehusksurfacechemistrywerefoundtobedifferent as someof thefunctionalgroupsdisappearedduetotheacti- vationprocess,whichshowsthatthesefunctionalgroupswere thermallyunstable.FromFigure7aftertreatmentcurve,there isadecreaseintheabsorptionbandofaromaticgroupfromCH duetooxidativedegradationofaromaticringsduringchemical impregnationandheattreatmentstages.
4. Conclusion
From the aboveresult it maybe concludingthat activated coffee husk is effective for cyanide adsorption.On the basis ofRSMapproachforexperimentaldesignandfitnessofpoly- nomialequation,optimalconditionsmaximum90.6%cyanide adsorptionwerefoundat10mg/linitialconcentration,1hcon- tacttime, 1g adsorbent dose andpH 8.Analysis of variance
showedahighcoefficientofdeterminationvalue(Ad-R2=0.73 andPre-R2=0.59),thus,ensuring asatisfactoryadjustmentof thesecond-orderregressionmodelwiththeexperimentaldata.
The FTIR study shows reduction infunctional group incof- feehuskafterapplication.Scanningelectronmicrographshows highporousstructureofcoffeehusk.TheXRDspectrashowthe availabilityofcellulose,hemicelluloseandotherelementonthe surfaceofcoffeehusk.Theoutcomeofthisresearchworkisthat thepresentbiosorbent(coffeehusk)canbeusefulforthecyanide adsorptionfromeffluentsdischargedatlowconcentration.
Conflictofinterest
Theauthorshavenoconflictsofinteresttodeclare.
References
Abbas,M.N.,Abbas,F.S.,&Ibrahim,S.A.(2014).Cyanideremovalfrom wastewaterbyusingbananapeel.JournalofAsianScientificResearch,4(5), 239–247.
Adhoum,N.,&Monser,L.(2002).Removalofcyanidefromaqueoussolution usingimpregnatedactivatedcarbon.ChemicalEngineeringandProcessing:
ProcessIntensification,41(1),17–21.
Afroze,S.,Sen,T.K.,&Ang,H.M.(2016).Adsorptionremovalofzinc(II)from aqueousphasebyrawandbasemodifiedEucalyptussheathianabark:Kinet- ics,mechanismandequilibriumstudy.ProcessSafetyandEnvironmental Protection,102,336–352.
Agarwal,B.,Balomajumder,C.,&Thakur,P.K.(2013).Simultaneousco- adsorptive removal of phenol and cyanide from binarysolution using granularactivatedcarbon.ChemicalEngineeringJournal,228,655–664.
Al-Asheh,S.,Banat,F.,Al-Omari,R.,&Duvnjak,Z.(2000).Predictionsof binarysorptionisotherms forthesorptionofheavymetalsbypinebark usingisothermdata.Chemosphere,41,659–665.
ASTM,AmericanSocietyforTestingandMaterials.(2006).Standardtestmeth- odsforcyanidesinwater,ASTMD2036-06.pp.82–101.
Anastopoulos,I.,&Kyzas,G.Z.(2014).Agriculturalpeelsfordyeadsorption:
Areviewofrecentliterature.JournalofMolecularLiquid,200,381–389.
Botz,M.M.(2001).Overviewofcyanidetreatmentmethods.MiningEnviron- mentalManagement,MiningJournal,28–30.
Dash,R.R.,Dash,R.R.,&Balomajumdar,C.(2014).Treatmentofcyanidebear- ingeffluentsbyadsorption,biodegradationandcombinedprocesses:Effect ofprocessparameters.DesalinationandWaterTreatment,52,3355–3366.
Dash,R.R.,Gaur,A.,&Balomajumder,C.(2009).Cyanideinindustrialwastew- atersandits removal:Areviewonbiotreatment. JournalofHazardous Materials,163(1),1–11.
Foo,K.Y.,&Hameed,B.H.(2010).Insightsintothemodelingofadsorption isothermsystems:Review.ChemicalEngineeringJournal,156(1),2–10.
Giraldo,L.,&Moreno-Piraján,J.C.(2010).Adsorptionstudiesofcyanide ontoactivatedcarbonand␥-aluminaimpregnatedwithcooperions.Natural Science,2(10),1066–1072.
Gupta,N.,Balomajumder,C.,&Agrawal,V.K.(2012).Adsorptionofcyanide onPressmudsurface:Amodelingapproach.ChemicalEngineeringJournal, 191,548–556.
Halet,F.,Yeddou,A.R.,Chergui,A.,Chergui,S.,Nadjemi,B.,&Ould-Dris, A.(2015).Removalofcyanidefromaqueoussolutionsbyadsorptionon activatedcarbonpreparedfromlignocellulosicby-products.JournalofDis- persionScienceandTechnology,36(12),1736–1741.
Hall,K.R.,Eagleton,L.C.,Acrivos,A.,&Vermeulen,T.(1966).Pore-andsolid- diffusionkineticsinfixed-bedadsorptionunderconstant-patternconditions.
IndustrialandEngineeringChemistryFundamentals,5(2),212–223.
Hanela, S., Durana, J., & Jacobob, S. (2015). Removal of iron–cyanide complexesfromwastewatersbycombinedUV–ozoneandmodifiedzeo- lite treatment. Journal of Environmental Chemical Engineering, 3(3), 1794–1801.
Jorgensen,H.,Kristensen,J.B.,&Felby,C.(2007).Enzymaticconversion oflignocelluloseintofermentablesugars:Challengesandopportunities?
BiofuelsBioproductandBiorefining,1(2),119–134.
Kilic,M.,Apaydin-Varol,E.,&Pütün,A.E.(2011).Adsorptiveremovalof phenolfromaqueoussolutionsonactivatedcarbonpreparedfromtobacco residues:Equilibrium,kineticsandthermodynamics.JournalofHazardous Materials,189(1–2),397–403.
Langmuir,I.(1918).Theadsorptionofgasesonplanesurfacesofglass,mica andplatinum.JournaloftheAmericanChemicalSociety,40(9),1361–1403.
Mekuto,L.,Jackson,V.A.,&Ntwampe,S.K.O.(2013).Biodegradationoffree cyanideusingbacillusspecificconsortiumdominatedbyBacillusSafensis, LichenformisandTequilensisStrains:Abioprocesssupportedsolelywith whey.JournalofBioremediesBiodegradation,18,02–07.
Mohammad,S.G.(2013).Biosorptionofpesticideonto alowcost carbon produced from apricot stone (Prunusarmeniaca): Equilibrium, kinetic andthermodynamicstudies.JournalofAppliedScienceandResearch,9, 6459–6469.
Mohana,S.,Shrivastava,S.,Divecha,J.,&Madamwar,D.(2008).Response surfacemethodologyforoptimizationofmediumfordecolorizationoftextile dyedirectblack22byanovelbacterialconsortium.BioresourceTechnology, 99(3),562–569.
Montgomery,D.C.(2001).Designandanalysisofexperiments(5thed.).New York,USA:JohnWileyandSons.
Moussavi,G.,&Khosravi,R.(2010).Removalofcyanidefromwastewaterby adsorptionontopistachiohullwastes:Parametricexperiments,kineticsand equilibriumanalysis.JournalofHazardousMaterials,183(1),724–730.
Myers,R.H.,Montgomery,D.C.,&Anderson-Cook,C.M.(2016).Response surface methodology:Processand productoptimizationusingdesigned experiments.JohnWiley&Sons.
Osobamiro,M.T.(2012).Determinationoftheconcentrationoftotalcyanide inwastewaterofatobaccocompanyinSouthwesternNigeria.Journalof AppliedScienceandEnvironmentManagement,16(1),61–63.
Palmer,S.A.K.,Breton,M.A.,Nunno,T.J.,&Sulivan,D.M.(1988).Super- natantmetal/cyanidecontainingwatertreatmenttechnologies.pp.07656.
Millroad,ParkRidge,NJ:NoyesDataCorporation.
Parga,J.R.,Shukla,S.S.,&Carrillo-Pedroza,F.R.(2003).Destructionof cyanidewastesolutionsusingchlorinedioxide,ozoneandtitaniasol.Waste Management,23(2),183–191.
Prasad,A.L.,&Santhi,T.(2012).Adsorptionofhazardouscationicdyesfrom aqueoussolutionontoAcacianiloticaleavesasanecofriendlyadsorbent?
SustainableEnvironmentResearch,22(2),113–122.
Sathishkumar, P., Arulkumar, M., & Palvannan, T. (2006). Utilization of agro-industrialwasteJatrophacurcaspodsasanactivatedcarbonforthe adsorptionofreactivedye.RemovalBrilliantBlueR.JournalofCleaner Production,87,899–905.
Singh,N.,Agarwal,B.,&Balomajumder,C.(2016).Simultaneoustreatmentof phenolandcyanidecontainingaqueoussolutionbyadsorption,biotreatment andsimultaneousadsorptionandbiotreatment(SAB)process.Journalof EnvironmentalChemicalEngineering,4(1),564–575.
Stavropoulos,G.G.,Skodras, G.S.,&Papadimitriou,K.G.(2015).Effect ofsolutionchemistryoncyanideadsorptioninactivatedcarbon.Applied ThermalEngineering,74,182–185.
U.S.EnvironmentalProtectionAgency.(1994).Technicalreport:Treatmentof cyanideheapleachesandtailings.Washington,DC,USA.
Vedula,R.K.,Dalal,S.,&Majumder,C.B.(2013).Bioremovalofcyanideand phenolfromindustrialwastewater:Anupdate.BioremediationsJournal,17, 278–293.
Xu,F.,Yu,J.,Tesso,T.,Dowell,F.,&Wang,D.(2013).Qualitativeandquan- titativeanalysisoflignocellulosicbiomass usinginfraredtechniques:A mini-review.AppliedEnergy,104,801–809.
