thin films Room-temperature ferromagnetism in Ni-doped TiO diluted magneticsemiconductor Technology Journal of Applied Researchand

Journal of Applied Research and Technology

www.jart.ccadet.unam.mx JournalofAppliedResearchandTechnology15(2017)132–139

Original

Room-temperature ferromagnetism in Ni-doped TiO ₂ diluted magnetic semiconductor thin films

Bushra Parveen

, Mahmood-ul-Hassan

, Zeeshan Khalid

, Saira Riaz

, Shahzad Naseem

aMaterialsGrowthandSimulationLaboratory,DepartmentofPhysics,UniversityofthePunjab,Lahore54590,Pakistan

bDepartmentofPhysics,LahoreGarrisonUniversity,Lahore54590,Pakistan

cCentreofExcellenceinSolidStatePhysics,UniversityofthePunjab,Lahore54590,Pakistan Received20September2016;accepted20January2017

Availableonline31March2017

Abstract

WereportundopedandNi-dopedTiO2 (xNi=0.00,0.50,1.00,1.50,2.00and2.50wt.%)thinfilmsfabricatedonglasssubstratesbyusinga combinationofsolid-statereactionanddipcoatingtechniques.ThestructuralpropertiesareobservedbyX-raydiffraction(XRD),whichhave depictedthatannealingat650^◦CresultsinrutileNi-dopedTiO2asamajorphasealongwithaminoranatasephase.Thesurfacemorphologyofthe depositedthinfilms,asmeasuredbyscanningelectronmicroscopy(SEM),indicatesgranularsphericalshapednanostructures.Room-temperature ferromagnetism(RTFM)hasbeenillustratedbyallthegrownthinfilms,aselucidatedbyvibratingsamplemagnetometer(VSM).AlthoughNi contenthasnopronounced effectonthecrystallinitythatindicatesasubstitutionalreplacementofNiinTiO2 lattice,however,Ni contentis observedtoinfluencetheferromagneticbehavior.Therefore,thepresentstudysignifiesthepotentialspintronicapplicationsofNi-dopedTiO2

dilutedmagneticsemiconductors,fabricatedbyalow-costmethod,asitexhibitsRTFMwithnanograinsatthesurface.

©2017UniversidadNacionalAutónomadeMéxico,CentrodeCienciasAplicadasyDesarrolloTecnológico.Thisisanopenaccessarticleunder theCCBY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/4.0/).

Keywords: Dilutedmagneticsemiconductors;Dipcoating;Solid-statereaction;Room-temperatureferromagnetism(RTFM);Vibratingsamplemagnetometer

1. Introduction

Inthelastfewyears,thinfilmtechnologyhasbeenplaying a vitalrole in the fabrication of solid-state magneticstorage devices to study and realize novel phenomena to meet the requirementsoftheintegrated-circuitindustry.Thecostoffab- ricationofthinfilmsisverylowascomparedtothatoftheirbulk counterparts.Moreover,thechargetransportandtheinforma- tionstoragecaneffectivelyberealizedinnanoscaledevicesby employingthinfilms.Dilutedmagneticsemiconductorshavea prominentplaceintraditionalsemiconductor-basedspintronic research.Traditionalsemiconductors,dopedwithasmallcon- centration of magnetic ions to associate magnetic degree of freedom, have become suitable for both current controlling andmagneticstoragedeviceapplications(Oganisian,Hreniak,

∗Correspondingauthor.

E-mailaddress:[email protected](Mahmood-ul-Hassan).

PeerReviewundertheresponsibilityofUniversidadNacionalAutónomade México.

Sikora,Gaworska-Koniarek,&Iwan, 2015;Paulauskasetal., 2013). TiO2 dopedwithvarious magneticions for achieving room-temperatureferromagnetism(RTFM)isoneoftheattrac- tiveaspectsinspintronicsresearch.Basically,TiO2appearsin three different crystallographicphases namely,anatase, rutile andbrookite.Amongwhichrutilephaseisthe preferredone.

Rutilephasehasbeenwidelystudied,asitisnontoxicwithhigh transparencyinthevisibleregion,ithashighrefractiveindex, highdielectricconstantandgoodchemicalstability(Oganisian etal., 2015;Paulauskas etal.,2013;Sahdan,Nayan,Dahlan, Mahmoud,&Hashim,2012);therefore,itisconsideredasone of the suitable photocatalystsin heterojunctions.Also, it has potentialapplicationsincludinggassensors,membranes,LEDs, fibers, light transparent electrodes, heatreflectorsandoptical gain media(Akpan& Hameed,2010;Chen, Yang, Wang,&

Jiang, 2006). Thereare many reports inthe literature, which indicates thatthebandgapofTiO2canbetaperedbydoping withtransitionmetals(Fe,Cr,Mn,Ni),nonmetals(F,I,P, N, S),rareearthelements(La,Gd,Nd,Er,Pr)andalkalineearth metals(Be,Mg)etc.,thatenhancethephotocatalyticactivityin

http://dx.doi.org/10.1016/j.jart.2017.01.009

1665-6423/©2017UniversidadNacionalAutónomadeMéxico,CentrodeCienciasAplicadasyDesarrolloTecnológico.Thisisanopenaccessarticleunderthe CCBY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/4.0/).

thevisiblelight(Ahmad etal., 2010;Al-Jawad,2014;Asahi, Morikawa, Irie, & Ohwaki, 2014; Ayieko, Musembi, Waita, Aduda,&Jain,2013;Buzbyetal.,2006;DiValentin,Pacchioni,

&Selloni, 2005; Gültekin, 2014; Hamal & Klabunde,2007;

Hermawan, Pranowo, & Kartini, 2011; Hernández-Martínez, Estevez, Vargas, Quintanilla, & Rodríguez, 2012; Li, Ding, Liu,&Ning,2013; Niemela,Yamauchi,&Karppinen,2014;

Seery,George,Floris,&Pillai,2007;Sobana,Muruganadham,

&Swaminathan,2006;Tian,Gao,Deng,etal.,2013;Tian,Gao, Kong,etal.,2013;Yangetal., 2007),whichindicatespoten- tialopto-electronicdeviceapplications.Asurprisingfactisthe presenceofroom-temperatureferromagnetisminundopedTiO2

(Hong,Sakai,Poirot, &Brizé,2006; Yoonet al.,2006).The ferromagnetismis reported to be originated because of oxy- genvacancies actingas n-type dopantand cause a magnetic order. Through theoretical calculation(Paxton & Thien-Nga, 1998)themagneticorderinghasbeendemonstratedtobeorig- inatedbecauseof the Tistates operating withinthe bandgap.

Transitionmetalelementsarepreferredas dopantsbecauseof theiralmostsimilarionicradiiwiththeTi-cationsinthehost TiO2lattice,andadditionallymediateferromagnetism.Among manytransitionmetaliondopants,Ni⁺²isconsideredtobemore efficientdopantinTiO2,asitcanimproveelectricalandmag- neticproperties(Ayieko etal.,2013;Li etal.,2013; Nosaka, Matsushita,Nishino,&Nosaka,2005).Oganisianetal.,synthe- sizedFe-dopedTiO2bysol–gelmethodandstudiedtheeffectof dopingconcentrationonthemagnetism,andsuggestedthefor- mationofmulti-domainstructurethatenhanceswithFecontents (Oganisianet al., 2015).Choudhary et al., havereported the roleofOxygendefectsinassistingparamagnetictoferromag- netic phase change activated through the Fe doping inTiO2

(Choudhury,Verma,&Choudhury,2014).Similarly,Ni-doped TiO2isemployedtostudytheeffectofdopantsontheferromag- neticpropertiesofTiO2stabilizedinrutilephase(Park,Choi, Lee,Kim,&Kim,2007).Ontheotherhand,J.Tianetal.fab- ricatedTiO2 thin films bysol–gel process andobserved that increaseintransitionmetalconcentrationreducesmagnetization (Tian,Gao,Deng,etal.,2013;Tian,Gao,Kong,etal.,2013).

Gultekin,havestudiedAu-dopedTiO2thinfilms,andreported thatTiO2showbodycenteredtetragonalstructure,andtheband gapofTiO2thinfilmsincreasesfrom3.74eVto3.89eVwith increase in Au nanoparticle concentrations (Gültekin, 2014).

Yang etal. characterized C andV – TiO2 photocatalystsby sol-gelprocessfordegradationof acetaldehyde,andobserved thatVdopantsinTiO2increasesthesurfacearea,makingitsuit- ablefor lightactivity(Yangetal.,2007).Similarly,Ni-doped TiO2thinfilmspreparedbyspraypyrolysishaveshownthatNi stabilizesTiO2inrutilephase,makingitmoreefficientinphoto- catalyticapplications(Al-Jawad,2014).Hermawanetal.(2011) synthesizedNi-dopedTiO2nanocrystalsbysol–gelprocessand observed that the crystallite size increases from 18.51nm to 20.35nmandtheband gapenergy decreasesfrom2.73eV to 2.51eV.Thestudy ofmagneticcharacterinTiO2oxidesemi- conductorsis motivated because of recent interesting reports onIn2O3,ITO,SnO2,inwhichmagneticdopinginducedmag- neticpropertieshavebeenillustrated(Babu&Kaleemulla,2016;

Babuetal.,2016;Borges,Scolfaro,Alves,daSilva,&Assali,

2012; Kuppanetal., 2016a,2016b;Xu etal.,2009).Asit is evidentintheliterature,noreportexistonNi-doped TiO2for spintronicsapplicationsasDMS,fabricatedbylowcostmethod, toexhibitferromagnetismatroomtemperature(RT).

In this work, Ni-doped TiO2 thin films are deposited by employingacombinationofsolid-statereactionanddipcoating techniqueswithafocusonmagneticdeviceapplications,with variousdopantconcentrations(xNi=0.00,1.00,1.50,2.00and 2.50wt.%).Ni isa suitable transition metal magneticdopant for TiO2 and thisdiluted magneticsemiconductor has worth applications inmagneticdata storagedevices. The structural, morphologicalandferromagneticpropertieshavebeenexplored usingXRD,SEMandVSM,respectively.Theresultsobtained arecorrelatedandcomparedwiththeliteratureandsignificant pointshavebeenelaboratedwithafocusontheutilizationofthe studiedthinfilmsfordatastorageapplications.

2. Experimentalprocedures

Inthisstudy,wehaveemployedaverysimpleandlowcost methodtoprepareNi-doped(xNi=0.00,0.50,1.00,1.50,2.00 and2.50atwt.%)TiO2Thinfilmsonglasssubstrate.Forthe fabricationofundopedthinfilm,3.2gTitaniumdioxidepowder was addedintoethanol solution andstirredusing amagnetic stirrer for 1h at70^◦C,to achieve awell-mixedand uniform solution.The TiO2 thin filmwasfabricatedontoaglasssub- strateusing dipcoating.In thismethodultrasonicallycleaned glasssubstrateweredippedintothepreparedsolutionfor2–3 timestogetasuitableanduniformthicknessof thefilm.The TiO2thinfilmdepositedontheglasssubstratewasdriedinan electricovenat150^◦Cfor15min.InthiswayonelayerofTiO2

wasdepositedontheglasssubstrate.Thisprocesswasrepeated 3timestogetappropriateTiO2layerthickness.Similarly,Ni- doped(xNi=0.50,1.00,1.50,2.00and2.50atwt.%)TiO2thin films were fabricatedbyadding appropriate amountsof NiO powders(accordingtotheirmolarmasses)intotheTiO2solu- tion.Allthepreparedthinfilmshavebeenannealedinadigital furnaceat650^◦Cfor60min,toenhancethecrystallinity.The temperatureof650^◦Cwasattainedintwosteps,from0^◦Cto 350^◦Cin90min,andthenfrom350^◦Cto650^◦Cin60min.The crystal structureof thin filmsis observedwithX-raydiffrac- tion (XRD).Thesurfacemorphological detailsareelucidated by employingscanningelectron microscopy(SEM). The fer- romagnetic behaviors have been studied by vibratingsample magnetometer(VSM).

3. Resultsanddiscussion 3.1. Structuralstudy

XRD patterns of undoped and Ni-doped TiO2 (xNi=0 to 2.50wt.%)thinfilmsdepositedonglasssubstratesandannealed inairfor1hat650^◦CareshowninFigure1.TheX-raysare diffracted at the angles of 25.2^◦, 27.4^◦, 36.1^◦, 39.2^◦, 41.3^◦, 44.1^◦,48.1^◦,54.4^◦,56.7^◦,62.8^◦,64.1^◦,69.1^◦correspondingto variouscrystallographicplanes,asFigure1shows.TheseXRD patternsarecomparedwithICDDfilesforanataseTiO2(ICDD

20 30 40 50 2θ (degrees)

Powder Tio₂

0.00%

0.50%

1.00%

1.50%

2.00%

2.50%

(112)

(221)

(310)

(002)

(220)

(211)

(200)

(210)

(111)

(200)(101)

(110)(101)A A

Intensity (a.u.)

60 70 80

Fig.1.X-raydiffractionpatternsoftheprecursorpowderTiO2andthinfilmsof TiO2withxNi=0.00,1.00,1.50,2.00and2.50%.Allthegrownthinfilmsamples indicaterutileTiO2asdominatingphasealongwithaminutecontributionfrom anataseTiO2phase.However,thepresenceofthepeaksfromtheanatasephase canbeobservedfromtheXRDpatternoftheprecursorTiO2powder;therefore, thepresenceoftheanatasephaseisnotduetoNiadditioninTiO2,instead,the anatasephaseisalreadypresentintheprecursorTiO2powder.

file number: 21-1272) and rutile TiO2 (ICDD File number:

87-0710).ForundopedandNi-dopedtitaniumdioxide,amajor phase of rutile TiO2 is evident; nevertheless, a minor phase of anatase TiO2 is also present for all compositions of Ni.

The 2θ peaks around 25.2^◦ and48.1^◦ belongs to the anatase phaseforundopedandNi-dopedTiO2.Thesmallercontentof anatasecomparedtotherutilephaseisevidentfromdiffraction intensities. It can be observed that the anatase phase is not inducedbecauseoftheNiadditionbutitisalreadypresentin theprecursorpowderofTiO2,asshowninFigure1.Thelattice constantsaandchavebeencalculatedfromtheXRDspectra for major rutile phases, using the following relation (Wilso, Matijasevich,Mitchell,Schulz,&Will,2006)

1

d² =(h²+k²) a² + l²

c² (1)

hereddenotestheinterplanardistanceandh,k,lshowthemiller indices.The calculatedlatticeconstantsaandc, forundoped andNi-doped TiO2,are displayed inTable1.The calculated latticeparametersareplottedagainstNi-concentrationasgiven inFigure2.Thecalculatedvalueslatticeconstantsareingood match with earlier reports. The crystallite size is extracted

Table1

Thelatticeconstants,crystallitesizeandstraincalculatedfromXRDpatterns.

xNi Latticeconstants Crystallitesize (nm)

Strain(10⁻³)

(%) a=b( ˚A) c( ˚A)

0.0 4.600 2.955 17.677 9.046

0.5 4.645 2.955 15.919 7.254

1.0 4.582 2.971 15.704 9.012

1.5 4.616 2.962 19.626 7.860

2.0 4.616 2.962 18.847 7.240

2.5 4.633 2.958 16.823 8.200

from the major (110) peak using Debye-Scherer’s relation (Kuznetsovetal.,2009)

D= kλ

βcosθ (2)

where D=crystallite size, k=shape factor that is equal to 0.9,λ=X-raywavelengthandβ=fullwidthathalfmaximum (FWHM). The undoped TiO2 thin film has a crystallite size of 17.677nm. Doping with xNi=0.50%, the crystallite size decreases to 15.919nm. When the dopantconcentration fur- ther increasesto1.00%,thecrystallitesizevaluedecreasesto 15.704nm.Thedecreasingcrystallitesizemightbejustifiedby thepresenceofNi–O–TibondsinNi-dopedTiO2,whichinhibits thegrowthofthecrystals.Nonetheless,astheNiconcentration increasesfurtherto1.50%,thecrystallitesizeagainincreasesto 19.677nm,andforthelasttwoconcentrations2.00%and2.50%, thecrystallitesizeagaindecreasesto18.847nmand16.823nm.

ThisshowsthattheNi⁺²ionreplacesTi⁺⁴ionssubstitutionally andresultsinlineardecayofthecrystallitesize.Thefluctua- tioninthecrystallitesizeisduetothelatticestrains,whichare producedduringthesynthesisofundopedandNi-dopedTiO2. Thecrystallitesizestartstoincreasewhenthenumberofdislo- cationsdecreaseswiththeincreaseinNicontent(Prabbu,Rao, Kumar,&Kumari,2013;Purushothan&Krishna,2015;Vijay

4.8 4.6 4.4 4.2 4.0 3.8 3.6 3.4 3.2

0.0 0.5 1.0 1.5 2.0

Ni content, %

a (Å) c (Å)

c (Å) a (Å)

2.5 2.85 3.00 3.15 3.30 3.45 3.60 3.75 3.90

Fig.2.Thelatticeparametersaandc,evaluatedfromXRDpatternsofundoped andNi-dopedTiO2thinfilmsareplottedagainsttheNicontent.Thegraphshows thatwithincreasingtheNicontentthevalueoflatticeconstantsaandcvaries from4.60 ˚Ato4.63 ˚Aand2.955 ˚Ato2.977 ˚A,respectively.

235

0.0 0.5 1.0 1.5

Ni content, %

CS size: Debye scherer CS size: W-H plot Grain size: SEM

Crystallite size: W-H (nm) Crystallite size: Debye scherer equation

Grain size (nm): SEM

2.0 2.5

20 15 16 17 18 19 20

40 60 80 100

240 245 250 255 260 265 270

Fig.3.CrystallitesizecalculatedfromtheXRDandW–Hanalysisandgrain sizeobservedbySEM(seeFig.5)areplottedagainsttheNicontent.Thegraph showsthatwithvaryingtheNicontentthereisasmallfluctuationinthegrain sizeasdeterminedfromtheSEMimages;however,ithasanoveralldecreasing tendency.ThecrystallitesizerandomlyvarieswiththeNicontent.

etal.,2009).Thecalculatedcrystallitesizeisplottedagainstthe Nicontent,asshowninFigure3.Although,theanatasephaseis alreadypresentintheprecursorTiO2powerbuttherutilephase stabilitycanalsobejustifiedbecauseofthecalculatedcrystallite sizethatis>14nm,whichisacriticalvalue,belowandabove,it stabilizestheanataseandrutilephases,respectively(Bahadur, Pasricha,G,Chand,&Kotnala,2012).Thestrainpresentinthe samplesbecauseoflatticeimperfectionsanddistortioniscalcu- latedbyemployingStokeWilsonformula(Wilsoetal.,2006)

ε= β

4 tanθ (3)

The undoped TiO2 thin film possesses strain value of 9.046×10⁻³. Upon doping with xNi=0.50%, strain value decreasedto7.254×10⁻³.Whendopantconcentrationwasfur- therincreasedto1.00%,strainvalueincreasedto9.012×10⁻³. However,asthe Niconcentration increasedfurtherto1.50%, associated strain value decreased to 7.86×10⁻³ andfor the last two concentrations 2.00% and 2.50%, strain values are 7.24×10⁻³ and8.20×10⁻³respectively.The smallervalues ofstrainanddislocationdensityenhancefreecarriermobilityin thehostlattice.Duringthethermalenergyrelaxationswithinthe sample,certaindeformationsareproduceddependingontem- peraturegradientindifferentdirectionswhichisthecausefor increaseanddecreaseinstrain.Secondly,sincestrainisinversely relatedwithcrystallitesizethereforevariationincrystallitesize willaltertherelateddeformationswithinthesample(Kuznetsov etal.,2009;Prabbuetal.,2013;Purushothan&Krishna,2015;

Wilsoetal.,2006).

ThestrainvaluesplottedversusNicontentaredepictedin Figure4.UsingWilliamson–Hall(W–H)plotmethod,thecrys- tallitesizeandthestrainarefurthercomputed.The following

0.2

0.0 0.5 10 1.5

Ni content, %

Strain: W-H plot Strain: W-S formula

Strain: Wilson stokes formula Strain: W-H plot

2.0 2.5

0.4 0.6 0.8 10 1.2

1.4 0.0095

0.0090

0.0085

0.0080

0.0075

0.0070

Fig.4.ThestraincalculatedfromWilson-stokesrelationandW–Hanalysisis plottedagainsttheNicontent.Theobserveddifferencemayarisebecauseofthe residualanisotropicstrain.

equationhasbeenutilizedtoelucidatetherequiredproperties (Patle,Labhane,Huse,&Chaudhari,2015).

β cosθ

λ =4εsin θ

λ + k

D (4)

where,D=Crystallitesize,ε=strain.Eq.(3)iscalledasW–H equation.Wehaveplottedgraphsfor allconcentrationsof Ni by taking 4sinθ/λ and βcosθ/λ along x- and y-axis, respec- tively. By linear fitting the calculated data, the crystallite size and strain is extracted from y-intercepts and slopes of the linear fit lines,respectively. A comparative investigations between crystallite size and strain, calculated from Debye- Scherer’s relation, Wilson-stokes relation and W–H plot are giveninFigures3and4,respectively.Thestraincalculatedfrom Wilson–Stokesrelationhassomehowdifferentvaluesascom- paredtothatcalculatedbyusingW–Hplotmethod,whichmight beduetotheinvolvementof variouspeaksinW-Hplotwith differentFWHM,therefore,differentstrainvaluesappear.The mostintense(110)peakhasbeenusedtocalculatestrainusing Wilson-stokesrelation. Accordingtoadislocationmodel, the defectsmayinduceanisotropicstrainwithinthecrystals,there- fore,magnitudeofstrainmayvaryforvariouscrystallographic planesandthismightbethereasonforobservedconsiderable differences in the strain calculated by two methods (Ungar, 2008).The presence of strain indicates the imperfections are presentinthefabricatedsamples,whichcouldeitherbearising fromthefilm-substrateinterface,Nidopantintroducingstresses into thehostTiO2 latticeorbecauseof thenativeiondefects arisingbecause of theemployed growthconditions,inwhich theemployedthermalenergyisnotsignificantlyrelievedfrom thehostlattice,andtherefore,exertsstresses.

3.2. Surfacemorphology

TheSEMmicrographsof Ni-dopedTiO2(xNi=0.00,0.50, 1.00,1.50,2.00and2.50atwt.%)thinfilmsaregiveninFigure5.

Ingeneral,allthesethinfilmsampleshaveexhibitedagranular

X_Ni =0.0% X_Ni =0.50% X_Ni =1.0%

X_Ni =2.5%

X_Ni =2.0%

X_Ni =1.5%

Fig.5.SEMmicrographsofundopedandNi-dopedTiO2thinfilmsdepositedonglasssubstrates.Thegranularstructurealongwithporesbetweenthegrainsare present.ThemeasuredvaluesofthegrainsizefromtheSEMimagesdepictthatitdecreasescontinuouslyastheNicontentinTiO2thinfilmsisincreased.

structure.Thedifferentnanograinsonthesurfacedonothave largesizevariations.However,aporousstructurebetweendif- ferentgrainsisalsoevident.Thecalculatedgrainsizedecreases from 266.19nm to222.17nm withincrease inNi concentra- tion(xNi=0.00,0.50,1.00,1.50,2.00and2.50wt.%).Wehave comparedthecrystallitesizecalculatedfromXRDpatternsand thegrainsizecalculatedfromSEM,forvaryingNicontentas given in Figure 3. There is a small fluctuation in grain size withincreasingNicontent;howeverithasanoveralldecreasing tendency. The XRDcrystallite size decreases as the Ni con- tent increasesaboveup toxNi=1.00%, furtherincreasingthe Nicontent reducesthecrystallite size.Moreover,the crystal- lite size calculated from XRD and the Williamson analysis are plottedinFigure3, inwhichitis evidentthatwithvary- ingthe Nicontentasmallfluctuationingrainsizeoccurs,as determinedfromtheSEMimages.Thedifferencebetweenthe crystallitesizeandgrainsizevariationsareduetothefactthat bothshouldbeconsideredasindependentparametersbecause anambiguousrelationexistsamongthem.Because,abiggrain shownby anSEMimagemaycontain severalcrystallites, as depictedbyXRDstudies.The crystallitesizeextractedusing DebyeSchererformulaismuchsmallerthanthegrainsizeelu- cidatedfromSEM, whichisduetothepolycrystallinenature of the grownthin film.A grainmaybe composedof several crystallites because of which, while calculating the average crystallitesizeusingXRD,wesumupallthecrystallitescon- tributingtovariousgrains(Hasan&Nasir,2015).However,we canexpectcomparablegrainandcrystallitesize ifthe grown materialis singlecrystallineinnature. Therefore,anobvious relationbetweenbothcouldnotbeexpected(Bushroa,Rahbari,

Masjuki,&Muhamad,2012;Tomaszewski, 2013),andtrans- missionelectronmicroscopy(TEM)investigationscanfurther clarifytherelation.ThecrystallitesizecalculatedfromW-Hplot andDebyeSchererformulashaveasimilartrend, whichindi- catestheaccuracyofthepresentwork.Ithasbeenreportedthat thesurfacemorphologyofTiO2thinfilmsamplesshowscolum- nargrainsandroughnessofthesurfaceincreaseswithincreasing doping(Gültekin,2014).Similarly,AhmadMKetal.reported thatTiO2thinfilmsdepositedonsiliconsubstratehavesmaller poresonthesurface,andthegrainsizeincreaseswithincreasing theannealingtemperature(Ahmadetal.,2010).Theevidence ofthepresenceofnanograins,asshownbySEMimages,with rutileTiO2majorphasedeterminedbyXRD,indicatespotential applicationsofNi-dopedTiO2thinfilmsinthemicroelectronic industry.

3.3. Ferromagneticproperties

The magnetic behaviors of Ni-doped TiO2 thin films are investigatedwithVibratingSampleMagnetometer(VSM).The VSMresultshaveshownthatferromagneticbehaviorisexhib- itedbyallthinfilms.Thegraphsshowtypicalhysteresisloops duetoeachspecimenexhibitingcoercivityandremanence, as shown in Figure6. The diamagnetic effects dueto the glass substrates are suppressedbythe ferromagnetic signalsof the undopedandNi-dopedTiO2thinfilms.Thehysteresisloopsof TiO2 thinfilmsfor undoped anddopedwithxNi=0.50%and xNi=1.00%, shows that in addition to ferromagnetism, there isalsoaparamagneticbehaviorthatcanbeseenfromthelin- ear variationof magnetizationathigher valuesof theapplied

0.0012

0.0008

0.0004

0.0000

–0.0004

–0.0008

–0.0012

–10 000 –8000 –6000 –4000 –2000 0

Applied magnetic field (Oe)

0.5% 1.0% 1.5% 2.0% 2.5%

Magnetization (emu/g)

2000 4000 6000 8000 10 000 8000 4000 0 –4000 –8000 –0.0006

0.0000 0.0006 1500

1000 500 0 –500 –1000 –1500 –0.00016 –0.00008 0.00000 0.00008 0.00016

300K

300K

Undoped TiO₂

Fig.6.Thehysteresisloopsmeasuredat300KforTiO2thinfilmdopedwithxNi=0.50%,1.00%,1.50%,2.00%andxNi=2.50%showingferromagneticbehavior.

Theinsetfigureontheupperleftsideisplottedaround±2000Oetoobservethedetailsaroundtheorigin.Theinsetfigureinthelowerrightcornershowsthe hysteresisloopmeasuredfortheundopedTiO2thinfilm.

magneticfield.ThehysteresisloopsofTiO2thinfilmdopedwith xNi=1.50%,xNi=2.00%andxNi=2.50%,showsferromagnetic behavior,andfromthesegraphsitcanbeseenthatsaturation magnetizationMsat enhances with Ni content. The values of remanentmagnetizationMr,saturationmagnetizationMS and coercivity Hc of Ni-doped and undoped thin films are listed inTable2.Remanentmagnetizationincreasescontinuouslyup tothe sample withxNi=1.50%, however, above thisconcen- tration, a sharp increase has been observed. The saturation magnetizationdecreasesuptoxNi=2.00%andthenasudden increaseappearsforTiO2thinfilmwithxNi=2.50%.Thecoer- civityincreasesuptothesamplewithxNi=1.50%,however,it decreasessharplyforthesampleswithxNi=2.00–2.50%,which suggestthatNi-dopedTiO2thinfilmsshowsoftferromagnetic behavior athigher Ni concentrations and hardferromagnetic behavioratlowerNicontent,indicatingthepotentialspintronic device applications of TiO2. There are many reports in the

Table2

Theremanentmagnetization,saturationmagnetizationandcoercivityextracted fromthemeasuredhysteresisloops.

xNi

(%)

Remanent magnetization (emu/g)

Saturation magnetization (emu/g)

Coercivity (Oe)

0.0 4.1E−5 8.16E−4 1079

0.5 4.4E−5 8.04E−4 1405

1.0 4.8E−5 6.35E−4 1625

1.5 4.7E−5 5.37E−4 1675

2.0 7.9E−5 2.40E−4 0455

2.5 1.4E−4 11.0E−4 0320

literature about the magnetic properties of rutileand anatase phaseofTiO2.Recently,Bahaduretal.reportedthatthemag- netic momentof TiNiO2 rises atlowerNiconcentrations but withincreasing Niconcentration itdecreases (Bahaduret al., 2012).Themagneticpropertiesaredeterminedat300K(room temperature),therefore,thepresenceof hysteresiseveninthe undopedTiO2indicatesthepresenceofroom-temperaturefer- romagnetism(RTFM)(Gao,Tian,Zheng,Tan,&Zhang,2015), whichcouldbejustifiedbythepresenceofoxygeninterstitial defects.Thepresenceofoxygendefectsintheundopedhostlat- ticeshavealreadybeenreportedtomediateroom-temperature ferromagnetism(Das,Kar,&Srinivasan,2014).Ithasalready beenreportedthatintrinsicTiO2isn-typesemiconductor,which become p-type because of Ni doping. The n-type to p-type change incharacteris duetothevacanciesintroducedby the Ni dopant. In order to determine the real origin of room- temperatureferromagnetism(RTFM)inthestudiedthinfilms, wecanconsidervariouspossiblemechanisms.Onepossibility oftheappearanceofRTFMisduetothepresenceofsecondary phases (Radovanovic & Gamelin, 2003; Schwartz, Norberg, Nguyen,Parker,&Gamelin,2003)whichcouldberuledout, accordingtheXRDthatexcludesthesecondaryphaseappear- ance.Anothermechanismistheformationofboundmagnetic polaron(BMP).IthasalreadybeenreportedthatTiO2exhibits robust polaronic effect.According toBMP,effective massof carriersisincreasedbecauseofthesturdyelectron-phononinter- action. A trapped electron at an oxygen vacancy causes an F-center,andbecause ofitsorbital,it willeffectivelyoverlap surroundingNiions.Hencethepossiblemechanismfororigin ofmagnetisminTiO2isduetotheF-centerinducedBMP,which

ismediatedbylocalizedelectrons,surroundingtheimpurities (Tian,Gao,Deng,etal.,2013;Tian,Gao,Kong, etal., 2013;

Choudhuryetal.,2014).Similarly,RKKYmodel(i.e.theindi- rectexchangeinteractions),canalsobeusedtoexplaintheorigin causingferromagneticbehaviorinTiO2hostlattice.InRKKY interactions, Ni⁺² ionsare ferromagneticallycoupled through thefreecarriers(Bahaduretal.,2012).

Transition-metal-dopedTiO2 has alreadybeen reported to exhibitferromagnetismmediatedbysuper-exchanged-dmech- anism,whichincreaseswithincreasingdopantconcentrations.

Therefore,the increase indopant concentration shouldresult inincreaseinmagnetizationofthehostlattice,however,adif- ferenttrendwehavefoundinthestudiedNi-doped TiO2thin films, as evident in Table 2, where up to xNi=2%, remnant magnetizationincreasesbutsaturationmagnetizationdecreases.

As observed in Table 2, the coercivity is higher at smaller dopant concentration, but it decreases at higher Ni content (xNi=2.0–2.5%),indicatingashiftofrelativelyhardtosoftfer- romagneticbehaviorsofthefabricatedthinfilms,whichcould beassociatedwiththeresidualstrains,asdeterminedfromXRD, thatcanreducethemagneticanisotropyofthedopedTiO2thin films.

Thevariationinferromagneticcharacteristicsisrelatedwith theXRDandSEMstudies.Byincreasingthemagneticimpurity concentrationinTiO2,thereappearsnosecondaryphasebelong- ingtoNi,asdepictedinFigure1,whereamajorrutilephase of TiO2 thin filmsappear. Thisshowsthat Ni⁺² ion replaces Ti⁺⁴ionssubstitutionallyandresultinanearlylineardecayof SEMgrainsize,asis evidentinFigure3.Thecorresponding magneticpropertiesalso vary linearly, as illustratedbyVSM resultsinTable2,whichisinagreementwiththeearlierreports thatcontent ofaddedmagneticions,substitutingthe hostlat- ticecations,mayinduceproportionalferromagneticbehavior, inthehostlattice(Bahaduretal.,2012;Sabry,Al-Haidarie,&

Kudhier,2016).

4. Conclusions

Undoped andNi-doped TiO2 thin films have successfully been synthesized on glasssubstrates by a very simple solid- statereactionanddipcoatingtechnique.Ithasbeenobserved that Ni doping playsa vibrant role inthe exhibited physical characteristicssuchascrystalquality,crystallitesizeandmag- neticcharacteristicsofNi-dopedTiO2thinfilms.Theobserved averagecrystallitesizeis17.677nm.Goodagreementbetween values of the crystallite size and strain, extracted from the major XRD peaks and through the W-H analysis, has been found.The surfacemorphology of allsamples elucidated by SEM shows that there is an overall decreasing tendency in nanograinsize withincreasingNi content.The averagegrain size decreases from 266.19nm to 222.17nm withthe added Nicontent.Room-temperatureferromagnetism(RTFM)isillus- tratedinthepreparedthinfilms.Themeasuredcharacteristicsof thestudiedthinfilmsarefoundtodependupontheNicontent, whereitshows hardferromagneticbehavioruptoxNi=1.5%

andexhibitssoft ferromagnetictrendabovexNi=1.5%,there- fore,theobservedcharacteristics ofNi-doped TiO2thin films

aretunablebytheNicontentandcanbeemployedinvarious spintronicdeviceapplications.

Conflictofinterest

Theauthorshavenoconflictsofinteresttodeclare.

Acknowledgment

One of the authors, Mahmood-ul-Hassan (M. Hassan), is thankfultoUniversityofthePunjab,Lahore,forfinancialsup- portthroughorderno.D/1384/Est-I.

References

Ayieko,C.O.,Musembi,R.J.,Waita,S.M.,Aduda,B.O.,&Jain,P.K.(2013).

PerformanceofTiO2/In(OH)iSj/Pb(OH)xSyCompositeETAsolarcell fabricatedfromnitrogendopedTiO2thinfilmwindowlayer.International JournalofMaterialsEngineering,3(2),11–16.

Ahmad,M.K.,Halid,M.L.M.,Rasheid,N.A.,Ahmed,A.Z.,Abdullah,S.,&

Rusop,M.(2010).Effectofannealingtemperatureonsurfacemorphology andelectricalpropertiesofTiO2 thinfilmspreparedbysol–gelmethod.

JournalofSustainableEnergy&Environment,1,17–20.

Akpan,U.G.,&Hameed,B.H.(2010).Theadvancementsinsol–gelmethod ofdoped-TiO2photocatalysts.AppliedCatalysisA:General,375(1),1–11.

Al-Jawad,S.M.H.(2014).Effectofdopingonstructuralandelectricalproperties ofTiO2 thinfilmsforgassensors.InternationalJournalofEngineering Science,5(1),2171–2176.

Asahi,R.,Morikawa,T.,Irie,H.,&Ohwaki,T.(2014).Nitrogen-dopedtitanium dioxideasvisible-light-sensitivephotocatalyst:Designs,developments,and prospects.ChemicalReviews,114(19),9824–9852.

Babu,S.H.,Krishna,N.S.,Kaleemulla,S.,Rao,N.M.,Kuppan,M.,Krish- namoorthi,C.,...&Basheed,G.A.(2016).Synthesisandcharacterizations of(In0.90Sn0.05Ni0.05)2O3nanoparticlesusingsolidstatereactionmethod.

InAIPconferenceproceedings(Vol.1731,No.1,p.120003).

Babu,S.H.,&Kaleemulla,S.(2016).Roomtemperatureferromagnetismin ITOandNidopedITO.AdvancedMaterialsLetters,7(11),891–896.

Bahadur,N.,Pasricha,R.,Govind,Chand,S.,&Kotnala,R.K.(2012).Effectof Nidopingonthemicrostructureandhighcurietemperatureferromagnetism inSol–gelderivedtitaniapowders.MaterialsChemistryandPhysics,133(1), 471–479.

Borges,P.D.,Scolfaro,L.M.,Alves,H.W.L.,daSilva,E.F.,&Assali,L.

V.(2012).Studyoftheoxygenvacancyinfluenceonmagneticproperties ofFe-andCo-dopedSnO2dilutedalloys.NanoscaleResearchLetters,7(1), 540.

Bushroa,A.R.,Rahbari,R.G.,Masjuki,H.H.,&Muhamad,M.R.(2012).

ApproximationofcrystallitesizeandmicrostrainviaXRDlinebroadening analysisinTiSiNthinfilms.Vacuum,86(8),1107–1112.

Buzby,S.,Barakat,M.A.,Lin,H.,Ni,C.,Rykov,S.A.,Chen,J.G.,&Shah,S.I.

(2006).VisiblelightphotocatalysiswithnitrogendopedTiO2nanoparticles preparedbyplasmaassistedchemicalvapordeposition.JournalofVacuum Science&Technology,24(3),1210–1214.

Chen,D.,Yang,D.,Wang,Q.,&Jiang,Z.(2006).Effectsofborondopingon photocatalyticactivityandmicrostructureoftitaniumdioxidenanoparticles.

Industrial&EngineeringChemistryResearch,45(12),4110–4116.

Choudhury,B.,Verma,R.,&Choudhury,A.(2014).Oxygendefectassisted paramagnetictoferromagneticconversioninFedopedTiO2nanoparticles.

RSCAdvances,4,29314–29323.

Das,A.K.,Kar,M.,&Srinivasan,A.(2014).Roomtemperatureferromag- netisminundopedZnOnanofiberspreparedbyelectrospinning.PhysicaB:

CondensedMatter,448,112–114.

Gao,H.,Tian,J.,Zheng,H.,Tan,F.,&Zhang,W.(2015).EffectsofFedoping ontheopticalandmagneticpropertiesofTiO2filmsdepositedonSisub- stratesbyasol–gelroute.JournalofSol–GelScienceandTechnology,74(2), 521–527.