Enhancement in UV emission and band gap by Fe dopingin ZnO thin films

ANCHAL SRIVASTAVA*, NISHANT KUMAR, and SANJAY KHARE

Department of Physics, University of Lucknow, Lucknow−226007, India

Enhancement of the optical band gap of ZnO from 3.14 to 3.29 eV has been obtained using Fe dopant. Undoped and dopedZnO films are deposited by sol−gel spin coating. XRD patterns indicate polycrystalline nature and hexagonal wurtzite struc−ture of Zn1–x FexO films. EDX analysis confirms the presence of iron dopant. The photoluminescence spectra show an ultra−violet emission peak at 398 nm (NBE emission) and defect emission peak at 485 nm. Intensity of the NBE emission is muchhigher for the doped samples with its ratio to defect emission intensity highest for 2 at. % doping. The NBE emission shifts tohigher energy with increasing dopant concentration in a manner similar to that exhibited by the band gap. Surfacemorphology has been studied using FESEM.

Key words: Thin films, ZnO, sol−gel, Fe−dopant, band gap enhancement, UV−emission.

1. Introduction

Recent studies show that II–VI based semiconductors areexcellent materials for electronics, photonics and spintro−nics applications. Specifically, zinc oxide which has largedirect band gap (3.37 eV) and exciton binding (60 meV)energies is inexpensive and environmentally safe, and hasbeen identified as a promising host material [1–3]. Possibleapplications of ZnO films is in photonic devices such as UVlight emitting diodes, UV lasers and blue luminescentdevices, photodetectors, transparent conducting electrodes,surface acoustic wave devices, as well as in gas sensors.Doping of valency two metals viz. Mg[2,4], Ca[3], Sr[5];and Cd[6,7] has been reported which affects the opticalband gap of zinc oxide in exactly opposite ways as Cddecreases the band gap. Combined, this provides a consider−able range of band gap tuning, thereby opening a gatewayfor tuneable light sources. Ca and Sr doped ZnO has alsobeen reported as Love wave sensor [8]. Aluminium andIndium, valency three metals, have also been doped whichrespectively enhances [1,9] and decreases [10] the band gapof ZnO, as well as enhance its electrical conductivity bynearly two orders thus rendering ZnO a good candidate fortransparent electronics. Rare−earths like Ce, Nd, etc. havealso been doped which affects the band gap [11,12]. ZnOhas also been doped with anions to obtain a p−type ZnO andrealize homojunction LEDs although efforts in this direc−tion are still on. All these prove that zinc oxide is a versatilehost. ZnO is also considered as a host material for dilutedmagnetic semiconductor obtained by doping transition met−als. Cobalt and manganese have been comparatively exten−

sively studied. Dopants from group II, III, rare earth andtransition metals are found to change the properties of ZnOfilms drastically. Fe doped ZnO films are a subject of con−tinuous study as their magnetic, as well as optical propertiesmay take a U−turn depending heavily on the method of fab−rication. For example, as reported in Ref. 13 Polyakov et al.[14] found room temperature ferromagnetism by implantingFe ions in ZnO crystal grown by vapour phase while Janischet al. [15] reported that Fe−doped ZnO showed a paramag−netic behaviour. Also optical band gap of Fe doped ZnO hasbeen found to decrease by some authors [13,16–19],whereas it is reported to increase by others [20,21].

There are several methods for producing ZnO basedfilms, e.g., chemical vapour deposition, thermal evaporation, spray pyrolysis, successive ionic layer absorption and reac−tion, magnetron sputtering, pulsed laser deposition, laserchemical vapour deposition, sol−gel, screen printing, etc.[1–3,13,17,20,22]. Out of these methods sol−gel is simplerbesides producing good quality films. In this paper opticalband gap variation and photoluminescence properties alongwith structural, compositional and morphological analysesof Zn1–x FexO (x = 0 to 0.03) thin films prepared by sol−gelspin coating has been presented.

2. Experimental

The precursor for obtaining ZnO films is prepared by using0.1 M solution of zinc acetate dihydrate in ethanol anddiethanolamine. The mixture is magnetically stirred at 60°Cfor 30 minutes to get a homogeneous solution. To this solu−tion, appropriate volumes of 0.1 M solution of ferric nitrate[Fe(NO3)3 � 9H2O] in isopropanol is added to obtain 1, 2 and3 at.% doping of Fe. These solutions are again stirred for 30

68 Opto−Electron. Rev., 22, no. 1, 2014

OPTO−ELECTRONICS REVIEW 22(1), 68–76

DOI: 10.2478/s11772−014−0179−x

*e−mail: asrivastava.lu@gmail.com

min. Both the undoped and doped solutions are aged for 5days to achieve sol stability. The precursor solutions thusobtained are spin coated on properly cleaned glass slides.Spinning speed is kept at 3000 rpm while the spinning timeis 30 s. After each coating, the sample is heated to 400°C for1 h and then cooled back to room temperature. Spin coatingand heating is repeated 15 times for obtaining appreciablethickness. Finally all the films are annealed at 400°C for 4hrs. Thus there are four samples ZnO, ZnO: 1 at. % Fe,ZnO: 2 at. % Fe, and ZnO: 3 at. % Fe which are named assamples Z, ZF1, ZF2 and ZF3, respectively.

The crystal phase and crystallinity of the samples hasbeen investigated using X−Ray diffractometer (Model –PANalytical X'Pert PRO MRD X−ray diffractometer) for2� values ranging from 25 to 75° using CuK� radiation(� = 1.540598 �). FTIR and transmittance spectra havebeen recorded using spectrometer (Model− AlphaT,Bruker) and UV−Vis spectrophotometer (Model− V670,Jasco), respectively. The PL spectra have been obtainedusing fluorescence spectrometer (Model LS−55, PerkinElmer). The excitation wavelength is 325 nm coming froma 20 kW Xe discharge lamp. The surface morphologyis obtained using FESEM (Model− ZEISS) and withthe same equipment energy dispersive analysis (EDX)has also been done to ascertain the presence of iron dopant.All these measurements have been done at room tempe−rature.

3. Result and discussions

3.1. StructureXRD patterns (Fig. 1) of undoped and Fe− doped ZnO thinfilms deposited on duly cleaned glass substrates indicatepolycrystalline nature of all the film samples. No peak relat−ing to iron is found. This implies that doping of Fe up to3 at. % does not change the crystal structure. All the ZnObased films have a hexagonal wurtzite structure with majordiffraction peaks along (100), (002) and (101) planes. Smallpeaks along (110), (103), (102) and (112) are also seen [22].For undoped sample preferred orientation is along (101) but

as the dopant is introduced orientation along (101) decre−ases monotonically whereas it increases for (002) and (100)becoming maximum for 2at.% doping. However, further Fedoping reduces the orientation parameter both along (002)and (100) plane. The peak along c−axis, i.e., (002) planeoccurs at 2� = 34.575°, 34.525°, 34.525° and 34.575° forx = 0, 0.01, 0.02 and 0.03, respectively.

The particle size in the sample is estimated from theDebye−Scherer (DS) formula

tk

DS � �� �cos

,

where tDS is the particle diameter, k is the Scherer constantand is taken equal to one, � is the wavelength of the X−raysand � is the full width at half maximum (FWHM) of X−raydiffraction peaks in radians. Particle size along (100),(002) and (101) crystallographic plane as calculated by DSformula for all the samples is shown in Table 1 and in gen−eral it decreases with increasing Fe concentration in theZnO films. In order to distinguish the crystallite size−indu−ced broadening and strain−induced broadening of diffrac−tion peaks the Williamson and Hall (WH) plot has beenperformed and is shown in Fig. 2. The trend line equationsare found to be y = –0.029x + 0.018, y = –0.296x + 0.107,y = –0.690x + 0.234 and y = 0.132x – 0.017 for samples Z,ZF1, ZF2 and ZF3, respectively. The strain and particlesize are calculated by comparing the trend line equationwith

� � � � �cos sin� �C t WH 2 ,

where tWH is the particle size, � is the strain and C is the cor−rection factor which is taken equal to 1. The strain and parti−cle size determined from the above relation is shown inTable 1. The tensile and compressive strains are indicatedby positive and negative signs, respectively. The particlesizes obtained by DS formula and that by WH plot are quitedifferent indicating that the broadening of FWHM may bedue to strain existing in the films and not due to the particlesize.

Opto−Electron. Rev., 22, no. 1, 2014 Anchal Srivastava 69

Fig. 1. XRD pattern of Zn1–xFexO thin films. Fig. 2. WH plot of Zn1–xFexO thin films.

Table 1. Particle size determined by DS formula and WH plot; andstrain in Zn1–xFexO thin films.

SampletDS tWH Strain

(100) (002) (101)Z 13.48 15.46 14.94 8.55 –1.4 × 10–2

ZF1 5.29 7.69 8.98 1.43 –14.8 × 10–2

ZF2 3.08 5.15 6.82 0.65 –34.5 × 10–2

ZF3 8.26 5.20 6.68 9.06 6.6 × 10–2

The orientation parameter �(hkl) = I(hkl)/�[I(hkl)] are given inTable 2. These values of orientation parameters vary from0.0555 to 0.2927 indicating random orientation. The calcu−lated value of dislocation density which represents the amountof defects in the film randomly varies from 41.79 × 1014 to1051.51 × 1014 lines/m2 for the undoped and doped samples(Table 3).

Table 2. The orientation parameter �(hkl) along different planes forZn1–xFexO thin films.

SampleOrientation parameter �(hkl)

(100) (002) (101) (102) (110) (103) (112)Z 0.1875 0.2158 0.2927 0.0555 0.1015 0.0807 0.066

ZF1 0.1914 0.2001 0.2456 0.0818 0.1081 0.0966 0.076

ZF2 0.2101 0.2325 0.2452 0.0774 0.0892 0.0720 0.0733

ZF3 0.1773 0.1867 0.2342 0.0781 0.1148 0.1038 0.1048

Table 3. Lattice constant and dislocation density determined fromXRD data for Zn1–xFexO thin films.

SampleLattice Constant

c/aDislocation Density� × (1014 lines/m2)

a (�) c (�) (100) (002) (101)

Z 3.236 5.186 1.602 54.97 41.79 44.75

ZF1 3.236 5.194 1.605 357.19 168.72 123.95

ZF2 3.231 5.194 1.607 1051.51 376.15 214.72

ZF3 3.231 5.186 1.605 146.29 369.47 223.48

The lattice constants a and c of wurtzite structure of ZnOgiven in Table 3 have been calculated using the expression [2]

1 4

32

2 2

2

2

2d

h hk k

a

l

chkl

� � ��

�

�� �

�

�

�� ,

along with Bragg’s law. The lattice constant a of sample Z,i.e., ZnO thin film is larger than those of doped samples ZF2and ZF3. The lattice constant c of ZnO thin film is smallerthan doped samples ZF1 and ZF2 and is the same for sampleZF3. Bond length is also determined for all the sampleswhich varies between 1.967 to 1.970 �.

When Fe ion enters the ZnO lattice a distortion occurswhich manifests itself as strain. It is seen that with anincrease in Fe concentration the (002) peak initially shifts tolower angles as compared to that for undoped ZnO and, thenagain to higher angle being equal to that for undoped ZnO.The reason for such results may be assigned to the valencestate of Fe ion. A decrease in diffraction angle with increas−

ing Fe concentration is reported by Chen et al. [17] andWang et al. [13] where Fe ion exist as Fe2+. Increase in thediffraction angles is reported by Xu et al. [20] where Fe ionsexist as Fe3+. In the present paper dopant is introducedthrough ferric nitrate and therefore initially Fe ion would bein Fe3+. It is well known that Fe3+ ion reduces to Fe2+ in thepresence of zinc powder [23]. Here the host being zincoxide, conversion of Fe3+ to Fe2+ is very likely. In the lightof the (002) peak positions here and the results of references[13,17,20] it appears that Fe2+ and Fe3+ coexist in the sam−ple films presented in this paper. Coexistence of both ions isreported by Rambu et al. [18] where the films were preparedby spin coating and by other workers also [16].

Fe2+ and Fe3+ either of the ions with ionic radius 78 and68 pm, respectively is large enough to create strain if itlodges itself in the interstitials which may enhance the c−lat−tice constant as is happening up to 2 at.% dopant concentra−tion (Table 3). A large charge imbalance occurring due topresence of Fe3+ ions will enforce occurrence of Zn vacan−cies to maintain overall charge neutrality. When the dopantconcentration is increased further, Fe ions substitutes at theZn2+ sites and since the radius of Fe3+ (68 pm) is smallerthan Zn2+ (74pm), the lattice shrinks thereby resulting inreduction in the lattice constants as can be seen (Table 3) inthe present case.

3.2. FTIR

The FTIR spectra of undoped and doped samples are shownin Fig. 3. The peak at 545cm–1 is attributed to Zn−O stretch−ing vibrations. This peak shifts to 538, 533 and 547 cm–1 asdopant increases to 1, 2, and 3 at.% Fe, respectively. This

Enhancement in UV emission and band gap by Fe doping in ZnO thin films

70 Opto−Electron. Rev., 22, no. 1, 2014 © 2014 SEP, Warsaw

Fig. 3. FTIR spectra of Zn1–xFexO thin films.

shift is related to the change in Zn−O bond length. The bondlength lying between 1.967 and 1.970 � does not vary mo−notonously with the increasing dopant concentration. Theabsence of monotony can be attributed to the strain existingin the film. Such variation in the bond length has been alsoreported by earlier workers [24].

Position of various peaks corresponding to differentvibrations [25–28] vary with change in dopant concentra−tion and are summarized in Table 4.

Table 4. Absorption peak positions obtained from FTIR spectra ofZn1–xFexO films.

Observed absorption peakposition (cm–1) Attributed to

(with references)Z ZF1 ZF2 ZF3

545 538 533 547 Zn–O stretching vibrations[25,28]

717 724 749 727 Zn–O–Zn antisymmetricvibration [26]

– 1055 1099 – C=C stretching [25]

1152 1165 1137 1141 C–O stretching vibrationof ester group [27]

1301 1299 – 1279Complex structure of

COO−group with Zn ion, i.e.,carboxylic ionic complex [27]

1425 1447 1457 1442 C–O stretching vibrationsof the acetate group [27]

1616 1597 1599 1620 OH bending [28]1834 1826 1844 – C=O bond

3.3. Transmission and optical band gap

The transmission spectra of the Zn1–xFexO films depositedon glass slides obtained in the wavelength region 300 to 900nm is shown in Fig. 4. As Fe dopant is introduced in theZnO film the transmittance T rises to nearly 96% from 78%.Maximum transmittance in the visible region occurs for2at% doping and width of the transmission window (T 80%) is also the largest for this sample. The cut−off occurs atshorter wavelengths as the dopant concentration increasesindicating an increase in the optical band gap Eg which isalso determined quantitatively later in this section.

Since the Eg which is higher than the energy in visibleregion, the sample is being more and more transparent forvisible region of the spectrum, being maximum for the2at.% doped sample. With further doping, conduction tohigher conduction inter band and intra conduction bandoptical absorption processes [29] may start playing a rolethereby reducing the transmission for 3 at.% doping.

The absorption coefficient is calculated from Beer’slaw I = I0e–t, where I is the transmitted intensity, I0 is theincident intensity, and t is the thickness of the film. It hasbeen determined using stylus profilometer and is 490, 510,500 and 540 nm for undoped, 1, 2 and 3 at. % Fe doped ZnOfilms, respectively. The absorption coefficient for directband gap materials relates with optical band gap energy inaccordance with the expression = (hv –Eg)1/2, where h isthe Planck’s constant and v is the frequency of incident pho−ton. Intercept on the energy axis obtained by extrapolatingthe linear portion of the Tauc’s plot, i.e., (hv)2 vs. hv plot,as shown in Fig. 5, determines the band gap Eg. The bandgap for undoped 1, 2 and 3 at. % Fe doped films are ob−tained as 3.14, 3.24, 3.26 and 3.29 eV, respectively. Thismonotonous change in band gap with dopant concentrationx is shown in the inset of Fig. 5, linear fit (dotted line) towhich follows the equation

E x xg ( ) . . .� �4 7 3162

Thus, the optical band gap of ZnO can be tuned between3.14 and 3.29 eV by doping iron up to 3 at.%. Similar trendsof band gap enhancement in Fe doped ZnO films and nano−crystals have been reported by Linhua Xu et al. [20] andParra−Palomino et al. [21], respectively. Gradual changes inoptical band gap with increasing Fe doping have been repor−ted by several earlier workers. A summary of few relatedreferences is given below where positive and negative val−ues of change in a band gap correspond to a band gapenhancement and decrement, with respect to an undopedZnO sample, respectively.

In Ref. 20 and our present paper XRD results indicatethe existence of Fe3+ ions in the ZnO lattice. Enhancementin the optical band gap where the dopant exists in +3 statehad been explained through Burstein Moss shift when the

Opto−Electron. Rev., 22, no. 1, 2014 Anchal Srivastava 71

Fig. 4. Transmittance spectra of Zn1−xFexO thin films.

Fig. 5. Tauc’s plot, i.e., (hv)2 vs. photon energy for Zn1–xFexO thinfilms. The inset shows variation in band gap with Fe concentration

along with the linear fit (dotted line).

carrier concentration had been found to increase viz. Fe3+

doped ZnO by Xu et al. [20], Al doped ZnO by Shukla et al.[1] and Das et al. [9]. For Mg doped ZnO band gap enhance−ment is well known but reason is hardly reported. For Fedoped ZnO where iron exists in +2 state, at least five papers[13,16–19] have reported decrease in band gap. Nd (a lan−thanide) doped ZnO leads to a band gap narrowing althoughthe dopant is in +3 state. Douyar et al. [12] has reported thataccording to Zheng et al. [32] Nd 4f electrons introduce newstates close to the conduction band of ZnO. A newer LUMOis therefore formed which leads to a reduction of the bandgap [12].

Mn and Co (again transition metals) dopant increase theband gap of ZnO, in spite of cobalt being in +2 state. Bylsmaet al. [33] have dealt with exchange interaction related modi−fication of band gap in absence of applied magnetic field.A shifting of band edge energy due to exchange interaction ofelectrons in conduction band and valence band with the delectrons of Mn (dopant) was invoked to explain the variationin the band gap with dopant concentration. Diouri et al. [34]and Bylsma et al. [33] have shown that for an alloy in itsparamagnetic phase, as a result of second order perturbationof exchange interaction between the Mn+2 ion and the bandelectron, i.e., s−d and p−d interaction gives rise to negative and

positive correction to the energy of conduction and valenceband, respectively and causes a red−shift of the band gap withincreasing x. For antiferromagnetic phase the effect of ex−change interaction changes so as to produce a blue shift [35].Matsui et al. [30] have showed (VB) PES−spectra of ZnO andZn1–xCoxO reflecting the effect of Co (3d) states, which re−mains unchanged with increasing x and the band gap inZn1–xCoxO has been explained in terms of sp hybridizationbetween ZnO and CoO. Furdyna [36] stressed the close rela−tionship of s and p electron bands in AII

1–xMnxBVI material tothe band structure of nonmagnetic AIIBVI materials. Incorpo−ration of Mn into ZnO also increases the band gap [37].

Qin Guoqiang et al. [38] have reported that the band gapof wurtzite ZnO changes under triaxial strains, a new type ofstrain where band gap increases [39] along with a and c lat−tice constants, although ZnO still remains a direct band gapsemiconductor. Comparing with the unstrained ZnO, the Eg

increases under compressive strain but decreases under ten−sile strain. The strain dependence of Eg may lie in thechanges in the interaction between electrons. However, inour case where triaxial strain do not appear to exist, Eg

increases with an increase in a compressive strain (up to2 at % doping) and it continues to increase even for tensilestrain (3 at. % doping).

Enhancement in UV emission and band gap by Fe doping in ZnO thin films

72 Opto−Electron. Rev., 22, no. 1, 2014 © 2014 SEP, Warsaw

Dopant/DopantSource Sample/Substrate Method

of PreparationChang in Band gap

(meV) State of ion Explanation Reference

Fe/– Thin Film/Si(100) Sol−gelspin coating �+18 Fe3+ Burstein−

−Moss shift Xu et al. [20]

Fe/Fe(II) acetate Nanocrystals Chemical route +340 – QuantumConfinement

Para−Polominoet al. [21]

Fe/iron pieces Thin Film/Al2O3(001) RFM Sputtering –500 Fe2+ sp−d exchange

interaction Chen et al. [17]

Fe/iron pieces Thin Film/ Si(100) RFM Sputtering –86 Fe2+ sp−d interaction C. Wang et al. [13]

Fe/– Thin Film/Al2O3(0001) RFM Sputtering –70 Fe2+ and Fe3+ sp−d interaction Kim at al. [16]

Fe/iron nitrate Thin Film/Glass Spin coating –190 Fe2+ and Fe3+ sp−d interaction Rambu et al. [18]

Fe/FeCl3 Thin Film/Glass Spray Pyrolysis –340 Fe2+ s−d and p−d Alver et al. [19]

Co/ZnCoOceramic target

Thin Film/Sapphire (0001) PLD +625 Co2+ sp hybridization Matsui et al. [30]

Mn/– Nanocrystals Thermolysis �+100 – – Wang et al. [31]

Nd/ NdCl3 Thin Film/Glass Spray Pyrolysis –80 Nd3+ – Douayar et al. [12]

Mg/MgO Thin Film/Glass PLD +156 Mg2+ – Shan et al. [2]

Mg/MgO ThinFilm/Al2O3(001) PLD +330 Mg2+ – P. Mishra et al. [4]

Cd/CdO ThinFilm/Al2O3(001) PLD –380 Cd2+ – P. Mishra et al. [6]

Cd/cadmium ace−tate

Thin Film/quartz Spray Pyrolysis –260 Cd2+ – Vijaylaksmi

et al. [7]

Al/Al2O3Thin Film/

Glass PLD +400 – BM shift Shukla et al. [1]

Al/Al2O3Thin Film/

Sapphire (0006) PLD +350 – BM shift Das et al. [9]

In the light of the above discussion, it can be inferredthat sp hybridization is responsible for the increase in theband gap of Fe doped ZnO with increase in dopant concen−tration in the present paper.

3.4. Photoluminescence spectra

PL emission spectra of all the films are recorded with anexcitation wavelength of 325 nm and are shown in Fig. 6.Emission spectra of the sample Z, i.e., undoped ZnO filmshows UV emission peak at 398 nm (3.111 eV) followed bydefect related peaks in lower energy region; a broad onecentred about 460 (2.695 eV) and a sharp one at 485 nm(2.556 eV). Emission peak at 398 nm which is a near bandedge (NBE) emission has intensity higher than the other twopeaks occurring in the visible region.

The PL emission intensities are seen to depend upon theamount of Fe doping. In general with an increase in dopantconcentration UV emission intensity increases nearly 2.4fold, emission near 460 nm gets suppressed and emission at485 nm increases only slightly. The intensity ratio IU/IV forUV emission to defect emission (485 nm) is presented in theinset of Fig. 6. It is highest for 2 at. % Fe doped ZnO thinfilm. Density of excitons in ZnO is the major factor [40,41]affecting the intensity of ultraviolet emission at room tem−perature. Increase of iron dopant increases the number offree excitons but excess doping (beyond 1at. %) quenchesthe PL emission partially as due to larger concentration ofdopant ions they interfere with the activity sphere of neigh−bouring dopant ion. Simultaneously it is seen that the defectemission at 485 nm for ZF2 sample decreases, due toimproved crystalline quality of this sample as seen by theorientation parameter along (002) plane, thereby resulting inhighest value of I IU V for this sample i.e. 2 at.% doping.

Besides intensity, an increase in UV−emission energy isalso seen with an increase in dopant concentration. TheNBE emission shows successive blue shift occurring at

3.111, 3.179, 3.183 and 3.187eV for 0, 1, 2 and 3 at.% Fedoping in ZnO thereby facilitating a tuning over 76 meV.Nature of change in band gap energy and the NBE emissionenergy with dopant concentration conform to each other asseen in Fig. 7. Similar trend has been reported by Xu et al.[20]. NBE emission occurs at energies (3.111 to 3.187 eV)lower than the band gap (3.14 to 3.29eV) determined byTauc’s plot, for the corresponding samples. Similar obser−vations have been reported by earlier workers [42,10] alsosince the excitionic level is lower than the conduction band.

Peaks in visible region occur because of defect statescreated during deposition of films. No shift in this emissionindicates that they are consistent in all film samples. PLemission at 485 nm (2.556 eV) has been reported earlieralso [5,43]. This blue emission at 485 nm may be due to theelectronic transition probability from the donor level of zincinterstitial (Zni) to the acceptor level of zinc vacancy (VZn)[43,44] as the coexistence of these two defects has highprobability [45]. VZn is also likely to be formed in the vicin−ity of Fe3+ ions to maintain the charge neutrality. Kukreja etal. [46] has reported that according to Ahn et al. [47] andJanotti et al. [48] emission in the ~2.4 to 2.9 eV region isdue to transitions corresponding to the deeper intra−bandlevels of oxygen interstitials (Oi) and/or zinc vacancies(VZn

–). Theoretical calculations by Janotti and Van de Walle[48] and experiments by Reynolds et al. [49,50] Kohan etal.[51] and Sekiguchi et al.[52] showed that the transitionsfrom conduction band or a shallow donor level to singlyionized Zn vacancy (VZn

–) are possible and these transitionshave energies close to ~2.5 eV.

3.5. Surface morphology and EDX

The surface morphology of sol gel spin coated Zn1–xFexOthin films has been studied using Field Emission ScanningElectron Microscopy (FESEM) and is shown in Fig. 8. Thesample Z, i.e., undoped ZnO of molarity 0.1 M has rod−likestructures spread throughout with a granular surface in thebackground. As the dopant Fe is introduced in 1 at. % in

Opto−Electron. Rev., 22, no. 1, 2014 Anchal Srivastava 73

Fig. 7. Change in Eg and NBE− emission energy with Fe−dopant con−centration.

Fig. 6. PL emission spectrum of Zn1–xFexO thin films. Inset showsthe ratio (IU/IV) of PL intensities for the UV and the visible.

ZnO these rods start coalescing resulting in broader rodswith continuous flakes in the background. For 2 at. % Fedoped sample rods disappear and the flake−like structuresgrow huge which, with further doping, take shape of oystershell poised to roll up. Doping of Fe has influenced the sur−face structure of ZnO based films. Surface morphology of

undoped ZnO sample consists of well−defined structures,i.e., rods over a granular background. With gradual intro−duction and enhancement of Fe dopant, which is coexistingin both the ionic states, the host lattice is distorted inhibitingformation of well−defined structures thereby leading to theirultimate disappearance and formation of huge flakes.

Enhancement in UV emission and band gap by Fe doping in ZnO thin films

74 Opto−Electron. Rev., 22, no. 1, 2014 © 2014 SEP, Warsaw

Fig. 8. FESEM of Zn1−xFexO thin films.

Fig. 9. EDX spectra of Zn1–xFexO thin films for x = 0 (a) and x = 0.03 (b). (c) shows the elemental mapping of Zn, O and Fe for x = 0.03.

The EDX spectra of Zn1–xFexO thin films are shown inFig. 9. Figure 9(a) shows the presence of Zn and O as theonly elementary species in the sample Z (x = 0), i.e., undo−ped ZnO. For x = 0.03 presence of Fe is also observed,besides Zn and O, with a concentration of approximately 3at. % as expected [Fig. 9(b)]. Insets show the concentrationof Zn, O and Fe in at. % present in Zn1–xFexO thin films andFigure 9(c) shows the elemental mapping of Zn, O and Fefor 3 at. % Fe doped sample.

Conclusions

Enhancement in band gap and blue shift in UV (NBE) emis−sion is obtained in the spin coated transparent Zn1–xFexOfilms and nature of their change with dopant concentrationconforms to each other. All the films are polycrystallinehaving hexagonal wurtzite structure with no impurity phase.EDX confirms the presence of iron in the doped films. Theoptical band gap can be tuned from 3.14 to 3.29 eV, anenhancement of 4.77%, by increasing the Fe concentrationfrom x = 0 to x = 0.03. The near−band−edge PL emissionshows blue shift of 76 meV as dopant concentration isincreased to 3 at. %. Intensity of the NBE emission is muchhigher for the doped samples with its ratio to defect emis−sion intensity highest for 2 at. % doping. Rod−like structuresspread throughout over a granular surface in the backgroundstart coalescing resulting in broader rods and disappearanceultimately as the dopant concentration increases succes−sively. Grains in the background turn into continuous fla−kes, growing huge and taking shape of oyster shell ready toroll up.

Acknowledgement

Award of Research Assistantship to one of the authors,Nishant Kumar from CSTUP India vide project no.CST/D−1129, financial assistance from DST New Delhi India videproject no.SR/S2/CMP−0028/2010 and UGC through Spe−cial Assistance Program is gratefully acknowledged. Theauthors also gratefully acknowledge the insightful com−ments and suggestions made by the anonymous reviewers.

References

1. R.K. Shukla, A. Srivastava, and K.C. Dubey, “Growth oftransparent conducting nanocrystalline Al doped ZnO thinfilms by pulsed laser deposition”, J. Cryst. Growth 294, 427(2006).

2. F.K. Shan, B.I. Kim, G.X. Liu, Z.F. Liu, J.Y. Sohn, and W.J.Lee, B.C. Shin, and Y.S. Yu, “Blue shift of near band edgeemission in Mg doped ZnO thin films and aging” , J. Appl.Phys. 95, 4772 ( 2004).

3. K.P. Misra, R.K. Shukla, and A. Srivastava, “Blue shift inoptical band gap in nanocrystalline Zn1–xCaxO films depos−ited by sol−gel method”, Appl. Phys. Lett. 95, 031901 (2009).

4. P. Misra, P. Bhattacharya, K. Mallik, S. Rajagopalan, L.M.Kukreja, and K.C. Rustagi, “Variation of band gap with oxy−

gen ambient pressure in MgxZn1–xO thin films grown bypulsed laser deposition”, Solid State Comm. 117, 673 (2001).

5. T.A. Vijayan, R. Chandramohan, S. Valanarasu, J. Thiru−malai, and S.P. Subramanian, “Comparative investigation onnanocrystal structure, optical, and electrical properties ofZnO and Sr−doped ZnO thin films using chemical bath depo−sition method”, J Mater Sci. 43, 1776 (2008).

6. P. Misra, P.K. Sahoo, P. Tripathi, V.N. Kulkarni, R.V.Nandedkar, and L.M. Kukreja, “Sequential pulsed laser de−position of CdxZn1–xO alloy thin films for engineering ZnOband gap”, Appl. Phys. A 78, 37 (2004).

7. S. Vijayalakshmi, S. Venkataraj, and R. Jayavel, “Character−ization of cadmium doped zinc oxide (Cd: ZnO) thin filmsprepared by spray pyrolysis method”, J. Phys. D: Appl. Phys.41, 245403 (2008).

8. W. Water, S−F Wang, Y−P Chen, and J−C Pu, “Calcium andstrontium doped ZnO films for love wave sensor applica−tions”, Integrated Ferroelectrics 72, 13 (2005).

9. A.K. Das, P. Misra, A. Bose, S.C. Joshi, R. Kumar, T.K.Sharma, and L.M. Kukreja, “Structural, electrical and opticalcharacteristics of Al doped ZnO films grown by sequentialpulsed laser deposition”, Phys. Express 3, 5 (2013).

10. H. Li, Y. Zhang, X. Pan , H. Zhang , T. Wang, and E. Xie,“Effects of In and Mg doping on properties of ZnO nano−particles by flame spray synthesis”, J Nanopart Res. 11,917(2009).

11. Q. Luo, L.S. Wang, H.Z. Guo, K.Q. Lin, Y. Chen, G. H. Yue,and D. L. Peng, “Blue luminescence from Ce− doped ZnOthin films prepared by magnetron sputtering”, Appl. Phys.A108, 239 (2012).

12. A. Douayar, P. Prieto, G. Schmerber, K. Nouneh, R. Diaz, I.Chaki, S. Colis, A. El Fakir, N. Hassanain, A. Belayachi, Z.Sekkat, A. Slaoui, A. Dinia, and M. Abd−Lefdi, “Investiga−tion of the structural, optical and electrical properties ofNd−doped ZnO thin films deposited by spray pyrolysis”, Eur.Phys. J. Appl. Phys. 61, 10304 (2013).

13. C. Wang, Z. Chen, Y. He, L. Li, and D. Zhang, “Structure,morphology and properties of Fe−doped ZnO films preparedbyfacing−target magnetron sputtering system”, Appl. Surf.Sci. 255, 6881 (2009).

14. A.Y. Polyakov, A.V. Govorkov, N.B. Smirnov, N.V.Pashkova, S.J. Pearton, K. Ip, R.M. Frazier, C.R. Abernathy,D.P. Norton, J.M. Zavada, and R.G. Wilson, “Optical andmagnetic properties of ZnO bulk crystals implanted with Crand Fe”, Mater. Sci. Semicond. Process. 7, 77 (2004).

15. R. Janisch, P. Gopal, and N.A. Spaldin, “Transition metal−−doped TiO2 and ZnO present status of the field”, J. Phys.:Condens. Matter 17, R657 (2005).

16. K.J. Kim and Y.R. Park, “Optical investigation of Zn1–xFexOfilms grown on Al2O3 (0001) by radiofrequency sputtering”,J. Appl. Phys. 96, 4150 (2004).

17. Z.C. Chen, L.J. Zhuge, X.M. Wu, and Y.D. Meng, “Initialstudy on the structure and optical properties of Zn1–xFexOfilms”, Thin Solid Films 515, 5462 (2007).

18. A.P. Rambu, V. Nica, and M. Dobromir, “Influence ofFe−doping on the optical and electrical properties of ZnOfilms”, Superlattices and Microstructures 59, 87 (2013).

19. U. Alver, T. K�l�n�, E. Bacaks�z, and S. Nezir, “Structure andoptical properties of Zn1–xFexO thin films prepared by ultra−sonic spray pyrolysis”, Mat. Sci. Eng. B138, 74(2007).

Opto−Electron. Rev., 22, no. 1, 2014 Anchal Srivastava 75

20. L. Xu and X. Li, “Influence of Fe−doping on the structuraland optical properties of ZnO thin films prepared by sol−gelmethod”, J. Cryst. Growth 312, 85 1(2010).

21. A. Parra−Palomino, O. Perales−Perez, R. Singhal, M. Tomar,J. Hwang, and P.M. Voyles, “Structural, optical, and mag−netic characterization of monodisperse Fe−doped ZnO nano−crystals”, J. Appl. Phys. 103, 07D121 (2008).

22. A. Srivastava, R.K. Shukla, and K.P. Misra, “Photolumines−cence from screen printed ZnO based nanocrystalline films”,Cryst. Res. Technol. 46, 949 (2011).

23. A.S. Russell and J.C. Carver, “Reduction of ions of nickel,cobalt, iron and other metals by zinc amalgam”, Nature 142,210 (1938).

24. K. Jayanthi, S.V. Manorama, and S. Chawla, “Obserbationof Nd3+ visible line emission in ZnO: Nd3+ prepared bya controlled reaction in the solid state”, J. Phys. D: Appl.Phys. 46, 325101 (2013).

25. R.N. Gayen, K. Sarkar, S. Hussain, R. Bhar, and A.K. Pal,“Effect of annealing temperatures on properties of sol−gelgrown ZnO−ZrO2 films”, Indian J. Pure Appl. Phys. 49, 470( 2011).

26. S.L. Patil, M.A. Chougule, S,G. Pawar, S. Sen, and V.B.Patil, “Effect of camphor sulfonic acid doping on structural,morphological, optical and electrical transport properties onpolyaniline−ZnO nanocomposites”, Soft Nanoscience Lett. 2,46 (2012).

27. X−W. Du, Y−S. Fu, J. Sun, X. Han, and J. Liu, “Complete UVemission of ZnO nanoparticles in a PMMA matrix”, Semi−cond. Sci. Technol. 21, 1202 (2006).

28. T. Ivanova, A. Harizanova, T. Koutzarova, and B. Vertruyen,“Effect of annealing temperatures on properties of sol−gelgrown ZnO−ZrO2 films”, Cryst. Res. Technol. 45, 1154(2010).

29. P.D.C. King and T.D. Veal, “Conductivity in transparent ox−ide semiconductors”, J. Phys.: Condens. Matter. 23, 334214(2011).

30. H. Matsui and H. Tabata, “Lattice, band, and spin engineer−ing in Zn1−xCoxO”, J. Appl. Phys. 113, 183525 (2013).

31. Y.S. Wang, P.J. Thomas, and P. O’Brien, “Optical propertiesof ZnO nanocrystals doped with Cd, Mg, Mn, and Fe Ions”,J. Phys. Chem. B110, 21412 (2006).

32. J.H. Zheng, J.L. Song, Z. Zhao, Q. Jiang, and J.S. Lian, “Op−tical and magnetic properties of Nd−doped ZnO nanopar−ticles”, Cryst. Res. Technol. 47, 713 (2012).

33. R.B. Bylsma, .M. Becker, J. Kossut, and U. Debska, “De−pendence of energy gap on x and T in Zn1–xMnxSe: The roleof exchange interaction”, Phys. Rev. B33, 8207 (1986).

34. J. Diouri, J.P. Lascaray, and M. El Amrani, “Effect of themagnetic order on the optical−absorption edge in Cd and Mn,Te”, Phys. Rev. B31, 7995 (1985).

35. Y.R. Lee and A.K. Ramdas, “Energy gap, excitonic, and “in−ternal” Mn2+ optical transition in Mn−based II–VI dilutedmagnetic semiconductors”, Phys. Rev. B38, 10600 (1988).

36. J.K. Furdyna, “Diluted magnetic semiconductors”, J. Appl.Phys. 64, R29 (1988).

37. T. Fukumura, Z. Jin, A. Ohtomo, H. Koinuma, and M. Kawa−saki, “An oxide−diluted magnetic semiconductor: Mn−dopedZnO”, Appl. Phys. Lett. 75, 3366 (1999).

38. Q. Guoqiang, Z. Guanglei, Y. Jinhui, Y. Gang, F. Hua, and J.Fengqiu, “Electronic band gap of ZnO under triaxial strain”,J. Wuhan University of Tech.−Mater. Sci. Ed. 28, 48 (2013).

39. R.E. Marotti, D.N. Guerra, C, Bello, G. Machado, and E.A.Dalchiele “Band gap energy tuning of electrochemicallygrown ZnO thin films by thickness and electrodeposition po−tential ”, Sol. Energ. Mat. Sol. C82, 85 (2004).

40. S.S. Kurbanov, G.N. Panin, T.W. Kim, and T.W. Kang, “Im−pact of visible light illumination on ultraviolet emission fromZnO nanocrystals”, Phys. Rev. B78, 045311 (2008).

41. H.D. Li, S.F. Yu, A.P. Abiyasa, Clement Yuen, S.P. Lau,H.Y. Yang, Eunice S.P. Leong, “Strain dependence of lasingmechanisms in ZnO epilayers”, Appl. Phys. Lett. 86, 261111(2005).

42. L. Irimpan, V.P.N. Nampoori, P. Radhakrishnan, A. Deep−thy, and B. Krishnan, “Size dependent fluorescence spectros−copy of nanocolloids of ZnO”, J. Appl. Phys. 102, 063524(2007).

43. A. Ghosh and R.N.P. Choudhary, “Structural evolution andvisible photoluminescence of Zn–ZnO nanophosphor”,Phys. Status Solidi A206, 535 (2009).

44. B. Lin, Z. Fu, and Y. Jia, “Green luminescent center inundoped zinc oxide films deposited on silicon substrates”,Appl. Phys. Lett. 79, 943 (2001).

45. P.S. Xu, Y.M. Sun, C.S. Shi, F.Q. Xu, and H.B. Pan, “Theelectronic structure and spectral properties of ZnO and its de−fects”, Nucl. Instrum. Methods Phys. Res. B199, 286 (2003).

46. L. M. Kukreja, P. Misra, J. Fallert, D.M. Phase, and H. Kalt,“Correlation of spectral features of photoluminescence withresidual native defects of ZnO thin films annealed at differ−ent temperatures”, J. Appl. Phys. 112, 013525 (2012).

47. C.H. Ahn, Y.Y. Kim, D.C. Kim, S.K. Mohanta, and H.K.Cho, “A comparative analysis of deep level emission in ZnOlayers deposited by various methods”, J. Appl. Phys. 105,013502 (2009).

48. A. Janotti and C.G. Van de Walle, “Native point defects inZnO”, Phys. Rev. B76, 165202 (2007).

49. D.C. Reynolds, D.C. Look, B. Jogai, and H. Morkoc, “Simi−larities in the band edge and deep−centre photoluminescencemechanisms of ZnO and GaN”, Solid State Commun. 101,643 (1997).

50. D.C. Reynolds, D.C. Look, B. Jogai, J.E. Van Nostrand, R.Jones, and J. Jenny, “Source of the yellow luminescenceband in GaN grown by gas−source molecular beam epitaxyand the green luminescence band in single crystal ZnO”,Solid State Commun. 106, 701 (1998).

51. A.F. Kohan, G. Ceder, D. Morgan, and C.G. Van de Walle,“First−principles study of native point defects in ZnO”, Phys.Rev. B61, 15019 (2000).

52. T. Sekiguchi, N. Ohashi, and Y. Terada, “Effect of hydroge−nation on ZnO luminescence”, Jpn. J. Appl. Phys. Part 2 36,L289 (1997).

Enhancement in UV emission and band gap by Fe doping in ZnO thin films

76 Opto−Electron. Rev., 22, no. 1, 2014 © 2014 SEP, Warsaw