Band Gap Reduction and Improved Ferromagnetic Ordering via Bound Magnetic Polarons in Zn(Al, Ce)O Nanoparticles

ABSTRACT Polycrystalline nanoparticles of Al-doped and Ce-co-doped ZnO were processed through sol-gel co-precipitation. Crystallite sizes determined from XRD were in the range of 10.90 to 48.77 nm. FTIR spectra indicated a stretching mode of Zn-O bond. The insertion of Al and Ce at Zn-O site created Al-O and Ce-O bonds. UV-vis spectra favoured the formation of impurity levels by doping and co-doping. Their overlapping with the conduction band edge led to the reduction of band-gap. Blue-green emission which arises from radiative recombination of a photogenerated hole with an electron occupying oxygen vacancy was observed in photoluminescence spectra. FESEM suggested granular growth in undoped ZnO which changed to diverse structures such as nano-bar and cluster of elongated grains. Significant improvement in room temperature ferromagnetism (RTFM) was noticed by co-doping of Ce in Al-doped ZnO. Formation of bound magnetic polarons (Ce3+-Vo-Ce3+) was the primary mechanism responsible for the improvement in RTFM.


Introduction
In the last few decades, the band gap engineering has emerged as an important tool in the applied study of metal oxide nanoparticles.In the list of metal oxides, important examples are TiO 2 and ZnO.Many studies have shown greater interest in ZnO because of ZnO is an important versatile material because of excellent optical, electrical, optoelectronic, gas sensing, piezoelectric and photochemical properties [1,2].ZnO has large band gap of 3.37 eV and exhibits large exciton binding energy of 60 meV that makes it attractive for its emission tendency in UV and full colour lighting [3][4][5][6].Also, ZnO is less expensive and allows the production of highly pure and ultrafine nanoparticles with a high level of transparency in visible region [7][8][9][10][11].It belongs to II-VI semiconductor family that have flexible magnetic properties and good electron mobility.
Impurities can be introduced into semiconductors to create energy levels within the band gap.This is the reason as to why doping in semiconductor has been widely studied experimentally and theoretically for multiple applications such as improved luminescence, photocatalytic activity, band gap tuning and better electron transport [12][13][14][15][16][17][18][19].Apart from various optoelectronic characteristics, ZnO-based materials can be considered as diluted magnetic semiconductors (DMS).There are studies related to existence of room temperature ferromagnetism (RTFM) caused by grain boundaries and various point defects such as Zn interstitial and oxygen vacancies.Doping ZnO with elements such as Al, Cu, Ga, Fe, Co, Mn and Cr incorporates addition carriers which also contribute to RTFM [20,21].The interaction between local magnetic moments of the dopants with free charge carriers plays a significant role in developing optimum RTFM.The defects created by doping also help in optimising RTFM in ZnO [20,21].
Doping of rare earth (RE) elements is another area which is widely explored in connection with ZnO.Particularly, the ferromagnetic ordering in RE-doped ZnO arises via coupling of partially filled 4 f shells of RE ions with delocalised charge carriers of ZnO [22].Doping of single element in semiconductors has several limitations like recombination of carriers at the doping sites and creation of uncompensated defects which eventually limit some of the important features of semiconductors including photocatalytic activity, electrical, electronic and magnetic behaviour [22].To address the issue, co-doping is generally studied.The co-doping provides relatively lesser formation energy than single doping, indicating that the co-doped system is more stable, and the system is easier to form under O-rich condition [23].ZnO has demonstrated some of the important properties using a co-dopant along with a suitable dopant.Fe and Ce co-doping in ZnO has recently shown improved photocatalytic behaviour [22].
Assadi et al. showed enhancement of ZnO optical and magnetic properties by co-doping [24].Amiri et al. also demonstrated that (Ce, Cu) co-doped ZnO created donor impurity bands producing a superexchange interaction between the two atoms [25].Density functional theory based on first principles has also been widely employed to predict and analyse the properties of doped and co-doped systems of ZnO.Hence, the experimental investigation of such materials constitutes a challenge from the perspective of tuning the optoelectronic variables of ZnO and optimising its RTFM.
Accordingly, we have selected here a transition metal, i.e.Al as dopant in the matrix of ZnO along with a RE co-dopant i.e.Ce, based on the anticipation that the band gap tuning and optimum ferromagnetism will be simultaneously obtained.A facile and convenient method of preparation, i.e. sol-gel coprecipitation was adopted to develop Zn(Al, Ce)O nanoparticles (NPs).Structural, optical, morphological and magnetic characterisation were carried out to understand various properties of synthesised NPs.

Synthesis
The undoped, Al-doped and Ce-co-doped ZnO nanoparticles were synthesised by sol-gel co-precipitation.To prepare Al-doped ZnO, analytical grade zinc acetate dihydrate (99.9%,Sigma-Aldrich) and aluminium chloride hexahydrate (99.9%,Alfa Easer) were used as a host and dopant compounds.Methanol was taken to act as a solvent.Zinc acetate di-hydrate was dissolved in methanol under high speed magnetic stirring for two hours for preparing the 0.08 M host precursor solution.Equimolar solution of aluminium chloride hexahydrate in methanol was simultaneously prepared.The solution of aluminium chloride in appropriate amount was added to the zinc acetate dihydrate solution to achieve 5%, 8%, 10%, 12% and 15% atomic doping of Al in ZnO.
The precursor solution of cerium chloride (III) hepta-hydrate was simultaneously prepared in methanol for co-doping.While adding the dopant solution to the host solution drop by drop, the temperature was maintained at 50°C under continuous stirring and after the addition of dopant, codopant solution (cerium chloride (III) heptahydrate) was added drop by drop under continuous stirring and heated for 45 minutes.Stock solution of 0.5 M NaOH was also prepared and added drop by drop till complete precipitation.
The resulting solution was aged for 1 h and the precipitate/slurry was collected and washed by distilled water and ethanol to remove the unreacted reagents.The collected slurry was dried in an oven at 80°C for about 10 h and finally annealed at 400°C for 2 h.Thus, we had seven samples as per configuration given in Table 1, for further characterisation.

Characterisation
The powder samples were subjected to multiple characterisations to understand their structural, optical, magnetic and morphological features.X-ray diffraction (XRD) patterns were recorded in 2θ range of 20 to 80° by Rigaku Ultima IV diffractometer (Japan) using CuK α radiation (1.5406A o ).Using XRD patterns, orientation parameter, lattice constant, and volume of unit cells, anion-cation bond length, and crystallite size were determined.The Williamson-Hall (W-H) plots that were used to determine crystallite size and strain were also based on the XRD data.Fourier Transform Infrared Spectroscopy (FTIR) (Bruker Alpha, USA) was used to study the various bonds present in the powder.Spectrophotometer (Shimadzu UV2600, Japan) provided the UV-vis spectra.Small amount of samples was dissolved in ethanol to prepare 500 ppm solution and ultrasonicated for 30 min.During the ultrasound treatment, the agglomerated nanoparticles broke into individual nanoparticles; thus, homogeneous dispersion was prepared and then the spectra were recorded in transmission mode for each sample in the wavelength range 300 nm to 800 nm.Corresponding Tauc's plots provided the band gaps.Photoluminescence (PL) spectrometer (Horiba Fluoromax 4c, Horiba, USA) was used to record the emission spectra of all the powder samples.Here also, the powder samples were dissolved and sonicated in ethanol for getting the PL spectra with excitation wavelength of 325 nm.The microstructural details of all the samples were observed
The peaks for the given samples did not show any substantial shift in 2θ values.The XRD pattern of distinct peaks of Figure 1 from which the major structural properties of ZnO were determined, are shown in Figure 2. Al doping sharpens the diffraction peak and Ce co-doping reduces of intensity and broadens the diffraction peak.
Graphs are plotted (Figure 3) between the sample number and angle of diffraction (2θ) for major three peaks i.e. (100), (002) and (101).For (100) and (101) peak, 2θ shows closely same and random variation.The random fluctuation of 2θ corresponding to these two peaks is related to the randomness of the occurrence of displacement of Zn by dopant and co-dopant or by the settlement of foreign entities at the interstitial sites.In the hexagonal structure, c-axis plane i.e. (002) was distinct and most significant, hence it was selected for further analysis.
The initial doping with Al slightly increased 2θ corresponding to (002) peak.When co-dopant Ce was introduced with next level of Al doping, value of 2θ shifted towards higher angle up to S5.The regular shift of 2θ values towards higher angle with introduction of dopant and co-dopant was noted.This can happen when the lattice parameter of doped and codoped powder samples is smaller than the undoped ZnO.The minor increment in 2θ values by Al +3 incorporation in ZnO (i.e.S2 sample) indicated smaller reduction in lattice parameter.It is because of the smaller size of Al +3 (0.053 nm) as compared to Zn +2 (0.074 nm).From S3 onwards 2θ values showed greater change, that meant both dopant and codopant started replacing significantly the Zn +2 ions.The highest level of shift in 2θ was noticed in S6 sample and thereafter it was again resumed in S7.The reduction potentials of Zn, Al and Ce were −0.763V, −1.68 V and −2.33 V respectively, and according to electrochemical series the metals which had lower reduction potential (higher negative value) can easily reduce the metal causing the higher probability of displacing Zn in the ZnO matrix [26][27][28][29].Hence, Al and Ce had lesser reduction potential than Zn, which can easily replace Zn +2 ions in crystal lattice.Thus, the value of 2θ was increased up to S5.In S6 it decreased because it is expected to lead to saturation in displacement reaction and Al and Ce start occupying the interstitial sites thereby increasing the lattice parameter.The abnormal change in S7 sample (increase in 2θ values) can be explained by EDS analysis, the co-doping percentage of Ce is less than S6 and is the reason why the value of 2θ is further increased.
It may also be noted that doping ZnO with Al atoms led to the increase in XRD peaks intensities indicating the increase in crystalline quality [30,31].This result implied, on the one hand, the occupation of Zn +2 (0.074 nm) vacancies by Al +3 ions (0.054 nm).On the other hand, the strong covalency of Al-O makes the interaction of Al with O stronger than Zn with O, leading to the decrease of oxygen defects in the lattice, which promotes the crystallinity of the nanopowder samples.The intensity of diffraction peaks decreased after the introduction of Ce atoms as compared to undoped and Al-doped ZnO because of random occupation of intrinsic defects in ZnO lattice by co-dopant ions and their difference in the ionic radii of zinc, aluminium and Ce +3 (0.114 nm) or Ce +4 (0.101 nm).
In all the samples, for a given plane with Miller indices (hkl), the corresponding orientation parameter (γ (hkl) ) was calculated using the following equation [32]: where I (hkl) represents the intensities of the corresponding planes with Miller indices (hkl).The γ (hkl) vales are given in Table 2.The random variation can be noticed in the data which implies the random distribution of various crystallographic planes in the samples.Furthermore, for a given plane with Miller indices (hkl) and interplanar spacing (d hkl ) the lattice parameter a (=b) and c were calculated using the following relationship [32]: The lattice parameters (i.e. a and c) are given in Table 3.The variation of a and c with sample number is shown in Figure 4.Although a did not follow any definite trend in its variation, however if we can see the variation of c, it systematically decreased until S5 and increased for S6 and finally reduced again for S7.Variation of c is consistent with changes in c-axis oriented XRD peak i.e. (002).The reason behind random variation of a is related to the random changes in 2θ values corresponding to other XRD peaks.The 'u' parameter, which depicts the atomic displacement, is given by [32]: The calculated values of 'u' parameters for all the samples are given in Table 3.It shows the randomness in its variation.
It is expected that the magnitude of length of Zn-O bond is affected by Al and Ce incorporation into the ZnO matrix.The anion-cation (Zn-O) bond length (L) is given by [32]: L ¼ ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi a 2 3 The magnitude of 'L' estimated for all the samples is given in Table 3.It shows variation in the third decimal place indicating minor random effects of doping and co-doping.The minor variation of 'L' with dopant and co-dopant incorporation should lead to insignificant change in the volume of the unit cell of ZnO.To confirm this, the volume 'V' of hexagonal unit cell was calculated by [32]: It is indeed interesting to note that the variation of 'V' with dopant and co-dopant concentration demonstrated the trend that was similar to the variation of 'L' with the dopant and co-dopant concentration.

Crystallite size and degree of crystallinity
The crystallite size (t s ) was estimated using the wellknown Scherrer equation given by [32,33]: where β is the full width at half maximum (FWHM) of the given peak, θ is the corresponding angle of Bragg diffraction, λ is the wavelength of X-ray used (1.5406A 0 ) and k is a constant (0.9) [34,35] The calculated values are given in Table 4.It confirmed the nano-crystalline nature of all the powder samples with crystallite size in the range of 10.90 nm to 48.77 nm.
From the above equation, it was observed that the crystallite size depends on FWHM of the XRD peak apart from the angle of diffraction and the wavelength of the x-rays.Although it was noted that the variation in particle size did not follow any order but based on the proceeding section (002) plane was selected because of c-axis orientation.There was a trend observed in the graph between crystallite size and samples in Figure 5.It was observed that the maximum magnitude of ts was for S2 and S3.It seems that the incorporation of Al and Ce led to ZnO towards clustering.When Al (dopant) and Ce (co-dopant) were added in higher amount, the clustering was decreased and the crystallite size was less than its maximum magnitude.
These results indicated that the presence of Ce atoms leads to the decrease of the crystallite size probably because of decrease in molecular concentration at the crystallite surface [29].Also, it has been reported that smaller size of highly co-doped sample is caused by the formation Zn-Ce-O bond in the crystal lattice, which plays an important role in hindering the crystal growth [30].
The degree of crystallinity (Xc) was calculated using the following empirical relation [31]: The calculated values of degree of crystallinity is tabulated in Table 5. Degree of crystallinity variation was evaluated for (002) plane, which had the same  variation as the crystallite size.Variation in the degree of crystallinity and particle size is presented in Figure 5 in which the introduction of Al increased the degree of crystallinity and the crystallite size.It was further increased by the introduction of lowest co-doping but as the degree of co-doping increased, the crystallite size and degree of crystallinity decreased.The observed trend in crystallite size was also supported by degree of crystallinity variation.The best crystallinity is found in S3 and the worst one in S7.It confirmed that the highest clustering occurred in sample S3 and clustering property was decreased as the doping and co-doping was increased.

Dislocation density
The addition of dopant causes generation of dislocations.Therefore, the dislocation density was calculated for all the samples by [32]: The δ (hkl) values are given in Table 6.It is seen that for all the crystallographic planes the δ (hkl) value showed

W-H analysis
W-H plot is generally used for qualitative peak profile analysis, as it ignores several factors like the profile size (Lorentzian or Gaussian), any size distribution, domain shape, the elastic anisotropy for the microstrain and its actual source etc [19].It is based on the experimentally recorded XRD patterns.The graph between βCosθ versus Sinθ was plotted and a linear fit to the curve was compared with following equation [32,33]: where t WH is crystallite size, ε is strain, λ is the wavelength used and β is FWHM.Constant C is a correction factor, taken as 0.9.Plots for W-H analysis are given in Figure 6.The equations of trend-line fit are given in the inset of W-H plot for each sample in Figure 6.The calculated values of strain and t WH are presented in Table 4.
It can be seen that samples S1 and S2 showed the tensile strain but in sample S3, the strain became compressive.This is indicated respectively by the positive and negative sign of the slope of linear fit equations.ZnO's unintentional n-type conductivity is the result of native point defects which is because the oxygen vacancies are inherently present in ZnO [34].Here high temperature annealing of powder samples and addition of dopant and co-dopant appears to contribute towards the generation of vacancies and that is the reason for tensile strain in S1 and S2.However, the co-dopant entry in the lattice of host material may cause the annihilation of vacancies, resulting in compressive strain, which is the case for S3.On increasing the amount of codopant, the strain became tensile because of the displacement of Zn by Ce and Al in a symmetric manner in samples S4 to S7.Generally, the shift of strain from tensile to compressive is because of the change in concentration of co-dopant at certain amount or at the saturation point.
The relatively higher amount of strain is the primary reason behind the discrepancy in the crystallite size (t S and t WH ) obtained from the two different methods.Despite the smaller amount of strain, if the crystallite size does not match, this means that the data points are distributed on the graph.The trend -line fit is shown by R 2 value and is presented in the inset of graphs.

FTIR Analysis
The FTIR spectra, recorded in transmission mode for all the powder samples are given in Figure 7.A sharp dip is observed in all the samples at 527 cm −1 related to stretching of Zn-O bond [19].Intensity of peak shows the random variation.Samples S2 to S7 had a dip at 686 cm −1 , which is related to the stretching mode of Al-O bond [19].It indicated that Al is present in these samples.The absorption band between 970-980 cm −1 is due to the stretching of Ce-O bond that gives evidence in favour of the presence co-dopant i.e. cerium.The intensity of the dip corresponding to Ce-O bond increased with the increase of Ce in samples, which implied the increase of entities with Ce-O bond.In ZnO based nano-powder derived via chemical route there were some other absorption dips because of >C = O functional group [12].Here such dips are noticeable at 1528 and 1627 cm −1 .Position and intensity of these dips varied followed doping and co-doping in ZnO.The origin of these dips lies in the use of organic compounds during the sol-gel synthesis.

UV-Vis spectra
UV-Vis transmission spectra of undoped ZnO (S1), Al doped ZnO (S2) and Ce co-doped in Al doped ZnO (S3 to S7) are given in Figure 8.The absorption dip was visible in the wavelength range of 368 to 390 nm.Beyond 400 nm, the transparency level of NPs was increased in all the samples.The optical band gap analysis was done using Tauc plot (Figure 9), in which a curve of (αhν) 2  and hν was plotted and the linear portion of the curve was extrapolated to hν-axis.The intercept at hν axis provided the band gap values [34].The calculated optical band gap values are given in Table 7 and optical band gap variation for samples is shown in Figure 10.
From Figure 10, it is clear that the optical band gap continuously decreased with the introduction of dopant and co-dopant.The band gap reduction on doping with Al is well known but how Ce affects the Al-doped ZnO system is a matter of further analysis.Djouadi et al. reported that after the introduction of dopant in the ZnO lattice, absorption edge shifts to higher wavelength, thereby indicating reduction of band gap [31].This shift is due to the formation of impurity bands by dopant, which overlaps with the conduction band edge and leads to the decrease in the optical band-gap [31].The regular decrease in band gap in a systematic manner is supported by the above assumption.This means that Al doping creates overlapping in between impurity band and conduction band edge and after Ce co-doping the suggested overlapping attains the higher extent level which results in further decrease in the optical band gap.

Photoluminescence study
The room temperature PL emission spectra of nano powder samples (S1 to S7) recorded at the excitation wavelength of 325 nm are shown in Figure 11.It may be noted that a major emission peak was observed at 428 nm in all the samples, in the visible region.Generally, in ZnO based materials two peaks are observed one belongs to UV region due the near band-edge (NBE) emission which is generated from the recombination of exciton-exciton through collision process [34].Another emission belongs to visible range and is referred to as bluegreen-yellow emission that comes out of the radiative recombination of a photogenerated hole with an electron occupying the oxygen vacancy.This means, such type of blue emission (428 nm) is because of defects.This may be due to any one of them or by their combination; the oxygen vacancy (V o ), zinc vacancy, oxygen interstitial, zinc interstitial.Generally, oxygen vacancies arise in three different states; one is a neutral oxygen vacancy, second is a singly ionised and third is a doubly ionised.Vanheusden  singly ionised oxygen vacancies are responsible for the green luminescence in ZnO [36].In samples S1 to S7, the peak position belongs to the visible range that favours the above assumption that the recombination process of photogenerated hole with an electron occupying the oxygen vacancy results in blue emission.There was no variation in peak position from sample S1 to S7 with an exception S6.According to EDS spectra, highest co-doping was observed in S6 that resulted in the shift of peak position towards higher wavelength and shows the red shift.It is also previously reported by Li et al. that in PL spectra of Ce-doped ZnO nanoparticles new emission bands are developed which lead to the red shift in UV emission [37].

FESEM and EDS analysis
The FESEM images of all samples are shown in Figure 12.First image corresponding to S1 (undoped ZnO) shows the elongated spherical inter-linked chain of grain like structure.Al-doping (S2) yields rods near the grains.Here two images of S2 are displayed: in the first image the grains are settled near the rods but in the second image the elongated grains near the rod are visible.At the initial addition of Ce (S3), the grains appear to combine with each other and the size of grain is increased.Further addition of Ce, i.e. in S4, S5 and S7, the size of particle was changed and converted into clusters of diverse structures.The bar-like structures are pronounced and clear in S4.In S6 indefinite shape of grains was observed because of the clustering.The cluster, to some extent, was converted into semi-spherical hockey ball-like structure.This abnormal feature of S6 sample image was also noticed in the EDS analysis.
EDS spectra of sample S1 to S7 are given in Figure 13 and their atomic percentage are shown in the inset-tables in Figure 13.From the value of atomic percentage in all samples, it is clear that Zn and O are present as major elements in all the samples.S1 has only Zn and O that represent the undoped ZnO.In S2, minor content of Al was observed, consistent with the addition of dopant Al in ZnO.From S3 to S7, smaller amount of Ce was observed, which supported the presence of co- dopant.The amount of Ce increased from S3 to S6 and in S7 it was decreased.Here, S6 showed the highest atomic presence of Ce, which is the major reason for its exceptional behaviour in PL emission and lattice constant variation.The EDS spectra shown here is for a given specific area of the powder samples.Therefore, the amount of dopants and co-dopants shown here are just to confirm the presence of added elements, and not for the true representation of their content [38].

Magnetic studies
M-H curves for undoped, Al doped and Ce co-doped ZnO NPs are shown in Figure 14.Insets given in Figure 14   towards saturation.The plot of variation of coercivity and retentivity with sample number are shown in Figure 15.Both the plots followed similar trend.The coercivity was highest, i.e. 806 Oe for S3 and lowest at 202 Oe for S6.The highest and lowest amount of retentivity were 0.005 emu/mg and 0.0004 emu/mg for S3 and S1, respectively.The unsaturated magnetisation curves of S1 and S2 indicated the contribution of paramagnetic response.RTFM in undoped ZnO NPs is caused by grain boundaries and availability of free carriers in the form of intrinsic defect [39].Al doping in ZnO is expected to increase the carrier concentration thereby causing the RTFM.Initial addition of Al in ZnO increased both coercivity and retentivity, which implied that there was an increase in ferromagnetic order.However, the contribution of Al was not adequate to completely suppress the paramagnetic behaviour.Incorporation of co-dopant i.e.Ce, along with Al significantly enhanced the ferromagnetic ordering resulting in RTFM in S3.Further addition of Ce (S4) showed diamagnetic contribution in large magnetic field region, as indicated by the bending of curves.The diamagnetism in ZnO arises due to incomplete oxidation of Zn clusters [39].Also, we have noticed that FESEM image of S4 was characterised by a nano-bar structure, which was unique to S4 only.Hence, the specific nanostructure and un-oxidised Zn clusters both can be considered responsible for the presence of diamagnetic feature in this sample.
Next level of co-doping (S5-S7) resumed the ferromagnetic ordering similar to S3. Upon inclusion of Ce, the improvement in ferromagnetic ordering occurred due to exchange interaction among free delocalised charge carriers and 4 f spins of Ce ions [20].Such interaction results in the formation of bound magnetic polarons (BMPs).In Zn-RE system, the magnetic vacancies V o accept one electron forming H-like orbit with finite radius [21].The ferromagnetism with such radius arises because of relationship with the doped magnetic ions (Ce) with impurity electrons.When the defects are greater than the percolation limit, the Ce 3+ ions are ferromagnetically coupled.This coupling is mediated via oxygen vacancies through the formation of Ce 3+ -V o -Ce 3+ groups.Such group is referred as BMP.Similar mechanism has been reported earlier for Sm 3+ doped ZnO [21].

Conclusions
Polycrystalline Al-doped and Ce-co-doped ZnO nanoparticles were obtained via sol-gel coprecipitation with crystallite size in the range of 10.90 to 48.77 nm.The optical band gap decreased with the introduction of dopant and co-dopant.Aldoping induced overlapping in between the impurity band and conduction band edge and Ce co-doping induced overlapping to higher level leading to the decrease in the optical band gap.The PL emission spectra was characterised by a sustainable blue-green emission, arising from the recombination process of photogenerated hole with an electron occupying the oxygen vacancy.FESEM images revealed the granular growth in undoped ZnO which changed to diverse structures such as nano-bar and cluster of elongated grains.Significant improvement in room temperature ferromagnetism (RTFM) was osberved by codoping of Ce in Al-doped ZnO.The formation bound magnetic polarons (Ce 3+ -V o -Ce 3+ ) was the primary mechanism responsible for the improvement in RTFM.The band gap engineering with improved RTFM makes Zn(Al, Ce)O nanoparticles a suitable candidate for different optoelectronic devices and magnetic applications.

Figure 1 .
Figure 1.X-ray diffraction pattern of undoped (S1), Al doped (S2) and Ce co-doped (S3-S7) ZnO nanoparticles.The presence of multiple peaks in the XRD patterns indicates the polycrystalline nature of the nanoparticles.

Figure 3 .
Figure 3. Variation of angle of diffraction (2θ) corresponding to three major crystallographic planes i.e. (100), (002) and (101) for different samples.The 2θ values vary randomly with increase in dopant and co-dopant in the ZnO nanoparticles.
represent the M-H curves in the low magnetic field region.All the samples showed ferromagnetic behaviour.The hysteresis loops can be clearly seen in the inset.The undoped and Al-doped ZnO samples (S1 & S2) did not show the saturation.However, Ce co-doping (S3-S7) led the M-H curves

Figure 15 .
Figure 15.Variation of coercivity and retentivity in with dopant and co-dopant (sample number).

Table 1 .
Details of the samples.

Table 2 .
Orientation parameters for different crystallographic planes.

Table 3 .
Lattice parameter, u parameter, bond length and volume of unit cells as obtained from XRD patterns.

Table 4 .
Strain and crystallite size obtained using Scherrer equation along different planes and W-H plot.

Table 7 .
Band gap of different samples.