Highly Visible Photoluminescence from Ta-Doped Structures of ZnO Films Grown by HFCVD

Tantalum-doped ZnO structures (ZnO:Ta) were synthesized, and some of their characteristics were studied. ZnO material was deposited on silicon substrates by using a hot filament chemical vapor deposition (HFCVD) reactor. The raw materials were a pellet made of a mixture of ZnO and Ta2O5 powders, and molecular hydrogen was used as a reactant gas. The percentage of tantalum varied from 0 to 500 mg by varying the percentages of tantalum oxide in the mixture of the pellet source, by holding a fixed amount of 500 mg of ZnO in all experiments. X-ray diffractograms confirmed the presence of zinc oxide in the wurtzite phase, and metallic zinc with a hexagonal structure, and no other phase was detected. Displacements to lower angles of reflection peaks, compared with those from samples without doping, were interpreted as the inclusion of the Ta atoms in the matrix of the ZnO. This fact was confirmed by energy dispersive X-ray spectrometry (EDS), and X-ray diffraction (XRD) measurements. From scanning electron microscopy (SEM) images from undoped samples, mostly micro-sized semi-spherical structures were seen, while doped samples displayed a trend to grow as nanocrystalline rods. The presence of tantalum during the synthesis affected the growth direction. Green photoluminescence was observed by the naked eye when Ta-doped samples were illuminated by ultraviolet radiation and confirmed by photoluminescence (PL) spectra. The PL intensity on the Ta-doped ZnO increased from those undoped samples up to eight times.


Introduction
In recent years, ZnO has become one of the most closely studied materials, due to its very interesting properties in optoelectronic devices applications, such as room temperature lasers [1], light emitting diodes [2,3], ultraviolet (UV) detectors [4], field-emission displays [5,6], photonic crystals [7], solar cells [8][9][10], and sensing in the nano-size range [11,12].The control of its properties is critical in the context of novel applications.For this reason, a better understanding of the synthesis of microstructures/nanostructures with different morphologies is important and necessary to achieve objectives and applications on materials for the modern world [13].
Several techniques have been employed to make ZnO in thin films and nano-or microstructures; among them are radio-frequency magnetron sputtering, pulsed laser deposition, spray pyrolysis, and thermal evaporation flame synthesis [14][15][16][17].The properties of the materials seem to be dependent on the process and technique used to synthesize them.Developing new or modified methods to obtain those type of materials could report new insights on topics that are not yet understand for those types of oxides.
A profitable, simple, and easily scalable method to synthesize the material the hot filament chemical vapor deposition (HFCVD).This technique allows us to produce multifunctional structures on a larger scale.In addition, it is possible to control the material properties by only changing the synthesis conditions.
This technique has been used to grow a wide variety of films, including diamond [18], as well as amorphous silicon nitride [19,20], semiconductor compounds [21], and more.Recently, the technique has been used to grow several nanostructures of different materials, such as silicon-rich oxides [22], nanocrystalline silicon [23], graphene [24], molybdenum selenide [25], silicon carbide [26], and zinc sulfide [27] among others, for applications in new or improved devices, such as solar cells, sensors, and metal oxide semiconductors (MOS) structures [18,23].HFCVD technology is compatible with the current silicon-based technology, and it has advantages over other methods because it can produce materials at low cost, and on a wide substrate surface.
An important issue regarding applications in electronic and optoelectronic technologies is doping.ZnO exhibit naturally n-type conductivity because the presence of punctual defects, such as zinc interstitial and/or oxygen vacancies in its matrix [28,29].P-type is not easy to obtain, due to the compensation effect that impedes this [3,29].However, some other properties, such as optical emission can be changed when ZnO is doped.It is reported light emission, in the UV-visible range, can be obtained when the material is doped with transition metals.
It has been reported that green emission occurs at around 510 nm when copper atoms are incorporated in ZnO films [30][31][32].The luminescence was ascribed mainly to the existence of oxygen vacancies.Other transition metals when introduced to ZnO, such as Al, Li, Co, Mn, and Zn, have also been reported to exhibit emissions in different ranges; however, the wavelengths of emission and intensity depended on the process and technique of the growth [10,[33][34][35].Similarly, yellow emission was been observed when the ZnO synthesis was performed by chemical techniques.This emission was assigned to the existence of oxygen interstitials [36].
Ta is an atom that has been introduced in ZnO to change the properties of the material.It was found that the Ta atoms can substitute Zn atoms in the ZnO matrix, modifying the electrical and optical properties of the films.Some uses include photoactinic and TCO (transparent and conductive oxide) uses, among others.However, information on the changes on the photoluminescent properties with the incorporation of Ta in the ZnO material is missing.
To investigate on the effect of ZnO doping with Ta atoms, core-shell structures doped with tantalum were fabricated by the HFCVD method, and the optical (photoluminescence), structural, morphological, and electrical properties were studied and reported.By comparison, the undoped structures were also prepared, to observe the effect of the dopant.
The results showed that the properties of ZnO core-shell structures are a bit different from those of pure ZnO crystals.Photoluminescence (PL) intensity on Ta-doped ZnO increased from those undoped samples by up to eight times.This effect was ascribed to a change in the morphology of the structures, with the incorporation of tantalum in the process.The morphology of the structures in this work was found to be modified with the incorporation of tantalum in the process; this effect increases the specific area of the material, increasing the number of radiative centers.
Tantalum atoms do not appear to introduce additional radiative centers that contribute to the PL, but they affect the electricals of the ZnO grains.Ta is a donor in ZnO that contributes to reduce the resistivity of the ZnO grains in the shell.The results are in concordance with other authors.

Materials and Methods
Films obtained in this work were grown using a hot filament chemical vapor deposition system (HFCVD).The experimental homemade reactor consisted of elements showed in Figure 1.As can be seen in the figure, the system has a 2-way valve to introduce a reactant gas into the chamber, such as hydrogen or nitrogen; in this work, molecular hydrogen was only used as reactant to grow the films.Also, the system has a flowmeter and a pressure valve to control gases into the chamber; in this work, the reaction chamber was saturated with hydrogen, and the temperature in the substrate was kept at 800 • C. hydrogen or nitrogen; in this work, molecular hydrogen was only used as reactant to grow the films.Also, the system has a flowmeter and a pressure valve to control gases into the chamber; in this work, the reaction chamber was saturated with hydrogen, and the temperature in the substrate was kept at 800 °C.The reaction chamber (red box in Figure 1) consists of a cylindrical tube of quartz that was selected to support elevated temperatures in its interior.The tube has metal caps of stainless steel and O-rings at each end to ensure complete isolation between the inside of the tube under atmosphere, owing to elevated temperatures, reactant gases are highly pyrophoric.Principal components inside the reaction chamber are shown in Figure 2, and a sprinkler of stainless steel was used to fill the reaction chamber with the reactant gas.A filament of tungsten was used to raise the temperature inside the reaction chamber; in this work, the filament was fixed at 2000 °C.A pellet fabricated with a mixture of powders of ZnO (Mallinckrodt Chemicals CAS 1314-13-2, St. Louis, MO, USA) and Ta2O5 (ALDRICH CAS 1314-61-0, St. Louis, MO, USA) was used as a precursor to obtain the ZnO and ZnO:Ta doped films.The amount of ZnO powder was kept in 500 mg, and the amount of tantalum oxide was varied by a percentage of 0%, 10%, 20%, 40%, 50%, and 60%, for 500 mg of ZnO, i.e., 500 mg ZnO + 0 mg Ta2O5, 500 mg ZnO + 50 mg Ta2O5, 500 mg ZnO + 100 mg Ta2O5, and so on.After having the required weights of each powder, both powders of ZnO and Ta2O5 were mixed until a homogenous powder was obtained.After that, each mix was 1 ton-pressed.A disk shape of 10 mm in diameter and 5 mm in height were the result.All films were grown on a silicon n-type substrate.
The grown process was as follows: the reactor was sealed, and a flow of hydrogen was passed twice through it; next, the hydrogen flow was maintained inside the chamber and the filament is turned on; this process was kept by 3 min.
At a filament temperature of 2000 °C, the molecular hydrogen gas was dissociated into radicals that were highly reactive, which impinged on the pellet that was used as the source to produce intermediaries, such as Zn (vapor), Ta (vapor), and OH principally, although the additional reactions could occur with the gas and the substrate [37].Finally, by the reactor dynamics, the gases are transported to the substrate where the film is obtained, optical, and structural properties of this were studied by X-ray, SEM, energy dispersive X-ray spectrometry (EDS); the Hall Effect, and PL techniques.
The crystalline structure and the orientation of the crystallites were studied using X-ray diffraction (BRUKER D8 with Cu Kα radiation (λ = 1.541Å), Billerica, MA, USA); Raman spectra were taken with a RAM HR800 Raman spectrometer (HORIBA, Les Ulis, France) equipped with an Olympus BX41 microscope (Olympus, Tokyo, Japan); a 632.8 nm He-Ne laser was used as the excitation source.The surface morphology and composition of the films were examined with two systems; the low resolution was taken with a system SEM Jeol model 6610LV (JEOL Ltd., Tokyo, Japan) equipped with an INCA energy dispersive X-ray attached to this, and the high resolution was The reaction chamber (red box in Figure 1) consists of a cylindrical tube of quartz that was selected to support elevated temperatures in its interior.The tube has metal caps of stainless steel and O-rings at each end to ensure complete isolation between the inside of the tube under atmosphere, owing to elevated temperatures, reactant gases are highly pyrophoric.Principal components inside the reaction chamber are shown in Figure 2, and a sprinkler of stainless steel was used to fill the reaction chamber with the reactant gas.A filament of tungsten was used to raise the temperature inside the reaction chamber; in this work, the filament was fixed at 2000 • C. A pellet fabricated with a mixture of powders of ZnO (Mallinckrodt Chemicals CAS 1314-13-2, St. Louis, MO, USA) and Ta 2 O 5 (ALDRICH CAS 1314-61-0, St. Louis, MO, USA) was used as a precursor to obtain the ZnO and ZnO:Ta doped films.The amount of ZnO powder was kept in 500 mg, and the amount of tantalum oxide was varied by a percentage of 0%, 10%, 20%, 40%, 50%, and 60%, for 500 mg of ZnO, i.e., 500 mg ZnO + 0 mg Ta 2 O 5 , 500 mg ZnO + 50 mg Ta 2 O 5 , 500 mg ZnO + 100 mg Ta 2 O 5 , and so on.After having the required weights of each powder, both powders of ZnO and Ta 2 O 5 were mixed until a homogenous powder was obtained.After that, each mix was 1 ton-pressed.A disk shape of 10 mm in diameter and 5 mm in height were the result.All films were grown on a silicon n-type substrate.
The grown process was as follows: the reactor was sealed, and a flow of hydrogen was passed twice through it; next, the hydrogen flow was maintained inside the chamber and the filament is turned on; this process was kept by 3 min.
At a filament temperature of 2000 • C, the molecular hydrogen gas was dissociated into radicals that were highly reactive, which impinged on the pellet that was used as the source to produce intermediaries, such as Zn (vapor), Ta (vapor), and OH principally, although the additional reactions could occur with the gas and the substrate [37].Finally, by the reactor dynamics, the gases are transported to the substrate where the film is obtained, optical, and structural properties of this were studied by X-ray, SEM, energy dispersive X-ray spectrometry (EDS); the Hall Effect, and PL techniques.
The crystalline structure and the orientation of the crystallites were studied using X-ray diffraction (BRUKER D8 with Cu Kα radiation (λ = 1.541Å), Billerica, MA, USA); Raman spectra were taken with a RAM HR800 Raman spectrometer (HORIBA, Les Ulis, France) equipped with an Olympus BX41 microscope (Olympus, Tokyo, Japan); a 632.8 nm He-Ne laser was used as the excitation source.The surface morphology and composition of the films were examined with two systems; the low resolution was taken with a system SEM Jeol model 6610LV (JEOL Ltd., Tokyo, Japan) equipped with an INCA energy dispersive X-ray attached to this, and the high resolution was analyzed with a field emission scanning electron microscope, FESEM JSM 5400LV (JEOL Ltd., Tokyo, Japan) equipped with a NORAN energy dispersive X-ray spectrometer.The profilometry and roughness analyses of the films were taken with a surface profiler Dektak 150 (Veeco Instruments, Inc., Plainview, NY, USA).Photoluminescence measurements were performed at room temperature with a spectrofluorometer FluroMax 3 (Horiba, Ltd., Kyoto, Japan) that was equipped with an emission detector with high sensibility and a 150 W xenon lamp to excite the samples.The electrical properties of the films were determined with a Hall Effect system from Ecopia Company, model HMS-5000 (Ecopia, Waterloo, ON, Canada), using a magnetic field of 0.5 T. analyzed with a field emission scanning electron microscope, FESEM JSM 5400LV (JEOL Ltd., Tokyo, Japan) equipped with a NORAN energy dispersive X-ray spectrometer.The profilometry and roughness analyses of the films were taken with a surface profiler Dektak 150 (Veeco Instruments, Inc., Plainview, NY, USA).Photoluminescence measurements were performed at room temperature with a spectrofluorometer FluroMax 3 (Horiba, Ltd., Kyoto, Japan) that was equipped with an emission detector with high sensibility and a 150 W xenon lamp to excite the samples.The electrical properties of the films were determined with a Hall Effect system from Ecopia Company, model HMS-5000 (Ecopia, Waterloo, ON, Canada), using a magnetic field of 0.5 T.
The diffraction peaks from the ZnO:Ta samples are presented in Figure 3b-f.Some differences can be seen from that of reference sample (Figure 3a).As can be observed, the peak of ZnO in 2θ = 69.3°was could not peak in the other range; to get a better view of other peaks, we plotted the same diffraction patterns in a range from 2θ = 30° to 66° in Figure 4.In this graphs, it was possible to observe an additional diffraction peak arising at 2θ = 66.98° corresponding to the structure of ZnTa2O6, suggesting some that Ta atoms were incorporated into the ZnO lattice (pdf card 049-0746) [38].Because the atomic radii of Ta (0.78 Å) are very similar to Zn atoms (0.82 Å) these atoms easily replace some zinc atoms in the ZnO network without making abrupt changes in the ZnO lattice.In us case, a small level expansive stress in the ZnO lattice can be observed in 2θ = 63.1, corresponding to the plane (103); this stress caused this peak to shift to lower angles (red arrows in Figure 4c,d), but this peak could not be observed clearly in all diffractions [39,40].

Structural Characterization by XRD
The diffractograms of the films of polycrystalline zinc oxide doped with tantalum are shown in Figure 3, and a pure spectrum of ZnO is presented as a reference.For the pure ZnO material (Figure 3a), the presence of the diffraction peaks indicated that the material possessed a polycrystalline structure.Three well-distinguished diffraction peaks placed at 2θ = 32 • , 34.6 • , and 36.4 • , corresponding to planes (100), (002), and (101) were identified and matched with a hexagonal (wurtzite) ZnO crystalline structure (pdf card 075-0576).Additionally, some other diffractions peaks could be seen at 2θ = 47.The diffraction peaks from the ZnO:Ta samples are presented in Figure 3b-f.Some differences can be seen from that of reference sample (Figure 3a).As can be observed, the peak of ZnO in 2θ = 69.3• was could not peak in the other range; to get a better view of other peaks, we plotted the same diffraction patterns in a range from 2θ = 30 • to 66 • in Figure 4.In this graphs, it was possible to observe an additional diffraction peak arising at 2θ = 66.98 • corresponding to the structure of ZnTa 2 O 6 , suggesting some that Ta atoms were incorporated into the ZnO lattice (pdf card 049-0746) [38].Because the atomic radii of Ta (0.78 Å) are very similar to Zn atoms (0.82 Å) these atoms easily replace some zinc atoms in the ZnO network without making abrupt changes in the ZnO lattice.In us case, a small level expansive stress in the ZnO lattice can be observed in 2θ = 63.1, corresponding to the plane (103); this stress caused this peak to shift to lower angles (red arrows in Figure 4c,d), but this peak could not be observed clearly in all diffractions [39,40].It is interesting to note that for the principal peak of ZnO at 2θ = 36.4• in Ta-doped samples corresponding to (101), the plane changed strongly in intensity as the concentration of Ta was increased in the process (Figure 4).This result suggests that the high content of Ta may lead to the degradation of ZnO crystallinity.The other ZnO peaks also became small and broad, which could be an indication that tantalum atoms in the HFCVD process could modify the crystalline orientation of the growth, and from here modify the film morphology.
Diffraction peaks were analyzed with software HighScore Plus of PANalytical, every profile was approximated using a standard Caglioti function with a Pseudo-Voight profile fitting; the results are shown in Table 1.Analyzing the diffraction chart, it can be see the three important diffraction peaks of (100), (002), and (101), the planes of ZnO were diminished, with the introduction of Ta in the ZnO process, which it could mean the growth in that direction is inhibited while the growth along (201) is favored.According to the data obtained and presented in Table 1, it can be seen that the crystallite size of ZnO is bigger than the crystallites of ZnO:Ta doped, because when ZnO is doped with tantalum atoms, the crystallite size decrease strongly, and after this, the crystallite size decreases slowly, in this case, this fact could be an indication that the tantalum atoms are creating this effect.
According to the other reviewed reports, when ZnO is doped with tantalum, the crystallite size is modified, but there is no consensus of the increment or decrement in the crystallite size in ZnO:Ta-doped films; this fact seems to be more greatly affected by the technique used to grow the film; even when ZnO is doped with other materials, the crystallite size sometimes increases, while at other times it decreases [38,40,41].For example, Bang et al. reported the effect of lithium doping in ZnO films, they found that the (002) planes of ZnO were reduced as the Li content was increased; this was interpreted as a degradation of the ZnO crystallinity, while an increase in the grain size was observed as the concentration of lithium increased [42].Also, it has been reported that the presence of metallic zinc is associated with the existence of the core-shell structures, where the metallic zinc is the core and the ZnO nanocrystals are comprise the shell [37].

Raman Measurements
The room temperature Raman spectra for undoped and Ta-doped ZnO in the range of 100 to 1000 cm −1 are presented in Figure 5. Three main bands were observed for pure ZnO (Figure 5a).A peak labeled E 2 (High) at 437 cm −1 is known as the active optical phonon mode.The vibrational modes at 97 cm −1 , known as E 2 (LO) and 570 cm −1 , known as A 1 (LO), are also observable on the spectra.The presence of the three observed vibrations, in our measurements, confirmed that the material synthesized had wurtzite hexagonal structure [43][44][45][46].
It was reported that a broad Raman band at about 570 cm −1 may originate from defects related to the phonon mode, O-vacancy-related stable complexes [47], zinc interstitials, or from the disorder-activated B1 (High) silent mode [48] of w-ZnO.
It is well known that when metallic zinc particles are exposed to air or/and to an oxygen environment, a Zn 1+X O-defective zinc oxide with a Zn ion in the interstitial position is formed at around the metallic zinc [49], and a broad Raman band with a peak centered at 560 cm −1 was seen on the spectra.If it was assumed that the Raman band centered at 563 cm −1 , as seen in Figure 5a, had a contribution due to the formation of a defective oxide, then this result suggests us that our material was a core-shell type, where the core was metallic zinc and the shell consisted of ZnO grains.
Moreover, the undoped and Ta-doped ZnO samples were grown under an oxygen-deficient environment, and therefore, the synthesized material contained a great number of oxygen vacancies and zinc interstitials [29,50,51], and they could also contribute to the broadening of the Raman band observed at 563 cm −1 .
contribution due to the formation of a defective oxide, then this result suggests us that our material was a core-shell type, where the core was metallic zinc and the shell consisted of ZnO grains.
Moreover, the undoped and Ta-doped ZnO samples were grown under an oxygen-deficient environment, and therefore, the synthesized material contained a great number of oxygen vacancies and zinc interstitials [29,50,51], and they could also contribute to the broadening of the Raman band observed at 563 cm −1 .Raman spectra from the Ta-doped ZnO are shown in Figure 5b-d.It can be seen that all of the peaks were found to correspond to that from the undoped sample (Figure 5a); the only difference was the reduction in the intensity of the E2 (high) vibration, and the broadening of both the second order and A1 (LO) modes.
This observation could be an indication that Ta is encrusted in the ZnO matrix, and that it induces small disorder in the lattice.The reduced intensity in the E2 (high) peak might be related to the degradation on the crystalline quality as observed by X-ray diffraction; we can expect those tantalum atoms, when incorporated in the film, to induce change in the bonds strength between zinc and oxygen; therefore crystallinity is affected by this process, as have been observed in XRD results.

SEM and EDS Studies
Analysis by SEM was carried out to observe the film morphology, and the images are presented in Figure 6.It can be seen the surface morphology changed as the percentage of Ta2O5 increased in the pellet source.The picture for undoped ZnO film (Figure 6a) shows a morphology with a high density of big spherical shapes, and both the density and size of these structures reduced or even disappeared as the Ta2O5 content increased in the process (Figure 6b-f).Magnification of these samples (Figure 7) showed the shape details of their morphology.Raman spectra from the Ta-doped ZnO are shown in Figure 5b-d.It can be seen that all of the peaks were found to correspond to that from the undoped sample (Figure 5a); the only difference was the reduction in the intensity of the E 2 (high) vibration, and the broadening of both the second order and A 1 (LO) modes.
This observation could be an indication that Ta is encrusted in the ZnO matrix, and that it induces small disorder in the lattice.The reduced intensity in the E 2 (high) peak might be related to the degradation on the crystalline quality as observed by X-ray diffraction; we can expect those tantalum atoms, when incorporated in the film, to induce change in the bonds strength between zinc and oxygen; therefore crystallinity is affected by this process, as have been observed in XRD results.

SEM and EDS Studies
Analysis by SEM was carried out to observe the film morphology, and the images are presented in Figure 6.It can be seen the surface morphology changed as the percentage of Ta 2 O 5 increased in the pellet source.The picture for undoped ZnO film (Figure 6a) shows a morphology with a high density of big spherical shapes, and both the density and size of these structures reduced or even disappeared as the Ta 2 O 5 content increased in the process (Figure 6b-f).Magnification of these samples (Figure 7) showed the shape details of their morphology.
Micrographs taken on the Ta-doped ZnO samples (Figure 6b-f) showed that the roughness of the surface decreased as the content of Ta 2 O 5 was increased in the pellet source; measurements of profilometry were taken over the surface of each film, and the results obtained showed an irregular surface, as can be seen in that Figure 6; the results obtained showed depths with a minimum of 10 µm and a maximum of up to 150 µm, but surface roughness (RMS) analysis performed with this technique showed that surface tended to form a softer surface when the amount of tantalum was greater in the growth process; this RMS factor was plotted and it can be observed in Figure 8. Micrographs taken on the Ta-doped ZnO samples (Figure 6b-f) showed that the roughness of the surface decreased as the content of Ta2O5 was increased in the pellet source; measurements of profilometry were taken over the surface of each film, and the results obtained showed an irregular surface, as can be seen in that Figure 6; the results obtained showed depths with a minimum of 10 µm and a maximum of up to 150 µm, but surface roughness (RMS) analysis performed with this technique showed that surface tended to form a softer surface when the amount of tantalum was greater in the growth process; this RMS factor was plotted and it can be observed in Figure 8.A detailed SEM image from those samples with a 500× zoom is shown in Figure 7.The surface was composed by spheres and rods in a random form like a porous sponge.Additionally, a FESEM analysis of Figure 7a and b displayed the same type of spherical structures, but of different size.Some of those structures seemed hollow or even broken, showing a shell type structure, this shapes could be observed with a high resolution for ZnO in Figure 9(  Micrographs taken on the Ta-doped ZnO samples (Figure 6b-f) showed that the roughness of the surface decreased as the content of Ta2O5 was increased in the pellet source; measurements of profilometry were taken over the surface of each film, and the results obtained showed an irregular surface, as can be seen in that Figure 6; the results obtained showed depths with a minimum of 10 µm and a maximum of up to 150 µm, but surface roughness (RMS) analysis performed with this technique showed that surface tended to form a softer surface when the amount of tantalum was greater in the growth process; this RMS factor was plotted and it can be observed in Figure 8.A detailed SEM image from those samples with a 500× zoom is shown in Figure 7.The surface was composed by spheres and rods in a random form like a porous sponge.Additionally, a FESEM analysis of Figure 7a and b displayed the same type of spherical structures, but of different size.Some of those structures seemed hollow or even broken, showing a shell type structure, this shapes could be observed with a high resolution for ZnO in Figure 9(  A detailed SEM image from those samples with a 500× zoom is shown in Figure 7.The surface was composed by spheres and rods in a random form like a porous sponge.Additionally, a FESEM analysis of Figure 7a and b displayed the same type of spherical structures, but of different size.Some of those structures seemed hollow or even broken, showing a shell type structure, this shapes could be observed with a high resolution for ZnO in Figure 9(   From EDS analysis carried out on undoped ZnO films, only oxygen, and zinc elements were detected (Figure 10a).The relation of Zn/O of greater than one indicated that the material was out of stoichiometry, which meant a zinc-rich ZnO material.Our results were in agreement with those found by other authors for this type of systems [37,52,53].From EDS analysis carried out on undoped ZnO films, only oxygen, and zinc elements were detected (Figure 10a).The relation of Zn/O of greater than one indicated that the material was out of stoichiometry, which meant a zinc-rich ZnO material.Our results were in agreement with those found by other authors for this type of systems [37,52,53].From EDS analysis carried out on undoped ZnO films, only oxygen, and zinc elements were detected (Figure 10a).The relation of Zn/O of greater than one indicated that the material was out of stoichiometry, which meant a zinc-rich ZnO material.Our results were in agreement with those found by other authors for this type of systems [37,52,53].EDS results on Ta-doped ZnO samples (Figure 10b) indicate that it was a Zn-rich ZnO material, but there was also a clear contribution of the tantalum element around of 0.1% and 0.15% in atomic percentage (±0.5% in weight), this could be owing to the fact that tantalum atoms are mainly incorporated inside the lattice of ZnO, but no over surface of the structures obtained.
The data above could shed some light about the growth process.For undoped ZnO samples, the spheres could be formed as follows.The atomic hydrogen impacted on the pellet source producing OH − (gas) and Zn (gas) according to the reaction proposed in [37,39].After that, Zn atoms condensed in a vapor phase, and some Zn liquid drops were formed and they acted as nucleation sites, and OH ions are attached to the surface, forming a mixture of zinc in the liquid state, covered with a thin film of ZnO (solid), and they were deposited on the substrate, where it eventually coalesced and this drop increased in size.Further, due to the presence of OH, the zinc oxide cover increases in thickness, and a core(liquid)-shell(solid) structure was formed because of differences in the melting points 420 °C and 1975 °C for zinc and ZnO respectively.
As the growth was performed at a high temperature, the nuclei remained in a liquid state, producing a Zn gas pressure, due to the vapor-liquid equilibrium, and the Zn atoms diffused through the grain boundaries of the shell to release the internal pressure.Eventually the shell broke and Zn gas escaped, producing holes on the shell, as seen in Figure 9.This effect has been reported in the literature, and concentric core-shell (Zn/ZnO) balls have been synthesized by oxidizing ZnO powders at low temperatures in air, followed by annealing at temperatures above the zinc melting point [39,54,55].
The existence of this crystalline zinc forming the core were detected by X-ray diffraction results (Figure 3) and by EDS, where it obtained a relation Zn/O of 69% at 31%, this means a zinc-rich ZnO material.
On the other hand, for Ta-doped ZnO samples; the formation of these structures could be explained as a similar form as undoped but instead of forming Zn liquid drops, it forms Zn-Ta alloy liquid drops.Because the difference of the fusion point between Zn (420 °C) and Ta (2996 °C) the properties of the alloy changes and it can affect the size and form of the deposited material, as seen in the sequence in Figure 6b-f.The higher the percentage of Ta2O5, the small structures are obtained.The shell is composed of small polyhedral crystals of ZnO material according to SEM, X-ray, and Raman measurements.
The crystals forming the shell were also affected by the introduction of Ta, as can be observed from FESEM in Figure 11b.The ZnO composing the shell had sheet-like polycrystalline shapes with a smaller crystallite size than the undoped ZnO (see Table 1).
The effect of both the decreased grain size and the change in the shape of the crystals in the ZnO:Ta-doped films was also observed and reported by different authors [40,41,[56][57][58], and it was in agreement with our experiments.This effect could be owing to the fact that tantalum atoms create a EDS results on Ta-doped ZnO samples (Figure 10b) indicate that it was a Zn-rich ZnO material, but there was also a clear contribution of the tantalum element around of 0.1% and 0.15% in atomic percentage (±0.5% in weight), this could be owing to the fact that tantalum atoms are mainly incorporated inside the lattice of ZnO, but no over surface of the structures obtained.
The data above could shed some light about the growth process.For undoped ZnO samples, the spheres could be formed as follows.The atomic hydrogen impacted on the pellet source producing OH − (gas) and Zn (gas) according to the reaction proposed in [37,39].After that, Zn atoms condensed in a vapor phase, and some Zn liquid drops were formed and they acted as nucleation sites, and OH ions are attached to the surface, forming a mixture of zinc in the liquid state, covered with a thin film of ZnO (solid), and they were deposited on the substrate, where it eventually coalesced and this drop increased in size.Further, due to the presence of OH, the zinc oxide cover increases in thickness, and a core(liquid)-shell(solid) structure was formed because of differences in the melting points 420 • C and 1975 • C for zinc and ZnO respectively.
As the growth was performed at a high temperature, the nuclei remained in a liquid state, producing a Zn gas pressure, due to the vapor-liquid equilibrium, and the Zn atoms diffused through the grain boundaries of the shell to release the internal pressure.Eventually the shell broke and Zn gas escaped, producing holes on the shell, as seen in Figure 9.This effect has been reported in the literature, and concentric core-shell (Zn/ZnO) balls have been synthesized by oxidizing ZnO powders at low temperatures in air, followed by annealing at temperatures above the zinc melting point [39,54,55].
The existence of this crystalline zinc forming the core were detected by X-ray diffraction results (Figure 3) and by EDS, where it obtained a relation Zn/O of 69% at 31%, this means a zinc-rich ZnO material.
On the other hand, for Ta-doped ZnO samples; the formation of these structures could be explained as a similar form as undoped but instead of forming Zn liquid drops, it forms Zn-Ta alloy liquid drops.Because the difference of the fusion point between Zn (420 • C) and Ta (2996 • C) the properties of the alloy changes and it can affect the size and form of the deposited material, as seen in the sequence in Figure 6b-f.The higher the percentage of Ta 2 O 5 , the small structures are obtained.The shell is composed of small polyhedral crystals of ZnO material according to SEM, X-ray, and Raman measurements.
The crystals forming the shell were also affected by the introduction of Ta, as can be observed from FESEM in Figure 11b.The ZnO composing the shell had sheet-like polycrystalline shapes with a smaller crystallite size than the undoped ZnO (see Table 1).
The effect of both the decreased grain size and the change in the shape of the crystals in the ZnO:Ta-doped films was also observed and reported by different authors [40,41,[56][57][58], and it was in agreement with our experiments.This effect could be owing to the fact that tantalum atoms create a compression stress inside the lattice of ZnO, and they prevents an increase in these big shapes, as generated by the difference of ionic radii between Zn and Ta, as commented before in the XRD results.In fact, this phenomenon also has been observed when ZnO is doped with tantalum by other techniques [56], or when ZnO is doped with other materials such as aluminum and lithium [35,42].compression stress inside the lattice of ZnO, and they prevents an increase in these big shapes, as generated by the difference of ionic radii between Zn and Ta, as commented before in the XRD results.
In fact, this phenomenon also has been observed when ZnO is doped with tantalum by other techniques [56], or when ZnO is doped with other materials such as aluminum and lithium [35,42].The synthesis of smaller structures means that somehow, the relation between the surface/volume of the film will increase, and this fact will be reflected in an increment of reactions between the existing species on the surface of structures, and also a possible increment in the adsorption and resorption of the species in these kind of porous films, and moreover, an increment of cations or anions on the surface of the film, as has been reported in other works; in fact, the role of the surface area has been observed to affect various mechanisms such as gas sensors, biological cells, and the effects of fluorescence emission [28,[59][60][61].

Photoluminescence
The photoluminescence (PL) technique is a useful way to obtain information about the energy states of the impurities and defects of a material.Figure 12 shows the PL properties from our samples.As can be observed in that figure, all films exhibited two main bands, one sharp curve of around 380 nm, and another one with a more widened curve centered around 495 nm and extending from 425 nm to 625 nm.The PL intensity first increased as the percentage of Ta increased, and after that it decreased.
Photoluminescent emissions in ZnO were grouped into two main bands denominating the emissions near-band-edge (NBE), and deep-level emissions (DLE) corresponding to the transition band-to-band and the transition through to the deep centers, with energy levels in the forbidden gap, respectively.Emissions NBE in ZnO have been attributed to free excitation recombination, and they are observed at around the ultraviolet range.
Nevertheless, emissions DLE in ZnO have been assigned to the presence of various point defects, either intrinsic or extrinsic, which introduce deep levels in the bandgap and are assumed to be responsible for PL band broadening in the range between the blue and red emission [3,28,29,34,62].
As can be seen in Figure 12, the main PL band is the DLE band, and it increases in intensity by up to 60% as the Ta2O5 is incorporated into the growth process, and after that it reduces.Moreover, all bands are peaked at the same wavelength, independently of the percentage of Ta2O5, which means the doping with Ta does not seem to play a role in the generation of radiative deep centers other than for those of pure ZnO.The synthesis of smaller structures means that somehow, the relation between the surface/volume of the film will increase, and this fact will be reflected in an increment of reactions between the existing species on the surface of structures, and also a possible increment in the adsorption and resorption of the species in these kind of porous films, and moreover, an increment of cations or anions on the surface of the film, as has been reported in other works; in fact, the role of the surface area has been observed to affect various mechanisms such as gas sensors, biological cells, and the effects of fluorescence emission [28,[59][60][61].

Photoluminescence
The photoluminescence (PL) technique is a useful way to obtain information about the energy states of the impurities and defects of a material.Figure 12 shows the PL properties from our samples.As can be observed in that figure, all films exhibited two main bands, one sharp curve of around 380 nm, and another one with a more widened curve centered around 495 nm and extending from 425 nm to 625 nm.The PL intensity first increased as the percentage of Ta increased, and after that it decreased.
Photoluminescent emissions in ZnO were grouped into two main bands denominating the emissions near-band-edge (NBE), and deep-level emissions (DLE) corresponding to the transition band-to-band and the transition through to the deep centers, with energy levels in the forbidden gap, respectively.Emissions NBE in ZnO have been attributed to free excitation recombination, and they are observed at around the ultraviolet range.
Nevertheless, emissions DLE in ZnO have been assigned to the presence of various point defects, either intrinsic or extrinsic, which introduce deep levels in the bandgap and are assumed to be responsible for PL band broadening in the range between the blue and red emission [3,28,29,34,62].
As can be seen in Figure 12, the main PL band is the DLE band, and it increases in intensity by up to 60% as the Ta 2 O 5 is incorporated into the growth process, and after that it reduces.Moreover, all bands are peaked at the same wavelength, independently of the percentage of Ta 2 O 5 , which means the doping with Ta does not seem to play a role in the generation of radiative deep centers other than for those of pure ZnO.To investigate the possible origin of the green luminescence observed in this type of films, many reports that try to describe this phenomena exactly, were reviewed, but there is no consensus, as the green emission has been observed on several films doped with Cu or Co atoms, on films with zinc vacancies (VZn), oxygen vacancies (VO), interstitial zinc ions (Zni), oxygen antisites (OZn), and transitions Zni→VZn, for these reasons, researchers came to the conclusion that this emission is a combination of several deep levels [62].For example, in reference [63], it was reported that when ZnO films were annealed at temperatures higher than 800 °C, oxygen vacancies and zinc vacancies are generated, producing green emissions at 490 and 530 nm.In addition, a good correlation between the green emission and singly ionized oxygen vacancies was observed in commercial ZnO powders, annealed in forming gas (N2:H2) or O2 at temperatures ranging from 500 to 1050 °C [64].
By other side, has been reported that either zinc vacancy and oxygen vacancy could contribute to green emission through shallow donors, and even Rodnyi and Khodyuk [62] reported that it is possible to assume the existence of donors with two levels (ground level and excited one) instead of two kinds of shallow donors.It has also been observed that in films of ZnO with an excess of oxygen, the maximum of the green luminescence is around 2.35 eV, and that zinc vacancies are responsible for the emission, but in ZnO films with an excess of zinc, the maximum of luminescence is around 2.53 eV, and in this case, oxygen vacancies acting as deep acceptor centers are responsible for the observed emission.
In our case, the DLE emission observed was centered at 2.53 eV, and according to the results presented above, the films contained both zinc in excess and oxygen vacancies; therefore, those defects could be involved in the PL phenomena.
As can be observed in Figure 12, all DLE emissions obtained in these films were centered at 2.50 eV (495 nm), and except for curve f, all films presented a broadening from 414 nm to 630 nm, and as mentioned above, the most intense peak could be attributed mainly to oxygen vacancies, and the broadening of curve could be assigned to a combination of different deep levels, either intrinsic or extrinsic.
Tantalum atoms introduced shallow donors in ZnO and modified its electrical conductivity, but in this case, those centers did not seem to play any effect on the PL phenomena, because zinc interstitials and oxygen vacancies are the defects involved in the PL phenomena.
Because the effect of the introduction of Ta atoms in ZnO is to increase the ratio of the volume to surface area as the structures area become smaller, the increase on the PL intensity could be attributed to a variation in the surface area of these films, as has been shown in the SEM section.To investigate the possible origin of the green luminescence observed in this type of films, many reports that try to describe this phenomena exactly, were reviewed, but there is no consensus, as the green emission has been observed on several films doped with Cu or Co atoms, on films with zinc vacancies (V Zn ), oxygen vacancies (V O ), interstitial zinc ions (Zn i ), oxygen antisites (O Zn ), and transitions Zn i →V Zn , for these reasons, researchers came to the conclusion that this emission is a combination of several deep levels [62].For example, in reference [63], it was reported that when ZnO films were annealed at temperatures higher than 800 • C, oxygen vacancies and zinc vacancies are generated, producing green emissions at 490 and 530 nm.In addition, a good correlation between the green emission and singly ionized oxygen vacancies was observed in commercial ZnO powders, annealed in forming gas (N 2 :H 2 ) or O 2 at temperatures ranging from 500 to 1050 • C [64].
By other side, has been reported that either zinc vacancy and oxygen vacancy could contribute to green emission through shallow donors, and even Rodnyi and Khodyuk [62] reported that it is possible to assume the existence of donors with two levels (ground level and excited one) instead of two kinds of shallow donors.It has also been observed that in films of ZnO with an excess of oxygen, the maximum of the green luminescence is around 2.35 eV, and that zinc vacancies are responsible for the emission, but in ZnO films with an excess of zinc, the maximum of luminescence is around 2.53 eV, and in this case, oxygen vacancies acting as deep acceptor centers are responsible for the observed emission.
In our case, the DLE emission observed was centered at 2.53 eV, and according to the results presented above, the films contained both zinc in excess and oxygen vacancies; therefore, those defects could be involved in the PL phenomena.
As can be observed in Figure 12, all DLE emissions obtained in these films were centered at 2.50 eV (495 nm), and except for curve f, all films presented a broadening from 414 nm to 630 nm, and as mentioned above, the most intense peak could be attributed mainly to oxygen vacancies, and the broadening of curve could be assigned to a combination of different deep levels, either intrinsic or extrinsic.
Tantalum atoms introduced shallow donors in ZnO and modified its electrical conductivity, but in this case, those centers did not seem to play any effect on the PL phenomena, because zinc interstitials and oxygen vacancies are the defects involved in the PL phenomena.
Because the effect of the introduction of Ta atoms in ZnO is to increase the ratio of the volume to surface area as the structures area become smaller, the increase on the PL intensity could be attributed to a variation in the surface area of these films, as has been shown in the SEM section.

Hall Measurements
It is well-established that ZnO has natural type-n conductivity because of the formation of intrinsic defects, such as oxygen vacancies and zinc interstitials, during the synthesis.Oxygen vacancies can be easily formed and can contribute with two free electrons to the conduction band.However, it has been shown that those defects are actually deep donors, and they cannot contribute to the n-type conductivity.Therefore, zinc interstitials seem like the most probable defect acting as a donor in undoped ZnO films.
Carrier concentration, mobility, and the resistivity of films measured by the Hall-Vander Paw technique, are presented in Figure 13.The majority carrier mobility first increases up to a maximum value, and afterwards it reduces, as the Ta 2 O 5 percentage is increased in the growth, while the carrier concentration increases as the Ta 2 O 5 percentage increases.

Hall Measurements
It is well-established that ZnO has natural type-n conductivity because of the formation of intrinsic defects, such as oxygen vacancies and zinc interstitials, during the synthesis.Oxygen vacancies can be easily formed and can contribute with two free electrons to the conduction band.However, it has been shown that those defects are actually deep donors, and they cannot contribute to the n-type conductivity.Therefore, zinc interstitials seem like the most probable defect acting as a donor in undoped ZnO films.
Carrier concentration, mobility, and the resistivity of films measured by the Hall-Vander Paw technique, are presented in Figure 13.The majority carrier mobility first increases up to a maximum value, and afterwards it reduces, as the Ta2O5 percentage is increased in the growth, while the carrier concentration increases as the Ta2O5 percentage increases.The minimum value of the carrier concentration is for undoped samples (Figure 13a); therefore, the increase of the concentration was for the incorporation of the Ta atoms in the Zn positions of ZnO, and they contributed with electrons to the conduction band.The carrier concentration increased as the amount of Ta 2 O 5 was incremented in the source pellet; this could be a clear reason for the increment in carrier concentration in ZnO:Ta-doped films, since if a tantalum atom replaces a zinc atom, this tantalum could have an oxidation state of Ta1 + , Ta2 + , or Ta3 + with regards to Zn; this means that tantalum is capable of contributing with free electrons.
The mobility (Figure 13a) initially increased, with Ta 2 O 5 reaching a maximum value in the film fabricated with 50 mg of Ta 2 O 5 in the source; after that, it started to decrease.This effect has also been observed by Subha et al. [57,58], and is attributed to an excess of carriers in the ZnO lattice.A small increment observed in mobility, was owing to a high number of free carriers that were generated by tantalum atoms, and the relatively greater crystal size of the shell.Nevertheless, this mobility was reduced as the carrier concentration increased, because this carrier created a high number of collisions between them, besides the increase in the number of grain boundaries, due to a reduction in the crystal size in the shell.There are reports of ZnO with other dopants, in which it has been observed that foreign atoms inside the lattice of ZnO could act as charge trapping sites [33,42], which prevents the mobility of these free chargers.In our case, Ta atoms could be creating a similar effect, and therefore, the mobility in our films decreases with the amount of Ta 2 O 5 in the source pellet.
With regard to the resistivity of the films, it can be appreciated in Figure 13b that initially for ZnO film, this had the highest value, but when these were doped with tantalum atoms, resistivity was around 0.5-1.0Ωcm, but this tended to decrease, and this effect was attributed to an excess of carriers generated by donors from Ta atoms incorporated in the ZnO lattice.
For the conduction process in these type of films and to attempt to explain the results above the following, a model is proposed in Figure 14.The minimum value of the carrier concentration is for undoped samples (Figure 13a); therefore, the increase of the concentration was for the incorporation of the Ta atoms in the Zn positions of ZnO, and they contributed with electrons to the conduction band.The carrier concentration increased as the amount of Ta2O5 was incremented in the source pellet; this could be a clear reason for the increment in carrier concentration in ZnO:Ta-doped films, since if a tantalum atom replaces a zinc atom, this tantalum could have an oxidation state of Ta1 + , Ta2 + , or Ta3 + with regards to Zn; this means that tantalum is capable of contributing with free electrons.
The mobility (Figure 13a) initially increased, with Ta2O5 reaching a maximum value in the film fabricated with 50 mg of Ta2O5 in the source; after that, it started to decrease.This effect has also been observed by Subha et al. [57,58], and is attributed to an excess of carriers in the ZnO lattice.A small increment observed in mobility, was owing to a high number of free carriers that were generated by tantalum atoms, and the relatively greater crystal size of the shell.Nevertheless, this mobility was reduced as the carrier concentration increased, because this carrier created a high number of collisions between them, besides the increase in the number of grain boundaries, due to a reduction in the crystal size in the shell.There are reports of ZnO with other dopants, in which it has been observed that foreign atoms inside the lattice of ZnO could act as charge trapping sites [33,42], which prevents the mobility of these free chargers.In our case, Ta atoms could be creating a similar effect, and therefore, the mobility in our films decreases with the amount of Ta2O5 in the source pellet.
With regard to the resistivity of the films, it can be appreciated in Figure 13b that initially for ZnO film, this had the highest value, but when these were doped with tantalum atoms, resistivity was around 0.5-1.0Ωcm, but this tended to decrease, and this effect was attributed to an excess of carriers generated by donors from Ta atoms incorporated in the ZnO lattice.For the conduction process in these type of films and to attempt to explain the results above the following, a model is proposed in Figure 14.
First of all, we consider that the conduction of carriers in core shell structures could be a bit different from the pure crystals.As seen in the SEM pictures (Figure 9), the core shell structures were composed by a metallic core, and a shell formed by the oxide crystals.Due to the high temperature of the processes (above of the melting point of the zinc ~420 °C), the zinc atoms diffused through the grain boundaries to release the internal pressure, creating zinc-rich regions between grains.These zinc atoms became suboxides with a higher concentration of zinc interstitials at the surface than in the grain volume.The effect created a surface path for electric conduction, while the grain volume with lower concentrations of interstitials had a low contribution to the conductivity.
When tantalum was introduced into the ZnO matrix, the concentration of the carriers in the volume increased, and the grain was capable of being a transport carrier, and the resistivity of the film decreased, as shown in Figure 13.First of all, we consider that the conduction of carriers in core shell structures could be a bit different from the pure crystals.As seen in the SEM pictures (Figure 9), the core shell structures were composed by a metallic core, and a shell formed by the oxide crystals.Due to the high temperature of the processes (above of the melting point of the zinc ~420 • C), the zinc atoms diffused through the grain boundaries to release the internal pressure, creating zinc-rich regions between grains.These zinc atoms became suboxides with a higher concentration of zinc interstitials at the surface than in the grain volume.The effect created a surface path for electric conduction, while the grain volume with lower concentrations of interstitials had a low contribution to the conductivity.
When tantalum was introduced into the ZnO matrix, the concentration of the carriers in the volume increased, and the grain was capable of being a transport carrier, and the resistivity of the film decreased, as shown in Figure 13.

Figure 1 .
Figure 1.Schematic diagram of the hot filament chemical vapor deposition (HFCVD) homemade system used to grown films in this work.

Figure 1 .
Figure 1.Schematic diagram of the hot filament chemical vapor deposition (HFCVD) homemade system used to grown films in this work.

Figure 2 .
Figure 2. The principal components inside of the reaction chamber.

Figure 2 .
Figure 2. The principal components inside of the reaction chamber.

Figure 7 .
Figure 7. Zoom of 500× on the film of (a) ZnO, and (b) ZnO:Ta doped with 300 mg of Ta2O5 on pellet.

Figure 7 .
Figure 7. Zoom of 500× on the film of (a) ZnO, and (b) ZnO:Ta doped with 300 mg of Ta2O5 on pellet.

Figure 7 .
Figure 7. Zoom of 500× on the film of (a) ZnO, and (b) ZnO:Ta doped with 300 mg of Ta 2 O 5 on pellet.

Figure 8 .
Figure 8. Roughness dependence in films obtained according to the amount of Ta in the source pellet.

Figure 8 . 18 Figure 8 .
Figure 8. Roughness dependence in films obtained according to the amount of Ta in the source pellet.

Figure 10 .
Figure 10.X-ray spectrometry (EDS) analysis for (a) undoped ZnO film, and (b) ZnO:Ta-doped film grown with 300 mg of Ta2O5 in the pellet source.

Figure 10 .
Figure 10.X-ray spectrometry (EDS) analysis for (a) undoped ZnO film, and (b) ZnO:Ta-doped film grown with 300 mg of Ta 2 O 5 in the pellet source.

Figure 11 .
Figure 11.FESEM measurements on the shell from (a) undoped ZnO film, and (b) ZnO:Ta-doped film obtained with 300 mg of Ta2O5 in the pellet.

11 .
FESEM measurements on the shell from (a) undoped ZnO film, and (b) ZnO:Ta-doped film obtained with 300 mg of Ta 2 O 5 in the pellet.

Figure 14 .
Figure 14.A model of conductivity through the crystal in the film.Two possible paths for conduction: intergrain (R1) and through the volume of the grains (R2).

Figure 14 .
Figure 14.A model of conductivity through the crystal in the film.Two possible paths for conduction: intergrain (R 1 ) and through the volume of the grains (R 2 ).

Table 1 .
Lattice parameters and crystallite size for ZnO and ZnO:Ta-doped films obtained.