Tuning of Structural and Morphological Characteristics of V2O5 Thin Films Using Low Energy 16 keV N + for Optical and Wetting Applications

Effect of nitrogen (N+) ion implantation on the morphological, structural, optical, and compositional properties of vanadium pentoxide (V2O5) thin films grown on glass substrates is studied. Surface morphology shows the formation of grains and the growth dynamics is governed by roughness (α) and growth (β) exponents. X-ray diffraction studies reveal that V2O5 exists in a hybrid form, with properties of both the orthorhombic and tetragonal phases. Ion implantation induces defects and strain in V2O5 thin films causing a reduction in crystalline properties and deformation in the β-phase with a corresponding change in crystallite size. Contact angle wetting properties are found to be co-related with fractal growth of the films under ion implantation. Oxygen vacancies and electron scattering/trapping centres are revealed to have increased after N+ implantation, leading to a smaller bandgap in the thin films. The benefits of decreasing the optical band-gap of V2O5 thin films for optical applications are outlined in the present work.

Vanadium pentoxide (V 2 O 5 ) is a transition metal oxide that has received a lot of attention lately as it is the most stable oxide because of the saturated oxidation state of V 5+ . It occurs in different polymorphs such as α, β, γ, and δ-V 2 O 5 , out of which, orthorhombic α-V 2 O 5 is the most stable phase. 1 Recently, it has become a promising material for various applications including chemical sensing, optoelectronics, cathode material in batteries, electrochromic devices and catalysis 1-4 due to its multi-layer structure, wide optical range gap, good chemical and thermal stability, excellent electrical thermal properties, etc Solar cells use V 2 O 5 as an electrode material, while nano-electronics use it as a dielectric material. In optoelectronics V 2 O 5 thin films are used for making light-emitting diodes (LEDs) and other devices. Also, high electrical conductivity, high optical reflectivity, and low dielectric constant makes them useful for making energy-saving windows or smart clothing. 5 V 2 O 5 undergoes a reversible change from being an insulator to a conductor when exposed to light or heat which makes it useful for applications where switching between these two states is desired. For example, V 2 O 5 -based window coatings can reflect heat when the outside temperature is high, preventing the building from using air conditioning much. In the winter season when outside temperature is low, it may allow Sunlight to enter, thus heating the interior naturally. This same property can also be utilized in smart clothing. Such characteristics and rationale compelled us to concentrate on a deeper investigation of the surface and microstructure characteristics of V 2 O 5 as most of the above-mentioned applications depend on these two characteristics. Moreover, it is necessary to monitor crucial optical factors that are connected to diverse technological applications.
Nanostructured vanadium oxide thin films have improved performance in devices due to their distinct chemical and physical properties when compared to bulk because of their unique morphology and large surface area. 5 Structural, optical, electrical properties, and the oxidation states of vanadium oxide thin films depend on ambient conditions and preparation techniques.V 2 O 5 thin films can be deposited through many different techniques such as thermal evaporation, 6 pulsed laser deposition 7 and d.c. magnetron sputtering. 8 The deposited thin films are amorphous and become crystalline if subjected to heat treatment. 9 Ion implantation is a process in which ions are accelerated and then injected into a solid. These ions may alter the material's physical, chemical, or electrical properties. The ion implantation is achieved by means of ion beams which are generated through electron collisions in a plasma, transforming the atoms into ions and then concentrated into a stream using magnets and propelled towards the substrate by a voltage gradient. With this method, an ion beam (between 5 and 200 keV) can be focused on the surface of the component. The depths of implants typically range from 0.1 to 0.3 μm. Ion implantation is extensively used to modify the surface of a material. 10 Surface dimensions and fractals are also affected by the structure of crystallites following ion implantation, in addition to the films' physical characteristics. Changes in surface morphology have a direct effect on a solid's wettability. Wetting qualities on such a surface have been investigated by Carbone et al., 11 who used a basic sinusoidal wave profile to investigate the surface's roughness. Yang et al. 12 noticed that the wettability of silicon surfaces responds to ripple patterns in an anisotropic fashion. If the contact angle (CA) of a liquid droplet over a surface is larger than 90 degrees, we say that the surface is hydrophobic for that liquid. Surface wetting qualities are known to be a function of interface width, co-relation length, and fractal dimension, all of which are intrinsic to nanostructures. Contact angle measurements on micron-to nanometer-scale rough surfaces have been carefully studied, revealing that increased surface roughness improves wettability. Wenzel, in 1936, 13 presented a compelling case for the significance of surface roughness in the equation cos θ r = r cos θ f , where the contact angles of a liquid droplet on a rough and flat surface are given by θ r and θ f , and the roughness factor, denoted by "r," is the ratio of the actual rough to the expected flat area. According to Wenzel's model, the roughness of a surface has a greater effect on how hydrophilic (hydrophobic) it is.
In the present paper, we have studied the effect of ion implantation of different doses on structural, optical and surface properties of V 2 O 5 thin films using glancing angle X-ray diffraction (GAXRD), z E-mail: tanuj.nsm@cujammu.ac.in Fourier Transform Infra-Red (FTIR), UV-vis spectroscopy, atomic force microscopy (AFM), fractal characterizations and contact angle measurements. This paper might be of interest to the broad research community working in the area as well as to the industries mentioned above. Experiment V 2 O 5 thin films were deposited on corning glass substrates in a vacuum chamber using thermal evaporation technique. The source material V 2 O 5 (purity 99.997%) was evaporated from the tungsten small boat. The chamber pressure during deposition was maintained at 5 × 10 −6 mbar. The distance between the source material and the substrate was fixed at 35 cm. The deposited films had thicknesses of 300 nm. After the deposition, the films were thermally annealed at 500°C for 6.5 h. The thermal annealing process was achieved by using muffle furnace system (upto1200°C) with changing the annealing temperature from 100°C to 500°C under air pressure conditions. The 16 keV N + ion implantation at fluencies of 5 × 10 11 , 5 × 10 12 , 1 × 10 13 and 1x10 14 ions/cm 2 into crystallised V 2 O 5 thin films was carried out using an indigenously made tabletop accelerator at the Inter-University Accelerator Centre, New Delhi, India. The structural phase analysis of the annealed and implanted samples were carried out using an X-ray diffractometer Rigaku Miniflex 600 by using a Cu K α radiation source (λ=1.5406 Å ) operated in 40 kV and 20 mA. The FTIR analysis was carried out by using Shimadzu IR infinity in KBR mode in the range 400-4000 cm −1 . The optical characteristics and band gap of the thin films were examined by using a UV-vis-NIR spectrophotometer (Model-Shimadzu 3600i Plus) at Central University of Jammu, Jammu. The V 2 O 5 thin films' thickness was determined using a diamond needle stylus profilometer (Mitutoyo SJ-301).

Methodology
Relation between roughness exponent (α) and fractal dimension (D f ).-A fractal geometrical dimension (or fractal dimensionality) is a statistical indicator used in mathematics that offers information on the evolution of surface patterns as a function of size. It represents the ability of an embedded design to completely fill the surrounding surface. It is given by is the dimension of the surface area. Roughness exponent α = 1 describes a perfectly smooth surface. If the roughness exponent is decreased, the fractal dimensions grow and the surface takes up a larger percentage of the embedding space.
Fractal dimension and contact angle.-The relationship between the contact angle θ 0 of a liquid drop and a flat surface is well explained by the Young's equation, 14 which is given by where the interfacial free energies per unit area for the solid-liquid, liquid-vapor, and solid-vapor interfaces are denoted byγ sl , γ lv, and γ sv , respectively. It can be obtained from the total free energy of the system. If γ sl > γ sv , the contact angle is greater than 90 0 and less than 90 0 if γ sl > γ sv .
For a rough surface, the effective area of the solid surface increases by a factor r which was claimed by Wenzel. 13  where the contact angle for the rough surface is denoted by θ 0 . Thus, as suggested by the Wenzel model, surface becomes more wettable on increasing the roughness of a surface.
Some vapour remains trapped between the liquid drop and the cavities of a rough surface as suggested by the Classie model. The solid and liquid are in contact by the fractional area which is denoted by φ. Thus, Eq. 3 gets modified into On comparing Eq. 5 with Eq. 3, Now, let the area of the solid-liquid surfaces in contact and the area of the solid-vapour surfaces in contact be denoted by f 1 , f 2 respectively. On considering this assumption, Eq. 5 gets modified into giving, Onda et al. 15 suggested that, to a first-order approximation for a fractal surface of dimension D, r in Eq. 7 is given by where L and l are the upper and lower cut-offs of the fractal behavior. Then Eq. 7 becomes The right-hand side of this equation can be quite big, whereas the left-hand side is limited, hence this formula can only be a first approximation. In order to make additional adjustments to Eq. 9, we Figure 1. Schematic of the experimental path followed. must take into account the vapour trapped between the solid-liquid interface and the liquid trapped between the solid-vapour interface. This means that the contact angle is determined by the relative areas of the various interfaces as well as the range of fractality and fractal dimension.

Results and Discussion
XRD study.-X-ray diffraction (XRD) patterns are shown in The peak's orientation parallel to the surface and the observed decrease in relative intensity when ion fluence is increased from 1 × 10 13 to 1 × 10 14 ions cm −2 are both indicative of a loss of crystalline structure. The XRD patterns of the irradiation V 2 O 5 samples were different from the annealed sample, both in terms of the relative intensities of the peaks and their placements. An incremental and persistent reduction in the interplanar gap, or peak-to-trough movement, has been detected. Samples implanted with a fluence of 1 × 10 14 ions/cm 2 show the greatest reduction of peak intensity. Degradation of the crystal structure caused by the creation of defects (O, V-vacancies) after 16 keV N + ion implantation results in a reduction in peak intensities. Analysis of XRD shows the minor variation in grain size with fluence. The annealed V 2 O 5 : Gl film has a mean grain size of about 70.5 nm, according to the Debye-Scherrer formula, while samples treated with N + at ion fluencies 1 × 10 13 and 1 × 10 14 ions/cm 2 have grain sizes of 70.2 nm and 70.3 nm respectively. According to Debye-Scherrer formula: Where D is the crystallite (grain) size, β is the full width at half maximum (FWHM) of a diffraction peak with θ angle, λ=1.54056 Å is the X-ray wavelength, and k is the shape factor. Figure 3 displays the W-H plot, or β cosθ Vs 4 sinθ plot, for annealed V 2 O 5 thin film. Synthesized particles contain tensile strains, as evidenced by the positive slope of the WH plot. The crystallite size and lattice microstrain are defined by the intercept and slope values obtained from the linear fit of the plot, and these values are displayed in Table I. Similar calculations are done for the N + implanted samples as well (not shown here). It seems that the formation of grains in thin films of V 2 O 5 is facilitated by N + ions.
Williamson and Smallman suggested the equation: δ = D 1 , 2 16 where δ is the dislocation density which represents the degree to which the synthesised particles have crystallised. Using the above equation, we may extract information about the defects present in the samples. In the Eq. 12, it is evident that the dislocation density is inversely related to the square of the crystallite size, thus a larger δ means a lower D, which points to the presence of defects in the samples. [12][13][14] As-formed crystalline materials typically feature dislocations due to stresses (mechanical, thermal, etc) incurred during the formation process. New dislocations form at the borders of grains or on the surfaces of crystalline materials. Large numbers of interstitials and vacancies are formed when ions are implanted into thin film and these defects tend to recombine and condense throughout the annealing process, giving rise to a wide variety of defects. Different sorts of extended defects might be seen depending on the experimental settings (implant dosage and energy, thermal budget). Crystallite size affects the measured properties of produced nanoparticles, which are listed in Table I. With N ± implantation, microstrain and dislocation marginally increase, leading to modest variations in microstructure and a loss of crystallinity due to fine particle size variation. Peak broadening in XRD spectra can be attributed to several factors, such as non-uniform atomic displacements during film formation which can affect crystallite size, lattice structure, and internal strain (ε) in a lattice.
Since the crystallite size in the ion-implanted films was smaller than in the annealed samples, this suggests that the ion implantation boosted the dislocation density due to the high defect concentration and poor crystallinity. Films that were implanted had a smaller grain size than those that were deposited normally. Furthermore, implantation resulted in an increase in internal strain, which was attributed to the development of non-uniform grains and an amplified degree of   lattice imperfection. Ion implantation often results in more defects and dislocations than without it, mostly as a result of implantationinduced damage. Vacancies induce both compressive and tensile stress, and an increase in the strain distribution of samples is connected to a high dislocation density or the impact of grain refinement due to ion implantation. [14][15][16] It has been found that the implantation of nitrogen ions increases the tensile residual stress for metals or noble gas ions. 17,18 Interstitial defects are created by the implants, whereas vacancies are produced by bombardment. This is because the strain associated with interstitials is an order of magnitude more than that caused by vacancies.
Atomic force microscopic study.-Using atomic force microscopy, we were able to see the surface both before and after ion implantation and get a better overall sense of the growth. Figure 4 displays AFM pictures of V 2 O 5 thin film before and after N + implantation on samples at different fluencies. Considering that the average surface roughness is around 1 nm across all samples, it can be concluded that V 2 O 5 thin film was grown smoothly using thermal evaporator and that implantation did not cause any surface roughening. After annealing and successive ion implantations, the AFM pictures indicate the formation of grains throughout all the substrate surfaces without much ordering in an arrangement. 19 In addition, as shown in Fig. 5, we have computed for each surface an interface-width (w) parameter, which is the root-meansquare (RMS) departure of the surface heights from the mean surface height value (s), and is given by where S is the area over which w is determined and h(s) is the height at a point "s" on the surface. A matrix of surface height estimates is used to determine interface width, though. Interface width is determined by the following discrete calculation: Here, h(i, j) is the AFM-estimated surface height at the coordinates (i, j) across the sample surface, and 〈h(i, j)〉 is the average surface height over the NxN locations.  A surface's lateral correlation length (ξ) is the maximum distance across which the surface's height characteristics are substantially associated. So, the height-height correlation function (r) is governed by the following two sets of scaling rules, one for short co-relation length scales and the other for long co-relation length scales. Here, the roughness exponent, 0 < α < 1. 20 The value α = 1 indicates a perfectly smooth surface, whereas a lower value indicates an increasingly rough surface. The correlation and self-affine features of the surfaces were not identified since the (w) values of the interface only represent the global properties of surface height fluctuations. To get the corelation length (ξ) and roughness exponent (α), one must undertake further calculations using autocorrelation A(r) or the height-height correlation function H(r). The autocorrelation A(r) or height-height correlation function H(r) for unirradiated and irradiated V 2 O 5 thin films on glass substrates is shown in Fig. 6. In addition, the  additional fractal analysis yields far more in-depth data on the processes involved in surface generation and provides richer information on surface morphology. When comparing annealed and ion implanted surfaces, the statistical heights at the sites of surface matrix are not independent. The autocorrelation function is used to look for evidence of a laterally-correlated shift in the surface height fluctuations. Because of the crystallite mounds covering the surface, the function A(r) exhibits a non-linear spatial response, as seen in Fig. 6a. Depending on the surface alterations, these mounds can expand laterally and/or vertically. One important parameter for gauging the degree to which statistical surfaces are correlated is the lateral correlation length (ξ), defined as the value of r at which the autocorrelation function A(r) is 1/e of its original value. The surface heights of two points are said to be associated if the distance between them is within ξ; otherwise, the characteristics are said to be unrelated. Fractal analysis has been employed in recent studies to correlate surface spatial features with material qualities 21,22 where ZnO thin film surfaces were proposed to have certain sensing qualities by a fractal growth study. 22 Table II displays the calculated values of ξ for all of the sample surfaces. As expected, annealed films have larger estimated values of ξ, when compared with implanted films. As seen in AFM images, smaller crystallite structures have lower lateral ordering, which accounts for the smaller values of ξ.
The height-height correlation function H(r), whose dependence on the autocorrelation function is provided by Eq. 13, can be used to more precisely quantify surface shape. Surfaces implanted with a fluences of 5 × 10 11 ions cm −2 fall into two separate zones on a plot of log H(r) vs log r (Fig. 6b). Calculated values of log H(r) show similar results for various irradiation surfaces (not marked here). For small values of r, the power-law H(r) ∼ r 2α is followed by a linear response of log H(r) vs log r, which transforms into oscillatory behaviour towards 2w 2 . In this case, the roughness exponent "α", also known as the scaling exponent. The results of these computations for various samples are displayed in Table II. Roughness exponent values characterise the evolution of roughness or local surface irregularity on a very short time frame. Small-scale mass redistribution on the surface may account for the observed variance in the values of "α". This means that it can only assess the surface's immediate roughness. If α = 1, then the surface is smooth; otherwise, it is rough. In this approach, the fractal dimension D f = d+1-α is connected to the fractal characteristic, which specifies the fractal nature of the surface. Table II displays the calculated values of D f for all V 2 O 5 thin films on Glass substrate at varying ion fluencies. D f describes the degree of surface complexity which can be quantified using the above formula. It was observed that the D f decreases with increasing N + ion implantation fluencies. This is because the crystallites become more uniform in size, leading to a simpler boundary at the V 2 O 5 -substrate contacts and, ultimately, fewer surface voids.
Contact angle.-The wetting characteristics of V 2 O 5 films on glass substrates have also been investigated in this work. Figure 7a displays the contact angle (θ) at the three phase boundary following N + implantation at 5 × 10 11 , 5 × 10 12 , 1 × 10 13 and 1 × 10 14 ions cm −2 (b-f). 5 × 10 12 ions cm −2 and the annealing process reveal a similar pattern (not shown here). The value of CA provides information on the surface's wettability (hydrophilic/hydrophobic). 15 In this work, we observe the surface crystallite structures of varying dimensions and find that surface roughness plays a significant role in the wetting behavior of the liquid-solid interface. Li and coworkers have done extensive research on the impact of ripple patterns and RMS roughness on surface wettability. 23 When a droplet is two to three orders of magnitude larger than the roughness scale, 13 liquid can penetrate the roughness grooves, according to the Wenzel model. Inside the valley of crystallites, water penetration is desired because the droplet radius (0.1 mm) is substantially larger than the roughness (up to 90.50 nm). Therefore, the interfacial area within the contact perimeter is crucial to the wettability of the surface. As the surface's roughness and contact area with water droplets grow, so do the  interactions between the liquid and solid phases. Because of this, higher surface interactions cause a decrease in the droplet's contact angle with the crystallite surface.
FTIR study.-FT-IR spectra of annealed and N + implanted V 2 O 5 thin films are shown in Fig. 8. Bands at 754, 890, 1052 and 1537 cm −1 are seen in the spectrum. The absorption bands at 890 cm −1 and 1052 cm −1 are due to the crystalline V 2 O 5 . 24 The band at 1052 cm −1 is assigned to unshared V = O stretching vibration. Specifically, the 754 cm −1 band is ascribed to V-O-V asymmetric stretching vibration and 890 cm −1 corresponds to V-O-V symmetric stretching vibration. 25 With the N-O stretch region spanning from 1475 to 1550 cm-1, the IR peak of symmetric N-O stretch occurs at 1537 cm −1 . 26 As observed that the bands detected in the range of 1060-500 cm −1 are all slightly moved to higher wavenumber with reduction in intensity and all these band values are listed in the Table III matched with the earlier works. The presence of N + close to oxygen and the distortion of V=O bonds in N + implanted V 2 O 5 nanoparticles cause the V=O band's wavenumber to shift with decreasing intensity.
UV-vis study.-Using UV-vis spectroscopy, the optical characteristics of thin films of vanadium pentoxide with thicknesses of 300 nm at different fluencies of 5 × 10 11 , 5 × 10 12 , 1 × 10 13 and 1 × 10 14 ions cm −2 were investigated. In Fig. 9 we see the UV-vis absorption spectrum of annealed and N + implanted V 2 O 5 samples grown on glass substrates. It is clear that following ion implantation, thin films absorbs less light than it did before. There are no new peaks created in the absorption curve, but the curve itself remains consistent. Researchers have found that ion implantation leads to interstitial, substitutional and vacancy defects in the material as a result of surface and bulk damage brought on by "V" and "O" vacancies. 26,27 Increased absorption is the result of faulty energy levels being created within the bandgap of V 2 O 5 . The transmission spectra show the transmission in the visible region (400-700 nm) making V 2 O 5 an excellent material for a variety of device applications. As seen in the Fig. 8, samples display a dramatic drop in transparency around the UV region.    provides a better linear fit as compared to indirect transition at the same band gap. It is less likely for the indirect transition to occur since there is no linear regime in (αhν) 1/2 vs hv. The results demonstrate that at an ion fluence of 1 × 10 14 , the band gap of the V 2 O 5 sample decreases from its maximal value (E g = 2.21 eV) to a more manageable level (E g = 2.01 eV). Figure 10 represents the Energy Vs (αhν) 2 plot of annealed and irradiated V 2 O 5 thin films and the inset shows the variation of band gap with ion fluence. On calculating it was observed that the E g values lie in the range of 2.21-2.01 eV, which is consistent with the values found in the literature. E g decreases with increasing ion fluence. The different phases that may produce growth related stress account for the little variation in E g values. Figure 11a represents the predicted damage range caused by 16 keV N + ion implantation is about 104 nm within the V 2 O 5 layer. The sputtering of the target is also caused by the N + implantation. The sputtering process involves the removal of atoms from a target's near surface. When a cascade imparts energy to an atom that exceeds its "surface binding energy," the atom may sputter. An atom's energy normal to the surface must be greater than the surface binding energy for it to sputter as it crosses the surface's plane. The degree to which a surface is sputtered is measured in terms of the "Sputtering Yield," which is the average number of sputtered target atoms per incident ion. If the target consists of many elements, each one will provide a unique sputtering yield. The present research estimates that the sputtering yield is 0.87 atoms/ion for O. The increased O sputtering results in more O vacancies (Fig. 11c) up to the depth of N + implantation. The weakening of peaks in XRD and FTIR experiments is attributed to a cascade of oxygen vacancies, which in turn breaks V=O and V-O-V bonds. Thus, the increase in fluence results in the broken bonds which offer many trapping and/or scattering centres for the passage of electrons and holes. Figures 11b  and 11d) shows the band structure inwhich the electron transitions can occur at the edges of valance band (VB) to conduction band (CB). Subsequent irradiation using N + can induce the prominent "O" vacancies inside V 2 O 5 under preferential sputtering, as observed using SRIM-TRIM simulations. 28 The bandgap narrows as a result of an increase in oxygen vacancies, which is known to increase ZnO' absorption of visible light. 29 So, as illustrated in Fig. 11d, the existence of "O" vacancies can cause the formation of the shallow donor zone above the valance band. Thus, the UV-vis investigation has shown that the optical band gap narrows when N + ion irradiation fluence increases, as a result of an increase in O-vacancies.

Conclusions
The structural, optical and compositional characteristics of vanadium pentoxide (V 2 O 5 ) thin films produced on glass substrates are investigated in relation to the impact of nitrogen (N + ) ion implantation. The thickness of the film measured using stylus profilometer was found to be matched with the value observed on DTM of thermal evaporator during deposition. The mixed orthorhombic and tetragonal phases of V 2 O 5 are seen in XRD investigations. A reduction in crystalline characteristic of V 2 O 5 has been found by N + implantation on all the samples due to defects development. V 2 O 5 thin films undergo deformation in the β-phase due to ion implantation-induced strain. Thin films' bandgap are shown to have reduced as a result of an increase in oxygen vacancies   and electron scattering/trapping centres following the N + implantation. The AFM pictures were used to conduct a surface fractal growth investigation. With 16 keV N + implantation doses, the estimated values of co-relation length (ξ) and fractal dimension (D f ) decrease, indicating a smoothing of the crystallites' surface boundary. In accordance with the Wenzel model of wettability, a measurement of the water contact angle reveals that the hydrophilic character of V 2 O 5 thin film increases with ion fluence compared to annealed V 2 O 5 thin film. This is owing to the higher fractal dimension (D f ) and interface width (w).The current study provides the advantages of lowering the optical band-gap and the conductance of V 2 O 5 thin films for optical applications and may lead to the development for an econometric model to control liquid wetting of fractal structures in micro or nano-fluidics and microelectronic devices.

Achnowledgment
Author would like to express deepest appreciation and a sense of gratitude toDr. Vinay Kumar, HOD DNSM, CUJ for his unwavering support. Mr. Raj Kumar, IUAC, New Delhi for allowing to use table top accelerator. Satish Dhawan Space centre facility, CUJ isachnowledged for characterization facilities.