Influence of different annealing ambient on terbium oxide passivation layers sputtered using the RF sputtering on silicon substrate

This study investigates the influence of different annealing ambient on terbium oxide (Tb4O7) passivation layers sputtered using radio frequency (RF) sputtering on silicon (Si) substrates. The passivation layers were subjected to annealing in various ambient, including oxygen (O2), nitrogen (N2), argon (Ar), and nitrogen-oxygen-nitrogen (NON). The structural, morphological, compositional, topological, and optical properties of the passivation layers were characterized using various techniques. The obtained results indicate that the annealing ambient has a significant impact on the properties of Tb4O7 passivation layers. Annealing in Ar ambient leads to the formation of Tb4O7 with improved crystallinity close to 49.75 nm and higher surface roughness at (2.32 nm). In contrast, annealing in the O2 ambient results in broad GIXRD peaks with the lowest surface roughness around (1.34 nm). Notably, annealing in N2 ambient exhibits an intermediate behavior, with partial crystallized size values (31.80 nm) compared to the Tb4O7 passivation layer annealed in Ar ambient and moderate surface roughness. The optical bandgap (Eg) was estimated by applying the Kubelka–Munk (KM) approach and the obtained values were 3.28, 3.17, 2.37, and 2.27 eV for annealed in O2, N2, Ar, and NON ambients, respectively. The investigation of Tb4O7 as a passivation material expands the range of materials available for semiconductor device fabrication, offering potential advancements in optoelectronics applications. Therefore, the significance of this study lies in its contribution to the optimization of Tb4O7 passivation layers in the field of semiconductor device technology. Hence, the sample annealed in an Ar ambient demonstrated the best results in terms of structural, morphological, compositional, topological, and optical properties of Tb4O7 passivation layers as compared to other samples.


Introduction
Modern science and technology would not be complete without semiconductor devices, which are crucial to a variety of applications ranging from optoelectronics to microelectronics [1].The silicon (Si) substrates are the fundamental components of these devices.As the second most abundant element in the earth's crust, it is available for large-scale semiconductor production [2].The Si has semiconducting characteristics, meaning it can conduct electricity under certain conditions while behaving as an insulator under others [3].This feature is vital for regulating the movement of electrons and holes in semiconductor devices.With the ongoing reduction in the size of semiconductor devices, experts are now asserting that silicon (Si) substrates are approaching their maximum capacity.Consequently, there is a growing focus on investigating alternative wafers like silicon carbide (SiC), gallium nitride (GaN), and sapphire to facilitate the development of novel devices.However, concerns about the poor thermal characteristics of sapphire substrates [4], the high price of SiC substrates [5], and the high mismatch of GaN substrates [6], have limited their potential utilization.Therefore, the Si substrate often requires good surface passivation to mitigate the detrimental effects of defects and surface states to cope with the increasing market demand [7].The issue of surface passivation becomes essential to mitigate these detrimental effects and enhance the performance of semiconductor devices [8].Recently, the use of thin-layer passivation techniques has received much attention [9].The surface passivation made up of Tb 4 O 7 on silicon substrates is important in the semiconductor field due to their ability to effectively decrease the number of interface traps and enhance carrier mobility, hence resulting in enhanced device performance [10].A lot of various techniques have been successfully used for passivation layer deposition, including dip-coating [11], spincoating [12], electron beam evaporation [13], physical vapor deposition (PVD) [14], atomic layer deposition (ALD) [15], electrochemical deposition [16], and radio frequency magnetron (RF) sputtering [17].One of the most accessible methods is RF sputtering, which offers numerous advantages like uniform deposition of conductive and non-conductive materials over a large area, affordable, reproducibility, strong adhesion, and long-term stability [18].Moreover, low pressure was maintained throughout the passivation layer's deposition, which makes it possible to avoid coming into contact with any potential sources of contamination [19].
Previously, a variety of materials were sputtered as passivation layers for various applications in semiconductors, such as Sb 2 Se 3, CdTe, CeO 2 , and Ga 2 O 3 , but pose several demerits as follows.Despite the binary nature of Sb 2 Se 3 , faces challenges in passivation due to defects at the p-n junction, interfacial defects, and narrow band gap.As a result, it may react with other materials in semiconductor device fabrication processes.Moreover, the intrinsic chemical instability of Sb 2 Se 3 could lead to device performance degradation, reliability issues, and reduced operational lifetimes [20].The challenge of using CdTe as a passivation layer lies in its limited open-circuit voltage (VOC) due to defects related to its polycrystalline nature, carrier concentration, narrow band gap, interface defects, and susceptibility to chemical reactions [21].The use of Ga 2 O 3 as a passivation layer on silicon substrates presents several challenges.While it has been shown to effectively passivate both n and p-type silicon surfaces, prolonged annealing at high temperatures can lead to depassivation [22].The interface between Ga 2 O 3 and the silicon substrate can also suffer from intrinsic brittleness, flat band voltage instability, and grain boundary defects [23].Additionally, the surface recombination velocities of Ga 2 O 3 thin films decrease with annealing temperature, indicating a potential saturation point [24].While Ga 2 O 3 can provide excellent passivation for graphene, further processing like plasma-enhanced atomic layer deposition of Al 2 O 3 may pose challenges [25].Furthermore, using CeO 2 as a passivation layer on a silicon substrate presents numerous issues.Among the issues it was reported that the CeO 2 passivation layers annealed at high temperatures, such as 1000 °C, can lead to the formation of an excessive SiO 2 interfacial layer, causing permanent dielectric breakdown [26].Also, the post-deposition annealing (PDA) in ammonia gas ambient can lead to the formation of mixed oxidation states in the CeO 2 film, affecting its passivation properties [27].It further highlighted the presence of crystallographic defects and an amorphous layer at the CeO 2 /Si interface, which can compromise the effectiveness of the passivation layer [28].Thus, additional progress in the passivation layer remains extremely important.
Terbium oxide passivation layers (Tb 4 O 7 ) are highly desirable for a wide spread of semiconductor applications, including chemical [29], optical [30], ceramic [31], and electronics [32], and so forth, owing to their excellent properties for instance a large bandgap (>2.23 eV) [33], hydrophobic nature [34], high-k (24 ± 2) [35], enhanced stability with Si substrate [36], low flat band voltage (0.007 V) [37], low lattice misfit (∼2%) [38], high refractive index [35] and strong luminescence [39].Tb 4 O 7 is comprised of Tb +4 ions, which is ascribed to the remarkable steadiness of the formation of a half-filled 4f 7 subshell in the electronic configuration [40].The Tb 4 O 7 passivation layers have been deposited using different techniques, including metal-organic chemical vapor deposition (MOCVD) [41], reactive physical vapor deposition (RPVD) of metallic terbium [42], and physical vapor deposition (PVD) deposition [43].Among these, a PVD is a technique that shows promise for creating a consistent and accurately controlled thickness passivation layer.Moreover, the use of Ar gas flow in the formation of passivation layers by RF sputtering has been extensively studied in numerous research papers [44].This is due to its superior properties such as inert chemistry, high atomic weight, relatively low cost, and the ability to avoid oxidation by ceramic targets or silicon substrates [45].Therefore, the physical properties of Tb 4 O 7 passivation layers are strongly influenced by deposition parameters such as RF sputtering power, deposition time, gas type, and the post-deposition annealing (PDA) processes.Hence, this study expands on previous research in the following way.The green electroluminescence from Tb 4 O 7 films on a silicon substrate.The occurrence involves the impact excitation of Tb 3+ ions by hot carriers with the Tb 4 O 7 film sputtered on Si (n-type and p-type) substrates via the RF sputtering of (110 W).The sputtered film was subjected to postdeposition annealing at 900 °C for 5 min in an Argon environment [33].The metal-organic chemical vapor deposition (MOCVD) synthesis of terbium oxide (Tb 2 O 3 ) films.The films are then annealed in air ambient at a temperature of 800 °C, which leads to the formation of Tb 4 O 7 [35].The kinetics of terbium oxide film growth from the decomposition of Tb(dpm) 3 vapor in argon flow at different temperatures [41].The focus of most of the research was centered on the utilization of Tb 4 O 7 oxide in various applications.This motivated us to further investigate the structural, morphological, compositional, topological, and optical properties of the Tb 4 O 7 oxide.
Therefore, our study was innovated by specifically focusing on a detailed investigation of the effect of postdeposition annealing processes on the formation of Tb 4 O 7 passivation layers.The RF sputtering was used to sputter the ceramic target of Tb 4 O 7 on an n-type Si(100) in an argon (Ar) ambient.The deposited samples were subjected to post-deposition annealing (PDA) treatment at a temperature of 800 °C for 30 min in O 2 , N 2 , Ar, and NON ambients.The properties of Tb 4 O 7 passivation layers, including crystal structure, thickness, morphology, composition, topology, and optical properties by grazing incidence x-ray diffraction (GIXRD), Field emission scanning electron microscopy (FESEM) equipped with energy dispersive x-ray analysis (EDX), Atomic force microscopy (AFM) as well as Ultraviolet-visible spectrophotometer (UV-vis) were investigated.At the end of this work, the results obtained will significantly highlight the importance of post-deposition annealing treatments using different annealing ambient and provide insight into the formation of optimal Tb 4 O 7 passivation layers.Furthermore, the current work is based on materials study, but subsequent experiments are planned to include applications of the Si-based MOS capacitors and/or the Photodetectors.

Experimental methods
A high-purity ceramic target of Tb 4 O 7 (99.999%)with a purity level of 99.99% and a diameter of approximately 2 inches was acquired from Changsha Xinkang Advanced Materials Co., Ltd The silicon substrate of n-type with a (100) orientation was cleaned using the conventional Radio Corporation of America (RCA) process.Subsequently, it was immersed in a diluted solution of hydrofluoric acid (HF) and water (H 2 O), as shown in figure 1.The RCA procedure was carried out under the following conditions: the silicon substrate was submerged in a solution mixture of H 2 O/NH 4 OH/H 2 O 2 (200:40:40 ml) at a temperature not exceeding 75 °C for 10 min.The Si substrate was taken out and rinsed in deionized (DI) water before the next step.The Si substrate was also dipped in the solution of HF/H 2 O (15:150 ml), not exceeding 75 °C for 10 min.Subsequently, the substrate was then immersed in a solution of H 2 O/HCl/H 2 O 2 (240:40:40 ml) and heated again at 75 °C for 10 min to remove organic residues, an oxide layer, and metallic impurities from the wafer surface.Finally, the substrate was dried and rinsed with DI water and then subjected to a nitrogen (N 2 ) gas drying system, respectively, before the sputtering process.The radio frequency magnetron (RF) sputtering was used for the deposition of the ceramic target using a sputtering power of (80 W /30 min and then raised to 100 W /15 min) at a pressure of 6.61 × 10 − 3 mbar.The rationale for these choices of RF sputtering power is based on the factors related to the quality of the film, the deposition rate, and the thickness of the deposited passivation layer.
The chamber was evacuated to a base pressure of 5 × 10 −5 mbar before sputtering, and an argon gas flow rate of 15 sccm was maintained throughout the processes.The deposited samples were then annealed at 800 °C for 30 min in O, N, Ar, and NON ambients.A constant gas flow (50 ml min −1 ) was maintained throughout the entire ramp-up to (800 °C), dwelling time (30 min), and during the temperature ramp-down processes.Furthermore, during the annealing process in a NON ambient, nitrogen was continuously passed through the annealing chamber as the temperature increased.Once the temperature reached 800 °C, the nitrogen was turned off and replaced with oxygen for the duration of the annealing period (dwelling time).After 30 min, the oxygen was turned off and nitrogen was reintroduced, which was maintained throughout the temperature ramp-down process.

Characterization
The structural studies were performed in a grazing incidence x-ray diffractometer using a Cu-Kα radiation source λ = 1.5406Å (model: Bruker D8, MA, USA).The type of test performed was the Grazing incidence x-ray diffraction (GIXRD), which involves directing the incident x-ray beam at a lower angle.It improves the sharpness and visibility of the diffraction pattern.Thus, the experiment was carried out to specifically investigate the characteristics of the thin film, without any influence from the substrate.The scanning range was from 10 to 90, with a uniform step size of 0.5 for all the studied samples.Surface morphology was carried out using fieldemission scanning electron microscopy FESEM (Model: FEI Nova Nano SEM 450, Oregon, USA) instrument equipped with energy-dispersive x-ray analysis (EDX) to detect the components that make up the samples.Surface roughness changes in the sample surface were also studied using atomic force microscopy (AFM) (Model Dimension-edge Bruker, MA, USA).The reflectance properties of the samples were investigated in the wavelength range of 190-1000 nm using a UV-vis spectrophotometer (Model: Cary 5000, Agilent Technologies, Santa Clara, USA).200), (220), and (311) planes, which were in line with the International Centre for Diffraction Data (ICDD) file no.00-013-0387.There are no diffraction peaks of additional stoichiometric terbium oxides like TbO 2 and Tb 2 O 3 or non-stoichiometric oxides, TbO 1.714 , TbO 1.750 , and TbO 1.818, or impurities identified in the GIXRD pattern (see figure 2(A)).The obtained results were similar to those of earlier studies available in the literature [33].As observed, all the patterns exhibited strong orientation corresponding to the (111) plane [46].The results of these findings indicate that the Tb 4 O 7 passivation layer annealed in Ar ambient achieve the highest peak oriented in the (111) plane with improved crystallinity, followed by the Tb 4 O 7 passivation layers annealed in N 2 and NON ambients.However, a broad peak can be seen on the samples subjected to annealing in O 2 ambient with a relatively lower intensity as compared to the Tb 4 O 7 passivation layers annealed in other ambient (see figure 2(B)).The observation mentioned above is primarily due to the chemisorption of oxygen (O 2 ) on the surface of the Tb 4 O 7 passivation layer following annealing in an O 2 ambient, which reduces the peak intensity of grain boundaries and distorts crystallites.The findings were consistent with the reported literature elsewhere [47].However, a slight shift of the peak oriented in (111) planes to lower diffraction angles was observed (see figure 3), compared to the Tb 4 O 7 peaks reported in ICDD file no.00-013-0387.

Structural features
The shift of the peaks to the lower diffraction angle from the GIXRD pattern analysis might be understood as a transformation that occurred in lattice parameter (a) of the investigated Tb 4 O 7 passivation layers at different annealing ambient.Thus, considering the shifting in the diffraction peak of the Tb 4 O 7 (111) plane, the lattice parameter (a) of the investigated Tb 4 O 7 passivation layers can be computed by applying the following equations (1) and (2) for a cubic system: where hkl, θ, and d hkl , represent Miller's index, diffraction angle, and interplanar spacing, respectively.Figure 4 shows the computed lattice parameter (a) values of all the examined Tb 4 O 7 passivation layers annealed in different ambients.The shift of the peaks observed from the GIXRD patterns to lower 2θ value (diffraction angles) can be understood as an increase in the lattice parameter (a) with regards to different annealing ambient.Therefore, the above equations (1) and (2) extracted by fitting the Gaussian function of the reflection peak of the (111) surface can be employed to calculate lattice parameters (a).As can be seen, in O 2 , N 2 , Ar, and NON ambients, the diffraction peaks oriented in (111) plane shift to 28.72°, 28.92°, 29.21°, and 28.85°( see figure 3), and the calculated lattice parameter (a) values were 0.5380, 0.5343, 0.5291, and 0.5355 nm, respectively (see table 1).The passivation layer annealed in the O 2 ambient achieved the highest lattice parameter (a) value around 0.5380 nm, which may be due to O 2 elements replacing O 2 space in the structure as a result of the thermal annealing process [48].Moreover, the increase observed in the lattice parameters (a) confirms that the peaks of the investigated Tb 4 O 7 passivation layers shifted to lower diffraction angles.This increase may suggest the presence of smaller-size dopants in the sample, which would modify the lattice of the Tb 4 O 7 passivation layer.It is worth noting that the Tb 4 O 7 passivation layers deposited on the Si substrate do not involve the integration of additional foreign atoms during the sputtering process.The observed shifting in lattice parameter (a) concerning post-deposition annealing in different ambient was probably due to the progressive oxidation of the Tb layer from Tb to oxygen-deficient Tb 4 O 7 in one annealing ambient and then to oxygen-rich Tb 4 O 7 in another annealing ambient at high temperature.This is very possible since all Tb 4 O 7 passivation layers were annealed at high temperatures (800 °C).As a result, it was anticipated that the increase observed in the computed lattice parameter (a) values was attributed to phase change in the Tb 4 O 7 passivation layer increasing from oxygen-deficient Tb 4 O 7 (Tb 3+ to Tb 4+ ) with regards to different annealing ambient in which ionic radius of Tb 4+ (0.88 nm) was less than that of Tb 3+ (0.104 nm) [49].Furthermore, the results obtained from this work indicate that Tb 4 O 7 passivation layers annealed in (N 2 , Ar, and NON) ambients showed a very small shift in the calculated lattice parameter (a) value as compared to Tb 4 O 7 passivation layer annealed in (O 2 ) ambient.Hence, it is worth mentioning that the computed lattice parameter (a) values for all investigated Tb 4 O 7 passivation layers annealed in different ambient were closer to that of the Tb 4 O 7 reference pattern (0.5290 nm) in line with the ICDD file no.00-013-0387.
To further study the influence of different annealing ambient on the formation of the Tb 4 O 7 passivation layers, the Scherrer equation was utilized to compute the changes in terms of growth of the crystallite size (D) using the following equation (3) [50], and equation (4) could be used to compute microstrain (ε) [51], where D represents the crystallite size, λ the x-ray wavelength, θ is the diffraction angle, β is the integral breadth of the x-ray peak on the 2θ axis, widely calculated as the full width at half maximum (FWHM in radians), ε is the (Microstrain), and K is the (Scherrer constant).The shape of the crystal, its size distribution, the diffraction line index, and the exact definition of β (FWHM or integral width) are the factors that determine K [52].K can have values ranging from 0.62 to 2.08 in different forms.The K value used in this investigation was 0.9.
Figure 5 shows the calculated crystallite size (D) and microstrain (ε) of the Tb 4 O 7 passivation layers subjected to annealing in different ambient.The computed D values for the studied Tb 4 O 7 passivation layers annealed in different ambient (O 2 , N 2 , Ar, and NON) were found to be 12.57, 31.80,49.75, and 19.17 nm, respectively.As observed, the Tb 4 O 7 passivation layer annealed in an O 2 ambient achieved the lowest crystallinity, close to 12.57 nm anticipating that this layer could have oxidized during the annealing process.These demonstrate that annealing in an O 2 -rich ambient can fill the oxygen vacancies in the Tb 4 O 7 passivation layer, resulting in oxidation and a decrease in D (see figure 5).An increase in the D was noticed for the Tb 4 O 7 passivation layer annealed in Ar ambient close to 49.75 nm.The observed increase may be related to the presence of oxygendeficient Tb 4 O 7 and/or oxygen-rich Tb 4 O 7 in the investigated Tb 4 O 7 passivation layer or to the occurrence of layer growth owing to the coalescence of small crystallites.However, a decrease in the D was observed for the  On the other hand, opposite results were achieved for microstrain (ε), as shown in figure 4.Even though the sputtered Tb 4 O 7 passivation layer was found to be crystalline, however, there are several grains with varying relative locations and orientations, which causes the phase difference between waves scattered by various grains to differ.These actions will create lattice mismatches identified as microstrains (ε) [53].The microstrain (ε) of the Tb 4 O 7 passivation layers is shown in table 1, which was consistent with previous research in which this microstrain (ε) decreases with increasing crystal size [54].As observed, the lowest microstrain (ε) was achieved for the Tb 4 O 7 passivation layer annealed in Ar ambient (see figure 4) as compared to Tb 4 O 7 passivation layers annealed in (O 2 , N 2 , and NON) ambients.However, the lowest microstrain (ε) obtained for the Tb 4 O 7 passivation layer annealed in an Ar ambient agrees with the highest D obtained for this Tb 4 O 7 passivation layer (see figure 5).This could be associated with enough energy adsorbed and occupied oxygen vacancies (V o ) in the lattice to promote the coalescence of smaller crystallites into larger crystallites.
To further understand the structural transformation caused by dislocations in the investigated Tb 4 O 7 passivation layers, as shown in figure 4, Williamson and Smallman's relationship was employed to compute the dislocation density (δ) using equation (5) below [55]: in which D denotes the size of the crystallites in the investigated passivation layers that were extracted using the Scherrer approach.
The variations noticed in the crystallite size of the Tb 4 O 7 passivation layers are due to differences in dislocation density (δ).The δ values calculated for the samples subjected to different annealing ambient are presented in table 1.The highest δ value was achieved at (6.3263 × 10 −3 nm −2 ) for the Tb 4 O 7 passivation layer annealed in O 2 ambient.Subsequently, a decrease was observed for the Tb 4 O 7 passivation layer annealed in N 2 (9.8858 × 10 −4 nm −2 ) and Ar (4.0396 × 10 −4 nm −2 ) ambients.However, an increase in the δ value for the Tb 4 O 7 passivation layer annealed in NON ambient to (4.5031 × 10 −4 nm −2 ) was observed.The increase in the Tb 4 O 7 passivation layer annealed in NON ambient may be attributed to the increase in O 2 adsorption and diffusion activity while decreasing oxygen defect/vacancies (V o ) available for N 2 attachment reported elsewhere [56].Hence, it was hypothesized that the increase in crystallite size occurs due to the decrease in lattice parameters, microstrain, and dislocation density of the investigated materials [57].The above phenomena are consistent with the increase in crystallite size observed in the GIXRD pattern analysis (see figure 5).The values achieved for lattice parameters, dislocation density, crystallite size, and microstrain are displayed in table 1.

FESEM characterization
To demonstrate the larger crystallite size observed in the Tb 4 O 7 passivation layer annealed subjected to annealing in an Ar ambient, cross-sectional field emission scanning electron microscopy (FESEM) analysis was done, as illustrated in (figure 6).The total oxide thickness (t TOT ) was determined using ImageJ at ten (10) separate places across each of the investigated Tb 4 O 7 passivation layers.The measured t TOT values for a standard   (51.46 nm) between the passivation layer and the Si surface as shown in figure 6(a).In contrast, the t TOT values exhibited an increase when the investigated samples were subjected to annealing in different ambients in comparison to the standard sample (see figures 6(b)-(e)).Noticeably, among the annealed samples, the Tb 4 O 7 passivation layers subjected to annealing in O 2 ambient achieved the lowest t TOT values at (58.97 nm) when compared to other annealed Tb 4 O 7 passivation layers (see figure 7).This can be attributed to oxygen atoms diffusing into the Si substrate during the post-deposition annealing process, leading to the formation of the SiO 2 interfacial layer.Hence, at high annealing temperatures, the O-deficient Tb 4 O 7 passivation layer can completely oxidize by absorbing oxygen atoms from both the formed silicate interfacial layer and the oxygen ambient.Similar explanations were reported on the structural properties of sputtered HfO 2 films on Si (100) substrate and Annealing of Al 2 O 3 thin films prepared by atomic layer deposition deposited on Si(111) substrate at high temperature [58,59].Thus, the subsequent experiment will investigate in detail the development of an interfacial layer between the Si substrate and the passivation layers for this material (Tb 4 O 7 ).Furthermore, the highest thickness at (75.39 nm) achieved for the Tb 4 O 7 passivation layer annealed in Ar ambient was supported by the lesser dislocation density and microstrain shown by the GIXRD pattern analysis (see figures 4 and 5).
Figure 8(a)-(d) shows the top view morphologies taken at 100,000x magnifications of the Tb 4 O 7 passivation layers subjected to annealing at different annealing ambient.Noticeably, distinct images were achieved, indicating that the grain size of the studied Tb 4 O 7 passivation layers differed in the annealing ambient.The Tb 4 O 7 passivation layer annealed in the N 2 ambient has a homogeneous and smooth surface consisting of small grains with some voids.In contrast, an improvement in grains was seen on the surface of the sample annealed in  the NON ambient.The circular grains seen in the N 2 annealed sample were no longer visible after the passivation layer was annealed in Ar ambient.This means that individual grains grow with clear boundaries during the annealing process in this ambient while neighboring grains coalesce with small micropores.This growth process was associated with the formation of pore space, which may be related to stress relaxation effects for this passivation layer at high temperatures.The formation of pore space observed during this growth process may have been associated with those mentioned above experienced at high annealing of 800 °C (see figure 8(c)).However, the formation of large micropores on the layer surface was also observed on the passivation layer annealed in O 2 ambient (see figure 8(d)).These anticipated that the coalescence of neighboring grains during the annealing process in this ambient could generate stress that the layer could not withstand.Due to this, cracks occurred on the surface of the passivation layer, as evidenced by cross-sectional analysis using FESEM (see figure 6(a)).Figure 10 shows the spectrum obtained from energy dispersive x-ray spectroscopy (EDX) techniques as well as the elemental composition constituted on the surface of studied Tb 4 O 7 passivation layers annealed in different ambient (inserted in figures 10(a)-(d).The observed, typical EDX disclosed spectra that verify the presence of terbium (Tb), oxygen (O), and silicon (Si) on the Tb 4 O 7 passivation layers subjected to annealing at different ambient.However, the detected Si content revealed a high-intensity peak in comparison to O and Tb, with intensity corresponding to their concentrations in various spectra.No additional peaks related to other elements were detected in the pattern, and the substrate was responsible for the visible Si peak in the spectrum.From the observed pattern, the passivation layer annealed in an Ar ambient achieved the highest atomic percent (at%) of O at (17.84 at%), followed by the passivation layer annealed in an N 2 ambient at (16.56 at%) and then the Tb 4 O 7 passivation layer annealed in NON ambient at (16.96 at%).Therefore, the formation of the interfacial layer was considered very little since the observed at% of O was higher for this Tb 4 O 7 passivation layer and did not differ significantly.However, the attainment of lowest at% of O at (10.22 at%) for a passivation layer annealed at in O 2 ambient was associated with the prediction that O anions from the Tb 4 O 7 passivation layer may have diffused away from the surface of the layer to the interface to create an interfacial layer, which increases the degree of lattice distortion and thus lead to a high dislocation density (see figure 4).

AFM characterization
Figure 11 shows the atomic force microscopy (AFM) technique.The technique was acquired to observe the three-dimensional (3D) surface topography of the studied Tb 4 O 7 passivation layer's response to annealing in different ambient.The scanning area has dimensions of 5 μm × 5 μm.
The surface topography and the root mean square (RMS) roughness were observed for the Tb 4 O 7 layers subjected to post-deposition annealing at different ambient.The observed RMS values attained for the Tb 4 O 7 passivation layers subjected to annealing in (N 2 , Ar, and NON) ambients were (1.34, 1.85, 2.32, and 1.69 nm), respectively, as shown in figure 12.The Tb 4 O 7 passivation layer subjected to annealing in O 2 ambient revealed a larger protrusion in the upper part of the three-dimensional (3D) topography.In contrast, the smaller part of the 3D topography was composed of smaller protrusions with a huge difference in dimensions (see figure 11(a)), in which a lower RMS value of around 1.34 nm (see table 2) and the improper protrusions was observed (See figure 12).Therefore, this observation could be due to the crack on the Tb 4 O 7 passivation layer surface related to the formation of large mesopores on FESEM morphological analysis (see figure 6(a)).However, the passivation layer annealed in N 2 ambient indicates a smooth surface, with some voids that could be seen, whereby protrusion with a lower size located at the upper part of the 3D topography forms protrusions with some obvious peaks (see figure 11(b)).Topographical transformations were seen for the passivation layer annealed in Ar ambient, which revealed a formation of more protrusions with obvious peaks (see figure 11(c)).The observed  changes would anticipate that during the annealing process, the protrusion with a lower size has acquired an adequate amount of energy to merge and form protrusions with more clear peaks, which play a part in the highest RMS values for this passivation layer compared to other Tb 4 O 7 passivation layers.
Furthermore, the interaction between the growth of the Tb 4 O 7 layer, the adsorption of oxygen, the binding of nitrogen to oxygen vacancies, and the diffusion of oxygen to the Tb 4 O 7 /Si interface would unavoidably contribute to an irregular surface topography.This could translate into the formation of voids in the sample that was subjected to the annealing in a NON ambient, as demonstrated by the AFM analysis (figure 11(d)).The nitrogen incorporation in the investigated Tb 4 O 7 passivation layer annealed in NON has contributed to the degradation of the crystal quality of the Tb 4 O 7 passivation layer (see figure 5) as well as the lower root-meansquare (RMS) roughness in comparison to the passivation layer annealed in Ar ambient (see figure 12).The results obtained were consistent with the similar findings published in the literature for annealing in a nitrogenoxygen-nitrogen (NON) atmosphere [23,26], Therefore, it is worth mentioning that the observed 3D topographies of the Tb 4 O 7 passivation layers subjected to annealing in (O 2 , N 2 , Ar, and NON) ambients agreed with surface morphologies acquired from FESEM analysis (see figures 8(a)-(d)).

Optical analysis
The band gap energies (Eg) of studied Tb 4 O 7 passivation layers subjected to annealing in different ambient were computed by using the Kubelka-Munk (KM) function approach based on the obtained information from the ultraviolet-visible (UV-vis) spectroscopic measurement data.The KM function is expressed mathematically in equation (6) below [60]: in which R stands for diffuse reflectance, and where hv is multiplied by the F(R) function with the appropriate coefficient (n), where n was assumed to be 1/2 for the direct band gap (Eg) transition and 2 for the indirect band gap (Eg) transition.As a result, the value of n is assumed to be two (2) when applying equation (5), this is because the previous work indicates that terbium oxide has only a direct band gap [35].Based on the obtained (F(R) x hv) 2 against energy (hv) plots as depicted in (Figure 14  NON were found to be 3.28, 3.17, 2.37, and 2.27 eV, respectively.These observed Eg values fall within the expected range of bandgap values reported for Tb 4 O 7 bandgap values (1.94-4.10eV) [61,62].The Tb 4 O 7 passivation layer subjected to annealing in an O 2 ambient revealed the largest Eg, followed by a decreasing trend for the Tb 4 O 7 passivation layers annealed in N 2 > Ar > NON ambients.The observed decreasing trend in the direct E g value is offset by the improvement in crystallinity, increased thickness, and increased surface roughness of the investigating material [63].The results obtained for the investigated Tb 4 O 7 passivation layers subjected to annealing in different ambient (O 2 , N 2 , Ar, and NON) agree with this assumption as these improvements were

Conclusion
In this study, we have successfully deposited Tb 4 O 7 passivation layers on Si substrates via the RF magnetron sputtering.The influence of annealing in different annealing ambient (O 2 , N 2 , Ar, and NON) ambients on the good and high-quality formation of Tb 4 O 7 passivation layers is investigated using various characterization techniques.The GIXRD patterns of the annealed Tb 4 O 7 passivation layers exhibited crystalline structures with (111) crystallographic planes.However, the formation of Tb 4 O 7 passivation layers was enhanced under N 2 and Ar ambient and affected under O 2 and NON ambients.The FESEM morphology showed a uniform and homogeneous surface without any impurities on the Tb 4 O 7 passivation layer annealed in Ar ambient.EDX spectrum confirmed the presence of Tb, O, and Si peaks with intensity proportional to their respective concentrations in different spectra.The optical bandgap (Eg) estimated for O 2 , N 2 , Ar, and NON ambients were 3.28, 3.17, 2.37, and 2.27 eV, respectively.As a result of our research, we anticipate that the formation of a Tb 4 O 7 passivation layer using Ar ambient may be the most effective method for possible applications in microelectronics and optoelectronics.

Figure 1 .
Figure 1.Experimental schematic illustration of the procedure.

Figure 2
shows the grazing incidence x-ray diffraction (GIXRD) patterns analysis of the examined Tb 4 O 7 passivation layers subjected to post-deposition annealing in different ambients.The occurrence of an increase and decrease in diffraction peak density revealed the formation of the crystalline Tb 4 O 7 passivation layers.The diffraction peaks seen in the GIXRD patterns belong to the cubic Tb 4 O 7 phase, space group Fm-3m (225), and are oriented in the (111), (

Figure 2 .
Figure 2. (A) GIXRD patterns plot and (B) The smooth diffraction peaks of the Tb 4 O 7 passivation layers annealed in different ambient.

Figure 4 .
Figure 4.The lattice parameter and dislocation density of Tb 4 O 7 passivation layers annealed in O 2 , N 2 , Ar, and NON ambient.

Figure 5 .
Figure 5.The Crystallite size and Microstrain of Tb 4 O 7 passivation layers annealed in O 2 , N 2 , Ar, and NON ambient.

Figures 9 (
a)-(f) depicts the observed elemental composition analysis using field emission scanning electron microscopy equipped with energy dispersive x-ray (FESEM-EDX) mapping performed on the studied Tb 4 O 7 passivation layers in different ambient in comparison with the FESEM image (see figure 9(a)) of the Tb 4 O 7

Figure 7 .
Figure 7.Total oxide thickness for a standard Tb 4 O 7 passivation layer and Tb 4 O 7 passivation layers annealed in different ambient.

Figure 9 .
Figure 9. (a) FESEM image of the Tb 4 O 7 passivation layers annealed in Ar ambient, and (b)-(f) elemental mapping of Ar, Tb, O, and Si elements belonging to sample annealed in Ar ambient.
(A)) for the Tb 4 O 7 passivation layers annealed in (O 2 , N 2 , Ar, and NON) ambients.An extrapolation of (F(R) x hv) 2 against photon energy (hv) = 0 would enable the determination of direct band gap energy (Eg) values of the studied Tb 4 O 7 passivation layers.

Figure 13 (
B) depicts the direct band gap (Eg) plots for Tb 4 O 7 passivation layers obtained at different annealing ambient.The calculated direct Eg values for the Tb 4 O 7 passivation layers annealed in O 2 , N 2 , Ar, and

Figure 12 .
Figure 12.Total oxide thickness for the Tb 4 O 7 passivation layers annealed in different ambient.

Table 1 .
Shows the Lattice parameter, dislocation density, crystallite size, and microstrain.Tb 4 O 7 passivation layer annealed in NON ambient.Thus, the formation of larger crystallites during coalescence was hindered for Tb 4 O 7 passivation layers annealed in a NON ambient.This is because the presence of a high concentration of oxygen vacancies and nitrogen atoms attached to these vacancies made it difficult for neighboring crystallites to merge.
Tb 4 O 7 passivation layer and passivation layers annealed in (O 2 , N 2 , Ar, and NON) ambient were 51.46, 58.97, 67.56, 75.39, and 65.71 nm, respectively as outlined in table 2. It can be noted that the standard sample has the lowest t TOT value

Table 2 .
Presents the Thickness, RMS roughness, and band gap energy values.