N-type H2-doped amorphous silicon layer for solar-cell application

In this work, we report that hydrogen (H2) doped in n-type a-Si:H thin films strongly influences the electronic correlation in increasing the conversion output power of solar cells. Type n a-Si:H thin films were grown using PECVD on ITO substrates with various H2-doping, to obtain various thin films for solar-cell applications. N-type a-Si:H thin films were prepared, and then characterized using ellipsometric spectroscopy, atomic force microscopy, Fourier transform infrared spectroscopy, and transmission electron microscopy. The addition of doped-H2 to the thin layer shows a decrease in optical conductivity, while the energy gap in the thin layer shows a significant increase in the a-Si:H-type thin layer. Our results show that H2 doping plays a very important role in the electronic structure, which is indicated by the significant energy gap difference. On the other hand, the bond structure of each H2-doped thin film showed a change from amorphous to nanocrystalline structures which were evenly distributed in each H2-doped bonding. Overall, we believe that the addition of doped-H2 to our findings could help increase the power conversion output of the solar cell due to the modification of the electronic structure.


Introduction
To create high-quality thin films for large-area devices like photovoltaic (PV) cells, a popular technology called plasmaenhanced chemical vapor deposition (PECVD) has emerged [1].The formation of an a-Si:H thin film was found to be influenced by a number of deposition factors in the PECVD process, including gas flow rate, excitation power, pressure, substrate temperature, and dilution of the source gas (silane) with another gas hydrogen [2].The glow-discharge decomposition of mixed SiH 4 + H 2 gas was used to create the a-Si:H films [3], which were then examined for how hydrogen dilution affected their structural and optical characteristics [4].For the single p-type and n-type thin layers, a dopant is used in the form of mixing diborane gas (B 2 H 6 ) for the p-type layer and phosphine gas (PH 3 ) for the n-type coating into the gas source [5].The thin-layer growth process is carried out using silane gas (SiH 4 ) and hydrogen (H 2 ) as gas sources [6].However, the high localized state density in the extrinsic layer of amorphous silicon reduces the effectiveness of the doped process.The localized state density can be reduced by adding hydrogen gas [7], or what is popularly known as the hydrogenation process [8].
Electrical conductivity, layer thickness, and band gap are the most important parameters in p-i-n solar cells, because these three parameters influence each other, so each layer will affect the p-i-n structure [9].Therefore, the quality of each layer (p-type, i-type, and type-n) must be considered, so that the optimum conditions for each layer will be obtained and the resulting efficiency in the application of solar cells will increase [10].Film thickness and optical energy gap can also be determined from both optical transmission spectroscopy ellipsometry [11].Although optical transmission is the most convenient and precise approach for analyzing thin films formed on transparent substrates [12], spectroscopy ellipsometry is the most suitable and effective alternative method [13].This is because the substrate has no effect on ellipsometry [14].Additionally, it has been utilized to analyze the kinetics of hydrogen interaction with a-Si:H during H 2 plasma treatment as well as hydrogen diffusion into the films [15].Through the analysis of the dielectric function using suitable models like the Tauc-Lorentz (TL) [16], Cody-Lorentz [17], Forouhi-Bloomer [18], Tetrahedral [19], Harmonic oscillator [20], and Bruggman Effective Medium Approximation (BEMA) [21].To acquire accurate and reliable results, it is crucial to select an appropriate model and verify that best fitting.
Here, we present a novel approach to generating amorphous structures of silicon from SiH 4 with H 2 doped by the PECVD technique and reveal the detailed evolution of the electronic structure in n-type a-Si:H and their relationship to the performance of photovoltaic devices.The impact of intermittent hydrogen plasma treatment on the microstructure and optoelectronic characteristics of a-Si:H films is the topic of the current article.To explore the microstructure of the films, ex-situ spectroscopy ellipsometry measurements were carried out.The data were then analyzed using the TL and TL + BEMA models to find the optical constants, thickness, surface roughness, and volume fractions.Optimizing the deposition parameters of the PECVD system, which includes an increase in the rate of hydrogen gas, is expected to produce a single thin layer of a-Si:H type-n with optimum quality to improve the performance of solar cells based on a-Si:H p-i-n structure.

Experiment method
The a-Si:H thin layer is intrinsically deposited on corning 1737 glass and ITO substrates that do not use plasma waves plus chemical vapor chemical deposition (PECVD) techniques in key-based UHV space loads with a key electrode area of 19.62 cm 2 and 4 cm electrode separation.The total deposition time for each film was kept constant at 30 min and 60 min.The deposition parameters for the doped of hydrogen SiH 4 /H 2 are 0, 16, and 36, RF power 10 W, substrate temperature 270 °C, and process pressure (PP) 2000 mTorr.For N-0, no doped-H 2 was carried out during precipitation, while for N-16 and N-36 hydrogen was diluted for 30 min and 60 min, as shown in Table 1.No plasma treatment is carried out on the film's top surface.The scheme of precipitation and doped hydrogen for the sample is shown in Fig. 1.
For the characterization of hydrogenated a-Si:H, hydrogenated nanocrystalline silicon (nc-Si:H), and hydrogenated microcrystalline silicon (c-Si:H) films produced by RF-PECVD, numerous researchers have used the spectroscopy ellipsometry technique shown in Fig. 2 for the observed values of Ψ and Δ.The spectroscopy ellipsometry provides more precise structural information than Raman and can distinguish between nc/c-Si:H films and a-Si:H films [22].With in-situ and ex-situ methods, the transition from the amorphous to the nc/c-Si:H phase can be qualitatively examined [23].The real-time spectroscopy ellipsometry technique has provided a clear explanation for the structural phase transitions from a-Si:H to c-Si:H [24], the growth process of the films throughout various stages of film deposition [25], and the surface morphology of the films [26].It has been effective to use spectroscopy  The current article discusses the impact of doped H 2 plasma treatment on the microstructure and optoelectronic characteristics of a-Si:H films.To study the microstructure of the films.The data were then analyzed using the TL + Gauss model and TL + BEMA models to determine the optical constants [27], thickness [28], surface roughness [29], volume fractions of the amorphous/microcrystalline phase [30], and voids in the films [31].It was discovered that the estimated values agreed with those obtained from FTIR, transmission electron microscopy, and other methods.

Results and discussion
High-resolution TEM images on a-Si:H doped-H 2 thin films are shown by electron diffraction patterns; a-Si:H doped-H 2 thin film shows a dark contrast with a significant growth direction on H 2 -doped and a corresponding electron diffraction pattern is shown from the diffraction ring, as well as spots indicating mixed-phase amorphous and nanocrystal morphologies shown in Fig. 3.The enlarged TEM image also shows the formation of a nanocrystalline structure with the same nanocrystalline electron diffraction pattern for the a-Si:H doped-H 2 thin film.Therefore, the a-Si:H doped-H 2 thin film exhibits nanocrystals embedded in the amorphous thin film [32].Here, there are morphological differences between a-Si:H thin films without H 2 -doped and with H 2 -doped a-Si:H which can cause appropriate differences in structural properties [33].We further obtained an a-Si:H thin film without H 2 -doped and with H 2 -doped a-Si:H included in its hydrogen-related electronic structure characteristics.
Topographic measurements of a-Si:H surface thin films were carried out using the manufactured instrument by AIST-NT (Smart SPM 1000).This measurement uses tapping mode on a scan area of 1 μm by 1 μm, and the results of topographic measurements were evaluated using the Gwyddion 79 software including data leveling, and background reduction [34].Figure 4 shows the topography and surface morphology of the sample with roughness values of 8.95, 6.20, and 3.50 nm, as shown in Table 2.The root means that the squared roughness values for the three a-Si:H thin films are around 0.1 μm.Based on these results, the surface roughness was not modeled for the ellipsometry spectroscopic evaluations as a separate layer.This observation shows an increase in structural orders with the doped-H 2 [35].It was also observed that the volume fraction of the amorphous and void phases decreased and the nanocrystalline phase increased with increasing plasma hydrogen phase.
To support the optical and electronic analysis of each sample, the surface roughness of the N-1 and N-2 samples was obtained, as shown in Fig. 4, which shows the surface morphology of N-1 and N-2 on a 2 × 2 μm 2 film area.It is clear that this sample is very uniform across samples, small grain size is observed in sample N-1, and grain size increases in sample N-2.It is caused by changes in the nucleation and growth process on samples with an increase in gas-doped H 2 [36].The surface roughness value Root-Mean-Square (RMS) of the samples was obtained from AFM measurements of 5.45 and 6.12 nm according to the samples N-1 and N-2 samples.The a-Si:H n-type layer is measured at room temperature in the optical energy range of 0.6-6.6 eV.The results of the N-1 and N-2 fitting samples obtained information and parameters from the optical properties of the material, including refractive index, dielectric constant, thickness, and roughness of the n-type layer a-Si:H [37].
It has been possible to express the dielectric function of a-Si:H N-type films using a variety of models.The Tauc-Lorentz model, the Harmonic Oscillator model, the Gauss model, and a mixture of these models are the most often utilized models among these [38].To calculate the amorphous, nanocrystalline, and void fractions in a-Si:H, the TL Model, and BEMA were also used.To fit the dielectric function of amorphous films N-type in our situation, we utilized TL and TL and Gauss model, respectively [39].The  band gap, optical constants, and thickness of these films are estimated using a two-layer system consisting of an ITO substrate, bulk layer, surface roughness layer, and air (Fig. 5a).
The measured experimental ψ and Δ of photon energy and those obtained using model fitting are shown in Fig. 6a.
For films with the hallmark of nanocrystalline N-1, the Gauss model is used to fit the two band-to-band transition near 3.2 and 4.6 eV along with the TL model for fundamental transition near 3.6 eV.In contrast, for films with amorphous films N-0, the TL model alone could give a good fitting to the experimental data with a single peak in < ε 2 > corresponding to a-Si:H around 1.0-4.6 eV shown in Fig. 7a.By deconvoluting the peak in the < ε 2 > spectra of these samples using the peak fit algorithm of the software ORIGIN 8.5, the initial values of various peaks for the fitting were produced.Table 3 contains the values of the fitted parameters A, E 0 , C, and E g .The values of the thickness and band gap were determined via UV-vis-NIR transmission.
We have also estimated the amorphous (f a ), crystalline (f c ), and void (f v ) fractions as well as the thickness, band gap, and optical constants of the films using the TL model in conjunction with BEMA.To achieve this, we used a two-layer structure that included a corning substrate, bulk layer, surface roughness layer, and air (Fig. 5b).A blend of amorphous, crystalline, and void phases makes up both the bulk layer and the surface layer [40].Figure 6b displays the experimentally measured and calculated photon energies as well as results from model fitting with TL + BEMA.Table 3 also includes the values of the fitted parameters A, E 0 , C, and Eg as determined by TL + BEMA.Along with the roughness values acquired from AFM, the computed values of the amorphous, crystalline, and void percent for both the bulk layer and rough surface layer are reported in Table 2.This observation shows an increase in structural orders with dilution of H 2 .It was also observed that the volume fraction of the amorphous and void phases decreased and the crystalline phase increased with increasing plasma hydrogen phase (Fig. 8).The thickness, band gap, and refractive index (at 632 nm) calculated using the two models also agreed with the values obtained from the UV-VIS transmission data.The roughness of the film from the AFM study is also very similar to the thickness of the rough surface layer (Fig. 9).The table clearly shows that the values of A, E 0 , C, and Eg between the two models (TL/TL + G and TL + BEMA) are in good agreement.With an increase in the number of H 2 plasma doped steps, the value of the amplitude parameter (A) is found to rise, while the broadening parameter (C) is found to fall.
Our results show that with increasing hydrogen dilution not only does the total hydrogen content in the film increase, but the amount of hydrogen bound as strong Si-H also increases.Band gap increases, because the Si-H bond has a  greater energy than the Si-Si bond and the valence band is lowered in the film.Figure 10 shows an increase in band gap with total hydrogen content in the film.It was observed that the band gap increased almost linearly with the total hydrogen content.The increase in hydrogen content can be understood as follows: During hydrogen plasma treatment, a small part of the amorphous phase is engraved and the nc-Si seed layer is formed [41].This facilitates the movement of hydrogen atoms on the surface of the film and allows it to diffuse in large numbers and replace weak Si-Si bonds with strong Si-H bonds.With more plasma treatment steps, more hydrogen enters most of the film as the amorphous phase is etched and what remains is the more regular nc-Si phase [42].This results in a change in peak position at < ε 2 > toward higher energy.The peak amplitude < ε 2 > was also found to increase with H 2 and the widening parameter decreased (Fig. 10).
The thickness of the film decreases with increasing hydrogen dilution; the growth rate is lower than the nc-Si: H. phase.Initially, when most of the amorphous phase will be scratched, its thickness decreases rapidly [43].However, when most of the a-Si:H phase is carved with subsequent plasma treatments and the nc-Si phase remains stable, the rate of decrease in total thickness decreases.
Where there is a change in E x-1 energy in the N-2 sample which changes the energy to a smaller one compared to the N-1 sample with a large change of 0.5 eV.Likewise, with the E x-2 value, this proves that at this energy, the N-2 sample can absorb more energy than the N-1 sample and ultimately can improve the performance of the n-type layer on a-Si:H p-i-n solar cell.It is explained that along with the addition of H 2 gas to the N-1 sample in the deposition process, the Si atoms that will bind to H atoms will also increase in addition to bonding with fellow Si atoms [44].The more Si atoms that are bonded, the less empty Si hands will be; in other words, the dangling bond which causes defects in the forbidden band gap is also reduced [45].The reduced defects in the band gap will make it easier for the electrons in the valence band to jump or move to the conduction band, so that the electrical conductivity will increase with increasing doped of H 2 .Because H 2 gas acts as a reduction in the defect condition in the band gap by passivating the dangling bond so as to bring an effective dopant [46].However, the N-2 sample shows different results, where the electrical conductivity of the coating decreases.In addition to lowering disability, H 2 incoming interstitials in tetrahedral bonding configuration can also make sure that the active dopant is PH 3 , so that H 2 gas can limit the atoms that appear in the active dopant configuration (P 4 + ) [47]. Figure 11 shows the FTIR spectra of the a-Si:H thin-film samples deposited at different film growth rates on the gas mixture of SiH 4 and H 2 , respectively.The infrared absorption measurements of both samples showed very similar spectral results.The bonding that occurs in a-Si:H thin film shows relatively low Si-Hn-bond characteristics.The Si-H dangling bonds can be correlated to the passivation of the internal surfaces of microscale voids by trapping H 2 molecules [48].However, the relatively dense a-Si:H thin film N-0, as compared to the a-Si:H thin film N-2, showed a large intensity from the Si-H bond.This indicates that the a-Si:H thin film N-2 forms a reduced internal structure of nanoscale voids including the Si-H bonds.The structural change from microvoids to nanovoids induces an increase in the Si-H bond length, so that the band gap of the a-Si:H thin film decreased to that of the a-Si:H thin film N-2.
Figure 12 shows describe the model of a-Si:H type-n energy levels as a result of increased doped H 2 gas, where the atomic state of H can diffuse into a new dangling bond.The dangling bond is then left behind by the H atom, which then recombines with D − to form a normal Si-Si bond [49].The observed conductivity reduction was thought to be caused by the active dopant activation mechanism by H 2 , and it was found that the addition of a high doped of H 2 made the doping ineffective, thereby reducing the electrical conductivity value of the a-Si:H type-n layer.Meanwhile, the increasing number of H 2 atoms results in the re-formation of defects in the forbidden band gap but not as a dangling bond but in the form of another defect.This excessive number of H atoms results in the formation of a gap state in the energy band gap [50].The existence of a gap state in the energy band gap will prevent electrons from jumping from the valence band to the conduction band.Electrons that should be able to move from the valence band must first pass through the gap state and then jump to the conduction band [51].
The effect of increasing H 2 doped on the energy band gap width in the a-Si:H type-n layer, where the bandwidth of the energy band gap increases with increasing doped of H 2 .Increased doped of H 2 plays a role in reducing the defects in the bandgap of amorphous silicon materials.The mechanism of defect reduction can occur where the atoms H will make the bonds Si empty into passive.Along with the addition of the H 2 gas rate, the bonds of Si are empty and will be filled by the atoms H 2 , so that defects in the energy band gap will be reduced resulting in the shift of the valence band to the bottom and the valence band to the top.This will result in the widening of the energy band gap, so that the band gap will get bigger.Likewise, increasing the doped of H 2 will slow down the rate of silane gas by most researchers associated with an increase in hydrogen concentration, which will reduce the thickness of the layer formed.

Conclusion
In summary, the effect of increasing H 2 gas doped on the optical characteristics of the n-type layer a-Si:H deposition of the PECVD technique shows that the change in the energy gap increases along with the increase in H 2 doped, but the increasing number of hydrogen atoms results in defects in the energy gap but not as a dangling bond but in the form of another defect.We observed that the microstructure factor is directly related to a void fraction in the films and to the amplitude of the imaginary part of the pseudo-dielectric function.The single TL model could not give accurate results in the case of mixed-phase films and TL + G model must be used to calculate the optical constants, thickness, and band gap of the films.Further, results obtained from the TL + G model have been compared with those obtained using TL + BEMA for the accuracy of the model.Meanwhile, the effect of increasing the H 2 gas doped shows a decrease in the optical conductivity value of the a-Si:H type-n layer.On the other hand, the bond structure of each H 2 -doped thin film showed a change from an amorphous structure to a nanocrystal which was evenly distributed at each H 2 -doped binding.Overall, we believe that the addition of doped-H 2 to our findings can help increase the power conversion output of solar cells due to modifications in the electronic structure.

Fig. 1
Fig. 1 Sample preparation scheme a-Si:H solar cells.a RF-PECVD techniques and b Schematic diagram chamber.c Thin-film a-Si:H solar cells

Fig. 5 aFig. 6
Fig. 5 a Schematic two-layer model for TL, TL + G and b TL + BEMA model

Fig. 7
Fig. 7 The imaginary part of the complex dielectric function using the ellipsometry spectroscopic.a N-type films (fitted with the Tauc-Lorentz model and Gauss model) and b N-type films (fitted with Tauc-Lorentz and BEMA model)

Fig. 8 Fig. 9
Fig. 8 Void fraction and crystalline fraction plotted as function of H 2 plasma steps

Fig. 11 FTIR
Fig. 11 FTIR absorption spectra of the a-Si:H-doped H 2 thin films grown by PECVD.The a-Si:H thin film shows the larger infrared light absorptions for the hydrogen-related chemical bonds of Si H at 2000 cm −1 and the a-Si:H thin film shows the chemical bonds of Si H2 at 2100 cm −1 as well as Si H at 2000 cm −1 , indicating the existence of nanocrystalline grain boundaries saturated by H 2 molecules