Abstract: Background and Objective: Light capturing is an essential part of many optical devices such as optoelectronic devises and solar cells. This study aimed at modifying surface reflectivity of silicon to improve light trapping. A simple and easily controllable etching technique was used to achieve this goal. Methodology: The surface topography of (100) P-type silicon wafers was modified by etching a controllable pyramid structure on these surfaces. Potassium hydroxide (KOH) solution was used to etch the silicon surface; the concentrations of KOH were varied between (20-36 wt%)±0.14 wt% with 3±0.1 wt% of isopropyl alcohol (IPA) at temperatures between (60-80°C) ±0.1°C and a varying etching time between 20-40 min, the mean and standard deviation of the pyramids size was calculated by taking five SEM images per case. Results: The optimal etching condition was determined as a 20±0.14 wt% aqueous KOH solution with a 3±0.1 wt% IPA for an etching time of 40 min at an etching bath temperature of 80°C. The pyramids resulting under this condition has a size of 1.7±0.2 μm. The surface reflectivity at these optimal conditions was measured to be 11±0.2% in the wavelength range 550-840 nm. Conclusion: The optical conditions for etching and modifying the morphology and reflectivity of (100) silicon surfaces were determined. Surface reflectivity modification is one the effective methods that can be used to control light trapping and light scattering needed for proper functioning of many optical applications and optical devices.
INTRODUCTION
Silicon with various structural morphologies is widely used for optoelectronic devices1. The reflection of light at the surface is one of the main loss mechanisms; where many efforts have been made to reduce this optical loss2. The texturization of the silicon surface leads to an increase in its topographic irregularities which results in an increased surface area. This will enhance its responsivity and therefore, it increases its efficiency of converting light into electric current3. Anisotropic etching of silicon has found a broad acceptability in fabrication of industrial silicon solar cells in order to reduce reflection losses from the front surface4-6.
Anisotropic etchants for silicon are aqueous alkaline solutions, where the main component can be either an organic or an inorganic compound7. Among the inorganic solutions are sodium hydroxides (NaOH) or potassium hydroxide (KOH) with isopropyl alcohol (IPA) and water; which are generally used in industry.
In this study, texturization based on alkaline anisotropic etching was investigated by using KOH as alkaline etchant and IPA as a surfactant. Adding IPA can improve the wettability of silicon surface and control the etching rate by preventing an explosive reaction between the silicon surface and the OH-ions8-10. Etching of silicon in KOH solution has the advantages of simplicity, ease of handling, low-cost and homogeneous etching rate of the (100) crystal plane11. Silicon is one of the basic materials for the semiconductor industry and (100) orientation was chosen to be evaluated in this study done to its popularity and ease of handling. This study aimed at optimizing the process of surface texturization to obtain the lowest reflectance where the reflectance depends on the shape of the pyramid structure9. Good control of the pyramid micro structure shape requires careful design of the etching conditions such as, solution concentration, temperature and etching time.
This study aimed at finding a simple and cost effective method for controlling the surface reflectivity of one of the most popular materials in electronic and optoelectronic applications. Modifying surface reflectivity can help in controlling light trapping properties.
MATERIALS AND METHODS
Mono-crystalline P-type (100) silicon with a resistivity of 3-30 Ω cm and a thickness of 675±25 μm were cleaved into 4 cm2. Before preparing the pyramid structure, the silicon wafers were cleaned using acetone, ethanol and deionized water to remove the adsorbed dust and surface contaminants, then silicon wafers were immersed into 20% NaOH solution at 80°C for 5 min and again they were cleaned using acetone, ethanol and deionized water.
To fabricate the pyramids, different concentration KOH solutions were prepared. The concentrations were varied between (20-36 wt%)±0.14 wt% with 3±0.1 wt% of isopropyl alcohol (IPA) at temperatures between (60 - 80°C) ±0.1°C and a varying etching time between 20-40 min. After obtaining the pyramid structure, the silicon wafers were washed in deionized water followed by boiling in dilute hydrofluoric acid (5% HF) to remove the metallic impurities and finally they were washed again in deionized water and dried out using hot air. The different pyramid sizes were accomplished by changing the chemical concentration, temperature and the etching time resulting in a surface with small, big and mixed pyramids on a completely and not completely covered waver surface. These textured surfaces have been studied by FEI Inspect F50 (Eindhoven, Netherlands) Scanning Electron Microscopy (SEM) and the percent reflectivity spectra were recorded using a FilmTek 3000 Scientific Computing International Carlsbad, CA 92011 spectrophotometer in the wavelength range of 250-840 nm.
Statistical analysis of data: In order to define the size of the pyramids after texturing, intercept method was used. First, draw the diagonal line on the SEM image for example (Fig. 1). As the scale bar indicates 4 μm, the length of diagonal line was 20 μm. Then the number of pyramids caught by the diagonal lines was counted. The length of the diagonal line was divided by number of pyramids to get the average value of pyramids size10. The mean and standard deviation of the pyramids size was calculated by taking five SEM images per case and the mean and standard deviation of the pyramids percent reflectivity was calculated by taking three spectra for the wavelength range of 550-840 nm. Errors in the values of concentration were calculated by Eq. 112:
| (1) |
where, CKOH and ΔCKOH are the concentration and the error in the concentration, mKOH and ΔmKOH are the mass and the error in the mass of solution KOH and mIDW and ΔmIDW are the mass and the error in the mass of solvent (deionized water).
RESULTS AND DISCUSSION
The reflectivity of (100) silicon surface was reduced substantially through out the visible spectrum. The simple and cost effective technique used was shown to work perfectly in surface modification of silicon wafers.
| Fig. 1: | Intercept method using SEM image |
| Fig. 2(a-b): | Reflection of incident light from (a) Bare planar Si and (b) Si micro pyramid texture |
The surface morphologies of different samples were investigated after varying the texturization process. Reflection of incident light from a bare polished plane of silicon and a silicon plane with micro pyramid texture is presented in Fig. 2. Light suffers a specular reflection from polished silicon surface, while it suffers from a mixture of diffuse and specular reflection from the pyramid structure. The measurements in this study were all taken at normal incidence.
Scanning Electron Microscope (SEM) micrographs of the different surface topographies resulting from different texturization times at 80°C in 20 wt% aqueous KOH solution with 3 wt% IPA are shown in Fig. 2. All the micrographs were taken using the same magnification. The pyramids grew in size with longer texturization time and they became regular and more uniformly distributed. The pyramid size obtained was between 0.9±0.2 μm and 1.7±0.2 μm. A big pyramid size indicates a lower density of pyramids.
The SEM micrographs of the pyramidal texture for the wafer etched for 20 min is shown in Fig. 3a. Small pyramids were visible and there were areas not yet covered with pyramids, this observation could be attributed to the short etching time. The pyramidal texture for the wafer etched for 30 min is seen in Fig. 3b, where, the pyramids have grown in size. After 40 min of etching time (Fig. 3c), the whole surface was covered with small pyramids and some pyramids grew larger and a more homogeneous distribution of pyramids could be seen. The size of pyramids increased as the texturization time increased. The surface texturization reduced the light reflection of the silicon wafer to around 11±0.2% in the wavelength range of 550-840 nm.
The percent reflectivity at normal incidence for the bare silicon wafer and for silicon wafers textured for 20, 30 and 40 min in 20 wt% aqueous KOH with 3 wt% IPA at 80°C are shown in Fig. 4. In the 550-800 nm range, the reflectivity of polished silicon surface dropped by 68.6% after being textured, where at a 30 min etching time the percent reflectivity was 12.6% at wavelength 650 nm and it is in agreement with the results reported by Bachtouli et al.13, where a percent reflectivity of 12.0% with NaOH low concentration was observed. The percent reflectivity versus texturization times for different wavelengths are shown in Fig. 5.
| Fig. 3(a-c): | SEM images of silicon surface after etching in a 20 wt% KOH + 3 wt% IPA at 80°C for different durations (a) 20 min, (b)30 min and (c) 40 min |
The percent reflectivity reduced with increased texturization time. The percent reflectivity dropped by 67.6% for the 650 nm wavelength (visible red), while a reduction of 78.6% was seen for these samples at 270 nm wavelength (ultra violet). The results are in a good agreement with the results reported by Singh et al.7, Sparber et al.14 and Wang et al.15, where a decrease in percent reflectivity was observed as the texturization time increased.
The SEM images for silicon surfaces textured with different KOH concentrations, 20, 24, 28 and 36 wt%, with 3 wt% IPA and heating at 80°C for 40 min are shown in Fig. 6. The shape and coverage of pyramids depend on KOH solution concentration. At relatively lower concentrations, 20 and 24%, the pyramid structure could be clearly observed as seen in Fig. 6a and b. At higher KOH concentrations, the etching rate was reduced and the pyramids were small and didn’t fill the whole substrate’s area as seen in Fig. 6c. When the KOH concentration increased to 36% the pyramid structure disappeared as seen in Fig. 6d. The disappearance of the pyramid structure was argued as follows, when the concentration of KOH increased the dissolution rate of silicon atoms from the crystal surface was so high such that the transport of the formed Si (OH)4 complex could not maintain following its formation, as the concentration of this complex became too high and it polymerized and covered the silicon surface preventing the etchant from reaching the surface of the silicon substrate16.
| Fig. 4: | Percent reflectivity of silicon wafers under various texturization times |
| Fig. 5: | Percent reflectivity versus texturization time for wavelengths 270, 360, 450, 550 and 650 nm |
Percent reflectivity of silicon surfaces textured at different concentrations, 20, 24, 28 and 36 wt%, with 3 wt% IPA at 80°C for 40 min are shown in Fig. 7. The percent reflectivity reduced by 57.7% as the KOH concentration decreased from 36-20%.
The percent reflectivity versus KOH concentration for different wavelengths, 270, 360, 450, 550 and 650 nm are offered in Fig. 8. The percent reflectivity increased as the KOH concentration increased, for example at 650 nm wavelength the percent reflectivity increased by 56.7% as the KOH concentration increased from 20% to 36%.
| Fig. 6(a-d): | SEM images of silicon surface after etching with KOH+3 wt% IPA mixture solution (a) 20 wt% KOH, (b) 24 wt% KOH, (c) 28 wt% KOH and (d) 36 wt% KOH, heating at 80°C for 40 min |
These results, specifically the 11±0.2% reflectivity at 20% KOH concentration, are in good agreement with the results reported by Basu et al.17, where they reported a reflectivity of 13% at low KOH concentrations. The increase of percent reflectivity is a direct indication of the loss of the pyramid structure and the smoothness of the surfaces of the treated wafers. The etching kinematics and the formation of the protective layer prevented the KOH from etching the surface of the wafer as mentioned earlier. The results are in a good agreement with the results reported by Wang et al.8 and Powell and Harrison18. Where an increase in percent reflectivity was observed as the concentration increased.
The SEM images of the pyramid textures at 20 wt% KOH and 3 wt% IPA based solution for 40 min etching time at different temperatures, 60, 70 and 80°C are presented in Fig. 9. The SEM images showed clearly the influence of temperature on the morphology of the pyramidal texture. Lower temperatures, 60 and 70°C, prevented the formation of complete pyramids on the wafer’s surface. At 80°C, below the boiling point of IPA 82.4°C, the pyramids were more homogenous with lower reflectivity. These results are comparable to those obtained by Wang et al.8, who reported similar pyramid sizes.
The variations of the percent reflectivity of (100) silicon textured at different etching temperatures, 60, 70 and 80°C, in KOH based solution of 20 wt% aqueous KOH with 3 wt% IPA for 40 min are presented in Fig. 10. An average decrease of 60% as more is observed for the reflectivity of all treated samples.
The percent reflectivity values versus temperature for wavelengths 270, 360, 450, 550 and 650 nm are shown in Fig. 11. The percent reflectivity decreased as the texturization temperature increased and the etching rate increased with increasing texturization temperature19,20.
| Fig. 7: | Percent reflectivity of silicon versus wavelength under various KOH concentrations |
| Fig. 8: | Percent reflectivity of silicon wafers under various KOH concentrations for wavelengths, 270, 360, 450, 550 and 650 nm |
At 650 nm wavelength the percent reflectivity decreased by 31.2% as the etching temperature increased from 60-80°C, while for the 270 nm wavelength (ultra-violet) the percent reflectivity decreased by 32.4% for the same texturization temperatures. This shows an almost monotonic change in reflectivity ’Between the etching temperatures’ over the whole studied spectral range as can be seen in Fig. 10 and 11.
Through this study, a low reflectivity of silicon surface was obtained by a simple, low-cost technique that has positive effects industrial applications. This study also recommends working on reducing the reflectivity of the silicon surface area more than the reported values by introducing nanopores on the surface of the pyramid structure.
| Fig. 9(a-c): | SEM images of silicon surface after etching in a 20 wt% KOH+3 wt% IPA based solution during 40 min at different temperatures (a) 60°C, (b) 70°C and (c) 80°C |
| Fig. 10: | Percent reflectivity of silicon wafers under various temperatures |
| Fig. 11: | Percent reflectivity versus temperature for wavelengths 270, 360, 450, 550 and 650 nm |
CONCLUSION
This study presented the results of the different factors effecting surface texturing for the purpose of producing pyramidal texture on (100) silicon surface. The surface morphology shows that the size of the pyramidal texture varies according to the etching conditions such KOH concentration, temperature and etching time. Low reflection of pyramid base was found between 1.5 and 1.8 μm. The optimized conditions were found to be at 20 wt% aqueous KOH with 3 wt% IPA, 40 min etching time and 80°C temperature. The surface reflectivity of silicon reached values as low as 11% for the wavelength range 550-840 nm; which is an advantageous result for solar cells and photovoltaic applications.
SIGNIFICANCE STATEMENT
This study provides a simple and effective way to control surface reflectivity. This can be beneficial for the optoelectronic industrial application that uses light trapping as part of their devices.
ACKNOWLEDGMENTS
Author would like to express their gratitude to Mr. Waddah Faris for assistance in the SEM observation at Hamdi Mango Center for Scientific Research building, Department of Geology, University of Jordan. We also acknowledge Dr. Walid Hamoudi for his helpful discussion, at Applied Sciences Departments, University of Technology.