Effect of Laser Processing Parameters on Polycrystalline Silicon Surface Texturing

Effect of Laser Processing Parameters on Polycrystalline Silicon Surface Texturing

On this page
As mentioned earlier, the laser processing is influenced by environmental factors and processing methods, which can result in varying texturing effects. The quality is largely determined by the method used in the laser texturing process. Environmental conditions also play a role. In a standard processing environment, laser parameters such as wavelength, power, scanning speed, fill spacing, number of scans, and others, including frequency, spot overlap, pulse rate, and pulse width for pulsed lasers, also affect the texturing outcome. In particular, the power level directly impacts the material removal rate, and different materials absorb light at different wavelengths at varying rates. Additionally, spot overlap at different frequencies also influences the processing outcome.
 

3.1 Effect of Laser Wavelength on Processing Polycrystalline Silicon Wafers

Laser texturing involves ablating the material's surface through photothermal or photochemical effects to remove material and create a textured surface. Materials absorb light at different rates depending on the wavelength. As shown in Figure 6, the effects of different laser wavelengths on polysilicon materials vary. Shaohu Pei and others used an Nd:YAG solid-state laser with a pulse width of less than 10 ns to drill holes in the polysilicon surface at wavelengths of 1064 nm, 532 nm, 355 nm, and 266 nm. The results show that polycrystalline silicon wafers have the highest absorption efficiency at a wavelength of 532 nm. When using a 1064 nm wavelength laser, the diameter of the heat-affected zone increases with the number of pulses in a star-shaped distribution, and staggered lines appear at the bottom of the pit. Bineti and others used an Nd:YAG laser with a 10 ps pulse width and second and third harmonic generators to convert the primary 1064 nm wavelength (infrared) into 532 nm (green) and 355 nm (ultraviolet). A 1/4 wave plate was also used to convert linearly polarized light into circularly polarized light, which was then used to texture the surface of the polycrystalline silicon wafers. The results show that the infrared laser texturing of multi-grade silicon wafers achieves the lowest reflectivity, less than 8%, while ultraviolet and green laser texturing achieve average reflectance values of less than 11% and 13%, respectively.

The varying effects of different laser wavelengths on the final texture of polycrystalline silicon wafers can be explained by the fact that other parameters must remain constant when using different wavelengths to etch the material's surface. For lasers with wavelengths ranging from 1064 nm to 355 nm, the damage threshold and light penetration depth of polysilicon decrease as the wavelength shortens. Additionally, with shorter wavelengths, the energy is primarily deposited in a very shallow surface layer. As the wavelength decreases, the absorption coefficient of polycrystalline silicon increases, making short-wavelength lasers more likely to ablate the material under the same energy density. At a wavelength of 532 nm, polycrystalline silicon experiences the maximum degree of melting, while at 1064 nm, the melting is minimal, and at 355 nm, it is the least.
 
Currently, based on the reviewed literature, most studies use infrared lasers with a wavelength of 1064 nm for texturing, while green and ultraviolet lasers are primarily employed in theoretical research. This may be because, although green and ultraviolet lasers, generated by frequency doubling of infrared wavelengths, produce fine processing results, the efficiency of light conversion is significantly reduced, making them less suitable for industrial production. Therefore, when selecting lasers with different wavelengths, both processing performance and factors like efficiency and cost should be considered.
 

3.2 Effect of Pulsed Laser Spot Spacing in Various Directions on Polycrystalline Silicon Surface Texturing

Pulsed lasers emit discontinuous light. During processing, each pulse is influenced by parameters such as frequency, scanning speed, and fill spacing, resulting in the formation of microholes with varying distances per unit area. The spacing between these microholes can be categorized into overlap holes, array holes, and scanning pitch holes based on their characteristics. Overlap holes refer to the microscopic morphology formed by the positions of adjacent holes, ranging from tangent (Figure 7a) to overlapping (Figure 7b). Array holes are characterized by light spots that transition from being tangent to being separated by a specific distance. Scanning pitch holes refer to the microscopic pattern formed when different fill spacings are used during line-by-line scanning in laser processing.

Morphology of overlapping holes formed by adjusting pulse laser parameters
Figure 7 Morphology of overlapping holes formed by adjusting pulse laser parameters
 
Radfar and others used a nanosecond infrared laser with an output power of 30 W to texture the surface of polycrystalline silicon wafers. They varied the frequency (30 kHz / 55 kHz) to adjust the spot overlap rate (72% / 82%) and observed the resulting morphological changes in the overlapping holes formed on the silicon surface. The optimal reflectivity was found when the repetition frequency was 55 kHz and the overlap rate was 82%. Tiandai Jia and studied the effect of spot spacing on the reflectivity of textured surfaces. They developed a theoretical model using the finite difference time domain (FDTD) method, simulating the interaction between incident light with a wavelength of 700nm and the holes, and determined that the optimal array hole pitch was 30μm. They then used a laser with a wavelength of 515nm, a focused spot diameter of 30μm, and a pulse width of less than 15ps to conduct single-point drilling experiments on the polycrystalline silicon wafer surface. The etched images showed hole spacings of 30μm, 40μm, 50μm, and 60μm (Figure 8). They compared the morphology of array holes with different hole spacings and their surface reflectance in the wavelength range of 350–1050nm under these processing parameters. The results showed that a hole pitch of 30μm produced the optimal processing distance, with a microstructure that effectively reduced the surface reflectivity of polysilicon and improved the photoelectric conversion efficiency of polysilicon solar cells.
 
The surface of multi-grade silicon samples etched by laser under different hole spacing and the corresponding quilt surface
Figure 8 The surface of multi-grade silicon samples etched by laser under different hole spacing and the corresponding quilt surface
 
Xiaozhan Lu and colleagues studied the effects of scanning path spacing and the number of scans on polycrystalline silicon surface texturing. They used a picosecond laser with a pulse width of less than 20ps and a wavelength of 1064nm to process polycrystalline silicon wafers. The laser scanning path spacing was set to a series of closely packed straight lines, with spacings of 20μm, 30μm, and 40μm, as shown in Figure 9. When the scanning spacing was 30μm, relatively uniform and regular hole diameters and depths were achieved. Based on this spacing, they also studied the impact of varying the number of scans (20, 60, 85, and 140) on texturing. With a scanning spacing of 30μm and 85 single-line scans, the texture quality was good, with micropores evenly and densely distributed. No fused residue or cracks were observed around the holes or in the surrounding area, providing a solid foundation for subsequent chemical etching. The primary goal of adjusting micropore spacing is to create a more uniform and light-trapping pit texture on the polycrystalline silicon wafer. As shown in Table 2, by adjusting the spacing between different types of holes, the density of pits per unit area is also altered. Research on light reflection from polycrystalline silicon surfaces has demonstrated that pit density is a key factor influencing the light-trapping properties of the surface.

Microscopic morphology of laser scanned samples under different scanning spacings
Figure 9 Microscopic morphology of laser scanned samples under different scanning spacings
 
Table 2 Effect of spot spacing in different directions on the light absorption properties of multi-crystalline silicon wafers
Source Overlap Hole (Spot Overlap Rate, %) Array Hole (Hole Spacing, µm) Scanning Pitch Holes (Scanning Path Spacing, µm) Reflectivity (%) Reflectivity After Chemical Etching (%)
Radfar and others 82 6.59
Tiandai Jia and others 30 5.17 6.95
Xiaozhan Lu and others 30 About 8 <5
 
Note: "—" means that this content is not indicated in the literature.
 

3.3 The Influence of Parameters such as Beam Incident Angle, Energy Distribution, and Pulse Width on Polycrystalline Silicon Surface Texturing

Weihong Zhang and his team etched the surface of polysilicon wafers at various laser incident angles of 0°, 30°, 45°, and 60° to investigate their impact on the texturing of the polysilicon surface. The results show that the groove etched by the laser at a 45° incident angle produces the best texturing effect, with a reflectance as low as 8%. Additionally, as shown in Figure 10, the energy distribution varies with different beam profiles, leading to distinct microstructure morphologies. Compared to a Gaussian laser, the microstructure etched by a flat-top laser has a more uniform structure height and distribution. Furthermore, parameters such as pulse width, pulse count, and single-pulse energy also significantly impact the laser texturing effect. Taking a pulsed laser as an example, when the pulse duration is reduced from nanoseconds to picoseconds, and even to femtoseconds, the peak energy of the laser increases accordingly. Therefore, compared to a nanosecond laser, using picosecond or femtosecond lasers for etching results in more significant surface damage. According to the current literature, nanosecond lasers are most commonly used for texturing. Under typical processing conditions, increasing the number of pulses and single-pulse energy improves the texturing effect. Within a certain range, the light-trapping properties of the processed surface increase with the number of laser pulses.

(a) Gaussian beam profile and (b) flat-top beam profile energy distribution 
Figure 10 (a) Gaussian beam profile and (b) flat-top beam profile energy distribution
 

Nickname*:
E-mail*:
Rate*:
Comments*:


About the author
Teresa
Teresa
Teresa is a skilled author specializing in industrial technical articles with over eight years of experience. She has a deep understanding of manufacturing processes, material science, and technological advancements. Her work includes detailed analyses, process optimization techniques, and quality control methods that aim to enhance production efficiency and product quality across various industries. Teresa's articles are well-researched, clear, and informative, making complex industrial concepts accessible to professionals and stakeholders.