Surface Modification of Wide Bandgap Semiconductor GaN Using Femtosecond Laser Induced Periodic Surface Structuring LIPSS

doi:10.21203/rs.3.rs-8239438/v1

Download PDF

Research Article

Surface Modification of Wide Bandgap Semiconductor GaN Using Femtosecond Laser Induced Periodic Surface Structuring LIPSS

https://doi.org/10.21203/rs.3.rs-8239438/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

The constant growth of power electronics field of technologies and applications demands ever growing research to find novel materials with characteristics to stand harsher work conditions and wider range of application, with lower costs and energy consumption [1-3]; which in turn requires the search for faster, cheaper and more effective machining techniques of these new materials. Currently we are witnessing the huge replacement of Silicon-based electronics with its better alternative, wide bandgap semiconductors, such as Silicon Carbide (SiC), Aluminum Nitride (AlN), and Gallium Nitride (GaN), to name a few. In our experiment, we investigate the surface modification of wide bandgap semiconductor crystal GaN by femtosecond laser irradiation, in different experimental parameters. The goal is to obtain a data base for optimal experimental conditions to achieve highly reproducible laser induced surface structures, also ca ,lled ripples, by means of femtosecond laser radiation. The results obtained and recorded are useful for future experiments involving micromachining of wide bandgap semiconductors, and can be applied for wide range of applications in industrial, medical and military fields.

The main objective of our research is to better study the characteristics of wide bandgap (WBG) semiconductors, namely GaN, and its interaction with ultrafast laser irradiation, to create an active, useful applicable database of experimental conditions. The availability of such database will allow the creation of highly periodic, wide areas of LIPSS on the surface of WBG GaN, with minimal preliminary experiments, to save time and effort for other current and future researchers who work with GaN. Hoping such database will facilitate and speed up the industrial implementation of GaN for wider range of applications.

In our study, we investigate the laser micromachining of WBG semiconductor GaN, by studying the femtosecond laser radiation (λ = 1030 nm, τ ≈ 180 fs) interaction with the surface of transparent WBG crystal GaN under different experimental conditions. By definition, WBG semiconductors have direct energy gap equal or higher than 2eV[4]. GaN is a WBG semiconductor belonging to group III-V Nitrides, which has a wurtzite crystal structure and direct energy gap of 3.5 eV, providing it with special abilities to withstand higher voltages, energies and temperatures, compared to other conventional semiconductors. The bandgap width of GaN make it ideal for optoelectronic application, especially manufacturing blue and UV LEDs and laser diodes, and for high-power and high-frequency devices [4–6]. Its hardness, high thermal conductivity and high chemical inertness in addition to its high voltage and thermal-breakdown limits all make GaN a perfect candidate for applications in fields of telecommunication (5G devices), material coating (protective coatings for solar cell devices), and military and space applications [3, 7].

Owning to these superb chemical and mechanical characteristic that makes GaN so attractive, the high chemical inertness and hardness make it insusceptible to the conventional fabrication methods such as etching, photolithography, and deposition, so laser micro- and nano-machining became a promising alternative.

Femtosecond laser induced periodic surface structures (LIPSS) is a one step, no contact, low-cost micromachining technique used for surface modifications of variety of solid materials, ranging from metals to semiconductors and dielectrics [7–9]. It is characterized by the periodic ripples created on the irradiated surface with almost no depth and of period equal or smaller than the wavelength of laser radiation used [7, 10], performed in air. LIPSS can be used for surface functionalization by changing and controlling the properties of the surfaces irradiated. These properties include optical, chemical, electrical and mechanical, tribological and wettability [11–12].

As the purpose of our study focuses on the experimental parameters involved in creating and controlling the LIPSS on GaN, the mechanisms of LIPSS formation will not be discussed in details. It is noteworthy to mention that so far there is not a single, unified theory behind LIPSS formations, but there are many works in literature that were dedicated to studying the several LIPSS formation mechanisms proposed [13–14], such as [10] the excitation of surface plasmon polaritons SPP [15–17], second harmonic generation [18, 19], and parametric decay process [17, 20], with results being consistent with experiments.

Laser machining involves a number of variable parameters concerning the experimental conditions, these parameters can be divided into four distinguishable categories: laser beam parameters, scanning parameters, sample characteristics, and processing conditions. For our study, sample characteristics and processing conditions remain the same during all experiments, while we focus on investigating the variation of laser beam parameters and scanning parameters and their effect on the characteristics of the induced structures. The goal is to obtain a data base of optimal experimental parameters to produce highly periodic and homogenous LIPSS on large surfaces of GaN. The investigated laser beam parameters are: laser fluence, laser energy, laser frequency, laser focus position and laser polarization, and the investigated scanning parameters are: scanning speed, repetition and scanning direction. The results are discussed regarding the potential applications of LIPSS, such as: surface characteristic control (wettability), antibacterial properties (surface topography alteration for medical uses), fabrication of quantum dots and quantum-wires (LEDs, solar cells), and enhancement of optical properties (photoluminescence, absorption) [4–7, 10].

The experiments to determine the optimal parameters for laser irradiation of GaN surface to produce LIPSS is performed by utilizing an easily controllable setup (see Fig. 1). The setup consists of ultrafast laser source, a diode-pumped Yb: KGW femtosecond laser source Pharos PH2 (Light Conversion Inc.) generating laser pulses of wavelength λ = 1030 nm and duration τ ≈ 180 fs, with variable frequency ranging from few Hz to 100 KHz, and maximum power of 10 W, with maximum Energy in pulse reaching 1 mJ. Guiding optics are used to guide the laser beam through a beam expander to the Laser scanner intelliSCAN 14 iii (Scanlab Gmbh.), containing an F-theta lens which focuses and controls the scanning process, by moving the laser beam in straight lines over the surface of GaN, at an angle of incidence i = 0°. The GaN used is in the form of an Undoped, N-type, C-plane (0001) Gallium Nitride Single Crystal Substrate with double side polishing, of dimensions 10x10.5 (mm) and thickness 350 µm. The software LaserDESK is used to synchronize the laser radiation with the scanning process to control the dimensions c parameters of the machining process (number of pulses, scanning speed and direction, shape and dimensions of the area to be irradiated) for accurate control over the different range of the combination of parameters tested during the experiments. The control of measurements and machining parameters was done by observing the LIPSS obtained at different set of parameters separately using the confocal laser scan microscope of Ziess LSM 900. The results were recorded and compared to gather the best outcome.

At the start of our experiments, for easier calibration of the initially investigated parameters and fine tuning of said parameters, and to reduce the wasted material, we observe the LIPSS in single spots on smaller areas. Later on, after establishing the best laser parameters to work with, we move on making the LIPSS continuous on larger surface area, by turning them from singular spots to consecutive, overlapping spots to form parallel and equally-distanced, adjacent lines (canals), forming squares of dimensions 1 mm by 1 mm. These squares allow enough surface area to be treated to observe how larger surface areas of LIPSS will present and be affected by the laser radiation process ongoing, which is more practical for implementing LIPSS for actual applications, compared to only studying single spots of LIPSS. The table below shows the full range of each parameter tested in our experiments.

Table 1

shows the full range of each parameter tested in our experiments.
Range of experimental parameters studied
Wavelength λ	1030 nm
Pulse duration τ	180 fs
Frequency F	100-1KHz
Energy E	10–20 µJ
Position of laser focus relative to GaN surface	0-500 µm
Laser spot D	10–40 µm
Scanning Speed V	5–10 mm/s
Scanning repetition	1–8 times
Scanning direction in respect to LIPSS orientation	Parallel/ perpendicular
Polarization	Linear/ Circular

3.1. LIDT

It is important to note that LIPSS are produced when a surface is irradiated with laser radiation of fluence equal or slightly higher than its laser induced damage threshold (LIDT), but smaller than its ablation threshold [14]. Reaching the LIDT is important for micromachining the surface to obtain permanent changes, which alters the surface topological, chemical and optical characteristics of the material, which is the main goal of micromachining in our case. To calculate the LIDT of our material, we irradiated the material with single spot irradiation at different energies. LIDT energy was determined at the point where a permanent change occurs. According to our calculations, LIDT fluence of GaN at 1030 nm is F_th = 1.04 J/cm².

3.2. Laser Focus Position

In our experimental setup, we use a beam expander to minimize the focus diameter of the laser, from about 40 µm to ultimately achieve a minimal diameter of about 10 µm, in which the formed LIPSS, or ripples, have a period Λ = 1 µm. This is called a low spatial frequency LIPSS, or LSFL, which periodicity is equal or slightly shorter than the wavelength Λ ≤ λ [7], as it is in our case. For a laser spot of such small size, we realize the intensity of laser radiation is quite large, and would easily go beyond the ablation threshold, even at low energies and frequency settings of the laser. For that purpose, we decided that ideally the working position of the laser radiation to be in the vicinity of, but different from its focus position. The criteria to choose the best working position of laser focus relative to GaN surface are: preserving the circularity of the laser spot and having better control of laser fluence. In addition to that, having a larger area of laser irradiation upon the surface is an advantage when treating larger surfaces, as it maintains better surface periodicity and reduces machining time.

To determine the best working position, we decided to irradiate the surface of GaN with same laser radiation settings at different positions from focus, by moving the laser scanner head (containing the f-theta lens) vertically along the z-axis, in steps of 100 µm. Figure 3 below shows the results, concluding the best working position to be at 300 µm below the laser focus position. At that position the laser diameter is measure to be about 20 µm.

4.1 Energy fluence

By definition, laser fluence is the intensity of laser irradiation energy on a surface area of the material irradiated. Laser fluence can be changed by varying the laser energy or the spot size. In our case, we varied the laser fluence by changing the laser energy used. At that fluence, some surface structures were produced, but they were scarce and didn’t cover the whole surface of the irradiated area, and had low periodicity. By gradually increasing the energy of the laser irradiation, we determined that fluence best working for LIPSS was observed at energy E = 4 µJ, at frequency f = 1 KHz, where the created periodic structures had clear ripple forms and high homogeneity and periodicity. Lower frequency required lower scanning speed or more scanning repetitions, which prolongs the machining process, which would be a disadvantage in industrial settings. Higher frequency causes more instability and lower accuracy during the scanning process. For the sake of comparison, we continued irradiating at same conditions at higher energies to see what will happen. We noticed the formation of holes in the material with significant depth ≈ 15 µm, which shows we reached the ablation threshold of the crystal. Some of these holes had periodic structures visible in depth, as shown in Fig. 4 below, which could prove useful for certain applications, but that is out of the scope of our current study.

4.2 Scanning speed and overlapping

To get a uniformly organized LIPSS canals with smooth edges, the position between consecutive laser pulses should be smaller or equal to radius of laser spot. And since the laser spots are circular, the same distance is applied between the adjacent canals. Experimentally the formula for scanning speed is derived by: V≤ \(\:\frac{D}{2}f\) where V is the scanning speed of the scan head, D is laser spot diameter and f is the laser frequency used. Our experiments showed that if the distance between two consecutive spots is shorter than D/2, which is in our case 10 µm, the LIPSS created will not fully overlap and there will present areas where ripples are scarce or nonexistent (Fig. 5,a). On the other hand, if the distance is longer than D/2, then the overlapped ripples will lose their periodicity, and the surface will reach ablation threshold.

So we established the proper scanning speed for our case is V = 10 mm/s.

4.3 Scanning Repetition

Scanning repetition depends on laser frequency and scanning speed and is strongly related to them. As previously shown, for our experiments we use laser frequency of 1 KHz and scanning speed of 10 mm/s. For these values, the most uniform and periodic LIPSS were formed after 4 repetitions, example shown in Fig. 6,a).

After establishing this combination of parameters, an experiment was done with half speed with half the repetitions, respectively 5 mm/s for 2 repetitions, to test if the ripples resulted would present the same. After observation, we notice that, while in the central part of the irradiated surface area, the LIPSS appear to be uniform and periodic, the edges were quite rough and the ripples were scare and disorganized, as visible on Fig. 6,b).

4.4 Scanning direction

According to literature, irradiating with linearly polarized laser radiation causes the formed structures to be aligned perpendicular to the incident electric field vector; this is also the direction of orientation of LSFL, which is the most dominant in our study. To find which scanning direction is best for our experiments, we tested two scanning directions, moving the scanning head in direction parallel to the formation of LIPSS (scanning axis marked in red in Fig. 7), and in direction perpendicular to formation of LIPSS (scanning axis marked in blue in Fig. 7). After comparison it is clear that for longer, parallel, smoother edges and better organized LIPSS, the scanning direction should be perpendicular to the direction of LIPSS formation.

This is also proved true for irradiation of larger surfaces, as shown on fig. 8.

4.5 Polarization

Up till now in our study, all the experiments were done using linearly polarized laser beam. Employing the laser parameters we established so far, we tested the effect of polarization on the formed LIPSS. According to Almeida et al. [23], polarization determines the geometry and orientation of the produced ripples. Most researches done on LIPSS formed with circularly polarized irradiation where done on metals [24, 25], so our purpose is to study how circular polarization affects formation of LIPSS on GaN as a wide bandgap semiconductor. Only a brief review of our preliminary experiments are shown in this work, for the future there is more to be researched concerning the LIPSS formed on GaN with polarization different than linear.

To change the polarization of our initially linearly polarized beam, we added a quarter-wave plate on the optical path of the laser beam, before its entrance into the scanning head. The quarter-wave plate was placed at an angle of 45° to the optic axis, to make the polarization of the beam circular. The results obtained show that LIPSS is formed in circular polarization, but its structure is no longer linear, but rather changed into dot like, or pillar structure (Fig. 9,a). We rotated the quarter-wave plate 90° degrees additionally, so it is 135° from its axis of propagation, and we noticed that the LIPSS formed have rotated with 90° compared to then LIPSS formed at 45° (Fig. 9,b). This effect is yet to be explained and cleared, whether it is an anomaly, or the rotation of the quarter-wave plate actually affects the direction of LIPSS formed with circularly polarized laser beam.

Another curious effect observed is that, despite using the same experimental parameters proved optimal for LIPSS on GaN by linearly polarized laser beam, with circular polarization we notice that most of the area irradiated didn’t form LIPSS, which shows the parameters should be re-adjusted for the new polarization.

The multiple experiments done in the context of this study, observing the femtosecond laser interaction with WBG semiconductor crystal GaN, allow us to establish a database for the optimal combination of experimental parameters used in order to achieve highly periodic and reproducible LIPSS on GaN. These parameters were further tested and confirmed to be optimal for micromachining of large surface areas of GaN. Results recorder and discussed are summarized in the table 2 below.

The correlation between some of the parameters can be used and adapted for machining of wide range of media, as well, with minimal adjustments. This study proved yet again the efficacy and advantage of LIPSS as an easily adjusted and cost effective micro-machining technique for surface functionalization of wide range of materials, including wide bandgap semiconductors. LIPSS resulting from circularly and elliptically polarized laser beams have shown to be currently a trending topic of study, with promise for even wider range of newer applications and possibilities.

The obtained optimal value of each of the parameters studied
Wavelength λ	1030 nm
Pulse duration τ	180 fs
Frequency F	1KHz
Energy E	10 µJ
Position of laser focus relative to GaN surface	300 µm
Laser spot D	≈ 10 µm
Scanning Speed V	10 mm/s
Scanning repetition	4 times
Scanning direction in respect to LIPSS orientation	Perpendicular
Polarization	Linear

Author Contribution

M.S and D.T wrote the main manuscript. T.P provided the main idea for research. All authors worked on the experiments and contributed to the results presented and prepared in the tables. M.S, D.T and L.S prepared figures 1 to 7, T.P prepared figures 8 and 9 a&b. All authors reviewed the manuscript.

Acknowledgments

6. The research is funded by projects: Bulgarian Science Fund DN – 18/7 -10.12.2017, BG05M2OP001-1.001-0008 National center of Mechatronics and Clean Technologies Operational Program: Science and Education for Smart Growth 2014–2020 and ELI "Extreme Light" (Extreme Light Infrastructure BG) D01– 401/18.12.2020.

Data Availability

All data supporting the findings of this study are available within the paper and its Supplementary Information. All data used and generated is shown in the figures and in the tables of the manuscript.

R. W. Keyes: Proceedings of the IEEE89, 3, 227-239 (2001)
M. Zhao, C. Xu, W. H, W. Wang, L. Guo, X. Chen: Chinese Phys. Lett. 33.10, 104201 (2016)
Q. Hua, B. Ma and W. Hu: Encyclopedia of Materials: Technical Ceramics and Glasses (Elsevier 2021)
K. Czelej: Chemistry of Materials 36 (13): 6392–6409 (2024)
Y. Arakawa: IEEE Journal of Selected Topics in Quantum Electronics 8 (4): 823–832 (2002)
A. Di Carlo: Physica Status Solidi A. 183 (1): 81–85 (2001)
J. Bonse, S. Höhm, S. V. Kirner, A. Rosenfeld, J. Krüger: IEEE Journal of selected topics in quantum electronics 23(3) (2016)
X. Shi, L. Jiang, X. Li, K. Zhang, D. Yu, Y. Yu, Y. Lu: J. Appl. Phys. 116, 033104 (2014)
X. Shi, X. Xu: Applied Physics A, 125:256 (2019)
R. Miyagawa, O. Eryu: Journal of Applied Physics 58, SCCB01 (2019)
K. M. T. Ahmmed, C. Grambow, A. Kietzig: Micromachines, 5(4), 1219-1253 (2014)
M. Mastellone,, M.L Pace, M. Curcio, N. Caggiano, A. De Bonis, R. Teghil, P. Dolce, D. Mollica, S. Orlando, A. Santagata, V. Serpente, A. Bellucci, M. Girolami, R. Polini, D. M. Trucchi: Materials 15, 1378 (2022)
F. Keilmann and Y. H. Bai, Appl. Phys. A, (29), 9–18, (1982)
J. E. Sipe, J. F. Young, J. S. Preston, H.M. van Driel: Phys. Rev. B (27) 1141–1154, (1983)
G. Miyaji, K. Miyazaki, K. Zhang, T. Yoshifuji , J. Fujita: Opt. Express 20 , 14848 (2012)
T. Tomita, R. Kumai, H. Nomura, S. Matsuo, S. Hashimoto, K. Morita, T . Isu : Appl. Phys. A 105, 89 (2011)
T. J. Y. Derrien , T. E. Itina , R. Torres, T. Sarnet , M. Sentis : J. Appl. Phys . 114, 083104 (2013)
R. L. Harzic , D. Dörr , D. Sauer, F. Stracke, and H. Zimmermann, Appl. Phys . Lett. 98, 211905 (2011)
A. Borowiec and H. K. Haugen: Appl. Phys. Lett. 82, 4462 (2003)
S. Sakabe , M. Hashida , S. Tokita , S. Namba , and K. Okamuro : Phys. Rev. B 79 , 033409 (2009)
P. Gregorcic, M. Sedlacek, B. Podgornik, J. Reif: Applied Surface Science 387, 698-706 (2016)
S. Durbach, N. Hampp: Appl. Phys. Lett. 121, 251601 (2022)
M. Almeida, F. Udo, B. Voisiat, I. Tabares, F. Ränke, A. F. Lasagni: Scientific Reports 15, 20285 (2025)
S. Graf, F. A. Muller: Applied Surface Science, 331, 150-155 (2015)
S. Georges, A. Nakhoul, V. Fedorov, P. Saby, N. Faure, H. Bruhier, N. Compère, Y. Di Maio, Yoan, X. Sedao, J. and Colombier, Jean-Philippe: Nanophotonics, 14, 18, 2993-3007 (2025).

No competing interests reported.

Download PDF

Reviewers agreed at journal
29 Dec, 2025
Reviewers invited by journal
18 Dec, 2025
Editor assigned by journal
07 Dec, 2025
Submission checks completed at journal
01 Dec, 2025
First submitted to journal
29 Nov, 2025

You are reading this latest preprint version

Surface Modification of Wide Bandgap Semiconductor GaN Using Femtosecond Laser Induced Periodic Surface Structuring LIPSS

Status:

Version 1

Abstract

Figures

1. Introduction

2. Experimental Setup

3. Preliminary Studies

3.1. LIDT

3.2. Laser Focus Position

4. Results

4.1 Energy fluence

4.2 Scanning speed and overlapping

4.3 Scanning Repetition

4.4 Scanning direction

4.5 Polarization

5. Conclusion

Declarations

Author Contribution

Acknowledgments

Data Availability

References

Additional Declarations

Status:

Version 1