English

Scientists used femtosecond lasers to increase the rate of silicon ablation by 23 times

Categories:
Laser
224
1
See translation
2023-04-13

Scientists working on laser applications at the RIKEN Center for Advanced Photonics (RAP) have employed a new technique that uses gigahertz burst femtosecond laser pulses, which are packaged in a MHz envelope called BiBurst mode, to greatly increase the rate of silicon ablation without reducing the ablation quality.

 

Published in the International Journal of Extreme Manufacturing (IJEM), a team led by researchers from the Advanced Laser Machining Research Group has used BiBurst mode to successfully increase the throughput of silicon micromachining through ablation for practical applications.

The team has shown that the BiBurst mode can etch silicon at an ablative rate 23 times faster than the monopulse mode without reducing the ablative quality, while avoiding air ionization. Increasing the ablation rate is very important for increasing throughput in practical applications. Therefore, these findings have great implications not only for basic scientists but also for people in industry.

 

In previous work, the team reported that BiBurst mode femtosecond laser pulses improved the ablative efficiency and quality of crystalline silicon compared to a monopulse mode. In GHz bursts, femtosecond laser pulse sequences have extremely short pulse-to-pulse intervals of hundreds of picoseconds (ps), controlling temporal energy deposition on silicon to improve ablative efficiency and quality. The team then further explored the ability to perform high-throughput micromachining of silicon at significantly higher BiBurst pulse energies corresponding to the integrated energy of each pulse in the BiBurst pulse.

 

Importantly, when the same total laser energy is transmitted, the energy per femtosecond laser pulse (within the pulse) in the BiBurst pulse is significantly less than the pulse energy in the monopulse mode. For a monopulse mode, the ablative surface is severely damaged at strength exceeding critical values due to air ionization. In contrast, the BiBurst mode can provide higher total energy to ablate silicon without causing air ionization due to its lower in-pulse intensity.

 

Therefore, compared with the monopulse mode, the BiBurst mode achieves 23 times higher ablation rate due to the synergistic effect of higher total energy and higher ablation efficiency under the condition of avoiding air ionization. In addition, BiBurst's time control of energy deposition can maintain high ablation quality even at such high total energy.

 

The team proposes that higher ablation efficiency of GHz bursts can be achieved through collaborative contributions within successive pulses. Specifically, due to the production of free electrons, the previous intrapulse in the burst will generate transient absorption sites for subsequent intrapulse in order to improve the ablation efficiency.

 

On the other hand, ablation efficiency decreases gradually as the energy within the pulse exceeds the ablation threshold energy of the monopulse mode, since ablation can be directly induced within a single pulse. The improved ablation efficiency due to the collaborative contribution within successive pulses is no longer expected in this regime. The lower in-pulse energy in BiBurst mode not only avoids air ionization, but also maintains a higher ablation efficiency, thus achieving a higher ablation rate than a single pulse in a GHz burst or monopulse mode.

 

As a result, BiBurst mode ablation has the potential to provide higher throughput while maintaining high quality for practical applications of silicon micromachining.

 

Higher intensity femtosecond laser pulses often suffer collateral damage due to air ionization and overheating. We show for the first time that the GHz/MHz BiBurst mode of femtosecond lasers has the potential to significantly improve throughput without reducing the ablation mass. We expect that BiBurst mode femtosecond laser processing will change the common sense of femtosecond laser processing and overcome the bottleneck in industrial applications."

 

Source: Laser Net

C
    • jessie

      2023-04-14
      Thanks for sharing
        See translation
        Reply
      Collect
      Report