# Sub-surface modifications in silicon with ultrashort pulsed lasers above 2 microns

## TL;DR

This study investigates how ultrashort pulsed lasers with wavelengths above 2 microns induce sub-surface modifications in silicon, revealing optimal wavelengths around 2000-2200 nm due to nonlinear optical effects.

## Contribution

It provides experimental and numerical analysis of silicon modifications at wavelengths > 2 μm, highlighting the influence of nonlinear effects on laser processing.

## Key findings

- Wavelengths 2000-2200 nm are optimal for sub-surface modifications.
- Nonlinear Kerr-effect enhances self-focusing at ~2100 nm.
- Wavelength dependence differs from that at 1550 nm.

## Abstract

Nonlinear optical phenomena in silicon such as self-focusing and multi-photon absorption are strongly dependent on the wavelength, energy and duration of the exciting pulse. Thus, a pronounced wavelength dependence of the sub-surface modifications with ultra-short pulsed lasers exists, especially for wavelengths > 2 $\mu$m. This wavelength dependence is investigated for wavelengths in the range of 1950-2400 nm, at a pulse duration between 0.5-10 ps and the pulse energy varying from 1 $\mu$J to 1 mJ. Numerical and experimental analyses have been performed on both the surface and sub-surface of Si wafers processed with fibre-based lasers built in-house that operate in this wavelength range. The results have been compared to the literature data at 1550 nm. The analysis carried out has shown that due to a dip in the nonlinear absorption spectrum and a peak in the spectrum of the third-order non-linearity, the wavelengths between 2000 - 2200 nm are more favourable for creating sub-surface modifications in silicon. This is the case even though those wavelengths do not allow as tight a focusing as those at 1550 nm in the linear regime. This problem is compensated by an increased self-focusing due to the nonlinear Kerr-effect around 2100 nm at high light intensities, characteristic for ultra-short pulses.

---
Source: https://tomesphere.com/paper/1907.13186