EXSEED researches primarily relied on the studying of femtosecond laser interactions at non-conventional driving wavelengths.
The range of nonlinear responses accessible by radiation tuning allowed revisiting questions as important as the achievable precision in laser surface machining technologies. In particular, we have established that the concept of nonlinear resolution is not applicable for femtosecond laser ablation [Optics Letters 45, 952-955 (2020)]. This is because any observable based on a threshold-based response simply ruins all potential benefits that could be expected on resolution from the nonlinear confinement of absorption. Another important consequence is that the achievable precision and repeatability can be directly derived from the level of determinism of the interaction. By comparing the results of a simple 'noise' model accounting for laser fluctuations, we have quantified the precision limits [Applied Physics Letters 117, 171604. (2020)] proposed an innovation to improve processing performances [Optics Letters 47 (2022) 993].
At the opposite side of the spectrum, ultrashort infrared laser pulses opened a way for internal structuring of semiconductor materials that are extremely challenging to process in the three dimensions (3D writing). Our first proposed solution used hyper-focused beams to demonstrate microplasmas up to permanent material modifications in the bulk of silicon with sub-100-fs pulses [Nature Communications 8, 773 (2017).]. For more practical alternatives, we worked on optimizations in the time domain. We investigated the picosecond regime limiting the nonlinearities and improving the process reliability [Physical Review Applied 12, 024009 (2019); Optics Express 28, 26623 (2020)]. We revealed that an even more promising approach is to rely on transient accumulation strategies. To this aim, we generated and applied ultrafast trains of pulses at up to Terahertz repetition-rates [Research 8149764 (2020), International Journal of Extreme Manufacturing 4 (2022) 045001]. An important aspect also addressed by our experiments is the unusually high sensitivity of 3D writing in semiconductors to the temporal contrast of the pulses [Physical Review Research 2, 033023 (2020)]. This causes laser-technology dependent results and represents an important finding for a comprehensive reading of the literature on this topic [Laser and Photonics Reviews (2021) 2100140]. Taken these approaches together, we recently introduced unique multi-timescale control parameters holding promises for future exploitation in the semiconductor industry. As an illustration of innovative process solutions deriving from this knowledge, we can note the first demonstration a transmission laser welding method applicable to semiconductors [Laser and Photonics Reviews 16 (2022) 2200208]. Among the general performance advances, this work culminates with a method for enhanced precision (incl. resolution and material change controllability) writing deep inside silicon chips. This holds promises for commercial exploitations and is today the subject of a patent application [EPEP21184898.1 R37868EP, PCT/EP2022/068835].
