Search

International audience; Ultrafast laser welding is an advanced manufacturing technology capable of bonding materials by selectively irradiating the interface between them. In this process, the beam is focused at the interface through the transparent top material, resulting in melting and subsequent bonding after resolidification (Fig. 1a). This has allowed to weld several materials in different configurations including through-glass, polymers or ceramics. However, it is striking to note that this process remains difficult to apply on silicon (Si) or other semiconductors, which would most certainly benefit the microelectronics industry.The intrinsic properties of Si (high refractive index and narrow bandgap) impede producing internal modifications by common laser processing configurations since nonlinear propagation effects and prefocal plasma avoid reaching high enough intensities at the focal spot to induce modification. To circumvent these effects, recently the first demonstration of laser welding of silicon on copper was made with picosecond pulses, obtaining relatively modest ~2 MPa bonding strengths [1]. After examining pulse durations from 200 fs up to ~20 ps, we confirm important physical limitations with intense pulses. In view of this, and inspired by stealth dicing in which the thermal properties of nanosecond pulses are used to create below-surface modification for subsequent dicing, we propose to use a similar approach to melt the interface of two Si pieces and join them.In this work, by disruptively moving to longer pulses of 5 ns at 1550 nm, we present the first demonstration of laser welding of Si. First, by single-spot irradiations (Fig. 1b), we identify that small imperfections in the contact constitute a Fabry-Perot cavity that leads to no modification of the bottom piece. Relying on regions in optical contact, we later vary the irradiations conditions (focus position and scan speed) to determine the most material mixture between wafers without the appearance of detrimental cracks, cavities or porosities. We then raster-scan the optimized beam to produce a large welded area (Fig. 1c). By shear force measurements, we demonstrate a joining strength of 32±10 MPa for the Si-Si configuration. Later, we extend this optimization work to gallium arsenide (GaAs) and test different configurations alongside with Si finding shear joining strengths >10 MPa in all processed cases [2].To our knowledge, this provides not only the first demonstration of laser welding of similar semiconductors but also dissimilar ones (Si and GaAs). The bonding strengths are similar to those obtained by molecular wafer bonding and ultrafast laser welding of dielectrics, leading to unprecedented process efficiency, given the use of a very compact fiber laser source. To illustrate this, we produced a few millimeters Si-GaAs weld (3.5×7 mm2), allowing to hang the employed laser equipment (Fig. 1d). We anticipate that the process can be advantageously scaled up to large wafers with the use of industrial high-power sources, holding great potential for a unique additive manufacturing solution applicable to semiconductors.

National audience

International audience; An important challenge in the field of three-dimensional (3D) ultrafast laser processing is to achieve permanent modifications in the bulk of silicon (Si) and narrow-gap materials. High-energy infrared femtosecond lasers pulses fail when conventional laser machining configurations are used. By comparisons between ultrafast plasma images, 3D energy density maps inside the samples and nonlinear propagation simulations, we show the strong nonlinear and plasma effects causing a strict clamping of the intensity that can be delivered inside Si with tightly focused infrared pulses. To circumvent this limitation, we will describe a solution inspired by solid-immersion microscopy to achieve hyper-focusing of the pulses. From a proof-of-concept experiment, we demonstrate the first micro-modifications achieved inside Si with sub-100fs pulses. For more practical alternatives, we will describe today’s effort on optimizations in the time domain. We will discuss the picosecond regime limiting the detrimental nonlinearities and provoking progressive thermal band gap closure to assist energy deposition. We also perform multi-pulse irradiation experiments where femtosecond, picosecond and nanosecond pulses are synchronized. This identifies breakdown channels seeded by pre-ionization or local thermal pre-stimulation. Another approach is to rely on accumulation strategies. To this aim, we rely on ultrafast trains of pulses at the highest achievable repetition-rates (up to THz). From these approaches, we will introduce multi-timescale control parameters exploited for improved energy deposition and for demonstrations of reliable 3D laser writing deep inside silicon chips that would not be possible otherwise.

International audience

International audience; Volume laser inscription (VLI) inside materials have been widely investigated because of its unique possibility to create precise and functional 3D microstructures under the surface. Recent years witness an increasing interest of shifting similar applications from dielectric materials to more challenging semiconductors that have different material properties (Fig. 1a). In particular, much stronger nonlinear propagation effects are expected that can strongly distort the laser focus conditions when working deep inside the materials (Fig. 1b) [1]. Recently, tremendous efforts have concentrated on silicon (Si), the basis material for electronics industry. However, there are so far very few works on the specificity of other important semiconductors, especially Gallium Arsenide (GaAs) which also has important applications in electronics sector such as integrated circuits, light-emitting diodes and photovoltaic cells. GaAs has very different properties compared to Si. GaAs exhibit a direct bandgap of 1.44 eV whereas Si has an indirect bandgap of 1.12 eV. The nonlinear refractive index of GaAs is also more than 5 times higher than the one for Si. These and other differences must lead to significant differences for the VLI conditions.

International audience; In this study, we investigate the conversion of femtosecond laser energy deposition from plasma into a shockwave in ambient air. The experiments are carried out using a 380fs pulsed laser at 1.03μm, with laser intensities below the filamentation threshold. The measurements of this dynamic phenomenon are carried out with the help of a time-resolved transmission microscope, and the pressure and temperature space-time evolution are evaluated using a theoretical model. In our conditions we generate shockwaves with initial pressure loading in the range of GPa and maximum propagation velocity in the order of a few km/s.

Supplementary material of experimental setup, Liu's method, beam characterization, statistics, and process stability

International audience

International audience

International audience

International audience

Direct laser writing of amorphous lines in crystalline silicon has the potential for becoming a flexible alternative to silicon-on-insulator technology for photonic integrated circuits. Yet, the maximum amorphous layer thickness achieved is 60 nm, which is below the requirements for waveguiding at telecom wavelengths. Here, the authors report on different strategies to push the layer thickness beyond today's limit. To this end, irradiation with femtosecond laser pulses covering an extremely broad wavelength range (515 nm–4 µm) up to the yet unexplored near- and mid-infrared region of silicon transparency is investigated. The results show that much thicker amorphous layers can be obtained upon multipulse irradiation at 3-µm wavelength. The deepest amorphization is achieved in silicon wafers covered with a thick silicon dioxide layer that strongly assists the heat extraction, yielding steep index profiles with a maximum amorphous layer thickness of 128 nm. This superior thickness is compatible with single mode waveguiding for a symmetric waveguide configuration. This study also contributes to a better understanding of the mechanisms involved in laser-induced amorphization., This work was partly funded by Laserlab-Europe (Grant Agreement No. 284464, EU’s Seventh Framework Programme), Project No. CNRS-LP3002581. J.S. acknowledges financial support through the national research grant UDiSON (TEC2017-82464-R) from the Spanish Research Agency (AEI, Ministry of Research and Innovation) and the European Regional Development Fund (ERDF), as well as the Consejo Superior de Investigaciones Científicas for the intramurales project (201850E057). N.C. acknowledges a predoctoral fellowship from the MICINN assigned to project UDiSON. D.G. acknowledges financial support from the European Research Council (ERC), H2020 program (Grant Agreement No. 724480).

International audience

International audience; Laser-based nanofabrication plays today an important role in the most advanced technology developments. The necessity for increased versatility and precision has led to the development of the approach so-called Double-Pulse Laser Induced Forward Transfer (DP-LIFT) [1]. DP LIFT is a potential solution to print 2D-3D nanostructures with increased resolution [1,2].

Laser direct writing is a widely employed technique for 3D, contactless, and fast functionalization of dielectrics. Its success mainly originates from the utilization of ultrashort laser pulses, offering an incomparable degree of control on the produced material modifications. However, challenges remain for devising an equivalent technique in crystalline silicon which is the backbone material of the semiconductor industry. The physical mechanisms inhibiting sufficient energy deposition inside silicon with femtosecond laser pulses are reviewed in this article as well as the strategies established so far for bypassing these limitations. These solutions consisting of employing longer pulses (in the picosecond and nanosecond regime), femtosecond-pulse trains, and surface-seeded bulk modifications have allowed addressing numerous applications.

International audience; We generate and apply ultrafast trains of infrared pulses with durations ranging from 200 fs to 3.5 ns. We show how this gives unique multi-timescale control parameters for improved energy deposition and reliable 3D laser writing deep inside silicon.