Fabrication processes that microelectronic developed for Integrated circuit (IC) technologies for decades, do not meet the new emerging structuration requirements, in particular for non-IC related technologies ones, such as MEMS/NEMS, Micro-Fluidics, photovoltaics, micro lenses manufacturing. Actually complex 3D structures (pattern with several levels) requires either altering several steps of layering and etching or the use of complex lithography patterning approaches such as gray-scale electron beam lithography, laser ablation, focused ion beam lithography, two photon polymerization. It is now challenging to find cheap and easy technique to achieve 3D structures.

In this work, we propose a straightforward process to realize 3D structuration, intended for silicon based materials (Si, SiN, SiOCH …). This structuration technique is based on nano-imprint lithography (NIL), ion implantation and selective wet etching.

In this paper, we will present 2D and 3D structures obtained on silicon. The process scheme is described in figure 1, 2. In a first step a pattern is performed by nanoimprint lithography on a substrate (Figure 1a, 2a). Then ion implantation is realized through the (2D/3D) resist mask in order to create localized modifications in the material and transfer the pattern. The subjacent layer is protected partially or completely by the resist according to different levels of the pattern, so the ions are stopped differently, which makes the transfer possible. Finally, after the resist stripping (Figure 1b, 2b), a selective wet etching is carried out to remove selectively the modified material regarding the non-modified one (Figure 1c, 2c). Many chemistries have been tested, the most efficient and selective one were silicon anodization in acid bath.

The achieved selectivity comes from modifications induced into silicon during ion implantation. By using heavy ions, the material is amorphised, which has been confirmed by Raman spectroscopy. Despite this morphological defects, Scanning Spreading Resistance Microscopy (SSRM) shows that implanted areas has a lower resistivity than non-modified ones, which can be explained by the appearance of states in the forbidden gap of Si, generated by induced defects. Photoluminescence spectroscopy has been performed and bears out this idea. We also noticed, thanks to SSRM, a depletion zone in the interface between implanted and non-implanted Silicon which may guide electric field during the electrochemical process. All these elements makes the dissolution of implanted silicon more selective regarding non implanted one.

The 3D structuring process that we are proposing leans on several parameters, some are related to dissolution mechanism like illumination, current and electrolyte composition, others concern ion implantation process such as type of ion, energy and dose, and finally lithography parameters for instance type and thickness of the resist. We will show that a key combination between all this parameters is required to achieve pattern transfer.