On the assessment of aspheric refractive microlenses

Refractive microlenses and microlens arrays are key components for many applications such as optical data communication, laser and medical devices, or cameras. In particular, refractive micro-optics enables the miniaturization of high-tech systems but also offers novel optical functionalities. The success of this technology lies in the wafer-level fabrication technique, using the method of photoresist reflow with a subsequent pattern transfer into the substrate by reactive ion etching. Indeed, it allows the manufacturing in parallel of thousands of spherical or aspheric microlenses smaller than 1 mm. Their characterization is usually performed by measuring their surface, as this allows at the same time the evaluation of the microlens performance and feedback for fabrication process optimization. In this thesis, we assess this characterization approach to understand the fabrication process better and to improve the microlenses performance. Concretely, we first study surface form tolerancing, which is crucial to ensure the microlenses quality. However, the link between the surface form of a microlens and its performance is not straightforward, usually resulting in over-restrictive tolerances. Here, we investigate this connection for simple cases and then compare different approaches to tolerance typical micro-optical systems. Practical guidelines are proposed based on the results. Secondly, we present methods to improve surface measurements. For this, we develop an original calibration procedure that takes into account the aberrations of the imaging system. In the presented example, the accuracy is increased by a factor 7, rendering the characterization of diffraction-limited microlenses with high numerical apertures possible. Thirdly, we model the fabrication process to find correlations with the manufactured surface. Thereby, the fabrication optimization is made faster and more accurate. We validate this approach by increasing the uniformity of a large (100mm x 100mm) microlens array by a factor ~3. Finally, we evaluate another microlens characterization that consists of probing the optical functionalities in transmission and compare it to surface measurements. Particularly, we give the reasons for our doubts about its application to wafer-level microlenses. In conclusion, we show that a quantitative analysis of the microlens characterization allows for a significant improvement of the microlens quality and a better understanding of the fabrication process, resulting in lower production cost. For this reason, we believe that the results presented in this thesis will help to render wafer-level refractive micro-optics a more mature technology and build its bright future.