An effective light trapping system is required in silicon solar cells in order to collect a large amount of photons. That is why we focus our investigation on the fabrication and evaluation of two types of optical systems introduced on the back side of solar cells. The aim of these structures is to enhance the light trapping of the long wavelength photons (above 1000 nm). On the one hand, we evaluated a Si/SiO2 linear nanograting; on the other hand, hexagonal nanostructures fabricated with SiO2 nanoparticles and a filling matrix are under investigation. In this paper, we describe the fabrication processes developed for both approaches and we present the solar cell results and characterisation. For the first approach, we show a reflectance reduction on test structures, which occurs at the same wavelength as the increase of absorption induced by the simulated gratings. Moreover, we demonstrate the feasibility of the fabrication of silicon solar cells with the hexagonal nanostructures as a diffractive back reflector. Although no short circuit current increase has been observed due to a poor rear side passivation, a current gain up to 0.3 mA/cm2 is possible in the wavelength range of 1050-1150 nm due to these nanostructures. Finally, we also comment on the advantages and drawbacks of each approach and on the feasibility to introduce these systems in the solar cell process flow

The highest efficiency silicon solar cells are fabricated using defined texturing schemes by applying etching masks. However, for an industrial production of solar cells the usage of photolithographic processes to pattern these etching masks is too consumptive. Especially for multicrystalline silicon, there is a huge difference in the quality of the texture realized in high efficiency laboratory scale and maskless industrial scale fabrication. In this work we are describing the topography of a desired texture for solar cell front surfaces. We are investigating UV-nanoimprint lithography (UV-NIL) as a potential technology to substitute photolithography and so to enable the benefits resulting of a defined texture in industrially feasible processes. Besides the reduced process complexity, UV-NIL offers new possibilities in terms of structure shape and resolution of the generated etching mask. As mastering technology for the stamps we need in the UV-NIL, interference lithography is used. The UV-NIL process is conducted using flexible UV-transparent stamps to allow a full wafer process. The following texturisation process is realized via crystal orientation independent plasma etching to tap the full potential of the presented process chain especially for multicrystalline silicon. The textured surfaces are characerised optically using fourier spectroscopy

Predicting clinical response to anticancer drugs remains a major challenge in cancer therapy research. The current treatment of patients with ovarian cancer involves surgery and platinum-based chemotherapy, which still result in recurrence on advanced stages of the disease. During tumor progression, cancerous cells continuously accumulate mutations giving rise to a heterogeneous population of cells. Additionally, spatial distribution of stromal, endothelial and immune cells in the microenvironment affects tumor progression, cell morphology, and ultimately drug response. In order to predict clinical outcome of chemotherapy in vitro, tumor heterogeneity and microenvironment constituents must be conserved. Current models of tumor biology and microenvironment consist in xenografts of human tumors implanted in immunodeficient mice. These models allow studying systemic treatment response, but its application is limited to small developmental studies. Here, we present a high throughput in vitro 'grafting' platform where we co-culture blood vessels with tumor explants. Each unit in this platform is composed of two parallel microfluidic channels and a central chamber. Two endothelial tubules are generated in the microfluidic channels and cultured in presence of a gradient of angiogenic factors (S1P, VEGF, bFGF and PMA) added to the central chamber of the culture unit. Angiogenic tubules form vascular beds within 3-5 days, after which tumor explants are loaded to the central chamber on top of the vascular beds. The model shown in this study consists of ovarian serous papillary adenocarcinoma collected after xenograft growth in immunodeficient mice. This ovarian cancer explant was previously characterized as resistant to Paclitaxel. In this study, we observe how the vascular bed remodels in the presence of the explant and closely interacts with ovarian tumor tissue. Vessel perfusion and stabilization of vascular bed was monitored by real time imaging of 150 kDa FITC-Dextran. Cultures were evaluated by assessment of morphology and presence of endothelial and tumor cell biomarkers. Moreover, co-culture response to Sorafenib (anti-angiogenic), Palbociclib and Paclitaxel was detected by distinctive proliferation rates as compared to control conditions. The established ovary cancer-on-a-chip platform enables the study of fundamental aspects of tumor disease and progression. In addition, these co-cultures serve as a platform for understanding tumor-endothelial cell crosstalk and its consequences for tumor aggressiveness. Moreover, these models constitute a suitable platform for drug screenings of anti-cancer and anti-angiogenic compounds, making them a powerful translational tool for drug selection in personalized medicine applications. Citation Format: Silvia Bonilla, Jean-François Mirjolet, Pauline Berger, Elodie Rajon, Erik Walinga, Remko van Vught, Henriëtte Lanz, Jos Joore, Paul Vulto, Fabrice Viviani, Karla Queiroz. Ex vivo modelling of vascularized pdx-derived ovarian tumors [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 2795.