Deterministic lateral displacement (DLD) has been extensively implemented in the last decade for size-based sample preparation, owing to its high separation performances for a wide range of particle dimensions. However, separating particles from 1 μm to 10 μm in one single DLD device is challenging because of the required diversity of pillar dimensions and inherent fabrication issues. This paper presents an alternative approach to achieve the extraction of E. coli bacteria from blood samples spiked with prostate cancer cells. Our approach consists in cascading individual DLD devices in a single automated platform, using flexible chambers that successively collect and inject the sample between each DLD stage without any external sample manipulation. Operating DLD separations independently enables to maximize the sorting efficiency at each step, without any disturbance from downstream stages. The proposed two-step automated protocol is applied to the separation of three types of components (bacteria, blood particles and cancer cells), with a depletion yield of 100% for cancer cells and 93% for red blood cells. This cascaded approach is presented for the first time with two DLD modules and is upscalable to improve the dynamic range of currently available DLD devices.

Materials selected for microfluidic technology exhibit mechanical properties that can be a source of innovation. For instance, devices that take advantage of rigid (glass, silicon) or soft (elastomer, PDMS) materials, as well as porous materials, such as paper, are widely reported in the literature. In this paper, we illustrate the potentialities of hyper elastic materials for lab-on-chip developments. Two breakthrough examples are reported: (i) a new digital microfluidics approach based on a stretchable membrane for addressing a large range of liquid volumes in complex protocols and (ii) a new low-cost approach for prototyping fully deformable microfluidic devices based on a polymeric foam.

No abstract