Bertrand Selva

An optimized resistor pattern for temperature gradient control in microfluidics

In this paper, we demonstrate the possibility of generating high-temperature gradients with a linear temperature profile when heating is provided in situ. Thanks to improved optimization algorithms, the shape of resistors, which constitute the heating source, is optimized by applying the genetic algorithm NSGA-II (acronym for the non-dominated sorting genetic algorithm) (Deb et al 2002 IEEE Trans. Evol. Comput. 6 2). Experimental validation of the linear temperature profile within the cavity is carried out using a thermally sensitive fluorophore, called Rhodamine B (Ross et al 2001 Anal. Chem. 73 4117–23, Erickson et al 2003 Lab Chip 3 141–9). The high level of agreement obtained between experimental and numerical results serves to validate the accuracy of this method for generating highly controlled temperature profiles. In the field of actuation, such a device is of potential interest since it allows for controlling bubbles or droplets moving by means of thermocapillary effects (Baroud et al 2007 Phys. Rev. E 75 046302). Digital microfluidics is a critical area in the field of microfluidics (Dreyfus et al 2003 Phys. Rev. Lett. 90 14) as well as in the so-called lab-on-a-chip technology. Through an example, the large application potential of such a technique is demonstrated, which entails handling a single bubble driven along a cavity using simple and tunable embedded resistors.

Temperature-induced migration of a bubble in a soft microcavity

We perform studies of pancake-like shaped bubbles submitted to a temperature gradient in a micrometric height Hele-Shaw cell. We show that under the experimental conditions, usually found in microfluidic devices, the temperature-induced dilation of the cavity overcomes the thermocapillary convection due to surface tension variation, effectively driving the bubble toward the cold side of the cavity. The bubble velocity is experimentally characterized as a function of the bubble radius, the temperature gradient, and the initial Hele-Shaw cell thickness. We propose a theoretical prediction of the bubble velocity, based on the analytical resolution of the hydrodynamical problem. The equations set closure is ensured by the pressure value near the bubble and by the dissipation in the moving meniscus.