ABSTRACT
Attention of both producers and consumers is growing towards improved food quality and safety, which can be conveniently investigated by means of biosensors. Miniaturization of biosensors resulting in microfluidic lab-on-chips leads to an enormous reduction of analysis costs and times. When developing a new biosensor, optimization is recommended. Computational fluid dynamics (CFD) models are a valid tool to avoid expensive and time-consuming optimization processes. In this chapter, the governing equations for modeling fluid motion, mass transport and enzymatic reactions in microfluidic biosensors are presented. Three different mechanisms for fluid motion are analyzed and advantages and disadvantaged are highlighted. Classical CFD is also compared to the reduced order modeling (ROM) approach, which consists of representing complex 3D microchannel lay-outs as simplified 1D integrated circuits where every channel is replaced by its equivalent resistance. Advantages of computational modeling, and in particular of ROM, are further demonstrated through an example concerning optimization of an electrokinetically driven lab-on-a-chip for simultaneous detection of glucose, sucrose and fructose in food samples; quantification of these sugars in food products is greatly important due to their effect on taste and health. The effects of operational and geometrical parameters on the assay performances are investigated and the lab-on-a-chip is optimized to ensure the maximal amount of final product requiring the shortest assay time. Finally, future perspectives in the field of microfluidics and biosensors modeling are reported. Particular attention is paid to the interesting features of capillary flow that would allow the production of completely stand-alone biosensors. Porous, monolithic materials are the optimal candidates to achieve such a flow. Unfortunately, computational modeling of porous materials is computationally expensive and cumbersome, which is why simplified models, called pore-network (PN) models, are usually employed. Future advances in fabrication of and modeling of biosensors, as well as improvements of computational resources, are expected. This will lead to faster and cheaper biosensor designs and optimization that, together with the reduction of biosensor limitations, will help these convenient devices reach the commercialization stage.
