System-level Modeling
Electrophoresis
iMSEL has developed system-level models for analyzing the dispersion of electrophoretic transport of charged analyte molecules in a general-shaped microchannel. The models are based on the method of moments to describe analyte dispersion (including both the skew and broadening of the band) and hold for analyte bands of virtually arbitrary initial shape, and offer orders-of-magnitude improvement in computational efficiency over full numerical simulations. The model is used to perform a systematic parametric study of serpentine channels consisting of a pair of complementary turn microchannels. The results indicate that dispersion in a particular turn can contribute to either an increase or decrease of the overall band broadening. The efficiency and accuracy of the model is further demonstrated by its application to general-shaped channels that occur in practice.
Mixing
iMSEL has also developed a hierarchical model for the efficient and accurate simulations of laminar diffusion-based electrokinetic passive micromixers by representing them as a system of mixing components of relatively simple geometry. Parameterized and analytical models for constituent component are obtained, which are valid for general sample concentration profiles and arbitrary flow ratios at the component inlet. A lumped-parameter, system-level model is constructed by an appropriate set of continuous parameters at the component interface to link adjacent components. The system-level model, which simultaneously computes electric circuitry and sample concentration distributions in the entire micromixer network, agrees with numerical and experimental results, and offers orders-of-magnitude improvements in computational efficiency over full numerical simulations.
A particular sub-domain of mixing network is the concentration gradient generation that has recently gained significant attraction in cellular assay and high-through screening. iMSEL has investigated a systematic modeling methodology for microfluidic concentration gradient generators. The generator is decomposed into a system of microfluidic elements with relatively simple geometries, and parameterized models for such components are analytically derived and are applicable to concentration gradient generators that rely on either complete or partial mixing (the latter renders most electric analogy-based methods invalid). The system model is verified by numerical analysis and experimental data and accurately captures the overall effects of network topologies, component sizes, flow rates, and reservoir sample concentrations on the generation of sample concentration gradient. This modeling methodology has been used to design novel and compact microfluidic devices able to create concentration gradients of complex shapes by juxtaposing simple constituent profiles along the channel width.
Biochemical Assay
iMSEL has presented a “mixed-methodology” based system-level modeling and simulation for biochemical assays in Lab-on-a-Chip (LoC) devices. The methodology uses a combination of numerical schemes and analytical approaches to simulate biological and physicochemical processes, specifically, an integral approach for fluid flow and electric field, Method of Lines (MOL) and two-compartment models for biochemical reactions, and Fourier series-based model for analyte mixing. The solution procedure begins with decomposing the lab-on-a-chip device into a system of inter-connected components (e.g., channels and junctions) and the models are solved in a network fashion. Models are developed to accurately capture the multi-physics (e.g., flow, mixing, and reaction) behavior of individual components. The assembly of the components is facilitated via exchange of fluid flux and Fourier series coefficients (or average concentration) of analytes between various components, which enables network solution of the models. The system models are validated against both experimental and numerical models on various biochemical assays (e.g. immunoassays and enzymatic reactions), showing significant computational speedup (100 -10,000-fold depending on the assay) without appreciably compromising accuracy (
Liquid Filling
Liquid filling in microfluidic channels is a complex process that depends on a variety of geometric, operating, and material parameters such as microchannel geometry, flow velocity/pressure, liquid surface tension, and contact angle of channel surface. Accurate analysis of the filling process can provide key insights into the filling time, air bubble trapping, and dead zone formation, and help evaluate tradeoffs among the various design parameters and lead to optimal chip design. iMSEL has developed a parameterized dynamic model for the system-level analysis of liquid filling in 3D microfluidic networks based on the system decomposition approach, which tracks the liquid front in the microchannels. The principle of mass conservation at the junction is used to link the fluidic parameters in the microchannels emanating from the junction. The models are used to simulate the transient liquid filling process in a variety of microfluidic constructs and in a multiplexer, representing a complex microfluidic network. The accuracy (relative error less than 7%) and orders-of-magnitude speedup (30,000X – 4,000,000X) of our system-level models are verified by comparison against 3D high-fidelity numerical studies. Our findings clearly establish the utility of our models and simulation methodology for fast, reliable analysis of liquid filling to guide the design optimization of complex microfluidic networks.
