No abstract

A microfluidic diffusion diluter was used to create a stable concentration gradient for dose response studies. The microfluidic diffusion diluter used in this study consisted of 128 culture chambers on each side of the main fluidic channel. A calibration method was used to find unknown concentrations with 12% error. Flow rate dependent studies showed that changing the flow rates generated different gradient patterns. Mathematical simulations using COMSOL Multi-physics were performed to validate the experimental data. The experimental data obtained for the flow rate studies agreed with the simulation results. Cells could be loaded into culture chambers using vacuum actuation and cultured for long times under low shear stress. Decreasing the size of the culture chambers resulted in faster gradient formation (20 min). Mass transport into the side channels of the microfluidic diffusion diluter used in this study is an important factor in creating the gradient using diffusional mixing as a function of the distance. To demonstrate the device's utility, an H2O2 gradient was generated while culturing Ramos cells. Cell viability was assayed in the 256 culture chambers, each at a discrete H2O2 concentration. As expected, the cell viability for the high concentration side channels increased (by injecting H2O2) whereas the cell viability in the low concentration side channels decreased along the chip due to diffusional mixing as a function of distance. COMSOL simulations were used to identify the effective concentration of H2O2 for cell viability in each side chamber at 45 min. The gradient effects were confirmed using traditional H2O2 culture experiments. Viability of cells in the microfluidic device under gradient conditions showed a linear relationship with the viability of the traditional culture experiment. Development of the microfluidic device used in this study could be used to study hundreds of concentrations of a compound in a single experiment.

This paper presents a numerical steady-state model of ion transport in micro-and nanofluidic devices with widely varying geometric scale, such as transitions between micro-and nanochannels. Finite element or finite volume simulation of such problems is challenging, due to the number of elements needed to produce a satisfactory mesh. Here, only the lengthwise channel dimension is meshed; standard analytical approximations are used to incorporate cross-channel properties. Singularly perturbed cases are built up by continuation. The method is shown to reproduce our previously reported measurements of electric double-layer effects on conductivity, ion concentration, and ion enhancement and depletion. Comparison with 2-D simulations reported in the literature shows that effects on accuracy due to the 1-D approximation are small. The model incorporates analytical models of surface charge density taken from the literature. This enables predictive simulation with reasonable accuracy using published parameter values, or these values may be tuned based on experiment to give improved results. Use of the model for iterative design and parameter estimation is demonstrated.