Single-Layered Microfluidic Network-Based Combinatorial Dilution for Standard Simplex Lattice Design.

Kangsun Lee,Kwang Oh,Choong Kim

doi:10.3390/mi9100489

Abstract

In this paper, we presented a straightforward strategy to generate 15 combinations of three samples based on an experimental simplex lattice design using a single-layer microfluidic network. First, we investigated the performances of the plain structural and the groove structural combinatorial devices by computational simulation (CFD-ACE+). The simulated output concentrations were extremely close to the desirable values within an absolute error of less than 1%. Based on the simulated designs, polydimethylsiloxane (PDMS) devices were fabricated with soft lithography and tested with fluorescent dye (sodium salt). The mixing results for 15 combinations showed good performance, with an absolute error of less than 4%. We also investigated two liquid handling methods (bottom–up and top–down) for high-throughput screening and assay. The liquid-handling methods were successfully accomplished by adding the systematic structured groove sets on the mixing channels.

Highlights

One-factor-at-a-time experiments are inefficient when compared with experiments in which factors are changed simultaneously and systematically
The design of experiments (DOE) methodology has recently been adopted by scientists and engineers who work in such domains as medical devices, pharmaceuticals, foods, materials science, and semiconductors [4,5,6,7,8]
We have demonstrated a very simple continuous-flow microfluidic device that can generate 15 combinatorial mixture outputs with the relative concentrations of three-sample inputs (A, B, and C), representing a combinatorial mixture DOE configuration of a simplex design

Summary

Introduction

One-factor-at-a-time experiments are inefficient when compared with experiments in which factors are changed simultaneously and systematically. The incorporation of sophisticated pneumatic microfluidic design and fabrication technology is mandatory in order to produce a large number of combinatorial mixtures or dilutions This often requires significant expertise that is not part of life-science training [14]. The use of continuous-flow microfluidic networks is appropriate for generating simple binary-based combinatorial mixtures because it has the prerequisite of much complicated multilayer design and fabrication technology, it has limited ability to produce complex combinatorial dilutions with multiple samples at multiple concentrations. We have demonstrated a very simple continuous-flow microfluidic device that can generate 15 combinatorial mixture outputs with the relative concentrations of three-sample inputs (A, B, and C), representing a combinatorial mixture DOE configuration of a simplex design.

Microfabrication and Fluorescence Measurement

Computational Modeling of Concentration Profiles

Results and Discussion

Fluorescent Experimental Results