Offline Washing Schemes for Residue Removal in Digital Microfluidic Biochips

Abstract

A digital microfluidic biochip (DMB) is often deployed for multiplexing several assays in space and in time. The residue left by one assay may contaminate the droplets used for subsequent assays. Biochemical assays involving cell culture and those based on particle microfluidics also require sweeping of residual media from an active droplet on-chip. Thus, fluidic operations such as washing or residue removal need to be performed routinely either to clean contamination from the droplet pathways or to rinse off certain droplets on the chip. In this work, several graph-based techniques are presented for offline washing of biochips that may have either a regular geometry (e.g., a 2D array of electrodes), or an irregular geometry (e.g., an application-specific layout). The schemes can be used for total washing, that is, for cleaning the entire biochip or for selective washing of sites or pathways located sparsely on the chip. The problem of reducing the path length and washing time of the droplets is investigated with or without capacity constraints. The proposed algorithms for offline washing make use of several techniques such as graph traversal, integer linear programming (ILP) modeling, and customized heuristics based on the nature of the geometric distribution of the contamination profile. The contaminated pathways are assumed to be Manhattan or curved, and hence the techniques are applicable to the conventional field-actuated DMBs as well as to the emerging classes of light-actuated and active-matrix DMBs. These techniques will be useful in enhancing the reliability of a wide class of emerging digital microfluidic healthcare devices