2Thies.pdf Computer-Aided Design for Microfluidic Chips Based on Multilayer Soft Lithography

Abstract
Microfluidic chips are emerging as a powerful platform for automating biology experiments.
In this paper, we develop design automation techniques for microfluidic chips based on multilayer soft lithography

I. INTRODUCTION
Microfluidic chips are “lab-on-a-chip” systems that can automate biology experiments by programmatically manipulating small quantities of fluids [1], [2].

Despite these advances, the design methodology for microfluidic chips relies on many manual steps and represents a serious barrier to scaling the design complexity to the
limits allowed by the underlying technology

(page 1 col 2)

Our vision is to bring the same automation and discipline to the microfluidic design process that electronic CAD brought to circuit design.

As a first step towards this vision, in this paper we address the problem of generating the control layer on a microfluidic chip.

(page 2)
1. Control Inference
2. Routing
3. GUI Generation.

Implemented as an AutoCAD plugin and freely available
online [14], the operation of Micado is illustrated by
example in Figure 2.

(page 3)
II. SPECIFYING A MICROFLUIDIC ISA
Previous research has established the notion of a microfluidic
instruction set architecture (ISA) as the primitive
set of operations supported by a microfluidic device

(page 3 col 2)
In addition to points on the actual flow layer, each chip
contains an abstract source point and sink point which should
be considered to reside off-chip.

In addition to flows, the language allows one to specify
pumping and mixing functionality

Example. Figure 2 provides an example of a flow network.

III. GENERATING CONTROL LOGIC
After defining an ISA, the next step of the microfluidic
designer is to implement a control layer that induces the
desired flows.

We describe the first method to automatically generate the
control layer, including the placement of valves, the logic of
valve operation, the sharing of control lines across valves,
and the routing of valves to external control ports.

A. Problem Definition
We assume that the source specified in the flow ISA
is pressurized, such that fluids will flow naturally in the
direction of the sink.

A formal description of the problem of generating control
logic appears in Problem Definition 1

B. Complexity of Control Minimization