See Comments Next Page - GWR

See comments next page - GWR

Multilayer Microfluidics

______

Department of Materials Science and Engineering

University of Maryland, College Park

ENMA490

Fall 2003

Susan Beatty, Charles Brooks, Shawna Dean, Mark Hanna, Dan Janiak, Chen Kung, Jia Ni, Bryan Sadowski, Anne Samuel, Kunal Thaker

Special Thanks to Dr. Gary Rubloff and Theresa Valentine

INSTRUCTOR’S COMMENTS ON FINAL REPORT:

Please see the comments I have made throughout the report – I think there are important lessons to learn from them.

Introduction: This section is weak in not motivating the project very well, and in not putting enough materials information into the project. As I have reminded you before, this is a materials course despite the fact that we have concentrated a lot more on how to integrate materials into design of products and systems. I don’t think this section was critically reviewed and revised, at least not enough.

Device Design Stages:These sections are generally very good, nicely distinguishing motivation for each design, masking levels, test sites, process flow, and experiments. Constituting the bulk of the report, they reflect good progress in the project. The primary weakness in these parts are a number of places where the writing is unclear or confusing, a problem I attribute to inadequate proofreading and refinement of the report.

Modeling: This is good work and important for the project. However, the variables and units for the equations were not adequately explained, and references to where the equations came from (since most were not derived here) are missing. This section would also benefit from some brief descriptions of the meaning of the equations.

Alternative valve designs: This section is interesting, and you came up with some interesting ideas. In a sense it was too bad we didn’t have time to look at these in more depth.

Preferred design elements: This section is a little short but reasonable, although the perspective is not all that critical an analysis.

Conclusion: This section is weak. If the section on Preferred Design Elements had been more critically evaluated, then the broader perspective from it could have been put into the conclusion, or the two combined altogether.

References: The handling of referencing in your work is a SERIOUS shortcoming. At first there were basically no references (one web site). Of course this is unacceptable for any report or paper. I basically returned the report to you, and now there are 16 references, of which 4 are web sites and 2 are magazines (Industrial Physicist). This might be considered acceptable, but mediocre, particularly since I know you found and read numerous other references. However, many places in the report should have noted references but did not (see my detailed comments), so that the referencing is at best a patch job. I believe this flaw is an unanticipated consequence of inadequate planning for completion of the report. But I must admit to my disappointment after my efforts to give guidelines about referencing on the class web site at

and the time we spent a the beginning of the course discussing this. This is an issue of professionalism and engineering ethics.

Writing style: You will notice some places where imprecision in wording either makes the discussion seriously unclear, or even wrong. When it comes to explaining research results, you must try to be precise in both content and writing. The mindset should be to explain things to a novice, and thereby to consciously avoid imprecision. Furthermore, with a team project, at least a second person should be charged to read and revise the first draft of each portion, so that a cross-check on clarity is done. Virtually all writing has this activity somewhere along the line.

Instructor’s summary:

The accomplishments of this project are very good. You investigated many dimensions of a real materials-based technology challenge, pursued experimental and modeling work, analyzed designs and process flows, and most of you contributed actively in class. All this made a stimulating experience for me, and I think for you.

Unfortunately, the quality of the final report falls notably short of reflecting the quality of the work you did. Integration of the report - particularly proofreading with a keen eye to consistency, clarity, and logic – was inadequate. The absence of referencing was a major shortcoming, which should have been easily avoided with proper planning.

We emphasized from the beginning that there is a firm end date, so all work has to be planned and scheduled to make those deadlines. This planning was the responsibility of the class, particularly the development team, and if this had been done correctly the proofreading and referencing would have been far better. I might have required the report earlier so that there would be time for iteration and repair, but I considered this part of your responsibility as a project team.

During the course, and despite my attempts to highlight the issues, I sensed several times that the importance of project planning and the role of some kind of development team were not understood or appreciated. I believe this impression was validating by the referencing problem, the inadequate attention to proofreading and consistency in the report, and the anxiety accompanying the last-minute practice just before the final presentation. I hope your learning from this course includes an appreciation of the importance for each team member to make sure that he/she understands the overall project and how the pieces fit together (even if not in great technical detail), to help keep the project moving toward the intended goal along an appropriate timeline, and to meet the deadline with a quality product.

Table of Contents

Introduction…………………………………………………………………………………………………...3

Problem Definition……………………………………………………………………………………..3

Problem Scope…………………………………………………………………………………………3

Initial Materials Information…………………………………………………………………………...3

Initial Literature Research Results……………………………………………………………………..3

Device Design Overview………………………………………………………………………………4

Device Design Stage 1: Initial Microchannel Test Design……………………………………………….....4

Device Objective……………………………………………………………………………………….4

Device Logic…………………………………………………………………………………………...4

Device Dimensions…………………………………………………………………………………….5

Materials………………………………………………………………………………………………..6

Processing method with Mask Design………………………………………………………………....6

Stage 1 Summary……………………………………………………………………………………....8

Device Design Stage 2: Modified Microchannel Test Design………………………………………………8

Device Objective……………………………………………………………………………………….8

Device Logic…………………………………………………….…………………………………...... 9

Device Dimensions…………………………………………………….………..……………………10

Materials…………………………………………………….……………………………...…………10

Processing Method with Mask Design………………………………………………………………..10

Fabrication Step: SU-8 Molds…………………………………………………….…………………..13

Fabrication Steps: PDMS Microchannels…………………………………………………………….13

Experimental Trials…………………………………………………….………………………..……14

Future Work …………………………………………………….……………………………………16 Stage 2 Summary…………………………………………………….……………………………….16

Device Design Stage 3: Pressure Actuated Valve Test Design……………………………………………16

Objective……………………………………………………………………………………………...16

Device Logic…………………………………………………….……………………………………17

Device Dimensions…………………………………………………….…………………………..…17

Materials…………………………………………………….………………………………………...18

Design Problems…………………………………………………….………………………………..18

Processing Method with Mask Design………………………………………………………………..18

Additional Issues…………………………………………………….………………………………..19

Stage 3 Summary…………………………………………………….……………………………….19

Membrane Deflection Modeling…………………………………………………….……………………...20

Fluid Flow Modeling…………………………………………………….…………………………………..21

Alternative Actuated Valve Designs…………………………………………………….…………………22

Piezoelectric Valves………………………………………………………………………………..…22

Electrochemical Valves……………………………………………………………………………….23

Thermally Activated Valves…………………………………………………….……………………24

Preferred Design Elements…………………………………………………….……………………………25 Channels…………………………………………………….………………………………………...25

Valves…………………………………………………….…………………………………………...25

Scaling…………………………………………………….………………………………………...... 25

Conclusion…………………………………………………….……………………………………………...26

Appendix…………………………………………………….……………………………………………….26

Gantt Chart…………………………………………………….…………………………………...…26

Fluid Flow Modeling………………………………………….…………………………………...…26

References…………………………………………………………………………………………………..27

Introduction

Problem Definition

To use micro processing techniques to address the problems associated with multi-level channel routing in bio-micro fluidic applications [1]

To integrate materials application for building the layers of a multilevel micro fluidics system

To use a control system to arrange fluidic flow through the multilevel micro fluidics

[GWR1]

Problem Scope

The mission of this project is[GWR2] to create a multi-level micro-fluidics system for bio-micro fluidic application. [GWR3]

[GWR4]

The packaging of this device should be efficient, feasible and versatile because we would want the fluid flow to reach multi-levels instead of remaining on a single layer. Active control devices will control the fluid flow. To flow from one layer to another layer we would have vertical vias or interconnects from the first layer to the next. Therefore to process this we would need the basic knowledge of materials that are feasible and current research accomplished on micro fluidics. These are mentioned in the Materials Information and Literature Research section.

Due to time budget our group decided[GWR5] to neglect the biochemistry interactions of the fluid and the interior walls of the channels. We will[GWR6] only consider the fluid flow and how to transport the fluid from one reservoir to another within the system.[GWR7] We will be looking at many control systems that will manage the fluid flow throughout the channels and interconnects. All the control systems we will be discussing will be internally integrated within the micro fluidics system.[GWR8] The design of the control system will be discussed more thoroughly in stage 3 of the Devices Design Stages. Therefore the biochemistry interactions will not be discussed in our report due to time constraints, and we will not use external control systems.[GWR9]

Initial Materials Information

Materials considered for our micro-fluidic design consisted of Pyrex and silicon substrates with polydimethylsiloxane (PDMS), (SU-8), and (PMMA) layers [2]. [GWR10] Piezoelectric materials were also researched as possible materials for pressure actuated control valves. Our final design utilized silicon as a substrate, with PDMS to form channels and a flexible membrane layer, and SU-8 layers to fabricate rigid fluid flow control gates. We narrowed down our list of potential materials by determining the desired material properties in our design as well as the ease of manufacturing of each material. [GWR11]

Initial Literature Research Results

We divided our group into teams researching different areas of interest including microchannels and control devices. The microchannel team researched multilayer micro-fluidic designs. The control device team researched various control valve designs.[GWR12]

Single level microfluidic devices are limited to fluid flow in two-dimensions. To explore the advantages of microfluidic devices having more than a single level, we examined Prof. Stephen R. Quake’s work on microfluidic multiplexors that are combinatorial arrays of binary valve patterns [3,4].[GWR13] Their work focused on increasing the processing power of a networkby allowing complex fluid manipulations with a minimal numberof controlled inputs. The multiplexors worked as a binary tree and allowedcontrol of n fluid channels with only 2log2n control channels. The integration of additional microfluidic levels was shown to overcome the limitations of single level microfluidics.

In the effort to control fluid flow in microfluidic devices, an attempt is being made to phase out check valves and other mechanisms that slow down the frequency response of the pumping system. The control device team researched many controlled valve designs including pressure, bubble, and PZT actuated valves [5]. The easier and cheaper the valve is to fabricate the more likely it will be used. From the literature on various valve designs, the pressure actuated valve seemed to be the most feasible design for our project.

Device Design Overview

In each device design stages we will have have objectives, device logic, device dimensions, materials used in the design, the processing method, additional issues, manufacturing results and experimental results.

The initial microchannel design consists of only 2 layers with interconnects. The initial design purpose is only used to test if the fluid flows through channels. Controls are neglected in this design because if fluid cannot flow through the channels then adding controls will not necessary. The second stage is the modified version of stage one which is designed to fit the packaging that will be used during testing. The third stage and final design stage consists of an actuated valve that will allow control over fluid flow. Within each stage are fabrication and experimental results that leads to transition from one stage to the next.

Device Design Stage 1: Initial Microchannel Test Design

Device Objective

Once the design requirements and assumptions were finalized, the group determined that testing preliminary designs on the path to a final design was necessary to ensure constant feedback to assess the practicality of design choices. In this path, the testing of the fluidic channels was of primary importance, as the option of including control elements would be mute moot [GWR14]if the fluid itself was unable to pass through the channels designed. Therefore, the group generated the Initial Microchannel Test Design[GWR15]. The purpose of this device was to allow the group to test the experimental capabilities available to us, as well as establish a base upon which more complicated devices could be modeled. More specifically, the device was designed to test the viability of basic multi-level micro-fluidic devices with the equipment and materials currently available.

Device Logic

The design that was chosen consisted of the simplest two-channel horizontal/lateral interconnect layer three-dimensional channel geometry that could be constructed, and at the same time test the practicality of multi-level microfluidics. Moreover, the group decided to use sequential layers of PDMS molds[GWR16] to build up the desired structure. These molds were to be stacked on a bare silicon wafer, which would provide rigidity for transport and testing. There were several reasons for this choice, which included the following:

1. The PDMS layers would allow for heightened design flexibility, as the molds could be re-used and the layers created from these molds could be stacked numerous times in several different orientations.[GWR17]
2. The existing knowledge available to the group based on prior experimental tests done on similar processes by students in the department.
3. The known material compatibility between PDMS and many biological agents that could be used in multi-level micro fluidic devices.
4. The equipment and material constraints based on availability, or lack thereof, of potential materials and equipment for other, less common forms of microprocessing.
5. The PDMS and SU8 molding process and fabrication of isolated, patterned PDMS layer from SU-8 molds were was known to have a relatively fast turnaround time, and this was critical due to the time constraints on the semester.

Figure 1 below shows a schematic of the proposed design. The colors (blue, green, and yellow) signify voids within the PDMS (white). The design consists of three layers: one interconnect layer, and two microchannel layers. The lower microchannel layer (shown in blue) connects to the interconnect pegs (shown in green), which in turn, connect to the top microchannel layer (shown in yellow). The circles located at the ends of the microchannels are the reservoirs which run from the top to the bottom layer, and provide top down access points to all the microchannels, thus allowing for fluid access to all microchannels to assist in testing. Fluid flow into anyone of the 12 reservoir inputs would allow the fluid to enter the device and test to ensure the fluid was able to stay within the pre-determined microchannels that were constructed, as opposed to being forcedforcing between the layers and resulting in layer delamination. [GWR18]

Micro-Channel Layer 1

Micro-Channel Layer2

Interconnect Layer

Figure 1: A Schematic of the Initial Microchannel Test Design (Top View)

Device Dimensions

Based on the logic of the design, the group then determined some appropriate dimensions. Given that this was the first design stage of the semester, there were only a few constraints on the dimensions that could be chosenknown or understood, so the dimensions were chosen based on the approximate sizes encountered in most of the literature. The only constraints that were considered during the device dimensioning were the overall silicon wafer size of diameter = 4 inches and the fact that the PDMS layers could not be molded to a thickness greater then approximately 100m, given past experimental use with the material. Below in Table 1 is a summary of all the critical dimensions that were determined for this Initial Microchannel Test Design. These dimensions were chosen very loosely as the purpose of this design was to test the general performance of the design proposal and not the specifics of the device geometry.

Critical Dimension / Value
PDMS Layer Height / 100m
Microchannel Width / 150m
Microchannel Length / 45mm
Interconnect Width / 150m
Interconnect Depth / 150m
Reservoir Diameter / 300m
Distance Between Channels / 300m

Table 1: A Table of the critical dimensions for the Initial Microchannel Test Design

The cross sectional shapes of the microchannels were rectangularprocess limited, as the SU8 and PDMS molding process does not easily allow for the creation of ridges or grooves that are non-rectangular. Therefore, the cross-section the microchannels were made rectangular. Given the different orientation of the reservoirs and interconnects relative to the micro channels, they could have been made any number of shapes, however, for simplicity, the interconnects were made square in vertical cross-section and the reservoirs were made circular as seen in layout viewin cross-section. [GWR19] These dimensions and geometry constituted, what the group thought asconsidered, the most basic design option to test the viability of multi-level micro-fluidic devices.

Materials

In the introductory sections, we gave a list of materials that are candidates. In this section, we will be discussing the materials that are used, as well as why they are used.

At this stage, the materials used for fabrication of our device are silicon, SU-8 and PDMS (polydimethyl siloxane). We selected a silicon wafer as our substrate because it is cheap and convenient for most of the fabrication processes like lithography. PDMS is a soft polymer that has attractive physical properties[GWR20], in addition to a low cost. Fabricating PDMS [GWR21]involves a lithographic process. Its physical properties include elasticity, conformality, optical transparency, etc.[GWR22] Devices made of PDMS can be integrated with other components, since PDMS conforms to materials like silicon or glass easily[GWR23]. This conformal property makes both reversible and irreversible sealing possible. It is non-toxic to biological agents, such as proteins, and it is gas permeable.[GWR24] Also, since it is transparent in the visible/UV region, it is compatible with many optical detection methods.