An automated microfluidic nucleic acid extraction system for sample preparation with an integrated polymerase chain reaction module

This project will produce an automated microfluidic system capable of extracting and purifying nucleic acids from raw samples for detection and analysis. The first step will be the development and characterization of microfluidic components and fabrication methods that will be implemented into the final device. The gas permeability properties of PDMS will be utilized to demonstrate integrated components for pumping, gas bubble trapping and removal, and enhanced mixing. Next, a three-layer PDMS with silicone membrane microfluidic platform will be developed to control fluid flow for nucleic acid purification processes. This microfluidic chip will be capable of taking a raw biological sample through the steps of cell lysis and solid phase nucleic acid extraction to deliver purified DNA or RNA for testing and analysis. The microfluidic chip will be mounted on a portable, desktop control system to allow automated device operation in clinics, laboratories, or the field. Finally, a disposable oscillatory flow PCR chip will be made from polycarbonate to amplify low concentrations of nucleic acid. The PCR module will also be controlled by the same instrument used for nucleic acid extraction. Temperature control will be provided by external heating blocks, and internal chip fluid temperature will be determined by numerical simulations. This device will be a step towards having a universal nucleic acid purification device to fill the much-needed niche in sample preparation for lab-on-a-chip applications.

AN AUTOMATED MICROFLUIDIC NUCLEIC ACID EXTRACTION SYSTEM FOR SAMPLE PREPARATION WITH AN INTEGRATED POLYMERASE CHAIN REACTION MODULE by Michael Aaron Johnson A dissertation submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Mechanical Engineering The University of Utah May 2018 Copyright © Michael Aaron Johnson 2018 All Rights Reserved The University of Utah Graduate School STATEMENT OF DISSERTATION APPROVAL The dissertation of Michael Aaron Johnson has been approved by the following supervisory committee members: , Chair Bruce Gale 6/6/2015 Date Approved , Member Tim Ameel 6/19/2015 Date Approved , Member Kuan Chen Date Approved , Member Debra Mascaro 6/18/2015 Date Approved , Member Hanseup Kim Date Approved and by the Department/College/School of , Chair/Dean of Tim Ameel Mechanical Engineering and by David B. Kieda, Dean of The Graduate School. ABSTRACT This project will produce an automated microfluidic system capable of extracting and purifying nucleic acids from raw samples for detection and analysis. The first step will be the development and characterization of microfluidic components and fabrication methods that will be implemented into the final device. The gas permeability properties of PDMS will be utilized to demonstrate integrated components for pumping, gas bubble trapping and removal, and enhanced mixing. Next, a three-layer PDMS with silicone membrane microfluidic platform will be developed to control fluid flow for nucleic acid purification processes. This microfluidic chip will be capable of taking a raw biological sample through the steps of cell lysis and solid phase nucleic acid extraction to deliver purified DNA or RNA for testing and analysis. The microfluidic chip will be mounted on a portable, desktop control system to allow automated device operation in clinics, laboratories, or the field. Finally, a disposable oscillatory flow PCR chip will be made from polycarbonate to amplify low concentrations of nucleic acid. The PCR module will also be controlled by the same instrument used for nucleic acid extraction. Temperature control will be provided by external heating blocks, and internal chip fluid temperature will be determined by numerical simulations. This device will be a step towards having a universal nucleic acid purification device to fill the much-needed niche in sample preparation for lab-on-a-chip applications. TABLE OF CONTENTS ABSTRACT ...................................................................................................................... iii LIST OF TABLES ........................................................................................................... vii Chapters 1. INTRODUCTION .........................................................................................................1 1.1 Motivation ..........................................................................................................2 1.2 PDMS-Based Microfluidic Components and Fabrication Background.............6 1.3 Microchip-Based Nucleic Acid Extraction ........................................................8 1.3.1 Solid Phase Extraction Introduction and Background ............................8 1.3.2 Current State of Microfluidic Nucleic Acid Sample Preparation ............9 1.4 Nucleic Acid Amplification ............................................................................15 1.4.1 Microchip Based PCR ...........................................................................15 1.4.2 Oscillatory Flow PCR ..........................................................................16 1.5 Project Overview ............................................................................................20 1.5.1 Summary ..............................................................................................20 1.5.2 Chapter Overviews ...............................................................................21 1.6 References ........................................................................................................23 2. BUBBLE INCLUSION AND REMOVAL USING PDMS MEMBRANEBASED GAS PERMEATION FOR APPLICATIONS IN PUMPING, VALVING AND MIXING IN MICROFLUIDIC DEVICES ....................................36 2.1 Abstract ............................................................................................................37 2.2 Introduction ......................................................................................................37 2.3 Background ......................................................................................................38 2.3.1 Pumping ................................................................................................38 2.3.2 Bubble Injection and Removal ..............................................................38 2.3.3 Bubble Enhanced Mixing .....................................................................40 2.4 Materials and Methods .....................................................................................40 2.4.1 Pumping ................................................................................................40 2.4.2 Bubble Injection and Removal ..............................................................41 2.4.3 Mixing ..................................................................................................41 2.5 Results and Discussions ..................................................................................42 2.5.1 Glass Channels .....................................................................................42 2.5.2 Bubble Removal ...................................................................................42 2.5.3 Mixing ..................................................................................................43 2.6 Conclusion ......................................................................................................44 2.7 References .......................................................................................................44 3. NUCLEIC ACID EXTRACTION SYSTEM .............................................................46 3.1 Introduction ......................................................................................................47 3.1.1 Extraction System Overview .................................................................47 3.2 Materials and Methods .....................................................................................48 3.2.1 Chip Design and Fabrication .................................................................48 3.2.2 Solid Phase Extraction Chip Fabrication ...............................................54 3.2.3 Nucleic Acid Quantification Methods ..................................................57 3.2.4 Extraction Protocols .............................................................................60 3.2.5 Microfluidic Control System .................................................................61 3.2.6 Surface Modification .............................................................................63 3.2.7 RNA Extraction .....................................................................................63 3.2.8 DNA Extraction .....................................................................................65 3.3 Results ..............................................................................................................65 3.3.1 Microfluidic Chip Fabrication ...............................................................65 3.3.2 Silicate Solid Phase Membrane Extraction Yield Characterization and Improvement .......................................................67 3.3.3 Nucleic Acid Quantification Methods ...................................................70 3.3.4 Microfluidic Control System .................................................................70 3.3.5 Surface Modification .............................................................................70 3.3.6 Nucleic Acid Extraction ........................................................................72 3.4 Conclusions ......................................................................................................79 3.5 References ........................................................................................................80 4. OSCILLATORY FLOW PCR WITH EVAPORATION CONTROL ON A POLYCARBONATE CHIP.........................................................................................81 4.1 Abstract ............................................................................................................82 4.2 Introduction ......................................................................................................82 4.3 Materials and Methods .....................................................................................83 4.3.1 Chip Fabrication ....................................................................................85 4.3.2 Temperature Control ............................................................................89 4.3.3 Temperature Modeling .........................................................................90 4.3.4 Temperature Testing..............................................................................96 4.3.5 Pumping/Integration .............................................................................97 4.3.6 PCR Process Parameters .......................................................................97 4.3.7 Gradient PCR .......................................................................................98 4.3.8 Melting analysis ...................................................................................99 4.3.9 Evaporation Control .............................................................................99 4.3.10 Chip Coating .....................................................................................100 4.4 Results ...........................................................................................................101 4.4.1Chip Fabrication ...................................................................................101 v 4.4.2 Numerical Simulations ........................................................................101 4.4.3 Evaporation Control ............................................................................108 4.4.4 Amplification.......................................................................................111 4.5 Conclusions ....................................................................................................114 4.6 References ......................................................................................................115 5. CONCLUSIONS AND FUTURE WORK ................................................................116 5.1 Conclusions ...................................................................................................117 5.2 Contributions..................................................................................................119 5.3 Future Work ..................................................................................................120 Appendices A: MICROFLUIDIC SAMPLE PREPARATION: CELL LYSIS AND NUCLEIC ACID PURIFICATION .................................................................124 B: PROTOCOLS FOR EXTRACTION OF NUCLEIC ACID USING LABVIEW CONTROLLED INSTRUMENT ............................................138 C: MATLAB CODE FOR 1D STEADY STATE AND 1D TRANSIENT SIMULATIONS FOR INTERNAL TEMPERATURES OF MULTILAYER POLYCARBONATE OSCILLATORY FLOW PCR CHIP ..........151 vi LIST OF TABLES Tables 1.1 Current state of reported microfluidic nucleic acid sample preparation systems ..........5 3.1 Summary of protocol steps as performed by the microfluidics automated control program ........................................................................................................................71 3.2 RNA extraction yields from four samples of E. Coli cells ..........................................75 3.3 List of microfluidic extraction runs and RNA quantitation for each run .....................76 3.4 Nucleic acid extracted from multiple sample types .....................................................79 4.1 Description of PCR reaction chemistry .......................................................................98 4.2 Simulated and calculated values for heat flux and convection coefficients based on chip surface temperature ............................................................................................104 4.3 Calculated and measured chip surface temperatures .................................................105 4.4 Effect of heat sink compound on temperature uniformity .........................................106 Table 1 Design considerations for a cell lysis component ...............................................131 Table 2 Comparison of nucleic acid techniques ..............................................................134 B.1 Protocol steps for basic extraction of DNA ..............................................................139 B.2 Protocol steps for basic extraction of RNA...............................................................140 B.3 Protocol steps for extraction of RNA from viral samples .........................................142 B.4 Steps for extracting RNA from bacteria samples ......................................................145 B.5 Protocol steps for extracting DNA from blood samples ...........................................149 CHAPTER 1 INTRODUCTION 2 1.1 Motivation Biological assays based on nucleic acid (DNA and RNA) analysis have been explored for a wide range of applications: pathogen detection,1-8 genetic analysis,9-12 environmental monitoring,13-18 terrorism or biowarfare agent detection,19, 20 foodborn illness prevention,21-24 and others. In each of these cases rapid on-site detection is desired, but current laboratory-based nucleic acid detection methods are generally not compatible with this goal. Such methods require highly trained laboratory staff, specialized equipment, and clean environments to reduce the risk of contaminating samples. As an alternative to lab-based methods, microfluidic lab-on-a-chip devices may be able to deliver rapid, on-site detection and are appealing for their ability to consolidate the steps needed for nucleic acid analysis onto a single, compact device. Since its inception over two decades ago, the field of microfluidics has surged as people have seen the potential benefits of miniaturized systems capable of performing rapid biological assays. Microfluidic systems progressing towards the goal of reducing complex laboratory analysis processes onto a single lab-on-a-chip device often rely on nucleic acids as the main analysis target because of their ubiquity, stability, and specificity. Many different sensing systems have been developed using techniques such as melting analysis after polymerase chain reaction (PCR),25 surface plasmon resonance (SPR),26, 27 microarrays,28-31 and other technologies.32-36 While work in microfluidics has been ongoing for many years, there have been few significant advances towards widespread commercialization of microfluidic systems in recent years and some have expressed doubt about whether microfluidic devices will be able to reach this potential. George Whitesides, one of the leading microfluidic authorities, illustrates this issue and 3 provides potential reasons for this lack of progress in the following statement: [Microfluidics] offers so many advantages and so few disadvantages. . . . But it has not yet become widely used. Why not? . . . Revolutions in technology require both a broad range of different types of components and subsystems, and their integration into complete, functional systems. . . that can be used by non-experts. . . . The original hope for microfluidics, and that which still motivates many of us working in the field, is that it will be a practical technology. . . Above all, it must become successful commercially, rather than remain a field based on proof-of-concept demonstrations and academic papers. 37 This statement aptly summarizes the motivation behind the research presented in this work. As supported by the above statement, the first requirement for a microfluidic technological revolution is an increase in the number and variety of microfluidic components and subsystems. Work on new components, and their application to lab-ona-chip fluid handling, increases the number of tools available to researchers. These components form the basic units which can be used to assemble more complex systems capable of performing the fluid handling tasks for on-chip assays. Once sufficient components and subsystems have been identified, there remains the task of integrating these components into functional systems. Primary research on improved microscale pumps, valves and mixers does little to expand the use of microfluidics if they cannot be integrated into operative systems. Such integration needs to consider the compatibility of these components with each other as well as to the samples, chemical reagents, and control architecture. The last requirement is that the integrated microfluidic system must be usable by nonexperts. This need highlights a problem faced by a large number of microfluidic devices where the expertise and equipment required to operate the microfluidic device prohibit its widespread use. Such systems have come to be known as "chip-in-a-lab" 4 devices, a critical term derived from the "lab-on-a-chip" designation, drawing attention to the failure of microfluidic systems to simplify laboratory processes. Even with some of these unsolved challenges in microfluidics, there are several specific application areas that have benefited from microfluidic technologies. An example of one such application is single cell analysis,38, 39 where the economies of scale available with microfluidics permit higher throughput with lower cost, while simultaneously simplifying the automation of analysis. In this manner single cells have been the target for protein analysis40 and gene expression.41 Microfluidic devices have also been successfully used to synthesize carriers for drug delivery systems.42, 43 But perhaps the largest field of success in current microfluidic research and technology is biomolecule detection. Microfluidic devices have the ability to isolate, amplify, or detect target molecules, most often DNA and RNA. While such techniques show great promise in expanding the role of microfluidic systems in reducing human intervention in performing bioassays, there is still a problem. These sensing platforms demand purified nucleic acids, which typically require manual genetic extraction and purification methods to remove unwanted background material from the sample. Incomplete integrated sample preparation has thus limited the use of lab-on-a-chip devices to specialized laboratories with the capability to perform the sample preparation steps. To achieve the widespread development of bench-top and portable microfluidic analysis systems with true sample-in, answer-out capabilities, development of highly integrated and efficient nucleic acid purification sample preparation devices will be needed. The need for sample preparation devices is especially evident in cases where 5 analysis chips neglect the important sample preparation step. Instead, off-chip laboratorybased nucleic acid sample preparation methods are used as preliminary steps to provide purified nucleic acid as the sample. Current bench-top nucleic acid sample preparation techniques are highly labor intensive, requiring skilled human interaction at multiple junctions to perform steps such as: filtration, cell lysis, centrifugation, nucleic acid separation, etc.44 The incorporation of these steps into a single lab-on-a-chip device would yield many benefits including the elimination of the need to have highly skilled laboratory personnel perform the operations, a reduction in the chance of sample cross contamination, and the minimization of the amounts of costly reagents. A wide variety of researchers have been working to achieve these goals. The current state of microfluidic nucleic acid sample preparation systems is summarized in Table 1.1, with further discussion in section 1.3. Table 1.1 demonstrates a current lack of microfluidic sample preparation systems capable of starting with multiple sample types to Table 1.1 Current state of reported microfluidic nucleic acid sample preparation systems. Only the Utah system reports successful integration of the features listed. References Extract Extract Integrated Automated Multiple DNA RNA PCR or Sample Module Types 15, 47-61  12, 14, 62-65   66-71   9, 13, 72-79    80-82   29, 83, 84  85, 86   87, 88   89-91   92    93-95    96-98     99     Utah System      6 extract the target molecule (DNA or RNA) in an automated fashion with an optional PCR module for amplification. There are a large number of devices capable of extracting DNA or RNA, but targeting both types of nucleic acid is less common. Of those that can extract both DNA and RNA, there are even fewer that report automation, the capability of handling multiple sample times, or integrated PCR capability. However, all of these processes are needed in a generic nucleic acid sample preparation to be fully successful and enable further development and commercialization of microfluidic nucleic acid analysis systems. Accordingly, this work introduces a system capable of filling this need and is a step towards solving the problems which have labeled nucleic acid sample preparation the weak link in microfluidic biodetection,45 and the most important obstacle to a revolution in biomolecule analysis.46 1.2 PDMS-Based Microfluidic Components and Fabrication Background The adaptation of the multiple fluid handling steps of sample preparation to a labon-a-chip system requires on-chip components such as pumps, microchannels, valves, and mixers. There have been many reports in the literature on component improvements and fabrication methods for these components in substrates of glass,100-102 silicon,47, 103 and polymers.104-106 One method for fabricating microfluidic chips evolved from integrated circuit (IC) silicon micromachining. This technique creates microchannels and functional microfluidic components using microfabrication processes such as photolithography, wet or reactive ion etching, chemical vapor deposition, and others. Devices created in this manner exhibit complexity and versatility in performing microfluidic fluid handling 7 operations. Another benefit derives from the ability to seamlessly integrate control circuitry on the same chip using IC fabrication protocols. However, production complexity and cost are deterrents for silicon microfluidic chips to find widespread use as disposable medical diagnostic devices. In addition, microfluidic devices tend to be relatively large, compared to typical silicon structures, so silicon processing technologies tend to be expensive and over-capable when considered for microfluidic technologies. Polymers have emerged as alternative substrates to silicon-based microfluidics with the promise of providing inexpensive disposable chips. Fabrication of these devices is accomplished using injection molding, hot embossing, layered stacks and casting. Active components are difficult to achieve with polymers, so these chips usually rely on external actuators and pumps to perform on-chip operations. One polymer of note that finds extensive use in the microfluidic community is polydimethyl siloxane (PDMS), a silicone-based flexible elastomer that has desirable chemical properties and can be rapidly molded around structures to create submicrometer features. Publications reporting the use of replica molding of PDMS as a substrate for microfluidic applications began to appear in the late 1990s.107-109 An increase in the popularity of PDMS was precipitated by several landmark papers: PDMS as a microfluidic substrate and a rapid prototyping process introduced by the George Whitesides group,110, 111 and monolithic valves and large scale integration by the Stephen Quake group.106, 112 Advancements in PDMS-based components and integration have continued with new valve designs101, 113 and pumping schemes,114 allowing other researchers to create more complex microfluidic platforms.115 Further study of the capabilities of PDMS as a microfluidic substrate will help in the development of devices 8 with the capabilities to handle the complex fluid handling for nucleic acid extraction. 1.3 Microchip-Based Nucleic Acid Extraction 1.3.1 Solid Phase Extraction Introduction and Background In order to use nucleic acids as target molecules for bioanalysis, the nucleic acid first needs to be released from the cells, isolated, and purified by purging unwanted cellular debris. While there are other means (e.g., phenol-chloroform based precipitation, isotachophoresis) for completing this extraction process, a large portion of the recent methods for both bench-top and microfluidic extraction have relied on the principle of solid phase extraction. Solid phase extraction relies on three main steps: binding, washing, and elution. In the binding step, nucleic acid is bound to the surface of a solid phase under certain buffer conditions. For washing, cellular debris and unwanted contaminants are rinsed off of the solid phase while the nucleic acid remains attached. Finally, the nucleic acid is eluted from the solid phase with a low ionic strength elution buffer. One of the first demonstrations of solid phase extraction was performed in 1979 by Vogelstein and Gillespie and used a glass surface as the solid phase to extract DNA. Vogelstein and Gillespie used a chaotropic salt solution to selectively bind DNA to glass surfaces (in powdered form) to elute the DNA molecules from the surrounding agarose separation medium.116 Under the conditions reported, they found that 1 μg of DNA would bind to a glass surface of about 750 mm2. To increase the surface area, they employed a glass that had been ground to powder. Glass fiber filters were also used, but they found that recovery was too small to quantify. In their experiment, over 99% of the DNA was bound to the glass, with a recovery of about 90%. It should be noted that they 9 were using prepurified DNA that had previously been separated in agarose gel, and their DNA extraction was simply a separation from agarose, not from original cellular material. This method was expanded upon in 1990 by Boom et al.117 They demonstrated that a similar procedure could be used to purify nucleic acids (both DNA and RNA) from cell-rich samples similar to clinical samples. In their process, they introduced the use of two new chaotropic salt solutions to act as the binding buffer. Guanadinium thiocyanate (GuSCN) and guanadinium hydrochloride (GuCHl) proved to be powerful tools for the purification of nucleic acids because they have the ability to lyse target cells and subsequently bind the released nucleic acid to the solid particles. GuSCN also has the beneficial property of deactivating nucleases, making it especially useful in purifying more fragile RNA molecules. Nucleic acid extraction from human urine and serum was shown to have over 50% efficiency using their method, despite evidence that nucleic acid was being lost at each major step of the extraction process. Baker et al. used glass fiber filters to extract pure DNA from whole blood, but their process does not depend on chemical buffers to promote reversible DNA adhesion to the glass solid phase.118 Instead, white blood cells are trapped in the filter matrix and lysed using a detergent solution. The DNA is then also physically trapped in the filter until released in a special incubation vessel and eluted. This process demonstrates the ability of glass filters to be used to purify DNA, but does not explore using the filters as the binding surface using solid phase extraction techniques with chaotropic salts. 10 1.3.2 Current State of Microfluidic Nucleic Acid Sample Preparation The NA purification processes described thus far involve several steps requiring manual manipulation by personnel. Commercial systems often used in laboratories today use variations on this process. These and other similar methods require pipetting, mixing, centrifuging, vortexing, transferring containers, incubation and other time consuming and labor intensive steps. Labor reduction has been achieved by the implementation of robotics to complete the fluid handling steps, but such systems are too expensive to see much use outside of centralized laboratories. Microfluidic chip analysis methods lend themselves well to such situations by reducing the required human involvement by integrating the essential steps onto small benchtop platforms. The literature shows that nucleic acid extraction is a process that has been adapted in several variations to microfluidic chips. Performing nucleic acid extraction on-chip has been the focus of much attention in the literature. Appendix A is a published literature review of microfluidic advancements in on-chip cell lysis and nucleic acid purification. An updated review including emphasis on devices related to the system presented in this work is discussed in the remainder of this chapter. The stability and specificity of DNA, as compared to RNA, has made it an attractive option as the target for microfluidic based bioassays, leading to a large number of on-chip DNA extraction devices.15, 47-61 A few of the most recent and most powerful will be discussed. The research group of James P. Landers has published extensively on methods of performing solid phase based nucleic acid extraction on microchips. In 2003 Wolfe et 11 al.61 introduced the idea of using a sol-gel to immobilize silica beads inside microfluidic channels to form a porous, high-surface area solid phase on which nucleic acids can bind. This method eliminated the difficulties of including a weir to hold the beads in place. A guanidine-based binding solution with DNA was passed through the porous matrix, washed with ethanol, and eluted with TE buffer. This process generated extraction efficiencies ranging from 8.7% to 70.6%. Breadmore et al. expanded on this work by using biological samples, including whole blood and bacteria.95 Extraction efficiency data were not presented, but the effectiveness of extraction was qualitatively demonstrated by amplifying the extraction product by PCR and performing gel electrophoresis. As with their previous work, no mention was made on how pumping and interfacing with samples and reagents were done. Similar results were later obtained by the same group80, 81 with the distinction that the solid phase was fabricated as a porous monolith instead of immobilized beads. Other research groups have reported similar results using an amino silica monolith.55 One final method of note introduced by the Landers group deviates from their common practice of using guanidine-based buffers to bind nucleic acid to silica surfaces. Instead, Hagan et al.91 coat the silica surface of glass microchannels with chitosan. The nucleic acid can then be bound to and eluted from the chitosan using buffers at different pH that create induced-charge electrostatic means. Using this method, a slight improvement in extraction efficiency was observed (65% vs 40% for chitosan vs plain silica, respectively), and reagents that can inhibit later PCR amplification reactions were avoided. Other efforts to perform nucleic acid purification on the microscale include using 12 optimized microfabricated silica pillar structures as the solid phase to extract DNA from E. Coli bacteria.47 Mesoporous silica was used to demonstrate that pore sizes of 2-5nm were effective at binding duplex DNA without needing high ionic strength chaotropic salt solutions.119 Similarly, DNA was extracted on a nanoporous aluminum oxide membrane with NaCl acting as the binding promoter.53, 120 Xing et al. fabricated a porous silicon dioxide matrix on a silicon chip to serve as the solid phase.121 In a step towards more disposable chips, Bhattacharyya and Klapperich presented a thermoplastic microfluidic chip for nucleic acid purification fabricated by hot embossing.3, 49 To create the solid phase, a porous plastic monolith was impregnated with silica particles. Successful purification of RNA and DNA was demonstrated. Another plastic-based chip used layered double hydroxides (LDH) on the surface of polycarbonate chips to bind DNA, which was released in a slight acid etch of the LDH.52 Yet another approach to solid phase extraction employs a lab-on-valve system with a renewable silica microcolumn packed inside a channel.51 A finding of note from this setup is that longer residence times of the target in proximity with the silica surface increase the elution efficiency. Nanassy et al. also avoided microchip extraction by extracting DNA directly onto glass microscope slides.50 A highly parallel system was shown by Park et al. where a polycarbonate 96-well microfabricated titration plate was used to extract DNA from bacteria using a NaCl immobilization buffer, ethanol wash and DI water elution.122 The works discussed above are very valuable at exploring the methods and effectiveness of microchip-based solid phase extraction of nucleic acids. However, little effort has been made in the publications discussed to address the interface between the 13 microchip and the outside world. Expensive and complicated fluid pumping and analysis apparati are required to run such microchips, demanding the same kind of technical expertise required with current benchtop protocols. To make a practical difference, microfluidic extraction systems need to be highly integrated and automated, as well as effective. One more integrated system was reported in 2007, which used a silicon/glass microfabricated chip as the heart of the extraction process.69, 123, 124 The chip included a size-based exclusion filter to remove red blood cells in order to extract DNA and viral RNA from blood samples. The integrated system was shown to be effective by performing PCR and RT-PCR on DNA and RNA samples, respectively, but separate, portable systems had to be made for the DNA and RNA protocols. Beier et al. devised a sample preparation chip with integrated components such as inlet, filter, solid phase extraction chamber, reagent storage, valves, etc.87 The chip is operated by an instrument, eliminating the need for other external equipment. While this device demonstrates hands-free automation of sample preparation, the target application is very specific: extracting viral RNA from cervical liquid for the detection of HPV infections. Another popular method of performing solid phase extraction is to use silica superparamagnetic particles, which allows the mobile solid phase to be manipulated using magnetic fields. One device uses these particles, immobilized by permanent magnets, to purify DNA from blood and perform PCR.12, 125 Commercial extraction magnetic bead systems have also been used in conjunction with microfluidic systems to combine sample preparation and hybridization for microarrays.28 14 Chen et al. presented a disposable polycarbonate cassette that contains reagents, a silica membrane modified from a commercial extraction kit, and a PCR chamber.97 The temperature of the stationary PCR reaction chamber is controlled using thermoelectric units at a rate of approximately 5˚C/sec. While their work is impressive in providing a self-contained chip that performs extraction, PCR, and RT-PCR for RNA and DNA detection, the types of samples used were limited to spiked saliva samples. No efforts were reported to extract nucleic acids from blood or other more complex matrices. Another fully integrated sample preparation chip was demonstrated by RitziLehnert et al. with application towards analysis of respiratory viruses from nasal swabs.96 They present a lab-on-a-chip nucleic acid purification and amplification system focused on viral RNA detection based on the one-tube RT-PCR QIAplex method. The on-chip RT-PCR and PCR thermal cycling is achieved using rotating constant temperature clamps on the top and bottom of the PCR reaction chamber with ramp rates of approximately 5˚C/sec for large volume (120 μl) PCR reactions. While this device demonstrates the ability to extract both RNA (target) and DNA (control), the sample preparation method is explicitly assay specific for RNA respiratory viral detection and no effort was reported to expand the types of sample inputs to create a more universal sample preparation device. Previously reported work from Bruce Gale's lab at the University of Utah has produced a microfluidic chip for extraction of RNA and DNA from blood and bacteria samples as a step towards universal sample preparation.99 However, this device required several off-chip lysis steps that are specific to the target cell prior to loading in the chip. Also, no attempt has been made to integrate PCR control into the automated extraction 15 process. The current system is an improvement in that it integrates cell lysis on-chip and a PCR module for nucleic acid amplification. 1.4 Nucleic Acid Amplification Using microfluidics to perform nucleic acid extraction and analysis presents a barrier in that only small amounts of sample are used, which makes the detection of scarce target molecules especially difficult because there can be very low concentrations of the target available. Using polymerase chain reaction (PCR) to amplify the number of molecules available for analysis shows great promise in overcoming this difficulty. PCR makes copies of targeted DNA strands by using enzymatic primers to locate and replicate specific nucleotide sequences. The process depends on having the primers go through a temperature cycle where the DNA molecule denatures (unzips the double-strand into two single strands) at a high temperature, anneals (the primers attach to the specific sequence of interest) at a low temperature, and extends (the primer moves along the strand, inserting complementary base pairs to again create a double-stranded DNA molecule) at a moderate temperature. Thus, the amount of target DNA can be doubled simply by putting the prepared sample through a single temperature cycle. 1.4.1 Microchip Based PCR Adapting PCR to the microchip has seen much attention because of the rapid thermal cycling possibilities when using small liquid volumes, the reduced reagent usage, and the ability to reduce contamination between processing steps. There are several substrates that have been used in the manufacture of microfluidic PCR systems. The most prolific substrates are silicon,2, 126-137 glass11, 138-150 or a silicon/glass hybrid.83, 151-156 These substrates are attractive as they take advantage of well-established micromachining 16 practices and allow easy integration with optical and electrical components. Polymers have seen accelerated interest as substrates for PCR chips as the less expensive manufacturing processes enhance the possibility of making the chips disposable. Polymer substrates used include PDMS,73, 157-163 polycarbonate,75, 98, 164-168 or other plastics.169-173 1.4.2 Oscillatory Flow PCR While there are a wide range of techniques that can be applied to performing PCR, oscillatory-flow PCR is of particular interest and will be used extensively in this dissertation. Oscillatory flow PCR operates on the principle of pumping the sample fluid through distinct zones held at the required cycle temperature. With this system architecture, the plug of fluid moving to the new temperature zones is the only component in the device that undergoes temperature transitions, thus eliminating the need to unnecessarily heat and cool the chip itself. Oscillatory flow PCR has the added advantage over continuous flow PCR devices due to its flexibility in cycle times and temperatures. Specifically, many PCR protocols require a long incubation period at a set temperature for initialization or finalization of the PCR process. This incubation period is especially evident when dealing with RNA, which requires an initial reverse transcription step to convert the RNA strand to a more stable complementary DNA strand before PCR operation. These advantages make an oscillatory flow PCR chip advantageous as a module for a sample preparation device that deals with both DNA and RNA. Such an oscillatory PCR device was presented by Bu et al. to provide the flexibility in thermal cycling control with the thermal advantages of flow based PCR.174 They presented a microfabricated silicon chip with integrated heaters, a peristaltic pump, and optical sensors to detect fluid droplet position. However, the expensive, complex 17 fabrication methods make this chip a poor candidate for widespread use as a disposable chip. Hardt et al. and Muchow et al. presented polymer chips (PMMA and cyclo olefin copolymer) that rest on three heater blocks with embedded heater cartridges.175, 176 A pneumatically coupled ferrofluid pump was used as the actuation chamber for droplet pumping. Fluid droplet temperatures were predicted by CFD simulations based on temperature settings and infrared measurements. Evaporation was observed, but no attempt at evaporation control was reported. Auroux et al. presented the first results obtained by oscillatory flow PCR.177 Previous to their efforts, oscillatory flow PCR as a technology was theoretical or computational in nature. Channels were formed using patterned SU-8 on PMMA substrates. Pumping was done by an external syringe pump. The flexibility of the oscillatory flow PCR concepts was demonstrated by running several PCR programs with modifications to the fluid residence times. The nonoptimized system demonstrated amplification, although at a significantly reduced efficiency (25%) compared to commercial systems. A radial temperature gradient oscillatory PCR geometry was presented by Cheng et al. where a PMMA fluid chip was placed in contact with a circular ITO/glass heater substrate.178 They also attempted to mitigate evaporation loss by increasing the pressure in the channel as the fluid was pumped over the high temperature zone, which was accomplished by having the high temperature zone near the termination of a dead-end channel causing the air in the channel to be pressurized as the fluid approaches, thus raising the vapor pressure. Using this method, they reported 30% evaporation after 30 18 cycles. However, the high pressures produced by this method require pumping by an external syringe pump. Another silicon-based microfluidic chip reported by Wang et al. demonstrated amplification of a human papilloma virus target sequence after 35 cycles in 15 minutes.179 While this chip demonstrated good amplification, faster than conventional instruments, it suffered from a high degree of fabrication complexity that will make it difficult to be integrated as a disposable component. Ohashi et al. performed PCR in a droplet-in-oil system where sliding permanent magnets were used to move magnetic particles containing PCR droplets across a temperature gradient.166 Another paradigm for oscillatory PCR is based in digital microfluidics as presented by Sista et al.1 In this system, fluid droplets are moved by electrowetting manipulation. Their device performs PCR on a printed circuit board substrate by shuttling a droplet back and forth between two temperature zones heated by aluminum heater bars. While providing flexibility, this chip requires high voltages for electrowetting and laborious glass drilling and bonding processes. Sciancalepore et al. cast a capillary tube in PDMS and positioned the tube over three patterned microheaters.180 A droplet of PCR reagents was suspended in an oil phase and pumped by syringe pumps over the temperature zones. A nested PCR cycle is achieved four times faster (50 min) than with conventional systems. A later report showed the capability of this design to greatly reduce evaporation of the fluid droplet during the PCR process.181 Drawbacks to this design include a 4 ˚C temperature variation in the temperature zones and a slightly slower thermal response time due to 19 interference from layers of glass and PDMS. Sugumar et al. developed a glass chip with serpentine channels that pass over three temperature zones for the amplification of salmonella DNA.182 While their use of oil, a sealed channel and elevated pressure kept evaporation to about 10% (lower than the rate seen in other polymer chips up to this point in this discussion), the difficulty in fabricating glass channels and bonding glass substrates makes polymers a more attractive option for disposable chips. Zhang et al. presented a multichannel oscillatory flow PCR device with applications in detecting food-borne pathogens. PCR was carried out in PTFE capillary tubes which were set into small channels cut into copper blocks. A glass cover was then bound over the top of the channels.183 Temperature zones were heated by cartridge heaters set into the copper blocks; pumping was done by an external syringe pump; and the temperature was controlled by a computer interface and monitored by a thermocouple embedded in the copper. Successful multiplexed amplification of food-borne pathogens from bacteria was demonstrated. Evaporation control was achieved using oil plugs on either side of the sample plug. Mechanical pumping of a PCR fluid droplet between temperature zones on a PDMS/glass hybrid chip was demonstrated by Chia et al.184 The electromagnetic coil actuator depresses a plastic flake into the PDMS chamber, forcing the fluid out into a different temperature zone. While they reported the time it takes heat up the temperature zones, there was no discussion of the time required for the fluid to heat and cool when pumped to different zones. Also, no discussion of evaporation rates was included. The above discussion of the state of the art in microfluidic sample preparation 20 systems shows the need for a device like the Utah system, capable of performing automated extraction of both DNA and RNA. PDMS has proven a viable substrate for microfluidic devices, but further expansion of component capabilities will lead to more complex devices. Solid phase extraction has been successfully applied to microfluidic devices, but further work is needed to demonstrate automated systems with enough protocol adaptability to handle different sample types and target molecules. Microfluidic PCR devices are effective at rapidly amplifying nucleic acids. The oscillatory flow PCR chips provide flexibility in temperature, cycle number, and residence times with accompanying rapid cycles. An oscillatory flow PCR device on a plastic substrate that can control evaporation will be a step forward in disposable PCR chips integrated with extraction systems. 1.5 Project Overview 1.5.1 Summary The research presented here is a step towards the realization of the goal to have functional microfluidic systems that can be operated by nonexperts by implementing practical solutions to problems that have hindered widespread adoption of microfluidic systems. Specifically, this work focuses on the generation of a nucleic acid sample preparation system for the purification of nucleic acid from biological samples of blood, bacteria cells, and cultured virus samples. This system incorporates on-chip chemical lysis of blood and bacteria cells and extracts purified DNA or RNA using solid phase extraction techniques. The system incorporates automated control of all process steps, so users need only load the sample. A polymerase chain reaction (PCR) module is also incorporated to allow amplification of target nucleic acids to increase assay sensitivity. 21 1.5.2 Chapter Overviews Chapter 2 of this work presents findings on how PDMS membranes can be incorporated into microfluidic devices to perform essential fluid handling tasks. Specifically, the chapter outlines how the gas permeability of PDMS membranes is turned from a weakness to a strength to perform controlled pumping at low flow rates, reduce air bubble interference, and enhance mixing. These component advancements were instrumental in proving the versatility of three-layer PDMS chips at fluid handling tasks. Consequently, a three-layer PDMS chip platform was chosen for development of the main fluid control chip for the nucleic acid sample preparation platform. Fabrication methods and microfluidic components are explored that will aid in reducing the complexity of device fabrication, making a final microfluidic system feasible. A thin PDMS membrane will be included into microfluidic devices to perform pumping, valving, bubble removal, and mixing tasks. All of these components will be fabricated using a simple three-layer technique, where a thin membrane is bonded between two thicker substrates (PDMS or glass) that house the fluid channels and packaging/control mechanisms. Having such a simple fabrication method with the ability to control many fluid-handling tasks can help reduce the complexity of manufacturing microfluidic platforms and minimize the number of external controls required for a labon-a-chip device. In Chapter 3 a masked corona discharge bonding method185 will be explored to improve pumping and valving for the more complex fluid handling applications that will be required based on the choice of cell lysis and extraction methods. Chemical cell lysis and subsequent solid phase extraction of nucleic acids require a complex sequence of 22 reagent mixing to efficiently recover the nucleic acid. A major focus of Chapter 3 is on the generation of a single PDMS microfluidic chip with the capacity to control all needed fluid handling tasks for nucleic acid extraction. A range of possible extraction protocols will be considered in the design in order to make the device applicable to multiple situations. The chip will be able to draw from a bank of reagents to perform different extraction protocols. The system will be automated, controlled by a programmable LabView interface. Different extraction sequence programs will be run to selectively purify DNA or RNA from different samples. Following extraction, the purified nucleic acid samples will be delivered to several different detection platforms to verify that the device adequately interfaces between raw sample and genetic analysis. Specifically, DNA will be extracted and quantified using fluorescence, spectroscopic absorption, and PCR. RNA will similarly be extracted and analyzed with fluorescence and RT-PCR with applications for viral detection. Advances in this field will facilitate an increase in the spread and scope of nucleic acid analysis, leading to more insight and discovery of underlying biological phenomena. The proposed work intends to generate these insights by increasing the ability of next generation analysis systems to rapidly process multiple samples using microfludic-based extraction systems. The work presented in Chapter 4 demonstrates a chip that couples the flexibility of oscillatory flow PCR in a disposable polymer chip with good temperature uniformity and low evaporation. A thin multilayer polycarbonate chip will be fabricated from laminates. The chip will contact an external heater block module with PID temperature controllers to create fixed temperature zones on the PCR chip. Extra chip layers that 23 surround the fluid channel with pressurized water will provide evaporation control at the high temperatures experienced during PCR. Numerical simulations will be used to determine internal chip temperatures and required residence times of the fluid droplet at the temperature zones. Rapid on-chip PCR will be demonstrated controlled by a pumping module controlled by the same instrument that performs the nucleic acid extraction discussed in Chapter 3. In total, these advancements in microfluidic sample preparation technology combine to demonstrate a system that improves on current reported devices by performing the complex fluid handling required to extract DNA and RNA from different sample types with an integrated PCR module for amplification. 1.6 References 1. R. Sista, Z. Hua, P. Thwar, A. Sudarsan, V. Srinivasan, A. Eckhardt, M. Pollack and V. Pamula, Lab on a Chip 8 (12), 2091-2104 (2008). 2. P. Belgrader, W. Benett, D. Hadley, G. Long, R. Mariella, F. Milanovich, S. Nasarabadi, W. Nelson, J. Richards and P. Stratton, Clinical Chemistry 44 (10), 21912194 (1998). 3. A. Bhattacharyya and C. M. Klapperich, Sensors & Actuators: B. Chemical 129 (2), 693-698 (2008). 4. E. M. Elnifro, A. M. Ashshi, R. J. Cooper and P. E. Klapper, Clinical Microbiology Reviews 13 (4), 559-570 (2000). 5. D. N. Fredricks and D. A. Relman, Clinical Infectious Diseases 29 (3), 475-486 (1999). 6. V. Srinivasan, V. K. Pamula and R. B. Fair, Lab on a Chip 4 (4), 310-315 (2004). 7. S. Yang, S. Lin, G. D. Kelen, T. C. Quinn, J. D. Dick, C. A. Gaydos and R. E. Rothman, Journal of Clinical Microbiology 40 (9), 3449-3454 (2002). 8. S. Yang, P. Ramachandran, A. Hardick, Y.-H. Hsieh, C. Quianzon, M. Kuroki, J. Hardick, A. Kecojevic, A. Abeygunawardena and J. Zenilman, Journal of Clinical 24 Microbiology 46 (4), 1386-1390 (2008). 9. L. A. Legendre, C. J. Morris, J. M. Bienvenue, A. Barron, R. McClure and J. P. Landers, JALA - Journal of the Association for Laboratory Automation 13 (6), 351-360 (2008). 10. P. Liu and R. A. Mathies, Trends in Biotechnology 27 (10), 572 (2009). 11. C. J. Easley, J. M. Karlinsey, J. M. Bienvenue, L. A. Legendre, M. G. Roper, S. H. Feldman, M. A. Hughes, E. L. Hewlett, T. J. Merkel and J. P. Ferrance, Proceedings of the National Academy of Sciences 103 (51), 19272-19277 (2006). 12. Y. Zhang, S. Park, S. Yang and T.-H. Wang, Biomedical Microdevices 12 (6), 1043-1049 (2010). 13. C. J. Bruckner-Lea, N. C. Anheier Jr, D. A. Holman, T. Tsukuda, M. T. Kingsley, F. J. Brockman, J. M. Price, J. W. Grate and D. P. Chandler, presented at the Environmental and Industrial Sensing, 2000 (unpublished). 14. C. Delattre, C. P. Allier, Y. Fouillet, D. Jary, F. Bottausci, D. Bouvier, G. Delapierre, M. Quinaud, A. Rival and L. Davoust, Biosensors and Bioelectronics 36 (1) (2012). 15. S. Isabel, M. Boissinot, I. Charlebois, C. M. Fauvel, L.-E. Shi, J.-C. Lévesque, A. T. Paquin, M. Bastien, G. Stewart and É. Leblanc, Applied and Environmental Microbiology 78 (5), 1505-1512 (2012). 16. J. C. Jokerst, J. M. Emory and C. S. Henry, Analyst 137 (1), 24-34 (2012). 17. E. A. Ottesen, J. W. Hong, S. R. Quake and J. R. Leadbetter, Science 314 (5804), 1464-1467 (2006). 18. Y. Li, C. Zhang and D. Xing, Analytical Biochemistry 415 (2), 87-96 (2011). 19. D. Ivnitski, D. J. O Neil, A. Gattuso, R. Schlicht, M. Calidonna and R. Fisher, Biotechniques 35 (4), 862-869 (2003). 20. S. Yang, R. E. Rothman, J. Hardick, M. Kuroki, A. Hardick, V. Doshi, P. Ramachandran and C. A. Gaydos, Academic Emergency Medicine 15 (4), 388-392 (2008). 21. A. Cifuentes, ISRN Analytical Chemistry 2012 (2012). 22. P.-G. Lantz, B. Hahn-Hägerdal and P. Rådström, Trends in Food Science & Technology 5 (12), 384-389 (1994). 25 23. A. G. Sciancalepore, E. Mele, V. Arcadio, F. Reddavide, F. Grieco, G. Spano, P. Lucas, G. Mita and D. Pisignano, Food Microbiology 35 (1) (2013). 24. V. Velusamy, K. Arshak, O. Korostynska, K. Oliwa and C. Adley, Biotechnology Advances 28 (2), 232-254 (2010). 25. N. Crews, C. T. Wittwer, J. Montgomery, R. Pryor and B. Gale, Analytical Chemistry 81 (6), 2053-2058 (2009). 26. P. Englebienne, A. Van Hoonacker and M. Verhas, Spectroscopy: An International Journal 17 (2), 255-273 (2003). 27. L. Malic, T. Veres and M. Tabrizian, Biosensors and Bioelectronics 24 (7), 22182224 (2009). 28. T. Chien-Hsuan, S. Jyh-Wei, C. Tsuey-Yu, H. Suz-Kai, L. Chun-Che and L. Gwo-Bin, Microfluidics and Nanofluidics 10 (5), 999-1009 (2011). 29. C.-H. Tai, J.-W. Shin, T.-Y. Chang, S.-K. Hsiung, C.-C. Lin and G.-B. Lee, Microfluidics and Nanofluidics 10 (5), 999-1009 (2011). 30. D. Rose, Microarray Biochip Technology, (M. Shena. Ed.), Eaton Publishing, Natick, MA, 19-38 (2000). 31. C. Situma, M. Hashimoto and S. A. Soper, Biomolecular Engineering 23 (5), 213231 (2006). 32. C. Ziegler, Analytical and Bioanalytical Chemistry 379 (7-8), 946-959 (2004). 33. J. Li, H. T. Ng, A. Cassell, W. Fan, H. Chen, Q. Ye, J. Koehne, J. Han and M. Meyyappan, Nano Letters 3 (5), 597-602 (2003). 34. D. Branton, D. W. Deamer, A. Marziali, H. Bayley, S. A. Benner, T. Butler, M. Di Ventra, S. Garaj, A. Hibbs and X. Huang, Nature Biotechnology 26 (10), 1146-1153 (2008). 35. Z. Xu, T. Esumi, N. Ikuta and T. Hirokawa, Journal of Chromatography A 1216 (17), 3602-3605 (2009). 36. K. M. Ririe, R. P. Rasmussen and C. T. Wittwer, Analytical Biochemistry 245 (2), 154-160 (1997). 37. G. M. Whitesides, Nature 442 (7101), 368-373 (2006). 38. A. R. Wheeler, W. R. Throndset, R. J. Whelan, A. M. Leach, R. N. Zare, Y. H. Liao, K. Farrell, I. D. Manger and A. Daridon, Analytical Chemistry 75 (14), 3581-3586 26 (2003). 39. V. Lecault, A. K. White, A. Singhal and C. L. Hansen, Current Opinion in Chemical Biology 16 (3), 381-390 (2012). 40. M. Wu and A. K. Singh, Current Opinion in Biotechnology 23 (1), 83-88 (2012). 41. V. Sanchez-Freire, A. D. Ebert, T. Kalisky, S. R. Quake and J. C. Wu, Nature Protocols 7 (5), 829-838 (2012). 42. F. S. Majedi, M. M. Hasani-Sadrabadi, S. H. Emami, M. A. Shokrgozar, J. J. VanDersarl, E. Dashtimoghadam, A. Bertsch and P. Renaud, Lab on a Chip 13 (2), 204207 (2013). 43. R. R. Hood, E. L. Kendall, D. L. DeVoe, Z. Quezado, M. Junqueira, J. C. Finkel and W. N. Vreeland, presented at the Microsystems for Measurement and Instrumentation (MAMNA), 2013, 2013 (unpublished). 44. C. W. Price, D. C. Leslie and J. P. Landers, Lab on a Chip 9 (17), 2484-2494 (2009). 45. R. Mariella, Jr., Biomedical Microdevices 10 (6), 777-784 (2008). 46. D. Xia, J. Yan and S. Hou, Small 8 (18), 2787-2801 (2012). 47. N. C. Cady, S. Stelick and C. A. Batt, Biosensors & Bioelectronics 19 (1), 59-66 (2003). 48. C. Xing, D.-F. Cui and L. Chang-Chun, presented at the Nano/Micro Engineered and Molecular Systems, 2006. NEMS '06. 1st IEEE International Conference on, 2006 (unpublished). 49. A. Bhattacharyya and C. M. Klapperich, Analytical Chemistry 78 (3), 788-792 (2006). 50. O. Z. Nanassy, P. V. Haydock and M. W. Reed, Analytical Biochemistry 365 (2), 240-245 (2007). 51. X.-W. Chen, Z.-R. Xu, B.-Y. Qu, Y.-F. Wu, J. Zhou, H.-D. Zhang, J. Fang and J.H. Wang, Analytical and Bioanalytical Chemistry 388 (1), 157-163 (2007). 52. C.-H. Chan, J.-K. Chen and F.-C. Chang, Sensors and Actuators, B: Chemical 133 (1), 327-332 (2008). 53. J. Kim and B. K. Gale, Lab on a Chip 8 (9), 1516-1523 (2008). 27 54. D.-W. Park, M. Hupert, M. Witek, B. You, P. Datta, J. Guy, J.-B. Lee, S. Soper, D. Nikitopoulos and M. Murphy, Biomedical Microdevices 10 (1), 21-33 (2008). 55. L. Liu, S. Yu, S. Yang, P. Zhou, J. Hu and Y. Zhang, Journal of Separation Science 32 (15-16, Sp. Iss. SI), 2752-2758 (2009). 56. J. Kim, H. Kido, J. V. Zoval, D. Gagné, R. Peytavi, F. J. Picard, M. Bastien, M. Boissinot, M. G. Bergeron and M. J. Madou, presented at the MOEMS-MEMS 2006 Micro and Nanofabrication, 2006 (unpublished). 57. E. Hoppmann and I. White, presented at the 26th Southern Biomedical Engineering Conference SBEC 2010, April 30-May 2, 2010, College Park, Maryland, USA, 2010 (unpublished). 58. A. Govindarajan, S. Ramachandran, G. Vigil, P. Yager and K. Böhringer, Lab on a Chip 12 (1), 174-181 (2012). 59. M. M. Packard, M. Shusteff and E. C. Alocilja, Biosensors 2 (4), 405-416 (2012). 60. K.-Y. Hwang, S. H. Kwon, S.-O. Jung, H.-K. Lim, W.-J. Jung, C.-S. Park, J.-H. Kim, K.-Y. Suh and N. Huh, Lab on a Chip 11 (21), 3649-3655 (2011). 61. K. A. Wolfe, M. C. Breadmore, J. P. Ferrance, M. E. Power, J. F. Conroy, P. M. Norris and J. P. Landers, Electrophoresis 23 (5), 727-733 (2002). 62. E. A. Oblath, W. H. Henley, J. P. Alarie and J. M. Ramsey, Lab on a Chip 13 (7) 1325-1332 (2013). 63. J. A. Lounsbury, A. Karlsson, D. C. Miranian, S. M. Cronk, D. A. Nelson, J. Li, D. M. Haverstick, P. Kinnon, D. J. Saul and J. P. Landers, Lab on a Chip 13 (7), 13841393 (2013). 64. M. Mahalanabis, J. Do, H. ALMuayad, J. Y. Zhang and C. M. Klapperich, Biomedical Microdevices 12 (2), 353-359 (2010). 65. H. Becker, N. Hlawatsch, R. Klemm and C. Gärtner, presented at the SPIE MOEMS-MEMS, 2011 (unpublished). 66. L. Xie, C. Premachandran, M. Chew and S. C. Chong, Advanced Packaging, IEEE Transactions on 32 (2), 528-535 (2009). 67. L. Xie, C. Premachandran, M. Chew, S. C. Chong, L. C. Wai and J. Lau, presented at the Electronic Components and Technology Conference, 2008. ECTC 2008. 58th, 2008 (unpublished). 68. C.-H. Weng, K.-Y. Lien, S.-Y. Yang and G.-B. Lee, Microfluidics and 28 Nanofluidics 10 (2), 301-310 (2011). 69. W. C. Hui, L. Yobas, V. D. Samper, H. Chew-Kiat, S. Liw, J. Hongmiao, C. Yu, C. Lin, L. Jing and L. Tit Meng, Sensors and Actuators A (Physical) 133 (2), 335-339 (2007). 70. H. M. Ji, V. Samper, Y. Chen, W. C. Hui, H. J. Lye, F. B. Mustafa, A. C. Lee, L. Cong, C. K. Heng and T. M. Lim, Sensors and Actuators A: Physical 139 (1), 139-144 (2007). 71. L. A. Marshall, L. L. Wu, S. Babikian, M. Bachman and J. G. Santiago, Analytical Chemistry 84 (21), 9640-9645 (2012). 72. J.-H. Wang, L. Cheng, C.-H. Wang, W.-S. Ling, S.-W. Wang and G.-B. Lee, Biosensors & Bioelectronics 41, 484-491 (2013). 73. C. Chiou, D. Shin, S. Hosmane, Y. Zhang and T. Wang, presented at the Micro Electro Mechanical Systems (MEMS), 2013 IEEE 26th International Conference on, 2013 (unpublished). 74. S. Petralia, R. Verardo, E. Klaric, S. Cavallaro, E. Alessi and C. Schneider, Sensors and Actuators B: Chemical 187, 99-105 (2013). 75. N. Hlawatsch, R. Klemm, C. Carstens, T. Brandstëtter, H. Becker, R. Elbracht and C. Gärtner, presented at the SPIE MOEMS-MEMS, 2012 (unpublished). 76. R. Klemm, N. Hlawatsch, T. E. Hansen-Hagge, H. Becker and C. Gärtner, presented at the Proceedings of SPIE, 2012 (unpublished). 77. A. J. Hopwood, C. Hurth, J. Yang, Z. Cai, N. Moran, J. G. Lee-Edghill, A. Nordquist, R. Lenigk, M. D. Estes and J. P. Haley, Analytical Chemistry 82 (16), 69916999 (2010). 78. M. D. Estes, J. Yang, B. Duane, S. Smith, C. Brooks, A. Nordquist and F. Zenhausern, The Analyst 137 (23), 5510-5519 (2012). 79. Y. Sun, J. Høgberg, T. Christine, L. Florian, L. G. Monsalve, S. Rodriguez, C. Cao, A. Wolff, J. M. Ruano-Lopez and D. D. Bang, Lab on a Chip 13, 1509-1514 (2013). 80. Q. Wu, J. M. Bienvenue, B. J. Hassan, Y. C. Kwok, B. C. Giordano, P. M. Norris, J. P. Landers and J. P. Ferrance, Analytical Chemistry 78 (16), 5704-5710 (2006). 81. J. Wen, C. Guillo, J. P. Ferrance and J. P. Landers, Analytical Chemistry 78 (5), 1673-1681 (2006). 82. J. Siegrist, R. Gorkin, M. Bastien, G. Stewart, R. Peytavi, H. Kido, M. Bergeron 29 and M. Madou, Lab on a Chip 10 (3), 363-371 (2010). 83. S. Bhattacharya, S. Salamat, D. Morisette, P. Banada, D. Akin, Y.-S. Liu, A. K. Bhunia, M. Ladisch and R. Bashir, Lab on a Chip 8 (7), 1130-1136 (2008). 84. C. N. Hughes-Chinkhota, M. Banda, J. M. Smolinski, R. A. Thomas, D. M. Petibone, J. D. Tucker and G. W. Auner, Sensors and Actuators B: Chemical 155 (2), 437-445 (2011). 85. K. A. Hagan, C. R. Reedy, M. L. Uchimoto, D. Basu, D. A. Engel and J. P. Landers, Lab on a Chip 11 (5), 957-961 (2011). 86. Q. Cao, M. Mahalanabis, J. Chang, B. Carey, C. Hsieh, A. Stanley, C. A. Odell, P. Mitchell, J. Feldman and N. R. Pollock, PLoS One 7 (3), e33176 (2012). 87. T. Baier, T. E. Hansen-Hagge, R. Gransee, A. Crombe, S. Schmahl, C. Paulus, K. S. Drese, H. Keegan, C. Martin, J. J. O'Leary, L. Furuberg, L. Solli, P. Gronn, I. M. Falang, A. Karlgard, A. Gulliksen and F. Karlsen, Lab on a Chip 9 (23), 3399-3405 (2009). 88. J. Kim, J. Elsnab, M. Johnson and B. K. Gale, presented at the MOEMS-MEMS, 2010 (unpublished). 89. R. Zhang, H.-Q. Gong, X. Zeng, C. Lou and C. Sze, Analytical Chemistry 85 (3), 1484-1491 (2012). 90. S. M. Berry, E. T. Alarid and D. J. Beebe, Lab on a Chip 11 (10), 1747-1753 (2011). 91. K. A. Hagan, W. L. Meier, J. P. Ferrance and J. P. Landers, Analytical Chemistry 81 (13), 5249-5256 (2009). 92. G. Xu, D. Y. San Lee, H. Xie, D. Chiew, T.-M. Hsieh, E. M. Ali, X. L. Looi, M.H. Li and J. Y. Ying, Biomicrofluidics 5, 034107 (2011). 93. X. Zhao and T. Dong, Analytical Chemistry 84 (20), 8541-8548 (2012). 94. T. Poeckh, S. Lopez, A. O. Fuller, M. J. Solomon and R. G. Larson, Analytical Biochemistry 373 (2), 253-262 (2008). 95. M. C. Breadmore, K. A. Wolfe, I. G. Arcibal, W. K. Leung, D. Dickson, B. C. Giordano, M. E. Power, J. P. Ferrance, S. H. Feldman, P. M. Norris and J. P. Landers, Analytical Chemistry 75 (8), 1880-1886 (2003). 96. M. Ritzi-Lehnert, R. Himmelreich, H. Attig, J. Claußen, R. Dahlke, G. Großhauser, E. Holzer, M. Jeziorski, E. Schaeffer and A. Wende, Biomedical 30 Microdevices 13 (5), 819-827 (2011). 97. D. Chen, M. Mauk, X. Qiu, C. Liu, J. Kim, S. Ramprasad, S. Ongagna, W. R. Abrams, D. Malamud, P. L. A. M. Corstjens and H. H. Bau, Biomedical Microdevices 12 (4), 705-719 (2010). 98. X. Qiu, D. Chen, C. Liu, M. G. Mauk, T. Kientz and H. H. Bau, Biomedical Microdevices 13 (5), 809-817 (2011). 99. J. Kim, M. Johnson, P. Hill, R. S. Sonkul, J. Kim and B. K. Gale, Journal of Micromechanics and Microengineering 22 (1), 015007 (2012). 100. L.-M. Fu, W.-J. Ju, R.-J. Yang and Y.-N. Wang, Microfluidics and Nanofluidics, 1-9 (2013). 101. W. H. Grover, A. M. Skelley, C. N. Liu, E. T. Lagally and R. A. Mathies, Sensors and Actuators B (Chemical) B89 (3), 317-325 (2003). 102. C. Yamahata, F. Lacharme, Y. Burri and M. A. Gijs, Sensors and Actuators B: Chemical 110 (1), 1-7 (2005). 103. T. Lemke, G. Biancuzzi, H. Feth, J. Huber, F. Goldschmidtböing and P. Woias, Sensors and Actuators A: Physical 168 (1), 213-222 (2011). 104. K. Li, C. Miville-Godin, F. Normandin, E. Roy and T. Veres, presented at the Proceedings of the 2nd European Conference on Microfluidics, 2010 (unpublished). 105. D. Snakenborg, H. Klank and J. P. Kutter, Microfluidics and Nanofluidics 10 (2), 381-388 (2011). 106. M. A. Unger, H. P. Chou, T. Thorsen, A. Scherer and S. R. Quake, Science (New York, N.Y.) 288 (5463), 113-116 (2000). 107. C. S. Effenhauser, G. J. Bruin, A. Paulus and M. Ehrat, Analytical Chemistry 69 (17), 3451-3457 (1997). 108. E. Delamarche, A. Bernard, H. Schmid, B. Michel and H. Biebuyck, Science 276 (5313), 779-781 (1997). 109. K. Hosokawa, T. Fujii and I. Endo, Analytical Chemistry 71 (20), 4781-4785 (1999). 110. D. C. Duffy, J. C. McDonald, O. J. Schueller and G. M. Whitesides, Analytical Chemistry 70 (23), 4974-4984 (1998). 111. J. C. McDonald and G. M. Whitesides, Accounts of Chemical Research 35 (7), 31 491-499 (2002). 112. T. Thorsen, S. J. Maerkl and S. R. Quake, Science 298 (5593), 580-584 (2002). 113. N. L. Jeon, D. T. Chiu, C. J. Wargo, H. Wu, I. S. Choi, J. R. Anderson and G. M. Whitesides, Biomedical Microdevices 4 (2), 117-121 (2002). 114. M. A. Eddings and B. K. Gale, Journal of Micromechanics and Microengineering 16 (11), 2396 (2006). 115. Y. Li, W. Jones, F. Rasti, I. Blaga, G. Bogdan, D. Eberhart, B. Kobrin, D. Lee, B. Nielsen and E. van Gelder, Lab on a Chip 11 (15), 2541-2550 (2011). 116. B. Vogelstein and D. Gillespie, Proceedings of the National Academy of Sciences of the United States of America 76 (2), 615-619 (1979). 117. R. Boom, C. J. A. Sol, M. M. M. Salimans, C. L. Jansen, P. M. E. Wertheim-Van Dillen and J. Van Der Noordaa, Journal of Clinical Microbiology 28 (3), 495-503 (1990). 118. M. P. Baker, A. Mitchell, C. Bridge, S. Martin, F. Holder, N. D. Pathirana, A. Jarvis and N. Butt, Biotechniques 31 (1), 142-145 (2001). 119. M. Fujiwara, F. Yamamoto, K. Okamoto, K. Shiokawa and R. Nomura, Analytical Chemistry 77 (24), 8138-8145 (2005). 120. M. G. Elgort, M. G. Herrmann, M. Erali, J. D. Durtschi, K. V. Voelkerding and R. E. Smith, Clinical Chemistry 50 (10), 1817-1819 (2004). 121. C. Xing, C. Da-Fu and L. Chang-Chun, presented at the 2006 1st IEEE International Conference on Nano/Micro Engineered and Molecular Systems, 18-21 Jan. 2006, Piscataway, NJ, USA, 2006 (unpublished). 122. D. S. W. Park, M. Hupert, M. Witek, B. You, P. Datta, J. Guy, J. B. Lee, S. Soper, D. Nikitopoulos and M. Murphy, Biomedical Microdevices 10 (1), 21-33 (2008). 123. L. Yobas, J. Hongmiao, H. Wing-Cheong, C. Yu, L. Tit-Meng, H. Chew-Kiat and K. Dim-Lee, IEEE Journal of Solid-State Circuits 42 (8), 1803-1813 (2007). 124. J. Hong Miao, V. Samper, C. Yu, H. Wing Cheong, L. Hui Jen, F. B. Mustafa, L. Ai Cheng, C. Lin, H. Chew Kiat and L. Tit Meng, Sensors & Actuators: A. Physical 139 (1-2), 139-144 (2007). 125. Z. Yi and W. Tza-Huei, presented at the 23rd IEEE International Conference on Micro Electro Mechanical Systems (MEMS 2010), 24-28 Jan. 2010, Piscataway, NJ, USA, 2010 (unpublished). 32 126. M. A. Burns, B. N. Johnson, S. N. Brahmasandra, K. Handique, J. R. Webster, M. Krishnan, T. S. Sammarco, P. M. Man, D. Jones and D. Heldsinger, Science 282 (5388), 484-487 (1998). 127. P. Neuzil, C. Zhang, J. Pipper, S. Oh and L. Zhuo, Nucleic Acids Research 34 (11), e77-e77 (2006). 128. M. A. Northrup, B. Benett, D. Hadley, P. Landre, S. Lehew, J. Richards and P. Stratton, Analytical Chemistry 70 (5), 918-922 (1998). 129. D. Trau, T. M. Lee, A. I. Lao, R. Lenigk, I.-M. Hsing, N. Y. Ip, M. C. Carles and N. J. Sucher, Analytical Chemistry 74 (13), 3168-3173 (2002). 130. P. Wilding, L. J. Kricka, J. Cheng, G. Hvichia, M. A. Shoffner and P. Fortina, Analytical Biochemistry 257 (2), 95-100 (1998). 131. J. Wu, R. Kodzius, K. Xiao, J. Qin and W. Wen, Biomedical Microdevices 14 (1), 179-186 (2012). 132. N. R. Beer, B. J. Hindson, E. K. Wheeler, S. B. Hall, K. A. Rose, I. M. Kennedy and B. W. Colston, Analytical Chemistry 79 (22), 8471-8475 (2007). 133. N. R. Beer, E. K. Wheeler, L. Lee-Houghton, N. Watkins, S. Nasarabadi, N. Hebert, P. Leung, D. W. Arnold, C. G. Bailey and B. W. Colston, Analytical Chemistry 80 (6), 1854-1858 (2008). 134. Z. Guttenberg, H. Müller, H. Habermüller, A. Geisbauer, J. Pipper, J. Felbel, M. Kielpinski, J. Scriba and A. Wixforth, Lab on a Chip 5 (3), 308-317 (2005). 135. D. Verdoy, Z. Barrenetxea, J. Berganzo, M. Agirregabiria, J. M. Ruano-López, J. M. Marimón and G. Olabarría, Biosensors and Bioelectronics 32 (1), 259-265 (2012). 136. D.-S. Lee, S. H. Park, H. Yang, K.-H. Chung, T. H. Yoon, S.-J. Kim, K. Kim and Y. T. Kim, Lab on a Chip 4 (4), 401-407 (2004). 137. E. T. Lagally, C. A. Emrich and R. A. Mathies, Lab on a Chip 1 (2), 102-107 (2001). 138. H. Lee, N. Han, I.-H. Choi and K.-H. Han, Biomedical Microdevices 15 (1), 9-15 (2013). 139. Y. Yu, B. Li, C. A. Baker, X. Zhang and M. G. Roper, Analytical Chemistry 84 (6), 2825-2829 (2012). 140. C. J. Easley, J. M. Karlinsey and J. P. Landers, Lab on a Chip 6 (5), 601-610 (2006). 33 141. M. U. Kopp, A. J. De Mello and A. Manz, Science 280 (5366), 1046-1048 (1998). 142. E. Lagally, J. Scherer, R. Blazej, N. Toriello, B. Diep, M. Ramchandani, G. Sensabaugh, L. Riley and R. Mathies, Analytical Chemistry 76 (11), 3162-3170 (2004). 143. L. A. Legendre, J. M. Bienvenue, M. G. Roper, J. P. Ferrance and J. P. Landers, Analytical Chemistry 78 (5), 1444-1451 (2006). 144. M. N. Slyadnev, Y. Tanaka, M. Tokeshi and T. Kitamori, Analytical Chemistry 73 (16), 4037-4044 (2001). 145. P. Vulto, C. Hermann, P. Zahn, U. Maier, G. Dame and G. Urban, presented at the Solid-State Sensors, Actuators and Microsystems Conference, 2009. TRANSDUCERS 2009. International, 2009 (unpublished). 146. J. Pipper, Y. Zhang, P. Neuzil and T. M. Hsieh, Angewandte Chemie 120 (21), 3964-3968 (2008). 147. S. K. Jha, G.-S. Joo, G.-S. Ra, H. H. Lee and Y.-S. Kim, Sensors Journal, IEEE 11 (9), 2065-2070 (2011). 148. J. J. Chen, C. M. Shen and S. J. Chen, Applied Mechanics and Materials 284, 883-887 (2013). 149. J. Y. Choi, Y. T. Kim, J. Ahn, K. S. Kim, D.-G. Gweon and T. S. Seo, Biosensors and Bioelectronics (2012). 150. S. Li, D. Y. Fozdar, M. F. Ali, H. Li, D. Shao, D. M. Vykoukal, J. Vykoukal, P. N. Floriano, M. Olsen and J. T. McDevitt, Microelectromechanical Systems, Journal of 15 (1), 223-236 (2006). 151. Y.-F. Lee, K.-Y. Lien, H.-Y. Lei and G.-B. Lee, Biosensors and Bioelectronics 25 (4), 745-752 (2009). 152. R. Pal, M. Yang, R. Lin, B. Johnson, N. Srivastava, S. Razzacki, K. Chomistek, D. Heldsinger, R. Haque and V. Ugaz, Lab on a Chip 5 (10), 1024-1032 (2005). 153. N. J. Panaro, X. J. Lou, P. Fortina, L. J. Kricka and P. Wilding, Biomolecular Engineering 21 (6), 157-162 (2005). 154. S. S. Yeung, T. M. Lee and I.-M. Hsing, Analytical Chemistry 80 (2), 363-368 (2008). 155. P. K. Yuen, L. J. Kricka, P. Fortina, N. J. Panaro, T. Sakazume and P. Wilding, Genome Research 11 (3), 405-412 (2001). 34 156. T. G. Kang, H. M. Ji, S. P. M. Tan, G. K. I. Tay, M. Y. D. Ang, S. Rafei, K. C. Tang, L. Zhang, X. Zhang and Y. Chen, presented at the Electronics Packaging Technology Conference (EPTC), 2010 12th, 2010 (unpublished). 157. W. Wan and J. T. Yeow, Biomedical Microdevices 14 (2), 337-346 (2012). 158. S. Quake, Electrophoresis 23, 1531-1536 (2002). 159. M. M. Kiss, L. Ortoleva-Donnelly, N. R. Beer, J. Warner, C. G. Bailey, B. W. Colston, J. M. Rothberg, D. R. Link and J. H. Leamon, Analytical Chemistry 80 (23), 8975-8981 (2008). 160. L. Mazutis, A. F. Araghi, O. J. Miller, J.-C. Baret, L. Frenz, A. Janoshazi, V. Taly, B. J. Miller, J. B. Hutchison and D. Link, Analytical Chemistry 81 (12), 4813-4821 (2009). 161. S. Kumar, M. A. Cartas-Ayala and T. Thorsen, International Journal of Thermal Sciences 67, 72-86 (2013). 162. J. M. Karlsson, T. Haraldsson, S. Laakso, A. Virtanen, M. Maki, G. Ronan and W. van der Wijngaart, presented at the Solid-State Sensors, Actuators and Microsystems Conference (TRANSDUCERS), 2011 16th International, 2011 (unpublished). 163. G. Kokkoris, D. Moschou, E. Mavraki, S. Chatzandroulis and A. Tserepi, presented at the 3rd Micro and Nano Flows Conference, Thessaloniki, Greece, 22-24 August 2011, 2011 (unpublished). 164. M. Hashimoto, F. Barany and S. A. Soper, Biosensors and Bioelectronics 21 (10), 1915-1923 (2006). 165. Y. Liu, C. B. Rauch, R. L. Stevens, R. Lenigk, J. Yang, D. B. Rhine and P. Grodzinski, Analytical Chemistry 74 (13), 3063-3070 (2002). 166. T. Ohashi, H. Kuyama, N. Hanafusa and Y. Togawa, Biomedical Microdevices 9 (5), 695-702 (2007). 167. M. Hashimoto, P.-C. Chen, M. W. Mitchell, D. E. Nikitopoulos, S. A. Soper and M. C. Murphy, Lab on a Chip 4 (6), 638-645 (2004). 168. R. H. Liu, J. Yang, R. Lenigk, J. Bonanno and P. Grodzinski, Analytical Chemistry 76 (7), 1824-1831 (2004). 169. Y.-K. Cho, J.-G. Lee, J.-M. Park, B.-S. Lee, Y. Lee and C. Ko, Lab on a Chip 7 (5), 565-573 (2007). 170. S. R. Jangam, A. K. Agarwal, K. Sur and D. M. Kelso, Biosensors and 35 Bioelectronics 42, 69-75 (2013). 171. J.D. Kim, C.Y. Park, S.Y. Kim, O.D. Gwak, D.J. Lee, Y.S. Kim and H.J. Song, International Journal of Multimedia and Ubiquitous Engineering, 7 (2), 395-402 (2012) 172. N. Pak, D. C. Saunders, C. R. Phaneuf and C. R. Forest, Biomedical Microdevices 14 (2), 427-433 (2012). 173. Q. Cao, M. C. Kim and C. Klapperich, Biotechnology Journal 6 (2), 177-184 (2011). 174. M. Bu, T. Melvin, G. Ensell, J. S. Wilkinson and A. G. Evans, Journal of Micromechanics and Microengineering 13 (4), S125 (2003). 175. S. Hardt, D. Dadic, F. Doffing, K. Drese, G. Münchow and O. Sörensen, Nanotechnology 1, 55-58 (2004). 176. G. Münchow, D. Dadic, F. Doffing, S. Hardt and K.-S. Drese, Expert Review of Molecular Diagnostics 5 (4), 613-620 (2005). 177. P. Auroux, P. Day and A. Manz, NSTI-Nanotechnol 1, 67-69 (2004). 178. J.-Y. Cheng, C.-J. Hsieh, Y.-C. Chuang and J.-R. Hsieh, Analyst 130 (6), 931-940 (2005). 179. W. Wang, Z.-X. Li, R. Luo, S.-H. Lü, A.-D. Xu and Y.-J. Yang, Journal of Micromechanics and Microengineering 15 (8), 1369 (2005). 180. A. G. Sciancalepore, A. Polini, E. Mele, S. Girardo, R. Cingolani and D. Pisignano, Biosensors and Bioelectronics 26 (5), 2711-2715 (2011). 181. A. Polini, E. Mele, A. G. Sciancalepore, S. Girardo, A. Biasco, A. Camposeo, R. Cingolani, D. A. Weitz and D. Pisignano, Biomicrofluidics 4, 036502 (2010). 182. D. Sugumar, A. Ismail, M. Ravichandran, I. Aziah and L. X. Kong, Biomicrofluidics 4 (2), 024103 (2010). 183. C. Zhang, H. Wang and D. Xing, Biomedical Microdevices 13 (5), 885-897 (2011). 184. B. Chia, S.-A. Yang, M.-Y. Cheng, C.-W. Lin and Y.-J. Yang, Journal of Mechanics 27 (03), 357-364 (2011). 185. Y. Chenying, W. Wei and L. Zhihong, presented at the 2009 4th IEEE International Conference on Nano/Micro Engineered and Molecular Systems (IEEENEMS 2009), 5-8 Jan. 2009, Piscataway, NJ, USA, 2009 (unpublished). CHAPTER 2 BUBBLE INCLUSION AND REMOVAL USING PDMS MEMBRANE-BASED GAS PERMEATION FOR APPLICATIONS IN PUMPING, VALVING AND MIXING IN MICROFLUIDIC DEVICES Journal of Micromechanics and Microengineering, (2009) 19, Bubble inclusion and removal using PDMS membrane-based gas permeation for applications in pumping, valving and mixing in microfluidic devices. M. Johnson, G. Liddiard, M. Eddings, B. Gale. © Owned by IOP, published by IOP, 2009. Reprinted with kind permission of IOP Publishing. 37 38 39 40 41 42 43 44 45 CHAPTER 3 NUCLEIC ACID EXTRACTION SYSTEM 47 3.1 Introduction Bioassays for detection or quantification of organisms often rely on nucleic acid detection. The varied conditions of the sources of target genetic material (i.e., prokaryotic or eukaryotic cells in blood, water, stool, soil, and others) and the stringent purity requirements of analysis techniques most often force system-specific sample preparation operations, which often require time consuming and expensive manual handling.1 Self-contained lab-on-a-chip systems capable of preparing genetic material out of raw biological samples have been achieved with varying levels of success as presented in Chapter 1. The benefits envisioned by making nucleic acid analysis more readily available will be closer to realization when steps are taken towards universal sample preparation with the ability to take samples from various sources (blood, bacteria cultures, and virus samples) and selectively extract and purify nucleic acids (DNA or RNA). 3.1.1 Extraction System Overview In the work described in this chapter, efforts to build and test a system for automated NA purification from a variety of biological samples are described. The nucleic acid is purified using an integrated microfluidic fluid handling chip and a solidphase extraction chip. The overall extraction process operates by pumping the sample and various reagents using integrated metering/mixing reservoirs through an extraction filter on the extraction chip where the nucleic acid is selectively bound and released depending on buffer conditions. Altering the amounts, types, and sequences of reagents allows selective adjustment of the protocol to handle different sample types and control the output as DNA or RNA. These adjustments are easily accomplished by modifying a 48 control program and input reagents. This chapter is composed of several sections that report on the development of the microfluidic NA extraction system. First, the design and fabrication of the individual components are discussed. Details are provided of the fabrication of a three-layer polydimethylsiloxane (PDMS) microfluidic chip designed to perform the main fluid handling tasks. A multilayer mold and rapid prototyping technique is employed in the fabrication of integrated microchannels, valves, and metering diaphragm pumps. Second, a separate module is fabricated to incorporate the silica membrane used as the disposable solid phase for nucleic acid extraction. Calculations and experiments are performed to develop the filter chip architecture. The nucleic acid extraction efficiency of the solid phase filters are characterized with respect to parameters such as flow rate and filter material. Third, the ability of the system to handle different sample inputs is demonstrated by extracting DNA from both a stock DNA solution and whole blood, and by recovering RNA form known RNA solution standard, living E. Coli cells, and FMDV virus cultures. Following extraction, the purified genetic material can be mixed with other reagents required for analysis techniques such as electrochemical detection or transfer to a PCR module as will be discussed in Chapter 4. 3.2 Materials and Methods 3.2.1 Chip Design and Fabrication 3.2.1.1. PDMS chip overview The heart of the NA extraction system is a microfluidic fluid handling chip fabricated using a three-layer approach with a fluid layer, a flexible membrane, and a pneumatic control layer see Figure 3.1. The pneumatic layer forces membrane actuation 49 Fluid Layer Membrane Pneumatic Control Layer Figure 3.1: Diagram of 3-layer microfluidic chip. to operate the embedded pumps and valves. The chip design incorporates eight fixedvolume, on-chip metering diaphragm pumps similar to those reported by Grover et al.2 These pumps interface with 17 fluid inlet ports and two module access ports to facilitate adaptable protocols for nucleic acid extraction. Different pump volumes allow a range of input volumes to accommodate the mixing ratios of extraction protocols. Figure 3.2 is a schematic of where the valves and pumps are located on the chip layout. Locations of other features important to chip function are also indicated. 3.2.1.2 Valve design and operation On-chip valves operate by applying pressure to the control layer which pushes the membrane against the valve seat, sealing the channel. Vacuum applied to the control channel pulls the membrane away from the valve seat, and fluid is then free to flow through the open channel. Valve function is described by the schematic in Figure 3.3. 50 Figure 3.2. Diagram of functional features incorporated onto PDMS extraction chip. Numerically labeled features are on-chip valves, with numbers indicating the associated off-chip solenoid valve. Features A-J are the metering pumps and mixing chambers, KM are sample and reagent input ports, N is the waste output, O highlights postextraction processing section, P indicates extraction process output and external module connection, Q is the location of the solid phase chip, and P is the direct pressure line for drying and channel clearing steps. 51 Figure 3.3: Diagram of on-chip valve and metering pump operation. A) chip with membrane at equilibrium. B) Pressure applied to the Pump, P and valves V1 and V2. C) Vacuum applied to V2 and P2 opens a channel for fluid to enter the inlet port and fill the pump reservoir. D) V2 is closed with pressure, and V1 is opened with Vacuum. E) Fluid in the reservoir is forced out through the channel passing V1. On-chip valves are actuated by switching the control layer portion of the valve between pressure and vacuum using an off-chip electrically actuated three-way solenoid valve. The valve design has the channel opening into a circular area with a gap in the middle. The widened circle increases the area exposed to the pressure difference and reduces the pressure required to overcome the innate stiction of the membrane to the valve seat. The control channel side of the valve is made a slightly larger circle than in the fluid channel side to reduce alignment tolerance requirements to produce a high yield 52 of operational valves over the large are of the PDMS chip. 3.2.1.3 Pumps Metering diaphragm pumps pull fluid into the chip through the access ports. In the default state of the pump, control pressure pushes the flexible membrane against the top of the fluid reservoir, reducing the volume in the fluid side of the reservoir to the space around the sharp corners where the membrane cannot extend. When vacuum is then applied to the control layer, the membrane is pulled to the bottom of the reservoir, and fluid is drawn in to fill the evacuated space. The fluid is forced out of the reservoir pump by again applying pressure to the control layer, forcing the membrane up and displacing the fluid which exits through an open channel. The fluid layer and the pneumatic control layer are created by casting PDMS (Slygard 184, Dow Corning). Molds are made using a xurographic technique3 with features being cut from a vinyl tape (Instachange, 3M, St. Paul, MN) on a knife-plotter (Graphtec, Irvine CA). After the features were cut in the vinyl tape, they were transferred to a PMMA substrate. This rapid prototyping fabrication technique allows the formation of molds with fluid channel widths as low as 200 μm, without the need for conventional microfabrication or micromachining methods, in as little as 10 minutes. This fabrication method was also improved for the fabrication of the fluid handling chip by including larger mesoscale 3D structures to interface with a range of larger fluid volumes (50-1000 μl) required for the extraction process. The pneumatic control layer mold was designed with features of different heights to increase diaphragm pump volumes and improve valve performance. Different mold heights for valve controls were achieved by stacking cut tape layers. Large reservoirs for the on-chip fixed 53 volume diaphragm pumps were made by bonding laser-cut PMMA features to the corresponding tape features using a cyanoacrylate adhesive (421, Loctite, Westlake, OH). Following mold fabrication, PDMS was mixed at a ratio of 10:1 with curing agent, degassed in a vacuum chamber for 15 minutes, and poured into the molds. The filled molds were cured in a 60 ˚C oven until fully cured (more than 2 hours). The cured fluid and pneumatic control layers were removed from the molds, and pneumatic and fluid connection ports were cored using a 1.5 mm diameter biopsy punch (Uni-Punch, Clear Lake, WI). The fluid and control layer were aligned optically under the microscope prior to any bonding operations and four alignment holes were cored through the PDMS to allow alignment of the two layers through the opaque membrane. A thin silicone sheet (Bisco HT-6135, Stockwell Elastomerics, Philadelphia, PA) was used for the membrane layer. The membrane was first bonded to the pneumatic control layer using plasma treatment bonding which operates by creating hydrophilic, chemically active functional groups on the PDMS surface that bind to other activated groups on the mating surface.4 The areas to be bonded on the membrane and control layer were exposed to the discharge from an air plasma surface treater (Enercon Industries, Menomonee Falls, WI) so that all surfaces were exposed for approximately 1 minute with the treater head approximately 0.5 inches above the surface. The treated surfaces were then brought into contact and pressed together with a rubber roller. The control layer/membrane assembly was then incubated for 20 minutes at 60 ˚C. Following the first bonding step, holes were cored through the membrane to open the alignment holes and to produce three cross-membrane vias. One provides the port through which pressurized air can directly enter the chip for drying steps or to aid in 54 emptying channels. The other two vias route the waste channel to the other side of the membrane in order to cross another channel without interference. When bonding the fluid layer to the membrane/pneumatic control layer assembly, steps were taken to prevent unwanted bonding at the valve seat and over the large surface areas of the metering diaphragm pumps. Covering the PDMS layer to be bonded with a masking material during corona discharge treatment blocks areas from exposure to the corona plasma so that only desired surfaces are activated, as shown in Figure 3.4. Several masking materials were shown to be effective: rubylith electrostatic film, 3M Instachange tape, mineral oil, and liquid spin-on glass. Subsequent removal of the mask and contact with another activated PDMS surface is shown to produce selective bonding, with the masked areas remaining unbound. 3M Instachange tape was used as the mask for this device with the mask shapes cut on the knife plotter. Masks were manually placed over the areas that were not to be bonded. Again the surfaces were treated with air plasma for approximately 1 minute. Following treatment, the mask material is removed and alignment pins are placed through the alignment holes. The two layers to be bonded are then brought into contact, compressed, and incubated as described above. 3.2.2 Solid Phase Extraction Chip Fabrication The purification of target nucleic acid from background compounds that would impede analysis is performed using solid phase extraction techniques with a silicate substrate acting as the solid phase. For nucleic acid purification on a microfluidic chip, the solid phase needs to be easy to integrate into the flow control structure of the chip and be disposable to avoid cross-contamination between extraction assays. The first generation of the integrated solid phase module (SPE chips) consists of a 55 Figure 3.4. Masked plasma bonding process. The PDMS face is covered with a mask that exposes the surfaces to be bonded. The surface is then activated by exposing it to corona discharge. The mask is removed and the surface is brought into contact with another activated PDMS surface. Areas where two activated surfaces contact are irreversibly bonded, while other areas remain nonbonded, even if they are in direct contact. PDMS casing enclosing a stack of silica beads (Sigma-Aldrich Corp. St. Louis, MO). The PDMS casing is molded in four layers, each layer bonded to the others using a plasma surface treatment activation method. The beads are held in the solid phase chamber between two glass microfiber filters (GF/D, Whatman, Maidstone, Kent, UK) at the inlet and outlet of the solid phase module. The next generation of silica chips omitted the beads in favor of silicate filter membranes only because of fabrication difficulty and the calculation that the microfiber filter material had approximately an order of magnitude more surface area per volume than the beads (5.6 x 104 vs 8.8 x 103 m2/m3). Disposable filter chips shown in Figure 3.5 were fabricated by sandwiching the 56 Figure 3.5. Diagram of the construction of the glass filter solid phase chip. filter material between two laser-cut PMMA blocks. A CO2 laser (Universal Laser Systems) was used to cut the access ports, filter pocket, and microchannel in the PMMA blocks. The filter pocket was created by laser etching a well into the PMMA with a shallower ring around the lip which serves to seal off the edges and force the fluid to pass through the filter. The two halves of the block are sealed together using a double-sided pressure sensitive adhesive (467 MP, 3M). Tubing connections are made by inserting 22 gauge stainless steel tubing into the access ports and bonding using a UV cure adhesive (3106, Loctite, Westlake, OH). These connectors are press-fit into access ports located on the PDMS fluidic control chip. 3.2.2.1 Filter material Two types of glass filters were used as the solid phase membrane: borosilicate glass and quartz (GF/D and QM/A, respectively). To test the extraction efficiency of the different filter materials, 100 ng and 200 ng samples of prepurified human genomic DNA were added to a chaotropic salt buffer (Qiagen, Venlo, Netherlands) and passed through the filters at 100 μl/min. The filters were then rinsed with 80% ethanol and dried for 2 57 minutes with pressurized air at 10 psi. Finally, the DNA is eluted from the filters with 100 μl of purified water. The collected DNA was quantified using the fluorescence method described in later sections. 3.2.2.2 Flow rates To determine the effect of flow rate on the extraction efficiency of the NA extraction chip, a syringe pump was loaded with a known concentration of DNA sample in a binding buffer. The syringe was connected to the solid phase filter chip and the fluid pumped through the filter at different flow rates. The residence time was kept the same for all flow rates by pumping the samples at higher flow rate samples several times through the filter. Consequently, the molecules at all flow rates would spend the same amount of time in proximity to the filter surface. Following the binding step, the filters were washed with ethanol to remove excess salt and impurities. The filter was then dried by pumping air through the filter at a pressure of 10 psi for 2 minutes to remove residual ethanol, and then the DNA was eluted with 100 μl of purified water. The quantity of DNA eluted was determined using fluorescence as described in later sections. 3.2.3 Nucleic Acid Quantification Methods To determine the effectiveness of the microfluidic nucleic acid extraction, the quantity of nucleic acid recovered from the system must be measured. Several methods have been developed to quantify the amount of nucleic acid, and three of the primary techniques were used to verify results in this work. The first method uses a spectrophotometer (Nanodrop, Wilmington, DE) to measure the nucleic acid absorption of UV light. Nucleic acid has a specific UV 58 absorption spectrum, with the wavelength of interest being 260 nm. Light is passed through a sample and the instrument measures the absorption and relates it to the concentration of nucleic acid in the sample. Because of the specific absorption profile, the spectrum can also be used to determine the purity of the nucleic acid by examining the ratio of characteristic absorption wavelengths. For the case of nucleic acids, a ratio of absorption at 260 nm to 280 nm approximately between 1.8 and 2.0 indicates sufficiently pure nucleic acid. Contaminants left in the sample that absorb UV light in this spectral range can negatively impact readings, and care must be taken to ensure that the absorption spectrum from the carrier buffer be subtracted from the sample spectrum so that only the absorption of nucleic acids is taken into account. A second method of nucleic acid quantification uses fluorescent dyes. Ribogreen and Picogreen (both from Invitrogen) are two such fluorescent dyes that give a linear response to the presence of nucleic acid and were used to quantify DNA and RNA, respectively. The quantification process begins with making a standard curve by diluting a known concentration of DNA or RNA to concentration values that surround the region of interest. The linear relation that results can be used to directly determine the nucleic acid concentration of an unknown sample from its level of fluorescence. For the studies in this work, fluorescence was measured on a 96-well microplate reader (BioTek). The final quantification method used in this work is using real-time PCR. When a fluorescent dye is added to the PCR reagent mix, the fluorescence emission of the amplified product can be monitored after every amplification cycle. An increase in fluorescence indicates more strands have been duplicated, with the fluorescence making an exponential increase in intensity after a certain number of cycles, referred to as the 59 crossing point. Samples that begin with a larger quantity of target nucleic acid molecules will have an earlier crossing point than samples with lower concentration. Nucleic acid concentrations can be determined by comparing the cycle number of unknown sample crossing points to crossing points from known concentrations. The instrument used for these measurements is the LightCycler (Roche) which monitors the amplification of the DNA in real time using an intercalating fluorescent dye. All three quantification methods have drawbacks and benefits. Spectrophotometry has low signal to noise ratios at low concentrations and is thus unreliable for samples that are dilute. However, spectrophotometry is very rapid and requires little handling, modification, or extra reagents. Fluorescence readings are adversely affected by the presence of autofluorescing agents and chemicals that interfere with the binding of the dyes to the molecules of interest, but fluorescence is superb at handling very low concentrations. PCR is very sensitive to contamination by PCR inhibitors and requires specialized equipment and strict laboratory techniques, but gives the best indication of the usability of the extracted sample for downstream analysis since much of the nucleic acid analysis for microfluidic systems includes PCR amplification of the target prior to detection and study. To ensure the accuracy of the three measurement systems described above, samples with known quantities of nucleic acids were compared using the three nucleic acid quantification methods: spectrophotometry, fluorescence, and real-time PCR crossing point analysis. 60 3.2.4 Extraction Protocols The standard chemical protocols to extract nucleic acid from a biological sample follow the same basic steps, independent of sample source and target molecule. Differences will be discussed later. First, the sample is mixed with reagents that work to release the nucleic acid into solution, which usually involves the lysis of target cells. The lysis buffer also contains a high concentration of a chaotropic salt (guanidinium thiocyanate or guanadinium hydrochloride) which will act as the selective binding agent in subsequent steps. Following lysis, the sample mixture with binding salts contacts the solid phase silica surface. The nucleic acid is reversibly bound to the solid phase and the remaining sample mixture is discarded. A wash step is then incorporated where an alcohol-based solution is flowed over the solid phase to rinse off any remaining cellular debris or proteins that might interfere with downstream amplification or analysis. After the wash step, the filter is dried, leaving only nucleic acid bound to the surface of the filter. The nucleic acid is then released from the solid phase using a buffer or pure water. For extracting RNA from viruses, the basic protocol is followed with a few additions. Usually a stabilizing reagent (such as B-mercaptoethanol) is included with the lysis buffer to protect the more fragile RNA from degrading during the extraction process. Also, following the wash step, RNAse-free DNAse is used to digest any DNA that might have been extracted during the purification process, leaving only RNA for analysis. To extract RNA from bacteria, the cell lysis step also needs to include a method for breaking the more robust cell wall. In this case lysozyme is used to perform an enzymatic process to disrupt the cell wall, enhancing the bacterial lysis process. 61 Extracting nucleic acid from blood samples provides an additional challenge. Proteins that are abundant in blood samples interfere with the extraction process and hinder amplification of target DNA strains by PCR. Consequently, blood extraction protocols begin with a proteinase digestion to remove proteins. 3.2.5 Microfluidic Control System Performing multiple purification protocols on a single chip is accomplished by using a programmable instrument that controls the on-chip flow sequence. A manifold that interfaces directly with the PDMS chip is mounted onto an instrument that electrically triggers the pneumatic operations for the on-chip fluid handling tasks. The manifold is a plate embedded with barbed fittings at locations that match the pneumatic control ports on the bottom side of the PDMS chip (see Figure 3.6). The chip is loaded onto the instrument by pushing the chip down onto the manifold, creating press-fit seals between the manifold fittings and pneumatic ports. The instrument contains 24 electrically actuated mini solenoid valves (Lee Company). Control of the valve actuation is accomplished using a USB I/O digital controller (Elexol) connected to a PC running a LabView control program. The LabView program (see Figure 3.7) has a mode for manual manipulation of fluid on the chip used for protocol development and troubleshooting. Up to 4 protocol sequences can also be stored and implemented in an automatic mode, allowing for hands-free extraction of nucleic acids. A cleaning step was included in the program to allow for multiple tests on a single PDMS chip. A solution of RBS Neutral (Sigma-Adrich) is flushed twice over the system followed by three flushes of nuclease-free water (Qiagen). The programmable nature of 62 Figure 3.6: Barbed fitting manifold for interface between PDMS chip and control instrument facilitating rapid microfluidic chip replacement. Figure 3.7: Screen shot of LabView control program showing manual mode (left) and 4 selectable automated protocol sequences (right). 63 the device allows for other disinfectants and cleaning agents to be introduced into the PDMS device, and the entire PDMS chip is also manifold-mounted for easy replacement for sensitive tests where disposable units are desired. 3.2.6 Surface Modification Further optimization of the RNA extraction system was done by employing several surface coatings to PDMS chip to mitigate nonspecific binding of target molecules to the channel surfaces. Surfaces were coated with sodium polyphosphate (NaPP), P-108 pluronic, P-20, and polyvinyl periladone (PVP). To perform these coatings, approximately 1% w/v solutions of the coating chemicals were continuously pumped through the extraction chip for several hours, and finally the chip filled with the coating solution was put in a 60 ˚C oven overnight. Following coating of the channels, known 100 ng samples of RNA were extracted in the microfluidic chips and compared to standard extraction techniques. 3.2.7 RNA Extraction Initial extraction protocols developed for the system were developed for extraction of RNA. Optimization of protocol parameters progressed by adapting the steps contained in the RNeasy Mini Kit (Qiagen 74104). Volumes were adjusted to keep mixing ratios the same and process steps were altered to incorporate chip-based pumping through the filter membrane instead of centrifugation. For initial verification of RNA extraction process parameters, known concentrations of RNA from a standard stock (Invitrogen) were used as the sample. An RNA-containing solution was placed at the sample inlet port of the PDMS device. A protocol (Appendix B) for on-chip mixing of the sample with reagents was used to bind 64 the RNA to the glass filter, wash off any contaminants, and subsequently release the RNA in an elution of nuclease-free water. RNA extraction efficiencies were determined using a fluorescent Ribogreen RNA quantification kit obtained from Invitrogen. Fluorescence data were obtained using a 96-well plate reader (Bio-TEK). RNA extraction was determined to be successful if the system produced a linear response to the input concentration of RNA at levels on the order of what was seen by the control (Qiagen commercial spin kit). It was not anticipated that the microfluidic system would exceed the extraction efficiency of the commercial kit. Rather, the innovation of the microfluidic system would be demonstrated in its versatility and ease of use. Following successful demonstration of RNA from prepurified samples from the system, RNA from raw biological samples was demonstrated by performing lysis and RNA extraction from E. Coli bacteria cells (New England Biolabs). The cells were incubated in cell culture medium and growth was monitored using optical density measurements to ensure viability. The E. Coli cells were then loaded into the automated system where they were enzymatically lysed, and RNA was extracted and quantified. Viral RNA was also extracted from cultured samples for the detection of Foot and Mouth Disease Virus (FMDV). FMDV samples in tissue culture fluid were used as the sample. The extraction system was programmed according to the virus protocol listed in Appendix B. The presence of extracted viral RNA was verified using spectroscopy on a nanodrop instrument. Success in this case was determined by the ability to amplify the purified RNA by means of RT-PCR, and detected VIA gel electrophoresis. 65 3.2.8 DNA Extraction The DNA extraction process was also established with prepurified human genomic DNA used as the input sample. For these tests 100 μl samples of purified DNA with total DNA mass of 0, 50, 100, 200, and 400 ng were drawn into the device and mixed with the appropriate reagents. Purified genomic DNA samples were provided as a gift from Carl Wittwer's lab in the Department of Pathology at the University of Utah. Other reagents were obtained from the DNeasy Blood & Tissue Kit (69504 Qiagen). To perform the extraction tests, 100 µl of the known sample was drawn into a pipette tip, which was then inserted into the inlet port of the microfluidic chip. The DNA extraction protocol (Appendix B) was initiated, pulling the sample into the microfluidic chip, mixing it with reagents, and extracting the DNA using the same basic binding, washing, and eluting steps described for RNA extraction. The DNA extraction was quantified using fluorescence (Picogreen, Invitrogen) and verified by performing PCR on the eluate using a LightCycler instrument (Roche). Following verification of the ability to recover DNA, whole blood from a volunteer was used as the extraction sample and DNA was again extracted in the system and quantified using fluorescence and PCR. The criterion for successful DNA extraction on the microfluidic system again was set as the ability to elicit a linear response to input concentration, at levels on the order of those seen in commercial standards. 3.3 Results 3.3.1 Microfluidic Chip Fabrication The microfluidic chip was successfully fabricated from PDMS. The chip and control platform are shown in Figure 3.8. Forty-one pneumatic valves are connected to 66 Figure 3.8: Completed PDMS extraction chip and control mechanism. This compact device uses electric vacuum and pressure pumps along with a bank of solenoid valves to control a PDMS microfluidic chip. A LabView control program allows easy adjustment of the protocol to handle a variety of sample inputs. the on-chip membrane valves through a manifold connector. An on-board compressor and vacuum generator supply the different air pressures required to operate the on-chip pumps and valves. Reagent and waste reservoirs also connect to the PDMS chip through tubing with barbed fittings. A notebook computer with a LabView program control the processing sequence. Masked corona bonding was successful at selectively bonding the layers together except at the interface between the membrane and the valve seat. All 41 on-chip valves were successful at sealing the channels for valve supply pressures greater than pumping pressures. Volumes pumped by the metering pumps were measured using a micropipettor and were found to vary from the desired volume by a maximum of 5% for the larger pumps and a minimum of 2.7% for the smaller pumps. 67 3.3.2 Silicate Solid Phase Membrane Extraction Yield Characterization and Improvement In order to improve the extraction efficiency of nucleic acid recovery in the microfluidic purification system, the effect of several parameters were adjusted and studied: solid phase material, flow rates, elution volume. 3.3.2.1 Filter material Two different types of filter material were used as the solid phase in the construction of the solid phase extraction chips: borosilicate glass microfiber filters and quartz filters. The borosilicate glass filters consistently had higher extraction efficiencies than the quartz as shown in Figure 3.9. 3.3.2.2 Flow rates The effect of sample flow rate through filters made of both borosilicate glass and quartz was tested. Flow rates from 1 to 50 μl/min were tested and the results are displayed in Figure 3.10. Surprisingly, the lower flow rates yielded significantly less DNA than higher flow rates. As the flow rate increases, from zero, the yield increases to a point and then seems to approach an asymptote or even decrease at the high end of the flow rates tested. As this seems to disagree with results published in the literature; it is likely that the physical geometry of the filter chip inhibits DNA binding at low flow rates, perhaps by allowing the fluid to pass through a single, low-resistance flow path through the filter and reducing the effective surface area onto which the nucleic acid can be bound. More investigation is recommended to determine the cause of the discrepancy, but for the current work, the flow rates achieved by the PDMS chip through the filter corresponded to higher extraction efficiencies, so no further optimization was attempted. 68 DNA recovered (ng) Extraction yield by filter type 100 90 80 70 60 50 40 30 20 10 0 quartz glass 200 ng 200ng 100 ng 100ng Figure 3.9: Comparison of extraction efficiencies of quartz and glass filters. Effect of Flow Rate on DNA Recovery 40 DNA out (ng) 35 30 25 quartz 20 glass 15 quartz outlier 10 glass outlier 5 0 0 10 20 30 40 50 60 flow rate (ul/min) Figure 3.10: Figure showing affect of elution flow rate and different solid phase material on extraction efficiency. 69 3.3.2.3 Elution volume When the DNA is eluted from the solid phase filter, only a certain fraction of the bound DNA is released. Elution profiles have shown 5, 6 that earlier elution fractions have a higher concentration of nucleic acid, with the rate of nucleic acid release decreasing as more fluid passes the solid phase. Elution fractions were examined by collecting two sets of five 35 μl (approximate) elution volumes from the extraction chip. The first elution volume was significantly less as the internal dead volume of the system had to be filled before the fluid actually went through the filter. Results in Figure 3.11 indicate that the peak concentration of elution happens between 15 and 70 μl with the amount of DNA decreasing after that. Using only two volumes from the elution pump chamber will thus yield the highest concentration of nucleic acid, but additional 35 μl fractions will continue to elute more DNA from the SPE filter. Elution fraction recovery DNA Recovered (ng) 3 2.5 2 1.5 1 0.5 0 14 36 71 131 185 Series1 0.812380345 2.010848756 1.660497766 1.133375877 0.629865986 Series2 0.734524569 2.781110402 1.634971283 1.01531589 1.21952776 Figure 3.11: Elution fractions. Bars represent the amount of DNA recovered with each successive 35 μl elution volume. 70 3.3.3 Nucleic Acid Quantification Methods Fluorescence, spectroscopy, and PCR crossing point analysis all correctly identified the amount of DNA, and can thus be used interchangeably to quantify the extraction results. The inclusion of samples extracted using the microfluidic system and their successful quantification using the different methods indicates that the system adequately eliminates chemical contaminants, as shown in Figure 3.12. We see that all three quantification methods give strong qualitative evidence to the presence and purity of nucleic acids, as well as correlated, accurate quantitative measurements. 3.3.4 Microfluidic Control System An instrument was constructed that contains a vacuum pump, a pressure pump, 24 solenoid valves, and a digital I/O controller. The controller connects via USB to a computer running a LabView control program. Protocol sequences were developed to adapt to the established chemistry used to extract RNA or DNA from different samples as explained above. Table 3.1 highlights the differences between extraction protocols on using the automated system. 3.3.5 Surface Modification Binding of the nucleic acid to the PDMS surfaces was minimized using various coatings and solution stabilizers. Results from such coating tests are presented in Figure 3.13 and extraction yields are compared to the commercial system. RNA recovery yields from a sodium poly-phosphate (NaPP) coated system approached those from the commercial spin kit. However, these high-yield products displayed some NaPP contamination that interfered with reverse-transcript polymerase chain reaction (RTPCR) analysis methods. Other coatings actually reduced the extraction efficiency of the 71 Measured DNA concentration (ng/μl) DNA Measurement Method Verification PCR* Spectr.* Fluor.* PCR Spectr. Fluor. 90 80 70 60 50 40 30 20 10 0 0 10 20 30 40 50 60 70 Known DNA concentration (ng/μl) Figure 3.12: Quantification of known quantities of DNA by three different methods shows a close 1:1 relation between known concentration and measured concentration for all measurement techniques. Blue data points were measured with known concentrations spiked into a negative control extraction system output to verify that the system does not output trace contaminants that affect nucleic acid measurement. The dotted line indicates known values. Table 3.1: Summary of protocol steps as performed by the microfluidics automated control program Protocol RNA extraction from E. Coli DNA extraction RNA extraction (with DNA extraction from blood from FMDV hybridization) RNA extraction Number of process steps 42 64 112 166 (224) 85 Process time (minutes) 7.2 21.6 17.5 53.2 (62) 9.3 4 6 6 8(12) 6 Number of reagents required 72 Nanograms RNA Extracted PDMS Chip Surface Coating Comparison 50 45 40 35 30 25 20 15 10 5 0 Spin Kit NaPP Uncoated P-20 PVP f-108 Pleuronic Figure 3.13: RNA recovery yields from automated extraction system with different chip surface coatings. chip. NaPP could still be beneficial for applications that do not require PCR as the detection method. It was used with some success in tests performed by collaborating researchers (Integrated Explorations, Guelph, ON, Canada) using the microfluidic device to extract RNA for electrochemical detection. Coatings for the microfluidic chip were not used for further nucleic acid extraction tests because the improvement with NaPP came with an unacceptable incompatibility with PCR. 3.3.6 Nucleic Acid Extraction 3.3.6.1 Extraction of RNA The device was used to demonstrate successful RNA recovery at yields approaching results attained by a commercial spin kit (Qiagen). Known concentrations 73 of RNA from an RNA standard stock (Invitrogen) were used as the sample. RNA extraction results are determined using a fluorescent Ribogreen RNA quantification kit obtained from Invitrogen. Fluorescence data are obtained using a 96-well plate reader. The device was used to demonstrate successful automated recovery of prepurified RNA at yields approaching results attained by a commercial spin kit. RNA is extracted from different starting concentrations with the microfluidic system demonstrating an extraction efficiency of approximately 50% compared to the commercial standard as seen in Figure 3.14. 3.3.6.2 Extraction of RNA from E. Coli bacteria The ability of the extraction system to purify bacterial RNA was tested by using E. Coli 0157:H7 verotox negative bacteria as the sample. Bacteria were obtained from Figure 3.14: Nucleic acid extracted from the automated microfluidic system (diamonds with error bars) vs a commercial kit (squares). 74 New England BioLabs and cultured in cell growth medium until they reached the exponential growth phase. The number of bacterial cells was measured using optical density techniques. Seven million bacteria cells were used as the sample input to the microfluidic extraction devices. An automated protocol listed in Appendix B performed the necessary steps to lyse, extract, and purify the bacterial RNA, similar to the abovementioned protocols, with the exception of the addition of a lysozyme to break the cell walls and the need for RNA protection reagents. A total of 193 ng of RNA was recovered as quantified using fluorescence techniques. As a secondary evaluation of the ability of the system to purify RNA from bacteria samples, the micfoluidic system was tested at Integrated Explorations Inc., a biolab in Guelph, Ontario Canada. The robust nature of the system was demonstrated by extracting RNA from raw biological samples (E. Coli cells) as shown in Table 3.2. A 10fold increase in the number of cells proved to produce less than double the recovery of RNA, indicating a nonlinear relationship between amount of bacteria supplied and the extraction yield. Factors that could possibly contribute to this include incomplete lysis of all cells, saturation of the solid phase during binding, RNA loss during washing step, or incomplete elution of target molecules. Further investigation and optimization would be required to use the extraction chip as a quantitative measurement tool, but the current results are adequate for detection purposes. 3.3.6.3 RNA extraction from virus samples The automated RNA extraction protocol was modified to follow established steps for extracting viral RNA from biological samples. The system was tested in India by Indian Immunologicals LLC. Cultured FMDV samples were used as system inputs. 75 Table 3.2: RNA extraction yields from four samples of E. Coli cells that were lysed onchip. Successful extraction demonstrates that the system is capable of extracting nucleic acid from raw biological samples Sample E. Coli 0157:H7 E. Coli 0157:H7 E. Coli 0157:H7 E. Coli 0157:H7 Cells Input 70 million cells 70 million cells 56 million cells 7 million cells Total RNA Retrieved 429ng 374ng 253ng 276ng Successful extraction was analyzed by both spectroscopic and RT-PCR detection. Table 3.3 shows Nanodrop readings of the concentration of RNA in the 35 μl elutions. The first column indicates the strain of virus and the run number identification. The last column is the ratio that signifies the purity of the extracted RNA. The 260/280 ratios in the table are consistently above the desired 1.8 which indicates a pure sample. To verify that the readings do, in fact, represent the presence of viral RNA, RTPCR was carried out on a plate cycler. Figure 3.15 shows results where RNA extraction results from the microfluidic device (MFD) were amplified using RT-PCR and verified by performing gel electrophoresis. Lane 6 is a positive control where the viral RNA was extracted using the Qiagen RNEasy Kit, and amplified on a plate cycler. Lanes 1-4 have distinct bands in the correct location indicating that the microfluidic system successfully extracted viral RNA. The weak band in lane 2 and the nonexistent band in lane 5 are signs that there are intermittent failures in the system. 3.3.6.4 DNA extraction The nucleic acid extraction capabilities of the microfluidic system were again explored by using the solid phase extraction principles to purify DNA as well. Similar to RNA extraction, known quantities of DNA were obtained and used as sample inputs to the system. Figure 3.16 shows the DNA extraction efficiency compared to a commercial 76 Table 3.3: List of microfluidic extraction runs and RNA quantitation for each run using Nanodrop measurements Target/Microfluidic Extraction Run Nucleic Acid Conc. (µg/ml) A260 A280 260/280 O IND R2/75 MFD1 O Manisa MFD2 O TNN 24/84 MFD3 O TNN 24/84 MFD4 O R2/75 MFD5 O R2/75 MFD5 O TNN 24/84 MFD 7 O TNN 24/84 MFD 9 O IND R2/75 MFD 10 O TNN 24/84 MFD12 O TNN 24/84 MFD13 O TNN 24/84 MFD14 O IND R2/75 MFD15 O IND R2/75 MFD16 O IND R2/75 MFD17 O IND R2/75 MFD21 O IND R2/75 MFD22 O IND R2/75 MFD23 O IND R2/75 MFD24 69.300 86.200 18.900 15.400 27.900 27.700 6.900 11.100 7.200 0.183 0.031 0.087 0.100 0.126 0.097 38.0 60.6 48.7 26.3 1.732 2.154 0.473 0.386 0.698 0.691 0.174 0.278 0.179 4.577 0.78 2.162 2.487 3.149 2.412 0.951 1.516 1.217 0.658 0.372 0.471 0.103 0.108 0.21 0.202 0.025 0.023 0.034 1.332 0.31 0.885 1.054 0.83 0.557 0.441 0.72 0.531 0.292 4.65 4.58 4.58 3.57 3.33 3.42 7.02 11.96 5.23 3.44 2.52 2.44 2.36 3.79 4.33 2.16 2.11 2.29 2.25 Figure 3.15: Gel electrophoresis of RT-PCR amplification of extracted RNA products from FMDV samples using automated microfluidic system. M: 100 bp DNA Ladder, Lane 1 to 6 RT & PCR with standard kit. Lane 1: O TNN 24/84 MFD Run 4. Lane 2: O IND R2/75 MFD Run 5. Lane 3: O TNN 24/84 MFD Run 3. Lane 4: O IND R2/75 MFD Run 1. Lane 5: O IND R2/75 RNA recovery in MFD. Lane 6: O IND R2/75 Positive control. 77 Recovery of prepurifed DNA 140 120 ng out 100 80 60 glass 40 spin kit 20 0 0 100 200 300 400 500 ng in Figure 3.16: DNA recovery using microfluidic automated nucleic acid extraction system compared to commercial system. laboratory kit (Qiagen DNA kit). To ensure that the eluent is free of contaminating chemicals introduced by the extraction process, nucleic acid-free control samples sent through the extraction system were spiked with varying amounts of purified DNA in known quantities. Different amounts of whole human blood were used as the samples for nucleic acid extraction process, and results were again quantified using real-time PCR crossing-point analysis. Results were compared to extractions performed on standard commercial centrifugation kits. The glass filter from one of our devices was used to replace the filter in a commercial spin column, and the filter from the spin column was placed in one of our filter casings in order to compare our solid phase filter to the one used in the commercial kit. Results in Figure 3.17 show that while the extraction efficiency of the microfluidic device is slightly lower than commercial standards, the device is capable of extracting PCR-amplifiable DNA from a whole blood sample in an automated system. When the 78 Blood DNA Extraction 600 Kit DNA Extracted (ng) 500 Microfluidics 400 Kit 2 Microfluidics 2 300 200 100 0 0 20 40 60 80 100 Blood Sample Volume (μl) Figure 3.17: Extraction of DNA from whole blood samples performed on-chip compared with standard spin kit extraction progress. Kit 2 series indicates a spin kit test with the column filter replaced with the filter material used in the microfluidics. Microfluidics 2 is a test performed on the microfluidic chip with the filter material replaced with the material from the spin kit extraction columns. filters were switched, both systems extracted slightly less genetic material than what would be expected based on the linear response observed in other tests. The glass microfiber filter solid phase proved effective and is easily incorporated into an automated nucleic acid extraction system to extract PCR-amplifiable product from raw blood samples in less than 20 minutes. 3.3.6.5 Summary of nucleic acid extraction from multiple sample types A system capable of varied extraction conditions and targets is desirable for the creation of a versatile sample preparation device. Table 3.4 gives a summary of total amounts of nucleic acid recovered from the representative extraction tests performed on the automated microfluidic system described in previous sections. Successful extraction of nucleic acids was demonstrated more by the ability of the system to produce detectible 79 Table 3.4: Nucleic acid extracted from multiple sample types Total nanograms Sample NA Retrieved E. Coli 0157:H7 429 ng RNA FMDV virus 539 ng RNA Stock RNA 35.7 ng RNA Whole Blood 22.0 ng DNA Stock DNA 118 ng DNA levels of the target NA, as opposed to having a mandated extraction efficiency or quantity threshold. Input amounts and conditions varied for each sample type in the table, so no conclusions should be drawn as to the relative quantitative ability of the system to process the different sample types. Rather this table illustrates the qualitative ability of the system to readily extract nucleic acids form different input samples and target molecules in an automated fashion. 3.4 Conclusions The developed NA extraction system has proven capable of handling a variety of sample preparation tasks in an effort to make automated nucleic acid sample preparation more practical and available. Such work is a step towards a universal sample preparation unit that can handle many kinds of samples from viruses to tissue samples and provide rapid, meaningful results from a lab-on-a-chip platform. The RNA detection abilities were demonstrated for two main applications. The first was the detection of E. Coli RNA to monitor drinking water supplies for harmful bacteria where RNA is a more appropriate genetic material for detection. RNA was also extracted from viral samples in tissue 80 culture fluid to demonstrate another form of pathogen detection. These applications will open the door for future pathogen screening and detection tests. The same system was shown to be capable of extracting DNA from whole blood samples. The project results therefore show that the developed system meets the need for fast, simple, and adaptable sample preparation systems. 3.5 References 1. A. J. de Mello and N. Beard, Lab on a Chip 3 (1), 11N-20N (2003). 2. W. H. Grover, A. M. Skelley, C. N. Liu, E. T. Lagally and R. A. Mathies, Sensors and Actuators B (Chemical) B89 (3), 317-325 (2003). 3. D. A. Bartholomeusz, R. W. Boutté and J. D. Andrade, Microelectromechanical Systems, Journal of 14 (6), 1364-1374 (2005). 4. Y. Chenying, W. Wei and L. Zhihong, presented at the 2009 4th IEEE International Conference on Nano/Micro Engineered and Molecular Systems (IEEENEMS 2009), 5-8 Jan. 2009, Piscataway, NJ, USA, 2009 (unpublished). 5. T. Poeckh, S. Lopez, A. O. Fuller, M. J. Solomon and R. G. Larson, Analytical Biochemistry 373 (2), 253-262 (2008). 6. X.-W. Chen, Z.-R. Xu, B.-Y. Qu, Y.-F. Wu, J. Zhou, H.-D. Zhang, J. Fang and J.H. Wang, Analytical and Bioanalytical Chemistry 388 (1), 157-163 (2007). CHAPTER 4 OSCILLATORY FLOW PCR WITH EVAPORATION CONTROL ON A POLYCARBONATE CHIP 82 4.1 Abstract An oscillatory flow PCR module is presented for integration with the automated nucleic acid extraction system. A disposable, multilayer polycarbonate oscillatory flow PCR chip is fabricated with evaporation control to keep fluid loss to under 10% over 40 cycles. Temperature zones on the chip are created by making contact with external aluminum heating blocks. Fluid temperatures are computed using numerical simulations. Successful on-chip PCR is demonstrated. 4.2 Introduction Nucleic acid amplification is a process by which a specific sequence of a strand of DNA or RNA is copied multiple times to produce a larger quantity of the target sequence. Of the methods of amplification, PCR has grown in popularity since its inception in 19871 because of its simplicity. In PCR, DNA sequence primers attach to the beginning and end of specific sequences of nucleotide bases of a single strand of DNA in an annealing process that takes place at about 60 ˚C. Elevating the temperature to an intermediate range around 72 ˚C activates an extension process where a polymerase enzyme begins at the primers and zips along the strand, inserting corresponding nucleotide bases to convert the single strand into a double strand of DNA. The temperature is then elevated above the melting temperature of the DNA at around 95 ˚C where the DNA molecule denatures into two single strand segments, and the process is repeated by following the same temperature cycle. PCR is well suited for application to microfluidic devices because of the direct realization of several proposed microfluidic benefits: lower volumes of costly reagents and less thermal inertia for rapid thermal cycling. There have been several reviews 83 published in the last decade that explore the progress made in the adaptation of PCR to the microchip. There are hundreds of microfluidic devices, and most microfluidic PCR chips can be placed into two broad categories: stationary chamber-based chips with integrated temperature controllers and continuous flow PCR chips. Stationary chips offer good flexibility in temperature and cycle control but generally rely on more expensive silicon processing manufacturing techniques. Continuous flow chips generally have faster temperature ramp rates because only the fluid temperature needs to change, eliminating the thermal inertia of surrounding chip components. However, continuous flow chips have a fixed number of cycles based on channel geometry, offering little room for process adaptation. A subset of continuous flow PCR chips that relies on oscillatory flow between temperature zones has arisen that attempts to include the flexibility of the stationary chamber devices with the low thermal mass seen in continuous flow chips. Using this architecture, a fluid droplet is shuttled back and forth between different temperature zones on the chip. This chapter presents a disposable polycarbonate oscillatory flow PCR chip that allows control over cycle times and temperatures. The chip is integrated with the nucleic extraction system presented earlier by using the same instrument and control software for pumping the PCR droplet between temperature zones. 4.3 Materials and Methods The overall approach to developing an oscillatory PCR chip involved 3 stages: (1) Fabrication of an appropriate, cost effective chip, (2) Modification of the chip to overcome evaporation and sample loss, (3) Simulation of the chip to determine appropriate operating conditions. The motivation and rationale for these steps follows. 84 The creation of this oscillatory PCR chip began with the choice of substrates. Inexpensive, disposable devices are desirable for PCR chips to reduce the risk of cross contamination between tests. Plastics present fabrication cost advantages and can also have favorable optical properties. However, the elevated temperatures associated with PCR present a stability problem for many common thermoplastics. Polycarbonate was selected as the material for the chip substrate because it is optically transparent, relatively inexpensive, and has a sufficiently high glass transition temperature (150˚C) to withstand the elevated PCR annealing temperature. Another disadvantage to using a plastic substrate is that it is an insulator. The relatively low thermal conductivity (k=0.2 w/m^2 K) impedes the heat transfer desired to quickly ramp the fluid temperature through the PCR thermal cycle. This low thermal conductivity can be circumvented by having the heat source in contact with the fluid, but it is difficult to embed on-chip heaters and temperature probes and maintain low production costs. Fluid temperature feedback in microchannels has been achieved using embedded temperature probes,2, 3 temperature dependent fluorescent dyes,4 and noncontact interferometry.5 The difficulty in creating an inexpensive rapid plastic oscillatory PCR chip lies in having the ability to quickly and accurately adjust the internal fluid temperature using an external heat source. The approach used in this work mitigates the thermal insulating effect of the polycarbonate substrate by using thin walls in good thermal contact with external heating blocks. Verification of internal fluid temperatures without direct feedback is achieved using numerical simulations to model the transient thermal state of the fluid. Parameters such as wall thickness, material thermal properties, and external heater temperature will 85 be varied in the simulation to determine optimum chip characteristics to achieve the fluid temperatures needed for PCR. Transient analysis will determine the cycle times required. Numerical results will be compared with measured surface and fluid temperatures gathered using topographical thermal imaging and embedded thermocouples. The parameters developed using these models are then used experimentally and successful PCR is demonstrated. 4.3.1 Chip Fabrication Microfluidic devices for PCR need to be disposable to prevent crosscontamination between assays. Even trace amounts of target nucleic acids left from previous test runs can be amplified during the PCR process to detectable levels. Consequently, a simple laminated plastic chip design is used in this work to produce inexpensive, disposable PCR chips. The microfluidic PCR chip is then mounted onto a temperature controlled heater manifold to produce the regions at temperatures required for PCR. Two versions of the microfluidic chip were created: a simple three-layer chip with a serpentine chip, and a more complex version with seven layers that incorporates an evaporation control scheme. 4.3.1.1 Three-layer chip fabrication The microfluidic PCR chip is fabricated using three laminate layers of polycarbonate (Lexan FR83, Sabic Polymershapes Pittsfield, MA). As noted earlier, polycarbonate is an ideal material because its glass transition temperature is well above the maximum temperature required for PCR (~95 ˚C), so it can be used in PCR applications. Even though polycarbonate has a high glass transition temperature, 86 multilayer devices can still be easily created using heat to laminate layers together for rapid prototyping. Figure 4.1 illustrates the process of creating a multilayer polycarbonate chip. The first step in chip fabrication is to cut the channel from the middle layer of polycarbonate. The serpentine nature of the channel geometry adds difficulty to cutting this middle layer as the interdigitated thin polycarbonate fingers that are produced are easily displaced from their intended position. Even small finger displacement leads to device failure in the final chip because the small cross-section channels are easily sealed by a slight dislocation of the channel walls. To prevent this dislocation, the middle layer is supported by a layer of adhesive during the cutting step and subsequent bonding to the top layer. The adhesive tape layer is applied to a sheet of polycarbonate (a). The Figure 4.1: Schematic of multilayer chip fabrication method. 87 polycarbonate/tape assembly is then cut using a knife plotter (Graphtec) to cut through the polycarbonate but leave the supporting adhesive tape layer intact (b). The polycarbonate left where the fluid channel will be is then manually removed with tweezers, leaving an empty channel. Following the cutting of the fluid channel, the top layer of polycarbonate is thermally bonded to the middle layer using a heat press set to 175˚C (c). The top polycarbonate layer is placed in contact with the polycarbonate side of the middle layer polycarbonate/tape assembly. The layers are sandwiched between two pieces of smooth paper backing and put into the heat press, with the tape layer down. The heat press is kept at 175 ˚C, just over the glass transition temperature of 150 ˚C, and the two layers are bonded for 2 minutes with slight pressure. This produces a good bond between the two layers (d), but minimizes melting to maintain the cross-section geometry of the fluid channel. Access holes are then cored into the top layer for the fluid inlet and outlet (e). The adhesive tape support layer can then be removed from the middle layer because the bond between the middle and top polycarbonate layers now provides the support for the thin channel walls. A bottom layer of polycarbonate is then placed in contact with the now exposed middle layer(f), and the chip assembly is again placed in the heat press in the manner stated before to bond the bottom layer to the middle(g). The resulting chip has an enclosed serpentine channel in a disposable polycarbonate threelayer chip assembly. Custom access ports made from laser-cut acrylic are then attached to the cored port holes in the top layer with UV-cured adhesive (h-i). 88 4.3.1.2 Seven-layer evaporation control chip fabrication The seven-layer version of the oscillatory flow PCR chip is fabricated using the same process as the three-layer device, but with two extra polycarbonate layers on the top and bottom surfaces shown in Figure 4.1 (j) and Figure 4.2. The intermediate layers create an extra fluid layer on either side of the sample fluid layer in the middle of the chip. The new fluid layers share a common inlet and outlet port. 4.3.1.3 Chip surface inoculation Following chip fabrication, the channel walls were passivated to reduce nonspecific adsorption of target molecules to the polycarbonate surface. To passivate the surface of the channels in the polycarbonate chip, 0.1 % solution (w/v) of bovine serum albumin (BSA) was pumped through the chip at a linear rate of approximately 1 mm/sec Figure 4.2: Diagram of seven-layer stack for fabrication of evaporation control chip. 89 for 1 hour. The BSA-filled chip was then incubated at 40 ˚F for 24 hours to allow BSA to coat all binding sites on the interior channel walls. Prior to use in experiments, the BSA was pumped out and the channels dried by a flow of air. 4.3.2 Temperature Control The temperature zones for the PCR chip are controlled using three aluminum heater blocks. Each block is attached to a flexible film resistive heater (Omega) and has an embedded thermocouple for temperature feedback. The heater and thermocouple are connected to a commercial PID to monitor and control the heater block assembly temperature. Voltage to the heaters is provided by a DC power supply. The heater blocks are housed in a polycarbonate casing which separates the heater blocks with a small air gap (approximately 1 mm). The chip is held in contact with the heater blocks by a polycarbonate lid layer attached with bolts. Figure 4.3 is a photograph of the PCR Figure 4.3: Photograph of heater control setup. 90 chip module setup including the 3 commercial PID controllers and the chip holder assembly. 4.3.3 Temperature Modeling The critical step in performing PCR is ensuring that the fluid reaches the proper and precise temperature to facilitate the three steps of PCR: denaturation of the target DNA strand, annealing of the primer to the target sequence, and extension of the primer along the strand to create a copy of the target sequence. The precise requirements for the temperature at which these steps occur dictates a rigorous analysis of the PCR chip temperature profile and transient analysis to optimize the PCR protocol and ensure that the fluid has sufficient residence time in each temperature zone to reach the proper temperature and complete each step in the PCR cycle. To help determine the best operating parameters, first, a 3-dimensional steady state simulation of the thermal conditions outside the chip was performed. This provided the boundary conditions needed to determine the internal temperature conditions of the chip. Next, 1-dimensional, steady state solutions are determined for a range of intrachannel fluid temperatures to yield a first approximation for the initial conditions applied in the transient temperature simulation. A 1-dimensional transient numerical solution is then implemented to determine the minimum residence time necessary for the fluid in each temperature zone to reach the desired temperature. 4.3.3.1 Determination of convection coefficient: h The first modeling step is the determination of the boundary conditions needed to calculate the internal fluid temperature; specifically the rate of heat leaving the top surface of the chip, which is dependent on the convection coefficient, h. A common 91 method for determining h for natural convection involves using Nusselt number correlations based on the geometry and fluid properties. These empirical correlations were generally developed for larger systems with higher Rayleigh numbers than what is found in the system discussed in this work. Also, the geometric conditions of the PCR chip assembly require determination of the natural convection coefficient specific to this system. As the empirical convection coefficients are inappropriate for the conditions in this work, convection coefficients here are determined from 3-dimensional multiphysics modeling using Comsol multiphysics software. The fluid of interest for determining the convection coefficient is the air above the chip surface. The heated chip surface warms the air which lowers the air density, causing it to rise from buoyancy forces. The geometry modeled for this system was a block of air shown in Figure 4.4. . g f e h i a b c Figure 4.4: Geometry used for natural convection modeling. d 92 The size of the block of air to be modeled was chosen based on guidance from the software manufacturer to place boundaries sufficiently far away from the heating surface. These recommended boundary distances were verified by incrementally increasing the size of the geometry until simulated heat flux results from the surface of interest remained constant. The final geometry used for the simulations was stable with a 0.74% change in heat flux with a 16.7% change in geometry dimensions. Symmetrical flow and adiabatic boundary conditions were used on the boundary representing the plane of symmetry (face h). Fixed temperatures were used at the surface of the chip temperature zones (faces a, b, and c). The other solid surfaces were modeled as thermal insulators (face i and all unlabeled faces). Those solids surfaces were constrained as no-slip walls for laminar flow. The top air boundary (face g) was a heat outflow and a laminar outlet at ambient pressure. The side air boundaries (faces d, e, and f) are set to ambient temperature and laminar inflow with no pressure. A body force based on the gravity acting on the temperature-induced density gradient was added to the air in the simulated domain to create the density driven flow in a natural convection environment. The model was used to calculate I for a series of surface temperatures that cover the range expected for PCR experiments. The numerical simulation was iterated with successively finer mesh elements until the change in the results varied less than 1%. 4.3.3.2 1D steady state solution For a first approximation of the temperature state in the PCR chip and the fluid inside the channel, a1D steady state analytical solution for a composite wall is used. Multiple layers are used to represent the polycarbonate layers, fluid layers, and the heater 93 block. The thicknesses of the layers and material properties are adjusted to attain optimal PCR temperature control. The boundary conditions of the composite wall are a fixed temperature at the boundary representing the contact between the chip and the heater control manifold and a free convection boundary at the top surface exposed to room temperature. The variable of interest is the temperature of the heater control manifold. The goal is to determine what temperature to set as the manifold operating temperature to achieve the proper fluid temperature inside the channel. To calculate the temperature at points in the wall, we need to find the heat flux q, which can be determined using the thermal resistance of the entire composite wall coupled with the boundary conditions. The heat equation for the composite wall is: 𝑞 ′′ = 𝑈(𝑇𝑖 − 𝑇𝑜 ) (3.1) where 𝑞 ′′ is the heat flux per unit area, 𝑇𝑖 is temperature at the inner surface of the wall, 𝑇𝑜 is the temperature at the outer surface, and 𝑈 is the overall heat transfer coefficient in the wall given by 𝑈= 1 𝑡1 ⁄𝑘1 + 𝑡2 ⁄𝑘2 + ⋯ + 𝑡𝑛 ⁄𝑘𝑛 (3.2) where 𝑡1−𝑛 and 𝑘1−𝑛 are, respectively, the thickness and thermal conductivity of each layer of the PCR chip. Any layer can be omitted in the simulation by setting the thickness of that layer to zero. The temperature at the outer surface is found by combining the heat equation in the wall with the convection heat equation at the surface: 𝑞 ′′ = ℎ(𝑇𝑜 − 𝑇∞ ) (3.3) where 𝑇∞ is the ambient temperature and h is the convection coefficient. Solving for the 94 temperature at the outer surface yields: 𝑇𝑜 = ℎ𝑇∞ + 𝑇𝑖 𝑈 ℎ+𝑈 (3.4) In the numerical simulation, the unknown temperature 𝑇𝑖 at the inner surface is given a guessed initial value below the expected value for the temperature at the heater manifold/PCR chip interface. The temperature at the outer surface is then calculated from equation 3.4. Then the heat flux 𝑞 ′′ can be calculated from equation 3.1, which can then be used to calculate the temperature at all interior points. The average temperature in the middle PCR chip layer is then compared to the desired fluid temperature. If the temperatures are not equal to within the desired tolerance, the guessed inner surface temperature is adjusted and the process repeats until the correct fluid temperature is reached. This approach provides the steady-state temperature profile for the temperature through all layers of the PCR chip. This temperature profile (like that in Figure 4.5) provides the initial conditions for solving the transient temperature profile to determine the required residence time for the fluid to reach the desired temperature. 4.3.3.3 Transient temperature analysis A transient temperature analysis of the PCR chip is performed by using a finite difference approximation of the differential heat equation with no heat generation: 𝜕 𝜕𝑇 𝜕 𝜕𝑇 𝜕 𝜕𝑇 𝜕𝑇 (𝑘 ) + (𝑘 ) + (𝑘 ) = 𝜌𝐶𝑝 𝜕𝑥 𝜕𝑥 𝜕𝑦 𝜕𝑦 𝜕𝑧 𝜕𝑧 𝜕𝑡 (3.5) The form often seen of this equation assumes constant properties throughout the material. In the case of the composite wall in this study, there are abrupt changes in the thermal conductivity at the boundaries of the different materials. Thus, the differential heat 95 Aluminum Layer PC Layer Fluid PC Layer Ambient Layer Layer Figure 4.5: Graph of a steady-state temperature profile solved for a 3-layer PCR chip. equation must be solved assuming that the thermal conductivity varies with x. The equation can also be simplified because the planar geometry of the system allows the assumption that heat flows only in the x direction (normal to the chip surface) to remove any y and z dependence. These assumptions result in the following form of the heat equation: 𝜕 2 𝑇 𝜕𝑇𝜕𝑘 𝜕𝑇 𝑘 2+ = 𝜌𝐶 𝑝 𝜕𝑥 𝜕𝑥 2 𝜕𝑡 (3.6) Equation (3.6) is then converted to a finite difference approximation using a forward time, center space configuration. 96 𝜎𝑖𝑚+1 𝑚 𝑚 𝜎𝑖−1 − 2𝜎𝑖𝑚 + 𝜎𝑖+1 = 𝛼𝑖 ∆𝑡 ( ) ∆𝑥 2 + (3.7) 𝑚 𝑚 𝛼𝑖 ∆𝑡 𝑘𝑖 − 𝑘𝑖−1 𝜎𝑖+1 − 𝜎𝑖−1 ( )( ) 𝑘𝑖 ∆𝑥 2∆𝑥 where σ is the temperature at position i and time m, ∆t is the time step and ∆x the distance between position nodes. The variable thermal conductivity term does not use a center space configuration in equation 3.7. In the composite wall configuration, using the centered space finite difference approximation for the thermal conductivity derivative does not accurately reflect the step function nature at the material interfaces. Consequently a backward space finite difference scheme is applied for the thermal conductivity derivative only. Boundary conditions for the transient thermal model included a constant temperature at the heater block and a convection coefficient at the chip surface. 4.3.4 Temperature Testing Verification of the numerical temperature simulations was performed using infrared thermography. An infrared thermal camera (FLIR 420) was used to capture the surface temperature of the chip. The camera was mounted to look down at the surface of the PCR chip on the heater block assembly. The chip surface was coated with a thin layer (approximately 30 μm) of matte black acrylic paint. The emissivity of the paint was measured following the camera manufacturer's instructions to be 0.95. The temperatures of the heater blocks were adjusted from 50 ˚C to 100 ˚C and surface temperatures of the three-block heater assembly were recorded at 10 ˚C increments. These measured surface temperatures were compared to calculated surface temperatures based on the simulated values for heat flux and convection coefficients. 97 Temperature uniformity to within approximately 1°C over the surface of interest was desired. Heat sink compound was employed between the heater block surface and the plastic chip to ensure good thermal contact and reduce temperature deviations. 4.3.5 Pumping/Integration The pumping method used for moving the plug of fluid between temperature zones uses the same metering diaphragm pumps used for the nucleic acid extraction chip. The pumping chip is fabricated in PDMS using the three-layer technique already described. Inlet and outlet tubes from the polycarbonate PCR chip are connected to two ports on the pumping chip. Diaphragm pump volumes are calculated such that a single stroke pushes the PCR fluid from one temperature zone to the next. A pressurized air line provides an elevated back pressure to the entire pump/PCR chip system to keep the PCR fluid from boiling at the high temperature zone. The pumping module for the PCR chip is controlled using the same instrument and software used for the nucleic acid extraction chip, creating an integrated extraction/amplification system. 4.3.6 PCR Process Parameters The ability of the oscillatory flow PCR system to amplify nucleic acids was demonstrated by thermal cycling a PCR sample based on an engineered, relatively short DNA strand obtained as a gift from BioFire Diagnostics Inc. PCR chemistry conditions are indicated in Table 4.1. The template was an engineered DNA sequence of 184 base pairs in the following sequence: CTCGCAAATGACTGGCTAAAAGTCTGATTAACCGATAGGTCATTCA GATTCATTTATGTAGCTTTATTATCCTAAATCTCCACTTCCACTGAGCTGGGTC 98 Table 4.1: Description of PCR reaction chemistry used for functional testing of polycarbonate oscillatory flow PCR chip PCR Mix Reagent Description Volume (ul) 10X dntp's 30mM MgCl2 10X PCR Buffer 10X LC Green 10X Primer Mix (5uM) Klen Taq Template Water total 1.5 1 1 1 0.8 2 2.7 10 Final concentration 1.5 X 1 X 1 X 0.5 μM 0.4 U/μl GTCCCTATCCTTATTGCACTACTATCTTTAGTTTACGAATATGGTTGTTTAAAT ATGAGTCCGTGTTGACCTGAATAGCGAGG. The forward primer used had sequence ATAGGTCATTCAGATTCAGTTTATGTA and the reverse primer was CAGGTCAACACGGACTCATATT. The melting temperature for the sequence is 83 ˚C. For thermal cycling on the chip, the first temperature zone is set to 60 ˚C and the second and third temperature zones are both set to 95 ˚C. The short nature of this specific DNA target sequence does not require an intermediate extension temperature. The fluid is pumped between the two temperatures, 20 seconds at 60 ˚C and 10 seconds at 95˚C for each cycle. 4.3.7 Gradient PCR To verify the chemistry, eight samples of the PCR mix with template were placed in the wells of a 96-well plate. The plate was thermocycled on a gradient thermocycler (BioRad S1000) with annealing temperatures ranging from 55 ˚C to 70 ˚C. The plate underwent 40 cycles of 30 seconds at the annealing temperatures followed by 10 seconds 99 at the denaturation temperature. Following temperature cycling, the plate was fluorescently imaged and analyzed using high resolution melting analysis (Light Scanner, Idaho Technologies). 4.3.8 Melting Analysis Melting analysis is performed by including a dye that fluoresces in the presence of DNA. As the temperature increases to the point where the DNA denatures, the dye fluorescence declines rapidly. For the 184 base-pair engineered DNA sequence, the melting temperature was previously calculated to be at 83 ˚C and experimentally verified by performing real time PCR and melting analysis on a commercial instrument (LightCycler, Roche). 4.3.9 Evaporation Control While the small volume of PCR fluid on the microscale is advantageous for decreasing thermal cycling times, evaporation can be a major concern. The high temperatures associated with PCR (especially with the denature step) and the high surface to volume ratios of the droplet cause significant evaporation rates. Also, a thin polycarbonate film is not a perfect barrier to water vapor. Polycarbonate has a permeability of 1400 barrers6 at STP, which correlates to a loss of approximately 1 μl (10% of a 10 μl sample) over the approximate time span of a 40 cycle PCR amplification run using the oscillatory flow PCR chip. The fluid loss is expected to be even higher at elevated temperatures. While information on the permeability of the polycarbonate to water vapor at elevated temperature is scarce, Hanada et al. indicate an approximate 10 fold increase in water vapor transmission rates for a 40˚C increase in temperature for 120 100 μm polycarbonate membranes.7 The chip conditions at the elevated PCR temperatures (especially the denature temperature) predict a significant loss of fluid by permeation through the thin polycarbonate chip layer. Losing even small percentages of fluid can affect the delicate PCR chemistry balance and reduce PCR efficiency. Several attempts have been made to mitigate the evaporation effects in microchip PCR including surrounding the droplet in oil,8 embedding capillary channels in PDMS,9 and increasing the pressure in the channel.10 Evaporation control was attempted using several oil-based methods in relation to the chip under development. Effective evaporation control was achieved on the polycarbonate chip by including an evaporation control layer on the top and bottom surfaces of the polycarbonate chip. The evaporation control layers are filled with water and pressurized with the same back pressure used in the fluid layer of the chip. 4.3.10 Chip Coating With evaporation effectively reduced, on chip DNA amplification still demonstrated primer-dimerization, an ineffective amplification of the target sequence. To improve PCR performance, the channel surface of a seven-layer evaporation control chip was coated with bovine serum albumin (BSA) to inoculate nonspecific binding sites and reduce possible target molecule or PCR reagent adsorption to the channel walls. For PCR on the coated, evaporation control chip, samples were loaded and pumped between two temperature zones. Forty cycles were performed with denaturation taking place for 10 seconds at a temperature zone set for 97.5 ˚C, annealing and extension taking place for 20 seconds at 60 ˚C. 101 4.4 Results 4.4.1 Chip Fabrication Three-layer (Figure 4.6) and seven-layer (Figure 4.7) chips were fabricated using the lamination method described above. Tubing was connected to the access ports to load samples and interface with the control system. The chips held approximately 20 μl of fluid per temperature zone. A laser-cut polycarbonate stack was fabricated to house the aluminum heater blocks for the heating module. Figure 4.8 shows the heater module with wire leads for the thermocouple embedded in the aluminum and the thin film heater attached to the aluminum blocks. 4.4.2 Numerical Simulations 4.4.2.1. 3D modeling to determine convection coefficient Simulations were performed on COMSOL multiphysics modeling software to determine the convection coefficient h for the chip geometry with multiple temperature zone configurations. Flow and body force profiles for one such configuration are visualized in Figure 4.9. It was observed that the middle temperature zone produced less upward flow relative to the neighboring temperature zones when held at the same temperature. This Figure 4.6: Polycarbonate PCR chip with serpenting channels to transmit fluid to three distinct temperature zones. 102 Figure 4.7: Photograph of fabricated seven-layer evaporation control chip with wider channels for pressurized water above and below the sample fluid layer. 1 inch Figure 4.8: Heater block module for temperature zone creation. 103 Figure 4.9: Flow and body force profiles for the fluidic chip (a) Velocity magnitude vectors of air flow above the fluidic chip and (b) body force based on buoyancy in cross section at line of symmetry. 104 effect is likely due to the outside zones being exposed to inflows of cool air from three sides, while the middle zone is insulated on two sides by the other temperature zones. However, this effect was found to have negligible impact on the determined convection coefficient, with differences of less than 0.5% observed from zones of the same temperature at different locations on the chip. The average convection coefficient for free convection heat transfer from the chip surface was calculated from the heat flux per unit area q"conv leaving the surface as determined by the simulations. Results for h and q"conv are summarized in the first three columns of Table 4.2 for several chip surface temperatures. The small amount of heat loss from the chip surface due to natural convection makes radiation effects important for this system. Indeed, calculated heat loss to the environment based on radiation, q"rad proves to be a larger proportion of the total heat flux than that caused by convection effects. For such calculations, ambient temperatures (T∞) were set to 20°C. An effective convection coefficient heff was calculated based on the combined heat flux q" total from both radiation and convective effects. While heff thus includes a 4th power function of temperature with the radiative heat flux, the total heat Table 4.2: Simulated and calculated values for heat flux and convection coefficients based on chip surface temperature. 6 q" conv (W/m2 ) 240.6 q" rad (W/m2 ) 278.5 q" total (W/m2 ) 519.1 heff (W/m2 K) 13 72 6.3 328.7 384.5 713.2 13.7 90 6.8 474.6 565.8 1040.4 14.9 100 7 559.9 678.8 1238.7 15.5 Surface Temp ˚C 60 h (W/m2 K) 105 flux can be closely approximated by a linear function over the temperature range of interest. This will be shown to be advantageous when applying boundary conditions to calculate the transient temperature conditions in the system. 4.4.2.2 Agreement with calculated temperature Table 4.3 shows good agreement between thermographically measured surface temperature and the calculated surface temperature to within 1˚C over the range of interest. 4.4.2.3 Temperature uniformity over chip surface Thermography temperature data from the total area of each chip temperature zone provide information on the temperature uniformity. Thermal images were taken of the PCR chip contacting the heater blocks in two different manners: direct clamping from the top plate of the heater assembly, and direct clamping coupled with a layer of heat sink compound. Chips without heat sink compound were shown to demonstrate a larger variance in temperatures, both between temperature zones set at the same temperature, and within a single temperature zone. The middle temperature zone showed better temperature uniformity than the outside temperature zones regardless of the presence of heat sink Table 4.3: Agreement between calculated and measured chip surface temperatures Set Block Temperature (˚C) 60 70 80 90 100 Measured surface temp (˚C) 59.8 69.3 78.5 87.9 97.4 Calculated surface temp (˚C) 59.62 69.05 78.83 88.59 98.36 Difference (˚C) (absolute value) 0.18 0.25 0.33 0.69 0.96 106 compound, likely due to more uniform physical contact based on the distribution of the clamping bolts. Figure 4.10 shows the temperature distribution on the surface with and without heat sink compound. Table 4.4 summarizes the statistics for the temperature zones all set to the same temperature (110 ˚C) with or without heat sink compound. Figure 4.11 shows the simulated effects of changing geometric parameters in the system. The first graph shows the temperature of the sample fluid over time under heating and cooling conditions with changes to the thickness of the sample fluid layer. The second Figure 4.10: Thermographic image showing temperature uniformity over the chip surface whith and without heat sink compound. Table 4.4: Effect of heat sink compound on temperature uniformity Zone 1 Zone 2 Zone 3 Ave. Temp ˚C (no compound) St. dev. 100.4 106.0 98.4 2.7 .4 3.3 Ave. Temp ˚C (with compound) St. dev. 105.7 106.6 105.3 0.7 0.5 0.7 107 Fluid Temperature, Transient Analysis 370.00 Temperature (K) 365.00 360.00 355.00 350.00 345.00 340.00 335.00 330.00 0 0.5 1 1.5 2 2.5 Time (s) Cooling, .000125 Cooling, .000250 Cooling, .000375 cooling 0.000500 Heating, 0.000125 Heating, 0.000250 Heating, 0.000375 Heating, 0.000500 Temperature (K) Fluid Temperature, Transient Analysis 370 365 360 355 350 345 340 335 330 0 0.5 1 1.5 2 2.5 Time (s) Figure 4.11: Transient heating and cooling analysis of sample fluid with adjusted parameters. Legend in (a) indicates whether it was a heating or cooling operation and the thickness of the fluid layer in meters. 108 graph shows the effect of adjusting other parameters such as insulating layer thickness, evaporation control layer thickness, convection coefficient, etc. These parameters each had a much smaller effect on the heating and cooling rates of the target fluid. The simulation results indicate that keeping the sample fluid layer as thin as possible is the best way to keep the temperature ramp rate high, with other layer thicknesses being less important. The sample fluid can cool from 95˚C to the desired 64.7 ˚C in less than .5 seconds. Heating rates are similarly rapid. 4.4.3 Evaporation Control Several approaches were used to reduce evaporation in the polycarbonate chip. First, mineral oil droplets of two different sizes (5 μl and 20 μl) were placed on either side of a plug of water inside a three-layer polycarbonate chip. In another test, the top and bottom surfaces of a chip were coated with mineral oil. These two efforts showed at best a 25% evaporation rate over 35 cycles, which was little improvement over control cycles (see Figure 4.12) The seven-layer evaporation control PCR chip was with fraction of fluid remaining Fluid evaporation 1.2 fast pumping (back pressure leak) 1 short oil plugs 0.8 0.6 large oil plugs 0.4 0.2 oil coated chip 0 0 20 40 60 80 100 evaporation control Number of cycles Figure 4.12: Fraction of fluid remaining inside PCR chip after multiple cycles. 109 equalized pressure and humidity on either side of the polycarbonate membrane was shown to greatly reduce evaporation: less than 10% over as many as 60 cycles. This shows that oil droplets do not work as effective evaporation deterrents for systems made of materials that are not good vapor barriers. However, equalizing the pressure and humidity on either side of the fluid channel prevents fluid from evaporating through the thin channel wall of the polymer chip. Initial Tests of PCR amplification on the polycarbonate oscillatory flow PCR chip were carried out on three-layer chips. Melting curves for amplicon produced by the chip are shown in Figure 4.13. The red line (02-08-13-t1) is a PCR reaction control performed on commercial PCR equipment (LightCycler). The single sudden drop in fluorescence and single peak in the derivative plot at the expected temperature of approximately 81°C indicates proper PCR amplification. The other lines in Figure 4.13 have multiple fluorescent drops indicated by multiple peaks in the derivative plot. These peaks indicate different lengths in the strands of DNA molecules present. The likely cause of this is primer-dimerization, where primers bind to each other and the PCR reaction replicates small primer chains, rather than the sequence of interest. Dimerization can be caused by multiple factors including improper temperature conditions, chemical imbalances, or other forms of interference with the annealing or extension PCR steps. Good temperature control in the current system evident by the agreement between simulations and measurements indicated that chemical problems with the PCR reaction rather than temperature anomalies were likely the main cause of these incorrect amplification products. 110 Figure 4.13: Fluorescence and fluorescence derivative plots for high resolution melting analysis The fluorescence plot (top) shows the relative intensity of the emitted fluorescence (y-axis) as the temperature is increased (x-axis). The fluorescence decreases as the DNA strands unwind at elevated temperature. The fluorescence derivative plot (bottom) is the negative derivative of the fluorescence plot. This serves to highlight abrupt changes in fluorescence caused by strands of the same molecule denaturing at the same melting temperature. 111 4.4.4 Amplification 4.4.4.1 Gradient PCR Melting analysis performed on the gradient PCR test samples indicates that the template amplified at all temperatures tested, with the highest amplification coming on samples with annealing at a temperature of 64.7 ˚C (see Figure 4.14). Figure 4.14 also shows that the sample at the annealing temperature of 64.7 ˚C had the lowest small melting curve derivative peaks at other temperatures. These other peaks indicate primer dimerization, where the primers attach to a section of a DNA strand (or primer) other than the one targeted and amplify incorrect sequences. This condition is detrimental to PCR efficiency and can lead to false positive tests and is caused by temperature or chemical variations from the PCR recipe standard. Based on these results, PCR for this molecule is effective with annealing temperatures below 64.7 ˚C. However, all temperatures produced relatively minor dimerization peaks, indicating a relatively stable process with respect to temperature. 4.4.4.2 Amplification on evaporation control chip Because temperature was ruled out as the cause of PCR reaction failures, the next step in producing good PCR product was to decrease evaporation rates to stabilize the reagent concentrations. This was done using the seven-layer evaporation control chip. Following BSA coating of a seven-layer evaporation control chip, amplification was performed on chip with 40 cycles alternating between 95 ˚C for 10 seconds and 65 ˚C for 20 seconds. Successful amplification was demonstrated in 20 minutes with the proper melting temperature observed without dimerization products as evidenced by Figure 4.15. 112 Figure 4.14: High resolution melting curve and negative derivative plot of gradient PCR performed at a range of annealing temperatures. 113 Figure 4.15: Plot of negative derivative of melting curve indicating dimer-free amplificaiton of DNA template for two amplification tests. Verification that the amplicon produced by the on-chip PCR is correct was also carried out by performing a gel electrophoresis shown in Figure 4.16. Bright bands were evident with the same number of base pairs as a positive control amplified using a water bath PCR method. Amplification of the target molecule was successful using the BSA-coated chip as evidenced by the electropherogram lanes 2 and 6 matching the positive control in lane 8. For those tests, the samples were pumped through the temperature zones and cycled by hand, prior to the instigation of full automation. Though fainter, lanes 1 and 7 which correspond to totally automated also showed bands at the correct position. The smearing of material at the bottom of the lanes indicates some primer dimerization. The likely cause is changes in the pumping distance for each cycle that expose portions of the fluid 114 M 1 2 3 4 5 6 7 8 M M: DNA Ladder Lane 1: Automated shuttle PCR Lane 2: Evaporation control, BSA-coated Lane 3: Automated shuttle PCR Lane 4: Not amplified Lane 5: negative control Lane 6: Evaporation control, BSA-coated Lane 7: Automated shuttle PCR Lane 8: Positive control Figure 4.16: Electropherogram of DNA amplicon showing successful on-chip amplification. plug to regions of incorrect temperature. However, the amplification was successful to the point that a melting curve would detect the presence of the target molecule. 4.5 Conclusions A disposable oscillatory flow PCR chip was fabricated using multilayered thermally laminated polycarbonate sheets. Temperature conditions were numerically simulated and experimentally verified. A unique scheme for evaporation control was implemented. Successful amplification of DNA was achieved in 40 PCR cycles in approximately 20 minutes. 115 4.6 References 1. K. B. Mullis, Scientific American 262 (4), 56-61 (1990). 2. M. Bu, T. Melvin, G. Ensell, J. S. Wilkinson and A. G. Evans, Journal of Micromechanics and Microengineering 13 (4), S125 (2003). 3. Q. Xiang, B. Xu, R. Fu and D. Li, Biomedical Microdevices 7 (4), 273-279 (2005). 4. D. Ross, M. Gaitan and L. E. Locascio, Analytical Chemistry 73 (17), 4117-4123 (2001). 5. D. C. Leslie, E. Seker, L. A. Bazydlo, B. C. Strachan and J. P. Landers, Lab on a Chip 12 (1), 127-132 (2012). 6. S. Metz, W. Van de Ven, J. Potreck, M. Mulder and M. Wessling, Journal of Membrane Science 251 (1), 29-41 (2005). 7. T. Hanada, I. Shiroishi, T. Negishi and T. Shiro, presented at the OPTO, 2010 (unpublished). 8. Z. Yi and W. Tza-Huei, presented at the 23rd IEEE International Conference on Micro Electro Mechanical Systems (MEMS 2010), 24-28 Jan. 2010, Piscataway, NJ, USA, 2010 (unpublished). 9. A. Polini, E. Mele, A. G. Sciancalepore, S. Girardo, A. Biasco, A. Camposeo, R. Cingolani, D. A. Weitz and D. Pisignano, Biomicrofluidics 4, 036502 (2010). 10. J.-Y. Cheng, C.-J. Hsieh, Y.-C. Chuang and J.-R. Hsieh, Analyst 130 (6), 931-940 (2005). CHAPTER 5 CONCLUSIONS AND FUTURE WORK 117 5.1 Conclusions A system for automated extraction and amplification from a variety of samples was developed. The system built off of advancements to 3-layer PDMS microfluidic chip manufacturing to provide a device capable of complex fluid handling tasks. The programmability of the device allowed for hands-off processing of nucleic acids. A novel integratable shuttle PCR chip design was fabricated and successfully reduced evaporation rates and amplified DNA. Reviewing the state of the art in microfluidic nucleic acid extraction devices and investigating the capability to create an automated system provides several clear conclusions:  There is a need for versatile nucleic acid sample preparation systems that can process pure RNA or DNA from a variety of sample types in a way that can be done by nonexperts.  Three-layer PDMS components and devices were fabricated and proved to be effective in portable microfluidic systems with minimal external equipment.  The gas permeability of the membrane provided a unique pumping, bubble injection, and bubble removal mechanism that was effective at low-flow pumping in dead-end channels, trapping air bubbles, and enhancing mixing by over 40%.  A system was created that performs automated extraction of RNA or DNA from multiple sample types to make sample preparation more practical and available.  The Utah system is capable of extracting detectible levels of RNA from E. Coli bacteria and FMDV virus samples. It likewise can be used to extract human genomic DNA from blood. 118  The LabView control interface for the extraction system is capable of handling multiple protocols to adjust the specifics of the test to be performed.  A multi-layer laminate oscillatory flow PCR chip was fabricated for integration with the extraction system.  The oscillatory flow PCR scheme provides flexibility in cycle parameters while keeping thermal inertia low for rapid cycle times.  Thin polycarbonate PCR chips are prone to evaporation of sample fluid at the high PCR temperatures. This evaporation can effectively be minimized by layering the fluid channel between layers of water pressurized equal to the sample fluid.  BSA coating of polycarbonate PCR chips prevents PCR reagent loss and permits dimer-free PCR amplification. Advancements in membrane-based component integration were instrumental in the selection of PDMS as the substrate for the main fluid handling chip for the nucleic acid extraction system. This system has proven capable of handling a variety of sample preparation tasks in an effort to make automated nucleic acid sample preparation more practical and available. Such work is a step towards a universal sample preparation unit that can handle many kinds of samples from viruses to tissue samples and provide rapid, meaningful results from a lab-on-a-chip platform. The RNA detection abilities were demonstrated for two main applications. The first is the detection of E. Coli RNA to monitor drinking water supplies for harmful bacteria where RNA is a more appropriate genetic material for detection. RNA was also extracted from viral samples in tissue culture fluid to demonstrate another form of pathogen detection. These applications will open the 119 door for future pathogen screening and detection tests. The same system was shown to be capable of extracting DNA from whole blood samples. This project satisfies global demand for fast, simple, and adaptable genetic testing systems. Detection of DNA or RNA often requires amplification through PCR. To integrate an amplification module with the extraction system, a disposable oscillatory flow PCR chip was fabricated using multilayered thermally laminated polycarbonate sheets. Temperature conditions were numerically simulated and experimentally verified to produce temperature ramp rates as high as 60 ºC/s. A unique scheme for evaporation control was implemented. Successful amplification of DNA was achieved in 40 PCR cycles in approximately 20 minutes. 5.2 Contributions Contributions made over the course of completing this project are summarized in the following list: ▪ Utah system extraction unit for nucleic acid purification ▪ Multiscale rapid prototyping PDMS casting mold creation process with masked corona bonding ▪ Disposable glass fiber extraction chips ▪ Matlab code for solving 1D steady state, and 1D transient temperature analysis for a fluid layer in a composite wall geometry with varying layer thickness and material properties ▪ Numerical temperature analysis to determine convection coefficient for free convection for low Rayleigh number horizontal heated surface ▪ Numerical temperature analysis for transient temperature profiles inside 120 multilayer, variable material microfluidic device ▪ Novel sample evaporation control for polymer film chips to reduced evaporation to less than 6% over 40 cycles ▪ LabView program for interface and control of extraction unit ▪ Successful automated amplification of DNA template in 40 cycles in ~20 minutes. 5.3 Future Work While the system presented in this work has proven effective at extracting nucleic acids from multiple sample types, there are still several areas recommended for improvement in future iterations of the system and related research. Multiple sample types were used in the extraction system, but they were all in liquid form. Work still needs to be done in demonstration of the current system's ability to extract nucleic acids from an increased number of sample types. Some raw sample mediums of interest such as stool, soil, or other solid material present difficult challenges for use in microfluidic systems. Small channels could easily be blocked by excessive amounts of solid matter. A filtration mechanism added to the sample input will further increase the versatility of the sample preparation system. Filter material or integrated fabricated filter features could be chosen such that large particles in impure samples could be screened out in the sample input well, while smaller particles and organisms that don't pose the risk of channel blockage would pass through. Demonstration of extraction of the target molecule from soil, food, or stool samples would indeed demonstrate versatility of the sample preparation system. Akin to filtration, a sample preconcentration step would help detect minute 121 quantities of the target molecule when it would normally get lost among the more prevalent DNA and RNA molecules in the sample. As an illustration, contrast the demonstrated ability of the current system to extract human genomic DNA from blood versus the goal to find the nucleic acid signature of a scarce pathogen found in the blood stream. The more plentiful genetic material from the host and other organisms in the blood stream would compete for solid phase binding sites. The relatively dilute target (often with orders of magnitude fewer molecules) could end up fading into the background noise of the detection system. A preconcentration step that would rid the sample of unwanted organisms or genetic material would greatly increase the specificity of the extraction system. The difficulty in implementing a preconcentration step for a system that purports to be agnostic to sample type is that it is not feasible to sort mechanically solely on cell size. The size of the target would change based on the application as would the size of the unwanted background debris. Fortunately, the control chip is already designed for modular components, and perhaps future work could focus on interchangeable mechanical filter modules with different pore sizes that can be used as the initial sample input section. The mechanical separation method seems a better fit with the goal of universality than other preconcentration technologies such as functionalized magnetic beads that attach to target molecules using specific antibodies. While the functionalized beads perform well, they are highly specific to a single organism and require skilled handling for implementation. Another recommended advancement to improve the extraction system would be 122 to fully integrate the extraction and PCR modules into a single disposable device. The current modular method has a benefit in that some customers may have a detection or analysis method that does not require amplification. However, the sensitive nature of biomolecular tests begs for extreme care in avoiding cross-contamination, so reusable chips pose a threat. The current PDMS extraction chip could technically be considered a disposable, but the materials and fabrication process are more costly than what would be seen in a thermoplastic chip. Other suggested work for the future does not include such drastic changes to the system's over-all design. Instead work can focus on realizing an expanded role of features already in place in the current system. For example, the flexibility of the oscillatory PCR chip has not been fully explored. The fluid temperature behavior has been characterized, but the ability of the system to amplify nucleic acids has only been proven on a single molecule. The flexible nature of the automated PCR chip could be explored by performing RT-PCR on the chip. Also longer DNA strands could be amplified and protocols with tighter temperature tolerances could be performed to explore the limits of the PCR chip. Finally, integration of an analysis or detection module onto the system would make it a complete sample-in answer-out system. The main focus of the research to this point has been solely on extraction, purification, and amplification of the target, as there is already a wide range of devices that detect or analyze nucleic acids. However, simply because the main focus is on extraction does not mean that analysis cannot be an offered module for the system. Analysis could be integrated with the PCR module, where real time monitoring of fluorescence of the PCR process would be able to indicate the 123 presence of amplified target. The capability to perform such analysis would require little more than a light transmitter and fluorescence detector. LabView, which already controls the instrument function, can readily control and monitor the detection equipment at the same time. These recommendations will further improve a system that has demonstrated an ability to simplify nucleic acid extraction. This system has demonstrated improvements in components and subsystems that can be used by nonexperts. It has also proven to be a practical technology, used by nonexperts in the field in Canada and India, far from the supervision of the system creators. As the future work continues to expand on the past demonstrations of versatile nucleic acid extraction, this sample preparation system will be on the cutting edge of the transition to a commercially viable technology as it benefits researchers, clinicians, patients, and the public at large. APPENDIX A MICROFLUIDIC SAMPLE PREPARATION: CELL LYSIS AND NUCLEIC ACID PURIFICATION Kim, Jungkyu, Michael Johnson, Parker Hill, and Bruce K. Gale. "Microfluidic sample preparation: cell lysis and nucleic acid purification." Integrative biology 1, no. 10 (2009): 574-586. - Reprinted by permission of the Royal Society of Chemistry 125 126 127 128 129 130 131 132 133 134 135 136 137 APPENDIX B PROTOCOLS FOR EXTRACTION OF NUCLEIC ACID USING LABVIEW CONTROLLED INSTRUMENT 139 Stepwise Control Sequences Used to Perform Nucleic Acid Extraction Table B.1 Protocol steps for basic extraction of DNA Step 1 Description TE buffer pretreatment of filter Solenoid 4 80:00:00 006000 80:00:40 12 003000 00:E0:40 21 030000 00:E0:01 003000 00:02:01 14 005000 00:02:00 1 003000 10:00:00 006000 10:08:00 4 19 3 19 10 11 4 10 11 12 5 21 14 7 Load buffer pretreatment Hexcode 003000 2 6 Time (ms) 8 1 16 9 16 10 12 003000 00:A8:00 10 10 12 21 030000 00:A0:01 11 21 14 003000 00:02:01 12 14 005000 00:02:00 13 003000 80:00:00 14 4 15 006000 80:04:00 15 15 1 003000 10:04:00 16 15 1 16 007000 10:0C:00 17 15 16 11 7 8 003000 0C:4C:00 8 005000 0C:48:02 007000 0C:40:02 18 Sample mixing with load buffer 16 22 11 7 19 22 11 7 8 20 22 7 8 003000 0C:00:02 21 7 8 16 008000 0C:08:00 22 7 8 22 008000 0C:00:02 23 7 8 16 008000 0C:08:00 24 7 8 22 015000 0C:00:02 22 9 8 12 003000 08:90:02 9 8 12 21 060000 08:90:01 21 14 25 Mixing 4 Pass through filter 26 27 Out to waste 28 29 Wash 003000 00:02:01 14 010000 00:02:00 2 003000 20:00:00 30 2 16 006000 20:08:00 31 16 10 12 003000 00:A8:00 32 10 12 21 030000 00:A0:01 21 14 003000 00:02:01 010000 00:02:00 060000 01:82:00 003000 80:00:00 33 out to waste 34 14 35 Dry 5 36 Elution 4 12 14 37 4 19 008000 80:00:40 38 19 10 11 6 003000 02:60:40 39 10 11 6 20 030000 02:60:80 40 20 11 12 003000 00:C0:80 41 11 12 015000 00:C0:00 003000 00:00:00 42 140 Table B.2 Protocol steps for basic extraction of RNA Step 1 Description Mix 50uL of sample with 100uL of BPR (becomes Mix 1) 3 2 draw sample into CH3 3 17 17 6 6 7 5 3 15 6 3 17 15 7 17 6 7 15 3 4 move to CH4 8 Solenoid Time (ms) Hexcode 002000 40:00:00 005000 40:00:10 7 002000 06:00:10 15 008000 06:04:00 002000 40:04:00 005000 40:04:10 002000 06:04:10 006000 06:04:00 003000 10:04:00 15 6 7 9 Introduce and mix 350uL of B-ME-RLT to MIX 2 (now Mix 3) 15 1 10 draw RLT 15 1 16 010000 10:0C:00 11 prepare to move Lysed sample to CH5 15 16 11 7 8 003000 0C:4C:00 12 Move mixtures to CH5, prepare to move to CH6 11 16 7 8 22 010000 0C:48:02 22 13 16 7 8 003000 0C:08:02 14 move all mixtures to CH6 7 8 22 010000 0C:00:02 15 move to CH5 7 8 16 003000 0C:08:00 16 move all mixtures to CH6 7 8 22 003000 0C:00:02 17 move to CH5 7 8 16 003000 0C:08:00 18 move all mixtures to CH6 7 8 22 003000 0C:00:02 19 move to CH5 7 8 16 003000 0C:08:00 20 7 8 22 015000 0C:00:02 21 move all mixtures to CH6 Introduce and mix 250uL of Ethanol to MIX 3 (now Mix 4) 22 4 002000 80:00:02 22 Draw EtOH to CH4 22 4 15 006000 80:04:02 22 15 11 002000 00:44:02 22 16 11 007000 00:48:02 22 16 7 002000 0C:08:02 23 24 send to CH5 25 8 26 move to CH6 22 7 8 006000 0C:00:02 27 move to CH5 16 7 8 003000 0C:08:00 28 Move to CH6 22 7 8 003000 0C:00:02 29 FWD 7 8 16 003000 0C:08:00 30 BCKWD 7 8 22 003000 0C:00:02 31 FWD 7 8 16 003000 0C:08:00 32 BCKWD 7 8 22 010000 0C:00:02 33 Pass MIX 4 through the silica filter (prepare) 22 9 8 12 001000 08:90:02 34 Pass MIX 4 through the silica filter (prepare) 22 9 8 12 002000 08:90:02 35 FWD 21 9 8 12 020000 08:90:01 36 BCKWD 22 9 8 12 020000 08:90:02 37 FWD 21 9 8 12 045000 08:90:01 38 Send waste to waste bottle…..Ready…. 21 14 003000 00:02:01 39 Go! 14 020000 00:02:00 40 drain to waste 14 5 008000 01:82:00 41 14 12 003000 00:82:00 42 14 001000 00:02:00 003000 08:90:00 010000 08:90:01 43 44 Make sure everything is empty (open valves) Open ch 7 12 8 9 12 8 9 12 21 141 Table B.2 continued Step Description 45 Send waste to waste bottle…..Ready…. 21 Solenoid 46 Go! 14 47 drain to waste 14 14 5 48 14 12 49 14 12 Time (ms) Hexcode 003000 00:02:01 008000 00:02:00 008000 01:82:00 003000 00:82:00 001000 00:02:00 003000 40:00:00 010000 40:08:00 50 Introduce 350uL of rpe buffer and pass thru silica filter 3 51 draw rpe 3 16 52 prepare to move thru filter 16 10 12 002000 00:A8:00 53 GO! 21 10 12 025000 00:A0:01 54 Send waste to waste bottle 21 14 003000 00:02:01 010000 00:02:00 008000 01:82:00 003000 00:82:00 55 56 14 drain to waste 57 58 14 5 14 12 12 14 001000 00:02:00 59 Introduce 350uL of Rpe buffer and pass thru silica filter 3 003000 40:00:00 60 draw Rpe 3 16 010000 40:08:00 61 prepare to move thru filter 16 10 12 003000 00:A8:00 62 GO! 21 10 12 025000 00:A0:01 63 Send waste to waste bottle 21 14 003000 00:02:01 010000 00:02:00 060000 01:82:00 64 65 14 14 5 66 14 12 003000 00:82:00 67 14 001000 00:02:00 4 3000 80:00:00 8000 80:00:40 68 drain to waste Elution Step 12 69 4 19 70 19 10 11 6 3000 02:60:40 71 10 11 6 20 3000 02:60:80 72 19 10 11 6 3000 02:60:40 73 10 11 6 20 3000 02:60:80 74 19 10 11 6 3000 02:60:40 75 10 11 6 20 3000 02:60:80 76 19 10 11 6 3000 02:60:40 77 10 11 6 20 3000 02:60:80 78 19 10 11 6 3000 02:60:40 79 10 11 6 20 3000 02:60:80 80 19 10 11 6 3000 02:60:40 81 10 11 6 20 6000 02:60:80 20 10 11 6 4000 02:E0:80 20 11 12 3000 00:C0:80 11 12 6000 00:C0:00 3000 00:00:00 82 83 output 84 85 end 12 142 Table B.3 Protocol steps for extraction of RNA from viral samples Step 1 Description Mix 50uL of sample with 100uL of BPR (becomes Mix 1) 2 draw sample into CH3 Solenoid 3 6 7 3 6 7 15 17 6 7 15 6 7 15 5 3 15 6 3 17 15 7 17 6 7 15 3 4 move to CH4 8 6 7 Introduce and mix 350uL of B-ME-RLT to MIX 2 (now Mix 3) 15 1 10 draw RLT 15 1 16 11 prepare to move Lysed sample to CH5 Move mixtures to CH5, prepare to move to CH6 15 16 11 11 16 16 7 9 12 13 17 15 Time (ms) Hexcode 003000 00:00:00 008000 46:00:00 003000 46:04:10 008000 06:04:10 003000 06:04:00 005000 40:04:00 003000 40:04:10 009000 06:04:10 003000 06:04:00 010000 10:04:00 7 8 003000 10:0C:00 7 8 22 010000 0C:4C:00 8 22 003000 0C:48:02 14 move all mixtures to CH6 7 8 22 010000 0C:08:02 15 move to CH5 7 8 16 003000 0C:00:02 16 move all mixtures to CH6 7 8 22 003000 0C:08:00 17 move to CH5 7 8 16 003000 0C:00:02 18 move all mixtures to CH6 7 8 22 003000 0C:08:00 19 move to CH5 7 8 16 003000 0C:00:02 20 7 8 22 020000 0C:08:00 21 move all mixtures to CH6 Introduce and mix 250uL of Ethanol to MIX 3 (now Mix 4) 22 4 003000 0C:00:02 22 Draw EtOH to CH4 22 4 15 010000 80:00:02 22 15 11 003000 80:04:02 22 16 11 010000 00:44:02 22 16 7 003000 00:48:02 8 006000 0C:08:02 23 24 send to CH5 25 8 26 move to CH6 22 7 27 move to CH5 16 7 006000 0C:00:02 28 Move to CH6 22 7 8 006000 0C:08:00 29 FWD 7 8 16 006000 0C:00:02 30 BCKWD 7 8 22 006000 0C:08:00 31 FWD 7 8 16 006000 0C:00:02 32 7 8 22 020000 0C:08:00 22 9 8 12 001000 0C:00:02 34 BCKWD Pass MIX 4 through the silica filter (prepare) Pass MIX 4 through the silica filter (prepare) 22 9 8 12 003000 08:90:02 35 FWD 21 9 8 12 030000 08:90:02 36 BCKWD 22 9 8 12 030000 08:90:01 37 FWD 21 9 8 12 060000 08:90:02 38 Send waste to waste bottle…..Ready…. 21 14 003000 08:90:01 39 Go! 14 030000 00:02:01 40 drain to waste 008000 00:02:00 003000 01:82:00 001000 00:82:00 33 14 5 41 14 12 42 14 8 12 143 Table B.3 continued Step 43 Description Make sure everything is empty (open valves) Solenoid 8 9 12 44 Open ch 7 8 9 12 45 Send waste to waste bottle…..Ready…. 21 14 46 Go! 14 47 drain to waste 14 5 48 14 12 49 14 12 21 Time (ms) Hexcode 003000 00:02:00 025000 08:90:00 003000 08:90:01 015000 00:02:01 008000 00:02:00 003000 01:82:00 001000 00:82:00 003000 00:02:00 010000 20:00:00 50 Introduce 350uL of RW1 buffer and pass thru silica filter 2 51 draw RW1 2 16 52 prepare to move thru filter 16 10 12 003000 20:08:00 53 GO! 21 10 12 010000 00:A8:00 54 16 10 12 010000 00:A0:01 55 21 10 12 010000 00:A8:00 56 16 10 12 010000 00:A0:01 57 21 10 12 040000 00:A8:00 21 14 003000 00:A0:01 025000 00:02:01 008000 00:02:00 003000 01:82:00 001000 00:82:00 003000 00:02:00 010000 20:00:00 58 Send waste to waste bottle 59 60 14 drain to waste 14 5 61 14 12 62 14 12 63 Introduce 350uL of RW1 buffer and pass thru silica filter 2 64 draw RW1 2 16 65 prepare to move thru filter 16 10 12 003000 20:08:00 66 GO! 21 10 12 040000 00:A8:00 67 Send waste to waste bottle 21 14 003000 00:A0:01 010000 00:02:01 008000 00:02:00 003000 01:82:00 001000 00:82:00 003000 00:02:00 010000 40:00:00 68 69 14 14 5 70 drain to waste 14 12 71 14 12 72 Introduce 350uL of rpe buffer and pass thru silica filter 3 73 draw rpe 3 16 74 prepare to move thru filter 16 10 12 003000 40:08:00 75 GO! 21 10 12 010000 00:A8:00 76 16 10 12 010000 00:A0:01 77 21 10 12 010000 00:A8:00 78 16 10 12 010000 00:A0:01 79 21 10 12 040000 00:A8:00 21 14 003000 00:A0:01 010000 00:02:01 008000 00:02:00 003000 01:82:00 14 001000 00:82:00 3 003000 00:02:00 80 Send waste to waste bottle 81 82 14 drain to waste 83 84 85 Introduce 350uL of Rpe buffer and pass thru silica filter 14 5 14 12 12 144 Table B.3 continued Step Description 86 draw Rpe 87 88 Solenoid 3 16 prepare to move thru filter 16 10 GO! Time (ms) Hexcode 010000 40:00:00 12 003000 40:08:00 21 10 12 010000 00:A8:00 89 16 10 12 010000 00:A0:01 90 21 10 12 010000 00:A8:00 91 16 10 12 010000 00:A0:01 92 21 10 12 040000 00:A8:00 21 14 003000 00:A0:01 010000 00:02:01 008000 00:02:00 003000 01:82:00 001000 00:82:00 003000 00:02:00 93 Send waste to waste bottle 94 95 14 drain to waste 96 97 14 5 14 12 12 14 98 Introduce 350uL of Rpe buffer and pass thru silica filter 3 99 draw Rpe 3 16 010000 40:00:00 100 prepare to move thru filter 16 10 12 003000 40:08:00 101 GO! Send waste to waste bottle 21 10 12 40000 00:A8:00 102 103 104 14 3000 00:A0:01 10000 00:02:01 120000 00:02:00 3000 01:82:01 14 6000 00:82:01 4 3000 00:02:00 8000 80:00:00 14 Long drying time (in current MF protocol its step #40) 105 106 107 21 Elution Step 5 12 21 14 12 21 14 108 4 19 109 19 10 11 6 3000 80:00:40 110 10 11 6 20 5000 02:60:40 111 19 10 11 6 5000 02:60:80 112 10 11 6 20 5000 02:60:40 113 19 10 11 6 5000 02:60:80 114 10 11 6 20 5000 02:60:40 115 19 10 11 6 5000 02:60:80 116 10 11 6 20 5000 02:60:40 117 19 10 11 6 5000 02:60:80 118 10 11 6 20 5000 02:60:40 119 19 10 11 6 5000 02:60:80 120 10 11 6 20 5000 02:60:40 20 10 11 6 6000 02:60:80 20 11 12 3000 02:E0:80 11 12 10000 00:C0:80 3000 00:C0:00 0 00:00:00 121 122 output 123 124 125 end 12 145 Table B.4 Steps for extracting RNA from bacteria samples. Step 1 2 Description Open Valve for Port I5 draw 50 ul sample into CH3 3 Solenoid 3 17 10000 40:00:10 6 7 3000 06:00:10 6 7 15 10000 06:04:00 3000 20:04:00 5000 20:04:10 10000 06:04:10 3000 06:04:00 3000 20:04:00 5000 20:04:10 10000 06:04:10 3000 06:04:00 300000 00:04:00 3000 10:04:00 3000 10:04:10 6000 06:04:10 3000 06:04:00 3000 10:04:00 3000 10:04:10 6000 06:04:10 3000 06:04:00 600000 00:04:00 3000 10:04:00 5 Open valve for port I4 15 2 6 draw 50 ul BPR to ch3 15 2 17 7 move BPR to CH4 15 17 6 7 15 6 9 Open valve for port I4 15 2 10 draw 50 ul BPR to ch3 15 2 17 11 move to CH4 15 17 6 15 6 7 Incubate Mixture for 5min at room temp (20-25C) Open valve for port I3 15 15 Draw in 50 ul Lysozyme to ch3 15 1 17 16 move to CH4 15 17 6 15 6 7 13 14 17 Open valve for port I3 15 15 1 Draw in 50 ul Lysozyme to ch3 15 1 17 20 move to CH4 15 17 6 15 6 21 22 23 Incubate Mix 2 for 10min at ROOM TEMP (20-25C) Open valve for port I8 24 draw 350 ul RLT 25 prepare to move Lysed sample to CH5 Move mixtures to CH5, prepare to move to CH6 26 27 7 7 1 19 18 40:00:00 3 move to CH4 12 Hexcode 3000 17 4 8 Time (ms) 7 7 7 15 15 15 1 1 16 10000 10:0C:00 16 11 7 8 3000 0C:4C:00 11 16 7 8 22 10000 0C:48:02 16 7 8 3000 0C:08:02 10000 0C:00:02 15 22 28 move all mixtures to CH6 7 8 22 29 move to CH5 7 8 16 3000 0C:08:00 30 move all mixtures to CH6 Open valve for port I7 7 8 22 20000 0C:00:02 22 4 3000 80:00:02 22 4 15 10000 80:04:02 22 15 11 3000 00:44:02 22 16 11 10000 00:48:02 22 16 7 3000 0C:08:02 31 32 Draw 250 ul EtOH to CH4 33 34 send to CH5 35 8 36 move to CH6 22 7 8 6000 0C:00:02 37 move to CH5 16 7 8 6000 0C:08:00 38 22 7 8 20000 0C:00:02 39 Move to CH6 Pass MIX 4 through the silica filter (prepare) 22 9 8 12 3000 08:90:02 40 FWD 21 9 8 12 30000 08:90:01 41 BCKWD 22 9 8 12 30000 08:90:02 42 FWD 21 9 8 12 60000 08:90:01 43 Send waste to waste bottle…..Ready…. 21 14 3000 00:02:01 146 Table B.4 continued Step Description Solenoid 14 45 Go! Make sure everything is empty (open valves) 8 9 12 46 Open ch 7 8 9 12 47 Send waste to waste bottle…..Ready…. 21 14 48 Go! 14 49 Open valve for I9 2 50 draw 350 ul RW1 2 16 51 prepare to move thru filter 16 10 52 GO! 21 10 53 Send waste to waste bottle 21 14 44 21 Time (ms) Hexcode 30000 00:02:00 3000 08:90:00 25000 08:90:01 3000 00:02:01 15000 00:02:00 3000 20:00:00 10000 20:08:00 12 3000 00:A8:00 12 40000 00:A0:01 3000 00:02:01 14 25000 00:02:00 55 Open Valve for I1 4 3000 80:00:00 56 Draw water…. 4 19 5000 80:00:40 57 prepare to send to CH5 19 10 11 3000 00:60:40 58 send to CH5 10 11 16 10000 00:68:00 7 8 16 3000 0C:08:00 7 8 22 10000 0C:00:02 22 13 3000 00:01:02 13 21 10000 00:01:01 21 14 54 59 60 send to CH6 61 62 send to CH7 63 3000 00:02:01 64 send to waste 14 15000 00:02:00 65 Repeat RW1 well rinse 4 3000 80:00:00 66 Draw water…. 4 19 5000 80:00:40 19 10 11 3000 00:60:40 10 11 16 10000 00:68:00 7 8 16 3000 0C:08:00 7 8 22 10000 0C:00:02 22 13 3000 00:01:02 13 21 20000 00:01:01 21 14 67 68 send to CH5 69 70 send to CH6 71 72 send to CH7 73 74 75 76 send to waste Open valve for port I2 14 draw DNAse into ch2 1 77 1 18 18 3 3000 00:02:01 40000 00:02:00 3000 10:00:00 10000 10:00:20 3000 40:00:20 10000 40:08:20 3000 00:38:20 10000 00:38:00 78 Draw RDD to CH5 18 3 16 79 DNAse moves to CH5 18 16 9 16 9 10 16 10 12 3000 00:A8:00 10 12 21 40000 00:A0:01 21 14 3000 00:02:01 80 81 82 move to filter 83 84 85 86 87 10 junk to waste Incubate DNAse 1 on silica filter for 5min 20-25C Open valve for port I2 14 20000 00:02:00 0 300000 00:00:00 3000 10:00:00 draw DNAse ONLY 1 10000 10:00:20 1 18 147 Table B.4 continued Step Description 88 Solenoid 18 3 89 draw RDD ONLY 18 3 16 90 DNAse moves to CH5 Hexcode 3000 40:00:20 10000 40:08:20 18 16 9 3000 00:38:20 91 16 9 10 10000 00:38:00 92 16 10 12 3000 00:A8:00 10 12 21 45000 00:A0:01 21 14 3000 00:02:01 93 move to filter 94 10 Time (ms) 14 20000 00:02:00 96 junk to waste Incubate DNAse 1 on silica filter for 5min 20-25C 0 300000 00:00:00 97 Open valve to port I9 2 98 draw RW1 2 16 16 10 95 99 100 send to filter 21 10 101 Send waste to waste bottle 21 14 3000 20:00:00 10000 20:08:00 12 3000 00:A8:00 12 40000 00:A0:01 3000 00:02:01 14 15000 00:02:00 103 Open valve for port I6 4 3000 80:00:00 104 draw RPE 4 17 10000 80:00:10 17 6 7 3000 06:00:10 15 10000 06:04:00 10000 80:04:00 3000 00:04:00 3000 00:44:00 102 105 106 move to CH4 6 7 107 Draw EtOH to CH4 4 15 108 15 109 15 11 11 16 8000 00:48:00 16 10 12 3000 00:A8:00 10 12 21 25000 00:A0:01 21 14 3000 00:02:01 15000 00:02:00 110 send to CH 5 111 112 send through filter 113 14 115 send to waste Suck in 50uL of RPE buffer and 200uL of ethanol and mix 116 draw RPE 4 17 17 6 114 117 4 118 move to CH4 6 7 119 Draw EtOH to CH4 4 15 3000 80:00:00 10000 80:00:10 7 3000 06:00:10 15 10000 06:04:00 10000 80:04:00 3000 00:04:00 3000 00:44:00 120 15 121 15 11 11 16 8000 00:48:00 16 10 12 3000 00:A8:00 10 12 21 25000 00:A0:01 21 14 122 send to CH 5 123 124 send through filter 125 14 127 send to waste Suck in 50uL of RPE buffer and 200uL of ethanol and mix 128 draw RPE 4 17 17 6 126 129 4 130 move to CH4 6 7 131 Draw EtOH to CH4 4 15 3000 00:02:01 15000 00:02:00 3000 80:00:00 10000 80:00:10 7 3000 06:00:10 15 10000 06:04:00 10000 80:04:00 148 Table B.4 continued Step Description Solenoid 132 15 133 15 11 11 16 16 10 12 10 12 21 21 14 134 send to CH 5 135 136 send through filter 137 14 139 send to waste Suck in 50uL of RPE buffer and 200uL of ethanol and mix 140 draw RPE 4 17 17 6 138 141 4 142 move to CH4 6 7 143 Draw EtOH to CH4 4 15 145 146 send to CH 5 147 148 send through filter 149 150 send to waste Long drying time (in current MF protocol its step #40) 00:04:00 3000 00:44:00 8000 00:48:00 3000 00:A8:00 25000 00:A0:01 3000 00:02:01 15000 00:02:00 3000 80:00:00 80:00:10 7 3000 06:00:10 15 10000 06:04:00 10000 80:04:00 3000 00:04:00 15 11 3000 00:44:00 11 16 8000 00:48:00 16 10 12 3000 00:A8:00 10 12 21 25000 00:A0:01 21 14 3000 00:02:01 15000 00:02:00 14 5 12 21 152 12 21 14 153 21 14 154 14 155 151 Hexcode 3000 10000 15 144 Time (ms) 14 300000 01:82:01 3000 00:82:01 3000 00:02:01 3000 00:02:00 0 3000 00:00:00 3000 80:00:00 156 open valve for port I1 4 157 Draw in 150 ul H20 4 19 10000 80:00:40 158 open valves 19 10 11 6 3000 02:60:40 159 pass through filter 10 11 6 20 15000 02:60:80 3000 00:00:80 160 20 149 Table B.5 Protocol steps for extracting DNA from blood samples Step 1 Description proteinase solution in Solenoid Time (ms) 4 Hexcode 002000 80:00:00 010000 80:00:10 2 4 17 3 17 7 6 002000 06:00:10 4 7 8 15 010000 0C:04:00 15 3 002000 40:04:00 004000 40:04:10 002000 06:04:10 004000 06:04:00 002000 40:04:00 004000 40:04:10 002000 06:04:10 004000 06:04:00 002000 20:04:00 010000 20:04:10 5 100 ul sample in 6 15 3 17 7 15 17 6 8 15 6 7 7 9 15 3 10 15 3 17 11 15 17 6 15 6 7 15 2 14 15 2 17 15 15 17 6 002000 06:04:10 16 15 6 7 010000 06:04:00 17 6 7 003000 06:00:10 18 15 6 7 003000 06:04:00 19 17 6 7 003000 06:00:10 20 15 6 7 003000 06:04:00 21 17 6 7 003000 06:00:10 15 6 7 003000 06:04:00 600000 00:04:00 002000 10:04:00 004000 10:04:10 002000 06:04:10 004000 06:04:00 002000 10:04:00 004000 10:04:10 12 13 17 50 ul AL buffer in Mix 22 23 incubate for 10 minutes 15 24 input 100 ul of etoh 15 7 7 1 25 15 1 17 26 15 17 6 27 15 6 7 28 15 1 29 15 1 17 30 15 17 6 7 11 002000 06:44:10 31 15 6 7 16 11 004000 06:4C:00 32 6 7 11 16 004000 06:48:00 33 16 11 002000 00:48:00 15 11 003000 00:44:00 35 16 11 003000 00:48:00 36 15 11 003000 00:44:00 37 16 11 003000 00:48:00 34 38 Mix 16 13 12 002000 00:89:00 39 13 12 21 040000 00:81:01 40 21 14 002000 00:02:01 41 14 002000 00:02:00 42 14 004000 01:82:00 43 14 008000 00:02:00 002000 10:00:00 015000 10:08:00 44 45 Pass through filter 7 wash with AW1 5 1 1 16 12 150 Table B.5 continued Step Description Solenoid Time (ms) Hexcode 46 16 13 12 002000 00:89:00 47 12 13 21 040000 00:81:01 48 21 14 002000 00:02:01 49 14 010000 00:02:00 2 002000 20:00:00 50 wash with AW2 51 2 16 015000 20:08:00 52 16 13 12 002000 00:89:00 53 12 13 21 040000 00:81:01 54 21 14 002000 00:02:01 55 14 010000 00:02:00 120000 01:82:00 002000 80:00:00 004000 80:00:40 002000 02:60:40 010000 02:60:80 001000 02:E0:80 002000 00:C0:80 003000 00:C0:00 001000 00:00:00 56 Dry 5 57 elution 4 12 14 58 4 19 59 19 10 11 6 60 10 11 6 20 61 20 11 6 10 62 20 11 12 63 11 12 64 12 APPENDIX C MATLAB CODE FOR 1D STEADY STATE AND 1D TRANSIENT SIMULATIONS FOR INTERNAL TEMPERATURES OF MULTILAYER POLYCARBONATE OSCILLATORY FLOW PCR CHIP 152 %% 1D trasient simulation of temperature in multi-layer thin chip clear it1=0; for tbot2=[0.000127]%0.00024,0.000376,0.000501] clear T Tinitial Ts a1 alpha chamberx cs ks r rhos thicknesses times ts u uold x; it1=it1+1; tmax = 1; % end time ti=.01; % time interval to sample answer for plotting "every ti seconds" Tinf=20+273; % Ambient Air Temperature Tgoal=273+60; % temperature you want the fluid to reach h0=.0628*(Tgoal-273)+9.2062+10; % Set convection coefficient q0=18*(Tgoal-273)-571.21 %heat flux from 3d simulation w/m^2 fluidtemp=273+95; % Set temperature of fluid to be pumped into channel tins1=0.000127; % Thickness of insulating layer on top of chip tinsw1=0.000126; % Thicnkess of top water layer ttop=.000127; % Thicnkess of top chip layer tmid=.000127; % Thicnkess of middle chip layer *sample* tbot=.000127; % Thicnkess of bottom chip layer tbotw=.000127; % thickness of bottom water layer tbot2=0.000127; % thickness of bottom insulating layer tins2=0.0003; % Thicnkess of layer between chip and heater %heater between these two layers tins3=0; % Thicnkess of layer under heater points=25; % number of points per layer (assuming no layer is zero) thicknesses=[tins1,tinsw1,ttop,tmid,tbot,tbotw,tbot2,tins2,tins3] ; % alternative number of points per layer if min(thicknesses)>0 si=min(thicknesses)/points; % space increment calculate else si=.00002; % manually set alternative space increment end x=[0:si:sum(thicknesses)]; % x vector distance through chip % Thermal conductivity of materials database kpdms=.16; ksilicond=.21; kh20=.58; kpolyc=.2; 153 kalum=250; % Density of materials database rpdms=965; rsilicond=965; rh20=1000; rpolyc=1220; ralum=2700; % Heat capacity of materials database cpdms=120; csilicond=120; ch20=4181; %4181 cpolyc=1200; calum=910; %Assign properties to layers kins1=kpolyc; rins1=rpolyc; cins1=cpolyc; kinsw1=kh20; rinsw1=rh20; cinsw1=ch20; ktop=kpolyc; rtop=rpolyc; ctop=cpolyc; kmid=kh20; rmid=rh20; cmid=ch20; kbot=kpolyc; rbot=rpolyc; cbot=cpolyc; kbotw=kh20; rbotw=rh20; cbotw=ch20; kbot2=kpolyc; rbot2=rpolyc; cbot2=cpolyc; kins2=kalum; rins2=ralum; cins2=calum; kins3=kpdms; rins3=rpdms; cins3=cpdms; % Create vectors of the properties associated with the entire geometry ks=ones(1,length(x)); ks(find(x<tins3))=kins3; ks(find(x>=tins3 & x<tins3+tins2))=kins2; ks(find(x>=tins3+tins2 & x<tins3+tins2+tbot2))=kbot2; ks(find(x>=tins3+tins2+tbot2 & x<tins3+tins2+tbot2+tbotw))=kbotw; ks(find(x>=tins3+tins2+tbot2+tbotw & x<tins3+tins2+tbot2+tbotw+tbot))=kbot; ks(find(x>=tins3+tins2+tbot2+tbotw+tbot & x<tins3+tins2+tbot2+tbotw+tbot+tmid))=kmid; ks(find(x>=tins3+tins2+tbot2+tbotw+tbot+tmid & x<tins3+tins2+tbot2+tbotw+tbot+tmid+ttop))=ktop; ks(find(x>=tins3+tins2+tbot2+tbotw+tbot+tmid+ttop & x<tins3+tins2+tbot2+tbotw+tbot+tmid+ttop+tinsw1))=kinsw1; ks(find(x>=tins3+tins2+tbot2+tbotw+tbot+tmid+ttop+tinsw1 & x<tins3+tins2+tbot2+tbotw+tbot+tmid+ttop+tinsw1+tins1))=kins1; %ks(find(x>=tins3+tins2+tbot+tmid+ttop))=kins1; 154 rhos=ones(1,length(x)); rhos(find(x<tins3))=rins3; rhos(find(x>=tins3 & x<tins3+tins2))=rins2; rhos(find(x>=tins3+tins2 & x<tins3+tins2+tbot2))=rbot2; rhos(find(x>=tins3+tins2+tbot2 & x<tins3+tins2+tbot2+tbotw))=rbotw; rhos(find(x>=tins3+tins2+tbot2+tbotw & x<tins3+tins2+tbot2+tbotw+tbot))=rbot; rhos(find(x>=tins3+tins2+tbot2+tbotw+tbot & x<tins3+tins2+tbot2+tbotw+tbot+tmid))=rmid; rhos(find(x>=tins3+tins2+tbot2+tbotw+tbot+tmid & x<tins3+tins2+tbot2+tbotw+tbot+tmid+ttop))=rtop; rhos(find(x>=tins3+tins2+tbot2+tbotw+tbot+tmid+ttop & x<tins3+tins2+tbot2+tbotw+tbot+tmid+ttop+tinsw1))=rinsw1; rhos(find(x>=tins3+tins2+tbot2+tbotw+tbot+tmid+ttop+tinsw1 & x<tins3+tins2+tbot2+tbotw+tbot+tmid+ttop+tinsw1+tins1))=rins1; cs=ones(1,length(x)); cs(find(x<tins3))=cins3; cs(find(x>=tins3 & x<tins3+tins2))=cins2; cs(find(x>=tins3+tins2 & x<tins3+tins2+tbot2))=cbot2; cs(find(x>=tins3+tins2+tbot2 & x<tins3+tins2+tbot2+tbotw))=cbotw; cs(find(x>=tins3+tins2+tbot2+tbotw & x<tins3+tins2+tbot2+tbotw+tbot))=cbot; cs(find(x>=tins3+tins2+tbot2+tbotw+tbot & x<tins3+tins2+tbot2+tbotw+tbot+tmid))=cmid; cs(find(x>=tins3+tins2+tbot2+tbotw+tbot+tmid & x<tins3+tins2+tbot2+tbotw+tbot+tmid+ttop))=ctop; cs(find(x>=tins3+tins2+tbot2+tbotw+tbot+tmid+ttop & x<tins3+tins2+tbot2+tbotw+tbot+tmid+ttop+tinsw1))=cinsw1; cs(find(x>=tins3+tins2+tbot2+tbotw+tbot+tmid+ttop+tinsw1 & x<tins3+tins2+tbot2+tbotw+tbot+tmid+ttop+tinsw1+tins1))=cins1; % plot(ks) n1=0; Th=Tgoal; err=1; tcorr=.1; % while err>0.001 % n1=n1+1; % Th=Th+tcorr; % % U=1/(ttop/ktop+tmid/kmid+tbot/kbot+tins2/kins2); % calculating 155 % % the thermal resistance to find q across the entire chip % %Ttop=(h0*Tinf+Th*U)/(h0+U); % find temperature on boundary with ambient % q=q0; % calculate heat flux % % % use q to find the temperature at each interface % Tins2=-q*tins2/kins2+Th; % Tbot2=-q*tbot2/kbot2+Tins2; % Tbotw=-q*tbotw/kbotw+Tbot2; % Tbot=-q*tbot/kbot+Tbotw; % Tmid=-q*tmid/kmid+Tbot; % Ttop=-q*ttop/ktop+Tmid; % Tinsw1=-q*tinsw1/kinsw1+Ttop; % Tins1=-q*tins1/kins1+Tinsw1; % % % fluid chamber is the area of interest, find the average temperature of % % the fluid in the chamber % Tchamber(n1)=(Tmid+Tbot)/2; % err=abs(Tgoal-Tchamber(n1)); % tcorr=Tgoal-Tchamber(n1); % end q=q0; % calculate heat flux % use q to find the temperature at each interface Tins2=-q*tins2/kins2+Th; Tbot2=-q*tbot2/kbot2+Tins2; Tbotw=-q*tbotw/kbotw+Tbot2; Tbot=-q*tbot/kbot+Tbotw; Tmid=-q*tmid/kmid+Tbot; Ttop=-q*ttop/ktop+Tmid; Tinsw1=-q*tinsw1/kinsw1+Ttop; Tins1=-q*tins1/kins1+Tinsw1; % fluid chamber is the area of interest, find the average temperature of % the fluid in the chamber n1=1; Tchamber(n1)=(Tmid+Tbot)/2; %%% locate the section of the x-vector associated with the fluid chamber 156 chamberx=find(x>tins3+tins2+tbot2+tbotw+tbot & x<=tins3+tins2+tbot2+tbotw+tbot+tmid); %Plot the calculated temperature profile in the chip ts=[0,tins2,tins2+tbot2,tins2+tbot2+tbotw,tins2+tbot2+tbotw+tbot, tins2+tbot2+... tbotw+tbot+tmid,tins2+tbot2+tbotw+tbot+tmid+ttop,tins2+tbot2+ tbotw+tbot+tmid+ttop+tinsw1,... tins2+tbot2+tbotw+tbot+tmid+ttop+tinsw1+tins1]; Ts=[Th,Tins2,Tbot2,Tbotw,Tbot,Tmid,Ttop,Tinsw1,Tins1]; plot(ts,Ts) disp(['heatertemp temp: ',num2str(Th-273)]) disp(['chamber temp: ',num2str(Tchamber-273), ' +/- ', num2str(Tchamber-Tmid)]) disp(['surface temp: ',num2str(Ttop-273)]) % %% % Calculate Biot number to determine if lumped capacitance method can be % used % Rcond=(tins1/kins1+ttop/ktop+tmid/kmid+tbot/kbot+tins2/kins2); % Rconv=1/h0; % Bi=Rcond/Rconv % %% transient % --- Define constants and initial condition L = max(x); % length of domain in x direction alpha=ks./(rhos.*cs); nx = length(x); % number of nodes in x direction dx = L/(nx-1); nt = ceil(max(alpha)*tmax/(.5*dx.^2)); % number of time steps dt = tmax/(nt); % rhos=1000*ones(1,length(x)); % cs=1200*ones(1,length(x)); % alpha=ones(1,length(x))*kh20./(rh20.*ch20); % r=.25*ones(1,length(x)) r = alpha.*dt./dx.^2; plot(r) % %% set initial temperature to what was calculated above 157 Tinitial=ones(1,length(x)); Tinitial(1)=Th; for k=2:length(x) if Tinitial(k-1)>=Tins2; Tinitial(k)=Tinitial(k-1)-(Th-Tins2)/(tins2/dx); elseif Tinitial(k-1)>=Tbot2; Tinitial(k)=Tinitial(k-1)-(Tins2-Tbot2)/(tbot2/dx); elseif Tinitial(k-1)>=Tbotw; Tinitial(k)=Tinitial(k-1)-(Tbot2-Tbotw)/(tbotw/dx); elseif Tinitial(k-1)>=Tbot; Tinitial(k)=Tinitial(k-1)-(Tbotw-Tbot)/(tbot/dx); elseif Tinitial(k-1)>=Tmid; Tinitial(k)=Tinitial(k-1)-(Tbot-Tmid)/(tmid/dx); elseif Tinitial(k-1)>=Ttop; Tinitial(k)=Tinitial(k-1)-(Tmid-Ttop)/(ttop/dx); elseif Tinitial(k-1)>=Tinsw1; Tinitial(k)=Tinitial(k-1)-(Ttop-Tinsw1)/(tinsw1/dx); else Tinitial(k-1)>=Tins1; Tinitial(k)=Tinitial(k-1)-(Tinsw1-Tins1)/(tins1/dx); end end plot(x,Tinitial) % %% % --- Loop over time steps t = 0; % u = Th*ones(1,length(x));% initial condition % u(1)=Th; % u(length(u))=Tinf; u=Tinitial; % initial condition as solved by 1D steady state u(chamberx)=fluidtemp; chambertemp(1)=mean(u(chamberx)) times(1)=0; T(1,:)=u; dtdx2=dt/dx^2; figure (2) a1=(alpha(2:nx-1)*dtdx2./(ks(2:nx-1)*4).*(ks(3:nx)-ks(1:nx-2))); plot(a1) % %% figure(3) tic; 158 % subplot(2,2,1) for m=1:nt uold = u; % prepare for next step t = t + dt; for j=2:nx-1 %u(j) = r(j-1)*uold(j-1) + r2(j)*uold(j) + r(j)*uold(j+1); u(j)=dtdx2*alpha(j)*(uold(j-1)2*uold(j)+uold(j+1))+uold(j)+... alpha(j)*dtdx2/(ks(j)*2)*(ks(j+1)ks(j))*(uold(j+1)-uold(j)); end u(end)=(ks(end)/(h0*dx)*uold(nx1)+Tinf)/(1+ks(end)/(h0*dx)); %convective boundary condition % vectorized % u(2:nx-1)=dtdx2.*alpha(2:nx-1).*(uold(1:nx-2)-2*uold(2:nx1)+uold(3:nx))+uold(2:nx-1)+... % alpha(2:nx-1).*dtdx2./(ks(2:nx-1)*4).*(ks(3:nx)ks(1:nx-2)).*(uold(3:nx)-uold(1:nx-2)); % % u(end)=(ks(end)/(h0*dx)*uold(nx1)+Tinf)/(1+ks(end)/(h0*dx)); %convective boundary condition if mod(t,ti)<mod(t-dt,ti) index1=round(t/ti)+1 chambertemp(it1,index1)=mean(u(chamberx)); times(index1)=t; T(index1,:)=u; plot(x,u-273) title('1-D Transient Temperature, Multiple Layers') xlabel('Distance (m)') ylabel('Temperature (K)') axis([0, L, 50, 100]) text(0,60,['Time: ', num2str(t)]) % display time on graph text(0,65,['Fluid temp: ', num2str(chambertemp(index1)273)]) % display time on graph M(round(t/ti))=getframe(gcf); disp(['Current time: ',num2str(t), '. End time: ', 159 num2str(tmax)]) %pause(.01) if abs(mean(u(chamberx))-Tchamber)<.1 break % break when the fluid reaches the desired temperature end end %modtest(m)=mod(t,ti); end toc subplot(2,2,2) surf(x,times,T-273) subplot(2,2,3) plot(times,chambertemp-273) subplot(2,2,4) plot(x,u) end %% filename=['chambertemp_',num2str(fluidtemp),'_',num2str(Th),'_',n um2str(tmax)] movie(gcf,M,1,10); %% movie2avi(M,'60_97_3sec.avi','COMPRESSION','None') %% filename1=['chambertemp_',num2str(fluidtemp),'_',num2str(Th),'.da t'] save(filename,'chambertemp', '-ascii') load filename %%