Lise BILHAUT 2004 class
Third year internship report Engineering Degree of l’École Nationale Supérieure de Physique de Strasbourg
Pumped Elastomeric Microfluidic Systems for High Resolution Printing March-August 2004
Training supervisor John Rogers Material Science & Engineering Department 1304 W. Green Street Urbana, IL 61801 USA
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
Content Acknowledgements .................................................................................................................... 4 Abstract ...................................................................................................................................... 5 Résumé ....................................................................................................................................... 5
1
Project environment and motivation .................................................................................. 6 1.1 The Nano-Chemical-Electrical-Mechanical Manufacturing Systems Center ............ 6 1.1.1 Motivation of Nano-CEMMS Center................................................................. 6 1.1.2 Goal of Nano-CEMMS Center........................................................................... 6 1.2 Internship within the scope of Nano-CEMMS Center ............................................... 8 1.2.1 Building a new type of printhead ....................................................................... 8 1.2.2 Building proof-of-principle devices ................................................................... 9 1.3 Internship management .............................................................................................. 9 1.3.1 Human resources ................................................................................................ 9 1.3.2 Internship schedule........................................................................................... 10
2
Printhead prototype .......................................................................................................... 12 2.1 Introduction .............................................................................................................. 12 2.1.1 Inkjet printer technology .................................................................................. 12 2.1.2 Advantages of polydimethylsiloxane regarding microfluidic devices ............. 13 2.1.3 Pressure pump .................................................................................................. 15 2.1.4 Heater pump ..................................................................................................... 17 2.2 Heater pump ............................................................................................................. 17 2.2.1 General overview of the system ....................................................................... 17 2.2.2 Modeling each layers in PDMS ....................................................................... 18 2.2.3 Heater fabrication ............................................................................................. 20 2.2.4 Bonding ............................................................................................................ 22 2.2.5 Nozzle drilling.................................................................................................. 24 2.2.6 Tubing system .................................................................................................. 27 2.3 Different generations of printhead prototypes ......................................................... 29 2.3.1 First generation (G1) ........................................................................................ 29 2.3.2 Second generation (G2).................................................................................... 32 2.3.3 Third generation (G3)....................................................................................... 35 2.3.4 Fourth generation (G4)..................................................................................... 39 2.4 Printhead prototype improvement ............................................................................ 42
3
Organic transistors............................................................................................................ 44 3.1 Gold electrodes organic thin film transistors ........................................................... 44 3.1.1 Poly(3-hexylthiophene)-2,5-diyl: a semiconductor polymer ........................... 44 3.1.2 Fabrication (with nano transfer printing) ......................................................... 44
Third year internship report
2/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
3.1.3 Transistor characteristics.................................................................................. 46 3.2 PEDOT electrodes.................................................................................................... 48 3.2.1 PEDOT/PSS ..................................................................................................... 48 3.2.2 Resistance of PEDOT/PSS............................................................................... 49 3.2.3 Fabrication........................................................................................................ 50 3.2.4 Characteristic.................................................................................................... 51
Conclusion................................................................................................................................ 52
Appendix 1 Stamp processing.................................................................................................. 53 Appendix 2 Valves system recipe ............................................................................................ 54 Appendix 3 Heaters fabrication................................................................................................ 55 Appendix 4 PDMS/PDMS bonding ......................................................................................... 57 Appendix 5 PDMS/Kapton bonding ........................................................................................ 58 Appendix 6 Metallic electrodes OFET fabrication .................................................................. 59 Appendix 7 PEDOT/PSS electrodes OFET fabrication........................................................... 60 Appendix 8 Resistance of PEDOT/PSS ................................................................................... 61 Bibliography – Webography .................................................................................................... 64
Third year internship report
3/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
Acknowledgements I take this opportunity to thank my supervisor John Rogers for his guidance and motivation throughout the internship even in the long periods when I could not get any striking results. Many thanks to everyone under John for their patience help friendship and their fruitful discussions, especially to my co-workers Jang-Ung Park, Shraddha Avasty, Erica Lin, Tomazs Biegala and Etienne Menard. A special thanks to Dhal-Young Khang to remind me that all people in this group are “crazy crazy crazy”. I would also like to thank Tony Banks from the Frederick Seitz Materials Research Laboratory for his help on the microfabrication facility, Roger Smith from the School of Chemical Sciences Machine Shop from his availability, Michael Toepke and Matthew Cole from Paul Kenis group and Matthew Stewart and Seung-Hyun Hur from Ralph Nuzzo group for their help in fields unknown of me. All people from Metagamers, especially Robert Zimmerman, are welcomed any time in France so I could show them that DnD is played the same way around the world. And finally, I would like to thank my grand-mother for her weekly postcard, my parents for their updated news of France and Florian for having come.
Third year internship report
4/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
Abstract This training period took place in the Ralph Nuzzo group in the School of Chemical Sciences at the University of Illinois at Urbana-Champaign. After a short introduction to the project environnement and motivation, especially to the Nano-Chemical-ElectricalMechanical Manufacturing Systems Center which aims to build a tool capable to make nanoscale devices in a production line, the second and third parts of this report describe the accomplished work. The second part explains the core goal of the internship, which is the conception of a printhead prototype in polydimethylsiloxane. The use of this soft elastomer should make it possible for the printhead to print electro-organic devices with nanometric resolution. Here we also explain how we managed to create the first working prototypes: what were the systems imagined to pump the fluid out from the nozzles, how did the prototype devices were realized, the evolution of the design guided by the results we got and suggestions to improve the current prototype. The third part deals with the fabrication of organic transistors, which will be the ultimate use of the aforementiond printhead. Some transistors with gold electrodes were patterned and studied in order to fully understand the physics and fabrication process of organic thin film transistors. Then, transistors with organic source and drain electrodes were roughly built and characterized.
Résumé Ce stage s’est déroulé dans le groupe de recherche de Ralph Nuzzo de l’École des Sciences Chimiques à l’Université de l’Illinois à Urbana-Champaign (USA). Après avoir présenté l’environnement et les motivations du stage, notamment du Nano-ChemicalElectrical-Mechanical Manufacturing Systems Center dont le but est de concevoir un outil de production capable de fabriquer des systèmes à l’échelle nanométrique, nous expliquons en détail le travail accompli. La deuxième partie approfondit le but central du stage, qui était de concevoir un prototype de tête d’impression en polydimethylsiloxane. Ce polymère flexible est utilisé afin de permettre l’impression de composants électro-organiques avec une résolution nanométrique. Nous expliquons comment les premiers prototypes ont vu le jour : quels sont les systèmes imaginés pour éjecter le fluide des buses, les différentes étapes de fabrication, l’évolution du modèle en fonction des résultats obtenus ainsi que quelques suggestions à même d’améliorer le prototype actuel. La troisième partie de ce rapport porte sur la fabrication de transistors organiques qui est le but de la tête d’imprimante mentionnée ci-dessus. Des transistors organiques avec des électrodes en or ont été tout d’abord fabriqués et étudiés afin de se familiariser avec la physique et les techniques de fabrication, puis des transistors rudimentaires aux électrodes de source et de drain en polymère ont été fabriqués et caractérisés.
Third year internship report
5/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
1 Project environment and motivation 1.1 The Nano-Chemical-Electrical-Mechanical Manufacturing Systems Center 1.1.1 Motivation of Nano-CEMMS Center A stunning breadth of advancement has occurred in nanoscience that might radically change the way humans live in the world. However, the manucfacturing technology used to create systems based on all-nanoscale component does not yet exist. The nano-Chemical-Electrical-Mechanical Manufacturing Systems Center (NanoCEMMS Center) regroups three institutions (California Institute of Technology, North Carolina Agricultural & Technical State University and the University of Illinois at UrbanaChampaign) in order to address a central problem in the development of nanotechnology: how to assemble nano-scale structures than can be seen and manipulated. Making threedimensional, nanoscale devices and systems from millions to trillions of different types of molecules is incredibly difficult. The Center’s goal is to develop a reliable, robust and costeffective nanomanufacturing system to make nanostructures from multiple materials ([1]). This technology will allow advancements and discoveries in nanoscience to move from the laboratory to production.
1.1.2 Goal of Nano-CEMMS Center The envisioned manufacturing system is schematized on Figure 1, from normal scale tooling, to microsteppers, to nanopositioners, electronics sensing feedback arrays for computer control, all driving the micronanofluidic network toolbit for chemical processing at the attoliter scale. This system should be able, in reasonable time and cost scales, to build 3-D structures from the nano- to micro- to macro- length scales, simultaneously handle liquids and solids, and utilize multiple types of raw materials. Several physical systems are necessary as we travel from the macro-scale to the nanoscale in the realization of the nano-CEMMS technologically critical components. It is instructive to have, at least in a schematic, what physical system we envision as the realization of the nano-CEMMS technnologically critical components as we travel from the macro-scale to the nano-scale in the system. The elements operating at these different length scales include:
Third year internship report
6/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
•
Macro-scale Stepper and Nanopositioner: These steppers are required to position the nano-CEMMS tool relative to a worktable holding the workpiece.
•
ToolHolder and Interface: This is the interface that provides a physical interface for mounting the Nano-CEMMS toolbit on to the stepper. It provides for fluids entering or leaving the toolbit.
•
Nano-CEMMS Toolbit: The nano-CEMMS toolbit is at the heart of the research to be conducted in the center. We envision this ‘toolbit’ to contain the following subcomponents arranged in layers. o Macro-Micro Interconnect: An interface for locating the toolbit, providing pressurized fluids, & electrical connections. o Microfluidic Array: A 3-D array of micro channels interconnected by valves to transport and mix reactants. o Micro-Nano Interconnect: A 3-D array of micro-channels interconnected by nano-gates for metering and functionalizing materials. o Tool-Workpiece Interface: A compliant nano-tube array for transcription, transduction and nano-position adaptation
Figure 1
– Schematic of the manufacturing system envisioned (Nano-CEMMS, Abridged Project)
Third year internship report
7/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
1.2 Internship within the scope of Nano-CEMMS Center The Ralph Nuzzo group intervenes in two areas of the Nano-CEMMS toolbit: microfluidics array and micro-nano interconnect. The first objective of the group is to build a micron-scale fluidic system which is capable of reliability and accurately patterning water soluble electroactive organics (1.2.1). The second step will be to use this device to pattern functional devices such as Light Emitting Diode or transistor (1.2.1). Both these parts must maintain the final objective, which is to scale down the system doing the micro-nano interconnection and widen the use of the system to other solvents than water.
1.2.1 Building a new type of printhead In addition to the conventional use for document printing on paper, inkjet printing is currently being explored for applications such as printing electronic devices like organic light emitting devices and transistors ([2] and [3]). The resolution of fluid dispensing from current printers significantly limits the use of inkjets for biochemical and electronic device manufacturing processes. A new inkjet technology is thus required in order to improve resolution and cost. The Ralph Nuzzo group proposes the development of a flexible inkjet printhead made in a soft polymer called polydimethylsiloxane (PDMS) for improved resolution and reduced cost of printing miniaturized devices. PDMS is now widely use in many research groups to build 3-D microfluidic devices ([4] [5] and [6]), thus a lot of useful processing methods are available. The use of a flexible printhead would allow it to get closer to the printing surface without damaging the surface. PDMS has the ability to enter into conformal contact with most of substrate, created atomic bonding. With this method, we can get the size of printed droplets quite close to the nozzle diameter thus bringing the printhead close to the printing surface and reducing deflection of ink droplets before they strike the print surface (the distance between current printhead and substrate is about 100 µm, which causes the drop to spatter and creates satellite droplets). Furthermore, PDMS has a range of interesting properties like biocompatibility, which would made the use of this printhead possible in a wide range of application. The development of devices using such a printhead would be less costly since, besides the low cost of the raw material (~$0.05/cm3), a PDMS printhead would not require the especially clean environment demanded by currently silicon-based processing. As the fabrication of the printhead has been the main part of the internship, the second part of this report is devoted to describing initial progress.
Third year internship report
8/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
1.2.2 Building proof-of-principle devices The second objective is to use the PDMS printhead as a first generation nano-CEMMS toolbit to build proof-of-principle functional devices. In particular, a conducting polymer, poly(3,4-thylenedioxythiophene), doped with polystyrene sulfonic acid (PEDOT/PSS) will be used to print electrodes of two classes of functional organic devices : 1. An array of organic thin film transistors using printed source and drain electrodes of PEDOT/PSS with Poly(3-hexylthiophene)-2,5-diyl semiconductor polymer; 2. A passive matrix organic light emitting diode array that uses printed PEDOT/PSS with an uniform layer of the electroluminescent polymer poly[2-methylxy-5-(2’ethylhexyloxy)-1,4-phenylenevinylene]. Some work has been conducted on second objective in order to clear possible issues (cf. 3).
1.3 Internship management 1.3.1 Human resources For the first month, I worked in close collaboration with Jang-Ung Park (visiting researcher) in order to find a way of doing nozzles in PDMS. Later, once we envisioned two possible kinds of ejecting system for the printhead (cf.2.1), we decided that I lead on the heating pump (cf.2.1.4 and 2.2) and that Shraddha Avasty (graduate student) leads on the pressure pump (cf. 2.1.3). Shraddha also worked on the modeling of fluid in PDMS channel in order to explain and get useful experimental parameters such as pressure which has to be applied to push a liquid in a small PDMS channel and the deformation which occurs in the material. Two undergraduate students, Erica Lin and Tomazs Biegala spent the summer in Nuzzo’s group: Erica focused mainly on alternative materials to pattern arrays of nozzles and Thomas worked on characterization the sagging of PDMS. For the outreach portion of the Nano-CEMMS Center, we organized a workshop for high-school teachers in collaboration with Paul Kenis’ group. We designed an experimental protocol using PDMS as a diffraction grating, which is easily reproduced by high-school students. This protocol was intended to show molding properties of PDMS and the action of environmental parameters such as heat or solvent on it.
Third year internship report
9/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
1.3.2 Internship schedule The Gantt chart of the internship is presented Figure 2. Besides the bibliographical search done throughout the project and the weekly reports and group meetings, my time has been dedicated to: Training on facilities: the first month was devoted to be training in clean-room processing (photolithography) and machine use (Electron-beam evaporator, Reactive Ion Etching system, Scanning Electron Microscope); Nozzles design: about 6 weeks were spent trying to design nozzles by building 3D PDMS structures; G1, G2, G3 and G4: correspond to four successive generations of printhead prototypes (cf. 2.3) Au electrodes transistors: building these transistors for use in the process and to have references (cf. 3.1) All-organic transistors: PEDOT/PSS electrode transistors (cf. 3.2) May
April
March
July
June
August
Training on facilities Nozzles design
1 G1
G2
Bonding issues
G3
G4
3
2 Au electrodes 4 transistor All-organic electrodes transistor
Bibliography Weekly reports and group meetings
PDMS Printhead fabrication Striking point Devices fabrication Figure 2 – Internship Gantt chart
Third year internship report
10/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
A few striking points marked out the project: 1. The micromachining facility became available, leading to the best solution to the nozzle design problem (cf. 2.2.5); 2. Two methods were proven successful in bonding the different layers of the printhead prototype (cf. 2.2.4); 3. A drop of water was ejected using the thermal pump (cf. 2.3.3); 4. Meaningful characteristics were obtained for gold electrodes transistors (cf. 3.1.3).
Third year internship report
11/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
2 Printhead prototype 2.1 Introduction We will present an introduction consisting of inkjet printer technologies currently used in industry (2.1.1). Because the use of polydimethylsiloxane as the principal printhead material is the main innovation in the project, we will look at its specific properties and chemistry (2.1.2). We will then explain the two chosen approaches to build a system enabling a precise control of the ejected ink quantity (2.1.3 and 2.1.4).
2.1.1 Inkjet printer technology Although inkjet printers only appeared on the consumer market in the late 1980s, they had been under development for more than twenty years before their release. In the mid1970s, printer companies realized the potential of the inkjet technology, which would make dot matrix printers obsolete. The challenge, however, was to develop a means to create an affordable inkjet printer that would reliably create high-quality printouts. The technical challenges were to control accurately the flow of ink from the printhead onto the page and to prevent the printhead from becoming clogged with dried ink. Furthermore, some research had to be conducted on the relationship between the ink, the printhead and the paper, with all these factors intervening in the quality of the printed page. We will see that in the case of a PDMS printhead, each of these issues needs to be studied and solved again. Essentially two different styles of printing appeared: •
Continuous inkjet printers (developed by IBM) use electrically-charged droplets to coat the page with ink very quickly but also waste a lot of ink. The cost of the very complex ink system used in recirculation of unused fluid and the limited resolution (the drops are nearly double the size of the orifice) made the continuous inkjet printer technology limited its consumer viability. It is mainly used today in industrial settings, for labeling cartons and addressing direct mail.
•
Drop-on-demand inkjet printers (invented by Siemens in 1977) spray ink only where needed and are slower than continuous inkjet printers, but less expensive. This kind of printer, producing droplets nearly the same size as the orifice, assure a good resolution. The inkjet printer has come a long way since it became available almost twenty years ago: Hewlett Packard's DeskJet printer, which was among the first available to the public, was priced at $1,000 in 1988! Drop-on-demand printers use two different technologies to push the drops of ink out of the print head :
Third year internship report
12/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
Piezoelectric technology (patented by Epson) uses piezo crystals located at the back of the ink reservoir of each nozzle. The crystal receives a tiny electric charge that causes it to vibrate. When the crystal vibrates inward, it forces a tiny amount of ink out of the nozzle. When it vibrates out, it pulls some more ink into the reservoir to replace the ink sprayed out (more information about piezoelectric technology can be found in [7]). Although the piezoelectric system offers a better resolution, the hard materials used to build the channel which ejects the ink prevent us from using it in our system. Thus, we chose the second technology as our initial approach. Thermal technology (used by manufacturers such by Hewlett Packard, Canon, or Lexmark): this method, commonly referred to as bubble jet, is described in the following section 2.1.4. Currently, inkjet can reproducibly dispense spheres of fluid with diameters of 25 µm (10 pL) at rates of up to 4 kHz for single droplets in the case of drop-ondemand and up to 1 MHz for continuous droplets. •
Nowadays, a lot of research is done on printing systems, and other technologies have appeared, mainly using pressure technology to push the ink out from the nozzle, for example using thermo-deformable material ([8]) or MEMS element ([9]). These new ejection systems and innovations made in the microfluidics domain inspired the second ejection system designed (2.1.3).
2.1.2 Advantages of polydimethylsiloxane regarding microfluidic devices Although silicon and glass are attractive materials for fabricating microfluidic devices, polymers are less expensive and involve simpler and less expensive manufacturing processes ([4]). In particular, polydimethylsiloxane (PDMS)1 attracts more and more attention in microfluidics research because of its attractive physical, chemical and processing properties such as: 1. Elasticity (the Young modulus ranges from 8.7.105 Pa to 3.6.105 Pa, depending on the PDMS composition); 2. Optical transparency, at least in the visible spectrum, it is therefore easy to monitor the fluid flow; 3. Flexible surface chemistry: the repeated units of -O-Si(CH3)2-groups of PDMS lead to a hydrophobic surface. This surface can be made hydrophilic by exposing it to an oxygen plasma (exposure to plasma induces silanol (Si-OH) groups and destroys methyl groups (Si-CH3)). Furthermore, when PDMS makes conformal contact with smooth plastic or glass substrates, both reversible and irreversible sealing are possible. We took advantage of this flexible surface chemistry to stick PDMS to itself or to polyimide (2.2.4); 1
The PDMS used was Sylgard 184 silicone elastomer from Dow Corning.
Third year internship report
13/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
4. Low permeability to water; 5. Low electrical conductivity; 6. Easy to process using micromolding techniques: as it is liquid before it is cured, it can mold any relief forming what is called a stamp (cf. Figure 3, the full outline of creating a stamp with a silicon master is given in Appendix 1); 7. Low cost compared with Si ($50/1 lbs for PDMS); this low cost combined with the cheap and economical processing (for example, once a master is made, it can be used several time to do hundreds of stamp) allow lower production costs.
B
A
C
Figure 3 – Stamp processing : A - A mold in photoresist is created in a substrate; B - Liquid PDMS is pour over the mold and cured; C - The stamp, replicating inverted resist relief, is peeled away from the mold
PDMS chemistry PDMS is cured by an organometallic crosslinking reaction: the siloxane base oligomers contain vinyl groups (Figure 4, A), and the cross-linking oligomers contain at least 3 silicon hydride bonds each (Figure 4, B). The curing agent contains a proprietary platinumbased catalyst that catalyzes the addition of the SiH bond across the vinyl groups, forming SiCH2-CH2-Si linkages. The multiple reaction sites on both the base and crosslinking oligomers allow for three-dimensional crosslinking (Figure 4, C). One advantage of this type of addition reaction is that no waste products such as water are generated. If the ratio of curing agent to base is increased, a harder, more cross-linked elastomer results. Heating will also accelerate the crosslinking reaction (for a piece of 5mm depth, it takes about one hour in an oven at 70°C).
Third year internship report
14/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
Figure 4 – PDMS chemical base (A) and curing agent (B). The 3-D crosslinking reaction (C)
2.1.3 Pressure pump The pressure pump is inspired by an article by Quake and Co. [10]. The principle is explained in Figure 5. A two layered structure is build: in the first layer, the flow layer, a channel is made to allow fluid circulation. The second layer, the control layer, also has a channel, which is positioned perpendicular to the flow channel. This channel will function as a valve: when some nitrogen pressure is injected inside, the channel will expand, thus creating pressure, which will close the flow channel. In Figure 5 A, no pressure is applied to the valve, which is off (cf. also Figure 6 A). When we inject some nitrogen pressure (30 psi), the valve turns on and closes the channel (Figure 5 B and Figure 6 B).
Third year internship report
15/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
A
B
Control layer
Air
Air pressure Flow layer
Ink Pressure
Nozzle layer
Figure 5 – Pressure pump principle: A - System in rest; B - Some air is injected in the control layer, deforming the channel and applying some pressure in the flow layer A
100 µm
Valve
B
100 µm
Channel Figure 6 – A: Valve is inactive (OFF). B: valve is active (ON)
Simulations have shown that the shape of the flow channel is important for proper actuation of the valve. Rectangular shaped channels will not close completely under pressure from above. Flow channels with a round cross section close completely, the round shape transfers force from above to the channel edges and causes the channel to close from the perimeter inwards. To obtain round shape using the classical photolithographic techniques, we hard-bake the master after the development so the photoresist can reflow (Figure 9 and Figure 10 in 2.2.2). As the shape of channel is noy perfectly round, the valve cannot close completely and it may be one of the reasons we could not use them as originally intended (cf. 2.3.3).
Third year internship report
16/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
2.1.4 Heater pump The heater pump is based on the heater system used in common inkjet printer: in a thermal inkjet printer, tiny resistors create heat, and this heat vaporizes ink to create a bubble. As the bubble expands, some of the ink is pushed out of a nozzle onto the paper. When the bubble collapses, a vacuum is created. This pulls more ink into the print head from the cartridge (cf. Figure 7). Some useful information was found in the patent US 6,332,677 ([11]) such as the pulse width of 2.3 µm used to eject the fluid out from the nozzle. In the patented design, as in most industrial designs, the heater elements are localized above the nozzle. However, these designs use hard materials like silicon to build the heater elements and the channel structure. Using a soft polymer like PDMS was the challenge of the project. In the imagined design, we decided to put the heater on the edge of the nozzle (cf. 2.3 for the design specification). As my work was mostly on the heater pump system, all the work and experiments conducted are detailed in the following two sections (2.2 and 2.3). Ink supply
Heater
Pressure !
Vapor
Nozzle
Figure 7 – Thermal pump principle
2.2 Heater pump 2.2.1 General overview of the system In the course of the internship, the printhead prototype design underwent a lot of different changes that we will explain in the section 2.3. However, the usual fabrication steps, which are described more accurately on the parts 2.2.2, 2.2.3 and 2.2.4, are summarized in the Figure 8: the ink flows in the flow layer (A) in channels and reservoirs. In some designs, a control layer (B) is added, allowing control over whether the ink crosses the channel. After bonding these two layers (or just using the flow layer), a heater layer (D) is added to finalized Third year internship report
17/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
the printhead prototype (E). The nozzles were drilled using laser micromachining (cf. 2.2.5) before or after the bonding of the heater layer, depending on the alignment requirement.
A. Control layer Valve C. PDMS upper stamp
reservoir B. Flow layer Channel
Heater Nozzle
E. Printhead prototype D. Heater layer
Figure 8 – General overview of the printhead prototype
The ink used was a conducting polymer, poly(3,4-thylenedioxythiophene), doped with polystyrene sulfonic acid. This polymer, from Aldrich Corporation, is diluted in water (cf. 3.2.1). The nozzle layer is made from Kapton (cf. 2.2.3) and also from a PDMS layer (about 50 µm thick). This layer can be bound to Kapton using spin-on-glass (cf. 2.2.4) or just oxydation of Kapton since it will not undergo the same pressure than in the flow layer. However, the addition of this last PDMS layer can be done at the end of the process fabrication which is why the printhead prototypes do not have this layer, since the focus is for now to the reliability of the ejection of the drops.
2.2.2 Modeling each layers in PDMS PDMS channels are fabricated by soft lithography using PDMS molding, a technique that allows rapid prototyping of microfluidic devices. The masters created to mold the channel can be used several times, hence reducing the cost and the time of the printhead fabrication. First, the design is created using computer-aided design software (Adobe Illustrator CS). This design is then printed on a 5080dpi transparency which serves as a photomask in the photolithographic process. This process was carried out in the class-100 clean-room of the
Third year internship report
18/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
Frederick Seitz Materials Research Laboratory. The following are the main steps to build the master (cf. Figure 9) (the detailed outline procedure is in Appendix 2): Step 1: SRJ 5740 is spin-coated on a silicon wafer at 2,400 rpm (round per minute) for 60 s (thickness of 15 µm) then it is prebaked at 105°C for 90 s on a hotplate; Step 2: the master is exposed to an ultraviolet wavelength (λ = 365 nm and power of 16.5 mW.cm-²) under the chosen with an aligner for 26 s; Step 3: the master is developed in a solution of 2041 developer diluted in deoxidized water and rinsed with deoxidized water; Step 4: for the flow layer, the master is hard-baked for 30 min at 100°C in order to reflow the resist and make a round shape (Figure 10X shows Scanning Electron Microscopic (SEM)2 images of the cross-section of a channel before hard-baking (A) and after (B)). Step 5: a self assembled monolayer of (Tridecafluoro-1,1,2,2 tetrahydrooctyl)-1trichloro silane is formed on the master to prevent the polymer from sticking to the silicon; Step 6: for the flow layer, PDMS (8:1)3 is spin-coated at 2500 rpm for 30 s (thickness of 50 µm) and cured at 100°C for two hours. For the control layer, about 5 mm of PDMS is poured over the top and cured in an oven at 70°C for one hour; Step 7: for the control layer, the stamp is released from the master by cutting and carefully peeling it away from the master surface. The flow layer remains on the silicon master until it is stuck to another slab of PDMS (cf. 2.2.4).
Step 1 – Coating and prebaking SRJ 5740
Step 2 – Exposure
2
All the SEM images were taken with a Dual-Beam Focused Ion Beam Microscope (FEI Strata DB-235) from the Center for Microanalysis of Materials. 3
As thin PDMS layers are difficult to fully cure, using more curing agent helps to fasten the cross-linking reaction.
Third year internship report
19/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
Step 3
March-August 2004
Step 4 – Hard-baking
– Development
Steps 5-6 – Silanizing master and pouring PDMS over; curing
Step 8 – Releasing the stamp
Figure 9 – Master and channel prototyping
A
B
Figure 10 – SEM pictures of a SRJ 5740 channel before (A) and after (B) hard-baking
2.2.3 Heater fabrication As the heater could not have been made on PDMS because of the softness of the material (PDMS can expand under heating, it could generate some cracks in the metal layer leading to failure of the device), Kapton E, a polyimide film from Dupont Company that presents also other advantages, was chosen as a heater support.
Third year internship report
20/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
This polyimide film is synthesized by polymerizing an aromatic dianhydride and an aromatic diamise. The physical, electrical and mechanical properties of Kapton are maintained over a wide range of temperatures. Its chemical resistance (it has no know organic solvents) and the wide range of temperatures it can stand (over 354°C), along with the fact that its thermal expansion coefficient is similar to copper’s (dimensional stability at 150°C: 0.05 %), made Kapton the primary choice for our application. The particular property of Katpon E type is its high modulus (tensile strength of 50 Kpsi, Young’s modulus 780 Kpsi and elongation of 50%) which helps to rigidify the printhead prototype without taking away the softness: sticking Kapton to PDMS prevent any deformation and sagging of the PDMS that would occur and lower the resolution of the printhead (cf. Figure 11) ([12]). Kapton sheet B
C
A
PDMS stamp Substrate Figure 11 – Sagging of PDMS A – Original stamp shape; B – Sagging of PDMS : the recessed area enters in conformal contact with the substrate; C – Composite stamp: the Kapton sheet backslide rigidifies the stamp, diminishing sagging
Moreover, the Kapton E film has also very low moisture absorption (1.8% at 100%RH, water vapor permeability of 5 gms/m2/day) and is laser ablatable. All the characteristics of Kapton E film are summarized in [13]. The lift-off process was used to build the heaters. The mains steps are summarized below and in Figure 12 (the full outline is described in Appendix 3): Step 1: the desired design is patterned with a S1805 photoresist on a 25 µm thick Kapton E film (cf. Appendix 3 for the processing of S1805 PR); Step 2: 3 nm of titanium and 100 nm of gold are evaporated with an e-beam evaporator on the Kapton sheet (titanium is used to improve the adhesion of gold on Kapton); Step 3: S1805 is removed in an acetone ultrasonicator bath; the metal deposited in the resist is removed in the same time whereas the metal which was deposited on the developed area (ie onto Kapton) remains on the sheet.
Third year internship report
21/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
The heaters design was improved by changing the thickness of the heater doing a twostep lift-off in order to localize the resistance in a desired small area (cf. Appendix 3 for the full process). In that case, the main wires are 400 nm thick and the heating part is 100nm thick. Ti + Au
S1805
Substrate
(Si)
Step 1 – Patterning the design
Step 2 – Evaporation of Ti/Au
Step 3 – Lift-off in acetone bath Figure 12 – Lift-off processing
2.2.4 Bonding The bonding between the different layers was a huge issue that had to be faced. As the printhead prototypes are eventually intended to work for a long time frame and under a wide range of experiments, the reliability of the bonding is essential. We tried several methods listed below before selecting two of them, which work with a satisfactory reliability. •
Bonding PDMS to PDMS : o Partial curing : the PDMS is preparing by mixing one part (the base, A) containing polydimethylsiloxane bearing vinyl groups and a platinium catalyst and another part (the curing agent, B) containing silicon hybrid groups, which form a covalent bond with vinyl group. Whereas PDMS is usually used at a ratio of 10:1 (A:B), for bonding, one layer is made with 20:1 and the other with 5:1. As each layer has an excess of one of the two components, reactive molecules remain at the interface between the layers. Further curing causes the two layers to irreversibly bond and the strength of the interface equals the strength of the bulk. ([10]). Although this method is reliable, the UV/Ozone method selected later is simpler. o Oxidation: this method uses oxygen plasma, to produce silanol groups on PDMS and -OH-containing functional groups on the other materials; these polar groups form covalent -O-Si-O-bonds with oxidized PDMS when these surfaces are brought into contact ([5]) (method reliability of 50 %).
Third year internship report
22/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
o UV/Ozone: the final selection, this method uses the same chemical reaction as the one above, but is more reliable ([14]). The UV/Ozone process can be explained with the following steps (cf. Figure 13 and full outline in Appendix 4): Step 1: the flow layer is oxidized with an oxygen plasma; Step 2: the thick layer is put under a low pressure mercury lamp (173 µW.cm2 ) for 3 min; Step 3: both pieces are brought into contact and put in an oven at 150°C for 40 min, then removed from the substrate.
UV light
Step1 – Oxidation of flow layer
Step 2 – Treatment of control layer with UV light
Step 3 – Both pieces are brought in contact, put in an oven, then removed from the substrate Figure 13 – UV/Ozone processing
•
Bonding PDMS to Kapton : o Cold-welding: the principle is to evaporate a very thin layer (~10 nm) of gold on both surfaces, and then to bring these surfaces in contact immediately after the evaporation when the gold is still fresh and uncontaminated: in this state, interactomic forces will take over to produce the weld. Although reliable, this method is more time and resource consuming than the other one ([15]). o Spin-on-Glass (SoG): SoG is a solution of glass forming compound. When applied to a surface, it dries to yield a pure film of SiO2. In this method, however, the SoG is not fully cured, producing, when oxidized, silanol groups (same bonding principle as oxidation).
The process of using SoG to bond Kapton to PDMS is summarized in the following steps (cf. Figure 14 and full outline in Appendix 5):
Third year internship report
23/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
Step 1: a very thin layer of titanium (about 10 nm) is evaporated onto Kapton to promote adhesion of SoG; Step 2: SoG is spin-coated at 3000 rpm for 30 (thickness about 220 nm) and baked at 140°C for 2 hours in a hotplate; Step 3 : both the coated Kapton sheet and the PDMS stamp are oxidized then brought in contact; Step 4: the sample is heat on a hotplate at 85°C for 10 min to promote adhesion. The thickness of gold electrodes, titanium layer and SoG layer is small enough (about 630 nm all together) to allow the PDMS to hug the substrate surface.
Step 1 – Evaporation of Ti
Step 2 – Spin-coating and baking of SoG
Step 3 – Oxidation of both pieces
Step 4 – Contacting and heating
Figure 14 – Bonding Kapton to PDMS using Spin-on-Glass
2.2.5 Nozzle drilling Laser micromaching was the simplest and quickest way found to drill nozzles though Kapton film. It was done in the School of Chemical Sciences Machine Shop with an LMT 5000 manufactured by Potomac Photonics. The laser used to ablate polymers is a one watt 193 nm argon/fluorine laser. For more information about laser micromachining, the reader can refer to [16][17] [18] or [19]. Usually, nozzles were 30 µm square holes drilled on the Kapton film. However, besides film damages caused by heat diffusion (cf. Figure 15), leftover is generated. In Figure 16, one can see that the nozzle drilled after the different layers were assembled is less clean that the one drilled before. In the first case, it may be because the laser beam also heat and damage the above PDMS. But we are sure that is does not go through the
Third year internship report
24/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
thin PDMS layer between the flow layer and the control layer, because we never see an occurrence of fluid going in the valve. Figure 17 and Figure 18 show that the holes were made all through Kapton layer but that the vertical edges are slightly tilted and that the melted polymer forms drapery around the edge. One limitation of the laser micromachining is its resolution of about 5 µm. When used this small, the shape of the nozzle is hard to control resulting in a round hole, rather than a square (cf. Figure 19).
Figure 15 – 30 µm square hole : laser damage done to the surrounding area A
B
Figure 16 – In A, the nozzle was drilled after the assembly of the different layers. In B, it was drilled before the assembly
Third year internship report
25/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
Top view 45° side view Figure 17 – 30 µm square hole, top view
Bottom view
45° bottom side view
Figure 18 – 30 µm square hole, bottom view
Third year internship report
26/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
Figure 19 – 5 µm square hole (top view)
2.2.6 Tubing system Holes were poked out with a needle sharpened with sand paper. An iron wire was used to remove the leftover PDMS (cf. Figure 20 A). Figure 21 and Figure 22 show the edge of the hole. The connection between this inlet and the syringe pump containing ink is made with a metal tube embedded in the inlet and a plastic tube (cf. Figure 20 B). A
Plastic tube
B
Metallic tube
Figure 20 – ink intlet :
1 cm
Third year internship report
A - Poking holes in PDMS with a needle B – A metallic tube is inserted in PDMS
27/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
Figure 21 – SEM image of inlet (top view)
Figure 22 – SEM image of inlet (12° side view)
The experimental set-up is shown in Figure 23: 1. Printhead prototype is placed under the microscope (an imaging system allow to save pictures and movies); 2. Syringe and syringe pump are used to infuse ink at controlled infuse rate (0.001 mL.min-1 to 0.005 mL.min-1, a higher infuse rate would lead to a failure in Kapton and PDMS);
Third year internship report
28/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
3. The electrical generator generates a 10 ms pulse of 5 volts, monitored by an oscilloscope. 1
Microscope and imaging system
Generator
Syringe pump
Printhead prototype
Figure 23 – Experimental set-up
2.3 Different generations of printhead prototypes 2.3.1 First generation (G1) G1 design The first generation device had one reservoir where the inlet hole was punched, then a 50 µm wide channel that carries ink to another reservoir. In this second reservoir are three heaters. The heaters (2.5 nm of Ti and 100 nm of Au thick) are made of two electrodes and wire, composed of two parts : one 250 µm width and another 50 µm width. This thinner part encloses the nozzle (cf. Figure 24 and Figure 25) drilled after bonding the upper stamp to the heaters layer. G1 results 1. Some vapor can indeed be generated by applying a voltage of 5 V in the electrodes (Figure 26);
Third year internship report
29/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
2. This vapor creates a pressure and the droplet is ejected from of the nozzles located away from the actuated heater (Figure 27); 3. The vapor does not remain in the reservoir, but goes back through the channel (back-flow); 4. Ink accumulates in the reservoir without any possible escape (Figure 28). As the solvent (here water) can diffuse though the PDMS, this will eventually lead to an increasing concentration of solid ink component; 5. Under heating, PEDOT/PSS dries (Figure 29).
Ink inlet
Channel 50 µm
3 mm 250 µm Reservoir
Gold electrodes Figure 24 – First generation design
Third year internship report
30/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
250 µm
50 µm
30 µm Nozzle
Figure 25 – Microscopic images of G1 design
Figure 26 – Applying a voltage of 5 V generate some vapor around the heater
Figure 27 – The pressure generated by the vapor expulses a PEDOT/PSS droplet out of a nozzle
Figure 28 – Ink is filling the reservoir : the difference in color shows that ink is accumulating in the reservoir
Third year internship report
31/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
Figure 29 – Dried PEDOT/PSS
2.3.2 Second generation (G2) G2 design Design is changed to solve issues arisen with G1 (cf. Figure 30 and Figure 31): - There is only one channel per nozzle to avoid ejecting ink from a nozzle while heating another nozzle; - Each channel has a triangular shape to avoid the back-flow (60 µm out the inlet reservoir to 230µm in nozzle reservoir). Because pressure is larger in a smaller area, the pressure gradient should prevent the vapor from going back to the channel. In this design, as the dimensions are a few hundred micrometers, manual alignment is possible and the nozzle could be drilled before bonding the upper stamp and Kapton. G2 results 1. Experimental data shows that the pressure gradient is not high enough to prevent the back-flow (cf. Figure 32); 2. The nozzle locations are changed to see if the expelled vapor could push PEDOT/PSS through the new holes (cf. Figure 33) but this proved unsuccessful.
Third year internship report
32/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
m 1.5 mm
Figure 30 – Second generation design
250 µm
50 µm 30 µm
Figure 31 – Microscopic images of G2 design
Third year internship report
33/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
Leakage
Vapor is going back
Figure 32
– This picture also shows an example of leakage between the PDMS stamp and the Kapton film
Figure 33 – Nozzles are drilled in different locations in the channels
Third year internship report
34/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
2.3.3 Third generation (G3) G3 design In order to block the back-flow, a valve is incorporated into the design as a second PDMS layer (the control layer, cf. 2.1.3) (Figure 34 and Figure 35); A 50 µm wide channel conducts ink from the inlet to a triangular reservoir, smaller than the previous system channel and reservoir. G3 results 1. The valve does not work as intended (cf. Figure 37); the vapor pressure is higher than the pressure exerted by the valve on the channel (a maximal nitrogen pressure of 30 psi), which causes the PEDOT/PSS to travel back through the channel. 2. Using water and food color4 instead of PEDOT/PSS enables the system to work as intended. (cf. Figure 38): a droplet of colored water is ejected under the pressure created by the vaporization of water (the valve did not need to be implemented – cf. Figure 39). In spite of this first success, there were still some issues to be resolved: 3. The droplet size is hard to control accurately: in some cases (around 7 out of 10), the expulsed droplet continued to expand beyond the desired diameter; 4.
The printhead prototype still did not work with PEDOT/PSS (cf. Figure 40) which dried too fast and blocked the nozzle.
Figure 34 – Third generation design 4
Color food is a solution of water, propylene glycol, artificial colors (FD&C: Yellow #5, Blue #1, Red #40, Red #3) and propylparaben.
Third year internship report
35/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
500 µm
Valve (control layer)
90 µm
Channel (flow layer)
50 µm
Heater
100 µm Figure 35 – Microscopic images of G3 design
Actuated valve Figure 36 – In the left picture, the valve is OFF, in the right one, it is ON (nitrogen pressure makes the PDMS valve expends, blocking the channel)
Third year internship report
36/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
Vapor going through the channel
Valve ON
Figure 37 – The valve is ON, but the vapor pressure is higher than the pressure exerted by the valve on the channel : back-flow occurs
Third year internship report
37/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
B
C
Figure 38 – A – The valve is OFF B – The valve is ON C – Applying a voltage in the heater ejected some colored water out from the nozzle
Figure 39 – Colored water was ejected without the use of the valve
Third year internship report
38/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
Figure 40 – It does not work with PEDOT/PSS
2.3.4 Fourth generation (G4) G4 design Considering the difficulty encountered trying to control the amount of ejected ink and the drying of PEDOT/PSS, another design was needed. A fourth generation prototype was created, that again incorporated valves (thus both pressure and heating pumps can be tested) (cf. Figure 41 and Figure 42): - Ink is not confined anymore in a reservoir, but is allowed to exit from the channel by an outlet; - Three valves are incorporated: two outside valves are intended to trap ink in a confined area and a middle valve acts as the pressure pump itself. When activated, the middle valve should push ink through the nozzle; - Heater design was improved by doing a two-step lift-off thus localizing the heating area; - Nozzle location was chosen in order that the heater pump be situated in the middle of the heater (resistance increases more in this drilled area) and also such that the pressure pump be placed at the intersection of the middle valve and the channel.
Third year internship report
39/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
G4 results 1. As with the third design, the heater pump worked again with colored water but PEDOT/PSS still dried before any ejection (cf. Figure 43); 2. A solution of PEDOT/PSS dissolved in isopropanol (1:4) is tested; the boiling point of isopropanol is 78.5°C, thus this solvent should evaporate before the water, preventing the original solution from drying;
Figure 41 – Fourth generation design
Third year internship report
40/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
50 µm
1.5 mm
March-August 2004
500 µm
Channel
250 µm 50 µm
Heater
Valve
90 µm
Figure 42 – Microscopic images of G4 design
Figure 43 – PEDOT/PSS dried by the heater pump
Third year internship report
41/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
2.4 Printhead prototype improvement Heater efficiency improvement One striking issue was the electromigration ([20]) which happened when the applied voltage was too high. When too high, the momentum transfer between conducting electrons and diffusing gold atoms resulted in some gold atoms being moved through the wire (cf. Figure 44). As the applied voltage was limited, the pulse used had to be around 10 ms to have any visible effect. In that case, the amount of vapor generated was large enough that it remained in the channel instead of disappearing immediately after the pulse. Changing the metal in the heater for example substituting aluminium for gold, would allow the voltage to increase, leading to a shorter pulse width.
Figure 44 – Electromigration
Faster prototyping Another design in two dimensions is possible and perhaps more pratical to study, since 2D-structures would require less time for fabrication processing (no need of laser ablation to drill the nozzles). PEDOT/PSS solution modification Some work should be conducted about the dissolution of PEDOT/PSS in solvent in order to pursue the experiment of diluting PEDOT/PSS in isopropanol. Pressure pump Using ethanol or isopropanol instead of nitrogen would increase the amount of pressure actuating the valve, hence improving its role as a barrier.
Third year internship report
42/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
Scaling o
When the scaling approaches nanometrique scale, Gelest should be used instead of PDMS (the resolution of this polymer is about one nanometer whereas it is about 100 nm with conventional PDMS);
o
To make smaller nozzle, it is possible to use a focused ion beam or, when the resolution of this technique will be reached, to go back to multilayer 3D PDMS structure.
Third year internship report
43/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
3 Organic transistors Building functional devices such as organic thin film transistors was the natural progression from building a PDMS printhead prototype, since this prototype should eventually be used to pattern such transistors. Before building all-organic transistors, thin film organic transistors using metallic electrodes were made in order to get accustomed to working with organic semi-conductor polymers such as poly(3-hexylthiophene) and serve as a wellknow reference for future work.
3.1 Gold electrodes organic thin film transistors 3.1.1 Poly(3-hexylthiophene)-2,5-diyl: a semiconductor polymer Being easily processable and compatible with plastic substrate material, organic materials have been the focus of an explosion in research and industry for the last twenty years. In terms of cost, liquid phase processing polymers such as poly(3-hexylthiophene)-2,5diyl (P3HT, cf. Figure 45) are the most promising. As this polymer processing is beginning to be well-known, it was chosen as semi-conductor channel transistor material. As with most of the organic semiconductor polymers, high transistor performance of P3HT depends strongly on the morphology of thin film ([21] and [22]). We thus had to pay particular attention to the processing conditions (solvent, curing, etc.). C6H13
S
n Figure 45 – Chemical structure of P3HT
3.1.2 Fabrication (with nano transfer printing) The structure of the transistor is shown in Figure 46. Si and SiO2 compose the gate and the gate dielectric respectively. The channel is made of P3HT and the source/drain gold electrodes are deposited using the nanotransfer printing method ([24]).
Third year internship report
44/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
A general overview of the steps to build organic thin film transistors is given below (cf. Figure 47 and Appendix 6 for the detailed guide): Step 1: a solution of P3HT diluted in chlorobenzene (6mg of P3HT for 1 mL of CB) is prepared and filtered with 450 µm filters; Step 2 : the P3HT solution is spin-coated on an oxidized wafer (SiO2 thickness of 300 nm) at 3000 rpm for 60 s (P3HT thickness of 25 nm); Step 3 : the wafer is cured in a vacuum oven at 110°C for 5 min (vacuum = 20 Hg); Step 4 : Au / Ti / Au is evaporated on an appropriate PDMS stamp (10 nm / 5 nm / 25 nm) which is then brought in contact with the coated wafer; Step 5: the stamp, while still in contact with the coated wafer, is put in an oven at 70°C for one hour to increase metal transfer efficiency, then is taken off, leaving patterned electrodes. Throughout the processing, the P3HT solution and coating have to be kept out of the light. Some ways to improve the device performance are indicated in Appendix 6. Drain
Source
Gate
Figure 46 – Organic thin film transistor structure
Stamp P3HT + CB Ti Au
SiO2 Si
Steps 1 to 3 – Spin-coating P3HT on SiO2/Si wafer and curing
Steps 4 and 5 – Evaporate on a stamp Au/Ti/Au, then bring in contact and heat it
Figure 47 – Gold electrodes P3HT transistor fabrication steps
Third year internship report
45/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
3.1.3 Transistor characteristics Transistor characterization has been made with a Hewlett Packard Semiconductor Parameter Analyzer 4155A. The effective mobility of an organic semiconductor in a transistor is typically calculated either in the regime where the source-drain current depends linearly on the sourcedrain voltage (linear regime) or where the source-drain current is independent of source-drain voltage (saturation regime). Figure 48 shows IDS versus VDS for different gate bias.
Figure 48 – IDS vs VDS for different VG for gold source/drain electrodes transistor
o In the linear regim (for example for VDS = -10 V), we have the equation dI DS W = µlin Ci (VG − VT ,lin ) dVDS L where µlin is the effective mobility, VT,lin is the effective threshold voltage, and Ci is the capacitance of the dielectric (which is for wafer with 300 nm of SiO2 1.15102.10-8). In Figure 49 is plotted the slope of IDS versus VG. Here, the slope and the intercept of the linear fit to the experimenta data determine the effective mobility and threshold voltage, respectively.
Third year internship report
46/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
We have W W µlinCi = 8.10−9 and µlinCiVT ,lin = −6.10−7 L L ie with with W = 800 µm and L = 10 µm :
µlin = 8.69.10−3 cm 2 .V −1.s −1 and VT ,lin = −75 V
Figure 49 – Linear regim : IDS vs VG
o In the saturation regime, the gradual channel approximation and the assumption that the field effect mobility is independent of the gate voltage yields the following expression for the saturation current IDS, sat: I SDS , sat =
W µ sat Ci (VG − VT , sat )2 2L
where µsat is the effective mobility, VT,sat is the effective threshold voltage, and Ci is the capacitance of the dielectric. Thus, the squarre root of the saturation current is linearly dependt on the gate bias. Figure 50 shows a plot of |IDS,sat|1/2 at VDS = -80V versus gate voltage. In that case, we have: W µ sat Ci = 9.10−5 2L
Third year internship report
47/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
ie with the same geometry than above: µ sat = 1.76.10−2 cm 2 .V −1.s −1
Figure 50 – Saturation regim : (IDS)1/2 vs VG
3.2 PEDOT electrodes 3.2.1 PEDOT/PSS Poly(3,4-ethylenedioxythiophene)-polystyrenesulfonic acid (PEDOT/PSS) (cf. Figure 51) is among the most successful conjugated polymers to be used commercially since it was synthesized in the late 1980s. Its popularity is due to its many advantageous properties, such as high stability in its p-doped form, high conductivity, good film-forming properties and excellent transparency in the doped state. The reader can refer to [23] and [25] for an overview of possible PEDOT/PSS applications. In our case, we used this polymer dissolved in water as the source and drain electrodes.
Third year internship report
48/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
PSS
PEDOT
Figure 51 – Chemical structure of PEDOT/PSS
3.2.2 Resistance of PEDOT/PSS Experiments were made to determine resistance of four different sheets of PEDOT/PSS deposited on a glass substrate and dried on a hotplate at 110°C. The different samples were made of: -
Pure PEDOT/PSS droplets (thickness around 300 µm); Pure PEDOT/PSS spin-coat at 500 rpm for 30 s (thickness around XX); Diluted PEDOT/PSS droplets (thickness around 100 µm); Diluted PEDOT/PSS spin-coated at 500 rpm for 30 s (thickness around XX);
The dilution process, taken from [27] is in Appendix 8, as well as the original I/V curves. Resistance values are listed in the following table: Pure PEDOT/PSS Spin-coated Droplets (Figure XX) (Figure XX) Resistance (in Ω)
9287.57
372.67
Diluted PEDOT PSS Spin-coated Droplets (Figure XX) (Figure XX) 68.18
19.18
Since it has been shown that the granularity and disorder play a major role in the conduction of conjugated polymers, like for P3HT, film morphology of PEDOT/PSS is an
Third year internship report
49/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
important parameter in the conductivity ([26] [27] and [28]). It is therefore not surprising to note that diluted PEDOT/PSS exhibits a lower resistance than pure PEDOT/PSS: the solvent helps polymer molecules to rearrange in a better way. As the above table shows, another important parameter for conductivity is the thickness of PEDOT/PSS sheet.
3.2.3 Fabrication Rough all-polymers transistors were fabricated using PEDOT/PSS as source/drain electrodes (see the structure of this transistor in Figure 52). The general overview of the fabrication is explained here (cf. Figure 53 and the full procedure in Appendix 7): Step 1 to step 3: the first three steps are the same as in metal electrode transistor fabrication (cf. 3.1.2); Step 4: two drops of the PEDOT/PSS solution are deposited with a syringe to form the source and the drain, and dried on a hotplate at 110° for around 2 min; Step 5: a shadow mask is cut out (for example in Kapton) and taped on the sample so as to place the holes above PEDOT/PSS drops; 50 nm of Au is evaporated with an e-beam. These gold pads will improve the contact between the needle of the probe-station and the PEDOT/PSS electrodes.
Gold contact pad
Figure 52 – Structure of all-organic transistor
P3HT + CB
PEDOT/PSS drop
SiO2 Si
Steps 1 to 3 – Spin-coating P3HT on SiO2/Si wafer and curing
Third year internship report
Step 4 – Drop PEDOT/PSS on a substrate and bake it on a hotplate at 110°C until all the solvent evaporates
50/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
Shadow mask
March-August 2004
Au
Step 5 – Evaporate gold contact pad by shadow masking
Figure 53 – PEDOT/PSS electrodes P3HT transistor fabrication steps
3.2.4 Characteristic Figure 54 shows characteristic of a PEDOT/PSS transistor. The current very low may be due to the very big channel (width about 1 mm) and some problem of contact with the organic. The bad characteristic we obtainted prevent us to do more analysis.
Figure 54 – IDS vs VDS for different VG for PEDOT/PSS source/drain electrodes transistor
Third year internship report
51/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
Conclusion The challenges of the Nano-CEMMS Center are very ambituous but it is very essential to fulfill them if we want one day to be able to exploit all the discoveries in the nanotechnology field and move them out of laboratories to everyday life. Six months were spent imaging, designing, buiding and trying PDMS printhead prototypes. The first success, a water droplet ejection, happened four months after the beginning of the project. During that time, we had to face and overcome challenges like the bonding of the different layers. The next steps will be the ejection of a PEDOT/PSS drop and accurately controlling the amount of fluid ejected out from the nozzle to the printed surface. In the same time, some work has been done about all-organic transistors (ie source/drain electrodes and canal in polymers) in order to get a better understanding of the physics and the processing problems that will be faced using the printhead as a production tool. This internship was very interesting for me because it corresponded to my own background: general engineering, useful for example in the conception of heaters or the characterization of transistor, and solid state physics, whose theoritical knowledge helped me having a better understanding of the chemical processes taking place at the atomic level while I was building the printhead or the transistors at a macroscale level. Finally, the environnement of the project, working in a team, being part of a bigger picture, handing knowledge to people outside my field, gave me an understanding of what a general engineer is able and requested to do.
Third year internship report
52/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
Appendix 1 Stamp processing I. Pre-treatement If the master is made of silicon, a self assembled monolayer of (Tridecafluoro-1,1,2,2 tetrahydrooctyl)-1-trichloro silane has to be formed on the master in order to avoid the polymer to stick on the silicon5.
II. Preparation of PDMS (Product from Dow Corning) 1. Put on a polystyrene weighing dish the willing quantity of Sylgard 184 Silicon elastomer base; 2. Weight it; 3. Put 10% of this weigh of Sylvard 184 Silicon elastomer curing agent; 4. Mix; 5. Put in a vesicator for one hour before to use in order to degas bubbles generated by mixing.
III. Stamp preparation 1. Put the master in a polystyrene Petri dish; 2. Put the PDMS on the master (about 5 mm height); 3. Cure the stamp for 2 hours at 60°C; 4. The stamp is released from the master by cutting and carefully peeling it away from the surface.
5
(Tridecafluoro-1,1,2,2 tetrahydrooctyl)-1-trichloro silane is a long alkanechain molecule that is perfluorinated and has a trichlorosilane group on the end. The chlorines react with the hydrogen of –OH groups on the SiO2 surface, yielding a little HCl and strong Si-O-Si bonds that link the molecules to the surface.
Third year internship report
53/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
Appendix 2 Valves system recipe The photoresist used is SJR 5740 from Shipley Corp.
1. Clean the wafer with isopropanol and acetone, rinse it with deionized water (DI water) and dry it on a hotplate at 110°C for 10 min; 2. Spin-coat SJR 5740 at 3,000 rpm for 60s (slowly increase the speed); 3. Bake on a hotplate 105°C for 90s; 4. Expose on ultraviolet light (λ = 365 nm, power of 16.5 mW.cm-²) for 26 s; 5. Development in 2041 developer diluted in DI water (1:5); 6. Rinse with DI water in a bowl (not under the tap otherwise the resist can lift-off); 7. Post-bake on a hotplate at 100°C the flow layer for 30 min, rotating the water every 10 min (to make tchannels round).
Notes: the adhesion of SJR5740 can be improved by using an adhesion precursor such as HMDS.
Third year internship report
54/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
Appendix 3 Heaters fabrication I. First layer photolithography 1. Cut Kapton and tape it to a glass slide; 2. Spin coat S1805 at 3,000 rpm for 30 s (be careful to recover all the Kapton surface with the resist to have a nice uniform layer);
I.
S1805
3. Bake at 110°C for 3 min and 15 s; 4. Expose to UV light (λ = 365 nm, power of 16.5 mW.cm-²) for 6 s;
Substrate
(Si)
5. Develop in MF-26 A for about 10 s; 6. Rinse with DI water.
II. First evaporation The evaporation was made with the temescal electron beam deposition system in a vacuum of 5.10-6 mTorr.
II.
Ti + Au
1. Evaporate 25 Å of Ti (Ti improves adhesion between Kapton and gold); 2. Evaporate 1000 Å of Au.
III.
III. Lift-off Do the lift-off in acetone using the ultrasonicator bath (the lift-off should last about 8 min) by checking on microscope for the details.
Third year internship report
55/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
IV.
IV. Second layer of photolithography Steps are the same than in I. except that before the exposure, the second mask has to be aligned with the first pattern.
V.VCCcV.
V. Second evaporation Same condition than II. 1. Evaporate first 30 Å of Ti to improve the adhesion of Au on the oxidized gold; 2. Evaporation 3000 Å of Au.
VI.
VI. Second lift-off Same process than III.
Third year internship report
56/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
Appendix 4 PDMS/PDMS bonding 0. Punch holes for the control layer as air inlet; 1. Put on the UV lamp to heat it (at least 20 min prior to use); 2. Wash both PDMS piece with isopropanol and blow up with nitrogen gun; 3. Oxidize the flow layer (the thinner layer) with the Uniaxis 79° series Reactive Ion etching system with the following parameters: Pressure = 100 mTorr O2 flux = 10 sccm Power = 30 W Time = 35 s 4. Put the control layer (or the thicker layer) under a low pressure mercury lamp (173 µW.cm-2) for 3 min; 5. Bring the two surfaces in contact within one minute! 6. Put the sample in the oven at 150°C for 40 min with a piece of metal to press it; 7. Let cool down and cut.
Third year internship report
57/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
Appendix 5 PDMS/Kapton bonding This bonding is used to stick PDMS to Kapton or to Mylar using Spin-on-Glass.
I. Preparation of the substrate 1. Put the substrate in a glass-side attached with tape; 2. If the substrate is Kapton, evaporate 10 Å of Ti with the e-beam evaporator to allow spin-on-glass sticks to the Kapton; 3. Spin-coat spin-on-Glass on Kapton (resp. Mylar) at 3000 rpm for 30 s; 4. Bake at 140°C for 2 h (resp. 100°C for 2:30 h).
II. Bonding 0. Punch some holes in the flow layer as fluid inlet; 1. Wash the PDMS with isopropanol and blow it with the nitrogen gun. Blow also the spin-on-glass; (The cleaness is very important at this step, since any dust between the 2 parts will prevent the bonding.) 2. Oxidize both part with the RIE with the following parameters : Pressure = 30 mTorr O2 flux = 20 sccm Power = 20 V Time = 30 s 3. Immediately after the oxidation, bring the two pieces in contact; 4. Put them on a hotplate at 85°C for 10 min.
Third year internship report
58/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
Appendix 6 Metallic electrodes OFET fabrication 1. Prepare a P3HT (from Aldrich Corp.) solution diluted in chlorobenzène (6mg of P3HT for 1 mL of CB). Stir the solution for one day and filtered it with 450 µm filter; 2. Clean an oxidized wafer (SiO2 thickness of 300 nm) with isopropanol and acetone, rinse it with deionized water (DI water) and dry it on a hotplate at 110°C for 10 min; 3. Spin-coat the P3HT solution on the wafer at 3000 rpm for 60 s; 4. Cure the wafer on a vacuum oven at 110°C for 5 min (vacuum = 20 Hg); 5. Evaporate Au/Ti/Au on a PDMS stamp (10 nm / 5 nm / 25 nm) and bring the stamp in contact with the coated P3HT (evaporation made with a temescal electron beam deposition system in a vacuum of 5.10-6 mTorr); 6. Put in an oven at 70°C for one hour and remove the stamp.
Improving transistor performance: In the step 6, after removing the stamp from the oven but before removing it, oxidize the sample (20 mTorr, 20 sccm O2, 100 W, 60 s) in order to get ride of P3HT.
Third year internship report
59/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
Appendix 7 PEDOT/PSS electrodes OFET fabrication 1. Prepare a P3HT (from Aldrich Corp.) solution diluted in chlorobenzène (6mg of P3HT for 1 mL of CB). Stir the solution for one day and filtered it with 450 µm filter; 2. Clean an oxidized wafer (SiO2 thickness of 300 nm) with isopropanol and acetone, rinse it with deionized water (DI water) and dry it on a hotplate at 110°C for 10 min; 3. Spin-coat the P3HT solution on the wafer at 3000 rpm for 60 s; 4. Cure the wafer on a vacuum oven at 110°C for 5 min (vacuum = 20 Hg); 5. Step 4 : put two drops of PEDOT/PSS solution to form the source and the drain; 6. Step 5 : cut a shadow mask (for example in Kapton) and tape it on the sample so as to place the holes above PEDOT/PSS drops and evaporate 50 nm of Au : these gold pads will improve the contact between the needle of the probe-station and the PEDOT/PSS electrodes.
Improving transistor performance: P3HT can be wiped right after it has been spin-coating, in order to hold it only on the channel area and limited the leakage through the SiO2 layer;
Third year internship report
60/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
Appendix 8 Resistance of PEDOT/PSS •
Dilution recipe
o o o o
10 g of PEDOT/PSS 10 g of isopropanol 0.5 g of 1-Methyl-2-pyrrolidone 0.3 g of D-sorbitol
• I/V curves
I-V characteristics of pure spin-coated PEDOT/PSS
Third year internship report
61/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
I-V characteristics of pure PEDOT/PSS drop
I-V characteristics of diluted spin-coated PEDOT/PSS
Third year internship report
62/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
I-V characteristics of pure PEDOT/PSS drop
Third year internship report
63/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
Bibliography – Webography [1] Nano-Chemical-Electrical-Mechanical Manufacturing systems center (Nano-CEMMS), Abridged projet description from proposal, submitted to NSF February 14, 2003 [2] H. Sirringhaus, T. Kawase, R. H. Friend, T. Shimoda, M. Inbasekaran, W. Wu and E. P. Woo, High-Resolution Inkjet Printing of All-Polymer Transistor Circuits, Science, Vol 290, 2000, 2123-2126 [3] G. Perçin, T. S. Lundgren and B. T. Khuri-Yakub, Controlled ink-jet printing and deposition of organic polymers and solid particles, Appl. Phys. Lett., Vol 73, No 16, 1998, 2375-2377 [4] J. M. K. Ng, I. Gitlin, A. D. Stroock, G. M. Whitesides, Components for integrated poly(dimethylsiloxane) microfluidic systems, Electrophotoresis 2002, 23, 3461-3473 [5] J. R.Anderson, D. T.Chiu, R. J. Jackman, O. Cherniavskaya, J.C. McDonald, H. Wu, S. H. Whitesides and G. M.Whitesides , Fabrication of Topologically Complex Three-Dimensional Microfluidic Systems in PDMS by Rapid Prototyping, Anal.Chem.2000,72,3158-3164 [6] H. Wu, T.W. Odom, D. T. Chiu, G.M. Whitesides, Fabrication of Complex ThreeDimensional Microchannel Systems in PDMS, J. Am. Chem. Soc., Vol. 125, No 2, 2003, 554559 [7] K. Yoshimura, M. Kishimoto, T. Suemune , Inkjet Printing technology, Oki technical Review, Vol. 64, 41-44 [8] US 6,712,986 patent [9] US 6,652,074 patent [10] M. A. Unger, H-P. Chou, T. Thorsen, A. Scherer, S. R. Quake, Monolithic Microfabricated Valves and Pumps by Multilayer Soft Lithography, Science, Vol 288, 113116 [11 ] US 6,332,677 patent [12] E. Menart, L. Bilhaut, J. Zaumseil and J. A. Rogers, Improved Surface Chemistries, Thin film deposition technique, and stamp designs for nanotransfer printing, Langmuir, 2004, Vol 20, 6871-6878 [13] http://www.dupont.com/kapton/products/H-78305.html [14] W. R. Childs, R.G. Nuzzo, Decal Transfer Microlithography : a New-Soft-Lithographic Patterning Method, J. Am. Chem. Soc., Vol. 124, No 45, 2002, 13583-13596
Third year internship report
64/65
Pumped Elastomeric Microfluidic Systems for High Resolution Printing
March-August 2004
[15] Y-L. Loo, R. L. Willett, K. W. Baldwin and J. A. Rogers, Interfacial chemistries for nanoscale transfer printing, J. Am. Chem. Soc. 2002, Vol 124, 7654-7655 [16] P. K. Subrahmanyan, Laser Micromachining in the microelectronics industry: emerging applications [17] http://www.usc.edu/org/techalliance/Anthology2003/Final_Thelen.pdf [18] http://www.esi.com/products/pdf/cft/micromachining.pdf [19] http://www.esi.com/products/pdf/cft/laser_micromachining_2.pdf [20] Electromigration for designers, an introduction for the non-specialist [21] M. Mas-Torrent, D. den Boer, M. Durkut, P. Hadley and A. P. H. J. Schenning, Field effect transistors based on poly(3-hexylthiophene) at different length scales, Nanotechnology, 15, S265-S269 [22] Z. Bao, A. Dodabalapur and A.J. Lovinger, Soluble and processable regioregular poly(3hexylthiophene) for thin film field-effect transistor applications with high mobility, Appl. Phys. Lett., Vol. 69, 4108-4110 [23] J. Lu, N. J. Pinto and A.G. MacDiarmid, Apparent dependence of conductivity of a conducting polymer on an electric field in a field effect transistor configuration, Journal of Applied Physic, Vol 92, No 10, 6033-6038 [24] S-H. Hur, D-Y. Khang, C. Kocabas and J.A. Rogers, Nanotransfer printing by use of non-covalent surface forces : Application to thin film transistors that use single walled carbon nanotube networks and semiconducting polymers, Appl. Phys. Lett, submitted [25] L. B. Groenendaal, F. Jonas, D. Freitag, H. Pielartzik and J. R. Reynolds, Poly(3,4ethylenedioxythiophene) and its derivatives : past, present and future, Adv. Mater., 2000, Vol. 12, No 7, 481-494 [26] J. Y. Kim, J. H. Junng, D. E. Lee and J. Joo, Enhancement of electrical conductivity of poly(3,4-ethylenedioxythiophene)/poly(4-styrenesulfonate) by a change of solvents, Synthetic Metals 126 (2002) 311-316 [27] S. K. M. Jönsson, J. Birgerson, X. Crispin, G. Greczynski, W. Osikowicz, A. W. Denier ver de Gon, W. R. Salaneck and M. Fahlman, The effects of solvents on the morphology and sheet resistance in poly(3,4-ethylenedioxythiphene)-polystyrenesukfonic acid (PEDOT-PSS) films, Synthetic Metals, 139 (2003) 1-10
Third year internship report
65/65
