Lise BILHAUT 2004 class
Second year internship report Engineering Degree of l’École Nationale Supérieure de Physique de Strasbourg
Nano-Transfert Printing studies July, August & September 2003
Training supervisor John Rogers Material Science & Engineering Department 1304 W. Green Street Urbana, IL 61801 USA
Nano-Tranfert Printing Studies
July-September 2003
Content Acknowledgements .................................................................................................................... 4 Abstract ...................................................................................................................................... 5 Résumé ....................................................................................................................................... 5
1.
Training period environment.............................................................................................. 6 1.1. The University of Illinois at Urbana-Champaign....................................................... 6 1.2. The Department of Materials Science and Engineering............................................. 6 1.3. Rogers Research Group.............................................................................................. 7
2.
About nano-Transfert Printing ........................................................................................... 9 2.1. Why we need nano-Transfert Printing? ..................................................................... 9 2.2. For what is nano-Transfert Printing? ......................................................................... 9 2.3. How nano-transfert printing works? ........................................................................ 10
3.
Improvement of printed lines profile ............................................................................... 11 3.1. A nano-Transfert Printing problem and solution ..................................................... 11 3.2. Method used ............................................................................................................. 11 3.3. Results ...................................................................................................................... 14 3.3.1 Observation means ........................................................................................... 14 3.3.2 Comments......................................................................................................... 15
4.
Study of resolution of nano-Transfert Printing ................................................................ 20 4.1. Master....................................................................................................................... 20 4.2. Stamps ...................................................................................................................... 23 4.2.1 Stamp processing.............................................................................................. 23 4.2.2 Influence of making a stamp on the master...................................................... 24 4.3. Molded stamps ......................................................................................................... 29 4.3.1 Norland Optical Adhesive molded stamp ........................................................ 29 4.3.2 Gelest molded stamp ........................................................................................ 32 4.4. Comparison .............................................................................................................. 32
5.
Self-induced grooved alignment of carbon nanotubes..................................................... 35 5.1. Carbon nanotubes onto 4 µm width lines................................................................. 35 5.1.1 Carbon nanotubes onto resist lines................................................................... 35 5.1.2 Carbon nanotubes onto Gelest.......................................................................... 36 5.2. Carbon nanotubes onto 100 nm lines ....................................................................... 37 5.2.1 Carbon nanotubes onto 100 nm Gelest lines.................................................... 37
Second year internship report
2/49
Nano-Tranfert Printing Studies
5.2.2 5.2.3
July-September 2003
Carbon nanotubes spin-coated onto 100 nm Gelest lines ................................ 38 Oxidation of the 100 nm Gelest lines............................................................... 40
Conclusion................................................................................................................................ 41
Appendix 1 Recipe for LOR master......................................................................................... 42 Appendix 2 Stamp processing.................................................................................................. 43 Appendix 3 E-beam evaporation & printing ............................................................................ 45 Appendix 4 Nordland Optical Adhesive process ..................................................................... 46 Table of figures ........................................................................................................................ 47 Bibliography – Webography .................................................................................................... 49
Second year internship report
3/49
Nano-Tranfert Printing Studies
July-September 2003
Acknowledgements I take this opportunity to thank my supervisor John Rogers for having given me a chance. His guidance and motivation were very helpful and it was great to learn how to do research under his supervision. I would also like to thank Scott Maclaren from the Centre for Microanalysis of Materials, Tony Banks from the Frederick Seitz Materials Research Laboratory and Scott Robinson from the Imaging Technology Group for their respective help on the Atomic Force Microscope, the clean-room use and the Scanning Electron Microscope. Thanks again to Scott Maclaren for the revelation and the discussion about the space elevator. Thanks to Joana Maria, Yangxin Zhou, Seokwoo Jeon, Jang-Ung Park, Anshu Gaur, Matthew Meitl and Etienne Menard from my research group for their patience, help and friendship. A special thank to Etienne for the accommodation not always accommodating, the car and the fun and to Matthew for the jumper cables and the way he made me discover young American people.
Second year internship report
4/49
Nano-Tranfert Printing Studies
July-September 2003
Abstract This training period took place in the John Rogers group in the Material Science and Engineering Department at the University of Illinois at Urbana-Champaign. After a short introduction to the training period environment, the second part of this report gives some details about the nano-Transfert Printing (nTP) method. To use nTP technique, a mold, called master, has first to be made (usually with conventional lithographic techniques) and then polymeric stamps can be molded on this master. Each stamp reproduces the inverse of the master patterns which are then transferred when the stamp is brought into conformal contact with a substrate. The third part deals with the improvement of the profile of few microns width lines printed with nTP. We used for this purpose a particular resist to design a master in such a way that the stamp has a re-entrant profile: this profile prevents deposition of metal on side walls (of the stamp’s raised patterns) and as a result the edges of printed patterns are improved. The fourth part is about the research of the nTP technique ultimate resolution. We used a master covered with carbon nanotubes in order to have stamps with patterns as small as possible. These stamps were characterized with an Atomic Force Microscope, which has a vertical resolution of few angstroms. In the final part, a method used to try to align carbon nanotubes is briefly presented: the tubes were put on grooves in order to self-induce their alignment.
Résumé Ce stage s’est déroulé dans le groupe de John Rogers qui appartient au Département de Science des Matériaux et d’Ingénierie de l’Université de l’Illinois à Urbana-Champaign (USA). Après avoir présenté l’environnement du stage, nous consacrons la deuxième partie de ce rapport à une courte explication de la méthode d’impression par nano-transfert (nTP). C’est en effet sur cette technique de lithographie additive qu’a porté le travail de recherche effectué. On fabrique tout d’abord un moule appelé master (le plus souvent par des techniques de lithographie classique) à partir duquel on peut générer plusieurs tampons en polymère qui reproduisent le motif inverse que celui que l’on veut imprimer. Le transfert du motif se fait simplement en posant le tampon sur le substrat. La troisième partie porte sur l’amélioration du profil des motifs imprimés par nTP. Nous avons à cette fin utilisé une résine spéciale pour la fabrication du master afin d’avoir un profil ré-entrant sur le tampon. Ce profil empêche le dépôt de métal sur les bords du tampon et améliore ainsi grandement le profil des motifs transférés. La quatrième partie traite du problème de la résolution de la technique de nTP: il s’agissait de fabriquer des tampons dont les motifs soient les plus fins possibles. Nous avons choisi de travailler avec des masters faits à partir de nanotubes de carbone et nous les avons caractérisés avec un Microscope à Force Atomique (dont la résolution verticale est de l’ordre de quelques angströms). Enfin, dans la cinquième et dernière partie, nous présentons quelques essais effectués pour essayer d’aligner des nanotubes de carbone en les déposant sur des substrats présentant des cannelures.
Second year internship report
5/49
Nano-Tranfert Printing Studies
July-September 2003
1. Training period environment 1.1. The University of Illinois at UrbanaChampaign The University of Illinois at Urbana-Champaign (the UIUC) [1] is one of the original 37 public land-grant institutions created within 10 years of the signing of the Morrill Act by Abraham Lincoln in 18621. It was chartered in 1867 as the Illinois Industrial University and opened in 1868. Its current president is James W. Stukel. The university is located in the twin cities of Champaign and Urbana, situated about 140 miles south of Chicago, USA. The UIUC greet 38,291 students: 28,271 undergraduate and 10,020 graduate and professional. Students are typically from 50 states and 100 nations. There are 1,908 tenured faculty members, 873 professors, 498 associate professors and 531 assistant professors. There are 11 Nobel laureates and 18 Pulitzer Prize winners among the alumni. The total budget was $1.244 billion in 2003. The incomes come both of public and private support: auxiliary enterprises and miscellaneous funds ($351.6 million), tuition and fees ($197.9 million), private grants and contracts ($114.4 million), state and federal grants and contracts and federal appropriations ($268 million) and state funds (292.3 million). The UIUC enjoy the largest public university collection in the world with more than 40 departmental libraries and divisions. The library has more than 22 million items, including nearly 9.8 million volumes, more than 13 million printed and no printed materials and over 90,000 periodicals and journals. Its online catalogue is accessed more than one million times weekly! The campus research is ranked 16th of all the universities in America in spending on research and development in science and engineering in 1999 ($358 million). There are more than 80 centers, laboratories and institute which perform research for government agencies, industry and campus units.
1.2. The Department of Materials Science and Engineering 1
Passed on July 2, 1862, this act made possible for new western states to establish colleges for their citizens. The new land-grant institutions, which emphasized agriculture and mechanic arts, opened opportunities to thousands of farmers and working people previously excluded from higher education.
Second year internship report
6/49
Nano-Tranfert Printing Studies
July-September 2003
The Department of Materials Science and Engineering (MatSE) [2] was established in 1987 with the merger of the Departments of Ceramic Engineering and Metallurgy & Mining Engineering. Although a relatively young department, MatSE actually was built on a tradition of excellence that dates back to the founding of the University of Illinois in 1867, when the university was required to have a mining program as part of its mission as a land-grant institution. The Mining Engineering Department later included metallurgy and petroleum engineering, and in 1940 the Department of Metallurgy & Mining Engineering was established. The undergraduate curriculum in mining engineering was phased out in the late 1960s. The Department of Ceramic Engineering was established in 1905. MatSE is organized into five areas of concentration—biomaterials, ceramics, electronic materials, metals, and polymers. Students take common core courses, then specialize in one area according to their research interests. The department is one of the largest in the nation, with about 30 full-time faculties and over 300 undergraduate and graduate students. MatSE at Illinois has earned a reputation as one of the top three materials programs in the nation (the department ranked first for the undergraduate program and second for the graduate program in a recent poll of U.S. News and World Report2).
1.3.
Rogers Research Group
Rogers research Group [3] seeks to understand and exploit interesting characteristics of “soft” materials, such as polymers, liquid crystals, and biological tissues. Its aim is to control and induce novel electronic and photonic responses in these materials; it also develops new “soft lithographic” and biomimetic approaches for patterning them and guiding their growth. This work combines fundamental studies with forward-looking engineering efforts in a way that promotes positive feedback between the two. Its current research focuses on soft materials for flexible “macroelectronic” circuits, nanophotonic structures, microfluidic devices, and microelectromechanical systems. These efforts are highly multidisciplinary, and combine expertise from nearly every traditional field of technical study. Some highlights of its recent work include the first: • flexible paper-like displays • tunable microfluidic optical fiber • stamping techniques with nanometer resolution • liquid crystal modulators built on optical fiber There are 4 current research projects:
2
Ranking computed in January 2003.
Second year internship report
7/49
Nano-Tranfert Printing Studies
July-September 2003
Plastic and Molecular Electronics: This field is interesting partly because it has the potential to enable useful devices - flexible paperlike displays, woven electrotextiles, low cost identification tags, etc - that might be difficult to achieve with established silicon technologies. In addition, many of the materials and processing approaches developed for this area are important for a variety of emerging carbon-based nanoelectronic systems that might have roles in future high density memories or processors. This project focuses on fundamental and applied aspects related to the active organic materials and the lithographic methods that are used to pattern circuits out of them. The work involves materials ranging from small molecule semiconductors, to polymeric electroluminescent materials, to single wall carbon nanotubes, to organic self-assembled monolayers. Patterning these materials into active electronic components allows their chemistry and other basic properties to be correlated to device performance. Microfluidics and Liquid Crystals for Photonics This project seeks to exploit pumped microfluidics for new classes of tunable photonic devices. It includes basic study and development of means to fabricate microfluidic networks and phenomena, such as electrowetting, that can be used to pump the fluids. With proper designs, the motion of the fluids can be coupled to the optical properties of basic photonic elements such as planar waveguides and optical fiber. In another approach, it is possible to construct directly these and other elements (e.g. microlenses) out of fluidic structures whose shapes can be dynamically adjusted. These types of technologies have capabilities that can complement those of conventional systems. A closely related effort attempts to understand fundamental issues and practical considerations that define upper limits for the operating speed of liquid crystal based modulators and switches. It includes a component that focuses on inventing unusual means to use liquid crystals for tunable fiber and integrated optical devices. Unconventional Techniques for Nanofabrication New tools for fabricating structures with micron and nanometer dimensions are critical to the progress of nanoscience and nanotechnology. This project seeks to develop soft lithographic methods for nanofabrication, and to use them for building structures that are needed for basic and applied studies. Our recent efforts focus on methods for building 2D and 3D nanophotonic systems and for constructing organic transistors and diodes that have nanometer or molecular scale dimensions. Microstructural Acoustics and Picosecond Ultrasonics: Picosecond pulsed lasers provide a convenient source of acoustic waves with frequencies in the GHz range and with wavelengths between one and several hundred microns. This project seeks to develop and use these laser-based tools to study the high frequency acoustic responses of structures with characteristic dimensions that are similar to the acoustic wavelengths: thin films and membranes, multilayer stacks, microfluidic networks, etc. Analysis of these measurements yields intrinsic mechanical and thermal properties on micron length scales. These methods can also be used for basic studies of phononic bandgaps and other interesting acoustic phenomena in micro and nanofabricated structures.
Second year internship report
8/49
Nano-Tranfert Printing Studies
July-September 2003
2. About nano-Transfert Printing 2.1.
Why we need nano-Transfert Printing?
New methods for micro- and nanofabrication will be essential to scientific progress in many areas of biology, physics, chemistry and materials science. They will also form enabling technologies for applications ranging from microfluidic devices to micro-optical components to molecular diagnostics to plastic electronics to nanoelectromechanical systems. In many cases, advances will be aided by the highly engineered and spectacularly successful lithographic techniques that are used for microelectronics. These methods have certain drawbacks, however, that will limit their applicability to new devices and fields of study. For example, photolithographies cannot be used with many organic and biological materials due to their chemical incompatibility with typical photoresists and developers; they cannot easily pattern features with dimension of less than 100 nm; they require expensive capital equipment and facilities; they have difficulty forming features on curved, uneven or rough objects, they can only directly pattern a small set of specialized, photosensitive materials, they cannot reproduce features with complex, threedimensional shapes, and they can only pattern small areas in a single step. This situation creates a need for research into alternative patterning methods with capabilities that can complement those of photolithography and other established approaches.
2.2.
For what is nano-Transfert Printing?
Nano-Transfert Printing (nTP) is a low-cost technique for rubber stamping invented in 2002 ([4] and [5]). It combines the high spatial resolution of sophisticated forms of photolithography with capabilities that are not present in other approaches (e.g. single-step patterning of large areas and nonplanar surfaces). NTP will be useful for applications where established methods are ineffective. Two areas are particularly promising: Plastic electronics, where the chemical incompatibility of the constituent materials with common photoresists and developers can preclude the use of photolithography, and where nTP with rotating cylindrical stamps forms an excellent match with the type of reel-to-reel processing that is envisioned for these systems; New classes of optical-fiber and microcapillary-based devices, where nTP allows high resolution circuits, photomasks and actuators to be printed directly on the highly curved surfaces of cylinders with submillimeter diameters.
Second year internship report
9/49
Nano-Tranfert Printing Studies
2.3.
July-September 2003
How does nano-Transfert Printing work?
The Figure 1 shows the basic steps of the method: 1 – Some polydimethylsiloxane (PDMS) is used to fabricate a stamp from a master template which can be formed using conventional lithographic methods; 2 – The stamp is removed from the master by peeling away the cured polymer. Some gold is then evaporated by e-beam on the stamp; 3 – After the evaporation, the stamp is brought in contact with the substrate. Although the whole exposed surface on the stamp is covered with gold, only the raised regions that come into contact with gold coated substrate are transferred. In addition it does not require routine access to clean-room and photolithographic equipment, once a master is made, it can be used to make multiple stamps.
PDMS
Resist pattern
Master
Au Printed pattern
1
2
Substrate 3
Figure 1 – Schematic description for the transfer of Au patterns using a PDMS stamp
Second year internship report
10/49
Nano-Tranfert Printing Studies
July-September 2003
3. Improvement of printed lines profile 3.1. A nano-Transfert Printing problem and solution Patterns printed with nTP show profiles which are not very well defined: the edges of the lines roll up (see Figure 13). The first objective of this training period was to improve this profile. We choose to use some Lift-Off Resist (LOR) developed by MicroChem. Used in combination with conventional positive resists, LOR has the propriety to develop isotropically, creating a bi-layer re-entrant sidewall profile.
3.2.
Method used
The first objective was to find the recipe to have good masters with good undercut (see Appendix 1 for the whole recipe). The lithography was made in a class-100 clean-room in the Frederick Seitz Materials Research Laboratory. We choose to work on a mask with lines width from 2 µm up to 100 µm. The main steps of the process are the following (see Figure 2): Step 1: we spin-coated some LOR A (the A-series have relatively low dissolution rates and offer good undercut control) on a silicon wafer at 3,000 rpm (round per minute) for 30 s. Then we prebaked the LOR for 5 min at 130°C on a hotplate (the prebake temperature is the parameter with the greatest influence on undercut rate); Step 2: we spin-coated some S1818 resist on the LOR and then we baked it for 5 min at 110°C on a hotplate; Step 3: we exposed the imaging resist to an ultraviolet wavelength (wavelength of 365 nm and power of 16.5 mW.cm-²) under the chosen mask with an aligner; Step 4: we developed the resist and the LOR with MF-319 developer for 70 s: the LOR layer creates a re-entrant profile under the S1818 resist.
Second year internship report
11/49
Nano-Tranfert Printing Studies
July-September 2003
LOR
S1818
Substrate (Si) Step 1: coat and prebake LOR
Step 2 : coat and brebake S1818
UV light
Mask
Undercut
Step 3: expose imaging resist
Step 4: develop resist and LOR in MF-319
Figure 2 – Main steps for the master preparation
The second objective was to make the stamp. We choose to make a bi-layer stamp to have a better resolution of the pattern [6]. The whole recipe is in Appendix 2 but the main steps are (see Figure 3): Step 5: a self assembled monolayer of (Tridecafluoro-1,1,2,2 tetrahydrooctyl)-1trichloro silane was first formed on the master to prevent the polymer from sticking; Step 6: we put a layer of a polymer nick-named Gelest (this is this layer which gives a better stamp resolution) and put the master in a Petri dish; Step 7: we poured over the poly(dimethylsiloxane) (PDMS) and cured the stamp in an oven; Step 8: we released the composite stamp from the master by cutting and carefully peeling it away from the master surface.
Second year internship report
12/49
Nano-Tranfert Printing Studies
July-September 2003
Gelest
(Tridecafluoro-1,1,2,2 tetrahydrooctyl)-1-trichloro silane
Step 5: self assembled monolayer of silane deposition
6 : coat Gelest and prebake
PDMS
Step 7: coat PDMS
Step 8: remove from the master Figure 3 – Stamp’s preparation
The third objective was to print lines. We used the Temescal Electron Beam Deposition System at the Frederick Seitz Materials Research Laboratory to make the e-beam evaporation. The process is (see Figure 4 and also Appendix 3): Step 9 : we first oxidized the stamp to prevent any cracks on the printed lines; Step 10: we evaporated 3 nm of titanium at 3 Å.s-1 then 20 nm of gold at 10 Å.s-1 with the e-beam evaporator; Step 11: we transferred the lines by cold welding on a gold coated substrate: we just put the stamp on top of the substrate, the contact between the substrate and the relief region propagates by wetting, the gold on the relief lines is hence transferred.
Second year internship report
13/49
Nano-Tranfert Printing Studies
July-September 2003
Step 9: oxidation of the stamp
Step 10: evaporation of titanium and gold by e-beam
Step 11: bring the stamp in contact with the substrate (the lines are transferred by cold welding)
Figure 4 – Evaporation of metal and lines printing
3.3.
Results 3.3.1
Observation means
We use two different means of observation to look at masters, stamps and transferred lines: •
Scanning Electron Microscopy (SEM) from the Imaging Technology Group. The SEM used was a Field-Emission Environmental Scanning Electron Microscope (model Philips XL30 ESEM-FEG, manufactured by FEI Company). This microscope creates the magnified images by using electron instead of light waves used in conventional light microscope. The samples have to be prepared to be conductive: we sputtered about 40 Å of gold-paladium on the master and on the stamp. We used a sputter-coater on this purpose (model Desk II TSC, manufactured by Denton Vacuum). As the coated stamps and the gold lines were covered by gold, we observed them directly.
•
Atomic Force Microscopy (AFM) from the Centre for Microanalysis of Materials was use to look at the printed lines. The AFM used was a Digital Instruments
Second year internship report
14/49
Nano-Tranfert Printing Studies
July-September 2003
Dimension 3100 AFM. We used the contact mode which measures topography by sliding a probe tip across a surface.
3.3.2
Comments
On Figure 5 and Figure 6, we can see a line on a master made with (LOR master) and without LOR (usual master). The LOR layer and the undercut are clearly visible on Figure 7. Figure 11 and Figure 12 are the corresponding stamps: the undercut creates a re-entrant profile on the stamp (see Figure 8 for a magnified view). Figure 9 and Figure 10 show stamps after titanium and gold coating. The re-entrant profiles on the LOR stamp noticeably prevent deposition of metal on side walls. Figure 13 and Figure 14 show the edge of a printed line made with both usual stamp and LOR stamp. In Figure13, the edge of the line rolls up and is not well defined whereas on Figure 14, the line edge is nearly perfect. By using a layer of LOR between the substrate and the resist, we were able to greatly improve the edge definition of the patterns printed by nTP.
Figure 5 – Master made without LOR (usual master)
Second year internship report
15/49
Nano-Tranfert Printing Studies
July-September 2003
Figure 6 – Master made with LOR (LOR master)
Figure 7 – Magnification on the undercut (LOR master)
Figure 8 – Magnification on the re-entrant profile (stamp)
Second year internship report
16/49
Nano-Tranfert Printing Studies
July-September 2003
Figure 9 – Stamp after Ti/Au coating (from an usual master)
Figure 10 – Stamp after Ti/Au coating (from a LOR master)
Second year internship report
17/49
Nano-Tranfert Printing Studies
July-September 2003
Figure 11 – Stamp after printing (from an usual master)
Figure 12 – Stamp after printing (from a LOR master)
Second year internship report
18/49
Nano-Tranfert Printing Studies
July-September 2003
Figure 13 – Edge of a line printed with an usual stamp
igure 14 – Edge of a line printed with a LOR stamp
F
Second year internship report
19/49
Nano-Tranfert Printing Studies
July-September 2003
4. Study of resolution of nanoTransfert Printing The second project of this training period was to determinate the resolution of nTP method. The approach was to use some master with patterns as small as possible and cast stamps from them. We use some silicon substrate with carbon nanotubes, as diameter of single-walled carbon nanotubes (swnt) is about 2 nm.
4.1.
Master
We first tried to use three different kinds of carbon nanotubes: •
Nanotubes made by High-Pressure Carbon Monoxide conversion synthesis, not purified, suspended in dichloromethane (HiPCO tubes, Figure 15);
•
Carbon nanotubes produced by laser ablation bought to Dupont suspended in dichloromethane (laser tubes, Figure 16);
•
HiPCO tubes purified by a high-temperature vacuum anneal (1600 degrees at 10-8 Torr), suspended in dichlorobenzene (purified HiPCO tubes, Figure 17). We prepared masters as follow: Step 1: we put the tubes solution in an ultrasonicator for one hour : it allows the clusters of tubes to break up; Step 2: we spun a silicon wafer at 10,000 rpm and dropped on at the same time about 30 drops of the tubes solution.
Second year internship report
20/49
Nano-Tranfert Printing Studies
Figure 15 – HiPCO tubes
July-September 2003
Figure 16 – Laser tubes
Figure 17 – Purified HiPCO tubes
To image the master, we used the AFM in tapping mode (ie the measure of topography is made by oscillating the probe tip: it reduces the sample damage). We tried to use some ultrasharp tips manufactured by MikroMasch to get better images. Whereas the radius of curvature of usual tip (BS – Tap300AI from BudgetSensors) is about 20 nm (this is the main limit to the horizontal resolution), the radius of curvature of these tips is about 2-3 nm, which should improve the horizontal resolution. However, these tips appeared not improve the resolution of the AFM images: in Figure 18 and Figure 19, we can see a section analysis of a master made with an usual tip and an ultrasharp tip. As the height does not strongly depend on the radius of curvature of the tip, we can suppose the height information is quite accurate: in both images, the height is about 2 nm (± 0.5 nm) which allows us to tell we are in presence of a swnt. So, as a nanotube is cylindrical, the horizontal cross-section should be about 2 nm too. However, in both images, the horizontal cross-section remains about 48 nm (± 2 nm), which is undoubtedly due to the radius of curvature of the tip.
Second year internship report
21/49
Nano-Tranfert Printing Studies
July-September 2003
Figure 18 – swnt (HiPCO tubes) (image taken with a BS Tap300AI tip from Budget sensors tip (usual tip)), section analysis
Figure 19 – swnt (HiPCO tubes) (image taken with an Urchin NSC16 ultrasharp tip), section analysis
Second year internship report
22/49
Nano-Tranfert Printing Studies
4.2.
July-September 2003
Stamps 4.2.1
Stamp processing
We show that the way a stamp is made is very important: dropping some Gelest on the master and directly pouring some PDMS over is the worse way to make a bi-layer stamp (Figure 20). Gelest does not remain on the master but intermix with the PDMS back layer. Brebaking the thin layer of Gelest before pouring PDMS over seems to improve the resolution of thin details (Figure 21) whereas the spin-coating of Gelest seems to improve the depth of molded patterns (Figure 22). We hence decided to use both spin-coating and pre-baking of Gelest to make composite stamp (Figure 23).
Figure 20 – Stamp made without any pre-treatment of Gelest before pouring Gelest over
Figure 21 – Stamp made by brebaking Gelest before pouring PDMS over
Figure 22 – Stamp made by spin-coating Gelest before pouring PDMS over
Figure 23 – Stamp made by spin-coating and brebaking Gelest before pouring PDMS over
Second year internship report
23/49
Nano-Tranfert Printing Studies
4.2.2
July-September 2003
Influence of making a stamp on the master
Although one advantage of nTP is that when a master is made, it can be used to make multiple stamps, it seems that making these stamps has some altering influences on the master surface. Nanotubes are always with more or less amorphous carbon, depending on the process they are manufactured. Making a stamp is a way to clean (but not completely) the master of the amorphous carbon: Figure 24 and Figure 26 show the master before we made the stamp. We can see a lot of amorphous carbon has been removed from the master after having released the stamp (Figure 25 and Figure 27).
Figure 24 – Master with laser tubes before making a first stamp
Figure 25 – Master with laser tubes after having made one stamp
Figure 26 – Master with purified HiPCO tubes before making a first stamp
Figure 27 – Master with purified HiPCO tubes after having made one stamp
Although making a stamp seems to clean the master, it also seems that making several stamps from one master contaminate it.
Second year internship report
24/49
Nano-Tranfert Printing Studies
July-September 2003
Figure 28 shows the roughness analysis of a silicon wafer, Figure 29 shows the one of the master (after having spin-coated the nanotubes solution but before putting the silane layer), Figure 30 shows the one of the master after having made one stamp (Figure 31 is the same image but it shows the phase. As the phase of an AFM picture is more sensible to the change of material, we can see better the contamination of the substrate). Figure 32 shows a master after having made several stamps from it: the stains we can see on the Figure 30 are more visible. This contamination introduces some roughness on the molded stamp. The relief of the stains is transferred on the stamp (Figure 33 and Figure 34 show the master and the associated stamp): these stains have a similar height than the pattern; the resolution is hence limited. To verify that it is Gelest which contaminates the master, some X-ray Photoelectron Spectroscopy analysis should be made (XPS gives a spectrum with a series of photoelectron peaks whose binding energy are characteristic of each element : the peak area can be used to determine the composition of the surface).
Figure 28 – Roughness analysis of a Si wafer
Second year internship report
25/49
Nano-Tranfert Printing Studies
July-September 2003
Figure 29 – Roughness analysis of a master (after having spin-coated HiPCO tubes)
Figure 30 – Roughness analysis of a master (after having made one stamp)
Second year internship report
26/49
Nano-Tranfert Printing Studies
July-September 2003
Figure 31 – Phase image of Figure 30
Figure 32 – Master after having made several stamps
Second year internship report
27/49
Nano-Tranfert Printing Studies
July-September 2003
Figure 33 – Contaminated master
Figure 34 – Stamp made with a contaminated master
Second year internship report
28/49
Nano-Tranfert Printing Studies
4.3.
July-September 2003
Molded stamps
We replicated master’s patterns by casting a new stamp from the stamp. The principle is showed on Figure 35: we made a stamp (2) from the master (1) and then, we used this stamp as a master to make a new stamp called the molded stamp (3).We used two ways to make the molded stamp: first, we made it using optical adhesive (4.3.1) and secondly, we made it using the same method we used to make a stamp (4.3.2).
Nanotube
Substrate (Si) (1) Master
(2) Stamp
(3) Molded stamp
Figure 35 – Molded stamp principle
4.3.1
Norland Optical Adhesive molded stamp
We firstly used some Norland Optical Adhesive (NOA): this adhesive is optically clear and liquid. It cures when exposed to long wavelength ultraviolet light (the maximum absorption is between the range of 350-380 nm). We choose to work with the NOA 73 because it has the lowest viscosity and hence, suited to reproducing stamp’s details. The recipe we used is in Appendix 4. The main steps are showed in Figure 36: Step 1: we spin-coated the NOA on a glass substrate; Step 2: we put the stamp on the NOA (the stamp enters into conformal contact by wetting) and then, we exposed it to UV light (wavelength of 365 nm and power of 16.5 mW.cm-²) for 5 min; Step 3: we removed the stamp from the molded stamp.
Second year internship report
29/49
Nano-Tranfert Printing Studies
July-September 2003
NOA layer UV light
Original stamp
Substrate (glass) Step 1 : spin-coat NOA on the substrate
Step 2 : put the stamp on the NOA and expose to UV light
Step 3 : remove the stamp from the molded stamp
Figure 36 – Method to make a NOA molded stamp
Figure 37 shows the AFM picture of the master and Figure 38 shows the one of the NOA molded stamp on the same place.
Second year internship report
30/49
Nano-Tranfert Printing Studies
July-September 2003
Figure 37 – Master
Figure 38 – NOA molded stamp
Second year internship report
31/49
Nano-Tranfert Printing Studies
4.3.2
July-September 2003
Gelest molded stamp
We proceeded the same way we used to make a stamp (3.2. Steps 5-8 and Appendix 2): then we used this stamp exactly like a master. When the Petri dish is removed from the oven, both original and molded stamps can be seen in the PDMS: we just cut the stamp inside the edges. The two stamps separate from each other without any problem. Figure 39 shows the image of a Gelest molded stamp at the same place than the master on Figure 37.
Figure 39 – Gelest molded stamp
4.4.
Comparison
The section analyses show that: •
The resolution of Gelest molded stamp is worse than the one of NOA molded stamp: Figure 42 shows that we cannot measure the height of a pattern which was about 1.5 nm on the master (Figure 18). However, even if the pattern is not accurately
Second year internship report
32/49
Nano-Tranfert Printing Studies
July-September 2003
replicated, we showed that it was possible to make such a molded stamp with a resolution of about ten nm. •
Dimensions of the pattern are smaller on the molded stamp than on the master (it was observed on all the samples). We think it is due to the stamp which shrinks.
•
Although the horizontal distance still remains about 50 nm, the vertical distance, the most accurate information about the dimension we can have with an AFM, is less than 2 nm. This shows that we can reproduce patterns with a resolution of at least 2 nm with nTP.
Horizontal distance Vertical distance
Master 47.058 nm 1.480 nm
NOA molded stamp 40.195 nm 1.099 nm
Figure 40 – Characteristic of a same place in the master and NOA molded stamp (information from the Figure 18 and Figure 41)
Figure 41 – NOA molded stamp, section analysis
Second year internship report
33/49
Nano-Tranfert Printing Studies
July-September 2003
Figure 42 – Gelest molded stamp, section analysis
Second year internship report
34/49
Nano-Tranfert Printing Studies
July-September 2003
5. Self-induced grooved alignment of carbon nanotubes Nanoscale electronic devices made from carbon nanotubes, such as transistors and sensors, are much smaller and more versatile than those that rely on conventional microelectronic chips, but their development for mass production has been thwarted by difficulties in aligning and integrating the millions of nanotubes required. The idea was here to make the swnt align themselves by using a grooved induced alignment. If the nanotubes align themselves on the relief region of a stamp, then they could be transferred by nTP.
5.1.
Carbon nanotubes onto 4 µm width lines 5.1.1
Carbon nanotubes onto resist lines
First, we just tried to drop some tubes in a solution on a silicon wafer with 4 µm width resist (S1818) lines. The solvent evaporates itself for about 2 minutes. The optical microscope images (Figure 43 and Figure 44) show that the tubes mostly put themselves perpendicularly in the recessed areas (ie on the silicon), between the lines.
Figure 43 – HiPCO tubes onto ~ 4µm resist lines
Second year internship report
Figure 44 – Laser tubes onto ~ 4µm resist lines
35/49
Nano-Tranfert Printing Studies
5.1.2
July-September 2003
Carbon nanotubes onto Gelest
We dropped after that the tubes solution on stamps made with masters prepared in 3 and we waited again for the solvent to evaporate. A first look was promising (Figure 45 and Figure 46): the tubes seem to put themselves onto the relief lines, mostly parallel to the lines. But when we do the same experiment again and look at the stamp by SEM (Figure 47 and Figure 48), the tubes appeared not to be mostly on the relief area and parallel.
Figure 45 – HiPCO tubes onto ~ 4µm Gelest lines
Figure 46 – Laser tubes onto ~ 4µm Gelest lines
Figure 47 – HiPCO tubes onto ~ 4µm Gelest lines
Figure 48 – Purified HiPCO tubes onto ~ 4µm Gelest lines
Second year internship report
36/49
Nano-Tranfert Printing Studies
5.2.
July-September 2003
Carbon nanotubes onto 100 nm lines 5.2.1
Carbon nanotubes onto 100 nm Gelest lines
We also used some stamps with 100 nm-lines. We proceeded the same way than in 5.1.2, and looked at the samples first with a SEM (Figure 49, Figure 50 and Figure 51): the ropes of nanotubes put themselves onto the lines, mostly perpendicular. However, we look at the sample with an AFM in order to see what about the smaller tubes (Figure 52): the AFM image shows that there is not a real alignment of the tubes onto the lines.
Figure 49 – HiPCO tubes onto 100 nm Gelest lines
Figure 50 – Laser tubes onto 100 nm Gelest lines
Figure 51 – Purified HiPCO tubes onto 100 nm Gelest lines
Second year internship report
37/49
Nano-Tranfert Printing Studies
July-September 2003
Figure 52 – Purified HiPCO tubes onto 100 nm Gelest lines
5.2.2 lines
Carbon nanotubes spin-coated onto 100 nm Gelest
We deposited tubes on the same stamp but this time by spin-coating: we made the stamp turn at 1,000 rpm (Figure 53) and 10,000 rpm (Figure 54) and in the same time, we dropped some tubes solution drops; but the AFM images do not show that the tubes are more aligned than without spin-coating.
Second year internship report
38/49
Nano-Tranfert Printing Studies
July-September 2003
Figure 53 - Purified HiPCO tubes deposited by spin-coating at 1,000 rpm onto 100 nm Gelest lines
Figure 54 - Purified HiPCO tubes deposited by spin-coating at 10,000 rpm onto 100 nm Gelest lines
Second year internship report
39/49
Nano-Tranfert Printing Studies
5.2.3
July-September 2003
Oxidation of the 100 nm Gelest lines
Finally, we tried to align the tubes by chemically altering the stamp surface [7]: we oxidized the stamp by using the same recipe than the one used to oxidize the stamp in 3.2 (step 9, see also Appendix 3. I) and then looked with the AFM (Figure 55). The first image was promising (the tubes seem to but mostly on the relief region and the majority is aligned perpendicularly to the lines), but we had not the time to investigate further in this direction.
Figure 55 - Purified HiPCO tubes onto 100 nm oxidized Gelest lines
Second year internship report
40/49
Nano-Tranfert Printing Studies
July-September 2003
Conclusion The challenge of the first project was to improve the definition of the edges of the patterns printed by nTP: it has been taken up as we have successfully printed some gold lines with very well defined edges. This first project was also the occasion to get familiarized with the use of the clean-room, the Scanning Electron Microscope and the Atomic Force Microscope. The second project cleared the way for the study of the resolution limit of the nTP method: we established that the Gelest polymer can mold patterns of about 2 nm high, but the means of measurement (an AFM) limited the study: of about 20 nm for the horizontal resolution and of about a few angstroms for the vertical one. This study is being continued. Further measures should be performed using an AFM tip ending with a single nanotube or using the future ultra-vide AFM that the Frederick Seitz Materials Research Laboratory should receive in the near future (this one should have atomic resolution!). It would also be interesting to compute a simulation of the Gelest behavior inside small molds. The last project consisted in exploring a way of aligning nanotubes. As very simple means (ie put nanotubes in grooves and let them align by themselves) did not work, some further investigation implying surface treatment should be carried out.
Nano-Transfert Printing is already used in Rogers Group to build some devices like single-nanotube transistors. This technique is very industrially interesting since it is a low cost additive technique. Furthermore it can be used to pattern sub-micron features in a single step over large areas. But nTP is still a young soft lithographic technique (the first publication about it has been made in 2002) and there is still a lot to do to improve the technique.
Second year internship report
41/49
Nano-Tranfert Printing Studies
July-September 2003
Appendix 1 Recipe for LOR master This process explains how to make master with a LOR layer between the silicon substrate and an usual (here S1818) positive resist. The LOR masters were made in a class100 clean-room at the Frederick Seitz Materials Research Laboratory
1. Clean the wafer with acetone; 2. Bake on a hotplate at 150°C for 10 min to dehydrate the surface; 3. Blew up with a nitrogen gun a lot (it enables a better spreading of LOR); 4. Spin-coat LOR 1A at 3,000 rpm for 30 s; 5. Prebake hotplate 130°C for 5 min; 6. Spin-coat S1818 3,000 rpm for 30 s; 7. Bake hotplate 110°C for 5 min; 8. Expose on ultraviolet light (λ = 365 nm, power of 16.5 mW.cm-²) for 7 s; 9. Develop in MF-319 developer for 75 s.
Comments: The prebake temperature (step 5) is the parameter with the greatest influence on undercut rate. It would be better to work on full wafer to have a very good contact between mask and wafer: it improves the edges of the lines. We used the MF-319 developer because of it was available. However, MicroChem advices to use developer like Shipley’s CD 26 to have a better control of the undercut rate.
The datasheet of the Lift-Off resist 1A can be found at this web-address: http://www.microchem.com/products/pdf/lor_data_sheet.pdf
Second year internship report
42/49
Nano-Tranfert Printing Studies
July-September 2003
Appendix 2 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 silicon3.
II. Polymers preparation • Preparation of Gelest (All the products are from Gelest Corp.) 1. Mix in a polystyrene weighing dish: ~ 3.4 g of (7-8 %Vinylmethylsiloxane)-(Dimethylsiloxane) copolymer; ~ Two drops of 1,3,5,7-tetravinyl – 1,3,5,7 – tetramethylcyclotetrasiloxane; ~ One drop of platinum divinyltetramethyldisiloxane; 2. Put one half an hour in a vesicator to degas the mixture; 3. Put 1 g of (25-30% Methylhydrosiloxane)-(Dimethylsiloxane) copolymer; 4. Mix; 5. Put 5 min in a vesicator. (Be careful: after about 10 min, the mixture becomes to dry!) • 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
(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.
Second year internship report
43/49
Nano-Tranfert Printing Studies
July-September 2003
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.
III. Stamp preparation For a bi-layer stamp: 1. Drop Gelest on the master and spin-coat (spin-coat process: 5 s ramp to pass from 0 to 500 rpm, 500 rpm for 5 s, then 5 s ramp to go from 500 to 1,000 rpm, then 1000 for 40 s); 2. Put the master in a polystyrene Petri dish and cure it for 2 min at 60°C in the oven; 3. Put the PDMS on the master (about 5 mm height); 4. Cure the stamp for 2 hours at 60°C; 5. The composite stamp is released from the master by cutting and carefully peeling away the stamp from the surface.
Second year internship report
44/49
Nano-Tranfert Printing Studies
July-September 2003
Appendix 3 E-beam evaporation & printing I. Stamp oxidation The stamp has to be oxidized before evaporating any metal. This treatment prevents any cracks in the gold printed pattern. We oxidized it in a Uniaxis 79° series Reactive Ion etching system. We created a O2 plasma (pressure 30 mT, O2 flux of 20, DC =10 V for 13 s). The e-beam deposition has to be done short (a few minutes) after this oxidation.
II. Evaporation The evaporation of Ti/Au was made in a Temescal Electron Beam Deposition System. 1. The vacuum is made (about 3.10-6 Torr); 2. Evaporation of titanium at 3 Å.s-1 for 10 s; 3. Evaporation of gold at 10 Å.s-1 for 20 s.
Comments: In the same time we evaporate the metal on the stamp, we evaporate some gold on a glass slide used as substrate to print patterns.
III. Printing Just after the e-beam deposition, the stamp is brought in contact with the coated glass slide. There is a wetting which allows the gold deposited on stamp relief to be transferred on the gold layer of the glass slide without any pressure on the stamp. The stamp is then gently taken off the slide: the gold lines are transferred.
Second year internship report
45/49
Nano-Tranfert Printing Studies
July-September 2003
Appendix 4 Nordland Optical Adhesive process This process allows reproducing the inverse patterns of a stamp with some Nordland Optical Adhesive 73 (NOA). It was made in the class-100 clean room at the Frederick Seitz Materials Research Laboratory.
1 – Put the substrate (here a glass slide) on the spin-coater, and then gently press the NOA bottle to but some NOA on the substrate (gentler the pressure is, fewer bubbles are on the substrate and better is the deposition); 2 – Spin-coat at 6,000 rpm for 40 s; 3 – Bring the stamp in contact with the NOA (be careful to the face brought in contact: it has to be the face with the relief structure!); 4 – Expose on UV light (λ = 365 nm, power of 16.5 mW.cm-²) for 300 s; 5 – Remove the stamp from the substrate.
More information about NOA 73 can be found on Norland’s web-site: http://www.norlandprod.com/adhesives/NOA%2073.html
Second year internship report
46/49
Nano-Tranfert Printing Studies
July-September 2003
Table of figures Figure 1 – Schematic description for the transfer of Au patterns using a PDMS stamp 10
Figure 2 – Main steps for the master preparation .............................................................. 12 Figure 3 – Stamp’s preparation ............................................................................................ 13 Figure 4 – Evaporation of metal and lines printing ............................................................ 14 Figure 5 – Master made without LOR (usual master)........................................................ 15 Figure 6 – Master made with LOR (LOR master).............................................................. 16 Figure 7 – Magnification on the undercut (LOR master) .................................................. 16 Figure 8 – Magnification on the re-entrant profile (stamp) ............................................... 16 Figure 9 – Stamp after Ti/Au coating (from an usual master)........................................... 17 Figure 10 – Stamp after Ti/Au coating (from a LOR master) ........................................... 17 Figure 11 – Stamp after printing (from an usual master) .................................................. 18 Figure 12 – Stamp after printing (from a LOR master)..................................................... 18 Figure 13 – Edge of a line printed with an usual stamp ..................................................... 19 Figure 14 – Edge of a line printed with a LOR stamp ........................................................ 19
Figure 15 – HiPCO tubes....................................................................................................... 21 Figure 16 – Laser tubes.......................................................................................................... 21 Figure 17 – Purified HiPCO tubes........................................................................................ 21 Figure 18 – swnt (HiPCO tubes) (image taken with an usual tip), section analysis......... 22 Figure 19 – swnt (HiPCO tubes) (image taken with an ultrasharp tip), section analysis 22 Figure 20 – Stamp made without any pre-treatment of Gelest.......................................... 23 Figure 21 – Stamp made by bre-baking Gelest before pouring PDMS over .................... 23 Figure 22 – Stamp made by spin-coating Gelest before pouring PDMS over .................. 23 Figure 23 – Stamp made by spin-coating and bre-baking Gelest ...................................... 23 Figure 24 – Master with laser tubes before making a first stamp...................................... 24 Figure 25 – Master with laser tubes after having made one stamp ................................... 24 Figure 26 – Master with purified HiPCO tubes before making a first stamp .................. 24 Figure 27 – Master with purified HiPCO tubes after having made one stamp ................ 24 Figure 28 – Roughness analysis of a Si wafer ...................................................................... 25 Figure 29 – Roughness analysis of a master (after having spin-coated HiPCO tubes).... 26 Figure 30 – Roughness analysis of a master (after having made one stamp) ................... 26 Figure 31 – Phase image of Figure 30................................................................................... 27 Figure 32 – Master after having made several stamps ....................................................... 27 Figure 33 – Contaminated master ........................................................................................ 28 Figure 34 – Stamp made with a contaminated master ....................................................... 28 Figure 35 – Molded stamp principle..................................................................................... 29 Figure 36 – Method to make a NOA molded stamp............................................................ 30 Figure 37 – Master ................................................................................................................. 31
Second year internship report
47/49
Nano-Tranfert Printing Studies
July-September 2003
Figure 38 – NOA molded stamp............................................................................................ 31 Figure 39 – Gelest molded stamp.......................................................................................... 32 Figure 40 – Characteristic of a same place in the master and NOA molded stamp ........ 33 Figure 41 – NOA molded stamp, section analysis ............................................................... 33 Figure 42 – Gelest molded stamp, section analysis ............................................................. 34
Figure 43 – HiPCO tubes onto ~ 4µm resist lines................................................................ 35 Figure 44 – Laser tubes onto ~ 4µm resist lines .................................................................. 35 Figure 45 – HiPCO tubes onto ~ 4µm Gelest lines .............................................................. 36 Figure 46 – Laser tubes onto ~ 4µm Gelest lines ................................................................. 36 Figure 47 – HiPCO tubes onto ~ 4µm Gelest lines .............................................................. 36 Figure 48 – Purified HiPCO tubes onto ~ 4µm Gelest lines ............................................... 36 Figure 49 – Hipco tubes onto 100 nm Gelest lines............................................................... 37 Figure 50 – Laser tubes onto 100 nm Gelest lines ............................................................... 37 Figure 51 – Purified HiPCO tubes onto 100 nm Gelest lines ............................................. 37 Figure 52 – Purified HiPCO tubes onto 100 nm Gelest lines ............................................. 38 Figure 53 - Purified HiPCO tubes deposited by spin-coating at 1,000 rpm ..................... 39 Figure 54 - Purified HiPCO tubes deposited by spin-coating at 10,000 rpm ................... 39 Figure 55 - Purified HiPCO tubes onto 100 nm oxidized Gelest lines............................... 40
Second year internship report
48/49
Nano-Tranfert Printing Studies
July-September 2003
Bibliography – Webography [1] – University of Illinois at Urbana-Champaign web site: http://www.uiuc.edu/index.html [2] – The Department of Materials Science and Engineering web site: http://www.mse.uiuc.edu/ [3] – Rogers research Group web-site: http://rogers.mse.uiuc.edu/http://rogers.mse.uiuc.edu/
[4] – Additive nanoscale patterning of metal films with a stamp and a surface chemistry mediated transfer process: Applications in plastic electronics Y.-L. Loo, R.L. Willett, K.W. Baldwin and J.A. Rogers, Appl. Phys. Lett. 2002, 81, 562 [5] – Interfacial Chemistries for Nanoscale Transfer Printing Y.L. Loo, R.L. Willett, K.W. Baldwin, J.A. Rogers, Am. Che. Soc., 124, 7654, 2002. [6] – Improved Pattern Transfer in Soft Lithography using Composite Stamps, T.W. Odom, J.C. Love, D.B. Wolfe, K.E. Paul and G.M. Whitesides, Am. Che. Soc., April 2002 [7] – Large-scale assembly of carbon nanotubes M. W. Miller and E. S. Williams, Nature, September 2003, vol. 425, p. 36
Second year internship report
49/49
