Final Report: Dry Lithography: Environmentally Responsible Processes for High Resolution Pattern Transfer and Elimination of Image Collapse using Positive Tone ResistsEPA Grant Number: R829586
Title: Dry Lithography: Environmentally Responsible Processes for High Resolution Pattern Transfer and Elimination of Image Collapse using Positive Tone Resists
Investigators: DeSimone, Joseph M.
Institution: University of North Carolina at Chapel Hill
EPA Project Officer: Richards, April
Project Period: November 1, 2001 through November 1, 2004
Project Amount: $347,898
RFA: Technology for a Sustainable Environment (2001) RFA Text | Recipients Lists
Research Category: Sustainability , Nanotechnology , Pollution Prevention/Sustainable Development
Conventional lithography (the means by which integrated circuits and memory devices are produced) requires over 1 kg of organic solvent and aqueous waste to produce a 2 g chip. The objective of this research project was to explore the use of environmentally responsible, “green” solvents for photoresist synthesis and microelectronics processing in an effort to reduce the consumption/disposal of potentially eco-toxic substances. Additionally, we have developed a new method for solvent-free fabrication of microelectronic structures, which we call Pattern Replication in Non-wetting Templates (PRINT), that dramatically reduces or eliminates solvents and waste in nanometer-scale fabrication.
Carbon dioxide has been shown to be an environmentally responsible alternative for aqueous and organic solvents in a variety of processes. Previous research in our lab has shown the utility of carbon dioxide in the lithographic process at 193 nm. We currently are exploring new CO2-processable photoresist formulations that optimize fabrication parameters, such as etch resistance and CO2 spin coating and development. Industry continues to move to smaller wavelengths to decrease feature size, therefore, we also have started developing a 157 nm positive tone image for a carbon dioxide-based lithographic system. In addition to the environmental benefits of switching from aqueous and organic solvents to carbon dioxide, dramatic performance benefits also should be realized. Industry is moving to larger wafers that must be spin coated with a photoresist. By utilizing carbon dioxide, larger surfaces can be coated at lower spin rates and concentrations. Furthermore, during the development of small images (< 70 nm) with aqueous base, the potential for image collapse is very high given the high surface tension of water. By replacing water with liquid or supercritical carbon dioxide, a low surface tension fluid, the potential for image collapse should be eliminated.
The development of PRINT offers a solvent-free fabrication of sub-100 nm structures that are suitable for next-generation microelectronics. Nanofabrication using PRINT satisfies several of the principles of an environmentally friendly manufacturing system, including waste prevention, high “atom efficiency,” minimal energy usage, and minimal hazard. PRINT belongs to a family of lithographic techniques called “imprint lithography,” where a patterned surface is used to “mold” or “emboss” features in another material. We have developed new materials, perfluoropolyether (PFPE) elastomers that can be used as molds for imprint lithography. Using PRINT, we easily can fabricate sub-100 nm patterned thin films or sub-100 nm isolated objects. By molding photocurable liquids, we can eliminate solvents from the fabrication process altogether. These molds can be used for multiple fabrication steps, which avoids the reagent-intensive processes associated with the microelectronics industry and the toxic chemicals and gases associated with large-scale nanofabrication.
Development of CO2-Processable Photoresists
The astonishing advances in computer technology over the past several decades can be attributed to the continuing development of faster computer microchips. This is because of the ability to pattern the chips with smaller feature sizes. The complex patterns are generated by photolithography with the minimal feature size being directly proportional to the wavelength of the imaging light. The integrated circuit industry would like to use shorter wavelengths of light to produce faster microchips, and 193 nm and 157 nm have emerged as the leading candidates. This necessitates the development of new photoresists that are transparent at these wavelengths. We have successfully synthesized a range of fluoropolymers in CO2 that have potential applications as optical materials in advanced photolithographic processes. The materials are based on Teflon® AF, an amorphous copolymer of tetrafluoroethylene, and 2,2-bis(trifluoromethyl)-4,5-difluoro-1,3-dioxole (PDD), which was previously identified as a material with extremely low absorbance at both 193 nm and 157 nm, and on highly-strained alicyclic polymers.
The Teflon® AF-based photoresists (Figure 1) have high glass transition temperatures (> 120°C) and the protecting groups are stable thermally (up to 200°C). In addition, the photochemical cleavage of the protecting groups has been spectroscopically verified. Moreover, the deprotected polymer is soluble in the industry standard developing solution, whereas the protected polymer is not; this provides the necessary contrast for high-resolution imaging. Uniform thin films of the materials also have been successfully spin coated onto 6-inch silicon wafers. The materials are extremely promising photoresists for emerging next generation photolithographic techniques and currently are being evaluated using a state of the art 193 nm step and scan system ($10 million) located in the Triangle National Lithography Center. Preliminary data suggests that these materials are capable of submicron imaging. Moreover, these polymers are synthesized in CO2, which affords a range of environmental as well as chemical advantages.
In addition to the synthesis of Teflon® AF-based materials, we also have synthesized CO2 processable photoresists using highly strained alicyclic polymers. Addition polymers of highly fluorinated norbornene derivatives, NB-FOA and NB-HFA, were synthesized using an allylpalladium dimer catalyst (Figure 2), then protected with tert-butoxycarbonate (tBOC). Thermogravimetry (TGA) measurements of the tBOC-protected polymer showed a high cleavage temperature of the tBOC at 175°C, and polymer decomposition beginning at 300°C. Such high decomposition temperatures ensure that a resist based on this material will not thermally degrade during baking steps. The polymers were soluble in condensed CO2, as well as fluorinated solvents, with a significant difference in CO2 solubility, thus providing an accessible region for CO2 development (Figure 3). Plasma reactive ion etching studies of this material under a CHF3/O2 etch showed a moderate etch rate of 1.4 times faster than Novolac. Using this resist, 3 micron features were observed after exposure with a 193 nm lithography tool and development in CO2.
Figure 1. Representative Structures of Resists Synthesized in CO2: (a) TFE-Based Materials and (b) Non-TFE-Based Materials
Figure 2. Synthesis of Fluorinated Addition Copolymer of NB-FOA and NB-HFA
Figure 3. Cloud Point Curves of Poly(NB-FOA/NB-HFA), Poly(NB-FOA/NBHFA-tBOC). Both curves obtained at 2.5 wt% polymer in CO2.
Although 3 micron images are a good starting point for this novel material, state-of-the art images are typically less than 100 nm. Further work in this area will focus on the optimization of CO2 development, as well as obtaining lower molecular weight material to yield smaller images using 193 nm lithography.
Solvent-Free Nanofabrication Using PRINT
PRINT uses highly fluorinated, minimally adhesive PFPE materials as molds for imprint lithography. Previously, our group has reported on these materials as novel materials for microfluidics and soft lithography. For this research project, we expanded on our previous results, demonstrating that PRINT can be for solvent-free nanofabrication of a variety of materials on many length scales.
To perform PRINT, a patterned silicon wafer (called the “master”) is first used as a template to structure the PFPE molds. To accomplish patterning of these PFPE elastomers, a pourable liquid PFPE-dimethacrylate precursor (Figure 4) is pooled over the patterned silicon wafer and photopolymerized to form a cross-linked, rubbery material. These rubbery molds are then used to replicate the master pattern in another material, usually a liquid monomeric composition suitable for solvent-free, bulk polymerization. In practice, PFPE molds are pressed against a curable liquid that adopts the topography of the PFPE mold and is polymerized to form a solid, patterned replica of the PFPE elastomer. Because these molds are highly fluorinated, and, therefore, non-swelling and minimally-adhesive, it is possible to make structures with smaller dimensions and/or fewer replication errors than using other imprint lithography materials such as silicone rubber, glass, or silicon. Furthermore, fabrication of patterned PFPE molds inherently is environmentally friendly because PFPE-dimethacrylate polymerization is 100 percent solvent-free and 100 percent atom efficient. Using these molds as templates for further replication, we can fabricate embossed films or isolated objects, depending on the applied pressure and the substrate that is used for molding. Embossed films are fabricated using untreated silicon or glass surfaces, but if a highly fluorinated, “non-wetting” surface is used, isolated objects are formed because the non-wetting surface and mold “squeeze out” residual material between shapes (Figure 4). In fact, PRINT offers the most general route to the fabrication of these types of shape-specific, isolated nano-objects.
Figure 5 shows results that demonstrate the feature resolution that can be achieved using PRINT. Here, we have fabricated sub-100 nm lines by molding and photopolymerizing a solvent-free resist formulation consisting of triacrylate resin, bisphenol-A diacrylate, and N-vinyl pyrrolidinone. Because these patterns were formed by bulk polymerization of these monomers within the PFPE mold, PRINT can significantly reduce the consumables, waste, and energy costs associated with nanofabrication. Whereas conventional photolithography requires significant solvent processing, etching, and baking steps to produce similar features, PRINT can make these types of structures in a one-step, solvent-free manner. Note that the sizes of the features on the silicon master correspond to the sizes on the PFPE mold and on the replicate. This demonstrates the fidelity and resolution that can be achieved using PRINT. We believe that we can move even to smaller feature sizes than is demonstrated here, toward the sub-10 nm regime.
Figure 4. Top: Structure of the PFPE-Dimethacrylate Mold Precursor. Middle and bottom: Fabrication of embossed films and isolated objects using PRINT.
Figure 5. High Resolution AFM Images of a Silicon Master (left) With 140 nm Lines Separated by 70 nm PFPE Mold (middle) With 70 nm Lines Separated by 140 nm, Poly (triacrylate) Replicate Made With PFPE Mold (right) With 140 nm Lines Separated by 70 nm. Beneath each picture is a representative height profile.
Figure 6 shows isolated particles that have been fabricated using PRINT nanofabrication. This figure demonstrates the composition and shape-control that can be obtained. Because the shapes of the particles are determined by the morphology of the PFPE mold, it is easy to control nanoparticle shape simply by controlling the shape of the mold. Additionally, by changing the material that is molded, one can easily produce particles with different compositions. We have fabricated particles successfully out of many materials using solvent-reduced or solvent-free (bulk) polymerizations, including poly (ethylene glycol)-diacrylate, poly (lactic acid), and poly (pyrrole). Most traditional nanofabrication processes, such as surface-energy driven colloidal synthesis, are extremely chemically intensive, requiring solvents, surfactants, and multiple purification steps. In contrast, PRINT uses direct-contact molding with reusable PFPE molds to control nanostructure shape, making it much more environmentally friendly.
Figure 6. SEM Images of PEG, PLA, and PPy Particles Printed Using PRINT. PFPE molds were generated from the original silicon masters and particles were fabricated in three shapes: 3 μ arrows, 500 nm conical structures that are < 50 nm at the tip, and 200 nm trapezoidal structures.
In summary, PRINT offers a solvent-free, environmentally friendly route to nanofabrication. We have demonstrated that we can routinely fabricate structures with sub-100 nm shape control from a variety of materials, including resist formulations, biocompatible materials, biodegradable materials, and conducting polymers. Because PRINT forms nanostructures using direct-contact imprint molding rather than exposure and/or solution processing, it is more environmentally benign than other nanofabrication techniques such as photolithography and chemical synthesis. These attributes make PRINT a desirable route for environmentally friendly manufacturing.
Journal Articles:No journal articles submitted with this report: View all 22 publications for this project
Supplemental Keywords:green chemistry, alternatives, clean technologies, innovative technology, waste reduction, environmentally conscious manufacturing, carbon dioxide, pattern transfer, image collapse, imprint lithography, nanofabrication, solvent-free, nanoparticles, nanowires, polymerization,, RFA, Scientific Discipline, Sustainable Industry/Business, Chemical Engineering, cleaner production/pollution prevention, Environmental Chemistry, Sustainable Environment, Technology for Sustainable Environment, Economics and Business, Environmental Engineering, supercritical carbon dioxide (SCCO2) technology, in-process waste minimization, cleaner production, clean technologies, green design, high resolution pattern transfer, environmentally benign solvents, alternative materials, supercritical carbon dioxide, alternative solvents, engineering, solvent substitute, microelectronics, environmentally benign alternative, dry lithography, pollution prevention, green chemistry
Progress and Final Reports:Original Abstract
2002 Progress Report
2003 Progress Report