Patterning of Polymer Brushes. A Direct Approach to Complex, Sub

Sep 3, 2010 - Address: 310 Bard Hall, Ithaca, NY 14853. .... Adriana Pavía-Jiménez , Emile A. Schweikert , James W. Thackeray , Peter Trefonas , and...
1 downloads 0 Views 3MB Size
Download PDF
pubs.acs.org/NanoLett

Patterning of Polymer Brushes. A Direct Approach to Complex, Sub-Surface Structures Marvin Y. Paik,† Youyong Xu,† Abhinav Rastogi,†,‡ Manabu Tanaka,† Yi Yi,† and Christopher K. Ober*,† †

Department of Materials Science and Engineering and ‡ Department of Chemistry and Chemical Biology, Cornell University, Ithaca, New York 14853 ABSTRACT We report a unique method to directly fabricate complex polymer brush structures with nanometer scale features by means of electron beam lithography. Polymer brushes for direct patterning were grown from surface-anchored initiator sites using atom transfer radical polymerization. Selected monomers (poly(2-hydroxyethyl methacrylate) and poly(methyl methacrylate)) were used based on their ability to readily scission when exposed to radiation. Single step direct patterning of polymer brushes is attractive as this eliminates many process steps, reducing the possibility of contamination and possibly improving resolution. In addition, we report a method to form subsurface polymer brush channels with nanometer-scale features. With the chains tethered to a surface, a diblock copolymer brush with a negative tone upper layer (polystyrene) and a positive tone under layer (poly(methyl methacrylate)) or (poly(2-hydroxyethyl methacrylate) were patterned to create channels. In the work presented, the direct electron beam patterning behavior of the brushes was studied and fabrication of nanochannels was demonstrated. Imaging of the nanopatterned surfaces was carried out using atomic force microscopy and fluorescence microscopy. KEYWORDS Subsurface, patterning, polymer brushes, electron beam exposure, direct

P

atterned polymer brushes can be exploited as a means to tailoring the surface properties of materials for a desired functionality, shape, and feature dimensions. Polymer brushes in our study refer to tethered polymer chains that are immobilized by one end to a surface or an interface.1,2 The crowded tethering imparts sufficient constraints on the chains forcing them to stretch away from the surface or interface to avoid occupying the same volume. Polymer brushes first attracted attention in the 1950s when Van der Waarden showed that grafting polymer molecules to colloidal particles helped prevent flocculation.3 In more recent studies, they have been considered for use in microelectronics as thin layer dielectric materials4 and as a means to achieving extremely small feature sizes.5 Other applications where they can be found useful are in bioselective surfaces,6 nanofluidic devices,7 microreaction vessels and drug delivery,8 biomimetic material fabrication,9 and cell growth control.10 High-resolution lithography of polymer brushes is typically carried out by first patterning an initiator layer on a substrate, followed by surface initiated polymerization of a suitable monomer from the initiator sites.6 Several techniques (see Figure 1a) have been used to fabricate patterned initiator monolayers such as microcontact printing,11 initiator decomposition,12 scanning probe microscopy techniques,13 nanoimprint lithography,14 and chemical lithography.15 However, growth of patterned brushes in this way

can lead to lower resolution features when the brush height is comparable in length to the pattern width due to chain relaxation into the voided regions during growth.16 If initiator is damaged during patterning, then brush growth can lack uniformity. Also, these approaches can involve timeconsuming, labor intensive steps that make them unattractive for practical use and also increase chances of surface contamination. To avoid these shortcomings, we have studied the direct patterning of preformed polymer brushes. A single step patterning process not only remedies the drawbacks to

* To whom correspondence should be addressed. Address: 310 Bard Hall, Ithaca, NY 14853. Tel: 607-255-8417. Fax: 607-255-2365. E-mail: [email protected]. Received for review: 04/21/2010 Published on Web: 09/03/2010 © 2010 American Chemical Society

FIGURE 1. (a) Conventional patterning methods of polymer brushes and (b) direct patterning method of polymer brush by ebeam lithography. 3873

DOI: 10.1021/nl102910f | Nano Lett. 2010, 10, 3873–3879

SCHEME 1. Surface-Initiated ATRP Reaction Scheme for the Synthesis of the PMMA and PHEMA Homopolymer Brushes As Well As the PMMA-b-PS and PHEMA-b-PS Diblock Copolymer Brush

scale (∼20 nm patterns) through self-assembly.20 By patterning a diblock copolymer brush using a negative tone upper block and a positive tone lower block, we demonstrate that subsurface nanochannels can be formed by directly patterning these diblock copolymer brushes in a single step. No previously reported patterning technique can be used to produce these arbitrarily shaped subsurface patterns, making the direct patterning process even more appealing. These patterned materials could be useful in such applications as nanofluidic devices,21 waveguide materials,22 or directed nanowire assembly.23 The covalent attachment of the silane initiator onto the silicon substrate was first carried out to create a uniform and dense layer (see Supporting Information). The formation of a monolayer of initiator was confirmed by ellipsometry and atomic force microscopy (AFM). The dry ellipsometric thickness of the covalently attached monolayer was measured to be 1.6 ( 0.3 nm (the error is due to uncertainties present during the calculation of the film thickness from the optical data). This value is in good agreement with the theoretical height of the initiator containing self-assembled monolayer (SAM). The surface topography and roughness was measured by AFM. The rms roughness of the initiator coated silicon substrate was 0.2 nm in a 0.5 × 0.5 µm2 scanning area. This value is similar to the rms roughness measured for clean bare silicon pieces. PMMA and PHEMA homopolymer brushes were synthesized via ATRP based on the procedure described by Huck et al.24 in an aqueous medium at room temperature (see Supporting Information). The reaction for the synthesis of the ATRP silane initiator and preparation of polymer brushes is shown in Scheme 1. No sacrificial initiator was added to the solvent mixture and the polymerization was surface initiated and surface confined, preventing the formation of undesired polymer in solution and allowing clean surfaces to be obtained simply by washing the polymer brushes with water, acetone, and ethanol. The polymerization appeared to be well controlled as reported previously by Jones and

conventional methods but is expected to help improve the resolution of the patterned features as well. In this paper, we first report the direct patterning of polymer brushes in a single step using electron beam (ebeam) lithography. By using ebeam lithography to fabricate patterned brushes instead of growth from a patterned initiator layer, chain collapse at the pattern edges can be avoided resulting in higher resolution features.17 As chain relaxation into the voided regions during growth would be avoided, direct patterning would produce structures with far less edge distortions with the use of a good developer (a solvent that would wash away the scissioned polymer fragments but not dissolve the unexposed regions). We have demonstrated in previous work how the choice of developer is crucial in preserving the pattern fidelity in polymer brushes.18 By choice of the developer, one could control the edge collapse of the polymer chains to vary the pattern sizes after direct patterning. A schematic representation of the direct patterning process is shown in Figure 1b. Poly(methyl methacrylate) (PMMA) and poly(2-hydroxyethyl methacrylate) (PHEMA) polymer brushes were grown on silicon surfaces via controlled atom transfer radical polymerization (ATRP). These PMMA and PHEMA homopolymer brushes were directly patterned via e-beam lithography to produce patterned polymer brush surfaces in a single step.18 We have successfully formed polymer brush patterns with critical dimensions as small as 20 nm using this single step approach. In addition, we demonstrate the ability to create subsurface patterns in polymer brushes. Poly(methyl methacrylate)-b-polystyrene (PMMA-b-PS) and PHEMA-b-PS block copolymer brushes were grown by surface-initiated atom transfer radical polymerization (ATRP) on preformed either PMMA or PHEMA brushes. While neither is considered a modern high-resolution resist, PMMA and PHEMA are known positive tone electron beam resists, and PS exhibits negative tone behavior.19 These characteristics have been used by the block copolymer patterning community to create very small © 2010 American Chemical Society

3874

DOI: 10.1021/nl102910f | Nano Lett. 2010, 10, 3873-–3879

Huck and as indicated by a linear increase in thickness of the brushes with polymerization reaction time.24 The homopolymer brushes were characterized by ellipsometry and tapping mode AFM, which indicated that the polymer brushes were homogeneous throughout the silicon substrate. PMMA brushes of ellipsometric thicknesses between 19-50 nm were used in this study. AFM of the PMMA brush surfaces revealed that the surfaces were very smooth with an rms roughness of 0.34 nm. PHEMA brushes with ellipsometric thickness of 50 nm were used in this study. Tapping mode AFM of the PHEMA brushes showed an rms roughness of 0.48 nm. All brushes showed increase in the ellipsometric thickness after reinitiation with styrene. To investigate the feasibility of patterning polymer brushes in a single step, brushes of PMMA and PHEMA were directly patterned using electron beam lithography. As PMMA and PHEMA are known positive tone electron beam and UV resists,25,26 we chose these two corresponding monomers to demonstrate that direct patterning of these polymer brushes is also possible. The patterned polymer brushes were characterized and imaged through the use of tapping mode AFM. PMMA brushes were successfully patterned in a single step on exposure to an electron beam. A 0.5 nA beam current, 100 kV accelerating voltage, and 5 nm pixel size was used for the generation of the contrast curves and for higher-resolution patterning. Doses ranging from 10 to 1200 µC/cm2 were used in this study. The PMMA and diblock brushes were developed in a 1:3 methyl isobutyl ketone (MIBK) to isopropyl alcohol (IPA) mixture for 90 s followed by thorough rinsing in deionized water and then dried under a stream of nitrogen. Pattern depths were measured using AFM and a resulting contrast curve was generated as shown in Supporting Information Figure S1. Using the appropriate dose ascertained from the contrast curve, higher-resolution lines of patterned PMMA brushes were imaged using AFM as shown in Figure 2. Figure 2b shows a directly patterned PMMA brush with 20 nm lines with a pitch of 60 nm, demonstrating the ability of this polymer brush system to achieve high resolution patterned features in a single step. Following the successful patterning of PMMA brushes, we also demonstrated the patterning of PHEMA brushes and compared their sensitivity toward e-beam patterning with that of PMMA brushes. PHEMA is a polymer that has also been demonstrated to exhibit positive tone patterning behavior under e-beam exposure conditions.26 This polymer has pendent hydroxyl groups that can be easily functionalized through simple organic chemistry.27 Thus, PHEMA brushes are a good candidate to prepare surfaces for applications that require controlled immobilization of chemical or biological moieties on patterned brushes. To understand the PHEMA brush character under electron beam exposure, similar patterning conditions were used to that for the PMMA brushes. PHEMA brushes ∼50 nm in thickness were used in this study. The brushes were developed in a 0.09 N TMAH © 2010 American Chemical Society

FIGURE 2. AFM height images of a patterned 19 nm thick PMMA brush. (a) Fifty nanometer exposed lines with 100 nm pitch. (b) Twenty nanometer exposed lines with a 60 nm pitch.

(tetramethylammonium hydroxide) solution for 60 s followed by a rinsing in DI water and then dried under a stream of nitrogen. A contrast curve was generated and is shown in Supporting Information Figure S1. Higher-resolution patterns of PHEMA brushes are shown in Figure 3. Although the PHEMA brush appears to be more sensitive than the PMMA brush, its contrast is much poorer. Thus, the ultimate resolution of the patterned lines in this system appears to be lower in comparison as well. As the brushes are much lower in molecular weight in comparison to the typically 3875

DOI: 10.1021/nl102910f | Nano Lett. 2010, 10, 3873-–3879

with an electron beam. Polystyrene has been shown to be a cross-linkable, negative tone electron beam resist.19 By patterning a brush with a negative tone upper layer and a positive tone under layer, it should be possible to directly pattern stable nanochannels with precise control of the channel size and location. PMMA-b-PS and PHEMA-b-PS brushes were synthesized using surface initiated ATRP (see Supporting Information). The diblock copolymer brushes were grown from well-characterized PMMA brushes (30 nm thick) and PHEMA brushes (50 nm thick), respectively. Ellipsometry clearly showed a change in thickness in going from 30 nm PMMA brushes to PMMA-b-PS block brushes as thick as 120 nm and from 50 nm PHEMA brushes to PHEMA-b-PS block brushes of 80 nm. The rms roughness of the PMMA-b-PS diblock copolymer brushes (30 nm PMMA and 50 nm PS) was measured to be 0.7 nm. The PHEMA-bPS brushes were much rougher and appeared to form nodular structures on the surface. This structure is common in diblock copolymer brushes with large incompatibility between the two blocks.28,29 Although the surface is rougher, the ability to form subsurface patterns within the brush is still possible. PMMA patterning conditions were used for the PMMA-bPS brush as PS is much more sensitive than the underlying PMMA layer. PMMA development conditions typical for e-beam patterning were also used with immersion in a 1:3 MIBK/IPA mixture for 90 s. MIBK is known to be a very good solvent for polystyrene and can easily swell the cross-linked polymer.30 This developer will penetrate the top PS layer and solubilize any scissioned PMMA fragments, allowing their removal. Previous work involving monitoring the diffusion rate of long polymer chains (Mn ) 920 kg/mol) through a swollen cross-linked polystyrene network was done by Wu et al.31 and showed that a polymer chain diffusion underwent three stages. In the first stage, the effective diffusion coefficient was even faster than the same chains in a free dilute solution and that the diffusion of longer chains through a small pore are accelerated by shorter chains that are present. Given that our system involves short scissioned polymer fragments with an original brush molecular weight of ∼20 kg/mol, it is reasonable to assume the diffusion through the cross-linked PS upper layer to allow full development within the 90 s time frame. Using ebeam patterning and development under the same conditions, the patterning of three PMMA-b-PS block copolymer brushes with the same PMMA thickness (30 nm) and different PS thicknesses (30, 50, and 90 nm, respectively) were compared. It was observed during AFM measurement that patterns could be directly observed optically for PMMA-b-PS with a relatively thin PS layer (30 nm), while no patterns could be seen for the other two brushes with thicker PS layers. AFM measurements shown in Supporting Information Figure S2 revealed that the irradiated portion was recessed about 15 nm into the region left by the scissioned PMMA according to a cross-sectional analysis of

FIGURE 3. AFM height images of a patterned 50 nm thick PHEMA brush. (a) Two hundred nanometer exposed lines with 400 nm pitch. (b) Eighty nanometer exposed lines with a 160 nm pitch.

spin-coated electron beam resists, the polymer chains may be more susceptible to collapsing into the void areas. This increased susceptibility in combination with the loss of the constraint to force the chains to stretch away near the pattern edges suggests that additional study of development conditions could further improve resolution. However, many applications such as those in the biomaterials area require only pattern features down to the micrometer regime where line uniformity of the patterned brushes is not very critical. To study the versatility of the direct patterning technique, the single step patterning approach was investigated using diblock copolymer brushes of PMMA-b-PS and PHEMA-b-PS © 2010 American Chemical Society

3876

DOI: 10.1021/nl102910f | Nano Lett. 2010, 10, 3873-–3879

FIGURE 4. An illustration representing the nanochannel fabrication process with the corresponding surface topography as imaged by AFM. A 200 nm channel with 600 nm pitch was patterned. (a) The diblock brush before patterning. (b) The diblock brush after both electron beam exposure and development. (c) The diblock brush after etching the polystyrene layer away using oxygen plasma.

a 300 nm pattern line. Since the PMMA layer is only ∼30 nm high, there may still exist a void ∼15 nm high between the sunken PS layer and the wafer. The probable reason for this retained layer of PS is that the cross-linking of PS spills over to the neighboring PS brush segments. When the PS layer is thin, this cross-linking and support is not sufficient to prevent sagging of the cross-linked PS. However, for the other two PMMA-b-PS block copolymer brushes with thicker PS layers, the neighboring PS chains are sufficiently crosslinked that a free-standing PS film reveals no patterned structures on the surface observable optically or by AFM. These results are outlined in the Supporting Information, Scheme S1a,b as well as Figure S2b. To observe and confirm the resulting nanochannel layer in the PMMA-b-PS brushes with thicker PS layers, etching of the PS was done for the brush with 30 nm PMMA and 50 nm PS, using the SLR-720 Plasmatherm Etcher (see Supporting Information). Figure 4 shows the nanochannel patterning process and the corresponding topology during each step of the procedure. Not until the thicker PS layer was etched away did any topological features appear under AFM. Even after development, no topological change was observed, even with the cross-linked polystyrene undergoing swelling and deswelling (Figure 4). The absence of topographical changes is understandable as Berger et al. studied the mechanical effects taking place after multiple swelling cycles of cross-linked polystyrene beads and saw that the beads retained their shape.32 To eliminate the possibility that the etching process removed the scissioned PMMA instead of the development step, etching was done on a PMMA brush that was exposed to an electron beam but not developed. We found that the polymer fragments from the exposed regions did not get removed significantly faster than the unexposed brush as shown by comparisons between © 2010 American Chemical Society

FIGURE 5. (a) The PHEMA-b-PS diblock brush before patterning. (b) One micrometer channels with 1 µm pitch patterned into the diblock brush. (c) Three hundred nanometer channels with 600 nm pitch patterned into the diblock brush.

optical microscope images of the exposed, etched brush before and after development (see Supporting Information Figure S3). Once the polystyrene layer was removed we were able to confirm the existence of the directly patterned nanochannels with AFM as seen in Figure 4c. For the PMMA-b-PS block brush with 30 nm PMMA and 90 nm PS segments, no patterned structures were observed after ebeam patterning and development, either. Only a flat surface was observed by AFM measurement (See Figure S4a,b in Supporting Information). However, when the same brush was put in a solvent vapor of dichloromethane (a good solvent for both blocks) for 30 min or immersed in the same solvent and dried afterward, clear patterns could be observed optically or by AFM. (See Figure S4c,d in Supporting 3877

DOI: 10.1021/nl102910f | Nano Lett. 2010, 10, 3873-–3879

To conclude, we have successfully demonstrated the ability to directly pattern polymer brushes using electron beam lithography. Homopolymer brushes of PMMA and PHEMA exhibited electron beam resist behavior more sensitive to their spun coat counterparts. Also, the ability to directly pattern polymer brushes in a single step brings with it the advantages of simplicity, environmental benefits, and lower surface contamination risk. We speculate that improvements to the ultimate resolution achievable can be made by investigating different development conditions. Furthermore, we showed that this direct patterning technique could uniquely create subsurface patterned brushes. By exploiting the fact that the polymer chains are tethered to the substrate, using a system where the upper layer is a cross-linkable, negative tone resist with a scissionable, positive tone under layer, nanochannels could be directly written. We also find the selection of a suitable development solvent is very crucial for nanochannel formation.

FIGURE 6. Fluorescent microscopy images of (a) PMMA-b-PS (30 nm PMMA and 30 nm PS); (b) pure PMMA brush; and (c) pure PS brush after immersion in fluorescein/isopropanol solution and washing with deionized water.

Acknowledgment. This work was supported by the National Science Foundation under agreement DMR-0518785. We also acknowledge Cornell NanoScale Science and Technology Facility (CNF) and the Cornell Center for Materials Research (CCMR) for use of their facilities.

Information). Since the PMMA layer is 30 nm thick, there should be still about 20 nm channel space left. These findings indicate the importance of developer selection. A relatively poor solvent can wash away the chain fragments while keep the neighboring chains “frozen”, thus retain the nanochannel structures. However, a good solvent not only develops the small fragments but also softens the neighboring chains and weakens the channel structures. For the PHEMA-b-PS brushes, we did not succeed in identifying a suitable developer due to the high-polarity difference of the PHEMA and PS chains. When DMF, a good solvent for both blocks, was used for the development this not only washed away the PHEMA chain fragments but also softened the supporting PS and PHEMA chains for the nanochannels’ PS cover. Figure 5 shows the AFM images of the patterned PHEMA-b-PS brush. The cross-linked PS layer sunk into the empty channel below, as was observed in the PMMA-b-PS brushes developed in dichloromethane. Attempts were made to use staining by fluorescent dyes to reveal more information about the nanochannels. The polymer brushes were immersed in a fluorescein solution in isopropanol for 3 h and subsequently washed with water. As an example, fluorescence microscopy of the PMMA-bPS brush with 30 nm PMMA and 30 nm PS is shown in Figure 6a and reveals that the patterned lines (ebeam exposed regions from 100 nm to 1 µm) are darker than the unexposed regions. Control experiments with pure PMMA and PS brushes treated in the same conditions are shown in Figure 6b,c. The dye is not retained by the PS brush but is readily absorbed by the PMMA brush. This may be caused by the very polar nature of the fluorescein dye and its attraction to the more polar PMMA chains. Fluorescence therefore reveals the developed subsurface regions where PMMA chains have been removed. © 2010 American Chemical Society

Supporting Information Available. Further information regarding the materials used and purchased, polymer brush synthesis procedure, patterning process, and characterization techniques is provided. This material is available free of charge via the Internet at http://pubs.acs.org. REFERENCES AND NOTES (1) (2) (3) (4)

(5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19)

3878

Milner, S. T. Science 1991, 251, 905. Zhao, B.; Brittain, W. J. Prog. Polym. Sci. 2000, 25, 677. Van der Waarden, M. J. J. Colloid Sci. 1950, 5, 317. Pinto, J. C.; Whiting, G. L.; Khodabakhsh, S.; Torre, L.; Rodriguez, A. B.; Dalgliesh, R. M.; Higgins, A. M.; Andreasen, J. W.; Nielsen, M. M.; Geoghegan, M.; Huck, W. T. S.; Sirringhaus, H. Adv. Funct. Mater. 2008, 18, 36. von Werne, T. A.; Germack, D. S.; Hagberg, E. C.; Sheares, V. V.; Hawker, C. J.; Carter, K. R. J. Am. Chem. Soc. 2003, 125, 3831. Andruzzi, L.; Senaratne, W.; Hexemer, A.; Sheets, E. D.; Ilic, B.; Kramer, E. J.; Baird, B.; Ober, C. K. Langmuir 2005, 21, 2495. Adiga, S. P.; Brenner, D. W. Nano Lett. 2005, 5, 2509. Shen, Y.; Zhan, Y.; Tang, J.; Xu, P.; Johnson, P. A.; Radosz, M.; van Kirk, E. A.; Murdoch, W. J. AIChE J. 2008, 54, 2979. Chen, M.; Briscoe, W. H.; Armes, S. P.; Cohen, H.; Klein, J. ChemPhysChem 2007, 8, 1303. Gupta, B.; Plummer, C.; Bisson, I.; Frey, P.; Hilborn, J. Biomaterials 2002, 23, 863. Jones, D. M.; Smith, J. R.; Huck, W. T. S. Adv. Mater. 2002, 14, 1130. Tsujii, Y.; Ejaz, M.; Yamamoto, S.; Fukuda, T.; Shigeto, K.; Mibu, K.; Shinjo, T. Polymer 2002, 43, 3837–3841. Kaholek, M.; Lee, W.; LaMattina, B.; Caster, K. C.; Zauscher, S. Nano Lett. 2004, 4, 373. Liu, Y.; Klep, V.; Luzinov, I. J. Am. Chem. Soc. 2006, 128, 8106. He, Q.; Kller, A.; Grunze, M.; Li, J. Langmuir 2007, 23, 3981. Patra, M.; Linse, P. Nano Lett. 2006, 6, 133. Jonas, A. M.; Hu, Z.; Glinel, K.; Huck, W. T. S. Nano Lett. 2008, 8, 3819–3824. Rastogi, A.; Paik, M. Y.; Tanaka, M.; Ober, C. K. ACS Nano 2009. Austin, M. D.; Zhang, W.; Ge, H.; Wasserman, D.; Lyon, S. A.; Chou, S. Y. Nanotechnology 2005, 16, 1058. DOI: 10.1021/nl102910f | Nano Lett. 2010, 10, 3873-–3879

(20) Ruiz, R.; Kang, H.; Detcheverry, F. A.; Dobisz, E.; Kercher, D. S.; Albrecht, T. R.; de Pablo, J. J.; Nealey, P. F. Science 2008, 321, 936. (21) Cannon Jr, D. M.; Flachsbart, B. R.; Shannon, M. A.; Sweedler, J. V.; Bohn, P. W. Appl. Phys. Lett. 2004, 85, 1241– 1243. (22) Matsui, T.; Komatsu, K.; Sugihara, O.; Kaino, T. Opt. Lett. 2005, 30, 970–972. (23) Kim, E.-U.; Baeg, K.-J.; Noh, Y.-Y.; Kim, D.-Y.; Lee, T.; Park, I.; Jung, G.-Y. Nanotechnology 2009, 20, 355302. (24) Jones, D. M.; Huck, W. T. S. Adv. Mater. 2001, 13, 1256. (25) Yasin, S.; Hasko, D. G.; Ahmed, H. Appl. Phys. Lett. 2001, 78.

© 2010 American Chemical Society

(26) Vasilopoulou, M.; Boyatzis, S.; Raptis, I.; Dimotikalli, D.; Argitis, P. J. Mater. Chem. 2004, 14, 3312. (27) Diamanti, S.; Arifuzzaman, S.; Elsen, A.; Genzer, J.; Vaia, R. A. Polymer 2008, 49, 3770. (28) Tomlinson, M. R.; Genzer, J. Langmuir 2005, 21, 11552–11555. (29) Wang, J.; Mu¨ller, M. Macromolecules 2009, 42, 2251–2264. (30) Lei, C. P.; Long, T.; Obendorf, S. K.; Rodriguez, F. Polym. Mater. Sci. Eng. 1992, 66, 113–114. (31) Song, Q.; Zhang, G.; Wu, C. Macromolecules 2007, 40, 8061– 8065. (32) Zhang, R.; Cherdhirankorn, T.; Graf, K.; Koynov, K.; Berger, R. Microelectron. Eng. 2008, 85, 1261–1264.

3879

DOI: 10.1021/nl102910f | Nano Lett. 2010, 10, 3873-–3879