Grayscale Patterning of Polymer Thin Films with Nanometer Precision

Jul 22, 2008 - Conventional photolithography using a photomask has also been widely employed to produce patterns over wide areas in photoresist films;...
0 downloads 0 Views 633KB Size
Langmuir 2008, 24, 8939-8943

8939

Grayscale Patterning of Polymer Thin Films with Nanometer Precision by Direct-Write Multiphoton Photolithography Xiao Yao, Takashi Ito,* and Daniel A. Higgins* Department of Chemistry, Kansas State UniVersity, Manhattan, Kansas 66506 ReceiVed March 20, 2008. In Final Form: May 18, 2008. ReVised Manuscript ReceiVed May 14, 2008 The fabrication of arbitrary grayscale patterns in poly(ethylene dioxythiophene):poly(styrene sulfonate) (PEDOT: PSS) thin films is demonstrated. Patterns are formed by ablative direct-write multiphoton lithography using a sample scanning microscope and 870-nm light from a mode-locked Ti:sapphire laser. The surface profiles of all etched samples are characterized by atomic force microscopy. Grayscale patterns are produced by modulating the laser focus during etching. Quantitative models describing the etch depth as a function of laser power and focus are presented and employed to reproducibly control film patterning. PEDOT:PSS is found to be etched by a combination of linear and nonlinear optical processes. Sensitization by PEDOT in the composite is concluded to facilitiate removal of PSS. An ultimate etch depth precision of 1 nm is achieved.

Introduction The ability to pattern polymer films on micrometer to nanometer length scales is required for the development of many organic electronic and optoelectronic devices, chemical and biological sensors, and nanofluidic chemical separations systems. As described in recent reviews,1–3 stamping, molding, embossing, and printing techniques have been widely used for polymer film patterning. These methods allow for rapid replicate production of binary and grayscale structures over large substrates, but alteration of the structures to be prepared requires fabrication of a new master and stamp. Conventional photolithography using a photomask has also been widely employed to produce patterns over wide areas in photoresist films; but again, fabrication of a new mask is required each time the pattern is altered.3 Lithographic methods based on direct laser writing in photoresist films offer distinct advantages over these methods in that they can be used to produce arbitrary binary and three-dimensional structures by simply changing the preprogrammed illumination pattern.3–12 However, most resist-based techniques employ chemical development steps to obtain the final structures. * Authors to whom correspondence should be addressed. E-mail: ito@ ksu.edu; [email protected]. (1) Xia, Y.; Rogers, J. A.; Paul, K. E.; Whitesides, G. M. Chem. ReV. 1999, 99, 1823–1848. (2) Geissler, M.; Xia, Y. AdV. Mater. 2004, 16, 1249–1269. (3) Menard, E.; Meitl, M. A.; Sun, Y.; Park, J.-U.; Shir, D. J.-L.; Nam, Y.-S.; Jeon, S.; Rogers, J. A. Chem. ReV. 2007, 107, 1117–1160. (4) Li, L.; Fourkas, J. T. Mater. Today 2007, 10, 30–37. (5) Baldacchini, T.; LaFratta, C. N.; Farrer, R. A.; Teich, M. C.; Saleh, B. E. A.; Naughton, M. J.; Fourkas, J. T. J. Appl. Phys. 2004, 95, 6072–6076. (6) Wu, E. S.; Strickler, J. H.; Harrell, W. R.; Webb, W. W. Proc. SPIE-Int. Soc. Opt. Eng. 1992, 1674, 776–782. (7) Witzgall, G.; Vrijen, R.; Yablonovitch, E.; Doan, V.; Schwartz, B. J. Opt. Lett. 1998, 23, 1745–1747. (8) Cumpston, B. H.; Ananthavel, S. P.; Barlow, S.; Dyer, D. L.; Ehrlich, J. E.; Erskine, L. L.; Heikal, A. A.; Kuebler, S. M.; Lee, I.-Y. S.; McCord-Maughon, D.; Qin, J.; Rockel, H.; Rumi, M.; Wu, X.-L.; Marder, S. R.; Perry, J. W. Nature 1999, 398, 51–54. (9) Kawata, S.; Sun, H.-B.; Tanaka, T.; Takada, K. Nature 2001, 412, 697– 698. (10) Juodkazis, S.; Mizeikis, V.; Seet, K. K.; Miwa, M.; Misawa, H. Nanotechnology 2005, 16, 846–849. (11) Ka¨rkka¨inen, A. H. O.; Tamkin, J. M.; Rogers, J. D.; Neal, D. R.; Hormi, O. E.; Jabbour, G. E.; Rantala, J. T.; Descour, M. R. Appl. Opt. 2002, 41, 3988– 3998. (12) Tamkin, J. M.; Bagwell, B.; Kimbrough, B.; Jabbour, G.; Descour, M. R. Proc. SPIE-Int. Soc. Opt. Eng. 2003, 4984, 210–218.

Laser ablation methods13–15 eliminate the need for photomasks and additional chemical development steps. In laser ablation, material is directly removed from the substrate surface by photothermal or photochemical processes triggered by irradiation of the polymer film.14 Such methods are most commonly employed for fabricating micrometer-scale binary patterns. In contrast, grayscale patterning by laser ablation has been reported in only a very limited number of articles16 and is difficult to achieve for a number of reasons. Most importantly, ablation usually involves nonlinear processes that exhibit strong thresholding and can be difficult to control.14 Ablation by nanosecond lasers is especially hard to control, because of the delivery of significant quantities of energy in a very short period of time (i.e., a single laser pulse). Mode-locked femtosecond lasers offer a distinct advantage over nanosecond lasers in that each pulse delivers a much smaller quantity of energy and many pulses are required to deliver the energy necessary for ablation. The result is frequently a less explosive removal of material and much better control over the ablation process. Our groups have recently demonstrated that submicrometer-scale binary patterns can be prepared in a number of common polymer films by ablative multiphoton photolithography, using femtosecond pulses of near-IR laser light from a mode-locked Ti:sapphire laser.15,17,18 It was shown that polymers possessing UV-absorbing chromophores and high glass transition temperatures yield the greatest lateral etching resolution, with etched poly(methyl methacrylate) exhibiting edge sharpnesses of ∼120 nm in films of ∼80-nm thickness.17 In contrast, visibleabsorbing polythiophene films exhibited significantly worse etching resolution. These differences in etching resolution were attributed to differences in the order of the nonlinear optical processes involved in each case. While high-resolution binary structures are best prepared using materials that are etched by high-order nonlinear processes, the fabrication of grayscale structures is difficult in such materials (13) Srinivasan, R.; Sutcliffe, E.; Braren, B. Appl. Phys. Lett. 1987, 51, 1285– 1287. (14) Lippert, T.; Dickinson, J. T. Chem. ReV. 2003, 103, 453–485. (15) Ibrahim, S.; Higgins, D. A.; Ito, T. Langmuir 2007, 23, 12406–12412. (16) Naessens, K.; Ottevaere, H.; Baets, R.; van Daele, P.; Thienpont, H. Appl. Opt. 2003, 42, 6349–6359. (17) Higgins, D. A.; Everett, T. A.; Xie, A.; Forman, S. M.; Ito, T. Appl. Phys. Lett. 2006, 88, 184101. (18) Xie, A.; Ito, T.; Higgins, D. A. AdV. Funct. Mater. 2007, 17, 1515–1522.

10.1021/la8008877 CCC: $40.75  2008 American Chemical Society Published on Web 07/22/2008

8940 Langmuir, Vol. 24, No. 16, 2008

because of the strong dependence of the etching rate on incident laser intensity. The associated difficulties were manifested in our previous studies as increased film roughness in regions where the polymer was not completely removed to the underlying substrate.15 The preparation of grayscale structures is expected to be much simpler in materials that are etched by lower-order nonlinear processes. This article demonstrates the controlled fabrication of grayscale surface relief patterns in ∼80-nm-thick poly(ethylene dioxythiophene):poly(styrene sulfonate) (PEDOT:PSS) films by ablative multiphoton photolithography.19 PEDOT:PSS was chosen because etching is expected to involve a low-order nonlinear process and because it is of profound technological importance,20 having broad applications in the fabrication of organic photovoltaics,21 LEDs,22,23 and other organic electronic components.24 Film etching is accomplished by focusing 170-fs pulses of near-infrared light (centered at 870 nm) into the thin films, using a sample scanning confocal microscope with a high numerical aperture oil immersion objective.15,17,18 Here, multiphoton-excited fluorescence emitted by the sample is utilized as a means to accurately position and precisely control the laser focus for reproducible grayscale etching of the films. The dependence of the etched depth on laser power at fixed focus proves the etching process involves a combination of linear and nonlinear optical processes. The power-dependent etching data is used to model the focus dependence of PEDOT:PSS etching. This model is subsequently used to accurately control the laser focus during production of grayscale patterns in ∼80-nm-thick PEDOT:PSS films. An ultimate precision in etching depth of 1 nm is deduced from the rms roughness of films etched to a fraction of their full thickness.

Experimental Section Materials. PEDOT:PSS was obtained from H. C. Starck as an aqueous dispersion (Baytron-P). The solution obtained was diluted 1:3 (by volume) with methanol to aid in film formation. As received, the dispersion was ∼1.3% by weight PEDOT:PSS with a PEDOT: PSS weight ratio of 1:2.5. Aqueous solutions of PEDOT:PSS for use in spectroscopic experiments were prepared by further diluting the methanol-containing dispersions 10:1 in deionized water. PSS was obtained from Aldrich as a sodium salt and was used as received. PSS solutions for film deposition were prepared by dissolving 400 mg in 10 mL of deionized water. A small amount of rhodamine B dye was added (final dye concentration was 100 nM) to aid in focusing the microscope during film etching. For spectroscopic experiments, the aqueous PSS solution was further diluted by 100:1 in water. Glass coverslips (1′′ × 1′′; 200-µm thick; Fisher Premium) were used as substrates for all film samples. Before use, all such substrates were cleaned in a Harrick plasma cleaner for 5 min. Preparation of Polymer Films. PEDOT:PSS and PSS films were prepared by spin-coating (at 2000 rpm) directly from the solutions described above. The films were subsequently cured at 110 °C in an oven for ∼20 min. Film thickness in each case was measured by removing a portion of the film using a razor blade and imaging the resulting film edge by atomic force microscopy (AFM). Film thicknesses obtained in this manner were found to average 86 ( 5 (19) McDonald, J. P.; Hendricks, J. L.; Mistry, V. R.; Martin, D. C.; Yalisove, S. M. J. Appl. Phys. 2007, 102, 013107. (20) Kirchmeyer, S.; Reuter, K. J. Mater. Chem. 2005, 15, 2077–2088. (21) Arias, A. C.; Granstro¨m, M.; Thomas, D. S.; Petritsch, K.; Friend, R. H. Phys. ReV. B 1999, 60, 1854–1860. (22) Brown, T. M.; Kim, J. S.; Friend, R. H.; Cacialli, F.; Daik, R.; Feast, W. J. Appl. Phys. Lett. 1999, 75, 1679–1681. (23) de Kok, M. M.; Buechel, M.; Vulto, S. I. E.; van de Weijer, P.; Meulenkamp, E. A.; de Winter, S. H. P. M.; Mank, A. J. G.; Vorstenbosch, H. J. M.; Weijtens, C. H. L.; van Elsbergen, V. Phys. Status Solidi A 2004, 201, 1342–1359. (24) Sirringhaus, H.; Kawase, T.; Friend, R. H.; Shimoda, T.; Inbasekaran, M.; Wu, W.; Woo, E. P. Science 2000, 290, 2123–2126.

Yao et al. nm for PEDOT:PSS films prepared as described above. PSS films were found to be of a similar thickness (90 nm). Direct-Write Multiphoton Photolithography. The experimental setup used for direct-write multiphoton photolithography was described in our previous publications.15,17,18 Briefly, the polymercoated coverslips are mounted on a piezoelectric sample scanning stage that is attached to an inverted epi-illumination microscope. The light source used for etching was a mode-locked Ti:sapphire laser outputting 170-fs pulses of light centered at 870 nm and having a repetition rate of 76 MHz. Light from this laser was first passed through polarization optics to control the incident power, through an electronic shutter to control irradiation of the sample, and through a 630-nm long-pass filter to block light in the UV and visible regions. The laser light was then reflected by a dichroic beam splitter into the back aperture of a 1.3 numerical aperture (NA), 100× oil-immersion objective. The objective produced a diffraction-limited focused spot of light having ∼550 nm 1/e2 diameter in the polymer film. Average incident laser powers given below are estimates of the power at the objective focus and were measured outside the microscope, just before the dichroic beam splitter. Etching of binary patterns was accomplished by raster scanning the polymer film over the focused laser spot (pixel size: 100 nm; pixel time: 40 ms; scan rate: 2.5 µm/s). Under the conditions employed, 1 mW of incident power corresponds to an average intensity at the laser focus of ∼4 × 105 W/cm2 and individual sample regions were exposed for a total time of ∼1 s during raster scanning. In most such experiments, the laser focus was determined as described below. Irradiation of the sample was controlled by a pattern fed into the microscope control software (written in house) that controlled sample scanning and the state of the electronic shutter. Etching of grayscale patterns was accomplished by modulating the laser focus in a controlled fashion, using a closed-loop piezoelectric objective mount (Physik Instrumente). In all such experiments, the laser focus was initially determined by detecting fluorescence excited in the sample as the focus position was scanned. Grayscale patterns with 64 bit resolution were then produced by varying the laser focus, using the model described below. All power-dependent and focus-dependent experiments were performed at least three times and were observed to yield similar results within experimental and curve fitting errors. Atomic Force Microscopy Measurements. Detailed characterization of the etched samples was performed using contact-mode AFM in air. A Digital Instruments Multimode AFM with Nanoscope IIIa electronics was employed. Contact-mode pyramidal AFM tips were purchased from Veeco. The rms surface roughness values were obtained from the etched regions by two methods. For squares etched at fixed focus or fixed power (i.e., Figures 1B and 3A), the rms roughness was determined from approximately the full etched area in each case. The mean rms roughness and standard deviation derived from the data in Figure 1B were determined from eight different regions etched at intermediate powers. The mean rms roughness values reported for the data in Figure 3A,C were obtained from three replicate experiments, performed at the same power and focus positions. The roughness values for the pyramids and spiral ramp were obtained by extracting at least three line profiles from each, line fitting these data, and calculating the rms roughness after leveling by subtraction of the line.

Results and Discussion Etching Mechanism of PEDOT:PSS Films. Etching of PEDOT:PSS is readily accomplished by focusing only a few milliwatts of near-IR light from the Ti:sapphire laser into the film. Figure 1A demonstrates that arbitrary patterns can readily be produced. PEDOT:PSS etching is believed to involve multiphoton-induced depolymerization and vaporization of the polymer fragments from the film surface. To determine the effective order of the nonlinear process(es) involved in PEDOT: PSS etching, the dependence of film etch depth on incident laser power was determined. In these experiments, a series of 5 × 5 µm2 regions were etched (at fixed focus) using different laser powers. Figure 1B presents representative AFM data from such

Grayscale Patterning of Polymer Thin Films

Langmuir, Vol. 24, No. 16, 2008 8941

Figure 2. (A) Absorbance (solid line) and fluorescence (dashed line) from a dilute aqueous PEDOT:PSS solution. The fluorescence spectrum (peaked near 375 nm) was obtained by exciting at 210 nm. (B) Dependence of multiphoton-excited PEDOT:PSS film fluorescence on focus position. Plotted are the experimental data (9), their fit to a Lorentzian function (solid line), and the expected profile for one photon fluorescence excitation (dashed line). The experimental curve is broadened significantly by bleaching and etching of the polymer film.

Figure 1. (A) AFM topography of a binary text pattern etched into a PEDOT:PSS film. Dark regions depict areas where the polymer has been at least partially removed; light regions designate unetched areas. (B) Topographic image of squares (5 × 5 µm2) etched into PEDOT:PSS at the average powers shown, in milliwatts. (C) Line profile taken across three of the etched regions (designated by the white line) in B. (D) Plot of the mean etch depth from each square in B. (E) Power-dependent etch depth data obtained in similar fashion from a PSS film.

experiments, while Figure 1C shows a line profile taken across these data. The etch depth at each power was determined by measuring the change in film height for each etched region. Figure 1D plots the results obtained from the data in Figure 1B. As is readily apparent from these data, etching of the film begins at very low incident powers (