Langmuir 2008, 24, 12077-12084
12077
Regulation of Pattern Dimension as a Function of Vacuum Pressure: Alkyl Monolayer Lithography Om P. Khatri, Hikaru Sano, Kuniaki Murase, and Hiroyuki Sugimura* Department of Materials Science and Engineering, Kyoto UniVersity, Sakyo-ku, Kyoto 606-8501, Japan ReceiVed July 8, 2008. ReVised Manuscript ReceiVed August 18, 2008 Photopatterning of a hexadecyl (HD) monolayer has been demonstrated using vacuum ultraviolet (VUV; λ ) 172 nm) light under controlled vacuum pressure with the objective of minimizing the pattern dimension. X-ray photoelectron spectroscopy (XPS) and lateral force microscopy (LFM) studies reveal that photodegradation of the HD monolayer not only is limited to the regions exposed to VUV but also spreads under the masked regions. The strong oxidants generated by VUV irradiation to atmospheric oxygen and water vapor diffuse toward the masked regions through the nanoscopic channels and photodissociate the monolayer under the masked area, near the photomask apertures, resulting in broadening of the photopattern. Such broadening decreases with decreased vacuum pressure inside the VUV chamber, associated with a decrease of oxidant concentration and reduction of their diffusion. Gold nanoparticles (AuNPs) were immobilized on the VUV patterned features to probe the dimension of the chemically active pattern. Field emission electron microscopy reveals the construction of 565 nm wide pattern features at a vacuum pressure of 10 Pa. This pattern widens to 1030 nm at 104 Pa using the same size apertures (500 nm) as printed on the photomask. This study provides insight for fabricating submicron patterns with high reproducibility and its exploitation for different applications, which includes the patterning of nanoparticles, biopolymers, and other nano-objects at submicron dimensions.
Introduction Chemical and morphological characteristics of material surfaces play important roles in emerging applications such as molecular electronics, information storage devices, optical and photonic devices, sensor chips, bimolecular platforms, and so forth.1,2 There has been a great deal of interest in the tailoring of such characteristics at micro- to nanoscale dimensions with spatial distribution of organic molecules, biopolymers, and nanoparticles.3-5 Several techniques including photolithography, microcontact printing, electron and ion beam lithography, dippen writing, and scanning probe microscopy (SPM) based patterning have been demonstrated for surface modification; each technique is specific to pattern type and dimension.6-9 Photolithography that employs a single step photodegradation process is an elegant approach for uniform micropatterning on a large scale using an organic monolayer as a photoresist thin film.10-13 Increasing interest in reducing the pattern size with optimization of pattern quality requires monitoring the photochemistry of the photoresist thin film with spatial resolution. Many factors such as the wavelength of UV light, quality of the photoresist thin film, proximal gap between the photomask and sample surface, * To whom correspondence should be addressed. Telephone: +81 75 753 9130. Fax: +81 75 753 5484. E-mail:
[email protected]. media.kyoto-u.ac.jp. (1) (a) Fan, H.; Lu, Y.; Stump, A.; Reed, S. T.; Baer, T.; Schunk, R.; PerezLuna, V.; Lopez, G. P.; Brinker, C. J. Nature 2000, 405, 56–60. (b) Whitesides, G. M.; Ostuni, E.; Takayama, S.; Jiang, X.; Ingber, D. E. Annu. ReV. Biomed. Eng. 2001, 3, 335–373. (c) Liu, C. AdV. Mater. 2007, 19, 3783–3790. (2) (a) Shipway, A. N.; Katz, E.; Willner, I. ChemPhysChem 2000, 1, 18–52. (b) Daniel, M.-C.; Astruc, D. Chem. ReV. 2004, 104, 293–346. (3) Smith, R. K.; Lewis, P. A.; Weiss, P. S. Prog. Surf. Sci. 2004, 75, 1–68. (4) Veiseh, M.; Zareie, M. H.; Zhang, M. Langmuir 2002, 18, 6671–6678. (5) (a) Hoeppener, S.; Maoz, R.; Cohen, S. R.; Chi, L.; Fuchs, H.; Sagiv, J. AdV. Mater. 2002, 14, 1036–1041. (b) Sugimura, H.; Nakagiri, N. J. Am. Chem. Soc. 1997, 119, 9226–9229. (6) Xia, Y.; Rogers, J. A.; Paul, K. E.; Whitesides, G. M. Chem. ReV. 1999, 99, 1823–1848. (7) Gates, B. D.; Xu, Q.; Stewart, M.; Ryan, D.; Willson, C. G.; Whitesides, G. M. Chem. ReV. 2005, 105, 1171–1196. (8) Innocenzi, P.; Kidchob, T.; Falcaro, P.; Takahashi, M. Chem. Mater. 2008, 20, 607–614. (9) Geissler, M.; Xia, Y. AdV. Mater. 2004, 16, 1249–1269.
vacuum pressure of the UV chamber, and the roughness of the photomask and sample surface are significantly accountable in order to determine the photodegradation process and resultant micropatterning. Several mechanisms on the photodegradation of organic monolayers have been addressed based on the types of monolayers and wavelengths of the light. Borguet et al. have demonstrated that a combination of UV photons (λ ) 254 nm with 3% of λ ) 183 nm) and oxygen is necessary for alkylsiloxane monolayer degradation.14,15 The dissociation of the alkyl monolayer proceeds effectively in the presence of oxygen, particularly at a wavelength of less than 175 nm.16 The photodegradation process initiates from the terminal part (methyl) of the monolayer and gradually approaches the headgroup.11,15 In contrast, a thiolate monolayer undergoes direct photodissociation of the Au-S linkage by either ozone or UV irradiation.17 The Uosaki group has reported that a covalently attached alkyl monolayer on a silicon (Si-C) surface dissociates by UV light (254 and 185 nm) through two different mechanisms based on the presence and absence of oxygen.18 In the presence of oxygen, the alkyl monolayer dissociates by UVgenerated ozone, and under an argon atmosphere photodegra(10) (a) Sugimura, H.; Ushiyama, K.; Hozumi, A.; Takai, O. Langmuir 2000, 16, 885–888. (b) Sugimura, H.; Sano, H.; Lee, K.-H.; Murase, K. Jpn. J. Appl. Phys. 2006, 45, 5456–5460. (c) Sugimura, H.; Nakagiri, N. Appl. Phys. A: Mater. Sci. Process. 1998, 66, S427-S430. (11) Sugimura, H.; Saito, N.; Maeda, N.; Ikeda, I. B.; Ishida, Y.; Hayashi, K.; Hong, L.; Takai, O. Nanotechnology 2004, 15, S69-S75. (12) Sun, S.; Leggett, G. J. Nano Lett. 2007, 7, 3753–3758. (13) (a) Zhou, C.; Walker, A. V. Langmuir 2007, 23, 8876–8881. (b) Huang, J.; Dahlgren, D. A.; Hemminger, J. C. Langmuir 1994, 10, 626–628. (14) Ye, T.; Wynn, D.; Dudek, R.; Borguet, E. Langmuir 2001, 17, 4497– 4500. (15) Ye, T.; McArthur, E. A.; Borguet, E. J. Phys. Chem. B 2005, 109, 9927– 9938. (16) Ronald, R. P.; Bolle, M.; Anderson, R. W. Chem. Mater. 2001, 13, 2493– 2500. (17) (a) Zhang, Y.; Terrill, R. H.; Tanzer, T. A.; Bohn, P. W. J. Am. Chem. Soc. 1998, 120, 2654–2655. (b) Norrod, K. L.; Rowlen, K. L. J. Am. Chem. Soc. 1998, 120, 2656–2657. (c) Brewer, N. J.; Rawsterne, R. E.; Kothari, S.; Leggett, G. J. J. Am. Chem. Soc. 2001, 123, 4089–4090. (18) (a) Uosaki, K.; Quayum, M. E.; Nihonyanagi, S.; Kondo, T. Langmuir 2004, 20, 1207–1212. (b) Takakusagi, S.; Uosaki, K. Jpn. J. Appl. Phys. 2006, 45, 8961–8966.
10.1021/la8021613 CCC: $40.75 2008 American Chemical Society Published on Web 09/27/2008
12078 Langmuir, Vol. 24, No. 20, 2008
Khatri et al.
Table 1. Physical, Chemical, and Morphological Characterization of the HD Monolayer Assembled on a Hydrogen-Terminated Silicon Surface through Covalent Bonding (Si-C) chemical quantification (XPS) [%]
sample
water contact angle [°]
monolayer thickness [nm]
rms roughness (over 2 × 2 µm2) [nm]
O 1s
C 1s
Si 2p
HD monolayer
110 ( 1
2.07 ( 0.02
0.03 ( 0.02
5.8
33.9
60.3
dation proceeds via cleavage of the Si-C bonds by photogenerated electrons and holes on the silicon surface.18a Despite a lot of studies on photolithography, vacuum ultraviolet (VUV; λ ) 172 nm) irradiation at controlled vacuum pressure with monitoring of oxidant precursors and the reduction of pattern size to a submicron scale still remain unexplored. In the present study, we report the fabrication of a submicron pattern by VUV photodegradation of the hexadecyl (HD) monolayer under controlled vacuum pressure. The compact and conformationally ordered structure of a HD monolayer attached to a silicon surface through a Si-C covalent linkage works as an excellent ultrathin photoresist film and provides a high reproducibility in submicron pattern fabrication. Here, we demonstrate the VUV patterning of a HD monolayer through a chromium-coated quartz photomask. The concentration of oxidant precursors (oxygen and water molecules) and the diffusion of oxygen-derived active species such as O(1D), O(3P), O3, and OH in the masked regions were controlled by changing the vacuum pressure of the VUV chamber. The roughness of the contact surfaces of the photomask and HD monolayer plays a vital role in the diffusion of these oxidants toward the masked regions. The diffusion of oxygen-derived active species increases with increasing the vacuum pressure, resulting in damage to the HD monolayer under the masked area. The pattern width broadens with increasing the vacuum pressure. Gold nanoparticles (AuNPs) were immobilized on the VUV patterned features to probe the dimensions of the chemically active pattern. This versatile patterning approach can be used to fabricate submicron patterns of nanoparticles, nano-objects, and biopolymers on different materials for electronic, optical, and sensor applications at the large scale.
Experimental Section A. Hexadecyl Monolayer Formation. Phosphorus-doped, n-type silicon (111) was used as a substrate to synthesize the hexadecyl monolayer using hexadecene (HD; >99%) as a monolayer precursor. First, the silicon samples were separately sonicated with ethanol and ultrapure water (UPW) for 15 min each. Subsequently, the samples were kept in a VUV treatment chamber for 20 min to burn off carbonaceous contaminants. A hydrogen-terminated silicon surface (Si-H) is formed by treating these samples with 5% HF at room temperature for 300 s in a light shield and then with 40% NH4F at 80 °C for 30 s. The HD monolayer on a hydrogen-terminated silicon surface was prepared by visible light activation with continuous flow of nitrogen in a monolayer precursor.19,20 The visible light from a xenon lamp (high power xenon, MAX-1000, Asahi Spectra Co. Ltd.) was focused on the silicon surface using a collimating lens and passed through a short wavelength cut filter (λ > 422 nm, LU422, Asahi Spectra Co. Ltd.). The measured light intensity at the sample was 290 mW cm-2. After the 8 h visible light exposure, the silicon sample was sonicated with n-hexane for 10 min to remove any excess physisorbed monolayer precursor. The HD monolayer was characterized by water contact angle, X-ray photoelectron spectroscopy (XPS), and ellipsometry, as summarized in Table 1. The water contact angle on the HD-modified silicon surface was found (19) Sun, Q.-Y.; De-Smet, L. C. P. M.; Lagen, B. V.; Giesbers, M.; Thune, P. C.; Engelenburg, J. V.; De-Wolf, F. A.; Zuilhof, H.; Sudholter, E. J. R. J. Am. Chem. Soc. 2005, 127, 2514–2523. (20) Sano, H.; Maeda, H.; Matsuoka, S.; Lee, K.-H.; Murase, K.; Sugimura, H. Jpn. J. Appl. Phys. 2008, 47, 5659–5664.
to be 110 ( 1°, suggesting the formation of a conformationally ordered methyl-terminated monolayer.20,21 The uniform structure of the HD monolayer was confirmed by the recurrence of atomically flat terraces after the monolayer formation on the hydrogen-terminated silicon surface. B. VUV Patterning. The VUV patterning of the HD monolayer was performed using an excimer lamp (Ushio Inc., UER20-172V, invariable light intensity at the lamp window ) 10 mW cm-2) as the VUV (λ ) 172 nm) source. The samples were irradiated with VUV light through a photomask contacting on the sample surfaces. The distance from the lamp window to the surface of the quartz plate was maintained as 5 mm during VUV patterning. The photomask was fabricated from 100 nm thick chromium patterns on a 2 mm thick quartz plate (93% transparency at 172 nm). All patterning experiments were performed in a photolithography chamber at a monitored vacuum pressure ranging from 10 to 105 Pa under atmospheric conditions. C. Gold Nanoparticles Immobilization. To examine the chemically active pattern width, an aminopropyltriethoxysilane (APS) monolayer was deposited on the VUV patterned samples. The presence of oxide/hydroxide moieties attracts the APS molecules on the selected (patterned) sites. Subsequently, the sample was exposed to citrate-functionalized gold nanoparticles (AuNPs, φ ) 20 ( 3 nm). The deposition of AuNPs on the predefined APS monolayer originates from the electrostatic attraction between the citrate coating of the AuNPs and the amino groups of the APS monolayer.22 D. Instrumentation. The surface features of the VUV patterned HD monolayers were scanned using an atomic force microscope (AFM, SII Nanotechnology SPA-300HV + SPI-3800N) at ambient conditions. All topographic and lateral force images were taken in the contact mode using rhodium-coated silicon cantilevers (tip radius: >30 nm; spring constant: 1.6 N/m). A JEOL JSM-7400F field emission scanning electron microscope (FESEM) was used to examine the AuNPs arrays. X-ray photoelectron spectroscopy (XPS, Kratos Analytical Ltd., ESCA 3400) measurements were performed to monitor the chemical changes of the HD monolayer caused by VUV exposure. All XPS measurements were executed using a Mg KR line as an X-ray source.
Results and Discussion For this study, we designed a photomask with 500 nm wide apertures separated by a 4 µm wide chromium coating on a 2 mm thick quartz. The roughness of the chromium coating was measured by AFM and was found to be 0.4 ( 0.05 nm (rms roughness) over a 2 × 2 µm2 area. To achieve the maximum possible contact between the photomask and the HD-functionalized silicon surface (rms roughness: 0.03 nm over a 2 × 2 µm2 area), the accessory was attached by screws with modest pressure. Although both contacting surfaces were very smooth, the nanoscale asperities on the photomask (over the chromium-coated regions) did not permit close packing with the HD monolayer and provided nano- and subnanoscopic channels as illustrated in Figure 1a. Figure 1b and c, which shows 3D topographic images of the HD monolayer and the chromium coating on the photomask, respectively, demonstrates the possibilities of nanoscale channels. The apertures (500 nm wide linear features) on the photomask are distinguished by a 100 nm thick chromium coating on a quartz plate, and these regions function as localized (21) Nihonyanagi, S.; Miyamoto, D.; Idojiri, S.; Uosaki, K. J. Am. Chem. Soc. 2004, 126, 7034–7040. (22) Khatri, O. P.; Murase, K.; Sugimura, H. Langmuir 2008, 24, 3787–3793.
Regulation of Pattern Dimension by Vacuum Pressure
Langmuir, Vol. 24, No. 20, 2008 12079
Figure 1. (a) Schematic illustration of VUV patterning of a HD monolayer. VUV photons are irradiated through a photomask to photodissociate the HD monolayer. Three-dimensional topographic images of a (b) HD monolayer assembled on silicon surface and (c) chromium film coated on quartz to construct the apertures on the photomask. Contact between the HD monolayer and the photomask was determined by the roughness of these surfaces.
factories to generate the oxidants (oxygen active species) under VUV exposure. Oxygen and water molecules in these areas experience photodissociation at λ ) 172 nm and form strong oxidants such as O(1D), O(3P), O3, and OH as follows:16,23
O2 + hV f O(1D) + O(3P)
(1)
O2 + O + hV f O3
(2)
H2O + hV f OH + H
(3)
Electronically excited atomic oxygen [O(1D)] is considered to have strong oxidative reactivity toward alkyl monolayers and acts as a major component to degrade the monolayer.24 The methyl group (terminal part) of the HD monolayer is believed to be oxidized first by these strong oxidants and turn to oxygen containing polar functionalized groups.11,15 Further scission of the alkyl chains progressed by two possible mechanisms: (a) The VUV generated oxidants gradually photodissociate the alkyl chains until complete degradation, and, simultaneously, (b) the probability of direct absorption of VUV photons by oxygen containing polar functionalized groups results in photoetching of the alkyl chains.25 The presence of nanoscopic channels (23) Finlayson-Pitts, B. J.; Pitts, J. N. Atmospheric Chemistry: Fundamentals and Experimentals Techniques; John Wiley & Sons: 1986; Chapter 3, p 139. (24) Paz, Y.; Trakhtenberg, S.; Naaman, R. J. Am. Chem. Soc. 1994, 116, 10344–10345. (25) Truica-Marasescu, F.-E.; Wertheimer, M. R. Macromol. Chem. Phys. 2005, 206, 744–757.
between the photomask and sample surfaces facilitate the diffusion of these oxidants toward the masked area. Such diffused oxidants can damage the alkyl chains under the masked area through mechanism (a). The photopatterning of the HD monolayer was executed by irradiating the sample surface with VUV light under controlled vacuum pressure. Figure 2 shows the lateral force and corresponding topographic images of the HD monolayer after VUV irradiation for 10 min at a vacuum pressure of 103 Pa. The bright and dark linear features in the lateral force image correspond to the irradiated and masked regions, respectively. The pristine HD monolayer under the masked area shows low friction, which is associated with the closely packed structure of the hexadecyl molecules with their highly oriented methyl groups at the terminal. Upon VUV irradiation, selected regions of the monolayer undergo photodissociation and form the hydrophilic oxide/hydroxide layer, which shows high friction with a hydrophilic tip.11 The topographic image (Figure 2b) shows a 1.2 ( 0.2 nm height difference between the VUV patterned and the masked regions, which is less than the height of the HD monolayer (2 nm), although the monolayer is completely degraded in the VUV exposed regions. This could be due to (a) the twisting of the AFM probe induced by the lateral force differences when the probe travels from the hydrophobic (pristine monolayer) to the hydrophilic (VUV patterned) regions, distorting the AFM signal,26 and (b) (26) Lee, B.-W.; Clark, N. A. Langmuir 1998, 14, 5495–5501.
12080 Langmuir, Vol. 24, No. 20, 2008
Khatri et al.
Figure 2. (a) Lateral force and (b) corresponding topographic images of a VUV patterned HD monolayer. High friction (bright linear features) and low height (dark linear features) in the LFM and topographic images, respectively, correspond to the VUV patterned features. VUV patterning of the HD monolayer was performed at a vacuum pressure of 103 Pa. VUV exposure time: 10 min.
Figure 3. Lateral force image (5 × 5 µm2) of a VUV patterned HD monolayer. Bright and dark features correspond to the patterned and pristine HD monolayers, respectively. The linear profile over the LFM image shows the gradual decrease of friction force with increasing distance from the core of the patterned line. The core of the pattern feature with maximum friction force, illustrated as (a), is equal to the width of the aperture printed on a photomask. The fwhm of pattern line, illustrated as (b), shows high friction regions. VUV patterning of the HD monolayer was performed at a vacuum pressure of 103 Pa. VUV exposure time: 10 min.
the development of a thin layer of silicon oxide estimated to be ∼0.2 nm thick.10 The width of the pattern lines formed on the HD-functionalized silicon surface was observed to be larger than the size of the aperture printed on the photomask. This indicates the involvement of the masked area in the photodegradation process to some extent, resulting in the broadening of the pattern features. To gain a better understanding, we scanned a friction image of a single pattern line, as shown in Figure 3, and drew a line profile across both the patterned and masked features. The core of the pattern feature [illustrated as (a)] is associated with the maximum friction, as shown in the friction profile; the observed width of this core is ∼500 nm, which is equal to the size of the photomask aperture. Instead of the distinct boundary between the patterned and masked areas, the lateral force microscopy (LFM) signal decreases gradually as it departs from the core of the pattern feature, revealing the partial dissociation of the HD monolayer near the aperture edge under the masked area. This broadening increased at high vacuum pressure (105
Pa) with further involvement of the masked area. At a vacuum pressure of 103 Pa, the full width at half-maximum (fwhm) of the pattern lines was estimated to be 1300 ( 50 nm, which widened to 1750 ( 100 at 105 Pa using the same size of apertures (500 nm) as printed on the photomask (detail has been given in the Supporting Information). The photochemical modification of the HD monolayer and diffusion of the oxidants in the masked regions were examined in more detail based on XPS studies. To control the concentration and diffusion of the oxidants, we varied the vacuum pressure inside the VUV chamber from 105 to 10 Pa. The rate of photodegradation of the alkyl monolayer depends on the concentration of oxygen and water vapor inside the VUV chamber, which are determined by the vacuum pressure. Therefore, the optimum time to decompose the alkyl monolayer differs with vacuum pressure.27 The photodegradation of HD monolayers was examined as a function of VUV irradiation time at different vacuum pressures using a 2 mm thick quartz plate instead of a
Regulation of Pattern Dimension by Vacuum Pressure
Langmuir, Vol. 24, No. 20, 2008 12081
Figure 4. XPS spectra of the C 1s region for the VUV patterned HD monolayer. Photopatterning was performed at different vacuum pressures: (a) 10 Pa, (b) 103 Pa, and (c) 105 Pa.
2 mm thick photomask. At a low vacuum pressure (10 Pa), the photodegradation of the HD monolayer becomes slow due to the decreased concentration of oxygen-derived active species, but at high vacuum pressure (105 Pa) monolayer degradation is accelerated by increased concentration of oxygen-derived active species. For the vacuum pressures of 10, 102, 103, 104, and 105 Pa, the optimized VUV irradiation time were estimated 15, 12, 10, 6, and 6 min, respectively (detail has been given in the Supporting Information). These results suggest that the efficiency of VUV degradation increased with increasing vacuum pressure, resulting in the decrease in VUV irradiation time (optimized) to photodegrade the HD monolayer. Here, we benchmarked the optimized irradiation time as a function of vacuum pressure and then examined the spreading of oxygen-derived active species under the masked regions. Figure 4 shows a representative set of C 1s signals for VUV irradiated HD monolayers through the photomask at different vacuum pressures: (a) 10 Pa, (b) 103 Pa, and (c) 105 Pa. All C 1s signals were processed with a Gaussian-Lorentzian fitting program and were deconvoluted to examine the different types of carbon. The signal at 285.1 eV, which is associated with aliphatic carbon (C-C/C-H), represents the HD monolayer; the signal at 283.7 eV is associated with carbon covalently linked to the silicon surface thorough C-Si. The signals at 286.6 and 288.8 eV correspond to the oxygen containing functionalities (CHO, CO, and COO, etc).28 Figure 5a shows the percent reduction in C 1s intensity for the VUV irradiated monolayer compared to the pristine HD monolayer as a function of vacuum pressure. The percent reduction in C 1s intensity for the VUV patterned samples was calculated using the following equation:
% reduction of C 1s intensity ) (IC 1s)pristine monolayer - (IC 1s)VUV patterned monolayer × 100 (4) (IC 1s)pristine monolayer At 10 Pa, we observed 12.1% reduction in the intensity of C 1s, which is slightly higher than the aperture area (11.1%) of the photomask, and the percent reduction increased with increasing vacuum pressure. At atmospheric pressure (105 Pa), plenty of (27) Sugimura, H.; Lee, K.-H.; Sano, H.; Toyokawa, R. Colloids Surf., A 2006, 284-285, 561–566. (28) Briggs, D. Surface Analysis of Polymers by XPS and Static SIMS; Cambridge University Press: New York, 1998; p 65.
Figure 5. (a) Change in % reduction of C 1s intensity for the VUV exposed HD monolayer relative to that for the pristine HD monolayer and (b) variation in atomic concentration of C-C/C-H and CO/COO in the VUV exposed HD monolayer as a function of vacuum pressure. The fwhm of the C 1s signal increases with vacuum pressure. The fwhm is determined by the presence of different types of carbon on the VUV patterned surface. Here, all values are extracted from the XPS spectra of C 1s signals.
oxidants are generated owing to the presence of a higher concentration of oxygen and water molecules. These oxidants easily diffuse through the nanoscopic channels and degrade the monolayer under the masked regions. Hence, the maximum reduction of the C 1s intensity was found at 105 Pa. The changes
12082 Langmuir, Vol. 24, No. 20, 2008
Khatri et al.
Figure 6. XPS spectra of the Si 2p region for the VUV patterned and pristine HD monolayers. Photopatterning was performed at different vacuum pressures: (a) 10 Pa, (b) 103 Pa, and (c) 105 Pa.
in the atomic concentrations of C 1s, assigned to aliphatic carbon (C-C/C-H) and oxygenated carbon (CO/COO), are separately shown in Figure 5b as a function of vacuum pressure. It is believed that the terminal methyl groups of the HD monolayer were first photooxidized by VUV generated oxidants and formed intermediate products which contain CHO or COOH groups.11,18 Further attacks of oxidants on these polar groups advance the gradual scission of the hexadecyl chains until complete degradation. At the vacuum pressure of 10 Pa, the formation of oxygenderived active species lessened due to the lack of oxygen and water molecules, and the diffusion of these oxidants in the masked area also became very small. Additionally, there is a possibility of direct interaction of VUV photons with intermediate products, which contain COOH/CHO groups, and this evolves the photoetching of the HD monolayer.25 The direct interaction of VUV photons is limited to the irradiated regions. This suggests that the photodgradation of the HD monolayer at low vacuum pressure proceeds by both (a) the active oxygen species mediated mechanism with minimum spreading of oxidants and (b) the direct absorption of VUV photons by intermediate products. As a result, a very small fraction under the masked area near the aperture edge is partially oxidized and the contribution from the oxygenated groups (CHO/COOH) is very low, as shown in Figure 4a. The active oxygen species mediated mechanism dominates with increasing vacuum pressure. At 103 Pa, a moderate quantity of oxidants is generated, and hence, there is a greater probability for them to diffuse under the masked area. The increased contribution of the oxygenated groups and the greater reduction in C 1s intensity (Figure 5) suggest that the damage to the HD monolayer under the masked area has increased compared to that at 10 Pa. A gradual decrease in the friction force on the pattern line, as illustrated in Figure 3, distinctly demonstrates the partial degradation of the HD monolayer under the masked area. The availability of plenty of the oxidants at atmospheric pressure drastically facilitates the diffusion process. The 55% decrease in the intensity of C 1s and the large contribution of the oxygenated groups reveal that the HD monolayer under the masked area at the aperture edge is probably completely damaged and shows
Figure 7. FESEM images of AuNP arrays immobilized on the VUV patterned HD monolayer. The different chemical templates were fabricated by VUV irradiation at vacuum pressures of (a) 102 Pa, (b) 103 Pa, and (c) 104 Pa. The bright and dark linear features correspond to the AuNPs and HD monolayer, respectively. Self-assembly of the APS monolayer on the VUV exposed sites attracts the AuNPs. The observed widths of the AuNP arrays were more than the aperture size (500 nm) of the photomask, and it broadened with increased vacuum pressure during VUV irradiation.
partial degradation as the distance from the core of the pattern feature increases. An increase in the fwhm (Figure 5b) of the C 1s signal with increasing vacuum pressure illustrates the increasing contribution of CHO/COOH groups or partial degraded molecules. The high friction on VUV irradiated regions and reduction of the C 1s intensity reveals that the HD monolayer has been photodegraded from the VUV exposure regions and formed the silicon oxide/hydroxide, which is corroborated by the appearance of a new Si 2p peak (SiO2) at 103 eV apart from the core Si 2p
Regulation of Pattern Dimension by Vacuum Pressure
Langmuir, Vol. 24, No. 20, 2008 12083
Figure 8. Variation in the chemically active pattern width as a function of vacuum pressure. The widths of the AuNP arrays on the VUV pattered HD monolayers are considered as chemically active pattern widths; each value is the average of four different measurements.
signal (Si0) at 99.5 eV as shown in Figure 6. The intensity of Si 2p signal corresponds to SiO2, which increased with increasing the vacuum pressure. In addition, we examined the broadening of the chemically active pattern features by site-selective assembly of AuNPs. Prior to AuNPs immobilization, the VUV patterned templates were functionalized with an APS monolayer. Deposition of the AuNPs originates from the electrostatic attraction between the citratestabilized AuNPs and the amino groups of the APS monolayer.22,29 Figure 7 shows representative images of the resulting AuNP arrays on three different types of templates patterned at different vacuum pressures ranging from 102 to 104 Pa. The oxides/hydroxides generated by the VUV irradiation at selective sites are expected to have a strong affinity toward the APS molecules and form an amino-terminated monolayer through covalent linkage, while the HD covered surface remains untreated. As observed from the XPS and LFM studies, the HD monolayer near the edge of the aperture under the photomask was partially degraded and formed the -COOH functional groups at the terminal part. Probably, these carboxyl groups also interact with the APS molecules and form the amino-terminated monolayer, which attracts the AuNPs. Consequently, using this method, we can define the chemically active pattern width, which includes not only the complete degraded sites but also the partially degraded regions containing carboxyl and hydroxyl groups as the terminal part of the HD monolayer. Since both areas have an affinity to form the APS monolayer, this results in the immobilization of the AuNPs. The width of the AuNP arrays increases with increased vacuum pressure, as illustrated in Figure 8. At 10 Pa, the observed pattern width (565 ( 15 nm) is slightly higher than the size of the aperture (500 nm) on the photomask. This indicates that a very small fraction of the HD monolayer near the aperture edge experiences partial degradation, resulting in broadening of 30-40 nm from each side of the pattern core. The LFM profile of the VUV pattern line (Figure 3) fabricated at 103 Pa shows that friction decreases gradually as the distance from the core of the (29) Sato, T.; Brown, D.; Johnson, B. F. G. Chem. Commun. 1997, 1007– 1008.
pattern line increases. The total width of the region [illustrated as (b) in Figure 3] with higher friction was observed to be more than 1300 nm, which is larger than the chemically active pattern width (755 ( 15 nm), as shown in the inset of Figure 7b. These results reveal that the HD monolayer near the aperture edge undergoes severe photodegradation and forms the carboxyl and hydroxyl groups, which attract the APS molecules, resulting in the immobilization of the AuNPs in these areas. The HD monolayer next to these regions undergoes little photooxidation and probably forms a polar group such as aldehyde (-CHO). These molecules show high friction but do not contribute to the formation of the APS monolayer. The randomly distributed, small number of AuNPs near the pattern line, as shown in the inset of Figure 7b, reveals that at some places the HD molecules photodissociate to hydroxyl or carboxyl groups, resulting in the immobilization of AuNPs. The number of these particles decreased as the distance from the pattern core increased. This phenomenon is associated with the dimensions of the available nanoscopic channels between the photomask and sample surface. At a higher vacuum pressure (104 and 105 Pa), the monolayer not only under the apertures but also near the apertures was completely degraded (based on XPS and LFM results), resulting in a vast broadening of the pattern width. Figure 7c shows such broadening of the pattern at 104 Pa as well as that more AuNPs are randomly deposited near the pattern lines. At a high vacuum pressure, the availability of a higher concentration of strong oxidants and the ease of their diffusion under the mask area broaden the pattern width.
Conclusions In summary, we report the fabrication of a submicron pattern using VUV light at controlled vacuum pressure. The highly ordered and compact HD monolayer assembled on a silicon surface through Si-C covalent linkage was used as a photoresist thin film. Strong oxidants such as O(1D), O(3P), O3, and OH were generated by photodissociation of atmospheric oxygen and water vapor at 172 nm, degrading the HD monolayer in the VUV exposed area through the intermediate products containing COOH/CHO groups. The nanoscopic channels between the photomask and HD monolayer, and the vacuum pressure inside
12084 Langmuir, Vol. 24, No. 20, 2008
the VUV chamber play important roles in controlling the pattern dimension. The rate of oxidant generation and their diffusion increases with increasing vacuum pressure, and the resulting pattern broadens as revealed by XPS and LFM analysis. At low vacuum pressure, the photodissociation of the HD monolayer is governed by two possible pathways: (a) active oxygen mediated photooxidation with minimum spreading of oxidants and (b) direct absorption of VUV photons by intermediate products. However, at high vacuum pressure, the active oxygen mediated mechanism dominates the photodissociation, which is supported by the broadening of the pattern width. The chemically active pattern widths were examined by site-selective immobilization of citrate-stabilized AuNPs. Both the completely photodegraded VUV exposed regions and the partially degraded HD monolayer near the aperture edge, containing carboxyl and hydroxyl terminal groups, show strong affinity toward APS molecules and, as a result, form the AuNP monolayer on these regions. We observed 565 nm wide AuNP arrays for a chemical template photopatterned at a vacuum pressure of 10 Pa. The pattern was widened to 1030 nm at 104 Pa using the same size aperture (500 nm), due to an
Khatri et al.
increase in the concentration of the oxidants with their unrestrained diffusion under the masked area. This photopatterning approach can be used to pattern nanoparticles, biopolymers, nanofibers, and so forth on a large scale to develop devices for electronic, optical, and sensor applications. Acknowledgment. The authors are grateful to Prof. M. Oyama, International Innovation Center (IIC), Kyoto University, for providing the FESEM facility and acknowledge the support of Mr. A. Ito in utilizing the FESEM facility. They also acknowledge the financial support on Priority Area No. 19049010 “Strong photons-molecules coupling fields (470)” from the Ministry of Education, Culture, Sports, Science and Technology of Japan. O.P.K. acknowledges the Japan Society for the Promotion of Science for a JSPS postdoctoral fellowship. Supporting Information Available: Details on (a) the lateral force line width of the VUV pattern as a function of vacuum pressure with LFM images and (b) the VUV irradiation time. This material is available free of charge via the Internet at http://pubs.acs.org. LA8021613
