pubs.acs.org/Langmuir © 2009 American Chemical Society
Enhanced Uniformity in Arrays of Electroless Plated Spherical Gold Nanoparticles Using Tin Presensitization Phillip Blake,† Wonmi Ahn,‡ and D. Keith Roper*,†,‡ †
‡
Department of Chemical Engineering, University of Arkansas, Bell 3202, Fayetteville, Arkansas 72701 and Department of Materials Science and Engineering, 304 CME, University of Utah, Salt Lake City, Utah 84112 Received October 20, 2009. Revised Manuscript Received December 1, 2009
Gold nanoparticle arrays created with electroless gold plating provide a unique means of transforming nanocylinders usually formed in electron beam lithography to spherical nanoparticles. Alone, electroless gold plating is not selective to the substrate and results in the formation of a gold film on all exposed surfaces of an electron beam patterned sample, including the electron resist. Undesired gold plating occurred near patterned features on the substrate surface, which was reduced by increasing post-spin-coat cure time. When the electron resist is removed, some nanocylinders break off with the gold film, leaving partial cylinders or holes in the patterned elements. By presensitizing the substrate surface with tin, gold cylinders may be selectively deposited to the substrate surface without forming a film on the electron resist. Tin presensitized arrays were produced with 47.1 ( 7.4 nm radius gold nanoparticles with an interparticle distance of 646.0 ( 12.4 nm. Defects from sheared, missing, and redeposited Au particles associated with the resist removal were minimized, resulting in enhanced size and shape uniformity of pillars and arrays. Hollow particles were eliminated, and relative standard deviation in particle size was reduced by 7.4% on average, while elongation was reduced 12.3% when astigmatism was eliminated.
Introduction Ordered arrays of metallic nanoparticles (NPs) combine diffractive coupling with localized surface plasmons, resulting in unique optical properties.1-3 These unique optical properties are being utilized in applications ranging from waveguides,4 nanoantennae,5 and nanometer resolution optical microscopy6 to enhanced fluorescence sensors,7 surface enhanced Raman spectroscopy (SERS),8,9 and biosensors.10-12 The optical characteristics that make these devices possible are highly dependent on NP shape,13 size,14,15 composition,16 interparticle spacing,17,18 and surrounding *To whom correspondence should be addressed. E-mail: dkroper@ uark.edu. (1) Kravets, V. G.; Schedin, F.; Grigorenko, A. N. Phys. Rev. Lett. 2008, 101, 087403. (2) Hicks, E. M.; Zou, S; Schatz, G. C.; Spears, K. G.; Van Duyne, R. P. Nano Lett. 2005, 5, 1065–1070. (3) Chu, Y.; Schonbrun, E.; Yang, T.; Crozier, K. B. Appl. Phys. Lett. 2008, 93, 181108. (4) Radko, I. P.; Evlyukhin, A. B.; Boltasseva, A.; Bozhevolnyi, S. I. Opt. Express 2008, 16, 3924–3930. (5) Dorfm€uller, J.; Vogelgesang, R.; Weitz, R. T.; Rockstuhl, C.; Etrich, C.; Pertsch, T.; Lederer, F.; Kern, K. Nano Lett. 2009, 9, 2372–2377. (6) Smolyaninov, I. I. J. Opt. A: Pure Appl. Opt. 2005, 7, S165–175. (7) Pompa, P. P.; Martiradonna, L.; Della Torre, A.; Della Sala, F.; Manna, L.; De Vittorio, M.; Calabri, F.; Cingolani, R.; Rinaldi, R. Nat. Nanotechnol. 2006, 1, 126–130. (8) Gopinath, A.; Boriskina, S. V.; Reinhard, B. M.; Dal Negro, L. Opt. Express. 2009, 17, 3741–3753. (9) Lee, K.; Irudayaraj, J. J. Phys. Chem. C 2009, 113, 5980–5983. (10) Yu, J. S.; Kim, M.; Kim, S.; Ha, D. H.; Chung, B. H.; Chung, S. J.; Yu, J. S. J. Nanosci. Nanotechnol. 2008, 8, 4548–4552. (11) Shon, Y. S.; Choi, H. Y.; Guerrero, M. S.; Kwon, C. Plasmonics 2009, 4, 95–105. (12) Zhang, J.; Atay, T.; Nurmikko, A. V. Nano Lett. 2009, 9, 519–524. (13) Mirin, N. A.; Halas, N. J. Nano Lett. 2009, 9, 1255–1259. (14) Willets, K. A.; Van Duyne, R. P. Annu. Rev. Phys. Chem. 2007, 58, 267–297. (15) Auguie, B.; Barnes, W. L. Phys. Rev. Lett. 2008, 101, 143902. (16) Ekinci, Y.; Christ, A.; Agio, M.; Martin, O. J. F.; Solak, H. H.; L€offler, J. F. Opt. Express 2008, 16, 13287–13295. (17) Stodolka, J.; Nau, D.; Frommberger, M.; Zanke, C.; Giessen, H.; Quandt, E. Microelectron. Eng. 2005, 78-79, 442–447. (18) Zhao, L.; Kelly, K. L.; Schatz, G. C. J. Phys. Chem. B 2003, 107, 7343–7350. (19) Yang, T.; Crozier, K. B. Opt. Express 2008, 16, 8570–8580.
Langmuir 2010, 26(3), 1533–1538
dielectric environment.2,19 Recent theoretical analyses show that, for spherical arrays of Au NPs, the electromagnetic field enhancement can be up to 107 relative to an isolated NP.20 Fabrication methods for regular NP arrays include dip-pen nanolithography (DPN),21,22 nanosphere lithograhpy (NSL),10,14,23 and electron beam lithography (EBL).1,7,15 Most of these methods rely on thermal evaporation or sputtering to form the metallic NPs. The directional nature of this approach limits NP shapes that can be created. Current fabrication methods form disks, triangles, or cylindrical particles, not spherical particles. Using electroless Au plating on EBL template samples allows the formation of spherical NPs. Additionally, an adhesive layer of chromium is usually required for good surface adhesion, which changes the local dielectric and affects optical response characteristics.24 We recently reported thermal transformation of Au island thin films created by electroless (EL) Au plating to random assemblies of spherical Au NPs on planar silica substrates25 and on the inner walls of rectangular borosilicate glass capillaries.26 Electroless plating of gold uses sequential metal depositions of tin (Sn2þ), silver (Agþ), and gold (Au) with a chemical reducing agent to give strong adhesion of Au particles to the substrate. After thermal annealing in a 250-800 °C oven, EL plated Au thin films transform to thermally stable, spherical NP assemblies with tunable particle size and interparticle spacing. EL plating, however, is not specific to the substrate, and when applied to masked surfaces templated by EBL Au is deposited in an interconnected sheet both on the surface of the electron resist as well as (20) Zou, S.; Schatz, G. C. Chem. Phys. Lett. 2005, 403, 62–67. (21) Zhang, H.; Chung, S.-W.; Mirkin, C. A. Nano Lett. 2003, 3, 43–45. (22) Zhang, H.; Lee, K.-B.; Li, Z.; Mirkin, C. A. Nanotechnology 2003, 14, 1113– 1117. (23) Zheng, Y. B.; Juluri, B. K.; Mao, X.; Walker, T. R.; Huang, T. J. J. Appl. Phys. 2008, 103, 014308. (24) Simsek, E. Plasmonics 2009, 4, 223–230. (25) Ahn, W.; Taylor, B.; Dall’Asen, A. G.; Roper, D. K. Langmuir 2008, 24, 4174–4184. (26) Ahn, W.; Roper, D. K. J. Phys. Chem. C 2008, 112, 12214–12218.
Published on Web 12/09/2009
DOI: 10.1021/la903985m
1533
Letter
Blake et al.
Scheme 1. Fabrication Process for EBL Samples on ITO Glass Substratea
(MIBK)/isopropyl alcohol (IPA), rinsed with IPA, and then immersed in IPA for 1 or 2 min. Development was performed either at room temperature or at 4 °C (cold development) to reduce NP size.27
Creation of Regular Arrays of Spherical Au NPs Using Electroless Au Plating. EL Au plating was performed as a three step metal plating procedure: (1) immersion in a solution of tin (Sn2þ) for 3 min, (2) ammoniacal AgNO3 for 2 min, and (3) sodium gold sulfite (Na3[Au(SO3)2]) for 10 min. Au deposition occurs due to differences in reduction potentials between Ag and Au.25 After Au plating, PMMA was lifted off by immersion in 50 °C acetone for 5 min followed by 50 min of gentle sonic agitation in 50 °C acetone, resulting in regular arrays of Au nanostructures. Presensitized Substrate Preparation. The above procedure was modified to test the effects of presensitization of the substrate. Sn and Ag presensitized substrates were first EL plated with Sn and Sn/Ag prior to spin-coating and stored under nitrogen gas until PMMA coating was performed. Prebake and methanol/ acetone cleaning steps were not performed prior to spin-coating. EBL patterning and development remain unchanged, but EL plating was only performed for the metals which had not already been coated onto the substrate. Liftoff for Sn sensitized samples was performed using a 5 min immersion (no sonication). Fabrication conditions are summarized in Table 1. Scanning electron microscope images were taken with the same ESEM used to create the EBL patterns.
a (I) Regular processing, (II) Sn presensitization, and (III) Sn/Ag presensitization. (a) PMMA spin-coating on substrate. (b) electron beam patterning. (c) Cold development (4 °C) in MIBK/IPA (1:3) followed by pure IPA. (d) Sn sensitization. (e) Ag plating. (f) EL Au plating via galvanic displacement. (g) PMMA liftoff in 50 °C acetone. (h) Au nanostructure pattern remaining after PMMA liftoff. (i) Thermal annealing to transform Au nanostructures to Au nanoparticles.
on the substrate. We have modified the EL plating procedure to include substrate presensitization which makes EL plating substrate specific. Herein we present and characterize this presensitization EL plating method and its usefulness in producing more uniform regular arrays of spherical Au NPs with EBL with respect to NP size and shape.
Experimental Section Scheme 1 illustrates the fabrication process for (I) regular processing, (II) Sn presensitization, and (III) Sn/Ag presensitization. Substrate Preparation. Conductive indium-tin-oxide (ITO)-coated polished glass slides (SPI Supplies, West Chester, PA) were used to improve electron beam lithography (EBL) writing. ITO slides were prepared for EBL by prebaking on a 150 °C hot plate for 1 min, followed by cleaning with methanol and acetone. ITO slides were spin-coated using a CEE 100 spincoater (Brewer Science, Inc. MO) with 950 kamu poly(methylmethacrylate) (PMMA) and post (spin-coat) cured at 150 °C for 0 s, 30 s, 1 min, or 2 min. Rotational speed was increased from 1500, 2500, to 4000 rpm as the PMMA concentration increased from 2, 2.5, to 3%, to produce a PMMA layer with a thickness of ∼130, 220, and 280 nm, respectively. Pattern Creation by Electron Beam Lithography. EBL was performed using a Philips XL 30 environmental scanning electron microscope (ESEM) (FEI, Hillsboro, OR) equipped with a Nanometer Pattern Generation system (JC Nabity Lithography Systems, Bozeman, MT) to create patterns, control electron beam dosage, and control stage movement. The pattern consisted of arrays of wheels and 100.5 100.5 μm2 square arrays of dots with a horizontal and vertical spacing of 670 nm. After beam exposure, samples were developed in a 1:3 solution of methyl isobutyl ketone 1534 DOI: 10.1021/la903985m
Results and Discussion EL Plating on EBL-Templated Substrates (EL-EBL). EL deposited Au is not directionally restricted, as is the case in thermal evaporation, and will form an Au film on all exposed surfaces of the patterned PMMA as well as the ITO substrate, forming Au pillars that are attached to the substrate and a layer of Au on the surface of the PMMA. During liftoff, these pillars break from the top Au layer for successful liftoff. The location where the pillar breaks is uncontrollable and results in some of the pillars being sheared from the surface, leaving only small NPs or rings. Figure 1a shows a Au layer from the PMMA surface that was only partially removed during liftoff. Figure 1b shows typical pillar morphology for non-presensitized substrates. EL plated Au nanostructures that are not spherical can be thermally transformed into spherical NPs stable to a wide range of temperatures and solvent conditions. Figure 2 shows images of Au nanocylinders before (a-c) and after (d-f) thermal annealing. Au nanostructures before thermal annealing were 192.5 ( 6.8 nm in diameter. Thermal annealing caused the nanocylinder structure to fold in on itself to create spherical NPs with a diameter of 156.4 ( 4.9 nm. This ability is not limited to nanocylinder structures, and it has been demonstrated in Au island films.26 Typical Au pattern defects that occur with this method have been highlighted in Figure 2a. These defects include partial Au NPs due to shearing during PMMA removal, entirely removed Au NPs, and redeposition of Au after liftoff. Extraneous Gold Layer Adjacent to Patterned Features in EL-EBL. An extraneous surface layer of Au was observed to form on the substrate around patterned features on the ITO substrate when EL plating EBL patterned samples. Three factors were identified that influence the formation of extraneous Au films in EL-EBL: post (spin-coat) cure time, electron beam dose, and IPA development time. Extraneous Au formation decreased as the post cure time increased from 0 (Figure 3A,C) to 30 s (Figure 3B,D), but it did not appear to be directly affected by the PMMA thickness since both 130 and 220 nm samples showed (27) Hu, W.; Sarveswaran, K.; Lieberman, M.; Bernstein, G. H. J. Vac. Sci. Technol., B 2004, 22, 1711–1716.
Langmuir 2010, 26(3), 1533–1538
Blake et al.
Letter Table 1. Summary of Experimental Conditions for EBL Samples
samplea
presensitized
PMMA thickness (nm)
PMMA conc (%)
spin-coater speed (RPM)
pre-spin-coat bake (min)
post-spin-coat cure (s)
IPA time (min)
A B
-
130
2
1200
1
30
2
C D E F
-
220
2.5
2500
1
30 30
2
G H I J K L
280 3 Sn Sn and Ag a Sample labels correspond to images shown in Figure 3.
4000
1
-
30 180 300 60
1 2
Figure 1. Electroless Au plating on patterned electron beam lithography samples shows Au plating is nonselective and Au plates to all exposed surfaces. (a) Au film from top of PMMA layer peeling away from ITO substrate surface after partial liftoff with Au cylinders still attached. (b) Au cylinder remaining on ITO surface after shearing away Au-coated PMMA layer. Scale bars are 500 and 200 nm, respectively.
similar levels of extraneous Au. When the post cure time was further increased to 3 and 5 min (Figure 3I and J), it became evident that post cure time effects on extraneous Au film formation reach a critical point after which no further reduction is seen. The second factor in extraneous surface Au formation was electron beam dose during EBL. The electron beam dose required to create a pattern is dependent upon a number of factors including PMMA thickness, post cure time, proximity of pattern features in EBL pattern, and development conditions. Table 2 summarizes the required electron beam dose as thickness and post-spin-coat cure time increases. The minimum dose required was defined as the lowest dose for which the dot array pattern was clearly visible in SEM images after development. Minimum dose levels were established by creating 25-30 100.5 100.5 μm2 square arrays of dots at incremental values of dose with the initial dose being below that required for patterning, as determined by previous experiments. Pattern element dimensions were observed to increase as the dose was increased (not shown). As PMMA thickness was increased, the dose required to draw a pattern element also increased for postcured samples (Table 2). Non-postcured samples do not appear to show the same trend, but considering that the electron beam has to interact with more material as the thickness increases suggests that dose requirements should increase with increasing PMMA thickness for postcured and non-postcured samples.28 (28) Olzierski1, A.; Vutova, K.; Mladenov, G.; Raptis, I.; Donchev, T. Supercond. Sci. Technol. 2004, 17, 881–890.
Langmuir 2010, 26(3), 1533–1538
Improved EL-EBL with Tin-Presensitization. We hypothesized that extraneous surface Au and pattern defects seen in EL-EBL samples could be reduced or eliminated by performing segments of EL plating before the substrate was coated with PMMA, as shown in Scheme 1 II and III. Sn presensitization was performed to eliminate the formation of Ag on the PMMA surface as shown in Figure 2. Figure 4 shows SEM images of regular arrays of Au NPs created on a Sn presensitized ITO substrate. Original image magnification was changed from (a) 20 000, (b) 80 000, to (c) 200 000 to show top-down images of regular NP patterns, and from (d) 25 000, (e) 100 000, to (f) 350 000 with a 45° tilt angle to better show Au NP structural morphology. Au NPs with a radius (r) of 47.1 ( 7.4 nm were patterned into regular square arrays with an interparticle distance (d) of 646.0 ( 12.4 nm in Figure 4a-c. Some of the Au NPs were not fully developed, resulting in a deposition of Au traces. However, creation of small Au NPs (e100 nm in diameter) was achieved by tin presensitization of the ITO substrate that allows more selective EL Au plating. Sn presensitization avoided forming thick Au films on the PMMA surface. These Au films increased the difficulty of PMMA removal and led to pattern variations in the remaining NPs due to broken Au pillars or redeposition of Au particles onto the ITO substrate during the PMMA lift-off process. Sn presensitization also resulted in more uniformly filled Au nanostructures, as clearly seen by comparing the tilted SEM images in Figure 4d-f with that of Figure 1b. Thermal treatment is expected to result in smaller and more spherical NPs with smoother surfaces as previously shown DOI: 10.1021/la903985m
1535
Letter
Blake et al.
Figure 2. SEM images of typical nanoparticle structures without presensitization after PMMA liftoff: before (a-c) and after (d-f) thermal annealing. Typical defects in Au deposition associated with this processing method are boxed in (a). Thermal annealing transforms the hollow cylinder structures into spherical nanoparticles. Scale bars for each column represent 1 μm, 500 nm, and 100 nm, respectively. Table 2. Dose Requirements As a Function of Thickness and PostSpin-Coat Cure Time minimum dose (μC/cm2) thickness (nm) 0 cure time 130 5.46 ( 11.38 ( 220a 220 6.94 ( 280 5.70 ( a 1 min IPA development.
Figure 3. Wheel structures created using EBL for three different PMMA thicknesses and four post (spin-coat) cure times (at 150 °C) show the extraneous surface Au is reduced by increasing post cure time. An upper limit exists where post cure time no longer reduces the extraneous Au, as seen in the 180 and 300 s samples. EBL patterning conditions for these samples are located in Table 1. Scale bars are 2 μm.
(Figure 2). Thin background Sn films were observed in higher magnification SEM images such as Figure 4c and f. Presensitization with Sn-followed-by-Ag presensitization was also tested. Ag films appeared to undergo a thermal transformation during the postcure process. This change was observable by a darkening in the substrate color, attributable to creation of a rough surface as the thin Ag film began to thermally transform into Ag NPs. SEM images of the patterned Au NPs are shown for increasing magnification in Figure 4 from (g) 35 000, (h) 50 000, to (i) 200 000. Au NPs formed via Sn/Ag presensitization were 1536 DOI: 10.1021/la903985m
1.48 1.48 1.48 0.18
30 s cure time 3.98 ( 1.48 24.7 ( 1.48 8.42 ( 1.48 14.34 ( 1.48
flatter and less coherent than those formed from Sn presensitization. Also, the substrate surface was rough from the Ag NPs formed during the postcure step (Figure 4i). Comparing substrate surface features from nonsensitized, Sn presensitized, and Sn/Ag presensitized methods shows that presensitization yields more background surface particulates. Figure 2f shows the ITO surface forms into particles ∼15 nm in size when thermally treated. Sn presensitization, as shown in Figure 4c and e, results in an additional Sn layer. Thermal annealing does not remove the additional Sn, producing larger Sn particles than those from ITO without presensitization. Effects of these particulates on optical responses of fabricated arrays are minimal. Sn/Ag presensitization, as shown in Figure 4h and i, clearly forms a Ag film and particulates in the substrate background of the Au NP array. These particles are ∼25 nm in size, which absorb near 355 nm, as confirmed by MiePlot. Circularity of NPs in EL-EBL. The spherical nature of Au NPs formed with regular and Sn presensitized EL-EBL were compared using Matlab (Mathworks Inc., Natick, MA) from SEM images using standard measures of particle diameter, circularity, and elongation.29 Particle diameter was determined by the regionprops function built into Matlab with the formula rffiffiffiffiffiffi 4A D ¼ π
ð1Þ
(29) Almeida-Prieto, S.; Blanco-Mendez, J.; Otero-Espinar, F. J. J. Pharm. Sci. 2004, 93, 621–634.
Langmuir 2010, 26(3), 1533–1538
Blake et al.
Letter
Figure 4. SEM images of regular Au NP arrays on a Sn presensitized substrate prior to annealing (a-f) in order of increasing magnification, from (a) 20 000, (b) 80 000, to (c) 200 000 (0° tilt), and from (d) 25 000, (e) 100 000, to (f) 350 000 (45° tilt). Au NP arrays on a Snfollowed-by-Ag presensitized substrate (g-i) in order of increasing magnification from (g) 35 000, (h) 50 000, to (i) 200 000, showing flatter, less coherent Au NPs and small Ag particles in the background. Scale bars in each column represent 1 μm, 200 nm, and 100 nm, respectively.
where D is the circle-equivalent particle diameter and A is the NP area. Circularity is the ratio 4πA P2
ð2Þ
where P is the NP perimeter. Circularity approaches 1 for circular objects. Elongation is calculated by 1 - aspect ratio f 1 -
minor axis length major axis length
ð3Þ
Elongation approaches 0 for circular objects. The value of elongation can be transformed to approach 1 for circular objects by subtracting elongation measured in eq 3 from unity to obtain “1 - elongation”. The value of 1 - elongation will be reported herein. Figure 5 compares particle size, circularity, and 1 - elongation of three regular EL-EBL arrays (samples 1-3) and two Sn presensitized arrays (samples 4, 5). Sample 1 represents a regular sample developed (MIBK/IPA (1:3) and IPA) at room temperature, while all other samples were developed at 4 °C. Standard deviations for each sample were based on analysis of N = 30, 90, 33, 61, and 50 particles in SEM images of the sample. Respective electron resist thicknesses for the samples were 280, 280, 220, 280, and 280 nm. The left-hand frame of Figure 5 shows that circleequivalent particle diameters were calculated to be 207.9 ( 16.6, 76.7 ( 13.3, 97.2 ( 21.6, 93.7 ( 8.3, and 107.5 ( 8.6 nm. Comparing sample 1 with sample 2, and later samples, confirmed Langmuir 2010, 26(3), 1533–1538
Figure 5. Particle shape analysis for regular (samples 1-3) and Sn presensitized (samples 4-5) EL-EBL fabricated arrays. Values of circularity and 1 - elongation are unity (1.0) for circular particles. Sample distributions are based on (from left to right) N = 30, 90, 130, 33, 61, and 50 particles. Circle-equivalent particle diameter variations are smaller for Sn presensitized samples than for regular samples.
that cold development reduces NP size.27 Sn presensitized samples had more narrow particle diameter distributions than those of regular samples. Circularity values determined for the respective samples were 0.86 ( 0.12, 1.08 ( 0.23, 0.91 ( 0.11, 0.79 ( 0.09, and 0.92 ( 0.14. While the circularity values differ between Sn presensitized samples, standard deviation values were comparable to those of the regular samples values. The differences in circularity may be attributed to the small size of the particles in SEM images, since DOI: 10.1021/la903985m
1537
Letter
fewer pixels per particle result in larger changes in circularity for similar shape perturbations. Smaller NPs will be more sensitive to shape defects than their larger counterparts in terms of circularity and elongation. Sample 1 was included because it represented the best particle morphology for regular EL-EBL samples, but the dose was higher for sample 1 (∼20 μC/cm2) than for Sn sensitized sample 4 (12.86 μC/cm2). This variation accounts for some of the size difference between samples 1 and 2. Calculated values of 1 - elongation ranged from 0.76 ( 0.06, 0.77 ( 0.11, 0.77 ( 0.09, 0.69 ( 0.09, to 0.89 ( 0.07. The average 1 - elongation value for Sn presensitized samples was similar to elongation values for regular samples, with similar standard deviations for both. The 1 - elongation value for Sn presensitized sample 5 was closer to ideal (unity), with a lower standard deviation than sample 4 due to lack of astigmatism during EBL patterning. Close inspection of the SEM images (not shown) for the Sn presensitized sample 4 showed the elongation was directional; that is, the major axes were nearly all vertically aligned. This directionality indicates the deviation in circularity was likely caused by insufficient elimination of astigmatism when writing the pattern via EBL. Sn presensitized samples have smaller particle size distributions than those of regular EL-EBL samples. Relative standard deviations for Sn presensitized samples were 7.4% smaller than those in regular samples while maintaining similar circularity, in addition to forming more coherent particles before thermal annealing (spheroids versus hollow cylinders). When astigmatism was eliminated, Sn presensitized samples
1538 DOI: 10.1021/la903985m
Blake et al.
exhibited improved 1 - elongation (0.89 ( 0.07) relative to regular samples (average 0.77 ( 0.09).
Conclusions EL Au plating has been successfully integrated with EBL to create regular arrays of spheroid Au NPs for the first time. Extraneous Au plating on the substrate adjacent to patterned features was reduced by increasing the post cure time. Performing Sn plating on the substrate prior to coating with electron resist reduced the number of sheared, missing, and redeposited Au particles compared to regular EL-EBL, and eliminated hollow particle morphology. Sn presensitized arrays were produced with 47.1 ( 7.4 nm radius Au NPs with an interparticle distance of 646.0 ( 12.4 nm. Heat stability of the EL plating process allows these Au NP arrays to be transformed from nanocylinders, typically formed in EBL, to spherical NPs. The relative standard deviation in particle size was reduced by 7.4% on average, while elongation was reduced 12.3% when astigmatism was eliminated. Acknowledgment. This work was supported in part by NSF CMMI-0909749 and University of Arkansas Foundation. The authors would like to acknowledge helpful discussions with Dr. Morgan Ware and Dr. John Shultz; helpful critique by reviewers; instrumentation assistance from Alan Toland with the Electron Optics Facility at the University of Arkansas; and use of sample preparation facilities at the Arkansas Institute for Nanoscale Materials Science and Engineering at the University of Arkansas.
Langmuir 2010, 26(3), 1533–1538
