Journal of Colloid and Interface Science 298 (2006) 713–719www.elsevier.com/locate/jcis

Direct-write fabrication of colloidal photonic crystal microarraysby ink-jet printing

Jungho Park a, Jooho Moon a,∗, Hyunjung Shin b, Dake Wang c, Minseo Park c

a Department of Materials Science and Engineering, Yonsei University, Seoul 120-749, South Koreab School of Advanced Materials Engineering, Kookmin University, Seoul 136-702, South Korea

c Department of Physics, Auburn University, Auburn, AL 36849, USA

Received 10 October 2005; accepted 16 January 2006

Available online 3 February 2006

Abstract

An array of the colloidal photonic crystals was directly fabricated using an ink-jet printing. The colloidal ink droplets containing the monodis-persed polystyrene latex particles were selectively deposited on a hydrophobic surface. Solvent evaporation from each ink droplet leads to aformation of microdome-shaped colloidal assembles of close-packed structures. Microspectroscopic analysis has confirmed that the individualassembly serves as a photonic crystal and its optical properties can be correlated with the microstructural features. Unlike other techniques ofpatterned growth of colloidal photonic crystal, the substrate does not need to be patterned first and no template is needed in the direct writing bythe ink-jet printing. Using our strategy, we have rapidly produced the colloidal photonic crystal microarrays composed of different-sized spheresaddressably patterned on the same substrate.© 2006 Elsevier Inc. All rights reserved.

Keywords: Self-assembly; Colloid crystal; Ink-jet printing

1. Introduction

The ability of monodispersed colloidal spheres to self-assemble into highly ordered structures has promised a varietyof technological applications of the ordered colloidal spheres.These structures can serve as building blocks such as tem-plates for porous materials that can act as catalysts in chemicaland biological processes, masks for nonlithographic patterning,and biochemical sensors [1]. Ordered arrays of colloidal par-ticles are particularly interesting as photonic band gap (PBG)materials which can be utilized for manipulating light in opti-cal and optoelectronic devices [2]. However, this promise willnot be fulfilled unless methods of positioning and addressingcrystalline object individually with specific geometries can bedeveloped.

Colloidal particles can be assembled into close-packed ar-rays by many different techniques. The most common ones are

* Corresponding author. Fax: +82 2 365 5882.E-mail address: jmoon@yonsei.ac.kr (J. Moon).

0021-9797/$ – see front matter © 2006 Elsevier Inc. All rights reserved.doi:10.1016/j.jcis.2006.01.031

self-assembly under the influence of external fields—gravita-tional sedimentation [3], vertical deposition by evaporation [4],and the microfluidic cell method [5]. On the other hand, onlya few approaches to control the specific placement of colloidson surfaces by self-assembly have been reported. These meth-ods involve the assembly of colloids onto chemically patternedsurfaces by using (1) electrostatic forces [6], (2) differencesin wettability [7], or (3) ability of colloids to assemble underspatial confinement on a topographically structured substratewith holes, grooves, or microchannels [8]. However, these ap-proaches all suffer from some deficiency: they do not provideprecise positioning, require specific susceptibilities, have lowthroughput, and/or are not scalable. In this communication, wehave demonstrated “directly writing” a photonic crystal arrayby self-assembling colloidal droplets produced using an ink-jetprinting.

Direct-write fabrication allows us to design and rapidly fab-ricate materials in complex structures and shapes without theneed for expensive tooling, dies, or lithographic masks. Ink-jetprinting is the one of direct-write methods that involves an add-on process of materials. Ink-jet printing has recently emerged,

714 J. Park et al. / Journal of Colloid and Interface Science 298 (2006) 713–719

beyond image transfer capability, as an attractive technique forflexible electronics and displays, controlled release drug de-livery devices in pharmaceuticals, and refractive microlensesmade of hybrid organic–inorganic materials [9]. It relies onthe selective deposition of colloid-, biomolecule-, or organic-based inks on arbitrary surface in the form of droplets. A majorchallenge in applying ink-jet processes for directly writing ma-terials is formulating suitable inks. For example, colloidal inksmust be dilute fluids having low viscosity. The inks are capableof flowing through the print-head nozzle without clogging. Inkchemistry and formulations not only dictate the quality of theprinted image and minimum feature size, but they also deter-mine the drop ejection characteristics and the reliability of dropformation.

Recently, we demonstrated a rapid formation of macroscop-ically patterned nanostructures of colloidal particles by ink-jetprinting. This approach was similar to the method of printingvisual information onto paper [10,11]. Our approach involvedthe use of submicrometer-scale, self-assembling colloidal inkconsisted of monodispersed silica microspheres. Ink-jet print-ing method was capable of placing ink droplets with a nearlyuniform size on a selective area of a solid surface. Silica par-ticles contained in each ink droplet underwent self-assemblyupon evaporation, producing colloidal aggregates with an in-ternally ordered structure. By varying interfacial properties ofthe surfaces with which the ink interacts, it was demonstratedthat the shape and size of the self-assembled colloidal aggre-gates can be controlled. Microspectroscopic technique allowedus to confirm that these aggregates serve as a colloidal photoniccrystal individually and to address theoretical aspects of theiroptical features. In this study, we have focused on the fabri-cation of the photonic crystal microarrays composed of dome-shaped colloidal aggregates by ink-jet printing. Particle sizedependency on structural and optical properties of the colloidalphotonic crystals was investigated. Using this direct-write strat-egy, we have also demonstrated the capability of rapid ad-dressable patterning, where colloidal crystal arrays comprisedof different-sized spheres can be patterned on the same sub-strate.

2. Experimental

2.1. Ink preparation

Colloidal ink was prepared using a mixed solvent of 70weight % water and 30 weight % formamide. Monodispersepolystyrene microspheres with different size were obtainedfrom Bangs Laboratories, Inc. The mean particle diameterswere 190, 210, and 270 nm with a polydispersity less than3%. The pH of the as-received PS suspensions was measuredto ∼5.5. The zeta (ζ ) potential of the PS latex particles was inthe range of −34 to −32 mV, as examined by Brookhaven ZetaPlus. The PS particles were dispersed in the premixed solvent ata solid loading of 1.6 volume %. After stirring for 60 min, octylalcohol was added to control surface tension. The formulatedink was treated ultrasonically (Model 550 Sonic Dismembrator,Fisher Scientific) followed by filtration through a 5-µm nylon

mesh. The ink has the viscosity of ∼7 mPa s at shear rate of10 s−1, as measured by a cone and plate viscometer (DV-III+,Brookfield Engineering).

2.2. Substrate preparation

P-type Si wafer of [100] orientation (LG Siltron, thickness505–545 µm, resistivity 1–30 � cm) was used as a substrate.Silicon wafer was modified with a self-assembled monolayer(SAM) of octadecyltrichlorosilane (OTS) (Aldrich ChemicalCo.) to produce a hydrophobic surface. OTS-SAM was grownby immersing the Si substrate cleaned in a piranha solution inanhydrous toluene solution containing 0.1 vol% OTS for 2 h un-der ambient condition. The wafer was rinsed with toluene anddeionized water for removal of physically adsorbed OTS, fol-lowed by drying with a stream of N2. Surface roughness of theOTS-SAM coated substrates was measured to be 4 nm (RMS)by atomic force microscopy (AFM, SPA 400, Seiko). The hy-drophobic surface gave a good reproducibility on the contactangle measured for colloidal droplets; the mean value of θ is105◦ and the standard deviation is ±2◦.

2.3. Ink-jet printing of colloidal crystal array

The colloidal ink containing monodisperse polystyrene la-tex spheres was printed by an ink-jet printer. The printer set-up consisted of a drop-on-demand (DOD) piezoelectric ink-jetnozzle manufactured from Microfab Technologies, Inc. (Plano,TX) with a 30-µm orifice. The print head was mounted onto acomputer-controlled three-axis gantry system capable of move-ment accuracy of ±5 µm. The gap between the nozzle and thesurfaces was maintained at 0.5 mm during printing at 25 ◦C and40% relative humidity. The uniform ejection of the droplets wasperformed by applying 35 V impulse lasting 30 µs at a fre-quency of 1 kHz. CCD camera equipped with a strobe-LEDlight was employed to watch individual droplet by which thephysical properties of the droplets were analyzed.

2.4. Characterization

The microstructure of the individual colloidal assemblesafter drying was investigated by scanning electron micro-scope (JEOL, JSM6500F), confocal laser scanning microscopy(LSM 5 Pascal, Carl Zeiss), and atomic force microscope.Microreflectance spectroscopy was carried out using a Zeissmicroscope coupled with a light source and a spectrometer.The sample was irradiated using a UV/visible light source(DT-1000 Ocean Optics). The light from the broadband lightsource was focused onto a single droplet of colloidal aggre-gate via a microscope objective lens. The microreflectancespectrum was collected by Jobin-Yvon’s TRIAX 550 spectrom-eter equipped with a thermoelectrically cooled charge coupleddevice (CCD) detector. The spectral position and the half-width-at-half-maximum (FWHM) of the reflectance peaks weredetermined by fitting the spectrum with a Lorentzian func-tion.

J. Park et al. / Journal of Colloid and Interface Science 298 (2006) 713–719 715

3. Results and discussion

The colloidal ink containing negatively charged monodis-persed polystyrene (PS) latex spheres was printed onto a hy-drophobic surface of a self-assembled monolayer (SAM) of oc-tadecyltrichlorosilane (OTS) on silicon wafer (denoted as OTS-SAM/Si) by a DOD-type ink-jet printer. The 1.6 volume %(vol%) of PS latex spheres (190–270 nm in diameter) in anaqueous suspension were used. The composition of the ink wascarefully controlled to ensure the evaporation-driven colloidalcrystallization with a high degree of crystal quality. The uni-form droplets were ejected from a nozzle of a 30-µm orificesize by applying 35 V impulse that lasts 30 µs at a frequency of1 kHz. Jetting conditions such as the droplet volume and sizeprior to impact for each colloidal ink with different sizes aresummarized in Table 1. We have produced a 10 × 10 array ofphotonic crystals as schematically depicted in Fig. 1. Such amicroarray was rapidly deposited with precise position controlwithin 5 min, followed by drying at 25 ◦C for 5 h under ambientwith a 40% relative humidity.

Fig. 2 shows the array of photonic crystals that was preparedon the OTS-SAM/Si using the ink-jet printing method. Eachdome-shaped colloidal aggregate was produced from a singledroplet (∼60 pL) of the colloidal ink. All the colloidal crys-tals printed on the substrates displayed excellent homogeneityin shape and size. The number of the latex particles contained ineach is approximately 10–25 × 104. The droplet placed on hy-drophobic surface played as templates where the colloidal par-ticles were self-assembled to form colloidal assemblies. Sincethe impinged ink droplet usually reaches an equilibrium statewithin 50–2000 ms, it is reasonable to assume that no signifi-cant solvent evaporation occurs in an ambient atmosphere forsuch a short period. Configuration of the droplet templates is

then determined solely by the wettability of the ink on the sub-strate. The droplet placed on the hydrophobic OTS-SAM/Siretains a hemispherical shape (neglecting gravitational effect)with a high contact angle (∼60◦). The presence of thick liquidlayer at the contact line of the droplet permits slow evapora-tion to occur uniformly throughout the liquid/gas interface. Insuch a case, the contact line rather retracts without pinning asthe droplet shrinks while maintaining a hemispherical shape.The particles suspended in the evaporating droplet are graduallyconcentrated as the solvent slowly evaporates. When the parti-cle concentration reaches a certain critical value at which theirmobility is significantly reduced with the decrease in dropletvolume, the particles begin to crystallize in a close vicinity tothe external surface, and the incoming particles become crys-tallized when they contact the ordered arrays of the particles,resulting in the growth of a colloidal crystal [11]. Such anevaporation-induced assembly process yielded smooth assem-blies with ∼22 µm in diameter and ∼6 µm in height. Macro-scopic dimensions of each microdome as a function of the PSparticle size are summarized in Table 1.

Low magnification optical microscopy shows brightly col-ored reflecting patches on the surface of these dome-shapedphotonic crystals as shown in Fig. 3. The curvature of the mi-crodome results in different orientation of the diffracting planesof the aggregates being presented to the observer. The ob-served colors depend on the size of the latex particles used;blue, bluish-green, and red for 190, 210, and 270 nm mi-crospheres, respectively. The colors arise from light diffraction(opalescence) from the colloidal arrays, indicating a long-rangeordering of the latex particles. The long-range order of thecolloidal photonic crystal can be visualized directly by scan-ning electron microscopy (SEM) and atomic force microscopy(AFM), as shown in Fig. 4. The PS microspheres were arranged

Table 1Physical characteristics of the droplet prior to impact and the resulting colloidal assemblies with different sizes after drying

Colloidal diameter used in an ink 190 nm 210 nm 270 nm

Diameter of the droplet prior to impact (µm) 49.1 ± 0.3 48.5 ± 0.4 48.2 ± 0.2Volume of the droplet prior to impact (pL) 62.9 ± 1.0 59.7 ± 1.3 58.8 ± 0.8Diameter of the dried colloidal assemblies (µm) 23.0 ± 0.7 22.5 ± 0.8 22.3 ± 0.6Height of the dried colloidal assemblies (µm) 6.2 ± 0.2 6.3 ± 0.4 5.8 ± 0.2Average particles center-to-center distance on the 174 ± 7 204 ± 6 260 ± 4surface of colloidal assemblies (nm)

Fig. 1. Schematic of ink-jet printing process for generating uniform droplets containing colloidal latex particles that are self-assembled upon drying on OTS-SAMhydrophobic surface.

716 J. Park et al. / Journal of Colloid and Interface Science 298 (2006) 713–719

in a three-dimensional long-range ordered structure, displayinglarge domains of a hexagonally packed colloidal structure anda few point and line defects. Each domain is composed of hun-dreds to thousands of microspheres. The size of the monocrys-talline domains varied with the particle sizes of the constituentPS latex. The average domain sizes are 2–3, 3–4, and 5–6 µmfor 190, 210, and 270 nm microspheres, respectively. Smallerdomain sizes in the microdome of smaller particles are proba-bly linked to stronger capillary forces associated with smallerpore size of the close-packed sphere. It is believed that thesecapillary forces are also responsible for the deformation of thesoft PS latex during the self-assembly, which results in highlydense ordered colloidal aggregates in which periodic particlescenter-to-center distances on the surface measured by AFM dif-fer from the original particle sizes. The deviations in particlediameters become larger with decreasing particle sizes, as sum-marized in Table 1.

Surface structure of the colloidal crystal in the first visiblelayer may not exactly reflect its internal structure. However,the structure of the latex particles inside the microdome isnot easily observed because of its small size (∼22 µm). UV–vis spectroscopy is often employed to study a photonic bandgap effect for a photonic crystal as well as a complementaryway of probing the colloidal crystal quality [12]. For colloidalcrystal deposited on an opaque material such as silicon, onlyreflectance spectroscopy is feasible. Due to the small size ofthe PS microsphere, it is necessary to use microreflectancespectroscopy to probe each individual microdome. In our in-

vestigation, we have used a microreflectance spectroscopy, andsuccessfully collected reflectance spectrum from an individ-ual colloidal crystal. Fig. 5a shows the reflectance spectra ofa single photonic microcrystal which is composed of colloidalparticles with different sizes. As the particle size increases, theposition of the peak shifts to a longer wavelength, implying thatthe photonic band gap decreases as the colloidal particle size in-creases.

Reflection of light from a plane in a crystalline structure isexpected when the following condition is satisfied [13]:

λc/2 = neffdhkl sin θ,

where λc is the wavelength at the center of the reflection peak,neff is the effective refractive index of the system (air and col-loidal particles), d(hkl) is the interplanar spacing, and θ is theBragg angle. The wavelength in agreement with the Braggequation is selectively reflected from the plane of the colloidalcrystal. The colloidal crystal appears colored upon reflection.A variation in the interparticle lattice constant (i.e., the spherediameter) and/or refractive index contrast can result in a changein the observed reflection peak of the colloidal crystal. The col-loidal photonic crystals are usually arranged into a FCC latticewith the (111) plane parallel to the substrate, which is con-firmed both theoretically and experimentally [14]. For a nor-mal incidence on the (111) plane whose interplanar spacingis d(111) = √

2/3φ (φ is the sphere diameter), the Bragg’s lawcan be expressed as λc = 2 × 0.816neffφ. In Fig. 5b, the reflec-tion maximum (λc) was plotted against the two different sphere

Fig. 2. SEM images of the array composed of the hemispherical colloidal aggregates.

Fig. 3. Optical micrographs of structured hemispherical assemblies obtained from (a) 190 nm, (b) 210 nm, and (c) 290 nm polystyrene latex particles.

J. Park et al. / Journal of Colloid and Interface Science 298 (2006) 713–719 717

diameters (φ): the original PS diameter prior to assembly andthe diameter measured on the colloidal crystal surface by AFM.The observed peak positions fit better with the sphere diametersbased on the measured values than the original ones. From theslope of the fitted curve, we obtain neff = 1.519 for the originaldiameter and neff = 1.451 for the measured diameter. These canbe used to calculate packing factors of the colloidal assemblies.

According to Bruggeman’s effective medium theory, effec-tive refractive indices (neff) can be expressed as [15]

fεps − εeff

εps + 2εeff+ (1 − f )

εair − εeff

εair + 2εeff= 0,

where f is the packing factor of the colloidal/air composite; εps(= n2

ps = 1.592), εair (= n2 = 12), and εeff (= n2 ) are dielec-

air eff

tric constants of polystyrene, air, and effective dielectric con-stant of the colloidal/air composite, respectively. Using Brugge-man’s equation, we have f = 0.882 for the original diameter,which is unrealistically high for the close-packed sphere. Onthe other hand, we obtain f = 0.771 for the measured diame-ter, which is slightly higher that theoretical FCC closed packing(f = 0.74). This result indicates that the polystyrene particlesdo not behave as hard spheres, rather deform to be denselypacked during self-assembly.

We also observed the broadening of the reflection peak asthe size of the polystyrene microsphere increases. If the col-loidal crystals have the same degree of crystal quality, the shapeof the reflection spectra remains the same while the wavelengthcorresponding to the stop bands increases with the sphere di-

Fig. 4. SEM and AFM images of structured hemispherical assemblies obtained from (a) 190 nm, (b) 210 nm, and (c) 290 nm polystyrene latex particles.

(a) (b)

Fig. 5. (a) UV–visible reflectance spectra of a single photonic crystal composed of colloidal particles with different sizes and (b) reflection peak position variationas a function of the diameter of the latex particles.

718 J. Park et al. / Journal of Colloid and Interface Science 298 (2006) 713–719

ameter. This broadening may be related to the higher densityof defects in the photonic crystal made of larger beads, or itcan be attributed to the fact that there are less repeating units inthe photonic crystal made of larger beads than the one made ofsmaller beads. The relative stop bandwidth, �λ/λc, where �λ

is the half-width at half-maximum and λc is the wavelength atthe center of the reflection peak, is inversely proportional to thequality of the colloidal crystal [8b]. For the microdome madefrom 190 nm microspheres, the stop band is narrow (1.7%). The�λ/λc value increases a little (4.8%) for 210 nm microspheresand more markedly for 270 nm microspheres (8.5%), possiblyimplying an increase in disorder.

The presence and number density of the crystal defects de-pends on the way the ordered colloidal assemblies are formed.Our colloidal ink droplet contains the PS latex of similarmonodispersity, surface charge density, and concentration ofthe particles, but varying diameters. Under this circumstance,the solvent flow through the self-assembled structure can play amajor role in determining the quality of the evaporation-drivencolloidal crystals [16]. During self-assembly, spheres that areconvected or sedimented are carried into the growing crystalfront by the evaporation-induced solvent flow. The rate at whichnew particles arrive at the growth site must be slow enough toallow the system to crystallize. Otherwise, either a defectivecrystal or an amorphous solid results. The particle arrival rateis controlled by the solvent flow rate through the pore channelsof close-packed spheres. According to Rideal–Washburn equa-tion, the flow rate increases with pore size that is proportionalto the diameter of the spheres [17]:

L =(

rγlv cos θ

2ηt

)1/2

,

where L is the liquid penetration distance, r the pore radius,η the viscosity, γlv the surface tension, t the penetration time,and θ the contact angle of liquid on the pore surface. Therefore,we suggest that the colloidal assembly composed of a smallerPS latex undergoes a slower crystallization process, leading toa formation of the ordered structures with a higher structuralperfection.

In our direct patterning by the ink-jet printing, the substratedoes not need to be patterned first and no template is needed. In

that sense, our method is different from many other techniquesof patterned growth of colloidal photonic crystal. The most un-conventional feature of our method is the capability of rapidaddressable patterning and/or positioning of the self-assembledcolloidal building blocks, which is desirable for integratingvarious elements into photonic device and optical circuit. Wehave rapidly fabricated the colloidal crystal arrays comprisedof different-sized spheres patterned on the same substrate. Twodifferent colloidal inks containing PS particles with 190 and270 nm in diameter, respectively, were alternatively depositedon the same substrate. Upon drying, an array of microdome-shaped colloidal islands with similar macroscopic dimensionswas formed. However those two photonic crystals with differentcolloidal sizes exhibit two different colors, as is demonstratedin Fig. 6. Blue and red colors arise from the reflection by thecolloidal islands composed of 190 and 270 nm microspheres,respectively. Our strategy is very simple and can be applied toany material system that can be dispersed in the liquids. There-fore, we can anticipate that this technique will enable us torapidly produce hierarchically organized functional structuresof an arbitrary design, which may be of practical importance todirectly writing photonic devices and sensor array.

4. Conclusions

We have fabricated a photonic crystal microarray composedof hemispherical colloidal assemblies by ink-jet printing. A mi-crospectroscopic investigation combined with microstructuralobservations indicate that these colloidal assemblies behave asan individual photonic crystal whose photonic band gap shiftsas a function of the size of the constituent colloidal particle. Itwas observed that the crystal quality of the photonic crystal de-pends on the colloidal particle size, which can be attributed todifference in nucleation and growth kinetics which are involvedin the evaporation-induced self-assembly of the colloidal crys-tals. The particle size determines the pore size of the colloidalassemblies, which in turn influences the evaporation-inducedsolvent flow through the pore and the magnitude of the capillarystress, both of which play important roles in determining thequality of the resulting colloidal crystal. We have also demon-strated the rapid addressable patterning of the photonic crystals

Fig. 6. Confocal and optical microscopic images of the colloidal photonic crystal microarray comprised of different-sized spheres (190 and 290 nm) patterned onthe same substrate.

J. Park et al. / Journal of Colloid and Interface Science 298 (2006) 713–719 719

arrays comprised of different-sized spheres on the same sub-strate by the ink-jet printing, which could not be easily achievedwith other techniques.

Acknowledgments

This work was supported by a grant No. R01-2002-000-00318-0 and the National Research Laboratory (NRL) Pro-gram, both from Korea Science and Engineering Foundation.Portions of the work performed at Auburn University were alsosupported by Auburn University Detection and Food SafetyCenter (AUDFS) through a United States Department of Agri-culture (USDA) grant.

References

[1] (a) Y. Xia, B. Gates, Y. Yin, Y. Lu, Adv. Mater. 12 (2000) 693;(b) C. López, Adv. Mater. 15 (2003) 1679.

[2] (a) Y. Xia, B. Gates, Z.-Y. Li, Adv. Mater. 13 (2001) 409;(b) S.M. Yang, H. Miguez, G.A. Ozin, Adv. Funct. Mater. 12 (2002) 425.

[3] (a) K.E. Davis, W.B. Russel, W.J. Glantschnig, Science 245 (1989) 507;(b) R. Mayoral, J. Requena, J.S. Moya, C. López, A. Cintas, H. Míguez,F. Meseguer, L. Vázquez, M. Holgado, A. Blanco, Adv. Mater. 9 (1997)257.

[4] (a) P. Jiang, J.F. Bertone, K.S. Hwang, V.L. Colvin, Chem. Mater. 11(1999) 2132;(b) Y.A. Vlasov, X.Z. Bo, J.C. Sturm, D.J. Norris, Nature 414 (2001) 289;(c) L.M. Goldenberg, K. Wagner, J. Stumpe, B.R. Paulke, E. Gornitz,Langmuir 18 (2002) 3319.

[5] S.H. Park, Y. Xia, Langmuir 15 (1999) 266.[6] (a) K.M. Chen, X. Jiang, L.C. Kimerling, P.T. Hammond, Langmuir 16

(2000) 7825;(b) H. Zheng, I. Lee, M.F. Rubner, P.T. Hammond, Adv. Mater. 14 (2002)569.

[7] (a) M. Allard, E.H. Sargent, P.C. Lewis, E. Kumacheva, Adv. Mater. 16(2004) 1360;(b) F. Fan, K.J. Stebe, Langmuir 20 (2004) 3062.

[8] (a) K.Y. Suh, A. Khademhosseini, G. Eng, R. Langer, Langmuir 20 (2004)6080;(b) R.E. Schaak, R.E. Cable, B.M. Leonard, B.C. Norris, Langmuir 20(2004) 7293;(c) Y. Yin, Y. Lu, B. Gates, Y. Xia, J. Am. Chem. Soc. 123 (2001) 8718;(d) T. Kanai, T. Sawada, A. Toyotama, K. Kitamura, Adv. Funct. Mater. 15(2005) 25.

[9] (a) S.-C. Chang, J. Liu, J. Bharathan, Y. Yang, J. Onohara, J. Kido, Adv.Mater. 11 (1999) 734;(b) P. Calvert, Chem. Mater. 13 (2001) 3299;(c) T. Shimoda, K. Mori, S. Seki, H. Kiguchi, MRS Bull. 28 (2003) 821;(d) J.Z. Wang, Z.H. Zheng, H.W. Li, W.T.S. Huck, H. Sirringhaus, NatureMater. 3 (2004) 171;(e) B.-J. de Gans, P.C. Duineveld, U.S. Schubert, Adv. Mater. 16 (2004)203.

[10] (a) H.Y. Ko, J. Park, H. Shin, J. Moon, Chem. Mater. 16 (2004) 4212;(b) D. Wang, M. Park, J. Park, J. Moon, Appl. Phys. Lett. 86 (2005)241114.

[11] (a) O.D. Velev, A.M. Lenhoff, E.W. Kaler, Science 287 (2000) 2240;(b) J.R. Millman, K.H. Bhatt, B.G. Prevo, O.D. Velev, Nature Mater. 4(2005) 98;(c) Y. Masuda, T. Itoh, K. Koumoto, Adv. Mater. 17 (2005) 841;(d) J.H. Moon, G.-R. Yi, S.-M. Yang, D.J. Pine, S.B. Park, Adv. Mater. 16(2004) 605.

[12] (a) S. Wong, V. Kitaev, G.A. Ozin, J. Am. Chem. Soc. 125 (2003) 15589;(b) P. Jiang, M.J. McFarland, J. Am. Chem. Soc. 126 (2004) 13778.

[13] (a) V.N. Astratov, A.M. Adawi, S. Fricker, M.S. Skolnick, D.M. Whittaker,P.N. Pusey, Phys. Rev. B 66 (2002) 165215;(b) H. Míguez, C. López, F. Meseguer, A. Blanco, L. Vázquez, R. May-oral, M. Ocaña, V. Fornés, A. Mifsud, Appl. Phys. Lett. 71 (1997) 1148.

[14] (a) L.V. Woodcock, Nature 385 (1997) 883;(b) S.-C. Mau, D.A. Huse, Phys. Rev. E 59 (1999) 4396.

[15] (a) M. Patrini, M. Galli, M. Belotti, L. Andreani, G. Guizzetti, J. Appl.Phys. 92 (2002) 1816;(b) M. Qiu, Appl. Phys. Lett. 81 (2002) 1163.

[16] D.J. Norris, E.G. Arlinghaus, L. Meng, R. Heiny, L.E. Scriven, Adv.Mater. 16 (2003) 1393.

[17] (a) E.W. Washburn, Phys. Rev. 17 (1921) 273;(b) E.K. Rideal, Philos. Mag. 44 (1921) 1152.