A

msZh©

K

1

sowti

edtsutpco

cibDc

0d

Colloids and Surfaces A: Physicochem. Eng. Aspects 308 (2007) 123–128

Characterization of polymer brushes in capillaries

A.N. Constable, W.J. Brittain ∗Department of Polymer Science, University of Akron, 170 University Ave, Akron, OH 44224, United States

Received 5 February 2007; received in revised form 7 May 2007; accepted 24 May 2007Available online 2 June 2007

bstract

The synthesis of various polymer brush coatings were performed in 0.075 mm i.d. fused silica capillary tubing by atom transfer radical poly-erization (ATRP). Although characterization of polymer coatings inside of capillaries is extremely difficult, we will show how fluorescence

pectroscopy can be used as a technique to verify that covalently attached polymer brushes are present. Capillary rise measurements show theisman critical surface energy (γc) changes as the polymer coatings are changed from either hydrophilic or hydrophobic. The hydrophilic andydrophobic nature of polymers leads to selective flow behavior in micro-capillaries that will allow passive flow control. 2007 Elsevier B.V. All rights reserved.

ace e

updapWfldci

2

2

A9(

eywords: Coatings; Capillary rise; Atom transfer radical polymerization; Surf

. Introduction

Polymer brushes have been widely studied due to their uniquetimuli-responsive nature [1–5]. By functionalizing the surfacef a material, it is possible to create chemical gates [6–9] thatill lead to flow manipulation. It has been shown that tailoring

he surface properties of micro-capillaries facilitates separationsn capillary electrophoresis [10–13].

Polymer brushes can be described as polymer chains teth-red to a surface or interface with a sufficiently high graftingensity such that the chains are forced to stretch away fromhe tethering site [14]. In this work, the polymer brushes wereynthesized via atom transfer radical polymerization (ATRP)sing the “grafting from” approach. An immobilized ATRP ini-iator is first covalently attached to the surface, followed by theolymerization of the desired polymer. ATRP is known to be aontrolled/“living” polymerization that allows for the formationf complex molecular architecture [15,16].

Characterization of the polymer brush on the inside of theapillary is challenging compared to the analysis of flat or spher-cal surfaces. However, coated capillaries can be characterized

y their wetting properties as suggested by de Gennes [17].epending on the liquid chosen, polymers can be partially or

ompletely be wetted.

∗ Corresponding author. Tel.: +1 585 338 6263; fax: +1 585 338 0042.E-mail address: William j brittain@bausch.com (W.J. Brittain).

9a2a3(t

927-7757/$ – see front matter © 2007 Elsevier B.V. All rights reserved.oi:10.1016/j.colsurfa.2007.05.059

nergy; Polymer brushes

In this work, the Zisman critical surface energy (γc) wassed to estimate if the capillary rise corresponding to a particularolymer coating relates well with its γc on a flat wafer which wasetermined by measuring the polymers advancing water contactngel. Herein, we characterize surface coatings that have beenolymerized via ATRP inside of capillaries using capillary rise.e will relate the contact angels that have been measured on

at surfaces and the calculated Zisman critical surface energyata with the height measurements that were measured fromapillary rise. This data will be used to study the effects of flown capillaries and how it can be manipulated.

. Experimental

.1. Materials

Unless otherwise stated, all reagents were purchased fromldrich Chemical Co. 2-Hydroxyethyl methacrylate (HEMA,8%), methyl acrylate (MA, 99%), methyl methacrylateMMA, 99%), pentafluorostyrene (PFS, 99%), styrene (S,9%), and tert-butyl acrylate (t-BA, 98%) were passed throughcolumn of activated basic alumina prior to use. (11-(2-Bromo--methyl)propionyloxy)undecyltrichlorosilane was synthesized

s described in the literature [18]. 2-(8-Methacryloyloyx-,6-dioxaoctyl)thioxantheno[2,1,9-dej]isoquinoline-1,3-dioneHostasol) was donated by Professor Dave Haddleton fromhe University of Warwick. Purification of copper(I)bromide

1 es A:

(N19paowofptS

3

3

tcccwwisdItwasatTcsTtt

3

j5sTtsbawsnv

a(stafftawe3wal[i[i[i[i[i[0[m

3

itpPotic

3a

pihwap

24 A.N. Constable, W.J. Brittain / Colloids and Surfac

98%) was done according to literature methods [19].,N,N′,N′,N′′-Pentamethlydiethylenetrimine (PMDETA, 99%),,1,4,7,10,10-hexamethyltriethylenetetramine (HMTETA,7%), ethyl 2-bromoisobutyrate (E2Br-iB, 98%), cop-er(II)bromide (99%), anhydrous anisole (99%), anhydrouscetone (99.8%), anhydrous methyl alcohol (99.8%), anhydrous-xylene (97%), and reagent plus ethylene glycol (≥99%)ere used without further purification. 75 �m i.d. × 363 �m.d. undeactivated fused silica capillary was purchasedrom Western Analytical Products, Inc., silicon wafers wereurchased from Wafer World Inc., while silicon ATR crys-als (25 mm × 5 mm × 1 mm) were purchased from Harrickcientific.

. Methods

.1. Substrate cleaning and initiator deposition

ATR crystals, silicon wafers, and capillaries were cleaned byreatment with freshly prepared “piranha” solution (70/30, v/v,oncentrated H2SO4/30% aqueous H2O2) at 90 ◦C for 2 h. Theapillary was fitted with a water jacket while one end of theapillary was placed into the “piranha” solution and the otheras placed in an empty Schlenk flask. The “piranha” solutionas introduced to the capillary by reduced pressure while heat-

ng with circulating, hot water through the water jacket. Theamples were then removed and rinsed with copious amounts ofeionized (DI) water followed by drying in a nitrogen stream.t should be noted that the “piranha” solution is extremely reac-ive and should be handled with care. Into a dried Schlenk flaskere placed a freshly cleaned end of the capillary, silicon wafer,

nd ATR crystal. Anhydrous toluene (15 mL) and a 25 vol.%olution of the trichlorosilane initiator in toluene (0.5 mL) weredded to the flask. The opposite end of the capillary was placedhrough a rubber septum that was sealed around a Schlenk flask.he flow of solution was induced through the capillary via aombination of positive nitrogen and reduced pressure at oppo-ite ends. The reaction proceeded for 18 h at room temperature.he silicon wafer, ATR crystal, and capillary were washed with

oluene, sonicated in toluene for 15 min, followed by rinsing inoluene and nitrogen drying.

.2. Typical procedure for surface initiated ATRP

The capillary was fitted and sealed in a cylindrical wateracket, leaving the two ends to be fitted into Schlenk flasks. A0 mL Schlenk flask was used to hold one end of the capillary,ilicon wafer, and ATR crystal and was placed in the oil bath.he oil bath and the circulating water bath were set to the desired

emperature. The opposite end of the capillary was fitted into amall Schlenk flask and sealed. The flasks were degassed andack-filled with nitrogen three times and left under a nitrogentmosphere. CuBr, solvent, monomer, and magnetic stir bar

ere added to a separate 100 mL Schlenk flask and sealed. The

olution was subjected to three freeze-pump-thaw cycles usingitrogen as the backfill gas. PMDETA was added to the solutionia a syringe, and it was stirred at the polymerization temper-

3

s

Physicochem. Eng. Aspects 308 (2007) 123–128

ture until it became homogeneous (∼15 min). Free initiatorE2Br-iB) was added via a syringe to the flask containing theilicon wafer, ATR crystal, and capillary, followed by transfer ofhe CuBr/ligand solution via cannula. The polymerization wasllowed to proceed at room temperature for 3 h (HEMA), 90 ◦Cor 6 h (MA), 60 ◦C for 6 h (MMA), 90 ◦C for 19 h (PFS), 90 ◦Cor 8 h (S), and 60 ◦C for 6 (t-BA). At the end of the reaction,he wafers and capillary were removed and rinsed with CH2Cl2nd THF. To remove untethered polymer chains, the siliconafer and ATR crystal were placed in a Soxhlet extractor and

xtracted with THF for 24 h followed by sonication in THF for0 min followed by drying with an air stream. The capillaryas rinsed with THF for 24 h and then sonicated for 15 min

nd dried with nitrogen. Final concentrations were as fol-ows: [HEMA]0 = 123 mM, [methyl alcohol]0 = 370 mM,HMTETA]0 = 0.51 mM, [CuBr]0 = 0.36 mM, [E2Br-B]0 = 0.15 mM; [MA]0 = 111 mM, [anisole]0 = 184 mM,PMDETA]0 = 0.90 mM, [CuBr]0 = 0.49 mM, [E2Br-B]0 = 0.46 mM; [MMA]0 = 47 mM, [anisole]0 = 92 mM,PMDETA]0 = 0.94 mM, [CuBr]0 = 0.47 mM, [E2Br-B]0 = 0.47 mM; [PFS]0 = 36 mM, [o-xylene]0 = 164 mM,PMDETA]0 = 0.53 mM, [CuBr]0 = 0.27 mM, [E2Br-B]0 = 0.27 mM; [S]0 = 118 mM, [anisole]0 = 152 mM,PMDETA]0 = 0.77 mM, [CuBr]0 = 0.38 mM, [E2Br-B]0 = 0.16 mM; [t-BA]0 = 89 mM, [acetone]0 = 230 mM,PMDETA]0 = 0.45 mM, [CuBr]0 = 0.30 mM, [E2Br-iB]0 =.15 mM; fluorescence experiments: [Hostasol]0 = 0.047 mM,MMA]0 = 47 mM, [anisole]0 = 92 mM, [PMDETA]0 = 0.94M, [CuBr]0 = 0.47 mM, [CuBr2]0 = 0.047 mM.

.3. Typical procedure for diblock copolymer synthesis

Preparation of diblock copolymers was conducted in a sim-lar manner to that for the homopolymer brushes. The order ofhe blocks in the diblock brush was determined by the order ofolymerization from the surface. For example, a Si/SiO2//PS-b-MA brush was synthesized by first forming Si/SiO2//PS brushn the surface, followed by the polymerization of MA fromhe Si/SiO2//PS brush. The same concentrations that were usedn the homopolymer brush synthesis were used in the diblockopolymer synthesis.

.4. Typical procedure for deprotection of tert-butylcrylate to acrylic acid

A P(t-BA)-modified silicon wafer and ATR crystal werelaced in a crystallization dish while the capillary was placednside a glass column. These were then added to an oven pre-eated to 190–200 ◦C for 2 h under reduced pressure. The siliconafer, ATR-crystal, and capillary were rinsed with DI water

nd then submerged in DI water overnight following previousrocedure [4].

.5. Characterization methods

FTIR-ATR spectra were recorded with a Bruker Tensor 27pectrometer using a modified 4XF beam condenser (Harrick

A.N. Constable, W.J. Brittain / Colloids and Surfaces A: Physicochem. Eng. Aspects 308 (2007) 123–128 125

//PS-b

Ssabcio(tsP[(stAwtTt

4

4h

adb

ptowiotsubstrate by self-assembly techniques. In this research, (11-(2-bromo-2-methyl)propionyloxy)undecyltrichlorosilane was syn-thesized according to literature techniques [18] and used asthe initiator. Fig. 1a displays the ATR-FTIR spectrum of the

Scheme 1. Synthesis of Si/SiO2

cientific). Spectra were recorded at 4 cm−1 resolution, and 32cans were collected. Contact angles were determined using

Rame Hart NRL-100 goniometer equipped with a tiltingase mounted on a vibrationless table. Advancing and recedingontact angles of a 10 �L drop were determined using the tilt-ng stage method. Ellipsometry measurements were performedn a Gaertner model L116C ellipsometer with a He–Ne laserλ = 632.8 nm) and a fixed angle of incidence of 70◦. For thehickness layer calculations, refractive indices of n = 1.455 (forilicon oxide) [20], n = 1.508 (for initiator layer), n = 1.512 (for(HEMA)) [20], n = 1.480 (for PMA) [20], n = (for PMMA)20], n = 1.473 (for PPFS) [21], n = 1.589 (for PS) [20], n = 1.464for P(t-BA)) [20], and n = 1.527 (for PAA) [20]. Fluorescencepectroscopy measurements were made on a Jobin Yvon T6400riple monochromator that was equipped with an argon ion laser.n excitation wavelength of 457.8 nm and a power of 0.67 mWere used. Capillary rise measurements were recorded using a

raveling telescope cathetometer M912 from Gaertner Scientific.he probe fluid was allowed to rise overnight in order to come

o its equilibrium height.

. Results and discussion

.1. Synthesis and characterization of tetheredomopolymer and diblock copolymer brushes

Scheme 1 illustrates the general ATRP “grafting from”pproach used to synthesize the tethered homopolymer andiblock copolymer brushes. Here, the synthesis of Si/SiO2//PS--PAA diblock copolymer brush is shown. Similar synthesis

FS

-PAA brush on a silica surface.

rocedures were used to polymerize other polymer brushes. Dueo the challenges in characterizing the polymer brushes formedn the inside of the capillaries, silicon wafers and ATR crystalsere used to follow the reaction. Fig. 1 shows the addition of

nitiator to the surface followed by the synthesis of each blockf the copolymer brush, Si/SiO2//PS-b-PAA. A solution deposi-ion of a bromoisobutyrate ATRP initiator is tethered to a suitable

ig. 1. ATR-FTIR spectra of (a) Si/SiO2//bromoisobutyrate initiator, (b)i/SiO2//PS, (c) Si/SiO2//PS-b-P(t-BA), and (d) Si/SiO2//PS-b-PAA.

126 A.N. Constable, W.J. Brittain / Colloids and Surfaces A: Physicochem. Eng. Aspects 308 (2007) 123–128

Table 1Contact angle and ellipsometry data for the PS-based diblock copolymer brush on a flat substrate

Brush structure Water contact anglea (◦) Ethylene glycol contact anglea (◦) Thicknessb (nm)

θa θr θs

Si/SiO2//OH 3c 9 –Si/SiO2//ATRP initiator 89 72 55 5Si/SiO2//PS 92 74 57 22Si/SiO2//PS-b-P(t-BA) 91 72 68 10Si/SiO2//PS-b-PAA 61 47 44 4

le; θr, receding contact angle, and θs, static contact angle.easurements is ±1 nm.

isbtasio2sidtiwA1ddahiw

4

dcmSiaosstc

ptPr

St

spectrum of a freshly cleaned inner capillary having a sur-face covered with hydroxyl groups, and the capillary in Fig. 2b(Si/SiO2//PMMA) is a PMMA homopolymer brush coveringthe inner surface. Both spectra show no fluorescence, but

a The standard deviation of contact angles was <2◦. θa, advancing contact angb Thickness was determined by ellipsometry, and typical error on thickness mc Static contact angle was recorded for this measurement.

mmobilized initiator with peaks at 2927 and 2854 cm−1 (C–Htretching and CH2 stretching vibrations) and 1732 cm−1 (car-onyl stretching vibration of the ester group). Upon synthesis ofhe PS homopolymer brush, Fig. 1b shows the presence of peakst 3030, 3061, and 3084 cm−1, all attributed to aromatic C–Htretching vibrations. Table 1 shows the representative advanc-ng and receding contact angles for the PS homopolymer brushf 92◦ and 74◦, respectively, which has a brush thickness of2 nm. With the addition of the second block, P(t-BA), Fig. 1chows the carbonyl stretching peak at 1726 cm−1 correspond-ng to the addition of t-BA and a doublet at 1367 and 1392 cm−1

ue to the symmetric methyl deformation mode. There was lit-le change in the contact angles, but Table 1 shows a 10 nmncrease in brush thickness. The Si/SiO2//PS-b-P(t-BA) chainsere converted to Si/SiO2//PS-b-PAA chains via pyrolysis [4].fter pyrolysis (Fig. 1d), there is a broadening of the peak at726 cm−1 and a loss of the peaks associated with the pen-ent methyl groups. The advancing and receding contact anglesecreased from 91◦ and 72◦ for P(t-BA) to 61◦ and 47◦ for PAAs shown in Table 1. The outer most portion of the polymer brushas decreased from 10 to 4 nm after pyrolysis. This is approx-mately a 40% decrease in the outermost polymer brush layerhich correlates well with literature [4].

.2. Characterization of capillary coatings

Due to the inability to chemically characterize the inneriameter of capillaries, analysis has been challenging. Commonharacterization techniques typically employed to study poly-er brushes on flat surfaces and nanoparticles could not be used.ilicon wafers and ATR crystals were used to follow the surface-

nitiated polymerizations. Fluorescence spectroscopy was useds way to prove that polymer brushes were attached to the insidef the capillary, while capillary rise experiments were used tohow changes in the surface energy of the capillaries after eachynthetic step took place. With this technique, it was possibleo evaluate if the polymerization of homopolymers and diblockopolymers occurred.

Scheme 2 illustrates the structure of the PMMA-co-Hostasol

olymer brush, while Fig. 2 displays the fluorescence spec-roscopy data obtained from various systems. To prepare theMMA-co-Hostasol polymer brush, we used a monomer feedatio of 1000:1, MMA:Hostasol. Fig. 2a, Si/SiO2//OH, is a

F(

cheme 2. Structure of the copolymer brush used in the fluorescence spec-roscopy experiments.

ig. 2. Fluorescence spectroscopy of (a) Si/SiO2//OH, (b) Si/SiO2//PMMA, andc) Si/SiO2//PMMA-co-Hostasol.

A.N. Constable, W.J. Brittain / Colloids and Surfaces A: Physicochem. Eng. Aspects 308 (2007) 123–128 127

Table 2Physical properties for polymer coated capillary surfaces

Brush structure Advancing water contact anglea, θa (◦) Zisman critical surface energyb, γc (mN/m) Capillary rise (cm)

Si/SiO2//OH 4 95.2 24.17Si/SiO2//ATRP Initiator 89 31.8 7.92Si/SiO2//PS 90 31.1 6.64Si/SiO2//PPFS 94 28.1 12.61Si/SiO2//P(HEMA) 49 61.6 23.01Si/SiO2//P(t-BA) 92 29.6 5.63Si/SiO2//PAA 32 74.3 23.55Si/SiO2//PMA 72 44.5 10.93Si/SiO2//PS-b-P(t-BA) 92 29.6 4.04S

±0.1

Ssei

[t

γ

wgiagtwhTarwc

l(vaf‘wortwSrppiih

aahcStSmSbpcanmtthicken the layer) affect the thickness of the brush which in-turn,affect the height of the capillary rise [17]. These effects, whencomparing the same polymer coated capillaries polymerized atdifferent times, show unexpected results in capillary rise exper-

i/SiO2//PS-b-PAA 61

a The standard deviation of contact angles was <2◦.b γc was calculated by plotting θa vs. γc from the data in Ref. [25]; error was

i/SiO2//PMMA-co-Hostasol (Fig. 2c) shows a large peak emis-ion between 510 and 540 nm. This large emission peak providesvidence that the copolymer has been physically attached to thenside walls of the capillary.

Eq. (1) relates capillary rise and surface tension to one another22]. From their work, the treatment of capillary rise and surfaceension (γ) has been deduced:

= �ρghr

2 cos θ(1)

here �ρ is the difference in densities of liquid and gas phases,the acceleration due to gravity, h height that the liquid rises

n the capillary, r the radius of the capillary, and cos θ is thengle between the capillary wall and the meniscus. Ethylenelycol was used as the probe fluid due to its medium/high surfaceension of 49.89 mN/m [23] when compared to surface tension ofater at 73.56 mN/m [23], while most common organic solventsave a surface tension ranging from 17.98 to 34.96 mN/m [23].hese latter organic solvents have a tendency to completely wetpolymer surface. In capillary rise, the equilibrium height is

eached when the forces due to surface tension are equal to theeight of the probe fluid; further discussion about capillary rise

an be found in a number of publications and text books [24].A number of polymer brush systems were studied via capil-

ary rise as a way to characterize the various polymer coatingsTable 2). As the inner capillary surface was modified from aery hydrophilic surface to a hydrophobic surface and back tohydrophilic surface, these changes in surface energy could be

ollowed as shown in Fig. 3. A freshly cleaned surface, using apiranha solution,’ left the silica covered with hydroxyl groupshich allowed the ethylene glycol to completely wet the surfacef the capillary. This enhanced wetting leads to a large capillaryise (24.17 cm) corresponding to a small contact angle betweenhe capillary wall and the probe fluid. As the silica surfaceas modified to become more hydrophobic (Si/SiO2//PS andi/SiO2//PS-b-P(t-BA)), the capillary rise (6.64 and 4.04 cm,espectively) was lower due to less compatibility between theolymer and the probe fluid. Upon pyrolysis, the outer most

olymer layer was converted to PAA which produced a large cap-llary rise (23.86 cm) for the Si/SiO2//PS-b-PAA modified cap-llary value. The large capillary rise is a result of the probe fluidaving a greater compatibility with the PAA coated capillary.

Fec

52.7 23.6

.

As seen in Table 2, the data follows the expected patternpart from the Si/SiO2//PPFS system. The lowest water contactngles (and thus, the more hydrophilic compositions) have theighest Zisman critical surface energy which corresponds to theapillary system with the greatest capillary rises (Si/SiO2//OH,i/SiO2//P(HEMA), and Si/SiO2//PAA). The polymer systems

hat are more hydrophobic, Si/SiO2//PS, Si/SiO2//PPFS, andi/SiO2//P(t-BA), have higher water contact angles, lower Zis-an critical surface energy, and have a lower capillary rise.i/SiO2//PPFS does not follow the expected trend and coulde a result of the polymer brush thickness being low and/or theolymer brush grafting density being sufficiently lower whenompared to the other capillary systems. These differences canlso be seen on flat surfaces which can cause the surface rough-ess to increase and cause a change in the contact angle that iseasured. de Gennes et al. described how gravity forces (tend

o thin the layer) and long range Van der Waals forces (tends to

ig. 3. Capillary rise data for modified capillaries which has anrror of ± 0.05 mm. aSi/SiO2//OH, bSi/SiO2//bromoisobutyrate initiator,Si/SiO2//PS, dSi/SiO2//PS-b-P(t-BA), eSi/SiO2//PS-b-PAA.

1 es A:

iws

e[

5

itcpeaaib

A

dHflAr

R

[

[

[

[

[[

[

[

[

[[

[

[

[[

[

[

28 A.N. Constable, W.J. Brittain / Colloids and Surfac

ments. The important finding from these experiments is thate can follow changes in surface energies following multi-step

urface modification.Currently, we are using this data to design capillaries that will

xhibit passive flow control in response to flow compositions26,27].

. Conclusion

Various polymer brush systems have been synthesized in cap-llaries. Using fluorescence spectroscopy, we are able to showhat the polymer brush is physically attached to the wall of theapillary. By measuring the capillary rise of each system, it isossible to follow the change in Zisman critical surface energy ofach layer of our polymer brush. The height of the rise changess we go from a hydrophilic to a hydrophobic surface and back tohydrophilic surface. From this work, we can relate the changes

n surface energies that can be measured by capillary rise andegin to study how flow can be manipulated in capillaries.

cknowledgements

This research was supported by the National Science Foun-ation (DMR-0072977). We would like to thank Professor Daveaddleton from the University of Warwick for donating theuorescent monomer and Professor Alexei Sokolov and Dr.lexandor Kisliuk from the University of Akron for the fluo-

escence spectroscopy measurements.

eferences

[1] W.J. Brittain, S.G. Boyes, A.M. Granville, M. Baum, B.K. Mirous, B.Akgun, B. Zhao, C. Blickle, M.D. Foster, Surface rearrangement of diblockcopolymer brushes-stimuli responsive films, Adv. Polym. Sci. 198 (2006)125–147.

[2] R.C. Advincula, W.J. Brittain, K.C. Caster, J. Ruhe (Eds.), PolymerBrushes: Synthesis, Characterization, Application, Wiley-VCH, Wein-heim, Germany, 2004.

[3] P. Uhlmann, L. Ionov, N. Houbenov, M. Nitschke, K. Grundke, M.Motornov, S. Minko, M. Stamm, Surface functionalization by smart coat-ings: stimuli-responsive binary polymer brushes, Prog. Org. Coat. 55 (2006)168–174.

[4] N.D. Treat, N. Ayres, S.G. Boyes, W.J. Brittain, A facile route topoly(acrylic acid) brushes using atom transfer radical polymerization,Macromolecules 39 (2006) 26–29.

[5] C. de las Heras Alarcon, T. Farhan, V.L. Osborne, W.T.S. Huck, C. Alexan-

der, Bioadhesion at micro-patterned stimuli-responsive polymer brushes,J. Mater. Chem. 15 (2005) 2089–2094.

[6] A.M. Granville, W.J. Brittain, Stimuli-responsive semi-fluorinated polymerbrushes on porous silica substrates, Macromol. Rapid Commun. 25 (2004)1298–1302.

[

Physicochem. Eng. Aspects 308 (2007) 123–128

[7] Y. Ito, Y.S. Park, Y. Imanishi, Visualization of critical pH-controlled gatingof a porous membrane grafted with polyelectrolyte brushes, J. Am. Chem.Soc. 119 (1997) 2739–2740.

[8] M. Ejaz, Y. Tsujii, T. Fukuda, Controlled grafting of a well-defined polymeron a porous glass filter by surface-initiated atom transfer radical polymer-ization, Polymer 42 (2001) 6811–6815.

[9] Y.S. Park, Y. Ito, Y. Imanishi, pH-controlled gating of a porous glass fil-ter by surface grafting of polyelectrolyte brushes, Chem. Mater. 9 (1997)2755–2758.

10] X. Huang, L.J. Doneski, M.J. Wirth, Surface-confined living radical poly-merization for coatings in capillary electrophoresis, Anal. Chem. 70 (1998)4023–4029.

11] M.D. Miller, G.L. Baker, M.L. Bruening, Polymer-brush stationary phasesfor open-tubular capillary electrochromatography, J. Chromatogr. A 1044(2004) 323–330.

12] A. Feldmann, U. Claußnitzer, M. Otto, Optimization of capillary coating byhydroxyethyl methacrylate for capillary zone electrophoresis of proteins,J. Chromatogr. B 803 (2004) 149–157.

13] N. Idota, A. Kikuchi, J. Kobayashi, Y. Akiyama, K. Sakai, T. Okano,Thermal modulated interaction of aqueous steroids using polymer-graftedcapillaries, Langmuir 22 (2006) 425–430.

14] S.T. Milner, Polymer brushes, Science 251 (1991) 905–914.15] K. Matyjaszewski, J. Xia, Atom transfer radical polymerization, Chem.

Rev. 101 (2001) 2921–2990.16] M. Kamigaito, T. Ando, M. Sawamoto, Metal-catalyzed living radical

polymerization, Chem. Rev. 101 (2001) 3689–3745.17] F. Brochard-Wyart, P.G. de Gennes, Capillary rise of a liquid on a solid

grafted with long polymer chains, J. Adhes. Sci. Technol. 7 (1993) 495–502.

18] K. Matyjaszewski, P.J. Miller, N. Shukla, B. Immaraporn, A. Gelman, B.B.Luokala, T.M. Siclovan, G. Kickelbick, T. Vallant, H. Hoffmann, T. Pakula,Polymers at interfaces: using atom transfer radical polymerization in thecontrolled growth of homopolymers and block copolymers from siliconsurfaces in the absence of untethered sacrificial initiator, Macromolecules32 (1999) 8716–8724.

19] R.N. Keller, H.D. Wycoff, Copper(I) chloride, Inorg. Synth. 2 (1946) 1–4.20] J. Brandrup, E.H. Immergut, E.A. Grulke (Eds.), Polymer Handbook, Wiley

& Sons, New York, 1999.21] C. Pitois, S. Vukmirovic, A. Hult, D. Wiesmann, M. Robers-

son, Low-loss passive optical waveguides based on photosensitivepoly(pentafluorostyrene-co-glycidyl methacrylate), Macromolecules 32(1999) 2903–2909.

22] P.C. Hiemenz, Principles of Colloid and Surface Chemistry, M. Dekker,New York, 1977, p. 516.

23] C.L. Yaws, Chemical Properties Handbook, McGraw Hill, New York, 1999.24] (a) B. Legait, P.G. de Gennes, Capillary rise between closely spaced plates:

effect of van der Waals forces, J. Phys. Lett. 45 (1984) 647–652;(b) A.W. Adamson, A.P. Gast, Physical Chemistry of Surfaces, 6th ed.,John Wiley & Sons, Inc., New York, 1997.

25] E.A. Vogler, Structure and reactivity of water at biomaterial surfaces, Adv.Colloid Interface Sci. 74 (1998) 69–117.

26] D.J. Beebe, J.S. Moore, J.M. Bauer, Q. Yu, R.H. Liu, C. Devadoss, B.

Jo, Functional hydrogel structures for autonomous flow control insidemicrofluidic channels, Nature 404 (2000) 588–590.

27] N. Idota, A. Kikuchi, J. Kobayashi, K. Sakai, T. Okano, Microfluidicvalves comprising nanolayered thermoresponsive polymer-grafted capil-laries, Adv. Mater. 17 (2005) 2723–2727.