Journal of Colloid and Interface Science 388 (2012) 112–117

Contents lists available at SciVerse ScienceDirect

Journal of Colloid and Interface Science

www.elsevier .com/locate / jc is

Dispersion polymerization of styrene using a polystyrene/poly(L-glutamic acid)block copolymer as a stabilizer

Tomomichi Itoh ⇑, Seiji Komada, Eiji Ihara, Kenzo InoueDepartment of Materials Science and Biotechnology, Graduate School of Science and Engineering, Ehime University, 3 Bunkyo-cho, Matsuyama, Ehime 790-8577, Japan

a r t i c l e i n f o

Article history:Received 18 June 2012Accepted 19 August 2012Available online 29 August 2012

Keywords:Polymer particlesBlock copolymer stabilizerPoly(L-glutamic acid)PolystyreneDispersion polymerizationStimuli-responsive polymer

0021-9797/$ - see front matter � 2012 Elsevier Inc. Ahttp://dx.doi.org/10.1016/j.jcis.2012.08.040

⇑ Corresponding author.E-mail address: itoh@ehime-u.ac.jp (T. Itoh).

a b s t r a c t

A block copolymer (PS-b-poly(L-Glu)) composed of polystyrene and poly(L-glutamic acid) was used as astabilizer for dispersion polymerization of styrene. When dispersion polymerization of styrene was con-ducted at 70 �C in 80% dimethylformamide–water with 0.5 wt% PS-b-poly(L-Glu), spherical polystyreneparticles with Dn = 0.72 lm and narrow size distribution were obtained. Whereas AIBN concentrationdid not have any effects on particle size, molecular weight of the polystyrene particles was stronglydependent on the initiator concentration. As concentration of the PS-b-poly(L-Glu) increased from 0.2to 1.0 wt%, particle size decreased from Dn = 0.91 to 0.69 lm with keeping surface area occupied byone poly(L-glutamic acid) chain about S = 50 nm2. On the other hand, an increase in initial concentrationof styrene from 2 to 20 wt% caused an increase in particle size from Dn = 0.48 to 1.36 lm and a decrease insurface area per poly(L-glutamic acid) block from S = 91 to 45 nm2. Colloidal stability of the polystyreneparticles in aqueous solution was responsive to pH due to the surface-grafted poly(L-glutamic acid). Fordispersion polymerization of styrene, the PS-b-poly(L-Glu) functions as both a stabilizer and a surfacemodifier.

� 2012 Elsevier Inc. All rights reserved.

1. Introduction

Polymer particles play important roles in various fields, such asbiomedical analysis, cosmetics, electronics, and chemical industry.Particle morphology, size, size distribution, and surface propertiesof the particles are important for these applications. Dispersionpolymerization is an attractive technique to prepare micrometeror submicrometer polymer particles with narrow size distribu-tions. Because of the direct participation of stabilizer in formationof the particles, type and concentration of stabilizer are key ele-ments in control of dispersion polymerization [1,2]. Thus, severalkinds of amphiphilic homopolymers [3–5], macromonomers[2,6,7], and block copolymers [8–14] have been reported as effec-tive stabilizers. In our previous publication, dispersion polymeriza-tion of styrene using block copolymers composed of polystyrene(PS) and poly(aminomethyl styrene) gave narrowly-distributedPS particles, which were covered with highly functional aminogroups in the poly(aminomethyl styrene), although a poly(amino-methyl styrene) homopolymer was not an effective stabilizer [13].The results demonstrated that PS block copolymers with a hydro-philic polymer block would have considerable potential as a stabi-lizer for PS particle preparation, at the same time capable ofimparting a functional surface to the resulting particles. Accord-

ll rights reserved.

ingly, we have been prompted to develop PS containing blockcopolymers with a hydrophilic polymer block to be utilized as astabilizer.

Poly(a-L-glutamic acid) (poly(L-Glu)) is a typical stimuli-responsive polypeptide to form a right-handed a-helix and adisordered random coil in acidic and alkaline aqueous solution,respectively [15–18]. In addition, the poly(L-Glu) has biocompati-bility, biodegradability, and side-chain functionality. Hence,poly(L-Glu) particles have a great potential to be applied inbiomedical applications. Dibbern et al. prepared poly(L-Glu) core–shell particles whose stability was affected by pH because thepoly(L-Glu) was non-covalently attached to the core [19]. Heiseet al. reported another method using star-polymer formation; thatis, solution polymerization of divinylbenzene with a block copoly-mer (PS-b-PBLG) composed of PS and poly(c-benzyl-L-Glutamate)(PBLG) with an active nitroxide or a chain-transfer agent at thechain terminus of the PS block, followed by hydrolysis of theside-chain ester group in the PBLG block [20,21]. They obtainedhighly-stable nanoparticles, where poly(L-Glu) chains werecovalently bound to a cross-linked core. However, size of the nano-particles was varied and limited less than 100 nm. Rodríguez-Hernández et al. carried out precipitation polymerization ofstyrene–divinylbenzene mixture in presence of a block copolymercomposed of PS and poly(L-Glu) blocks (PS-b-poly(L-Glu)) andsuccessfully obtained highly cross-linked particles with 3–4 lmin size [22]. The surface of the particles showed pH-responsive

T. Itoh et al. / Journal of Colloid and Interface Science 388 (2012) 112–117 113

behavior, while the block copolymer chains were only incorpo-rated in the particles. For other polypeptides, homopolymer ormacromonomer of random coil–forming sodium poly(a,b-asparticacid) synthesized by polycondensation of L-aspartic acid was usedas a stabilizer for dispersion polymerization of styrene to give sub-micrometer particles [6,23]. Although polypeptide-functionalizedpolymer particles have been prepared, each of these methods hassome limitations on size, size distribution, core species, and confor-mation of the polypeptides. In addition, quantitative information ofsurface properties has never been provided by these studies;nevertheless, precise control of surface structure in these particlesis one key factor to achieve advanced functionality.

This study is conducted in an alternative approach to preparepolymer particles modified with a-helix-forming poly(L-Glu) insubmicrometer size range (0.1–1 lm). We employed the poly(L-Glu) as a hydrophilic block to be combined with PS and investi-gated the ability of the block copolymer as a stabilizer for disper-sion polymerization of styrene to obtain PS particles on whichthe poly(L-Glu) chains were bound. The PS-b-poly(L-Glu) was pre-pared by a combination of atom transfer radical polymerizationof styrene [24] and ring-opening polymerization of N-carboxyanhydride (NCA) of amino acid [25,26]. Because of well-definedstructure of the PS-b-poly(L-Glu), we can discuss surface structureof the PS particles in detail. Effects of reaction parameters on sizeand surface structure are also investigated with a view to precisecontrol of the poly(L-Glu)-functionalized PS particles.

2. Experimental section

2.1. Materials

Chloroform, dimethylformamide (DMF), tetrahydrofuran (THF),triethylamine (TEA), and styrene (Nacalai Tesque, Kyoto, Japan,98%) were purified by conventional methods before use. Anisole(Wako Pure Chemical Industries, Osaka, Japan 95%), benzyl alcohol(Wako Pure Chemical Industries, 99%), (1-Bromoethyl)benzene(Tokyo Chemical Industry, Tokyo, Japan, 95%), tert-buthylamine(Tokyo Chemical Industry, 98%), CuBr (Wako Pure ChemicalIndustries, 95%), 1,2-dichloroethane (Nacalai Tesque, 82–84%),L-glutamic acid (Wako Pure Chemical Industries, 99%), 25%HBr–CH3COOH (Wako Pure Chemical Industries), hydrazine mono-hydrate (Nacalai Tesque, 80%), phthalimide potassium salt (NacalaiTesque, 98%), trifluoroacetic acid (TFA; Nacalai Tesque, 99%),triphosgene (Tokyo Chemical Industry, 98%), and (–)-sparteine(Tokyo Chemical Industry, 95%) were used as received.

2.2. Methods

Molecular weight (Mn) and molecular weight distribution (Mw/Mn) of PS homopolymers were determined using gel-permeationchromatography (GPC) on a Jasco-Borwin system (version 1.50;Jasco, Tokyo, Japan) equipped with PS-calibrated Tosoh TSKgel(G3000HHR, G4000HHR and G6000HHR; Tosoh, Tokyo, Japan) usingTHF as an eluent. Analysis of a GPC curve was performed by usinga Origin Pro software equipped with peak-fitting module (version7.5J; OriginLab, Northampton, MA). 1H nuclear magnetic resonance(1H NMR; 400 MHz) spectra were recorded in a Bruker Avance 400spectrometer (Bruker, Rheinstein, Germany). The 1H NMR mea-surements for a block copolymer (PS-b-PBLG) composed of PSand PBLG and a PS-b-poly(L-Glu) were taken in 20% TFA–CDCl3

and dimethylsulfoxide-d6 solutions, respectively. From the 1HNMR measurement, Mn of the PBLG block in the PS-b-PBLG wasdetermined on the basis of Mn of the PS precursor. Scanning elec-tron microscopy (SEM) studies were conducted using a JEOL JSM-5310 (Japan Electron Optics Laboratory, Tokyo, Japan). Average

diameter of particles (Dn) and particle size distribution (Dw/Dn)were determined by following equations:

Dn ¼Pn

i¼1Di

n;

Dw ¼Pn

i¼1D4iPn

i¼1D3i

;

where Di is diameter of a particle estimated from a SEM photographand n is the number of particles (>100). CD spectra were recordedon a Jasco J-820 spectrometer using a 0.1 mm jacked quartz cellin the wavelength region 200–260 nm. Titration of the particleswas carried out by a TOADKK AUT-701 automatic titrator (TOA-DKK, Tokyo, Japan). In a typical case, PS particles (100 mg) wereadded to 5.0 mM NaOH (8 mL) and sonicated to obtain turbid mix-ture which was titrated with 10 mM HCl.

2.3. Synthesis of PS-b-poly(L-Glu)

A x-amino-functionalized PS macroinitiator (Mn = 2100, Mw/Mn = 1.15) was synthesized by atom transfer radical polymeriza-tion of styrene followed by Gabriel method as described in a previ-ous paper [27]. A mixture of N-carboxy-c-benzyl-L-glutamateanhydride (3.00 g, 11.4 mmol), the PS macroinitiator (200 mg,0.09 mmol), and dry chloroform (150 ml) was placed in a flaskand stirred for 3 days. Resulting polymer was purified by repeatedreprecipitations in diethyl ether to give a PS-b-PBLG (yield: 2.24 g,70%). 1H NMR (20% TFA–CDCl3): d = 7.9 (s, NH, 1H), 7.4–7.1 (m, aro-matic H), 5.1 (t, benzyl, 2H), 4.6 (s, Ca H, 1H), 2.5 (s, Cc H2, 2H), and2.1 (t, Cb H2, 2H) for the PBLG block and 7.4–7.1 (m, aromatic H), 6.6(br, aromatic H, 2H), 1.9 (br, CH2–CH, 1H), and 1.5 (br, CH2–CH, 2H)for the PS block.

A solution of the PS-b-PBLG (1.60 g) in dichloroethane (15 ml)was added dropwise to 25% HBr–CH3COOH (10 ml). The mixturewas stirred at room temperature for 24 h. Precipitated polymerwas washed with distilled water by applying centrifugation-decan-tation cycles repeatedly and thoroughly dried under vacuum at45 �C. 1H NMR (dimethylsulfoxide-d6): d = 8.2–8.0 (br, NH, 1H),4.2–3.9 (br, Ca H, 1H), 2.3 (br, CcH2, 2H), and 1.9 (br, Cb H2, 2H)for the poly(L-Glu) block. Signals for the PS block were negligiblysmall due to this block forming micelle cores in dimethylsulfox-ide-d6.

2.4. Dispersion polymerization of styrene

In a typical run, a solution of styrene (200 mg, 1.9 mmol), AIBN(3 mg, 0.02 mmol), and PS-b-poly(L-Glu) (10 mg) in 80% DMF–water mixture (1.8 g) in a glass tube was degassed by freeze–pump–thaw cycles and sealed off under vacuum. Polymerizationwas performed at 70 �C for 17 h without any stirring. The resultingpolymer was filtered over a 0.20 lm membrane filter, repeatedlywashed with methanol and dried under vacuum at 45 �C to givePS particles (yield: 165 mg, 82%).

3. Results and discussion

3.1. Synthesis and characterization of a PS-b-poly(L-Glu)

A stabilizer, PS-b-poly(L-Glu), was synthesized as shown inFig. 1. A PS-b-PBLG prepolymer was prepared as described in ourprevious paper [27]. Molecular weights of the PS and PBLG blockswere determined as Mn = 2000 and 40,000 (the number of repeat-ing units = 19 and 180), respectively. The PS-b-PBLG was treatedwith 25% HBr–CH3COOH for hydrolysis of the benzyl ester. Fig. 2shows 1H NMR spectrum where a signal at 5.1 ppm assigned to

Br

ATRP

CH2 CH Brn 1. Potassium phthalimide

2. 80%NH2NH2 • H2O

CH2 CH NH2n 1. BLG-NCA

2. 25%HBr/CH3COOH

CH2 CH NHn

CO CH NHCH2

m

CH2

COOHPS-b-poly(L-Glu)

Fig. 1. Synthesis of PS-b-poly(L-Glu).

114 T. Itoh et al. / Journal of Colloid and Interface Science 388 (2012) 112–117

benzyl protons of the PBLG block completely disappeared and a Ca

H proton signal was found at 4.0 ppm without any shoulders. Thisindicates that hydrolysis of side-chain groups in the PBLG blockwas achieved without any racemization, and the PS-b-poly(L-Glu)stabilizer was successfully synthesized.

It is well known that poly(L-Glu) forms an a-helix and a randomcoil in acidic and alkaline aqueous solution, respectively. Confor-mation of the poly(L-Glu) block in the PS-b-poly(L-Glu) was esti-mated from CD measurement. In an HCl aqueous solution at pH5.2, an a-helix was clearly indicated from a large negative mini-mum at 222 nm with mol ellipticity [h]222 = �35,000. The molellipticity increased with increasing pH value to reach [h]222 = 0at pH 6.6, indicating formation of a random coil. This helix–coiltransformational behavior was almost identical with that reportedfor poly(L-Glu) homopolymers [17].

Fig. 3. A SEM photograph for PS particles. Polymerization conditions: styrene(10 wt%); AIBN (0.16 wt%, [styrene]0/[AIBN]0 = 100); PS-b-poly(L-Glu) (0.5 wt%);DMF/water = 80:20; 70 �C; 17 h.

3.2. Dispersion polymerization of styrene with the PS-b-poly(L-Glu)

A dispersion polymerization was performed in a homogeneoussolution of styrene (10 wt%), AIBN (0.16 wt%; [styrene]0/[AIBN]0 = 100/1), PS-b-poly(L-Glu) (0.50 wt%), and 80% DMF–watermedium at 70 �C. After 17 h, spherical and narrowly-distributed PSparticles were obtained with Dn = 0.72 lm and Dw/Dn = 1.01 at 82%styrene conversion as shown in Fig. 3. The formation of such nar-rowly-distributed PS particles is in sharp contrast to that ofwidely-distributed ones (Dn = 1.90 lm, Dw/Dn = 1.18) whenpoly(L-Glu) homopolymer (Mn = 21,000) was used as a stabilizer,demonstrating the ability of the PS-b-poly(L-Glu) to control theparticle size. The ability could be attributed to the enhanced

Fig. 2. 1H NMR spectra for PS-b-PBLG in 20% TFA–CDCl3 (upper) and PS-b-poly(L-Glu) in dimethylsulfoxide-d6 (lower).

adsorption of the block copolymer on the particles because of thepresence of the PS block.

While the PS-b-poly(L-Glu) played an important role in stericrepulsion of the PS particles, colloidal stability is conferred via sub-tly different mechanism compared to conventional stabilizers. It isknown that polymer particles are prevented from coagulation bystabilizer chains adsorbed at the surface. For conventional stabiliz-ers such as PVP forming a random coil, steric repulsion is attrib-uted to loss of conformational entropy accompanied bycompression or stretch of the stabilizer chains on surface whenthe particles get into touch [1,2]. In contrast, the entropic loss ofthe poly(L-Glu) was negligible because of formation of an a-helixin 80% DMF–water [28]. To the best of our knowledge, this is thefirst experimental study to utilize an a-helical polypeptide as a ste-ric stabilizer for dispersion polymerization. Although further stud-ies are required, we tentatively propose that the steric repulsion ofthe PS particles is partly ascribed to an excluded volume betweenrigid-rod a-helices.

We briefly investigated pH-responsive behavior of the PS parti-cles which were dispersed by ultrasonication in aqueous solutionof sodium hydroxide (pH 10), hydrochloric acid (pH 2), and dis-tilled water (pH 7) to give turbid mixtures, respectively. After12 h, as shown in Fig. 4, the PS particles maintained colloidal sta-bility at pH 10, but flocculated at pH 2. In the distilled water,slightly dispersed PS particles were observed. The pH-responsivebehavior of the colloidal stability confirms adsorption of thepoly(L-Glu) block onto the PS particles and indicates that the useof the PS-b-poly(L-Glu) is effective in preparation of narrowly-dis-tributed and surface-functionalized PS particles.

The PS particles were added into excess 5.0 mM NaOH and backtitrated by 10 mM HCl aqueous solution. A titration curve showstwo inflections corresponding to neutralization of free NaOH andtitration of carboxylate in side chain of the poly(L-Glu). On theassumption that the junction of the two blocks in the PS-b-poly(L-Glu) is located at the particle surface, surface area occupied

Fig. 4. Colloidal stability of PS particles in hydrochloric acid (pH 2), distilled water(pH 7), and aqueous solution of sodium hydroxide (pH 10), respectively.

(a)

(b)

Fig. 5. (a) A GPC curve for PS particles prepared with 0.08 wt% AIBN ([styrene]0/[AIBN]0 = 200) (solid line) and fitting curves (dashed line). (b) AIBN concentrationdependence of molecular weight of PS particles (solid circle) and ratio of peak areaof the fitting curves: high (open circle), middle (open triangle), and low (opensquare) molecular weight components.

Table 1Effects of polymerization time on the dispersion polymerization of styrene.a

Time (h) Yield (%) Dn (lm) Dw/Dn N (1012)

1 4 0.31 1.02 2.68 79 0.92 1.00 1.9

48 82 0.95 1.00 1.8

a Polymerization conditions: styrene (10 wt%); AIBN (0.16 wt%, [styrene]/[AIBN] = 100); PS-b-poly(L-Glu) (0.5 wt%); DMF/water = 80:20 medium; 70 �C.

T. Itoh et al. / Journal of Colloid and Interface Science 388 (2012) 112–117 115

by one poly(L-Glu) chain (S) was estimated from the followingequation:

S ¼ 6WP0mCHClDnqPSVT NA

;

where WP0 is the weight of particle feed for titration, m is the num-ber of repeating units in the poly(L-Glu) block (=180), CHCl is theconcentration of the HCl aqueous solution used for titration(=10 mM), qPS is the density of PS particles (=1.05 g cm�3), VT isthe difference of titer between the two inflections in the titrationcurve, and NA is the Avogadro number, giving S = 51 nm2. We com-pare the S value with the size of a poly(L-Glu) block. For an a-helix,long axis is calculated to be 27 nm from m = 180 and 0.15 nm perrepeating unit. Interchain distance of a-helical poly(L-Glu) chainswas reported to be between 1.2 nm in crystals [29,30] and 1.8 nmin solution [29,31]. Consequently, cross-sectional area of thepoly(L-Glu) block is assumed to be 32–49 nm2. The value is compa-rable or smaller than the S = 51 nm2. However, anisotropic nature ofthe a-helix requires alignment of the poly(L-Glu) blocks in the samedirection or lying over one another as illustrated in Fig. 5. Accord-ingly, it appears that the surface of the PS particles was crowdedwith the poly(L-Glu) blocks.

From the titration, the ratio of the PS-b-poly(L-Glu) consumedfor surface coverage to that used for the reaction mixture (h) wascalculated from the following equation, leading to h = 11%:

h ¼ CHCIVT Mn;block

mWP0YC0;block;

where Mn,block is the total molecular weight of PS-b-poly(L-Glu), Y isthe yield of PS particles, and C0,block is the initial concentration ofthe PS-b-poly(L-Glu) to styrene monomer (=0.05). Although the sur-face of the PS particles was densely functionalized, a large part ofthe PS-b-poly(L-Glu) stayed in the reaction medium. Previously,we reported for PS-b-poly(aminomethyl styrene)-stabilizing disper-sion polymerization that block length of the PS and the poly(amino-methyl styrene) strongly influenced the h value as well as particlesize and surface structure [13]. Compared with the large composi-tion of the poly(L-Glu) block, which was highly soluble to the med-ium, the anchor PS block was designed to be short so that theadsorption of the PS-b-poly(L-Glu) onto the PS particles was slightlyenhanced. The design of the relative block length is responsible forthe low efficiency of the PS-b-poly(L-Glu) as a surface modifier.

3.3. Effect of polymerization time

Effect of polymerization time on the PS particles is listed in Ta-ble 1. As the dispersion polymerization time was varied from 1 to8 h, yield of the PS particles increased from 4% to 79% and particlesize from Dn = 0.31 to 0.92 lm. Further increases in the yield andthe size were insignificant up to 48 h, indicating that the dispersionpolymerization of styrene was almost complete within 8 h.

It is well known that polymer particles are formed via stabiliza-tion and growth during dispersion polymerization [1–3]. When a

Table 2Effects of initiator concentration on the dispersion polymerization of styrene.a

AIBN (%) [Styrene]/[AIBN] Yield (%) Dn (lm) Dw/Dn N (1012) Mnb

1.6 10 84 0.79 1.01 9.7 50000.32 50 88 0.93 1.00 6.3 180000.08 200 95 0.86 1.00 8.5 650000.03 500 83 0.90 1.01 6.5 1100000.02 750 85 0.81 1.01 9.1 240000

a Polymerization conditions: styrene (10 wt%); PS-b-poly(L-Glu) (0.5 wt%); DMF/water = 80:20 medium; 70 �C; 17 h.b Molecular weight of polystyrene particles.

Table 3Effects of stabilizer concentration on the dispersion polymerization of styrene.a

PS-b-poly(L-Glu) (wt%) Yield (%) Dn (lm) Dw/Dn N (1012) S (nm2) h (%)

0.20 88 0.91 1.01 2.1 50 200.50 82 0.72 1.01 4.0 51 111.0 78 0.69 1.01 4.3 56 5

a Polymerization conditions: styrene (10 wt%); AIBN (0.16 wt%); DMF/water = 80:20 medium; 70 �C; 17 h.

116 T. Itoh et al. / Journal of Colloid and Interface Science 388 (2012) 112–117

reaction mixture is heated, dead polymer chains are generated andaggregated to form unstable particles. Until the point of the stabil-ization, generation and coagulation of the unstable particles suc-cessively occur. The particles are stabilized when surface iscovered with minimum amount of stabilizer chains required forsteric repulsion. In the following stage of particle growth, the sta-bilized particles are enlarged by capture of dead polymer chainsand by polymerization of incorporated monomer.

The number of particles produced from 1 g of styrene monomer(N) is estimated from following equation:

N ¼ 6Y

D3npqPS

;

giving N = 2.6 � 1012 at 1 h. Although the yield increased with poly-merization time, the number of the particles remained constant atN = (2.2 ± 0.4) � 1012. Size distribution of the PS particles was verynarrow within Dw/Dn = 1.02 at each polymerization time. This indi-cates that the stabilization was completed within 4% conversion,and after that point, each PS particle had an equal chance to grow.In spite of a-helical conformation, there is no reason to considerthat the mechanism of particle formation with the PS-b-poly(L-Glu) stabilizer is qualitatively different to that with conventionalstabilizers.

3.4. Effect of AIBN concentration

Concentration of AIBN in the initial solution was varied at 0.02–1.6 wt% ([styrene]/[AIBN] = 750/1–10/1), while PS-b-poly(L-Glu)(0.5 wt%), styrene (10 wt%), reaction time (17 h), and temperature(70 �C) were fixed. As summarized in Table 2, PS particles are foundat Dn = 0.86 ± 0.07 lm and Dw/Dn 6 1.01 regardless of the AIBN con-centration. This indicates that the AIBN concentration did not havea remarkable effect on the particle size. In contrast, it has been re-ported for conventional stabilizers of homopolymer and macromo-nomers that higher initiator concentration gave larger particlesbecause short anchor blocks are formed in the stabilizer by chaintransfer and random copolymerization, respectively, to lead toweak adsorption to unstable particles [1,2]. In this study, the negli-gible effect of the AIBN concentration on the particle size suggeststhat chemical composition of the PS-b-poly(L-Glu) did not changethroughout the dispersion polymerization. As shown in Fig. 5,molecular weight of the particle core was strongly affected by theAIBN concentration: the Mn value for the PS particles prepared with1.6 wt% AIBN (Mn = 5000) was 48 times smaller than that from

0.02 wt% AIBN (Mn = 240,000). A heavy dependence of the molecu-lar weight on the AIBN concentration with Mn / ½AIBN��0:9

0 was ob-served, while a relationship of Mn / ½initiator��0:5

0 is well known forradical polymerization. Fig. 5a shows a GPC curve for PS particlesprepared with 0.08 wt% AIBN. Using a peak-fitting software, thebroad curve is divided into three peaks in high, middle, and lowmolecular weight components. With decreasing the AIBN concen-tration, each component shifted to higher molecular weight region.In addition, fraction of peak area for the high component increasedfrom 6% to 67% as shown in the bottom half of Fig. 5b. This clearlyillustrates the strong effect of the AIBN concentration on the molec-ular weight of the PS particles.

In the Fig. 5a, the peaks of the three components are centered at18.2, 20.0, and 22.4 mL of elution volume which correspond toMn = 670,000, 240,000, and 13,000, respectively. Among them,there is no relationship derived from bimolecular radical couplingand disproportionation reaction. Since the particles were stabilizedat very low conversion, most of the PS chains forming particlecores were prepared in the growth stage. This indicates that thehigh, middle, and low components reflect three mechanisms ofparticle growth. It has been explained for PVP-stabilizing disper-sion polymerization that polymer particles are enlarged by (i) cap-ture of dead polymer chains generated in the reaction medium, (ii)capture of oligomeric radicals and subsequent propagation in theparticles, and (iii) polymerization initiated and propagated in theparticles [1]. In general, bulk polymerization leads to higher molec-ular weight than solution polymerization due to absence of chaintransfer of propagating radicals to solvent molecules and reductionin bimolecular termination in high-viscose solid phase (gel-effect).Hence, it is reasonable that the high, middle, and low componentswere prepared in the manners of the above (i), (ii), and (iii) mech-anisms, respectively.

3.5. Effects of stabilizer and monomer concentrations

Control of both surface structure and particle size is of impor-tance for direct preparation of surface-modified polymer particles.In this experiment, initial concentration of PS-b-poly(L-Glu) stabi-lizer and that of styrene monomer were varied because they di-rectly participate in formation of particle surface and core,respectively.

PS-b-poly(L-Glu) concentration was increased from 0.2 to 1 wt%,while styrene concentration (10 wt%), AIBN concentration(0.16 wt%), reaction time (17 h), and temperature (70 �C) were

Table 4Effects of monomer concentration on the dispersion polymerization of styrene.a

Styrene (wt%) Yield (%) Dn (lm) Dw/Dn N (1012) S (nm2) h (%)

2 75 0.48 1.01 12 91 105 85 0.62 1.01 6.5 77 8

13 86 0.74 1.01 3.6 71 720 88 1.36 1.02 0.63 45 6

a Polymerization conditions: PS-b-poly(L-Glu) (5 wt% relative to styrene); AIBN(1.6 wt% relative to styrene); DMF/water = 80:20 medium; 70 �C; 17 h.

T. Itoh et al. / Journal of Colloid and Interface Science 388 (2012) 112–117 117

fixed, and the results are listed in Table 3. The increase in the PS-b-poly(L-Glu) concentration caused a decrease in size of PS particlesfrom Dn = 0.91 to 0.69 lm and an increase in the number of parti-cles from N = 2.1 � 1012 to 4.3 � 1012. This result was expectedfrom an adsorption–desorption equilibrium where high concentra-tion of the PS-b-poly(L-Glu) enhanced adsorption of the stabilizeronto particles. However, surface area occupied by one poly(L-Glu)chain was about S = 50 nm2 regardless of the PS-b-poly(L-Glu) con-centration. One possible explanation is that the enhancement ofthe PS-b-poly(L-Glu) adsorption was in balance with an increasein required amount of PS-b-poly(L-Glu) chains to stabilize total sur-face of the PS particles, which was enlarged along with the de-crease in particle size. The total surface area of the PS particlesprepared with 1 wt% PS-b-poly(L-Glu) is only 1.3 times larger thanthat with 0.2 wt% PS-b-poly(L-Glu), leading to a decrease in surfacecoverage efficiency from h = 20% to 5%.

Initial concentration of the styrene monomer increased from 2to 20 wt%, while AIBN and PS-b-poly(L-Glu) concentration werekept at 1.6 and 5 wt% relative to styrene (0.03–0.32 and 0.1–1.0 wt% to solution), respectively. As summarized in Table 4, theincrease in the styrene concentration caused enlargement in sizefrom Dn = 0.48 to 1.36 lm along with shrinkage in surface areaper poly(L-Glu) chain from S = 91 to 45 nm2. This can be explainedin connection with change in composition of reaction medium dur-ing the polymerization. Until the stabilization point, high concen-tration of initial styrene allowed more coalescence of unstableparticles because solubility of the PS-b-poly(L-Glu) was enhancedby the styrene monomer, a good solvent for PS. In the followingstage, an increase in total surface area accompanied by particlegrowth demanded further adsorption of the PS-b-poly(L-Glu),which was fulfilled by a decrease in solubility of the PS-b-poly(L-Glu) as styrene monomer was consumed. At the end of the poly-merization, solvency of the reaction medium was irrelevant tothe initial styrene concentration. The PS particles prepared from20 wt% styrene have 2.8 times smaller total surface area than thosefrom 2 wt% styrene. The PS-b-poly(L-Glu) chains were required tobe adsorbed to the small area, which is responsible for the morecrowded surface with twice the number of poly(L-Glu) blocks asindicated from a decrease in the S value. A decrease in the surfacecoverage efficiency from h = 10% to 6% was also caused by the in-crease in the initial concentration of styrene, indicating that fur-ther adsorption of the PS-b-poly(L-Glu) chains became difficultdue to the crowded surface.

Particle size had scaling relationships to PS-b-poly(L-Glu) andstyrene concentrations with Dn / [PS-b-poly(L-Glu��0:3

0 ½styrene�0:40

and surface area per PS-b-poly(L-Glu) chain S / [PS-b-poly(L-GluÞ�00 ½styrene��0:3

0 . These results indicate possibility of indepen-dent control of size and surface by direct preparation of surface-modified polymer particles.

4. Concluding remarks

This study focuses on an alternative method to preparepolypeptide-modified polymer particles. A block copolymer

(PS-b-poly(L-Glu)) composed of polystyrene (PS) and a-helicalpoly(a-L-glutamic acid) (poly(L-Glu)) was an effective stabilizerfor dispersion polymerization of styrene in dimethylformamide–water medium to give micrometer or submicrometer, well-defined, surface-functionalized, and very narrowly-distributed PSparticles. To the best of our knowledge, this is the first experimen-tal study of an a-helical polypeptide as a steric stabilizer fordispersion polymerization. The use of the well-controlled PS-b-poly(L-Glu) enables us to estimate surface structure of the resultingparticles by comparing surface area occupied by one poly(L-Glu)chain and cross-sectional area of an a-helix, indicating crowdedsurface with the poly(L-Glu) chains. Initial concentration of AIBNin the reaction medium had no effect on particle size, whereasmolecular weight of the PS core depended on the AIBN concentra-tion. The effect of initial concentrations of the PS-b-poly(L-Glu) andstyrene monomer had scaling relationships of [PS-b-poly(L-GluÞ��0:3

0 ½styrene�0:40 in size and [PS-b-poly(L-GluÞ�00 ½styrene��0:30

in surface area per poly(L-Glu) chain. These results indicate thatparticle size and surface were separately controllable by directpreparation of PS core–poly(L-Glu) shell particles using the PS-b-poly(L-Glu) stabilizer. The pH-responsive behavior of thepoly(L-Glu) at the surface was reflected by colloidal stability andinstability of the PS particles in an aqueous solution of sodiumhydroxide at pH 10 and in hydrochloric acid at pH 2, respectively.The PS particles covered with poly(L-Glu) are highly attractive inadvanced applications such as latex for immunodiagnosticreagents.

References

[1] A.J. Paine, Macromolecules 23 (1990) 3109–3117.[2] S. Kawaguchi, K. Ito, Adv. Polym. Sci. 175 (2005) 299–328.[3] K.P. Lok, C.K. Ober, Can. J. Chem. 63 (1984) 209–216.[4] H. Bamnolker, S. Margel, J. Polym. Sci., Part A: Polym. Chem. 34 (1996) 1857–

1871.[5] A.J. Paine, W. Luymes, J. McNulty, Macromolecules 23 (1990) 3104–3109.[6] K. Tomita, T. Ono, Colloid Polym. Sci. 287 (2009) 109–113.[7] K.L. Thompson, S.P. Armes, D.W. York, J.A. Burdis, Macromolecules 43 (2010)

2169–2177.[8] G. Riess, C. Labbe, Macromol. Rapid Commun. 25 (2004) 401–435.[9] C.L. Winzor, Z. Mrázek, M.A. Winnik, M.D. Croucher, G. Riess, Eur. Polym. J. 30

(1994) 121–128.[10] F.L. Baines, S. Dionisio, N.C. Billingham, S.P. Armes, Macromolecules 29 (1996)

3096–3102.[11] Y. Ma, W. Yang, J. Polym. Sci., Part A: Polym. Chem. 42 (2004) 2678–2685.[12] J.M. Lee, K. Lee, K. Min, S. Choe, Curr. Appl. Phys. 8 (2008) 732–735.[13] T. Itoh, K. Fukutani, M. Hino, E. Ihara, K. Inoue, J. Colloid Interface Sci. 330

(2009) 292–297.[14] H.V. Penfold, S.J. Holder, B.E. McKenzie, Polymer 51 (2010) 1904–1913.[15] M. Nagasawa, A. Holtzer, J. Am. Chem. Soc. 86 (1964) 538–543.[16] J.Y. Cassim, E.W. Taylor, Biophys. J. 5 (1965) 573–589.[17] Y.P. Myer, Macromolecules 2 (1969) 624–628.[18] A. Holtzer, J. Am. Chem. Soc. 116 (1994) 10837–10838.[19] E.M. Dibbern, F.J.-J. Toublan, K.S. Suslick, J. Am. Chem. Soc. 128 (2006) 6540–

6541.[20] R.J.I. Knoop, M. de Geus, G.J.M. Habraken, C.E. Koning, H. Menzel, A. Heise,

Macromolecules 43 (2010) 4126–4132.[21] F. Audouin, R.J.I. Knoop, J. Huang, A. Heise, J. Polym. Sci., Part A: Polym. Chem.

48 (2010) 4602–4610.[22] A. Bousquet, R. Perrier-Cornet, E. Ibarboure, E. Papon, C. Labrugère, V.

Héroguez, J. Rodríguez-Hernández, Biomacromolecules 9 (2008) 1811–1817.[23] T. Nakashima, Y. Yamada, H. Yoshizawa, Colloid Polym. Sci. 285 (2007) 1487–

1493.[24] K. Matyjaszewski, J. Xia, Chem. Rev. 101 (2001) 2921–2990.[25] H.R. Kricheldorf, Angew. Chem. Int. Ed. 45 (2006) 5752–5784.[26] N. Hadjichristidis, H. Iatrou, M. Pitsikalis, G. Sakellariou, Chem. Rev. 109 (2009)

5528–5578.[27] T. Itoh, T. Hatanaka, E. Ihara, K. Inoue, Polym. J. 44 (2012) 189–194.[28] T. Mori, M. Koga, Chem. Express 4 (1989) 289–292.[29] H.D. Keith, G. Giannoni, F.J. Padden Jr., Biopolymers 7 (1969) 775–792.[30] T. Mori, A. Morofuji, R. Tanaka, Polym. J. 11 (1979) 721–729.[31] P. Saludjian, C. de Lozè, V. Luzzati, C. R. 256 (1963) 4514–4516.