polyelectrolyte block copolymer and oppositely charged surfactant

Colloids and Surfaces B: Biointerfaces 99 (2012) 127–135

Contents lists available at SciVerse ScienceDirect

Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb

Salt effect on microscopic structure and stability of colloidal complex obtained from neutral/polyelectrolyte block copolymer and oppositely charged surfactant Masahiko Annaka ∗ Department of Chemistry, Kyushu University, 6-10-1, Hakozaki, Higashi-ku, Fukuoka-shi, Fukuoka 812-8581, Japan

a r t i c l e

i n f o

Article history: Available online 2 December 2011 Keywords: Diblock copolymer Surfactant Colloidal complex Salt-effect Laser light scattering Small angle neutron scattering

a b s t r a c t The salt effect on complex formation of poly(acrylamide)-block-poly(acrylic acid) (PAM-b-PAA) and dodecyltrimethylammonium bromide (DTAB) at different NaBr concentrations, CNaBr , was investigated by laser light scattering (LLS) and small angle neutron scattering (SANS). LLS and SANS clearly indicates that the aqueous solution of PAM-b-PAA and DTAB associate into colloidal complexes. For low surfactant-topolymer charge ratio Z lower than the critical value ZC , the colloidal complexes are single DTAB micelles dressed by a few PAM-b-PAA. Above ZC , the colloidal complexes form a core–shell microstructure. The complex formation in the PAM-b-PAA/DTAB is enhanced by addition of salt: ZC decreases with increasing CNaBr . This is considered to similar to the cmc behavior for the pure surfactant system. The core of the complex consists of densely packed surfactant micelles (DTA+ ), and PAA block chains bind to these micelles, displace their counteranions (Br−) and bridge them. The corona of the complex is constituted from the PAM. Since the interaction between polyelectrolyte and oppositely charged surfactant is primarily electrostatic in nature, the core radius and the intermicellar distance of the DTA+ micelles inside the core depend on CNaBr . The addition of salt screens the electrostatic attraction between oppositely charged PAA block and DTAB, which weakens the interaction. With increasing CNaBr , therefore, the core of the colloidal complex is considered to swell, which leads to the increases in the core radius RC and the intermicellar distance of the DTA+ micelles inside the core. The aggregation number expressed in terms of DTA+ micelles per complex is also evaluated using the analogy with the homopolyelectrolyte/surfactant system. © 2011 Elsevier B.V. All rights reserved.

1. Introduction During the past decade, there has been a widespread interest for the design and synthesis of new polymer-based colloidal particles of high stability in aqueous solutions. Among these particles, colloidal complexes have emerged as a new type of microstructure with potential applications in drug delivery, and gene delivery [1]. Colloidal complexes involving oppositely charged polyelectrolyte or polyelectrolyte and low-molecular-weight surfactant constitute a very active field of research in recent years [2–6]. These systems attract scientific attention due to a number of interesting characteristics. From the scientific point of view, such systems present the possibility of the formation of various self-assembled nanostructures, whose characteristics can be tuned by a large number of parameters, including total concentration, charge ratio, ionic strength, and pH. These systems also show similarities in structure and behavior with more complex biological macromolecular self-assembled systems, such as lipoproteins, and protein/DNA

∗ Corresponding author. Tel.: +81 92 642 2594; fax: +81 92 642 2607. E-mail address: [email protected] 0927-7765/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfb.2011.11.021

complexes [7–9]. Since the interaction between polyelectrolyte and oppositely charged surfactant is primarily electrostatic in nature, electrostatic factors, such as macromolecular charge densities and ionic strength, are the most important factors. Whether or not association takes place seems to depend mainly on the micelle surface charge density, the polymer linear density, and the Debye–Hückel ion-atmosphere thickness [10,11]. The addition of salt, therefore, should have a significant effect on the complex formation between polyelectrolyte and oppositely charged surfactant. Addition of salt, thus far, is considered to weaken the strength of complex formation: that is, the critical surfactant concentration for the onset of complex formation increases with increasing salt concentration [10–15]. Furthermore, the addition of excess amount of salt completely suppresses the formation of complexes [16–19]. These salt-reducing effects on complex formation are generally explained in terms of the reduction or complete screening of the electrostatic attraction between polyelectrolyte and surfactant. Stoichiometric polyelectrolyte–surfactant complexes usually precipitate from aqueous solution. Recently, however, Berret and co-workers [20–22] started to explore colloidal complexes result from a self-assembly mechanism between polyelectrolyte–neutral block copolymer and oppositely charged surfactant. The overall

128

M. Annaka / Colloids and Surfaces B: Biointerfaces 99 (2012) 127–135

size and stability of the colloid depend on the nature of the electrostatic charges, on the molecular weight, and on the flexibility of the chains. Berret and co-workers [22–24] have, respectively, performed small-angle neutron scattering studies of colloidal complexes resulting from the self-assembly of poly(acrylamide)block-poly(acrylic acid) (PAM-b-PAA) as a neutral–anionic block copolymer and dodecyltrimethylammonium bromide (DTAB) as a cationic surfactant. They proposed that the complex exhibited a core–shell structure, of which the core is a dense and disordered microphase made of surfactant micelles connected by the polyelectrolyte blocks. The corona of the core–shell complex was a diffuse shell of the neutral chains and it ensured steric stability. Recently, our group revealed the two contrary salt effect on the complex formation of PAM-b-PAA/DTAB system depending on the salt concentration: the salt-enhancing effect on the complex formation between PAM-b-PAA and DTAB in lower salt concentration range, and the salt-reducing effect in higher salt concentration range. On one hand, the addition of salt favors the formation and growth of surfactant micelles [23–28]. Therefore, in the polyelectrolyte/surfactant system, surfactant molecules tend to form larger micelles at higher salt concentration, which have a strong tendency to bind on polymer chain [10,12,29,30]. This salt effect can be designated as the increasing interaction. On the other hand, the addition of salt screens the electrostatic attraction between PAM-b-PAA and DTAB, which weakens the interaction. This salt effect can be designated as the screening of interaction. Thus, the overall effect of salt depends on the competition of the increasing of interaction with the screening of interaction. However, such a salt-enhancing effect on the complex formation has not yet well understood. More precise study, therefore, must be needed to reveal the salt effect on the complex formation between polyelectrolyte and oppositely charged surfactant. In this study, we focused on the salt effect on complex formation and microscopic structures of complex composed of PAM-b-PAA and DTAB under various conditions, specifically the effects of the addition of NaBr, and DTAB concentration (surfactant-to-polymer charge ratio) by laser light scattering (LLS), and small-angle neutron scattering (SANS). 2. Experimental 2.1. Materials Acrylamide (AAm) was recrystallized from ethyl acetate and dried in vacuo. Acrylic acid (AAc, Wako Pure Chemicals) was distilled under reduced pressure prior to use. 4,4 -azobis(4-cyanopentanoic acid) (ABCA, Wako Pure Chemicals) was recrystalized from methanol. Dimethylsulfoxide (DMSO, Wako Pure Chemicals) and methanol (Wako Pure Chemicals) were distilled over drying agent under dry nitrogen atmosphere prior to use. A RAFT agent, 2-cyanopropyldithiobenzoate (CPDB) was prepared according to literature procedure [31,32]. Nphenyl-1-naphthylamine (PNA, Aldrich) was used without further purification. 2.2. Preparation of PAM-b-PAA was prepared by reversible additionPAM-b-PAA fragmentation chain transfer (RAFT) polymerization using 2-cyanopropyldithiobenzoate as a chain transfer agent (CTA). PAM was prepared by reversible addition-fragmentation chain transfer polymerization using CPDB as a chain transfer agent. AAm, CPDB and ABCA as initiator were dissolved in DMSO. The solution was deoxygenated by three freeze–pump–thaw cycles, and polymerization was carried out at 70 ◦ C. The polymer was isolated by

precipitation in an excess of diethyl ether, and purified by repeated precipitations, followed by drying in vacuo. RAFT polymerization of diblock copolymers, PAM-b-PAA, were performed with PAM as the macro-CTA. Acrylic acid (AAc), AIBN, and PAM were dissolved in acetate buffer (pH 5.0), and solutions were deoxygenated by three freeze–pump–thaw cycles, and polymerization was carried out at 67 ◦ C. After polymerization, polymers were precipitated in an excess diethyl ether and then dried in vacuo. 2.3. Preparation of PAM-b-PAA/DTAB complexes The colloidal complexes in salt-free solutions might be out of equilibrium state. Stop-flow experiments conducted by Kavanov and co-workers [28] showed that polyelectrolyte complexes formation takes place in less than 5 ␮s, nearly corresponding to the diffusion collision of the polyion coils. In salt-free solutions, frozen structures in a non-equilibrium state are considered to form. In this study, the total concentration of the system CT , defined as CT = CP + CDTAB , where CP and CDTAB are the weight concentrations of polymer and of DTAB, is fixed to 0.5 wt%. As standard condition, 0.25 wt% PAM-b-PAA in NaBr were put into the cell, and equal volume of DTAB solution and a solution of 0.5 wt% PAM-b-PAA containing same NaBr concentration were added up to the desired charge ratio Z. The dosage was carried out continuously with a slow rate under gentle stirring at 25 ◦ C. Solutions were made dust-free by filtration through a Millipore membrane (0.45 ␮m pore size). A polymer–surfactant charge ratio Z is defined as Z = [S]/˛n[P], where [S] and [P] are the molar concentrations for the surfactant and for the polymer, respectively. ˛ and n denote the degree of neutralization and the average number of monomers in the polyelectrolyte block, respectively. Z = 1 describes the isoelectric solution, characterized by the same number densities of positive (DTA+ ) and negative (COO− ) charged ions. 2.4. Characterization 2.4.1. Gel permeation chromatography (GPC) GPC of CPDB-PAM and PAM-b-PAA was performed using a TOSOH HLC-8220GPC apparatus with TSKgel G5000PWXL + G2500PWXL columns. 50 mM NaHCO3 /100 mM NaNO3 /20 mM triethanol amine/0.03% NaN3 was used as the mobile phase at a flow rate 1.0 mL/min at 25 ◦ C. Monodisperse poly(ethylene glycol) standards was used for calibration. 2.4.2. 1 H NMR Number average molecular weight, Mn , of CPDB-PAM (macroCTA) was determined by 1 H NMR spectra using JEOL JNMAL300 spectrometer operating at a frequency of 300 MHz in DMSO-d6 . The probe temperature was kept constant ±0.5 ◦ C by the passage of thermostatically regulated air during accumulation. The temperature was measured with a calibrated thermocouple. The number-averaged molecular weight is estimated by comparing the 1 H NMR peak area of phenyl protons in the terminal dithiobenzoate with that of methine protons of PAM. 2.4.3. Potentiometric titration Potentiometric titration was carried out to determine the number average molecular weight of PAA block in a titration vessel in which the temperature was controlled to within ±0.1 ◦ C of 25 ◦ C under a nitrogen atmosphere. Titration of PAM-b-PAA in its acidic form was carried out with 200 mM NaOH aqueous solution as a titrant at 0.2 wt% polymer concentration, and the solution pH was monitored with a ORION 720A pH meter with ORION 8102BN electrode precalibrated with pH 4.01, 7.0 and 10 buffers. The titration was carried out slowly for several hours to allow for proper equilibration.

M. Annaka / Colloids and Surfaces B: Biointerfaces 99 (2012) 127–135

2.5. Laser light scattering Static and dynamic light scattering (DLS) experiments were conducted with an ALV DLS/SLS-5000 light scattering system equipped with an ALV-5000 multiple  digital correlator for the measurements of the Rayleigh ratio R␪ (q,c) and the collective diffusion coefficient D(c). The wavevector q is defined as q = (4n/) sin(/2), where n is the refractive index of the solution,  the wavelength of the incident beam ( = 632.8 nm), and  the scattering angle. The excess Rayleigh ratio (Eq. (1)) was determined from the scattering intensity at wavevector q = 1.87 × 10−3 A˚ −1 ( = 90◦ ) normalized with respect to that of benzene.



n2 n2Ben

Species

bN (10−12 cm)

V0 (A˚ 3 )

N (1010 cm−2 )

CH2 CHCONH2 CH2 CHCOOH CH2 CHCOONa DTAB DTA+ Br− D2 O

1.336 1.66 2.40 −1.138 −1.817 0.679 1.92

2141 79.4 54.8 495.5 456.2 39.3 30

0.711 2.09 4.37 −0.23 −0.40 1.73 6.38

3. Results and discussion

 RBen, 

(1)

where RBen ,  is the Rayleigh ratio of benzene at scattering angle , with the value of 8.51 × 10−6 cm−1 at 632.8 nm; I, I0 , IBen are the scattered intensities of the solution, the solvent, and benzene, respectively; and nBen is the refractive index of benzene. The autocorrelation functions were analyzed by performing the inverse Laplace transform using the routine CONTIN assuming the superposition of exponentials for the distribution of relaxation times. The decay rate  of each process is calculated as the inverse of its relaxation time  −1 (= ). In the case of diffusive process, its diffusion coefficient D is obtained from the slope of  vs q2 by  = Dq2 . From this value the hydrodynamic radius was calculated according to the Stokes–Einstein equation: RH =

Table 1 Coherent neutron scattering length (bN ), molecular volume (V0 ), and length density (N ) of the surfactant and the polymers studied in this work.

kB T 6 0 D

(2)

where kB is the Boltzmann constant, 0 is the viscosity of the solvent, and T is the temperature of the sample. The polymer solutions were clarified by filtering through a Millipore membrane (0.45 mm pore size).

The PAM-b-PAA was prepared by RAFT polymerization using 2cyanopropyldithiobenzoate as a chain transfer agent as shown in Scheme 1. The molecular characteristics of the macro-CTA and the block copolymers, as they have been determined by 1 H NMR, GPC and potentiometric titration, are presented in Table 2. The obtained polydispersity is low. The living/controlled character of the polymerization is supported by the appearance of the characteristic UV signal at 500 nm due to the absorbance of the dithiobenzoate Ph S(C S) chromophore of the CTA for macro-CTA and block copolymers. Another evidence of this feature is given by characteristic 1 H NMR signals for both the dithiocarbenzoate and 2-cyanopropyl end groups of the polymer. The use of CPDB leads to the formation of ␻-dithiobenzoate homopolymers. These are subsequently utilized as macro-CTA in order to prepare diblock copolymers. The representative GPC trace of the PAM-b-PAA clearly shows the formation of block copolymer: GPC trace shifts to the higher molecular weight region after polymerization of acrylic acid from PAM macro-CTA (Fig. 1). This indicates that the PAM precursor efficiently participates as a macro-CTA via RAFT, therefore well-defined AB-type block copolymers are produced. 3.2. Potentiometric titration

2.6. Small-angle neutron scattering The small-angle neutron scattering experiments were performed using SANS-II spectrometer at the Swiss spallation neutron source SINQ, Paul Scherrer Institute, in Switzerland [33]. Each sample was kept at 25 ± 0.5 ◦ C in a quartz cell with optical length of 2 mm. The observed scattered intensity was corrected for cell and solvent scattering, incoherent scattering, and transmission by conventional procedures. The two-dimensional isotropic scattering spectra were azimuthally averaged, converted to an absolute scale, and corrected for detector efficiency by dividing by the incoherent scattering spectrum of pure water, which was measured with a 1 mm cell. We recall that the expression of the intensity scattered by a dispersion of spherical and homogeneous particles of volume V and radius R. d

(q, c) = n(c)V 2 2 S(q, c)F(q, R) d˝

3.1. Polymerization and molecular characterization

(3)

where n(c) is the number density of particles at concentration c, and  (=N − S ) is the difference of the scattering length density with respect to the solvent. Here, N and S are the scattering length density of the chemical species and the solvent, respectively. F(q) and S(q,c), respectively, are the form factor and the structure factor of the particles. The list of the coherent scattering lengths, molecular volumes, and coherent scattering length densities of the chemical species studied in this work are given in Table 1 [34,35]. In D2 O (S = 6.38 × 1010 cm−2 ) all these components are contributing to the total cross section.

The pH dependence on the degree of neutralization, ˛ = [Na+ ]/[COOH]0 , where [COOH]0 is the total concentration of carboxyl and carboxylate groups and [Na+ ] assigns the amounts of added NaOH, for PAM92 PAA156 under the presence of 100 mM NaCl at 25 ◦ C is presented in Fig. 2. From the titration curve, the apparent pKa value is found to be 5.4. This value agrees well with that obtained for PAA titrated in the presence of 100 mM NaCl

80

Response / mV

I − I0 R = IBen

129

PAM122AAc58

60

macro-CTA PAM

40

20

0 12

13

14

15

16

17

Elution Volume / mL Fig. 1. GPC trace for macro-CTA (CPDB-PAM) and PAM122 PAA58 in 50 mM NaHCO3 /100 mM NaNO3 /20 mM triethanol amine/0.03% NaN3 measured at 25 ◦ C.

130

M. Annaka / Colloids and Surfaces B: Biointerfaces 99 (2012) 127–135

S

CH3

+

S C

C

NC

CH2

CH3

CH C O

NC C

benzene 70ºC

NH2

S

CH3

AIBN

( CH2

CH ) S C n C O

CH3

NH2

CPDB

Acrylamide

COOH

S

CH3

ABCA

CH2 CH

+

PAM

PAM

NC C

Acetate buffer (pH 5.0) 67ºC

( CH2

CH3

Acrylic acid

CH ) n C O NH2

( CH2

CH ) S C m COOH

PAM-PAA diblock copolymer (PAMnPAA m) Scheme 1. Preparation of PAM-b-PAA.

Table 2 Characterization of PAA (macro-CTA) and PAM-b-PAA diblock copolymer. Code

[M] (mol L−1 )

Feed ratio [M]/[CTA]/[I]a

Reaction time (h)

CPDB-PAM PAM122 PAA58

4.00 3.00

4000/10/1 2500/10/2

6 6

a b c d

Mnb 8670 12,800

Monomer unit [AAm]/[AAc]c

PDId

122/– 122/58

1.15 1.18

Concentrations of monomer [M], RAFT agent or macro-CTA [CTA], and initiator [I]. Number-averaged molecular weight Mn estimated by 1 H NMR. Estimated by 1 H NMR and potentiometric titration. Polydispersity index determined by GPC.

(pKa = 5.4) [36]. This value is considerably lower than that for linear PAA in the absence of salt, reported to be 6.3 [37]. This is likely that, in the absence of salt, the negative charges created upon deprotonation of same AA units make the subsequent deprotonation of AA units more difficult. In the presence of sufficient of salt, those charges are screened, faciliating the deprotonation of all AA units. In this study, we investigated the sample, coded as PAM122 PAA58 with degree of neutralization ˛ = 0.8. 3.3. Laser light scattering For the light scattering experiments, PAM122 PAA58 /DTAB/NaBr solutions with the degree of neutralization ˛ = 0.8 were prepared

in H2 O with Z comprised between 0.1 and 10 under the presence of CNaBr = 0, 50, and 100 mM. The excess Rayleigh ratios R90 measured at q = 1.87 × 10−3 A˚ −1 ( = 90◦ ) for aqueous solutions of PAM122 PAA58 /DTAB with different NaBr concentrations are shown in Fig. 3. At low values of Z, the scattering intensity is independent of surfactant concentration and it remains at the level of the pure polymer (Z = 0). With increasing Z, there exists a critical charge ratio noted ZC , and comprised between 0.5 and 1.0 depending on the NaBr concentration, above which the Rayleigh ratio increases by several orders of magnitude, indicating the formation of colloidal complexes. The Rayleigh ratio for solutions with NaBr, then, decreases slowly at higher Z-values. For the samples with

10 10 -2

NaCl

9

ΔR 90 / cm-1

0mM 50mM 100mM

8

pH

7 6 5

10 -3

10 -4 0 mM NaBr 50 mM NaBr 100 mM NaBr

4 3

10 -5

2 0.0

0.2

0.4

0.6

0.8

1.0

Degree of neutralization, α Fig. 2. Potentiomatric titration curve for 1.0 g/L PAM122 PAA58 under the presence of 0, 50 and 100 mM NaCl at 25 ◦ C.

10-1

100

101

Charge Ratio, Z Fig. 3. The excess Rayleigh ratios of 0.5 wt% PAM92 PAA156 /DTAB with different charge ratio Z in aqueous solutions of (a) CNaBr = 0 mM, (b) CNaBr = 50 mM and (c) CNaBr = 100 mM obtained at q = 1.87 × 10−3 A˚ −1 ( = 90◦ ) and 25 ◦ C.

M. Annaka / Colloids and Surfaces B: Biointerfaces 99 (2012) 127–135

(b) CNaBr = 50 mM

G(Γ)

(a) CNaBr = 0 mM

10-4

10-2

100

131

(c) CNaBr = 100 mM

Z = 10

Z = 10

Z = 10

Z=5

Z=5

Z=5

Z=2

Z=2

Z=2

Z=1

Z=1

Z=1

Z = 0.7

Z = 0.7

Z = 0.7

Z = 0.5

Z = 0.5

Z = 0.5

Z = 0.3

Z = 0.3

Z = 0.3

Z = 0.1

Z = 0.1

Z = 0.1

102

104 10-4

10-2

τ / ms

100

102

104 10-4

τ / ms

10-2

100

102

104

τ / ms

Fig. 4. Dependence of the relaxation time distributions of 0.5 wt% PAM122 PAA58 /DTAB on the charge ratio Z in aqueous solutions of (a) CNaBr = 0 mM, (b) CNaBr = 50 mM and (c) CNaBr = 100 mM obtained at q = 1.87 × 10−3 A˚ −1 ( = 90◦ ) and 25 ◦ C.

10 2

RH / nm

CNaBr = 50 and 100 mM, higher electrostatic inter-micellar repulsion and inter-complex repulsion induced by a large amount of bound micelles may lead to a dissolution of large complexes into small complexes, corresponding to a decrease in the Rayleigh ratio. Fig. 4 shows the distributions of relaxation times of 0.5 wt% PAM122 PAA58 /DTAB solutions with different NaBr concentrations obtained at q = 1.87 × 10−3 A˚ −1 ( = 90◦ ) and 25 ◦ C. For solutions at Z < ZC , two relaxation processes are observed. At Z = 0.1 and CNaBr = 0 mM, the fast process with diffusivity Dfast = 4.5 × 10−7 cm2 s−1 , which corresponds to a hydrodynamic ˚ is attributed to single polymer chain, radius RH,fast = 55 ± 1.1 A, while slower process with diffusivity Dslow = 0.38 × 10−7 cm2 s−1 , ˚ which corresponds to a hydrodynamic radius RH ,slow = 650 ± 25 A, is also observed. This process may be attributed to some kind of polymer aggregation. Both processes exhibit low scattering intensities; this is expected for the unimer chains due to their small size, whereas for the polymer aggregates, it suggests that they are either very few in number or highly hydrated and, therefore, have a low scattering contrast. It is worthy to mention that since the CONTIN analysis renders the intensity-weighted distribution, the proportion of the large particle is strongly exaggerated, as the scattering intensity is dependent on the particle radius (∼R6 for spherical particle). The formation of these aggregates is still not well understood. The observed behavior is almost identical to those up to Z = 0.5 for all NaBr concentrations. Above ZC < Z, a remarkable changes in relaxation processes are observed for all samples. Here the normalized intensity autocorrelation function exhibits single exponential function with diffusive character of the concentration fluctuations. At Z = 0.7 and CNaBr = 0 mM, the single process with diffusivity D = 1.3 × 10−7 cm2 s−1 is observed, which corresponds to a hydro˚ which is likely to be attributed to dynamic radius RH = 190 ± 2.1 A, the complex formed by the association of PAM-b-PAA and DTAB. For sample solutions with NaBr, the critical charge ratio ZC is shifted to larger Z-value, the distribution of relaxations time exhibit narrow single peaks above Z = 1.0 as observed in Fig. 4b and c. At CNaBr ≥ 0.1 M, the shift of ZC to higher Z with increasing CNaBr is consistent with the generally observed shielding effect of salt on the complex formation. It is interesting to note that the polydispersity is narrower than that of single diblock copolymer chain. Fig. 5 exhibits the evolution of the hydrodynamic radius RH for a 0.5 wt% PAM122 PAA58 /DTAB with different concentrations of NaBr, CNaBr as a function of charge ratio Z at 25 ◦ C. Three regimes can be ˚ which is consistent distinguished. At low Z (
10 1 0 mM NaBr 50 mM NaBr 100 mM NaBr

10 0 10 -1

10 0

10 1

Charge Ratio, Z Fig. 5. The evolution of the hydrodynamic radius RH for a 0.5 wt% PAM92 PAA156 /DTAB in different concentrations of aqueous NaBr solutions as a function of charge ratio Z at 25 ◦ C.

increase in hydrodynamic radius with increasing Z. Beyond ZC , RH has larger value at higher CNaBr at same Z. For the NaBr-free sample, ˚ On the other hand, above ZC , RH is rather constant in Z around 300 A. by addition of NaBr, the core of the colloidal complex is considered to swell due to the screening of interaction between PAA blocks and DTAB micelles inside the core, which leads to the increase in RH . For the samples with CNaBr = 50 and 100 mM, further increase in the charge ratio Z leads to decrease in RH , which is consistent with the dependence of the excess Rayleigh ratio R90 on the charge ratio Z. The addition of salt screens the electrostatic attraction between PAM122 PAA58 and DTAB, which weakens the interaction. This saltscreening effect may lead to dissolution of large complexes into small complexes, corresponding to a decrease in RH . 3.4. Microstructure of colloidal complex Fig. 6 compares the scattering cross sections of the colloidal complexes obtained by SANS for 5 g/L aqueous solutions of PAM122 PAA58 /DTAB with different values of Z and CNaBr at 25 ◦ C. The pure polymer solution (Z = 0) exhibits an overall weak scattering (Fig. 6a), as is expected for dilute polymer solutions in good solvent. The scattering intensity from the pure DTAB solution (Z = ∝) is also modest (Fig. 6a); however, it exhibits a broad correlation-bump around 0.05 A˚ −1 attributed to electrostatic

132

M. Annaka / Colloids and Surfaces B: Biointerfaces 99 (2012) 127–135

102 (a) NaBr = 0 mM 1

10

Intensity / cm-1

(b) NaBr = 50 mM

Z = 0 (pure polymer) Z = 0.5 Z = 0.7 Z = 1.0 Z = 2.0 Z = ∞ (pure DTAB)

(c) NaBr = 100 mM

Z = 0.5 Z = 0.7 Z = 1.0 Z = 2.0

Z = 0.5 Z = 0.7 Z = 1.0 Z = 2.0

100 q0

q0

10-1

q0

10-2 10-3 10-2

10-1

Scattering vector /

Å-1

10-2

10-1

Scattering vector /

10-2

Å-1

10-1

Scattering vector / Å-1

Fig. 6. SANS profiles for 0.5 wt% PAM122 PAA58 /DTAB with different charge ratio Z in (a) D2 O, (b) 50 mM NaBr D2 O solution, and (c) 100 mM NaBr D2 O solution at 25 ◦ C. The arrow indicates the position of the structure peak at q0 ∼ 0.16 A˚ −1 . The solid line represents the form factor of polydisperse and homogeneous DTAB micelles of average radius ˚ In (a), SANS profiles for pure 0.5 wt% PAM122 PAA58 (Z = 0) and 0.1 wt% pure DTAB (Z = ∝) are included for comparison. RMic = 19 A.

Intensity / cm-1

interactions between DTAB micelles. For the PAM122 PAA58 /DTAB solutions, there exists a critical charge ratio above which the scattering intensity is dominated by two remarkable features: a strong forward scattering and the appearance of a structure peak at high wavevectors. This peak is located around q0 = 0.16 A˚ −1 and shown by arrow in Fig. 6. The amplitude of this peak depends on Z and/or CNaBr . It is worthy to note that for NaBr-free PAM122 PAA58 /DTAB at Z = 1 and Z = 2, there is a damped oscillation in the intensity decay around 0.05 A˚ −1 . Berret et al. [21] reported that these features are correlated, and they are characteristic to the core–shell aggregates. For Z ≤ 0.5, the peak becomes a weak bump, which seems to be superimposed to a power law dependence of the form I(q)∼q−3/2 . In copolymer micelles, the contribution of the corona manifests itself at a high wavevector as a power law of the form I(q)∼q−1/ , where is the Flory exponent. For Gaussian chains = 1/2, which gives a q−2 decrease of the intensity, whereas for chains in good solvent = 3/5 and intensity decreases as q−5/3 . Scaling exponents of the order of −3/2 are close to those for polymers in good solvent. This critical Z-value is consistent with that observed by DLS. All the scattering intensities obtained for PAM122 PAA58 /DTAB aqueous solutions with different concentrations of NaBr present typically the same characteristics, therefore we selected the ones for CNaBr = 50 mM to analyze the data. The neutron scattering data in the range Z < ZC are plotted in Fig. 7 using a Guinier representation

10-1

NaBr = 50mM Z = 0.5 Z = 0.7 Z = ∞ (pure DTAB)

10-2 0.00

0.01

for spherical particles. Here we consider the data only at high q region (0.07 A˚ −1 < q < 0.17 A˚ −1 ). All the data exhibit an exponential decrease of the intensity against q2 . The straight lines result from the best-fit calculations using the expression

 I(q) = I(0) exp

−

Rg2 q2

 (4)

3

where I(0) is the intensity extrapolated at zero wavevector, Rg the radius of gyration of the particles. These two quantities are adjustable parameters in the fitting. In Table 3, the parameters obtained from Guinier analysis for all the scattering data are listed. The solid line in Fig. 7a (Z = 0.5 < ZC ) is calculated using a form factor of polydisperse and homogeneous spheres of aver˚ In the q-region age radius R = 19 A˚ (standard deviation, R = 2.5 A). of 0.07 A˚ −1 < q < 0.17 A˚ −1 , therefore, we could assume that polymer chains do not contribute to the scattering intensity. For CNaBr = 0 mM and Z = 0.5 (
0.02

0.03

q2 / Å-2 Fig. 7. Guinier representation of the SANS data obtained from 0.5 wt% PAM122 PAA58 /DTAB aqueous solutions with CNaBr = 50 mM at Z = 0.5 and 0.7. The data obtained from the 0.1 wt% DTAB micelle D2 O solution (Z = ∝) is included for comparison.

NaBr (mM)

0 50 100

Z = 0.5

Z = 0.7

˚ Rg (A)

˚ RMic (A)

˚ RC (A)

˚ RMic (A)

16.9 ± 1.0 17.8 ± 0.7 19.5 ± 00

21.8 23.1 25.2

18.0 ± 0.4 18.9 ± 0.6 21.7 ± 0.7

23.2 24.4 28.0

M. Annaka / Colloids and Surfaces B: Biointerfaces 99 (2012) 127–135

133

1.2

102

100 10

-1

10-2

(b) q4 x I(q) / 10-6 cm-1 Å-4

Intensity, I(q) / cm-1

(a) 101

Z = 1.0 CNaBr = 0 mM RC = 111 Å

1.0 0.8 0.6 0.4 Z = 1.0

0.2

0 mM NaBr 50 mM NaBr 100 mM NaBr

0.0

10-3 10-2

10-1

0

Scattering vector, q / Å-1

0.01 0.02 0.03 0.04 0.05 0.06 Scattering vector, q / Å-1

Fig. 8. (a) SANS intensity for 0.5 wt% PAM122 PAA58 /DTAB aqueous solutions with CNaBr = 0 mM at Z = 1.0. The continuous line corresponds to the scattering from slightly ˚ (b) Porod plot of the SANS intensity for 0.5 wt% PAM122 PAA58 /DTAB at Z = 1 in different polydisperse spherical particles of radius RC = 111 A˚ and standard deviation R = 11 A. concentrations of NaBr D2 O solution.

The salt-enhancing effect is due to the larger increasing of interaction in comparison with the screening of interaction. This behavior is considered to similar to the cmc behavior for the pure surfactant system [2,39]. In PAM122 PAA58 /DTAB system, most of bromide counterions of DTAB are replaced by PAA blocks. Since the SANS experiments clearly identify spherical micelle below the cmc, the polymer–surfactant solution is above the critical aggregation concentration (cac). The SANS intensity at intermediate and low q-region have been interpreted as arising from spherical and homogeneous particles. To determine the core size of the colloidal complexes formed by the system PAM122 PAA58 /DTAB at ZC < Z, the SANS curves are fitted assuming a Gaussian distribution of homogeneous spheres with an average core radius RC and a standard deviation R [20]. For homo2

geneous spheres, F(q, R) = [3(sinX − X cosX)/X 3 ] where X = qR. In this regime, for micelles distributed according to a Gaussian distribution function, the scattering cross section becomes d

(q, c) = n(c)2 d˝



∞

V 2 G(R, RC , R )F(q, R)dR 0

 

1 exp − G(R, RC , R ) = √

R 2

(R − RC )2

(5)

 (6)

2 R2

where RC and R are the average core radius and standard deviation of the Gaussian function. In Fig. 8a is plotted the experimental intensity obtained at Z = 1 and CDTAB = 0 mM, together with the scattering function associated to slightly polydisperse spherical particle ˚ In the interof radius RC = 111 A˚ and a standard deviation R = 11 A. mediate q range, the agreement between the experimental and calculated intensity is satisfactory. The data for the PAM122 PAA58 /DTAB system at =0.5 wt% and Z = 1 with different CNaBr is shown in Fig. 8b using the Porod representation (q4 × I(q) versus q). The porod representation aims at emphasizing the local curvature of interfaces separating at the elementary scatterers and solvent molecules. The samples exhibit damped oscillations by fitting assuming a distribution of spherical

and homogeneous particles with average radius RC and a standard deviation R . The agreement between the data and the calculated curves are excellent. The values for RC at CNaBr = 0, 50, 100 mM are summarized in Table 4 for Z-values of 0.7, 1.0 and 2.0. The core radius RC increases with CNaBr . It is also confirmed that the oscillations of the form factor in the Porod representation moved toward higher q with increasing CNaBr . We can deduce the radius of the core from the position of the maximum and minimum in the Porod representation using the following relationship [42]. √ 3 3 q1max RC = (7) , q2max RC = 2, qmin RC =  2 2 The RC has values ranging from the micellar radius ˚ and the hydrodynamic radius (RH ∼ 200–300 A). ˚ (RMic ∼ 20–30 A) This result is due to the fact that the PAM122 PAA58 /DTAB forms a core–shell structure. The core dominates the SANS cross section at intermediate wavevector (0.01 A˚ −1 < q < 0.1 A˚ −1 ), whereas the corona which is composed by the PAM blocks is associated with the hydrodynamic size of the colloidal complex [39]. With increasing CNaBr , the structure peak moved toward higher q. This is likely that the core of the colloidal complex is considered to swell due to the screening of interaction between PAA blocks and DTAB micelles inside the core, which leads to the increase in RC . At large wavevectors (0.1 A˚ −1 < q), and charge ratios above the critical value (ZC < Z), the scattering is characterized by a structure peak at a wavevector noted q0 . For CNaBr = 0 mM, this peak located around q0 = 0.16 A˚ −1 , which corresponds to the charac˚ and is independent of the charge teristic distance 2/q0 ∼ 39 A, ratio. The intensity of the peak grows until a stoichiometric PAA block-surfactant complex (Z = 1) is formed. This indicates that the stoichiometry of the structural units is not significantly changed, since a charge imbalance would lead to a swelling due to repulsive double-layer force. The structure peak exhibits the two noticeable features. First, the position of q0 is found to shift to lower wavevectors with increasing CNaBr as shown in Fig. 9a. Second, when the aliphatic chains of DTAB are deuterated (C12 D25 N+ (CD3 )3 Br−,

Table 4 List of the parameters used in the SANS data assuming a distribution of spherical particles of radius RC and standard deviation R . NaBr (mM)

0 50 100

Z = 0.7

Z = 1.0

Z = 2.0

˚ RC (A)

˚

R (A)

˚ RC (A)

˚

R (A)

˚ RC (A)

˚

R (A)

145 ± 0.4 106 ± 0.6 –

25.8 ± 1.2 29.7 ± 1.4 –

127 ± 0.4 106 ± 0.6 96.9 ± 0.7

11.9 ± 1.6 10.6 ± 2.1 17.9 ± 2.4

113 ± 4.2 107 ± 2.9 135 ± 5.7

11.0 ± 7.0 19.3 ± 1.6 40.2 ± 3.1

134

M. Annaka / Colloids and Surfaces B: Biointerfaces 99 (2012) 127–135

102

6.0

(b) Z = 1 NaBr = 0 mM

0 mM NaBr 50 mM NaBr 100 mM NaBr

5.0

1

10

Intensity / cm-1

Intensity / 10-2 cm-1

(a) Z = 2

4.0 3.0

100 DTAB-d34

10-1 10-2

2.0 1.0 0.05

DTAB

10-1 0.10

0.15

0.20

0.25

10-2

10-1

Scattering vector / Å-1

Scattering vector / Å-1

Fig. 9. (a) Shift in the position of the structure peak at large wavenumber for PAM122 PAA58 /DTAB in different concentrations of NaBr D2 O solutions. (b) Comparison of the SANS profiles for 0.5 wt% D2 O solutions of PAM122 PAA58 /DTAB (Z = 1) and PAM122 PAA58 /DTAB-d34 (Z = 1). The structure peak q0 shown by arrow in the figure disappears when the aliphatic chains of DTAB are deuterated.

DTAB-d34 ), the peak disappears (Fig. 9b), indicating a change in the scattering contrast. From our observations, therefore, using the analogy with the homopolyelectrolyte/surfactant concentrated phases [43–46], the core is constituted of densely packed surfactant micelles (DTA+ ), and PAA block chains bind to these micelles, displace their counteranions (Br−) and bridge them together. The intermicellar distance of the DTA+ micelles decorated by PAA block ˚ By the addition of NaBr, the core of chains in the interior is ∼39 A. the colloidal complex is considered to swell due to the screening of interaction between PAA blocks and DTA+ micelles inside the core, which leads to the increase in the intermicellar distance of the DTA+ micelles and the shift of the position of q0 to lower wavevectors. We evaluate the aggregation number of micelles Nagg per colloidal complex in the line with the arguments made by Berret and co-workers [20,39]. Nagg can be written: Nagg = Mic

 R 3 C RMic

(8)

where Mic is the micellar volume fraction in the core. Mic is related to the homopolyelectrolyte/surfactant system. Poly(sodium acylate) and DTAB in solutions undergo a phase separation, giving rise to a solid precipitate which displays the scattering features of a cubic phase of micelles. The sequence of Bragg peaks is interpreted as arising from a cubic structure associated with the space group of symmetry Pm3n [22,47,48], meaning 8 micelles per unit cell and a capacity of 0.524. Since the core of the colloidal complex does not form any crystal structure, we can conclude that the upper limit of the volume fraction of micelle in the core must be less than 0.524. Here we implicitly assume that the mechanism of formation of the colloidal complexes is similar to that driving phase ˚ Values for Nagg separation, and thus take Mic = 0.5, and RMic = 20 A. of 0.5 wt% PAM122 PAA58 /DTAB at CNaBr = 0 mM are 128 for Z = 0.7, and 85 for Z = 1.0. On the bases of SANS results, a schematic representation of proposed structure of the colloidal complex formed by the association of PAM-b-PAA and DTAB is given in Fig. 10. 4. Conclusions

Fig. 10. Schematic representation of (a) DTA+ micelle decorated by PAM-b-PAA (Z < ZC ), and (b) a colloidal complex formed by the association of PAM-b-PAA and DTAB (ZC < Z).

PAM-b-PAA with low polydispersity was prepared by RAFT polymerization in methanol. DLS and SANS clearly indicates that the aqueous solution of PAM-b-PAA and DTAB associate into colloidal complexes. For low surfactant-to-polymer charge ratio Z lower than the critical value ZC , the colloidal complexes are single DTAB micelles dressed by a few PAM-b-PAA. Since DTAB concentration is low, these dressed micelles are very few and coexist with unassociated PAM-b-PAA. Above the critical value ZC , the colloidal complexes form a core–shell microstructure. The core of the complex consists of densely packed surfactant micelles (DTA+ ), and PAA block chains bind to these micelles, displace their counteranions (Br−) and bridge them. The corona of the complex is constituted from the PAM. A semi-quantitative interpretation of the scattering data allows us to determine the core dimension and the intermiceller distance inside the core. Since the interaction between polyelectrolyte and oppositely charged surfactant is primarily electrostatic in nature, the core radius and the intermicellar distance of the DTA+ micelles inside the core depend on CNaBr . The addition of salt screens the electrostatic attraction between oppositely charged PAA block and DTAB, which weakens the interaction. With increasing CNaBr , therefore, the core of the colloidal complex is considered to swell, which

M. Annaka / Colloids and Surfaces B: Biointerfaces 99 (2012) 127–135

leads to the increases in the core radius RC and the intermicellar distance of the DTA+ micelles inside the core. The aggregation number expressed in terms of DTA+ micelles per complex is also evaluated using the analogy with the homopolyelectrolyte/surfactant system. Acknowledgements The work was partly supported by a Grant-in-Aid (No. 22350053) and a Grant-in-Aid for Scientific Research on Priority Areas “Soft Matter Physics” (No. 19031024) and by a Grant-in-Aid for the Global COE Program, “Science for Future Molecular Systems” from the Ministry of Education, Culture, Science, Sports and Technology of Japan. This work is based on experiments performed at the Swiss spallation neutron source SINQ, Paul Scherrer Institute, Villigen, Switzerland (Proposal No. 20101224). References [1] T. Okano, Biorelated Polymers and Gels, Academic Press, Boston, 1998. [2] B. Lindman, K. Thalberg, in: E.D. Goddard, K.P. Ananathapadmanabhan (Eds.), Interactions of Surfactants with Polymers and Proteins, CPC Press, Boca Raton, 1993. [3] K. Thalberg, B. Lindman, K. Bergfelt, Langmuir 7 (1991) 3. [4] J. Klotz, S. Kosmella, T. Beitz, Prog. Polym. Sci. 26 (2001) 1199. [5] V.A. Kabanov, Russ. Chem. Rev. 74 (2005) 1. [6] S.J. Pispas, Phys. Chem. B 111 (2007) 8351. [7] C.L.F. Woodcock, L.L.Y. Frado, J.B. Rattner, J. Cell Biol. 99 (1984) 42. [8] H. Schiessel, J. Phys. Condens. Matter 15 (2003) R699. [9] A. Harada, K. Kataoka, Macromolecules 30 (1997) 7810. [10] P. Hansson, M. Almgren, J. Phys. Chem. 93 (1995) 16684. [11] C. Wang, K.C. Tam, Langmuir 18 (2002) 6484. [12] K. Hayakawa, J.C.T. Kwak, J. Phys. Chem. 86 (1982) 3866. [13] K. Hayakawa, J.C.T. Kwak, J. Phys. Chem. 87 (1983) 506. [14] A. Malovikova, K. Hayakawa, J.C.T. Kwak, J. Phys. Chem. 88 (1984) 1390. [15] C. Wang, K.C. Tam, R.D. Jenkins, C.B. Tan, J. Phys. Chem. B 107 (2003) 4667. [16] K. Thalberg, B. Lindman, J. Phys. Chem. 93 (1989) 1478. [17] A. Herslöf-Björling, M. Björling, L. Sundelöf, Langmuir 15 (1999) 353. [18] K. Thalberg, B. Lindman, K. Bergfelt, Langmuir 7 (1991) 2893. [19] M.A. Villetti, R. Borsali, J.S. Crespo, V. Soldi, K. Fukuda, Macromol. Chem. Phys. 205 (2004) 907.

135

[20] P. Hervé, M. Destarac, J.-F. Berret, J. Lal, J. Oberdisse, I. Grillo, Europhys. Lett. 58 (2002) 912. [21] J.-F. Berret, P. Hervé, O. Aguerre-Chariol, J. Oberdisse, J. Phys. Chem. 107 (2003) 8111. [22] J.-F. Berret, B. Vigolo, R. Eng, P. Hervé, I. Grillo, L. Yang, Macromolecules 37 (2004) 4922. [23] T. Imae, S. Ikeda, J. Phys. Chem. 90 (1986) 5216. [24] E. Roelants, F.C.D. Schryver, Langmuir 3 (1987) 209. [25] M.R. Böhmer, L.K. Koopal, J. Lyklema, J. Phys. Chem. 95 (1991) 9569. [26] B. Lindman, H. Wennerström, Top. Curr. Chem. 84 (1980) 744. [27] S. Hayashi, S. Ikeda, J. Phys. Chem. 84 (1980) 744. [28] A.V. Kavanov, T.K. Boronich, V.A. Kavanov, K. Yu, A. Eisenberg, J. Am. Chem. Soc. 120 (1998) 9941. [29] P. Hansson, Langmuir 17 (2001) 4167. [30] K. Hayakawa, J.P. Snterre, J.C.T. Kwak, Macromolecules 16 (1983) 1642. [31] J. Chiefari, Y.K. Chong, F. Elcore, J. Krstina, J. Jeffery, T.P.T. Le, R.T.A. Mayadunne, G.F. Meijs, C.L. Moad, G. Moad, E. Rizzardo, S.H. Thang, Macromolecules 31 (1998) 5559. [32] S. Perrier, C. Barner-Kowollik, J.F. Quinn, P. Vana, T.P. Davis, Macromolecules 35 (2002) 8300. [33] P. Strunz, K. Mortensen, S. Janssen, Physica B 350 (2004) e783. [34] B.L. Bales, R. Zana, J. Phys. Chem. B 106 (2002) 1926. [35] J.-F. Berret, J. Chem. Phys. 123 (2005) 164703. [36] M. Nagasawa, T. Murase, K. Kondo, J. Phys. Chem. 69 (1965) 4005. [37] F.A. Plamper, H. Becker, M. Lanzendörfer, M. Patel, A. Wittermann, M. Ballauff, A.H.E. Müller, Macromol. Chem. Phys. 206 (2005) 1813. [38] V. Bütün, S.P. Armes, N.C. Billingham, Z. Tuzar, A. Rankin, J. Eastoe, R.K. Heenan, Macromolecules 34 (2001) 1503. [39] J.-F. Berret, G. Cristobal, P. Hervé, J. Oberdisse, I. Grillo, Eur. Phys. J. E 9 (2002) 301. [40] J. Xia, H. Zhang, D.R. Rigsbee, P.L. Dubin, T. Shaikh, Macromolecules 26 (1993) 2759. [41] D. Chu, J.K. Thomas, J. Am. Chem. Soc. 108 (1986) 6270. [42] M. Filali, R. Aznar, M. Svenson, G. Porte, J. Appell, J. Phys. Chem. B 103 (1999) 7293. [43] P. Ilekti, L. Piculell, F. Tournilhac, B. Cabene, J. Phys. Chem. B 102 (1998) 344. [44] P. Ilekti, T. Martin, B. Cabene, L. Piculell, J. Phys. Chem. B 103 (1999) 9831. [45] A. Svensson, L. Piculell, P. Ilekti, B. Cabene, J. Phys. Chem. B 106 (2002) 1013. [46] R.M. Claesson, M. Bergström, A. Dedinaite, M. Kjellin, J.-F. Legrand, I. Grillo, J. Phys. Chem. B 104 (2000) 11689. [47] E. Sokolov, F. Yeh, A. Khokhlov, V.Y. Grinberg, B. Chu, J. Phys. Chem. B 102 (1998) 7091. [48] J.M. Seddon, J. Robins, T. Gulick-Krywicki, H. Delacroix, Phys. Chem. Chem. Phys. 2 (2000) 4485.