Colloid Polym Sci (2009) 287:1295–1304 DOI 10.1007/s00396-009-2096-y

ORIGINAL CONTRIBUTION

Polypeptide hybrid copolymers as selective micellar nanocarriers in nonaqueous media Guillermo Orts Gil & Sylvain Prévost & Magdalena Łosik & Florian Hermes & Helmut Schlaad & Thomas Hellweg

Received: 5 March 2009 / Revised: 13 July 2009 / Accepted: 13 August 2009 / Published online: 15 September 2009 # Springer-Verlag 2009

Abstract The self-assembly of polystyrene-block-poly (L-lysine) (PS-PLLys·HCl) copolymers with different block lengths has been studied in toluene. The obtained spherical micelles exhibit size variations upon addition of acids or bases, as indicated by light and neutron scattering studies. It is shown that pyridine induces a shrinking of the polystyrene chains in the corona region of the micelles, decreasing the aggregate solvent interface. The addition of benzoic acid, on the other hand, leads to a swelling of the copolymer micelles proportional to the molar fraction of polypeptide. This behavior suggests a selective permeability of the PS-PLLys micelles and the possibility to encapsulate G. Orts Gil : S. Prévost Stranski-Laboratorium für Physikalische und Theoretische Chemie, Institut für Chemie, Technische Universität Berlin, Strasse des 17. Juni 124, 10623 Berlin, Germany M. Łosik : F. Hermes : H. Schlaad (*) Max-Planck-Institut für Kolloid-und Grenzflächenforschung, Wissenschaftspark Golm, 14424 Potsdam, Germany e-mail: [email protected] S. Prévost Helmholtz-Zentrum-Berlin, Glienicker Str. 100, 14109 Berlin, Germany T. Hellweg (*) Universität Bayreuth, Physikalische Chemie I, Universitätsstr. 30, 95447 Bayreuth, Germany e-mail: [email protected] Present Address: G. Orts Gil Bundesanstalt für Materialforschung und -prüfung (BAM), Unter den Eichen 87, 12205 Berlin, Germany

organic compounds in toluene depending on their chemical nature. Keywords Block copolymer . Biohybrid . SANS . DLS

Introduction One of the most important features of amphiphilic block copolymers is their ability to self-assemble and to form a plethora of different structures in bulk systems as well as in solutions [1–3]. Also, with respect to technical applications, block copolymers are of increasing importance [4–6]. Another application is their use as templates in the synthesis of porous inorganic materials [7]. A major goal is to control the morphology or the structure in block copolymer solutions, which can be achieved by controlling the packing parameter. One approach to do this is mixing the block copolymer with surfactant in water [8–11]. Other rather easy ways to control the packing parameter are given by the variation of the total molecular weight, of the block composition and length, and of the molecular architecture and by the use of polymers with responsive blocks [12, 13]. In this context, there is a growing interest in the application of block copolymers comprising a biological block (e.g., polypeptides [14]) as selective nanocarriers [15]. Water is the most important solvent with respect to applications in life science like for instance drug delivery. Hence, many studies on block copolymers were done using aqueous systems, and less is known about nonaqueous systems [16]. In order to shed light on this topic, the present contribution deals with inverse micelles of polypeptide-based hybrid block copolymers in toluene. Poly(L-lysine) (PLLys) is an important water-soluble polypeptide which contains amine groups on the side

1296

Colloid Polym Sci (2009) 287:1295–1304

chains. Similar to natural peptides, the polypeptide block is charged and the degree of ionization depends on solution pH. PLLys can adopt three different conformations, i.e., random coil, α-helix, and β-sheet, depending on external parameters including pH, surfactant concentration, temperature, etc. [17]. In this paper, we describe the self-assembled structures formed by polystyrene-block-poly(L-lysine) (PS-PLLys·HCl) block copolymers in toluene and the selective encapsulation of basic and acidic compounds in these aggregates.

Experimental section Sample preparation Previous studies concerning block copolymers composed of PS-PLLys demonstrated that the molar ratio, as well as the total number of monomer units, plays an important role in determining the preferred structure in solution (micelles, vesicles, spheres, rod-like aggregates) [18] and the solubility in selective solvents [19]. Here, PS-PLLys block copolymers with three different block lengths were synthesized by anionic polymerization and subsequent ring opening polymerization of (Z)-L-lysine N-carboxyanhydride. The Z-protecting groups were removed by acidic hydrolysis. For details of the synthesis and the characterization, see [18–20]. The characteristics of the final products are listed in Table 1. Solutions of the copolymers were prepared by dissolving the dry polymer powder in toluene (99%, Merck) up to a concentration of 0.1 wt.%. The solutions were sonicated for 3 h (Sonorex super RK 52 H, with a max. power of 240 W) at high temperatures between 340 and 355 K. The obtained solutions were optically clear and stable. Based on dynamic light scattering (DLS) experiments, it was found that the outcome of the preparation procedure was reproducible within experimental precision. Always the same particle size distribution was obtained. Deuterated toluene (99.5%, Merck) was used for small-angle neutron scattering (SANS) measurements in order to generate the contrast with the hydrogenated polymer.

Table 1 Characteristics of the three studied polystyrene-block-poly (L-lysine) copolymers: molecular weight number average (Mn), mole fraction of lysine (xL-Lys), and polydispersity index (PDI) No.

Polymer

1 2 3

PS258-PLLys58•HCl PS258-PLLys107•HCl PS388-PLLys138•HCl

Mn/kg mol−1

xL-Lys

PDI

36.4 44.5 63.1

18 29 26

1.2 1.2 1.2

Dynamic light scattering (DLS) DLS measurements were carried out at a temperature of 20°C using a setup consisting of an argon ion laser (λ= 514.5 nm, P=600 mW), a goniometer with fiber-optic detector, and two photon counters (ALV, Langen, Germany). The signal is fed into a multiple-τ correlator (ALV5000/Fast) operated in the cross-correlation mode. DLS provides a fast noninvasive characterization of the dynamic modes in a polymeric sample. The normalized electrical field autocorrelation function, gl(τ), which contains the information about the dynamics of the scattering system, can be computed from the experimentally obtained intensity time autocorrelation function g2(τ) by the Siegert relation [21]. In order to account for polydispersity of the samples, the correlation functions obtained in the experiment have to be described by a weighted sum of exponentials [22, 23].

Z1 g ðt Þ ¼

Gð*Þ expð
*t Þd *

1

ð1Þ

0

Here, G(Γ) is the distribution function of relaxation rates. For monodisperse samples, G(Γ) becomes a δ-function, and a single exponential decay is obtained from Eq. 1. An analysis of the distribution of relaxation rates can be performed using an inverse Laplace transform of gl(τ), which can be computed by applying the FORTRAN program CONTIN [22, 24, 25]. From the mean value of the relaxation rate, one obtains the translational diffusion coefficient according to Γ=Dq2 [21], and the hydrodynamic radius Rh can be calculated making use of the Stokes– Einstein equation [26] D¼

kT 6phRh

ð2Þ

with η being the viscosity of the solvent, k the Boltzmann constant, and T the temperature. It is often assumed that the size polydispersity follows a logarithmic normal distribution in DLS. Hence, all moments of the distribution function can be easily calculated analytically [27, 28]. DLS measures a z-averaged radius which is proportional to the square of its volume, i.e., proportional to R6 [29]. This means that Rh obtained by DLS can be influenced by only a few large particles leading to an overestimation of the real particle size. An alternative to obtain a more realistic particle size in polydisperse samples is the use of depolarized DLS, but due to the low depolarized scattering intensities this technique normally requires high-power laser sources [19].

Colloid Polym Sci (2009) 287:1295–1304

1297

For weakly interacting particles and low concentrations, the concentration dependence of the diffusion coefficient is usually expressed as [30]: D ¼ D0 ð1 þ kD cM Þ

ð3Þ

where D0 is the diffusion coefficient at infinite dilution, and cM is the molar concentration. The proportionality constant kD is called the dynamical virial coefficient and can be written as [31]: kD ¼ 2MA2
kf
2u

ð4Þ

where A2 is the second osmotic virial coefficient; kf is the relative concentration dependence of the friction coefficient, and v is the partial specific volume of the solute. The last two terms are usually much smaller than the first, and the sign of A2 provides information about the pair interaction between the particles [32].

the ideal case where nontransmitted neutrons are scattered uniformly over the full solid angle of 4π; this method is known to give inaccurate scaling but is easily performed and leads to reproducible results; therefore, we expect a small systematic coefficient between the theoretical and the experimental absolute intensity. Reduced data were always isotropic and were consequently azimuthally averaged, and spectra from different configurations for each sample were merged with no need of any arbitrary scaling factor. Theoretical models for SANS data analysis For a first preliminary evaluation of the data, a simple Guinier analysis was applied [35, 36]. However, to achieve a full description of the data over the total measured q range, two more sophisticated models were employed. Model 1: compact spheres + Ornstein Zernicke The scattered intensity for noninteracting spheres is expressed as:

Atomic force microscopy (AFM) Samples were prepared by drop deposition onto a hydrophilic substrate (silicon) and evaporation of the solvent in a dust-protected environment. The silicon wafers were cleaned prior to use following the standard RCA-1 method [33]. A silicon cantilever (f=300 KHz) was mounted onto tapping mode cantilever holder. The surfaces were analyzed using a Nanoscope III Multimode Scanning Probe Microscope. Small-angle neutron scattering (SANS) SANS data acquisition and reduction SANS spectra were accumulated on the V4 small-angle scattering machine, at the BER Reactor of the HelmholtzZentrum Berlin (HZB), Germany. Neutrons were recorded on a two-dimensional 3He gas detector of 128×128 pixels of 5×5 mm2. Wavelength was at 0.605 nm (QFWHM 10.5%). Sample-to-detector distances of 1, 4, and 16 m were selected with collimations at, respectively, 4, 8, and 16 m, covering a q range from 0.03 nm−l up to 3.6 nm−l. Samples were kept in 2-mm pathway quartz cuvettes (QS, Hellma, Germany), with an illuminated area of 7.5-mm diameter. Data reduction was performed on 2D patterns using the BerSANS software package from the HZB [34]. Raw data were corrected for the scattering of the empty cell; pixel efficiency and solid-angle variations were taken into account by dividing with the incoherent scattering pattern of pure water kept in a 1-mm cuvette. Background noise was furthermore accounted for by measurements with cadmium at the sample position. Absolute scale calibration was done by measuring the transmission of water, assuming

Isphere ðq; R; $hÞ ¼ fV $h2 3

sin qR
qR cos qR ðqRÞ3

!2 ð5Þ

where R is the particle radius, and ∆η is the scattering contrast between the particle and the solvent. The scattering contrast was calculated as the difference between the average scattering length densities of PS and PLLys and the deuterated toluene. The polydispersity was modeled by convolution of a logarithmic normal distribution of the particle size with Eq. 5. This choice of the distribution function is based on the DLS measurements showing this kind of distribution for the hydrodynamic radius in the present case. The Ornstein–Zernicke (OZ) contribution is added to model the high and intermediate q part of the experimental data. This approach was introduced for the concentration fluctuations in semidilute polymer solutions and was used here to describe the concentration fluctuations and the related correlation length in the diffuse corona of the block copolymer micelles [37, 38]. IðqÞ /

I0 1 þ q2 x 2

ð6Þ

where ξ defines the correlation length. Model 2: spherical micelles with dense homogeneous core and Gaussian corona chains A more complex model was used based on the approximate knowledge of the volumes of PS and PLLys blocks. This model describes block copolymer micelles (spherical), with a homogeneous core constituted by the poorly solvated units and Gaussian chains in the shell constituted by units for which the

1298

Colloid Polym Sci (2009) 287:1295–1304

solvent quality is good. The fitted parameters are the aggregation number Nagg (with a lognormal distribution) and the apparent radius gyration of the PS chains. The form factor of the micelle is written as a sum of four terms [39]: 2 Imic ¼ Nagg b2core Pcore ðqÞ þ Nagg b2brush Pbrush ðqÞ 2 þ 2Nagg bcore bbrush Sbrush
core ðqÞ
 þ N agg Nagg
1 b2brush S brush
brush ðqÞ

ð7Þ

Nagg is the aggregation number of diblock polymers forming the micelle and βbrush =Vbrush(ηbrush −ηsolv) and βcore =Vcore(ηcore −ηsolv) the excess scattering length of a block in the corona and in the core, respectively. Vbrush and Vcore are the total volume of a block in the corona and in the core. ηbrush and ηcore are the corresponding scattering length densities, and ηsolv is the scattering length density of the surrounding solvent. The corresponding form factors for the core and brush are defined as: Pcore ðq; RÞ ¼ Φ2 ðqRÞ

ð8Þ

where

ΦðqRÞ ¼ 3

sinðqRÞ
qR cosðqRÞ

! ð9Þ

ðqRÞ3

and Pbrush








expð
xÞ
1 þ x q; Rg ðshell Þ ¼ 2 x2

 ð10Þ

with x = R2g (shell)q2. Sbrush−brush and Sbrush−core are functions of (q, R, Rg (shell), d). For more details about this model, please see [40]. Also, other more sophisticated models for block copolymer micelles were suggested [41, 42], which in the present case are difficult to apply, probably due to the polydispersity of the micelles prepared by sonication.

Fig. 1 Mean relaxation rates calculated by CONTIN at different scattering angles for the copolymers 1 (empty squares), 2 (filled circles), and 3 (empty triangles) in toluene. From the slope of Γ vs q2, the translational diffusion coefficient is obtained and, hence, the hydrodynamic radius can be calculated

respectively. For all three samples, a rather high polydispersity is observed. For instance, in case of polymer 1, the influence of the polymer concentration on the aggregation behavior was probed by DLS, measuring solutions of sample 1 at concentrations between 0.0025 and 0.08 wt.%. In this range of concentrations, a slight linear decrease of the diffusion coefficient with the concentration is found (Fig. 2). As explained in the previous sections, the dependence of the diffusion coefficient on the concentration provides information about the interaction between particles in solution. In this case, the second osmotic virial coefficient (A2) is negative, which indicates an attractive interaction between the micelles in solution (see Eq. 4).

Results and discussion Micellar structure in toluene DLS Solutions of the copolymers 1–3 at 0.1 wt.% in toluene were prepared (“Sample preparation”) and investigated using DLS. For all samples, five scattering angles were studied. In Fig. 1, the respective mean relaxation rates Γ for all three polymers are plotted vs. q2. The diffusion coefficients were determined from the slopes of the plots, and the average hydrodynamic radii were found to be Rh =49, 59, and 68 nm, for the samples 1, 2, and 3,

Fig. 2 Dependence of the diffusion coefficient for micelles of PS258-b- PLLys58 (sample 1) in toluene on the polymer concentration. The negative slope indicates an attractive interaction between the particles in solution

Colloid Polym Sci (2009) 287:1295–1304

AFM Atomic force microscopy (AFM) was used, in addition to the DLS experiments, to study the shape of the formed micelles after deposition on a hydrophilic silicon substrate. Since the deposition of the micelles is done using a solution in toluene, the PS blocks are expected to form the outer part of the self-assembled structures. Therefore, on a hydrophilic substrate, the micelles should not wet the surface. Figure 3 shows an AFM image obtained for polymer 1 on a silicon wafer surface. Full wetting of the surface did not occur, and apparently spherical particles with radii in the range from 15 to 27 nm are found. This is significantly smaller compared to the DLS results. However, in DLS experiments, the observed “size” is usually rather large due to the attached solvent and also due to the fact that the scattered intensity is dominated by larger particles (I ∝ R6; see also [29, 43]). In the present case, the AFM images were taken in the dry state and most of the toluene, which is swelling the PS corona in solution, is evaporated. An analysis of the height profile leads to values of approximately 3 nm and, hence, the micelles adsorb in a rather flat shape on the hydrophilic Si surface. However, the polymer does not spread over the silicon, and no film is formed as was found in our previous study dealing with mixed micelles of polymer 3 in water in presence of surfactants on graphite [19]. In this previous work, it turned out that the block copolymer spreads to the substrate when the hydrophilicity or hydrophobicity of the “wetting” block is similar compared to the used substrate. SANS

1299

by SANS. A Guinier analysis of the experimental points can be performed in the low q region of the spectra [35, 36]. In Fig. 4, the linear fit of the experimental SANS data by the Guinier approximation (logarithmic form) for the three copolymers is shown (the intensity is scaled to compare the curves). From the slope of the curves, the Guinier radius Rg was determined to be 27, 33, and 36 nm for samples 1, 2, and 3, respectively. This observed increase is in good agreement with the growing polymer contour length. Based on the DLS and SANS results, ρ parameters (ρ=Rg/Rh) between 0.53 and 0.56 are computed. This is much lower than the value for a compact sphere and indicates that large PS-b-PLLys copolymers self-assemble into loosely packed spherical micelles in solution [44]. The PS corona is strongly swollen and fuzzy, since toluene is a good solvent for this block under the chosen experimental conditions. A complete description of the SANS scattering curves was achieved by fitting of the experimental scattering intensity data using the program SASfit written by Joachim Kohlbrecher (PSI, Switzerland) [40]. In a first attempt, model 1 described in “Model 1: compact spheres + Ornstein–Zernicke” was used to analyze the experimental scattering curves. The use of the simple form factor for spheres is justified since the deuterated solvent strongly penetrates into the PS corona and, hence, the main contrast is obtained between the core and toluene. The OZ contribution clearly improves the fits at high q. The SANS curves for polymers 1, 2, and 3 and the corresponding fits with the model 1 are shown in Fig. 5 (left). The fits (black lines) describe reasonably well the

Solutions of the copolymers 1 to 3 in deuterated toluene at a copolymer concentration of 0.1 wt.% were also measured

Fig. 3 AFM image of apparently spherical micelles of PS258-b-PLLys58 (sample 1) on silicon

Fig. 4 Guinier plots of the experimental SANS data of the copolymers 1 (empty squares), 2 (filled circles), and 3 (empty triangles) at low scattering angle. From the slope of the linear fits, the Guinier radius is calculated

1300

Colloid Polym Sci (2009) 287:1295–1304

Fig. 5 SANS curves for copolymers 1, 2, and 3 in deuterated toluene: experimental data (empty squares) and fits (solid lines). a Curves fitted using model 1. b Curves fitted using model 2. The absolute intensities are given in arbitrary units because the curves were multiplied by multiples of ten to achieve a better representation of the data

experimental data (empty squares) except for sample 3 in the intermediate q range. In order to improve the analytical fits, a second model (model 2) comprising a spherical core with attached Gaussian coils was used. However, for polymer 3, the description of the data is not significantly improved using this model compared to model 1 (see Fig. 5, right). The results of the fits are shown in Table 2. The description of the SANS curves is not perfect but is the best possible with all applied models. The apparent discrepancy between the Guinier radii and the dimensions obtained by using models 1 and 2 can be understood in the following way: both models provide a core radius but no shell thickness. The shell is treated as a kind of semidilute polymer solution, and only a correlation length or “blob size” is computed. However, with the used models, it is not possible to estimate the number of “blobs” needed to determine the shell thickness. All results from the different fits are summarized in Tables 2 and 3. Principal models like the ones proposed in [42] and [41] should be closer to a block copolymer micelle structure. However, in the present case, the fits using these models were not satisfying.

Table 2 Results of the fits of the SANS curves for copolymer samples 13 in deuterated toluene polymer 1 2 3

R [nm] 13.2 15.2 26.0

s 0.48 0.45 0.31

ξ [nm]

Nagg

Rg (shell) [nm]

2.7 4.8 3.0

201 558 1,264

2.28 10 2.56

R is the particle core radius as defined by Eq. 5, s the polydispersity, and ξ the correlation length in the Ornstein–Zernicke contribution (model 1). Nagg is the aggregation number, and Rg (shell) is the radius of gyration of the PS chains in the particle shell obtained using by the model 2

Addition of pyridine As explained in the “Introduction,” the response of PLLys in water to pH changes is well known, but nothing is known about the behavior in other solvents. In toluene, PS forms the shell of the PS-PLLys micelles. Therefore, the addition of a base capable to penetrate into the PS-PLLys micelles might induce changes in the PLLys block conformation. This could lead to changes of the packing parameter and, hence, affect the micellar size, shape, or aggregation number. Therefore, solutions of the polymer 1 at concentration of 0.1 wt.% were prepared in deuterated toluene (tol-D8) and also in mixtures tol-D8/pyridine (5%, 10%, and 20% in volume; Merck, 99%) by sonication and investigated by SANS and DLS. Fourier transform infrared (FTIR) experiments showed that pyridine interacts with PS-PLLys micelles (data not shown, see [45]). On the other hand, the analysis of the intensity time correlation functions obtained from DLS experiments reveals that the radius of the micelles of polymer 1 decreases in the presence of pyridine by approximately 6% (Table 3). However, this

Table 3 Results of the fits of SANS curves for copolymer 1 in deuterated toluene containing 0, 5, 10, 20 vol.% pyridine polymer 1 1 (5) 1 (10) 1 (20)

R [nm] 13.2 12.7 14.0 11.0

s 0.48 0.51 0.44 0.52

ξ [nm]

Nagg

Rg (shell) [nm]

2.7 1.86 1.47 0.58

558 751 714 1,264

2.20 2.08 2.23 0.61

R is the particle core radius as defined by Eq. 5, s the polydispersity, and ξ the correlation length in the Ornstein–Zernicke contribution from the fits of the SANS data using model 1. Nagg is the aggregation number and Rg (shell) is the apparent radius of gyration of the PS chains in the micellar corona as obtained from model 2

Colloid Polym Sci (2009) 287:1295–1304

1301

Fig. 6 SANS curves for copolymer 1 at different pyridine contents (percent in volume in parenthesis). The empty squares correspond to the experimental points and the dashed lines to the fits. c Fits with model 1. d Fits according to model 2

1. Assuming that all pyridine molecules stay in the solution and contribute to the scattering contrast, we also assume that the polydispersity and particle number remain constant at any pyridine content. 2. No assumptions about the scattering contrast are made, and the contrast is just treated as another adjustable parameter.

Addition of benzoic acid Also, the effect of the addition of benzoic acid to solutions of the copolymer in toluene has been studied by DLS. Solutions of the copolymer in toluene at concentration 0.1 wt.% were prepared, and different amounts of crystalline benzoic acid were added. In order to calculate the hydrodynamic radius of the aggregates in toluene in presence of benzoic acid, the absolute values for the solvent viscosity have been corrected. This information is only available for benzoic acid mixtures in water [46]. For this reason, it was necessary to experimentally determine the influence of the benzoic acid content on the viscosity of the toluene solutions (data can be found in [45]). This means that the

R g (shell) / nm

The comparison of the calculated and the fitted scattering contrast from the second analysis procedure shows that the values are in good agreement. Hence, it seems to be straightforward to conclude that the pyridine molecules stay in the solution and do not significantly penetrate the micelles. Figure 6 shows the SANS curves for solutions of copolymer 1 in deuterated toluene at different pyridine contents. The incoherent scattering intensity increases with the pyridine content since pyridine is not deuterated. Moreover, at pyridine content of 20%, the intensity at low q values also changes. This can be related to the fact that the PS corona is less swollen and therefore the amplitude of this contribution goes down. This is in line with the observation that the correlation length ξ obtained from the fits with model 1 for the SANS curves shows a decrease with pyridine content. This is also consistent with the observed decrease of the Rg (shell) value as obtained from model 2 (Fig. 7). Therefore, it can be concluded that the addition of pyridine leads to a shrinkage of the PS corona of the block copolymer micelles and a decrease of the particle size. This fact is also in good agreement with the

DLS observation and can be related to pyridine-induced changes of the solvent quality.

ξ / nm

does not mean that the pyridine goes into the PLLys core and induces a conformational change. Actually, pyridine appears to be a too weak base to deprotonate the PLLys to a large extent. A change in the solvent quality can fully explain why the size of the particles is going down upon addition of pyridine to solutions in toluene. The same solutions were also studied using small-angle neutron scattering. Now, the fit of the SANS curves was done in two different ways:

Fig. 7 Decrease of the correlation length ξ of the micellar corona (empty squares) and apparent radius of gyration of the PS chains in the shell as obtained using model 2 (filled squares) for polymer sample 1 at different pyridine contents. Pyridine changes the solvent quality leading to a shrinkage of the micelles

1302

viscosity values used for the calculation of the hydrodynamic radius correspond to the case that all benzoic acid molecules stay in the bulk solvent and not in the PS-PLLys micelles. At low benzoic acid concentrations, an increase of the hydrodynamic radius with the benzoic acid content is found, suggesting that the acid can penetrate in the core of the copolymer micelles leading to a swelling effect. However, for benzoic acid concentrations around 0.6 mol/L, the hydrodynamic radius of the aggregates remains almost constant, and phase separation between the organic phase and solid benzoic acid is observed. This is consistent with the limiting solubility for benzoic acid in toluene found in the literature [47]. This means that the solubility of benzoic acid in toluene is the limiting parameter for the micellar growth. In Fig. 8, the increase of the aggregates hydrodynamic radius with the benzoic acid content is shown for the three different copolymers. Obviously, the maximal growth in size of the aggregates is related to the PLLys content in the copolymer chain. This fact supports the idea that benzoic acid molecules are able to penetrate in the micellar core (PLLys) leading to swelling of the aggregates, due to electrostatic interactions. However, evidence for a change of conformation was not found.

Colloid Polym Sci (2009) 287:1295–1304

When pyridine is added to solutions of the copolymers in toluene, it is preferentially located around the PS-PLLys micelles (FTIR measurements in [45]). Light scattering measurements show that the size of the PS-PLLys micelles decreases in the presence of pyridine. This suggests that pyridine changes the solvent quality, yielding a shrinking of the polystyrene chains in the shell of micelles. SANS measurements corroborate this, showing that the correlation length as computed using the Ornstein–Zernicke contribution in the applied model 1 decreases with increasing pyridine concentration. However, pyridine is a too weak a base to significantly penetrate the micellar core and to induce changes of the degree of protonation and the secondary structure of PLLys. Light scattering measurements show that the addition of benzoic acid to solutions of the copolymers in toluene leads to a swelling of the micelles proportional to the polypeptide molar fraction in the block copolymer. If one assumes no significant difference in molecular volumes between benzoic acid and pyridine, a selective permeability of PS-PLLys micelles to external species depending on the chemical nature can be deduced and the PLLys core can be selectively swollen with acidic organic compounds. The size effect found upon addition of benzoic acid points to the fact that rather considerable payloads can be achieved.

Conclusion Large polystyrene-block-poly(L-Lysine) copolymers selfassemble into spherical micelles in toluene at all studied polypeptide molar fraction and concentrations.

Acknowledgments We wish to thank the DFG (Sfb 448, TP A12, and BIOSONS) for financial support. T.H. also would like to acknowledge financial support from the Sfb 481 (TP A15).

Appendix A: additional sample parameters Average molecular volumes of the polymers and the PLLys and PS blocks are displayed in Table 4. Assuming monomeric micelles with a dense PLLys core and a PS shell, the PLLys cores having volumes lying between 12 and 30 nm3 would fit into spheres of radii between 1.4 and 2 nm, and the whole “dry” polymer would fit into spheres of radii between 2.4 and 2.9 nm.

Table 4 Composition of the PS-b-PLLys copolymers 1-3. n is the average number of repeat units, v is the molecular volume or volume fraction, and m is the mass fraction

Fig. 8 Increase of the micellar size with benzoic acid content for samples 1 (empty squares), 2 (filled circles), and 3 (empty triangles). The maximal micellar growth is proportional to the polypeptide molar fraction in the copolymer suggesting that benzoic acid can penetrate into the micellar core. The dashed line indicates the solubility limit of benzoic acid in toluene

1 2 3

nPLLys

nPS

ntot

vPLLys [nm3]

vPS [nm3]

vtot [nm3]

vPLLys [%vtot]

mPLLys [%mtot]

58 107 138

258 258 388

316 365 526

13.7 25.3 32.6

42.4 42.4 63.8

56.1 67.7 96.4

24 37 34

26 40 36

Colloid Polym Sci (2009) 287:1295–1304

1303

Appendix B: scattering length densities and molecular volumes The values from Table 5 were used for the fits. From Tables 4 and 5, we can calculate the average SLD of the whole polymers, at 1.31, 1.25, and 1.27·10 l0 cm−2. The average value of 1.3·10 l0 cm−2 will be used.

12.

13. 14.

15. Table 5 Scattering length densities (SLD) and apparent molecular volumes (vm) of the materials Material

d-Toluene Pyridine PLLys PS

Formula

C7 D8 C5 H5N [C6 Hl2 N2O·HCl]n [C8 H8]n

Density [g cm−3] 0.946 0.982 1.157 1.050

vm [nm3] 0.1755 0.0806 0.2362·n 0.1645·n

SLD [10l0 cm−2] 5.68 1.79 1.07 1.41

References 1. Förster S (1997) Amphiphilic block copolymers. Ber. Bunsenges. Phys Chem 101(11):1671–1678 2. Bahadur P (2001) Block copolymers—their microdomain formation (in solid state) and surfactant behaviour (in solution). Curr Sci 80:1002–1007 3. Langevin D (1992) Micelles and microemulsions. Annu Rev Phys Chem 43:341–369 4. Jakobs B, Sottmann T, Strey R, Allgaier J, Willner L, Richter D (1999) Amphiphilic block copolymers as efficiency boosters for microemulsions. Langmuir 15:6707–6711 5. Endo H, Mihailescu M, Monkenbusch M, Allgaier J, Gompper G, Richter D, Jakobs B, Strey R, Grillo I (2001) Effect of amphiphilic block copolymers on the structure and phase behavior of oil-water-surfactant mixtures. J Chem Phys 115(1):580–600 6. Holderer O, Frielinghaus H, Byelov D, Monkenbusch M, Allgaier J, Richter D (2005) Dynamic properties of microemulsions modified with homopolymers and diblock copolymers: the determination of bending moduli and renormalization effects. J Chem Phys 122:094908 7. Monnier A, Schüth F, Huo Q, Kumar D, Margolese D, Maxwell RS, Stucky GD, Krishnamurty M, Petroff P, Firouzi A, Janicke M, Chmelka BF (1993) Cooperative formation of inorganic–organic interfaces in the synthesis of silicate mesostructures. Science 261:1299–1303 8. Hecht E, Hoffmann H (1994) Interaction of ABA block copolymers with ionic surfactants in aqueous solution. Langmuir 10:86–91 9. Nordskog A, Egger H, Findenegg GH, Hellweg Th, Schlaad H, von Berlepsch H, Böttcher C (2003) Structural changes of diblock PB40 PEO62 -copolymer micelles induced by addition of low molecular weight surfactants: Scattering and cryo-TEM studies. Phys Rev E 68:11406 10. Fütterer T, Hellweg Th, Findenegg GH, Frahn J, Schlüter AD, Böttcher C (2003) Self-assembled structures formed by amphiphilic polyparaphenylenes in the bulk, at the water/air interface and in non-ionic surfactant solutions. Langmuir 19:6537–6544 11. Nordskog A, Fütterer T, Schlaad H, Heinemann A, von Berlepsch H, Böttcher C, Hellweg Th (2004) Formation of mixed micelles of

16. 17.

18.

19.

20.

21. 22.

23.

24. 25.

26. 27.

28.

29.

30.

31.

32.

33.

PB40PEO62 and the anionic surfactant SDS in aqueous solution. Phys Chem Chem Phys 6:3123–3129 Bates FS, Fredrickson GH (1990) Block copolymer thermodynamics: theory and experiment. Annu Rev Phys Chem 41:525– 557 Gil ES, Hudson SM (2004) Stimuli-responsive polymers and their bioconjugates. Prog Polym Sci 29(12):1173–1222 Lecommandoux S, Achard M-F, Langenwalter JF, Klok H-A (2001) Self-assembly of rod-coil diblock oligomers based on α-helical peptides. Macromolecules 34:9100–9111 Matsumura Y (2008) Poly (amino acid) micelle nanocarriers in preclinical and clinical studies. Advanced Drug Delivery Reviews 60:899–914 Ponsinet V, Fabre P (1996) Flexibility of the membranes in a doped swollen lamellar phase. J Phys Chem 100:5035–5038 Wang Y, Chang Y (2003) Synthesis and conformational transition of surface-tethered polypeptide: poly(L-lysine). Macromolecules 26:6511–6518 Sigel R, Losik M, Schlaad H (2007) pH responsiveness of block copolymer vesicles with a polypeptide corona. Langmuir 23:7196–7199 Orts Gil G, Losik M, Schlaad H, Hellweg T (2008) Properties of pH-responsive mixed aggregates of polystyrene-block-poly (L-lysine) and nonionic surfactant in solution and adsorbed at a solid surface. Langmuir 24:12823–12828 Schlaad H, Smarsly B, Losik M (2004) The role of chain-length distribution in the formation of solid-state structures of polypeptide-based rod-coil block copolymers. Macromolecules 37:2210–2214 Berne BJ, Pecora R (1976) Dynamic light scattering. Wiley, New York Provencher SW (1982) A constrained regularization method for inverting data represented by linear algebraic or integral equations. Comput Phys Com 27:213–217 Hellweg T, Brûlet A, Sottmann T (2000) Dynamics in an oilcontinuous droplet microemulsions as seen by quasielastic scattering techniques. Phys Chem Chem Phys 2(22):5168–5174 Provencher SW (1976) A Fourier method for the analysis of exponential decay curves. Biophys J 16:27–41 Provencher SW (1982) CONTIN: a general purpose constrained regularization program for inverting noisy linear algebraic and integral equations. Comput Phys Com 27:229–242 Einstein A (1956) Investigation on the theory of the Brownian movement. Dover, New York Thomas JC (1987) The determination of log normal particle size distributions by dynamic light scattering. J Colloid Interf Sci 117:187–192 Diamond S, Dolch W (1972) Generalized log-normal distribution of pore sizes in hydrated cement paste. J Colloid Interface Sci 38:234–244 Rübe A, Hause G, Mäder K, Kohlbrecher J (2005) Core-shell structure of Miglyol/poly(D, L-lactide)/poloxamer nanocapsules studied by small-angle neutron scattering. J Controlled Release 107:244–252 Mes E, Kok W, Poppe H, Tijssen R (1999) Comparison of methods for the determination of diffusion coefficients of polymers in dilute solutions the influence of polydispersity. J Polymer Sci Part B: Polym Phys 37:593–603 Vink H (1985) Mutual diffusion and self-diffusion in the frictional formalism of non-equilibrium thermodynamics. J Chem Soc Faraday Trans I 81:1725–1730 Evans D, Wennerström H (1994) The colloidal domain: where physics, chemistry, biology, and technology meet. VCH, New York Bachman M (1999) RCA-1 silicon wafer cleaning. Technical report, INRF, California

1304 34. Keiderling U (2002) The new BerSANS-PC software for reduction and treatment of small angle neutron scattering data. Appl Phys A Mater Sci Process 74:1455–1457 35. Guinier A, Fournet G (1955) Small-angle scattering of X-rays. Wiley, New York 36. Glatter O, Kratky O (1982) Small angle X-ray scattering. Academic, London 37. de Gennes P-G (1979) Scaling concepts in polymer physics. Cornell University Press, Ithaca 38. Hellweg Th, Schemmel S, Rother G, Brûlet A, Eckerlebe H, Findenegg GH (2003) Demixing dynamics of a binary liquid mixture in a controlled pore glass. Eur Phys J E 12:S1–S4 39. Pedersen JS, Gerstenberg MC (1996) Scattering form factor of block copolymer micelles. Macromolecules 29:1363–1365 40. Kohlbrecher J (2008) SASfit: a program for fitting simple structural models to small angle scattering data. Paul Scherrer Institut, Laboratory for Neutron Scattering, Villigen 41. Svaneborg C, Pedersen JS (2001) Block copolymer micelle coronas as quasi-two-dimensional dilute or semidilute polymer solutions. Phys. Rev. E 64(1):10802–10805

Colloid Polym Sci (2009) 287:1295–1304 42. Iatrou H, Hadjichristidis N, Meier G, Frielinghaus H, Monkenbusch M (2002) Synthesis and characterization of model cyclic block copolymers of styrene and butadiene. Comparison of the aggregation phenomena in selective solvents with linear diblock and triblock analogues. Macro-molecules 35(14):5426–5437 43. Khlebtsov NG (2003) On the dependence of the light scattering intensity on the averaged size of polydisperse particles: comments on the paper by M. S. Dyuzheva et al. (Colloid J2002, vol. 64, no. 1, p. 39). Colloid Journal 65(5):652–655 44. Burchard W, Richtering W (1989) Dynamic light scattering from polymer solutions. Prog Colloid Polymer Sci 80:151–163 45. Orts Gil G (2008) Self-assembly and dynamics of complex amphiphilic systems: from block copolymers to phospholipids. Ph. D. thesis, Technische Univer-sität Berlin, Stranski-Laboratorium für Physikalische Chemie, Berlin. 46. Sun T, Teja A (2004) Density, viscosity, and thermal conductivity of aqueous benzoic acid mixtures between 375 K and 465 K. J Chem Eng Data 49:1843–1846 47. Chipman J (1924) The solubility of benzoic acid in benzene and in toluene. J Am Chem Soc 46:2445–2448