Control of the PEO Chain Conformation on Nanoparticles by

ARTICLE pubs.acs.org/Langmuir

Control of the PEO Chain Conformation on Nanoparticles by Adsorption of PEO-block-Poly(L-lysine) Copolymers and Its Significance on Colloidal Stability and Protein Repellency St
ephanie Louguet,†,‡,§ Anitha C. Kumar,†,‡,^ Nicolas Guidolin,†,‡ Gilles Sigaud,|| Etienne Duguet,§ S
ebastien Lecommandoux,*,†,‡ and Christophe Schatz*,†,‡ †

Univ. Bordeaux/IPB, ENSCBP, 16 avenue Pey Berland, 33607 Pessac, France CNRS, Laboratoire de Chimie des Polymeres Organiques (UMR5629), Pessac, France § CNRS, Univ. Bordeaux, ICMCB, 87 Avenue du Dr Albert Schweitzer, 33608 Pessac, France CNRS, Univ. Bordeaux, CRPP, 115 Avenue du Dr Albert Schweitzer, 33600 Pessac, France ^ Acharya Nagarjuna University, Nagarjuna Nagar, Guntur 522 510, Andhra Pradesh, India

)

‡

bS Supporting Information ABSTRACT: The physical adsorption of PEOn-b-PLLm copolymers onto silica nanoparticles and the related properties of poly(ethylene oxide) (PEO)-coated particles were studied as a function of the block copolymer composition. Copolymers adopt an anchor
buoy conformation at the particle surface owing to a preferential affinity of poly(L-lysine) (PLL) blocks with the silica surface over PEO blocks when a large excess of copolymer is used. The interdistance between PEO chains at particle surface is highly dependent on the size of PLL segments; a dense brush of PEO is obtained for short PLL blocks (DP = 10), whereas PEO chains adopt a so-called interacting “mushroom” conformation for large PLL blocks (DP = 270). The size of the PEO blocks does not really influence the copolymer surface density, but it has a strong effect on the PEO layer thickness as expected. Salt and protein stability studies led to similar conclusions about the effectiveness of a PEO layer with a dense brush conformation to prevent colloidal aggregation and protein adsorption. Besides, a minimal PEO length is required to get full stabilization properties; as a matter of fact, both PEO45-b-PLL10 and PEO113-b-PLL10 give rise to a PEO brush conformation but only the latter copolymer efficiently stabilizes the particles in the presence of salt or proteins.

’ INTRODUCTION The surface modification of nanoparticles (NPs) with poly(ethylene glycol) (PEG), also referred as poly(ethylene oxide) (PEO), has become a common strategy in drug delivery applications. The surface PEGylation accounts for an improved stability of the NPs in blood circulation and avoids their rapid clearance from the reticuloendothelial system. PEG chains may also act as flexible spacers for functionalizing the NPs with ligands of biological interest as peptides, carbohydrates, or aptamers. Protein resistance, which is the most known feature of PEG grafted surfaces, has been attributed to both steric and hydration repulsion forces of PEG molecules introducing a high activation energy barrier for proteins to adsorb.1
3 The thickness and the chain density of the grafted PEG layer are the most relevant parameters in protein resistance. Indeed, the thickness of the layer must be large enough to screen protein
particle interactions, and the PEG grafting density must be sufficient to block the protein diffusion to the particle surface.4 Therefore, the brush conformation, where PEG chains are densely end-grafted to the surface, is the preferred one for shielding the substrate surface from protein adsorption. According to the scaling theory proposed by de Gennes and Alexander,5
8 r 2011 American Chemical Society

the brush height (L) depends on the grafting density (σ) and the number of monomers per chain (N) and scales as L ∼ Nσ1/3 in a good solvent. The linear variation of the brush equilibrium thickness with the degree of polymerization is a distinctive feature, which is in contrast with free polymer chains in a good solvent where the coil size varies as R ∼ aN3/5, with a being the size of the repeating unit. Besides the chemical approach, which consists of grafting PEG chains onto nanoparticles through a covalent approach, the physical adsorption of polymer or copolymer chains represents an alternative and versatile approach for functionalizing inorganic or organic NPs.9
12 Especially, the adsorption of poly(L-lysine)graft-poly(ethylene glycol) (PLL-g-PEG) copolymers onto various metal oxides has been reported as a convenient method for controlling the density of grafted PEG segments and the related resistance to protein adsorption.4 In this case, the PLL backbone is anchored to the negatively charged surface of the oxides Received: July 31, 2011 Revised: September 15, 2011 Published: September 19, 2011 12891

dx.doi.org/10.1021/la202990y | Langmuir 2011, 27, 12891–12901

Langmuir through electrostatic interactions while PEG chains stretch out in solution. The adsorption mechanism of such graft-copolymers is far from simple, as the interaction with the surface is governed by a delicate balance of attractive (van der Waals, electrostatic) and repulsive (steric) forces, each with its particular magnitude and modulated by the presence of the adsorbing surface. Depending on the relative importance of these forces, either attraction or repulsion of the copolymer with the surface may dominate.13 Surprisingly, the adsorption of linear diblock copolymers of PLL and PEG has been less studied despite their ease of synthesis and their expected properties at the solid
liquid interface similar to PLLg-PEG copolymers. In a previous work, we showed that the adsorption of such a double hydrophilic copolymer, namely, the poly(ethylene oxide)-block-poly(L-lysine) (PEO-b-PLL) copolymer, onto silica NPs is a convenient approach to form a PEO layer with stabilizing properties at the particle surface.14 At low surface coverage, both PEO and PLL blocks adsorb with a more or less flat conformation, whereas at higher surface coverage the stronger affinity of PLL with both silanol and silanolate groups is sufficient to desorb PEO blocks from the surface. The copolymer adopts a typical anchor
buoy conformation where the grafting density of PEO blocks is directly related to the number of adsorbed PLL blocks per particle. Interestingly, we demonstrated that the addition of silica particles into a large excess of copolymer solution is the preferred method to obtain an adsorbed PEO layer in closeto-equilibrium conditions.14 Based on a similar approach, the present work aims at studying first how the structure of the adsorbed PEO layer on silica NPs can be controlled in terms of thickness and grafting density of PEO chains. Especially, we aim at studying how the respective sizes of the PEO and PLL blocks as well as the experimental adsorption conditions, mostly the ionic strength of the medium, can achieve such control. It is expected that the size of the PLL anchor block should strongly influence the conformation of the PEO blocks according to scaling predictions for polymers at interfaces.15 Then, the colloidal properties of the PEO-coated NPs will be evaluated with regard to their stability at different ionic strengths and in the presence of bovine serum albumin (BSA) protein. We expect a close correlation between the structure of the PEO layer and its interactions with proteins as it has been shown for various types of PEGylated surfaces.16
18

’ MATERIALS AND METHODS Materials. Dimethylformamide (DMF) (Scharlau, 99.9%) was dried over molecular sieves (3 and 4 Å) and cryodistilled prior to use. Aminoterminated PEO (Mn = 5000 g/mol and Mn = 2000 g/mol) purchased from Rapp Polymere was dissolved in dioxane and lyophilized to remove water traces. ε-Trifluoroacetyl L-lysine N-carboxyanhydride (TFA-Lys NCA) (Isochem, +96%), 4-(2-hydroxy-1-naphthylazo)benzenesulfonic acid sodium salt, also referred as acid orange II or C.I. acid orange 7 (Aldrich), silica particles (Ludox TM-50, Aldrich), and bovine serum albumin (BSA, Fraction V) were used as received. Polymer and Block Copolymer Synthesis and Characterization. Detailed information about the synthesis of PLL homopolymer and PEO-b-PLL copolymers as well as characterization procedures can be found in previous references.19
21 Briefly, PEO-b-PLL diblock copolymer was synthesized by ring-opening polymerization (ROP) of TFA L-lysine NCA initiated by an amino-terminated PEO macroinitiator. 1-Azido-3-aminopropane compound was synthesized following an already published procedure and used as initiator for the synthesis of poly(TFA L-lysine) homopolymer.22 The degree of polymerization

ARTICLE

Figure 1. Chemical structures of PEOm-b-PLLn copolymer and PLLn homopolymer.

Table 1. Molecular Characteristics of PEOm-b-PLLn Block Copolymers and PLL Homopolymera Mn (g/mol)

Mw/Mn

PLL22

2800

1.20

PEO45-b-PLL10

3300

1.20

PEO45-b-PLL37

6800

1.20

PEO45-b-PLL90 PEO113-b-PLL10

13 600 6300

1.15 1.10

PEO113-b-PLL270

39 800

1.05

a

Mn and Mw are the number-average and weight-average molecular weights, respectively. (DP) of the poly(TFA L-lysine) block was determined from the molar ratio of TFA lysine and ethylene oxide units obtained by 1H NMR analysis (Br€ucker AC 400 spectrometer) in deuterated dimethyl sulfoxide. For the poly(TFA L-lysine) homopolymer, the determination of the DP by 1H NMR was possible only after deprotection of the trifluoroacetyl groups since the peaks of the initiator were hidden by those of the polymer before deprotection.14 The molecular weight distribution and the polydispersity index (PDI = Mw/Mn) were determined by sizeexclusion chromatography performed in DMF with LiBr (1 g/L) as eluent (0.8 mL/min) at 60 C using a Waters apparatus (Alliance GPCV2000) equipped with a refractometer detector and two PLgel 5 μm Mixed-C columns calibrated with polystyrene standards. The removal of the labile trifluoroacetyl (TFA) protecting groups of L-lysine was achieved by treatment with KOH (1.5 equiv) in tetrahydrofuran (THF) at RT during 24 h. Once the reaction completed, the THF was removed by rotary evaporation and the (co)polymers were precipitated in cold diethyl ether and dried overnight under dynamic vacuum. The full deprotection was confirmed by 13C NMR in D2O with DCl through the disappearance of the signal of the fluorinated carbon (
CF3: δ = 110
120 ppm) and by 1H NMR through the shift of the ε-methylene protons of the lysine side chains (ε-CH2: δ = 2.70 ppm).23 The chemical structures of PLL homopolymer and PEO-b-PLL copolymers are depicted in Figure 1. Main structural parameters related to the polymers synthesized in this work are presented in Table 1. It is worth noting that both PLL and PEO-b-PLL compounds could not directly be solubilized in 10 mM phosphate buffer (PB) at pH 7.4 because of remaining traces of KOH. Therefore, the polymer powder was dispersed at first in buffer and the then pH was adjusted to 7.4 with 1 M HCl solution after overnight stirring.

Characterization of Ludox TM-50 Silica Nanoparticles. Ludox TM-50 silica NPs have a diameter of 22 nm according to the manufacturer. Dynamic light scattering (Malvern instrument ZetaSizer Nano ZS) performed in 10 mM phosphate buffer pH 7.4 gives a z-average hydrodynamic diameter of 58 nm and a polydispersity index of 0.20 obtained by fitting the autocorrelation curve with the cumulant algorithm. The specific surface area of silica NPs was determined by small-angle neutron scattering (PACE spectrometer at the Laboratoire L
eon Brillouin, CEA-Saclay, France) from the q4 region (q is the scattering vector) by averaging the Porod constant A = 2πΔF2(S/V)Φ(1
Φ), where ΔF2 is the square of the difference in neutron scattering length 12892

dx.doi.org/10.1021/la202990y |Langmuir 2011, 27, 12891–12901

Langmuir density, Φ is the particle volume fraction, and S/V is the specific surface area (m2/m3) that can be converted into m2/g by considering the density of particles (d = 2.2). Ludox NPs were analyzed in D2O at a concentration of 10 mg/mL in a 5 mm thick quartz cell. A specific surface area of 87 ( 5 m2/g was derived using the above equation. Adsorption Isotherms. Adsorption isotherms of PLL, PEO-b-PLL, and BSA onto silica NPs were performed according to a depletion method, which relies on the measurement of the amount of remaining polymer (protein) in the supernatant, leading to the amount of polymer (protein) effectively adsorbed at the particle surface. Polymer or protein solutions prepared at various concentrations in 10 mM phosphate buffer pH 7.4 were quickly added to the dispersion of silica NPs at 2.8 g/L for polymer adsorption and 1 g/L for BSA adsorption, respectively. After 24 h stirring at room temperature, NPs were centrifuged (11 000g, 30 min, 10 C). The titration of PLL and PEO-b-PLL in supernatants was carried out using a 10
4 M solution of acid orange II in 0.017 M acetic acid as titrating solution.24,25 To 200 μL of supernatant, 1.3 mL of dye solution was added. After 24 h, the absorbance was measured at λmax = 485 nm using a microplate reader (SpectraMax M2e, Molecular Devices). Quantification was performed from calibration curves obtained by adding acid orange II to PLL and PEO-b-PLL solutions of known concentrations. BSA was titrated at 280 nm in microplates using a calibration curve ranging from 0 to 2  10
5 M BSA. PEO-b-PLL Coated Silica Nanoparticles. PEO-b-PLL-coated NPs used for stability studies (salt, BSA) were prepared by mixing equal volumes of block copolymer solution and silica suspension, both prepared in 10 mM PB pH 7.4. After 24 h of gentle stirring, the suspensions were centrifuged (2300g, 30 min, 10 C) and resuspended in PB to remove the excess of unadsorbed block copolymers. Initial concentrations of copolymer and silica used for the preparation of hybrid NPs are reported in the Supporting Information. Then, PEO modified NPs were mixed with NaCl and BSA solutions varying in concentration, and the colloidal stability was assessed through dynamic light scattering measurements.

Dynamic Light Scattering (DLS) and Static Light Scattering (SLS) Analysis. Hydrodynamic sizes of copolymer coated silica NPs as well as their size variations in different conditions (salt, BSA) were determined by DLS analysis at 25 C using a Malvern ZetaSizer Nano ZS instrument with a detection angle at 90. Mean hydrodynamic diameters and their size distributions were determined using a cumulant analysis method. SLS experiments were performed using an ALV goniometer and a 35 mW HeNe linear polarized laser with a wavelength of 632.8 nm. Samples were kept at constant temperature (25 C) during all the experiments. The accessible scattering angle range was from 40 up to 140. A total of 2 mL of particle suspension (0.4% w/v in 10 mM PB pH 7.4) was introduced in 20 mm diameter cylindrical glass cells and immersed in a filtered toluene bath. Dilutions of particle suspensions were performed by adding the solvent directly into glass cells. Scattered intensities were acquired during 20 s at each angle, and the whole data were analyzed with ALVStat software using toluene as standard for the determination of the Rayleigh ratios. The differential refractive index increments dn/dc of bare and PEO113-b-PLL10-coated silica NPs in PB buffer were measured over a concentration range of 0.8
4 mg/mL by means of a differential refractometer (Wyatt Optilab rEX) operating at a wavelength of 658 nm and at 25 C. Microelectrophoresis. Electrophorectic mobilities of silica NPs coated with PEO-b-PLL and/or BSA were measured using a Malvern ZetaSizer Nano ZS instrument. In this study, the electrophoretic mobilitity (μ) was converted to the zeta potential (ζ) using the Smoluchowski approximation. All the measurements were the average of at least five runs at 25 C. Isothermal Titration Calorimetry (ITC). Binding experiments of copolymer-coated NPs with BSA protein were performed on a ITC200 instrument (MicroCal), which operated at a constant temperature of 25 C. To carry out a binding experiment, 205 μL of suspension of

ARTICLE

Figure 2. Adsorption isotherms of PEO-b-PLL block copolymers onto Ludox TM-50 silica nanoparticles (C = 2.8 g/L) in phosphate buffer (10 mM, pH 7.4). Isotherms were fitted by the Langmuir equation. hybrid NPs (∼0.1% w/v in 10 mM PB pH 7.4, with the exact concentration of NPs being determined by thermogravimetry analysis) was placed in the sample cell and the syringe was filled with the BSA solution (1.5 mM in 10 mM PB pH 7.4). The first injection was set to a very small volume of 1 μL and was not taken into account in the ITC analysis, due to possible dilution effect occurring during the equilibration time preceding the first injection.26,27 It was followed by 39 injections of 1 μL each. The equilibration time was set to 300 s to ensure that the thermodynamic equilibrium was reached between two consecutive injections. The heat of dilution of BSA solution was determined in a blank experiment where BSA solution was injected into the sample cell containing only PB. The dilution heat was then subtracted to obtain the binding heat. MicroCal Origin 7.0 software was used to integrate peaks and determine the heat change associated with each injection of BSA solution. The one-site binding model was used to fit the data, with the molar concentration in NPs being derived from the mass concentration and the size and density of the NPs.

’ RESULTS AND DISCUSSION Adsorption of PEO-b-PLL Copolymers onto Silica Nanoparticles. As previously reported, the adsorption of PEO113-

b-PLL10 copolymer onto silica NPs may be used as a noncovalent strategy to produce PEGylated particles.14 The particle coating by copolymer is not trivial, as both PEO and PLL have an affinity with the silica surface through hydrogen bonds and electrostatic interactions, respectively. However, we demonstrated that PLL blocks have a higher affinity with the silica substrate than PEO blocks.14 Therefore, when the adsorption is performed in the presence of a large excess of copolymer, namely, by adding the particle dispersion to a concentrated solution of copolymer, PLL blocks are preferentially adsorbed at the silica surface whereas PEO chains stretch into solution, giving rise to a typical anchor
buoy conformation of the adsorbed copolymer. The thickness of the copolymer layer as well as the PEO grafting density being two essential parameters regarding the colloidal properties of the NPs and their interactions with proteins, a full physicochemical investigation of the adsorption of PEO-b-PLL copolymers with different block lengths (Table 1) and at various ionic strengths has been performed in the present study. The adsorption isotherms of five PEO-b-PLL copolymers on silica NPs, obtained through the so-called depletion method (see Materials and Methods), are plotted in Figure 2. The same trend 12893

dx.doi.org/10.1021/la202990y |Langmuir 2011, 27, 12891–12901

Langmuir

ARTICLE

Table 2. Physicochemical Parameters Related to the Adsorption of PEO-b-PLL Copolymers onto Silica Nanoparticles (10 mM Phosphate Buffer, pH 7.4) PLL22 a Cmax ads max a Cads

PEO45-b-PLL10

PEO45-b-PLL37

PEO45-b-PLL90

PEO113-b-PLL10

PEO113-b-PLL270

mg/g SiO2

274

309

237

197

700

240

mmol lysine units/g SiO2

2.06

0.94

1.30

1.31

1.11

1.63

(co)polymer

0.64

0.65

0.24

0.10

0.77

0.04

L/mol

1.23  105

1.76  105

4.48  105

8.39  106

1.58  105

6.72  107

CR

%

100

46

63

64

54

79

Dd

nm




1.4

2.3

3.6

1.3

5.6

dense brush

brush

interacting

dense brush

interacting

surface density (σ)

chains/nm2 kb c

conformation of PEO chains

mushroom

mushroom

a

Maximum mass concentration of surface-adsorbed (co)polymer. b Adsorption constant derived from the Langmuir equation. c Surface coverage ratio (CR) derived from the number of adsorbed PLL units by considering a coverage ratio of 100% with poly(L-lysine)22. d Intermolecular distance between PEO blocks organized in a hexagonal lattice, D = 2(πσ)
1/2.

was observed for all copolymers, namely the adsorbed amounts increase sharply at very low concentrations of copolymers and rapidly level off. This behavior is characteristic of a Langmuir type adsorption isotherm, suggesting a rather high affinity of the block copolymer toward the silica surface. Langmuir’s model of adsorption assumes that intermolecular forces rapidly decrease with distance and consequently predicts the formation of a monolayer of adsorbate.28 In addition, this model assumes a homogeneous adsorbent where all adsorption sites are identical and energetically equivalent. The Langmuir adsorption equation was used to fit the experimental data: CAds ¼

kCmax ads Ceq 1 þ kCeq

where Cmax ads is the maximum mass concentration of surfaceadsorbed copolymer chains, Ceq is the equilibrium concentration in copolymer, and k is the adsorption constant (= kadsorption/ kdesorption) related to the free adsorption energy, k  (
ΔGads/RT) (see values in Table 2). The maximal amount of adsorbed copolymer was also expressed in terms of surface density (the number of adsorbed copolymer chains per unit area) by considering the specific surface of silica NPs (87 m2/g, see Materials and Methods). The close correlation between the PLL length and the surface density of the copolymer chains seems to indicate that copolymer chains preferentially adsorb through their PLL blocks; the shorter their length, the higher their surface density (Table 2). It is expected that PLL blocks adopt a flat conformation by charge compensation between cationic lysine residues and dissociated silanol groups at the surface of the particles. Afshar-Rad et al. reported that PLL adsorbs in flat conformation even for polymers with a high degree of polymerization (DP ∼ 600).29 However, the copolymer adsorption mechanism is probably more complex since PEO also has an affinity with silica surface through hydrogen bonding. Therefore, even if the adsorption was carried out in the presence of a large excess of copolymer, it is likely that both PLL and PEO blocks stick to particles in the first stages of the adsorption. Then, stronger interactions of PLL with silica account for PEO desorption and subsequent PLL adsorption.14 In the whole process, the PEO displacement by PLL has an energy cost that can be evaluated by considering the value of the Langmuir adsorption constant (k, see Table 2) for various copolymer compositions. The variation

of k over the ratio of the number of lysine units to ethylene oxide units for each copolymer shows an exponential behavior, which is line with an electrostatically driven adsorption of copolymers through their PLL blocks (see the Supporting Information). The behavior of copolymers at interfaces can be rationalized using scaling theory of copolymer adsorption, which considers two regimes, the anchor regime where the adsorbing block is relatively long compared to nonadsorbing block and the buoy regime where the adsorbing block is small.30,31 The transition between the regimes can be defined by considering the asymmetry ratio β of copolymers, which corresponds to the ratio of the projected areas of both blocks. For a good solvent for both blocks, β is given by the following equation:30,31
6=5 Rg, B 2 NB β¼ ¼ 2 Rg, A NA where Rg and N are, respectively, the radius of gyration and the degree of polymerization of the buoy (B) and the anchor (A) block. In the buoy regime, β > NA, while 1 < β < NA in the anchor regime. Here, all copolymers containing PEO45 blocks are in the anchor-dominated regime. Importantly, the experimental data of the surface density (σ) of these copolymers fit well (R2 = 0.99, see the Supporting Information) with the theoretical prediction of σ in the anchor regime, which scales as σ ∼ NA
1.30,31 Using the same formalism, we can show that PEO113-b-PLL270 is in the anchor regime and PEO113-b-PLL10 in the buoy regime. The above results were correlated to the surface coverage ratio of PLL blocks. For this purpose, the adsorption of a homopolymer of PLL containing 22 lysine units was used as reference to calculate the maximal number of lysine units that possibly adsorbed onto silica NPs. The surface coverage ratio, derived for each copolymer, corresponds to the ratio of the actual number of adsorbed lysine units within PLL blocks of different lengths to the number of adsorbed lysine residues in homo PLL. Again, this derivation was performed assuming that PLL blocks adsorb in flat conformation.29,32,33 The highest values of coverage ratios of PLL blocks were found for copolymers in the anchor regime (Table 2), except for PEO45-b-PLL10, but the asymmetry ratio of this copolymer (β = 6) is close to the crossover regime value (β = 8). It is worth noting that PLL surface coverage of 100% was not reached even in the anchor regime. This can be understood by considering that adsorbed copolymer chains hinder to some 12894

dx.doi.org/10.1021/la202990y |Langmuir 2011, 27, 12891–12901

Langmuir

ARTICLE

Figure 3. Rayleigh ratios of silica NPs (O) and PEO113-b-PLL10-coated silica NPs (b) as a function of the concentration of silica particles in phosphate buffer (10 mM, pH 7.4).

Figure 4. Maximal amounts of PEO45-b-PLL10 adsorbed on silica particles in 10 mM phosphate buffer at pH 7.4 as a function of the concentration of added sodium chloride.

Table 3. Refractive Index Increments (dn/dc), Optical Constants (K), and Weight-Average Molecular Weights (Mw) of PEO113-b-PLL10 Block Copolymer, Silica Nanoparticles, and PEO113-b-PLL10-coated Silica NPs in 10 mM PB, pH 7.4. (see ref 36 for the Determination of Mw of Silica Particles)

by varying the concentration of added sodium chloride in 10 mM phosphate buffer. Considering the uncertainty of the measurements, Figure 4 shows that the amount of adsorbed copolymer decreased when the salt concentration increased. Such behavior is typical for the adsorption of polyelectrolytes onto oppositely charged substrates when electrostatic interactions are the main driving force of the adsorption,37
42 and therefore, this confirms the preferential adsorption of PEO-b-PLL copolymers through their charged PLL blocks. Increasing the salt concentration in the adsorption medium screens electrostatic attractions of PLL units with SiO
groups and thus hinders the adsorption of the copolymer. In addition, the specific adsorption of sodium counterions may compete with the lysine residues for neutralizing the silica surface charge and therefore might also explain the low amount of adsorbed copolymer at high salt concentration.43 Conformation of the Adsorbed PEO-b-PLL Chains at the Particle Surface. The conformation of polymer chains adsorbed on a solid surface is important regarding their interaction with the surrounding medium, especially when the adhesion of proteins is a concern. The data obtained from adsorption isotherms have been considered in detail in order to determine the conformation of the adsorbed copolymer chains. Considering the total surface of silica particles as a 2D hexagonal compact lattice, the distance D between chain attachment points was calculated for all samples (Table 2). In the Alexander
de Gennes formalism, the conformation of the grafted chains and the equilibrium thickness of the polymer layer on a planar surface are dependent on the distance D and the coil size of the grafted block given by the Flory radius (RF).5
8 In a good solvent, the Flory radius scales as RF ≈ aN3/5, where N is the number of repeating units and a is the monomer size, whose value is 0.358 nm for PEO.44 RF values of 3.51 and 6.11 nm were calculated for PEO chains of 2000 g/mol (PEO45) and 5000 g/mol (PEO113), respectively. Then, the conformation of the nonadsorbing PEO blocks can be deduced by comparing the distance D and the Flory radius of PEO chains in bulk solution. Three regimes can be identified. For low surface density, the distance D is larger than RF; the chains do not interact laterally and adopt a “mushroom” conformation. On the contrary, for high grafting density, this distance decreases and chains interact strongly (D < RF). The overlap between chains forces them to stretch away from the surface, resulting in a “brush”

dn/dc cm3/g PE0113-b-PLL10 SiO2 PEO113-b-PLL10/SiO2

0.1122 0.0475 0.0511

K mol 3 cm2/g2

MW g/mol


8

7.55  103


8

1.78  107


8

3.51  107

9.10  10 1.63  10 1.88  10

extent further adsorption of copolymer molecules owing to increased osmotic repulsion within the PEO layer, especially at high copolymer surface density. Also, while PLL adsorbs stronger, that does not mean that it will prevent all PEO from providing part of the surface coverage, especially as the PEO chain is anchored to the interface and can be of greater length than the PLL block. Another approach to quantify the chain density at the surface of particles, previously reported by Berret and co-workers,34,35 was applied here with PEO113-b-PLL10 as a comparison with the classical approach based on adsorption isotherms. This method relies on determining the Rayleigh ratios of bare and copolymercoated silica NPs as a function of the silica mass concentration (Figure 3) using the following expression: !2 R SiO2 =PEO-b-PLL K SiO2 =PEO-b-PLL MwPEO-b-PLL ¼ 1 þ nads R SiO2 K SiO2 MwSiO2 where Ri, Ki, and Miw are the Rayleigh ratios, the optical constants (= 4π2n2(dn/dc)2/NAλ4), and the weight-average molecular weights, respectively, of silica NPs or PEO113-b-PLL10-coated silica NPs (Table 3). A value of 2300 chains per particle was computed from the difference in the two slopes of the Rayleigh ratios of copolymer coated particles and bare particles (Figure 3). From the adsorption isotherm of PEO113-b-PLL10, a value of 2180 chains/particle was found, hence showing a good agreement between both approaches. The influence of the ionic strength on the adsorption of PEO45-b-PLL10 copolymer onto silica NPs was studied at pH 7.4

12895

dx.doi.org/10.1021/la202990y |Langmuir 2011, 27, 12891–12901

Langmuir

ARTICLE

Table 4. Hydrodynamic Diameters of Bare and PEO-b-PLLCoated Silica Nanoparticles in Phosphate Buffer (pH 7.4, 10 mM) DH

silica NPs

(nm)a

PDI = μ/Γ2 b

58

0.20

Lexperimental

Ltheoretical

(nm)

(nm)c

11.8

PEO45-b-PLL10

860

0.36

aggregation

PEO45-b-PLL37

75

0.19

8.5

PEO45-b-PLL90 PEO113-b-PLL10

85 97

0.15 0.19

13.5 19.5 (19)d

8.1 21.2

PEO113-b-PLL270

93

0.25

17.5

11.8

9.6

a

Hydrodynamic diameter of copolymer coated silica NPs determined from dynamic light scattering. b Polydispersity index (PDI) derived from a cumulant analysis. c Theoretical thickness of the PEO layer derived from the expression of Birshtein et al.51 assuming a brush conformation. d Adsorption onto silica nanoparticles with DH = 82 nm and μ/Γ2 = 0.04 (see previous work14).

conformation where the segment
segment interaction energy and the elastic free energy are balanced. In between these two regimes, when D ∼ RF, the polymer chains are still interacting, giving rise to the so-called “extended mushroom” or “interacting mushroom” regime.5
8,30,45,46 The conformational assignments for various PEOn-b-PLLm copolymers adsorbed on silica NPs are summarized in Table 2 by assuming in a first approximation that the particle surface is locally planar. As expected, the length of PLL blocks determines the conformation of the PEO chains, with the shorter PLL blocks giving rise to a high grafting density and a corresponding brush conformation whatever the length of the PEO block. In our study, the mushroom regime was not reached even with copolymers having a long anchor block size, namely, PEO45-b-PLL90 and PEO113-b-PLL270. Both copolymers adopt a conformation at the frontier between the interacting “mushroom” and “brush” regime. It is worth reminding that all conformation regimes were evaluated from the maximal adsorbed concentration in copolymer at the particle surface. Lower copolymer surface densities and corresponding conformation regimes should be obtained by adsorbing less copolymer at the particle surface. However, such a partial coverage of the particle surface may lead to a marked colloidal aggregation as demonstrated previously.14 The thickness (L) of the adsorbed copolymer layer is another important parameter that was assessed by dynamic light scattering (DLS) considering the difference of hydrodynamic diameter before and after copolymer adsorption. At equilibrium, the thickness traduces the balance between the osmotic pressure in the polymer layer which tends to increase L and the elastic stretching forces of the chains which tend to decrease L. At first sight, the quite high size-polydispersity of Ludox silica NPs (PDI = 0.20) might impede an accurate determination of this parameter by simple DLS measurements. However, we found that the thickness of an adsorbed layer of PEO113-b-PLL10 onto silica NPs with a much lower size-distribution (PDI = 0.04) was comparable to the thickness measured on Ludox particles (see Table 4). Significant and reliable differences in size were obtained for copolymer-coated NPs, except for PEO45-b-PLL10 where a marked aggregation was observed (see below). In addition, copolymer-coated NPs have a size-polydispersity close to this of uncoated NPs. In these conditions, data were analyzed on the basis of the theoretical layer thickness derived for the mushroom

and brush conformation regime.47
50 At low grafting density, in the mushroom regime, the layer thickness L is simply given by the Flory radius. In the brush regime, the chains interact and are forced to stretch due to mutual repulsion. On a flat surface, the chains can be described as a connected linear sequence of subunits (“blobs”) of size D, each containing gD monomers (gD = (D/a)5/3), with the overall thickness L of the brush being then given by L = ND/gD.5,8 However, for a convex surface, the monomer density decreases with the distance to the surface. Birshtein et al. established that, for a brush layer formed onto such surface, the end-to-end distance of chains is given by51
2=5 3=5 R L ¼ aN D with R being the radius of the convex surface. By using the expression of the layer thickness in the brush regime, a good agreement was found between experimental and calculated thicknesses for copolymers having short PLL blocks, namely, PEO45-b-PLL37 and PEO113-b-PLL10, confirming that the PEO blocks of these two copolymers adopt a brush conformation. For copolymers with larger PLL blocks, namely, PEO45b-PLL90 and PEO113-b-PLL270, the experimental values of the layer thicknesses did not fit with the expression of Birshtein. Actually, for both of these copolymers, the distance D between grafting points is close to the Flory radius of the PEO blocks and therefore a true brush conformation could not be obtained (Table 2). In addition, it is likely that the thickness of the adsorbed PLL blocks contribute to increase the overall thickness of the copolymer layer at the particle surface. The aggregation observed upon adsorption of PEO45-b-PLL10 was unexpected. This cannot be only due to the relatively short length of the PEO block, as both PEO45-b-PLL37 and PEO45-b-PLL90-coated nanoparticles were found to be stable. In the same manner, it cannot be related to a difference in charge surface as all copolymer-coated particles have a slight positive zeta potential, which traduces an overcharging of the adsorbed PLL blocks.14 Actually, PEO45-b-PLL10 has the lowest coverage ratio (44%) (Table 2), and therefore, one may hypothesize that interparticle interactions mediated by PEO blocks have occurred in the course of the adsorption process. Indeed, and as mentioned previously, PEO is known to have an affinity with silica surface through hydrogen bonding with silanol groups.52
54 Such aggregation did not occur with PEO113-bPLL10 despite a low coverage ratio (CR = 52%, see Table 2) as the PEO blocks were probably large enough to provide the system with steric stabilization during the overall adsorption process. Salt and pH Stability of PEO-b-PLL-Coated Silica Nanoparticles. The salt stability of NPs fully coated by PEO-b-PLL copolymers was evaluated by determining the hydrodynamic radius of copolymer-modified particles at various concentrations of added sodium chloride in the medium (Figure 5). The RH/ RH0 ratio, where RH0 and RH correspond to the hydrodynamic radius of the NPs with and without added salt, was determined by DLS to assess the colloidal behavior at various ionic strengths. NPs coated with PEO113-based copolymers were stable up to CNaCl = 0.1 M, whereas NPs coated with PEO45-b-PLL37 start to flocculate from CNaCl = 0.1 M. A strong aggregation was observed at CNaCl = 0.5 M whatever the block copolymer composition. As all copolymer-coated NPs were almost electrically uncharged,14 the decrease of the Debye length related to the increase in salt concentration should not modify the colloidal stabilization to a 12896

dx.doi.org/10.1021/la202990y |Langmuir 2011, 27, 12891–12901

Langmuir

ARTICLE

Table 5. Colloidal State of PEO113-b-PLL10-Coated Silica Particles Prepared at Physiological pH and Incubated at Various pHs pH desorbed amount of copolymera

3.2 83%

5 14%

7.4 0%

dispersion stateb

aggregation

stable over

stable over

4 weeks

4 weeks

a

Determined from the initial adsorbed amount of copolymer onto silica nanoparticles at pH 7.4. b Evaluated by dynamic light scattering.

Figure 5. Colloidal stability of PEO-b-PLL-coated silica particles as a function of the concentration of sodium chloride added to the medium (10 mM phosphate buffer, pH 7.4). Hydrodynamic radii (RH) are determined by DLS analysis at 25 C 1 h after the addition of salt. RH0 refers to the hydrodynamic radius of copolymer-coated particles before addition of salt.

great extent. However, the salt can screen the polylysine/SiO
interactions which held the copolymer at the particle surface as evidenced in Figure 4. Therefore, increasing the salt concentration may cause a partial desorption of the copolymer from the surface, leading to the flocculation of particles. In addition, PEO solvency properties are known to decrease upon addition of salt, thus favoring the particle aggregation.55
57 For comparison, Wind and Killmann showed that silica NPs coated with simple PEO chains of various molecular weights begin to aggregate at a salt concentration of about 0.01 M; the smaller the molecular weight of PEO, the faster the flocculation.58 In this case, the steric stabilization brought by the PEO layer could not counterbalance to a large extent the screening of the free negative charges at silica surface upon addition of salt. In other words, the thickness of the PEO layer is not large enough (L < 10 nm for PEO with Mw < 105 g/mol) to balance the shrinking of the electrical double layer at moderate ionic strength. Here, in the present study, the shift of the flocculation point toward a higher salt concentration (CNaCl > 0.15 M) shows that electrostatic interactions between PLL and SiO
groups are quite strong once formed and that PEO chains with an extended conformation also enable an improved steric stabilization. Concerning the lower stability of PEO45-b-PLL37coated particles, this probably has to do with the relatively thin layer of PEO at the silica surface (L = 8.5 nm, see Table 4), which cannot provide the system with sufficient steric stabilization as the concentration of salt increases. One advantage of the adsorption of PEO-based copolymers lies with the fact that this noncovalent process is reversible upon experimental condition changes, which may find some interesting issues in biological applications. In the view of using these particles for drug delivery applications, one may question their pH stability since pH values as low as 3
4 can be found in endosomal compartments. Therefore, we investigated the stability of the assemblies in various pH media as seen in Table 5. Both the hydrodynamic sizes of the NPs and the desorbed amounts of copolymer molecules were evaluated. At neutral pH, where the adsorption occurred, PEO113-b-PLL10-coated silica NPs remained

Figure 6. Adsorption isotherms of BSA onto silica nanoparticles coated with PEO-b-PLL copolymers in 10 mM phosphate buffer at pH 7.4. Lines serve to guide the eye only.

stable over at least 1 month without any desorption of copolymer. By decreasing the pH, PLL blocks are removed from the silica surface as a result of the protonation of the SiO
groups. At pH around 3, almost all copolymer molecules desorbed from the surface, inducing a spontaneous aggregation of the particles, which were not sterically or electrostatically stabilized anymore. Protein Adsorption onto PEO-b-PLL Copolymer-Coated Silica Particles. The adhesion of proteins onto PEO modified NPs is expected to be affected by both the grafting density and the layer thickness of PEO chains.1,2,59,60 Herein we have studied the interactions of a model protein, bovine serum albumin (BSA), with NPs fully coated by PEO-b-PLL copolymers using a combination of analytical techniques. A set of three copolymers was chosen to cover a large range of separation distances between PEO chains (D), namely, PEO113-b-PLL10 (1.3 nm), PEO45-bPLL37 (2.3 nm), and PEO113-b-PLL270 (5.6 nm). Adsorption isotherms of BSA on particles coated with copolymer allowed us to quantify the amount of adsorbed protein. DLS and microelectrophoresis measurements were used to characterize the colloidal stability of the particles in the presence of BSA. Finally, isothermal titration calorimetry (ITC) was carried out to quantify the interactions of BSA with copolymer-coated NPs. The overall results give strong evidence that the three copolymer systems display very different behaviors in the presence of BSA. Only little adsorption of BSA occurred on PEO113-b-PLL10coated silica NPs, while for PEO113-b-PLL270 protein molecules continuously adsorbed (Figure 6). In between, PEO45-b-PLL37 gave rise to a Langmuir type adsorption isotherm where the adsorbed amount sharply increased prior to reaching a plateau. 12897

dx.doi.org/10.1021/la202990y |Langmuir 2011, 27, 12891–12901

Langmuir Interestingly, both PEO45-b-PLL37 and PEO113-b-PLL10 coated NPs adsorbed less BSA than naked silica NPs for which a maximum surface excess of 1.1 mg BSA/m2 was derived from the adsorption isotherm and the particle porosity (87 m2/g). This value is in good agreement with other ones reported in the literature under similar conditions of pH and ionic strength.61
63 DLS analysis also showed rather different behaviors in the presence of BSA (Figure 7). NPs coated with PEO113-b-PLL10 were found to be stable whatever the concentration of added BSA, whereas a strong aggregation was observed with PEO113-bPLL270 even at low BSA concentration. For PEO45-b-PLL37coated particles, a mechanism of aggregation/redispersion was observed when the concentration of BSA increased. ITC experiments performed on the same systems and in the same conditions evidenced that all copolymer-coated particles gave rise to exothermic binding upon addition of BSA (Figure 8). This is in

ARTICLE

line with an enthalpically driven adsorption of BSA onto NPs through hydrogen bonding and/or electrostatic interactions. However, the binding affinity of BSA molecules with NPs is quite low as revealed by the values of K below 104 M
1 whatever the copolymer system. The fitting procedure could not detect significant differences in enthalpies (ΔH), but the stoichiometry of binding (N), that is, the number of BSA molecules per silica particle, confirmed the results from adsorption isotherms, namely, an increase of the number of adsorbed BSA molecules as the distance D between PEO chains gets longer (Figure 6). The adsorption of BSA onto PEGylated particles can proceed through either short-range protein
surface contact (primary adsorption) or long-range van der Waals protein
surface attraction at the outer edge of the PEO layer (secondary adsorption).64,65 Primary adsorption implies that protein molecules are smaller than the separation distance between PEO chains, allowing the protein to diffuse into the PEO layer and make contact with the surface. Conversely, a secondary adsorption occurs if BSA molecules are larger than the distance between PEO chains. In this case, the protein cannot diffuse into the PEO layer because of excluded volume interactions but it may adsorb at the brush
solvent interface if the van der Waals attraction is sufficient. For a planar geometry, the van der Waals interaction energy per unit area can be calculated using V ðLÞ ¼


Figure 7. Colloidal stability of PEO-b-PLL-coated silica particles in 10 mM phosphate buffer at pH 7.4 upon addition of BSA. RH and RH0 are the hydrodynamic radii of particles in the presence or absence of BSA.

A 12πL2

where A is the Hamaker constant (A ∼ 10
21 J)65 and L is the thickness of the PEO layer. In our case, V(L) is of the order of
0.01kT per BSA molecule whatever the copolymer system, meaning that this adsorption mode is not realistic. By considering that BSA molecules have a hydrodynamic diameter of 5.3 nm at pH 7.4,66 the primary adsorption of BSA was probably significant with PEO113-b-PLL270-coated NPs, where the separation distance between PEO chains was the largest among all copolymers used in this study (D = 5.6 nm, Table 2). However, the coverage ratio of the copolymer at the particle surface being quite high (79%, Table 2), the direct adsorption of BSA on the silica surface was unlikely. Therefore,

Figure 8. ITC data for titration of BSA (c = 1.5 mM) onto silica particles coated with (a) PEO113-b-PLL10, (b) PEO45-b-PLL37, and (c) PEO113-bPLL270 in 10 mM phosphate buffer, pH 7.4. 12898

dx.doi.org/10.1021/la202990y |Langmuir 2011, 27, 12891–12901

Langmuir one has to consider that protein molecules have interacted with adsorbed lysine residues through hydrogen bonding, hydrophobic, or electrostatic interactions. Especially, BSA, which is negatively charged at pH 7.4, may interact with protonated amines from PLL that are not involved in ion-pairing with SiO
groups.67 The fact that BSA was continuously adsorbing on the particles is difficult to interpret (Figures 6 and 8c). One may hypothesize a possible desorption of PEO113-b-PLL270 copolymer molecules from the particle surface upon addition of BSA. However, this hypothesis has been ruled out in an independent experiment where micrometric silica particles were coated with rhodaminelabeled PEO113-b-PLL270 copolymer. Fluorescence microscopy evidenced that the level of fluorescence of the particles was not modified upon addition of BSA (see the Supporting Information), indicating that copolymer chains remained strongly anchored at the particle surface even at high concentration of BSA. Actually, the infinite adsorption of BSA onto particles may be understood by considering that the strong interactions of the protein with adsorbed lysine units rapidly destabilized the particles, as shown by DLS analysis (Figure 7), leading to the formation of large flocs on which the protein could still adsorb in a multilayer fashion. For PEO113-b-PLL10 and PEO45-b-PLL37-coated NPs, both primary adsorption of BSA at the particle surface and secondary adsorption at the outer edge of the brush were unlikely regarding the small distance between PEO chains (D = 1.3 and 2.3 nm) and the thickness of the PEO layer (L = 19.5 and 8.5 nm). However, experimental data provided some evidence that the PEO layer cannot completely suppress the BSA adsorption onto silica NPs, especially when coated with PEO45-b-PLL37 copolymer. The presence of defects within the PEO brush layer may allow the proteins to be in contact with the silica surface. In addition, BSA molecules can compress the PEO layer to some extent and hence adsorb at a distance from the particle surface. In such an adsorption mechanism, entropic changes associated with the deformation of the PEO layer would be less favorable for PEO113-b-PLL10 and could explain its greater protein repellency.1 Also, the occurrence of enthalpic interactions between the PEO layer and BSA molecules cannot be completely ruled out as evidenced by exothermic binding values found by ITC analysis (Figure 8). Even if ITC could not detect any interaction between PEO and BSA in bulk solution (see the Supporting Information), it was reported that the brush conformation of PEO chains may give rise to a so-called ternary adsorption where proteins weakly interact with EO residues.64,68,69 Indeed, the compression of the PEO brush by large BSA molecules might induce conformational changes of EO units from a protein-repulsive polar conformation to a protein-attractive apolar conformation as it was described with large proteins such as streptavidin.64,70 The most striking feature observed with particles coated with PEO45-b-PLL37 copolymer is an aggregation/redispersion upon addition of BSA (Figure 7). This behavior may be attributed to a bridging phenomenon where PEO coated particles are connected to each other by BSA dimers or oligomers as previously reported by Chern et al.71 Once the particles become fully covered by BSA molecules, the aggregates are dissembled due to electrostatic repulsions induced by negative charges carried by the protein. Here, such a mechanism is supported by the zeta potential (ζ) values which decreased from +3 to
12 mV upon addition of BSA (result not shown). PEO113-b-PLL10 copolymer is the best system in term of protein repellency. The small PEO separation distance (D = 1.3 nm) and the quite thick PEO layer (L = 19.5 nm) both account

ARTICLE

for an efficient shielding of the silica surface from the adsorption of BSA. These results are in good agreement with those obtained by Gref et al. who reported that a maximal reduction in protein adsorption onto poly(lactic acid)-b-polyethylene glycol) (PLA-b-PEG) nanoparticles was obtained with PEG 5000 (113 units of ethylene glycol) and a PEG interchain distance of 1.4 nm.72 Even if BSA is slightly adsorbing to PEO113-b-PLL10coated particles, this does not induce any particle aggregation. Actually, the addition of BSA even caused a slight disaggregation of the particles as evidenced by the decrease of the mean hydrodynamic size (Figure 7). It is hypothesized that BSA molecules formed a thin shell with stabilizing properties as supported by the negative zeta potential of NPs in the presence of BSA (ζ =
9 mV, result not shown). Finally, the colloidal stability of copolymer-coated particles was assessed in fetal calf serum. After 24 h of incubation in 10 v/v % fetal calf serum solution, only PEO113-b-PLL10-coated NPs remained stable. Their hydrodynamic size before and after incubation were the same as well as their size-polydispersity (results not shown). Conversely, silica NPs modified with PEO45-b-PLL37 and PEO113-b-PLL270 were instantaneously aggregated when incubated in the serum. Therefore, the behavior of the particles in serum is in line with the results obtained with BSA alone.

’ CONCLUSION The physical adsorption of positively charged PEO-b-PLL copolymers onto silica nanoparticles was investigated as a means to control the surface properties of particles. Importantly, this study demonstrated the predominant role of the size of the PLL adsorbing blocks over the length of PEO blocks on the resulting conformations of PEO chains at the silica surface. As a matter of fact, the grafting density of PEO chains was directly related to the size of the anchoring block, with small PLL blocks giving the highest density of PEO chains, while this parameter was only little affected by the size of the PEO blocks, in the range of PEO molecular weights investigated. By varying the PLL length from 10 to 270 residues, brush or interacting mushroom conformations were obtained for PEO chains at the particle surface. Conversely, the size of the PEO block impacts more the thickness of the PEO layer than the grafting density does. Considering the properties of the particles, both a high copolymer surface density and a relatively high PEO molecular weight (MW = 5000 g/mol) are needed to provide the system with improved colloidal stabilization. In these conditions, silica nanoparticles coated with PEO113-b-PLL10 copolymer, which is in the so-called buoy regime, were found to be stable in salty media (up to 150 mM in NaCl) and showed almost no interaction with BSA. The overall study emphasized that the physical adsorption of PEO-b-PLL copolymer varying in composition and molar mass is a reliable alternative to the covalent approach for grafting PEO chains onto nanoparticles with good control over their conformation and resulting properties. ’ ASSOCIATED CONTENT

bS

Supporting Information. Preparation of copolymer coated nanoparticles, adsorption models, fluorescence microscopy of rhodamine-labeled PEO-b-PLL copolymers adsorbed on micrometer silica particles and ITC analysis of PEO and BSA in bulk solution. This material is available free of charge via the Internet at http://pubs.acs.org.

12899

dx.doi.org/10.1021/la202990y |Langmuir 2011, 27, 12891–12901

Langmuir

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected] (C.S.); [email protected] (S.L.).

’ ACKNOWLEDGMENT This work has been supported by the National Research Agency (ANR) through TecSan 2006 (Nano-Bio Imaging, ANR06-TecSan-015-03) and Blanc 2007 (ITC-NanoProbe, ANR-07BLAN-0290-01) programs and the European Commission through the seventh Framework Program (FP7) for Research & Development (CP-IP 213631-2 NANOTHER). The “Precision Polymer Materials” RNP program from ESF is also acknowledged. The authors thank the referees for their helpful comments and suggestions to improve the paper. ’ REFERENCES (1) Halperin, A. Langmuir 1999, 15, 2525. (2) Jeon, S. I.; Lee, J. H.; Andrade, J. D.; de Gennes, P. G. J. Colloid Interface Sci. 1991, 142, 149. (3) Jeon, S. I.; Andrade, J. D. J. Colloid Interface Sci. 1991, 142, 159. (4) Pasche, S.; V€or€os, J.; Griesser, H. J.; Spencer, N. D.; Textor, M. J. Phys. Chem. B 2005, 109, 17545. (5) Alexander, S. J. Phys. (Paris) 1977, 38, 983. (6) Alexander, S. J. Phys. (Paris) 1977, 38, 977. (7) de Gennes, P. G. Macromolecules 1980, 13, 1069. (8) de Gennes, P. G. Adv. Colloid Interface Sci. 1987, 27, 189. (9) McPherson, T.; Kidane, A.; Szleifer, I.; Park, K. Langmuir 1998, 14, 176. (10) Frka-Petesic, B.; Fresnais, J.; Berret, J. F.; Dupuis, V.; Perzynski, R.; Sandre, O. J. Magn. Magn. Mater. 2009, 321, 667. (11) Pavey, K. D.; Olliff, C. J. Biomaterials 1999, 20, 885. (12) Bearinger, J. P.; Terrettaz, S.; Michel, R.; Tirelli, N.; Vogel, H.; Textor, M.; Hubbell, J. A. Nat. Mater. 2003, 2, 259. (13) Feuz, L.; Leermakers, F. A. M.; Textor, M.; Borisov, O. Langmuir 2008, 24, 7232. (14) Louguet, S.; Kumar, A. C.; Sigaud, G.; Duguet, E.; Lecommandoux, S.; Schatz, C. J. Colloid Interface Sci. 2011, 359, 413. (15) Belder, G. F.; ten Brinke, G.; Hadziioannou, G. Langmuir 1997, 13, 4102. (16) Kenausis, G. L.; Voros, J.; Elbert, D. L.; Huang, N.; Hofer, R.; Ruiz-Taylor, L.; Textor, M.; Hubbell, J. A.; Spencer, N. D. J. Phys. Chem. B 2000, 104, 3298. (17) Pasche, S.; De Paul, S. M.; V€or€os, J.; Spencer, N. D.; Textor, M. Langmuir 2003, 19, 9216. (18) Sofia, S. J.; Premnath, V.; Merrill, E. W. Macromolecules 1998, 31, 5059. (19) Agut, W.; Br^ulet, A.; Taton, D.; Lecommandoux, S. Langmuir 2007, 23, 11526. (20) Agut, W.; Taton, D.; Lecommandoux, S. Macromolecules 2007, 40, 5653. (21) Rodriguez-Hernandez, J.; Gatti, M.; Klok, H.-A. Biomacromolecules 2003, 4, 249. (22) Carboni, B.; Benalil, A.; Vaultier, M. J. Org. Chem. 1993, 58, 3736. (23) Harada, A.; Cammas, S.; Kataoka, K. Macromolecules 1996, 29, 6183. (24) Gummow, B. D.; Roberts, G. A. F. Macromol. Chem. Phys. 1985, 186, 1245. (25) Maghami, G. G.; Roberts, G. A. F. Macromol. Chem. Phys. 1988, 189, 2239. (26) Mizoue, L. S.; Tellinghuisen, J. Anal. Biochem. 2004, 326, 125. (27) Chiad, K.; Stelzig, S. H.; Gropeanu, R.; Weil, T.; Klapper, M.; M€ullen, K. Macromolecules 2009, 42, 7545.

ARTICLE

(28) Wong, Y. C.; Szeto, Y. S.; Cheung, W. H.; McKay, G. Process Biochem. 2004, 39, 695. (29) Afshar-rad, T.; Bailey, A. I.; Luckham, P. F.; Macnaughtan, W.; Chapman, D. Colloids Surf. 1987, 25, 263. (30) Fleer, G. J.; Cohen Stuart, M. A.; Scheutjens, J. M. H. M.; Cosgrove, T.; Vincent, B. Polymers at Interfaces; Chapman & Hall Inc: London, 1993. (31) Maas, J. H.; Cohen Stuart, M. A.; Fleer, G. J. Thin Solid Films 2000, 358, 234. (32) Furusawa, K.; Kanesaka, M.; Yamashita, S. J. Colloid Interface Sci. 1984, 99, 341. (33) Bonekamp, B. C.; Vanderschee, H. A.; Lyklema, J. Croat. Chem. Acta 1983, 56, 695. (34) Qi, L.; Sehgal, A.; Castaing, J. C.; Fresnais, J.; Berret, J. F.; Chapel, J. P. ACS Nano 2008, 2, 879. (35) Sehgal, A.; Lalatonne, Y.; Berret, J. F.; Morvan, M. Langmuir 2005, 21, 9359. (36) The weight-average molecular weight Mw of silica particles was derived from the definition of the polydispersity index, PDI = Mw/Mn. The number-average molecular weight of silica particles was estimated from their diameter (D = 27 nm by transmission electron microscopy) and their density (2.22 g 3 cm
3). The polydispersity index was determined from the reduced second cumulant μ/Γ2 = 0.20 obtained through DLS analysis of the particles according to the following equation: μ/Γ2 = PDI(PDI
1) See: Varga, I.; Gil
anyi, T.; M
esz
aros, R.; Filipcsei, G.; Zrínyi, M. J. Phys. Chem. B 2001, 105, 9071. (37) Hoogeveen, N. G.; Cohen Stuart, M. A.; Fleer, G. J. Colloids Surf., A 1996, 117, 77. (38) Hoogeveen, N. G.; Cohen Stuart, M. A.; Fleer, G. J. J. Colloid Interface Sci. 1996, 182, 133. (39) Liufu, S.-C.; Xiao, H.-N.; Li, Y.-P. J. Colloid Interface Sci. 2005, 285, 33. (40) Shubin, V. J. Colloid Interface Sci. 1997, 191, 372. (41) Shubin, V.; Linse, P. J. Phys. Chem. 1995, 99, 1285. (42) Van de Steeg, H. G. M.; Cohen Stuart, M. A.; De Keizer, A.; Bijsterbosch, B. H. Langmuir 1992, 8, 2538. (43) Cohen Stuart, M. A.; Fleer, G. J.; Lyklema, J.; Norde, W.; Scheutjens, J. M. H. M. Adv. Colloid Interface Sci. 1991, 34, 477. (44) Raphael, E.; Pincus, P.; Fredrickson, G. H. Macromolecules 1993, 26, 1996. (45) Bijsterbosch, H. D.; Cohen Stuart, M. A.; Fleer, G. J. Macromolecules 1998, 31, 8981. (46) Vermette, P.; Meagher, L. Colloids Surf., B 2003, 28, 153. (47) Steinmetz, N. F.; Manchester, M. Biomacromolecules 2009, 10, 784. (48) Damodaran, V. B.; Fee, C. J.; Popat, K. C. Appl. Surf. Sci. 2010, 256, 4894. (49) Garbuzenko, O.; Barenholz, Y.; Priev, A. Chem. Phys. Lipids 2005, 135, 117. (50) Thevenot, J.; Troutier, A. L.; David, L.; Delair, T.; Ladaviere, C. Biomacromolecules 2007, 8, 3651. (51) Birshtein, T. M.; Borisov, O. V.; Zhulina, Y. B.; Khokhlov, A. R.; Yurasova, T. A. Polym. Sci. USSR 1987, 29, 1293. (52) Rubio, J.; Kitchener, J. A. J. Colloid Interface Sci. 1976, 57, 132. (53) Mathur, S.; Moudgil, B. M. J. Colloid Interface Sci. 1997, 196, 92. (54) Liu, H.; Xiao, H. Mater. Lett. 2008, 62, 870. (55) Boucher, E. A.; Hines, P. M. J. Polym. Sci., Part B: Polym. Phys. 1976, 14, 2241. (56) Einarson, M. B.; Berg, J. C. Langmuir 1992, 8, 2611. (57) Napper, D. H. J. Colloid Interface Sci. 1970, 33, 384. (58) Wind, B.; Killmann, E. Colloid Polym. Sci. 1998, 276, 903. (59) Carignano, M. A.; Szleifer, I. Colloids Surf., B 2000, 18, 169. (60) McPherson, T.; Kidane, A.; Szleifer, I.; Park, K. Langmuir 1998, 14, 176. (61) Giacomelli, C. E.; Norde, W. J. Colloid Interface Sci. 2001, 233, 234. (62) Norde, W.; Anusiem, A. C. I. Colloids Surf. 1992, 66, 73. (63) Su, T. J.; Lu, J. R.; Thomas, R. K.; Cui, Z. F. J. Phys. Chem. B 1999, 103, 3727. 12900

dx.doi.org/10.1021/la202990y |Langmuir 2011, 27, 12891–12901

Langmuir

ARTICLE

(64) Bosker, W. T. E.; Iakovlev, P. A.; Norde, W.; Cohen Stuart, M. A. J. Colloid Interface Sci. 2005, 286, 496. (65) Norde, W.; Gage, D. Langmuir 2004, 20, 4162. (66) Carter, D. C.; Ho, J. X. In Advances in Protein Chemistry; David, S. E., Ed.; Academic Press: San Diego, CA, 1994; Vol. 45, p 153. (67) Eckenrode, H. M.; Dai, H.-L. Langmuir 2004, 20, 9202. (68) Halperin, A.; Kr€oger, M. Langmuir 2009, 25, 11621. (69) Halperin, A.; Fragneto, G.; Schollier, A.; Sferrazza, M. Langmuir 2007, 23, 10603. (70) Efremova, N. V.; Sheth, S. R.; Leckband, D. E. Langmuir 2001, 17, 7628. (71) Chern, C.-S.; Lee, C.-K.; Kuan, C.; Liu, K.-C. Colloid Polym. Sci. 2005, 283, 917. (72) Gref, R.; L€uck, M.; Quellec, P.; Marchand, M.; Dellacherie, E.; Harnisch, S.; Blunk, T.; M€uller, R. H. Colloids Surf., B 2000, 18, 301.

12901

dx.doi.org/10.1021/la202990y |Langmuir 2011, 27, 12891–12901