Fluorescence Correlation Spectroscopy Monitors the Hydrophobic

Sep 29, 2015 - Fluorescence correlation spectroscopy (FCS) was applied to directly monitor the hydrophobic collapse of pH-responsive hairy nanoparticl...
0 downloads 0 Views 5MB Size
Download PDF
Article pubs.acs.org/Macromolecules

Fluorescence Correlation Spectroscopy Monitors the Hydrophobic Collapse of pH-Responsive Hairy Nanoparticles at the Individual Particle Level Jing Xie,† Keita Nakai,‡ Sayaka Ohno,‡ Hans-Juergen Butt,† Kaloian Koynov,*,† and Shin-ichi Yusa*,‡ †

Max Planck Institute for Polymer Research, Ackermannweg 10, 55128 Mainz, Germany Graduate School of Engineering, University of Hyogo, 2167 Shosha, Himeji, Hyogo 671-2280, Japan

‡

S Supporting Information *

ABSTRACT: Fluorescence correlation spectroscopy (FCS) was applied to directly monitor the hydrophobic collapse of pH-responsive hairy nanoparticles at the individual particle level. To this end, fluorescent nanoparticles (hydrodynamic radius 20 nm) with polystyrene core and poly(N,Ndiethylaminoethyl methacrylate) (PDEA) shell were prepared and used as a model system. Dynamic light scattering and turbidity measurements showed that the hydrophobic collapse of the hairs at high pH values is associated with strong interparticle aggregation that hinders determination of individual particles size. However, at the ultralow concentrations assessable by FCS (less than one particle per femtoliter) the aggregation was prevented. Thus, the pH-induced change in the particles size caused by the swelling or the collapse of the PDEA hairs was systematically measured and compared with that of individual freely diffusing PDEA chains under similar conditions.

1. INTRODUCTION Stimuli-responsive polymer materials that can react to environmental changes,1−4 e.g. of temperature, pH, light, electric fields, presence of ions, etc., are currently explored for their various applications1,5−9 including drug delivery, smart biointerfaces, sensors, self-healing coatings, tunable catalysis, etc. Such materials can be used in different macromolecular configurations1 including cross-linked polymer films, polymer brush layers, core−shell nanoparticles, capsules, and nanogels. Among these configurations, hairy particles are used as drug carriers,10 catalyst supports,11 smart particulate emulsifiers,12 foam and liquid marble stabilizers,13 etc. In many of these applications the hairy particles are dispersed in aqueous solutions, and the response to environmental changes takes place in the form of a collapse of the polymer hairs as a result of a change in their behavior from hydrophilic to hydrophobic. The hydrophobic collapse is often associated with strong interparticle aggregation.14 For example, pH-responsive poly(N,N-diethylaminoethyl methacrylate) (PDEA) hairy particles are dispersed well in acidic aqueous media but flocculated in basic aqueous media.15 A controlled design of such systems and a precise control over the transition require proper characterization of the hairy particles and good understanding of the hydrophobic collapse. To achieve these goals, various characterization techniques such as X-ray photoelectron spectroscopy, transmission electron microscopy, atomic force microscopy, scanning probe oxidation lithography, variable temperature 1H NMR spectroscopy analysis, contact angle measurement, small-angle neutron scattering, and dynamic light scattering have been used.16−21 © 2015 American Chemical Society

However, while these techniques can provide valuable information on the particle morphology and on the stimuliinduced hydrophobic collapse in general, it is still a major experimental challenge to monitor and study the collapse on a single particle level, especially for very small nanoparticles. Thus, for many systems it is still unclear how exactly the conformation of the polymer “hairs” change upon the application of the external stimuli and how this change compares to systems where the same polymer is grafted as a brush on a solid surface or is a freely diffusing as an individual chain. In fact, even a measurement of the change of the individual hairy particle size upon the stimuli-induced transition is often difficult or impossible because the hydrophobic collapse is commonly associated with strong aggregation that prevents the evaluation of individual particle sizes with techniques such as DLS, for example. In this respect, fluorescence correlation spectroscopy (FCS) which is a powerful technique for studying the dynamics of fluorescent species such as small molecules, macromolecules, or nanoparticles in various environments offers an interesting alternative.22 In FCS, the fluorescent intensity fluctuations caused by the diffusion of the species through a very small confocal detection volume are recorded and analyzed to obtain their diffusion coefficients, hydrodynamic radius, and concentrations. Because of its high single molecule sensitivity, FCS can Received: June 30, 2015 Revised: September 17, 2015 Published: September 29, 2015 7237

DOI: 10.1021/acs.macromol.5b01435 Macromolecules 2015, 48, 7237−7244

Article

Macromolecules be used to measure the hydrodynamic radius of hairy nanoparticles in extremely dilute dispersions, in which the hydrophobic particle aggregation may be prevented. While initially developed and still predominantly used as a tool in molecular and cell biology, FCS has also found many applications in polymer and colloid science.23−25 Furthermore, the method was previously applied to investigate stimuliresponsive polymer systems. For example, the pH or salt concentration induced changes in the size of individual polymer chains,26,27 or in the interaction between small molecules and grafted polyelectrolyte brushes,28,29 as well as the temperature responsiveness of poly(N-isopropylacrylamide) (PNIPAM) microgel particles30 and grafted hydrogel films31,32 were studied. In this paper we demonstrate how FCS can be used to directly monitor the hydrophobic collapse of pH-responsive hairy nanoparticle at individual particle level. To this end we prepared small fluorescent core−shell hairy nanoparticles with polystyrene core and PDEA shell and used them as a model system. PDEA is a pH-responsive polymer with biocompatibility and antibacterial activity that has been previously considered for applications such as gene delivery, pHresponsive controlled release, etc.33−35 Using FCS, we have systematically measured the pH-dependent swelling and collapse on a single particle level and compared it to the behavior of individual freely diffusing PDEA chains under similar conditions.

Scheme 1. Synthetic Route of (a) PDEA-P(S/AEM), (b) P(DEA/AEM), and (c) Fluorescence Labeling with Bodipy at the Amino Group

2. EXPERIMENTAL SECTION Materials. N,N-Diethylaminoethyl methacrylate (DEA, 99%) was purchased from Aldrich and passed through basic alumina columns to remove inhibitors. Styrene (S, >99%) from Wako Pure Chemical Industries was washed with an aqueous alkaline solution and distilled from calcium hydride under reduced pressure. Divinylbenzene (DVB, mixture of isomers, 93%) from Wako Pure Chemical Industries, 2aminoethyl methacrylate hydrochloride (AEM, 95%) from Polysciences, Inc., 4,4′-azobis(4-cyanopentanoic acid) (V-501, 98%) purchased from Wako Pure Chemical Industries, and 4,4-difluoro5,7-dimethyl-4-bora-3a,4a-diaza-s-indacene-3-propionic acid, succinimidyl ester (Bodipy), from Life Technologies were used without further purification. 2,2′-Azobis(2-methylpropionitrile) (AIBN, >98%) from Wako Pure Chemical Industries was purified by recrystallization from methanol. 4-Cyanopentanoic acid dithiobenzoate (CPD) was prepared according to the literature.36 Methanol was dried over 4 Å molecular sieves and distilled. Silica nanoparticles (diameter 7 nm, LUDOX SM, Aldrich) were used without further purification. The monodispersed PS spheres (diameter 1 μm) were synthesized as described before.37 Round microscope glass slides with diameter of 25 mm and thickness of 170 μm were purchased from Menzel, Germany, and cleaned with a 2% Hellmanex solution (Hellma, Germany) prior to further use. Synthesis of PDEA Macro-Chain-Transfer Agent (PDEA Macro-CTA). To obtain cross-linked PS particles with pH-responsive PDEA shells, first the PDEA macro-chain-transfer agent (PDEA macro-CTA) was prepared via reversible addition−fragmentation chain transfer (RAFT) controlled living radical polymerization (Scheme 1). A methanol solution (28 mL) of V-501 (65.4 mg, 0.233 mmol) and CPD (164 mg, 0.585 mmol) was added to an aqueous solution (65 mL, pH 6.2) of DEA (12.0 g, 64.8 mmol). The mixture was stirred at 60 °C for 8 h under Ar. After the reaction, the conversion was estimated from 1H NMR (conversion = 84.2%). The reaction mixture was dialyzed against pure water for 2 days. PDEA macro-CTA was recovered by freeze-drying (9.93 g, 82.7%). The number-average molecular weight (Mn) and molecular weight distribution (Mw/Mn) were estimated by gel-permeation chromatography (GPC) to be 1.55 × 104 g/mol and 1.14, respectively. The

degree of polymerization (DP) of PDEA macro-CTA was calculated to be 102 from the 1H NMR spectrum. Synthesis of Cross-Linked Random Copolymer of S and AEM Particle with PDEA Hair (PDEA-P(S/AEM)). Styrene (S), divinylbenzene (DVB), and 2-aminoethyl methacrylate (AEM) were randomly copolymerized in methanol using PDEA macro-CTA. S monomer can dissolve in methanol; however, PS cannot dissolve in methanol. Therefore, after the polymerization cross-linked random copolymer of P(S/AEM) particle stabilized PDEA shells can be obtained. PDEA macro-CTA (1.00 g, 5.21 × 10−2 mmol, Mn(theory) = 1.76 × 104 g/mol, Mw/Mn = 1.14), S (1.00 g, 9.61 mmol), DVB (12.8 mg, 9.81 × 10−2 mmol), AEM (16.2 mg, 9.81 × 10−2 mmol), and AIBN (9.48 mg, 5.77 × 10−2 mmol) were dissolved in methanol (12.6 mL) ([PDEA]:[S]:[AEM]:[DVB] = 1:184:1.1:1.1, molar ratio). The mixture was stirred at 60 °C for 48 h under Ar. After the reaction, the conversion was estimated from 1H NMR (conversion: S including DVB = 63.0%, AEM = 33.0%). The reaction mixture was dialyzed against methanol for 1 day, water (pH 4) for 1 day, and pure water for 1 day. Cross-linked random copolymer of S and AEM particle with PDEA shell (PDEA-P(S/AEM)) was recovered by freeze-drying (0.50 g). The composition of the core, S:DVB:AEM = 98.9:0.6:0.5 (molar ratio), was estimated from the conversion. Preparation of Bodipy-Labeled PDEA-P(S/AEM) (PDEA-P(S/ Bodipy)). The fluorescent labeling was done by reacting the pendant primary amino groups in the shell with succinimidyl ester group containing Bodipy dye. A methanol solution (20 mL) of cross-linked PDEA-P(S/AEM) (0.2 g) was added to a methanol solution (1.0 mL, 2.5 g/L) of Bodipy. The mixture was stirred at 30 °C for 24 h. The reaction mixture was dialyzed for 1 day against methanol, water (pH 4), and pure water. PDEA-P(S/Bodipy) was recovered by freezedrying (0.17 g, 85.0%). 7238

DOI: 10.1021/acs.macromol.5b01435 Macromolecules 2015, 48, 7237−7244

Article

Macromolecules Preparation of Bodipy-Labeled P(DEA/AEM) Random Copolymer (P(DEA/Bodipy)). Random copolymer of P(DEA/AEM) was synthesized via RAFT polymerization and labeled with succinimidyl ester group containing Bodipy to obtain P(DEA/ Bodipy). An aqueous solution (8.1 mL) of DEA (1.50 g, 8.10 mmol) and AEM (12.1 mg, 7.35 × 10−2 mmol) was neutralized using a HCl aqueous solution to pH 6.6. A methanol solution (3.5 mL) of AIBN (9.62 mg, 5.86 × 10−2 mmol) and CPD (20.5 mg, 7.34 × 10−2 mmol) was added to the aqueous solution. The mixture was stirred at 60 °C for 8 h under Ar. After the reaction, the conversion was estimated from 1H NMR (conversion: DEA = 98.7% and AEM = 43.7%). The reaction mixture was dialyzed against pure water for 2 days. P(DEA/AEM) was recovered by freeze-drying (1.49 g, 83.1%). Mn and Mw/Mn were estimated by GPC to be 1.80 × 104 g/mol and 1.17, respectively. DP of P(DEA/AEM) was calculated to be 104 from the 1H NMR spectrum. In the 1H NMR spectra all signals attributed to AEM units in P(DEA/AEM) were overlapped with signals of DEA units (Figure S1). Therefore, it is difficult to determine the composition of P(DEA/AEM) from the 1H NMR spectrum. The composition can be determined from the conversions of each monomer; DEA:AEM = 99.6:0.4 (molar ratio). A methanol solution (50 mL) of P(DEA/AEM) (0.5 g) was added to a methanol solution (1.48 mL, 2.5 g/L) of Bodipy. The mixture was stirred at room temperature for 24 h. The reaction mixture was dialyzed against methanol for 2 days and against pure water for 1 day. Bodipy-labeled polymer (P(DEA/Bodipy)) was recovered by freezedrying (0.48 g, 95.0%). Preparation of Inverse Opals Structures. Inverse opals structures on a thin glass slide substrate were prepared as described in detail elsewhere.38,39 Briefly, the substrate was lifted with 400 nm/s lifting speed from a mixed aqueous dispersion of 1.5 wt % PS particles and 0.3 wt % silica nanoparticles at 20 °C environmental temperature and 50% RH environmental humidity as measured and controlled. The PS template was removed by pyrolysis in air at 500 °C for 5 h, leaving the silica nanoparticle stacked and fused in an ordered threedimensional structure. General Characterization. 1H NMR spectra were obtained with a Bruker BioSpin DRX-500 spectrometer. GPC measurements were performed using a refractive index detector equipped with a Shodex Ohpak SB-804 HQ column working at 40 °C with a flow rate of 0.6 mL/min. An acetic acid (0.5 M) solution containing sodium sulfate (0.3 M) was used as the eluent. The values of Mn and Mw/Mn were calibrated using standard poly(2-vinylpyridine) samples. UV−vis absorption spectra were measured using a Jasco V-630BIO UV−vis spectrophotometer. Fluorescence spectra were recorded on a Hitachi F2500 fluorometer. The percent transmittance (%T) at 600 nm for aqueous solutions of the polymers was also measured with a Jasco V630BIO UV−vis spectrophotometer at 25 °C. The polymer concentration (Cp) was 10 g/L. Static light scattering (SLS) measurements were performed using an Otsuka Electronics Photal DLS-7000HL light scattering spectrometer at 25 °C. A He−Ne laser (10.0 mW at 633 nm) was used as a light source. The weight-average molecular weight (Mw), z-average radius of gyration (Rg), and second virial coefficient (A2) values were estimated from the relationship KCp Rθ

=

⎞ 1 ⎛⎜ 1 1 + R g 2q2⎟ + 2A 2 Cp ⎠ Mw ⎝ 3

Zetasizer Nano ZS equipped with a He−Ne laser (4 mW at 633 nm) at 25 °C. Data were taken at a 173° scattering angle and analyzed by Malvern Zetasizer Software version 6.20. All samples were filtered through 0.2 μm filters prior to DLS measurements. Transmission electron microscopy (TEM) measurements were performed with a Jeol TEM-1200 electron microscope operated at an accelerating voltage of 200 kV. Samples for TEM were prepared by placing one drop of the aqueous solution on a copper grid coated with thin films of Formvar. Excess water was blotted using filter paper. The samples were stained by sodium phosphotungstate and dried under vacuum for 1 day. The silica inverse opal structure was characterized by scanning electron microscopy (SEM) using a LEO Gemini 1530 microscope at a voltage of 0.7 kV. The pH value of the studied solutions was measured by a Seven Excellence pH meter from Mettler Toledo. Fluorescence Correlation Spectroscopy (FCS). The FCS measurements were performed on a commercial setup (Carl Zeiss, Germany) consisting of the modules LSM510, ConfoCor 2 and an inverted microscope model Axiovert 200 with a C-Apochromat 40×, NA 1.2) water immersion objective. An argon laser (488 nm) was used for excitation, and the emission was collected after filtering with a LP505 long pass filter. For detection an avalanche photodiode operating in single photon counting mode was used. Aqueous dispersions of the cross-linked PDEA-P(S/AEMA) hairy particles were studied either in 8-well, polystyrene chambered cover glass (Laboratory-Tek, Nalge Nunc International) or in an Attofluor cell chamber (Life Technologies) in which a microscope cover glass supporting the inverse opal structure was mounted. The desired pH value was adjusted by addition of HCl or NaOH. For each studied sample a series of 15 measurements with a total duration 5 min were performed. The time-dependent fluctuations of the fluorescence intensity δI(t) caused by the diffusion of the fluorescent species through the confocal observation volume were recorded and analyzed by an autocorrelation function G(τ) = 1 + ⟨δI(t) δI(t+τ)⟩/⟨I(t)⟩2. Technically, this was done by single photon counting, followed by binning and correlating using the commercial (Carl Zeiss) software correlator of the setup. As has been shown theoretically, for an ensemble of identical freely diffusing fluorescence species, G(τ), has the following analytical form:22 ⎡ ⎤1 fT G(τ ) = 1 + ⎢1 + e−t / τT ⎥ ⎢⎣ ⎥⎦ N ⎡1 + 1 − fT ⎣⎢

1 τ ⎤ ⎥ τD ⎦

1+

τ S 2τD

(2)

Here, N is the average number of diffusing fluorescence species in the observation volume, f T and τT are the fraction and the decay time of the triplet state, τD is the diffusion time of the species, and S is the socalled structure parameter, S = z0/r0, where z0 and r0 represent the axial and radial dimensions of the confocal volume, respectively. Furthermore, the diffusion time, τD, is related to the respective diffusion coefficient D, through D = r02/4τ.22 The experimentally obtained G(τ) can be fitted with eq 2 yielding the corresponding diffusion times and subsequently the diffusion coefficients of the fluorescent species. Next, the hydrodynamic radius Rh can be calculated (assuming spherical particles) using the Stokes−Einstein relation Rh = kBT/6πηD, where kB is the Boltzmann constant, T is the temperature, and η is the viscosity of the solution. Furthermore, the molecular (or particle) fluorescent brightness can be determined as ⟨I(t)⟩/N. As the radial dimension r0 of the confocal probing volume is not known a priori, it was determined by performing calibration experiments using a fluorophore with known diffusion coefficient, i.e., Alexa 488 in water.40 It should be noted that when fluorescently labeled nanoparticles are studied, effects such as energy migration or changes in instantaneous orientation of the transition dipole moments due to rotation may complicate the shape of the FCS autocorrelation and affect the fitting procedure even in the absence of aggregations. However, all experimental autocorrelation curves reported in this work could be well represented with eq 2, indicating that the abovementioned effects do not play a significant role for the studied nanoparticles, possibly due to their fairly small size. The presence of aggregates may also complicate the FCS autocorrelation curves. The

(1)

where Rθ is the difference between the Rayleigh ratio of the solution and that of the solvent, q is the magnitude of the scattering vector, and K = 4π2n2(dn/dCp)2/NAλ4, with dn/dCp being the refractive index increment against Cp and NA being Avogadro’s number. The q value was calculated using q = (4πn/λ) sin(θ/2), where n is refractive index of the solvent, λ is wavelength of the light source (= 632.8 nm), and θ is scattering angle. By measuring Rθ for a set of Cp and θ, the values of Mw, Rg, and A2 were estimated from Zimm plots. The known Rayleigh ratio of toluene was used to calibrate the instrument. The values of dn/ dCp at 633 nm were determined with an Otsuka Electronics Photal DRM-3000 differential refractometer at 25 °C. Dynamic light scattering (DLS) measurements were performed using a Malvern 7239

DOI: 10.1021/acs.macromol.5b01435 Macromolecules 2015, 48, 7237−7244

Article

Macromolecules nanoparticles considered in this work are relatively bright (carry ∼10 fluorescent dyes), and upon aggregation they form even brighter aggregates that may dominate the corresponding autocorrelation curves. The presence of such aggregates is clearly manifested by the appearance of strong spikes in the fluorescent intensity time trace as experimentally observed at high pH values and high nanoparticle concentrations (data not shown in the article). However, at the concentrations used in the FCS experiments described below such spikes caused by aggregates were practically not present.

to solvent polarity and pH changes. UV−vis absorption and fluorescence spectra for PDEA-P(S/Bodipy) in water were measured to confirm the introduction of Bodipy (Figure S4). While the results clearly show that the dye is attached to the nanoparticles, they also indicate a strong decrease of fluorescence with the increase of pH. A similar effect was observed in the FCS studies (see Table S2). On the other hand, the spectra in Figure S4 show that the dye absorption properties are almost independent of the pH. The number of Bodipy dyes per nanoparticle was estimated by dividing the molar extinction coefficient of PDEA-P(S/Bodipy) nanoparticles to that of Bodipy molecules. This yielded a value of ≈9 which is consistent with estimates based on FCS measurements of the fluorescent brightness of the nanoparticles and the individual dye molecules (see Supporting Information for details). 3.2. Monitoring the pH-Responsive Behavior at High Nanoparticles Concentration. We first investigated the pHresponsive behavior of the cross-linked PDEA-P(S/AEM) in aqueous dispersions with relatively high concentrations in the range 1−10 g/L using classical techniques like optical transmission measurements and dynamic light scattering (DLS). Figure 1 shows the pH dependence of the transmittance (% T) of an aqueous dispersion with concentration of 10 g/L

3. RESULTS 3.1. Preparation of pH-Responsive Hairy Nanoparticles. To obtain cross-linked PS particles with pH-responsive PDEA shells, first the PDEA macro-CTA was prepared via RAFT radical polymerization (Scheme 1). S, DVB, and AEM were then randomly copolymerized in methanol using PDEA macro-CTA, which is a precipitation polymerization technique to prepare cross-linked P(S/AEM) particles stabilized with PDEA hairs (PDEA-P(S/AEM)). Figure S1 shows 1H NMR spectra for PDEA macro-CTA, PDEA-P(S/AEM), and P(DEA/AEM) in D2O at pH 4. It is difficult to dissolve PDEA in D2O at pH ≥ 7 because the pendant N,N-diethylaminoethyl groups are hydrophobic at pH ≥ 7 due to deprotonation. Therefore, the polymers containing DEA units were dissolved in D2O at pH 4. The pH value was adjusted to 4 using DCl. The degree of polymerization (DP) values of PDEA macroCTA and P(DEA/AEM) were 102 and 104, respectively, calculated from 1H NMR area integral intensity ratios of the pendant methylene protons at 4.4 ppm and the terminal phenyl protons at 7.4−8.0 ppm. The signals attributed to P(S/AEM) cannot be observed in 1H NMR of PDEA-P(S/AEM) (Figure S1b). This observation indicates that motion of the protons in P(S/AEM) was restricted due to cross-linked core formation by cross-linker as DVB. On the other hand, motion of PDEA chains was not restricted because PDEA chains form shells surrounding the surface of the cross-linked P(S/AEM) core. Figure S2 shows GPC elution curves for PDEA macro-CTA and P(DEA/AEM). Symmetrical GPC peaks with narrow distribution can be observed, suggesting that the polymers have controlled structure with narrow Mw/Mn. The theoretical Mn values (Mn(theory)) can be calculated from the feed ratio and conversion. Table S1 lists degrees of polymerization, content of AEM unit, Mn and Mw/Mn estimated from GPC, and Mn(theory) for the polymers. DP of PDEA shells in PDEAP(S/AEM) particles was 102 because the particle was prepared using PDEA macro-CTA with DP = 102. The content of AEM in the P(S/AEM) core was 0.5 mol % estimated from the conversion. Figure S3 shows a typical example of Zimm plots for PDEAP(S/AEM) particles in 0.1 M NaCl aqueous solution at pH 4. The apparent values of Mw, Rg, and A2 for PDEA-P(S/AEM) particles determined by SLS measurements were 2.67 × 106 g/ mol, 21.4 nm, and 2.47 × 10−5 mL mol/g2, respectively. The dn/dCp value at 633 nm for the PIC micelles determined using a differential refractometer was 0.124 mL/L. To enable fluorescence correlation spectroscopy, a part of the hairy nanoparticles were fluorescently labeled by reacting the pendant primary amino groups in the core with succinimidyl ester group-containing Bodipy dye. For comparison of the pH-responsive behavior of the grafted hairs to that of individual freely diffusing chains, Bodipy-labeled PDEA (P(DEA/Bodipy)) was also prepared. The Bodipy was chosen as a fluorescent label because it is hydrophobic and according to the supplier information its fluorescence is relatively insensitive

Figure 1. Percent transmittance (%T) at 600 nm for aqueous dispersions of PDEA-P(S/AEM) particles as a function of pH at Cp = 10 g/L. Insets are photographs of aqueous dispersions of PDEA-P(S/ AEM) particles at pH 4, 7.5, and 10.

measured at the wavelength of 600 nm. The pH value was varied by adding HCl or NaOH to the dispersion. Below pH 7, the %T value was virtually constant and close to 100%. In this pH range the pendent tertiary amine groups in PDEA shells are protonated, the polymer hairs are hydrophilic, and thus the small nanoparticles are well dispersed in the aqueous environment. However, %T decreased sharply as the pH increased above 7. This increase in turbidity is caused by interparticle aggregation occurring at higher pH due to hydrophobic interactions between the pendant deprotonated amine groups in PDEA shells. The data displayed in Figure 1 indicate that the transition from well-dispersed single particles to interparticle aggregates occurred within a narrow pH range. Furthermore, while at pH 7.5 the aqueous dispersion is simply turbid at pH 10 it shows precipitates. Next, an aqueous dispersion of the PDEA-P(S/AEM) nanoparticles with concentration of 1 g/L, i.e., 10 times lower than that used for the turbidity measurements, was characterized by dynamic light scattering (DLS). To remove 7240

DOI: 10.1021/acs.macromol.5b01435 Macromolecules 2015, 48, 7237−7244

Article

Macromolecules impurities and occasional large aggregates, the dispersions were filtrated through 0.2 μm filters prior to the DLS measurements. Typical results in the form of hydrodynamic radius (Rh) distributions are shown in Figure 2a. The corresponding

Figure 2. (a) DLS data for the distribution of the hydrodynamic radius (Rh) of PDEA-P(S/AEM) particles in aqueous dispersions with pH 4 () and pH 10 (- - -) at Cp = 1 g/L. (b) Average Rh of the particles as a function of pH.

Figure 3. TEM images of PDEA-P(S/AEM) particles after drying from aqueous dispersions with (a) pH 4 and (b) pH 10.

the DLS experiments. Typical FCS autocorrelation curves measured at various pH values are shown in Figure 4a. With increasing pH the curves shift to the shorter lag times, reflecting the decrease in particle size and thus the faster diffusion. The experimental autocorrelation curves could be well fitted with eq 2 (see Experimental Section) yielding the diffusion coefficients and, through the Stokes−Einstein relation, the hydrodynamic radii, Rh, of the hairy nanoparticles. Figure 4b shows the Rh values obtained over broad pH range. The value at low pH, Rh = 23 nm, is very similar to that measured by DLS (Figure 2), which confirms the accuracy of the FCS data. However, in contrast to the DLS studies, with the increase of pH, the measured hydrodynamic radius decreases continuously down to a value Rh = 10 nm at pH 12. These results show that at the very low concentrations assessable by FCS the particle aggregation caused by hydrophobic interactions can be strongly diminished and the pH-responsiveness of the PDEA hairs can be monitored on the individual nanoparticle level. Moreover, compared to the DLS result (Figure 2), the FCS data shown in Figure 4b reveal a less sharp transition that extends over a broad pH range. This effect may be related to the much lower concentration at which the FCS measurements were performed and possibly to the coexistence of PDEA hairs at different stage of the collapse at a given overall pH value. It is important to emphasize at this point that depending on the experimental conditions different techniques may provide different values for the nanoparticles size (Table S3 and the related discussion in the Supporting Information), and therefore the exact experimental conditions should be always carefully considered. Next, we studied the reversibility of the pH-induced collapse on the grafted PDEA chains. Therefore, we consecutively tuned the pH of the nanoparticle dispersion between 4.0 and 11. As shown in Figure 4c, the hydrodynamic radius of the particles changed reversibly from 23 nm at low pH to 11 nm at high pH values. The cycle was repeated only three times because the

average Rh values at pH 4 and 10 were 25.4 and 71.2 nm, respectively. To investigate the pH-responsive behavior of PDEA-P(S/AEM) in water, the pH dependence of Rh was measured (Figure 2b). Below pH 7, the Rh value was virtually constant at about 25 nm. However, Rh increased sharply as the pH increased from 7 to 8, reaching a new plateau value of around 70 nm. These results indicated that even at the relatively low particle concentration of 1 g/L used in the DLS experiments the PDEA-P(S/AEM) nanoparticles still aggregate at high pH values due to hydrophobic interactions. This is further evident from the TEM images (Figure 3) of dried nanoparticle dispersions with different pH values. While for pH 4 individual particles are observed (Figure 3a) for pH 10 the particles tended to aggregate together. From the TEM images we could estimate that the average radius of the cross-linked PDEA-P(S/ AEM) particles was 9.4 nm at pH 4.0. As can be expected, this value is significantly smaller than that of the hydrodynamic radius (Rh ≈ 25 nm) estimated by DLS due to the shrinking of the particles during drying and hydrodynamic effects. In spite of the aggregated morphology, the TEM image obtained for the dispersion with pH 10 (Figure 3b) reveals a decrease in the individual nanoparticle size compared to that at pH 4 (Figure 3a) and thus confirms the hydrophobic collapse of the PDEA hairs at high pH. 3.3. Monitoring the pH-Responsive Behavior at the Individual Particle Level. In order to monitor pH-dependent hydrophilic−hydrophobic transition of the PDEA hairs of individually dispersed nanoparticles, we performed FCS experiments with highly diluted dispersions of Bodipy-labeled nanoparticles. Again, the pH value was tuned by adding HCl and NaOH to the dispersions. The concentration of PDEAP(S/Bodipy) under different pH conditions was always kept at 0.02 g/L, which is 50 times less than the concentration used in 7241

DOI: 10.1021/acs.macromol.5b01435 Macromolecules 2015, 48, 7237−7244

Article

Macromolecules

Figure 5. (a) Normalized, experimental FCS autocorrelation curves measured for aqueous dispersion of Bodipy-labeled nanoparticles with PDEA hairs and for solution of PDEA chains at pH 3. The solid lines represent the corresponding fits with eq 2. (b) Hydrodynamic radius, Rh, as a function of pH for the hairy nanoparticles (0.02 g/L) and for the PDEA chains (0.003 g/L).

brightness of the Bodipy labels. Furthermore, the PDEA chains carry only one dye per chain (see Supporting Information), and thus their fluorescent brightness is significantly weaker than that of the nanoparticles (with ∼10 dyes per particle). Because of these reasons and possibly some additional hydrophobic effects, e.g., chain migration to the water/air interface of the sample chamber, we were not able to record reasonable FCS autocorrelation curves and determine the hydrodynamic radius of individual PDEA chains at high pH values. The data shown in Figure 5b indicate that when the pH was increased from 2.3 to 5.1, the hydrodynamic radius of the P(DEA/Bodipy) chains (Mn = 18 000 g/mol) decreased from 2.9 to 2.2 nm, i.e., with 0.7 nm (hydrodynamic diameter decreased with 1.4 nm). For comparison, the hydrodynamic radius of the nanoparticles processing grafted PDEA hairs with similar molecular weight (Mn = 15 500 g/mol) decreased by 1.6 nm in the same pH range (Figure 5b). While the estimated hydrodynamic radius (diameters) should be considered only as an approximation for the real chain dimensions, the obtained values indicate that the chains grafted to the PS nanoparticles contract in a similar way and to similar extend compared to freely diffusing chains. 3.4. Controlling the Mobility of the pH-Responsive Particles in Nanoporous Media. Some of the most important applications of the stimuli-responsive nanoparticles are related to the possibility to control their mobility in nanoporous media. In such environment even small changes of the particle size may have a profound effect on the mobility. Furthermore, other stimuli-dependent effects; e.g., change of surface chemistry of the porous media may affect the interaction with the nanoparticles and thus their mobility. With this in mind, we studied the diffusion of the PDEA-P(S/

Figure 4. (a) Normalized, experimental FCS autocorrelation curves measured for aqueous dispersions of Bodipy-labeled nanoparticles at Cp = 0.02 g/L and different pH values. The solid lines represent the corresponding fits with eq 2. (b) Hydrodynamic radius, Rh, of the nanoparticles as a function of pH. The solid line is to guide the eye. (c) Hydrodynamic radius, Rh, as a function of pH upon cycling the pH value between 4 and 11.

continuous addition of NaOH and HCl solutions to the dispersion increased its total volume and thus decreased the concentration of the nanoparticles to the edge of the FCS sensitivity. In an attempt to compare the pH-induced collapse of PDEA chains attached to the surface of a nanoparticle and that of freely diffusing chains, Bodipy-labeled polymer (P(DEA/ Bodipy)) was prepared as described in the Experimental Section, and the hydrodynamic radius of the individual chains was measured with FCS. The experiments were performed in extremely diluted solutions (0.003 g/L) at various pH. Under acid conditions, the polymer chains were well dispersed in the solution. FCS autocorrelation curves (Figure 5a) could easily be recorded, and the individual chain’s hydrodynamic radius could be calculated (Figure 5b). As expected, the hydrodynamic radius continuously decreased with the increase of pH, reflecting the hydrophobic collapse of the chains. However, as discussed above and in the Supporting Information, the increase of pH leads to a strong decrease in the fluorescent 7242

DOI: 10.1021/acs.macromol.5b01435 Macromolecules 2015, 48, 7237−7244

Article

Macromolecules

effect, we fitted the autocorrelation curves with eq 2. It should be noted that because the nanoparticles’ diffusion is confined in the inverse opal voids, which have sizes similar to that of the FCS probing volume, the use of eq 2 that is derived for free 3D diffusion is not trivial. However, as eq 2 represents very well the experimentally measured autocorrelation curves (Figure 6c), we still used it as a preliminary, approximate way to estimate the diffusion coefficients of the nanoparticles in the inverse opals. In free aqueous environment the diffusion coefficient increased from 1.0 × 10−11 to 2.0 × 10−11 m2/s when the pH was changed from 4 to 9. The corresponding values in silica inverse opal were from 4.0 × 10−12 to 1.05 × 10−11 m2/s, respectively. Thus, the pH-induced change in diffusion coefficient was notably larger in the silica inverse opal, reflecting the stronger effect of the pH on the nanoparticles mobility in the porous environment.

Bodipy) nanoparticles in a water filled silica inverse opal at various pH values. A silica inverse opal consists of a regular arrangement of spherical voids surrounded by solid silica walls with interconnecting circular pores. It was chosen as a model for well-defined nanoporous structure.38,39,41,42 The particular structure used in this study (Figure 6a) had a void diameter

4. CONCLUSION We showed that due to its very high sensitivity FCS can be used to measure the stimuli-responsiveness of individual hairy nanoparticles. Using this technique, we found that when the pH is increased, PDEA chains grafted to a small polystyrene core collapse similarly to freely diffusing PDEA chains. The particle concentration plays an important role on the general behavior of stimuli-responsive hairy nanoparticles. At high and moderately low concentrations, the pH-induced collapse leads to aggregation and thus lowering of the nanoparticles mobility. At very low concentrations, however, the hydrophobic collapse is associated with decrease of nanoparticles’ size and increase in their mobility. The latter effect is particularly evident and important when the nanoparticles are dispersed in a nanoporous environment.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.macromol.5b01435. Figures S1−S5 and Tables S1−S3 (PDF)

Figure 6. Scanning electron microscopy (SEM) of silica inverse opal: (a) side view; (b) top view. (c) Experimental autocorrelation curves for the diffusion of the Bodipy-labeled hairy nanoparticle in the inverse opal (red squares) and in free aqueous environment (green circles) at pH 9 (open symbols) and pH 4 (solid symbols). The solid lines represent the corresponding fits with eq 2. The inset show a confocal microscopy image (in reflection mode) of water infiltrated silica inverse opal.

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] (K.K.). *E-mail: [email protected] (S.Y.). Notes

The authors declare no competing financial interest.

of 920 nm and interconnecting pores diameter of 240 nm. It was infiltrated with aqueous dispersion of the PDEA-P(S/ Bodipy) nanoparticles. The particles’ diffusion coefficient was measured with FCS. Prior to the FCS measurements the structure was imaged using the reflection mode of the confocal microscope in order to confirm the regular porosity and the absence of defects (Figure 6c, inset) at the point where the FCS measurements were performed. Figure 6c shows typical FCS autocorrelation curves measured in the inverse opal at pH = 9.0 and pH = 4.0. For comparison, the autocorrelation curves measured in free aqueous environment (without inverse opal) are also shown. The curves measured in the inverse opal are shifted to the longer lag times, reflecting the significantly slower diffusion in the nanoporous structure. Furthermore, the change in pH from 4.0 to 9.0 affects stronger the mobility of the nanoparticles in the inverse opal. In an attempt to quantify this

■

ACKNOWLEDGMENTS We gratefully acknowledge the China Scholarship Council Fellowship #201206170166 (J.X.) and the financial support of DFG (SFB 1066, Q2).

■

REFERENCES

(1) Stuart, M. A. C.; Huck, W. T. S.; Genzer, J.; Müller, M.; Ober, C.; Stamm, M.; Sukhorukov, G. B.; Szleifer, I.; Tsukruk, V. V.; Urban, M.; Winnik, F.; Zauscher, S.; Luzinov, I.; Minko, S. Nat. Mater. 2010, 9, 101−113. (2) Jochum, F. D.; zur Borg, L.; Roth, P. J.; Theato, P. Macromolecules 2009, 42, 7854−7862. (3) Zhou, F.; Biesheuvel, P. M.; Choi, E. Y.; Shu, W. M.; Poetes, R.; Steiner, U.; Huck, W. T. S. Nano Lett. 2008, 8, 725−730. 7243

DOI: 10.1021/acs.macromol.5b01435 Macromolecules 2015, 48, 7237−7244

Article

Macromolecules

(39) Wang, J. J.; Ahl, S.; Li, Q.; Kreiter, M.; Neumann, T.; Burkert, K.; Knoll, W.; Jonas, U. J. Mater. Chem. 2008, 18, 981−988. (40) Petrasek, Z.; Schwille, P. Biophys. J. 2008, 94, 1437−1448. (41) Cherdhirankorn, T.; Retsch, M.; Jonas, U.; Butt, H.-J.; Koynov, K. Langmuir 2010, 26, 10141−10146. (42) Raccis, R.; Nikoubashman, A.; Retsch, M.; Jonas, U.; Koynov, K.; Butt, H. − J.; Likos, C. N.; Fytas, G. ACS Nano 2011, 5, 4607− 4616.

(4) Magnusson, J. P.; Khan, A.; Pasparakis, G.; Saeed, A. O.; Wang, W. X.; Alexander, C. J. Am. Chem. Soc. 2008, 130, 10852−10853. (5) Andreeva, D. V.; Fix, D.; Moehwald, H.; Shchukin, D. G. Adv. Mater. 2008, 20, 2789−2794. (6) Cavallaro, A.; Taheri, S.; Vasilev, K. Biointerphases 2014, 9, 029005−10. (7) Tokareva, I.; Minko, S.; Fendler, J. H.; Hutter, E. J. Am. Chem. Soc. 2004, 126, 15950−15951. (8) Urban, M. W. J. Macromol. Sci., Polym. Rev. 2006, 46, 329−339. (9) Lu, Y.; Mei, Y.; Drechsler, M.; Ballauff, M. Angew. Chem., Int. Ed. 2006, 45, 813−816. (10) Shen, Z. Y.; Wei, W.; Zhao, Y. Y.; Ma, G. H.; Dobashi, T.; Maki, Y.; Su, Z. G.; Wan, J. P. Eur. J. Pharm. Sci. 2008, 35, 271−282. (11) Mei, Y.; Lu, Y.; Polzer, F.; Ballauff, M.; Drechsler, M. Chem. Mater. 2007, 19, 1062−1069. (12) Amalvy, J. I.; Armes, S. P.; Binks, B. P.; Rodrigues, J. A.; Unali, G.-F. Chem. Commun. 2003, 39, 1826−1827. (13) Dupin, D.; Armes, S. P.; Fujii, S. J. Am. Chem. Soc. 2009, 131, 5386−5387. (14) Gibson, M. I.; O'Reilly, R. K. Chem. Soc. Rev. 2013, 42, 7204− 7213. (15) Kotsuchibashi, Y.; Faghihnejad, A.; Zeng, H. B.; Narain, R. Polym. Chem. 2013, 4, 1038−1047. (16) Fujii, S.; Suzaki, M.; Nakamura, Y.; Sakai, K.; Ishida, N.; Biggs, S. Polymer 2010, 51, 6240−6247. (17) Li, D. J.; Jones, G. L.; Dunlap, J. R.; Hua, F. J.; Zhao, B. Langmuir 2006, 22, 3344−3351. (18) Yu, F.; Tang, X.; Pei, M. S. Microporous Mesoporous Mater. 2013, 173, 64−69. (19) Chen, M. Q.; Serizawa, T.; Li, M.; Wu, C.; Akashi, M. Polym. J. 2003, 35, 901−910. (20) Adelsberger, J.; Meier-Koll, A.; Bivigou-Koumba, A. M.; Busch, P.; Holderer, O.; Hellweg, T.; Laschewsky, A.; Müller-Buschbaum, P.; Papadakis, C. M. Colloid Polym. Sci. 2011, 289, 711−720. (21) Kayitmazer, A. B.; Seeman, D.; Minsky, B. B.; Dubin, P. L.; Xu, Y. Soft Matter 2013, 9, 2553−2583. (22) Rigler, R.; Elson, E. S. Fluorescence Correlation Spectroscopy: Theory and Applications; Springer: Berlin, 2001. (23) Papadakis, C. M.; Košovan, P.; Richtering, W.; Wöll, D. Colloid Polym. Sci. 2014, 292, 2399−2411. (24) Wöll, D. RSC Adv. 2014, 4, 2447−2465. (25) Koynov, K.; Butt, H. − J. AIP Conf. Proc. 2012, 1518, 357−364. (26) Pristinski, D.; Kozlovskaya, V.; Sukhishvili, S. A. J. Chem. Phys. 2005, 122, 014907−9. (27) Jia, P. X.; Zhao, J. J. Chem. Phys. 2009, 131, 231103−4. (28) Daniels, C. R.; Tauzin, L. J.; Foster, E.; Advincula, R. C.; Landes, C. F. J. Phys. Chem. B 2013, 117, 4284−4290. (29) Zhang, C. F.; Chu, X.; Zheng, Z. L.; Jia, P. X.; Zhao, J. J. Phys. Chem. B 2011, 115, 15167−15173. (30) Lehmann, S.; Seiffert, S.; Richtering, W. J. Am. Chem. Soc. 2012, 134, 15963−15969. (31) Raccis, R.; Roskamp, R.; Hopp, I.; Menges, B.; Koynov, K.; Jonas, U.; Knoll, W.; Butt, H.-J.; Fytas, G. Soft Matter 2011, 7, 7042− 7053. (32) Vagias, A.; Kosovan, P.; Koynov, K.; Holm, C.; Butt, H.-J.; Fytas, G. Macromolecules 2014, 47, 5303−5312. (33) Dai, S.; Ravi, P.; Tam, K. C. Soft Matter 2008, 4, 435−449. (34) Tang, Y. Q.; Liu, S. Y.; Armes, S. P.; Billingham, N. C. Biomacromolecules 2003, 4, 1636−1645. (35) Tian, W.; Lv, X. Y.; Huang, L. B.; Ali, N.; Kong, J. Macromol. Chem. Phys. 2012, 213, 2450−2463. (36) Mitsukami, Y.; Donovan, M. S.; Lowe, A. B.; McCormick, C. L. Macromolecules 2001, 34, 2248−2256. (37) D’Acunzi, M.; Mammen, L.; Singh, M.; Deng, X.; Roth, M.; Auernhammer, G. K.; Butt, H. − J.; Vollmer, D. Faraday Discuss. 2010, 146, 35−48. (38) Wang, J. J.; Li, Q.; Knoll, W.; Jonas, U. J. Am. Chem. Soc. 2006, 128, 15606−15607. 7244

DOI: 10.1021/acs.macromol.5b01435 Macromolecules 2015, 48, 7237−7244