ARTICLE pubs.acs.org/Langmuir
Preparation of Monodisperse Poly(N-isopropylacrylamide) Microgel Particles with Homogenous Cross-Link Density Distribution Roberta Acciaro, Tibor Gilanyi, and Imre Varga* Institute of Chemistry, E€otv€os University, 1117 Budapest, Pazmany s. 1/A, Hungary ABSTRACT: Monodisperse microgel latex with homogeneous cross-link density distribution within the particles was prepared by feeding the monomer and cross-linker into the reaction mixture in a regulated way during the polymerization. To determine the appropriate monomer feeding parameters, the kinetics of the particle formation was investigated by HPLC. The swelling and optical characteristics of the prepared homogenously cross-linked microgel particles were compared to the properties of inhomogenously cross-linked microgels prepared by the normal precipitation polymerization method. The distribution of the cross-link density within the particles inserts a great influence on the characteristics of the system. The degree of swelling of the homogeneous particles is significantly higher than that of the heterogeneous microgel particles. Furthermore, at room temperature the pNIPAm latex containing the homogeneously crosslinked particles is transparent, while the heterogeneously cross-linked particles form a highly turbid system at the same 0.1 wt % concentration.
’ INTRODUCTION Microgels are polymer colloid particles, which swell in a good solvent. Though microgels have been known for more than 70 years,1,2 their investigation has attracted a greatly expanding interest since Pelton and Chibante discovered3 in 1986 that poly(N-isopropylamide) (pNIPAm) based microgels could be prepared. pNIPAm microgels are usually monodisperse, spherical particles with an average size ranging from 50 nm to a few micrometers. The exceptional interest attracted by the pNIPAm microgels is related to their temperature responsive nature; that is, they can undergo a reversible volume phase transition (VPT). The pNIPAm polymer has a lower critical solution temperature (LCST) at around 32 °C.4,5 Below this temperature the microgel particles are in a highly swollen state in aqueous environment. However, when the temperature is raised above the LCST, the polymer network collapses, expelling most of its water content and any potentially dissolved molecule from the gel structure. Because a wide range of monomers can be readily copolymerized into the pNIPAm particles, the collapse temperature of the microgels can be easily fine-tuned and responsive nature to, e.g., pH, ionic strength, solvent quality, and the presence of specific molecules, can be introduced.6 Due to their unique features, pNIPAm-based microgels have been suggested as promising materials in several potential applications, e.g., in drug delivery,716 biosensors,1719 chemo-mechanical devices,20 separation/purification technologies,21 and the formation of responsive interfaces.2225 The large number of investigations performed on pNIPAm microgels have been summarized in several excellent reviews.2629 Monodisperse pNIPAm-based microgel particles are usually prepared by precipitation polymerization.3 The major requirement for the preparation of the microgel particles is that the r 2011 American Chemical Society
reaction is conducted above the LCST of pNIPAm. Under these conditions the water-soluble monomers form water insoluble polymers, which in turn aggregate until they form a colloidally stable particle. The colloid stability of the growing microgel particles is provided by electrostatic stabilization ensured by the sulfate groups originating from the persulfate initiator. The intact particle nature of the pNIPAm microgels is ensured by chemical cross-linking achieved by the copolymerization of N,N0 -methylenebisacrylamide (BIS). McPhee et al. have also showed that the microgel size can be controlled by varying the concentration of sodium dodecyl sulfate (SDS) in the reaction mixture.30 Because the properties of the swollen microgel particles are closely related to their internal structure, a considerable effort has been made in the literature to gain insight into the underlying morphology of microgels. The earliest results related to the internal structure of pNIPAm microgels were reported by Wu et al.,31 who investigated the kinetics of particle formation. They found that during the particle formation the cross-linker monomer (BA) was incorporated into the microgel structure faster than the NIPAm monomer. This result implies that the different polymerization rates of the monomers lead to the formation of inhomogeneously cross-linked microgel particles. To shed light on the internal structure of pNIPAm microgels, several scattering studies have been performed. Varga et al. performed a combined static and dynamic light scattering investigation,32 which demonstrated that the pNIPAm microgel particles do not have a homogeneous internal structure. They Received: March 21, 2011 Revised: April 29, 2011 Published: May 18, 2011 7917
dx.doi.org/10.1021/la2010387 | Langmuir 2011, 27, 7917–7925
Langmuir also found that the structural inhomogeneity depends on the degree of cross-linking. The most highly cross-linked microgels retain a particle character and exhibit a Gaussian segment density distribution in their swollen state. However, with decreasing cross-link density the microgel particles can be better described as core/shell structures formed by a highly cross-linked core and a shell of dangling polymer chains. Small angle neutron scattering has also been used to get information about the microgel structure. First, Crowther et al. used SANS for the characterization of the microstructure of pNIPAm microgels as a function of temperature.33 They observed Porod scattering in the collapsed state of the microgel particles and a combination of Porod and OrnsteinZernike scattering below the phase transition temperature. These results indicate that microgels form compact particles with a sharp phase boundary in their collapsed state but a significant fraction of the polymer chains occupy a solution-like microenvironment below the volume phase transition (VPT). Kratz et al. investigated the effect of cross-link density34 and the effect of the chemical nature of the cross-linker monomer35 on the microstructure of microgels and they observed similar structural features. Later, Fernandez-Barbero et al.36 as well as Saunders et al.37 interpreted the SANS results in terms of the formation of a cross-linker rich core and a low segment density shell at the surface of the particles. These results were formally summarized in a coreshell form factor by Stieger et al.38 describing a uniform chain density in the microgel core and gradual decay in the segment density toward the surface of the microgel. Recently, Hoare et al. investigated the distribution of various carboxylic acid containing functional comonomers within the microgel particles via transmission electron microscopy by selectively staining the carboxyl groups.39 The experimentally determined radial density distributions varied significantly depending on the copolymerized functional comonomers. To interpret these results, they developed a polymerization model for the prediction of the microstructure of microgel particles.40 The model calculates the monomer consumption rates based on reactivity ratios determined experimentally between pairs of comonomers used in the microgel preparation. Assuming that the microgel particles grow from their core toward their outer shells, the monomer conversions (local microgel composition) can be directly linked to a radial position within the particle. The model calculations were in good agreement with the experimental results. Hoare et al. have also investigated how the functional group distribution (COOH) affects the swelling of microgel particles.41 They found that microgel particles that contained the same amount of cross-linker and functional monomer, but which had different radial functional monomer distribution exhibited more than a 5-fold difference in the magnitude of their pH induced swelling. These observations could be interpreted by using local swelling calculations based on the compositional profiles predicted by the copolymerization kinetics model.40 These results clearly highlight the fact that the comonomers polymerized in the pNIPAm microgel particles usually do not have a uniform distribution within the particles, which in turn have a profound effect on the microgel properties (e.g., mesh size distribution, swelling, stability, optical properties). Furthermore, the preparation of microgel particles with uniform internal structure is a major challenge due to the differences in the reactivity ratios of the applied comonomers. The aim of our study was to develop a method that allows the preparation of microgel particles with uniform cross-link
ARTICLE
distribution. To achieve this goal, we have investigated the kinetics of the cross-linker incorporation into the microgel particles. Then by controlling the composition of the reaction mixture, we were able to synthesize homogenously cross-linked pNIPAm particles. The particles have been characterized, and their properties are compared to the properties of microgel particles prepared by the conventional preparation method using the same average cross-link density.
’ EXPERIMENTAL SECTION Materials. All reagents including the N-Isopropylacrylamide monomer (NIPAm), the N,N0 -methylenebis(acrylamide) cross-linker (BIS) and the ammonium persulfate initiator (APS) for the microgel synthesis were purchased from Sigma-Aldrich and were used as received. Sodium dodecyl sulfate (SDS) was recrystallized twice from 1:1 benzene and ethanol mixture. Solutions were prepared in ultraclean Milli-Q water (total organic content = 4 ppb; resistivity = 18 mΩ 3 cm, filtered through a 0.2 μm membrane filter to remove particulate impurities). Reactor Design. Because the reaction rate of the microgel synthesis was found to be rather temperature sensitive, a double-wall Pirex glass reaction vessel was used for the microgel preparation. The reactor had a cylindrical shape. To ensure the constant temperature of the reaction mixture, the outer shell of the reactor was connected to a thermostat and controlled temperature water was circulated in it. This allowed us to keep the temperature of the reaction mixture constant within 0.1 °C. The reactor had a total volume of 400 mL, and it was equipped with five ground joints to ensure stirring with a glass stirring rod with a Tefflon paddle, the connection of a reflux condenser, nitrogen bubbling, and sample manipulation. Synthesis of pNIPAM Microgel Particles by Batch Synthesis. The batch synthesis of pNIPAm microgels was based on the method developed by Wu et al.31 Milli-Q water (280 mL) was transferred into the reaction vessel. The temperature of the reactor was set to 80 °C (unless it is stated otherwise), and the water was intensively stirred (∼1000 rpm). To remove oxygen, nitrogen gas was purged through the reactor for 60 min. Calculated amounts of NIPAm monomer and BA cross-linker were dissolved in 18 mL of Milli-Q water. To avoid the slow temperature initiated polymerization of the monomers in the reactor during the initial degassing and heating procedure, the monomer solution was degassed by vacuum at room temperature and injected into the reactor through a septum just before the initiation of the polymerization reaction. After the injection of the monomers, 1 mL of SDS solution was also injected into the reactor and then the reaction was initiated by the addition of 1 mL of APS solution. The solution was stirred intensively during the entire polymerization, while it was continuously purged with nitrogen. In a typical batch, pNIPAm synthesis the total monomer concentration was 130 mM, the SDS concentration was 0.65 mM, and the APS concentration was 1.2 mM in the reactor. The average cross-link density of the microgels (the ratio of the NIPAm and BA monomers) was 20 unless it is stated otherwise. Synthesis of Homogeneous pNIPAM Microgel Particles by Continuous Monomer Feeding. The preparation for the fed pNIPAm synthesis was done similarly to the batch synthesis; that is, 280 mL of Milli-Q water was transferred into the reaction vessel, heated to 80 °C, and purged with nitrogen for 60 min. Calculated amounts of monomers were dissolved in 18 mL of water, degassed by vacuum, and transferred into the reactor with 1 mL of SDS just before the initiation (1 mL APS). The total monomer concentration was 13 mM, the SDS concentration was 0.65 mM and the initiator concentration was 1.2 mM in the reactor at the beginning of the reaction. To prepare microgel particles with a homogeneous cross-link density of 20, the ratio of the NIPAm and BA monomers was set to 50 in the reactor (for details see Results and Discussion). One minute after the initiation of the polymerization the 7918
dx.doi.org/10.1021/la2010387 |Langmuir 2011, 27, 7917–7925
Langmuir
ARTICLE
feeding of NIPAm and BA monomers to the reaction mixture was turned on. To make monomer feeding possible, the monomers were dissolved in water. The total monomer concentration of the feeding solution was 1.364 mol/dm3, and the ratio of the NIPAm and BA monomers was 20 in the feeding solution. The solution was degassed, filled into a 60 mL syringe, and a total volume of 28.6 mL was fed into the reaction mixture with a feeding rate of 200 μL/min by means of a syringe pump (NewEra Pump Systems, NE-4000). After 143 min the feeding was stopped and the reaction was quenched by switching the temperature of the water circulated in the outer shell of the reactor from 80 to 15 °C. Methods. Monomer Conversion Measurements. Monomer conversion was measured by HPLC using an apparatus that consisted of a C18 column, a Gilson 305 piston pump and a Gilson 805 manometric module, a Model 7125 Rheodyne injector equipped with a 25 μL sample loop, and an LKB Bromma 2141 variable wavelength detector operating at a wavelength of 224 nm. The HPLC system was coupled with a computer equipped with the Data Apex Clarity software package. A mixture of 30% Milli-Q water and a 70% methanol was used as the mobile phase. To follow the monomer conversion, 3 mL samples were taken regularly from the reaction mixture. Sampling was done by sucking the reaction mixture directly into a syringe that was previously filled with 3 mL of 10 mM methyl hydroquinone to ensure the immediate quenching of the polymerization reaction in the sample. The unreacted monomers were separated from the polymeric reaction products by using Amicon Ultra-4 centrifugal filter devices containing regenerated cellulose membranes with 3 kDa molecular weight cut off. Centrifugation was done in a Hettich 220R centrifuge at 6000 rpm. Prior to their use the membranes were soaked in Milli-Q water to remove traces of glycerine contained by the membrane. It should be noted that to avoid the dilution of the separated monomer solution caused by the small amount of water contained in the dead volume of the filtration device the first few hundred microliter filtrate was discarded and only the following filtrate was used for the HPLC measurements. Monomer conversion was determined from at least three parallel measurements with standard deviation smaller than 1%. Dynamic Light Scattering Measurements. Dynamic light scattering (DLS) measurements were performed with a Brookhaven Instruments device, which consists of a BI-200SM goniometer and a BI-9000AT digital autocorrelator. An Argon-ion laser, Omnichrome, model 543AP, was used as light source. The laser was used at a wavelength of 488.0 nm and emitted vertically polarized light. The autocorrelator was used in a “multi τ” mode; i.e., the time axis was logarithmically spaced to span the required correlation time range. The autocorrelation functions were measured at an angle of 70° in 218 channels using a 100 μm pinhole size. The measured autocorrelation functions were analyzed by the CONTIN and the second-order cumulant methods. At finite concentrations and q-values, an apparent diffusion coefficient is obtained (Dapp); by using the EinsteinStokes equation, the apparent hydrodynamic radius of the complexes could be determined, Rh ¼
kT 6πηDapp
ð1Þ
where k is the Boltzmann constant, T is the absolute temperature, η is the viscosity of the medium, Rh is the hydrodynamic radius, and Dapp is the apparent diffusion coefficient. Before the DLS measurements, the latex samples taken during the reaction were purified from unreacted monomers and polymeric byproducts by centrifugation (Hettich 220R centrifuge, 16 000 rpm), decantation, and redispersion. The centrifuged microgels were redispersed in Milli-Q water and centrifugation was repeated three times. DLS measurements were performed on dilute solutions (50 ppm) that were temperature equilibrated for five minutes. At least ten consecutive measurements were done in each case, which did not show
Figure 1. (a) Total conversion of monomers as a function of reaction time. (b) Conversion of the NIPAm monomer (solid symbols) and the conversion of the BA cross-linker (open symbols) as a function of total monomer conversion. The synthesis were performed at 60 °C (black triangle) 70 °C (red square), and 80 °C (blue circle), respectively. (The average NIPAm to BA ratio is 20.) systematic changes in the size of the microgel particles, indicating that equilibrium swelling was achieved during the temperature equilibration.
’ RESULTS AND DISCUSSION The motivation of our work was to develop a method suitable for the preparation of homogeneously cross-linked pNIPAm microgel particles. To achieve this goal, first we carried out a detailed investigation of the kinetics of the microgel particle formation performed by the traditional preparation method described by Wu et al.31 As summarized in the Introduction, this method produces close to monodisperse microgels with inhomogeneous cross-linker distribution within the particles. In the second part of the discussion a new preparation method is proposed for the formation of homogeneous microgel particles. This method is based on the continuous feeding of the reacting monomers to maintain the constant reaction rate of each component. To distinguish the two preparation methods used in this study, the traditional and the presented new methods will be called batch method and feeding method, respectively. Batch Method. To facilitate the formation of homogenously cross-linked pNIPAm particles, first we addressed how the particle homogeneity is affected by the reaction conditions. In Figure 1a the total conversion of the monomers, Ctot = 1 {cNIPAm(t) þ cBA(t)}/{cNIPAm,t=0 þ cBA,t=0}, is plotted as a function of the reaction time (t) at different reaction temperatures (θ). The investigated temperatures range θ = 6080 °C, corresponds to typical experimental values used in the literature. As indicated by the figure, with increasing temperature the polymerization becomes faster. While at 60 °C the reaction is 7919
dx.doi.org/10.1021/la2010387 |Langmuir 2011, 27, 7917–7925
Langmuir
ARTICLE
homogeneity, we have calculated the ratio of the instantaneous consumption of the two monomers: Rparticle ðtÞ ¼
Figure 2. (a) Relative concentration of the unreacted monomers as a function of reaction time (total monomer concentration: 130 mM, θ = 80 °C, average NIPAm to BA ratio is 20). (b) Instantaneous ratio of the monomer concentrations in the reactor (Rreactor, blue squares) and Rparticle (see eq 2 and text; green circle) as a function of reaction time. The red dashed line indicates the average cross-link density of the microgel particles (the initial monomer ratio). The inset indicates how Rparticle varies with Rreactor during the course of the reaction.
completed in approximately 2 h, at 80 °C only about 30 min is needed to complete the polymerization. Furthermore, the figure also indicates that with increasing temperature the induction period of the polymerization diminishes on the experimental time scale. This is in agreement with the general expectation that with increasing temperature the decomposition rate of the persulfate initiator should considerably increase. In Figure 1b the conversion of the NIPAm monomer (CNIPAm = 1 cNIPAm(t)/cNIPAm,t=0) and the conversion of the BA crosslinker (CBA = 1 cBA(t)/cBA,t=0) are plotted as a function of the total conversion. In this representation the NIPAm conversion is a straight line because cNIPAm . cBA; thus the total conversion is dominated by the conversion of NIPAm monomer (Ctot ≈ CNIPAm). However, the BA conversion shows a positive curvature at each temperature, indicating that the BA cross-linker has a larger conversion than the NIPAm monomer throughout the reaction. This leads to the inhomogeneous cross-link density distribution within the particles as it has been highlighted in the literature.28,29,31 It should be noted that the deviation of the BA conversion curve from linearity (consequently, the inhomogeneity of the microgel particles) increases with decreasing reaction temperature. Therefore, in favor of particle homogeneity we performed all of our further microgel preparations at 80 °C. In Figure 2a the relative concentration (c(t)/c0) of the NIPAm and BA monomers is plotted as a function of the reaction time at 80 °C. As shown in the figure, the relative BA concentration decreases faster in the reactor; i.e., the consumption of the BA monomer in the reactor is faster than that of the NIPAm. This has been interpreted in terms of the higher reactivity of the BA monomer.40 To demonstrate how this affects particle
cNIPAm ðt þ ΔÞ cNIPAm ðtÞ cBAðt þ ΔÞ cBA ðtÞ
ð2Þ
This quantity describes the cross-link density of the forming hydrogel at any specific reaction time (t). If we adopt the hypothesis of the inside-out growth,28,29 which assumes that polymerization takes place in the outer shell of the particles because the growing particles are in a compact collapsed state at the reaction temperature, then Rparticle(t) describes the radial distribution of the cross-linker within the microgel particles, i.e., the change of the local NIPAm/BA ratio from shell to shell starting from the center (at the early stage of polymerization) to the periphery of the (completely grown) particle. Rparticle(t) should be constant if the particles are homogeneous. Rparticle(t) is calculated for the pNIPAm synthesis performed at 80 °C and it is plotted in Figure 2b. At the beginning of the polymerization Rparticle is much smaller (∼10) than the average monomer ratio in the reactor (20) but it rapidly increases with proceeding polymerization. The Rparticle function indicates that in the center of the gel particles the cross-link density is significantly higher than the average value and the most outer shell of the completely grown particle is practically not crosslinked. In Figure 2b the instantaneous ratio of the unreacted monomer concentrations (Rreactor = cNIPAm/cBA) is also plotted. As indicated in the figure, Rparticle and Rreactor seem to change in line with each other implying that the two quantities are not independent. To establish the relation between Rparticle and Rreactor, as a first approximation the reaction rate of the two monomers can be written as below, if we assume that the microgel formation can be considered as a homogeneous reaction of the growing particles and the monomers and the reactivity of the growing polymer chain is not affected significantly by the chemical nature of the last monomer incorporated into it (kPol-NIPAm*/BA ≈ kPol-BA*/BA and kPol-NIPAm*/NIPAm ≈ kPol-BA*/NIPAm, where kPol-NIPAm*/X denotes the rate constant of a NIPAm terminated polymer radical with monomer X and kPol-BA*/X denotes the rate constant of a BA terminated polymer radical with monomer X, respectively, and X is either a BA or a NIPAm monomer): dnNIPAm ðtÞ ¼ kNIPAm ½Pol½NIPAm dt
ð3Þ
dnBA ðtÞ ¼ kBA ½Pol½BA ð4Þ dt The square brackets indicate the equilibrium concentration of the components and Pol* denotes the polymeric radicals in the reaction mixture. Rparticle can be defined as the ratio of eqs 3 and 4: Rparticle ðtÞ ¼ ¼
dnNIPAm kNIPAm ½NIPAm ðtÞ ¼ ½BA dnBA kBA kNIPAm Rreactor ðtÞ kBA
ð5Þ
Thus, eq 5 indicates that there is simple linear relationship between the instantaneous ratio of the unreacted monomer concentrations in the reaction mixture (Rreactor) and the cross-link 7920
dx.doi.org/10.1021/la2010387 |Langmuir 2011, 27, 7917–7925
Langmuir density of the instantaneously forming hydrogel (Rparticle). In the inset of Figure 2b, Rparticle is plotted as a function of Rreactor. As indicated in the figure, the linear relationship indeed holds for a while during the reaction; however, at higher conversions a slight deviation can be observed from the linearity, indicating that at high conversion the consumption of the BA cross-linker became even more favored. This indicates that the assumptions used in the derivation of eq 5 are not strictly valid, though they can be a good approximation. Equation 5 has an important implication regarding the particle inhomogeneity. It suggests that the particle inhomogeneity is only an indirect consequence of the different monomer reactivities, the direct cause is the continuous change of the monomer ratio in the reactor. Thus, if Rreactor is kept at a constant value during the polymerization, then it can be expected that Rparticle will also be constant; i.e., homogeneous particles can be produced. Feeding Method. In principle a constant Rreactor value can be ensured by appropriate feeding of the monomers into the reactor during the polymerization reaction. Technically the simplest approach is the continuous replacement of the reacted monomers during the polymerization, thus keeping the absolute monomer concentrations constant in the reaction mixture. However, this means that the polymerization must be stopped abruptly at the end of the reaction (e.g., by the rapid cooling of the reaction mixture); otherwise, the synthesis would be completed by the formation of an inhomogeneous outer shell around the gel particles as in the case of the batch method. Because the monomer concentrations are kept constant throughout the reaction, to limit the amount of the unreacted monomers present at the end of the reaction, it is necessary to reduce the monomer concentrations significantly in the reactor compared to the initial concentration used in the case of the batch synthesis. For this reason we decided to reduce the initial monomer concentration in the reactor by an order of magnitude (from 130 to 13 mM). To prepare homogeneously cross-linked microgels, two important parameters have to be determined. One parameter is what value of the free monomer ratio in the reactor (Rreactor) gives rise to a desired cross-link density (Rparticle) value in the microgel particles. The other parameter is what the reaction rates of the monomers are at the applied reaction condition, so the monomers can be fed with an appropriate rate into the reaction mixture to ensure constant monomer concentrations during the reaction. To determine these parameters, we performed a series of batch pNIPAm synthesis. The total monomer concentration was decreased to 13 mM and the average cross-link density (the initial monomer ratio in the reactor, Rreactor,0) was varied from 15 to 50. The decreased monomer concentration had two important consequences. As expected (see eqs 3 and 4), the rate of the polymerization reaction (the monomer consumption) became slower. However, the reaction rate decreased for the NIPAm monomer in a larger extent; thus the synthesis performed at lower total monomer concentration gave rise to even more inhomogeneous microgel particles than in the case the original synthesis (data are not shown). This observation is in agreement with our previous result, what showed that at high monomers conversion (low monomer concentration) the consumption of the BA cross-linker became more favored (see the inset of Figure 2b). To establish the relationship between the ratio of the monomer concentrations in the reactor (Rreactor) and the cross-link density of the instantaneously forming hydrogel (Rparticle), both quantities were determined for each of the performed microgels
ARTICLE
Figure 3. Rparticle vs Rreactor relationship determined for a series microgel synthesis performed with different initial ratio of the NIPAm and BA monomers (total monomer concentration 13 mM, θ = 80 °C).
synthesis. The results are summarized in Figure 3. As indicated by the figure, the ratio of the incorporation rate of the NIPAm and BA monomers (Rparticle) is a linear function of the actual monomer ratio in the reactor (Rreactor = cNIPAm(t)/cBA(t)) independently of the initial monomer ratio used in the reaction and the extent of the monomer conversion. Thus, it can be concluded that when the polymerization is conducted at low monomer concentrations (100 min). This confirms that by the end of the reaction intact microgel particles form in the feeding polymerization. However, when the sampling time become shorter than ∼100 min, the microgel concentration suddenly
ARTICLE
sharply dropped in the purified samples and DLS measurements could not be performed. This implies that when the temperature was decreased below the LCST the primary polymer chains could dissolve from the particle in the lack of sufficient crosslinking and the resulting small polymer fragments remained in the supernatant during the centrifugation. To confirm this conclusion, the DLS measurements were repeated on reheated but nonpurified samples (open symbols in Figure 7). These measurements indicated smaller average particle size in the case of the reheated samples and diverging polydispersity. The Contin analysis of the measured autocorrelation functions indicated multimodal distribution in these samples. These observations confirm that the samples taken before 100 min reaction time dissolved upon cooling and formed new highly polydisperse, smaller particles when they were reheated. The above results indicate that in the case of the feeding polymerization the formation of the intact microgel particles requires orders of magnitude longer time than in the case of the batch reaction. Even more importantly, the formation of the intact microgels occurs only around 70% conversion in the case of the feeding reaction, while in the case of the batch reaction this happens at lower than ∼20% conversion. These differences are presumably related to the much higher cross-linker concentration in the initial core of the microgel particle in the case of the batch synthesis, which facilitates the interchain cross-linking of the aggregated primary polymer chains. Finally, we have also determined how the swelling of the microgel particles evolve during the particle growth. To characterize microgel swelling, the ratio of the volumes of the swollen particle (25 °C) and the collapsed particle (40 °C) were calculated using the particle sizes determined by DLS (Figures 6 and 7). The calculated swelling ratios are plotted in Figure 8 as a function of collapsed particle size for both the batch and the feeding polymerizations. For the batch synthesis the swelling increases with increasing particle size and then shows a maximum at large conversions, which is followed by a small drop as the particle size slightly increases further. These observations are in agreement with the results of Wu et al.31 The initial increase in swelling indicates the decreasing cross-link density of the outer shells of the particle, while the final drop in swelling was attributed to the additional cross-linking at the end of the reaction between the previously unreacted double bonds within the particle. If the swelling of the microgel particles prepared by the feeding method is compared to the swelling of the particles prepared by the batch method, two main differences can be identified. First, the swelling of these particles is much larger despite the equal average cross-link density used in the two microgel preparations; second, the swelling is independent of the particle size (conversion) after the formation of the intact microgel particles. Both differences indicate that the particles prepared by the feeding method have a different internal structure. The swelling results (large, size independent swelling) are in good agreement with the uniformly cross-linked, homogeneous internal particle structure that is also implied by the kinetic measurements and the optical properties of the formed particles.
’ CONCLUSIONS The kinetics of pNIPAm microgel formation has been investigated and a novel synthetic method is proposed for the preparation of homogeneously cross-linked microgel particles. 7924
dx.doi.org/10.1021/la2010387 |Langmuir 2011, 27, 7917–7925
Langmuir The proposed method is based on keeping the NIPAm monomer and the BA cross-linker monomer concentrations constant within the reaction mixture, thus facilitating the constant rate of monomer incorporation for both components. The constant monomer concentrations are ensured by the continuous feeding of the monomers to the reactor. The optical properties and the swelling of the prepared homogenously cross-linked microgel particles significantly differ from the characteristics of the inhomogenously cross-linked microgels prepared with identical average cross-link densities using the normal batch precipitation polymerization. Our results imply that by controlling the composition of the reaction mixture the internal structure, the properties of the forming microgel particles can thus be controlled. Further work in this direction is currently in progress.
’ AUTHOR INFORMATION Corresponding Author
*E-mail:
[email protected].
’ ACKNOWLEDGMENT This work is supported within the 7th European Community RTD Framework Program by the Marie Curie European Reintegration Grant, PE-NANOCOMPLEXES (PERG02-GA2007-2249) as well as by the Hungarian Scientific Research Fund, OTKA K68027 and K68434. I.V. is a Bolyai Janos fellow of the Hungarian Academy of Sciences, which is gratefully acknowledged. ’ ADDITIONAL NOTE a Calculations were done by using the assumptions that the volume of the particles is proportional to the amount of the NIPAM monomers built in the particles (V = πdh3/6 µ 1 cNIPAm/co,NIPAm) and that convertion is proportional to the reaction time (1 cNIPAm/co,NIPAm µ t). This leads to the equation 1 cNIPAm/co,NIPAm) = k2 3 t3, and the unknown k2 constant was calculated from the dh value measured by DLS at t = 140 min (where c = co). ’ REFERENCES (1) Staudinger, H.; Husemann, E. Ber. Dtsch. Chem. Ges. 1935, 68, 1618–1634. (2) Baker, W. O. Ind. Eng. Chem. 1949, 41, 511–520. (3) Pelton, R. H.; Chibante, P. Colloids Surf. 1986, 20, 247. (4) Heskins, J. E.; Guillet J. Macromol. Sci. Chem. A2 1968, 8, 1441–1455. (5) Shild, H. G.; Tirrell, D. A. J. Phys. Chem. 1990, 94, 4352–4356. (6) Pelton, R. Adv. Colloid Interf. Sci. 2000, 85, 1–33. (7) Hsiue, G. H.; Hsu, S. H.; Yang, C. C.; Lee, S. H.; Yang, I. K. Biomaterials 2002, 23, 457–462. (8) Choi, S. H.; Yoon, J. J.; Park, T. G. J. Colloid Interface Sci. 2002, 251, 57–63. (9) Murthy, N; Thng, Y. X.; Schuck, S.; Xu, M. C.; Frechet, J. M. J. J. Am. Chem. Soc. 2002, 124, 12398. (10) Lopez, V. C.; Snowden, M. J. Drug Deliv. Syst. Sci. 2003, 3, 19–23. (11) Nayak, S.; Lee, H.; Chmielewski, J.; Lyon, L. A. J. Am. Chem. Soc. 2004, 126, 10258. (12) Nolan, C. M.; Reyes, C. D.; Debord, J. D.; Garcia, A. J.; Lyon, L. A. Biomacromolecules 2005, 6, 2032. (13) Varma, M. V. S.; Kaushal, A. M.; Garg, S. J. Controlled Release 2005, 103, 499.
ARTICLE
(14) Lopez, V.; Castro, H. J.; Snowden, M. J. Int. J. Pharmaceut. 2005, 292, 137. (15) Nolan, C. M.; Gelbaum, L. T.; Lyon, A. L. Biomacromolecules 2006, 7, 2918–2922. (16) Hoare, T.; Kohane, D. S. Polymer 2008, 49, 1993–2007. (17) Morris, G. E.; Vincent, B.; Snowden, M. J. Prog. Colloid Polym. Sci. 1997, 105, 16. (18) Retama, J. R.; Lopez-Ruiz, B.; Lopez-Cabarcos, E. Biomaterials 2003, 24, 2965. (19) Guo, Z.; Sautereau, H.; Kranbuehl, D. E. Macromolecules 2005, 38, 7992. (20) Varga, I.; Szalai, I.; Meszaros, R.; Gilanyi, T. J. Phys. Chem. B 2006, 110, 20297–20301. (21) Bromberg, L.; Temchenko, M.; Hatton, T. A. Langmuir 2003, 19, 8675. (22) Nerapusri, V.; Keddie, J. L.; Vincent, B.; Bushnak, I. A. Langmuir 2007, 23, 9572–9577. (23) Schmidt, S.; Hellweg, T.; von Klitzing, R. Langmuir 2008, 24, 12595–12602. (24) Serpe, M. J.; Jones, C. D.; Lyon, L. A. Langmuir 2003, 19, 8759–8764. (25) Acciaro, R.; Aulin, C.; Wagberg, L.; Lindstr€om, T.; Claesson, P. M.; Varga, I. Soft Matter 2011, 7, 1369–1377. (26) Saunders, B. R.; Vincent, B. Adv. Colloid Interface Sci. 1999, 80, 1. (27) Nayak, S.; Lyon, L. A. Angew. Chem. 2005, 44, 7686–7708. (28) Hoare, T.; Pelton, R. Curr. Opin. Colloid Interface Sci. 2008, 13, 413–428. (29) Saunders, B. R.; Daly, N. L., E.; Hu, S. T., X.; Stepto, R. Adv. Colloid Interface Sci. 2009, 147148, 251–262. (30) McPhee, W.; Tam, K. C.; Pelton, R. J. Colloid Interface Sci. 1993, 156, 24–30. (31) Wu, X.; Pelton, R. H.; Hamielec, A. E.; Woods, D. R.; McPhee, W. Colloid Polym. Sci. 1994, 272, 467. (32) Varga, I.; Gilanyi, T.; Meszaros, R.; Filipcsei, G.; Zrinyi, M. J. Phys. Chem. B 2001, 105, 9071–9076. (33) Crowther, H. M.; Saunders, B. R.; Mears, S. J.; Cosgrove, T.; Vincent, B.; King, S. M.; Yu, G. E. Colloids Surf., A Physicochem. Eng Asp. 1999, 152, 327–33. (34) Kratz, K.; Hellweg, T.; Eimer, W. Polymer 2001, 42, 6631–9. (35) Kratz, K.; Lapp, A.; Eimer, W.; Hellweg, T. Colloids Surf., A Physicochem. Eng. Asp. 2002, 197, 55–67. (36) Fernandez-Barbero, A.; Fernandez-Nieves, A.; Grillo, I.; LopezCabarcos, E. Phys. Rev. E 2002, 66, 051803. (37) Saunders, B. R. Langmuir 2004, 20, 3925. (38) Stieger, M.; Richtering, W.; Pedersen, J. S.; Lindner, P. J. Chem. Phys. 2004, 120, 6197–6206. (39) Hoare, T.; McLean, D. J. Phys. Chem. B 2006, 110, 20327–20336. (40) Hoare, T.; McLean, D. Macromol. Theory Simul. 2006, 15, 619–632. (41) Hoare, T.; Pelton, R. J. Phys. Chem. B 2007, 111, 11895–11906. (42) Inomata, H.; Wada, N.; Yagi, Y.; Goto, S.; Saito, S. Polymer 1995, 36, 875. (43) Senff, H.; Richtering, W. Colloid Polym. Sci. 2000, 278, 821. (44) Debord, J. D.; Lyon, L. A. J. Phys. Chem. B 2000, 104, 6327–6331.
7925
dx.doi.org/10.1021/la2010387 |Langmuir 2011, 27, 7917–7925
