Polymer-Grafted Magnetic Nanoparticles in Polymer Melts PDF

Article pubs.acs.org/Macromolecules...

Article pubs.acs.org/Macromolecules Assembly of Polymer-Grafted Magnetic Nanoparticles in Polymer Melts Yang Jiao and Pinar Akcora* Department of Chemical Engineering & Materials Science, Stevens Institute of Technology, Hoboken, New Jersey 07030, United States * S Supporting Information ABSTRACT: Hydrophobic iron oxide nanoparticles grafted with hydrophobic See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles. polymer chains of varying molecular weights and graft densities are synthesized to underpin the role of brush entanglement and dipolar forces on creating nanostructures. Grafting density on magnetic nanoparticles is controlled in grafting- to method by changing the concentration of functionalized polymer in solution. The Downloaded via BROCK UNIV on October 21, 2024 at 14:25:45 (UTC). grafting density and brush length have varied systemically to observe the changes in nanostructures. Bridging between grafted chains and dipolar forces become effective only at low grafting density and result in long chains of particles. We demonstrate experimentally that structural transition of magnetic nanoparticles is controlled with the balance between grafted chain entanglements and dipolar forces. INTRODUCTION Controlling the state of nanoparticle dispersion in polymer the organization of polymer-grafted nanoparticles. While the effect of entanglement between grafted and free chains on matrices through functionalizing the particles with polymers mechanical properties has been reported in previous experi- and ligands is essential to enhance functional properties of ments,15−17 the direct influence of brush−brush entanglement polymer nanocomposites.1−5 Polymer-decorated magnetic on the formation and presumably on the transitions between nanoparticles can play an essential role in forming equilibrium different nanostructures has not been shown experimentally. structures that will lead to significant advances in their Particle size, length of polymer grafts and low grafting density macroscopic mechanical and electroactive properties when can play an important role to investigate the brush−brush and they are mixed with polymers. Moreover, magnetic polymer brush-matrix interactions. nanocomposites find potential applications in microelectronics, Assembly of magnetic nanoparticles in polymer solution and biotechnology, sensing, and catalysis.6−8 films has been a challenge as they tend to aggregate without In the context of stabilizing nanoparticles through polymer surface functionalization.7 It is known that ferromagnetic decoration, the length of the free and grafted chains, the nanoparticles can assemble into chains due to their directional grafting density and the size of the particle core are critical dipolar interactions in solution18−21 or between oil−water parameters that govern the dispersion state and morphology interphases.22−24 For superparamagnetic nanoparticles, chains diagrams of functional particles. It is known that nanoparticles can be formed by applying magnetic fields,25,26 however chains functionalized uniformly with long dense polymer grafts cannot keep their structures upon removal of fields. In the case present good dispersion in melt.3,9,10 On the contrary, of magnetic particles adsorbed with polymer, irreversible chain- polymer-grafted nanoparticles with low grafting density have like structures are obtained after magnetic fields are removed been found to disperse well when tethered polymers are shorter due to the strong entanglement (bridging) between polymer than the polymer matrix due to repulsive interactions between chains.27 To improve the dispersion of magnetic nanoparticles tethered nanoparticles.11,12 Furthermore, a recent simulation in polymers, grafting-from approach was utilized in previous work presented the structural transitions of solvent-free works which yielded high grafting density.28−30 Iron oxide oligomer grafted nanoparticles with variations of grafting nanoparticles tethered with high grafting density of polymer density.13 Recent experimental work showed that anisotropic resulted in well-dispersion of particles and with short brushes organization of spherical nanoparticles could only be achieved aggregation of particles was observed in homopolymers and by entropic−enthalpic balance between particle and polymer block copolymers,31 typical results expected for polymer parts.14 Thus, particles of amphiphilic character can organize decorated particles. into various morphologies where polymer graft length, density and entanglement between brushes and matrix determine the Received: January 8, 2012 boundaries between different structures. However, it has been Revised: March 30, 2012 difficult to show the effect of entanglement between grafts on Published: April 10, 2012 © 2012 American Chemical Society 3463 dx.doi.org/10.1021/ma3000576 | Macromolecules 2012, 45, 3463−3470 Macromolecules Article Figure 1. TEM image of bare iron oxide nanoparticles and particle size histogram of corresponding TEM. Small-angle X-ray scattering pattern of bare particles in 156 kDa PS matrix. In this work, we tethered poly(styrene) of various molecular be different and this difference will increase as the solvent quality weights at low and intermediate grafting densities on iron oxide decreases, leading to star polymers (tethered polymers) collapsing nanoparticles using grafting-to approach through a ligand first. To accomplish this, ethanol was added to the solution of tethered exchange reaction. Different grafting densities with grafting-to and free PS chains in toluene drop by drop until the homogeneous solution turns to cloudy. The clear upper liquid, containing the free method were obtained by varying the concentration of free polymer, was removed from the particles by centrifuging the solution chains in dilute solution which is an easy method to obtain at 11000 rpm for 12 min. The dark brown precipitate was collected different grafting densities with low polydispersity polymers. It and redissolved in toluene. After multiple washes, the upper solution is important to note that magnetic nanoparticles and grafted was decanted in methanol to confirm that there was no free polymer polymers are both hydrophobic in this work. Consequently, left in solution. The brown sediment was dispersed and stored in with the right choice of magnetic particle size and polymer toluene for future use. Sample characteristics used in this work are lengths, steric repulsion and magnetic interactions become listed in Table 1. effective in the case of low grafting density. Herein, we report the structures of PS grafted iron oxide nanoparticles dispersed Table 1. Characteristics of Synthesized PS Grafted Fe3O4 in polymer matrices and discuss the effect of brush−brush σ entanglement and dipolar interactions on assembly of nano- particles. M̅ w (kg/mol) PDI (chains/nm ) 2 (chains/particle) 15.6 1.06 0.052 7.6 EXPERIMENTAL SECTION 43.2 1.05 0.017 3.4 0.044 6.4 Materials. 2,2′-Azobis(isobutyronitrile) (AIBN; 98%) was recrys- 0.066 20.7 tallized from methanol. Styrene (99%) was distilled under reduced 68.25 1.15 0.023 4.6 pressure over CaH2 before use. Benzyl ether (98%), iron(III) 0.04 12.4 acetylacetonate (97%), oleic acid (90%), oleylamine (70%), 1,2- hexadecanediol (90%), and other chemicals were used as received. All 124.1 1.28 0.013 1.9 chemicals were purchased from Sigma-Aldrich. 0.052a 13.8a Synthesis of Functionalized PS. The oleic acid and oleylamine- a Sample is synthesized using grafting-from approach. stabilized Fe3O4 nanoparticle was synthesized by thermal decom- position method. 32 A RAFT agent, 2-[(dodecylsulfanyl)- carbonothioyl]sulfanyl propanoic acid (DCSPA), was synthesized Surface-Initiated RAFT Polymerization of PS from Fe3O4 following the reported procedure.33 Reversible addition−fragmenta- Nanoparticles Surface. DCSPA was linked to the oleic acid- and tion chain transfer polymerization of poly(styrene) with carboxyl oleyl amine stabilized Fe3O4 nanoparticles via the ligand exchange group containing RAFT agent was followed according to the reported reaction. First, the freshly prepared Fe3O4 nanoparticles (200 mg) procedure.34 Chain transfer agent (CTA) (27.5 mg, 78.4 μmol), AIBN were dispersed in 1,2-dichlorobenzene (20 mL), followed by addition (2.5 mg, 15.7 μmol), and styrene (10 mL, 86.4 mmol) were added to a of DCSPA (1 g). The mixture was stirred at 80 °C for 24 h. Then, the 20 mL Schlenk flask, and degassed through three cycles of freeze− reaction mixture was added into a large excess of ethanol, and the pump−thaw. The flask was then placed in an oil bath preset to 85 °C precipitate was separated via centrifuge (7000 rpm, 10 min). The for 24 h. Polystyrene with different molecular weight was synthesized precipitated particles were purified by three times of redispersing in by changing the ratio of styrene to CTA. The polymerization was toluene and reprecipitating in ethanol to remove all traces of free stopped by quenching the flask in ice−water. The reaction mixture was DCSPA. The DCSPA-anchored Fe3O4 nanoparticles were dried under dissolved in toluene and the polymer was purified by precipitating in vacuum at room temperature for overnight prior to polymerization. methanol. Product was dried in a vacuum oven for overnight. A stock solution in 1,2-dichlorobenzene (5 mL) comprising Immobilization of PS onto Fe3O4 Nanoparticles. Carboxyl DCSPA-anchored Fe3O4 nanoparticles (107 mg), styrene (8.2 g, group ended PS was associated with oleic acid and oleylamine- 78.7 mmol), free DCSPA (7.54 mg, 21.5 × 10−3 mmol), and AIBN stabilized Fe3O4 nanoparticles through a ligand-exchange mechanism (1.13 mg, 6.89 × 10−3 mmol) in a Schlenk flask was degassed through (as shown in Figure 1). The freshly prepared NPs (50 mg) dispersed three freeze−pump−thaw cycles. The flask was then placed in an oil in toluene (50 mL) were stirred with the addition of PS at 60 °C for bath preset to 85 °C for 48 h. The polymerization was stopped by 48 h. Removal of the free PS chains after ligand exchange is critical for quenching the flask in ice water. The purification of PS-tethered determining grafting densities. PS-tethered nanoparticles resemble star nanoparticles was identical to the procedure described above. Further, polymers where their solubility decreases with the increase in number polymer was cleaved from nanoparticles by adding a HCl solution (2 of arms. Therefore, the solubility of free chains and tethered ones will mL at 37 wt %) into PS grafted iron oxide nanoparticles (5 mL at 10 3464 dx.doi.org/10.1021/ma3000576 | Macromolecules 2012, 45, 3463−3470 Macromolecules Article Figure 2. Synthesis schema of CTA-anchored PS and PS-grafted Fe3O4 NPs by grafting-to method. mg/mL). After the organic phase was clear, an excess of methanol was they are dispersed in hydrophobic PS polymer matrix and added to the solution to precipitate PS chains. The recovered polymer behave as surface fractals. The peak at the low q region was collected by centrifugation and dried under vacuum for indicates interference within particle aggregates which was subsequent GPC analysis. Polymer-grafted particles in toluene were mixed with the matrix observed as separated clusters in TEM image (see Supporting polystyrene homopolymers and then sonicated for 20 s to prepare the Information, Figure S1). composite solutions. The polymer-grafted particle concentration was 5 Synthesis and Stability of Poly(styrene) Tethered Iron mass % in all of the samples. The composites, in solution, were cast Oxide Nanoparticles in Solution. The key to critical onto Teflon dishes, dried to remove the solvent in air, and then understanding of various interactions between components of annealed for 48 h and additional 7 more days in a vacuum oven at 150 polymer brush nanocomposites is the preparation of low-PDI °C. Sample thicknesses were approximately 300 μm; therefore, particle tethered polymers onto iron oxide nanoparticles with dispersions presented for composites are for the bulk samples. controlled brush chain length and the variation of grafting Characterization. Molecular weights were determined by gel permeation chromatography-light scattering (GPC/LS) using a system densities ranging from low to intermediate values. We employ equipped with a VARIAN PLgel 5.0 μm Mixed−C gel column (7.5 grafting-to method (Figure 2) which is accomplished by a mm ID), a light detector (miniDawn, Wyatt Technology) and a ligand exchange reaction of carboxylic acid functionalized refractive index detector (Optilab rEX, Wyatt). Pure CTA and poly(styrene) (PS) of molecular weights ranging between 15 carboxyl group ended PS were confirmed by H1 NMR, performed on a kDa-124 kDa. Different grafting densities (0.01−0.06 chains/ Bruker AV 500 spectrometer. Dynamic light scattering (DLS) nm2) were achieved by controlling the molar concentration of measurements were carried out with a Zetasizer NanoS, Malvern free PS chains in solution. This method has been used in Instruments. Solutions having particle concentration of 0.3−0.5 mg/ mL in toluene were bath sonicated for about 2 min. Measurement previous works on silicon surfaces to achieve high grafting duration was set to be 15 s, and data were averaged over 15 runs. density.36,37 Nanocomposite films were microtomed into 50−80 nm slices with a Iron oxide nanoparticles with oleic acid and oleylamine diamond knife at room temperature and examined in transmission surfactants are unable to sustain a secondary reaction to bring electron microscopy (FEI CM20 FE S/TEM) operated at 200 keV. new functionalities.38 Ligand exchange reaction has been widely Many cross-sectional TEM images were captured over different areas used as the first step to modify nanoparticle surfaces. This and the images presented are representative of observed nanostruc- surface modification involves adding an excess amount of new tures. ligand containing carboxyls,39,40 silanes and amines.38 The TGA measurements were performed on a Q50 TGA (TA Instruments) for grafting density calculations. Measurements were structure of carboxylic acid (or chain transfer agent, CTA) performed under a constant flow of nitrogen of 20 mL/min at a functionalized PS was confirmed by the appearance of the peak heating rate of 20 °C/min, starting from room temperature up to 580 for methyl in hydrocarbon chain of CTA in 1H NMR spectrum °C, and then held constant at maximum temperature for 30 min. All of PS at 0.90−0.86 ppm (not shown here). The attachment of samples were dried in a vacuum oven at 60−80 °C prior to each TGA PS to the NP surface was probed by FTIR analysis (Figure 3). measurement to remove remaining moisture. All reported TGA curves were normalized with respect to the weight at 100 °C to make sure that only the solid fraction was measured. SAXS measurements were performed at Beamline X27C at the National Synchrotron Light Source (NSLS), Brookhaven National Laboratory. RESULTS AND DISCUSSION Bare Iron Oxide Nanoparticles. TEM image of bare particles prepared from a solution drop on a copper grid and the corresponding histogram are presented in Figure 1. The average particle core diameter is measured to be 7.8 ± 1.8 nm. The particle size was also determined by small-angle X-ray scattering (SAXS) from bare particles dispersed in 156 kDa PS. SAXS pattern in Figure 1 was fit to the unified equation35 where the primary particle size was obtained as 6.8 nm at intermediate-q (see Supporting Information for SAXS analysis). Power law exponent for the larger structure at low q region was found to be 3.78, indicating particles are dispersed like surface fractals. The hydrophobic surfactants on nanoparticles prevent Figure 3. FTIR spectra of (a) oleic acid and oleylamine-stabilized the aggregation of particles into large spherical clusters when Fe3O4 nanoparticles, (b) PS tethered Fe3O4 nanoparticles, and (c) PS. 3465 dx.doi.org/10.1021/ma3000576 | Macromolecules 2012, 45, 3463−3470 Macromolecules Article The appearance of aromatic absorption bands at 3081−3025, equilibrium at semidilute concentration leading to fast 1600, 1492, 756, and 698 cm−1 (Figure 3b) demonstrates the adsorption−desorption of chains, and thus dependence of presence of PS on particle surface. The successful dispersion of grafting density on polymer concentration varies from what has the PS-grafted iron oxide nanoparticles in toluene but their been proposed by Auroy et al.43 We have achieved the grafting poor dispersion in hexane is also indicative of successful surface density of 0.1 chains/nm2 (30 chains/particle) for 15 kDa PS modification. It is worth to note that compared with the sample at high polymer concentration as high osmotic pressure spectrum for PS (Figure 3c), the peak ratio of aromatic was required to overcome the steric hindrance to form a dense stretches of C−H (3025−3081 cm−1) to CH2 (2923, 2850 grafted layer. For longer chains, this steric force cannot be cm−1) decreased after ligand exchange reaction indicating oleic compensated by increasing polymer concentration, thus acid and oleylamine remain after tethering polymers. We note resulted in grafting densities of ∼5−10 chains/particle. that particles are presumably uniformly covered with the oleic Additionally, we observed that polymer immobilized nano- acid and oleylamine prior to and after attaching polymers. particles are stable in concentrated solution (10 mg/mL), but Surface modifications by grafting-to method involves precipitated easily in dilute concentration (0.6 mg/mL), but coupling of functional groups on the surface and on the end adding a small amount (∼0.2 mg/mL) of carboxylic group of PS chains, which can be achieved by controlling the number functionalized PS enables the stability of brushes in solution, of functional groups on surfaces or the number of chains.41 suggesting the existence of the equilibrium between adsorp- Because it is difficult to determine the amount of functional tion−desorption of ligands. With the presence of excess groups on surfaces, changing the number of polymer chains amount of ligands, adsorption of chains are favorable and with functional ends is an easy approach in order to achieve desorption process becomes negligible. Therefore, the ligand controlled grafting densities. Here, we reported that the exchange process and the grafting density of chains on concentration of free chains can control the extent of ligand- nanoparticles can be controlled with the concentration of free exchange reaction of Fe3O4 nanoparticles and hence grafting chains. On another note, when CTA was attached to density of different molecular weights of chains. Figure 4 shows nanoparticle and grafting-from polymerization was proceeded, high grafting density with 124 kDa PS could not be achieved which might be due to the desorption of CTA molecules. Adding silane layer on iron oxide nanoparticles is another option to obtain more stable ligands,38 however silanized iron oxide nanoparticles will have solubility problems because of the short length of silane ligands and thus may cause nonuniformity of grafting density. Next, we have characterized the average size of polymer- tethered nanoparticles in DLS to confirm their stability in solution. Figure 5 shows the normalized hydrodynamic size, Figure 4. Effect of free polymer concentration in solution on polymer grafting density. that grafting density increases with the concentration of free PS chains, varying from dilute to semidilute solution. The existence of a plateau at low grafting density (∼5 chains/ particle) of 124 kg/mol PS suggests a limited accessibility of chains to the particle surface due to steric hindrance of long chains. It is seen that the grafting density decreases when the chain length increases (Figure 4). The dependence of grafting density on the polymer concentration has been suggested by Figure 5. Variation of normalized hydrodynamic size of PS-tethered deGennes42 and has been confirmed experimentally by Auroy iron oxide nanoparticles with grafting density. et al.43 for grafting 145 kg/mol PDMS onto 400 nm porous silica particles that the amount of grafted polymer per unit area which is the ratio of measured number-averaged size of clusters is equal to C7/8N1/2 where C is concentration and N is degree of (2Rh) to calculated brush size with its core, (particle core plus polymerization. In their case all chains were irreversibly grafted. 4Rg) for the tethered polymers at various grafting densities. It is The lines in Figure 4 are the power-law fits and the seen that all of the brushes except 124 kDa PS agglomerated concentration exponents (n) are indicated on each line. Our when the number of grafted chains were lower than 10 chains data deviates from this relationship for 15 kDa PS where the per particle. Well-dispersion was achieved at all of the grafting exponent of concentration is 0.69, and it decreases to 0.53 for densities in 124 kDa brush. 43.2 kDa, to 0.41 for 66.5 kDa, and to 0.40 for 124 kDa. We Structural Characterization: Dispersion of Polymer explain this behavior by suggesting that chains are attached Tethered Iron Oxide Nanoparticles in Polymer Melts. We reversibly to the surface; therefore, the grafted chains are not at investigate the effect of brush length and grafting density on the 3466 dx.doi.org/10.1021/ma3000576 | Macromolecules 2012, 45, 3463−3470 Macromolecules Article Figure 6. Dispersion of PS-tethered iron oxide nanoparticles (pure magnetic brushes) as cast from solution in toluene: (A) 124 kDa (0.013 chains/ nm2); (B) 124 kDa (0.052 chains/nm2); (C) 15 kDa (0.052 chains/nm2). Figure 7. Effect of brush chain length on dispersion of PS grafted magnetic nanoparticles at sparse grafting density in 15 kDa PS matrix: (A) 124 kDa (0.013 chains/nm2); (B) 68.25 kDa (0.023 chains/nm2); (C) 43.2 kDa (0.017 chains/nm2). dispersion of polymer tethered magnetic oxide nanoparticles To understand the strength of interactions, we have primarily to underpin the role of brush−brush entanglements calculated the van der Waals and dipole−dipole interactions on particle aggregation. It is important to note that PS-grafted between two superparamagnetic particles of 8 nm in diameter iron oxide nanoparticles do not have the amphiphilic character that are in contact with the separation distance of 2 nm.31 The as of PS-grafted silica system14 and their controlled aggregation van der Waals interaction energy is calculated to be around 1 will be different as magnetic particles provide different kT and dipole−dipole interaction is calculated as 0.8 kT (see interplays between dipolar, repulsion and van der Waals Supporting Infomation for equations). Thus, dipole interaction interactions. plays less significant role in the aggregation of particles. In our The spherical iron oxide nanoparticles with average diameter work, grafted polymers play the physical bridging role between of 7.8 ± 1.8 nm are superparamagnetic at room temperature particles through the entanglement of chains, thus dipolar which means that particles cannot organize into 1-D chains interactions become effective as particles are connected to form purely by dipole magnetic interactions. It has been shown that chains. We surmise that dipolar interactions, even though magnetic coupling occurs in chain-like structures, causing an structures are not mainly induced by them, help the formation increase in coercivity.44,45 For example, superparamagnetic of the equilibrium structures of polymer grafted iron oxide nanoparticles that are covalently linked through their pole nanoparticles. We have also examined the composite structures points are shown to form chains with induced magnetization after annealing 9 days at 150 °C and have observed similar moments.46 In this work, instead of using a molecular linker, we structures that are seen for 2 days annealed samples as shown in have decorated particles with polymers at low grafting density Figures 7−9, indicating that the structures are in equilibrium. where the conformation of brush chains enable the brushes and We have synthesized polymers of varying molecular weights lead to string formation. In addition to this entanglement (15−124 kDa) and tethered them mainly at two different factor, magnetic coupling among particles in a chain keeps their grafting densities: ∼0.01 and ∼0.05 chains/nm2 (shown in elongated string structures. Table 1). We start our discussion with the structures of pure 3467 dx.doi.org/10.1021/ma3000576 | Macromolecules 2012, 45, 3463−3470 Macromolecules Article polymer tethered iron oxide nanoparticles with 124 kDa PS with 2 chains and 14 chains per particle (∼0.013 and ∼0.052 chains/nm2) which are good representations of particles dispersed in solution as they are prepared by evaporating a drop of solution on a grid. As discussed in previous section, magnetic nanoparticles feel dipolar interactions when the steric repulsion is weaker in the case of 2 tethered chains and form long branched chains (Figure 6a). On the other hand, particles tethered with 15 kDa PS form spherical aggregates as the result of van der Waals interactions dominating the dipolar forces (Figure 6c). Figure 7 presents the effect of increasing brush length on the aggregation of particles that are grafted at low density. When particles come into contact with attractive interactions, entanglement between the grafted chains of 124 kDa molecular weight at a low grafting density (0.013 chains/nm2) plays the bridging role and elongated chains are protected with the brush entanglement (Figure 7a). Therefore, the formation of string is not controlled purely with dipolar interactions but also with the entanglement of long brushes (124 kDa) which allows the particles aggregate from the two ends of a string. A shorter brush (43.2 kDa), having the shorter range of repulsion of grafted chains, at the same grafting density has a thinner effective entangled brush layer and dipolar forces become insufficient to keep the chain-like structures, hence they form branched chains and clusters (Figure 7c). We conclude that long strings are stabilized when the effective entanglement is large which is achieved only at low grafting density and also due Figure 8. TEM images show transition of nanostructures from large to the magnetic coupling of uniaxial particles in 1-D chains. We elongated branched chains to spheres and to short chains with the discussed the structures in Figure 7 by simply considering that increase of grafting density from bottom to top. The first column is for they are pure brushes in a good solvent as the matrix molecular 43 kDa PS-grafted nanoparticles, the grafting density is 0.017 weight (15 kDa) was below the entanglement molecular weight (bottom), 0.044 (middle) and 0.066 chains/nm2 (top). The second of PS. Note that TEM images of pure PS-tethered particles column is for 124 kDa brush with 0.052 chains/nm2 (top) and 0.013 (Figure 6a) show similar structure when they are mixed with 15 chains/nm2 (bottom). Ratio of brush and matrix molecular weights is kDa PS. We discuss the effect of matrix molecular weight in the 0.35 for 43 kDa and 1 for 124 kDa. following section. Figure 9. Dispersion of PS (124 kDa)-grafted magnetic nanoparticles in matrices of increasing molecular weight for the two grafting cases: 0.052 and 0.013 chains/nm2. 3468 dx.doi.org/10.1021/ma3000576 | Macromolecules 2012, 45, 3463−3470 Macromolecules Article Figure 10. SAXS data of PS (124 kDa)-grafted magnetic nanoparticles in different PS matrix molecular weights: 15, 40, 124, and 156 kDa of (A) 0.013 chains/nm2 and (B) 0.052 chains/nm2. Data shifted vertically for clarity. Next, we look into the effect of grafting density for the two hydrophobic magnetic nanoparticles is shown in the structural brushes of 124 kDa and 43 kDa molecular weights. Sterically transition of branched structures into spherical aggregates and stabilized long strings observed at 0.013 chains/nm2 for 124 then into isolated particles with the increase of graft density. kDa brush break into small aggregates of tetramers and Moreover, in the context of amphiphilic nanoparticles, pentamers with the increase of grafting density to 0.05 chains/ changing the particle size can play an important role in phase nm2 (Figure 8, second column). It is expected that with denser separation of polymer tethered particles at different grafting brushes, brush−brush entanglement will not be effective for density. However, the size effect is irrelevant in the physical particles to conserve long strings and steric repulsion separates picture we present here as our particles and chains are both particles into short chains and isolated particles. We observed that the increase in grafting density has the most influence in 43 hydrophobic. kDa brush where the branched chains seen at the low grafting To resolve the effect of matrix chains, we varied the ratio of density (0.017 chains/nm2) collapse into spherical aggregates at grafted molecular weight and matrix molecular weight (B/M) 0.044 chains/nm2 and then are dispersed well as short chains at from 0.79 to 7.9. TEM images have shown clearly that the 0.06 chains/nm2 (Figure 8, first column). While at low grafting matrix chains play the solvent role more effectively for 124 kDa density of 43 kDa brush the dipolar forces are dominant brushes at the lower grafting density (Figure 9). Interestingly, resulting in anisotropic branched structures, at intermediate we observed that even at B/M < 1, matrix chains which are density dipolar forces are balanced with stronger steric entangled with the brushes at the low grafting density (0.013 repulsion, hence they collapse. We interpret this behavior as chains/nm2) can create short chains instead of long strings isotropic repulsion of the short brushes counteracting the expected to form at the dewetting interface. SAXS patterns of directional forces and form demagnetized structures through the low grafting density composites in varying matrix molecular isotropic aggregation. Previous magnetic studies on spherical weights confirm the strong interference from the ordering of aggregates of superparamagnetic nanoparticles have reported strings giving a main peak at q = 0.008 Å−1 (Figure 10a). The demagnetizing effect due to random dipole−dipole inter- average spacing between these observed nanostructures is actions.26,47 At increased grafting density of 0.06 chains/nm2, they are well-dispersed short chains mainly due to stronger calculated from d = 2π/q as 78 nm. It is apparent that with the repulsion and steric depletion of free chains around particles connectivity of strings (as clearly seen in TEMs of B/M = 7.9 (see Supporting Information for SAXS data and analysis of 43 and 2.62), the low-q feature disappears, indicating that the kDa brush, Figure S2). The transition in nanostructures is structures are large, which cannot be measured in SAXS. TEM attributed to the entanglement between brushes, which is images of particles at high grafting density (0.05 chains/nm2) controlled mainly with grafting density and is determined by show that the structures do not change with the increase of the balance between dipolar and steric repulsion forces. Our matrix molecular weight (top row of Figure 9). Their scattering results validate the recent predictions of Panagiotopoulos et patterns (Figure 10b) provide that structures are like surface al.13 where the collapse of string-like structures into isotropic fractals with power-law exponent values between 3 and 4. TEM colloidal clusters is obtained at high grafting density in their and SAXS results suggest that matrix-brush entanglements do simulations. not influence the structures at high grafting density. We demonstrate that increasing the brush length at sparse In summary, we demonstrate the structural transitions of grafting density creates chains (Figure 7), which seems to be polymer grafted iron oxide nanoparticles which depend on the analogous to what has been shown for the PS−silica grafted system.14 In that previous relevant work, amphiphilic nano- brush lengths and subtle changes of grafting density. Dispersion particles with 0.01 chains/nm2 graft density present a transition of nanoparticles was controlled at the particle size level where from connected structures into strings with the corresponding the formation of tetramers, short chains and long chains were brush lengths we used in this work. In addition, the amphiphilic obtained by tethering polymer chains onto magnetic nano- nanoparticles show transition from spheres to anisotropic particles. This system offers controllable aggregation of objects or dispersed particles as grafting density increases for structures in equilibrium which further brings additional given brush length. Here, the particular difference of having magnetic properties to the polymer nanocomposite systems. 3469 dx.doi.org/10.1021/ma3000576 | Macromolecules 2012, 45, 3463−3470 Macromolecules Article CONCLUSION (10) Wang, X.; Foltz, V. J.; Rackaitis, M.; Böhm, G. G. A. Polymer 2008, 49, 5683. Polymer grafted iron oxide nanoparticles of varying grafting (11) Smith, G. D.; Bedrov, D. Langmuir 2009, 25, 11239. densities have been synthesized for the first time using grafting- (12) Cheng, L.; Cao, D. J. Chem. Phys. 2011, 135. to method. Previous reported strategy on controlling the (13) Chremos, A.; Panagiotopoulos, A. Z. Phys. Rev. Lett. 2011, 107. grafting density of various molecular weights of brushes have (14) Akcora, P.; Liu, H.; Kumar, S. K.; Moll, J.; Li, Y.; Benicewicz, B. been applied to magnetic oxide nanoparticles and their C.; Schadler, L. S.; Acehan, D.; Panagiotopoulos, A. Z.; Pryamitsyn, V.; dispersion in solution and melts has been presented. We Ganesan, V.; Ilavsky, J.; Thiyagarajan, P.; Colby, R. H.; Douglas, J. F. showed that brush−brush entanglement plays an effective role Nat. Mater. 2009, 8, 354. only at low grafting density in the string formation. Dipolar (15) Choi, J.; Dong, H.; Matyjaszewski, K.; Bockstaller, M. R. J. Am. forces and entanglement between brushes result in long strings Chem. Soc. 2010, 132, 12537. (16) Chevigny, C.; Dalmas, F.; Di Cola, E.; Gigmes, D.; Bertin, D.; of particles. Structural transitions observed with increase in Boué, F.; Jestin, J. Macromolecules 2011, 44, 122. grafting density and brush length substantiate the interplay (17) Moll, J. F.; Akcora, P.; Rungta, A.; Gong, S.; Colby, R. H.; between steric repulsion which is enhanced with entanglement, Benicewicz, B. C.; Kumar, S. K. Macromolecules 2011, 44, 7473. and dipolar forces. Organization of functional magnetic (18) Teixeira, P. I. C.; Tavares, J. M.; Telo Da Gama, M. M. J. Phys.: nanoparticles can be tuned differently under applied fields Condens. Matter 2000, 12. than the previously reported magnetic colloids due to the (19) Cabuil, V. Curr. Opin. Colloid Interface Sci. 2000, 5, 44. strong repulsion of tethered polymers. Field-induced extended (20) Lalatonne, Y.; Richardi, J.; Pileni, M. P. Nat. Mater. 2004, 3, 121. structures of polymer grafted magnetic nanoparticles are (21) Klokkenburg, M.; Vonk, C.; Claesson, E. M.; Meeldijk, J. D.; currently under investigation. Erné, B. H.; Philipse, A. P. J. Am. Chem. Soc. 2004, 126, 16706. (22) Benkoski, J. J.; Bowles, S. E.; Jones, R. L.; Douglas, J. F.; Pyun, ASSOCIATED CONTENT J.; Karim, A. J. Polym. Sci., Part B: Polym. Phys. 2008, 46, 2267. (23) Khomutov, G. B. Adv. Colloid Interface Sci. 2004, 111, 79. * S Supporting Information (24) Benkoski, J. J.; Bowles, S. E.; Korth, B. D.; Jones, R. L.; Douglas, TEM of bare iron oxide nanoparticles in 156 kDa PS matrix, J. F.; Karim, A.; Pyun, J. J. Am. Chem. Soc. 2007, 129, 6291. interaction energy between two superparamagnetic nano- (25) Butter, K.; Bomans, P. H. H.; Frederik, P. M.; Vroege, G. J.; particles, SAXS data analysis by unified fits, Power law Philipse, A. P. Nat. Mater. 2003, 2, 88. exponents, cluster size and primary size of particles obtained (26) Gross, A. F.; Diehl, M. R.; Beverly, K. C.; Richman, E. K.; from the Unified fits. This material is available free of charge via Tolbert, S. H. J. Phys. Chem. B 2003, 107, 5475. (27) Goubault, C.; Leal-Calderon, F.; Viovy, J. L.; Bibette, J. the Internet at http://pubs.acs.org/ Langmuir 2005, 21, 3725. (28) Garcia, I.; Tercjak, A.; Zafeiropoulos, N. E.; Stamm, M.; AUTHOR INFORMATION Mondragon, I. J. Polym. Sci., Part A: Polym. Chem. 2007, 45, 4744. Corresponding Author (29) Garcia, I.; Zafeiropoulos, N. E.; Janke, A.; Tercjak, A.; Eceiza, A.; *Telephone: (201) 216-5060. E-mail: Pinar.akcora@stevens. Stamm, M.; Mondracon, I. J. Polym. Sci., Part A: Polym. Chem. 2007, edu. 45, 925. (30) Robbes, A. S.; Cousin, F.; Meneau, F.; Chevigny, C.; Gigmes, Notes D.; Fresnais, J.; Schweins, R.; Jestin, J. Soft Matter 2012, 8, 3407. The authors declare no competing financial interest. (31) Xu, C.; Ohno, K.; Ladmiral, V.; Composto, R. J. Polymer 2008, 49, 3568. ACKNOWLEDGMENTS (32) Sun, S.; Zeng, H.; Robinson, D. B.; Raoux, S.; Rice, P. M.; This work is supported by NSF-CAREER award Grant No. Wang, S. X.; Li, G. J. Am. Chem. Soc. 2004, 126, 273. (33) Yang, Y.; Yang, Z.; Zhao, Q.; Cheng, X.; Tjong, S. C.; Li, R. K. 1048865 from DMR. We acknowledge the Multiscale Imaging Y.; Wang, X.; Xie, X. J. Polym. Sci., Part A: Polym. Chem. 2009, 47, 467. Lab (LMSI) of Stevens for TEM. Portions of this work were (34) Jung, M. L.; Ok, H. K.; Sang, E. S.; Lee, B. H.; Choe, S. performed at Beamline X27C at the National Synchrotron Macromol. Res. 2005, 13, 236. Light Source (NSLS), Brookhaven National Laboratory. NSLS (35) Beaucage, G. J. Appl. Crystallogr. 1995, 28, 717. is supported by the U.S. Department of Energy, Office of (36) Taylor, W.; Jones, R. A. L. Langmuir 2010, 26, 13954. Science, Office of Basic Energy Sciences, under Contract No. (37) Huang, H.; Penn, L. S. Macromolecules 2005, 38, 4837. DE-AC02-98CH10886. (38) De Palma, R.; Peeters, S.; Van Bael, M. J.; Van Den Rul, H.; Bonroy, K.; Laureyn, W.; Mullens, J.; Borghs, G.; Maes, G. Chem. REFERENCES Mater. 2007, 19, 1821. (39) Xiao, Z. P.; Yang, K. M.; Liang, H.; Lu, J. J. Polym. Sci., Part A: (1) Pyun, J.; Matyjaszewski, K. Chem. Mater. 2001, 13, 3436. Polym. Chem. 2010, 48, 542. (2) Pyun, J.; Zhou, X. Z.; Drockenmuller, E.; Hawker, C. J. J. Mater. (40) Bourlinos, A. B.; Bakandritsos, A.; Georgakilas, V.; Petridis, D. Chem. 2003, 13, 2653. Chem. Mater. 2002, 14, 3226. (3) Lan, Q.; Francis, L. F.; Bates, F. S. J. Polym. Sci., Part B: Polym. (41) Flesch, C.; Delaite, C.; Dumas, P.; Bourgeat-Lami, E.; Duguet, Phys. 2007, 45, 2284. E. J. Polym. Sci., Part A: Polym. Chem. 2004, 42, 6011. (4) Zou, H.; Wu, S.; Shen, J. Chem. Rev. 2008, 108, 3893. (42) De Gennes, P. G. Macromolecules 1980, 13, 1069. (5) Zhao, Y.; Thorkelsson, K.; Mastroianni, A. J.; Schilling, T.; (43) Auroy, P.; Auvray, L.; Leger, L. Macromolecules 1991, 24, 5158. Luther, J. M.; Rancatore, B. J.; Matsunaga, K.; Jinnai, H.; Wu, Y.; (44) Zhang, L.; Manthiram, A. Phys. Rev. B: Condens. Matter Mater. Poulsen, D.; Fréchet, J. M. J.; Paul Alivisatos, A.; Xu, T. Nat. Mater. Phys. 1996, 54, 3462. 2009, 8, 979. (45) Bottoni, G.; Candolfo, D.; Cecchetti, A.; Masoli, F. J. Magn. (6) Sun, S. Adv. Mater. 2006, 18, 393. Magn. Mater. 1992, 116, 285. (7) Pyun, J. Polym. Rev. 2007, 47, 231. (46) Nakata, K.; Hu, Y.; Uzun, O.; Bakr, O.; Stellacci, F. Adv. Mater. (8) Laurent, S.; Forge, D.; Port, M.; Roch, A.; Robic, C.; Vander Elst, 2008, 20, 4294. L.; Muller, R. N. Chem. Rev. 2008, 108, 2064. (47) Franco, V.; Batlle, X.; Labarta, A.; O’Grady, K. J. Phys. D: Appl. (9) Corbierre, M. K.; Cameron, N. S.; Sutton, M.; Laaziri, K.; Phys. 2000, 33, 609. Lennox, R. B. Langmuir 2005, 21, 6063. 3470 dx.doi.org/10.1021/ma3000576 | Macromolecules 2012, 45, 3463−3470