Polypeptide-Coated Silica Particles Dispersed in ... - ACS Publications

Article pubs.acs.org/JPCB

Polypeptide-Coated Silica Particles Dispersed in Lyotropic Liquid Crystals of the Same Polypeptide Cornelia Rosu,†,§,⊥ Sreelatha Balamurugan,§,⊥ Rafael Cueto,§ Amitava Roy,∥ and Paul S. Russo*,†,‡,§ †

School of Materials Science and Engineering and Georgia Tech Polymer Network, GTPN and ‡School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia 30332, United States § Department of Chemistry and Macromolecular Studies Group and ∥Center for Advanced Microstructures and Devices, CAMD, Louisiana State University, Baton Rouge, Louisiana 70803, United States S Supporting Information *

ABSTRACT: When a particle is introduced into a liquid crystal (LC), it distorts the LC director field, leading to new arrangements of the particles. This phenomenon is ordinarily studied using >100 nm particles and ∼2 nm mesogens. Usually the particle surface and mesogens are chemically distinct, which adds an enthalpic effect, even though the more interesting interactions are entropic. To raise the structures to the visible regime, while minimizing chemical differences between the particle surface and mesogen, silica particles coated with an α-helical polypeptide have been prepared and dispersed in lyotropic polypeptide LCs. The polypeptide is poly(γ-stearyl-α,L-glutamate) or PSLG. To make the particles easy to manipulate and easy to find, the silica core included superparamagnetic magnetite (Fe3O4) and covalently attached dye. Two methods were used to place polypeptides on these magnetic, fluorescent particles: a multistep graf ting-to approach in which whole polypeptides were attached and a one-pot graf ting-f rom approach in which the polymerization of the monomers was initiated from the particle surface. These approaches resulted in sparse and dense surface coverages, respectively. The influence of surface curvature and polypeptide molecular weight on the design of sparsely covered particles was investigated using the graf ting-to approach. The aggregated graf ting-f rom particles when freshly dispersed in a PSLG/solvent matrix disrupted the orientation of the characteristic cholesteric LC (ChLC) phase directors. In time, the hybrid particles were expelled from some domains, enabling the return of the familiar helical twist of the cholesteric mesophase. The sparsely coated graf ting-to hybrid particles when inserted in the PSLG/solvent matrix assembled into stable islet-like formations that could not be disrupted even by an external magnetic field. The bulk particles aligned in chains that were easily manipulated by a magnetic field. These results indicate that polypeptide ChLCs can control and facilitate colloidal assembly of particles with matching surfaces.

■

INTRODUCTION Particles with an inorganic core and polymeric shell have been studied extensively. Their popularity arises from the ability to design composite materials with properties not available from either the shell or core alone.1−6 Among this class of hybrid composites, silica particles with a homopolypeptide shell are of special interest. Homopolypeptides are chiral, can undergo conformational transitions,7 and offer a wide range of properties from the same basic chemistry. The appeal of colloidal silica as the core, sometimes with an added magnetic nugget, derives from its low cost, ease of fabrication, resistance to oxidation, and easy surface functionalization. Many applications have been identified for such hybrid particles, especially in biosensing, separations, photonics, and drug delivery.7−10 Polypeptides in the solid state typically exist in well-defined, ordered secondary structures (α-helix, β-sheet). They often retain these conformations even in solution.11−13 Different © 2016 American Chemical Society

moieties anchored to the side chains make polypeptides responsive to external stimuli such as changes in pH, magnetic field, temperature, or solvent.14−17 The present study focuses on poly(γ-stearyl-L-glutamate) (PSLG).18−21 A defining feature of PSLG is the many hydrophobic C18 side chains that sprout from the α-helical polypeptide backbone like branches from a tree trunk. These side chains confer good solubility in a wide range of organic solvents. The stiffness of PSLG’s polypeptide backbone and the ability of the side chains to “self-solvate” the whole molecule make both lyotropic and thermotropic liquid crystals (LCs) possible.22−24 Attachment of PSLG to a colloidal particle enables the study of mesogen-coated particles dispersed in lyotropic PSLG cholesteric LCs (ChLCs). It is expected that the ChLC will exert forces to alter how the particles are Received: April 21, 2016 Revised: June 27, 2016 Published: July 19, 2016 7275

DOI: 10.1021/acs.jpcb.6b03863 J. Phys. Chem. B 2016, 120, 7275−7288

Article

The Journal of Physical Chemistry B

Scheme 1. Synthetic Route for the Preparation of PSLG−Silica Hybrid Particles via the (1) Multistep Graf ting-To and (2) OnePot Graf ting-From Proceduresa

a

Details are given in the Supporting Information (SI) (Section S2).

group observed a landscape of phases in mixtures of various rodlike viruses and spherelike particles. Isotropic, nematic, lamellar, columnar, smectic, and crystalline phases were found.29,30 Although polypeptides were the first syntheticpolymer LCs,11 there seems to be no study until now of polypeptide-coated particles immersed in a ChLC of the same polypeptide. Attachment of polypeptides and other biomolecules to particles can be achieved by graf ting-f rom or graf ting-to approaches. In the graf ting-f rom method, amine-functionalized silica particles initiate the polymerization of the chosen amino acid, N-carboxyanhydride (NCA), via ring-opening polymerization. Polypeptide composite particles (PCPs) produced in

dispersed. Several examples on similar systems support this hypothesis. Mushenheim et al.25 found that motile Proteus mirabilis bacteria form reversible dynamic and hierarchical cellular assemblies when dispersed in a lyotropic LC, in contrast to passive microparticles, which showed irreversible selfassembly behavior. Pendery et al. studied gold self-assembly moderated by a cholesteric matrix.26 The gold nanoparticles were used to map the defects in the LC matrix because they positioned themselves at the LC disclination lines. This particle arrangement was previously observed by Mitov et al.27 Vallooran et al. have used magnetic particles to align lipidbased lyotropic LC domains.28 The magnetite particles organized themselves into ordered architectures. The Fraden 7276


Article

The Journal of Physical Chemistry B this manner can form colloidal crystals31,32 and their homopolypeptide33 or copolypeptide34 shell thickness can be controlled by rationally choosing the initiator to monomer ratio. Magnetic PCPs have also been prepared;35 these materials find application as supports for catalytic processes36 or drug delivery.37 The advantage of the grafting-f rom method is the realization of a hybrid particle with a dense surface coverage. The main disadvantage resides in the difficulty of measuring the molecular weight of the polymer anchored to the surface, although methods to overcome such difficulties have been developed.34,38 Click chemistry,39,40 especially the alkyne−azide variant, is particularly attractive for a graf ting-to scheme, despite limited polymer loading due to steric hindrance and reduced mobility of the reactive groups at the chain end. These problems are expected to be especially severe at high molecular weights. Reports devoted to the attachment of polymers onto silica particles by alkyne−azide click coupling are limited,41−45 and even fewer studies have involved polypeptides. Balamurugan et al. attached 8 kDa PSLG onto silica particles.46 Kar et al. used N-trimethylsilyl propargyl amine as an NCA initiator in an effort to lower the polydispersity of the polypeptide and improve control over the polymerization.47 The highest molecular weight of alkyne−poly(L-glutamic acid) obtained after side chain deprotection of alkyne−poly(p-methoxybenzyl-L-glutamate) was 12 kDa (Mw/Mn = 1.04). The same method was chosen to synthesize poly(L-lysine) (PLL) (∼8 kDa by NMR, Mw/Mn = 1.05 by gel-permeation chromatography (GPC)) by side chain deprotection of the PCBL homolog.48 A block copolymer consisting of PLL and poly(Lleucine) (8 kDa, Mw/Mn = 1.1) was also used to prepare complex conjugate particles.48 Click particles proved to be good support materials for building three-dimensional mesoporous scaffolds.47 The more recent research trend in the field is to take advantage of both grafting approaches used in combination. For example, the graf ting-to + graf ting-f rom tandem proved fruitful in developing magnetic composites grafted with glycopolypeptides as T1-weighting agents for magnetic resonance imaging.49 The present report focuses on magnetic, fluorescent silica with a PSLG-functionalized surface. The goals of this work were: (1) preparation and characterization of sparsely and densely covered magnetic, fluorescent PSLG hybrid particles with same-sized cores via graf ting-to and graf ting-f rom approaches, (2) assessment of the influence of surface coverage and PSLG molecular weight on the grafting density via the graf ting-to method, and (3) investigation of the behavior of colloidal complex fluids made of magnetic, fluorescent silica− PSLG hybrids suspended in the lyotropic ChLC PSLG matrix. A word about terminology is in order. Magnetic particles made by the grafting-f rom approach are referred to as magnetic grafted from (MGF). Particles made by the graf ting-to approach are referred to as GT or magnetic grafted to (MGT). A trailing number (e.g., MGF2) distinguishes different preparations when necessary. These abbreviations appear in Scheme 1.

Preparation of Colloidal ChLC (CChLC) Mixtures. First, a PSLG LC mixture was prepared in tetrahydrofuran (THF) (0.5 mL, 40% w/v, Mn = 46 000 Da). After polymer dissolution (overnight), an appropriate amount of MGF1 (S2-II.7, SI) dispersion (c = 1 mg/mL) was added, and the concentration of MGF1 was fixed at 0.1 wt % (the excess solvent was evaporated). This solution was homogenized using a vortex agitator and then dispensed into a thin rectangular capillary tube (0.4 mm thick × 4 mm wide; VitroCom Inc.). Second, MGT (S2-I.11, SI) (2 mL original suspension) particles were washed three times with THF, using a magnet to gather them before exchanging the solvent, and finally, they were dispersed in 1 mL of THF. Nonmagnetic graf ted-to (GT2) particles were dispersed in THF by centrifugation/redispersion. The same protocol was followed when the solvent was toluene. An aliquot (50 μL) was used to determine the concentration of the final dispersions. CChLCs were prepared in 4 mL scintillation vials sealed with polytetrafluoroethylene (PTFE)-faced lids. A volume of 200 μL solvent and appropriate amounts of PSLG polymer (Mw = 70 000 and 60 000 Da) were used for concentrations of 30 and 40% (w/w), respectively. The polymer was allowed to dissolve overnight under gentle stirring until the solution became clear. These samples were allowed to rest for one more day to reach equilibrium. Vitrocom cells (rectangular, 0.4 mm thick × 4 mm wide) were used to investigate the morphology of the LC, which was further used as the matrix for CChLC. Vitrocom cells were flame-sealed at both ends. To the remaining ChLC matrix solution, volumes of particle dispersions were used to bring their concentrations to the desired values of 1, 3, or 5 wt % (MGT), and 5:1 wt % (MGT:GT1). The supplemental amount of solvent introduced into the mixture was evaporated under a slow stream of dry nitrogen. Samples were equilibrated for 2 days before insertion into the same Vitrocom cells described above for investigation by polarized light microscopy. Characterization Methods. Transmission electron microscopy (TEM) images of magnetic particles at different stages were obtained using a high-resolution TEM (HRTEM) (JEOL2010CX), operating at 200 kV, and environmental JEOL 100CX, with an accelerating voltage of 80 kV. A dilute sample solution was prepared by dispersing 20 μL of particles in 1 mL of appropriate solvent (water, ethanol, THF). From this solution a drop was placed on a 400-mesh carbon-coated copper grid (Electron Microscopy Sciences) and dried, either in air overnight or for 2−3 h under vacuum. MGT particles were sized using multiangle dynamic light scattering (MADLS). Experiments were carried out on a custom-built apparatus using an ALV-5000 digital autocorrelator and a 632.8 nm laser source. Measurements were made at multiple scattering angles, from θ = 30 to 120°. The samples were prepared in precleaned and dust-free vials by the following process. First, 1 mL of solvent (dodecane) was added via a filter combination: 0.02 μm Whatman Anotop with a guard 0.1 μm Whatman PTFE filter in series for organic solvents. The purpose of the guard is to capture debris shed by the brittle Anotop filters. Next, a drop (∼5−10 μL) of the particle dispersion was added without any filtration. The surface composition of the particles at all stages was investigated with a Kratos Analytical Axis 165 X-ray photoelectron spectrometer (XPS) with Al K radiation, with an energy of 1.48 keV and a takeoff angle of 90°. For each sample, high-resolution and survey spectra of individual elements were recorded, with pass energies of 40 and 80 eV, respectively. The peak locations were calibrated on the basis of the C1s signal

■

MATERIALS AND METHODS Materials. All chemicals and vendors are listed in the Supporting Information (Section S1). Preparation of PCPs. Details on the synthetic procedures for PSLG graf ting-to and graf ting-f rom hybrids appear in the SI (Section S2). 7277


Article


Figure 1. HRTEM micrographs of Fe3O4−SiO2−[FITC-SiO2]−SiO2 particles (A; the inset shows a Fe3O4−SiO2 particle) and TEM image of Fe3O4−SiO2−[FITC-SiO2]−NH2 particles (B). The TEM image of the magnetite core is shown in Figure S3.1.

PSLG-coated silica particles via the graf ting-to and graf ting-f rom methods. Magnetite particles, Fe3O4, two batches of the same size, were prepared by the co-precipitation method from Fe (II) and Fe (III) salts in the presence of either NaOH or NH4OH and sodium citrate used as a catalyst and stabilizer, respectively.50,51 One batch of magnetite was subjected to multistep functionalization for use in the graf ting-to method, whereas the other batch followed a one-pot procedure for use in the graf ting-f rom approach. Magnetite particles (Figure S3.1) were covered with a layer of silica, Fe3O4−SiO2, through a modified Stöber procedure.52−55 The silica-coated particles were reacted with a premade fluorescent adduct56 (fluorescein isothiocyanate-3-aminopropyltriethoxy silane (FITC−APTES); see Figure S3.2) and tetraethylorthosilicate (TEOS).57 This step yielded fluorescent magnetic particles covered with a supplemental silica layer. Listing all components from the center to the surface, the particles are named Fe3O4−SiO2− [FITC-SiO2] (Scheme 1). Additional TEOS was added in the one-batch procedure to increase the silica shell thickness. In the multistep process, because some of the dye molecules can be situated at the particle surface, a final silica protective shell was built by the regrowth procedure using the Fe3O4−SiO2− [FITC-SiO2] particles as seeds57 and yielded Fe3O4−SiO2− [FITC-SiO2]−SiO2 (Scheme 1). For the graf ting-to approach, the resulting particles (Fe3O4− SiO2−[FITC-SiO2]−SiO2) were functionalized with bromine moieties43,46,58−60 to form bromine-functionalized particles, Fe3O4−SiO2−[FITC-SiO2]−SiO2−Br (Scheme 1), which were converted to azide functional groups through nucleophilic substitution (Fe3O4−SiO2−[FITC-SiO2]−SiO2−N3) (Scheme 1).43,46,58 The polymer shell alkyne end-terminated PSLG was prepared separately by ring-opening polymerization of γstearyl-L-glutamate NCA (SLG-NCA), initiated by propargyl amine.61 In the final step, copper (I)-catalyzed azide−alkyne click chemistry between the azide-functionalized fluorescent magnetic cores and alkyne−PSLG yielded the PSLG-click magnetic composite particles (Fe3O4−SiO2−[FITC-SiO2]− SiO2−PSLG) (Scheme 1). As a reminder of the definitions established above, this particle code is abbreviated MGT for Magnetic Grafted To. For the graf ting-f rom preparation, the Fe3O4−SiO2−[FITCSiO2] particles were given an amine functionality designed to serve as the initiator in the ring-opening polymerization of SLG-NCA. The initiator particles are coded Fe3O4−SiO2− [FITC-SiO2]−NH2 (Scheme 1). The graf ting-f rom composite particles were prepared by ring-opening polymerization of SLG-NCA, initiated by amino-functionalized core particles.

occurring at 284 eV. Fourier transform infrared spectroscopy (FTIR) was used to confirm the characteristic adsorption bands. Spectra were recorded with a Bruker Tensor 27 FTIR instrument equipped with a Pike Miracle single-bounce attenuated total reflectance cell and a ZnSe single crystal. Powder X-ray diffraction (XRD) analysis of magnetite cores was performed with a custom-built instrument, using Co Kα radiation (λ = 1.7902 Å). The apparatus is equipped with a double-crystal monochromator with silica (111) and germanium (220) crystals, a Huber four-cycle goniometer capable of a step size as low as 0.0001°, and a Canberra high-purity germanium solid-state detector. The sample holders consisted of quartz cut 6° from 0001, with a 50 μm sample depth. The data were analyzed with Jade software from Materials Data, Inc. The crystalline structures of the grafting-f rom MGF2 particles were analyzed with a Panalytical Empyrean multipurpose diffractometer instrument, using Cu Kα radiation (λ = 1.540598 Å). Thermogravimetric analysis (TGA) of the samples was carried out with a TGA Q50 apparatus from TA Instruments under nitrogen flow and at a heating rate of 10 °C min−1. The molecular weight of the alkyne−PSLG was measured by GPC with multiangle light scattering (GPC/ MALS). An instrument equipped with an Agilent 1100 solvent degasser, Agilent 1100 pump, and Agilent 1100 autosampler was used for separation. The column set was a 10 μm, 50 × 7.8 mm2 guard column and two Phenogel 300 × 7.8 mm2 columns from Phenomenex (Torrance, CA): (1) 10 μm, 105 Å (10− 1000 kDa) and (2) 10 μm, MXM (100−10 000 kDa). The Wyatt Dawn DSP-F MALS detector used a He−Ne laser. A Hitachi L-7490 differential index detector (32 × 10−5, refractive index full scale) served as concentration detector. An injection volume of 100 μL was used for separation. The mobile carrier was THF (1 mL min−1) stabilized with 250 ppm butylated hydroxytoluene. The specific refractive index increment of PSLG was taken as 0.080 ± 0.002 mL g−1.20 A Bruker PROFLEX III matrix-assisted laser desorption ionization timeof-flight (MALDI-TOF) mass spectrometer was used to record the matrix-assisted laser desorption/ionization TOF mass spectra of alkyne−PSLG. The matrix used was dithranol, and chloroform served as the solvent. Magnetic measurements on magnetite particle dispersions in ethanol (0.1077 g total solids) were performed on a super-quantum interference device (SQUID) magnetometer (Quantum Design MPMS XL-7).

■

RESULTS AND DISCUSSION Preparation and Characterization of Sparsely and Densely Covered Magnetic, Fluorescent PSLG Hybrid Particles. Scheme 1 shows the steps in preparing hybrid 7278


Article

The Journal of Physical Chemistry B Two weight ratios of the amine-functionalized particle and SLG-NCA monomer, 1:1.63 (1 g of Fe3O4−SiO2−FITCSiO2−NH2 initiator particle with 0.153 mol L−1 SLG-NCA monomer) and 1:12 (1 g of Fe3O4−SiO2−FITC-SiO2−NH2 initiator particle with 0.602 mol L−1 of SLG-NCA monomer), were used to synthesize Fe3O4−SiO2−[FITC-SiO2]−PSLG-1 (MGF1) and Fe3O4−SiO2−[FITC-SiO2]−PSLG-2 (MGF2) respectively (Scheme 1); see also Section S2.II.7. (MGF stands for Magnetic Grafted From.) The shapes and sizes of the magnetic cores and PSLG composite particles were analyzed by HRTEM and environmental TEM, respectively. The first silica deposition (Fe3O4− SiO2) was 4.5 ± 1 nm thick, as shown in the inset of Figure 1A. The protective silica coverage performed in the multistep approach (Fe3O4−SiO2−[FITC-SiO2]−SiO2) had a thickness of 16 ± 3 nm (Figure 1A). The TEM image of Fe3O4−SiO2− [FITC-SiO2]−NH2 (Figure 1B) shows quasispherical core− shell structures, with an iron-oxide core and silica shell. The particles had an average size of 45 ± 3 nm. Only a small number of particles can be sized conveniently by TEM; thus, MADLS was used to measure the hydrodynamic radius, Rh, of the hybrid particles in a dispersed state. To reduce the effect of number fluctuations,62 these studies were conducted using optical settings designed to increase the scattering volume (and with it, the number of scatterers) at the expense of optical coherence. Even so, some curvature and flattening of the correlation functions was observed at large lag times, probably reflecting residual number fluctuations on the time scale required to traverse the detected volume, rather than the desired 2π/q (where q is the spatial frequency or scattering vector magnitude). At shorter lag times, the semilogarithmic representation of the correlation functions (Figure 2A) curves gently, an indication of modest polydispersity. The radius values (Figure 2B) correspond to clusters stabilized either by magnetic attractions or by the “stickiness” of the C18 side chains in the solvent used for these measurements, which was dodecane. The apparent size decreases with scattering angle, which is typical for polydisperse, large particles (the largest ones in the distribution contribute less to the total signal as the angle increases). The normalized variance (quotient of second cumulant μ2 to first cumulant or average decay rate Γ̅ 2) followed no particular trend, as shown in Figure 2B (inset). The average value (ca. 0.1) typifies well-dispersed but nonuniform particles. The elevated temperature at which the measurements were performed (60 °C) could not overcome the cluster formation. Clusters measured at lower angles had an average hydrodynamic diameter of 600 nm, equivalent in diameter to that of about 10 particles. The number of particles in the cluster depends on how they are distributed, and this is not known. Information on the Fe3O4 crystal and its magnetic properties was obtained by XRD and SQUID, as seen in Figure 3C,D, respectively. Fe3O4 has an inverse spinel structure described by a cubic close packing, with tetrahedral sites occupied by Fe3+ and octahedral sites randomly shared by either Fe3+ or Fe2+ in the same proportions. On the basis of this arrangement, the structural formula can be written as [Fe3+](Fe2+Fe3+)O4.63 The characteristic 311 peak was located at 2θ = 41.4°, which matched the standard values of pure magnetite taken from NIST file no. 19-0629. Calculations based on the Scherrer equation (eq 1) allowed for evaluation of the crystallite (coherent diffraction domain) size for both pure magnetite and silica-coated particles. The relation is

Figure 2. Dynamic light scattering results for MGT particles dispersed in dodecane at 60 °C: (A) semilogarithmic plot of the intensity autocorrelation function vs time at various angles, (B) apparent hydrodynamic radius Rh,app and polydispersity index μ2Γ−2 (inset) as a function of the squared scattering vector magnitude, q2. Run times were 200 s.

d=

Kλ β cos θ

(1)

where d is the crystallite diameter, K is the shape factor (approximately 0.93 for magnetite, 2[ln(2/π)1/2]),64 λ is the Xray wavelength (Co Kα = 1.7092 Å) , β is the line broadening at half the maximum intensity (full width at half-maximum) for the 311 peak, and θ is the Bragg angle. The measured crystallite values for Fe3O4 were 11 ± 1 nm (Figure 3C, black trace) and for Fe3O4−SiO2, 19 ± 1.5 nm (Figure 3C, red trace), respectively. The difference in 2θ values associated with the 311 peak for MGF when compared to that for Fe3O4, 35.6 versus 42°, is due to the different instrument sources; a Cu Kα was used for MGF. The relative intensities and peak positions of MGF1 (Figure 3A) also matched well with those of standard Fe3O4 (NIST file no. 19-0629). The broad diffraction peak between 10 and 30° is due to the amorphous silica coating of the Fe3O4 particle. The SQUID measurements revealed the magnetic nature of both pristine magnetite cores (Figure 3D) and MGF1 (Figure 3B). 7279


Article


Figure 3. XRD and SQUID data for MGF1 (A, B), magnetite (Fe3O4, black trace), and silica-coated magnetite (Fe3O4−SiO2, red trace) (C, D).

Table 1. Sample Code; Monomer-to-Initiator Ratio; Theoretical, Number-Average, and Weight-Average Molecular Weights; Polydispersity Index; and Length for Alkyne End-Terminated PSLG, PSLG-Pr

a

polymer

[M]/[I]

theoretical M/Da

Mn/Da

Mw/Da

Mw/Mn

L/Åa

PSLG-Pr

100:1

38 000

25 700

30 000

1.17

118

Based on Mw

The magnetic properties of the particles greatly depend on their sizes. A close inspection of the magnetization curve for pure magnetite shows no hysteresis loops at 300 K. This behavior characterizes superparamagnetic particles with a grain size of up to a critical value of 20 nm.65 Native magnetite showed a slow approach to saturation at a high field and no coercivity, Hc. The lack of hysteresis in the magnetization curve demonstrates the superparamagnetic nature of the magnetite core. The same trend was also observed for the hybrid MGF1 particle. Additional evidence of the magnetic behavior of the two hybrid particles appears in the SI (Figure S3.3). MGT particles were exposed to a magnetic field provided by a permanent magnet. In response, they aligned in straight chains, more or less uniform in length and thickness. When the magnet was removed, the chains quickly lost their alignment and adopted an undulating appearance. MGF, carrying a significantly higher load of polypeptide than MGT, also responded quickly to an external magnetic field and showed chaining behavior. This observation demonstrates that the polypeptide load did not destroy the ability of the magnetite core to respond to an applied magnetic field.

XPS was used to verify the surface composition of the products obtained at each step during hybrid particle preparation, as displayed in Figure S3.4 of the SI, along with details of the XPS data analysis. These results confirm the functionalization of the silica-protected magnetite surface with amine groups and bromine groups, the conversion to azide, and the grafting of polypeptide to and from the surface. One of the advantages of the grafting-to method compared to the graf ting-f rom approach is that the former allows in-depth characterization of the polymer destined to constitute the shell. The molecular weights and end groups of the alkyne−PSLG polypeptide before grafting onto the particle surface were determined by GPC/MALS (Table 1) and MALDI-TOF (Figure 4), respectively. Comparing the average molecular weight values determined by GPC/MALS to the monomer/initiator ratio shows that not all initiator molecules were fast enough to initiate uniformly growing polymer chains. The amine-initiated mechanism was confirmed by the MALDI-TOF analysis (Figure 4) (which, however, does not find any polypeptides as large as the GPC/ MALS measurements suggest). A range of single peaks was observed: from m/z = 1731 to 5715 Da, with a mass difference of 381 units, matching the SLG repeat units plus that of the 7280


Article


usual discrimination against large molecules by the mass spectrometer detector. Polypeptides are known11 to adopt different secondary structures depending on the solvent, temperature, and pH (for aqueous systems). PSLG is helical in solvents such as THF, toluene, and chloroform. A single turn of the helix requires 3.6 monomer units, with a length of 5.4 Å for a projection of 1.5 Å per monomer subunit. The total number of helical turns is ∼20 on the basis of a DP of about 78 computed from GPC/MALS. The formation of the helical secondary structure requires between 10 and 18 monomers, depending on the polypeptide.11 Consequently, the alkyne−polypeptide shell was expected to exist in the α-helical conformation prior to surface attachment. FTIR was used to investigate the conformation of the PSLG polypeptide, both untethered and tethered to the particle surface. Figure 5 illustrates the IR spectra of PSLG hybrid particles produced by both graf ting-to and graf ting-f rom procedures. In the case of the graf ting-to procedure, Figure 5A shows specific bands of amide I at 1654 cm−1 and amide II at 1546 cm−1, which characterize the amide α-helical structure. The N− H stretching of amide A and the N−H end-group stretch overlapped in the region from 3500 to 3200 cm−1. The peaks visible between 3000 and 2800 cm−1 correspond to asymmetric and symmetric vibrations of the hydrocarbon side chain (CH2)17CH3. The CO ester stretching of the polypeptide can be seen at 1730 cm−1. Figure 5B suggests that PSLG

Figure 4. MALDI-TOF MS of PSLG prepared with propargyl amine initiator (PSLG-Pr).

alkyne end groups (58 amu). All species recorded in the mass spectra have a supplemental 35−39 amu, which may be explained by the adherence of potassium ions (K+) from glassware. The corresponding peaks for the high-molecularweight polymer (equivalent to degree of polymerization (DP) = 78, as measured by GPC/MALS) were not observed in the MALDI-TOF spectrum because they became weaker due to the

Figure 5. FTIR spectra of MGT (A), alkyne end-terminated PSLG (B), Fe3O4−SiO2−[FITC-SiO2]−NH2 (C), MGF1 (D), and pristine PSLG polymer (E). Spectra were recorded in chloroform and corrected for solvent background. 7281


Article


Figure 6. TGA traces of all steps necessary to synthesize the grafting-to (MGT) (A) and grafting-f rom (MGF1, MGF2) magnetic fluorescent particles (B).

position range from 200 to 600 °C matches that of the tethered PSLG polypeptide. The mass difference calculation revealed a 6% loss associated with the polypeptide shell. This finding confirmed a low grafting density with polypeptide and suggested a sparse population of chains on the core surface when compared to that in the graf ting-f rom particles. Compared to that in the Fe3O4−SiO2−[SiO2-FITC]−NH2 particles, (Figure 6B), MGF1 and MGF2 particles show a rapid weight loss from 270 to 540 °C, attributed to the decomposition of the polypeptide. This was ascertained by comparing the decomposition curves of these particles with those of untethered PSLG polypeptide prepared separately. TGA analysis of MGF1 (Figure 6B) shows an increase in weight loss by ∼40% when compared to that in the aminofunctionalized core. This percentage originates from the contribution of the polypeptide to the total mass. Comparison of the TGA profiles of MGF1 and MGF2 reveals that as the monomer concentration increased 12 times the amount of grafted polypeptide on the particle surface increased from ∼40 to 70%, suggesting an increase in the molecular weight of the polypeptide chains. Influence of Surface Curvature and PSLG Molecular Weight on the Grafting Density of MGT particles. The most important feature of the graf ting-to procedure is the possibility to characterize the soft polymer shell before its attachment to the hard core particle. The disadvantage is the low grafting density. The graf ting-f rom method enables a higher polymer load, but measuring its molecular weight is difficult because the core must be dissolved, sometimes under harsh conditions, a process that may alter the shell. In this light, two new batches of graf ting-to particles were prepared to understand how surface curvature and polymer molecular weight can influence coverage. Details on preparation and TGA traces appear in Section S2, Figures S3.9−S3.11. The conversion of GPC/MALS and TGA data on grafting density, σ, follow the expression46

retained the helical conformation after grafting on the spherical surface. The characteristic peaks of the grafted polypeptide are visible at the same wavenumbers within experimental error. The supplemental peak centered at 2100 cm−1 is assigned to the azide groups, which is in agreement with XPS data showing free azide moieties on the surface as a possible prerequisite of sparse coverage. The specific signal of the Si−O−Si band is also visible at 1095 cm−1. Additional IR data are provided in the SI (Figure S3.8). Turning now to the graf ting-f rom products, Figure 5C shows small absorption peaks in the area of 1500−1700 cm−1 due to N−H bending of amine groups on the surface of the particles. In the spectrum of MGF1 particles (Figure 5D), the characteristic peaks of the PSLG polypeptide (Figure 5E) are present beside the peaks at 1061 cm−1 coming from Si−O−Si, stretching within the silica core. Compared to those in aminefunctionalized particles, the absorption signals at 3283 cm−1 (main chain amide), 2915 and 2847 cm−1 (side chain CH2), 1736 cm−1 (ester CO), 1652 cm−1 (amide I main chain), 1545 cm−1 (amide II main chain), and 1470 cm−1 (side chains) confirm that α-helical PSLG has grown from the initiator attached to the surface. FTIR was also used to calculate the average helix orientation, θ (tilt angle), to the particle surface normal. The tilt angle of MGT was 50° ± 2.25° and that of MGF1 was 51° ± 2°. Details on calculations appear in Section S4. TGA provides further evidence supporting sparse and dense polypeptide coverage for MGT and MGF particles, respectively (Figure 6). Figure 6A shows the TG traces of the steps involved in MGT synthesis, recorded under a nitrogen atmosphere. The TGA profile of Fe3O4−SiO2 (magenta curve, mass difference ∼ 1%) is in agreement with the TEM and XRD measurements, suggesting a thin protective layer of silica. The TGA curve of the dye-labeling step, Fe3O4−SiO2−[FITC-SiO2], indicates a robust fluorescent particle (∼6% mass difference when compared to that of the magnetite core), which is desired for specific applications such as probe diffusion and in vivo or in vitro bioimaging. The next trace of Fe3O4−SiO2−[FITCSiO2]−SiO2 confirms the presence of an exterior silica protective layer, indicated by an increase in the thermal stability of the fluorescent particles. The weight loss up to 200 °C in the case of MGT corresponds to adsorbed water and probably to unreacted azide functional groups. The decom-

σ=

−1 −1 f PSLG w H − wsilica × [1 − f PSLG w H]

M × Sspec × (1 − μmol m−2

−1 f PSLG w H)

× 106 , (2)

where f PSLG is a correction factor representing the fraction of unattached PSLG polymer that decomposes between 200 and 7282


Article


Table 2. Particle Code, Radius of the Core, Weight-Average Molecular Weight of PSLG Shells, PSLG Grafting Densities, and Number of PSLG Chains Per Particle and Per Square Nanometer code

Rcore/nm

Mw/kDa

grafting density, σ (μmol m−2)

PSLG chains/particle

PSLG chains/nm2

MGT GT1 GT2

30 ± 2 103 ± 2 103 ± 2

30 26 107

0.03 0.05 0.02

140 3500 1750

0.012 0.026 0.013

Figure 7. PSLG/THF LC solution containing MGF1 particles immediately after preparation (A) and after 5 days (C); the polarizer and analyzer are crossed by 90°. Epifluorescence images immediately after preparation (B) and after 5 days (D). All scale bars are 100 μm.

600 °C from a hypothetical 1 g of hybrid particles. It accounts for the fact that PSLG does not “burn” to null char (for details see Section S5). wsilica represents the weight loss of silica, wH the weight loss of the MGT hybrid in the 200−600 °C temperature interval, M the molecular weight of the polypeptide (g mol−1), and Sspec the surface specific area. Table 2 summarizes the values obtained for the three batches of graf ting-to hybrids. As reflected in Table 2, all hybrid particles had a low grafting density and a sparsely covered surface, in agreement with XPS and TGA data. Achieving a sparse population of biopolymers such as PSLG on a curved surface enables other natural molecules to be bound with minimal steric hindrance. Let us look now into the effect of surface curvature on graf ting-to efficiency. The core of MGT is 3-times smaller than that of GT1, whereas the shell polymers are similar in molecular weight. The MGT polymer has a polypeptide grafting density of 0.03 μmol m−2, corresponding to about 140 polymer chains per particle, whereas that of GT1 was 0.05 μmol m−2, corresponding to about 3500 chains per particle. These values indicate that grafting relatively short PSLG polymers onto flatter surfaces enables a higher polypeptide cargo, as though PSLG prefers a more planar surface. When the core sizes were the same, GT1 versus GT2, the hybrid carrying 4-times-shorter PSLG, GT1, had a higher grafting efficiency. This seems to suggest that long chains of polymer can bend down onto the surface during reaction time and hinder the

available functional azide groups from reacting with the incoming alkyne−PSLG chains. GT2 only carried 1750 PSLG chains, accounting for a grafting density of 0.02 μmol m−2. Behavior of the PSLG-Coated Magnetic Fluorescent Particles in Cholesteric Complex Fluids. PSLG-coated particles were dispersed in a PSLG/THF LC matrix to investigate the nature of interactions between the hybrid particles and a cholesteric host matrix. In an isotropic fluid, the perturbation attraction theorem66 guarantees that any two particles will feel a force of mutual attraction (not necessarily the net force, particularly for charged particles). Replacement of the isotropic medium with an anisotropic environment characterized by long orientational order changes the particle behavior drastically. The ChLC director becomes distorted near the particle. These local perturbations give rise to disclinations or defects of various geometries that depend on particle size and anchoring.67 PSLG, with its rigid α-helical backbone and long paraffinic side chains, supports both lyotropic and thermotropic LCs. 22 Figure 7 shows polarized optical microscopy images of a mixture of MGF1 particles in a 40% w/v PSLG in THF LC. The upper pair of images (Figure 7A,B) was taken immediately after introducing the mixture into a thin rectangular capillary tube. It was too early for the solution to display well-formed LC textures; in particular, the usual cholesteric banding68 is not evident. The epifluorescence 7283


Article


Figure 8. POM micrographs (color plate) of the PSLG (Mw = 70 kDa)/THF matrix solution (ChLC), 30% (w/w) PSLG70/THF (A); a mixture of 3 wt % MGT/30% PSLG70/THF after 2 days (B), 2 weeks (C, under a magnetic field), and 2 months (D, the Vitrocom cell was kept in the vertical position for at least 3 weeks (see Figure S3.12), white dashed circles mark defect zones); mixture of 5 wt % MGT/30% PSLG70/THF after 2 days (E, inset: a distorted cholesteric droplet (white dashed circle) surrounded by MGT particles seen at a higher magnification), 2 weeks (F, applied magnetic field), and 2 months (G, in the vertical position for ∼3 weeks, applied magnetic field), and fluorescence image of the mixture after 2 months (H); mixture of a 5:1 ratio of MGT (Dapp = 60 ± 3 nm) to GT1 particles (Dapp = 210 ± 10 nm) in ChLC after 2 days (I), 2 weeks (J, under a magnetic field, (inset) particles lying near the vial ceiling), and 2 months (K, under a magnetic field), and fluorescence image of a 2 month old mixture (L). The white arrows represent the positions of crossed polarizers (90°). The white “horse shoe” symbol near the letter B represents a permanent magnet (B = 425 mT) and its orientation with respect to the Vitrocom capillary.

image (Figure 7B) shows that the fluorescent MGF1 particle aggregates are dispersed throughout the sample. The lower set of figures (Figure 7C,D) shows the same sample after 5 days. Now a better-formed cholesteric phase occupies the lower right portion of the image. The cholesteric mesophase seemed to exclude all MGF1 particles/aggregates from the lateral glass walls to the middle. The clear MGF1-free areas were symmetrical with respect to the zone (∼2 mm wide) in which MGF1 particles were concentrated. The exclusion of MGF1 from the glass wall was confirmed by epifluorescence (Figure 7D). Presumably, the graf ting-f rom hybrids have a dense and thick corona that interacts with the PSLG chains dissolved within the ChLC matrix. The ability of a LC, in this case a polymer lyotropic LC, to exclude particles has been discussed at length by Lavrentovich67 and can be used to transport particles within LC. It is normal for impurities to be excluded from solid crystals of small molecules, and apparently the soft interactions supporting LCs, even these polymeric and lyotropic ones, can do the same. Like MGF analogs, MGT hybrid particles were suspended in a PSLG ChLC, as shown in Figure 8.

The pure PSLG/THF ChLC is displayed in Figure 8A and had pitch p of 40 ± 5 μm. After 2 days of equilibration, the 3 wt % MGT sample developed the fingerprint pattern characteristic to the ChLC phase (Figure 8B). A portion of the aggregated particles was repelled from the PSLG/THF matrix near the ChLC−glass interfaces. Particles were observed at both the top and bottom interfaces, so the effect is not merely sedimentation. Others inhabited the ChLC at different depth levels, between the glass boundaries and the middle of the vial. In the middle of the CChLC layer, MGT particles were excluded to certain defect zones (Figure 8C,D, the white circles). The cholesteric pitch of bulk PSLG increased to 60 ± 4 μm (2 days, no magnetic field) and stabilized at a significantly higher value of 85 ± 5 μm (2 months, magnetic field). The CChLC containing 5 wt % MGT particles shows no characteristic cholesteric fingerprint after 2 days of equilibration, Figure 8E. This complex LC mixture consisted of several phases: distorted cholesteric droplets with MGT particles “arrested” at the droplet surface and within (Figure 8E, inset), distorted cholesteric mesophases of the PSLG/THF matrix, and bulk MGT particles “deposited” at different locations from the middle to near the ChLC−glass interface. A higher amount 7284


Article


Figure 9. Epifluorescence images of islet assemblies found in MGT/30% PSLG/THF at weight concentrations of 3% (A), 5% (B), and MGT/GT2/ 30% PSLG/THF at a 5:1 ratio (C). See Figure S3.14 for a high magnification of the MGT islet. The bar marker is 100 μm in all figures. Here, Panel B is identical to Panel H in Figure 8.

Figure 10. Optical micrographs of 1 wt % MGT hybrid particles suspended in 40% PSLG (Mw = 60 000 Da)/toluene, TOL (ChLC) after (A) 2 days, linear arrangements (no applied magnetic field), (B) 4 days, chains of aligned particles (under a magnetic field), (C) 6 days, distorted cholesteric droplet surrounded by particles (no magnetic field), (D) 8 days, islet self-assemblies (no magnetic field). The analyzer and polarizer crossed at 90°. Scale bar 100 μm.

the islets in epifluorescence. The CChLC was exposed to an external magnetic field in an attempt to disturb the islet selfassemblies. These structures proved insensitive to the magnetic field provided by a magnet positioned near the vial for a long time (>1 month). The stability of the islets, as well as bulk particles/aggregates positioned at various locations within the PSLG host, is most probably due to the elastic forces of the system used to reach a new energy minimum and to compensate against the perturbations introduced by particles. Over a period of 2 months, the system progressively evolved to an equilibrium state, in which the LC director fields re-establish their helical twist orientation (p = 46 ± 4 μm) (see Figure 8G).

of particles introduced into the PSLG/THF ChLC host had the ability to perturb the orientation of the LC director fields within the bulk cholesteric phase, causing distortion of the characteristic fingerprint. As in the case of 3 wt % MGT, the balance between the CChLC elastic and gravitational forces caused a new repartition of most MGT particles, many “floating” near the glass−ChLC interfaces (top and bottom of the cell). In time, the distorted cholesteric droplets probably coalesced and a new phase developed, resembling islet-like formations of MGT particles. These islets were observed during the entire time of the experiment and caused the fuzzy appearance in Figure 8F,G. Figure 8H shows the appearance of 7285


Article


inset), especially in the vicinity of the ChLC meniscus. After 2 more days, the distorted ChLC droplets disappeared, leaving behind large islet formations (Figure 10D), the same behavior observed when THF was used as a solvent.

On the other hand, at the ChLC meniscus, where the distorted ChLC bulk phase depleted the islets, chains of particles aligned by the magnetic field could be seen (see Figure S3.13). The MGT chains survived attempts to remove the particles by a magnetic field, suggesting that they are strongly bound together. They may be “debris” from the MGT formations at the interface between the PSLG distorted cholesteric droplets and bulk PSLG LC matrix. The defects introduced into the PSLG matrix by these chains may explain the preservation of the distorted instead of characteristic cholesteric mesophase in that region.69 A possible explanation for the islet and chain stabilities is the formation of local LC regions in the areas in which particles interpenetrate their PSLG shells. Any alignment of the rodlike chains would be further supported by the sticky interactions of the waxy C18 side chains. Because an applied magnetic field alone could not change the morphology of the MGT islets, a hybrid particle with a larger core size, GT2, was introduced to learn whether an internal perturbation by another particle could impact the MGT selfassemblies. Figure 8I−L displays a 5:1 wt % mixture of magnetic MGT (Dh = 60 ± 3 nm) and GT1 (Dh = 210 ± 10 nm) in a 30% PSLG/THF ChLC matrix. The twist of the PSLG/THF cholesteric mesophase was apparent after 2 days of equilibration (Figure 8I) and stabilized over time (Figures 8J,K). The cholesteric pitch varied from 43 ± 7 μm (2 days) to 42 ± 2 μm (2 weeks) and stabilized at 40 ± 8 μm (2 months). Large clusters of particles were present near the ChLC−glass interfaces (Figure 8J, inset). The corresponding epifluorescence image (Figure 8L) suggests that the aggregated particles were well dispersed in the proximity of the ChLC−glass interfaces. Before taking these images, the Vitrocom cell had been stored in a vertical position; thus, sedimentation is not responsible for the particles being located near the glass walls. At first look, islets were not observed, as that in 5 wt % CChLC. Yet, zigzag formations of droplets were identified at the CChLC meniscus interface and beneath the particle layer positioned near the ChLC−glass interfaces. Figure 9 shows the appearance of islets found in 3 and 5% MGT (Figure 9A,B) and 5:1% MGT/GT2 (Figure 9C) CChLC using epifluorescence microscopy. To explore the effect of solvent on the physical behavior of CChLCs, MGT hybrid particles were suspended in a PSLG/ toluene ChLC solution (40 wt %, Mw = 60 000 Da) in toluene at 1 wt % mass fraction, as seen in Figure 10A−D. Figure 10A shows the texture of the MGT particles 2 days after preparation. “Ridges” made of particles seemed to arrange in a preferred direction without being exposed to an external magnetic field. The appearance of the CChLC mixture suggests that the PSLG/toluene matrix can “template” the assembly of the MGT particles using as a tool the perturbations in the ChLC director fields. In the next step, conducted 4 days after mixing, an external magnetic field (B = 425 mT) was applied to test whether the particles were able to respond and align. The particles aligned within 15−20 min, as shown in Figure 10B. The orientation of the magnetic field was changed to test whether the particles were “locked” in the chain-like texture (see Figure S3.15). Particles were able to reorient accordingly with the different directions of the applied magnetic field.70 After the magnetic field was removed, the system evolved to an equilibrium state in which distorted ChLC droplets started to develop and pushed the MGT particles to their surfaces. The agglomerated polypeptide particles appear in Figure 10C as thick rings around “clear” droplets. The reformation of the PSLG cholesteric pattern can be infrequently seen (Figure 10C,

■

CONCLUSIONS Hybrid composite particles featuring hydrophobic polypeptide shells of PSLG were prepared by both the graf ting-to (multistep) and graf ting-f rom (one-pot) methods. The responsive nature of the composites was achieved by insertion of a magnetic core consisting of magnetite. The fluorescent motif was realized by covalently labeling the interior silica layers with the APTES−FITC adduct. The surface of the MGT hybrids had a low population of polypeptide chains, whereas that on the surfaces of MGF1 and MGF2 was difficult to quantify. The effect of surface curvature and PSLG molecular weight on the grafting-to-payload efficiency was investigated. A high curvature favors a high grafting density, whereas short polymers get attached to the surface in a higher numbers than those of the long ones. By minimizing the composition difference between the LC matrix and surface of the colloidal particle, the systems developed here facilitate the study of how LCs can direct the assembly of colloidal matter when particle and mesogen have comparable sizes. Suspensions of the PSLG-coated particles in ChLCs displayed complex behaviors, all readily visible under an optical microscope. Insertion of PSLG-coated particles into the bulk PSLG cholesteric phase slowed the formation of the usual cholesteric twist, leading to changes in optical properties. Perhaps the most important property of these mixtures was their ability to generate islets of PSLG-coated particles, regardless of the solvent and weight fraction of the particles. The islets displayed remarkable stability; an applied magnetic field could not disturb their structures. This feature was demonstrated especially by CChLC mixtures in toluene. Initially, PSLG−silica particles could be manipulated by an external magnetic field, but after the islets started to inhabit the CChLC, they were no longer responsive. Local LC regions may form as the particle shells interdigitate and stabilize the structure.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpcb.6b03863. Section S1, Materials; Section S2, Detailed synthetic procedures; Section S3, Supplemental Graphical Representations; Section S4, Calculation of average helix orientation, θ (tilt angle); and Section S5, Calculation of polypeptide grafting density (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Tel: (404)-385-2607. Author Contributions ⊥

C.R. and S.B. contributed equally to this work.

Notes

The authors declare no competing financial interest. 7286


Article


■

(15) Knoop, R. J. I.; de Geus, M.; Habraken, G. J. M.; Koning, C. E.; Menzel, H.; Heise, A. Stimuli responsive peptide conjugated polymer nanoparticles. Macromolecules 2010, 43, 4126−4132. (16) Lowik, D.; Leunissen, E. H. P.; van den Heuvel, M.; Hansen, M. B.; van Hest, J. C. M. Stimulus responsive peptide based materials. Chem. Soc. Rev. 2010, 39, 3394−3412. (17) Quadir, M. A.; Martin, M.; Hammond, P. T. Clickable Synthetic PolypeptidesRoutes to New Highly Adaptive Biomaterials. Chem. Mater. 2014, 26, 461−476. (18) Daly, W. H.; Poché, D.; Negulescu, I. I. Poly (γ-Alkyl-α, Lglutamate)s, derived from long chain paraffinic alcohols. Prog. Polym. Sci. 1994, 19, 79−135. (19) Jamil, T.; Russo, P. S.; Daly, W. H.; Negulescu, I. Static and dynamic light-scattering of a random coil polymer in rodlike polymersolution. Macromol. Symp. 1994, 79, 81−85. (20) Poché, D. S.; Daly, W. H.; Russo, P. S. Synthesis and some solution properties of poly(γ-stearyl α,L-glutamate). Macromolecules 1995, 28, 6745−6753. (21) Jamil, T.; Russo, P. S.; Negulescu, I.; Daly, W. H.; Schaefer, D. W.; Beaucage, G. Light-scattering from random coils dispersed in solutions of rodlike polymers. Macromolecules 1994, 27, 171−178. (22) Schmidtke, S.; Russo, P.; Nakamatsu, J.; Buyuktanir, E.; Turfan, B.; Temyanko, E.; Negulescu, L. Thermoreversible gelation of isotropic and liquid crystalline solutions of a “sticky” rodlike polymer. Macromolecules 2000, 33, 4427−4432. (23) Watanabe, J.; Sakajiri, K.; Okoshi, K.; Kawauchi, S.; Magoshi, J. Columnar liquid crystals in polypeptides, 1 Columnar hexagonal phase observed in lyotropic solutions of poly(γ-octadecyl-L-glutamate) dissolved in amphiphilic solvents, such as octadecylamine, octadecyl alcohol, and octadecanoic acid. Macromol. Chem. Phys. 2001, 202, 1004−1009. (24) Watanabe, J.; Takashina, Y. Columnar liquid-crystals in polypeptides. 1. A columnar hexagonal liquid-crystal observed in poly(γ-octadecyl L-glutamate). Macromolecules 1991, 24, 3423−3426. (25) Mushenheim, P. C.; Trivedi, R. R.; Tuson, H. H.; Weibel, D. B.; Abbott, N. L. Dynamic self-assembly of motile bacteria in liquid crystals. Soft Matter 2014, 10, 88−95. (26) Pendery, J. S.; Merchiers, O.; Coursault, D.; Grand, J.; Ayeb, H.; Greget, R.; Donnio, B.; Gallani, J.-L.; Rosenblatt, C.; Felidj, N.; Borensztein, Y.; Lacaze, E. Gold nanoparticle self-assembly moderated by a cholesteric liquid crystal. Soft Matter 2013, 9, 9366−9375. (27) Mitov, M.; Portet, C.; Bourgerette, C.; Snoeck, E.; Verelst, M. Long-range structuring of nanoparticles by mimicry of a cholesteric liquid crystal. Nat. Mater. 2002, 1, 229−231. (28) Vallooran, J. J.; Bolisetty, S.; Mezzenga, R. Macroscopic alignment of lyotropic liquid crystals using magnetic nanoparticles. Adv. Mater. 2011, 23, 3932−7. (29) Adams, M.; Dogic, Z.; Keller, S. L.; Fraden, S. Entropically driven microphase transitions in mixtures of colloidal rods and spheres. Nature 1998, 393, 349−352. (30) Adams, M.; Fraden, S. Phase behavior of mixture’s of rods (tobacco mosaic virus) and spheres (polyethylene oxide, bovine serum albumin). Biophys. J. 1998, 74, 669−677. (31) Fong, B.; Russo, P. S. Organophilic colloidal particles with a synthetic polypeptide coating. Langmuir 1999, 15, 4421−4426. (32) Fong, B.; Turksen, S.; Russo, P. S.; Stryjewski, W. Colloidal crystals of silica-homopolypeptide composite particles. Langmuir 2004, 20, 266−269. (33) Soto-Cantu, E.; Turksen-Selcuk, S.; Qiu, J. H.; Zhou, Z.; Russo, P. S.; Henk, M. C. Silica-polypeptide composite particles: controlling shell growth. Langmuir 2010, 26, 15604−15613. (34) Borase, T.; Iacono, M.; Ali, S. I.; Thornton, P. D.; Heise, A. Polypeptide core−shell silica nanoparticles with high grafting density by N-carboxyanhydride (NCA) ring opening polymerization as responsive materials and for bioconjugation. Polym. Chem. 2012, 3, 1267−1275. (35) Liu, D.; Li, Y.; Deng, J. P.; Yang, W. T. Synthesis and characterization of magnetic Fe3O4-silica-poly(γ-benzyl-L-glutamate) composite microspheres. React. Funct. Polym. 2011, 71, 1040−1044.

ACKNOWLEDGMENTS This work has been supported by Grant DMR-1306262 from the National Science Foundation. C.R. is deeply grateful for the generosity of the Hightower Family (Georgia Institute of Technology). We thank Professor Elsa Reichmanis (School of Chemistry and Bioengineering, School of Materials Science and Engineering, and School of Chemistry and Biochemistry, Georgia Institute of Technology) and Prof. Carlos Rinaldi (Department of Chemical Engineering, University of Florida) for helpful suggestions. The authors also thank Dr. Amar Karki (Shared Instrumentation Facility, LSU) for SQUID measurements of the MGF1 particles (Fe 3 O 4 −SiO 2 −[FITCSiO2]−PSLG-1); Dr. Jeremiah Hubbard, Dr. Wayne Huberty, and Sourav Chatterjee for assistance with evaluations of magnetite cores conducted in the laboratory of Prof. Carlos Rinaldi at the University of Puerto Rico Mayagüez (currently at the University of Florida); and Dr. Dongmei Cao (Department of Chemical Engineering and Material Characterization Center), Dr. Connie M. David (Department of Chemistry), and Ying Xiao (Department of Biological Sciences), Louisiana State University, Baton Rouge, for help with HRTEM, XPS, TEM, MALDI-TOF, and SEM measurements.

■

REFERENCES

(1) Janczak, C. M.; Aspinwall, C. A. Composite nanoparticles: the best of two worlds. Anal. Bioanal. Chem. 2012, 402, 83−89. (2) Li, H. F.; Zhang, H. M.; Guo, X. L. Preparation of Micro-/NanoNonspherical Polymer Particles. Prog. Chem. 2011, 23, 1196−1210. (3) Milonjic, S. K. Colloid Particles and Advanced Materials. In Materials Science Forum; Uskokovic, D. P., Ed.; Trans Tech Publications Ltd, 1996; Vol. 214, pp 197−204. (4) Parveen, S.; Misra, R.; Sahoo, S. K. Nanoparticles: a boon to drug delivery, therapeutics, diagnostics and imaging. Nanomed. Nanotechnol. Biol. Med. 2012, 8, 147−166. (5) Zou, H.; Wu, S. S.; Shen, J. Polymer/silica nanocomposites: Preparation, characterization, properties, and applications. Chem. Rev. 2008, 108, 3893−3957. (6) Chaudhuri, R. G.; Paria, S. Core/Shell nanoparticles: classes, properties, synthesis mechanisms, characterization, and applications. Chem. Rev. 2012, 112, 2373−2433. (7) Rosu, C.; Selcuk, S.; Soto-Cantu, E.; Russo, P. Progress in silica polypeptide composite colloidal hybrids: from silica cores to fuzzy shells. Colloid Polym. Sci. 2014, 292, 1009−1040. (8) Rumenapp, C.; Gleich, B.; Haase, A. Magnetic nanoparticles in magnetic resonance imaging and diagnostics. Pharm. Res. 2012, 29, 1165−1179. (9) Balazs, A. C.; Emrick, T.; Russell, T. P. Nanoparticle polymer composites: Where two small worlds meet. Science 2006, 314, 1107− 1110. (10) Piao, Y.; Burns, A.; Kim, J.; Wiesner, U.; Hyeon, T. Designed fabrication of silica-based nanostructured particle systems for nanomedicine applications. Adv. Funct. Mater. 2008, 18, 3745−3758. (11) Block, H. Poly(γ-benzyl-L-glutamate) and Other Glutamic Acid Containing Polymers; Gordon and Breach: New York, 1983. (12) Kricheldorf, H. R. Polypeptides and 100 years of chemistry of αamino acid N-carboxyanhydrides. Angew. Chem., Int. Ed. 2006, 45, 5752−5784. (13) Deming, T. J. Polypeptide and Polypeptide Hybrid Copolymer Synthesis via NCA Polymerization. In Peptide Hybrid Polymers; Klok, H. A., Schlaad, H., Eds.; Springer: Berlin, 2006; Vol. 202, pp 1−18. (14) Huang, Y. C.; Yang, Y. S.; Lai, T. Y.; Jan, J. S. Lysine-blocktyrosine block copolypeptides: Self-assembly, cross-linking, and conjugation of targeted ligand for drug encapsulation. Polymer 2012, 53, 913−922. 7287


Article

The Journal of Physical Chemistry B (36) Marcelo, G.; Muñoz-Bonilla, A.; Fernãndez-Garcĩa, M. Magnetite−polypeptide hybrid materials decorated with gold nanoparticles: study of their catalytic activity in 4-nitrophenol reduction. J. Phys. Chem. C 2012, 116, 24717−24725. (37) Marcelo, G.; Munoz-Bonilla, A.; Rodriguez-Hernandez, J.; Fernandez-Garcia, M. Hybrid materials achieved by polypeptide grafted magnetite nanoparticles through a dopamine biomimetic surface anchored initiator. Polym. Chem. 2013, 4, 558−567. (38) Soto-Cantu, E.; Cueto, R.; Koch, J.; Russo, P. S. Synthesis and rapid characterization of amine-functionalized silica. Langmuir 2012, 28, 5562−5569. (39) Kolb, H. C.; Finn, M. G.; Sharpless, K. B. Click chemistry: diverse chemical function from a few good reactions. Angew. Chem., Int. Ed. 2001, 40, 2004−2021. (40) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B. A stepwise Huisgen cycloaddition process: Copper(I)-catalyzed regioselective “ligation” of azides and terminal alkynes. Angew. Chem., Int. Ed. 2002, 41, 2596−2599. (41) Xu, Z.; Deng, P.; Tang, S.; Li, J. Fluorescent boronic acid terminated polymer grafted silica particles synthesized via click chemistry for affinity separation of saccharides. Mater. Sci. Eng., C 2014, 40, 228−234. (42) Li, Y.; Benicewicz, B. C. Functionalization of silica nanoparticles via the combination of surface-initiated RAFT polymerization and click reactions. Macromolecules 2008, 41, 7986−7992. (43) Ranjan, R.; Brittain, W. J. Synthesis of high density polymer brushes on nanoparticles by combined RAFT polymerization and click chemistry. Macromol. Rapid Commun. 2008, 29, 1104−1110. (44) Chen, J. C.; Xiang, J. M.; Cai, Z. W.; Yong, H.; Wang, H. D.; Zhang, L. H.; Luo, W. Q.; Min, H. Synthesis of hydrophobic polymer brushes on silica nanoparticles via the combination of surface-initiated ATRP, ROP and click chemistry. J. Macromol. Sci., Pure Appl. Chem. 2010, 47, 655−662. (45) Yilmaz, G.; Kumbaraci, V.; Talinli, N.; Tatar, P.; Demirel, A. L.; Yagci, Y. Photoinduced grafting of polystyrene onto silica particles by ketene chemistry. J. Polym. Sci., Part A: Polym. Chem. 2012, 50, 2517− 2520. (46) Balamurugan, S. S.; Soto-Cantu, E.; Cueto, R.; Russo, P. S. Preparation of organosoluble silica-polypeptide particles by “Click” chemistry. Macromolecules 2010, 43, 62−70. (47) Kar, M.; Pauline, M.; Sharma, K.; Kumaraswamy, G.; Sen Gupta, S. Synthesis of poly-L-glutamic acid grafted silica nanoparticles and their assembly into macroporous structures. Langmuir 2011, 27, 12124−12133. (48) Kar, M.; Vijayakumar, P. S.; Prasad, B. L. V.; Sen Gupta, S. Synthesis and characterization of poly-L-lysine-grafted silica nanoparticles synthesized via NCA polymerization and click chemistry. Langmuir 2010, 26, 5772−5781. (49) Borase, T.; Ninjbadgar, T.; Kapetanakis, A.; Roche, S.; O’Connor, R.; Kerskens, C.; Heise, A.; Brougham, D. F. Stable aqueous dispersions of glycopeptide-grafted selectably functionalized magnetic nanoparticles. Angew. Chem.,-Int. Ed. 2013, 52, 3164−3167. (50) Massart, R. Preparation of aqueous magnetic liquids in alkaline and acidic media. IEEE Trans. Magn. 1981, 17, 1247−1248. (51) Yang, D.; Hu, J. H.; Fu, S. K. Controlled synthesis of magnetitesilica nanocomposites via a seeded sol-gel approach. J. Phys. Chem. C 2009, 113, 7646−7651. (52) Bogush, G. H.; Tracy, M. A.; Zukoski, C. F. Preparation of monodisperse silica particles - control of size and mass fraction. J. NonCryst. Solids 1988, 104, 95−106. (53) Deng, Y. H.; Wang, C. C.; Hu, J. H.; Yang, W. L.; Fu, S. K. Investigation of formation of silica-coated magnetite nanoparticles via sol−gel approach. Colloids Surf., A 2005, 262, 87−93. (54) Stober, W.; Fink, A.; Bohn, E. Controlled growth of monodisperse silica spheres in the micron size range. J. Colloid Interface Sci. 1968, 26, 62−69. (55) Zhang, J. H.; Zhan, P.; Wang, Z. L.; Zhang, W. Y.; Ming, N. B. Preparation of monodisperse silica particles with controllable size and shape. J. Mater. Res. 2003, 18, 649−653.

(56) Lee, M. H.; Beyer, F. L.; Furst, E. M. Synthesis of monodisperse fluorescent core−shell silica particles using a modified Stober method for imaging individual particles in dense colloidal suspensions. J. Colloid Interface Sci. 2005, 288, 114−123. (57) Van Blaaderen, A.; Vrij, A. Synthesis and characterization of colloidal dispersions of fluorescent, monodisperse silica spheres. Langmuir 1992, 8, 2921−2931. (58) Ranjan, R.; Brittain, W. J. Tandem RAFT polymerization and click chemistry: An efficient approach to surface modification. Macromol. Rapid Commun. 2007, 28, 2084−2089. (59) Ranjan, R.; Brittain, W. J. Combination of living radical polymerization and click chemistry for surface modification. Macromolecules 2007, 40, 6217−6223. (60) Zhao, L.; Zhao, Y. F.; Han, Y. Pore fabrication in various silicabased nanoparticles by controlled etching. Langmuir 2010, 26, 11784− 11789. (61) Daly, W. H.; Poche, D. The preparation of N-carboxyanhydrides of α-amino acids using bis(trichloromethyl)carbonate. Tetrahedron Lett. 1988, 29, 5859−5862. (62) Berne, B.; Pecora, R. Dynamic Light Scattering; Wiley: New York, 1976. (63) Jean, M.; Nachbaur, V.; Le Breton, J. M. Synthesis and characterization of magnetite powders obtained by the solvothermal method: Influence of the Fe3+ concentration. J. Alloys Compd. 2012, 513, 425−429. (64) Peternele, W. S.; Monge Fuentes, V.; Fascineli, M. L.; Rodrigues da Silva, J.; Silva, R. C.; Lucci, C. M.; Bentes de Azevedo, R. Experimental investigation of the coprecipitation method: an approach to obtain magnetite and maghemite nanoparticles with improved properties. J. Nanomater. 2014, 2014, 682985. (65) Marchegiani, G.; Imperatori, P.; Mari, A.; Pilloni, L.; Chiolerio, A.; Allia, P.; Tiberto, P.; Suber, L. Sonochemical synthesis of versatile hydrophilic magnetite nanoparticles. Ultrason. Sonochem. 2012, 19, 877−882. (66) Witten, T.; Pincus, P. A. Structured Fluids Polymers, Colloids, Surfactants; Oxford University Press: Oxford, 2004; Chapter 5. (67) Lavrentovich, O. D. Transport of particles in liquid crystals. Soft Matter 2014, 10, 1264−1283. (68) Miller, W. G. Stiff chain polymer lyotropic liquid-crystals. Annu. Rev. Phys. Chem. 1978, 29, 519−535. (69) Gharbi, M. A.; Nobili, M.; In, M.; Prevot, G.; Galatola, P.; Fournier, J.-B.; Blanc, C. Behavior of colloidal particles at a nematic liquid crystal interface. Soft Matter 2011, 7, 1467−1471. (70) Diestra-Cruz, H.; Rinaldi, C.; Acevedo, A. Rheological, optical, and thermal characterization of temperature-induced transitions in liquid crystal ferrosuspensions. J. Appl. Phys.2012, 11107B30810.1063/ 1.3672078

7288
