Quest for Thermoresponsive Block Copolymer Nanoparticles with Liquid-Crystalline Surfactant Cores.

This work reports on the preparation and characterization of anisotropic composition nanoparticles based on the electrostatic binding of dodecyltrimethylammonium surfactant to poly(acrylic acid) blocks of diblock copolymers with poly(ethylene oxide) (PEO) and poly(N-isopropyl acrylamide) (PNIPAm). These nanoparticles form kinetically stable dispersions and display liquid-crystalline cores with a micellar cubic structure, as determined by small-angle X-ray scattering. Mixtures with different proportions of the two block copolymers and stoichiometric amounts of C12TA+ were prepared and their behavior was compared with that of the parent nanoparticles. Upon heating, dilute dispersions (0.01 and 0.1 wt %) analyzed by dynamic light scattering display a slight decrease in the hydrodynamic radius, consistent with the dehydration of PNIPAm and mixed PNIPAm-PEO blocks at the shell. At higher concentrations, 2 wt %, the nanoparticles with pure PNIPAm shell undergo macroscopic phase separation above 32 °C. Nanoparticles with a pure PEO shell do not display temperature sensitivity. For the mixtures, no visual change is observed, but the dynamic light scattering results evidence the formation of clusters, whose size and reversibility depend on the PEO/PNIPAm proportion. This indicates the formation of mixed nanoparticles containing both PEO and PNIPAm blocks. Nuclear Overhauser enhancement spectroscopy NMR analyses of the mixtures do not show the correlation peak expected for PEO and PNIPAm blocks in close proximity, suggesting their segregation at the nanoparticle shell. On the basis of these results, we discuss the possibilities of the neutral blocks distribution on the shell of mixed nanoparticles. Overall, we have confirmed that these nanoparticles may display a temperature-controlled reversible aggregation while preserving their internal liquid-crystalline structures.

Introduction

Preparation and design of nanoparticles for applications as building blocks of highly ordered structures have attracted much attention in the last years, especially due to the possibility to design and control self-assembled systems whose structures are sensitive to external stimuli[1] such as temperature,[2,3] light,[2] pH,[4,5] redox potential, among others.

An additional level of organization can be achieved by systems with attractive domains on their surface such as Patchy particles or with a strictly biphasic geometry such as Janus nanoparticles.[6,7] These systems present directional interactions and offer an additional level of organization because they can self-assemble to form diverse supramolecular structures, creating complex three-dimensional architectures.[8] In addition, this anisotropic characteristic leads to novel morphologies and diverse potential nanotechnological applications for cosmetics, drug delivery,[9,10] or surface coating.[11,12]

Patchy and Janus nanoparticles have been explored by some theoretical studies and described in detail by the experimental studies involving different polymers. Walther et al.,[13] for example, prepared mixed aggregates with block terpolymers with two outer hydrophilic blocks of poly(ethylene oxide) (PEO) and the thermoresponsive poly(N-isopropyl acrylamide) (PNIPAm). Higher temperatures triggered the collapse of PNIPAm chains and anisotropic composition Patchy and Janus superstructures were observed when the contour length of the thermoresponsive block was longer than that of the PEO chains.

PNIPAm is one of the best-known temperature responsive polymers. Although most synthetic water-soluble polymers become more soluble upon heating, PNIPAm, among a few others, phase separates from water because of its lower critical solution temperature (LCST). When the temperature is lower than its LCST, PNIPAm is hydrophilic and soluble through hydrogen bonding with water molecules, but it becomes dehydrated and precipitates from water upon heating, displaying a sharp coil to globule transition around 32 °C.[14,15]

Another study from Gröschel et al.[16] reported the conversion of multicompartment micelles into Janus nanoparticles based on the self-assembly of ABC triblock terpolymers of polystyrene-b-polybutadiene-b-poly(methyl methacrylate), cross-linking of polybutadiene, and solvent selectivity. They also showed that the Janus balance could be tuned by varying the block lengths. Compared with triblock copolymers, the structures of complex aggregates formed by two different diblock copolymers are conveniently tunable because the ratio of the mixed shell can be adjusted by the relative content of the two diblock copolymers.[9] Then different ratios are achieved by the addition of different amounts of the diblocks instead of a synthetic step to increase the block length.

A similar approach was followed by Voets et al.[17] while studying the structure and behavior in a solution of complex coacervate core micelles and their temperature dependence. Through the coacervation of mixtures of oppositely charged diblock copolymers poly(N-methyl-2-vinyl pyridinium iodide)-b-poly(ethylene oxide), P2MVP38-b-PEO211, and poly(acrylic acid)-b-poly(N-isopropyl acrylamide), PAA55-b-PNIPAAm88, a core–shell structure was obtained. They proposed that the ionic blocks would form the cores by electrostatic interaction and that the neutral PEO and PNIPAm blocks would form their shells. They found that PEO and PNIPAm chains appear to be randomly mixed within the micellar shell at a low temperature but change to a segregated core–shell-corona arrangement above 32 °C. Along the same line, an earlier study from our group has shown that Janus gold nanoparticles covered by thiolated PEO and PNIPAm were obtained because of the spontaneous segregation of the polymers, even at room temperature.[18]

Association between polyelectrolytes and oppositely charged surfactant is based on both electrostatic and hydrophobic interactions. Svensson et al.[19] performed the first studies of complex salts of poly(acrylic acid) with cationic quaternary ammonium surfactants, observing different bulk liquid–crystalline phases depending on the chain length of the surfactant used and on the composition of the system. These complex salts are prepared at fixed stoichiometry with a complete charge neutralization.[20]

With the introduction of neutral polymeric blocks in the complex salts, also called neutral block complex salts, a core–shell aggregate was obtained by Vitorazi et al.[21] using poly(acrylamide)-b-poly(acrylic acid) copolymers. Their self-assembly produced the dispersions of nanoparticles with a liquid–crystalline core, either micellar cubic or hexagonal, depending on whether dodecyltrimethylammonium or hexadecyltrimethylammonium surfactants were used, that was exactly the same obtained for complex salts formed by these surfactants and poly(acrylic acid). More recently, we reported that the internal structure of neutral block complex salts could be tuned by the controlled addition of n-alcohols to produce a wide variety of mesophases such as micellar cubic, direct, and inverse hexagonal, lamellar, and inverted micelles.[22]

In the present study, we aim at preparing thermoresponsive nanoparticles with a liquid–crystalline core from the mixtures of neutral block copolymer salts containing PEO and PNIPAm blocks. The effect of temperature on the liquid–crystalline core structure and the aggregation of the nanoparticles is investigated. From the characterization of the nanoparticles, we discuss whether there is segregation among PEO and PNIPAm blocks and, if this happens, how their chains are distributed in the shell.

Experimental Section

Chemicals

Poly(N-isopropyl acrylamide-b-acrylic acid), PNIPAm212-b-PAA138, with nominal molar mass of 10 000 (PAA) and 24 000 (PNIPAm), with Mw/Mn = 1.33, and poly(ethylene oxide-b-acrylic acid), PEO136-b-PAA72, with nominal molar mass of 5200 (PAA) and 6000 methyl terminal (PEO), with Mw/Mn = 1.15, were purchased from Polymer Source, Inc. (Dorval, Quebec, Canada). Their gel permeation chromatography curves, as provided by the manufacturer, display only one elution peak and were used as received. The surfactant C12TABr was purchased from Sigma-Aldrich with 99% purity and used with no further purification in an ion-exchange step to obtain the hydroxide form C12TAOH. The water used throughout was of MilliQ grade. The structures of the polymers used in this work are presented in Figure .

Figure 1

Chemical structures of the diblock copolymers used in this study.

Preparation of Neutral Block Complex Salts

Neutral block complex salts of PNIPAm and complex salts of PEO were prepared by the titration of the hydroxide form of the cationic surfactant solution in water, with the acid form of the polyion to the charge neutralization point at pH 8.5, following the methodology proposed by Svensson et al.[19] For the mixtures, the surfactant was titrated over an aqueous solution of PEO-b-PAA and PNIPAm-b-PAA in the desired proportion of (PNIPAm:PEO wt %). Solutions were freeze-dried and the white powders obtained were kept in a desiccator before dispersion in water at the desired concentration. The resulting complex salts are identified in terms of their copolymer composition, as explained in Table .

Table 1

Nomenclature for Complex Salts Used in This Study, Prepared with Dodecyltrimethylammonium Surfactant and Different Polymers

acronym	complex salt description
C12E100	containing only the copolymer PEO136-b-PAA72
C12N100	containing only the copolymer PNIPAm212-b-PAA138
C12NxEy	containing mixtures of both copolymers, where x refers to PNIPAm and y refers to PEO weight percentage considering the sum of PNIPAm and PEO content in each mixture

Preparation of Dispersions

Complex salts and water were weighed at the desired composition (typically 2 wt %) and dispersed for about 60 s by a Vortex-Genie 2 mixer (Scientific Industries) operating at 3200 rpm. After that, they were left to equilibrate for at least 1 day prior to analyses. The samples with lower concentrations were prepared from the dilution of the previous ones. The pH of these complex salt dispersions remained around the equivalence point, as checked by direct pH measurements.

Samples Characterization

Dynamic Light Scattering (DLS)

DLS measurements were performed on samples using a scattering angle θ = 90° on a CGS-3-based compact goniometer system (ALV-GmbH, Langen, Germany), which was equipped with a detection system in a pseudo–cross-geometry, with a 22 mW He–Ne laser (λ = 632.8 nm) and an ALV 7004 multi-tau correlator. cis-Decalin was used as the refractive index-matching liquid. The temperature was controlled at 25.00 or 45.00 ± 0.01 °C. The samples with 0.01 and 0.1 wt % concentration were placed in capped test tubes previously washed with Hellmanex 3%.

A Malvern Nano Zetasizer instrument at 25 and 45 °C, with a 632.8 nm laser and a detector positioned at 173°, was used for analyzing the samples with a concentration of 2 wt % of neutral block complex salt. For these samples, due to their high turbidity, the NIBS (noninvasive backscatter optics)[23] was used to attenuate multiple-scattering effects related to a high concentration. The obtained intensity time-correlation functions were analyzed by the cumulant method, which provided the Z-average radius.[24,25]

ζ Potential

The measurements of 0.01 wt % samples were performed with a Malvern Nano Zetasizer instrument at an angle of 12.8°. The ζ potential was calculated from the measured electrophoretic mobility using the Smoluchowski factor, f = 1.5 and the Henry equation: Ue = εζf/6πη, where Ue is the electrophoretic mobility, ε is the dielectric constant of water, η is the viscosity of water, and ζ is the zeta potential.

Small-Angle X-ray Scattering (SAXS)

The SAXS measurements were performed using the D11A-SAXS1 beamline from the Brazilian Synchrotron Light Laboratory (LNLS, CNPEM, Campinas, Brazil). Samples of 2 wt % concentration were injected into a sample holder closed by two mica windows, except for the solid samples, which were sealed in a Kapton sandwich. The radiation wavelength was λ = 1.550 Å and the scattering patterns were recorded under vacuum. The scattered intensity as a function of the scattering vector, I(q), was obtained for a q range of 0.01–4 nm–1. FIT2D software was used to correct the intensities for the detector response and the dark current signal to subtract the solvent scattering and to convert the two-dimensional (2D) SAXS images to one-dimensional distributions of scattered intensity.

1H NOESY NMR

The experiment of the nanoparticles of 1 wt % of C12N57E43 in D2O were recorded at 25 °C on a Bruker Avance-500 spectrometer, operating at 500 MHz and at 45 °C on a Bruker Avance-400 spectrometer, operating at 400 MHz. NOESY spectra were acquired with 2048 × 256 data points recorded at 25 and 45 °C, using the standard Bruker software with mixing time of 400 and 800 ms.

High-Sensitivity Differential Scanning Calorimetry

The DSC measurements were performed with 1 wt % aqueous dispersions of the neutral block complex salts using a VP-DSC (MicroCal, Northampton, MA) calorimeter in the range of 20–60 °C at the scanning rate of 1.0 °C/min. The samples were thermally equilibrated for 10 min before each scan. Each experiment was conducted in triplicate with good agreement among all of the runs. The baseline reference (water vs water) was subtracted from the sample thermograms. For instrument control, acquisition, and data analyses, MicroCal Origin 5.0 and OriginPro 8 softwares were used.

UV–Visible (UV–Vis) Spectroscopy

The spectroscopy was used to determine the turbidity of neutral block copolymer salt dispersions as a function of temperature. The measurements were performed using an HP 8453 UV–vis spectrophotometer at the wavelength of 410 nm and the UV–vis ChemStation software. Samples of 0.1 wt % were analyzed at a heating rate of 1.5 °C/min by an HP 89090A temperature controller in the range of 20–50 °C.

Cryogenic Transmission Electron Microscopy

The cryo-TEM samples were prepared utilizing a controlled environment vitrification system (CEVS). The chamber was kept at the target temperature saturated with water vapor to prevent any water evaporation from the sample. A small droplet (3.2 μL/0.5 wt %) of the sample was applied on a Lacey Carbon 300 Mesh grid both kept at either 25 or 45 °C. After 20 s, the sample was blotted and plunged into liquid ethane at its freezing point. The vitreous specimen was kept under liquid nitrogen until loaded into a cryogenic sample holder. Imaging was performed at the Brazilian Nanotechnology National Laboratory (LNNano) with a FEI Talos F200C microscope operating at 200 kV equipped with a Ceta 16 Mpixel camera.

Results

Core Structure—SAXS Measurements

First, we investigated the internal structures of the aggregates via SAXS analyses. The SAXS curves show information related to the internal structure of the aggregates as shown in Figure . The relative positions of the diffraction peaks (21/2, 41/2, 51/2, 61/2, and 81/2) are associated with the cubic phase Pm3n present in all of the systems.

Figure 2

SAXS curves of 2 wt % dispersions of C12E100, C12N57E43, C12N75E25, C12N92E8, and C12N100 at 25 and 45 °C (respective wine curves).

The additional peaks appear at 1.74 and 2.28 nm–1 (relative positions 31/2 and 51/2) for samples containing PNIPAm-b-PAA copolymer at room temperature. Interestingly, they disappear when the samples are heated to 45 °C. We could not unequivocally ascribe those peaks to any of the common structures reported for complex salt mesophases. Increasing the temperature shifts the SAXS curves to slightly higher q values and favors the appearance of only cubic Pm3n liquid–crystalline phase. Overall, temperature has no major effect on the structure of Pm3n mesophase, with all of the systems displaying a micellar cubic phase (Figure ), essentially the same reported for the bulk mixture of the same surfactant, polyacrylate, and water.[26]

The cubic cell parameter a was calculated from the angular coefficient of plot of (q/2π)2 versus Miller index of Pm3n phase at 25 and 45 °C, as described in eq . The results are reported in Table .where q is the scattering vector and h, k, l are the Miller indexes of Pm3n phase.

Table 2

Cubic Lattice Parameter, Hydrodynamic Radii, and ζ Potential Values for the Core–Shell Aggregates of Neutral Block Complex Salts Dispersions

			RH (nm) ± SD		RH (nm) ± SD		RH (nm) ± SD			ζ (mV) ± SD
	a (nm)		C = 0.01 wt %a		C = 0.1 wt %a		C = 2 wt %b			C = 0.01 wt %
neutral block complex salt	25 °C	45 °C	25 °C	45 °C	25 °C	45 °C	25 °C	45 °C	25 °C–after heatingc	25 °C
C12E100	8.3	8.2	140 ± 20	130 ± 20	170 ± 8	162 ± 3	278 ± 7	270 ± 20	270 ± 10	–31 ± 4
C12N57E43	8.3	8.2	193 ± 3	160 ± 10	152 ± 6	126 ± 7	274 ± 9	162 ± 1	240 ± 10	–23 ± 4
C12N75E25	8.2	8.2	212 ± 5	162 ± 4	-e	-e	500 ± 20	700 ± 80	450 ± 30	–28 ± 1
C12N92E8	8.3	8.2	250 ± 10	150 ± 40	-e	-e	398 ± 5	1400 ± 300	1100 ± 200	–27 ± 1
C12N100	8.3	8.3	185 ± 4	135 ± 3	176 ± 9	90 ± 10	400 ± 30	N/Dd	N/Dd	–25 ± 8

Denotes that the DLS measurements were performed on the samples using a scattering angle θ = 90° on a CGS-3-based compact goniometer systems from ALV-GmbH and analyzed using the constrained regularization REPES algorithm.

Denotes the DLS measurements performed on the samples using a scattering angle θ = 173° on a Malvern Nanozetasizer and analyzed using the cumulant method and NIBS because of samples high turbidity.

Denotes the samples equilibrated at room temperature for at least 24 h.

N/D (not determined). At 2 wt %, these measurements were not performed due to macroscopic phase separation.

- Denotes the samples not measured.

Visual Observation of Temperature Effect

Samples of all of the complex salts at 2 wt % are turbid and white at room temperature (Figure a). The vials shown in Figure b were equilibrated for 10 min at 45 °C and then immediately inverted, revealing that only the sample of C12N100 macroscopic phase separates to form a viscous gel that sticks to the glass vial after increasing the temperature to above the LCST of PNIPAm. With increasing temperature, the dehydration of PNIPAm starts to occur, and the nanoparticles form clusters to avoid contact with water. No visual change was observed for C12E100 or C12NE samples, indicating that the presence of PEO chains stabilizes the formed structures against the effects of PNIPAm dehydration upon heating. Further tests increasing the equilibration time at 45 °C were performed up to 24 h and the results are the same shown for 10 min of equilibration.

Figure 3

Macroscopic effect of temperature increase from (a) 25 °C to (b) 45 °C on nanoparticles at 2 wt % of (1) C12E100, (2) C12N57E43, (3) C12N75E25, (4) C12N92E8, and (5) C12N100. In (b), the vials were inverted after 10 min of equilibration at 45 °C to reveal that the macroscopic phase-separated sample sticks to the glass vial (red arrow).

We should also mention that for 0.01 and 0.1 wt % and at 45 °C, no sign of macroscopic phase separation could be identified in any sample, even for C12N100 nanoparticles; but this may be difficult to visualize due to the low concentration of complex salt present.

DLS Measurements of Diluted Samples

To further investigate the nanoparticle response to temperature changes, we performed the DLS measurements as a function of temperature and at different concentrations. For the samples of concentrations between 0.01 and 0.1 wt %, the intensity of scattered light is not affected by multiple scattering, we observed a unimodal relaxation time distribution (Figures S1 and S2) and a nondiffusive regime represented by a nonlinear interdependence between Γ and q2 (Figure S4).

The calculated nanoparticle sizes are listed as apparent hydrodynamic radii, RH, in Table for the two concentrations and at temperatures below and above LCST. For all of the samples with concentrations of 0.01 and 0.1 wt %, the size is close to that reported for similar nanoparticles prepared with block copolymer complex salts containing poly(acrylamide) as the neutral block,[21] even with cores with geometry of different mesophases.[22] Data from the samples at 0.01 wt % in Table reveal that the size of nanoparticles increases as the amount of PNIPAm is increased except for the nanoparticles of C12N100, whose size is similar to that of C12N57E43. Moreover, at 0.01 and 0.1 wt %, there is no sign of nanoparticle aggregation or clusters when the temperature is raised above LCST of PNIPAm. On the contrary, there is a slight decrease in RH values, except for the samples of C12E100, which is consistent with the dehydration of nanoparticle shells containing PNIPAm.

Another factor that should be taken to account is that these nanoparticles bear a negative surface charge (Table ) that may be responsible for hampering their aggregation. For this reason, the DLS measurements were also performed with these diluted samples in the presence of varying amounts of NaCl (Table S1), but the values for the hydrodynamic radii remained essentially the same, even upon heating. Addition of NaCl slightly affects the nanoparticles’ internal structure, increasing the cubic cell parameters, possibly due to partial screening of the electrostatic attraction and, at much higher concentrations (above 0.1 mol L–1), destroys the internal liquid–crystalline structure.[21,27,28]

DSC and Turbidity Measurements

High-sensitivity DSC (Figure ) measurements of dispersions at 1 wt % revealed a clear endothermic peak around 32 °C, related to the PNIPAm dehydration for the nanoparticles containing only PNIPAm copolymer and their mixtures with PEO, confirming that PNIPAm chains undergo the expected transition.

Figure 4

DSC curves for C12N100 and C12N57E43 at 1 wt % analyzed in three consecutive runs, indicating PNIPAm dehydration upon heating (endothermic peak).

Sample turbidity was also followed as a function of temperature for dispersions at 0.01 wt % (Figure ). For the samples of C12N100, a clear turbidity increase is observed. On the other hand, no change in turbidity is observed for C12E100 nanoparticles. For their mixtures, a small turbidity increase is observed starting at 32 °C, consistent with the dehydration of PNIPAm chains detected by DSC. Because no aggregation of nanoparticles was observed at this range of concentration, this increase in turbidity is most likely associated with an increase in the scattered light due to changes in the refractive index as a result of PNIPAm dehydration, as already reported.[29]

Figure 5

Turbidity of C12N100 and C12NE at 0.1 wt % as a function of temperature.

DLS Measurements on Concentrated Samples (2 wt %)

To clarify the possible influence of concentration on the nanoparticles behavior, light-scattering measurements were performed on 2 wt % samples. At this concentration and low temperature (25 °C), the samples of C12E100 and C12N57E43 display similar nanoparticle sizes (Table ). As the fraction of PNIPAm increases, there is an increase in the nanoparticle size that might be related to the presence of more (and longer) PNIPAm chains, which additionally turn nanoparticles less hydrophilic.

At a higher temperature (45 °C), C12E100 nanoparticles remain at the same size, C12N57E43 nanoparticles shrink slightly, probably due to the collapse of PNIPAm chains, as observed for the dilute dispersions above, and nanoparticles with higher PNIPAm contents present a significant increase in the size, which must be associated with their temperature-induced aggregation. Interestingly, after cooling, most systems recover their initial sizes, except for C12N92E8 and C12N100, even after 24 h. These systems with large contents of PNIPAm display only partial reversibility of their aggregates, possibly due to entanglements of the larger PNIPAm blocks. For sample C12N75E25, the content of hydrophilic PEO blocks is sufficient to obstruct this effect and make the aggregation fully reversible.

SAXS Results at Low q Range

To understand the effects of the nanoparticle shell composition and the temperature on aggregation, we explored the SAXS data at a low q ranging from 0.12 to 0.7 nm–1. For that, the SAXS curves are presented in logarithmic scales, as shown in Figure a,b.

Figure 6

SAXS curves presented in logarithmic scales of aqueous samples C12E100, C12N100, and C12NE of 2 wt % at (a) 25 °C and (b) 45 °C.

C12E100 and C12N100 nanoparticles display distinct profiles at a low q range, though for the mixtures there is a gradual change from one limit to the other. For C12E100, a damped oscillation corresponding to the nanoparticle form factor is observed. For complex salts with increasing PNIPAm content, this oscillation gradually disappears. Berret et al.[30] have investigated similar systems of surfactants and block copolymers by small-angle neutral scattering and applied a standard Monte-Carlo algorithm to simulate a “cage” model of micelles for the nanoparticle core. For dilute samples, they observed the damped oscillation, which gradually disappears with increase in the concentration of nanoparticles. This behavior was justified by the occurrence of an interparticle structure factor, which culminates in the appearance of an interference peak around 0.09 nm–1 for very concentrated samples (>15 wt %). The observable range of q in Figure does not allow us to verify the appearance of an interference peak. However, the observed trend indicates that increasing the content of PNIPAm increases the structure factor between the nanoparticles, in the same way as observed by Berret et al. with increasing sample concentration. It is an indication that the higher the PNIPAm content, the higher the nanoparticle tendency to aggregation, in accordance with the DLS results for the samples with the same concentration (2 wt %).

Except for C12E100, the samples of all of the complex salts present a different scattering profile at a low q range upon heating to 45 °C (Figure b). The change must be due to both the shell contraction and the aggregation of the nanoparticles, evidenced by DLS for C12N100 and C12NE nanoparticles at temperatures higher than LCST of PNIPAm.

Colloidal Stability

Regarding the colloidal stability of these systems, the complex salts reported in this study form aqueous dispersions, which are observed to phase separate in a few weeks time for the samples prepared at 2 wt %. It is interesting to notice that their size does not seem to vary significantly with the preparation method, as also reported earlier for another neutral block complex salt[21] and that they always display a reproducible liquid–crystalline core. Nanoparticles present the ζ potential values around −30 mV (Table ) for all of the samples at 25 and 45 °C, the charge being an important factor for kinetic stability.[25] The origin of the negative charge has been discussed before[22] and is related to the surfactant dissociation into the aqueous continuous phase. In addition, our preliminary observations indicate that the dispersions of C12E100 complex salts are more stable than the ones containing both PEO and PNIPAm blocks (C12NE), as less phase separation is observed compared with the mixtures as a function of time.

Segregation of PEO and PNIPAm Chains—1H NOESY NMR Results

Another issue investigated is whether there is a segregation between the different polymer blocks, PEO, and PNIPAm. To clarify this question, we performed the NMR NOESY experiments that have been already applied to mixtures of these two polymer blocks.[31,32] This technique is useful to determine whether specific protons not covalently bonded are in close proximity (d < 0.5 nm). 1H NOESY NMR measurements were performed for C12N57E43 at 25 and 45 °C. If the PEO and PNIPAm chains are randomly distributed, cross-correlation peaks are expected to appear in the positions indicated by the red circles of Figure a,b at 3.64 ppm from PEO and at 1.09 ppm from PNIPAm. The obtained NMR spectra do not contain these correlation peaks as an indication of no close proximity between the different polymer chains.

Figure 7

Two-dimensional 1H–1H NOESY NMR spectra of C12N57E43 dispersed in D2O at (a) 25 °C and (b) 45 °C. Red circles indicate positions where the cross peaks corresponding to the chemical shifts of the most intense peak for each polymer (PEO and PNIPAm) would be expected, in case of chain proximity. Individual 1H NMR are shown in Figure S8.

At 45 °C, a downfield shift in the peak position is observed; besides, peaks attributed to PNIPAm are broadened due to low diffusion and formation of bigger objects; as a consequence, they are more difficult to attribute. Other than that, the absence of correlation peaks is the same observed at lower temperature.

Surface Tension Measurements

If the segregation of PEO and PNIPAm blocks leads to the anisotropic composition (Patchy or Janus) nanoparticles, one may expect that the nanoparticles display surface activity, as reported for other systems.[33] For this reason, we performed surface tension measurements using 0.25 wt % dispersions of nanoparticles with pure copolymers and their mixtures. The results (shown as Table S2) reveal low surface tension values, around 32 mN/m even for dispersion of C12N100 or C12E100, probably due to the dissociation of surfactant ions from the nanoparticle core, already mentioned to explain the negative value of ζ potential for these nanoparticles. As such, these results are not conclusive and may only indicate that the nanoparticles are not more surface active than the dissociated surfactant ions.

Cryo-TEM Analyses

To further investigate these nanoparticles, we also performed some cryo-TEM analyses of C12N75E25 nanoparticles from 25 to 45 °C. The resulting images are shown in Figure . These images do not allow enough resolution to observe the liquid–crystalline structure of the nanoparticles core, but they allow visualization of roughly spherical nanoparticles displaying sizes ranging from 20 to 120 nm. Overall, these observations agree with the DLS results listed in Table at 0.01 wt %, considering that RH values are expected to be larger. Particle size estimated from cryo-TEM images is expected to be smaller than the one derived from the DLS data because the latter contains contributions from the particles’ hydration shell and from scattering of larger particles. The present DLS data are consistent with a polydisperse system, but not bimodal, as can be seen in Figures S5 and S6). The images at 45 °C suggest slightly smaller nanoparticles forming bigger aggregates. Although PEO and PNIPAm show very low contrast in cryo-TEM, from Figure d we can confirm that PNIPAm transition should control the final assembly.

Figure 8

Cryo-TEM images of sample C12N75E25 dispersed at 0.5 wt % in H2O at (a, b) 25 °C and (c, d) 45 °C.

Discussion

Internal Core Structure

For the discussion on the properties of these core–shell nanoparticles, we will start with their internal liquid–crystalline structure. In our first study with this type of system,[21] we reported that the internal order was only achieved when titrating the surfactant to the opposite-charge polymer solution (complex salts), as opposed to directly mixing solutions of surfactants and polymers, even if at the exact composition to produce a charge neutrality.[34] The structure of the nanoparticles core reproduces exactly the one that was obtained with complex salts prepared with the cationic surfactant and polyacrylate.[26]

In the present study, we observed that the same cubic micellar Pm3n structure is also obtained when the block copolymer complex salt is formed by PEO or PNIPAm neutral blocks, suggesting an insensitivity of this ordering to either the presence or the nature of the hydrophilic block. Two other peaks from a nonidentified phase are present for the samples containing PNIPAm blocks at room temperature. Some of these peaks according to the sequence 31/2 and 51/2 are expected for an hcp mesophase, but other peaks are missing. The hcp mesophase was reported to appear for complex salts in systems based on dodecyltrymethyammonium surfactant and polyacrylate[35] and also ethoxylated complex salts of hexyltrymethylammonium surfactant.[36] In both cases, the hcp appearance was related to either shorter surfactant alkyl chains or lower charge density of copolymer. However, with the present results, we cannot unequivocally ascribe those peaks to an hcp mesophase.

If one assumes that these nanoparticles are close to spherical and, according to the DLS results, with radius over 100 nm, the core must contain a significant fraction of hydrophilic blocks. It is most likely that they remain hydrated, hence at the shell for the surfactant aggregates. In this respect, it seems more surprising that for PNIPAm-containing complex salts, the Pm3n internal structure remains unaltered at 45 °C, except for the peaks attributed to the nonidentified phase, which disappears above the LCST of PNIPAm. At this temperature, PNIPAm blocks are dehydrated, as confirmed by the DSC results, and are expected to be more incorporated into the hydrophobic interior of micelles.

Earlier studies with the incorporation of hydrophobic solutes into complex salt aggregates[37] indicate that compounds such as p-xylene or cyclohexane are incorporated into the surfactant micelles (in that case, cylinders of a hexagonal phase), causing an increase in their radius due to swelling. Our present results indicate that the center-to-center distance between adjacent micelles, defined by the cubic cell parameter, remains constant. This finding is not inconsistent with a swelling of the micelles due to the incorporation of PNIPAm blocks, only indicating that the balance between attraction–repulsion forces among micelles remains the same.

Nanoparticles Composition and Aggregation

With respect to nanoparticles composition, one could envisage two limit cases for the complex salt of a mixture of two different block copolymers: (a) the formation of pure individual nanoparticles or (b) the formation of fully mixed ones, as sketched in Figure .

Figure 9

Schematic representation of two limit possibilities for nanoparticles of C12NE. (a) Due to the segregation of PEO and PNIPAm blocks, there is the formation of two different populations of nanoparticles. (b) The different block copolymers are completely randomly distributed in all of the nanoparticles formed.

The best way to assess this issue is by analyzing the effects of temperature on these mixed nanoparticles, considering that the ones with only PEO are not affected by temperature, whereas ones with only PNIPAm are. The DSC results confirm that, even in the mixture, PNIPAm blocks dehydrate around 32 °C (see Figure ). This is also supported by the turbidity measurements that are consistent with an increase in the refractive index expected as a consequence of PNIPAm dehydration.

Results from visual observation and DLS and SAXS measurements as a function of temperature indicate that the mixed nanoparticles always display an intermediary behavior between those of the pure nanoparticles, sometimes with a trend of convergence as the PNIPAm content increases. Moreover, the relaxation time distributions with two populations of different sizes are not present for the samples of C12NE. If here was one population related only to C12E100 nanoparticles, it would not vary in size after temperature increase, whereas a second population related to C12N100 nanoparticles would change the size above 32 °C. These are strong evidences in favor of the formation of mixed nanoparticles, and that their composition follows that of the global mixture.

DLS measurements as a function of temperature do not reveal any increase in size induced by heating for the samples at low concentrations (up to 0.1 wt %), indicating that no aggregation occurs. In fact, the hydrodynamic radii display a slight decrease upon heating, which could be ascribed to dehydration and collapse of the PNIPAm chains. Due to the different molar masses of PNIPAm (M = 24000 g mol–1) and PEO (M = 5200 g mol–1), it seems reasonable to have C12E100 nanoparticles slightly smaller in size than PNIPAm ones, and that their collapse leads to an overall decrease in the nanoparticle size. It is not clear, at this stage, why pure PNIPAm nanoparticles at low concentration do not aggregate upon heating, even with the addition of NaCl to screen electrostatic repulsion. However, DSC results confirm dehydration of their PNIPAm blocks at 32 °C. The only possible explanation would be the reduced nanoparticle encounter due to their very low concentration. In an earlier study with thermosensitive brush macromolecules, Lee and co-workers[38] reported that aggregation above the LCST of copolymer was only observed for concentrated solutions (5 wt % in their case), whereas a decrease in size occurred with the increase in temperature for dilute systems (below 0.5 wt %), similar to what we observed. In all of the cases, the nanoparticles’ hydrodynamic radii are similar to those found for acrylamide–acrylate complex salt systems.[21]

At higher concentration (2 wt %), both DLS and low q SAXS results confirm the aggregation of C12N75E25, C12N92E8, and C12N100 nanoparticles, and this becomes more prominent above LCST of PNIPAm, confirming the contribution of the PNIPAm blocks to the thermosensitivity of the mixed nanoparticles. Although PNIPAm dehydration is reversible[14] according to DSC results at 1 wt %, the nanoparticle aggregation could be only fully reversed for C12N75E25, returning immediately to their original size at 25 °C with the decrease in temperature. As for C12N92E8 and C12N100 nanoparticles, only partial reversibility was observed within 24 h, possibly indicating a significant entanglement of the longer PNIPAm blocks. In any composition, their internal Pm3n cubic micellar structure remained unchanged. However, the nonidentified phase, present at 25 °C in all of the samples of C12NE and C12N100, disappears at 45 °C.

With respect to the arrangement of the PEO and PNIPAM blocks, NOESY NMR results indicate that they are segregated, even at low temperature. Following that, one could envisage a series of possibilities for the neutral blocks distribution at the nanoparticles shell, as we try to represent in Figure .

Figure 10

Schematic representation of the models proposed for the segregation of PEO (blue) and PNIPAm (green) blocks at the nanoparticle shell.

The following models from (a) to (c) should represent a progressive change on nanoparticles as the PNIPAm content is increased. In the situation represented by Figure a, the relative PEO:PNIPAm ratio is high, allowing the formation of a PEO-dominated shell. PNIPAm chains would be concentrated in or at the interface of the surfactant micelles, making the interior richer in PNIPAm, especially above their LCST. Micelles located closer to the nanoparticle core should still contain PEO blocks that could not physically reach the interface, although one cannot rule out the possibility of a copolymer preferential migration that would enrich the nanoparticle core with PNIPAm chains and the shell with PEO chains while keeping the balance between surfactant and acrylate groups. This condition is similar to that reported by Ma et al.[32] of the formation of what they described as a core–shell-corona arrangement. However, having the entire shell or most of its surface rich in PEO chains, the nanoparticles are not expected to form aggregates. Following this idea, the sample C12N57E43 is probably somewhere between a PEO-dominated shell and a patchy nanoparticle model because no aggregation is observed for 2 wt % samples, even at 45 °C.

Patchy nanoparticles are formed when domains appear at the surface. Such domains can either be randomly distributed or distributed over specific directions at the shell, resulting in anisotropy. As PNIPAm is less hydrophilic than PEO, we would expect PNIPAm to form domains in the continuous PEO surface as more PNIPAm is added to the system. Depending on the hydrophobic/hydrophilic balance, the nanoparticles can form (b), and this situation is most probable to happen when nanoparticles are anisotropic in composition. The last case (c) represents the surface segregation in which PEO and PNIPAm chains at the shell arrange themselves to produce two different domains, producing Janus particles.[39,40]

Then samples C12N75E25 and C12N92E8 would be classified between Patchy (b) and Janus nanoparticles (c). These situations would agree with the segregation between PNIPAm and PEO chains indicated by NOESY NMR results, although we cannot ascribe unequivocally how these domains arrange themselves.

The striking difference is that the phase separation was fully reversible upon cooling the dispersions of C12N75E25, whereas those of C12N92E8 and C12N100 displayed only partial immediate reversibility, suggesting kinetic effects due to nanoparticle aggregation.

To ensure thermosensitivity, an excess of PNIPAm chains is required, but the nanoparticles aggregate to produce macroscopic phase separation when the chains are too many (as for C12N92E8 and C12N100), although this process is not fully reversible. Because aggregation is observed at 2 wt % and 45 °C, PNIPAm should be present at the surface and lead to the formation of bigger clusters with a temperature increase due to increase in its hydrophobicity.

Some reports show that PEO and PNIPam chains can phase segregate even below the LCST of PNIPAm,[18,41,42] so the formation of anisotropic composition surfaces, either Patchy or Janus, can be expected at room temperature. Above the LCST, the hydrophobic domains of PNIPAm can lead to nanoparticle aggregation due to the different nature of the polymer blocks.

Furthermore, the lack of correlation peaks in NOESY NMR was the main evidence invoked by Voets and co-workers[43] to propose the formation of Janus particles in a system composed by PEO and PAAm as the hydrophilic polymer pair in aqueous solutions. However, in another study by the same group, with coacervates of mixed PEO and PNIPAm polymers,[30] NOESY NMR spectra showed the presence of correlation peaks. They observed temperature-induced aggregation and proposed PNIPAm would be located in the core surrounded by a PEO brush layer, similar to the situation in Figure a.

We could not find means, by using all of the experimental information we report here, to ascribe unequivocally the distribution of PEO and PNIPAm blocks at the shell of the mixed nanoparticles described here. They are most likely segregated at room temperature in a Patchy-like configuration, although Janus nanoparticles model cannot be discarded.

Regardless of the clear discrimination on how these chains arrange themselves at the nanoparticle shell, this approach is capable of producing kinetically stable nanoparticles with liquid–crystalline cores that display reversible temperature-induced phase separation. Therefore, for the purpose of controlled aggregation and macroscopic phase separation as means of facilitated nanoparticle removal (for instance, for recycling), full reversibility is desirable and an intermediate composition (in the present study, around 70 wt % of PNIPAm) is suggested. Consequently, the combination of an ionic surfactant with two diblock copolymers (only one of them being thermoresponsive) in the correct ratio is crucial to achieve a fully reversible aggregation induced by heating.

Conclusions

This study is the first one using mixtures of neutral diblocks on the preparation of complex salts to produce reversible thermosensitive nanoparticles with liquid–crystalline cores. They were successfully prepared and characterized, displaying suitable kinetic stability and core structure, which agrees with findings of earlier reports on similar systems. The use of two copolymers with different neutral blocks at an optimal ratio, only one being thermosensitive (PNIPAm), allowed successful control over their temperature-induced aggregation and phase separation.

Several techniques have been employed for the characterization of the nanoparticles. The most important findings reveal that they display sizes in the range of 150–200 nm at 0.01 wt %, with negative ζ potentials and cryo-TEM images suggesting roughly spherical nanoparticles. As the temperature increases, the PNIPAm chains dehydrate and collapse around 32 °C, decreasing the nanoparticle size. When the PNIPAm content and nanoparticle concentration increase, the heating leads to the formation of clusters of nanoparticles and, in some cases, to macroscopic phase separation.

The intermediary behavior of mixed nanoparticles was important to propose that PEO and PNIPAm chains are most likely segregated at room temperature in a patchy-like configuration.

These results should promote a better understanding of complex salts obtained from mixed polymer diblocks and future application as fast temperature–responsive system.