Study of the morphology and texture of poly(ε-caprolactone)/polyethylene oxide blend films as a function of composition and the addition of nanofillers with different functionalities

Abstract

Films of a blend of semi-crystalline polymers, namely poly(ε-caprolactone) (PCL) and poly(ethylene oxide) (PEO), are prepared by casting drops of ethyl lactate (EL) polymer solutions onto a glass substrate. The goal of this work is to assess how the surface structure of the blend films can be controlled by: (i) varying the relative amounts of the polymeric components, and (ii) adding inorganic nanoparticles (NPs) with different functionality. Specifically, four types of NPs are used here: bare and silanized titanium dioxide, hydroxyapatite and aluminum–magnesium layered double hydroxide. All the films exhibit a segregated surface morphology, characterized by either PEO-rich domains dispersed in a PCL continuous phase or discrete PCL domains embedded in a PEO-rich phase, depending on the composition of the blend. Phase inversion occurs within the 30–40% range of PEO weight fraction. The incorporation of NPs to the starting polymeric solution is found to significantly affect the final blend morphology, leading to either a coarsening or a refinement of the polymer phases mainly according to the chemical affinity between the NPs and the suspending medium.

---

1. Introduction

Polymer blending is a simple and cost-effective way to tailor the microstructure and properties of polymeric materials. In typical blending routes, the polymers are either mixed directly in the melt state or dissolved in a common solvent, which is then removed in order to induce polymer precipitation. In particular, solvent casting techniques (drop casting, spin coating) are extensively used to produce polymer blend films, whose main technological applications in electronics, optics, photovoltaics, and biotechnology require well-defined surface morphology and topography.^1–3

Due to the generally low mixing entropy of macromolecular materials, most polymer pairs are thermodynamically immiscible. As a consequence, phase separation phenomena typically occur during the blending process. The resulting morphology is determined not only by the composition of the blend and the properties of the single constituents, but also by the selected processing conditions.⁴ For binary blends, immiscibility brings about two main morphologies: (i) distributed morphologies, where one of the phases is continuous and encompasses dispersed domains of the other constituent; (ii) co-continuous morphologies, characterized by a mutual interpenetration of the polymeric phases. While the mechanisms of phase separation have been thoroughly investigated in molten binary blends of amorphous and crystalline polymers,^1,4–9 a lower level of understanding has been reached in the case of solvent-cast blend films mainly due to the complex phenomena that occur when polymers are forced to lie on a flat geometry.¹⁰

The surface morphology and texture of solvent-cast polymer blend films are correlated to the domain phase structure, which in turn depends on the surface features of the substrate, the composition of the starting solution and the evaporation of the solvent.^11–15 According to Walheim et al.,¹⁶ phase separation occurs after the spreading of the initial solution onto the substrate as a consequence of the evaporation of the solvent at the air/solution interface. This results in the formation of different polymer-rich phases, which are initially liquid due to the presence of large amounts of solvent. During solvent evaporation, polymer concentration progressively increases in each of the phases until supersaturation is reached, eventually leading to polymer precipitation. The different polymeric phases solidify at different stages of the casting process, depending on the solubility of the polymers in the common solvent. The size and shape of the domains and the topography of the resulting film are primarily determined by the rate and sequence of polymer precipitation. Moreover, when at least one of the polymers is semi-crystalline, the morphology and texture of the final film are also affected by the crystalline structure.¹¹

In the last two decades, inorganic nano-sized particles (NPs) have emerged as useful alternatives to system-specific chemical compatibilizers for the control of the morphology and structure of polymer blends.¹⁷ Indeed, the ability of NPs to induce suppression or promotion of domain coalescence,¹⁸ stabilization of non-spherical domains,¹⁹ clustering,²⁰ and co-continuity,^17,21 has been widely observed in both bulk polymer blends and low-molecular-weight multiphasic fluids (e.g. Pickering emulsions). A key aspect to exploit in order to use NPs to guide the morphology of multiphase systems is the uneven spatial distribution of the filler, which tends to enrich specific regions depending on its chemical affinity with the different constituents.^22,23 Due to the mobility of NPs even in highly viscous mediums, however, the development of the morphology of a NP-filled polymer blend during solvent casting is determined by a complex interplay among wetting, phase separation and solvent evaporation. Consequently, it is quite difficult to predict the localization of the filler in the blend on the basis of those thermodynamic considerations which, on the other hand, are successfully employed dealing with melt-compounded polymer systems.^24–28 Composto et al. have reported that small amounts (<2 wt%) of silica NPs segregate at the interphase of a poly(methyl methacrylate)/poly(styrene-co-acrylonitrile) blend due to the similar wetting character of the two polymers, eventually stabilizing the morphology of the system.^29,30 Moreover, Minelli et al. have observed a refinement of the microstructure and an increase of the surface roughness of poly(methyl methacrylate)/polystyrene films upon the addition of metal NPs.³¹ In this scenario, however, there is still plenty of room for exploring the potential of inorganic fillers to control the surface structure and properties of polymer blend films.

In this paper we focus on multiphase films of semi-crystalline polymers, namely poly(ε-caprolactone) (PCL) and poly(ethylene oxide) (PEO), obtained through casting from solutions in ethyl lactate (EL). PCL and PEO form an immiscible polymer pair which has relevant technological interest.^32,33 In particular, PCL/PEO blends have been used in biomedical applications such as controlled drug delivery and porous scaffolds for tissue engineering.^34–36 Moreover, the choice of a benign solvent such as EL, whose precursor is generated from biomass, makes the casting process completely sustainable, and therefore suitable for the production of polymer films for bio-related purposes. The main goal of this study is to assess how the morphology of the surface, the crystalline structure and the topography of the films can be controlled by (i) varying the relative amount of the polymer components and (ii) adding NPs with different size, shape and chemical functionality.

2. Experimental section

2.1 Materials

PCL (molecular weight M_w = 60 kDa) and PEO (molecular weight M_v = 100 kDa) were purchased from Sigma-Aldrich (Madrid, Spain). EL (purity grade ≥ 99.0 wt%) was supplied by Sigma-Aldrich (Madrid, Spain) and used without further purification.

Four different kinds of commercially available NPs were selected as fillers. Bare titanium dioxide (Aeroxide® TiO₂ P25) and silanized titanium dioxide (Aeroxide® TiO₂ T805, treated with octylsilane), both consisting of primary spheroidal particles with an average diameter of 21 nm, were supplied by Evonik Industries (Essen, Germany). Hydroxyapatite (HA, calcium phosphate) spheroidal particles with size <200 nm were purchased from Sigma-Aldrich (Madrid, Spain). Aluminum–magnesium layered double hydroxide (LDH, Perkalite® F100S modified with a hydrogenated fatty acid), in the form of individual platelets of ∼0.5 nm in thickness and ∼150 nm in width, was provided by AkzoNobel (Amersfoort, The Netherlands). The used fillers can be divided in two groups depending on their surface features: P25 and HA are hydrophilic, while T805 and LDH are hydrophobic.

2.2 Sample preparation

PCL and PEO at different weight ratios (see Table 1) were dissolved in EL under stirring at 60 °C for 4 h on a common laboratory heater (RCT basic, IKA, Staufen, Germany) to prepare solutions with a total polymer concentration of 5 wt/vol%.

Table 1 Designation and composition of the samples

Sample	Composition
	PCL [wt%]	PEO [wt%]	Filler [phr]
PCL	100	0	0
PEO	0	100	0
PEO10	90	10	0
PEO20	80	20	0
PEO30	70	30	0
PEO40	60	40	0
PEO40 + P25-3	60	40	3
PEO40 + T805-3	60	40	3
PEO40 + HA-3	60	40	3
PEO40 + LDH-3	60	40	3
PEO40 + P25-5	60	40	5
PEO40 + T805-5	60	40	5
PEO40 + HA-5	60	40	5
PEO40 + LDH-5	60	40	5

---

In the case of NPs-filled PCL/PEO blends, the polymer weight ratio was fixed to 60/40 and the materials were prepared by means of the following two-step procedure. First, each filler was separately dispersed in EL at a concentration of 5 wt/vol%; the suspension was then subjected twice to magnetic stirring for 5 min and subsequent ultrasonication for 15 min in order to reduce particle aggregation. Second, the concentration of NPs was adjusted by adding pure EL and the polymers were incorporated in order to obtain different solutions with a fixed polymer concentration of 5 wt/vol% and filler content of either 3 or 5 wt% relative to the polymer blend. Mixing was carried out at 60 °C for 4 h on a laboratory heater.

The preparation of polymer blend films, whose compositions are listed in Table 1, was carried out via drop casting. Circular glass slides, having a diameter of 12 mm and a thickness of 0.15 mm (Labbox Labware, Barcelona, Spain), were cleaned with acetone and pre-heated at 40 °C on a laboratory heater. Next, a 10 μL droplet of polymeric solution was deposited onto the glass slide, using and automatic pipette, and the solvent was allowed to evaporate by keeping the system at 40 °C overnight.

2.3 Characterization

The crystallinity of the polymers, as well as the crystalline structure and morphology of the surface of the formed films, were investigated with an optical microscope (Olympus BX51, Tokyo, Japan) used in both direct and transmission mode. The optical images were further analyzed by using the ImageJ® software to assess the area fraction and characteristic mean size of the different phase domains. At least five pictures, made over different regions of independent specimens and at different magnifications, were analyzed for each blend, following the procedure that is described in detail in the ESI.† The statistical significance of the results was assessed by one-way ANOVA. Tukey post-hoc test at the significance level p < 0.03 was used to identify statistically different data by means of the software OriginPro 8.

The morphology of the samples at the micron-scale was further inspected through scanning electron microscopy (SEM, QUANTA200F FEG ESEM, FEI, Eindhoven, The Netherlands). The surface of each sample was gold-sputtered before SEM observations. Energy dispersive spectroscopy (EDX, INCA by Oxford Instruments, High Wycombe, United Kingdom) was used to investigate the localization and distribution of the NPs.

The surface texture of the samples was further examined by means of confocal microscopy (Leica DCM 3D, Wetzlar, Germany). The analysis was carried out on the gold-sputtered surface of the samples with a 50× objective, and relevant topographical parameters (defined by ISO 25178 standard) were estimated. The root mean square height of the surface (Sq), corresponding to the standard deviation of the height distribution, was used to assess the surface roughness of the films. The developed interfacial area ratio (Sdr), defined as the percentage deviation of the actual surface from perfect flatness, was used to estimate the specific surface of the films. The analysis also enabled the assessment of the skewness of the height distribution (Ssk), which is indicative of the symmetry of the topography about the mean plane; in particular, positive (respectively negative) values of Ssk indicate that the bulk of the material is below (respectively above) the mean plane.

3. Results and discussion

3.1 Morphological and crystalline features of the homopolymer samples

The surface structure of the films of pure PCL and PEO is shown in the optical images (acquired in transmission mode) that are reported in Fig. 1. The precipitation of both polymers leads to the formation of spherulites, which have a lamellar structure and grow radially from nucleation sites. The growth front is hindered by the impingement of adjacent spherulites, which determines the final shape of the crystalline domains.

Fig. 1 Optical micrographs (acquired in transmission mode) of samples of pure PCL and PEO.

The spherulitic morphologies of PCL and PEO samples have significant differences in terms of crystal structure and size. The Maltese cross pattern of birefringence evidences that PCL spherulites consist of bundles of primary lamellae and have well-defined boundaries. In a recent paper, we have reported that the average spherulite diameter in the PCL film, estimated by analyzing the optical micrographs, is 97.1 ± 26.2 μm.¹¹ On the other hand, PEO spherulites are made of short fibrillar crystals whose arrangement is reminiscent of the “wheat-like” branched morphology observed by Hou et al.³⁷ and by Han et al.^8,9 Compared with PCL crystallites, PEO spherulites are much larger, the order of magnitude of their diameter being ∼10³ μm.

3.2 Morphology, crystalline structure and surface texture of the unfilled PCL/PEO blend films

The immiscibility of PCL and PEO leads to phase separation during the solvent casting process. As shown in the optical images of Fig. 2, the unfilled PCL/PEO blend exhibits phase-separated surface morphologies in the whole range of investigated compositions, with significant differences depending on the relative amount of the two polymers. At the lowest PEO content (sample PEO10), the surface structure resembles that of the neat PCL, while PEO is present in the form of small dispersed domains. As the PEO weight fraction in the blend increases, the density of PEO-rich domains increases while their size slightly augments. This was evidenced by image analysis, whose results are summarized in Fig. 3 in terms of characteristic size and area fraction of the PEO-rich phase. Further PEO addition up to 40 wt% determines a phase inversion regarding the surface structure of the films. The PEO40 sample exhibits isolated PCL domains immersed in a PEO-rich continuous phase. A closer inspection reveals that in all the blends the PEO-rich regions exhibit a “salami-like” morphology, where rounded PCL domains are embedded in the PEO phase. The results of image analysis indicate that the small PCL inclusions occupy between 10 and 20% of the surface area of the host phase. Further morphological details of the blend films can be appreciated in the SEM micrographs of Fig. 4. The PEO-rich regions appear brighter than the PCL-rich ones, and the presence of small-scale polymer inclusions in the various phases is clearly detectable. In particular, the PCL domains enclosed by the PEO-rich regions appear to be concave (see the high-magnification micrograph of the PEO20 sample). In the same image, the texture of the PCL and PEO crystals, respectively characterized by bundle-like and wheat-like lamellar structures, can be also appreciated.

Fig. 2 Optical micrographs (in direct and transmission mode) of PCL/PEO blend samples at different weight compositions.

Fig. 3 Area fraction and mean size of the PEO-rich (white columns) and embedded PCL-rich domains (grey columns) in the unfilled blends. The symbols (asterisks, diamonds, hashes and crosses) placed above the histogram bars designate sets of statistically significant data (p < 0.03).

Fig. 4 SEM micrographs of films of unfilled PCL/PEO blends at different compositions.

The existence of polymer structures with different size scales (10² to 10³ μm and less than 1 μm) in immiscible blend films has been evidenced by Liu et al.³⁸ for a blend of PCL and polyetherimide. The authors propose that the formation of small-scale inclusions occurs in the so-called “secondary phase separation”, which is an intermediate step of the solvent casting process of solution blending. In brief, after the spreading of the starting solution, the solvent evaporates at the air/solution interface and induces the formation of liquid phases that contain both polymers, but have different compositions. The further evaporation of the solvent results in an increase of the viscosity of the solution, hence, the polymer chains are only able to diffuse over short distances. Such hindered diffusion is mainly responsible for the development of small polymer domains, which result from the precipitation of the minor polymeric constituents trapped in each phase. The precipitation step is governed by the different solubility of the two polymers in their common solvent. Based on the Hildebrand solubility parameters, δ, of EL (δ_EL = 20.2 MPa^1/2),³⁹ PCL (δ_PCL = 19.7 MPa^1/2),⁴⁰ and PEO (δ_PEO = 20.2 MPa^1/2),⁴¹ it can be assumed that PEO is more soluble in EL than PCL. Taking this aspect into account, and considering that the weight fraction of PCL is always higher than that of PEO, it can be assumed that PCL is the first polymer to reach supersaturation and condense, forming structures with different characteristic size scales while PEO is still swollen by the solvent. At this point, further solvent evaporation is required for PEO to solidify, surrounded by the already vitrified PCL phase. The previous discussion highlights that the adopted casting procedure brings about non-equilibrium blend morphologies, kinetically controlled by the relatively rapid evaporation of the solvent. Such kinetically trapped morphologies can be driven to the equilibrium state by further annealing treatments (i.e. melting the samples after casting and quenching below the crystallization temperatures of the polymers). Studying the microstructural evolutions of the as-cast films toward equilibrium, however, goes beyond the target of the present work.

The different surface morphologies obtained varying the blend composition lead to dissimilar textural features of the final films, inspected through confocal microscopy. Representative images of the blend film surfaces are shown in Fig. 5, and the corresponding topographical parameters are reported in Table 2. We have recently showed that the different textures of the homopolymer films are marked by the different signs of Ssk, which is negative for PCL and positive for PEO.^11,42 The topography of the PCL film surface ensues from the combination of the lamellar structure of the single crystallite and the space arrangement of the crystallites. The inner structure of the large PEO spherulites is instead dominant in dictating the PEO film texture. Concerning the blends, both Sq and Sdr are found to increase with increasing PEO concentration, the values of Ssk being intermediate between those of the homopolymers. At low PEO contents, the film texture resembles that of the PCL matrix. As the PEO concentration increases, the topography of the surface becomes more complex. In particular, the considerable increase in film roughness at a high PEO content reflects the presence of a large number of concave, micron-sized PCL domains inside the PEO-rich phase (see Fig. 5).

Fig. 5 Surface topography of films of unfilled PCL/PEO blends at different compositions.

Table 2 Textural features of the unfilled and filled PCL/PEO blend samples

Sample	Sq [μm]	Sdr [%]	Ssk
PCL	0.03	0.05	−4.1
PEO	0.13	0.18	3.4
PEO10	0.03	0.05	−1.0
PEO20	0.30	0.48	−1.5
PEO30	0.72	1.8	0.01
PEO40	1.07	1.7	−0.18
PEO40 + P25-3	0.57	3.5	−0.19
PEO40 + T805-3	1.15	2.0	0.69
PEO40 + HA-3	0.69	3.7	0.51
PEO40 + LDH-3	0.80	3.1	−0.50
PEO40 + P25-5	0.78	1.9	1.1
PEO40 + T805-5	1.12	4.2	0.06
PEO40 + HA-5	0.78	4.2	0.44
PEO40 + LDH-5	0.76	2.5	0.42

---

3.3 Effect of nanoparticles

The previous investigation on the unfilled PCL/PEO system provides the bases for studying the effects of the incorporation of NPs on the surface structure of polymer blend films. Specifically, PEO40 samples filled with four kinds of NPs, differing among them in terms of geometrical and surface characteristics, have been selected.

The film morphology achieved after complete evaporation of the solvent is investigated by means of optical microscopy. Fig. 6 and 7 show representative optical images of PEO40 samples with 3 and 5 wt% of NPs, respectively. The results of domain size measurements are reported in Fig. 8. At the lowest filler content, T805, P25 and HA induce a reduction of the mean size of the PCL-rich phases, while minor effects are noticed upon addition of LDH. A refinement of the PCL phase are also detected in all the filled systems at higher NPs loading, the size distributions having a maximum around 40 μm. The narrowest distribution is exhibited by the samples filled with bare titanium dioxide NPs. In particular, the blends filled with P25 exhibit a peculiar surface structure, characterized by a pattern of alternated PEO-rich regions and clusters of round PCL domains, which keep their individuality due to the interposition of very thin PEO channels. Conversely, coarser morphologies are observed in the presence of silanized titanium dioxide. In such systems, the PEO-rich regions encompass large aggregates of NPs. Confocal analysis (see ESI† for representative micrographs) highlights that in the filled samples, the PCL-rich domains exhibit not only smaller sizes, but also increased height and numerical density in comparison to the unfilled blends. As inferable from the textural data reported in Table 2, such relief structure results in higher values of Sdr without significantly affecting Sq.

Fig. 6 Optical micrographs (in direct and transmission mode and different magnifications) of PEO40 blend samples filled with 3 wt% of TiO₂ (P25 and T805), HA and LDH.

Fig. 7 Optical micrographs (in direct and transmission mode and different magnifications) of PEO40 blend samples filled with 5 wt% of TiO₂ (P25 and T805), HA and LDH.

Fig. 8 Size distributions of the PCL-rich domains in the unfilled PEO40 blend (white columns) and in the PEO40 blends filled with 3 wt% (light grey columns) and 5 wt% of NPs (dark grey columns).

After elucidating the micron-scale arrangement of the polymer phases, the space distribution of the NPs is now investigated. The degree of dispersion of the different fillers in the pristine polymer solutions can be inferred from the optical images of Fig. 9, representative of systems at 5 wt% of filler concentration soon after the deposition of a drop of solution onto the glass substrate. The best level of dispersion is attained for the hydrophilic NPs (P25 and HA), while the hydrophobic ones (T805 and LDH) appear in the form of aggregates of tens of microns. The chemical affinity between the NPs and the solvent, which has a marked polar feature, is hence crucial in dictating the initial filler dispersion. Once the solvent begins to evaporate, the concentration of the polymers gradually increases. The consequent alteration in polarity and viscosity of the suspending medium affects the NPs dynamics, which in turn move and rearrange in the host medium, influencing polymer phase separation. As a result of this complex interplay, the final film morphology may deviate from what has been observed in the case of unfilled blend film (Fig. 3).

Fig. 9 Optical micrographs of nanofilled polymer solutions (NPs concentration: 5 wt% with respect to the polymer blend) acquired at the early stage of the casting process. NPs and their aggregates appear as dark zones dispersed into the brighter suspending fluid.

The phase separation process in the NPs filled systems has been monitored conducting solvent evaporation under the optical microscope (rather than at controlled temperature) and the final film structures are presented in Fig. 10. Representative optical images, taken at different stages of the process of blend film formation and containing titanium dioxide NPs, are reported in the ESI.† Despite being obtained in different casting conditions, these snapshots give an idea about the assembly of the fillers and their uneven distribution within the segregated polymer phases. Bare titanium dioxide NPs are able to arrange themselves into branched flocs apparently interconnected in a fractal space-spanning network, the PCL phase being segregated into small domains trapped in the mesh of NPs. On the other hand, the silanized titanium dioxide NPs form large agglomerates, likely reminiscent of those found in the pristine solution, and tend to sediment on the substrate during casting, acting as a solid obstacle for the evolution of the surrounding fluid phases. Even if characterized by a hydrophilic surface, such as P25 NPs, HA appears in the form of tiny, isolated particles while some large aggregates are detected in the case of LDH particles.

Fig. 10 Surface structure of nanofilled samples (filler concentration: 5 wt% with respect to the polymer blend) as achieved by solvent evaporation during microscope observation.

The final distribution of NPs in a polymer blend film is the outcome of various simultaneous phenomena, such as polymer phase separation, solvent evaporation, as well as the self-assembly of NPs driven by Brownian motion and van der Waals interactions.⁴³ EDX elemental mappings (phosphorus for HA, magnesium for LDH, titanium for P25 and T805), superimposed to the corresponding SEM micrographs of the filled blends in Fig. 11, have been used to localize the NPs on the surface of the films. The fillers exhibit a general tendency to enrich the PCL/PEO interphase regions. Specifically, the EDX mapping relative to the sample containing hydrophilic titania P25 at 3 wt% reveals that the NPs preferentially occupy the thin PEO-rich channels between contiguous PCL-rich domains. Keeping in mind the clustered morphology of the polymer phases, NP-induced bridging phenomena can be invoked. The latter can take place when the filler is preferentially wetted by the continuous phase, which is the present case.^44–46 At higher P25 concentrations, the surface structure is too refined to assess the space distribution of the filler. Considering now the hydrophobic titanium dioxide, at both investigated concentrations, the T805 is almost exclusively located in the PEO-rich phase. Taking into account the high polarity of PEO, the latter finding is quite unexpected. The distribution of the two remaining kinds of NPs is also surprising, at least in terms of mere thermodynamic arguments: both HA and LDH are equally partitioned in the two polymeric phases, the hydrophobic filler being more inclined to gather at the PCL/PEO interphase. The linear elemental spectra reported in the ESI† corroborate the previous findings (see Fig. S5†). Compared with melt compounding, solvent casting seems to bring about a less selective (equivalently, more widespread) distribution of the filler in the polymer phases. This occurrence is ascribable to the predominance of kinetic factors over thermodynamics, which promotes non-equilibrium space arrangements of the NPs. Specifically, the compounding procedure adopted in this work is much longer than the typical melt blending times in extruders and mixers, and hence allows to obtain relatively homogeneous distributions of the fillers. The evaporation of the solvent arrests the mobility of the NPs, which are constrained to remain in the host precipitating polymer phase. Such mobility hindrance is particularly pronounced if the NPs form agglomerates in the pristine solution, remarkably increasing the viscosity of the phase where they tend to partition.

Fig. 11 SEM micrographs of filled PEO40 blend films and superimposed X-ray maps of different elements: titanium (red) for P25 and T805, phosphorus (green) for HA and magnesium (blue) for LDH.

The cross-check of the results concerning the different kinds of fillers allows to draw some meaningful conclusions about the impact of the addition of NPs on the surface morphology of polymer blend films cast from a polar solvent, such as EL. A prime role is played by the filler polarity, which determines the quality of the dispersion of the NPs in the starting solution: a fine initial dispersion is achieved for hydrophilic fillers (P25 and HA), while hydrophobic fillers (T805 and LDH) tend to aggregate more in the suspending medium. Nonetheless, the final localization of the NPs and, ultimately, the morphology of the polymer phases are dictated not only by the polarity of the filler, but also by its tendency to flocculate in the suspending medium, which is related to the NP chemistry. In this sense, P25 has revealed a marked ability to assemble into a fractal space-filling network and, as a consequence, to bring about a significant refinement of the dispersed polymer domains. A comparison between fillers with different initial dispersion, but similar ultimate distribution (HA and LDH), highlights that an inadequate dispersion results in marginal morphological effects in spite of the uneven final localization of the NPs. Finally, relatively low filler loadings (3–5 wt% with respect to the overall polymers amount) are required to assess and discern the effects described above, related to different NPs physicochemical features. The results reported in this work, despite being focused on a specific multiphase system, provide useful guidelines for the control of relevant parameters, such as polymer composition and filler characteristics, to tune the morphological and textural features of polymer blend films.

4. Conclusions

This study reports the preparation of films of blends of semi-crystalline polymers, namely poly(ε-caprolactone) (PCL) and poly(ethylene oxide) (PEO), through solvent casting from solutions in ethyl lactate. The results demonstrate that, as in the case of bulk multiphase systems, the morphology of polymer blend films can be effectively controlled by either tuning the blend composition and/or adding small amounts of inorganic nanoparticles (NPs).

The incompatibility between the polymers results in a multi-step phase separation process, which is driven by both solvent evaporation and solubility of each of the polymers in the common solvent. These aspects lead to the formation of final segregated structures with different characteristic length scales. In PCL/PEO systems, adjusting the composition to 40 wt% of PEO evidences the transition from a dispersed to a continuous PEO-rich phase. This phase inversion has also relevant implications on the blend surface texture, which is essentially dictated by the topography of the continuous phase.

The presence of NPs suspended in the starting polymer solution considerably impacts on the evolution of the fluid phases during solvent casting and the morphological features of the resulting blend films. The crucial factors that determine how the NPs are dispersed in the initial polymer solution and distributed in the final system are the polarity of the filler and the ability of the latter to flocculate in the suspending medium. The different filler structures formed during the casting process, namely large aggregates, isolated micron-sized flocs and space-filling networks, govern the final blend morphology, determining either a coarsening or a refinement of the PCL-rich domains.

Acknowledgements

A. Salerno and C. Domingo acknowledge the financial support of the Generalitat de Catalunya through the project 2014SGR-377 and the Spanish Plan National Project CTQ2014-56324.

References

1. C. L. Tucker III and P. Moldenaers, Annu. Rev. Fluid Mech., 2002, 34, 177.
2. C. D. W. Wilkinson, M. Riehle, M. Wood, J. Gallagher and A. S. G. Curtis, Mater. Sci. Eng., C, 2002, 19, 263.
3. N. Corcoran, A. C. Arias, J. S. Kim, J. D. MacKenzie and R. H. Friend, Appl. Phys. Lett., 2003, 82, 299.
4. V. E. Ziegler and B. A. Wolf, Macromolecules, 2005, 38, 5826.
5. H. Tanaka and T. Nishi, Phys. Rev. Lett., 1985, 55, 1102.
6. L. A. Utracki and Z. H. Shi, Polym. Eng. Sci., 1992, 32, 1824.
7. J. Liu and B. J. Jungnickel, J. Polym. Sci., Part B: Polym. Phys., 2007, 45, 1917.
8. W. Shi and C. C. Han, Macromolecules, 2012, 45, 336.
9. W. Shi, J. Yang, Y. Zhang, J. Luo, Y. Liang and C. C. Han, Macromolecules, 2012, 45, 2012.
10. S. Y. Heriot and R. A. L. Jones, Nat. Mater., 2005, 4, 782.
11. A. Causa, G. Filippone, D. Acierno, C. Domingo and A. Salerno, Macromol. Chem. Phys., 2015, 216, 49.
12. S. Walheim, M. Boltau, J. Mlynek, G. Krausch and U. Steiner, Macromolecules, 1997, 30, 4995.
13. P. Mokarian-Tabari, M. Geoghegan, J. R. Howse, S. Y. Heriot, R. L. Thompson and R. A. L. Jones, Eur. Phys. J. E: Soft Matter Biol. Phys., 2010, 33, 283.
14. C. M. Dekeyser, S. Biltresse, J. Marchand-Brynaert, P. G. Rouxhet and C. C. Dupont-Gillain, Polymer, 2004, 45, 2211.
15. L. Li, S. Sosnowski, C. E. Chaffey, S. T. Balke and M. A. Winnik, Langmuir, 1994, 10, 2495.
16. S. Walheim, M. Ramstein and U. Steiner, Langmuir, 1999, 15, 4828.
17. H. Xiu, Y. Zhou, J. Dai, C. Huang, H. Be, Q. Zhang and Q. Fu, RSC Adv., 2014, 4, 37193.
18. P. Thareja, K. Moritz and S. Velankar, Rheol. Acta, 2010, 49, 285.
19. M. Si, T. Araki, H. Ade, A. L. D. Kilcoyne, R. Fisher, J. C. Sokolov and M. H. Rafailovich, Macromolecules, 2006, 39, 4793.
20. G. Filippone and D. Acierno, Macromol. Mater. Eng., 2012, 297, 923.
21. C. Gao, S. Zhang, B. Han, H. Sun, G. Wang and Z. Jiang, RSC Adv., 2014, 4, 42175.
22. D. Wu, D. Lin, J. Zhang, W. Zhou, M. Zhang, Y. Zhang, D. Wang and B. Lin, Macromol. Chem. Phys., 2011, 212, 613.
23. G. Filippone, A. Causa, M. Salzano de Luna, L. Sanguigno and D. Acierno, Soft Matter, 2014, 10, 3183.
24. K. H. Wu and S. Y. Lu, Macromol. Rapid Commun., 2006, 27, 424.
25. F. Fenouillot, P. Cassagnau and J. C. Majesté, Polymer, 2009, 50, 1333.
26. M. H. Al-Saleh, H. K. Al-Anid and Y. A. Hussain, Synth. Met., 2013, 175, 75.
27. S. Chandran, N. Begam, V. Padmanabhan and J. K. Basu, Nat. Commun., 2014, 5, 3697.
28. A. Taguet, P. Cassagnau and J. M. Lopez-Cuesta, Prog. Polym. Sci., 2014, 39, 1526.
29. S. Gam, A. Corlu, H. J. Chung, K. Ohno, M. J. A. Hore and R. J. Composto, Soft Matter, 2011, 7, 7272.
30. H. J. Chung, K. Ohno, T. Fukuda and R. J. Composto, Nano Lett., 2005, 5, 1878.
31. C. Minelli, C. Frommen, C. Hinderling, R. Pugin, H. Heinzelmann and M. Liley, Colloid Polym. Sci., 2006, 284, 482.
32. Z. Qiu, T. Ikehara and T. Nishi, Polymer, 2003, 44, 3101.
33. Z. Fang, C. Harrats, N. Moussaif and G. Groeninckx, J. Appl. Polym. Sci., 2007, 106, 3125.
34. N. R. Washburn, C. G. Simon, A. Tona, H. M. Elgendy, A. Karim and E. J. Amis, J. Biomed. Mater. Res., 2002, 60, 20.
35. S. J. Park, K. S. Kim and S. H. Kim, Colloids Surf., B, 2005, 43, 238.
36. J. G. Lyons, P. Blackie and C. L. Higginbotham, Int. J. Pharm., 2008, 351, 201.
37. W. M. Hou, J. J. Zhou, Z. H. Gan, A. C. Shi, C. M. Chan and L. Li, Polymer, 2007, 48, 4926.
38. T. Liu, R. Ozisik and R. W. Siegel, Thin Solid Films, 2007, 515, 2965.
39. A. F. M. Barton, in Handbook of Solubility Parameters and Other Cohesion Parameters, CRC Press LLC, 1991, p. 105.
40. C. Bordes, V. Fréville, E. Ruffin, P. Marote, J. Y. Gauvrit, S. Briançon and P. Lantéri, Int. J. Pharm., 2010, 383, 236.
41. Q. W. Yuan, in Polymer Data Handbook, ed. J. E. Mark, Oxford University Press, Inc, 1999, p. 545.
42. A. Causa, G. Filippone, D. Acierno, C. Domingo and A. Salerno, J. Supercrit. Fluids, 2015 DOI:10.1016/j.supflu.2015.03.021.
43. N. Dutta and D. Green, Langmuir, 2008, 24, 5260.
44. T. S. Horozov and B. P. Binks, Angew. Chem., 2006, 118, 787.
45. P. Thareja and S. S. Velankar, Rheol. Acta, 2008, 47, 189.
46. S. P. Nagarkar and S. Velankar, Soft Matter, 2012, 8, 8464.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra08864j

---