Strong Screening Effect of Polyhedral Oligomeric Silsesquioxanes (poss) Nanoparticles on Hydrogen Bonded Polymer Blends

In this study we used anionic living polymerization to prepare two different homopolymers: a poly(methyl methacrylate) (PMMA) and a PMMA derivative presenting polyhedral oligomeric silsesquioxane (PMA-POSS) units as its side chains. We then employed differential scanning calorimetry (DSC), Fourier transform infrared (FTIR) spectroscopy, and wide-angle X-ray diffraction (WAXD) to investigate the miscibility and specific interactions of PMMA and PMA-POSS with three hydrogen bonding donor compounds: poly(vinyl phenol) (PVPh), phenolic resin, and bisphenol A (BPA). DSC revealed that all of the PVPh/PMMA, phenolic/PMMA, and BPA/PMMA blends exhibited a single glass transition temperature, characteristic of miscible systems; FTIR spectroscopic analyses revealed that such miscibility resulted from hydrogen bonding interactions between the C=O groups of PMMA and the OH groups of these three hydrogen bonding donor compounds. In contrast, all of the PVPh/PMA-POSS, phenolic/PMA-POSS, and BPA/PMA-POSS blends were immiscible: DSC revealed two 927 glass transition temperatures arising from strong screening effects (FTIR spectroscopy) and high degrees of aggregation (WAXD) of the POSS nanoparticles. We propose that the value of the intramolecular screening effect (γ) should be very close to 1 for all PMA-POSS blend systems when POSS nanoparticles appear as the side chains of PMMA, such that the OH groups of the hydrogen bonding donor compounds cannot interact with the C=O groups of PMA-POSS.


Introduction
Polymer blending can be a powerful route toward materials exhibiting properties and cost performances superior to those of their individual components. To enhance the formation of single-phase, miscible polymer blends, it is necessary to ensure the presence of favorable and specific intermolecular interactions (e.g., hydrogen bonding, ion-dipole, π-π, and charge transfer interactions) [1][2][3][4][5]. The blending of poly(methyl methacrylate) (PMMA, a hydrogen bond acceptor polymer) with various hydrogen bond donor polymers, including poly(vinyl phenol) (PVPh) and phenolic resins, has been investigated widely over the years [6][7][8][9][10][11][12]. These systems generally form miscible blends as a result of intermolecular hydrogen bonding interactions between the C=O groups of PMMA and the OH groups of the hydrogen bond donor polymers [11][12][13][14][15]. The miscibility of polymer blends is commonly ascertained through the use of differential scanning calorimetry (DSC) to measure glass transition temperatures (T g ). Moreover, Fourier transform infrared (FTIR) spectroscopy has been a powerful tool for characterizing the detailed structures of polymers and their specific interactions, which affect local electron densities and, thereby, result in frequency shifts [16][17][18].
Painter and Coleman association model (PCAM) is modified by the Flory-Huggins theory to include a free energy change associated with hydrogen bonding that has both entropic and enthalpic contribution (ΔG H /RT). Furthermore, the inter-association equilibrium constant and its related enthalpy necessary to quantify this free energy can be calculated at various temperatures by FTIR according to PCAM [1]. Serman et al. [11] used DSC to study the phase behavior of PVPh blended with PMMA, finding them to be miscible; they determined the inter-association equilibrium constant (K A ) for the interactions of the OH groups of PVPh with the C=O groups of PMMA to be 37.4. Painter and Coleman proposed [19][20][21] that "intramolecular screening and functional group accessibility" can have significant effects on the number of hydrogen-bonded functional groups. They used FTIR spectroscopy to measure the fraction of hydrogen-bonded C=O groups present in miscible blends of PVPh/PMMA as functions of the composition and temperature and compared their results with those of analogous PVPh-co-PMMA copolymers and polymer solutions of PVPh and ethyl isobutyrate (EIB) [22]. Painter and Coleman found that there were significant differences in the equilibrium fractions of intermolecular hydrogen-bonded C=O groups that formed under otherwise identical concentrations and temperatures [22]. They introduced an intramolecular screening parameter (γ), defined as the fraction of same-chain contacts originating from polymer chain self-bending, primarily through local, but also through long-range, connectivity effects [19][20][21]. Painter and Coleman also examined the screening effect for a hyperbranched, dendrimer-like polyester in the presence of low-molecular-weight PVPh as a model compound. The values of γ for a linear chain structure blend (e.g., PVPh/PMMA) is approximately 0.3; for a dendrimer-like polyester blend (e.g., PVPh/dendrimer-like polyester) it is approximately 0.8 [19,23]. Scheme 1. Synthesis of PMMA and PMA-POSS homopolymers through living anionic polymerization.
Phenolic/PMMA systems are also miscible blends that have been studied extensively [24,25]. In a previous study, we determined a value of K A of 20 directly through a least-squares fitting procedure based on the fraction of hydrogen-bonded C=O groups measured experimentally in a binary phenolic/PMMA blend [25]. We also used atom transfer radical polymerization (ATRP) to prepare inorganic/organic polymer hybrids featuring polyhedral oligomeric silsesquioxane (POSS) moieties at the chain ends; we obtained well-defined PMMA and PMMA-POSS homopolymers with designed molecular weights (ca. 40,300 and 38,300 g/mol) and then blended them with phenolic resin. POSS is a new type of material capable of forming various nanocomposites [25]. POSS derivatives are interesting materials because one or more of their organic groups can be functionalized for polymerization, while their remaining unreactive groups can be designed to solubilize the inorganic core and control the interfacial interactions between the POSS units and the polymer matrix. Due to their potential for advanced performance (e.g., high thermal stability; low flammability; low dielectric constant; low surface energy) relative to nonhybrid counterparts, POSS-polymer hybrid materials have attracted great interest recently [26][27][28][29][30][31][32][33][34][35]. We observed a "screening effect" in phenolic/PMMA-POSS blends arising from the tethered POSS units at the chain ends, with the inter-association equilibrium sec-BuLi DPE [25]. The value of γ of 0.65 for the phenolic/PMMA-POSS blend was lower than that for the dendrimer-like polyester blend (γ = 0.8) [19], but higher than that for the linear chain blend system (γ = 0.3) [23], suggesting that the chain structure of PMMA-POSS is somewhere between those of dendrimer-like and linear chain structures [25]. When POSS moieties are located only at the chain ends of PMMA polymers, the amount of POSS in a sample would be extremely low and the screening effect should be minimal. Therefore, in this present study we wished to investigate the effect of grafting POSS nanoparticles to the side chains of PMMA (forming so-called "PMA-POSS") on the miscibility behavior and hydrogen bonding interactions of PVPh/PMA-POSS, phenolic/PMA-POSS, and bisphenol A (BPA)/PMA-POSS blends. First, we used living anionic polymerization to prepare (Scheme 1) pure PMMA and pure PMA-POSS with similar molecular weight and narrow polydispersity indices (PDIs). We then blended PMMA and PMA-POSS individually with PVPh, phenolic resin, and BPA, with a particular interest in determining whether the POSS side chains would affect the thermal properties, miscibility behavior, and hydrogen bonding interactions of these binary blends.

Materials
Methyl methacrylate (MMA, SHOWA, 99%) was distilled from finely ground CaH 2 prior to use. Hepta(isobutylpentacyclo-octasiloxan-1-yl)propyl methacrylate (MA-POSS, Hybrid Plastic) was dissolved in tetrahydrofuran (THF) and then precipitated into MeCN; the purified MAPOSS monomer was then recrystallized from MeOH and dried in a vacuum oven at 60 °C overnight. THF was freshly distilled over Na/benzophenone (deep purple color) after refluxing for 2 h under N 2 . 1,1-Diphenylethylene (DPE) was distilled from n-BuLi; LiCl was dried overnight at 160 °C in a vacuum oven. sec-BuLi (1.3 M in cyclohexane; Chemetall) was used as obtained. Bisphenol A (molecular weight: 228.29) was supplied by the Showa Chemical (Japan). The PVPh polymer (M w = 9000-10,000) was purchased from Polyscience (USA). The phenolic was synthesized through an H 2 SO 4 -catalyzed condensation reaction, producing an average molecular weight (M n ) of 500, using a procedure described previously [36][37][38]. Table 1 summarizes the molecular weights of all of the materials used in this study.

PMMA Homopolymer through Living Anionic Polymerization
A glass reactor containing dry LiCl (25 mg, 0.61 mmol) was charged with THF (40 mL) and then cooled to −78 °C under Ar. After 5 min, sec-BuLi was added until the color became light yellow. The reactor was removed from the cooling bath and warmed to room temperature until the solution becomes colorless. The reactor was again cooled to −78 °C and then 1.3 M sec-BuLi in cyclohexane (0.035 mL, 0.046 mmol) was added. After 5 min, DPE (0.04 mL, 0.23 mmol) was added, resulting in a deep red color. After 30 min, MMA (2 mL, 18.4 mmol) was added via cannula to the polymerization reactor with vigorous stirring. The deep red color disappeared to give a light yellow solution. After 1.5 h at −78 °C, an excess of MeOH was added to terminate the reaction. The polymer was precipitated into MeOH and dried overnight at 60 °C in a vacuum oven. Yield: 1.45 g (77%). Table 1. Self-association and inter-association equilibrium constants and thermodynamic parameters for poly(methyl methacrylate) (PMMA) and polyhedral oligomeric silsesquioxane (PMA-POSS) and their blends with poly(vinyl phenol) (PVPh), phenolic, and bisphenol A (BPA) at 25 °C.

PMA-POSS Homopolymer through Living Anionic Polymerization
A glass reactor containing dry LiCl (13 mg, 0.31 mmol) was charged with THF (40 mL) and then cooled to −78 °C under Ar. After 5 min, sec-BuLi was added until the color became light yellow. The reactor was removed from the cooling bath and warmed to room temperature until the solution becomes colorless. The reactor was again cooled to −78 °C and then 1.3 M sec-BuLi in cyclohexane (0.018 mL, 0.023 mmol) was added. After 5 min, DPE (0.021 mL, 0.12 mmol) was added, resulting in a deep red color. After 30 min, a solution of MA-POSS (2.00 g, 2.1 mmol) in THF (6.0 mL) was added from to the polymerization reactor via cannula with vigorous stirring. The deep red color disappeared to give a colorless solution. After 9 h at −78 °C, an excess of MeOH was added to terminate the reaction. The polymer was then precipitated into MeOH and dried overnight at 60 °C in a vacuum oven. Yield: 1.55 g (78%).

Blend Preparations
Blends of phenolic/PMMA, PVPh/PMMA, BPA/PMMA, phenolic/PMA-POSS, PVPh/PMA-POSS, and BPA/PMA-POSS at various compositions were prepared through solution-casting. A THF solution containing 5 wt% of the polymer mixture was stirred for 6-8 h and then the solvent was evaporated slowly at 50 °C for 1 day. The film of the blend was then dried at 80 °C for 2 days to ensure total removal of residual solvent.

Characterization
1 H, 13 C, and 29 Si NMR spectra were recorded at room temperature using a Bruker AM 500 (500 MHz) spectrometer, with the residual proton resonance of the deuterated solvent as the internal standard. Molecular weights and molecular weight distributions were determined through gel permeation chromatography (GPC) using a Waters 510 high-performance liquid chromatograph equipped with a 410 differential refractometer and three Ultrastyragel columns (100, 500, and 10 3 Å) connected in series, with THF as the eluent (flow rate: 1.0 mL/min). DSC was performed using a TA-Q20 instrument operated at a scan rate of 20 °C/min over a temperature range from 0 to 250 °C under a N 2 atmosphere. FTIR spectra of the polymer films, sufficiently thin to obey the Beer-Lambert law, were recorded using the conventional KBr disk method and a Bruker Tensor 27 FTIR spectrophotometer; 32 scans were collected at a spectral resolution of 1 cm −1 . As polymers containing OH groups are hygroscopic, pure N 2 gas was used to purge the spectrometer's optical box to ensure dry sample films. FTIR spectra of samples at elevated temperatures were recorded using a cell mounted within the temperature-controlled compartment of the spectrometer. WAXD data were collected using the BL17A1 wiggler beamline of the National Synchrotron Radiation Research Center (NSRRC), Taiwan. A triangular bent Si (111) single crystal was employed to obtain a monochromated beam having a wavelength (λ) of 1.33001 Å.

Synthesis of PMA-POSS
Hirai et al. have reported that PMA-POSS homopolymers can be prepared through living anionic polymerization when using sec-BuLi as the initiator in the presence of excess LiCl and 1,1-diphenylethylene (DPE) at −78 °C [39]. In this study, we used a similar method to prepare the PMA-POSS homopolymer. Figure 1 displays the 1 H, 13 C, and 29 Si NMR spectra and MALS analysis data of our pure PMA-POSS homopolymer. The 1 H NMR spectrum (Figure 1a) of PMA-POSS features the signal for the SiCH 2 (a) unit of the POSS cage at 0.59 ppm and for the OCH 2 (g) unit of the side chain at 3.76 ppm; the 13 Figure 1d suggests a number-average molecular weight (M n ) for the PMA-POSS sample prepared in this study of 38,300 g/mol and a PDI of 1.07. Taken together, the 1 H, 13 C, and 29 Si NMR spectra and the GPC analysis confirmed the successful synthesis of PMA-POSS. Figure 2a,b displays conventional second-run DSC thermograms of PMMA and PMA-POSS, respectively, in the presence of PVPh at various weight ratios. In this study, the glass transition temperatures of PMMA and PMA-POSS were 131 and 59 °C, respectively. This value of T g for PMA-POSS was close to that reported previously after living anionic polymerization [39]. The value of T g of poly(n-propyl methacrylate) (PPMA) has been reported to be approximately 35 °C; therefore, the increase in the glass transition temperature for PMA-POSS (to 59 °C) was presumably due to the nano-reinforcement effect of the rigid cubic silsesquioxane core of the POSS units on the PPMA segments, effectively restricting their molecular chain motion [40][41][42]. Figure 2a presents the second heating runs in our DSC analyses of PVPh/PMMA blends at various compositions; all blend compositions exhibited a single glass transition temperature over the entire range of compositions, indicating that they were fully miscible and featured a homogeneous amorphous phase. In contrast, the DSC thermograms of the PVPh/PMA-POSS blend (Figure 2b) displays two values of T g , implying that the components were phase-separated in the amorphous phase. The value of T g of the PMA-POSS domain decreased slightly upon increasing the PVPh content; in contrast, the value of T g of the PVPh domain increased upon increasing the PMA-POSS content.    Figure 3a,b presents the conventional second-run DSC thermograms of PMMA and PMA-POSS, respectively, with phenolic blends at various weight ratios. Similar to the results for PVPh/PMMA and PVPh/PMA-POSS, the phenolic/PMMA blend compositions exhibited a single glass transition temperature, implying miscible blends; in contrast, the phenolic/PMA-POSS blends provided two values of T g , implying phase-separation in the amorphous phase. The value of T g of the PMA-POSS domain also decreased slightly upon increasing the phenolic content, whereas the value of T g of the phenolic domain increased upon increasing the PMA-POSS content. Figure 4a,b summarizes the glass transition temperatures for the PVPh/PMA-POSS and phenolic/PMA-POSS blends, respectively. Next, we investigated the blending of bisphenol A (BPA)-a crystalline, low-molecular-weight, hydrogen bond donor compound-with PMMA and PMA-POSS. Although the molar mass of BPA is only 228, it features a single glass transition temperature at 42.5 °C during its second heating scan; pure BPA also possesses a melting temperature of 162.3 °C [43]. The binary blends of BPA/PMMA were totally miscible in the amorphous phase, as evidenced by a single glass transition temperature in the DSC analyses in Figure 5a. In addition, the signal for the melting temperature of the BPA component disappeared upon increasing the PMMA content. In contrast, the melting temperatures for the BPA/PMA-POSS blends decreased slightly upon increasing the PMA-POSS content, but plateaued at relatively high PMA-POSS contents, indicating that BPA did not dissolve completely in PMA-POSS (phase-separated blends). As the values of T g of BPA and PMA-POSS were similar, it was difficult to discern the miscibility behavior based on DSC analysis. We could, however, conclude that the intermolecular interactions in the BPA/PMMA and BPA/PMA-POSS blends were quite different because of the distinct crystallization behavior of the BPA phases within them.   of T g , indicating that they were all fully miscible. In general, the relationship between the glass transition temperature and the composition of a miscible polymer blend follows the Kwei equation [44]:

DSC Analyses
where W 1 and W 2 are the weight fractions of the components, T g1 and T g2 represent the corresponding glass transition temperatures, and k and q are fitting constants. The parameter q corresponds to the strength of specific interactions in the blend, reflecting a balance between the breaking of self-association interactions and the formation of inter-association interactions. As displayed in Figure 6, using a non-linear least-squares "best fit" approach we obtained values of k and q for the PVPh/PMMA blends of 1 and −20, respectively; for the phenolic/PMMA blends of 1 and −15, respectively; and for the BPA/PMMA blends of 1 and −120, respectively. The negative values of q indicate that the inter-association equilibrium constants (or average hydrogen bonding strengths) for the OH···C=O interactions were less than the self-association equilibrium constants for the OH···OH interactions in these three binary blends. Here, we compare only the results for PVPh/PMMA and phenolic/PMMA, because BPA is a low-molecular-weight compound having a crystal structure that is different from those of PVPh and phenolic resin. We conclude from this study that the average hydrogen bonding strength of phenolic/PMMA was larger than that of PVPh/PMMA, consistent with our previous finding that phenolic is a more acidic hydrogen bond donor polymer relative to PVPh [45].

FTIR Spectra of PMMA Homopolymer Blend Systems
Infrared (IR) spectroscopy can provide positive information regarding the specific interaction between polymers, both qualitatively and quantitatively. The OH stretching range in an IR spectrum is sensitive to the presence of hydrogen bonds. Figure 7 presents the OH stretching regions of the FTIR spectra of PVPh/PMMA, phenolic/PMMA, and BPA/PMMA blends recorded at room temperature. The spectra of pure PVPh, phenolic, and BPA feature two distinct bands in the OH-stretching region of the FTIR spectra: a very broad band centered at 3350, 3360, and 3330 cm −1 , respectively, representing the wide distribution of the hydrogen bonded OH groups and a sharp band near 3525 cm −1 , caused by the free OH groups. In general, the intensity of the signal for the free OH groups decreased upon increasing the PMMA content, as would be expected. Meanwhile, the broad signal for the hydrogen-bonded OH groups shifted to higher frequency-to 3450, 3430, and 3445 cm −1 for phenolic, PVPh, and BPA, respectively-upon increasing the PMMA content to 80 wt%. All these observed changes arose from transformations from strong intramolecular OH···OH bonds to weak intermolecular OH···C=O bonds.  These results are consistent with the negative values of q obtained using the Kwei equation. Coleman et al. [1] have suggested using the frequency difference (Δν) between the signals for the hydrogen-bonded and free OH units as a measure of the average strength of the intermolecular interactions. In this study, we found that the average strength of the hydrogen bonds between the C=O groups of PMMA and the OH groups of the three OH-containing compounds followed the order phenolic/PMMA (Δν = 95 cm −1 ) > BPA/PMMA (Δν = 80 cm −1 ) > PVPh/PMMA (Δν = 75 cm −1 ).
In addition to the OH stretching region, the C=O stretching was also sensitive to hydrogen bond formation. Figure 8 displays the C=O stretching region of the FTIR spectra of PVPh/PMMA, phenolic/PMMA, and BPA/PMMA recorded at room temperature. All of the C=O stretching frequencies were split into two bands, which could be fitted well to the Gaussian function, as displayed in Figure 9. One band, with absorption at 1730 cm −1 , corresponds to the free C=O groups; the other, located at 1708, 1706, and 1706 cm −1 for phenolic, PVPh, and BPA, respectively, represents the hydrogen-bonded C=O groups. The fraction of hydrogen-bonded C=O groups can be calculated using an appropriate absorptivity ratio (a R = a HB /a F = 1.5), as has been discussed previously [1]. Table 2 summarizes the results from curve fitting; Figure 10 displays the fraction of hydrogen-bonded C=O groups in the individual binary polymer blends of PVPh/PMMA, phenolic/PMMA, and BPA/PMMA.
On the basis of the Painter-Coleman association model, we determined the inter-association equilibrium constants for PVPh/PMMA, phenolic/PMMA, and BPA/PMMA, using a least-squares method, as has been described previously [46]. Table 1 lists all of the thermodynamic parameters in these three polymer blend systems. K 2 and K B represent the self-association equilibrium constants for hydrogen-bonded dimers and multimers, respectively, of PVPh, phenolic, and BPA; K A represents the equilibrium constant describing the inter-association of PMMA with PVPh, phenolic, or BPA. We determined the inter-association equilibrium constants (K A ) for the PVPh/PMMA, phenolic/PMMA, and BPA/PMMA blends to be approximately 10, 20, and 20, respectively, as shown in Figure 10.    Table 2. Curve-fitting of the Fourier transform infrared (FTIR) spectroscopic data for the C=O groups of PMMA and its blends with PVPh, phenolic, and BPA.   The value of K A for the phenolic/PMMA blend is similar to that we had determined in a previous study [25]; in contrast, the value of K A for the PVPh/PMMA blend is smaller than those found in previous studies [11], presumably because of the different solvents employed in the different blend preparations. The implication here is that miscibility may be achieved when THF is used to cast the blend, but not when methyl isobutyl ketone (MIBK) is used as the solvent, as in the previous studies [11]. This phenomenon can be explained by considering the compositional heterogeneities that arise from the different solvent molecules, as has been discussed widely in previous studies [6][7][8][9]22]. Dong et al. [9] indicate that in solvent-cast PVPh + PMMA limited hydrogen-bonding interaction exists when THF is used, due to the strong hydrogen-bonding affinity of THF that develops hydrogen bonds with the polymer molecules, compared to the weaker hydrogen-bonding affinity of MIBK. Here, we choose THF as our cast solvent because PMA-POSS was not soluble in MIBK. For comparison with the PMA-POSS blend system, we also choose to study the PMMA blend system in THF solution. Clearly, the inter-association equilibrium constants and relative ratios of K A /K B followed the order phenolic/PMMA > BPA/PMMA > PVPh/PMMA. In these three binary blends, however, the OH groups of PVPh, phenolic, and BPA could still interact readily with the C=O groups of PMMA, even though their donor abilities were different. Figure 11 displays the C=O stretching regions of the FTIR spectra of PVPh/PMA-POSS, phenolic/PMA-POSS, and BPA/PMA-POSS recorded at room temperature. The blending of the PMA-POSS homopolymer with these hydrogen bond donor compounds was clearly quite different from that of the PMMA homopolymer. We did not observe any hydrogen-bonded C=O groups of PMA-POSS when blending with the polymer PVPh or phenolic or even with the low-molecular-weight compound BPA. We rationalize this phenomenon in terms of the so-called "screening effect." The value of the intramolecular screening parameter (γ) for a linear chain structure blend (e.g., PVPh/PMMA) is approximately 0.3, whereas that of a dendrimer-like polyester blend (e.g., PVPh/dendrimer-like polyester) is approximately 0.8 [19,23]. From a previous study [25], we reported that a value of γ of 0.65 for a phenolic/PMMA-POSS block copolymer blend in which the POSS nanoparticles were positioned at the chain end of the PMMA polymer. In this present study, the values of γ were all very close to 1 for the PVPh/PMA-POSS, phenolic/PMA-POSS, and BPA/PMA-POSS blend systems, in which the POSS nanoparticles were located on the side chains of the PMMA polymer. As a result, the C=O groups of PMA-POSS could not interact through intermolecular hydrogen bonding with the OH groups of the hydrogen bond donor compounds. This behavior presumably arose for two main reasons: (i) the inter-association equilibrium constant between the OH groups and the POSS units (K A = 38.6) [47,48] being greater than that between the OH and C=O groups (K A = 20) [25]; (ii) strong aggregation of the POSS nanoparticles on the PMMA side chains. We also tested the effect of increased temperature on the hydrogen bonding interactions in the phenolic/PMA-POSS blend ( Figure 12). Again, we did not observe any C=O groups hydrogen bonded with the phenolic resin: only the signal for the self-association of the OH···OH groups shifted to higher wavenumber when we disrupted the strong aggregation of the POSS nanoparticles upon increasing the temperature.   [39,49]; the first is consistent with the size of a POSS unit and was produced by the rhombohedral crystal structure of the POSS units [50], where a d-spacing of 0.49 nm is the average distance between the COOCH 2 groups of the PMA-POSS segments [39]. No X-ray diffraction peaks were discernible for the pure phenolic resin; only an amorphous halo appeared at a value of 2θ of 16.23° in Figure 13a. In contrast, many strong diffraction peaks are evident in Figure 13b,c, indicating highly crystalline structures for pure BPA [51]. Clear signals at a value of 2θ of 7.50° were also present in the diffraction patterns of the phenolic/PMA-POSS and BPA/PMA-POSS blends, indicating that POSS crystals with one dimension approximately equal to 1 nm remained in these blend systems.

WAXD Analyses of PMMA-POSS Homopolymer Blend Systems
In Figure 13b, peaks for crystalline BPA appear in the X-ray diffraction pattern of the BPA/PMA-POSS blend, indicating that the BPA molecules could not dissolve completely in PMA-POSS (phase-separated blends). The results are quite different for the BPA/PMMA blends; the crystallinity of BPA decreased upon increasing of PMMA content (Figure 13c). These WAXD data for BPA blends with PMMA and PMA-POSS are consistent with our DSC data in Figure 5, where the signal for the melting temperature of the BPA component disappeared upon increasing the PMMA content. In contrast, the melting temperatures for the BPA/PMA-POSS blends decreased slightly upon increasing the PMA-POSS content, but plateaued at relatively high PMA-POSS contents. In general, the crystallinity measured using WAXD is an in situ measurement with no thermal history involved in preparing the sample. In contrast, the crystallinity detected by DSC depends on the thermal history; accordingly, we did not observe melting peaks for the BPA/PMMA blends through DSC analyses [52]. We also observed a weak diffraction peak at a value of 2θ of 3.0°, corresponding to a d-spacing of 2.6 nm, in the X-ray diffraction patterns of both the phenolic/PMA-POSS and BPA/PMA-POSS blends; we attribute this signal to the average distance between the main chains of PMA-POSS [39]. Jin et al. reported that PMA-POSS chains are randomly oriented, without any ordering, in solvent-annealed films, whereas thermally-annealed PMA-POSS forms regularly packed molecular cylinders in a helical conformation in an orthorhombic lattice unit cell as a result of crystallization of the POSS moieties; through simulation, they estimated the size of the PMA-POSS chains in the helical conformation to be approximately 2.5 nm [39]. As a result, the strong aggregation through crystallization of the POSS nanoparticles was also enhanced after blending with phenolic resin or BPA, implying that the OH groups of the phenolic or BPA did not have the opportunity to interact with the C=O groups of PMA-POSS (Scheme 2) and, thus, these binary blends appeared as immiscible blend systems during DSC analyses, due to both strong screening effects and the aggregation ability of the POSS nanoparticles. We also investigated the WAXD patterns of phenolic/PMA-POSS and BPA/PMA-POSS at various temperatures ( Figure 14). The intensity of the diffraction peak at a value of 2θ of 3.0° increased upon increasing the temperature, indicating that thermal annealing of the PMA-POSS main chains did indeed enhance the regularly packing as molecular cylinders in a helical conformation in an orthorhombic lattice unit cell. In addition, the crystal peak from BPA disappeared (Figure 14b) when the temperature (180 °C) was above the melting temperature of BPA (162 °C), consistent with DSC analyses in Figure 5b. From the FTIR spectra in Figure 12, we observe that only the signal for the OH···OH self-association shifted to higher wavenumber-the signal for the free C=O groups of PMA-POSS did not change-upon increasing the temperature, again indicating that this phenolic/PMA-POSS blend was an immiscible system in which the OH groups of phenolic did not have the ability to interact with the C=O groups of PMA-POSS. Here, we proposed that the values of γ should be very close to 1 for all of the PVPh/PMA-POSS, phenolic/PMA-POSS, and BPA/PMA-POSS blend systems when the POSS nanoparticles were present on the side chains of PMMA, because strong aggregation of the POSS units induced a strong "screening effect" for the C=O groups of PMA-POSS.

Conclusions
We have prepared a series of blends of PMMA and PMA-POSS with PVPh, phenolic resin, and BPA and investigated their properties using DSC, FTIR spectroscopy, and WAXD. All of these blends of PMMA with PVPh, phenolic, and BPA were totally miscible in the amorphous phase over their entire compositions, as evidenced by a single glass transition temperature (DSC) and intermolecular hydrogen bonding interactions (FTIR spectroscopy) between the C=O groups of PMMA and the OH groups of PVPh, phenolic, and BPA. FTIR spectra revealed the absence of hydrogen bonding interactions between the C=O groups of PMA-POSS (with POSS units present on the side chains) and the OH groups of PVPh, phenolic, and BPA. The strong "screening effect" in these blends induced by the tethered POSS side chains induced immiscibility with the hydrogen bond donor compounds. The strong aggregation, through crystallization, of the POSS nanoparticles (observed in WAXD analyses) was enhanced after blending with phenolic resin or BPA compounds, implying that the OH groups of PVPh, phenolic, and BPA could not interact with the C=O groups of PMA-POSS; accordingly, DSC analyses indicated that these binary blends were immiscible systems.