*2.1. Materials*

PVC (Scheme 1a) and tetrahydrofuran (THF) were purchased from Sigma–Aldrich. DINCH (UniHydro UN899, Scheme 1b)) was supported by UPC Technology Corporation. The average molecular weight of the PVC used in this study was approximately 90,000, with a polydispersity index (PDI) of 2.25 by Sigma–Aldrich. MA-POSS (Scheme 1c) was purchased from Hybrid Plastics; this compound did not have a crystalline structure, but featured a glass transition temperature of −55 ◦C (Figure S1) and was a pale-yellow heavy oil [33,34].

**Scheme 1.** (**a**–**c**) Chemical structures of (**a**) Poly(vinyl chloride) (PVC), (**b**) Dissononyl cyclohexane-1,2-dicarboxylate (DINCH), and (**c**) methacryl-polyhedral oligomeric silsesquioxane (MA-POSS). (**d**) **Scheme 1.** (**a**–**c**) Chemical structures of (**a**) Poly(vinyl chloride) (PVC), (**b**) Dissononyl cyclohexane-1,2 dicarboxylate (DINCH), and (**c**) methacryl-polyhedral oligomeric silsesquioxane (MA-POSS). (**d**) Hydrogen bonding interaction in PVC/MA-POSS blends.

#### Hydrogen bonding interaction in PVC/MA-POSS blends. *2.2. PVC*/*DINCH, PVC*/*MA-POSS, and PVC*/*DINCH*/*MA-POSS Blend Films*

*2.2. PVC/DINCH, PVC/MA-POSS, and PVC/DINCH/MA-POSS Blend Films*  Binary blends of PVC/DINCH and PVC/MA-POSS and ternary blends of PVC/DINCH/MAPOSS were prepared through solvent-blending, using THF as the solvent. A solution containing 10 wt% of the polymer mixture was stirred for 24 h and then the solution was poured into a Teflon mold. The solvent (THF) was evaporated slowly over 3 days at room temperature. The resulting blend film was heated under a vacuum at 50 °C for 2 days and then at 80 °C for 2 days to ensure the total removal of any residual solvent. Samples for DSC, TGA, DMA and tensile testing were pelletized, compression-Binary blends of PVC/DINCH and PVC/MA-POSS and ternary blends of PVC/DINCH/MAPOSS were prepared through solvent-blending, using THF as the solvent. A solution containing 10 wt% of the polymer mixture was stirred for 24 h and then the solution was poured into a Teflon mold. The solvent (THF) was evaporated slowly over 3 days at room temperature. The resulting blend film was heated under a vacuum at 50 ◦C for 2 days and then at 80 ◦C for 2 days to ensure the total removal of any residual solvent. Samples for DSC, TGA, DMA and tensile testing were pelletized, compression-molded into disks (160 ◦C, 3 min), and then machined into suitable specimens for testing.

#### molded into disks (160 °C, 3 min), and then machined into suitable specimens for testing. *2.3. Characterization*

*2.3. Characterization*  The glass transition temperatures of all the blend systems were measured using DSC and DMA. DSC was performed using a TA-Q20 instrument, operated at a heating rate of 20 °C/min between 25 and 150 °C under a N2 atmosphere, and the sample was quickly cooled to −90 °C after the first scan. The *T*g values were determined at the midpoint of the transition temperature, with a scan rate of 20 °C/min, in the range −90–200 °C. The dynamic mechanical properties were characterized using a DuPont 2980 dynamic mechanical analyzer, with the blend sample mounted on the single cantilever clamp. A constant strain amplitude of 2.0% and a frequency of 1 Hz were applied, with a heating rate of 2 °C/min between −60 and +120 °C. The thermal stabilities of the blend samples were determined using a TA Q-50 thermogravimetric analyzer, operating within the temperature range 30–800 °C at a heating rate of 20 °C/min, under N2. FTIR spectra samples of the blend were recorded using a Bruker Tensor 27 spectrophotometer. A total of 32 scans were collected at a spectral resolution of 4 cm−1. The FTIR samples were measured using conventional KBr disk method, similarly to the preparation of the bulk sample. The film used in this work was sufficiently thin to obey the Beer–Lambert law. Tensile tests were performed using an Instron machine. Five different blend samples of each The glass transition temperatures of all the blend systems were measured using DSC and DMA. DSC was performed using a TA-Q20 instrument, operated at a heating rate of 20 ◦C/min between 25 and 150 ◦C under a N<sup>2</sup> atmosphere, and the sample was quickly cooled to −90 ◦C after the first scan. The *T*<sup>g</sup> values were determined at the midpoint of the transition temperature, with a scan rate of 20 ◦C/min, in the range −90–200 ◦C. The dynamic mechanical properties were characterized using a DuPont 2980 dynamic mechanical analyzer, with the blend sample mounted on the single cantilever clamp. A constant strain amplitude of 2.0% and a frequency of 1 Hz were applied, with a heating rate of 2 ◦C/min between −60 and +120 ◦C. The thermal stabilities of the blend samples were determined using a TA Q-50 thermogravimetric analyzer, operating within the temperature range 30–800 ◦C at a heating rate of 20 ◦C/min, under N2. FTIR spectra samples of the blend were recorded using a Bruker Tensor 27 spectrophotometer. A total of 32 scans were collected at a spectral resolution of 4 cm−<sup>1</sup> . The FTIR samples were measured using conventional KBr disk method, similarly to the preparation of the bulk sample. The film used in this work was sufficiently thin to obey the Beer–Lambert law. Tensile tests were performed using an Instron machine. Five different blend samples of each composition were examined at a tensile rate of 1 cm/s.

#### composition were examined at a tensile rate of 1 cm/s. **3. Results**

#### **3. Results**  *3.1. Thermal Analysis of PVC*/*DINCH Blends*

*3.1. Thermal Analysis of PVC/DINCH Blends*  Because DINCH is among the most promising plasticizers used as a replacement for DEHP in Because DINCH is among the most promising plasticizers used as a replacement for DEHP in PVC, we first investigated the thermal properties of PVC/DINCH blends through TGA analyses. Figure 1

PVC, we first investigated the thermal properties of PVC/DINCH blends through TGA analyses.

displays the TGA traces of pure PVC, pure DINCH, and the PVC/DINCH = 100/60 blend, measured from 25 to 800 ◦C at a heating rate of 20 ◦C/min. Pure DINCH displayed its thermal decomposition at a temperature of *T*d10 = 213 ◦C, without providing a char yield, as displayed in Figure 1c. The pure PVC displayed two main thermal decomposition temperatures during the degradation process. The pyrolysis of PVC is complex when compared with those of other thermoplastic polymeric materials; it depends on the type and amount of plasticizer used [1,35]. The first thermal decomposition of PVC was initiated by the elimination of HCl at a relatively low temperature (300 ◦C and *T*d10 = 290 ◦C)—the so-called "dehydrochlorination" stage, during which polyene radicals and unsubstituted aromatics are released, as displayed in Figure S2a. Aromatic products and various degrees of char were formed during the second thermal decomposition stage at approximately 460 ◦C from DTG curves (Figure S2a) [35]. The solid residue of the char yield was approximately 2.7 wt% after completion of the PVC pyrolysis, corresponding to ash formation. After blending with DINCH, the value of *T*d10 was 265 ◦C, because of the lower thermal decomposition temperature of pure DINCH. The char yield increased to 6 wt%, presumably because of ash formation from the reaction of DINCH with PVC. The lower thermal stability of the PVC/DINCH blend system, compared with that of pure PVC, encouraged us to add MA-POSS to potentially improve the thermal stability. Because of the possibility of intermolecular hydrogen bonding between PVC and the MA-POSS nanoparticles, we used DSC, TGA, DMA, and FTIR spectroscopy to perform initial investigations into the miscibility, thermal properties, and hydrogen bonding interactions of binary blends of PVC/MA-POSS. Figure 1 displays the TGA traces of pure PVC, pure DINCH, and the PVC/DINCH = 100/60 blend, measured from 25 to 800 °C at a heating rate of 20 °C/min. Pure DINCH displayed its thermal decomposition at a temperature of *T*d10 = 213 °C, without providing a char yield, as displayed in Figure 1c. The pure PVC displayed two main thermal decomposition temperatures during the degradation process. The pyrolysis of PVC is complex when compared with those of other thermoplastic polymeric materials; it depends on the type and amount of plasticizer used [1,35]. The first thermal decomposition of PVC was initiated by the elimination of HCl at a relatively low temperature (300 °C and *T*d10 = 290 °C)—the so-called "dehydrochlorination" stage, during which polyene radicals and unsubstituted aromatics are released, as displayed in Figure S2a. Aromatic products and various degrees of char were formed during the second thermal decomposition stage at approximately 460 °C from DTG curves (Figure S2a) [35]. The solid residue of the char yield was approximately 2.7 wt% after completion of the PVC pyrolysis, corresponding to ash formation. After blending with DINCH, the value of *T*d10 was 265 °C, because of the lower thermal decomposition temperature of pure DINCH. The char yield increased to 6 wt%, presumably because of ash formation from the reaction of DINCH with PVC. The lower thermal stability of the PVC/DINCH blend system, compared with that of pure PVC, encouraged us to add MA-POSS to potentially improve the thermal stability. Because of the possibility of intermolecular hydrogen bonding between PVC and the MA-POSS nanoparticles, we used DSC, TGA, DMA, and FTIR spectroscopy to perform initial investigations into the miscibility, thermal properties, and hydrogen bonding interactions of binary blends of PVC/MA-POSS.

**Figure 1.** Thermogravimetric analyses (TGA) of (a) the pure PVC, (b) the PVC/DINCH = 100/60 blend, and (c) pure DINCH. **Figure 1.** Thermogravimetric analyses (TGA) of (a) the pure PVC, (b) the PVC/DINCH = 100/60 blend, and (c) pure DINCH.

#### *3.2. Thermal Properties and FTIR Spectra of PVC/MA-POSS Blends 3.2. Thermal Properties and FTIR Spectra of PVC*/*MA-POSS Blends*

Figure 2 presents the results of DSC and DMA analyses of PVC/MA-POSS blends in the range 30–110 °C. The pure PVC exhibited a value of *T*g of 89 °C, based on DSC analysis (Figure 2a), and 94 °C, based on DMA analysis (Figure 2b). It is not unusual for the value of *T*g determined using DMA to be higher than that from DSC analysis, typically by 5–10 °C. In addition, pure MA-POSS has a reported glass transition temperature of −55 °C (Figure S1), and is a pale-yellow heavy oil [33,34]. Figure 2 presents the results of DSC and DMA analyses of PVC/MA-POSS blends in the range 30–110 ◦C. The pure PVC exhibited a value of *T*<sup>g</sup> of 89 ◦C, based on DSC analysis (Figure 2a), and 94 ◦C, based on DMA analysis (Figure 2b). It is not unusual for the value of *T*<sup>g</sup> determined using DMA to be higher than that from DSC analysis, typically by 5–10 ◦C. In addition, pure MA-POSS has a reported glass transition temperature of −55 ◦C (Figure S1), and is a pale-yellow heavy oil [33,34].

*Polymers* **2019**, *11*, x FOR PEER REVIEW 5 of 13

**Figure 2.** (**a**) Differential scanning calorimetry (DSC) and (**b**) dynamic mechanical analysis (DMA) of various PVC/MA-POSS blends; the inset to (**b**) displays the transparent films obtained after DMA analyses. **Figure 2.** (**a**) Differential scanning calorimetry (DSC) and (**b**) dynamic mechanical analysis (DMA) of various PVC/MA-POSS blends; the inset to (**b**) displays the transparent films obtained after DMA analyses.

The values of *T*g, determined using both DSC and DMA, decreased significantly upon increasing the MA-POSS concentration. When 10 wt% of MA-POSS was incorporated into the PVC matrix, the value of *T*g decreased to 63 (DSC) and 68 (DMA) °C, approximately 26 °C lower than that of pure PVC, indicating that the plasticizing effect of MA-POSS towards PVC was significant. Further increasing the MA-POSS content to 20, 30, and 40 wt% caused the values of *T*g to decrease to 64, 58, and 47 °C, respectively, based on DMA analysis, confirming that MA-POSS is a promising plasticizer for PVC, consistent with previous results [33,34]. In addition, all the PVC/MA-POSS blends possessed only a single glass transition temperature, implying that these blends were fully miscible. Furthermore, the values of *T*g were between those of the two individual compounds, as expected. More importantly, all the PVC/MA-POSS blends were transparent and flexible, even after thermal treatment (DMA), as displayed in the inset to Figure 2b. Figure 3 presents the values of *T*g of the PVC/MA-POSS blends, as determined using DMA. Although the glass transition temperatures of the PVC/MA-POSS blends were higher than the values predicted by the Fox rule and the linear rule, they were predicted using the Kwei equation [36]: The values of *T*g, determined using both DSC and DMA, decreased significantly upon increasing the MA-POSS concentration. When 10 wt% of MA-POSS was incorporated into the PVC matrix, the value of *T*<sup>g</sup> decreased to 63 (DSC) and 68 (DMA) ◦C, approximately 26 ◦C lower than that of pure PVC, indicating that the plasticizing effect of MA-POSS towards PVC was significant. Further increasing the MA-POSS content to 20, 30, and 40 wt% caused the values of *T*<sup>g</sup> to decrease to 64, 58, and 47 ◦C, respectively, based on DMA analysis, confirming that MA-POSS is a promising plasticizer for PVC, consistent with previous results [33,34]. In addition, all the PVC/MA-POSS blends possessed only a single glass transition temperature, implying that these blends were fully miscible. Furthermore, the values of *T*<sup>g</sup> were between those of the two individual compounds, as expected. More importantly, all the PVC/MA-POSS blends were transparent and flexible, even after thermal treatment (DMA), as displayed in the inset to Figure 2b. Figure 3 presents the values of *T*<sup>g</sup> of the PVC/MA-POSS blends, as determined using DMA. Although the glass transition temperatures of the PVC/MA-POSS blends were higher than the values predicted by the Fox rule and the linear rule, they were predicted using the Kwei equation [36]:

$$T\_{\mathcal{S}} = \frac{W\_1 T\_{\mathcal{S}1} + k \mathcal{W}\_2 T\_{\mathcal{S}2}}{\mathcal{W}\_1 + k \mathcal{W}\_2} + q \mathcal{W}\_1 \mathcal{W}\_2 \tag{1}$$

1 2 *W kW* + where *W*1 and *W*2 are the weight fractions and *T*g1 and *T*g2 are the glass transition temperatures of the MA-POSS and the PVC, respectively, while *k* and *q* are fitting constants. We obtain values of *k* and *q* of 1 and 50, respectively, for the PVC/MA-POSS blends, indicating that specific intermolecular where *W*<sup>1</sup> and *W*<sup>2</sup> are the weight fractions and *T*g1 and *T*g2 are the glass transition temperatures of the MA-POSS and the PVC, respectively, while *k* and *q* are fitting constants. We obtain values of *k* and *q* of 1 and 50, respectively, for the PVC/MA-POSS blends, indicating that specific intermolecular interactions were occurring between the PVC and MA-POSS nanoparticles.

interactions were occurring between the PVC and MA-POSS nanoparticles.

*Polymers* **2019**, *11*, x FOR PEER REVIEW 6 of 13

**Figure 3.** Experimental values of *T*g of PVC/MA-POSS blends and those predicted by the Fox and linear rules and the Kwei equation. **Figure 3.** Experimental values of *T*<sup>g</sup> of PVC/MA-POSS blends and those predicted by the Fox and linear rules and the Kwei equation.

FTIR spectroscopy can provide information about the noncovalent interactions of both polar and nonpolar groups through hydrogen bonding or dipole–dipole interactions [37–40]. Figure 4 presents the characteristic absorption signals for the (a) C=O units of MA-POSS, (b) C–H deformations of the CH2 units of PVC, and (c) C–Cl stretching vibrations of PVC in various PVC/MA-POSS blends. The free C=O groups of MA-POSS provided an absorption at 1717 cm−1; this signal shifted to 1714 cm−1 at 90 wt% of PVC concentration, as displayed in Figure 4a. In addition, the signal for the CH2 units of PVC shifted from 1332 to 1321 cm−1 (Figure 4b), while its C–Cl units shifted significantly from 614 to 606 cm−1 (Figure 4c), after blending with the MA-POSS nanoparticles. These results indicate that intermolecular hydrogen bonding existed between the H–CCl units of the PVC and the C=O groups of the MA-POSS, as displayed in Scheme 1d. Such weak hydrogen bonding interactions have also been observed in many other binary blend systems [41–43]. FTIR spectroscopy can provide information about the noncovalent interactions of both polar and nonpolar groups through hydrogen bonding or dipole–dipole interactions [37–40]. Figure 4 presents the characteristic absorption signals for the (a) C=O units of MA-POSS, (b) C–H deformations of the CH<sup>2</sup> units of PVC, and (c) C–Cl stretching vibrations of PVC in various PVC/MA-POSS blends. The free C=O groups of MA-POSS provided an absorption at 1717 cm−<sup>1</sup> ; this signal shifted to 1714 cm−<sup>1</sup> at 90 wt% of PVC concentration, as displayed in Figure 4a. In addition, the signal for the CH<sup>2</sup> units of PVC shifted from 1332 to 1321 cm−<sup>1</sup> (Figure 4b), while its C–Cl units shifted significantly from 614 to 606 cm−<sup>1</sup> (Figure 4c), after blending with the MA-POSS nanoparticles. These results indicate that intermolecular hydrogen bonding existed between the H–CCl units of the PVC and the C=O groups of the MA-POSS, as displayed in Scheme 1d. Such weak hydrogen bonding interactions have also been observed in many other binary blend systems [41–43].

Figure 5 displays the TGA curves of various PVC/MA-POSS blends, recorded from 25 to 800 °C at a heating rate of 20 °C/min, and the corresponding DTG curves are displayed in Figure S3. The thermal stability of pure MA-POSS was greater than that of pure PVC, which possessed a value of *T*d10 of 421 °C and a char yield of 44.5 wt%. Upon increasing the content of MA-POSS, the first thermal decomposition temperature and the char yield both increased. When the MA-POSS concentration reached 50 wt%, the value of *T*d10 was close to 306 °C, approximately 16 °C higher than that of pure PVC, while the char yield also increased to 33.7 wt%, higher than the value [23.6 wt% = (44.5 + 2.7)/2] suggested by the simplified relationship of pure MA-POSS (44.5 wt%) + pure PVC (2.7 wt%)—again, probably because of ash formation from the reaction of MA-POSS with PVC. This result suggests that the rate of HCl emission during the first thermal decomposition procedure had decreased as a result of the presence of MA-POSS nanoparticles. The main thermal decomposition of MA-POSS overlapped with the second thermal decomposition process of PVC, making it difficult to identify, as displayed in Figure S3. In addition, the decomposition process of MA-POSS becomes obvious after increasing the MA-POSS concentration, becoming comparable with that of pure MA-POSS. The final step in the TGA curves corresponds to the pyrolysis of conjugated double bonds (e.g., of the polyenes formed in the first step of PVC decomposition), with the decomposition temperature and char yield both increasing upon increasing the MA-POSS concentration, because of the retarding effect of MA-POSS nanoparticles toward the possible residual ash [33,34]. Thus, our DSC, DMA, FTIR spectroscopic, and TGA analyses confirmed that the incorporation of MA-POSS nanoparticles could Figure 5 displays the TGA curves of various PVC/MA-POSS blends, recorded from 25 to 800 ◦C at a heating rate of 20 ◦C/min, and the corresponding DTG curves are displayed in Figure S3. The thermal stability of pure MA-POSS was greater than that of pure PVC, which possessed a value of *T*d10 of 421 ◦C and a char yield of 44.5 wt%. Upon increasing the content of MA-POSS, the first thermal decomposition temperature and the char yield both increased. When the MA-POSS concentration reached 50 wt%, the value of *T*d10 was close to 306 ◦C, approximately 16 ◦C higher than that of pure PVC, while the char yield also increased to 33.7 wt%, higher than the value [23.6 wt% = (44.5 + 2.7)/2] suggested by the simplified relationship of pure MA-POSS (44.5 wt%) <sup>+</sup> pure PVC (2.7 wt%)—again,probably because of ash formation from the reaction of MA-POSS with PVC. This result suggests that the rate of HCl emission during the first thermal decomposition procedure had decreased as a result of the presence of MA-POSS nanoparticles. The main thermal decomposition of MA-POSS overlapped with the second thermal decomposition process of PVC, making it difficult to identify, as displayed in Figure S3. In addition, the decomposition process of MA-POSS becomes obvious after increasing the MA-POSS concentration, becoming comparable with that of pure MA-POSS. The final step in the TGA curves corresponds to the pyrolysis of conjugated double bonds (e.g., of the polyenes formed in the first step of PVC decomposition), with the decomposition temperature and char yield both increasing upon increasing the MA-POSS concentration, because of the retarding effect of MA-POSS nanoparticles toward the possible residual ash [33,34]. Thus, our DSC, DMA, FTIR spectroscopic, and TGA analyses confirmed that the incorporation of MA-POSS nanoparticles could decrease the

decrease the value of *T*g of PVC, and increase both its thermal decomposition temperature and the char yield, while providing full miscibility with PVC, because of weak hydrogen bonding in the value of *T*<sup>g</sup> of PVC, and increase both its thermal decomposition temperature and the char yield, while providing full miscibility with PVC, because of weak hydrogen bonding in the PVC/MA-POSS blends. As a result, MA-POSS has the potential to be a suitable replacement for traditional plasticizers, e.g., DEHP, and could be used as a blood bag, to avoid the DEHP coming into contact with tissues or body fluids. *Polymers* **2019**, *11*, x FOR PEER REVIEW 7 of 13 PVC/MA-POSS blends. As a result, MA-POSS has the potential to be a suitable replacement for traditional plasticizers, e.g., DEHP, and could be used as a blood bag, to avoid the DEHP coming into contact with tissues or body fluids. *Polymers* **2019**, *11*, x FOR PEER REVIEW 7 of 13 PVC/MA-POSS blends. As a result, MA-POSS has the potential to be a suitable replacement for traditional plasticizers, e.g., DEHP, and could be used as a blood bag, to avoid the DEHP coming into contact with tissues or body fluids.

**Figure 4.** FTIR spectral analyses of the (**a**) C=O groups of MA-POSS, (**b**) C–H deformation of the CH2 units of the PVC, and (**c**) C–Cl stretching vibration of the PVC, for various PVC/MA-POSS blends recorded at room temperature. **Figure 4.** FTIR spectral analyses of the (**a**) C=O groups of MA-POSS, (**b**) C–H deformation of the CH<sup>2</sup> units of the PVC, and (**c**) C–Cl stretching vibration of the PVC, for various PVC/MA-POSS blends recorded at room temperature. **Figure 4.** FTIR spectral analyses of the (**a**) C=O groups of MA-POSS, (**b**) C–H deformation of the CH2 units of the PVC, and (**c**) C–Cl stretching vibration of the PVC, for various PVC/MA-POSS blends recorded at room temperature.

**Figure 5.** TGA analyses of PVC/MA-POSS blends of various compositions. **Figure 5. Figure 5.**  TGA analyses of PVC TGA analyses of PVC/MA-POSS blends of various compositions. /MA-POSS blends of various compositions.

#### *3.3. Thermal and Mechanical Properties of PVC*/*DINCH*/*MA-POSS Ternary Blends 3.3. Thermal and Mechanical Properties of PVC/DINCH/MA-POSS Ternary Blends*

To solve the problem of the lower thermal stability of PVC/DINCH blend systems compared to pure PVC, we tested the effect of adding MA-POSS nanoparticles. Figure 6 presents DMA analyses of PVC/DINCH/MA-POSS ternary blends at various MA-POSS contents. The value of *T*<sup>g</sup> of the binary PVC/DINCH blend was 13.4 ◦C. Increasing the MA-POSS content to 5, 10, and 15 phr caused the value of *T*<sup>g</sup> to decrease to 6.6, 4.7, and 2.6 ◦C, respectively, based on DMA analysis, suggesting that MA-POSS remains a promising plasticizer for PVC/DINCH blends. In addition, we observed transparent behavior for all the PVC/DINCH/MA-POSS ternary blends, with flexibility, even after thermal treatment through DMA analysis (see inset to Figure 6). To solve the problem of the lower thermal stability of PVC/DINCH blend systems compared to pure PVC, we tested the effect of adding MA-POSS nanoparticles. Figure 6 presents DMA analyses of PVC/DINCH/MA-POSS ternary blends at various MA-POSS contents. The value of *T*g of the binary PVC/DINCH blend was 13.4 °C. Increasing the MA-POSS content to 5, 10, and 15 phr caused the value of *T*g to decrease to 6.6, 4.7, and 2.6 °C, respectively, based on DMA analysis, suggesting that MA-POSS remains a promising plasticizer for PVC/DINCH blends. In addition, we observed transparent behavior for all the PVC/DINCH/MA-POSS ternary blends, with flexibility, even after thermal treatment through DMA analysis (see inset to Figure 6).

*Polymers* **2019**, *11*, x FOR PEER REVIEW 8 of 13

**Figure 6.** DMA analyses of various PVC/DINCH/MA-POSS blends: (**a**) 100/60/0, (**b**) 100/60/5, (**c**) 100/60/10, and (**d**) 100/60/15. The inset displays the transparent films obtained after DMA analysis of each composition. **Figure 6.** DMA analyses of various PVC/DINCH/MA-POSS blends: (**a**) 100/60/0, (**b**) 100/60/5, (**c**) 100/60/10, and (**d**) 100/60/15. The inset displays the transparent films obtained after DMA analysis of each composition.

Figure 7 presents the TGA curves of PVC/DINCH/MA-POSS ternary blends at various MA-POSS contents, measured from 25 to 800 °C at a heating rate of 20 °C/min. The corresponding DTG curves are also displayed in Figure S4. Increasing the MA-POSS content in the PVC/DINCH binary blend caused the first thermal decomposition temperature and the char yield to increase, as displayed in Figure S4. When the MA-POSS content reached 15 phr, the value of *T*d was close to 271 °C, approximately 6 °C higher than that of the PVC/DINCH binary blend. Furthermore, the char yield also increased to 17.3 wt%, 11.3 wt% higher than that of the PVC/DINCH binary blend. Figure 7 presents the TGA curves of PVC/DINCH/MA-POSS ternary blends at various MA-POSS contents, measured from 25 to 800 ◦C at a heating rate of 20 ◦C/min. The corresponding DTG curves are also displayed in Figure S4. Increasing the MA-POSS content in the PVC/DINCH binary blend caused the first thermal decomposition temperature and the char yield to increase, as displayed in Figure S4. When the MA-POSS content reached 15 phr, the value of *T*<sup>d</sup> was close to 271 ◦C, approximately 6 ◦C higher than that of the PVC/DINCH binary blend. Furthermore, the char yield also increased to 17.3 wt%, 11.3 wt% higher than that of the PVC/DINCH binary blend.

*Polymers* **2019**, *11*, x FOR PEER REVIEW 9 of 13

*Polymers* **2019**, *11*, x FOR PEER REVIEW 9 of 13

**Figure 7.** TGA analyses of PVC/DINCH/MA-POSS blends of various compositions. **Figure 7.** TGA analyses of PVC/DINCH/MA-POSS blends of various compositions. **Figure 7.** TGA analyses of PVC/DINCH/MA-POSS blends of various compositions.

Figure 8 displays the results of tensile tests of PVC/DINCH/MA-POSS ternary blends at various MA-POSS contents (a) before and (b) after thermal annealing at 100 °C for 7 days. All the PVC/DINCH/MA-POSS blends exhibited high elongation behavior, even after thermal annealing, and we could briefly observe that the modulus and tensile strength decreased, and elongation increased, upon increasing the content of MA-POSS, both before and after thermal annealing. Figure 8 displays the results of tensile tests of PVC/DINCH/MA-POSS ternary blends at various MA-POSS contents (a) before and (b) after thermal annealing at 100 ◦C for 7 days. All the PVC/DINCH/MA-POSS blends exhibited high elongation behavior, even after thermal annealing, and we could briefly observe that the modulus and tensile strength decreased, and elongation increased, upon increasing the content of MA-POSS, both before and after thermal annealing. Figure 8 displays the results of tensile tests of PVC/DINCH/MA-POSS ternary blends at various MA-POSS contents (a) before and (b) after thermal annealing at 100 °C for 7 days. All the PVC/DINCH/MA-POSS blends exhibited high elongation behavior, even after thermal annealing, and we could briefly observe that the modulus and tensile strength decreased, and elongation increased, upon increasing the content of MA-POSS, both before and after thermal annealing.

**Figure 8.** Stress–strain curves of PVC/DINCH/MA-POSS blends at various MA-POSS contents (**a**) before and (**b**) after thermal annealing at 100 °C for 7 days. The inset displays the tensile test samples used in this study (**a**) before and (**b**) after thermal annealing. **Figure 8.** Stress–strain curves of PVC/DINCH/MA-POSS blends at various MA-POSS contents (**a**) before and (**b**) after thermal annealing at 100 °C for 7 days. The inset displays the tensile test samples used in this study (**a**) before and (**b**) after thermal annealing. **Figure 8.** Stress–strain curves of PVC/DINCH/MA-POSS blends at various MA-POSS contents (**a**) before and (**b**) after thermal annealing at 100 ◦C for 7 days. The inset displays the tensile test samples used in this study (**a**) before and (**b**) after thermal annealing.

Figure 9 summarizes the mechanical properties. Elongation decreased, but the modulus and tensile strength increased, after thermal annealing. We also observed transparency for all the PVC/DINCH/MA-POSS ternary blends, as well as flexibility. They became pale yellow after thermal Figure 9 summarizes the mechanical properties. Elongation decreased, but the modulus and tensile strength increased, after thermal annealing. We also observed transparency for all the Figure <sup>9</sup> summarizes the mechanical properties. Elongation decreased, but the modulus andtensile strength increased, after thermal annealing. We also observed transparency for all the

PVC/DINCH/MA-POSS ternary blends, as well as flexibility. They became pale yellow after thermal

annealing at 100 °C for 7 days, as displayed in Figure 8.

annealing at 100 °C for 7 days, as displayed in Figure 8.

PVC/DINCH/MA-POSS ternary blends, as well as flexibility. They became pale yellow after thermal annealing at 100 ◦C for 7 days, as displayed in Figure 8. *Polymers* **2019**, *11*, x FOR PEER REVIEW 10 of 13

**Figure 9.** Mechanical properties of PVC/DINCH/MA-POSS blends at various MA-POSS contents (**a**,**c**,**e**) before and (**b**,**d**,**f**) after thermal annealing at 100 °C for 7 days: (**a**,**b**) elongation properties, (**b**,**d**) moduli, and (**c**,**f**) tensile strengths. **Figure 9.** Mechanical properties of PVC/DINCH/MA-POSS blends at various MA-POSS contents (**a**,**c**,**e**) before and (**b**,**d**,**f**) after thermal annealing at 100 ◦C for 7 days: (**a**,**b**) elongation properties, (**b**,**d**) moduli, and (**c**,**f**) tensile strengths.

#### **4. Conclusions 4. Conclusions**

We have examined the thermal properties, miscibility, hydrogen bonding interactions, and mechanical properties of PVC/MA-POSS binary blends and PVC/DINCH/MA-POSS ternary blends. Employing DSC, DMA, TGA, and tensile tests, we confirmed that the incorporation of MA-POSS nanoparticles into the blends occurred with full miscibility; they decreased the value of *T*g, the modulus, and tensile strength, but increased the thermal decomposition temperature, char yield, and elongation properties, because weak hydrogen bonding existed between the C=O groups of the MA-POSS and the H–CCl units of the PVC, as determined through FTIR spectral analyses. As a result, MA-POSS appears to be a suitable candidate for replacing traditional plasticizers, e.g., DEHP, in PVC We have examined the thermal properties, miscibility, hydrogen bonding interactions, and mechanical properties of PVC/MA-POSS binary blends and PVC/DINCH/MA-POSS ternary blends. Employing DSC, DMA, TGA, and tensile tests, we confirmed that the incorporation of MA-POSS nanoparticles into the blends occurred with full miscibility; they decreased the value of *T*g, the modulus, and tensile strength, but increased the thermal decomposition temperature, char yield, and elongation properties, because weak hydrogen bonding existed between the C=O groups of the MA-POSS and the H–CCl units of the PVC, as determined through FTIR spectral analyses. As a result, MA-POSS appears to be a suitable candidate for replacing traditional plasticizers, e.g., DEHP, in PVC matrices.

matrices. **Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4360/11/10/1711/s1.

**Supplementary Materials:** The following are available online at www.mdpi.com/xxx/s1.

**Author Contributions:** Y.-K.W. did the experiment of Figures 1–7, C.-C.M. and M.-L.W. did the experimental of Figures 8 and 9, F.-C.T., and S.-W.K. contributed to the literature review and to the writing of this paper.

**Funding:** This research was funded by the Ministry of Science and Technology, Taiwan, under contracts MOST 106-2221-E-110-067-MY3, MOST 105-2221-E-110-092-MY3, industry-academia-research cooperation project between UPC and NSYSU, and also supported by the National Natural Science Foundation of China (Grants U1805253 and 51773214).

**Conflicts of Interest:** The authors declare no conflict of interest.
