Effect of Silver Nanoparticles on the Microstructure, Non-Isothermal Crystallization Behavior and Antibacterial Activity of Polyoxymethylene

Silver (Ag) nanoparticles were synthesized by a facile route in the presence of oleic acid and n-propylamine. It was shown that the average primary size of the as-synthesized Ag nanoparticles was approximately 10 nm and the surface of as-synthesized Ag nanoparticles was capped with monolayer surfactants with the content of 19.6%. Based on as-synthesized Ag nanoparticles, polyoxymethylene (POM)/Ag nanocomposites were prepared. The influence of Ag nanoparticles on non-isothermal crystallization behavior of POM was investigated by differential scanning calorimetry (DSC). The Jeziorny, Jeziorny-modified Avrami, Ozawa, Liu and Mo, Ziabicki and Kissinger models were applied to analyze the non-isothermal melt crystallization data of POM/Ag nanocomposites. Results of half time (t1/2), crystallization rate parameter (CRP), crystallization rate function (K(T)), kinetic parameter (F(T)), the kinetic crystallizability at unit cooling rate (GZ) and the crystallization activation energy (∆E) were determined. Small amounts of Ag nanoparticles dispersed into POM matrix were shown to act as heterogeneous nuclei, which could enhance the crystallization rate of POM, increase the number of POM spherulites and reduce POM spherulites size. However, the higher loading of Ag nanoparticles were easily aggregated, which restrained POM crystallization to some degree. Furthermore, the POM/Ag nanocomposites showed robust antibacterial activity against Escherichia coli and Staphylococcus aureus.


Introduction
Ag nanoparticles have received significant attention owing to their unique volume effect and quantum size as well as their high conductivity, excellent catalytic performance and broad spectrum of antimicrobial activities [1,2]. Ag nanoparticles have been extensively used in nonaqueous conductive ink [3], hydrogen sorption and storage [4], antibacterial materials [5], bone tissue regeneration and wound repair [6,7], etc. Thus, polymer/Ag nanocomposites have widely been applied in a variety of areas, such as the microelectronics, optoelectronics, magnetic materials, catalysis, chemical sensors and biology [8][9][10]. One of the interesting studies was that of the effect of polyethylene (PE)/Ag nanocomposites on Ag ion release and antimicrobial properties by Zapata et al. [11]. It was found was added. The mixture was held at 50 °C and stirred for 3 h, and a dark brown organic Ag nanoparticle colloidal solution was obtained. After that, ethanol (500 mL) was added into the solution to precipitate the crude Ag nanoparticles. Finally, the precipitate was isolated by filtration, washed twice with acetone and vacuum-dried at 50 °C for 48 h to afford the final products in the form of blue powders. Synthesis procedure of Ag nanoparticles is shown in Figure 1.

Preparation of POM/Ag Nanocomposites
POM and the Ag nanoparticles were vacuum-dried at 80 °C for 24 h before use. Then POM/Ag nanocomposites were prepared by melt blending method using a torque rheometer. The contents of Ag nanoparticles were fixed at 0.1, 0.5, 1 and 2 wt%. The roller speed was 60 rpm and the process temperature was maintained at 180 °C. The mixing process lasted for 10 min to achieve homogeneous blending and was used for preparation of the POM/Ag nanocomposites.

Fourier Transform Infrared Spectroscopy (FTIR)
FTIR measurements were performed on an infrared spectrometer (Spectrum 100, PerkinElmer, Shanghai, China). All spectra were measured within the 500-4000 cm −1 region with KBr pellets.

Transmission Electron Microscopy (TEM)
The morphology of the Ag nanoparticles was investigated by TEM (Tecnai G2 F30, FEI, Hillsboro, OR, USA). All TEM samples were prepared by placing a drop of Ag nanoparticles in n-heptane onto an amorphous carbon-coated copper grid. The size distribution of the Ag nanoparticles was analyzed using Image J software.

Scanning Electron Microscopy (SEM)
Morphological and microstructural features of POM/Ag nanocomposites were investigated by SEM (Quanta 450, FEI, Hillsboro, OR, USA), and the cross-sectioned samples were prepared by fracturing the POM/Ag nanocomposites in liquid nitrogen. The Ag nanoparticles in the POM/Ag nanocomposites were also characterized by energy dispersive X-ray spectroscopy (EDS) analysis on the sample images.

Thermogravimetric Analysis (TGA)
Thermal stability experiments were performed by using a TG analyzer (Pyris I, PerkinElmer, Shanghai, China) under a nitrogen environment at a heating rate of 10 °C/min from 25 to 650 °C.

Preparation of POM/Ag Nanocomposites
POM and the Ag nanoparticles were vacuum-dried at 80 • C for 24 h before use. Then POM/Ag nanocomposites were prepared by melt blending method using a torque rheometer. The contents of Ag nanoparticles were fixed at 0.1, 0.5, 1 and 2 wt%. The roller speed was 60 rpm and the process temperature was maintained at 180 • C. The mixing process lasted for 10 min to achieve homogeneous blending and was used for preparation of the POM/Ag nanocomposites.

Fourier Transform Infrared Spectroscopy (FTIR)
FTIR measurements were performed on an infrared spectrometer (Spectrum 100, PerkinElmer, Shanghai, China). All spectra were measured within the 500-4000 cm −1 region with KBr pellets.

Transmission Electron Microscopy (TEM)
The morphology of the Ag nanoparticles was investigated by TEM (Tecnai G2 F30, FEI, Hillsboro, OR, USA). All TEM samples were prepared by placing a drop of Ag nanoparticles in n-heptane onto an amorphous carbon-coated copper grid. The size distribution of the Ag nanoparticles was analyzed using Image J software.

Scanning Electron Microscopy (SEM)
Morphological and microstructural features of POM/Ag nanocomposites were investigated by SEM (Quanta 450, FEI, Hillsboro, OR, USA), and the cross-sectioned samples were prepared by fracturing the POM/Ag nanocomposites in liquid nitrogen. The Ag nanoparticles in the POM/Ag nanocomposites were also characterized by energy dispersive X-ray spectroscopy (EDS) analysis on the sample images.

Thermogravimetric Analysis (TGA)
Thermal stability experiments were performed by using a TG analyzer (Pyris I, PerkinElmer, Shanghai, China) under a nitrogen environment at a heating rate of 10 • C/min from 25 to 650 • C.

X-ray Diffraction (XRD)
The Ag nanoparticles were examined by using an X-ray diffractometer (D8 Advance, Bruker, Beijing, China) with Cu-Kα radiation. The diffractograms were scanned in a 2θ range from 20 • to 90 • at a scanning speed of 6 • /min. The generator was operated at 40 KV and 200 mA.

Differential Scanning Calorimetry (DSC)
The DSC measurements of POM/Ag nanocomposites were performed on a DSC 204F1 (Netzsch, Selb, Germany). Samples weighing about 5-8 mg in an aluminum crucible were heated from 30 to 200 • C with the heating rate of 10 • C/min and kept at 200 • C for 5 min to remove the thermal history, then cooled from 200 to 30 • C with the cooling rates of 5, 10, 20, 30 and 40 • C/min. Degree of crystallinity (X C ) was calculated by the Equation (1) to examine the crystallinity change caused by the addition of the Ag nanoparticles [24]: where ∆H m is the enthalpy of melting in the second heating scan of the samples (J/g), ∆H o m is the enthalpy value of melting of a 100% crystalline form of matrix polymer and α is the weight fraction of Ag nanoparticles. The ∆H o m value of POM was 326 J/g [25].

Polarized Optical Microscopy
Polarized optical microscopy was performed to observe the crystalline morphology of POM/Ag nanocomposites on a polarizing microscope (Axio Imager A2, Zeiss, Shanghai, China) equipped with a digital camera (Axiocam Mrc 5, Zeiss, Shanghai, China). Each sample was sandwiched between two thin glass sides, kept at 180 • C for 5 min to remove any thermal history on the hot stage, then cooled to 25 • C at a rate of 10 • C/min. The photographs were then taken.

Antibacterial Tests
The paper-disk diffusion method was employed to assay the antibacterial activities of the POM/Ag nanocomposites. Escherichia coli (ATCC: 9739) and Staphylococcus aureus (ATCC: 6538) were selected as bacterial models. The medium used for growing and maintaining the bacterial liquid cultures was Luria-Bertani (LB) medium, and a solid medium was obtained by adding agar into the liquid medium. The POM/Ag nanocomposites were uniformly placed onto the filter paper discs (2 cm in diameter) and dispersed on the agar plates with either E. coli (5 mL, 10 5 colony forming units (CFU) per mL) or S. aureus (5 mL, 10 5 CFU per mL). The diameters of the inhibition zones were measured after incubation at 37 • C for 24 h.
In order to further elucidate the antibacterial efficiency of the POM/Ag nanocomposites and the inhibition rates of the POM/Ag nanocomposites with respect to the different Ag nanoparticle quantities, the bacterial densities of activated E. coli and S. aureus were first diluted from 10 8 CFU/mL to 10 6 CFU/mL with broth. Then, the POM/Ag nanocomposites were mixed with 50 mL of bacterial suspension and incubated at 37 • C for 24 h. UV absorbance (OD) of the strain broth at 630 nm was measured and used for the calculation of inhibition rate. The inhibition rate was calculated by the Equation (2) [26], where A 0 corresponds to the OD value of the culture broth before culture; A t represents the OD value of the test samples and A con is the OD value of the mixed solution of broth and saline after incubation for 24 h: Polymers 2020, 12, 424 5 of 22 Figure 2a shows the TEM images of Ag nanoparticles. It was clear that each Ag nanoparticle was well separated from the neighboring ones, indicating that the Ag nanoparticles were well surface-passivated by the stabilization action of the n-propylamine and oleic acid. Again, it could be found that the average size of Ag nanoparticles was mainly 10 nm, with a narrow size distribution from the corresponding size distribution of Ag nanoparticles (Figure 2a inset). In addition, the XRD pattern of the Ag nanoparticles is shown in Figure 2b. It was in good agreement with the literature value for the crystal structure of Ag nanoparticles [27]. It was believed that the prominent peaks of 2θ values at 38.1 • , 42.1 • , 64.5 • and 77.5 • represented the 111, 200, 220 and 311 Bragg's reflections of the face-centered cubic crystal structure of Ag. Furthermore, the FTIR spectra of oleic acid, n-propylamine and Ag nanoparticles are shown in Figure 3a. Two absorption peaks were observed at 790 cm −1 and 1623 cm −1 on the spectrum of the n-propylamine, which were attributed to the N-H vibration peak. Meanwhile, in the FTIR spectrum of pure oleic acid, the broad peak of 2500-3600 cm −1 belonged to the O-H stretching vibration peak in the carboxyl group, and 2970 cm −1 and 3005 cm −1 corresponded to the C-H stretching vibration peak beside the double bond and the C-H stretching vibration peak in the CH 2 , respectively [28,29]. As for the spectrum of Ag nanoparticles, this showed the characteristic adsorption peaks at 2890 cm −1 and 2950 cm −1 , which were attributed to C-H stretching vibration band of oleic acid. This indicated that the oleic acid molecules adsorbed on the surface of Ag nanoparticles changed the solid field of the Ag nanoparticles, and the blue shift of characteristic adsorption peak of C-H stretching vibration occurred [30]. In addition, the spectrum of Ag nanoparticles also showed the absorption band at 1560 cm −1 assigned to -COOstretching vibrations, which proved that the oleic acid and n-propylamine reacted and coated on the surface of the Ag nanoparticles. These results indicated that the surface of the Ag nanoparticles was a layer of surface-active agent which was composed of oleic acid and n-propylamine. Finally, the thermal analysis curves of Ag nanoparticles are shown in Figure 3b. Only a one-step mass loss process was observed in the temperature range from 200-400 • C, corresponding to the decomposition of the monolayer complex reacted by oleic acid and the n-propylamine, revealing the monolayer surfactant-capped structure of the Ag nanoparticles. In addition, the content of the monolayer surfactant on the surface of Ag nanoparticles was 19.6%.

Characterization of Ag Nanoparticles
Polymers 2020, 12, x FOR PEER REVIEW 5 of 21 Figure 2a shows the TEM images of Ag nanoparticles. It was clear that each Ag nanoparticle was well separated from the neighboring ones, indicating that the Ag nanoparticles were well surface-passivated by the stabilization action of the n-propylamine and oleic acid. Again, it could be found that the average size of Ag nanoparticles was mainly 10 nm, with a narrow size distribution from the corresponding size distribution of Ag nanoparticles (Figure 2a inset). In addition, the XRD pattern of the Ag nanoparticles is shown in Figure 2b. It was in good agreement with the literature value for the crystal structure of Ag nanoparticles [27]. It was believed that the prominent peaks of 2θ values at 38.1°, 42.1°, 64.5° and 77.5° represented the 111, 200, 220 and 311 Bragg's reflections of the face-centered cubic crystal structure of Ag. Furthermore, the FTIR spectra of oleic acid, n-propylamine and Ag nanoparticles are shown in Figure 3a. Two absorption peaks were observed at 790 cm −1 and 1623 cm −1 on the spectrum of the n-propylamine, which were attributed to the N-H vibration peak. Meanwhile, in the FTIR spectrum of pure oleic acid, the broad peak of 2500 -3600 cm −1 belonged to the O-H stretching vibration peak in the carboxyl group, and 2970 cm −1 and 3005 cm −1 corresponded to the C-H stretching vibration peak beside the double bond and the C-H stretching vibration peak in the CH2, respectively [28,29]. As for the spectrum of Ag nanoparticles, this showed the characteristic adsorption peaks at 2890 cm −1 and 2950 cm −1 , which were attributed to C-H stretching vibration band of oleic acid. This indicated that the oleic acid molecules adsorbed on the surface of Ag nanoparticles changed the solid field of the Ag nanoparticles, and the blue shift of characteristic adsorption peak of C-H stretching vibration occurred [30]. In addition, the spectrum of Ag nanoparticles also showed the absorption band at 1560 cm −1 assigned to -COOstretching vibrations, which proved that the oleic acid and n-propylamine reacted and coated on the surface of the Ag nanoparticles. These results indicated that the surface of the Ag nanoparticles was a layer of surface-active agent which was composed of oleic acid and n-propylamine. Finally, the thermal analysis curves of Ag nanoparticles are shown in Figure 3b. Only a one-step mass loss process was observed in the temperature range from 200-400 °C, corresponding to the decomposition of the monolayer complex reacted by oleic acid and the n-propylamine, revealing the monolayer surfactant-capped structure of the Ag nanoparticles. In addition, the content of the monolayer surfactant on the surface of Ag nanoparticles was 19.6%.

Microstructure and Morphology of POM/Ag Nanocomposites
SEM images of cross-section of POM/Ag nanocomposites are shown in Figure 4. Figure 4a shows the SEM image of pure POM; it can be observed that there were no nanoparticles in the POM matrix. However, it was observed that the Ag nanoparticles exhibited an average particle size of 200 nm dispersed in the POM matrix, as shown in Figure 4b-d. In addition, when the content of Ag nanoparticles was less than 1 wt %, Ag nanoparticles were well dispersed in the POM matrix, which was due to the stabilization action of the monolayer surfactants on the surface of the Ag nanoparticles. Nevertheless, when the content of Ag nanoparticles reached 2 wt % (Figure 4e), Ag nanoparticles would be easily aggregated together, which resulted in poor dispersion of Ag nanoparticles [31]. Furthermore, the Ag nanoparticles were characterized by EDS (Figure 4f). This proved that the chemical composition of Ag nanoparticles consisted of Ag atoms.

Melting and Non-Isothermal Crystallization Behavior of POM/Ag Nanocomposites
The non-isothermal crystallization behavior of the POM/Ag nanocomposites was studied and shown in Figure 5. The crystallization peak temperature (TP) and relative degree of crystallinity (XC) were adopted to describe the non-isothermal crystallization behavior of POM/Ag nanocomposites  Figure 4. Figure 4a shows the SEM image of pure POM; it can be observed that there were no nanoparticles in the POM matrix. However, it was observed that the Ag nanoparticles exhibited an average particle size of 200 nm dispersed in the POM matrix, as shown in Figure 4b-d. In addition, when the content of Ag nanoparticles was less than 1 wt %, Ag nanoparticles were well dispersed in the POM matrix, which was due to the stabilization action of the monolayer surfactants on the surface of the Ag nanoparticles. Nevertheless, when the content of Ag nanoparticles reached 2 wt % (Figure 4e), Ag nanoparticles would be easily aggregated together, which resulted in poor dispersion of Ag nanoparticles [31]. Furthermore, the Ag nanoparticles were characterized by EDS (Figure 4f). This proved that the chemical composition of Ag nanoparticles consisted of Ag atoms.

Microstructure and Morphology of POM/Ag Nanocomposites
SEM images of cross-section of POM/Ag nanocomposites are shown in Figure 4. Figure 4a shows the SEM image of pure POM; it can be observed that there were no nanoparticles in the POM matrix. However, it was observed that the Ag nanoparticles exhibited an average particle size of 200 nm dispersed in the POM matrix, as shown in Figure 4b-d. In addition, when the content of Ag nanoparticles was less than 1 wt %, Ag nanoparticles were well dispersed in the POM matrix, which was due to the stabilization action of the monolayer surfactants on the surface of the Ag nanoparticles. Nevertheless, when the content of Ag nanoparticles reached 2 wt % (Figure 4e), Ag nanoparticles would be easily aggregated together, which resulted in poor dispersion of Ag nanoparticles [31]. Furthermore, the Ag nanoparticles were characterized by EDS (Figure 4f). This proved that the chemical composition of Ag nanoparticles consisted of Ag atoms.

Melting and Non-Isothermal Crystallization Behavior of POM/Ag Nanocomposites
The non-isothermal crystallization behavior of the POM/Ag nanocomposites was studied and shown in Figure 5. The crystallization peak temperature (TP) and relative degree of crystallinity (XC) were adopted to describe the non-isothermal crystallization behavior of POM/Ag nanocomposites

Melting and Non-Isothermal Crystallization Behavior of POM/Ag Nanocomposites
The non-isothermal crystallization behavior of the POM/Ag nanocomposites was studied and shown in Figure 5. The crystallization peak temperature (T P ) and relative degree of crystallinity (X C ) were adopted to describe the non-isothermal crystallization behavior of POM/Ag nanocomposites (Tables 1 and 2). It was found that the T P shifted to lower temperature with the increase of the cooling rate. This may be attributed to that when the cooling rates increased, the time of the motion and rearrangement of POM macromolecular chains decreased, which made the formation of nuclei difficult. More interestingly, it was found that the T P and X C of POM/Ag nanocomposites were higher than that of the POM at the same cooling rate. This could be attributed to the fact that the efficient heterogeneous nucleation of Ag nanoparticles could enhance the crystallization rate of POM and increase the crystallinity of POM to a higher level. As the amount of Ag nanoparticles increased, the T P of the POM/Ag nanocomposites increased slightly [32]. However, when the amount of Ag nanoparticles was 2 wt%, the X C of the POM/Ag nanocomposites was lower than that of the POM/Ag nanocomposites with 1 wt% Ag nanoparticles, indicating that a high loading of Ag nanoparticles in POM/Ag nanocomposites could inhibit their crystallization to some extent [33]. This is because more Ag nanoparticles would be easily aggregated together in the POM matrix, which could affect the heterogeneous nucleation of Ag nanoparticles and crystallization performances of the POM. This was in agreement with the results from the SEM analysis.
Polymers 2020, 12, x FOR PEER REVIEW 7 of 21 (Tables 1 and 2). It was found that the TP shifted to lower temperature with the increase of the cooling rate. This may be attributed to that when the cooling rates increased, the time of the motion and rearrangement of POM macromolecular chains decreased, which made the formation of nuclei difficult. More interestingly, it was found that the TP and XC of POM/Ag nanocomposites were higher than that of the POM at the same cooling rate. This could be attributed to the fact that the efficient heterogeneous nucleation of Ag nanoparticles could enhance the crystallization rate of POM and increase the crystallinity of POM to a higher level. As the amount of Ag nanoparticles increased, the TP of the POM/Ag nanocomposites increased slightly [32]. However, when the amount of Ag nanoparticles was 2 wt%, the XC of the POM/Ag nanocomposites was lower than that of the POM/Ag nanocomposites with 1 wt% Ag nanoparticles, indicating that a high loading of Ag nanoparticles in POM/Ag nanocomposites could inhibit their crystallization to some extent [33]. This is because more Ag nanoparticles would be easily aggregated together in the POM matrix, which could affect the heterogeneous nucleation of Ag nanoparticles and crystallization performances of the POM. This was in agreement with the results from the SEM analysis.

Jeziorny Model Analysis
The calculation of the relative degree of crystallinity (X(T)) is displayed in Equation (3) [34]; the relationship between time (t) and temperature (T) was as given in Equation (4): where T 0 and T correspond to the onset and an arbitrary temperature, respectively. dH C is the enthalpy of crystallization released during an infinitesimal temperature range dT, ∆H C is the overall enthalpy of crystallization for a certain cooling rate, and Φ is the cooling rate. The relative degree of crystallinity (X(T)) with temperature and time are shown in Figures 6 and 7, respectively. In addition, the half-time of crystallization (t 1/2 ) values are given in Table 2. On one hand, the t 1/2 values decreased with the increasing cooling rate, indicating that all samples crystallized faster when the cooling rate increased. On the other hand, in a fixed cooling rate, as the content of Ag nanoparticles increased, the t 1/2 values decreased first, then increased. The t 1/2 value of POM/Ag nanocomposites with 1 wt% Ag nanoparticles was smallest. It was demonstrated that when the content of Ag nanoparticles was less than 1 wt%, the Ag nanoparticles enhanced the crystallization rate of POM due to their nucleation effect. However, when the content of Ag nanoparticles was reached 2 wt%, Ag nanoparticles easily aggregated together, which had a certain inhibitory action on POM crystallization.
Polymers 2020, 12, x FOR PEER REVIEW 9 of 21 when the content of Ag nanoparticles was less than 1 wt%, the Ag nanoparticles enhanced the crystallization rate of POM due to their nucleation effect. However, when the content of Ag nanoparticles was reached 2 wt%, Ag nanoparticles easily aggregated together, which had a certain inhibitory action on POM crystallization.   The crystallization rate parameter (CRP) was employed to quantitatively compare the non-isothermal crystallization rates of POM/Ag nanocomposites, which could be evaluated from the function of 1/t1/2 and Φ [35]. The plot of 1/t1/2 versus Φ is shown in Figure 8, and the CPR values are given in Table 2. The CRP values of POM/Ag nanocomposites with 0, 0.1, 0.5, 1 and 2 wt % Ag nanoparticles were 0.0575, 0.0717, 0.0764, 0.0805 and 0.0792, respectively. These CRP values clearly suggest that the crystallization rate of POM was enhanced by the Ag nanoparticles when the amount of Ag nanoparticles was less than 1 wt %, due to their nucleation effect. However, when the content of Ag nanoparticles was high, they had a certain inhibitory effect on the crystallization of POM. The crystallization rate parameter (CRP) was employed to quantitatively compare the non-isothermal crystallization rates of POM/Ag nanocomposites, which could be evaluated from the function of 1/t 1/2 and Φ [35]. The plot of 1/t 1/2 versus Φ is shown in Figure 8, and the CPR values are given in Table 2. The CRP values of POM/Ag nanocomposites with 0, 0.1, 0.5, 1 and 2 wt % Ag nanoparticles were 0.0575, 0.0717, 0.0764, 0.0805 and 0.0792, respectively. These CRP values clearly suggest that the crystallization rate of POM was enhanced by the Ag nanoparticles when the amount of Ag nanoparticles was less than 1 wt %, due to their nucleation effect. However, when the content of Ag nanoparticles was high, they had a certain inhibitory effect on the crystallization of POM. function of 1/t1/2 and Φ [35]. The plot of 1/t1/2 versus Φ is shown in Figure 8, and the CPR values are given in Table 2. The CRP values of POM/Ag nanocomposites with 0, 0.1, 0.5, 1 and 2 wt % Ag nanoparticles were 0.0575, 0.0717, 0.0764, 0.0805 and 0.0792, respectively. These CRP values clearly suggest that the crystallization rate of POM was enhanced by the Ag nanoparticles when the amount of Ag nanoparticles was less than 1 wt %, due to their nucleation effect. However, when the content of Ag nanoparticles was high, they had a certain inhibitory effect on the crystallization of POM.

Jeziorny-Modified Avrami Model Analysis
The Jeziorny-modified Avrami model was used to describe the non-isothermal crystallization process of POM/Ag nanocomposites. In the Jeziorny-modified Avrami model [36], Equation (5) was taken in the double-logarithmic form to obtain Equation (6): where Xt is the relative degree of crystallinity as a function of time and n is the Avrami exponent, reflecting the mechanism for the nucleation and crystal growth during crystallization of polymers. Kt is the crystallization rate constant. In order to eliminate the effect of cooling rate, ln(Kt) was normalized with Φ to obtain ln(KC). Therefore, plots of ln(−ln(1 − Xt)) versus ln(t) are shown in Figure  9, and ln(KC) values are given in Table 2. It was clearly shown that as the content of Ag nanoparticles increased, the ln(KC) values increased first, then decreased. The ln(KC) value of POM/Ag nanocomposites with 1 wt % Ag nanoparticles was the greatest. This indicated that the smaller

Jeziorny-Modified Avrami Model Analysis
The Jeziorny-modified Avrami model was used to describe the non-isothermal crystallization process of POM/Ag nanocomposites. In the Jeziorny-modified Avrami model [36], Equation (5) was taken in the double-logarithmic form to obtain Equation (6): where X t is the relative degree of crystallinity as a function of time and n is the Avrami exponent, reflecting the mechanism for the nucleation and crystal growth during crystallization of polymers. K t is the crystallization rate constant. In order to eliminate the effect of cooling rate, ln(K t ) was normalized with Φ to obtain ln(K C ). Therefore, plots of ln(−ln(1 − X t )) versus ln(t) are shown in Figure 9, and ln(K C ) values are given in Table 2. It was clearly shown that as the content of Ag nanoparticles increased, the ln(K C ) values increased first, then decreased. The ln(K C ) value of POM/Ag nanocomposites with 1 wt % Ag nanoparticles was the greatest. This indicated that the smaller amounts of Ag nanoparticles, i.e., less than 1 wt %, increased the crystallization rate of POM. It was clearly evident that the small amounts of Ag nanoparticles acted as the heterogeneous nuclei and raised the nucleation rate. However, when the amount of Ag nanoparticles was increased and reached 2 wt %, the Ag nanoparticles aggregated in the POM matrix and restrained crystallization of POM to some degree. In addition, the n value of POM was less than that of the POM/Ag nanocomposites, proving that Ag nanoparticles acted as heterogeneous nucleating agents and increased the crystal nuclei of POM.

Ozawa Model Analysis
Ozawa extended the Avrami equation for non-isothermal crystallization analysis [37]. The relative crystallinity (X T ) can be represented as a function of cooling rate, as in Equation (8): where m is the Ozawa exponent. K(T) is the Ozawa crystallization rate function.
raised the nucleation rate. However, when the amount of Ag nanoparticles was increased and reached 2 wt %, the Ag nanoparticles aggregated in the POM matrix and restrained crystallization of POM to some degree. In addition, the n value of POM was less than that of the POM/Ag nanocomposites, proving that Ag nanoparticles acted as heterogeneous nucleating agents and increased the crystal nuclei of POM.

Ozawa Model Analysis
Ozawa extended the Avrami equation for non-isothermal crystallization analysis [37]. The relative crystallinity (XT) can be represented as a function of cooling rate, as in Equation (8): where m is the Ozawa exponent. K(T) is the Ozawa crystallization rate function.
Plots of ln(−ln(1 − XT)) versus ln(Φ) are shown in Figure 10. The values of m and lnK(T) could be calculated from the slopes and intercepts of the fitted lines, as listed in Table 3. The results indicate that the lnK(T) values of POM/Ag nanocomposites decreased when the temperature increased from low temperature to high temperature. This suggested that the crystallization rate of POM/Ag nanocomposites decreased with increasing temperature. Furthermore, the lnK(T) values of POM/Ag  Figure 10. The values of m and lnK(T) could be calculated from the slopes and intercepts of the fitted lines, as listed in Table 3. The results indicate that the lnK(T) values of POM/Ag nanocomposites decreased when the temperature increased from low temperature to high temperature. This suggested that the crystallization rate of POM/Ag nanocomposites decreased with increasing temperature. Furthermore, the lnK(T) values of POM/Ag nanocomposites was greater than that of POM, which indicated that Ag nanoparticles enhanced the crystallization rate of POM.

Liu and Mo's Model Analysis
Liu and Mo's model [38], which is a combination of Avarima and Ozawa models, is given as Equation (10): where α is the ratio of Avrami exponent to Ozawa exponent: α = (m/n). F(T) = (K(T)/Z t ) 1/m , referring to a value of cooling rate which has to be chosen at a unit crystallization time when the samples reach a certain degree of crystallinity. At a fixed relative degree of crystallinity, F(T) and α were calculated from the intercept and slope of the linear relationship between ln(t) and ln(Φ), respectively ( Table 4). The F(T) values of POM/Ag nanocomposites were found to increase steadily with the increase in the relative degree of crystallinity, which indicated that a higher degree of relative degree of crystallinity could be obtained with a higher cooling rate. Furthermore, at a selected degree of crystallinity, it was clearly shown that as the content of Ag nanoparticles increased, the F(T) value decreased first, then increased, and the F(T) value of POM/Ag nanocomposites with 1 wt% Ag nanoparticles was least. This result was in good accordance with the results from analysis by Jeziorny, Jeziorny-modified Avrami and Ozawa models. Polymers 2020, 12, x FOR PEER REVIEW 12 of 21

Ziabicki Model Analysis
Ziabicki proposed a model to analyze non-isothermal crystallization kinetics of POM/Ag nanocomposites. Termed Ziabicki kinetics, it is given as Equation (11) [39]: where (dX/dT) Φ,max and D Φ are the maximum crystallization rate and the half-width of the derivative of relative crystallinity, respectively. T o m is the equilibrium melting temperature. T g is the glass transition temperature. Plots of (dX/dT) versus T are shown in Figure 11. The value of kinetic crystallizability at unit cooling rate (G Z ) can be obtained by normalizing G Z,Φ with Φ (Table 5). It was found that as the content of Ag nanoparticles increased, the F(T) value increased first, then decreased, and the G Z value of POM/Ag nanocomposites with 1 wt % Ag nanoparticles was greatest. This showed that crystallization rate of POM was enhanced by the small amount of Ag nanoparticles due to their nucleation effect. However, when the content of Ag nanoparticles reached 2 wt %, Ag nanoparticles easily aggregated, which had a certain inhibitory action on POM crystallization.
Polymers 2020, 12, x FOR PEER REVIEW 14 of 21 found that as the content of Ag nanoparticles increased, the F(T) value increased first, then decreased, and the GZ value of POM/Ag nanocomposites with 1 wt % Ag nanoparticles was greatest. This showed that crystallization rate of POM was enhanced by the small amount of Ag nanoparticles due to their nucleation effect. However, when the content of Ag nanoparticles reached 2 wt %, Ag nanoparticles easily aggregated, which had a certain inhibitory action on POM crystallization.     In order to calculate the activation energy of the POM/Ag nanocomposites' crystallization, Kissinger provided an equation, given as Equation (12) [40]: where ∆E and R are the crystallization activation energy and the gas constant, respectively. The fitted lines of ln(Φ/T 2 p ) versus 1/T p for POM/Ag nanocomposites are shown in Figure 12, and the values of r and ∆E are listed in Table 6. The ∆E values of the POM/Ag nanocomposites indicated that Ag nanoparticles induced the heterogeneous nucleation in POM by lowering the activation energy of POM. In addition, it is known that the magnitude of activation energy (|∆E|) is related to energy need for the motion of polymer chains during the transformation from the melt into the crystalline state. The absolute values of ∆E for POM/Ag nanocomposites were greater than that for the POM, indicating that Ag nanoparticles increased the difficulty for the transport of POM molecular chains to the crystal region [41]. On one hand, it was shown that the as-synthesized Ag nanoparticles had a strong interaction with POM. On the other hand, when the amount of Ag nanoparticles increased, Ag nanoparticles easily to aggregated and inhibited the movement of the POM molecular chain, which affected the crystallization properties of the POM. of POM. In addition, it is known that the magnitude of activation energy (|∆E|) is related to energy need for the motion of polymer chains during the transformation from the melt into the crystalline state. The absolute values of ∆E for POM/Ag nanocomposites were greater than that for the POM, indicating that Ag nanoparticles increased the difficulty for the transport of POM molecular chains to the crystal region [41]. On one hand, it was shown that the as-synthesized Ag nanoparticles had a strong interaction with POM. On the other hand, when the amount of Ag nanoparticles increased, Ag nanoparticles easily to aggregated and inhibited the movement of the POM molecular chain, which affected the crystallization properties of the POM.

Morphology of POM/Ag Nanocomposites Spherulites
The spherulite morphology of POM was observed by polarized optical microscopy to investigate the effect of the Ag nanoparticles on the dynamic crystallization behavior of POM. Figure 13 displays spherulite morphology of the POM/Ag nanocomposites. As seen in Figure 13a, the POM spherulites had considerable size due to the low nucleation density of the pure POM, and spherulites directly collided with each other. In the observed areas, there were a few nuclei and spherulites with a diameter of approximately 100 µm, along with clear interfaces. In addition, Figure 13b-e shows that the size of spherulites of POM/Ag nanocomposites became small and that the amount of the spherulites increased significantly. Furthermore, during the crystallization process, Ag nanoparticles generated a great number of nuclei and the maximum number of available nucleation sites [42]. Simultaneously, Ag nanoparticles grew in a limited space and led to the formation of smaller spherulites in the composites, indicating that the nucleation density was the dominant factor affecting the crystallization process of the POM/Ag nanocomposites [43]. However, when the content of Ag nanoparticles was increased further to reach 2 wt%, Ag nanoparticles easily aggregated, which reduced heterogeneous nucleation of Ag nanoparticles and restrained crystallization of POM. nucleation sites [42]. Simultaneously, Ag nanoparticles grew in a limited space and led to the formation of smaller spherulites in the composites, indicating that the nucleation density was the dominant factor affecting the crystallization process of the POM/Ag nanocomposites [43]. However, when the content of Ag nanoparticles was increased further to reach 2 wt%, Ag nanoparticles easily aggregated, which reduced heterogeneous nucleation of Ag nanoparticles and restrained crystallization of POM.

Antibacterial Activities of POM/Ag Nanocomposites
The optical images of inhibition zones of POM/Ag nanocomposites are displayed in Figure 14, and the inhibition zones and inhibition rates of POM/Ag nanocomposites are given in Table 7. Notably, POM did not present any antibacterial activity under the test condition, as shown in Figure 14a,A. Comparatively, it can be seen in Table 7 that POM/Ag nanocomposites with 0.1 wt % Ag nanoparticles showed very promising antibacterial performance and that the inhibition zones were 0.31 cm against E. coli and 0.30 cm against S. aureus. The inhibition zones of POM/Ag nanocomposites against E. coli widened from 0.31 to 0.57 cm, and those against S. aureus widened from 0.30 to 0.53 cm when the addition of Ag nanoparticles increased from 0.1 wt % to 2 wt %. In addition, when the content of Ag nanoparticles was 0.1 wt %, the inhibition rates of POM/Ag nanocomposites against E. coli and S. aureus were 87.23% and 86.59%, respectively. The inhibition rates of POM/Ag nanocomposites against E. coli and S. aureus were both found to increase with the increasing content of Ag nanoparticles. Furthermore, when the content of Ag nanoparticles reached 2 wt %, inhibition rates of POM/Ag nanocomposites against E. coli and S. aureus were 98.12% and 97.67%, respectively. The release of Ag atoms or ions from Ag nanoparticles had been considered as one major route to produce bactericidal effects [44]. Ag ions can interact with the cell membranes, nucleic acids and proteins of the bacteria via ligand exchange reactions to exert their activity [26]. Additionally, the direct interaction of Ag nanoparticles with the surface of bacterial cells also plays an important role in their antibacterial activity [45].
2 wt %, inhibition rates of POM/Ag nanocomposites against E. coli and S. aureus were 98.12% and 97.67%, respectively. The release of Ag atoms or ions from Ag nanoparticles had been considered as one major route to produce bactericidal effects [44]. Ag ions can interact with the cell membranes, nucleic acids and proteins of the bacteria via ligand exchange reactions to exert their activity [26]. Additionally, the direct interaction of Ag nanoparticles with the surface of bacterial cells also plays an important role in their antibacterial activity [45].

Conclusions
Ag nanoparticles were synthesized by a facile route in the presence of oleic acid and n-propylamine. The crystal structure of as-synthesized Ag nanoparticles was studied by XRD, and the average primary size of the Ag nanoparticles was approximately 10 nm. The surface of Ag nanoparticles was capped with monolayer surfactants which consisted of n-propylamine and oleic acid with the content of 19.6%, which was proved by FTIR and TGA. Based on as-synthesized Ag nanoparticles, POM/Ag nanocomposites were prepared by a simple melt compounding route. It was

Conclusions
Ag nanoparticles were synthesized by a facile route in the presence of oleic acid and n-propylamine. The crystal structure of as-synthesized Ag nanoparticles was studied by XRD, and the average primary size of the Ag nanoparticles was approximately 10 nm. The surface of Ag nanoparticles was capped with monolayer surfactants which consisted of n-propylamine and oleic acid with the content of 19.6%, which was proved by FTIR and TGA. Based on as-synthesized Ag nanoparticles, POM/Ag nanocomposites were prepared by a simple melt compounding route. It was found that when the content of Ag nanoparticles in the POM matrix was less than 1 wt %, the Ag nanoparticles showed good dispersibility in the POM. Furthermore, the effect of Ag nanoparticles on the crystallization behavior and non-isothermal kinetics of POM was investigated by DSC. Addition of Ag nanoparticles dramatically enhanced the crystallization peak temperature (T P ) and relative degree of crystallinity (X C ) of POM. All of the Jeziorny, Jeziorny-modified Avrami, Ozawa, Liu and Mo's analysis models were found to describe the non-isothermal kinetics of samples fairly well. The results showed that well-dispersed Ag nanoparticles in POM matrix could decrease its half-time of crystallization (t 1/2 ) and increase the crystallization rate parameter (CRP) values of POM. The values of the crystallization rate constant at unit cooling rate (ln(K C )) and the crystallization rate function (K(T)) for POM both were less than that for POM/Ag nanocomposites, and the values of ln(K C ) and K(T) of POM/Ag nanocomposites with 1 wt % Ag nanoparticles was the greatest. Furthermore, to demonstrate the characterization of the kinetic crystallinity of the samples for a certain cooling rate, the values of kinetic crystallizability at unit cooling rate (Gz) of the Ziabicki model analysis were found to be in the following order: Gz (POM/Ag nanocomposites) > Gz (POM). Similarly, Kissinger model analysis revealed that the crystallization activation energy (∆E) value of POM was −167.70 kJ/mol, greater than that of POM/Ag nanocomposites. Furthermore, all results regarding the kinetics of non-isothermal crystallization explained that when the amount of Ag nanoparticles was less than 1 wt %, Ag nanoparticles acted as heterogeneous nuclei of POM and generated a great number of nuclei, which reduced the spherulites size of POM and increased