Effect of Polystyrene on Poly ( Ethylene Terephthalate ) Crystallization

Industrial production of PET bottles is performed in two steps: injection molding of a preform (Figure 1a), and blow molding of the preform to obtain the finished product (Figure 1d). Crystallization may occur during cooling from the molten polymer, or during heating of a mostly amorphous solid sample, a process defined as cold crystallization. The degree of crystallinity and other crystallization parameters affect the final product properties. A tight control of crystallinity is a necessary condition to obtain high performance bottles. Excessive crystallinity may impair the stretching and blowing cycles, while very low crystallinity results in poor mechanical and barrier properties1-11. Although heating and cooling rates during industrial processing are very high and the material is subjected to severe dynamical stresses, differential scanning calorimetry (DSC) static tests, carried under far milder conditions, may be used as a guide to select the processing methods. Moreover, precise determination of optimal melting parameters is needed to avoid polymer degradation during processing and excessive energy consumption12-14. According to packing industry cold crystallized preforms are a serious waste of money; thus, the solution to this problem would be very welcome. Much research has been, and still is, dedicated to the thermodynamics and kinetics of phase change – both from a macroscopic and microscopic point of view – in highly crystalline polymers. From previous work we know that blending a small amount of an amorphous polymer may hinder unwelcomed cold crystallization of PET and prevent excessive crystallinity under processing conditions. Thus, polystyrene, an apolar amorphous polymer appears to be a good candidate12,15-21. The present work is concerned with an investigation of the melting and nonisothermal cold crystallization of PET and PET/PS blends with 1 to 60% of PS content. DSC and scanning electron microscopy (SEM) techniques were employed to investigate crystallization and melting parameters, and morphology. This work draws freely on previous publications of the author on the subject.


Introduction
Poly(ethylene terephthalate) (PET) is a semi crystalline thermoplastic widely used by the food and health care packaging industries in the production bottles and containers for soft drink beverages, oils, creams, etc.The PET molecule is made of alternating rigid aromatic terephthalate units and flexible aliphatic ethylene glycol units: Industrial production of PET bottles is performed in two steps: injection molding of a preform (Figure 1a), and blow molding of the preform to obtain the finished product (Figure 1d).
Crystallization may occur during cooling from the molten polymer, or during heating of a mostly amorphous solid sample, a process defined as cold crystallization.The degree of crystallinity and other crystallization parameters affect the final product properties.A tight control of crystallinity is a necessary condition to obtain high performance bottles.Excessive crystallinity may impair the stretching and blowing cycles, while very low crystallinity results in poor mechanical and barrier properties [1][2][3][4][5][6][7][8][9][10][11] .
Although heating and cooling rates during industrial processing are very high and the material is subjected to severe dynamical stresses, differential scanning calorimetry (DSC) static tests, carried under far milder conditions, may be used as a guide to select the processing methods.Moreover, precise determination of optimal melting parameters is needed to avoid polymer degradation during processing and excessive energy consumption [12][13][14] .
According to packing industry cold crystallized preforms are a serious waste of money; thus, the solution to this problem would be very welcome.Much research has been, and still is, dedicated to the thermodynamics and kinetics of phase change -both from a macroscopic and microscopic point of view -in highly crystalline polymers.From previous work we know that blending a small amount of an amorphous polymer may hinder unwelcomed cold crystallization of PET and prevent excessive crystallinity under processing conditions.Thus, polystyrene, an apolar amorphous polymer appears to be a good candidate 12,[15][16][17][18][19][20][21] .
The present work is concerned with an investigation of the melting and nonisothermal cold crystallization of PET and PET/PS blends with 1 to 60% of PS content.DSC and scanning electron microscopy (SEM) techniques were employed to investigate crystallization and melting parameters, and morphology.This work draws freely on previous publications of the author on the subject.

Materials
Bottle grade PET, trade name Rhopet S78 was supplied by Rhodia (Brazil), with an intrinsic viscosity [η] = 0.78 dL/g and weight-average molar mass M w = 48 kg/mol.Injection grade PS, trade name Styron 649D, was purchased from Dow (Brazil).Preliminary DSC scans of the polymers showed glass transitions at 70 °C and 95 °C for PET and PS, and the complete absence of crystallinity of the latter (Figure 2a).DSC scan of finished PET bottle is showed in Figure 2b; no cold crystallization peak is observed.

Methods
Before mixing the polymers were dried in oven with forced air circulation to remove humidity and prevent PET degradation during processing.PET was dried at 120 °C for 6 hours and PS at 80 °C for 14 hours.The blends were prepared in a Haake Rheomix 600 internal mixer fitted with high intensity (roller) rotors at 265 °C and 60 rpm for 10 min.Immediately after mixing the melt was quenched into iced water to prevent PET crystallization and to obtain substantially amorphous samples.
Nonisothermal cold crystallization and fusion were investigated using a differential scanning calorimeter Shimadzu DSC-50.Glassy samples, from 5 to 8 mg in  weight, of neat PET and PET/PS blends with 1 to 60% PS were heated from room temperature to 300 °C at 10 °C/ min.A detailed investigation was done employing heating rates ranging between 1 and 50 °C/min.All samples were tested wrapped in aluminum foil to get better heat transfer 22 .
Scanning electron microscopy analyses were performed in a Shimadzu SSX 550 Superscan.The samples were fractured in liquid nitrogen and covered with gold to avoid static charges.

Results and Discussion
Figure 3 shows raw DSC plots of heat flow versus temperature of neat PET obtained at different heating rates (a) and of PET/PS blends of different compositions, taken at 10 °C/min.
A preliminary look at the plots reveals a glass transition around 70 °C, an exothermic event at 100-200 °C, attributed to the cold crystallization of PET, and an endothermic event further down at 200-300 °C, attributed to the melting of PET.PS addition and increasing heating rates result in the displacement of the crystallization peak to higher temperatures, and contribute to a wider range of cold crystallization phenomenon which are probably linked to the segregation between PS and PET molecules and to diffusion mechanisms, however, they have little effect on the melting peak.At high heating rates (greater than 30 °C/min) a partial overlap of the cold crystallization and melting peaks was observed; it may result from the lack of time to complete the crystallization at these rates.Complex cold crystallization peaks were observed for PET/PS blends with less that 10% PS.According to literature, the shape of crystallization peaks is related to morphology of crystalline phase [23][24][25][26][27][28] .
Both peaks, exothermic crystallization peak and endothermic fusion peak, were integrated from starting to end points visually determined, using a straight virtual baseline between them, to obtain the mass fractions crystallized or fused, respectively: where J is the DSC output, J 0 is the virtual baseline during the event, t 1 and t 2 are the onset and end times of the event, and E 0 is the latent heat of phase change.The normalized peak, which is also the rate of phase change (melting or crystallization): It is convenient to measure times for the onset of the event, τ = t -t 1 .Both x and c may be considered as functions of either time (τ) or temperature (T), since temperature is a linear function of time at constant heating rate, T = T 0 + ϕτ, where ϕ is the heating rate (dT/dτ) and T 0 is the temperature at onset of the event, τ = 0.
Figure 4 shows the relative crystallinity developed during the cold crystallization of PET, and crystallization rate, as functions of temperature, at several heating rates.The crystallization process begins at very low rate, probably due to the small surface area of initial crystallites; this stage of crystal growth is sometimes associated in the literature with a nucleation process.Crystallization ends at very low rate, due to depletion of crystallizable polymer and (for high crystallinity materials) to a decrease of the interfacial area caused by interference between competing growing crystals, among other reasons.Between initial and final stages, the crystallization rate reaches a maximum, at a temperature identified as crystallization temperature of the polymer 1,[28][29][30][31][32][33][34][35][36][37][38] .
The cold crystallization temperature of PET samples tested increased from 110 °C when heated at 1 °C/min to 150 °C when heated at 50 °C/min.Within the same range of heating rates, the maximum crystallization rate changed from 0.1 min -1 to 2.0 min -1 .In contrast, the crystallinity, computed assuming a latent heat of fusion of the 100% crystalline PET equal to 140 kJ/kg, was found to be visually independent of the heating rate, remaining at 16 ± 1% for conditions tested 39,40 .
Figure 5 shows the relative crystallinity and the crystallization rate as functions of temperature for PET and PET/PS blends of different composition, measured at heating rate of 10 °C/min.In presence of PS, complex cold crystallization peaks are observed.At least two crystallization peaks, frequently uneven and partly superimposed, explicit or implicit (i.e., the minor peak manifested in the form of a "shoulder" on the major peak) are identified in Figure 5b for PS contents below 30%.It should be noted that, under the tests conditions PS is a completely amorphous polymer, immiscible in PET; the solubility limit of PS in PET is believed to be less than 1%, thus it is possible that once the solubility limit between PET and PS to be reached PS is segregated at amorphous zone then producing little effect on the crystalline phase of PET.In literature database several explanations are presented for the multiple crystallization phenomenon: two kinds of amorphous regions, inter-lamellar and inter-spherulite, different crystalline geometries, co-crystallization and fractional crystallization, secondary or recrystallization effects 6,9,11,12,14,28,[30][31]35,[41][42][43][44][45][46][47][48][49][50][51][52][53][54][55] .
PET crystallization in PET/PS blends proceeds at a higher temperature and lower crystallization rate than in the neat resin.These features, clearly observed in Figure 5, are quantitatively illustrated in Figure 6. Figure 6a shows crystallization temperature (T C ) along with initial (T 1% ) and final (T 99% ) values of the crystallization temperature range; Figure 6b shows the maximum crystallization rate (c Max ) and the average crystallization rates in the intervals 1% to 99% and 20%-80% of relative crystallinity.An interesting characteristic of PET/PS system is the mild dependence of crystallization temperature increases and crystallization rate dependences on the amount of PS present.The peak crystallization temperature varies between 124 °C and 150 °C depending on the PS content.Maximum crystallization rate is significantly depressed by the incorporation of small amounts of PS, from 1.1 min -1 for pure PET to 0.3-0.4min -1 for 1-5% PS, and partially recovers with increasing PS content, up to 0.8 min -1 for 60% PS.The total amount of PET crystallized, however, is independent of PS content, remaining at 16.7 ± 1.7%.The 70% drop in the crystallization rate of PET resulting from the addition of PS in such small amounts has been attributed to the interference of amorphous PS (dissolved or segregated) on the generation of crystal PET nuclei 14 .Hence, we speak of the antinucleating effect of PS on the crystallization of PET.
Figure 7 shows the molten fraction as a function of time measured from the onset of the event, and the rate of melting as a function of the molten fraction.In contrast with the crystallization process, melting shows a very regular, asymmetric simple peak, with peak melting temperature of 249 ± 1 °C, maximum melting rate of 0.7 ± 0.1 min −1 and crystallinity of 22 ± 2% in the interval 0-60% PS content.Comparing this value with the crystallinity recovered during cold crystallization reveals about 5% crystallinity for the "amorphous" samples.SEM images of fractured surface of PET/PS blends with 1% and 15% PS are shown in Figure 8.A two-phase morphology, with 0.8 μm to 2.4 μm diameter PS spherical particles, well dispersed in the amorphous PET matrix, is visible.The clean PS/PET interfaces suggest a poor particle/ matrix adhesion, compatible results with low solubility limit between phases.
Figure 9 presents two photographs taken during the industrial processing of PET bottles: (a) preform heating, (b) preform blowing and bottle confection.Two compositions were selected to make bottles at industrial scale, neat PET and 1% PS blend.
Apparently, 1% PS has a slightly yellowing tone, likely due to immiscible PET and PS resins, which have different refractive index; the large difference in the optimum processing temperature of PET and PS may have resulted in degraded specimens as well.In general, acceptable PET bottles are produced with both compositions.

Conclusions
Addition of amorphous PS to semi crystalline PET affects the nonisothermal cold crystallization process of the later, but not the total amount crystallized.The most important effect is the significant reduction of the PET cold crystallization rate by the presence of small amounts of PS.Melting of crystallized samples is not affected by the presence of PS.Neat PET and PET/PS blend preforms and bottles are successfully produced; both compositions may be applied at industrial scale.

Figure 1 .
Figure 1.PET injected preforms: (a) and (b); and blow molded bottles: (c) and (d).Unintended cold crystallization can be visualized in (b) as a whitening zone, resulting in a malformed bottle (d).

Figure 2 .
Figure 2. DSC scans of PET and PS, before blending (a); DSC scan of PET bottle after blowing (b).

Figure 3 .
Figure 3. DSC scans of neat PET at several heating rates (a) and of PET/PS blends of different compositions at heating rate 10 °C/min (b).

Figure 4 .
Figure 4. Relative crystallinity (a) and crystallization rate (b) of neat PET as function of temperature.

Figure 5 .
Figure 5. Relative crystallinity of PET and PET/PS blends (a) and crystallization rate (b), as functions of temperature, measured at a heating rate of 10 °C /min.

Figure 6 .
Figure 6.Crystallization temperature (a) and rate of crystallization (b) as functions of PS content.

Figure 7 .
Figure 7. Molten fraction as a function of time (a) and melting rate as a function of molten fraction (b).