Prevention of surface defects in calendered poly(vinyl chloride) sheets using a succinate‐capped poly(caprolactone) additive

Surface defects known as “gas checks” often mar the surfaces of poly(vinyl chloride) (PVC) calendered films. These defects are typically prevented through changes in the calender operating parameters, a costly exercise which also limits the sheet thickness and the production rate. Adding a low concentration of poly(caprolactone) (PCL)‐based star‐shaped compound can eliminate gas check defects in PVC calendering. The effects of a triheptylsuccinate‐terminated PCL with a PCL triol core and number average molecular weight of 540 g/mol (i.e., PCL540‐[(succ)‐C7]3) has been investigated on the material, thermal, and processing properties of PVC blends containing diisononyl phthalate (DINP) as a primary plasticizer and PCL540‐[(succ)‐C7]3 in low quantities (i.e., 0, 5, or 10 parts per hundred rubber (phr)) as a secondary plasticizer and processing aid. The most significant differences between PVC blends containing PCL540‐[(succ)‐C7]3 and those without are in the rheological properties of the PVC blends at higher temperatures and lower angular frequencies. Under these conditions, PVC blends containing 10 phr of PCL540‐[(succ)‐C7]3 have a complex viscosity nearly three times higher than those containing only DINP. PVC/PCL540‐[(succ)‐C7]3 blends had comparable tensile properties to those containing only DINP, with no significant change in maximum elongation and a small but significant increase of 28% in maximum stress. The addition of PCL540‐[(succ)‐C7]3 made it possible to produce calendered films without gas checks that were twice as thick as those produced in its absence. In addition to reduced wastage of marred films, the increased calender operating range for PVC films containing PCL540‐[(succ)‐C7]3 has the potential to significantly reduce energy costs for the calendering of thick PVC films.


Highlights
• Flaws called gas checks are produced in PVC films during the calendering process.• A low amount of additive, called PCL 540 -[(succ)-C 7 ] 3 , eliminated gas checks.
• Blends had comparable tensile and mixing properties relative to controls.
• The additive enables the direct production of thicker films.
• Potential benefits include reduced waste, energy, and operator interventions.

| INTRODUCTION
Poly(vinyl chloride) (PVC) is the fifth most produced plastic by weight with 38 million tonnes of PVC processed each year. 1 PVC products use more additives by volume than any other manufactured plastics [2][3][4] mainly due to the use of large volumes of plasticizers. 5Plasticizers account for nearly one third of the total plastic additives market 3,6 and are vital for making PVC processable and malleable. 7everal methods may be used to compound PVC with plasticizer.Calendering is an industrial process wherein thermoplastics are passed through a system of heated rollers to form a continuous sheet. 8PVC accounts for most calendered polymers, as relatively few other polymers are processed in this way. 8Calenders are capable of high production speeds, which can create cost benefits for the production of large volumes of films or sheets, and are also capable of producing uniform products with a low tolerance to variance in gauge thickness. 9However, several common defects can be produced when calendering PVC, such as air entrapments (i.e., gas checks), chevrons, and mattness. 9Gas checks are physical surface defects that are hypothesized to result from air being entrapped in the PVC sheet during calendering and are generally observed as blemishes varying in size in the direction of calender flow. 9,10While mainly seen as a visual defect, they can have localized negative effects on film quality, such as increased permeability in food wrappings and compromised mechanical properties. 9Current processing methods used to reduce gas checks include alterations in the calendering process, such as decreasing the calendering speed or decreasing the distance between rollers (i.e., nip roll width).This can complicate and slow the calendering process, as it can limit the functionality and thickness of the end product, as process parameters must be maintained within a limited range in order to prevent gas checks. 9For instance, calendered film thickness is generally limited to 0.05 to 0.5 mm, as it is difficult to maintain consistency and quality of films at thicker gauges. 8,9,11,12 has been previously reported that a series of poly(caprolactone) (PCL)-based compounds with diester linkers and alkyl chain cappers were effective at reducing or eliminating the formation of gas checks during PVC calendering at concentrations as low as 5 phr. 13,14otably, the use of an additive to avoid gas checks, thereby improving product quality, would also allow for the removal of processing steps in the production of thicker films.The currently used method to produce thick calendered PVC sheets involves calendering several thin sheets of PVC and subsequently pressing or laminating them together. 8,9,15Using additives to prevent gas checks could also improve existing manufacturing processes by reducing waste and improving operating line efficiency and consistency.
Here we investigate the effects of a triheptylsuccinateterminated PCL with a PCL triol core with a number average molecular weight, M n , of 540 g/mol (i.e., PCL 540 -[(succ)-C 7 ] 3 ) on the material, thermal, and processing properties of PVC blends containing diisononyl phthalate (DINP) as a primary plasticizer and PCL 540 -[(succ)-C 7 ] 3 in low quantities as a secondary plasticizer and processing aid (see Figure 1).The main goal was to determine whether the addition of PCL 540 -[(succ)-C 7 ] 3 significantly affects either the required processing conditions or the properties of the final product, other than the observed effect on gas check reduction.This work also explores whether the use of PCL 540 -[(succ)-C 7 ] 3 as a gas check reduction additive can effectively expand the PVC calendering operation range, thereby facilitating the direct production of thicker gauge films.

| Materials
Dry PVC mixtures, both with and without DINP, were obtained from Canadian General-Tower Ltd. (Cambridge, ON).Compositions of the standard blends are outlined in Table 1.PCL star-shaped plasticizer was synthesized according to procedures previously reported. 13The two standard mixtures were weighed and combined with 0, 5, or 10 parts per hundred rubber (phr) PCL 540 -[(succ)-C 7 ] 3 to create a total plasticizer concentration of 54 phr with DINP.

| Sample preparation
Fifty (50) g batches of PVC blends with a total plasticizer content of 54 phr and with PCL 540 -[(succ)-C 7 ] 3 in varying amounts of 0, 5, and 10 phr were prepared and loaded into a Rheocord System 40 double-arm internal batch mixer from Haake Buchler Instruments (Austin, TX).Batches were mixed at a temperature of 175 C and with a rotational speed of 50 rpm for 15 min.Mixing curves for the samples were tracked and used to identify the maximum torque achieved during blending, the final average mixing torque, the time to reach the final average torque, as well as the slope of the initial increase in torque.
Hot press molding was used to produce specimens of the PVC blends for tensile and rheological testing.Specimens were pressed in steel molds placed inside a heated Carver Manual Hydraulic Press (Wabash, IN).The previously mixed PVC blends were cut into small pieces and loaded into steel molds.Three geometries were produced using this method: namely, (1) ASTM D638 type IV test bars for tensile testing; (2) 25-mm diameter disks for rheology testing; and (3) 50 Â 10 Â 1.5 mm rectangular molds for dynamic mechanical thermal analysis (DMTA).The test bars and disks were pressed in two stages at a temperature of 175 C. A load of 5 metric tonnes (MT) for 5 min was applied followed by increasing the applied load to 10 MT and holding for 10 min.The rectangular bars were pressed at a temperature of 175 C in three stages and applied loads of 5 MT for 5 min, 10 MT for 10 min, and 15 MT for 30 min.After the heating stages were complete, the hot press was cooled with water until it reached a temperature of 75 C and then the applied load was released from the hot press.The steel mold was then detached from the hot press and specimens were removed.Calendered PVC films were produced by an industry partner, Canadian General-Tower Ltd. (Cambridge ON, Canada) as reported in previous work. 13All calendered blends contained 100 phr PVC suspension resin 70K, 7 phr antimony oxide Hi-Tint, 1 phr silica, 1 phr stearic acid, 4 phr barium/zinc stabilizer, and 1 phr acrylic processing aid.The treatment group contained the same base components while also containing 10 phr of PCL 540 -[(succ)-C 7 ] 3 .The premixture was blended using a Harteck (Guangzhou, China) two-roll mill HTR-300 (d = 120 mm, T = 160 C, 45 rpm).The mill was preheated for a minimum of 1 h, after which the premixture was added to the mill.Mixing was performed for 7 min, starting from the time of film formation on the mill rolls.The milled film was cut into four pieces, each of which was fed separately into the lab-scale calender (d = 180 mm, T = 160 to 170 C, P = 45 psi, 50 rpm).The calender nip distance was set to achieve film gauges of 0.4, 0.6, and 0.8 mm ± 0.05 mm.Each of the four pieces was mixed for 1 min on the calender before being removed.Each blend resulted in three or four sheets of film.Films at gauges of 0.4 and 0.8 mm were produced in triplicates, while only one 0.6 mm film was produced.

| Sample characterization
Tensile testing was performed using a Shimadzu (Kyoto, Japan) Easy Test instrument at a strain rate of 10 mm per minute.Specimens were pressed into ASTM D638 test bars and measurements of the length, width, and thickness of each test bar was taken with a digital caliper prior to testing.Six specimens were tested for each PVC/ PCL 540 -[(succ)-C 7 ] 3 blend.The maximum stress and strain for each specimen were determined from the resulting stress/strain curves and an average stress/strain curve was produced.
A parallel plate geometry was employed in oscillatory strain-controlled mode for the PVC/ PCL 540 -[(succ)-C 7 ] 3 blends using an Anton Paar (Graz, Austria) MCR 302 instrument using a PP25 configuration to determine the storage and viscous moduli.Disks (25-mm diameter)  were placed between parallel plates with a 1-mm gap at 175 C under nitrogen in a CTD 450 convection oven.Samples were tested at a shear strain of 1.0% and an angular frequency that was decreased logarithmically from 100 to 0.1 rad/s.
The DMTA was performed using an Anton Paar (Graz, Austria) MCR 302 instrument with an SRF12 configuration.Rectangular bars (50 Â 10 Â 1.5 mm) were loaded under tension into a CTD 450 convection oven at room temperature in nitrogen.Measurements of each specimen were taken prior to testing using a digital caliper.The specimens were applied with a constant oscillatory strain of 0.1% and a frequency of 1 Hz, as the temperature was increased to 175 C. Thermogravimetric analysis (TGA) was performed using a TGA 5500 from TA Instruments (New Castle, DE, USA) to determine the thermal stability of the PVC/PCL 540 -[(succ)-C 7 ] 3 blends.TGA was conducted under nitrogen with a flow rate of 25 mL/min from temperatures of 25-600 C at a heating rate of 10 C/min.The thermal degradation onset temperature was determined according to ASTM standard method E2550 using TA instruments TRIOS software.The glass transition temperatures (T g ) for each PVC blend were determined by performing differential scanning calorimetry (DSC) analysis with a TA Instruments DSC2500 (New Castle, DE, USA).DSC was performed over a temperature range of À90 to 100 C with a heating rate of 10 C/min.The T g values were estimated using the half-height method using the TA instruments TRIOS software.Complete thermographs are presented in the Appendix A, Figures A1 and A2.
Gas checks on calendered films were counted manually by sectioning off three 7 cm Â 7 cm squares on each film; i.e., from the bottom left corner, middle, and top right corner.The number of gas checks from each area were then counted and averaged and then used to estimate the total number of gas checks for each film.These counts were also normalized per square meter (m 2 ) of calendered film.
GraphPad Prism 9 software (Dotmatics, Boston, MA, USA) was used for statistical analyses of measurements using one-way and two-way ANOVA with Bonferroni post-tests.p-Values of less than 0.05 were interpreted as being statistically significant.

| Mixing properties
The results of the batch mixing experiments showing differences in the torque mixing curves for PVC blends with 0, 5, and 10 phr of PCL 540 -[(succ)-C 7 ] 3 are presented in Figure 2 and summarized in Table 2.There was a nonsignificant (p > 0.05, one-way ANOVA) increase in maximum achieved torque with PCL 540 -[(succ)-C 7 ] 3 loading, with the average maximum achieved torques being 13.4,14.3, and 14.6 NÁm for 0, 5, and 10 phr of PCL 540 -[(succ)-C 7 ] 3, respectively.There was also a non-significant increase in final average torque with the average final torques being 7.9, 8.1, and 8.5 NÁm for 0, 5, and 10 phr of PCL 540 -[(succ)-C 7 ] 3, respectively.The final average torque was taken by averaging the torque values over the range where the torque curve remained flat.The time to reach equilbrium had a non-significant decrease with PCL 540 -[(succ)-C 7 ] 3 loading, on average being 320, 300, and 280 s for 0, 5, and 10 phr of PCL 540 -[(succ)-C 7 ] 3, respectively.The time to reach equilibrium was measured in seconds from the time of the maximum achieved torque until the final average torque was reached.

| Thermal properties
Thermogravimetric analysis revealed comparable thermal stability between all mixes, with an initial degradation onset temperature of T ≈ 265 C for approximately 70% of total mass and a further degradation at T ≈ 450 C for a further 13% of total mass.There are two major thermal degradation stages for PVC, the first occurs due to the PVC dehydrochlorination with the formation of polyene sequences along the PVC polymer backbone, and the second is associated with the decomposition of the polyene sequences. 16The first stage (Change 1) occurred from approximately 260-325 C and resulted in an average thermal degradation of 72% for all samples.The second stage (Change 2) occurred from approximately 450-490 C and resulted in an average thermal degradation of 14.7%.A summary of the TGA data can be seen in Table 3, and the full thermographs are presented in the Appendix A, Figures A1 and A2.

| Mechanical properties
The average resulting stress-strain curve from tensile testing is seen in Figure 4.The relevant parameters from tensile testing are the maximum stress and the maximum strain, a summary of which is presented in Table 4. PCL 540 -[(succ)-C 7 ] 3 content did not have a significant effect on maximum strain (p > 0.05, one-way ANOVA),   17,18 Parallel plate tests in oscillatory mode showed no significant difference between complex viscosity (η*), storage modulus (G 0 ), or loss modulus (G 00 ) between 0 and 5 phr of PCL.However, when analyzing the rheological properties based on PCL 540 -[(succ)-C 7 ] 3 content and angular frequency, there was a very significant (p < 0.01, two-way ANOVA) difference in η* at 10 phr of PCL 540 -[(succ)-C 7 ] 3 , particularly at lower frequencies, driven mainly by a relative increase in storage modulus when compared to lower PCL 540 -[(succ)-C 7 ] 3 loading.The differences between the η* and G 0 of the 0 and 5 phr mixtures became statistically significant ( p < 0.05) compared to the 10 phr mixes at angular frequencies below 1.58 rad/s, while the G 00 achieved statistical significance (p < 0.05) at angular frequencies below 3.98 rad/s.The average η*, G 0 , and G 00 of the PVC blends are shown in Figure 5.
The DMTA was performed to examine the effect of temperature on the PVC blends with varying amounts of PCL.The average measured properties recorded from DMTA can be seen in Figure 6.The different blends exhibited similar G 0 and G 00 at elevated temperatures up until 100 C, at which point G 0 began to decrease at an increasing rate.[21]

| Film properties
Calendered films at three different gauges were produced to measure the ability of PCL 540 -[(succ)-C 7 ] 3 to reduce gas checks at increasing film thicknesses.As shown in Figure 7, a nearly complete elimination of gas check defects was observed with the addition of PCL 540 -[(succ)-C 7 ] 3, even at film gauges nearly double the upper limit of the operating range.As seen in Figure 8, PCL 540 -[(succ)-C 7 ] 3 effectively reduced gas checks at gauges of 0.4, 0.6, and 0.8 mm from an average of 3400, 2200, and 4200 gas checks per square meter (GC/m 2 ), respectively, without the PCL additive to effectively zero in nearly all cases with the PCL additive.

| DISCUSSION
Using PCL 540 -[(succ)-C 7 ] 3 as a secondary plasticizer and processing aid to prevent the formation of gas checks could expand the operating conditions of the PVC calendering process. 9,13Presently, the standard operating range for the thickness of calendered PVC sheets is between 0.05 and 0.5 mm, as it is difficult to maintain quality and consistency at thicker film gauges. 8,9,11,12To ensure that PCL 540 -[(succ)-C 7 ] 3 can be used as a gas  check reduction additive in a PVC calendering operation, it is important to know how it affects the physical properties of a typical PVC formulation.The effect of using PCL 540 -[(succ)-C 7 ] 3 as a secondary plasticizer was investigated at concentrations of 0, 5, and 10 phr with the primary plasticizer DINP to achieve a total plasticizer content of 54 phr.

| Effect on thermal properties
The unchanged degradation onset temperature with PCL 540 -[(succ)-C 7 ] 3 loading shows that a PVC mixture containing PCL 540 -[(succ)-C 7 ] 3 as a secondary plasticizer is thermally stable at typical PVC processing temperatures.Calendering typically takes place in a temperature range from 160 to 210 C, 7,8,22 which is still well below the onset degradation temperature for the PVC mixtures of 265 C. The increase in T g with PCL 540 -[(succ)-C 7 ] 3 loading (see Figure 3) indicates that this additive is less effective at plasticizing when compared to DINP at the same overall plasticizer loading.This is consistent with previous work which showed slightly higher T g s when compared to DINP at equivalent plasticizer concentrations for calendered PVC/PCL 540 -[(succ)-C 7 ] 3 mixtures. 14,23,24Nonetheless, the samples still had T g values well below room temperature, thereby showing the mixtures are still plasticized effectively with PCL as a secondary plasticizer.

| Effect on mechanical properties
The tensile properties of the PVC blends summarized in Table 4 did not change considerably with PCL 540 -[(succ)-C 7 ] 3 loading, with no statistically significant difference in the maximum strain or elongation at break, and only a small but significant increase in the maximum achieved stress at 10 phr of PCL 540 -[(succ)-C 7 ] 3 .Sample stiffness did increase slightly with PCL 540 -[(succ)-C 7 ] 3 loading, reflected in the secant modulus in the initial elastic region increasing with PCL 540 -[(succ)-C 7 ] 3 content.
Rheological testing revealed a significant change in complex viscosity at a PCL 540 -[(succ)-C 7 ] 3 concentration of 10 phr, particularly at lower angular frequencies (ω < 1 rad/s).This can be related to a shear viscosity (η(_ γ)) where _ γ is the shear rate, such as those experienced in calendering, using the Cox-Merz rule (jη*(ω)j = η(_ γ) j _ γ ffi ω). 25 Overall, the η* of the blends generally increased with PCL 540 -[(succ)-C 7 ] 3 loading, showing that PCL 540 -[(succ)-C 7 ] 3 is slightly less effective at plasticizing the PVC blend when compared to DINP (see Figure 5).At lower angular frequencies, the G 0 and G 00 for each blend decreases logarithmically, however at ω < 1 rad/s, the G 0 for the blend containing 10 phr of PCL 540 -[(succ)-C 7 ] 3 decreases far less than the blends with lower concentrations of additive (see Figure 5).This increase in G 0 may be due to an increase in entanglements or an aggregation of PCL 540 -[(succ)-C 7 ] 3 at lower frequencies, causing it to behave more like an elastic solid at low frequencies, and more like a viscous fluid at high frequencies.The increased η* of the 10 phr blend at low frequencies could be contributing to the removal of gas checks in calendering processes, as higher viscosity polymers produce fewer air inclusions. 9s shown in Figure 6B, the temperature at which G 0 crosses over G 00 tends to increase with increased PCL 540 -[(succ)-C 7 ] 3 loading.This crossover point, at which the sample begins to flow freely, increases from approximately 145 to 155 C for 0 and 5 phr of PCL 540 -[(succ)-C 7 ] 3 to +175 C for samples with 10 phr of PCL 540 -[(succ)-C 7 ] 3 .A higher melt flow temperature may require the calendering of PVC containing 10 phr of PCL 540 -[(succ)-C 7 ] 3 to occur at higher temperatures to ensure proper distribution of the polymer melt within the calender bank, which would require more energy to heat the polymer melt and may cause thermal degradation of some of the blend components.As the PVC samples with 10 phr of PCL 540 -[(succ)-C 7 ] 3 did not consistently transition into the melt flow region at temperatures below 175 C, PVC blends containing 10 phr of PCL 540 -[(succ)-C 7 ] 3 undergoing calendering would have viscosities approximately 10 times higher near operating temperatures.This increased viscosity at typical calender operating conditions would help to avoid air inclusions, as it would result in a higher pressure in the calender bank. 8,9,26

| Effect on film properties
A key focus of the film calendering experiments was to determine whether PCL 540 -[(succ)-C 7 ] 3 could effectively  reduce gas check defects at higher film gauges than previously reported. 13As mentioned earlier, the broadening of calendering operating conditions could have major benefits in reducing energy consumption in the calendering process given that thicker gauges of films could be produced without multiple calender runs and additional processing steps such as pressing or lamination. 8,11 visual assessment of the calendered films showed that those containing the PCL 540 -[(succ)-C 7 ] 3 additive had considerably smoother finishes when compared to those without, primarily due to the lack of air inclusions on their surface (see Figures 7 and 8).There was minor scratching on the surface of all calendered films, with and without PCL 540 -[(succ)-C 7 ] 3 , which is presumed to be imprints from the calender rolls.
There was no observable trend in the location of gas check defects on the surface of calendered sheets (i.e., at the beginning, middle, or end of the calender run), with gas checks being evenly distributed throughout the sheets that were not blended with PCL 540 -[(succ)-C 7 ] 3 .PVC films containing PCL 540 -[(succ)-C 7 ] 3 experiences a complete reduction in gas checks in nearly all cases, with the number of gas checks per unit area being reduced from 3400 GC/m 2 to effectively zero at a film gauge of 0.4 mm, and from 4200 to 23 GC/m 2 at a film gauge of 0.8 mm, as seen in Figure 8.

| CONCLUSION
Overall, the PVC blends containing PCL 540 -[(succ)-C 7 ] 3 performed comparably to those only containing DINP in both tensile and mixing properties.This shows that PCL 540 -[(succ)-C 7 ] 3 can be added to existing calender operations without requiring substantial changes to the process or significantly altering the PVC product.Rheological testing revealed differences in blends containing 10 phr of PCL 540 -[(succ)-C 7 ] 3 at higher temperatures and lower angular frequencies when compared to blends with lower PCL 540 -[(succ)-C 7 ] 3 loading.These differences could conceivably be contributing to the effect of PCL 540 -[(succ)-C 7 ] 3 on reducing gas checks in PVC calendering, as a higher viscosity at PVC calendering process conditions would lead to higher pressures in the calender bank and thus would result in fewer defects according to the current theories on gas check formation. 9,10The mechanism by which PCL 540 -[(succ)-C 7 ] 3 prevents the formation of gas checks, however, is still not well understood, and may not be purely due to physical phenomenon but may be dependent on chemical/physical interactions between the PCL 540 -[(succ)-C 7 ] 3 compound and the PVC matrix.This work shows a nearly complete reduction in gas check defects at a film gauge above the typical operating limits of PVC calendering.If this additive was utilized in a calendering operation, it could have benefits in terms of process simplification and reductions in waste produced and energy consumed.

F I G U R E 1
Molecular structures of additives in poly(vinyl chloride (PVC) blends: (A) PCL 540 -[(succ)-C 7 ] 3 and (B) diisononyl phthalate.T A B L E 1 Composition of PVC blends used in this study.Generic material name Quantity of components in blends at various phr
influence maximum stress ( p < 0.05, one-way ANOVA).Significant differences in the maximum achieved stress were only apparent at a PCL 540 -[(succ)-C 7 ] 3 concentration of 10 phr, with the average maximum achieved stresses being 14, 15, and 18 MPa for 0, 5, and 10.phr of PCL 540 -[(succ)-C 7 ] 3, respectively.Since the strain did not change significantly while the maximum stress increased, this indicates that the PVC blends with a higher PCL 540 -[(succ)-C 7 ] 3 content were stiffer than those with lower PCL 540 -[(succ)-C 7 ] 3 content.This was reflected in the secant modulus of the blends, which generally increased with PCL 540 -[(succ)-C 7 ] 3 loading from 0.046 MPa at 0 phr to 0.053 and 0.062 MPa at 5 and 10 phr, respectively.

F I G U R E 8
Number of gas checks per square meter (mean ± SEM) for calendered PVC with and without PCL 540 -[(succ)-C 7 ] 3 (n = 3 for film gauges 0.4 mm and 0.8 mm, and n = 1 for film gauge 0.6 mm, p < 0.01).