Fast cavity surface temperature evolution in injection molding: control of cooling stage and final morphology analysis

Abstract

Fast mold surface temperature evolution in injection molding improves the surface finishing and replicability of the molded parts, and may significantly reduce frozen-in orientation. In this paper the effect of a fast control of cavity surface temperature evolution on the morphology and processing conditions of iPP molded parts has been characterized. Phenomena not previously encountered, such as a double pressure packing step when the cavity surface heating lasts longer than the packing step, have been pointed out. Significant effects on the samples frozen-in orientation have been observed by optical microscopy and confirmed by X-ray analysis. AFM analysis shows that it is possible to achieve isotropic morphology with cavity surface temperature kept constant at 150 °C for a long heating time and low holding pressure.

---

1 Introduction

The rapid growth of micro and nano-device applications intensifies the need for fabrication techniques that allow obtaining polymer objects with high accuracy in terms of manufacturing precision and resolution.^1,2 Injection molding is probably the most interesting process from this point of view. Its benefits, namely, accurate control over part dimensions, a short processing cycle time, high automation, small waste yields, and use of inexpensive materials that have revolutionized medical and optoelectronic devices in recent decades.³ Despite its capability, injection molding is simple in principle: a thermoplastic material is melted, injected into a mold cavity that reproduces the negative of the desired shape, allowed to cool under controlled conditions and then removed when it has reached sufficient rigidity. During the cooling stage, which usually is the main part of the molding cycle, the possibility to manipulate the molecular chain configuration (i.e. molecular orientation, stretch and crystalline morphology) developed during the filling stage is quite limited. It generally poses severe limitations when high performance and tight tolerances are desired. Many authors demonstrated that process parameters such as temperature, pressure and cooling time are strictly dependent on design constraints, such as the type of polymer, mold cavity layout, and cooling system.^4–8 Many attempts to uncouple the complex dependence of all these variables is high desirable for the development of new high performance processes such as microinjection.

Mold temperature is probably the most important parameter affecting the internal morphology of the molded parts.⁹ It also has a direct influence on the surface micro-features replication, which is extremely important in several high added value applications.^10,11 In literatures it is well documented how a high mold temperature has a positive effect on the internal stress relaxation, level of crystallinity, surface appearance, strength, and replication accuracy of the molded part. Sha et al.¹¹ for example, investigated the effects of mold temperature on the surface quality of micro-features with three different polymer materials. They found that an increase of the mold temperature determines the decrease all polymer melt filling difficulties and improves replication in micro-cavities.¹¹ Tosello et al.¹² underlined the importance of mold surface control in weld line formation.

Wang et al.¹³ found that the rapid control of mold temperature allows to enhance the surface appearance of molded objects. Different techniques have been proposed to rapid control the mold temperature. Lucchetta et al.¹⁴ proposed a system to control mold temperature based on aluminum foam located just below the cavity surface in which heating and cooling fluids pass through, allowing a rapid evolution of mold temperature. Lin et al.⁵ proposed induction coils as system to reach a mold temperature increase rate of 8 °C s⁻¹. The time required by such technologies for the mold surface heating up to the desired temperature (and cooling afterwards) is too long because they all require temperature changes of considerable mold thicknesses, with consequently a not negligible increase of the overall cycle time. Yao et al.¹⁵ developed a device based on proximity effect that allows a mold surface temperature increase of about 250 °C in 5 seconds applying an electrical power density of 93 W cm⁻². In that case, significant additional design and tool cost are requested.

In order to keep low the cooling time increase and the tooling cost, Jansen and Flaman^16,17 developed a multilayer heating device consisting of an electrically heated layer hold between two insulation layers; they applied such a device to a conventional mold obtaining polystyrene molded parts with lower levels of birefringence.

Recently, De Santis and Pantani¹⁸ developed a thin electrical heating device to increase the cavity surface temperature inside a Haake Minijet. By keeping the cavity surface temperature at 120 °C as long as the pressure of the piston was active, they found that, in a cavity 200 μm thick and 5 mm width, the filling length of an iPP was about doubled.

The understanding of the effect of the heating conditions on the orientation relaxation requires the application of the heating system to real injection molding conditions and the measurement of temperature during the injection molding process. Liparoti et al.^19,20 reported about the effect of an asymmetric fast surface temperature evolution on the morphology of injection molded iPP samples. They characterized the development of a strong non-symmetrical sample morphology.^19,21,22 Despite the large body of available literature and the continuous interest on the subject, up to now, rapid heating and cooling of cavity surface does not represent a well-developed area. Indeed, to the best of the authors knowledge there are no papers trying to correlate the internal morphology of the molded parts with cavity surface fast (or pulsed) heating during standard injection molding process.

In this work, a thin heater has been adopted to control the temperature evolution of both cavity surfaces during an injection molding process. The devise adopted gives rise to fast cooling even if it is kept active for very long times, much longer than gate sealing time. A mold equipped with pressure and temperature sensors along the flow path has been adopted. It allows, for the first time, to correlate the cavity pressure evolution to the cavity surface temperature evolution, and in turn, to understanding their effects on the final sample morphology distribution.

2 Experimental

In this work a polypropylene (iPP, T30G) supplied by Montell (now Basell) is investigated. A deep investigation on rheology and crystallization kinetics, including the effect of flow and pressure on the crystallization behavior, can be found elsewhere.^23–27

A poly(amide–imide) (PAI) carbon black film, having a sheet electrical resistivity of 40 Ω square is adopted as heating element in the heater devices for driving the cavity surface temperature evolution. An accurate description of the heating element is reported elsewhere.²⁸

Molded samples were produced by a 70 ton Negri-Bossi reciprocating screw injection molding machine. All relevant information about the processing conditions and the geometry adopted are reported in the ESI.†

In particular, a list of all test carried out is summarized in the Table S1 reported in the ESI section.† Test code includes sequentially the holding pressure (“H” for high holding pressure, 720 bar, and “L” for low holding pressure, 360 bar), the asymptotic temperature (T_level) and the heating time t_h (after the 2 s of pre-heating).

Additional non-conditioned experiments were carried out for comparison. In particular, the P720 and P360 (Passive) experiments were carried out without activating the heater, whereas S720 and S360 (Steel) samples were carried out by replacing the heating device with a steel layer.

Morphological characterization has been carried out with optical microscopy in polarized light. For each sample were taken two optical images of the cross section in position P2 (see Fig. S1 of ESI† section for position P2): (i) with the slices oriented along the polarizer direction; (ii) with the slices rotated of 45° with respect to the polarizer direction. The change of brightness during a 45° rotation is generally proportional to the molecular orientation of the polymer.

In order to better characterize the morphological distribution in the cross sections of the injection molded samples, some of the slices were chemically etched according to the procedure reported by Bassett²⁹ and then observed by AFM. AFM investigations were conducted in air and at room temperature with a Bruker Dimension instrument coupled with a Nanoscope V controller operating in tapping mode. Commercial probe tips with nominal spring constants of 42 N m⁻¹, resonance frequencies of 300 kHz, and tip radius of 7 nm were used.

Wide-angle X-ray diffraction patterns, with nickel filtered CuKα radiation, were obtained, in reflection mode from the sample surfaces, with an automatic Bruker D8 Advance diffractometer.

X-ray patterns were analyzed by a deconvolution procedure performed according to a scheme reported in the literature³⁰ and summarized in the ESI section.†

3 Results and discussion

3.1 Temperature evolution

In Fig. 1, the surface temperature evolutions measured in position P2 (15 mm downstream the gate) with the heater operated with 9.5 W cm⁻² and for several times (t_h) during the injection experiments are reported. At large times, t_h > 6, the temperature evolutions reach asymptotic levels, which depend upon the power supplied to the heater and for the tests shown in Fig. 1 is 150 °C. The asymptotic temperatures of all the tests considered in this work are reported as T_level in the Table S1 of the ESI.† The contact with the molten polymer induces a suddenly surface temperatures increase; after a temperature peak the polymer surface starts to cool down toward the mold surface asymptotic temperature. At the heater deactivation the temperature decreased down very fast to the mold temperature (T_mold = 25 °C).

Fig. 1 Temperature evolutions in pos P2 for the experiments performed selecting an heating power of 9.5 W cm⁻² for different heating times (H-150.0; H-150.1, H-150.6, H-150.12, H-150.21). Temperature evolutions for P720 and S720 experiments are also reported for comparison.

The cooling rates at the heater deactivation are almost the same for all the considered conditions.

On turning down the heating devices, temperature curves show slight intermediate plateaus, or at least an inflection point, in the 50–70 °C temperature range (Fig. 1). These plateaus (or the inflection points) correspond to the detachment of the sample from the cavity surface which takes place starting soon after the holding pressure vanishes (see below for pressure evolutions). The detachments can even give rise to a small temperature increase due to the heat coming from the interior of the sample that is still hot. Higher values of holding pressures would have needed to avoid detachments from cavity surface.

By comparing surface temperature evolutions of non-conditioned samples (P720 and S720), the effect on cooling time delay determined by the heat flow resistance toward the mold can be evaluated. The delay of the Passive (P720) curve with respect to the Steel one (S720) changes with temperature (and time); in the temperature interval 50–60 °C it appears to be slightly smaller than the delay at lower temperature. The increase of the delay at temperature lower than 50 °C is due to the detachment from the mold surface of the Passive sample. Considering that the detachment of the Passive sample from the mold surface would be avoided by a small increase of holding pressure, the cooling delay of the Passive case with respect to the Steel one has to be considered in the temperature range 50–60 °C and can be estimated as slightly larger than 2 s.

The test carried out keeping active the heating device just during filling time (experiment H-150.0, see Table S1 of ESI†) is very interesting because it would be chosen in the attempt of decreasing filling pressure, and improving surface quality and replicability; the difference between the cooling time down to 50–60 °C of this test and the cooling time of the Passive case is smaller than 1 s, which means a delay of 3 s with respect to the Steel case.

Obviously, the cooling time delay increases as t_h increases. The data of Fig. 1 show that, at any temperature where the detachments are not encountered, the delay with respect to the Passive case is essentially proportional to t_h.

3.2 Pressure evolutions

In Fig. 2a, pressure evolutions measured in different positions along the flow path, during the test performed at 360 bar holding pressure and 150 °C mold temperature, are reported.

Fig. 2 (a) Pressure evolutions in different positions along the flow path of the experiment L-150.20; (b) pressure evolutions in pos P2 for the experiments performed at 360 bar holding pressure keeping the mold surface temperature at 150 °C for different heating times, t_h, as shown in the figure. t = 0 s is the time at which polymer melt reaches position P2.

At about 6 s, the cavity pressure in position P2 starts to decrease quickly and the pressure in P1 starts to increase toward a level slightly higher. This identifies the start of gate solidification. When gate solidification is completed the pressure in positions P1 reaches a new slightly higher level. Once the gate is solidified, the cavity remains completely isolated and its pressure evolves independently of what happened upstream.

The fact that pressures in P2 and in P3 evolve together after gate solidification indicates that, at the gate solidification time, the polymer that is confined in the volume between these two transduces positions is still nicely fluid and it experiences cooling under quiescent conditions. Meanwhile, both upstream and downstream (gate and cavity in pos. P4) the polymer is already solidified.

In Fig. 2b pressure evolutions in position P2, obtained keeping active the heating device for different times t_h at an electrical power of 9.5 W cm⁻² (T_level = 150 °C), are reported.

The pressure evolutions in position P2 reported in Fig. 2b clearly show that when t_h, the cavity surface heating time, is smaller than the end of gate solidification time (about 8 s) the pressure smoothly decreases directly toward zero; vice versa, if the cavity surface heating time is larger than the gate solidification time the pressure undergoes in sequence two steps down: the first step starts when gate is nearly solidified and lasts until the cavity surface temperature is held hot, the second step down toward the final pressure starts when the heating device is deactivated and corresponds to the cooling step from T_level to the mold temperature, T_mold. The two pressure cooling steps are well separated because, when the gate is sealed the transformation inside the cavity is essentially isochroous (except for minor cavity volume changes).

During the standard injection molding process there is only one cooling pressure step and, to the best of the authors' knowledge, a double cooling pressure step has not yet been mentioned in the literature.

The double cooling step allows to hold the materials at constant temperature for a time long enough to allow complete relaxation of the polymer chains before its solidification. Thus, morphology distribution in the final sample can be controlled according to the application purposes.

3.3 Morphology investigations

Cavity surface temperature evolutions strongly influence final morphology distribution along the sample thickness. The morphology distribution in position P2 of the tests whose cavity surface was heated for 12 s with different electrical powers is shown in Fig. 3. Corresponding surface temperature evolutions are reported in Fig. S5 of the ESI†. The optical micrographs reveal the morphology distribution typical of an injection molded semi-crystalline polymer.^32,33 This morphology, usually denoted as skin-core morphology (Fig. 4), is characterized by the presence of a series of distinct regions: a thin oriented skin layer, which includes the sample surface; a highly oriented fibrillar zone, the so called ‘shear layer’; a transitional area with smaller orientation and the spherulitic core.²⁰

Fig. 3 Optical micrographs of the slices cut from the samples produced at a heating times of 12 s and holding pressure of 720 bars at different electrical powers (a) flow direction at 45° respect to polarizer direction; (b) flow direction aligned along polarizer direction.

Fig. 4 Representation of skin-core morphology obtained by injection molding experiments.

The micrographs reported in Fig. 3 clearly show that the shear layer thickness decreases with the increase of electrical power, in particular, also applying a small electrical power (5 W cm⁻²) a significant decrease of the shear layer thickness is observed with respect to the Steel sample.

In Fig. 5, the optical micrographs of the samples obtained with 9.5 W cm⁻² electrical power (asymptotic cavity surface temperature 150 °C) for different heating times (1.3, 6, 12 and 21 s) are reported. In the same figure the optical micrographs of all non-conditioned samples (S720, S360, P720 and P360) are also shown.

Fig. 5 Optical micrographs of the slices cut close to position P2 from the samples produced with surface temperature T_level = 150 °C held for different heating times (0 s, 1.3 s, 6 s, 12 s and 21 s) and with packing pressures of 360 bar (a) flow direction at 45° with respect to polarizer direction; (b) flow direction aligned along polarizer direction.

The shear layer covers more than half sample thickness for the Steel samples (S360, obtained without heating device in the mold) and nearly one half of the cross section for the Passive (P360) samples obtained without activating the heaters (Fig. 5). The shear layer and the skin, generally characterized by high level of molecular orientation and stress frozen-in, are the main responsible for part warpage, premature failure and optical retardance.

The production of homogeneous, not-oriented and stress-free molded parts is not consistent with the presence of oriented layers. However, as evident from Fig. 5, keeping high the cavity surface temperature for long times, stress and molecular orientation and stretch have the possibility to relax. In particular, the thickness of the shear layer was found to decrease as the cavity surface heating time increases. Indeed, the samples produced with longest surface heating times (L-150.12 and L-150.20 in Fig. 5) seem to have an isotropic spherulitic morphology along the whole sample thickness.

AFM investigations allow to identify the details of different crystalline structures of the regions described above. In particular, Fig. 6 shows AFM height images obtained for samples produced at the lower holding pressure, 360 bar, without activating the heater (S360).

Fig. 6 AFM height images of the sample produced at a holding pressure of 360 bar without the heating system (S360) (half thickness of the optical micrographs).

In Fig. 6 it is possible to identify all the regions mentioned above: “1” is the skin layer, “2” is the shear layer, “3” is the transitional zone and “4” is the spherulitic core. Spherulites having mean dimension of about 10 μm can be observed in the spherulitic core, region “4”. The shear layer, region “2”, consists of thin fibrils of about 500 nm thickness. The transitional zone, region “3”, is also characterized by the presence of fibrils, however, those fibrils are ticker (about 2 μm thick) than those observed in the shear layer. In the skin layer, region “1”, sequences of globular elements, having mean dimensions lower than 1 μm, can be observed.

AFM analysis reveals the possibility to obtain isotropic structures by injection molding process. In Fig. 7 AFM micrographs of the sample L-150.20 are reported: the existence of a single spherulitic zone from the surface to the mid-plane was confirmed. The mean dimension of the spherulites present in the sample appears certainly larger with respect to the sample produced by conventional injection molding conditions, Steel (S360) analyzed in Fig. 6.

Fig. 7 AFM height images of the sample produced keeping active the heater for 21 s at 150 °C (L-150.20) (half thickness of the optical micrographs).

Furthermore, the spherulites in the zone region “4a” of Fig. 7, the core, appear to have a dimension of about 20 μm, certainly larger than the spherulites near the sample surface, region “1a”.

Somewhat different is the morphology distribution of the samples injection molded at higher holding pressures, with the same cavity surface temperature and heating time, 150 °C and 12 s; indeed, with the higher holding pressure, the shear layer does not completely disappear after 12 s surface heating, rather, it contracts into a thin transitional layer (Fig. 8).

Fig. 8 Optical micrographs of the slices cut close to position P2 from the samples produced with surface temperature T_level = 150 °C held for different heating times (0 s, 1.3 s, 6 s, 12 s and 21 s) and with packing pressures of 720 bar (a) flow direction at 45° with respect to polarizer direction; (b) flow direction aligned along polarizer direction.

The effect of holding pressure on the cross-section morphology distribution is related to the well-known increase of both viscosities and relaxation times with pressure:^31,34 at higher pressure the flow determines higher molecular stretch and also, after the flow interruption, a decrease of molecular relaxation rate.

The experimental information gathered with 150 °C cavity surface temperature allows to clarify if the crystallization takes place before or after the heating device de-activation. The reduction of the oriented thickness shown by the samples L-150-20 and H-150-20 with respect to those obtained with the experiments L-150-6 and H-150-12 excludes that the crystallization would take place before 6 s or 12 s with holding pressures equal to 360 bars and 720 bar respectively.

Furthermore, a crystallization, although limited, at times with the cavity surface held at 150 °C after gate solidification (8 s) would have given rise to a reduction of pressure values with a significant change of pressure evolution paths for both tests L-150-20 and H-150-20; vice versa, pressure evolutions maintain a very regular path as shown in Fig. 2a and b. Thus it can be concluded that for the tests carried out with T_level 150 °C, with both holding pressures, crystallization takes place after the heating device de-activation.

X-ray analysis were performed to correlate the sample surface morphology to the orientations as function of heating power (temperature) and heating time t_h.

Trotignon et al.³⁵ reported the orientation indices A₁₁₀ and A₁₃₀, which characterize the degree of orientation of α-PP crystallites. The determination of both indices is based on the vanishing (or decrease of intensity) of the reflections α₁₁₁ and α_131+041 in the equatorial measurement when the crystallites are oriented along the flow direction: the orientation indices are close to 1 for highly oriented α-crystallites, otherwise the orientation indices are lower than 1. These indices are calculated as follows:

where I is the intensity of each peak.

Table 1 shows the indices evaluated from X-ray diffraction patterns reported in Fig. 9.

Table 1 Orientation indices and crystalline percentage of the samples obtained with different t_h and mold temperature

	A110	A130	Crystallinity%
P720	0.96	0.94	0.59
P360	0.79	0.67	0.52
H-120.12	0.69	0.67	0.58
H-150.1	0.65	0.55	0.58
H-150.12	0.67	0.48	0.55
H-150.20	0.59	0.48	0.55

---

Fig. 9 Diffraction patterns for the experiments performed holding the cavity surface at 150 °C for different heating times. The X-ray pattern for the Passive sample P720 is also reported.

The orientation indices A₁₁₀ and A₁₃₀ reported in Table 1, consistently with the results of the optical micrographs and AFM observations reported from Fig. 3–7, show that the orientation at the sample surface decreases as electrical power density and heating time increase, namely when either the relaxation time decreases or there is more time to relax the molecular stretch and orientation.

4 Conclusions

A thin heating device operating by Joule effect has been adopted to achieve fast temperature evolution on the cavity surface during the injection molding process of an iPP. The cavity surface has been heated to desired temperature levels for different times. Due to the small thickness of the heating device, when the device is de-activated, the cooling toward the mold temperature is very fast, only about 3 s longer than in absence of the heating device.

If the heating device is kept active for sufficient time, the cooling process splits into two parts: the first from the injection temperature to the cavity surface temperature (held high by the heating device) and the second step from the heated surface temperature toward the mold temperature. Only for sufficient heating times double pressure cooling steps, parallel to the temperature steps, are observed.

The control of the cavity surface temperature evolution allows to calibrate the relaxation of the molecular stretch induced by the flow and thus to determine the final morphology of the moldings. In particular, due to the decrease of both viscosity and relaxation time with temperature, the thickness of the shear layer is found to decrease with surface temperature and heating time increases. The cross section of the samples molded with 360 bar holding pressure is found completely spherulitic, consistently with the decrease of both viscosity and relaxation time as pressure decrease. X-ray analysis confirms that the orientation at the sample wall generally decreases with the increase of cavity surface temperature, and also with the increase of the heating time.

References

1. C. T. Leondes, Mems/Nems: (1) Handbook Techniques and Applications Design Methods, (2) Fabrication Techniques, (3) Manufacturing Methods, (4) Sensors and Actuators, (5) Medical Applications and MOEMS, Springer Science & Business Media, 2007.
2. J. Giboz, T. Copponnex and P. Mélé, J. Micromech. Microeng., 2007, 17, R96–R109.
3. J. Greener and R. Wimberger-Friedl, Precision injection molding, Carl Hanser, Munich, 2006.
4. S. Menotti, H. N. Hansen, G. Bissacco, M. Calaon, P. T. Tang and C. Ravn, Int. J. Adv. Des. Manuf. Technol., 2014, 74, 907–916.
5. H. L. Lin, S. C. Chen, M. C. Jeng, P. S. Minh, J. A. Chang and J. R. Hwang, Int. Commun. Heat Mass Transfer, 2012, 39, 514–522.
6. M. R. Kamal, A. I. Isayev and S.-J. Liu, in Progress in polymer processing, ed. Hanser, Carl Hanser Verlag GmbH & Co. KG, 2009, ch. 18, p. 737.
7. R. Pantani, V. Speranza, A. Sorrentino and G. Titomanlio, Macromol. Symp., 2002, 185, 293–307.
8. U. Vietri, A. Sorrentino, V. Speranza and R. Pantani, Polym. Eng. Sci., 2011, 51, 2542–2551.
9. S. Liparoti, A. Sorrentino and G. Titomanlio, in Polymer Processing with resulting morphology and properties: Feet in the present and eyes in the future: Proceedings of the GT70 International Conference, AIP Conference Proceedings, Salerno, Italy, 2015, vol. 1695, p. 20052.
10. G. Lucchetta, M. Fiorotto and P. F. Bariani, CIRP Annals - Manufacturing Technology, 2012, 61, 539–542.
11. B. Sha, S. Dimov, C. Griffiths and M. S. Packianather, J. Mater. Process. Technol., 2007, 183, 284–296.
12. G. Tosello, A. Gava, H. Hansen and G. Lucchetta, in ANTEC-Conference Proceedings, 2007, vol. 4, p. 2002.
13. G. Wang, G. Zhao and Y. Guan, J. Appl. Polym. Sci., 2013, 128, 1339–1352.
14. G. Lucchetta and M. Fiorotto, Strojniški vestnik-Journal of Mechanical Engineering, 2013, 59, 683–688.
15. D. Yao, T. E. Kimerling and B. Kim, Polym. Eng. Sci., 2006, 46, 938–945.
16. K. M. B. Jansen and A. A. M. Flaman, Polym. Eng. Sci., 1994, 34, 894–897.
17. K. M. B. Jansen and A. A. M. Flaman, Polym. Eng. Sci., 1994, 34, 898–904.
18. F. De Santis and R. Pantani, J. Mater. Process. Technol., 2016, 237, 1–11.
19. S. Liparoti, A. Sorrentino and G. Titomanlio, AIChE J., 2016, 62, 2699.
20. S. Liparoti, A. Sorrentino, G. Guzman, M. Cakmak and G. Titomanlio, RSC Adv., 2015, 5, 36434–36448.
21. S. Liparoti, T. M. Hunag, A. Sorrentino, G. Titomanlio and M. Cakmak, in 30th International conference of Polymer Processing Society, AIP Conference Proceeding, Cleveland, Ohio, USA, 2014, vol. 1664, p. 110015.
22. S. Liparoti, A. Sorrentino, G. Guzman, M. Cakmak and G. Titomanlio, in Proceedings of PPS-31: The 31st International Conference of the Polymer Processing Society, AIP Conference Proceeding, 2016, vol. 1713, p. 170001.
23. I. Coccorullo, R. Pantani and G. Titomanlio, Polymer, 2002, 44, 307–318.
24. R. Pantani, F. De Santis, V. Speranza and G. Titomanlio, in Proceedings of PPS-29: The 29th International Conference of the Polymer Processing Society, AIP Conference Proceeding, 2014, vol. 1593, pp. 636–640.
25. R. Pantani, V. Speranza and G. Titomanlio, Macromol. Theory Simul., 2014, 23, 300–306.
26. R. Pantani, I. Coccorullo, V. Volpe and G. Titomanlio, Macromolecules, 2010, 43, 9030–9038.
27. F. De Santis, R. Pantani and G. Titomanlio, Polymer, 2016, 90, 102–110.
28. S. Liparoti, G. Landi, A. Sorrentino, V. Speranza, M. Cakmak and H. C. Neitzert, Adv. Electron. Mater., 2016, 2, 1600126.
29. H. M. White and D. C. Bassett, Polymer, 1998, 39, 3211–3219.
30. N. S. Murthy and H. Minor, Polymer, 1990, 31, 996–1002.
31. A. Sorrentino and R. Pantani, Polym. Bull., 2013, 70, 2005–2014.
32. R. Pantani, V. Speranza and G. Titomanlio, Polym. Eng. Sci., 2001, 41, 2022–2035.
33. R. Pantani, A. Sorrentino, V. Speranza and G. Titomanlio, Ind. Eng. Chem. Res., 2012, 51, 10840–10847.
34. I. Coccorullo, R. Pantani and G. Titomanlio, Int. Polym. Process., 2005, 20, 186–190.
35. J. P. Trotignon, J. L. Lebrun and J. Verdu, Plastics and Rubber Processing and Applications, 1982, vol. 2, pp. 247–251.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra22968a

---