Solution Crystallization and Dissolution of Polyolefins as Monitored by

Jul 5, 2013 - ... https://cdn.mathjax.org/mathjax/contrib/a11y/accessibility-menu.js .... Solution Crystallization Analysis by Laser Light Scattering ...
0 downloads 0 Views 2MB Size
Download PDF
Letter pubs.acs.org/ac

Solution Crystallization and Dissolution of Polyolefins as Monitored by a Unique Analytical Tool: Solution Crystallization Analysis by Laser Light Scattering Sadiqali Cheruthazhekatt,* Divann D. Robertson, Margaretha Brand, Albert van Reenen, and Harald Pasch Department of Chemistry and Polymer Science, University of Stellenbosch, 7602 Matieland, South Africa S Supporting Information *

ABSTRACT: For the first time, the solution crystallization and dissolution behavior of polyolefins in a variety of solvents was investigated by using a recently developed crystallization based analysis technique, solution crystallization analysis by laser light scattering (SCALLS). SCALLS results provide clear evidence that crystallization and dissolution of linear polyethylene (PE) and isotactic polypropylene (iPP) are greatly influenced by the type of solvent used. It was demonstrated for a blend of PE and iPP that cocrystallization effects are minimal in solvents such as TCB and o-DCB and are significantly more pronounced in xylene and decalin. Surprisingly, in xylene, individual dissolution curves (bimodal SCALLS profile) for both PE and iPP with minimal codissolution effects were observed while in TCB, o-DCB, and decalin both components dissolve simultaneously. These findings provide a novel and facile approach to understand the effect of solvents on cocrystallization and codissolution of chemically dissimilar components in preparative fractionations such as prep TREF (which normally uses xylene), by using TCB as the crystallization solvent and xylene as the eluent.

T

solvents that are used for fractionation and analysis. The recently developed solution crystallization analysis by laser light scattering (SCALLS)11 was found to be a suitable method to study both crystallization and dissolution of the polyolefins in a variety of solvents. This approach is different from analytical TREF and CEF, where only the dissolution can be monitored, and CRYSTAF, where only the crystallization can be monitored. In the present study, SCALLS measurements were performed on a linear polyethylene and an isotactic polypropylene in sample solvents such as TCB, o-DCB, xylene, and decalin. The solvents were found to be critical for the crystallization and dissolution of the polymer chains under similar experimental conditions.

emperature rising elution fractionation (TREF) and crystallization analysis fractionation (CRYSTAF) are the two most commonly used fractionation analysis methods for the chemical composition separation of polyolefins. Cocrystallization of polymer chains with different chemical structures having similar crystallizabilities, which reduces the resolution in separation processes, is one of the major problems commonly encountered in both techniques, where a continuous crystallization of polymer chains from dilute solutions takes place.1−9 Recently, a new technique, crystallization elution fractionation (CEF), has been introduced by Monrabal,10 which separates the polymer chains with significantly reduced cocrystallization effects compared to TREF and CRYSTAF. Several factors such as cooling rate, polymer chain length (molar mass), and chemical composition of the individual chains, are known to influence the cocrystallization of various components during their precipitation from solution. These have been extensively investigated by several authors.4,8,10 So far, no systematic reports on the effect of various solvents on the crystallization and dissolution of polyolefins in TREF and CRYSTAF have been presented in the literature up until now. This knowledge, however, is essential, in particular for the comparison of results obtained by separation methods which use different solvents, e.g., preparative fractionations such as prep TREF. Frequently, prep TREF is performed in xylene, while the separated fractions are analyzed by analytical TREF, CRYSTAF, CEF using solvents like 1,2,4- trichlorobenzene (TCB) and ortho-dichlorobenzene (o-DCB). Unfortunately, these techniques are limited with regard to the variety of © 2013 American Chemical Society

■

EXPERIMENTAL SECTION

Samples and Solvents. A sample of linear polyethylene (PE) with a weight average molar mass (Mw) of 115 kg mol−1 was purchased from Polymer Standards Service GmbH, Germany. Isotactic polypropylene (iPP) with a Mw of 110 kg mol−1 was purchased from American Polymer Standard Corp. Spectrophotometric grade TCB and o-DCB (>99% purity, Sigma-Aldrich), solvent grade xylene, and synthesis grade decalin (Merck) were used as purchased. Received: June 7, 2013 Accepted: July 5, 2013 Published: July 5, 2013 7019

dx.doi.org/10.1021/ac401700p | Anal. Chem. 2013, 85, 7019−7023

Analytical Chemistry

Letter

Figure 1. Schematic representation of the SCALLS instrument.

Figure 2. Normalized SCALLS detector response obtained from cooling (a) and heating (c) runs for PE in TCB, o-DCB, xylene, and decalin; parts b and d represent the corresponding normalized first derivative SCALLS plots.

Solution Crystallization Analysis by Laser Light Scattering (SCALLS). The experimental setup for the SCALLS instrument is shown in Figure 1. A significant advantage of this instrument is that it was constructed from relatively inexpensive, readily available components and does not require any micromachining.11−13 The development and layout of the instrument has been discussed in detail by van Reenen et al. in previous papers.10,13 As illustrated in Figure 1, the laser beam (blue laser at 405 nm) from a laser diode (a)

was focused in the center of the sample cell. Two photodiode detectors at 90° (b1) and 270° (b3) were used to detect the scattered light, and one in the forward direction at 180° (b2) was used to measure the decrease or increase in the intensity of the laser light as crystallization or dissolution occurs. A neutral density filter was put in the path of the detector b2 to protect this detector against signal saturation. The dissolution and crystallization of the sample was done by placing the quartz sample holder in the four port aluminum block, which was 7020

dx.doi.org/10.1021/ac401700p | Anal. Chem. 2013, 85, 7019−7023

Analytical Chemistry

Letter

Figure 3. Normalized SCALLS detector response obtained from cooling (a) and heating (c) runs for iPP in TCB, o-DCB, xylene, and decalin; parts b and d represent the corresponding normalized first derivative SCALLS plots.

The first approach involves the measurements of PE and iPP in solvents such as TCB, o-DCB, xylene, and decalin. In the second approach, measurements on a blend of PE and iPP were performed to determine the extent of cocrystallization and codissolution in different solvents. PE and iPP Homopolymers. Figures 2 and 3 compare the SCALLS profiles obtained for the PE and iPP in various solvents. It can be seen that the crystallization and dissolution temperature (Tc and Td) of PE in TCB and o-DCB are similar, around 84 and 94 °C, respectively. In contrast, in xylene and decalin the position of the peaks was shifted toward the low temperature side as shown in Figure 2. SCALLS measures the decrease or increase in the intensity of the laser light passing through the polymer solution, which results from the formation of the crystals from solution and their redissolution at respective temperatures during cooling and heating runs, respectively. In the case of iPP homopolymer (as illustrated in Figure 3a), except for xylene in all the other three solvents a sharp and maximum decrease in the detector signal was noticed. In xylene, a notable decrease in the signal intensity up to a certain temperature range (approximately at 55 °C) was obtained, followed by an abnormal behavior, a gradual increase in the signal after 52 °C. Here it was assumed that due to the low density of the solvent, crystal aggregation and precipitation at the bottom of the sample holder could occur, thereby decreasing the scattering of the incident light. It should be noted that the detector response (light scattering) strongly depends on the number and size of the crystals formed. An in-depth investigation is needed to explore the

placed on the top of the heater/stirrer. The heater coil was connected to the external temperature controller for controlled heating and cooling. For all SCALLS experiments, the crystallization and dissolution studies were conducted on sample solutions having a concentration of 1 mg/mL at a cooling rate of 0.5 °C/min and a heating rate of 1 °C/min with constant stirring (500 rpm) throughout the measurements.

■

RESULTS AND DISCUSSION

Very recently, van Reenen et al. demonstrated the use of SCALLS for studying the solution crystallization and dissolution behaviors of polyolefins.11,13 Crystallization profiles similar to the CRYSTAF results were obtained within short analysis times. In addition to the solution crystallization, the dissolution (melting) behavior of the crystallized components can also be measured by SCALLS. This is a highly useful approach to understand the crystallization and dissolution behavior of polymer chains since both steps are involved in fractionation methods such as TREF and CEF. An advantage is that relatively short analysis times and small amounts of the solvents are required and the instrument setup is simple. Effects of various experimental parameters such as cooling/heating rates, molar mass, and chemical composition of the samples on the SCALLS results has been discussed in previous papers.11,13 In the present paper, we report the use of SCALLS for studying the solution crystallization and dissolution of polyolefins and their blends in different solvents. 7021

dx.doi.org/10.1021/ac401700p | Anal. Chem. 2013, 85, 7019−7023

Analytical Chemistry

Letter

Figure 4. Overlay of the normalized SCALLS detector response obtained from cooling (a) and heating (c) runs for the PE-iPP blend in TCB, oDCB, xylene, and decalin; parts b and d represent the corresponding normalized first derivative SCALLS plots.

observed in both solvents, which implies the codissolution of PE and iPP. SCALLS analysis of the blend in xylene and decalin (see also Figures S3 and S4 in the Supporting Information) directly indicate a clear difference in the crystallization and dissolution behavior compared to the results obtained in TCB and o-DCB. In both solvents (xylene and decalin), the major part of the sample crystallized at higher temperatures, 80 and 75 °C, respectively. Surprisingly, individual dissolution peaks were observed in xylene, with minimal codissolution effects. In decalin, both the homopolymers and the blend are found to dissolve at similar temperature ranges with a slight shift in the peak maximum temperature values of approximately ±1 °C. As described above, a distinct shift in the crystallization and dissolution temperatures of the blend components was observed in various solvents by using SCALLS. These results show that the intermolecular interactions occurring during the crystallization and dissolution steps are different in various solvents. Co-crystallization effects are found to be minimal in TCB, indicating that this solvent is the best choice for the fractionation of linear PE from iPP according to their chemical composition with negligible cocrystallization. In contrast to this, solvents such as xylene and decalin were found to be strong cocrystallization promoting solvents. Nearly unimodal SCALLS profiles with clear shifts in the peak maximum temperature toward lower temperature ranges were observed. Surprisingly, when monitoring the dissolution behavior of the blend (see Figure 4c,d), a clear bimodality of the PE and iPP chains in xylene were noticed. These results can provide very important

effect of various solvents on the nature of the crystals formed, and therefore the effect on the SCALLS response will be discussed in a separate study. From Figure 3b, it is seen that for iPP in TCB and o-DCB the crystallization peak appears at 75 and 71 °C with a difference in the peak maximum temperature of 4 °C. In xylene and decalin, the crystallization and dissolution temperatures were found to be much lower; iPP chains crystallize around 58 and 43 °C, respectively. The observed cocrystallization effects could be the reason why low to medium molar mass iPP have been found in semicrystalline TREF fractions (60, 80, and 90 °C) of EP copolymers as reported by Cheruthazhekatt et. al.14−17 Blend of PE and iPP. As discussed above, SCALLS measurements revealed the effect of solvents on crystallization and dissolution behavior of PE and iPP. Taking these results into account, measurements on a blend of these two homopolymers were conducted under similar experimental conditions. SCALLS results for the PE-iPP blend are given in Figure 4 (overlay of the SCALLS profiles obtained for the homopolymers and blends in corresponding solvents are given in the Supporting Information, see Figures S1−S4). As can be observed from Figure 4 (see also Figures S1 and S2 in the Supporting Information), in TCB and o-DCB, the cocrystallization is found to be minimal and nearly baseline separated peaks were obtained for PE and iPP. A slight shift in the Tc value of iPP component was also detected. On the other hand, in the dissolution experiment a single peak around 108 °C was 7022

dx.doi.org/10.1021/ac401700p | Anal. Chem. 2013, 85, 7019−7023

Analytical Chemistry

Letter

Notes

information and a simple way for evaluating the effect of solvent on the crystallization and dissolution behavior of olefin homopolymers and their blends in a very short analysis time. It should be noted that the experimental conditions used for the present study were aimed at investigating the solvent effect on crystallization and dissolution. More extended measurements, including various heating/cooling rates, use of additional sample solvents such as cyclohexanone, 1-decanol, dimethyl sulphoxide and others, and a wide spectrum of polymers having varying chemical compositions, will be explored in future investigations.

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors would like to thank E. G. Rohwer and P. Neethling, Laser Physics Institute, University of Stellenbosch, for their contribution in the setting up of the instrument.

■

(1) Wild, L.; Ryle, T. Polym. Prepr., Am. Chem. Soc., Polym. Chem. Div. 1977, 18, 182. (2) Soares, J. B. P.; Hamielec, A. E. Polymer 1995, 36, 1639−1654. (3) Soares, J. B. P.; Anatawarskul, S.; Adams, P. M. W. Adv. Polym. Sci. 2005, 182, 1. (4) Anantawaraskul, S.; Soares, J. B. P.; Wood-Adams, P. M. J. Polym. Sci. Part B: Polym. Phys. 2003, 41 (14), 1762−1778. (5) Kissin, Y. V.; Fruitwala, H. A. J. Appl. Polym. Sci. 2007, 106, 3872−3883. (6) Pasch, H.; Brüll, R.; Wahner, U.; Monrabal, B. Macromol. Mater. Eng. 2000, 279, 46−51. (7) Monrabal, B.; Sancho-Tello, J.; Mayo, N.; Romero, L. Macromol. Symp. 2007, 257, 71−79. (8) Monrabal, B. J. Appl. Polym. Sci. 1994, 52, 491−499. (9) Soares, J. B. P.; Monrabal, B.; Nieto, J.; Blanco, J. Macromol. Chem. Phys. 1998, 199, 1917−1926. (10) Monrabal, B.; Mayo, N.; Cong, R. Macromol. Symp. 2012, 312, 115−129. (11) van Reenen, A. J.; Brand, M.; Rohwer, E.; Walters, P. Macromol. Symp. 2009, 282, 25−32. (12) Shan, C. L. P.; deGroot, W. A.; Hazlitt, L. G.; Gillespie, D. Polymer 2005, 46, 11755−11767. (13) van Reenen, A. J.; Rohwer, E. G.; Walters, P.; Lutz, M.; Brand, M. J. Appl. Polym. Sci. 2008, 109, 3238−3243. (14) Cheruthazhekatt, S.; Pijpers, T. F. J.; Harding, G. W.; Mathot, V. B. F.; Pasch, H. Macromolecules 2012, 45, 2025−2034. (15) Cheruthazhekatt, S.; Pijpers, T. F. J.; Harding, G. W.; Mathot, V. B. F.; Pasch, H. Macromolecules 2012, 45, 5866−5880. (16) Cheruthazhekatt, S.; Harding, G. W.; Pasch, H. J. Chromatogr., A 2013, 1286, 69−82. (17) Cheruthazhekatt, S.; Pijpers, T. F. J.; Mathot, V. B. F.; Pasch, H. Anal. Bioanal. Chem. 2013, DOI: 10.1007/s00216-013-6955-5.

■

CONCLUSION For the first time it was demonstrated that SCALLS experiments can be used to monitor the solution crystallization and dissolution of polyethylene and polypropylene in a wide range of solvents. It was remarkable to find that the type of solvent plays an important role on Tm and Td of the polymer. The following conclusions on the unique features of the SCALLS approach and the novel results obtained can be made: (1) The SCALLS method is cost-effective and convenient without any complicated operations, having excellent control of the heating and cooling rates. The instrument has a very simple setup and it utilizes only very small amounts of solvent compared to commonly used fractionation methods such as TREF, CRYSTAF, and CEF. (2) For a set of solvents (TCB, oDCB, xylene, and decalin) crystallization and dissolution temperatures are found to vary much depending on the solvents used. (3) SCALLS allowed a fast detection of the cocrystallization of the PE/iPP blend components (unimodal SCALLS profile) in xylene. This could be considered as a reason for the presence of significant amounts of low to medium molar mass iPP in semicrystalline TREF fractions of EP copolymers.14−17 Furthermore, it was clearly shown that fractionation according to crystallizability strongly depends on the solvent type. For an optimized experimental protocol it is necessary to investigate the effect of different solvents on the solution crystallization and dissolution behavior of polyolefins with different molecular structures. Only in this case, procedures for maximum separation efficiency can be established. It can be assumed that the present results will change the common concept of using xylene as the fractionation solvent in many preparative fractionation methods. The present results indicate that the combination of different solvents in preparative fractionation processes can be an advantage. Here a new method for improved TREF separation is proposed, using TCB as a crystallization solvent (since cocrystallization effects are minimal) and xylene as a suitable solvent for elution (since codissolution effects are found to be minimal). It is suggested to perform prep TREF for a wide spectrum of olefin homo- and copolymers in order to further elucidate the discussed solvent effects.

■

REFERENCES

ASSOCIATED CONTENT

S Supporting Information *

Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*Phone: (02721) 808-3902. E-mail: [email protected]. 7023

dx.doi.org/10.1021/ac401700p | Anal. Chem. 2013, 85, 7019−7023