Polymer CO 2 systems exhibiting retrograde behavior and formation of nanofoams

Transcription

1 Polymer International Polym Int 56:67 73 (27) Polymer CO 2 systems exhibiting retrograde behavior and formation of nanofoams A Victoria Nawaby, 1 Y Paul Handa, 2 Xia Liao, 1 Yamamoto Yoshitaka 3 and Mizumoto Tomohiro 4 1 Institute for Chemical Processes and Environmental Technology, National Research Council of Canada, Ottawa, Ontario, K1A R6, Canada 2 Pactiv Corporation, Technology Center, 2651 Brickyard Road, Canandaigua, NY 14424, USA 3 Methane Hydrate Research Laboratory, National Institute of Advanced Industrial Science and Technology, Tsukuba West, Tsukuba , Japan 4 Basic Chemicals Research Laboratory, Sumitomo Chemical Co. Ltd, 5-1, Sobiraki-cho, Niihama, Ehime , Japan Abstract: The sorption of compressed gases in polymers causing a reduction in the glass transition temperature (T g ) is well established. There is, however, limited information on polymer gas systems with favorable interactions, producing a unique retrograde behavior. This paper reports on using a combination of established techniques of in situ gravimetric and stepwise heat capacity (C p ) measurements using high-pressure differential scanning calorimetry (DSC) to demonstrate the occurrence of this behavior in acrylonitrile butadiene styrene copolymer (ABS) CO 2 and syndiotactic poly(methyl methacrylate) (spmma) CO 2 systems. The solubility and diffusion coefficient of CO 2 intherangeto65 C and pressures up to 5.5 MPa were determined, which resulted in a heat of sorption of 15.5 and 15 kj mol 1, and an activation energy for diffusion of 28.3 and 32.1 kj mol 1 in the two systems, respectively. The fundamental kinetic data and the changes in C p of the polymer gas systems were used to determine the plasticization glass transition temperature profile, its relationship to the amount of gas dissolved in the polymer, and hence the formation of nano-morphologies. 26 Society of Chemical Industry Keywords: retrograde behavior; CO 2 ; nanofoams; acrylonitrile butadiene styrene copolymer (ABS); syndiotactic poly(methyl methacrylate) (spmma) INTRODUCTION The substantial solubility of a gas in a polymer affects its most fundamental property the glass transition temperature (T g ) for amorphous polymers and the crystallization temperature (T c ) for semi-crystalline polymers. 1,2 The relationship between T g and the pressure of the gas (p) with which the polymer is in equilibrium gives information on the extent of plasticization and changes in the thermal transitions. In most polymer gas systems the T g p relationship follows a linear path; however, in certain polymer gas systems with favorable interactions between molecules, a retrograde path is observed, 2,3 and subsequently two transitions are detected under a constant pressure. Theoretical and experimental data have confirmed the occurrence of this behavior for two polymer gas combinations: poly(methyl methacrylate) (PMMA) CO 2 and poly(ethyl methacrylate) (PEMA) CO 2 systems. 4,5 In both cases, CO 2 has been reported to induce high degrees of plasticization resulting in a rubber-to-glass transition at a lower glass transition temperature (T g,l ) and a glass-to-rubber transition at higher glass transition temperature (T g,h ). TheoccurrenceofT g,l is due to the higher solubility of the gas in the polymer at lower temperature, and heating the polymer gas solution at a constant pressure will result in some gas loss, and formation of microcellular morphology. Upon further heating at the same constant pressure, more gas loss occurs, followed by formation of nanocellular structures. The techniques established to date for measuring T g,l and T g,h at a given pressure include gas chromatography (GC), creep compliance, stepwise DSC measurements, and systematic measurements of solubility data. 4 8 With a large database on polymer gas systems, prediction of retrograde behavior remains a challenging task and establishing a method by which systems can be identified is useful. As demonstrated for the PMMA CO 2 system, polymer gas sorption kinetic data can serve as an important guide and tool in forecasting this behavior. Further, theoretical predictions and calculations of T g p profiles involve applying an adjustable interaction parameter, Correspondence to: A Victoria Nawaby, Institute for Chemical Processes and Environmental Technology, National Research Council of Canada, Ottawa, Ontario, K1A R6, Canada victoria.nawaby@nrc.ca (Received 2 December 25; revised version received 3 March 26; accepted 1 April 26) Published online 4 September 26; DOI: 1.12/pi Society of Chemical Industry. Polym Int /26/$3.

2 AV Nawaby et al. the value of which can be accurately estimated from the gas solubility data. Higher than usual gas solubility in polymers results in a higher degree of plasticization, and a change from glassy-to-rubbery state as observed by sharp changes in the diffusion coefficient. 8 Therefore, systematic measurement of solubility data not only serves as a screening method for the occurrence of this phenomenon in polymers, but also aids in developing better theoretical models to calculate interaction parameters. Thus far, PMMA CO 2 is the only system that has been demonstrated through solubility data to exhibit a retrograde behavior. However, a recent investigation of the interaction of PMMA and its stereoisomers with CO 2 reports on the blending of isotactic PMMA (ipmma) and syndiotactic PMMA (spmma), their subsequent treatment with CO 2 at supercritical conditions, and formation of microcellular foams. 9 Stereoregular spmma has received much attention due to its ability to form helicoidal or double-helix structures found in biological systems and not normally observed in synthetic polymers. 1 Additionally, this study reports the morphologies obtained from stereocomplex mixtures to be superior to that obtained from pure PMMA, and therefore an interesting system to be further characterized for its interaction with CO 2. Current data for acrylonitrile butadiene styrene copolymer (ABS) CO 2 system indicate favorable interactions and a wider CO 2 processing window not previously explored. 11 Therefore, this paper reports on the phase behavior and phase transitions in ABS CO 2 and spmma CO 2 systems, the occurrence of the retrograde behavior, and the formation of nanofoams. EXPERIMENTAL Materials ABS polymer (T g = 17 C, density =.94 g cm 3 ) was supplied by Dow Chemicals (Magnum 346) and spmma (T g = 11 C, density = 1.47 g cm 3 ) was supplied by Sumitomo Chemical Co. Bone dry CO 2 was used. The samples were compression molded at 21 C into sheets 1 7µm thick and air quenched. Wideangle X-ray patterns of the samples were obtained using a Bruker AXS Inc. instrument at a radiation wavelength of Å, and a scan range of 2θ from 1to4 with a scan speed of.9 min 1. Sorption kinetic studies An in situ gravimetric technique (CAHN D11 microbalance) was used to investigate the sorption kinetics of CO 2 in the polymer samples in the temperature range 65 C, and pressures up to 5.5 MPa. The detailed description of this gravimetric technique has been reported elsewhere. 12 Approximately.3.5 g samples, in the form of 3 µm disks, were placed in the balance and the entire system was degassed for 48 h, or until no further change in weight was observed. Before each sorption run, CO 2 gas was preheated to the experimental temperature, sorption steps at various pressures and constant temperature were carried out, and weight changes over time were monitored. Blank runs were also performed under the same experimental conditions providing the balance zero-shift as a function of pressure. The equilibrium weight readings obtained in the experimental studies were then corrected for the balance zero-shift, and buoyancy effects. Buoyancy corrections were applied to the data since a small volume difference between the sample and reference side in the balance develops as the polymer dilates due to gas dissolution. Diffusion coefficients of the gas in the polymer were subsequently derived from the corrected sorption kinetic data. 12 High-pressure DSC The pressure dependence of T g can be established either from the diffusion coefficients obtained via sorption kinetic data, or directly by employing the stepwise C p measurement technique using highpressure DSC (Setaram DSC 121) as reported by Handa and Zhang. 2 The technique for obtaining high-pressure measurements with the DSC 121 involving polymers has been reported elsewhere. 7 The Setaram DSC 121 has a symmetrical furnace which houses both the sample and reference vessels. Each vessel has a pressure rating of 1.1 MPa and is connected to a gas handling manifold and reservoir. For each run the polymer sample was evacuated for a few hours in order to degas the specimen; subsequently the sample and reference sides were pressurized to the desired value. After allowing the system to reach equilibrium, the sample was scanned while it remained in contact with CO 2 gas. C p measurements of the system were carried out in the stepwise temperature mode in the range 12 C while holding the gas pressure constant. In order to establish the baseline at the initial temperature (T i ), the system was held at the desired starting temperature for 15 min. The step change in the temperature was set to 2 Cmin 1 followed by holding the system isothermally for 15 min to reach and establish the baseline for final temperature (T f ). Since the step changes were small, the holding time of 15 min was comparable to the CO 2 diffusive time scale for a 25 µm thick sample used in sorption and DSC studies. The temperature step change was continued until the entire test temperature range was captured at constant pressure. This procedure was repeated at several saturation pressures and the T g p profile was established. Data were collected for blank runs using the same experimental procedure and conditions, and used to correct the experimental data. T g was therefore taken as the onset temperature as described by Handa and Zhang Polym Int 56:67 73 (27) DOI: 1.12/pi

3 Polymer CO 2 systems: retrograde behavior and nanofoams Foaming experiments Polymer sheets, 65µm thick, were conditioned with either liquid or gaseous CO 2 for 24 h. The pressure in the vessel was released slowly and the gasladen polymer was rapidly heated by transferring the specimens to a water bath in the range 3 1 C where they stayed for 9 s. The cellular morphology was frozen by subsequently quenching the samples in an ice-water slurry. Foam samples were characterized for density and structural properties using the total immersion method and field emission scanning electron microscopy (Hitachi S-48), respectively. RESULTS AND DISCUSSION The sorption kinetics of CO 2 in ABS and spmma was investigated by recording the changes of the weight uptake in the polymer over time until a steady-state value was obtained over the desired pressure and temperature range. The higher rate of gas sorption and hence plasticization in both polymers resulted in a decrease in the time required to reach equilibrium with an increase in gas pressure. Sorption and desorption experiments carried out at 25 C resulted in a higher solubility observed during desorption runs in the case of the ABS CO 2 system. Due to irreversible dilation in glassy polymers, sorption and desorption of gas in a matrix frequently exhibit a hysteresis effect. 12 The irreversible swelling in the polymers during adsorption experiments increases the number of possible sorption sites at a given pressure during the desorption cycle, and hence the polymer exhibits hysteresis. 13 In the case where a sorptioninduced glass transition occurs during the sorption run, the sorption isotherms follow a linear relationship with pressure. ABS exhibited a greater hysteresis effect in comparison to spmma, although their solubility capacity for CO 2 during sorption was similar. The equilibrium solubility of gases in polymers in the rubbery state exhibits Henrian behavior and follows the dual sorption model in the glassy state. 8,12 In polymer gas systems with a higher gas solubility and enhanced plasticization effect, a reduction in T g is observed, and as the solubility increases the polymer behaves increasingly like a rubber. This observation is evident in the sorption curves obtained for ABS CO 2 and spmma CO 2 in the range 65 C(Fig.1).All sorption isotherms as a function of pressure are linear and follow Henrian behavior. The solubility is thus approximated by S = k d p (1) where S is the solubility, p the gas pressure, and k d Henry s law constant. The k d values determined for both polymer gas systems varied from 7.36 to 1.98 mg CO 2 (g polymer atm) 1 in the tested temperature range. The temperature dependence of the solubility data is given in Table 1 and by the following equation: ( ) HS S = S exp (2) RT S /mg CO 2 * g polymer -1 S /mg CO 2 * g polymer ABS - CO 2 spmma- CO 2 C 1 o C 25 o C 35 o C 45 o C 55 o C 65 o C C 1 o C o C 35 o C 45 o C 55 o C 65 o C Figure 1. Solubility of CO 2 below T g of the neat polymer. The curves are drawn through the points to show the trend in the data. Table 1. Heats of sorption and diffusion H S (kj mol 1 ) H D (kj mol 1 ) PMMA CO 2 (Handa et al. 8 ) ABS CO spmma CO where S is the pre-exponential factor and H S is the heat of sorption, for which values of 15.5 and 15 kj mol 1 for ABS CO 2 and spmma CO 2, respectively, were obtained. Diffusion coefficients at various pressures and temperatures were obtained by fitting the sorption kinetic data to a hybrid model that combines both short- and long-term Fickian diffusion equations. 12 The temperature dependence of diffusion coefficients is given by ( ) HD D = D exp (3) RT where D is the pre-exponential factor and H D is the activation energy for diffusion following a linear path in the tested temperature range, yielding values of 28.3 and 32.1kJmol 1 for ABS CO 2 and spmma CO 2 systems, respectively. The reported heat of sorption and activation energy for diffusion is different as observed in the PMMA CO 2 system (Table 1). 8 The Polym Int 56:67 73 (27) 69 DOI: 1.12/pi

4 AV Nawaby et al. heat of sorption is a representation of the heat of condensation of the gas in the polymer matrix and a similar value is obtained since the gas is common in the polymer gas systems investigated here and in the system reported by Handa et al. 8 The energy required for a gas molecule to overcome the barriers in the polymer matrix and to diffuse is given by the activation energy for diffusion. Due to different polymer structures, the barrier for the gas molecule to make a series of activated jumps is different and therefore a different activation energy for diffusion is observed. Plasticization of polymers by dissolved gas can cause a transition from glass to rubber, and therefore one can observe a sharp increase in the diffusion coefficient as a function of pressure. The pressure at which the sharp changes in the diffusion coefficient takes place (Fig. 2) is defined as the glass transition pressure (p g ). This sharp change in the diffusion coefficient is dependent on the gas pressure and the experimental sorption temperature. As such, the diffusion coefficient for each sorption isotherm as a function of gas pressure was obtained and values of p g for the systems were determined. Although sharp changes in the diffusion coefficient of the polymer gas system indicate a phase transition, other parameters such as the change in C p of the system can serve as confirmation of this observation. 2 The procedure as described by Handa and Zhang 2 for a stepwise high-ressure DSC temperature scan was followed and Fig. 3 shows a graphical representation of the results for the ABS system saturated with CO 2 at 2.9 Mpa. The corrected heat flow was obtained by subtracting the blank run under the same conditions from the sample run. Integration of the peaks for a step change in temperature (i.e. T f T i )resultedin the enthalpy change ( H). C p for the polymer gas system was determined from the above parameters at (T f + T i )/2. Figure 4 shows the heat capacity changes in the ABS saturated systems under a CO 2 pressure of 2.9 MPa. The change in the slope of the C p T curve represents a phase transition taking place in 1 8 D /cm 2 s p g p /MPa Figure 2. Diffusion coefficient of CO 2 in spmma at 45 C. Tangents to the curve indicate the glass transition pressure of the system. Heat Flow endo Corrected Heat Flow Temperature time / s T / C Figure 3. Stepwise DSC scans under a CO 2 pressure of 2.9 MPa in ABS. Each temperature jump was 2 Cmadeatarateof1 Cmin 1. Isothermal hold before and after each jump was 15 min. The corrected heat flow was obtained by subtracting the blank run from theabsrunat2.9mpa. C p / J g -1 C T g T / C Figure 4. C p of the ABS CO 2 system as a function of temperature for saturation condition of 2.9 MPa. Tangent to the curve is drawn to indicate the glass transition temperature of the system. the system and the point at which the change occurs is noted as T g. Data for the ABS CO 2 and spmma CO 2 systems were collected for a series of saturation conditions and T g wasthenplottedasa function of pressure (Fig. 5). The T p g values derived from the diffusion coefficients are plotted in Fig. 5 and compared to the plasticized T g obtained from the C p measurements and are found to be in good agreement. The polymer gas systems exist as a glass within the T g p envelope and as a rubber outside the retrograde envelope. One of the very interesting features of polymer gas systems exhibiting retrograde behavior is represented by the two distinct T g values observed under a constant gas pressure. For example, for the ABS CO 2 system (Fig. 5(a)), the lower glass transition temperature (T g,l ) occurs at about 37 C and the higher glass transition temperature (T g,h ) occurs at about 75 C for a pressure of 4.5 MPa. Upon heating the gas-laden polymer at a constant pressure of 4.5 MPa, the system will lose some gas and will switch from rubber to glass at 37 C. A further increase in temperature at constant pressure of 7 Polym Int 56:67 73 (27) DOI: 1.12/pi

5 Polymer CO 2 systems: retrograde behavior and nanofoams (a) (b) T g / C Tg / C rubber glass T (g,h) T (g,l) rubber glass T (g,h) T (g,h) Foam Density / gcm CO 2 saturation C and 3.4 MPa. CO 2 saturated at room and 5.5 MPa T / C Figure 6. Density of ABS foams as a function of CO 2 saturation conditions and foaming temperature. A solid curve is drawn through the points to show the trend Figure 5. T g of (a) ABS CO 2 and (b) spmma CO 2 systems as a function of saturation pressure:, solubility measurements; ž,dsc measurements. Solid lines are drawn through the points to show the trend. 4.5 MPa will result in more gas loss, a relaxation of the polymer as is normally observed on heating, and finally a transformation from glassy-to-rubbery state at 75 C. The retrograde behavior reported in this work is similar to that reported for the PMMA CO 2 system. 8 The results for spmma are not surprising since this polymer is the syndiotactic form of PMMA and X-ray diffraction patterns of the prepared samples indicate it be amorphous in nature. However, in the case of ABS, the finding is very interesting and it is anticipated that polymers showing CO 2 solubility and an extent of plasticization comparable to PMMA under subcritical conditions will exhibit a retrograde behavior, and thus will be of interest for further investigation. There are two methods of pressure quenching and temperature soaking used in the formation of cellular plastics with gases. In the process of pressure quenching, the polymer gas solution is depressurized rapidly; in the temperature soaking method, the solution is depressurized slowly and heated rapidly. In either method the polymer gas system should be above its plasticized T g for foaming to occur, and the foaming conditions can be determined from the T g p profile. 2 In the ABS CO 2 system, foaming occurred at temperatures above 6 Cwhen (a) (b) Figure 7. ABS CO 2 blown foams at 6 C: (a) CO 2 saturation conditions of room and 5.5 MPa; (b) CO 2 saturation conditions of C and 3.4 MPa. the solution was prepared at 25 C and 5.5MPa. Comparatively, for solutions prepared at C and 3.4 MPa, morphologies with high cell densities were Polym Int 56:67 73 (27) 71 DOI: 1.12/pi

6 AV Nawaby et al. The foam densities obtained in the range 3 1 C using two CO 2 saturation conditions for spmma are presented in Fig. 9. Even though foam densities (a) Foam Density / g cm CO 2 saturation condition of 25 C and 5.5 MPa. CO 2 saturation condition of C and 3.4 MPa Foaming Temperature / C Figure 9. Density of spmma foams as a function of CO 2 saturation conditions and foaming temperature. A solid curve is drawn through the points to show the trend. (b) Figure 8. ABS CO 2 blown foams at 9 C: (a) CO 2 saturation conditions of room and 5.5 MPa; (b) CO 2 saturation conditions of C and 3.4 MPa. obtained at 6 C. Foam characteristics such as cell density and cell size for this system were determined, and a minimum cell size and maximum cell density was obtained at 9 C when the polymer was saturated with CO 2 at room temperature and 5.5 MPa. As gas solubility increased with a decrease in temperature, the polymer gas solution prepared at C and 3.4 MPa resulted in cell nucleation occurring at much lower temperature, causing the occurrence of a minimum cell size at 6 C and a minimum density at 9 C (Fig. 6). Scanning electron microphotographs of ABS foam prepared at 6 C and different saturation conditions are presented in Fig. 7. With an increase in foaming temperature and an increase in the amount of dissolved gas in the polymer, more cells are nucleated and thus a nanocellular ABS morphology is obtained (Fig. 8). Although there is no significant decrease in cell size with an increase in the foaming temperature, above 6 C the cell density is still increasing and a maximum value of cells g 1 and a cell size of.47 µm were obtained at saturation conditions of Cand 3.4 MPa. (a) (b) Figure 1. spmma CO 2 blown foams at 7 C: (a) CO 2 saturation conditions of room and 5.5 MPa; (b) CO 2 saturation conditions of C and 3.4 MPa. 72 Polym Int 56:67 73 (27) DOI: 1.12/pi

7 Polymer CO 2 systems: retrograde behavior and nanofoams are similar, the cellular characteristics of the foams obtained at saturation conditions of Cand34atm are markedly different. Figure 1 shows scanning electron microphotographs of foams prepared at 7 and 9 C using the two saturation conditions. The average cell size and cell density obtained for this polymer gas system at saturation conditions of C and 3.4 MPa, and subsequently foamed at 9 C, are.8 µm and cells g 1, respectively. The corresponding values for the saturation condition of room temperature and 5.5 MPa are 5.5 µm and cells g 1. CONCLUSIONS The phenomenon of retrograde behavior in polymer gas systems has been established theoretically and experimentally for PMMA and PEMA while in contact with CO 2. This study demonstrates that this behavior also occurs for the ABS CO 2 and spmma CO 2 systems. Our finding was confirmed from T g measurements using a high-pressure DSC technique as well as systematic measurements of gas solubility. The sorption kinetic behavior in the systems investigated was found to be very similar to that of the PMMA CO 2 system. ACKNOWLEDGEMENTS The authors thank Mr. Gerry Pleizier and Mr. David Kingston for assistance with SEM work. The authors also thank Dr. Pamela Whitfield and Dr. Jiangfu Ding for their assistance with X-ray and GPC data, respectively. REFERENCES 1 Goel SK and Beckman E, J Polym Eng Sci 34:1148 (1994). 2 Handa YP and Zhang Z, J Polym Sci B: Polym Phys 38:716 (2). 3 Condo PD, Sanchez IC, Panayiotou CG and Johnston KP, Macromolecules 25:6119 (1992). 4 Wissinger RG and Paulaitis ME, J Polym Sci B: Polym Phys 29:631 (1991). 5 Condo PD and Johnston KP, Macromolecules 25:673 (1992). 6 Kikic I, Vecchione F, Alessi P, Cortesi A and Eva F, Ind Eng Chem Res 42:322 (23). 7 Zhang Z and Handa YP, J Polym Sci B: Polym Phys 36:977 (1998). 8 Handa YP, Zhang Z and Wong B, Cell. Polym 2:1 (21). 9 Mizumoto T, Sugimura N, Moritani M, Sato Y and Masuoka H, Macromolecules 33:6765 (2). 1 Saiani A and Guenet J-M, Macromolecules 3:966 (1997). 11 Murray RE, Weller JEandKumar V, Cell. Polym 19:413 (2). 12 Wong B, Zhang Z and Handa YP, J Polym Sci. B: Polym Phys 36:225 (1998). 13 Kamiya Y, Hirose T, Naito Y and Mizoguchi K, JPolymSciB: Polym Phys 26:149 (1988). Polym Int 56:67 73 (27) 73 DOI: 1.12/pi