The effect of penetrant activity and temperature on the anomalous diffusion of hydrocarbons and solvent crazing in polystyrene part I: Biaxially oriented polystyrene

The effects of temperature and penetrant activity on the sorption kinetics and equilibria of a series of alkanes in glassy, biaxially oriented polystyrene were studied. Normal isomers of pentane, hexane, and heptane cause crazing of polystyrene film samples at high penetrant activities (> 0.85). Crazing kinetics are identical to the kinetics of Case II transport. Transport of these normal hydrocarbons in glassy polystyrene in the temperature range 25 to 50°C is markedly non-Fickian; limiting Case II transport is observed at activities in exces of 0.6. Sorption appears to be controlled by highly activated relaxation processes including primary bond breakage at these high penetrant activities. Fickian diffusion behavior is approached, however, as penetrant activity is reduced. Sorption of the branched isomers of these compounds does not result in polymer microfailure. The sorption kinetics of the branched isomers, although time dependent, appear to be controlled primarily by thermally activated diffusion rather than large scale polymer relaxations which control Case II transport.     

Kinetics of vapor and liquid transport in glassy polyblends of polystyrene and poly (2,6-dimethyl 1,4-phenylene oxide) Part II

The kinetics of n-hexane vapor and n-hexane liquid sorption in solution cast films of polystyrene, poly(phenylene oxide), and blends of these homopolymers were studied over a significant range of temperature and penetrant activity. The kinetics of concomitant solvent crazing, apparent at high penetrant activities, were also monitored. In all cases the kinetics of sorption and crazing obeyed predominantly Case II or relaxation-controlled behavior. Although the rates of crazing and sorption at high activities was much more rapid in the homopolymers than in the blends, the sorption rate at lower penetrant activities increased monotonically with increasing poly(phenylene oxide) content. These kinetic results qualitatively superimpose upon the equilibrium relationships reported in Part I of this series. Specifically, at a fixed temperature, the rate of crazing was virtually a unique function of equilibrium n-hexane content independent of polymer composition and largely independent of penetrant activity. The coupling between rate of sorption and equilibrium penetrant content was dramatic. In limiting cases, the sorption rate increased by a factor of 10,000 apparently due to a twofold increase in equilibrium penetrant concentration. This seemingly complicated kinetic behavior is explained rather simply in terms of the equilibrium relationships between organic penetrants and polymeric glasses developed and interpreted in the first part of this series.     

The effect of molecular weight and orientation on the sorption of n-pentane by glassy polystyrene

The effects of polymer molecular weight, molecular weight distribution, and orientation on the rate of relaxation-controlled sorption of n-pentane by glassy polystyrene were studied. The sorption follows Case II kinetics but for films which sorb slowly the sorption rate increases at relatively long times until sorption is sharply terminated. This rate increase may be explained by the development of dispersed microvoids within the unrelaxed film core. Overshoot of the equilibrium n-pentane content occurs in sorption experiments in which accelerated sorption is pronounced. The sorption rate is independent of polymer molecular weight and molecular weight distribution per se over a broad wrange of these parameters. Essentially identical vapor sorption kinetics were observed for well annealed polystyrene films of different molecular weights and distributions. Conversely, for vapor sorption by uniaxially oriented films and for liquid sorption by partially annealed films, high molecular weight film (1,880,000) exhibits greater sorption rates than low molecular weight film (ca. 200,000). These differences in rate are not due to molecular weight differences per se, but are a consequence of the dissimilar response of free volume and strain development for films of different molecular weight prepared with a given time-temperature-strain history. Crazing of carefully annealed polystyrene films occurs during desorption of n-pentane from partially saturated films. The depth of craze penetration reflects the point of advance of the discontinuous Case II sorption boundary.     

Sorption kinetics and equilibria in annealed glassy polyblends

The sorption kinetics and equilibria of n-hexane in glassy polyblends of polystyrene and poly(2,6 dimethyl-1,4 phenylene oxide) were studied as a function of annealing conditions. Cast film samples were annealed 20°C above their respective glass transition temperatures for two hours and twenty-four hours. The rate of relaxation-controlled (Case II) sorption of n-hexane in these films was reduced markedly consequent to annealing. The effect of annealing on the sorption kinetics and the independently determined film densities was more pronounced for the poly(phenylene oxide)-rich samples. Although sorption rates were reduced by as much as a factor of 100, the sorption equilibrium was insignificantly affected by annealing. Super Case II transport was observed for the slow absorbing annealed samples whereas the more rapid sorption in the unannealed samples followed ideal Case II kinetics. The more pronounced effects of annealing for the poly(phenylene oxide)-rich samples on sorption rates and film densities were explained by considering the increasing difference between the film T_g and the drying temperature used in the original film preparation for the poly(phenylene oxide)-rich samples. These results suggest that glassy polymers, cast and dried well below their glass transition temperatures, will be subject to large long-term reductions in absorption rates and specific volume. Moreover, residual, excess free volume significantly affects relaxation-controlled absorption of vapors in partially annealed glassy polymers.     

Residual solvent removal and n-hexane sorption in blends of atactic and isotactic polystyrene

Atactic polystyrene, isotactic polystyrene, and intermediate polyblend films of the homopolymers were cast from solutions of 1,3,5-trimethylbenzene (mesitylene) or o-chlorotoluene. Residual solvent content, glass transition temperatures, and densities of the polyblends were determined. Consequent to these preliminary characterizations, n-hexane absorption kinetics and equilibria in unannealed blends were determined; maxima in the plots of weight change versus time were apparent for all compositions. The maximum is a consequence of the ‘unlocking’ of previously trapped residual casting solvent consequent to plasticizing invasion by the n-hexane in the unannealed film specimens. Neutron activation analysis for chlorine confirmed that the o-chlorotoluene content in the films was reduced dramatically consequent to n-hexane absorption. Residual solvent content was also reduced by thermally annealing the films at 110°C for 48 h under a vacuum and, therefore, no maximum in the sorption/time behaviour was observed in thermally annealed samples nor in films previously contacted with n-hexane vapour and subsequently vacuum stripped. Both the rate of sorption and the apparent sorption equilibrium are greatly reduced as the isotactic polystyrene content is increased. Predominantly Case II or relaxation controlled absorption kinetics were observed in the polyblends consequent to thermal annealing and/or ‘solvent’ annealing. Although the form of the sorption kinetics was quite similar for absorption of n-hexane in thermally annealed or ‘solvent-annealed’ films, the sorption rate and amount of penetrant sorbed at apparent equilibrium were both larger for the thermally annealed films. These results suggest that the history dependence, which so frequently affects polymeric glasses, appears to confound a simple interpretation of the sorption experiments reported here. Consistent with observations in related systems, Super Case II kinetics were observed for the more slowly sorbing films which, in this case, contained isotactic polystyrene. Moreover, film whitening related to micro-voiding accompanied n-hexane sorption in blends containing atactic polymer; however, the whitening in the isotactic polymer was less pronounced and densification, presumably related to solvent induced crystallization, occurred consequent to long term exposure of the isotactic polymer to n-hexane.     

Effect of particle size on the mechanism controlling n-hexane sorption in glassy polystyrene microspheres

The effect of particle size on the rate-determining transport mechanism controlling n-hexane sorption into polystyrene spheres was determined by monitoring the kinetics of n-hexane vapour absorption in two powder samples of distinctly different narrow particle size distributions. The microspheres differed in mean diameter by a factor of three hundred. In all cases, sorption in the smaller spheres was controlled principally by Fickian diffusion. In contrast, at temperatures and activities which result in relaxation-controlled (Case II) transport in films, the larger spheres (184 μm diameter) sorb by Case II kinetics. Under these identical boundary conditions, however, the small spheres (5340 Å diameter) absorbed n-hexane by Fickian kinetics. Presumably, there is insufficient time or space in the small spheres to develop the step concentration profiles associated with limiting Case II transport. A mathematical model describing Case II sorption in spheres cylinders, and slabs is developed. The kinetics describing absorption in the larger spheres are well described by this analysis. Desorption in all cases is adequately described by concentration-dependent diffusion. Repeated experiments confirmed, moreover, that at identical temperatures and penetrant vapour activities the apparent equilibrium concentration of n-hexane in the small spheres is significantly higher than the corresponding sorption equilibria in the large spheres. Although several possible explanations for this glassy-state anomaly are presented, the most satisfying explanation involves the development of a non-sorbing core in the large spheres consequent to the prior sorption in the surrounding shell. Desorption data are consistent with the notion of partial swelling of large spheres.     

Controllably crazed polystyrene: Morphology and permeability

The effects of n-heptane and heat treatment on the structural and transport properties of polystyrene films (biaxially oriented and unoriented) were studied to determine whether these treatments improve the film as selective barriers for separation of molecules differing only slightly in size and shape. n-Heptane treatment of biaxially oriented polystyrene produces a sandwich structure composed of expanded, crazed, surface layers surrounding an apparently unaffected central core. The crazed layers contain a continuous network of interconnected channels. The core provides the total resistance to gas permeation, hence, the overall effect of n-heptane treatment is fabrication of a thinner more permeable membrane. By restricting the stress-cracking treatment to one face of the film, it should be possible to make high flux, anisotropic membranes—a type of membrane which is required for successful development of membrane separation processes. n-Heptane treatment of cast, annealed polystyrene results also in a crazed polymer, but the crazing is in the form of spherical voids, and the films, even with a residual uncrazed core, are too weak to be useful as separation membranes. The crazing process in both types of polymer specimens is characteristic of case II non-Fickian diffusion in which the kinetics are apparently controlled by polymer relaxation processes. Sorption of isopentane into cast, annealed polystyrene does not cause visible crazing but the kinetics are again non-Fickian. Desorption of isopentane into n-heptane-treated polystyrene releases the appreciable residual n-heptane in the film which could not be removed by long-term exposure to vacuum. Analysis of D(0) values for isopentane in n-heptane treated films indicates that the polymer surrounding the visible voids in the film is essentially unaltered polystyrene with only a small fraction of the voids being interconnected by open channels.     

Sorption and transport of methanol in poly(ethylene terephthalate)

This paper reports the sorption and diffusion characteristics of methanol vapor in polyethylene terephthalate (PET). Amorphous PET, semicrystalline, biaxially oriented annealed and non-annealed samples have been studied for equilibrium sorption and kinetics of methanol. At activities of methanol less than 0.30, uptake shows Fickian kinetics and isotherm follows the Dual Mode model. Diffusion coefficients increase with penetrant concentration and are of the order of 10⁻¹⁰ cm²/s. Hysteresis during desorption and increase in solubility during resorption suggest methanol induced conditioning effects which may have detrimental effects on the barrier efficacy of PET. At activities greater than 0.30, swelling and relaxation effects occur and the isotherms show Flory-Huggins behavior for all three samples. Uptake follows two-stage kinetics fit by the Berens-Hopfenberg model. Greater polymer chain stability due to annealing reduces the extent of relaxation and improves the barrier efficacy over amorphous and non-annealed, oriented PET. For amorphous PET, at 80% activity and above, an induction time is observed which is absent in the semicrystalline films, suggesting strong relaxation effects in the amorphous phase of PET.     

The effect of orientation on the morphology and kinetics of solvent crazing in polystyrene

Unoriented, uniaxially oriented, and biaxially oriented polystyrene films were crazed by immersing samples in liquid n-hexane at 45°C. The craze morphology and crazing kinetics were studied as a function of the preorientation and thermal histories of the polymer films. The shape of the micropores was related to the degree of orientation of the film. Ellipsoidal microvoids were formed on the surfaces of uniaxially oriented films containing a residual glassy core. Unoriented samples displayed spherical microvoids 0.1 to 3 μ in diameter. The major-to-minor axis ratio of the micropores increased monotonically from 1/1 to 10/1 as the sample orientation was increased from 0% to 200%. The kinetics of the crazing process similarly increased with uniaxial orientation. Surfaces of unannealed films which crazed mainly during the sorption of liquid n-hexane displayed numerous micropores when examined by the SEM. Conversely, films which were extensively annealed crazed mainly during desorption regardless of subsequent orientation. These extensively annealed films exhibited surfaces with many fine cracks and few, if any, micropores.     

Sorption and transport of linear and branched ketones in biaxially oriented polyethylene terephthalate

Equilibrium sorption and kinetics of acetone, methyl ethyl ketone (MEK), methyl n-propyl ketone (MnPK), and methyl i-propyl ketone (MiPK) uptake in uniform, biaxially oriented, semicrystalline polyethylene terephthalate films were determined at 35 °C and low penetrant activity. Sorption isotherms for all penetrants were well described by the dual-mode sorption model. Sorption and desorption kinetics were described either by Fickian diffusion or a two-stage model incorporating Fickian diffusion at short times and protracted polymer structural relaxation at long times. Diffusion coefficients and equilibrium solubility at fixed relative pressure decreased in the following order: acetone>MEK>MnPK>MiPK. Diffusion coefficients for each penetrant increased with increasing penetrant concentration.     

Fluid sorption characterization of PEEK matrices and composites

Neat poly(ether-ether-ketone) (PEEK) and carbon fiber reinforced PEEK (APC-2) specimens were prepared using a variety of cooling rates to achieve a range of crystallinities. Amorphous specimens were exposed to a variety of fluids to determine the penetrant types which are able to strongly influence the material. This allowed the estimation of the solubility parameter and hydrogen bonding index for PEEK to be 9.5 and 3.1, respectively. Methylene chloride was used to investigate the kinetics of penetrant sorption. The data demonstrated Case II behavior, with the initial crystallinity having a pronounced effect on both the kinetic and equilibrium data. Accordingly, a model was proposed capable of describing the sorption level and penetration depth as a function of time given the sample crystallinity and sorption temperature. With Case II behavior there was no difference in the sorption kinetics of neat and fiber reinforced PEEK. Finally, the dynamic mechanical properties measured during sorption were found to be dependent on the sorption process.     

Network structure and methanol transport dynamics in poly(methyl methacrylate)

The effect of the polymer network structure on methanol transport dynamics in glassy polymers was investigated in both dry and plasticized disks of poly(methyl methacrylate) (PMMA) through gravimetric integral sorption studies. PMMA was synthesized by a controlled free radical polymerization mechanism, crosslinked in bulk with ethylene glycol dimethacrylate, and swollen in methanol under a variety of conditions. In the Case II transport regime, control over the transport rate was shown to depend on the glassy‐state properties of the polymer, and the Case II front velocity was found to be proportional to the square root of the crosslinking density. Similarities were observed in the penetrant transport behavior of both dry and plasticized samples at high degrees of crosslinking, and the activation energy of methanol transport at low degrees of crosslinking was found to be similar for both Fickian and Case II mechanisms. © 2011 American Institute of Chemical Engineers AIChE J, 2012     

The transport of methyl methacrylate monomer in poly(methyl methacrylate)

The absorption kinetics and equlibria of methyl methacrylate monomer into poly(methyl methacrylate) were studied over a range of penetrant activities. The interval sorption kinetics at elevated activities were determined, compared, and contrasted with the integral sorption experiments in previously unpenetrated film samples. The sorption kinetics in previously unpenetrated films were predominantly case II or relaxation controlled at high activities. A Fickian contribution to the overall kinetics was apparent at lower activities. In contrast, interval sorption, at elevated activities in previously equilibrated and plasticized samples, followed Fickian kinetics rather closely, whereas resorption, over an activity range which involved a traversal of the effective T_g, was characterized by more complicated kinetics involving a super case II mechanism at long times. These composite results reinforce the notion that the kinetics describing penetration of a single penetrant into a single polymer are extremely sensitive to the boundary condition imposed upon the polymeric sorbent.     

Time and temperature dependent sorption in poly-ether-ether-ketone (PEEK)

The possible multimodal sorption mechanisms in glassy amorphous poly-ether-ether-ketone (PEEK) are presented. By varying the penetrant-polymer affinity, experimental temperature, and external solvent activity, a broad range of sorption behaviors from ideal Fickian diffusion to limiting relaxation controlled kinetics is observed. In particular, water, methylene chloride, and n-heptane sorption kinetics are analyzed and interpreted on the basis of the multiple transport mechanisms. Low uptake liquid n-heptane sorption follows ordinary Fickian diffusion. Analogously, water vapor at low activity, is sorbed in small amounts in the same limiting mode while, at higher activities, the moderately higher penetrant uptakes induce slow relaxation coupled with ideal Fickian diffusion. The highly interacting methylene chloride leads to ideal Fickian diffusion only at very low activities, while anomalous non-ideal Fickian diffusion and limiting Case II and diffusion controlled swelling are observed at moderate and at high solvent activities, respectively. Limiting Case II sorption of methylene chloride in PEEK has been observed only at a very low temperature (?32°C). The optical microscopy observations of cryogenically fractured samples contacted with liquid methylene chloride at 5, 20 and 36°C revealed the presence of a sharp front moving linearly with the square root of time. Solvent induced crystallization in methylene chloride swollen samples was detected by means of differential scanning calorimetry (DSC) and wide angle X-ray scattering (WAXS). Finally, sorption from liquid methylene chloride/n-heptane solutions with varying compositions are presented. The progressive increase of the more high sorbing methylene chloride concentration in the solutions, leads to the same wide variety of sorption behavior observed in the methylene chloride vapor sorptions. The gas chromatographic (GC) analysis indicated that the presence of methylene chloride enhanced the n-heptane sorption in the polymer.     

INFLUENCE OF RHEOLOGICAL PROPERTIES IN MASS TRANSFER PHENOMENA: SUPER CASE II SORPTION IN GLASSY POLYMERS

The sorption rate of swelling penetrants in glassy polymers has been considered as controlled both by the swelling kinetics and the penetrant diffusion through the swollen layer. The stress exerted on the glassy core at the moving boundary is the driving force for the swelling, and results from an osmotic stress and a differential swelling stress contribution. During the sorption process, the osmotic stress at the moving boundary decreases, due to the increasing diffusion resistance; the differential swelling stress, on the contrary, increases giving rise to an acceleration of the front velocity (Super Case II).

The particular case of negligible diffusion resistance in the swollen region is here considered in more detail. It is shown that the rheological properties of both swollen and glassy phases crucially enter the mass transport problem; thestress relaxation in the swollen region must be taken into account in order to obtain a thickness dependent Super Case 11 effect     

Characterization and effects of n-alkane swelling of polystyrene sheets

Detailed penetration kinetics of a series of normal alkanes in 2 mm polystyrene sheets have been determined by an optical method over a broad range of temperatures encompassing the effective glass transition temperature of the resulting sheets. Whereas earlier thin film results, involving identical thermodynamic boundary conditions, have been characterized by ideal Case II transport, the results obtained here appear to be confounded at long times by a significant diffusional resistance in the swollen surface regions surrounding the unplasticized central core of these relatively thick sheets. At relatively low temperatures, the rate determining relaxations, at the boundary between swollen and unswollen polymer, are sufficiently slow to ensure ideal Case II transport over the entire course of the sorption history. These swelling experiments were complemented by Clash—Berg measurements of the temperature-dependence of the 10 sec torsional shear moduli of partly penetrated sheets, completely swollen sheets and unswollen sheets. The Clash—Berg results suggest that the T_g of the n-hexane swollen sheet is approximately 20°C and that the properties of the swollen regions of partly penetrated specimens are identical to the properties of the completely swollen sheet. A model based on the core-shell morphology, induced by the prior swelling, satisfactorily describes the observed mechanical properties of these swelling induced composites.     

The effect of sorbed penetrants on the aging of previously dilated glassy polymer powders. IV

Uniform, submicron-diameter polystyrene (PS) and poly(methylmethacrylate) (PMMA) microspheres were dilated by preswelling with pure organic vapors followed by rapid removal of the preswelling penetrant by protracted evacuation of the preswelling chamber to a pressure of 10^?3 mm Hg. Aging of the preswollen polymers was carried out both in vacuum and in the presence of various penetrants at sorbed concentrations typically less than 2 wt%. Inferences about relaxations of the polymers were based upon changes in concentrations of the penetrants within the microspheres, at a given temperature and penetrant activity, which result from aging in vacuum or in the presence of penetrant. The kinetics of the relaxations were monitored by probing the expanded glasses with relatively low concentrations of penetrants. In general, the continuous presence of these low concentrations of probe molecules either arrested or retarded the ensuing relaxation as compared with the aging that occurred in vacuum. A series of lower monohydric alcohols, lower n-alkanes, and some other similar penetrants were used as probes to test explicitly and systematically the effects of size and structure of the penetrant contacting the polymer during aging on the decay of excess sorption capacity of the preswollen glassy polymers. Decay of excess sorption in the presence of penetrant was evident only when molecularly small penetrants with interactive functional groups, alcohols for example, were sorbed into preswollen PMMA. This result suggested that the relaxation occurring in the presence of penetrant involved specific interactions between the penetrants and the carbonyl groups in the PMMA.     

Relaxation controlled (case II) transport of lower alcohols in poly(methyl methacrylate)

The remarkably consistent long term absorption data of Andrews, are shown to obey relaxation controlled (case II) transport kinetics for the absorption of methanol, ethanol, n-propanol, isopropanol, and n-butanol in glassy poly(methyl methacrylate). Andrews' data were obtained by monitoring gravimetric absorption, in 1 mm sheet specimens, under controlled conditions, for a period of years until absorption equilibria were established. Activation energies for the relaxation controlled (zero-order) absorption kinetics were calculated to be ～17 kcal/g mod and the enthalpy of absorption varied regularly and explicably with carbon number of the alcohol and with solubility parameter of the homologous penetrants. The dramatic monotonic increase of sorption rate with decreasing carbon number is analysed in terms of the increased concentration gradient which presumably develops as carbon number is decreased, rather than the more attractive theory that increased absorption rate is a direct function of decreased molar volume of the penetrant. Since the absorption kinetics are completely controlled by osmotic stress induced relaxations, molecular size of the penetrant does not directly affect the observed kinetics of penetration.     

Sequential vapor sorption of additional penetrants in preswollen polymers

Uptake and removal of a second penetrant in polymer samples preexposed to another vapor dictate many industrially important processes. Both the kinetics and equilibrium sorption can be strongly affected by the presence of the first permeant in the polymeric matrices. A novel experimental setup was constructed to study penetrant uptake in sequence. Sorption of the second vapor took place while the partial vapor pressure of the first vapor was maintained, so that diffusion into the preswollen polymer approximated transport in a pseudobinary system. Polybutylene was chosen in this work to illustrate the capability of this versatile experimental system. Both the rate of diffusion and equilibrium sorption of the second vapor were found to depend on the prevailing composition of the preswollen polymer for all penetrant pairs studied.