Ethanol PTMSP

Claes, S., Vandezande, R, Mullens, S. et al. 2010. High flux composite PTMSP-silica nanohybrid membranes for the pervaporation of ethanol/water mixtures, 351 ... 

Nagase et al. have studied the grafting of polyacetylene and its derivative onto poly(dimethylsiloxane) (PDMS) for membrane applications [118-121]. Poly(l-trimethyl-silyl-l-propyne) (PTMSP) is a polymer known to have excellent gas permeability but suffers from relatively low selectivity and a decrease in gas permeability over time. Graft copolymers of poly(l-phenyl-1-propyne) (PPP) onto PDMS were found to have improved gas permeability and selectivity, and the performance was related non-linearly to the PDMS content. A minimum oxygen permeability coefficient was found for a copolymer with 55 mol% PDMS [118]. The same membrane series was also foimd to be permselective for a range or organic liquids, including an ethanol/water mixture, and was used in pervaporation applications... 

Their applications and improved long-term stability for gas separation and pervaporation were further investigated. A maximum in the separation factor and rate was obtained for the PTMSP copolymer with 12 mol% PDMS. The membrane was able to convert a 7 wt% ethanol mixture to over 70 wt% in ethanol [120]. Good oxygen permeability and stability of over a month was also achieved for the PTMSP copolymers with over 60% PDMS [121]. In a separate study, PPP was foimd to have good permselectivity to water while PTMSP is alcohol permselective. These copolymers as well as a series of poly(phenyl acetylene) graft PTMSP copolymers were further studied for pervaporation applications [122]. 

Figure 9. Pure gas propane permeability and propane/methane selectivity for a series of selected organic liquids (O), rubbery siloxane-based polymers ( ), and glassy polymers ( ). The glassy polymers include PI, a polyimide (79), PC, polycarbonate (80), PS, polystyrene (81), and PTMSP (82), Data for the siloxane-based rubber polymers are from Stem et al (83), The solubility of propane and methane in selected organic liquids (hexane, heptane, octane, acetone, benzene, methanol, and ethanol) is from the compilation by Fogg and Gerrard (72). Diffusion coefficients of propane and methane in these liquids were estimated using the Tyn and Calus correlation (46 48),...

Five different penetrants were considered n-pentane (n-Cs), n-hexane (n-Ce), n-heptane (n-C ), ethanol (EtOH) and methanol (MeOH). For the cases of n-pentane, ethanol and methanol, solubility and difiiisivity isotherms at three different temperatures 285, 300 and 330 K were examined. Sorption experiments for n-hexane in PTMSP were performed both at 300 and 330 K, while only the highest temperature was used for n-heptane. 

In Figure 4 the solubility isotherms at three different temperatures are reported for ethanol in PTMSP. Unlike the case of n-alkanes, the solubility coefficient of ethanol in PTMSP is quite low in the range of low pressures it increases with penetrant content in the low concentration region and it decreases at higher penetrant fiigacity. The... 

Penetrant activity P/Pg Sorption isotherms for ethanol in PTMSP. 

The peculiar features shown by the sorption isotherms of ethanol in PTMSP are followed also by the solubility curves for methanol in PTMSP, which are presented in Figure 5. As in the case of ethanol, the methanol solubility coefBcient in the low pressure limit. So, is orders of magnitude lower than for alkanes, and the sorption isotherms are characterised by the previously mentioned S shape. 

The quantitative differences observed between the two families of penetrants are due both to enthalpy and to entropy contributions. Based on a qualitative consideration of the interaction energies between the polymer matrix and the aliphatic or the alcoholic groups of the different penetrants, the enthalpic term in Equation 1 is mostly responsible for the quantitative differences observed in the isotherms. This was also the conclusion drawn in a previous study (75) where the solubility of ethanol and n-pentane in PTMSP were qualitatively compared with each other. Based on the solubility isotherms at different temperatures for n-alkanes and alcohols, we can now directly evaluate the excess enthalpy of mbdng in Eq.(l). In a previous discusdon of the comparison of solubility data for ethanol and n-pentane in PTMSP (6), it was assumed that the difference in chemical potential for the two penetrants was essentially due to the enthalpic term. TMs hypothesis is reasonable since the interaction energy of penetrant molecules with PTMSP segments should be significantly different, relative to interaction energies of pure penetrant molecules, between polar and non polar penetrants. 

Diffusivity and Mobflity Isotherms. The temperature and penetrant concentration dependence of the diffusion coefficient, D, has been measured for binary mixtures of PTMSP with n-pentane, n-hexane, ethanol and methanol Indeed, for each sorption step in the sequences of isothermal sorption experiments, the kinetics of mass uptake have been analysed and estimates of the corresponding diffusivity have been obtained from both small and long time sorption data, as indicated in a previous study (6). The relative difference in the values of the diffusion coefficients obtained through the use of long time and short time sorption data is less than 5% in the great majority of cases and is less than 10% in the worst cases. Indeed, when one compares the diffusion coeffident results obtained from two different experiments performed under the same conditions, the diffusivity values are only determined to within approximately 10%. 

In Figure 7, diffusivity isotherms at 300 K are presented for n-pentane, n-hexane, ethanol and methanol. The measured values of D are orders of magnitude higher than the typical values of diffusion coeffidents in glassy polymers. The diffusivity of the two alkanes in PTMSP show similar trends as a function of the penetrant content. For both n-pentane and n-hexane, the diffusion coefficient exhibits a maximum variation of about 40% and a rather flat maximum value at intermediate concentrations. 

The diffusivity isotherms obtained for ethanol and methanol are similar to each other and quite different from those of the alkanes. Very large diffusion coefficients have been measured in the low penetrant concentration range. Then, the diffusion coefficients decrease to a minimum value at approximately 7% penetrant weight fraction. Afterwards, the diffusivity increases again and shows a local maximum value at o ll%. The latter trend is also observed for diffusivity coefficients measured at temperatures of 285 and 330 K. However, for the sake of conciseness, these data are not presented explicitly. The present data also confirm the diffusivity behaviour of ethanol in PTMSP reported previously (6). 

In Figure 8, the mobility isotherms at 300 K are reported for binary mixtures of PTMSP with n-pentane, n-hexane, ethanol and methanol. Despite some scatter in the data due to the uncertainty in the values of the diffusion coefficient and to some error propagation due to the use of Eq.(7), it is quite evident that all the isotherms show the same shape. In all cases the mobility coefficient decreases exponentially with increasing penetrant weight fraction. These results confirm the data obtained for... 

The solubility of several penetrants in PTMSP indicate relevant differences between polar and non-polar components. In particular, for the alcohols, unusual S shaped solubility isotherms are observed which cannot be explained by the dual mode model. Variations of solubility with temperature were also considered in this work, and the mixing enthalpy for n-pentane, n-hexane, ethanol and methanol in PTMSP was calculated. Significant negative mixing enthalpies were measured in all cases, and absolute values of the excess enthalpy of the sorption is much higher for the alkanes than for the alcohols. 

---