Membranes of Polymers of Intrinsic Microporosity (PIM-1) Modified by Poly(ethylene glycol)

Until now, the leading polymer of intrinsic microporosity PIM-1 has become quite famous for its high membrane permeability for many gases in gas separation, linked, however, to a rather moderate selectivity. The combination with the hydrophilic and low permeable poly(ethylene glycol) (PEG) and poly(ethylene oxides) (PEO) should on the one hand reduce permeability, while on the other hand enhance selectivity, especially for the polar gas CO2 by improving the hydrophilicity of the membranes. Four different paths to combine PIM-1 with PEG or poly(ethylene oxide) and poly(propylene oxide) (PPO) were studied: physically blending, quenching of polycondensation, synthesis of multiblock copolymers and synthesis of copolymers with PEO/PPO side chain. Blends and new, chemically linked polymers were successfully formed into free standing dense membranes and measured in single gas permeation of N2, O2, CO2 and CH4 by time lag method. As expected, permeability was lowered by any substantial addition of PEG/PEO/PPO regardless the manufacturing process and proportionally to the added amount. About 6 to 7 wt % of PEG/PEO/PPO added to PIM-1 halved permeability compared to PIM-1 membrane prepared under similar conditions. Consequently, selectivity from single gas measurements increased up to values of about 30 for CO2/N2 gas pair, a maximum of 18 for CO2/CH4 and 3.5 for O2/N2.


Synthesis within Paths 1, 2, 3, 4:
 Preparation of PIM-1 by polycondensation reaction (upper part of Scheme 1) was performed either by low temperature method following Budd and McKeown [2] and applying 1:2:K2CO3 in molecular equivalents 1:1:2.04 (in paths 2, 3) or by high temperature method following [18,21], applying identical amounts of the components at 150 °C and adding appropriate amounts of diethyl benzene to prevent a fast precipitation of the polymer. Reaction time is 0.5 h (PIM-1 in path 1).  Preparation of 2,4,5-trihydroxybenzoic acid(THBA)-PEG-mono-esters (3a-c) and 3,4dihydroxybenzoic acid(DHBA)-PEG-di-esters (5a-c) was performed by water extracting distillation of THBA resp. DHBA and PEG in boiling toluene in presence of catalytic amounts of p-toluene sulfonic acid. Reaction time 2 to 10 days, progress of reaction was followed by sampling, work up either by filtration/drying or by extraction, followed by evaporation of solvent under reduced pressure.
Measurements in the time lag apparatus primarily give the permeability for result, calculated by the increase of pressure on the permeate side per time after reaching a "quasi" steady state. Permeability coefficients P are calculated by eq. 1 (Vp is the permeate volume, l the actual membrane thickness, A the membrane area, pf the feed pressure (constant) and pp1 resp. pp2 permeate pressures at different times). Permeability is expressed in Barrer [1 × 10 −10 (cm³(STP) cm cm −2 s −1 cmHg −1 )].
Diffusivity coefficients D are calculated by Equation (2) using the so called time lag θ, the delay of pressure increase at start of every experiment, and the square of the membrane thickness l. Diffusivity is expressed in cm²/s.
The measured time lag θ of a PIM-1 membrane (thickness 80 µ m, 30 °C) varies from ca. 0.1 (He and H2), ca. 2 (O2), ca. 6 (CO2 and N2) to about 16 s for CH4. Although the time lag and permeability measurements are repeated several times, the experimental error for small gases like H2 and He within the highly permeable PIM-1 membrane is considerable high and the results are not really reliable.. Consequently, we concentrated on the results for the larger gas molecules under consideration being more reliable on account of their slower diffusion.
Following the solution-diffusion-model [15] permeability P is the product of diffusivity D and solubility S. Accordingly, the solubility coefficient S is calculated by eq. (3), dividing P by D.