No abstract

Intrinsically ultramicroporous (<7 Å) polymers represent a new paradigm in materials development for membrane-based gas separation. In particular, they demonstrate that uniting intrachain "rigidity", the traditional design metric of highly permeable polymers of intrinsic microporosity (PIMs), with gas-sieving ultramicroporosity yields highperformance gas separation membranes. Highly ultramicroporous PIMs have redefined the state-of-the-art in large-scale air (e.g., O 2 /N 2 ) and hydrogen recovery (e.g., H 2 /N 2 , H 2 /CH 4 ) applications with unprecedented molecular sieving gas transport properties. Accordingly, presented herein are new 2015 permeability/selectivity "upper bounds" for large-scale commercial membrane-based air and hydrogen applications that accommodate the substantial performance enhancements of recent PIMs over preceding polymers. A subtle balance between intrachain rigidity and interchain spacing has been achieved in the amorphous microstructures of PIMs, fine-tuned using unique bridged-bicyclic building blocks (i.e., triptycene, ethanoanthracene and Troger's base) in both ladder and semiladder (e.g., polyimide) structures. P olymer membrane technology is a simple low-energy intensity alternative to traditional gas separation technologies such as cryogenic distillation and absorption. 1,2 It is well established in various applications, with nearly two-thirds of the market comprising air enrichment (e.g., for nitrogen blanketing or oxygen-enhanced combustion) and hydrogen recovery (e.g., from ammonia purge gas and petrochemical refinery reactor streams). 3,4 Membrane materials performance and, thus, viability are gauged by the polymer permeability and selectivity. In 1991, Robeson 5,6 established that these two intrinsic material properties obey a trade-off relationship for polymers, whereby more permeable materials tend to be less selective and vice versa, and more recently, updated the database in 2008. Accordingly, the state of the art for a given gas pair is traditionally identified by a linear "upper bound" fit to the top performing materials on a log−log plot of available permeability/selectivity combinations. Freeman provided the fundamental theoretical basis of these upper bound gas pair relationships. 7 Highly selective but low-permeability commercially available polymers, such as cellulose acetate, polysulfone, polyimide, and polycarbonate, continue to be industrially employed in air and hydrogen separations since the 1980s, and a key challenge driving research has been to develop new polymers that defy the "upper bound" trade-off relationships and unite high selectivities with high permeabilities.Polymers of intrinsic microporosity (PIMs) are a rapidly expanding class of solution-processable amorphous glassy polymers designed for membrane separations. Traditionally, PIMs integrate microporosity (<20 Å) by virtue of rigid and contorted macromolecular architectures that pack inefficiently in the solid state. 8−25 The earliest PIMs were principally based on such "rigid" design metrics a...

Long-term physical aging and plasticization, two mobility-based phenomena that are counterintuitive in the context of "rigid" polymers of intrinsic microporosity (PIMs), were evaluated using pure-and mixed-gas permeation data for representative ladder and semiladder PIMs. PIMs between 1 and 4 years old retained from 10-to 1000-fold higher H 2 and O 2 permeabilities than commercial membrane materials with similar or higher selectivities. A triptycene-based ladder polymer (TPIM-1) exhibited very large selectivity gains outweighing permeability losses after 780 days, resulting in unprecedented performance for O 2 /N 2 (P(O 2 ) = 61 Barrer, α(O 2 /N 2 ) = 8.6) and H 2 /N 2 (P(H 2 ) = 1105 Barrer, α(H 2 /N 2 ) = 156) separations. Interestingly, TPIM-1 aged more and faster than its more f lexible counterpart, PIM-1, which exhibited P(O 2 ) = 317 Barrer and α(O 2 /N 2 ) = 5.0 at 1380 days. Additionally, the more "rigid" TPIM-1 plasticized more significantly than PIM-1 (i.e., TPIM-1 endured ∼93% increases in mixed-gas CH 4 permeability over pure-gas values compared to ∼60% for PIM-1). A flexible 9,10-bridgehead (i.e., TPIM-2) mitigated the enhancements induced by physical aging but reduced plasticization. Importantly, intra-chain rigidity alone, without consideration of chain architecture and ultra-microporosity, is insufficient for designing aging-and plasticization-resistant gas separation membranes with high permeability and high selectivity

A newly designed diamine monomer, 3,3,3′,3′-tetramethyl-1,1′-spirobisindane-5,5′-diamino-6,6′-diol, was successfully used to synthesize two types of polyimides for membrane-based gas separation applications. The novel polymers integrate significant microporosity and polar hydroxyl groups, showing the combined features of polymers of intrinsic microporosity (PIMs) and functional polyimides (PIs). They possess high thermal stability, good solubility, and easy processability for membrane fabrication; the resulting membranes exhibit good permeability owing to the intrinsic microporosity introduced by the highly contorted PIM segments as well as high CO2/CH4 selectivity that arises from the hydroxyl groups. The membranes show CO2/CH4 selectivities of >20 when tested with a 1:1 CO2/CH4 mixture for feed pressures up to 50 bar. In addition, the incorporation of hydroxyl groups and microporosity in the polymers enhances their affinity to water, leading to remarkable water sorption capacities of up to 22 wt % at 35 °C and 95% relative humidity.

Highly ultramicroporous, solution-processable polyimides bearing 9,10-bridgehead-substituted triptycene demonstrated the highest BET surface area reported for polyimides (840 m 2 g −1 ) and several new highs in gas selectivity and permeability for hydrogen (1630−3980 barrers, H 2 /CH 4 ∼ 38) and air (230−630 barrers, O 2 /N 2 = 5.5−5.9) separations. Two new dianhydrides bearing 9,10-diethyl-and 9,10-dipropyltriptycenes indicate that the ultramicroporosity is optimized for fast polymeric sieving with the use of short, bulky isopropyl bridgeheads and methyl-substituted diamines (TrMPD, TMPD, and TMBZ) that increase intrachain rigidity. Mechanically, the triptycene-based analogue of a spirobisindanebased polyimide exhibited 50% increases in both tensile strength at break (94 MPa) and elastic modulus (2460 MPa) with corresponding 90% lower elongations at break (6%) likely due to the ability of highly entangled spiro-based chains to unwind. To guide future polyimide design, structure/property relationships are suggested between the geometry of the contortion center, the diamine and bridgehead substituent, and the mechanical, microstructural, and gas transport properties.