Abstract
Two aromatic polyamides─poly(3,3'-dihydroxybenzidine terephthalamide) (DHTA) and poly(3,3'-dihydroxybenzidine isophthalamide) (DHIA)─are compared for their ability to remove salts from water. DHTA is linear and rigid whereas DHIA is nonlinear and semirigid. DHTA and DHIA were selected as they allow us to investigate the effect of polymer backbone geometry on salt exclusion in a non-crosslinked thin film membrane, independently of the backbone chemistry. Because of their differences in solution viscosity, spin coating parameters for DHTA and DHIA solutions were optimized separately to produce thin film composites (TFCs) with reproducible membrane properties. The resulting DHTA TFCs displayed salt rejections of 87.8% (NaCl), 97.0% (MgSO4), and 80.3% (CaCl2). In comparison, DHIA TFCs demonstrated poor salt rejections of 21.0% (NaCl), 29.3% (MgSO4), and 15.4% (CaCl2). Cross-sectional SEM images of DHTA and DHIA films reveal that DHTA has a stratified (layered) morphology whereas DHIA exhibits a dense, featureless morphology. Both DHTA and DHIA TFCs exhibit similar surface morphology, contact angle, surface charge, and water uptake. PEG rejection experiments indicate that the average pore size of DHTA TFCs is ∼2 nm while DHIA TFCs have an average pore size of ∼3 nm. Our findings illustrate that using a rigid, linear aromatic polyamide gives an active layer with a stratified morphology, uniplanar orientation, smaller pores, and higher salt rejection, whereas the nonlinear aromatic polyamide analogue results in an isotropic active layer with larger pores and lower salt rejection.