The plasma polymerization of organic compounds was used to prepare a composite reverse osmosis membrane which consists of an ultrathin semipermeable membrane formed by plasma polymerization of an organic compound or compounds and a porous substrate. Many nitrogen-containing compounds (aromatic amines, heteroaromatic compounds, aliphatic amines, and nitriles) were found to yield excellent reverse osmosis membranes by plasma polymerization directly onto porous substrates such as Millipore filters, porous polysulfone filters, and porous glass tubes. Factors involved in the preparation of reverse osmosis membranes by plasma polymerization were investigated and discussed. The plasma polymerized membranes have the following unique features: (1) very stable performance independent of salt concentration and applied pressure (practically no water flux decline was observed with many membranes): (2) salt rejection and water flux both increase with time in the initial stage of reverse osmosis (consequently, the performance of the membrane improves with time of operation); (3) very high salt rejection (over 99%) with high water flux (up to 38 gfd) can be obtained with 3.5% NaCl at 1500 psi (membranes perform equally well under conditions of sea water conversion and brakish water treatment).

The rates of polymer deposition from various olefinic monomers in an electrodeless glow discharge were studied. The previously found empirical relationship (with styrene in part I) between the rate of polymer deposition R , the monomer pressure p M , and gas pressure p x in a steady‐state flow system (i.e., R = a ( p M ) 2 [1 + b ( p x )], R being nearly independent of the discharge power) was also found with all monomers investigated. (The effect of gas was examined with nitrogen in this study.) However, it was found that the polymer deposition is controlled by the monomer flow rate and R o (in pure monomer flow) is proportional to the flow rate of monomer F w (based on the weight); i.e., R o = kF w , where k is a characteristic rate constant of the polymerization. Olefinic monomers can be generally classified into two major groups, i.e., type A monomers which predominantly polymerize, and type B monomers which decompose in a glow discharge. Type B monomers have smaller values of a and k compared to type A monomers. The values of a and k for type A monomers both increase with increasing molecular weight of the monomer. The values of k for all monomers investigated are within roughly an order of magnitude, indicating that the reactivity levels of monomers are very similar in a glow discharge polymerization.

The effects of gas plasma generated by electrodeless (inductive coupling) glow discharge on polymers were investigated as functions of gas pressure, discharge power, exposure time, and type of plasma gas. A remarkable similarity between the plasma susceptibilities of low molecular weight organic compounds and polymers was observed; i.e., polymers which have ether, carbonyl, ester, or carboxylic acid attached to a nonaromatic structure are very susceptible to plasma. The weight loss was proportional to the exposure time and exposed area. The discharge power and type of gas were found to have a great influence on both the rate of weight loss and the morphology of the exposed surface. The predominant effect of plasma on polymers was found to be degradation (manifested by weight loss). The crosslinking effect was found to be marginal with many polymers; however, significant crosslinking was observed with double bond-containing polymers. The crosslinking was examined by swelling the treated films. With copolymers of styrene–butadiene, 4-vinylpyridine–butadiene, methacrylic acid-butadiene, and acrylic acid–butadiene, the crosslinking was greatly dependent on the discharge power, the butadiene content of the copolymers, and the exposure time. Both degradation and crosslinking by gas plasma were generally limited to the exposed surface; however, the propagation of crosslinking in the direction of thickness was observed with copolymers of styrene–butadiene. The plasma of organic vapor also cause degradation of plasma-susceptible polymers, particularly at high wattage, although the deposition of polymer occurs simultaneously.

AbstractThe rates of polymer deposition from various organic compounds which do not contain an olefinic doublebond in an electrodeless glow discharge were studied. The polymerization rates of these unconventional monomers are by and large similar to those of olefinic monomers reported in the previous study (part II). The rate of polymer deposition R0 from pure monomer flow can be characterized, according to the analysis used in part II, by Ro = apM2 and R0 = kFw, where pM is the vapor pressure of the monomer, Fw is the weight basis flow rate of the monomer. Type A monomers which predominantly polymerize and type B monomers which decompose in a glow discharge were also found with these unconventional monomers. The effects of structural factors on the values of a amd k and on the classification of types A and B were examined. These structures and groups—aromatic, heteroaromatic, nitrogen‐containing (e.g., >NH,NH2,CN), Si‐containing, and olefinic doublebond—favor the polymerization. These structures and groups—oxygen‐containing chlorine, aliphatic hydrocarbon chains, and cyclic hydrocarbon chains—favor the decomposition of the monomer in a glow discharge. It is postulated that the polymerization of organic compounds proceed by the recombination of excited species (probably free radicals) created by glow discharge and reexcitation followed by further recombinations in the vapor phase and at the interface.

The transfer rates of oxygen via polymer membranes in gas–membrane–gas and gas–membrane–water (dissolved oxygen) were investigated with various porous membranes and compared with results of silicone rubber sheet (nonporous, homogeneous polymer membrane). With a nonporous membrane, the permeability constant obtained by gas–membrane–gas represents the true membrane permeability in gas–membrane–water system, and consequently the transport resistance due to boundary layer can be quantitatively estimated. With a porous membrane, the data in gas–membrane–gas system (under applied pressure) merely represent the gas effusion rate of the membrane and are not directly related to the dissolved oxygen transfer rate in gas–membrane–water system. The penetration of liquid water into the pores of porous membrane is the most important controlling factor for the dissolved oxygen transfer rate of a porous membrane. With a porous membrane in which liquid water does not penetrate into the pore, the overall transfer rate of dissolved oxygen reaches the level which corresponds to that of the boundary layer found with silicone rubber membrane.

The transport of water through a hydrophilic polymer and the molecular mobility of water in such polymers are considered in this study using systems that are as nearly homogeneous as can experimentally be achieved; these are hydrogels, which offer the added advantage that they can be prepared with widely varying water contents. Transport properties and molecular mobility can therefore be studied in dependence on the degree of hydration, that is the volume fraction of water in these gels. The homogeneity of the gels circumvents some of the difficulties in interpreting experimental results which are often encountered with polymeric systems having a superstructure.

The concept of a homogeneous polymer membrane, in which randomly coiled macromolecules pack the solid phase uniformly leaving some unoccupied space as free volume, has been applied in the study of permeability and selectivity of solutes through water (solvent)-swollen polymer membranes. Using this treatment, it became clear that fundamental principles of permeation through polymer film, i.e., a solution and diffusion mechanism of permeation, is also applicable to these swollen systems. However, the permeability characteristics of highly swollen systems resemble that of porous media due to the presence of the solvent, and the permeability of a swollen system to a solute cannot be characterized by a single permeability coefficient. The permeabilities based on diffusive flux (molecular flow) and on bulk flow of solvent, both of which occur under certain conditions, should be differentiated. Consequently permeability and permselectivity depend on the type of processes or conditions under which permeability is measured. Conformational entropies of macromolecules seem to play a predominantly important role in permselectivity and also in semipermeability (permselectivity between solute and solvent) of water-swollen membranes. Semipermeability depends on the combination of solute and solvent as well as on membrane characteristics such as swelling ratio. Membrane charge was found to have marked influence on permselectivities of ionized solutes. In general, the permselectivity decreases with increasing swelling ratio and consequent increasing permeability or flux of the solutes.

AbstractThe polymerization (polymer deposition) rate of styrene in an electrodeless glow discharge from styrene vapor and a mixture of styrene vapor and gas (H2, He, A, and N2) was investigated. The rate of polymerization, R, was found to be independent of the discharge power. The rate of polymerization of the pure monomer was found to be proportional to the square of monomer pressure pM. The addition of gas increased the rate of polymerization depending upon the partial pressure of the gas, px, and R can be generally expressed by R = a[pM]2{1 + b[px]}. The value of b is dependent of the type of gas and follows the order of N2, > A > He > H2. The distribution of polymer deposition was found to be nearly independent of the partial pressure of the gas and of the discharge power with N2 and H2 as plasma gas; however, with He and A, the distribution is highly dependent on the partial pressure of the gas and on the discharge power. The study strongly suggests that polymerization occurs in the vapor phase and that the growing polymer radicals deposit on the surface of the discharge vessel, yielding highly crosslinked polymer deposition.

AbstractThe diffusive permeability of water P, which relates to diffusive flux of water under a concentration gradient of water (measured by diffusion of tritiated water), and the hydraulic permeability of water K, which relates to the water flux under a hydraulic pressure gradient are defined. For the case of diffusive transport one has P = KRT/ν1, where ν1 is the molar volume of water. The relationship between P and K was investigated as a function of hydration H, i.e., the volume fraction of water in swollen polymer membranes. The following characteristic features of water permeability are revealed. (a) In the lowhydration region (H < 0.2), water permeates by diffusion even under an applied hydraulic pressure gradient and KRT/ν1 = P. (b) In the higher hydration region KRT/ν1 is greater than P, and the ratio ω = KRT/ν1P increases nearly exponentially with decrease of (1‐H)/H. Water in this region moves partly by bulk flow under an applied hydraulic pressure gradient but moves only by diffusion in the absence of a pressure gradient. (c) The dependence of log P on (1‐H)/H is nearly linear in regions of both high and low hydration but the slopes are different. The transition occurs in about the same H range where the discrepancy between P and KRT/ν1 becomes significant. Excellent agreement was found between the experimental data for P as a function of H and the theoretical prediction based on the free‐volume concept of diffusive transport in hydrated homogeneous membranes.

AbstractThe water permeability K1 [which is related to water flux J1 per unit membrane area by J1 = K1(Δp − ΔII)/ΔX, where Δp is the pressure difference, ΔII is the osmotic pressure of feed solution, and ΔX is the membrane thickness] of homogeneous ionic polymer membranes in reverse osmosis and their salt rejection Rs [which is given by Rs ≡ 1 − (C2″/C2′), where C2′ is the concentration of the salt in feed solution, and C2″ is the concentration of salt in effluent] were examined with cationic and anionic membranes of block and graft copolymers. For ionic membranes, Rs and K1 are related by K1 = A exp { − BRs}, where A and B are constants. This equation was found to be independent of the ion charge, the chemical nature of the polymer, and film morphology. The principle of salt rejection by ionic membranes was explained by the difference in the transport volumes (volume elements available for transport) for mobile co‐ions and water. The electric repulsive force between a fixed ion and a mobile co‐ion decreases the transport volume of the latter, thus creating a transport depletion of salt flux relative to water transport. This transport depletion is governed by the amount of water sorbed by a fixed ionic site, which also determines the water flux. Consequently, Rs and K1 for ionically charged membranes are related as described above. This relation significantly differs from that found between Rs and K1 for nonionic polymer membranes, where the size and the solubility of ions in the membrane are mainly responsible for the transport depletion. The decline of Rs with increasing K1 is much less in ionic membranes than in nonionic ones; however, in the high Rs region, K1 for both ionic and nonionic membranes become similar as the dominant mode of water transport changes from flow to diffusion.

AbstractMembranes of graft copolymers of polyethylene with poly(sodium styrene sulfonate), poly(4‐vinylpyridinium methyl bromide), and poly(sodium acrylate) were prepared by using the technique of peroxide grafting. The reverse osmosis characteristics of the membranes were examined as a function of grafting yield. In these membranes, the grafting can be considered as a process of introducing ionic sites, and it depends on the conditions of the grafting reaction, such as monomer concentration and temperature. However, the overall reverse osmosis characteristic is not only dependent on the number of ionic sites introduced but also on the swelling capability of the membrane. Consequently, the salt rejection of grafted membrane of a fixed graft yield depends on the conditions of the grafting reaction. All grafted membranes which have grafting yields above a certain value behave as normal ionic polymer membranes, and their interrelationship of salt rejection and water permeability follow the general dependence found for ionic polymer membranes.

AbstractThe difficulties involved in the measurement of the partition coefficient and diffusion coefficient of NaCl in hydrophilic polymer membranes are examined by the initial rate and late‐time analysis of desorption data for three nonionic polymer systems. The late‐time analysis, based on the plot of the logarithmic relative desorption versus time, was found to be useful when the diffusion is Fickian and the magnitude of D2/l2 is small. Depending on the magnitude of D2/l2, an accurate measurement of k2 and determination of D2 cannot be achieved without correction for the effect of surface salt solution, impossible with late‐time analysis. The initial rate measurement was found more versatile for general non‐Fickian diffusion and it gave a means of correction of the errors involved in estimating the partition coefficient. In general, the diffusion of NaCl in hydrophilic polymers is not ideal Fickian and the discrepancy between the two analyses is appreciable.

AbstractThe diffusive permeabilities, P, of sodium chloride, urea, creatinine, secobarbital and bis‐hydroxyethyl terephthalate through water swollen polymer membranes were studied. The linear dependence of log P on the reciprocal of the volume fraction of water in the water swollen membrane, 1/H, (reported in Part I) was also found for those solutes investigated. The negative slope of the plot was found to increase with increasing size of solute. Ionic charge of the membrane phase was found to decrease the permeability of ionized solutes from values for neutral membranes and changes the permeability ratios of solutes. The permeability of highly ionized solutes, such as sodium chloride, are reduced by either positive or negative charge on the membrane phase; however, the permeability of weakly ionized organic solutes is affected mainly by membrane charge identical to the organic ions, and the effect of counter ions was found to be small.

AbstractBased on an analysis by the free volume theory of diffusion, a study of the diffusion coefficients, D2, the distribution coefficients, κ2, and the permeability coefficients, P, of NaCl in various hydrophilic polymers was made with the assumptions that the free volume of a polymer/diluent system is linearly proportional to the volume fraction of diluent (H), and that NaCl permeates only through polymer/diluent systems in which the diluent of polymers is also a good solvent of the solute (NaCl).The values of D2 (over five orders of magnitude) were found to vary exponentially with changes in 1/H, as expected from the free volume theory of diffusion, in a wide range of H‐values and converge to the value of the diffusion constant of NaCl in pure water.The values of κ2 defined by g NaCl per cm3 polymer/g NaCl per cm3 solution were found to be more characteristic to the type of polymers particularly at a lower range of H‐values. At higher ranges of H‐values, k2 is nearly equal to H, however, at lower H, k2 values deviate significantly from the value of H and its extent was found to be greatly dependent on the chemical nature of polymers, whereas the diffusion coefficient follows the normal pattern predicted by the free volume theory.The values of P which are given by the product of D2 and k2 also were found to vary exponentially with changes in 1/H at higher range of H, however, at lower range of H it deviates from the relationship due to the varying dependence of factor k2 on 1/H.