The coagulation of cellulose from anisotropic solutions in the NH3/NH4SCN solvent system

Coagulation of cellulose has been studied in cellulose/ammonia/ammonium thiocyanate anisotropic solutions. The effect of coagulation variables such as coagulant, bath temperature, and cellulose concentration on the coagulation process is reported. The coagulation rate was measured by microscopic observation of the moving boundary associated with coagulation. Results indicate that the coagulation rate increases with increasing cellulose concentration and bath temperature. Methanol has the highest coagulation power among the coagulants employed. Mass transfer rate difference and equilibrium swelling were also measured. The results on the mass transfer rate differences show that the mass transfer rate of NH₃/NH₄SCN is greater than those of the respective coagulants under all coagulation conditions. The equilibrium swelling decreases with increasing bath temperature and cellulose concentration.     

Coagulation studies for cellulose in the ammonia/ammonium thiocyanate (NH3/NH4SCN) direct solvent system

An extensive study of the coagulation of cellulose from cellulose/ammonia/ammonium thiocyanate solutions is presented. The effect of major variables upon the coagulation process for cellulose solutions is reported. Microscopic observations of the moving boundary associated with the coagulation were performed on gelled cellulose solutions to determine the coagulation rate as a function of molecular volumes of coagulant, bath temperatures, bath compositions, and cellulose concentrations. The data were analyzed by means of a one-dimensional linear diffusion model based on Fick's law, thereby depicting the mechanism of the coagulation process, and obtaining the diffusion coefficients of mobile components involved in the coagulation.     

Control of the porosity of asymmetric TPX membranes

In the present work, the effect of adding nonsolvent in the casting solution on the porosity of asymmetric TPX (poly(4-methyl-1-pentene)) membranes was systematically investigated. A series of alcohols, with carbon number ranging from 2 to 14, was added in the casting solution (TPX/cyclohexane) to alter the porosity of two types of asymmetric TPX membranes, prepared by using ethanol and 1-propanol as the coagulation medium. It was found that the effect of nonsolvent on membrane porosity is different for the two types of membranes and the difference can be reasoned by considering the exchange rate between the polymer solvent and the coagulation medium during membrane formation. The results indicate that, for the membrane formation system with low exchange rate between coagulant and solvent, the membrane porosity is controlled by the coagulation value, defined as the volume of coagulant required to demix the casting solution. On the other hand, for the system with high exchange rate, the membrane porosity is not controlled by the coagulation value but by the penetration speed of the coagulant front moving through the casting solution.     

An improved model for mass transfer during the formation of polymeric membranes by the immersion-precipitation process

An analysis is presented, which describes the isothermal ternary diffusion process encountered in the formation of a cellulose acetate polymeric membrane by a direct immersion-precipitation of polymeric solutions in a nonsolvent bath. A material coordinate was employed to derive the mass transfer equations for the membrane solution and the convective mass transfer in the coagulation bath was taken into account by solving the hydrodynamic boundary layer equations. Diffusion coefficients were measured and used to deduce ternary phenomenological coefficients. The computed results are found to agree with the experimental precipitation time and membrane morphologies observed in scanning electron photomicrographs. © 1994 John Wiley & Sons, Inc.     

A new modeling of asymmetric membrane formation in rapid mass transfer system

Mass transfer process involved in the immersion precipitation of polyurethane/dimethylformamide (DMF)/water system was investigated. The set of diffusion equations describing the local composition of the membrane solution as a function of space coordinate and time were solved by numerical method, and the composition path in the phase diagram was obtained. Instead of boundary conditions based on the instantaneous equilibrium assumption between membrane solution and coagulation bath, new boundary conditions were set up by using mass transfer formalism at the interface which is especially valid in the condition that the mass transfer rate is extremely rapid. Phase separation phenomena during immersion precipitation were taken into account to continue the calculation after phase separation. The calculated results showed that the chance of phase separation via spinodal decomposition increases with the strength of nonsolvent, addition of nonsolvent to the dope solution, and the use of more hydrophobic polymer. The proposed model is the improvement of the previous works eliminating the equilibrium assumption at the interface and extending the calculation after phase separation.     

Hydrodynamically induced formation of cellulose fibers. III. Coagulation of droplets of cellulose solution in extensional flow

The formation of cellulose fibers by coagulation of drops of cellulose solution in extensional flow was studied theoretically and experimentally. In the theory, which applies for slow-motion conditions, homogeneity of the drop is assumed. The drop deforms according to a previously established deformation mechanism and becomes solid when a critical concentration of coagulant is reached at a certain position inside the drop. The theoretical predictions for the variation of fiber length with various parameters were tested experimentally in a four-roller mill for cellulose/dimethyl sulfoxide/paraformaldehyde with glycerol as coagulant. In agreement with the theoretical predictions, it was found that fiber length increases with shear rate and original droplet size, but decreases with the diffusion coefficient of the coagulant in the cellulose solution.     

Diffusion dynamics of 1-Butyl-3-methylimidazolium chloride from cellulose filament during coagulation process

The diffusion dynamics of 1-Butyl-3-methylimidazolium chloride ([BMIM]Cl) during coagulation process of cellulose filaments with H₂O as non-solvent were investigated in detail. The diffusion coefficients of [BMIM]Cl was calculated based on the Fick’s second law of diffusion according to the experimental data. Several factors which affect the coagulation process including polymer concentration, concentration and temperature of coagulation bath were discussed respectively. It is found that the diffusion rate of [BMIM]Cl decreased with the increasing polymer content in the spinning solutions and the initial concentration of [BMIM]Cl in the coagulation bath, while the diffusion coefficients increased largely with the coagulation temperature becoming higher. The diffusion coefficients of [BMIM]Cl is relatively lower, in contrast with the conventional solvent in the solution spinning process, which is coordinate with the result of polyacrylonitrile [BMIM]Cl system by Zhang et al. (Polym Eng Sci 48(1):184–190, 2008). Compared with the diffusion process of N-methylmorpholine-N-oxide (NMMO) from cellulose filament, the diffusion coefficients of [BMIM]Cl is lower, which suggested a stronger coagulation and washing conditions should be taken to produce regenerated cellulose fiber with [BMIM]Cl as solvent.     

Membrane formation by isothermal precipitation in polyamide-formic acid-water systems II. Precipitation dynamics

An analysis is presented, which describes the isothermal ternary diffusion process encountered in the formation of the aliphatic polyamide membranes such as Nylon-66 by direct immersion-precipitation of a polymeric solution in a nonsolvent bath. A material coordinate is employed to derive the mass transfer equations for the membrane solution. The convective mass transfer in the coagulation bath is taken into account by solving the hydrodynamic boundary layer equations. Diffusion coefficients were measured and used to deduce ternary phenomenological coefficients. The computed results are found to agree with measured precipitation times and with membrane morphologies observed by scanning electron photomicrographs. © 1995 John Wiley & Sons, Inc.     

Poly(vinylidene fluoride) membranes by phase inversion: the role the casting and coagulation conditions play in their morphology, crystalline structure and properties

Flat membranes with controlled morphology, pore dimensions, mechanical properties and crystal structure were prepared by wet and dry wet phase inversion from polyvinylidene fluoride (PVDF). The effects of several parameters such as precipitation temperature, composition of the polymer solution (concentration, type of solvent), exposure time before immersion in the coagulation bath, type of coagulant on the sequence and the extent of the two phase separation processes, i.e. liquid-liquid and liquid-solid demixing (crystallization), were studied.Using solvent/nonsolvent pairs with different mutual affinity (DMA/water, DMA/C1-C8 alcohols), different morphologies were obtained. High casting solution temperature plays important role to increase the rate of the liquid-liquid demixing on the crystallization, i.e. the type of crystallites formed (α-type) also by using a soft coagulation bath. Exposure time before immersion favours the first type of phase separation and therefore once again crystallites of α type were observed. At room temperature, using C1-C8 alcohols as nonsolvents, the presence of crystallites of α type can be related to molar volume of the coagulant.     

Hydrodynamically induced formation of cellulose fibers. I. Observations on the formation of cellulose fibers from stirred solutions

Like synthetic polymers, a natural polymer such as cellulose may crystallize in fibrous form from stirred solutions. In the present work, it is demonstrated that cellulose fibers can be formed by precipitation from dimethyl sulfoxide/paraformaldehyde solutions by two methods that involve different mechanisms of fiber formation, viz., (A) precipitation of cellulose by addition of nonsolvent to the stirred cellulose solution, and (B) precipitation of cellulose by coagulation of droplets of cellulose solution in a stirred precipitant. Both processes yield fibers with properties depending on the stirring speed and the coagulant strength. The molecular orientation and tensile strength of the fibers produced by method A was low, but increased with the stirring speed, while some fibers formed by method B reached extremely high orientation, depending on the thickness of the fibers. The two mechanisms of fiber formation are discussed on the basis of the experimental observations.     

Fiber formation via solution spinning of the cellulose/ammonia/ammonium thiocyanate system

Fiber formation via the cellulose/ammonia/ammonium thiocyanate system by wet spinning has been investigated. This report presents a characterization of the structure and tensile properties of fibers spun under various coagulation conditions. Microscopic observations showed that the molecular size of coagulant was the dominant factor governing the crosssectional shape of the fibers. Density, birefringence, and crystallinity data indicated that a higher cellulose concentration and lower coagulation temperature favored development of a fiber with a denser and more oriented structure. Under optimum conditions, a welldefined fibrillar structure was obtained. Fiber tensile property measurements suggested the existence of a linear relationship between the fiber breaking tenacity and the product of the square of the Hermans' orientation factor and the infrared crystallinity index.     

Synthesis and Characterization of Poly(ether-block-amide) Membranes

Poly(ether-block-amide) membranes were made via casting a solution on a nonsolvent (water) surface. In this research, effects of different parameters such as ratio of solvent mixture (n-butanol/isopropanol), temperature, composition of coagulation bath (water) and polymer concentration, on quality of the thin film membranes were studied. The mechanism of membrane formation involves solution spreading, solvent–nonsolvent exchange, and partial evaporation of the solvent steps. Solvent- nonsolvent exchange is the main step in membrane formation and determines membrane morphology. However, at higher temperature of polymeric solution greater portion of solvent evaporates. The results showed that type of demixing process (mutual affinity between solvent and nonsolvent) has important role in film formation. Also, addition of solvent to the nonsolvent bath is effective on membrane morphology. The film quality enhances with increasing isopropanol ratio in the solvent mixture. This behavior can be related to increasing of solution surface tension, reduction of interfacial tension between solution and nonsolvent and delayed solvent-nonsolvent demixing. Uniform films were made at a temperature rang of 60–80 °C and a polymer concentration of 4–7 wt%. Morphology of the membranes was investigated with scanning electron micrograph (SEM). Pervaporation of ethyl butyrate/water mixtures was studied using these membranes and high separation performance was achieved. For ethyl butyrate/water mixtures, It was observed that both permeation flux and separation factor increase with increasing ethyl butyrate content in the feed. Increasing temperature in limited range studied resulted in decreasing separation factor and increasing permeation flux.     

Investigation of the dynamics of poly(ether sulfone) membrane formation by immersion precipitation

Cast‐leaching experiments were carried out to investigate the dynamics of membrane formation by immersion precipitation, with an emphasis on the outflow of the solvent from casting solutions during the phase‐separation process. The casting solutions, consisting of poly(ether sulfone) as the polymer, N‐methyl‐2‐pyrolidone as the solvent, and water (H₂O), isopropyl alcohol, 1‐butanol, and diethylene glycol as nonsolvent additives (NSAs), were immersed in a coagulation bath. Two thermodynamically vastly different coagulants? H₂O, a strong coagulant, and ethylene glycol, a weak coagulant—were used to study the effect of the coagulant on the dynamics of membrane formation. The results showed that the outflow of the solvent during the initial stage of membrane formation was controlled by Fickian diffusion within the extremely wide range of conditions studied, that is, polymer concentrations of 10–38%, approaching ratios of 0–0.95, and thermodynamically vastly different NSAs and coagulants. The role of the initial state of the membrane‐forming solution, especially the conformational state of macromolecules in the membrane‐forming process, was examined. In contrast to those works on the behavior of small molecules, an attempt was made to qualitatively interpret membrane formation from the viewpoint of macromolecules. © 2005 Wiley Periodicals, Inc. J Polym Sci Part B: Polym Phys 43: 498–510, 2005     

Asymmetric membrane formation via immersion precipitation method. I. Kinetic effect

The kinetic effect of the phase inversion process on the membrane morphology is explored, with emphasis on the diffusion coefficient of the nonsolvent as a measure of the solvent/nonsolvent exchange rate. The diffusion coefficient is closely related to the nonsolvent tolerance of the polymer solution, which was estimated from a pseudo-ternary phase diagram of the following system: polymer: polysulfone; solvent system: a mixture of the solvent 1-methyl-2-pyrrolidinone and a solvent additive (formic acid, water or ethanol); and nonsolvent: ethanol. Regardless of the kind of solvent additive employed, when the diffusion coefficient of the nonsolvent is high for a given gelation medium, then the membrane consists of a smooth, defect-free surface and macrovoid-free cross section, and is highly permeable to oxygen. However, using a polymer solution with a low diffusion coefficient results in a membrane of a rather defective morphology. Therefore, it is concluded that the diffusion coefficient of the nonsolvent is a crucial parameter in controlling membrane morphology.     

Diffusion Mechanism of As‐spun Polyacrylonitrile Fiber in a Dimethyl Sulfoxide‐Water Coagulation Bath

In this paper, the diffusion mechanism of as‐spun PAN fiber was investigated in dimethyl sulfoxide‐water by determining the dynamic compositions of the fibers and the diffusion coefficients of solvent and nonsolvent during coagulation. The diffusion process could be divided into two stages. Results showed that the first stage of the diffusion process was the most important during the whole process, which was fundamental to further study on the formation mechanism. Also, compared with wet spinning, the dry‐jet wet spinning method had the advantage of mild coagulating at a high jet‐stretch. At high concentrations, the diffusion coefficients increased and the ratio of solvent diffusion coefficient to nonsolvent diffusion coefficient decreased; an increasing temperature resulted in the increase of both diffusion coefficients with a decrease in their ratios. To some extent, for the PAN‐DMSO‐water system, the more the ratios D_s*/D_n* tended to 1, the more the cross‐section shapes of as‐spun PAN fiber tended to be circular.     

Solid scintillation proximity membranes: I: Characterization of polysulfone-inorganic fluor morphologies precipitated from NMP solutions

Solid scintillation proximity membranes were prepared by coagulation of polysulfone polymer solutions containing cerium-activated yttrium silicate particles (CAYS) as a fluor. Membranes were formed by three different solidification processes: precipitation of the polymer-rich phase after liquid–liquid phase separation, vitrification via solvent evaporation, or rapid polymer collapse due to fast exchange between solvent in the solution and nonsolvent in the coagulation bath. The results indicate that when the coagulation process includes the liquid–liquid phase separation due to nucleation of the polymer-lean phase, the inorganic fluor particles are expressed out of the polymer-rich phase and separated from the polymer matrix. On the contrary, solidification by the vitrification of the polymer solution through solvent evaporation results in the inorganic fluors being surrounded by the polymer matrix, much like the dispersed state in the solution. In contrast, rapid collapse of the polymer also induces entrapment of the fluor in the polymer structure. However, fluor impregnation in the polymer matrix is distinguished from that in the vitrified membrane in that the impregnation is due to localized impingement of the fluor on the polymer structure rather than envelopment by polymer molecules.     

Effect of nonsolvents on properties of spinning solutions and polyethersulfone hollow fiber ultrafiltration membranes

The relationship among the presence of nonsolvent additives, the rheological behavior of spinning solutions and properties of hollow fiber membranes was studied. The additives tested were water, polyvinylpyrrolidone (PVP) and polyethylene glycol (PEG), and the base mixture was polyethersulfone/N-methyl-2-pyrrolidone (PES–NMP). In addition the effect of combining water and PVP or PEG was also studied. Membranes were prepared using a spinneret having two concentric orifices. The internal coagulant used as well as the nonsolvent from the coagulation bath were both water at 28°C and 30°C, respectively. Rheological properties of polymer solutions were evaluated using a rheometer Haake RV 20. Changes on composition of spin-solutions were also evaluated in terms of membrane water permeability, solute rejection and membrane structure observed using scanning electron microscopy (SEM). Experimental results from this work showed that spinning solutions containing any of the three additives behave as Newtonian fluids in the range of shearing rates tested. The addition of water, PVP or PEG to the base PES–NMP solution increased its viscosity and this effect was independent of the type of additive used. A direct relation between viscosity of casting solutions and membrane thickness was found. However, rheological properties (viscosity and normal stress difference) could not be used to explain differences on membrane water flux (MWF) when using different additives at the same concentration. The addition of any of the three additives generally increased MWF. The extent of this increment seemed to be more related to changes on membrane porosity than changes on pore sizes induced by the nature and concentration of the additive used.     

INFLUENCE OF ALCOHOL-BASED NONSOLVENTS ON THE FORMATION AND MORPHOLOGY OF PVDF MEMBRANES IN PHASE INVERSION PROCESS

Through the preparation of PVDF membranes using various nonsolvent coagulation baths following the phase inversion process, the influence of alcohol-based nonsolvents on the formation and structure of PVDF membranes were investigated. The light scattering and light transmission measurements were used to characterize the equilibrium phase diagram and the gelation speed, respectively. The locations of the crystallization-induced gelation boundaries for various systems and precipitation processes were explained from the corresponding thermodynamic and kinetic parameters. It was found that the better affinity between alcohol-based nonsolvents and DMAc solvent caused the gelation boundaries further away from the PVDF-DMAc axis with the coagulation bath varying from water, methanol, ethanol to iso-propanol. Due to the lower exchange rate of DMAc and alcohols, the delayed demixing took place for the membrane-forming using alcohols as baths, and the delayed time became longer when the coagulation bath was changed from methanol, ethanol to iso-propanol. The characterization results of membranes indicate that the influence of nonsolvents on the phase diagram and the precipitation process are in agreement with those on the membrane morphology. The better thermodynamic stability and a low exchange diffusion rate of PVDF/DMAc/alcohols favor the liquid-solid phase separation in gelation process, and therefore yield the membranes with a porous upper surface, a particular bottom surface and symmetrical structure.     

Fundamentals of polymer coagulation

Polymer coagulation is studied by a Monte Carlo diffusion model in which coagulant, solvent, and polymer particles move on nearest neighbor lattice sites according to the change in local interaction energy. Our approach has allowed, for the first time, to describe the coagulation process beyond the initial quench state and to reproduce the wide variety of different polymer structures that can be obtained, ranging from dust- to finger- to spongelike morphologies. We show that these morphologies are fully controlled by the coagulation rate which is itself strongly dependent on the degree of miscibility between solvent and coagulant. © 1995 John Wiley & Sons, Inc.     

Cellulose-based fibers from liquid–crystalline solutions. II. Processing and morphology of acetate/butyrate esters

Fibers were spun from isotropic and anisotropic dimethylacetamide solutions of cellulose esters. Take-up speeds of the dry jet/wet spinning process varied. Water served as the coagulant. The mechanical properties of the fibers increased as spinning progressed from the isotropic to the anisotropic state of the solution. A trade-off in solubility and fiber properties was noted as the butyryl acetyl ratio decreased. Whereas high butyryl content enhances both overall solubility and the formation of liquid–crystalline solutions at lower concentration, it results in lower fiber modulus and strength. Morphology of the fibers depended on the coagulation rate which was influenced by the concentration of the sppinning solution. The level of orientation and crystallinity of the fibers increased somewhat when they were spun from liquid-crystalline solutions. © 1993 John Wiley & Sons, Inc.