Polymer-solvent-precipitant systems selection

Figures 12.1.22 and 12.1.23 explain technical principles behind formation of efficient and selective membrane. Figure 12.1.22 shows a micrograph of hollow PEI fiber produced from N-methyl-2-pyrrolidone, NMP, which has thin surface layer and uniform pores and Figure 12.1.23 shows the same fiber obtained from a solution in dimethylformamide, DMF, which has a thick surface layer and less uniform pores. The effect depends on the interaction of polar and non-polar components. The compatibility of components was estimated based on their Hansen s solubility parameter difference. The compatibility increases as the solubility parameter difference decreases. Adjusting temperature is another method of control because the Hansen s solubility parameter decreases as the temperature increases. A procedure was developed to determine precipitation values by titration with non-solvent to a cloud point. Use of this procedure aids in selecting a suitable non-solvent for a given polymer/solvent system. Figure 12.1.24 shows the results from this method. Successfid in membrane production by either non-solvent inversion or thermally-induced phase separation requires careful analysis of the compatibilities between polymer and solvent, polymer and non-solvent, and solvent and non-solvent. Also the processing regime, which includes temperature control, removal of volatile components, uniformity of solvent replacement must be carefully controlled.

Solvent selection was also proven to affect the resulting structure in temperature-induced phase separation (TIPS) (Lloyd et al. 1990 Gu et al. 2006 Lu and Li 2009). Diluents such as phthalates promote the formation of irregular fuzzy structures (Lloyd et al. 1990). PVDF membranes prepared from dimethyl phthalate (DMP) as a solvent showed larger spherulite structures compared to that of membranes prepared from DBP, or mixtures of DMP with dioctyl sebacate and dioctyl adipate. It can be ascribed from these differences that the degree of polymer-solvents interactions influences the extent of PVDF crystallization (Lloyd et al. 1990 Gu et al. 2006). Using different types of solvents, the crystallization temperature of PVDF during TIPS can also be changed. A more complex mechanism is indeed expected for a combined solvent system. For instance, in the PVDF/DMAc/TEP system, TEP (weaker solvent) acts as a latent solvent in the immersion precipitation process (Liu et al. 2012). 

The polymer/additive system in combination with the proposed extraction technique determines the preferred solvent. In ASE the solvent must swell but not dissolve the polymer, whereas MAE requires a high dielectric solvent or solvent component. This makes solvent selection for MAE more problematical than for ASE . Therefore, MAE may be the preferred method for a plant laboratory analysing large numbers of similar samples (e.g. nonpolar or polar additives in polyolefins [210]). At variance to ASE , in MAE dissolution of the polymer will not block any transfer lines. Complete dissolution of the sample leads to rapid extractions, the polymer precipitating when the solvent cools. However, partial dissolution and softening of the polymer will result in agglomeration of particles and a reduction in extraction rate. 

For the analysis of organic additives in polymeric materials, in most cases, prior extraction will be necessary. Depending on the nature of the additive, many different approaches are employed. These include soxhlet extraction with organic solvent or aqueous media, total sample dissolution followed by selective precipitation of the polymer leaving the additive in solution, assisted extraction using pressurised systems, ultrasonic agitation and the use of supercritical fluids. In trace analysis, solid phase extraction (SPME) from solution or solvent partition may be required to increase the analyte concentration. 

In the bulk phase-separation approach, an organic solution of a polymer dissolved in a water-miscible solvent is injected into the tissue defect. After injection, the solvent diffuses away from the injection site, resulting in precipitation of the water-insoluble polymer. Selection of an appropriate solvent, which must be non-cytotoxic and not harmful to host tissue, is a key factor for success of the bulk phase-separation system. Two solvents that meet these criteria are N-methyl-2-pyrrolidone (NMP) and dimethyl sulfoxide (DMSO). In recent years, improved strategies for removal of the solvent and release of growth factors have been active areas of investigation. However, the requirement of a solvent to induce phase separation of the polymer limits the scale at which this approach can be applied in vivo. Even for relatively biocompatible solvents such as NMP and DMSO, injection of large volumes is anticipated to adversely affect host tissue, as well as the ability to eliminate the solvent from the body. 

At first glance, one might consider the effect of compressed CO2 on the phase behavior of multi-component polymer systems to be a simple combination of the known effects of liquid solvents and hydrostatic pressure. Solvent effects are primarily enthapic in nature and typically manifest in upper critical solution behavior. Common solvents mitigate unfavorable interactions between dissimilar segments and enhance miscibility. In blends, the addition of highly selective solvents, e.g. a non-solvent for one component, can lead to precipitation of the unfavored species at high dilution. In block copolymers, the effect of selective solvents is less clear, but studies to date reveal a collection of the solvent at the domain interface, selective dilation of one phase, and stabilization of the disordered phase via depression of the UODT. The systems we have studied each exhibit a lower critical transition. For these specific systems, previous work indicates the hydrostatic pressure suppresses free volume differences between the components and expands the region of miscibility. 

Although solvent-shift method enjoys simplicity and versatility in selection of the polymer, the outcome of this study is somewhat questionable as the organic solvent may interfere with precipitation kinetics and thermodynamics, and may not reflect amorphous solid system. In addition, supersaturation or precipitation inhibition potential of specific polymers is highly concentration dependent. Since local in vivo drug-polymer concentrations are difficult to predict, it is thus challenging to define suitable biorelevant polymer concentrations for in vitro experiments. 

---