Combined experimental and computer simulation study of the kinetics of solute release from a relaxing swellable polymer matrix. II. Release of an osmotically active solute

Abstract

ABSTRACT:The performance of a model monolithic controlled release device, consisting of a swellable polymeric matrix subject to structural relaxation (cellulose acetate), loaded with a simple osmotically active solute (NaCl) and activated by the ingress of solvent (water), was studied experimentally and by computer simulation. The former study involved detailed monitoring of the kinetics of both solute release and solvent absorption (followed in due course by desorption of osmotically imbibed excess solvent). The computer simulation study was based on extensive previous modeling work. Values of the relevant input model parameters were derived from independent experimental measurements of the sorption and diffusion properties of solvent (cf. Part I) and solute (reported in the present article). The resulting simulation was highly successful, considering that it proved possible to simulate closely and consistently the kinetics of both solute release (over practically the whole experimental range) and concurrent solvent absorption (including correct prediction of the magnitude of the osmotically induced excess swelling of the polymeric matrix). Simulation of the final desorption of osmotically imbibed water was facilitated by the realization that this process actually reflects the kinetics of a long-term deswelling relaxation of the polymer structure back to the state of normal hydration, the rate of which could be measured to a good approximation on the pure deswelling polymer. The results presented here are of obvious practical significance in relation to progress toward computer-assisted design of monolithic controlled release devices exhibiting relatively complex kinetic behavior. They should also prove useful by calling attention to an important caveat when desorption into water is used as a method for the straightforward determination of solute diffusivity in hydrated polymers, in cases of osmotically active solutes diffusing in nonhydrophilic polymers.

Performance of three-layer controlled release devices with uniform or non-uniform material properties: Experiment and computer simulation

“…This value is somewhat higher than those found in Ref. [20] for depleted matrices, previously equilibrated in 25% w/v NaCl (C o W,CA = 0.158-0.169 g/g, see Tables I and II of Ref. [20]), probably due to slightly different preparation conditions resulting in different degrees of polymer crystallinity [31].…”

“…This effect depends on solute load and on its osmotic strength relative to the sorptive capacity of the polymeric matrix for water. For relatively hydrophobic polymers, which absorb low or negligible amounts of water in the absence of the solute, the said effect will manifest itself as a significant amount of extra water uptake in the presence of solute, as for example in the case of NaCl-loaded matrices of silicone rubber [25], poly(ethylene-vinyl alcohol) [26], or cellulose acetate [20]. With increasing hydrophilicity of the polymer matrix, the intensity of the osmotic action of a particular solute will tend to decrease, as it will partly be masked by the high water uptake of the polymer matrix [17,19,27].…”

“…Both phenomena affect the NaCl permeability in the CA layers: the former tends to enhance the water uptake of the layer in the presence of NaCl, thus increasing both the diffusivity and the solubility of NaCl in the layer and the latter results in reduced NaCl solubility. Both effects have been studied previously [20,30] and the relative input parameters [defined in Eqs. (4) and (5), respectively] are derived from the detailed data of [20], taking into account the experimental degrees of hydration of CA, determined here.…”

“…Both effects have been studied previously [20,30] and the relative input parameters [defined in Eqs. (4) and (5), respectively] are derived from the detailed data of [20], taking into account the experimental degrees of hydration of CA, determined here. In particular:…”

“…The diffusion coefficient of solute D Ni in each layer i, is strongly dependent on C Wi , as follows [19,20]:…”