Elsevier

Journal of Membrane Science

Volume 325, Issue 2, 1 December 2008, Pages 758-771
Journal of Membrane Science

Interfacial polycondensation—Modeling of kinetics and film properties

https://doi.org/10.1016/j.memsci.2008.09.002Get rights and content

Abstract

Interfacial polycondensation (IP) is an important technique used in the encapsulation of a variety of active ingredients and synthesis of thin film composite membranes. The present work seeks to advance our understanding of the mechanisms underlying the reaction, phase separation and film formation in this process, and hence, of how the film properties are influenced by preparation conditions. The model presented here incorporates all the essential physicochemical processes at a fundamental level through simple phenomenologies: ionic equilibria in the aqueous phase, resistances due to external mass transfer, diffusion through polymer film, interfacial reaction, thermodynamics of phase separation, and formation of a coherent film. The model has been tested against the data previously communicated [S.J. Wagh, Studies in interfacial polycondensation. Ph.D. Thesis. IIT Bombay, 2004; S.J. Wagh, S.S. Dhumal, A.K. Suresh, An experimental study of polyurea membrane formation by interfacial polycondensation, Journal of Membrane Science, submitted for publication] on polyurea microcapsules. The influence of the model parameters and preparation conditions, on the properties of the polymer and film and their development during reaction, have been studied. The study provides important insights into the process and should help in designing synthesis methodologies to suit the application.

Introduction

Interfacial polycondensation (IP) is a technique of wide applicability for encapsulation of active ingredients (say for controlled release or containment), enzyme immobilization [1], and synthesis of thin film composite membranes (say for applications such as RO) [2]. IP offers the possibility of rapid production of polymers, under normal conditions of temperature and pressure, in an almost ready-to-use form. The mechanistic aspects of the process are, however, not well understood because of the difficulties in following the fast kinetics and the need to account for the interplay of several equilibrium and rate processes in any comprehensive modeling effort. As a result, only empirical information exists, and even this, only for particular systems, on how synthesis conditions (such as solvent used, concentrations employed, interfacial area available) affect the polymer film properties. Clearly, properties such as film thickness, molecular weight and its distribution, and the degree of crystallinity have an important bearing on the functional attributes of the product, and a predictive capability on how synthesis conditions influence such properties would go a long way in designing processes to deliver desired product characteristics. The present work is an attempt in this direction.

IP reaction involves a step growth polymerization between two monomers, each dissolved in one of a pair of immiscible phases. The reaction occurs at, or in a thin region adjacent to, the interface of the two immiscible phases, and the polymer product, being insoluble in both the phases, accumulates as a film at the surface of contact between the phases. Morgan [3] has described the salient features of the IP technique in detail for the preparation of films, fibers and coatings. While the exact locale of the reaction is not established in all cases, the balance of evidence is in favor of the organic side of the interface [3], [4], [5], [6]. Mechanistically therefore, IP can be considered as a process of heterogeneous mass transfer with chemical reaction, further complicated by the simultaneous occurrence of polymer phase separation and film formation.

Table 1 summarizes the literature, on the modeling of the IP process. The table shows that in general, the different physicochemical rate and equilibrium processes which have been considered are some or all of the following: (i) ionic equilibria for the aqueous phase monomer, (ii) transport of the aqueous phase monomer and/or the organic phase monomer from bulk phases to the site of reaction, (iii) the reaction between the two monomers, and finally, (iv) the phase separation of the formed oligomeric species. As for modeling the film formation, three different approaches can be seen. In the first, the reaction is assumed to occur at the interface (initially between the two liquid phases, and later between the already-formed film and organic phase) and the entire polymer formed is assumed to form the film [10], [11], [14]. In the second, the reaction is assumed to occur in a reaction zone which lies on the organic side of the interface mentioned above, and the polymer formed is excluded from the reaction zone as it forms; the reaction zone gets pushed into the organic phase as the film grows [12], [13], [15], [17]. In the third, the reaction is assumed to occur in a steady reaction zone having a finite thickness, in which the polymer forms and accumulates (the viscosity in the zone increasing as a result), ultimately taking on a gel-like form [16]. In the first two approaches, the film thickness is explicitly calculated, increases with time and increases the diffusion resistance with time. In the third approach, a film thickness is calculated based on the polymer concentration in the reaction zone, and the diffusion of monomers is modeled as taking place through a gel-like structure.

Experimental evidence [3], [5], [6], [11], [18], [19], [20], [21] points to a strong influence of the conditions employed in the preparation on the nature and properties of the film that forms. However, most of the models [7], [8], [9], [10], [11], [14], [15], [16], [17] focus on the kinetics and the variation of film thickness with time, and do not attempt to predict quantitatively the polymer properties as a function of process parameters. Properties such as molecular weight, polydispersity and crystallinity affect important characteristics of the polymer [20], [21], [22] such as mechanical properties, viscosity, ease of processing, permeability. Indeed, in their work on encapsulation, Yadav et al. [22] observed an order-of-magnitude variation in the permeability of the capsule wall because of variations in crystallinity. Karode et al. [12], [13] were the first to consider the detailed kinetics in their models and hence predict the molecular weight distribution. Their models however, have not been adequately tested against experimental data. For one, they were developed for the unstirred nylon 6–10 system and may require modifications for other systems. Even for the nylon system, the model assumes, as does much of the earlier literature, that the solvent effect is explained solely by the partitioning of the aqueous phase monomer. Recent work in this laboratory [5], [6] has shown the effect to be much more complicated, with solvent properties such as polarity playing a significant role. Properties such as crystallinity, while experimentally shown to be dependent on preparation conditions [11], [19], have not so far been modeled.

As already remarked, the fast kinetics of membrane formation makes it difficult to monitor the reaction and the development of structural attributes during IP. Tracking the reaction via reactant consumption or polymer film thickness has been attempted. The latter is particularly difficult in microencapsulation studies since the film is extremely thin and the amount of polymer formed, miniscule. Yadav et al. [9] used an on-line pH probe to follow the consumption of the aqueous monomer, since in their system, the reaction does not produce any species that changes the pH of the system. Chai and Krantz [23] proposed the techniques of light reflectometry and pendant drop tensiometry to follow the development of film thickness and rigidity. While these techniques have potential, they would need considerable refinement before quantitative information on kinetics can be obtained from them. Interpretation of kinetic data is another area that requires care. Since the overall mechanism of reaction involves physical transport and chemical reaction, issues of transport limitations and controlling regimes should be considered in estimating the kinetic parameters (reaction rate constant and/or diffusivity). While such considerations have often not been employed, an exception is the work of Yadav et al. [11] in which explicit criteria are established and used for regime identification, albeit with a simplified kinetic model. Table 2 gives a summary of the literature on experimental methods and parameters estimated.

With the above background, the present work describes a comprehensive modeling framework of unstirred IP for microcapsule formation, by extending the work of Karode et al. [13]. In addition to kinetics, the model predicts the evolution of film thickness, mass crystallinity and MWD with time. Our recently reported experimental data on the synthesis of polyurea membranes in two solvents [5], [6] have been used to test the model and estimate model parameters.

Section snippets

Experimental

Experiments on the IP reaction producing polyurea microcapsules were reported in our earlier work (Wagh et al. [5], [6]). The studies employed hexamethylene-1,6-diamine (HMDA) as the aqueous monomer and hexamethylene-1,6-diisocyanate (HMDI) as the organic monomer, the reaction being conducted at the drop-continuous phase interface of a oil-in-water dispersion to produce microcapsules a few microns in diameter. Experiments were conducted over a range of monomer mole ratio (R), moles of limiting

Theory

The model developed in this work takes that of Karode et al. [13] as the starting point. The model has been extended to non-buffered systems with the incorporation of ionic equilibria in the aqueous phase, and extended to enable predictions of crystallinity. While the model is developed in general terms, with the available data in mind, its applicability to the polyurea system, and in the production of self-supported films, is the specific point of focus in the following description. In

Polymer solution thermodynamics

Before proceeding to a prediction of kinetics and film properties with the model, we generate the polymer phase diagrams for each oligomeric species using the Flory–Huggins theory [24]. The relevant thermodynamic parameters for the four solvents used by Wagh et al. [5], [6] are tabulated in Table 6, and typical phase envelopes, generated for the oligomer A10 in cyclohexane and p-xylene using these parameters, are shown in Fig. 3. It may be expected that the ‘better’ solvents (with lower values

Conclusions

A comprehensive modeling framework has been proposed for the interfacial polycondensation reaction used in the microencapsulation and in the manufacture of thin film composite membranes. The model incorporates the salient physicochemical processes involved in the diffusion of monomers, polymerization reactions, phase separation and formation of a coherent membrane. The model, with parameters fitted from kinetics and end-of-run crystallinity values, is able to predict the observed trends in the

References (32)

  • S.J. Wagh, Studies in interfacial polycondensation. Ph.D. thesis. IIT Bombay,...
  • S.J. Wagh, S.S. Dhumal and A.K. Suresh, An experimental study of polyurea membrane formation by interfacial...
  • R. Pearson et al.

    Interfacial polymerization of an isocyanate and a diol

    Journal of Polymer Science, Part A, Polymer Chemistry

    (1985)
  • M. Sirdesai et al.

    A model for microencapsulation in polyurea shell by interfacial polycondensation

    Canadian Journal of Chemical Engineering

    (1988)
  • S.K. Yadav et al.

    Microencapsulation in polyurea shell by interfacial polycondensation

    AIChE Journal

    (1990)
  • S.K. Yadav et al.

    Microencapsulation in polyurea shell—kinetics and film structure

    AIChE Journal

    (1996)
  • Cited by (59)

    View all citing articles on Scopus
    1

    Present address: Department of Chemical Engineering, Dr. Babasaheb Ambedkar Technological University, Lonere 402103, India.

    View full text