Page 288 12 Mechanical properties of film coats Michael E.Aulton SUMMARY This chapter discusses the need for a film coat to possess the correct mechanical properties. One of the requirements of a film coat is that it should provide adequate protection to the dosage form. The capacity of the film coat to afford physical protection depends to a large extent on its mechanical characteristics. After considering those desirable properties, the chapter explains how to assess such properties. It also explains the need for a standardized approach to film preparation prior to testing. The main techniques that have been used successfully for the assessment of pharmaceutical film coat properties are indentation hardness and tensile testing. These techniques are described in detail and representative data for polymeric film coat formulations are presented. The source and consequences of internal stresses within a film coat are explained and the consequences with respect to film-coating defects are discussed. 12.1 INTRODUCTION 12.1.1 Desirable mechanical properties of polymeric film coats Tablets and pellets are film coated for many reasons. One of the requirements of a film coat is that it should provide adequate physical protection to the dosage form. The capacity of the film coat to afford this protection depends to a large extent on its mechanical characteristics. The coating must remain intact, be durable and be resistant to chipping and cracking during handling. Both the film itself and the composite system (i.e. film plus tablet or pellet substrate) should therefore possess suitable mechanical properties. Page 289 The mechanical characteristics of polymer film coats are an important parameter in dictating their performance in pharmaceutical dosage forms. A commercial film coat does not consist of polymer alone but contains many other ingredients. Additives are often included for a specific reason, either to assist processing or to improve performance. It should be appreciated that other materials added to a polymer system will almost invariably have an effect on the natural physical properties of that polymer. Often a material is added to a polymer specifically to improve its mechanical properties (plasticizers are a notable example), while on other occasions materials are added to the polymer to achieve one function, yet their addition often inadvertently changes its mechanical properties (here the classic example is the addition of insoluble pigments or opacifiers which tend to make the film much more brittle). It was mentioned above that in order to provide mechanical protection, film coats should have suitable mechanical properties. But how do we define suitable, and, having done so, how can it be quantified? It is advantageous to be able to quantify the mechanical properties of polymer films in order that performance predictions can be made at the development stage and that the effect of additives on these properties can be examined so that the formulator can limit any detrimental effects and enhance any beneficial changes. Banker (1966) considered that the mechanical strength and bonding ability of polymers arose from forces of cohesion within the material and adhesion between the material and its substrate. The magnitudes of these forces depend on the molecular size and structure of the polymer. Intra-molecular forces are generally very much weaker than inter-molecular forces, but polymers of sufficiently high molecular weight may give rise to large numbers of intra-molecular bonds resulting in high cohesive strength. The observed mechanical properties of polymers are a function of ‘free volume’ (see section 12.1.3) and thus will be modified by the presence of diluent molecules (e.g. plasticizer or residual solvent) and environmental temperature (Ferry, 1961). Depending on environmental temperature or composition, the mechanical properties of high molecular weight polymers may range between an almost perfect elastic state to an almost Newtonian viscous state. The observed properties will also be dependent on the test methodology, particularly strain rate. The deformation behaviour of high-molecular weight polymers has been categorized into five distinctly different regions by Tobolsky (1971): glassy, transition, rubbery, rubbery liquid and liquid. The five distinct regions of viscoelastic behaviour may be characterized by the type of stress-strain curve exhibited by the polymer at a particular temperature. While the change between regions is, in some respects, analogous to a phase change in true solids or liquids, it is not sharply defined and is gradual. 1. At low temperature, i.e. below the glass transition region, or at very high strain rates, a polymer behaves as an elastic glass. Tensile strength and elastic modulus are relatively high, but extensibility is low. 2. As the temperature is raised, the polymer enters the transition region. Tensile strength and elastic modulus are decreased, but extensibility is increased. Polymers in this region may show a ductile type of stress-strain curve, characterized by an elastic portion, a sudden fall-off of stress with increasing strain; then as strain is increased further the stress increases again. This process may be accompanied by the formation of a neck in the sample, and is called cold - drawing. Page 290 Whereas temperature and pressure are the independent variables for phase change, temperature and strain rate (or duration of strain) are responsible for viscoelastic transitions. True plasticization results in a lowering of the glass transition temperature (T g , see section 12.1.3) of the polymer-plasticizer blend. The influence of plasticization on the observed viscoelastic behaviour can therefore be interpreted in a manner analogous to the effect of increasing temperature. Later parts of this chapter will explain in detail how the desirable mechanical properties of a polymer can be defined and quantified, and how formulation variables can influence these properties. 12.1.2 Deformation of materials Fundamentally, most materials deform either elastically (i.e. they return to their original dimensions on removal of the deforming stress) or plastically (i.e. their deformation is permanent). When investigating the physical and mechanical properties of real solids or viscous liquids, it is frequently found that deviations occur from the classical theories of elasticity or viscous flow. Deviations involving stress (applied force divided by area of material over which force is applied), strain (deformation divided by original appropriate dimension of the material) and time are of two kinds. First, stress anomalies arise when the strain in a solid or the rate of strain in a liquid are not directly proportional to the applied stress but instead depend on the stress in a more complex manner. Secondly, time anomalies occur when the resultant stress depends not only on strain but also on the rate of strain: the observed deformation behaviour is both liquid-like and solid-like and is therefore termed viscoelasticity. Stress and time anomalies may coexist but in the absence of the former the behaviour is said to exhibit linear viscoelasticity. This implies that the ratio of stress to strain is a function of time alone and is independent of the magnitude of the applied stress. 3. As the temperature is increased further (or the strain rate is decreased) the polymer enters a region called the rubbery plateau. In this region long segments of the molecular chain are free to move, but they are constrained from slipping relative to each other by cross-links or entanglements. In the rubbery state, the polymer may be capable of undergoing considerable extension but, on removal of stress, it returns to its original state. 4. In the rubbery transition state, the polymer remains elastic and rubbery, but also has a finite component of plastic flow due to failure of cross - links or disentanglement. 5. Finally, at the highest temperatures or after long periods of straining, nearly all cross-links and entanglements are uncoupled and the polymer flows as a viscous liquid. Page 291 Linear viscoelastic behaviour applies, therefore, to cases where the elastic contribution is Hookean and the viscous contribution is Newtonian. To understand viscoelasticity it is necessary first to consider the extreme examples of deformation behaviour exhibited by an ideal elastic solid and an ideal viscous fluid. Elastic behaviour An ideal elastic solid is one which recovers its original strain after removal of an applied stress and thus obeys Hooke’s law. This states that the stress ( σ) is proportional to the linear strain (ε): σ=E.ε (12.1) where E is a constant of proportionality, effectively a measure of the stiffness or rigidity of the solid, known as the elastic modulus or Young’s modulus (thus E=σ/ε) or the rigidity modulus G. The implicit conditions of the above equation define elastic behaviour. First, the strain in response to an applied stress has an equilibrium value which is completely recoverable; secondly, deformation is ideally instantaneous, i.e. independent of time; and, thirdly, the relationship of strain to stress is linear. A simple metal-coiled spring exhibits typical Hookean behaviour. Fig. 12.1 shows stress versus time and strain versus time relationships. Fig. 12.1 The Hookean spring. Page 292 Viscous behaviour Application of a constant stress to a Newtonian viscous fluid results in a linear increase in strain with time until the stress is removed. The deformation is permanent and the original strain is not recovered. The linear strain-time relationship is characterized by the gradient of the stress versus strain-rate plot, i.e. η=σ/ε′ (12.2) where ε ′ is the rate of change of strain (the first differential of strain with respect to time, i.e. ε ′ =ε/t) and η is the Newtonian viscosity of the fluid. Newtonian viscous behaviour can be conveniently modelled by a piston and dashpot arrangement in which the dashpot cylinder is filled with a Newtonian fluid. Fig. 12.2 shows the deformation characteristics of such a material. Real materials Both stress anomalies and time anomalies result in deviations from the simplest case of Hookean elasticity giving rise to other modes of deformation. Norwick & Berry Fig. 12.2 The Newtonian dashpot. Page 293 (1972) have classified several types of mechanical behaviour according to the conditions obeyed by the stress-strain relationship (Table 12.1 ). Most pharmaceutical materials have combination properties which can only be described by a two- component system in which an ideal elastic phase is combined with an ideal viscous phase. In practice most materials do show, to some extent, both elastic and viscous characteristics (Davis, 1974; Lockett, 1972). Consequently, viscoelastic behaviour covers a wide range of mechanical properties from ideal elastic to ideal Newtonian behaviour. Mechanical models of linear viscoelasticity Mechanical models can also be used to represent the properties of a viscoelastic material. The simplest of these use a Hookean spring combined either in series or in parallel with a Newtonian dashpot. These are the Maxwell and Voigt models, respectively. Their properties have been reviewed extensively (see, for example, Castello & Goyan, 1964, and Barry, 1974) and will be be considered briefly here. The Maxwell model The Maxwell model is shown in Fig. 12.3. It consists of a Hookean spring in series with a Newtonian viscous dashpot. The strain response with time to an applied stress reflects both the viscous and elastic contributions to the resultant deformation. On application of the stress there is an instantaneous increase in strain associated with the deformation of the spring. This is followed by a time-dependent linear increase in strain due to the movement of the piston of the Newtonian dashpot. On removal of the stress the elastic strain alone is recovered. Under an applied external force the stress in the spring is equal to that in the dashpot. The total strain ( ε T ) in the Maxwell model is the sum of the strains in the spring ε S and in the dashpot ε D : ε T = ε S + ε D (12.3) Table 12.1 Types of mechanical behaviour classified according to the conditions obeyed by their stress-strain relationships (after Norwick & Berry, 1972) Condition Unique equilibrium relationship (complete recovery) Complete instant response Linearity Ideal elasticity Yes Yes Yes Non-linear elasticity Yes Yes No Instantaneous plasticity No Yes No Anelasticity Yes No Yes Linear viscoelasticity No No Yes Page 294 Fig. 12.3 The Maxwell model. By convention, viscoelastic deformations are studied by calculating compliance (J). Compliance is defined as the strain divided by the applied stress. Its use has the advantage of allowing the comparison of strain data obtained under different stress conditions, or allowing calculation of expected strain for a given applied stress. If the Maxwell model is maintained under conditions of constant strain, the initial stress in the Hookean spring will be reduced by a viscous deformation in the dashpot until the stress decays to zero. This phenomenon is termed stress relaxation. Measurement of stress relaxation in a material, therefore, provides a quantitative measurement of the ability of the material to undergo non-recoverable or plastic deformation. The Voigt model The Voigt model is shown in Fig. 12.4. It consists of a Hookean spring in parallel with a Newtonian dashpot. This provides a mechanical analogy for a material in which the response to an applied stress is not instantaneous but is retarded by viscous resistance. Removal of the stress results in a similarly retarded, but total, recovery of the strain. The Voigt model therefore exhibits the properties of creep and creep recovery. The change in strain with time is exponential, and the greater the apparent viscosity of the Newtonian dashpot the greater will be the retardation. In the Voigt model, on application of an external force, the strain at any time in the spring is equal to that in the dashpot, and the total stress ( σ T ) is the sum of the stresses in the spring ( σ S ) and in the dashpot (σ D ). Thus: Page 295 Fig. 12.4 The Voigt model. Total stress (σ T )=σ S +σ D (12.4) The Voigt model is capable of dissipating energy, a phenomenon known as internal friction. This parameter has the dimensions of viscosity and may be regarded as the apparent viscosity (η) of the Newtonian dashpot. G is the rigidity modulus of the Hookean spring. Unlike the Maxwell model, the Voigt model is incapable of stress relaxation. The quantity η/G is the retardation time (τ) for the unit; that is, the time required for strain to relax to 1/ e of its initial value on removal of stress. The retardation time is short and strain recovery is rapid where internal friction is small compared with the rigidity modulus. Thus at any time (t): σ T =G.ε(t)+ηε ′ (t) (12.5) In real materials there exist a number of molecular interactions resulting in more than one retardation time. The viscoelastic behaviour of such materials can be represented by the generalized Voigt model consisting of n Voigt units in series, where n is the number of discrete retardation times. For a viscoelastic solid exhibiting limited recoverable flow, the generalized Voigt model applies. If equation (12.5) is rearranged to include the retardation time (τ), then integration without limits for the ith element gives: (12.6) When time t=0, strain ε i =0 then k i =ln(σ i /J i ) and Page 296 (12.7) Thus, the strain in the ith element is (12.8) or, in terms of compliance, (12.9) The total strain ε(t) in the generalized Voigt model is the sum of the strains in the individual elements, and thus compliances in series are additive. For the case of n Voigt units in series: (12.10) Generalized linear viscoelastic model By combining a Maxwell model in series with one or more Voigt units, a generalized model for linear viscoelastic behaviour is obtained ( Fig. 12.5 ). Fig. 12.5 The generalized ‘ spring and dashpot ’ model for linear viscoelasticity. [...]... of practical coating conditions, and no information on drying air volumes or substrate temperatures was given Neither of the two model systems described above simulated the tumbling action of tablets within the coating pan A system for spraying film for testing within an actual coating pan was described by Porter (1980) who placed a vinyl-covered card inside a coating pan during an actual coating run... attempt to mimic a commercial film -coating process have been suggested Model systems The use of a model system which mimics conditions pertaining in commercial coating equipment has obvious advantages for research and development work Information on the coating process can be obtained without either the considerable capital cost, space or services required for commercial coating equipment or the need for... chains Differential scanning calorimetry (DSC) and thermomechanical analysis (TMA) are the most Page 298 commonly used methods to examine pharmaceutical film -coating systems Presented below is a very brief introduction to the application of thermal analysis in the study of pharmaceutically relevant polymers The reader is referred to the book in this series by Ford & Timmins (1989) for a comprehensive explanation... by the American Society for Testing and Materials (ASTM) In this text, many pieces of apparatus suitable for the testing of films are described, together with a brief description of their use In pharmaceutical technology, two tests have proved to be the most useful in the assessment of the mechanical properties of film coats: tensile testing and indentation hardness testing These two tests are discussed... substance (glass) to a rubber solid or vice versa on cooling Thus, at the Tg, a polymer undergoes a significant change in mechanical properties which may have implications in coating performance The Tg influences many physical properties of coating polymers including: elasticity, adhesion, viscosity, solvent release and permeability One theory of what happens at the glass transition temperature is the so-called... length/thickness/height of the sample Expansion coefficient measurements require a ‘zero’ load on the sample Thermomechanical properties of film -coating materials It is possible to determine the glass transition temperature (Tg) with some precision on a pure polymer sample, but very often coating polymers are mixtures of many ingredients and the addition of these other materials usually leads to a reduction in Tg and... scale spheronizer to apply the Gradient Matrix System films onto multiparticulate spheres (van Bommel et al., 1990) Residual solvent It could be assumed that once the coating solvent has been evaporated from a polymer film during the film -coating process it will have no residual effect on the mechanical properties of the resulting film Work on ethylcellulose films cast from different organic solvents... interactions Pickard (1979) found that a considerable loss of the plasticizer propylene glycol occurred both during the coating process and on storage This loss resulted in significant changes in film water vapour permeability, strength and elasticity Loss of propylene glycol during coating has also been reported by Skultety & Sims (1987) Plasticizers should not be volatile Thus water, while having a... 1977; Pickard, 1979) These papers illustrate the potential importance of the film coat application process in determining the properties of aqueous film-coated products Characterization of aqueous film -coating process variables and their effect on the properties of the resulting film coat has not been the subject of intensive study, although work has been carried out to try to isolate some of the more... volume occupied by a given number of molecules (VT) can be pictured as the sum of the ‘free volume’ (VF) (the voids) and the ‘occupied volume’ (VO) (the volume of the molecules themselves): VT=VF+VO (12 .12) It is assumed that as the temperature increases there is an increase in VF as thus VT will increase This will allow more movement of molecular groups and side chains As Tg is approached, VF increases . examine pharmaceutical film -coating systems. Presented below is a very brief introduction to the application of thermal analysis in the study of pharmaceutically. properties which may have implications in coating performance. The T g influences many physical properties of coating polymers including: elasticity,