Isidoro Caraballo


Among the advantages of the Controlled Release Systems we can mention the maintenance of drug levels within a desired range, the need for fewer administrations, optimal use of the drug in question, and increased patient compliance. Nevertheless, the potential disadantages cannot be ignored: the possible toxicity or nonbiocompatibility of the materials used, undesirable by-products of degradation, any surgery required to implant or remove the system, the chance of patient discomfort from the delivery device, and the higher cost of controlled-release systems compared to traditional pharmaceutical formulations. Since the patents of Gueret and Motta in 1993 and 1994, respectively, several experimental works have shown the advantages of the ultrasound-assisted compression to prepare pharmaceutical dosage forms. Nevertheless, this technique is still remaining at a experimental level (Levina et al., 2000). One of the reasons for this technique being not widely used in Pharmaceutical Technology can be the absence of a clear knowledge of the effect of the application of ultrasonic energy on the final properties of the dosage form. In this respect, the Percolation Theory is being applied to obtain a theoretical model able to predict the behaviour of the obtained systems. Recent studies have shown that in case of one component of the system undergoing thermoplastic deformation, the continuum percolation model can be used to predict the changes in the system with respect to a traditional pharmaceutical dosage form. The application of this model, together with a new parameter to quantify the Excipient Efficiency to control the drug release has been distinguished with one of the Prices of the Controlled Release Society in the 2001 Symposium in San Diego, California. In this work, the advantages of the application of the ultrasound-assisted compression for the preparation of controlled release systems have been demonstrated and quantified using a rational basis. In the following sections, the technique and the theoretical model derived from percolation theory are briefly described in the hope that this emerging technique can contribute to the reduction of the cost of preparation of the controlled release systems and consequently to their administration to an increasing number of patients.


The compression of a powder is a complex process that is usually affected by different kinds of problems. These problems have been widely investigated and mainly concern the volume reduction and the development of a strength between the particles of the powder sufficient to keep the tablet integrity (Leuenberger and Rohera, 1986). The application of ultrasonic energy is showing a great ability to reduce and even avoid these problems (Levina et al., 2000). Ultrasound is a term used to refer to mechanical waves with a frequency above 18 kHz (the approximate limit of the human hear). In an ultrasound compression machine, this vibration is obtained by means of a piezoelectric material (typically ceramics) that acts as transducer of alternate electric energy of different frequencies in mechanical energy. An acoustic coupler or “booster”, in contact with the transducer, increases the amplitude of the vibration before it is transmitted (usually in combination with mechanical pressure) to the material to be compressed. These elements are usually made of aluminium or titanium alloys, materials with high elastic module and high resistance to strength. Ultrasound assisted powder compression has been widely employed in metallurgy, as well as in plastics and ceramic industries (Kromp et al., 1985). Nevertheless, as it was previously mentioned, the first references in the pharmaceutical industry are two patents in 1993 (Gueret) and 1994 (Motta). Since them, only a few papers have presented experimental data in this field (Rodriguez et al., 1997; 1998; Fini et al., 1997; Millán et al. 1998a; Sancini et al., 1999; Caraballo et al. 2000). Interesting achievements are reported in these papers. For example, concerning the design of controlled release dosage forms, using a thermoplastic excipient (polymers of acrylic and metacrylic acid), an important decrease in the release rate has been found for tablets compressed with the assistance of ultrasonic energy, in comparison with traditional tablets. Although the effects of the ultrasonic energy on the material are not completely clarified, this slow release rate has been attributed to different phenomena:

Mechanical pressure exerted by the punches of the ultrasound-assisted tabletting machine. This is the main compression mechanism when low ultrasonic energies are employed (below 25 Joules in the mixtures studied by Rodriguez et al. (1997; 1998)) or when the materials used are not thermoplastic. In these cases the machine acts as a multiple impact mechanical press.

Thermal effects: Due to the poor conductivity for ultrasounds (low module of elasticity and high quantity of air trapped inside) that usually exhibit the materials included in pharmaceutical formulations, a fast decay of the ultrasonic energy to thermal energy is obtained. This process has been studied monitoring the temperature inside the compression chamber by means of a thermistor. In the studied mixtures (Rodriguez et al., 1997; 1998), a fast rise in temperature was obtained in tenths of a second, followed by a relatively fast decrease (see Figure 1). The peak temperature obtained for low ultrasonic energy (25 J) is below 80°C, whereas for high energies (125–150 J) is above 140°C. In mixtures of Ketoprofen with acrylic polymers (Sancin et al., 1999) the increase in temperature was slightly lower. In this respect it must be mentioned that a recent modification of the ultrasound-assisted tableting machine, that involves the suppression of Teflon isolators in contact with the powder, must result in a faster decrease in temperature inside the compression chamber. Thermal effects can cause the total or partial fusion of some components of the formulation. Nevertheless, in the assayed controlled release formulations, the components are usually below its melting points.

Plastic deformation, due to the combination of thermal and mechanical effects. The thermoplastic excipient was subjected to a temperature above its glass transition temperature (Tg) and to a high frequency mechanical pressure that can avoid the elastic recovery of the material.

Sintering. The combination of temperature, pressure and friction effects can result in the sintering of particles, so that the limits between them are no longer evident (Rodriguez et al., 1997; Millán et al., 1998a). Nevertheless, the ultrasound-assisted compression can still be considered as a novel technique in Pharmaceutical Technology (Levina et al., 2000). Two main reasons can be considered as responsible for the delay in the application of this technique: The lack of a clear physical knowledge of the effects of the ultrasonic energy on the material (in order to guarantee the absence of safety problems due to modifications of drug or excipient). In this respect, in some of the first papers dealing with ultrasound-assisted compression of pharmaceutical tablets, different analytical techniques as IR spectroscopy, Differential Scanning Calorimetry, Chromatography (HPLC, TLC), have been used and no important permanent modification of the drug has been found, with the exception of the lost of cristalinity (Fini et al., 1997; Sancin et al., 1999). The need of a theoretical model able to predict the biopharmaceutical and technological consequences of the application of this new compression method. In this sense, some experimental works (Millán, 1998a; Caraballo et al. 2000) are proposing the Percolation Theory as a first basis to interpret the technological and biopharmaceutical differences between the ultrasound compacted tablets and traditional tablets. These works deal with the difference between lattice percolation models and continuum percolation models to explain the influence of the ultrasound assisted compression. Furthermore, a new parameter proposed to quantify the efficiency of the excipient to control the drug release in both traditional and ultrasound tablets can help to develop such a model. In the next section, a brief summary of the main concepts of Percolation Theory and their application to explain the influence of the ultrasound-assisted compression is presented.


In 1957 Broadbent and Hammersley (Hammersley, 1957; Stauffer and Aharony 1992) presented a statistical theory -Percolation Theory- which was able to explain the behavior of disordered systems. Since these initial communications, Percolation Theory has been fruitfully applied to a great number of physical, chemical and biological problems as the flow of liquid in a porous medium, the polymer gelation, the glass transition or the confinement of quarks in nuclear matter (Zallen, 1983). Percolation Theory was introduced in the pharmaceutical field by Leuenberger and coworkers to explain the mechanical properties of compacts and the mechanisms of the formation of a tablet (Leuenberger et al., 1987; Holman and Leuenberger, 1988). After these initial applications, several investigations have demonstrated the usefulness of Percolation Theory in the pharmaceutical research (Leuenberger et al., 1989a; 1989b; Bonny and Leuenberger 1991;1993; Caraballo et al., 1993; 1996; 1999; 2000; Millán et al., 1998a; 1998b; Melgoza et al., 1998; 2000).

3.1. Concept

The Percolation Theory is a statistical theory, which deals with the formation of clusters and the existence of site and bond percolation phenomena. This theory supposes the existence of a regular lattice underlying the system. A cluster is defined as a group of neighbor occupied sites in the lattice, and the probability at which a cluster just percolates a system (a tablet in our case) is termed the percolation threshold. Figure 2 shows an example of a square lattice with an occupation probability of 34 %. The occupied sites are marked (X) in Figure 2a, and the clusters of size higher than 1 are shadowed in Figure 2b. In a binary mixture A/B, the sites -or cells- of this lattice can be occupied by the component A or by the component B. In random percolation models, the occupation of the sites is random, i.e. each site is occupied by the component A or B independent of the occupation status of its neighbors (Stauffer and Aharony, 1992).

3.2. Application of Percolation Theory to the characterization of Controlled Release Matrix Systems

In 1991, Bonny and Leuenberger explained the changes in dissolution kinetics of a matrix controlled release system over the whole range of drug loadings on the basis of Percolation Theory. For this purpose, the tablet was considered as a disordered system whose particles are distributed at random. Figure 3 shows a micrograph of an inert matrix tablet exhausted of drug, where the distribution of the pores due to initial porosity and to the dissolution of the drug can be observed. Furthermore, these authors derived a model for the estimation of the drug percolation thresholds from the diffusion behavior. The knowledge of the percolation thresholds or critical points of the system, allows a rational optimisation of the matrix formulation, avoiding the “trial and error method” usually employed in pharmaceutical industry. The ideal formulation of an inert matrix, following the Percolation Theory, must be above the drug percolation threshold (i.e., the drug plus the initial pores percolate the system). This fact guarantees the release of the total drug dose. On the other hand, the matrix must also contain an infinite cluster of excipient (i.e., the excipient must also be above its percolation threshold). This percolating cluster of excipient avoids the disintegration of the matrix during the release process and controls the drug release (Bonny and Leuenberger, 1991). This kind of systems, containing percolating clusters of both drug and excipient are called bicoherent systems. Furthermore, in order to decrease the variability in the biopharmaceutical behaviour of the matrices, due to little changes in the tablet composition, it is not convenient to formulate the matrices just at the percolation threshold. In this way, the knowledge of the percolation thresholds of drug and excipient suppose an important decrease in the cost of the optimisation process as well as in the time to market (Leuenberger, 1999). The percolation thresholds of different pharmaceutical powders have been already estimated, including drugs, as morphine hydrochloride (Melgoza et al, 1998), naltrexone hydrochloride (Caraballo et al., 1999), dextromethorphan hydrobromide (Melgoza et al, 2000) as well as matrix forming excipients, such as hydrogenated castor oil, ethilcellulose or acrylic polymers. On the other hand, the influence of different formulation factors and preparation methods on the percolation threshold has been studied in order to know if the estimated values can be valid for other formulations. For the moment, the percolation threshold behaves as a quite general parameter. Only two factors have demonstrated a clear influence on the percolation threshold: the relative particle size and the use of ultrasound-assisted compression, whenever a thermoplastic component is included in the formulation (Melgoza, 1999; Caraballo et al., 2000). With respect to the first factor, quantitative studies have shown a linear dependence between the relative drug/excipient particle size ratio and the drug percolation threshold (Caraballo et al., 1996; Millán et al., 1998). For the moment this relationship is holding for different kinds of drugs and excipients (Melgoza, 1999). The second factor, i.e., the effect of the application of ultrasonic energy during the compression process is discussed in the next section.

3.2.1. Study of the Ultrasound-Assisted tablets using percolation concepts

In a recent work (Caraballo et al., 2000), the influence of the compression method has been studied from the point of view of the Percolation Theory. For this purpose, matrix systems with different KCl content (as drug model) and an acrylic/metacrylic copolymer (as inert excipient) were prepared using an ultrasound-assisted press and a traditional eccentric machine. Excipient percolation threshold

Examination of the US tablets after the release assay indicated that the tablet integrity was lost only for tablets containing 90 % w/w of drug. Since tablet integrity indicates the existence of an excipient infinite cluster, pc Exc, able to develop mechanical resistance to the disintegration, the mechanical percolation threshold of the excipient was situated in the range of excipient volume concentration 13.4 % < pcExc < 20.2 %. As Figure 4 shows, the structure of the excipient inside the US tablets does not correspond to a particulate system but to an almost continuum medium. Therefore, there is no properly an excipient particle size inside these matrices. In order to have a scientific basis to explain the behaviour of the excipient in the US tablets, the continuum percolation model can be applied to these systems. The continuum percolation model dispenses with the existence of a regular lattice underlying the system, therefore the substance is not distributed into discrete lattice sites. This model deals with the volume ratio of each component and a continuum distribution function. The volume ratio is expressed as a space-occupation probability to describe the behaviour of the substance (Efros, 1994; Kuentz and Leuenberger, 1998). The percolation threshold of a substance in the continuum percolation model is situated at approximately 16 % v/v of occupation probability. So, the obtained result for the excipient percolation threshold (13.4 % v/v< pcExc < 20.2 % v/v) is in agreement with this continuum percolation model (pcExc = 16 %). Therefore, on the basis of Percolation Theory, the high decrease observed in the release rate can be attributed to a decrease in the excipient percolation threshold. As a consequence, the distance to the excipient percolation threshold (pExc - pcExc), becomes higher. Therefore, for a given concentration of excipient, a more extended infinite cluster of this hydrophobic substance will be found under US compaction. This fact can explain the lower release rate exhibited by these tablets with respect to those prepared using traditional techniques. According to the fundamental equation of Percolation Theory (Stauffer and Aharony, 1992), X = S (pExc – pcExc)q a critical property X , such as, the Higuchi’s constant, the effective diffusion coefficient, or the b-property, related to the transport through the system, decreases with pcExc (S being the scaling factor and q the critical exponent for this property). Drug percolation threshold The percolation thresholds for the drug in both US and traditional tablets, were estimated following the method of Leuenberger and Bonny modified by Caraballo et al. (1999). The value obtained for US tablets (58.61 - 61.01 % v/v) was higher than the calculated for the traditional ones (26.68 - 42.19 % v/v). This value for the drug threshold cannot be explained in terms of the continuum percolation model, due to the fact that the drug particles did not undergo thermoplastic deformation and therefore they did not form a continuum medium. Following a lattice percolation model, the high percolation threshold found for the US tablets could be interpreted as a consequence of the high relative particle size of the drug (Millán et al., 1998b). In this case, the thermoplastic deformation of the excipient and the formation of a continuum medium have to be interpreted as a reduction of the particle size for this component that tends to occupy a minor volume and can better fill the cavities among the KCl crystals, as it can be done by very small particles. Nevertheless, it must be emphasized again that there is not a true relative drug/excipient particle size in US tablets. Therefore the behavior of the drug dispersed in a quasi-continuum excipient is a new problem that has not been studied by any theoretical percolation model. New studies are needed in order to clarify the behavior of a substance (a drug in this case) distributed as discrete particles into a continuum phase (the excipient) that surrounds the drug almost completely.


The following conclusions can be derived from the mentioned studies on the application of the Ultrasound-Assisted Compression to the preparation of controlled release matrix tablets: Ultrasound-Assisted Compression resulted in a higher efficiency of thermoplastic excipients (about 200 % according to the Caraballo’s parameter to control the drug release process. This fact is especially interesting when a high drug dose has to be included in the dosage form (as frequently occurs in Controlled Release Systems). The matrices obtained using this compression method exhibited a lower variability in their release profiles. Application of the ultrasonic energy results in an increase in the temperature of the die during the compaction process. Consequences of this fact should be taken into account and cannot be neglected in the case of thermolabile drugs and/or excipients (Sancin, et al., 1999). Further research is needed in the area of Ultrasound-Assisted Compression of pharmaceutical powders, including a higher number of drugs and excipients. Nevertheless, on the light of the obtained results, the ultrasound-assisted compression can result in the development of a new kind of controlled release systems with a highly competitive cost/effectiveness ratio. In this way, a higher number of patients would benefit from the advantages of the controlled release, such as lower side-effects or higher effectiveness of the treatment.

Isidoro Caraballo
Dipartimento di Farmacia e Tecnologia Farmaceutica. Università di Siviglia. 41012 Siviglia, Spagna