George Shlieout is the technical managing director at Magnesia GmbH, Postfach 2168, D-21311 Lüneburg, Germany, tel. +49 4131 871 018, fax +49 4131 871 055.

George Shlieout

In summary, the investigation was limited to six fractions of ethyl cellulose 7cP (EC 7cP) with the following particle size ranges: ,50 mm, 90-125 mm, 160-250 mm, 250-350 mm, 350-400 mm and 400-500 mm. The fractions were compressed without any additional excipients. Using three criteria - compactibility, compressibility and the energy distribution of the force displacement curve - the mechanical behaviour of the different fractions was assessed. It was found that raising the compression force leads to an increase in crushing strength and the bulk density of the produced tablets. And, simultaneously, the ratio of plasticity to elasticity decreased. An increase in particle size showed an increase in the crushing strength of the tablets. This result did not correlate with classical theory. We propose that, during the initial loading into the press, the large ethyl cellulose (EC) particles behave mainly as brittle units, resulting in a greater number of interlocking binding points. As such, their combined strength exceeds the primary binding forces. Increasing the EC particle size also increased the effective energy of the force displacement curves. 	

A linear relationship was found between effective energy and crushing strength of the fractions. The elastic deformation energy was not affected by varying the particle size. The drug release rate of theophylline monohydrate when using EC as a matrix former was found to be controlled by varying the particle size. The drug release rate increased by increasing the particle size of EC. 	

A classic rule of tabletting states that particle size and crushing strength have an indirectly proportional relationship. Huttenrauch explains this relation by using the activation theory.1 To establish a connection between the classic and activation theories, smaller particles should be more activated than the coarse ones during compression. This assumption was confirmed by the compression behaviour of various particle sizes of lactose. Alderborn and Nystrom observed that the crushing strength decreased when the particle size of lactose was increased.2 Furthermore,  they observed an inverse effect when compressing sodium chloride; the crushing strength increased with increased particle size. This proportional relationship between particle size and crushing strength was observed in further studies. Landin et al. 	

Figure 1:  Relationship between mean particle size (d) and crushing strength of EC compressed to different compression forces (CF).				

investigated the compactibility of two microcrystalline celluloses (Avicel PH 101 and Avicel PH 102).3 They observed that the crushing strength of tablets made of Avicel 102 (100 mm) was greater than the crushing strength of tablets made of Avicel 101 (50 mm). Materials with interlocking deformation behaviour also correlate inversely with the classic theory. 	

The relationship between particle size and crushing strength was investigated by Shotten and Ganderton,4 who observed the opposite influence of particle size when using lubricants. That is, smaller particle sizes gave higher crushing strengths. York and Riempa et al. postulated that this effect could be observed by compressing materials with brittle deformation behaviour.5,6 The brittleness increases by increasing the particle size. Riempa et al. found a linear relationship between brittleness and particle size.7 Hwang et al. concluded that various types of lactose and di-calcium phosphate possessed different properties that affected compression and tabletting characteristics.8 The use of lactose with a large particle size (168 mm) resulted in softer tablets than fast-flow lactose (103 mm). Olsson and Nystrom postulated that tablet porosity and particle size are the structural features that best correlate with tablet strength.9 In further studies, the influence of EC and hydroxypropylmethylcellulose (HPMC) particle size on compressibility and compactibility were reviewed.10 They suggested that a decrease in particle size generally causes an increase in the tensile strength of HPMC and EC matrices.		

Table I:  Linear regression data of the crushing strength and square root of the mean particle size using different compression forces. 	

In previous studies, the influence of the molecular weight of EC on the 'tablettability' and drug release rate has been investigated.11,12 An increase in the EC molecular weight - that is, the viscosity of a 5% solution - leads to a reduction of the compressibility and compactibility. Dissolution studies indicated that the drug release rate of EC with high molecular weight was higher than that of low molecular weight.	

Materials: Ethyl cellulose 7cP (Hercules-Aqualon, Darmstadt, Germany, Batch No. 10058) and theophylline monohydrate Ph. Eur. (Carl Rot, Karlsruhe, Germany, Batch No. 1546638) were used as supplied.				

Methods: Ethyl cellulose 7cP was fractionated by using a Retsch sieve tower (Retsch GmbH & Co. KG, Haan, Germany) with a vibrating strength of 70%. Six fractions were investigated (,50, 90-125,160-250, 250-350, 350-400 and 400-500 mm). The bulk density was measured using the Ph. Eur. method.13	

True density. The true density of the materials used was determined (n55) using an Accupyc 1330 Pycnometer (Micromeritics, Dunstable, UK).		

Figure 2:  Relationship between effective energy (E2) and mean particle size using different compression forces (CF) and Figure 3:  The elastic deformation energy (E3) of different EC fractions using different compression forces.	

Scanning electron microscopy (SEM). The powder and tablets were evaporated with carbon and then sputtered with gold to make the samples electrically connected. The sputter conditions were as follows: the apparatus used was an SCD 050 (Balzers Union AG, Balzers, Liechtenstein) with a vacuum of 5310-2 mbar; carbon was layered to a thickness of approximately 10 nm; gold was layered to approximately 25 nm; and the SEM was an S-2400 (Hitachi Ltd, Tokyo, Japan).	

Tabletting. An instrumented single punch machine was used (EK0/DMS No. 1.0083.92, Korsch GmbH, Berlin, Germany). The upper punch and lower punch forces were measured. By using an inductive displacement transducer, the upper punch displacement was determined. The machine was connected to a DMS-Plus amplifier (HBM Company, Darmstadt, Germany) with a frequency carrier of 4.8 kHz. The materials were compressed with flat circular punches (9 mm diameter) at a rate of 10 tablets/min. The displacement of the upper punch was measured using an inductive transducer (W 20 TK, Hottenger Baldwin Meßtechnik, Darmstadt, Germany). The elastic deformation of the punches and tablet machine during the compression cycle was measured and evaluated. Details regarding the evaluation have previously been published.12 The calibration of the transducer was done by using steel slip gauges of 2, 3, 4 and 5 mm height. The tablet machine was connected to a DMC-Plus amplifier (Hottenger Baldwin Meßechnik) Ten tablets were produced at each condition.	

Crushing strength. The crushing strength of 10 tablets was measured using the Erweka crushing strength tester TBH 28 MDR (Erweka-Apparatebau GmbH, Offenbach, Germany). 	

Tablet bulk density. The bulk density of the produced tablets was determined by measuring their thickness, height and diameter (n510) using an 18E digital micrometer screw (Mitutoyo, Kanagawa, Japan). The weight of the tablet was determined using a digital balance (Sartorius Basic, Sartorius AG, Gottengen, Germany).		

Figure 4:  Relationship between mean particle size (d) and bulk density of the different EC fractions and Figure 5:  Effect on tablet density by changing the mean particle size (d). (Tablets were compressed using a compression force of 5 kN.)	

Drug release. The drug release from the produced tablets was determined using the Ph. Eur. method.13 Five tablets were used for each investigation. The following formulation was used to investigate the drug release: EC (100 mg) and theophylline monohydrate (100 mg). No lubricant was used, to prevent any interactions that could influence the drug release. The concentration of the released theophylline monohydrate was determined using a spectraphotometer (Spectronic 601; Milton Roy, Pennsylvania, USA) at a wavelength of 273 nm. 	

The tabletting properties of ethyl cellulose 7cP with different particle sizes were investigated by examining different parameters.				

Compactibility. The compactibility characterizes the relationship between the compression force used and the resulting crushing strength of the tablets. In Figure 1, the proportional relationship between the crushing strength and the square root of the mean particle size can be seen.		

Figure 6:  EC fraction at 3200 magnification.	

In Table I, the slope and correlation coefficient of the linear relationship between crushing strength and the square root of the particle size is shown. The slope of the linear regression increased as a result of raising the compression force. This means that the influence of the particle size on crushing strength increased as the compression force was raised.	

In Figure 1 and Table I, the proportional relationship between particle size and crushing strength can be seen. This result does not fit with the classic theory, which predicts an indirect proportional relationship between particle size and crushing strength. The reasons for this behaviour result from the fragmentation process and the amount of energy consumed for the fragmentation of large particles into smaller particles. This leads to an increase of the secondary binding points, resulting in tablets with a higher crushing strength. This result correlates with the results of Alderborn and Nystrom during the compression of sodium chloride.2 	

Katikaneni et al. investigated the particle size influence of EC 10cP (Dow Chemical Co., Midland, Michigan, USA) on the compressibility and compactibility of tablets.14 It was postulated that particle size has an indirectly proportional relationship with compressibility and compactibility. The discrepancy between our results and those of Katikaneni et al. probably results from the different EC sources.			

Figure 7:  EC fraction at 31000 magnification.	

Energy distribution of the force displacement diagram. Below, the friction energy (E1), the plastic deformation energy, that is, the effective energy (E2), and the elastic deformation energy (E3) of the force displacement diagram will be presented and discussed.	

The friction energy (E1) has a proportional relationship with particle size. This effect probably results from the bulk density increase, caused by increasing the particle size.15	

The effective energy increased as the particle size and compression force were raised (Figure 2). A linear relationship between the effective energy and the crushing strength could be observed, which correlates with the results of a previous study (Hersey et al.).15 	

We hypothesize that the higher effective energy of the larger particle size results from the amount of energy consumed for the fragmentation of large particles into smaller ones during compression.				

Figure 8:  Tablet surface of EC at 31000 magnification.	

Figure 3 shows the elastic behaviour of the different particle sizes of EC 7cP. It can be seen that the elastic deformation increases by increasing the compression force. We could not find statistically significant differences between the different fractions, whereas Katikaneni et al. stated that the elastic deformation increases by increasing the particle size.14	

Compressibility of different fractions of EC 7cP. The compressibility is the ability to consolidate under compression forces. To evaluate the compressibility of the different fractions, the bulk density of each fraction should be determined and discussed. In Figure 4 the bulk density of the different fractions versus the mean particle size can be seen. An indirect proportionality with a linear regression coefficient of 0.996 confirmed the linear relationship and the proportionality between particle size and bulk density.	

By increasing the particle size, the bulk density of the tablets decreases and the porosity increases. In Figure 5, the tablet bulk density versus the mean particle size can be seen. The tablets were compressed under the same compression force of 5 kN.	

According to Figure 5, an indirect proportionality between particle size and compressibility can be seen. The porosity of the produced tablets is dependent on the initial powder porosity.15 Our results replicate those of Katikaneni et al.14				

Table II:  Mean values and standard deviation of EC theophylline tablets.	

Scanning electron microscopy. Particle form of ethyl cellulose 7cP. The relationship between the particle size and shape was investigated using SEM. Two different fractions of EC 7cP (90-125 and 400-500 mm) were selected and the morphology of the particles was investigated. 	

Figure 6 shows the surface morphology of the particles using a 200-fold magnification. The particles of the 400-500 Âµm fraction seem to be more structured, and the surface is rough compared with the particles of the 90-125 Âµm fraction, which can activate the interlocking binding forces during compression. In Figure 7, using a 1000-fold magnification, the cavity of the larger fraction (400-500 mm) can be seen, whereas the small particle size shows a smooth shape.	

The primary form and size have a large influence on the interlocking binding forces.1,16 Huttenrauch proposed that the large particles are more able to wedge together than the small ones.1 Fuhrer stated that the plastic deformation process is much more important for tabletting ability than the primary particle form.17	

In this study, we can confirm the postulation of Huttenrauch and Fuhrer,17 because the large EC particles show higher plastic energy (Figure 2) and, simultaneously, they seem to have a better fragmentation behaviour compared with the small ones. Furthermore, the SEM photographs show that the particle cavities increased as the particle size increased. The cavity is a parameter that can give information regarding the pores in the particle. This can be quantified by measuring the specific surface area of the particles using an Hg-torsion pycnometer. The larger the cavity, the higher the tendency to build secondary binding points.				

Figure 9:  Theophylline release rate from EC matrix tablets using different particle size fractions.	

Morphology of the tablet surface. Figure 8 shows the surface of a tablet made with EC 7cP (90-125 mm) at a magnification of 1000. The marked fragmentation during compression can be seen. 	

Riempa et al.6,7 published that the fragmentation process increases the crushing strength of tablets. Furthermore, Riempa et al.7 proposed that the crushing strength can be affected by the degree of fragmentation and the specific surface of the tablet. The classic tabletting theory is not valid when the degree of fragmentation exceeds the influence of the specific area. Our results correspond to the previous research of Riempa et al.6,7	

Duberg and Nystrom have presented a theory concerning the volume-reduction of pharmaceutical materials, which often consist of aggregates of highly porous primary particles.18 These aggregates, or secondary particles, could behave as brittle units during the initial loading, with a negligible tendency to plastic deformation. This theory seems to apply to the behaviour of the large particles during compression. It is possible that the larger EC particles underwent fragmentation during compression, thus forming smaller, plastically deforming EC particles. This suggestion is in agreement with the scanning electron micrographs taken from the surface of the tablets (Figure 8). The original boundaries of the large EC particles had mostly disappeared, indicating a fragmentation of these particles. 	

Particle size influence on drug release. Two different fractions were used to investigate the influence of particle size on drug release. Fractions 90-125 mm and 400-500 mm were mixed with theophylline monohydrate (mean size 60 mm) and compressed at a force of 5 kN. The properties of the tablets produced can be seen in Table II. 	

As previously mentioned, the particle size of EC has a large influence on the subsequent tablet properties. A proportional relationship between particle size and crushing strength was found by compressing EC without any additional materials. 	

As can be seen in Table II, this relationship could not be found when EC was compressed with 50% theophylline monohydrate. We propose that the specific surface area in this instance exceeds the degree of fragmentation, which was found during the compression of pure EC. Landin et al. showed that the proportional relationship between MCC particle size and crushing strength was reversed when using 10% prednisone.3	

The drug release behaviour of both EC fractions can be seen in Figure 9. Increased particle size led to an increase of the drug release rate. The reason for the increased drug release rate can be found in the higher porosity and lower crushing strength of the tablets of both fractions (Table II). The porosity influences the diffusion rate and the crushing strength influences the erosion rate; the influence of the porosity on drug release has often been discussed19,20 and general opinion states that tablets with higher porosity resulted in a higher drug release rate.	

The particle size of ethyl cellulose 7cP has a large influence on the properties of the tablets produced. A directly proportional relationship between particle size and crushing strength could be observed by compressing the different fractions of EC. This relationship was reversed by using theophylline monohydrate (50%). For EC-matrix tablets, increasing the particle size led to an increase in the porosity of the produced tablets. Increasing the particle size improved the drug release rate, which was probably related to an increase in porosity and a reduction in the crushing strength of the tablets produced.						

Articles

The aim of this work was to investigate the compactibility, compressibility and drug release behaviour of different fractions of a commercially available ethyl cellulose.

Ethyl Cellulose as a Matrix Former: The Influence of Particle Size on Tabletting and Drug Release Behaviour

In this study, the authors investigated the compression behavior of excipients with different molecular weights using the modified Fraser-Suzuki function and correlated the S-parameter of this function with Young's modulus.