Back to the title page | E-thesis main page



2.1 Porosity

Porosity refers to the pore space in a material (Sing et al. 1985). Internal surface of the material comprises the pores and cracks that are deeper than they are wide. An open pore (Fig. 1.) is a cavity or channel that communicates with the surface of the particle. Closed pores (Fig. 1.) are inside the material. These open and closed pores are called intraparticular porosity of the material. A void is a space between particles, i.e. interparticular porosity. Powder porosity consists of the pores in and voids between the particles.

Pores are classified according to size into three categories; micropores (pore diameter smaller than 2 nm), mesopores (pore diameter 2 – 50 nm) and macropores (pore diameter larger than 50 nm) (Sing et al. 1985). With nitrogen gas adsorption, depending on the equipment used, pore diameter range of 0.3 – 300 nm, i.e. mesopores and macropores, are determined. Low-pressure mercury porosimetry determines macropores (pore diameter 14 – 200 m m), and high-pressure porosimetry mesopores and macropores (pore diameter 3 nm – 14 m m), depending on the equipment.


Figure 1. Pore types a) open pore b) closed pore c) ink-bottle pore d) cylindrical, open- ended pore.


2.2 Mercury porosimetry

2.2.1 Mercury porosimetry procedure

In mercury porosimetry, gas is evacuated from the sample cell, and mercury is then transferred into the sample cell under vacuum and pressure is applied to force mercury into the sample. During measurement, applied pressure p and intruded volume of mercury, V, are registered. As a result of analysis, an intrusion-extrusion curve is obtained (Fig. 2.). Parameters describing the pore structure of the sample can be calculated from the data obtained.


Figure 2. Intrusion-extrusion curve.



2.2.2 Washburn equation

Mercury porosimetry is based on the Washburn equation (Washburn 1921)

p × r = -2 × g × cos q ,(1)

where r is the radius of the pore where mercury intrudes, g surface tension of mercury and q contact angle of the mercury on the surface of a solid sample. Generally used values for surface tension and contact angle of mercury are 480 mNm -1 and 140 ° , respectively.


The Washburn equation (1) can be derived from the equation of Yang and Dupre

g SV = g SL + g LV × cos q ,(2)

where g SV is interfacial tension between solid and vapor, g SL interfacial tension between solid and liquid, g LV interfacial tension between liquid and vapor and q the contact angle of the liquid on the pore wall (Lowell & Shields 1991).


The work, W, required when liquid moves up the capillary during capillary rise when the solid-vapor interface disappears and solid-liquid interface appears is

W = ( g SL - g SV ) D A,(3)

where D A is the area of the capillary wall covered by liquid when its level rises.


According to equations (2) and (3),

W = -( g LV × cos q ) D A.(4)

The work required to raise a column of liquid a height h in a capillary with the radius r is identical to work that must be used to force liquid out of the capillary. When a volume V of liquid is forced out of the capillary with a gas at a constant pressure above ambient, D p gas , the work is presented as

W = V D p gas . (5)

When equations (4) and (5) are combined

D p gas V = -( g × cos q ) D A.(6)

When the capillary is circular in cross-section, parameters V and D A are given by p r 2 L and 2 p rL, when L is length of the capillary. When these terms are substituted to the equation (6), it yields the Washburn equation (1)

p × r = -2 × g × cos q .


2.2.3 Total pore volume and total pore surface area

Total pore volume (V tot ) is the total intruded volume of mercury at the highest pressure determined.

Total pore surface area (S) is calculated by Equation 7

. (7)

Total pore surface area is the area above the intrusion curve (Fig. 2.), and it is thus modelless and independent of the geometrical pore shape (Rootare & Prenzlow 1967).


2.2.4 Mean and median pore diameter

The mean pore diameter (d mean ) is calculated by Equation 8


based on an assumption of cylindrical shape of pores open at ends (Emmett & Dewitt 1943). Median pore diameter (d median ) is the pore diameter at which 50% of the total intruded volume of mercury is intruded into the sample (Dees & Polderman 1981). In general, mean pore diameter emphasizes the smaller pores rather more than median pore diameter.


2.2.5 Volume pore size distribution

Volume pore size distribution, D v (d), is defined as the pore volume per unit interval of pore diameter (d) by Equation 9


(Ritter & Drake 1945). Volume pore size distribution is based on a model of cylindrical pores (Fig. 1.).


2.2.6 Use of mercury porosimetry in pharmaceutical powder technology

Mercury porosimetry has been used in the studies presented in the following table.

Table 1. Examples of the use of mercury porosimetry in pharmaceutical powder technology.




Marshall & Sixmith 1975, Stanley-Wood 1978, Krycer et al. 1982, Carli & Motta 1984, Zouai et al. 1996, Tobyn et al. 1998


Strickland et al. 1956, Fujiwara et al. 1966, Nicholsson & Enever 1974, Opankule & Spring 1976, Stanley-Wood & Shubair 1979, Krycer et al. 1982, Veillard et al. 1982, Zoglio & Carstensen 1983, Juppo et al. 1994, Knight et al. 1998


Reich & Gstirner 1968, Selkirk & Ganderton 1970, Selkirk 1974, Sixmith 1977, Stanley-Wood 1978, Dees & Polderman 1981, Vromans et al. 1985, Riippi et al. 1992, Wikberg & Alderborn 1992, Landin et al. 1993a, Pourkavoos & Peck 1993, Faroongsarng & Peck 1994, Juppo 1996a, Juppo 1996b, Juppo 1996c, Zouai et al. 1996, Riippi et al. 1998

Cellulose beads

Ek. et al. 1994, Ek. et al. 1995


Millili & Schwartz 1990, Niskanen 1992a, Niskanen 1992b, O´Connor & Schwartz 1993, Bataille et al. 1993, Vertommen et al. 1998


Although mercury porosimetry has been widely used in determination of pharmaceutical samples, pretreatment of the samples before measurement or the effect of scanning speed on the results of pharmaceutical sample determinations has not been studied.


2.2.7 Advantages and limitations of mercury porosimetry

Mercury porosimetry is a relatively rapid method, with which a wide pore diameter range (3 nm - 200 m m) and variety of porosity parameters can be determined. However, the method is rarely used in quality control measurements, because the time used for a single analysis is 30 – 45 minutes (Webb & Orr 1997). The measurement itself is automatic, which allows personnel to engage other work at the same time. Dimensions of the sample cell limit the size of the sample. However, the diameter of the sample cell is commonly 1 cm, which normally allows the determination of pharmaceutical samples. With the method, only pores that reach the surface of the sample can be determined. The sample must be dry, because mercury cannot intrude into the sample when voids are filled with another liquid (Ek et al. 1995). Samples with a fine pore structure are difficult to degas, and adsorbed layers reduce effective pore diameter and pore radius values (Allen 1997).

During measurement, high pressures to force mercury into small pores may compress the sample (Palmer & Rowe 1974, Dees & Polderman 1981, Johnston et al. 1990, Ek. et al. 1994, Allen 1997, Webb & Orr 1997). This effect can be shown especially in samples containing closed pores (Webb & Orr 1997), and is observed as a too large volume of small or medium sized pores. Damage or compression of highly porous silica has been reported (Brown & Lard 1974, Johnston et al. 1990). However, no damage or sample compression of lactose, mannitol or glucose tablets or carbon black particles has been observed (Moscou & Lub 1981, Dees & Polderman 1981, Juppo 1995). In addition to compression of sample, also mercury, the sample cell or residual air may be compressed with increasing pressure (Allen 1997). These compressional effects and the effect of a rise in temperature (van Brakel et al. 1981) can be eliminated with the use of hydraulic oil as a medium for transferring pressure (Lowell 1980).

Usually, constant surface tension and contact angle values are used for mercury (Allen 1997). However, contact angle may differ due to differences in the surfaces of the samples. Contact angle can be determined for each material studied, and the corrected value can be used in determinations.

Mercury porosimetry overestimates the volume of the smallest pores (Auvinet & Bouvard 1989). This is due to ink-bottle shaped pores (Dees & Polderman 1981, Allen 1997) and interconnected pores (Allen 1997) that shift the volume pore size distribution towards smaller pores. The diameter of the pore opening into the surface of the sample determines when mercury is intruded into the sample. Large pores with a small opening are thus filled at high pressures, and detected as smaller pores than they actually are. Pore size distributions obtained with incremental and continuous mode differ (Allen 1997), and the results obtained with these two methods are thus not comparable. In incremental mode, the pressure is increased in steps. In continuous mode, the pressure is increased continuously at a predetermined rate.

Non-capillary pore structure and limitations of the Washburn equation in determination of the smallest pores are the reasons for the differences between pore size distributions determined with mercury porosimetry and nitrogen adsorption (De Wit & Scholten 1975). However, total pore volume and pore surface area results are not dependent on pore shape (Rootare & Prenzlow 1967), and the shape of pore size distribution is not remarkably different from the true distribution in spite of the assumption of a circular cross section of the pores (Ritter & Drake 1945). The pore size distribution obtained with mercury porosimetry has been a useful parameter in characterisation of tablets (Juppo 1995).


2.3 Nitrogen gas adsorption method

2.3.1 Total pore volume and volume pore size distribution

Total pore volume, i.e. volume of the pores in a pre-determined pore size range can be determined from either the adsorption or the desorption phase.

The volume pore size distribution is determined according to the BJH model (Barrett et al. 1951). The corrected Kelvin equation


is used to calculate the relative pressure of nitrogen in equilibrium with a porous solid, and applied to the size of the pores where capillary condensation takes place. The equation was presented in its original form by Thomson (1871).

In the Kelvin equation, p is the equilibrium vapor pressure of a liquid in a pore of radius r, p 0 the equilibrium pressure of the same liquid on a plane surface, g surface tension of the liquid, V L molar volume of the liquid, q the contact angle with which the liquid meets the pore wall, R the gas constant and T absolute temperature. When the meniscus of condensate is concave, capillary condensation will proceed in pores of radius r as long as the adsorptive pressure is greater than pressure p.

The equation is derived as follows. Liquid within the pore is in equilibrium with its vapour. A molar quantity of liquid ( d n) outside of the pore, where its equilibrium pressure is p 0 , is changed inside of the pore, where its equilibrium pressure is p. During the process, the total increase in free energy d G is the sum of three energies; d G 1 = evaporation of d n moles of liquid at pressure p 0 , d G 2 = changing d n moles of vapor from pressure p 0 to pressure p and d G 3 = condensation of d n moles of vapor to liquid at pressure p.

Condensation and evaporation are equilibrium processes, d G 1 = d G 3 = 0. Thus, the change in free energy during the process is presented as


when the vapor behaves as a perfect gas.

During condensation of vapor in the pores, the solid-liquid interface increases and solid-vapor interface ( d A) decreases. The change in free energy during this process is

d G 4 = d A ( g SV - g SL ).(12)

When g SV - g SL = - g LV cos( q ), wetting angle q is 0, and d G 4 = d G 2 , the equation can be presented as


The volume condensed in the pores is d V c = V L d n, where V L is molar volume. Thus, the equation can be presented as


The equation can be further organized to be


For cylindrical pores with radius r and length L, V c = p r 2 L and A = 2 p rL, and


which leads to the Kelvin equation (10)


Pore size distribution can be determined from the adsorption or desorption data of the isotherm. A cylindrical pore model is assumed, with the further assumption of open-ended pores and absence of pore networks. The pore size distribution determined from nitrogen desorption data and the distribution obtained from the intrusion phase of mercury porosimetry describe pore structure similarly (Conner et al. 1986).


2.3.2 Specific surface area

Specific surface area is calculated according to the BET equation (Brunauer et al. 1938)

, (17)

where V is volume adsorbed, V m volume of monolayer, p sample pressure, p 0 saturation pressure and c constant related to the enthalpy of adsorption (BET constant). The specific surface area (S BET ) is then calculated from V m by the following equation

, (18)

where n a is Avogadro constant, a m the cross sectional area occupied by each nitrogen molecule (0.162 nm 2 ), m weight of the sample and V L the molar volume of nitrogen gas (22414 cm 3 ). The theory is based on the assumption that the first adsorbed layer involves adsorbate/adsorbent energies, and the following layers the energies of the adsorbate/adsorbate interaction.


2.3.3 Use of nitrogen adsorption in pharmaceutical powder technology

Nitrogen adsorption has been used in studies listed in the following table.

Table 2. Examples of the use of nitrogen adsorption in pharmaceutical powder technology.




Marshall & Sixmith 1975, Stanley-Wood & Johansson 1978, Stanley-Wood & Shuibar 1979, Zografi et al. 1984, Niskanen et al. 1990, Landin et al. 1993a, Landin et al. 1993b, Stubberud et al. 1996


Stanley-Wood & Shuibar 1979, Stubberud et al. 1996


Sixmith 1977, Stanley-Wood & Johansson 1978, Vromans et al. 1988, Faroongsarng & Peck 1994, Riippi et al. 1998


Niskanen et al. 1990, Niskanen 1992a, Niskanen 1992b


2.3.4 Advantages and limitations of nitrogen gas adsorption

Many parameters that describe the pore structure of a sample, for example pore volume, specific surface area and pore size distribution, can be determined with this method. One drawback is that the time used for a single analysis can be hours. However, measurements can be done automatically for example during the night. The pore diameter range that can be determined is from 0.3 to 300 nm, a range not completely covered by mercury porosimetry.

With nitrogen adsorption, only open pores are determined, and the cylindrical pore model is assumed in pore size distribution measurements (Allen 1997). The desorption isotherm in the characterisation of pore size distribution is affected by the pore network; when pressure is reduced, liquid will evaporate from large open pores, but pores of the same size that are connected to the surface with narrower channels remain filled (Allen 1997). This changes the shape of the pore size distribution. The samples come into contact with the temperature of liquid nitrogen (-196 ° C) during analysis, which may destroy the sample.


2.3.5 Comparison of nitrogen adsorption and mercury porosimetry methods

Pore structure analysis by mercury porosimetry is faster than by nitrogen adsorption. In mercury porosimetry and nitrogen adsorption determinations, two different physical interactions take place. Both methods are based on surface tension, capillary forces and pressure. With mercury porosimetry, large pores at the intrusion phase are determined first, while with nitrogen adsorption, the smallest pores are measured first at the adsorption phase (Webb & Orr 1997). The determination range of high-pressure mercury porosimetry is wider (pore diameter 3 nm – 14 m m) than that of nitrogen adsorption (0.3 – 300 nm), and mercury porosimetry determines larger pores that are out of the detection range of nitrogen adsorption (Fig. 3.). With nitrogen adsorption, the smallest pores that are out of range of mercury porosimetry, can be determined. However, results of the two methods can be compared. The comparable parameters are total pore volume, volume pore size distribution and specific surface area/total pore surface area. Although the pore size range that can be determined with adsorption is narrower than that obtained with mercury porosimetry, it is more widely used (Allen 1997).


Figure 3. Pore diameter ranges determined with mercury porosimetry and nitrogen adsorption.


Milburn et al. (1991) have obtained similar pore volume values for silica samples with these methods. If the sample contains pores larger than 300 nm, the pore volume obtained with mercury porosimetry is larger than that determined with nitrogen adsorption (Webb & Orr 1997). Pore size distributions determined from mercury porosimetry intrusion data and nitrogen desorption data describe the pore structure similarly (Conner et al. 1986). Stanley-Wood (1978) and Conner et al. (1986) have formed almost the same pore size distributions with the two methods for uncompacted magnesium trisilicate and for Degussa aerosols, and Faroongsarn and Peck (1994) consistent pore size distributions for dicalcium phosphate dihydrate tablets. On the other hand, different pore size distributions have been obtained with these methods for silicas, iron oxide-chromium oxide catalyst, aerosil powder and chrysotile powder (Brown & Lard 1974, De Wit & Scholten 1975).

Even if the pore volume values obtained do not agree, surface area values may be similar (Webb & Orr 1997). This is because small pores have a larger effect on the surface area. According to Webb and Orr (1997), these two techniques are equal when pore size ranges from 3 to 300 nm are compared. Larger surface area values have been obtained with mercury porosimetry than with nitrogen adsorption for lactose tablets (Dees & Polderman 1981) and for silica samples (Milburn et al. 1991). Mikijelj & Varela (1991) obtained equivalent surface area results for magnesium oxide and diatomite compacts. Adkins and Davis (1988) have used a corrected contact angle in mercury porosimetry to make the surface area values similar. According to Milburn and Davis (1993), the correlation between surface areas obtained with these methods is poor if the samples have low surface area.


2.4 Water vapour adsorption

2.4.1 Adsorption of water on the surface of a solid sample

Water settles on the surface of a solid first as a monolayer and with increasing moisture as multilayers (Zografi 1988). The first layer is hydrogen bonded to the surface of the solid (Ahlneck & Zografi 1990) and is immobile (Ozeki et al. 1991). Additional layers can behave as a liquid (Ozeki et al. 1991), move along the surface of the sample (Zografi 1988, Ahlneck & Zografi 1990) and even cause dissolution of the solid (Ahlneck & Zografi 1990). According to Ozeki et al. (1991), at least the first layer of adsorbed water on the surface of chrysotile crystal has behaved like a solid, and water in fourth and higher layers behaved like liquid.


2.4.2 Behaviour of water in the pores

The diameter of a water molecule is 0.28 – 0.3 nm (Ozeki 1989). Micropore filling is a primary physisorption process, whereas physisorption in mesopores occurs in two stages; monolayer-multilayer adsorption and capillary condensation (Sing et al. 1985). At first, water is adsorbed into the surface of the pore wall (Fig. 4.), and then water is condensed and fills the core of the pore (Aharoni 1997). The reason why condensation occurs is that the surface of the condensed water in the pore is concave, and its vapor pressure is smaller than saturation pressure (Aharoni 1997). Relation between the diameter of the water-filled pores and the condensation pressure can be calculated with Equation (10), which is valid only in the pore radius range from 1.8 to 30 nm, part of which is measurable by high-pressure mercury porosimetry.


Figure 4. Adsorption and capillary condensation of water into a pore with radius r with increasing relative humidity RH1 < RH2 < RH3. At low relative humidity (RH1) water adsorbs as layers to the walls of the pore. With increasing relative humidity (RH2), thickness of adsorbed layers increases. Finally, at even higher relative humidity (RH3), water fills the pore.


At low relative humidities, water fills the smallest pores and adsorbs in layers to the surface of the sample. The capillary condensation model cannot be used for micropores. Also, the Kelvin equation handles the liquid-vapor interface within the pore as curvature and contact angle, which cannot be used with micropores (Aharoni 1997). When humidity of the surrounding air increases, water fills the larger pores. The first adsorbed layer is immobile, but in some cases water behaves like a liquid on flat surfaces and in wide pores (Ozeki et al. 1991). Ozeki et al. (1991) have studied the behaviour of water molecules on chrysotile crystal samples with cylindrical mesopores of 7 nm in diameter. Water molecules were adsorbed by capillary condensation to the mesopores, and formed a liquid-like phase. However, some pores of cement tiles have been reported at intermediate humidity to fill completely while others have remained empty (Bohris et al. 1998). Allen et al. (1998) observed with NMR technique that bulk water in the pores of silica forms puddles into the corners and cavities of irregular pores. This occurs even at low filling fractions of water together with physisorbed layers.


2.5 Mannitol

2.5.1 Characteristics of mannitol and its behaviour in wet granulation and tableting

Mannitol is a sugar alcohol and isomeric with sorbitol (Fig. 5.). Mannitol is used as a filler in conventional tablets. It is non-hygroscopic, and resists moisture sorption even at high relative humidities. Therefore, it has special value in tableting of moisture-sensitive drugs. Solubility of mannitol is 17 g/ 100 g of water at + 25 ° C.


Figure 5. Structural formula of mannitol.


Due to its needle-like shape and thus poor flowability, mannitol powder is often granulated. Mannitol has poor wettability in wet granulation, which is due to the electric charge and cohesivity of dry mannitol powder (Juppo et al. 1992). This has led to bimodal size distribution and angular shape of mannitol granules after wet granulation. Mannitol particles dissolve and recrystallise on the larger particles during wet granulation (Juppo 1995). The small particles also attach to each other by solid bridges formed by recrystallised mannitol or by binder. The granules produced have a high porosity percentage (Juppo & Yliruusi 1994). The needle-like particles form a fibrous network with a large number of small pores.

Compression of mannitol powder has been presented in few papers. Mannitol is characterised as ductile, that deforms plastically under loading (Roberts & Rowe 1987, Bassam et al. 1990). Evidently, hydrogen bonding due to the hydroxyl groups (Fig. 5.) is one bonding mechanism for mannitol powder (Juppo 1995). In addition, van der Waals attractions, electrostatic forces and mechanical interlocking takes place under compression. Mannitol tablets compressed from crystals have had lower strength than those compressed from granulated powder (Krycer et al. 1982).

Porous mannitol granules have deformed plastically and also fragmented under compression (Juppo et al. 1995). Under compression, porous mannitol granules with a fibrous structure interlock mechanically and undergo fragmentation and plastic deformation. When mannitol is compressed with low compression pressure, large pores vanish, the volume of smaller pores is reduced, indicating that the intragranular porosity of mannitol granules also decreases (Juppo 1996a).


2.6 Microcrystalline cellulose

2.6.1 Characteristics of microcrystalline cellulose and its mechanism of swelling

Microcrystalline cellulose powder is composed of porous particles. Microcrystalline cellulose is used as binder/diluent in wet granulation and direct compression. It is hygroscopic in nature, and insoluble in water, but swells when in contact with water. The structural formula of microcrystalline cellulose is presented in Figure 6. Glucose molecules are linked via beta-glucoside bonds. Intermolecular hydrogen bonds are formed between these cellulose polymers, glucan chains aggregate and form fibres. Thus, the structure has microcrystalline nature.


Figure 6. Structural formula of microcrystalline cellulose.


When microcrystalline cellulose is stored under humid conditions, water penetrates the amorphous structure (Zografi et al. 1984). Khan et al. (1988) have reported that water molecules are accommodated into the internal structure of microcrystalline cellulose in the spaces between the cellulose chains when the amount of water in the sample is below 3 wt%, and that no swelling occurs. According to Khan and Pilpel (1987), water disrupts the cellulose-cellulose bonds and forms new hydrogen bonds between them, which causes swelling of the sample and increases the volume of the particles. Figure 7 shows the mechanism of hydrogen bonding between water molecules and cellulose. At first, one sorbed water molecule is linked to two 6-OH groups in neighbouring cellulose chains (Fig. 7 a). When 3 wt% of moisture is present (Fig. 7 b), each water molecule is attached to the cellulose chain by only one hydrogen bond. When more moisture is absorbed, 6-OH groups in cellulose chain are hydrogen bonded with water, and weakly hydrogen-bonded water probably forms a bulk water phase (Fig. 7 c). This phase takes place when 6 wt% of water or more is absorbed to the structure of microcrystalline cellulose.


Figure 7. Absorption of water into the structure of microcrystalline cellulose (Khan & Pilpel 1987) 1 .


Similarly, Zografi and Kontny (1986) have explained water vapour sorption of microcrystalline cellulose with a three-step model. At low relative humidities, water is bound to available anhydroglucose units in the amorphous regions of cellulose with a stoichiometry of one water molecule per anhydroglucose unit. At relative humidities up to about 60 %, polymer-polymer hydrogen bonds are broken, which makes more primary binding sites available and allows water to begin to bind to other water molecules already bound to anhydroglucose units. Finally, at even higher relative humidities, water can also bind to other water molecules, including those not bound to primary sites.


2.6.2 Behaviour of microcrystalline cellulose in wet granulation and tableting

Microcrystalline cellulose works as a binder in wet granulation (Doelker 1993), but its good compactibility has been found to disappear due to loss of plasticity in wet granulation (Staniforth et al. 1988). Millili (1990) has reported that the degree of hydrogen bonding of microcrystalline cellulose is not responsible for harder pellets produced with water. He has explained densification of microcrystalline cellulose by autohesion, solid solid diffusion. Chatrath (1992) has called the reduced compactibility of microcrystalline cellulose after wet granulation ‘quasi-hornification’ to describe the increased intraparticle hydrogen bonding. Kleinebudde (1997) has explained the behaviour of microcrystalline cellulose in wet granulation, extrusion and spheronization by a crystallite gel model. In that model, the crystallites or their agglomerates of microcrystalline cellulose form a framework by crosslinking with hydrogen bonds at the amorphous ends. During drying, more hydrogen bonds are formed. No changes were observed at the level of individual crystallites. Increased internal hydrogen bonding in microcrystalline cellulose after wet granulation was observed with near IR –technique by Buckton et al. (1999). Ek and Newton (1998) have explained the deformation of microcrystalline cellulose during extrusion/spheronization with water by a sponge model. Various explanations for the behaviour of microcrystalline cellulose during processing with water have been put forward recently. However, increased internal hydrogen bonding appears to be the reason for the densified structure of microcrystalline cellulose granules after wet granulation.

Hydrogen bonding, large particle surface area, filamentous structure of the cellulose microcrystals and mechanical interlocking of irregular elongated particles are responsible for the excellent binding properties of microcrystalline cellulose in tableting (Bolhuis & Lerk 1973). Microcrystalline cellulose powder deforms plastically (Lamberson & Raynor 1976, David & Augsburger 1977, Shangraw et al. 1981, Staniforth et al. 1988). The modal pore radius of microcrystalline cellulose tablets compressed from powder has decreased with increasing compression pressure (Sixmith 1977). Strength of interparticle bonding was greater for the powder samples of microcrystalline cellulose than for granules (Staniforth et al. 1988). Staniforth et al. (1988) suggested that in the granules most of the compression force was used for breaking up the primary granule structure and hence did not establish areas of intimate contact to provide strong bonds between the cellulose particles. In compression of pellets, the dominating mechanism of compression has been permanent deformation in combination with densification of the pellets (Johansson et al. 1995, Johansson & Alderborn 1996, Johansson et al. 1998). Only limited fragmentation of pellets during compression has occurred. The effect of compression on specific surface area and porosity of microcrystalline cellulose tablets compressed from powder (Sixmith 1977, Zouai et al. 1996), from pellets (Johansson et al. 1995, Johansson & Alderborn 1996, Johansson et al. 1998) and also from granules (Chatrath 1992) has been studied. However, studies concerning the deformation of microcrystalline cellulose granules under compression based on pore structure with mercury porosimetry and nitrogen adsorption have not been thoroughly reported.


1 Reprinted from Powder Technology, 50, Khan & Pilpel, An investigation of moisture sorption in microcrystalline cellulose using sorption isotherms and dielectric response, p. 239, copyright (1987), Elsevier Science.


Back to the title page | E-thesis main page