DRUG DEVELOPMENT AND INDUSTRIAL PHARMACY, 2 1 (6), 643-661 (1995)
`
`RESEARCH PAPERS
`
`MORPHOLOGICAL, PACKING, FLOW AND TABLETING PROPERTIES OF
`NEW AVICEL TYPES
`E. Doelkerl, D. Maswellel, F. Veuillezl and P. Humbert-Droz2
`
`lSchool of Pharmacy, University of Geneva, Sciences II,
`Quai Ernest-Ansermet 30, CH-1211 Geneva (Switzerland)
`Labatec-Pharma S.A., Rue du Cardinal Journet,
`CH-1217 Meyrin (Switzerland)
`
`ABSTRACT
`
`The six Avicel products designed for compression - the classical grades
`PH-105, PH-103, PH-101 and PH-102, and the new Avicels PH-112 and PH-
`200 - have been submitted to a comparative investigation for both their basic
`and tableting properties. According to the manufacturer all these products differ
`by their nominal particle size and moisture content.
`Basic properties of the powders were first determined, namely moisture
`content (loss on drying and Karl Fischer titration), particle size and shape
`(sieving and image analysis), densities (true bulk and tap densities, Hausner
`ratio) and flow properties (vibratory hopper technique).
`As tableting properties, the compactibility of the powders and the effect of
`adding a hydrophobic lubricant (0.5% magnesium stearate) on the compact
`strength were evaluated by preparing compacts at a given applied pressure
`using a hydraulic press. Weight and dimensional variations were assessed by
`
`Note: The symbols for registered names and trademarks have been
`systematically omitted for sake of simplicity.
`
`Copyright 0 1995 by Marcel Dekker, Inc
`
`643
`
`DRUG DEVELOPMENT AND INDUSTRIAL PHARMACY, 21(6), 643-661 (1995)
`
`RESEARCH PAPERS
`
`MORPHOLOGICAL, PACKING, FLOW AND TABLETING PROPERTIES OF
`NEW AVICEL TYPES
`
`E. Doelker!, D. Massuelle!, F. Veuillez! and P. Humbert-Droz2
`
`1School of Pharmacy, University of Geneva, Sciences ll,
`Quai Ernest-Ansermet 30, CH-1211 Geneva (Switzerland)
`2 Labatec-Pharma S.A., Rue du Cardinal Journet,
`CH-1217 Meyrin (Switzerland)
`
`ABSTRACT
`
`The six Avicel products designed for compression - the classical grades
`
`PH-105, PH-103, PH-101 and PH-102, and the new Avicels PH-112 and PH-
`
`200 - have been submitted to a comparative investigation for both their basic
`
`and tableting properties. According to the manufacturer all these products differ
`
`by their nominal particle size and moisture content.
`
`Basic properties of the powders were first determined, namely moisture
`
`content (loss on drying and Karl Fischer titration), particle size and shape
`
`(sieving and image analysis), densities (true bulk and tap densities, Hausner
`
`ratio) and flow properties (vibratory hopper technique).
`As tableting properties, the compactibility of the powders and the effect of
`
`adding a hydrophobic lubricant (0.5% magnesium stearate) on the compact
`
`strength were evaluated by preparing compacts at a given applied pressure
`
`using a hydraulic press. Weight and dimensional variations were assessed by
`
`Note: The symbols for registered names and trademarks have been
`systematically omitted for sake of simplicity.
`
`643
`
`Copyright © 1995 by Marcel Dekker, Inc.
`
`MYLAN - EXHIBIT 1010
`
`MYLAN - EXHIBIT 1010
`
`

`

`644
`
`DOELKER ET AL.
`
`preparing tablets at a target crushing strength of 70 Newtons on a high speed
`machine.
`The comparison of the conventional Avicel PH grades showed that
`Avicel PH-105 differed markedly in its properties (high compressibility on
`tapping, high compactibility, inacceptable tablet weight variability and very
`poor disintegrating properties) from the other grades.
`As to the two new Avicel PH grades, conflicting results with the literature
`were obtained with the low-moisture product Avicel PH-112. We observed, like
`other authors but in contrast to manufacturer's data, values of compactibility
`and strength reduction ratio upon lubrication as well as of the coefficient of
`tablet weight variation similar to those of the standard Avicel PH-102, of
`comparable particle size.This can be certainly explained by an uptake of
`moisture of the Avicel PH-112 powder as proved experimentally. This would
`limit the use of this material to an air-conditioned room.
`The large particle size product Avicel PH-200 displayed a compactibility
`close to that of all the other Avicel PH grades (except PH-105), but the highest
`susceptibility to magnesium stearate. As expected, because it is free-flowing,
`Avicel PH-200 gave the lowest tablet weight variability. Additionnally, the two
`new grades showed disintegrating properties similar to those of Avicel PH-103,
`PH-102 and PH-101. Finally, one should bear in mind that the small
`differences reported here may not be significant because of substantial inter-
`batch variability.
`
`INTRODUCTION
`
`The preparation of microcrystalline cellulose (MCC) has been patented in
`the beginning of the 1960s by Battista and Smith of the American Viscose
`Company (1) and MCC has been put on the market as a pharmaceutical
`tableting excipient in 1963 under the trade name Avicelm (it is now sold by the
`FMC Corp.). Since then, its tableting properties, especially as a dry binder for
`direct compression, have been extensively investigated. Some of these
`studies compare the standard Avicel product with other types of Avicel and
`
`644
`
`DOELKER ET AL.
`
`preparing tablets at a target crushing strength of 70 Newtons on a high speed
`machine.
`
`The comparison of the conventional Avicel PH grades showed that
`
`Avicel PH-105 differed markedly in its properties (high compressibility on
`
`tapping, high compactibility, inacceptable tablet weight variability and very
`
`poor disintegrating properties) from the other grades.
`
`As to the two new Avicel PH grades, conflicting results with the literature
`
`were obtained with the low-moisture product Avicel PH-112. We observed, like
`
`other authors but in contrast to manufacturer's data, values of compactibility
`
`and strength reduction ratio upon lubrication as well as of the coefficient of
`
`tablet weight variation similar to those of the standard Avicel PH-102, of
`
`comparable particle size.This can be certainly explained by an uptake of
`
`moisture of the Avicel PH-112 powder as proved experimentally. This would
`
`limit the use of this material to an air-conditioned room.
`
`The large particle size product Avicel PH-200 displayed a compactibility
`
`close to that of all the other Avicel PH grades (except PH-105), but the highest
`
`susceptibility to magnesium stearate. As expected, because it is free-flowing,
`
`Avicel PH-200 gave the lowest tablet weight variability. Additionnally, the two
`
`new grades showed disintegrating properties similar to those of Avicel PH-103,
`
`PH-102 and PH-101. Finally, one should bear in mind that the small
`
`differences reported here may not be significant because of substantial inter-
`
`batch variability.
`
`INTRODUCTION
`
`The preparation of microcrystalline cellulose (MCC) has been patented in
`
`the beginning of the 1960s by Battista and Smith of the American Viscose
`
`Company (1) and MCC has been put on the market as a pharmaceutical
`
`tableting excipient in 1963 under the trade name Avicel® (it is now sold by the
`
`FMC Corp.). Since then, its tableting properties, especially as a dry binder for
`
`direct compression, have been extensively investigated. Some of these
`
`studies compare the standard Avicel product with other types of Avicel and
`
`

`

`PROPERTIES OF NEW AVICEL TYPES
`
`645
`
`since the 1980s with generic products (see ref. 2 for a complete review of the
`subject).
`For more than 20 years, only four Avicel grades were available for direct
`compression. These four grades were obtained by varying the hydrolysis,
`shearing and drying conditions :
`
`- Avicel PH-101 has typically an average particle size of 50 pm and a
`maximum moisture content of 5%,
`
`- Avicel PH-102 has a larger average particle size (100 pm) than that of
`Avicel PH-101 but a similar moisture content.
`
`- Avicel PH-103 has an average particle size similar to that of Avicel PH-101
`(50 pm) but a lower moisture content (max. 3%),
`
`- Avicel PH-105 has a smaller average particle size (20 pm) than Avicel PH-
`101 but a similar water content.
`
`Precise particle size distributions of these products can be found for instance
`in references 3 to 6. Though a controversial subject, all these Avicel PH
`products, with perhaps the exception of Avicel PH-102, suffer from poor flow
`properties. Further, and with the exception of Avicel PH-103, their water
`content is too high for making tablets containing water-sensitive drugs.
`To remedy these drawbacks, two new Avicel PH grades were launched
`these last years :
`
`- Avicel PH-1 12, having the typical particle size specifications of Avicel PH-
`102 (1 00 pm) but very low moisture content ( m a . 1.5%) which makes it an
`ideal excipient for moisture-sensitive substances (for this reason Avicel PH-
`112 was first named Avicel PH-102 SLM, Special Low Moisture),
`
`- Avicel PH-200 having a larger average particle size (200 pm) than Avicel
`PH-102, but a similar moisture content, was designed for improved flow.
`
`The aim of this work was to examine the properties of these new
`excipients (moisture content, morphology, packing, flowability) of importance
`
`PROPERTIES OF NEW AVICEL TYPES
`
`645
`
`since the 1980s with generic products (see ref. 2 for a complete review of the
`
`subject).
`
`For more than 20 years, only four Avicel grades were available for direct
`
`compression. These four grades were obtained by varying the hydrolysis,
`
`shearing and drying conditions :
`
`-
`
`-
`
`-
`
`-
`
`Avicel PH-101 has typically an average particle size of 50 um and a
`maximum moisture content of 5%,
`
`Avicel PH-102 has a larger average particle size (100 um) than that of
`Avicel PH-101 but a similar moisture content,
`
`Avicel PH-103 has an average particle size similar to that of Avicel PH-101
`(50 um) but a lower moisture content (max. 3%),
`
`Avicel PH-105 has a smaller average particle size (20 um) than Avicel PH-
`101 but a similar water content.
`
`Precise particle size distributions of these products can be found for instance
`
`in references 3 to 6. Though a controversial subject, all these Avicel PH
`
`products, with perhaps the exception of Avicel PH-102, suffer from poor flow
`
`properties. Further, and with the exception of Avicel PH-103, their water
`
`content is too high for making tablets containing water-sensitive drugs.
`
`To remedy these drawbacks, two new Avicel PH grades were launched
`
`these last years :
`
`-
`
`Avicel PH-112, having the typical particle size specifications of Avicel PH-
`
`102 (100 pm) but very low moisture content (max. 1.5%) which makes it an
`
`ideal excipient for moisture-sensitive substances (for this reason Avicel PH-
`
`112 was first named Avicel PH-102 SLM, Special Low Moisture),
`
`-
`
`Avicel PH-200 having a larger average particle size (200 um) than Avicel
`
`PH-102, but a similar moisture content, was designed for improved flow.
`
`The aim of this work was to examine the properties of these new
`
`excipients (moisture content, morphology, packing, flowability) of importance
`
`

`

`646
`
`DOELKER ET AL.
`
`for tableting and to relate these properties to the tableting performance of
`these new excipients (compactibility, susceptibility to hydrophobic lubricants
`and tablet weight variation). As MCC is known to possess disintegrating
`properties, the disintegration time of the tablets was also determined. The four
`conventional Avicel PH products were included in the study to judge the
`significance of the differences observed with the two new grades.
`
`MATERIALS
`
`The six Avicel PH grades were used as received from the supplier (FMC
`Corp., Brussels, Belgium) : Avicel PH-101 (lot No 1715), Avicel PH-102 (lot No
`7236), Avicel PH-103 (Lot No 8036), Avicel PH-105 (lot No 5005), Avicel PH-
`112 (lot No 91 11) and Avicel PH-200 (lot No X129). Magnesium stearate was
`purchased from Siegfried (Zofingen, Switzerland).
`
`METHODS
`
`Basic powder characteristics
`Moisture content was checked by two methods. The loss on drying was
`determined according to the NF monograph on microcrystalline cellulose (3 hrs
`at 105°C) and the water content was measured by the Karl Fischer method
`(Metrohm type 633, Herisau, Switzerland) using the one-component Hydranal
`Composite 5 reagent (Riedel-de-Hahn, Hannover, Germany).
`Size parameters were assessed using sieving, optical microscopy and
`scanning electron microscopy. Particle size distribution was first checked in
`duplicate using air-jet sieving (Alpine 200, Augsburg, Germany). The weight
`geometric mean diameter, dgw, and geometric standard deviation, og, were
`obtained from log-normal plots, and the arithmetic volume-surface mean
`diameter, dvs, was calculated using the appropriate Hatch-Choate equation to
`facilitate comparison with published data.
`
`646
`
`DOELKER ET AL.
`
`for tableting and to relate these properties to the tableting performance of
`
`these new excipients (compactibility, susceptibility to hydrophobic lubricants
`
`and tablet weight variation). As MCC is known to possess disintegrating
`
`properties, the disintegration time of the tablets was also determined. The four
`
`conventional Avicel PH products were included in the study to judge the
`
`significance of the differences observed with the two new grades.
`
`MATERIALS
`
`The six Avicel PH grades were used as received from the supplier (FMC
`
`Corp., Brussels, Belgium) : Avicel PH-101 (lot N° 1715), Avicel PH-102 (lot N°
`
`7236), Avicel PH-103 (Lot N° 8036), Avicel PH-105 (lot N° 5005), Avice! PH-
`
`112 (lot N° 9111) and Avicel PH-200 (lot N° X129). Magnesium stearate was
`
`purchased from Siegfried (Zofingen, Switzerland).
`
`METHODS
`
`Basic powder characteristics
`
`Moisture content was checked by two methods. The loss on drying was
`
`determined according to the NF monograph on microcrystalline cellulose (3 hrs
`
`at 105°C) and the water content was measured by the Karl Fischer method
`
`(Metrohm type 633, Herisau, Switzerland) using the one-component Hydranal
`
`Composite 5 reagent (Riedel-de-Hahn, Hannover, Germany).
`
`Size parameters were assessed using sieving, optical microscopy and
`
`scanning electron microscopy. Particle size distribution was first checked in
`
`duplicate using air-jet sieving (Alpine 200, Augsburg, Germany). The weight
`
`geometric mean diameter, daw: and geometric standard deviation, og, were
`
`obtained from log-normal plots, and the arithmetic volume-surface mean
`
`diameter, dq, was calculated using the appropriate Hatch-Choate equation to
`
`facilitate comparison with published data.
`
`

`

`PROPERTIES OF NEW AVICEL TYPES
`
`647
`
`Both size and shape of the particles were determined using a Wild M3Z
`macroscope (Heerbrugg, Switzerland) coupled to a Microscale TC image
`analyzer (Digithurst, Royston, U.K.). Four parameters were calculated :
`
`-
`
`-
`
`-
`
`-
`
`the average projected area diameter, dp
`
`the elongation ratio, defined as the quotient of the maximum diameter to the
`minimum diameter (7)
`
`the circularity, K, defined as (8) : K = 4 rt area / (perimeter)2
`
`the circularity, C, defined as (9) : C = 4 area / (IT d2max)
`
`(Eq. 1)
`
`(Eq. 2)
`
`where dmax is the maximum diameter.
`
`Scanning electron micrographs of the powders coated with gold were
`taken at a magnification of 600 X with a Jeol JSM-6400 apparatus (Tokyo,
`Japan) using an accelerating voltage of 15 kV.
`The true density was measured with the model 930 air comparaison
`pycnometer Beckman (Fullerton, USA). Determination of the bulk density and
`tap density was carried out in a 25-ml graduated cylinder using 6.0 g of
`material. The bulk density was calculated from the volume of powder after
`turning over the cylinder and the tap density was obtained from the volume
`after 100 tamps in the JEL volumenometer, model STAV 2003 (Engelsmann,
`Ludwigshafen, Germany). The Hausner ratio (1 0) was calculated from the
`quotient of tap to bulk density.
`As only Avicel PH-200 was a free-flowing material, a vibratory hopper
`technique was used. A DIN 12445 glass funnel with an efflux tube of 8 mm
`internal diameter and 10 cm length was fixed to a vibratory sieving apparatus
`(Fritsch Analysette, Idar-Oberstein, Germany).
`When set on intensity 6, a regular flow was observed for all Avicel
`grades except Avicel PH-105 which did not pass through the orifice. The flow
`rate was calculated from the time necessary for 50 g to pass through the
`funnel. Packing and flow properties were determined at least in triplicate.
`
`PROPERTIES OF NEW AVICEL TYPES
`
`647
`
`Both size and shape of the particles were determined using a Wild M32
`
`macroscope (Heerbrugg, Switzerland) coupled to a MicroScale TC image
`
`analyzer (Digithurst, Royston, U.K.). Four parameters were calculated :
`
`-
`
`-
`
`-
`
`-
`
`the average projected area diameter, dp
`
`the elongation ratio, defined as the quotient of the maximum diameter to the
`minimum diameter (7)
`
`the circularity, K, defined as (8) :
`
`K=4 narea/ (perimeter)2
`
`(Eq. 1)
`
`the circularity, C, defined as (9): C=4area/(n d?max)
`
`(Eq. 2)
`
`where dy gx
`
`is the maximum diameter.
`
`Scanning electron micrographs of the powders coated with gold were
`
`taken at a magnification of 600 X with a Jeol JSM-6400 apparatus (Tokyo,
`
`Japan) using an accelerating voltage of 15 kV.
`
`The true density was measured with the model 930 air comparaison
`
`pycnometer Beckman (Fullerton, USA). Determination of the bulk density and
`
`tap density was carried out in a 25-ml graduated cylinder using 6.0 g of
`
`material. The bulk density was calculated from the volume of powder after
`
`turning over the cylinder and the tap density was obtained from the volume
`
`after 100 tamps in the JEL volumenometer, model STAV 2003 (Engelsmann,
`
`Ludwigshafen, Germany). The Hausner ratio (10) was calculated from the
`
`quotient of tap to bulk density.
`
`As only Avicel PH-200 was a free-flowing material, a vibratory hopper
`
`technique was used. A DIN 12445 glass funnel with an efflux tube of
`
`8 mm
`
`internal diameter and 10 cm length was fixed to a vibratory sieving apparatus
`
`(Fritsch Analysette, Idar-Oberstein, Germany).
`
`When set on intensity 6, a regular flow was observed for all Avicel
`
`grades except Avicel PH-105 which did not pass through the orifice. The flow
`
`rate was calculated from the time necessary for 50 g to pass through the
`
`funnel. Packing and flow properties were determined at least in triplicate.
`
`

`

`648
`
`DOELKER ET AL.
`
`Tableting and tablet characteristics
`The study was carried out in two parts. Firstly, the compactibility of the
`powders and their sensitivity to a model hydrophobic lubricant were examined
`by preparing flat 1 2 m m diameter compacts with a hydraulic press (Specac,
`Sidcup, U.K.). A quantity of powder corresponding to a 2-mm thick compact at
`zero theoretical porosity was manually filled into the unlubricated die and
`compressed at 100 MPa during 10 seconds.
`The same operation was repeated with the powders mixed with 0.5%
`magnesium stearate for 5 min (Turbula type T2A, W.A. Bachofen, Basle,
`Switzerland) at 25 r.p.m.
`The diametral crushing force of those compacts was measured with a
`universal testing machine (Schenck-Trebel, type RM 50, Darmstadt, Germany)
`at a strain rate of 3 mmhnin. The result are the mean values of 10 compacts.
`As index of the lubricant sensitivity a strength reduction ratio was calculated,
`defined as the quotient of the compact tensile strength with and without the
`magnesium stearate addition (1 1).
`The second part of the study consisted in preparing flat bevel-edged 8
`mm diameter tablets with a breakline on a high speed 27-station rotary
`tableting machine (Manesty Unipress, U.K.), equiped with a forced feeding
`system and a compaction force monifor, and operating at 1400 tablets/min
`(12). A weight of 160 rng and a diametral crushing force of 70 N were targeted
`and the force necessary to prepare the tablets was estimated with the
`compaction force monitor.
`The mean weight and the coefficient of weight variation of the tablets
`were determined by weighing 30 tablets (Mettler AE163, Greifensee,
`Switzerland), The same tablets were also characterized for their uniformity of
`thickness and diameter. Porosity of the tablets was calculated from their
`dimensions at ejection and from the true density of the materials.
`The mean diametral crushing force and its coefficient of variation were
`measured on 10 tablets using the Heberlein hardness tester (Schleuniger,
`Zurich, Switzerland).
`The friability was determined from the weight loss of 20 tablets trumbled
`100 revolutions in a TAB Erweka friabilator (Heusenstamm, Germany). As a
`
`648
`
`DOELKER ET AL.
`
`Tableting and tablet characteristics
`
`The study was carried out in two parts. Firstly, the compactibility of the
`
`powders and their sensitivity to a model hydrophobic lubricant were examined
`
`by preparing flat 12-mm diameter compacts with a hydraulic press (Specac,
`
`Sidcup, U.K.). A quantity of powder corresponding to a 2-mm thick compact at
`
`zero theoretical porosity was manually filled into the unlubricated die and
`
`compressed at 100 MPa during 10 seconds.
`
`The same operation was repeated with the powders mixed with 0.5%
`
`magnesium stearate for 5 min (Turbula type T2A, W.A. Bachofen, Basle,
`
`Switzerland) at 25 r.p.m.
`
`The diametral crushing force of those compacts was measured with a
`
`universal testing machine (Schenck-Trebel, type RM 50, Darmstadt, Germany)
`
`at a strain rate of 3 mm/min. The result are the mean values of 10 compacts.
`
`As index of the lubricant sensitivity a strength reduction ratio was calculated,
`
`defined as the quotient of the compact tensile strength with and without the
`
`magnesium stearate addition (11).
`
`The second part of the study consisted in preparing flat bevel-edged 8
`
`mm diameter tablets with a breakline on a high speed 27-station rotary
`
`tableting machine (Manesty Unipress, U.K.), equiped with a forced feeding
`
`system and a compaction force monifor, and operating at 1400 tablets/min
`
`(12). A weight of 160 mg and a diametral crushing force of 70 N were targeted
`
`and the force necessary to prepare the tablets was estimated with the
`
`compaction force monitor.
`
`The mean weight and the coefficient of weight variation of the tablets
`
`were determined by weighing 30 tablets (Mettler AE163, Greifensee,
`
`Switzerland). The same tablets were also characterized for their uniformity of
`
`thickness and diameter. Porosity of the tablets was calculated from their
`
`dimensions at ejection and from the true density of the materials.
`
`The mean diametral crushing force and its coefficient of variation were
`
`measured on 10 tablets using the Heberlein hardness tester (Schleuniger,
`
`Zirich, Switzerland).
`
`The friability was determined from the weight loss of 20 tablets trumbled
`
`100 revolutions in a TAB Erweka friabilator (Heusenstamm, Germany). As a
`
`

`

`PROPERTIES OF NEW AVICEL TYPES
`
`649
`
`complement, disintegration testing was performed at 37°C in water using the
`EP/USP apparatus (Sotax DT-3, Basle, Switzerland). Six tablets were
`examined with the disks and six others without the disks. The reported
`disintegration times are the times when the sixth tablet of the test
`disintegrated. For all tests a delay of at least 24 hrs was respected before any
`measurement.
`
`RESULTS AND DISCUSS ION
`
`Basic powder characteristics
`The parameters related to the particle size and shape of the six Avicel
`PH grades tested are given in Table 1. Grades are listed by increasing order
`of particle size.
`As expected, big differences in particle size, both by sieving and
`microscopy, were found between the various Avicel PH grades. Some
`discrepancies are also noticeable with published data obtained using the same
`analytical methods but the data obtained are either unreliable or the
`experimental conditions were inadequately described. Thus, values of
`"median" particle size were calculated from air jet sieving for Avicel PH-105
`(17.5 pm), Avicel PH-101 (40 pm) and Avicel PH-102 (83 pm) by Roberts and
`Rowe (5). Excessively high values of weight geometric mean diameter were
`published for the Avicel grades PH-105, PH-103, PH-101 and PH-102 (6) and
`for Avicel PH-112 (181 pm) and Avicel PH-200 (213 pm) (13). This was
`obviously due to the sieving technique used. Median diameters were also
`given by Pande and Shangraw (14) for Avicel PH-103 (79 pm), Avicel PH-102
`(83 pm) and Avicel PH-112 (98 pm) using a Bradley Sonic Sifter apparatus.
`Finally an "average" particle size of 94 him was found for Avicel PH-102 and
`188 pm for Avicel PH-200, when using a Ro-Tap sieve shaker (1 5).
`It is well established that the sieving method much affects the results.
`values as different as 69
`For instance we obtained in another study (1 6) d,,
`and 108 pm for Avicel PH-101,85 and 144 pm for Avicel PH-102 and 118 and
`161 pm for Avicel PH-200 when using respectively air jet sieving (Alpine) and
`
`PROPERTIES OF NEW AVICEL TYPES
`
`649
`
`complement, disintegration testing was performed at 37°C in water using the
`
`EP/USP apparatus (Sotax DT-3, Basle, Switzerland). Six tablets were
`
`examined with the disks and six others without the disks. The reported
`
`disintegration times are the times when the sixth tablet of the test
`
`disintegrated. For all tests a delay of at least 24 hrs was respected before any
`
`measurement.
`
`RESULTS
`
`AND DI
`
`ION
`
`Basic powder characteristics
`
`The parameters related to the particle size and shape of the six Avicel
`
`PH grades tested are given in Table 1. Grades are listed by increasing order
`
`of particle size.
`
`As expected, big differences in particle size, both by sieving and
`
`microscopy, were found between the various Avicel PH grades. Some
`
`discrepancies are also noticeable with published data obtained using the same
`
`analytical methods but the data obtained are either unreliable or the
`
`experimental conditions were inadequately described. Thus, values of
`
`"median" particle size were calculated from air jet sieving for Avicel PH-105
`
`(17.5 um), Avicel PH-101 (40 um) and Avicel PH-102 (83 um) by Roberts and
`
`Rowe (5). Excessively high values of weight geometric mean diameter were
`
`published for the Avicel grades PH-105, PH-103, PH-101 and PH-102 (6) and
`
`for Avicel PH-112 (181 um) and Avicel PH-200 (213 um) (13). This was
`
`obviously due to the sieving technique used. Median diameters were also
`
`given by Pande and Shangraw (14) for Avicel PH-103 (79 um), Avicel PH-102
`
`(83 um) and Avicel PH-112 (98 um) using a Bradley Sonic Sifter apparatus.
`
`Finally an "average" particle size of 94 ym was found for Avicel PH-102 and
`
`188 um for Avicel PH-200, when using a Ro-Tap sieve shaker (15).
`
`It is well established that the sieving method much affects the results.
`
`For instance we obtained in another study (16) dg values as different as 69
`and 108 um for Avicel PH-101, 85 and 144 um for Avicel PH-102 and 118 and
`
`161 pum for Avicel PH-200 when using respectively air jet sieving (Alpine) and
`
`

`

`650
`
`DOELKER ET AL.
`
`TABLE 1
`Micromeritic Data of the Various Avicel PH Grades
`
`1
`
`Material
`
`S i e v i n g
`
`I
`
`M i c r o s c o p y
`
`dgw
`ag
`(v)
`
`37
`
`1.25 36
`
`ratio
`
`- 1
`
`d,,
`dp
`dma, Elongation Circularity Circularity
`(w) (rim) him)
`- I
`- 1
`
`K
`
`- 1
`
`C
`
`- 1
`
`Avicel PH-105
`Avicel PH-103
`Avicel PH-101
`Avicel PH-102
`Avicel PH-112
`Avicel PH-200
`
`1.46 60
`64
`1.61 54
`60
`1.63 77
`87
`1.59 77
`85
`145 1.95 116
`
`25
`41
`42
`30
`65
`
`40
`7
`66
`48
`96
`
`1.94
`2.39
`2.07
`1.90
`2.02
`
`0.73
`0.74
`0.70
`0.77
`0.80
`
`0.74
`0.55
`0.70
`0.73
`0.60
`
`~
`
`~~~~~~~~~~~
`
`I Could not be determined because the powder could not be properly disaggregated.
`
`~
`
`~~
`
`vibratory sieving (Fritsch). In the authors' opinion, only carefully selected
`sieving methods (such as air jet sieving or possibly Ro-Tap) can provide
`"clean" size fractions and consequently reliable results. When examining dgw
`or dvs values pesented in Table 1, only Avicel PH-105 and Avicel PH-200
`differ significantly from the other grades. Additionnaly, Avicel PH-200 shows
`the widest size distribution of all samples, as evidenced by the highest og
`value calculated.
`Optical microscopy gives size parameters related to the projected
`diameter which of course differs from the equivalent sieve diameter (Table 1).
`It also provides information on the shape of the particles. In this respect, all
`grades have close elongation ratios and circularity values, with Avicel PH-101
`having slightly the highest asymmetry as demonstrated by the SEM pictures
`shown on Figure 1. Incidentally, note that the circularity as defined by Eq. 2
`seems more discriminating than when calculated using Eq. 1.
`
`650
`
`DOELKER ET AL.
`
`TABLE 1
`
`Micromeritic Data of the Various Avicel PH Grades
`
`Material
`
`Microscopy
`
`dp
`
` dmax
`
`Elongation
`
`Circularity
`
`Circularity
`
`(um)
`
`(um)
`
`ratio
`
`K
`
`
`
`Avicel PH-105
`
`Avicel PH-103
`
`Avicel PH-101
`
`Avicel PH-102
`
`Avicel PH-112
`
`Avicel PH-200
`
`dd
`
`
`
`T Could not be determined because the powder could not be properly disaggregated.
`
`vibratory sieving (Fritsch). In the authors’ opinion, only carefully selected
`
`sieving methods (such as air jet sieving or possibly Ro-Tap) can provide
`
`“clean” size fractions and consequently reliable results. When examining dgw
`
`or dy values pesented in Table 1, only Avicel PH-105 and Avicel PH-200
`
`differ significantly from the other grades. Additionnaly, Avicel PH-200 shows
`
`the widest size distribution of all samples, as evidenced by the highest Sg
`
`value calculated.
`
`Optical microscopy gives size parameters related to the projected
`
`diameter which of course differs from the equivalent sieve diameter (Table 1).
`
`it also provides information on the shape of the particles. In this respect, all
`
`grades have close elongation ratios and circularity values, with Avicel PH-101
`
`having slightly the highest asymmetry as demonstrated by the SEM pictures
`
`shown on Figure 1. Incidentally, note that the circularity as defined by Eq. 2
`
`seems more discriminating than when calculated using Eq. 1.
`
`
`
`

`

`PROPERTIES OF NEW AVICEL TYPES
`
`65 I
`
`FIGURE 1
`Scanning electron micrographs of the Avicel PH grades
`
`Table 2 presents the moisture contents, densities and flow rates of the
`various powder samples. Loss on drying values comply with the NF
`requirement (max. 5%), but the two "low-moisture'' grades were beyond the
`manufacturer's specifications, namely Avicel PH-103 (LOD > 3%) and Avicel
`PH-112 (LOD > 1.5%). Karl Fischer data confirmed the loss on drying values,
`suggesting that all bound and free water was probably measured. Note that
`the values quoted for these two materials by the two batch certificates of
`analysis comply with the manufacturer's specifications. This discrepancy,
`which is due to the hygroscopicity of MCC, certainly comes from the fact that
`the moisture content was measured at the time of compression and not
`directly at the opening of the container.
`True densities are in agreement with published values and are quite
`close. In contrast, packing and flow properties differ significantly among
`products. Quite logically, the order of flow rates inversely rank that of Hausner
`ratios, with Avicel PH-105 and Avicel PH-200 showing extreme values.
`
`PROPERTIES OF NEW AVICEL TYPES
`
`651
`
`
`
`"w,
`
`Avicel PH-101
`
`Bi
`
`Re TI
`
`e.
`
`oy
`
`Avicel PH-105
`
`—
`100 um
`
`Avicel PH-102
`
`Avicel PH-112
`
`Avicel PH-200
`
`Scanning electron micrographs of the Avicel PH grades
`
`FIGURE 1
`
`Table 2 presents the moisture contents, densities and flow rates of the
`
`various powder samples. Loss on drying values comply with the NF
`
`requirement (max. 5%), but the two "low-moisture” grades were beyond the
`
`manufacturer's specifications, namely Avicel PH-103 (LOD > 3%) and Avicel
`
`PH-112 (LOD > 1.5%). Karl Fischer data confirmed the loss on drying values,
`
`suggesting that all bound and free water was probably measured. Note that
`
`the values quoted for these two materials by the two batch certificates of
`
`analysis comply with the manufacturer's specifications. This discrepancy,
`
`which is due to the hygroscopicity of MCC, certainly comes from the fact that
`
`the moisture content was measured at the time of compression and not
`
`directly at the opening of the container.
`
`True densities are in agreement with published values and are quite
`
`close. In contrast, packing and flow properties differ significantly among
`
`products. Quite logically, the order of flow rates inversely rank that of Hausner
`
`ratios, with Avicel PH-105 and Avicel PH-200 showing extreme values.
`
`
`
`

`

`652
`
`DOELKER ET AL.
`
`TABLE 2
`Moisture Contents, Densities and Flow Rates of the Various Avicel PH grades
`
`Material