`
`RESEARCH PAPER
`
`Effectiveness of Binders in Wet
`Granulation: A Comparison Using Model
`Formulations of Different Tabletability
`
`D. Becker, 1 T. Rigassi, 1 and A. Bauer-Brandl2
`
`1 Novartis, Inc., Building K-40 1.2.65, CH-4002 Basel, Switzerland
`2Universitat Freiburg, Pharmazeutische Technologie, Hermann-Herder-Str. 9,
`D-79104 Freiburg, Germany
`
`ABSTRACT
`
`Based on an analysis of model granulates and tablets, a comparison was made
`of the effectiveness of the binders PVP K30 PH, Cellulose HP-M 603, Lycatab
`DSH, Lycatab PGS, and L-HPC (type LH 11). A high shear mixer was used to
`prepare two model granulates (placebo and paracetamol) under processing con(cid:173)
`ditions which were, as far as possible, comparable. The binders were added as
`proportions of2%, 6%, and 10%. Water was used as the granulating liquid. The
`properties of the placebo granulates (particle size distribution, bulk and tapped
`density, granule strength, flow properties), and those of the tablets (crushing
`strength, friability) prepared from these granulates under different compaction
`forces, were generally good. However, with PVP, Cellulose HP-M603, and
`Lycatab, the disintegration time of the tablets did not meet pharmacopoeial require(cid:173)
`ments even though a "disintegrant" was used in the "outer phase. " The
`paracetamol formulations were prime examples of high-dose drug substances with
`particularly poor granulating and tabletting properties, well suited to reveal dif(cid:173)
`ferences between the binders. The paracetamol granulates were of higher friabil(cid:173)
`ity and less flowability than the placebo granulates. The tablets tended to cap, fri(cid:173)
`ability was (with few exceptions) high, and disintegration times were long. In the
`preparation of model tablets containing paracetamol, PVP K30 PH (6%), and
`Cellulose HP-M 603 (6%) turn out to be the binders of choice with respect to
`crnshing strength, but the disintegration times are too long. Lycatab PGS, Lycatab
`DSH, and L-HPC-LH 11 could not be used to produce paracetamol tablets that
`met the requirements.
`
`Copyright~ 1997 by Marcel Dekker, Inc.
`
`791
`
`I
`
`:
`
`I
`
`Par Pharm, Inc.
`Exhibit 1020
`Page 001
`
`
`
`792
`
`Becker, Rigassl, and Bauer·Bra.ndl
`
`Effecti·
`
`An assessment method involving calculation of averages for all granulates is used
`to evaluate the effectiveness of the binders.
`Key WordS: Wet granulation; High shear mixer; Binder; Tablet,· Hydroxypropyl(cid:173)
`methylcellutose; Polyvinylpyrrolidone; Lycatab; L-HPC.
`
`INTRODUCTION
`
`The properties of wet granulates, and of the tablets
`into which they are processed, are decisively influenced
`by binders. Not only are tlle type and amount of binder
`important, but also the processing procedure, e.g. the
`initial and then thorough wetting of the tablet mass (1).
`A standard method for wet granulation in a high shear
`mixer involves the dry addition of binder. followed by
`mixing, and then the addition of water. In this method
`a good correlation was found between granulate particle
`size and binder concentration (2), and in addition, this
`method does not require the preparation of a binder
`solvent.
`The aim of the present study is to compare the effec(cid:173)
`tiveness of different binders when used for wet granu(cid:173)
`lation in a high shear mixer.
`Commercial formulations of the binders polyvi(cid:173)
`nylpyrrolidone (PVP K30 PH) and hydroxypropyl(cid:173)
`methylcellutose (Cellulose HP-M 603) are widely used
`(3) and serve here as a reference. Lycatab rosn~ (a pre(cid:173)
`gelatinised maize starch), Lycatab DSHT>'I (a malto(cid:173)
`dextrin). and L-HPC, rype LH 11"" (a low-substituted
`hydroxypropy1cellulose) are used less frequently or are
`new.
`Two models, a placebo formulation and a drug for(cid:173)
`mulation, were assessed. The latter was given a very
`high content of paracetamol so that its tableting proper(cid:173)
`ties would be particularly unfavorable. The influence of
`different types and amounts of binders both on granu(cid:173)
`late properties (particle size distribution, bulk, and
`tapped density, granule strength, flow properties) and on
`tablet quality (crushing strength, friability, disintegration
`time) was investigated.
`The degree to which wetting affects the particle size
`of the agglomerates depends to a large extent on the ad~
`hesion properties between binder and powder (4,5).
`Powder wettability is particularly important for binder
`distribution in the granules (6) and for tlle mechanical
`properties of the tablets (7). For this reason, it was not
`possible in the present study to go to the literature for
`data on the reference substances. However, since par(cid:173)
`ticle enlargement is for the most part unrelated to the the
`
`particular machines used (4,8), these observations can
`be applied to other manufacturing situations.
`
`MATERIALS AND MEmODS
`Preparation of the Granulates
`
`The raw materials used in the preparation of the
`granulates are listed in Table 1. Of the 1800 g in each
`granulate. 89.3% of the final mixture was the "inner"
`phase and 10.7% was the "outer" phase. Table 2 sum(cid:173)
`marizes the compositions of the formulations.
`The manufacturing steps for tlle granulates are listed
`in Table 3.
`
`Properties of the Granulates
`
`Panicle size distribution of the granulates was deter(cid:173)
`mined twice in each case, on 50 g portions of granulate,
`using a laboratory VE 1000/s sieving machine (Kurt
`Retsch GmbH & Co. KG, 42781 Haan, Germany) set
`to run for 10 minutes at an amplitude of 1. 5 mm. The
`stack consisted of analytical-grade screens conforming
`to DIN/ISO 3310/1, with mesh sizes of 1000, 710, 630,
`500, 315, 250, 200, and 100 J.lill.
`The bulk and tapped density of the granulates were
`assessed in accordance with the Germany pbarmaco·
`poeia (DAB 1966) using a JEL tamped volume measur~
`ing apparatus (STAY 2003, J. Engelsmann AG, 6700
`Ludwigshafun, Germany). V250 is the result reported.
`The granule strength was determined by testing fri(cid:173)
`ability using the "Roche" oscillating friability testing
`machine. The testing drum, equipped with two steel
`rollers, alternately rolls 50 times to the left and right,
`rotating 170° each time. 10 g of granulate from the
`250-800 1-lffi sieve fraction was used for the test of gran(cid:173)
`ule strength. After the drum movement stopped, the
`granulate was sieved for 2 min through a 250 Jlill sieve,
`with an air throughput of (48-58) m31hr. using the AJ~
`pine 200 LS air-jet sieving machine (Alpine AG, 8900
`Augsburg, Germany). and the residue remaining on the
`sieve was weighed. The granule strength was calcu1ated
`using the following fonnula:
`
`grar
`
`The
`usin
`Alb
`grru
`diff
`rep
`(9),
`wh
`
`Pn
`
`(EJ
`
`Par Pharm, Inc.
`Exhibit 1020
`Page 002
`
`
`
`Effectiveness of Binders in Wet Granulation
`
`793
`
`Name
`
`Lycatah PGS
`Lycatah DSH
`L-HPC Typ: LH·ll
`Cellulose HP-M 603
`PVP K 30 PH
`Lactose, ground
`A vice! PH 102
`PVPXL
`Aerosil 200
`Magnesium stearate
`Paracetamol
`
`Table 1
`
`Materials
`
`Manufacturer
`
`Roquette Freres, F-Lme
`Roquette Freres, F-Lille
`ShinEtsu Chemicals, J-Tokio
`DOW Chemical USA, Midland, MI, USA
`T.S.P., Guildford, U.K.
`De Melkindustrie Holland, NL-Veghel
`FMC, Philadelphia, PA, USA
`I.S,P., Guildford, U.K.
`CABOT Corp., Tuscola, IL, USA
`FACT Italien ??
`Hou Zhou Syn. Phann. Fact.
`
`Batch
`
`E8213
`E5810
`501019
`JJ15012N23
`TX51028
`024448
`Y541
`S50529
`
`MGS-30159
`9512082(M)
`9512105
`
`granule strength
`
`fl.nal weight of sieve residue* 100
`weight of sample
`
`The flow properties of the granulates were assessed
`usjng a Flowtester FT 300 from Sotax (Sotax AG, 4123
`Allschwil, Switzerland). A single sample of 350 g of
`granulate was used for each of the 6 measurements (with
`differing funnel vibrations). The flow angle quotient is
`reported in each case as the result. According to Sotax
`(9), values in excess of 0.8 indicate that flow is good,
`while those below 0.6 indicate that it is poor,
`
`Pressing Into Tablets
`
`An EKO laboratory model eccentric tablet press
`(Emil Korsch, Berlin, Germany) was used to press 400
`
`mg tablets, 10 mm in diameter and with bevelled edges,
`at a rate of 52 tablets per min.
`The compaction forces and tolerances used in the
`preparation of batches of 100 tablets were: (5.0 ± 0.25)
`k.t'f, {7.5 ± 0.35) kN, (10.0 ± 0.50) kN, (12.5 ±
`0.60) kN, (15.0 ± 0.75) kN, (17.5 ± 0.90) kN, (20.0
`± 1.20) kN, (25.0 ± 1.80) kN, and (35.0 ± 2.00) kN.
`
`Tablet Properties
`
`Tablet friability was determined by placing 20 tablets
`each rime into a "Roche" friability testing machine and
`then setting the machine fur 500 revolutions. The frlabil~
`ity of the tablets was calculated using the following
`formula:
`
`Table 2
`
`Composition of Finished Blends
`
`Material
`
`1 paracetrunol
`2 lactose/Avicel 2.45 : 1
`4 binder
`
`6 Avicel
`7 PVP XL
`8 Aerosil 200
`9 magnesium stearate
`
`Placebo
`Proportion [M/M]
`
`l?aracetrunol
`Proportion [M/M]
`
`0%
`ad 100%
`0%, 2.0%,
`6.0%. 10%
`5.0%
`5.0%
`0.2%
`0.5%
`
`75%
`ad 100%
`0%, 2.0%,
`6.0%, 10%
`5.0%
`5.0%
`0.2%
`0.5%
`
`Inner phase
`
`Outer phase
`
`Par Pharm, Inc.
`Exhibit 1020
`Page 003
`
`
`
`794
`
`Processing Step
`
`1. dry mixing
`
`Becker, Rigassi, and Bauer-Brand!
`
`Ef
`
`Table 3
`
`Preparmion of the "Inner" Phase
`
`Machines
`
`Process Parameter
`
`high shear mixer Diosna PlO'"
`
`2 min impeller; 167 U min-1
`chopper: 3000 U mln·1
`30 sec; then scaping, ru:ldition of
`water, kneading impeller: 167 U min-•
`chopper: 3000 U min-1
`
`60°C 25-50 min as required
`lO g samples, 30 min. l05"C
`oszillating mode, screen: 1.25 mm
`mesh size; diamerer of wire 0.8 mm
`42 U min· 1 in 3 llidded drum; outer
`phase added via 0.8 mm manual
`screen, blending .for lO min.; then
`magnesium stearate via 0.8 mrn
`manual screen, blending for 5 min
`
`Bi
`
`L:
`
`L
`
`n
`51
`v:
`S(
`
`2. wetting and kneading
`
`high shear mixer Diosna .P 10"
`
`3. deagglomeration
`4, drying
`5. moisture content
`6. dry sieving
`
`7. finished blend
`
`3 rnrn manual screen
`fluidized bed dryer Strea 1 b
`Mettler infrared dryer LP 16•
`classifying screening machine
`Prewitt MGLd
`Turbula blender TlOB"
`
`"Dierks & Sohne, D-Osnabri.lck.
`bAeromatik AG, CH-Bubendorf.
`•Mettler Instrumente AG, CH-8606 Niinikon-Uster.
`dFrewitt AG, CH-Fribourg.
`OW. A. Bachofen Maschinenfabrik, CH Basel.
`
`friability
`
`weight of. sample - final weight x
`wetght of sample
`
`100
`
`The crushing strength of 1 0 tablets from each lot was
`determined using a Schleuniger 6 D tablet tester (Dr.
`Schleuniger & Co .• 4501, Solothum, Switzerland).
`The disintegration time for 6 tablets in each case was
`tested in accordance with DAB 1996 using the DT 3
`testing apparatus (Sotax AG, 4123 Allschwil, Switzer(cid:173)
`land).
`
`RESULTS AND DISCUSSION
`
`Preparation of the Granulates
`
`The time required for kneading and drying the pla(cid:173)
`cebo granulates is summarized in Table 4. The para(cid:173)
`cetamol granulates required markedly less granulating
`fluid than did the placebo granulate. The kneading and
`drying times are not directly comparable.
`A higher content of binder would be expected to
`accelerate formation of the granulate. thereby necessi(cid:173)
`tating shorter kneading times, and this is precisely when
`happens during granulation in the high shear mixer with
`most of the binders tested (Table 4), the only exception
`
`being L-HPC. L-HPC presumably differs in this regard
`due to the high swelling capacity (10) which causes it
`to absorb a large amount of water. thereby delaying the
`wetting of the particle surfaces of other substances. In
`addition, the particles of binder increase in volume as
`swelling progresses, physically separating the particles
`to be bound.
`Tbe drying time in the fluid-bed dryer also depends
`on the amount of binder used (Figure 1). In the case of
`PVP, HP-M, and Lycatab DSH, the higher the binder
`concentration, the shorter the drying time. On the other
`hand, the drying time remains constant or even increases
`slight1y when the amount of Lycatab PGS or L-HPC is
`increased. In both of these cases, this is also due to the
`large water-absorbing capacity of the binders, which
`precludes a rapid drying time. As a kinetic factor, dry(cid:173)
`ing time is nevertheless also highly dependent on par~
`ticle size distribution and other particle properties (po(cid:173)
`rosity, particle shape, and surface properties).
`
`Properties of the Granulates
`
`Particle Size Distribution
`
`Figure 2 (below) shows the partide size distribution
`of the placebo granulates. The values reported for total
`
`Par Pharm, Inc.
`Exhibit 1020
`Page 004
`
`
`
`Effectiveness of Binders in Wet Granulation
`
`795
`
`Table 4
`
`Kneading and Drying Times of Granulates
`
`Proportion
`of Binder
`
`Kneading
`Time
`(s}
`
`Placebo
`Drying
`Time
`(min)
`
`Water
`Content
`(%)
`
`Paracetamol
`Amount of Kneading Water
`Contem
`Water
`Time
`(%)
`(%)
`(s)
`
`Drying
`Time
`(min)
`
`2%
`6%
`10%
`2%
`6%
`10%
`2%
`6%
`10%
`2%
`6%
`10%
`2%
`6%
`10%
`
`240
`240
`90
`30
`180
`90
`60
`240
`120
`90
`240
`90
`30
`170
`150
`180
`
`49.0
`37.0
`34.5
`24.0
`32.0
`25.0
`25.0
`38.0
`37.0
`39.0
`45.0
`42.0
`40.0
`44.0
`40.0
`42.0
`
`2.5
`2.5
`2.5
`3,1
`2.0
`1.8
`2.6
`2.1
`2.5
`3.2
`2.4
`2.5
`2.8
`2-1
`2.9
`3.2
`
`29.5
`26.7
`24.4
`11.1
`18.1
`16.7
`16.7
`18.1
`18.1
`18.1
`13.9
`13.9
`13.9
`16.7
`16.7
`16.7
`
`30
`30
`20
`210
`360
`150
`90
`150
`150
`150
`120
`120
`120
`90
`90
`90
`
`0.3
`0.5
`0.9
`1.6
`0.4
`0.6
`0.5
`0.5
`0.8
`1.4
`0.5
`0.8
`0.8
`0.4
`0.7
`0.9
`
`32
`30
`30
`17
`28
`23
`23
`29
`29
`37
`24
`23
`25
`27
`33
`33
`
`Binder
`
`Without binder
`PVP K30 PH
`
`Cellulose HP-M 603
`
`Lycatab PGS
`
`Lycatab DSH
`
`L-HPC TY1Je: LHl L
`
`residues are 16% (oversize particles), the median (R
`50%), and the percentage of fine particles (R 84%). The
`values shown are the mean in each case for sieve analy(cid:173)
`ses carried out in duplicate. They can be compared, in
`
`Figure 3, with the corresponding figures for sieve analy(cid:173)
`sis of the paracetamol granulates.
`A rise in binder concentration would lead one to
`expect an increase in particle size, and a large upturn
`
`I
`'I
`'I I
`II I
`I' I:
`'I l
`I I
`
`I
`
`r :I
`
`I i , I
`
`.,
`.i
`
`'
`I
`
`: r'
`
`:.I
`
`:I
`
`I
`
`'I I ,,
`
`I
`
`50
`
`45
`
`40
`
`35
`
`30
`
`25
`
`0
`
`~~~~
`
`. ~ -•-PVPIOOPH
`-e-RP·M603
`-.&-Lyeatab PGS
`-T- Lycatab OSH
`-+- L-HPC LHil
`
`•
`
`• •
`
`lO
`proportion of binder in%
`
`Figure 1. Drying time of placebo granulates.
`
`·~
`.5:
`. .§
`·&
`
`t}
`
`Qj)
`
`"'='·
`
`Par Pharm, Inc.
`Exhibit 1020
`Page 005
`
`
`
`796
`
`E
`:::1..
`.s ...
`0 -;>
`~ :a
`0
`0
`'f
`!!
`
`1000
`
`800
`
`600
`
`400
`
`200
`
`0
`
`Becker, Rigassi, and Bauer-BrandJ
`
`I R 16%: coarse
`
`R 50%: median
`R 84%: fines
`
`Il
`
`no binder PVP
`2%6% H)%
`
`ros
`HPM
`DSH
`LHII
`2%6%10% 2%6%10% 2%6%10011. 2%6%10%
`
`Figure 2. Particle size distributions of placebo granulates.
`
`in t
`late
`Cel
`tior
`inc·
`ers
`me
`tim
`
`tiel
`inc
`J.liT
`
`re~
`ch;
`gn
`
`Gr
`
`ste
`
`R 16%: coarse
`
`I R 5Wio: median
`
`R 84%: fines
`
`:I.
`
`E 1000
`.5 ..
`* 800
`E
`~ :a
`0) u
`'f
`«<
`Q,.
`
`600
`
`400
`
`200
`
`0
`
`I I n
`
`DSH
`HPM
`PGS
`LHil
`2%6% 10% 2%6% !0% 2%6% lO% 2%6% 10%
`
`no binder
`
`PVP
`2%6% ID%
`
`Figure 3. Particle size distributions of paracetamol granulates.
`
`Par Pharm, Inc.
`Exhibit 1020
`Page 006
`
`
`
`Effectiveness of Binders in Wet Granulation
`
`797
`
`in the mean particle size (R 50%) of the placebo granu(cid:173)
`late can in fact be seen with the binders PVP K30 PH,
`Cetlulose HP-M 603. and Lycatab DSH at concentra~
`tions of up to 6% (Figure 2). Particle size cannot be
`increased further when the concentration of these bind(cid:173)
`ers is raised (to 10%), which suggests that the binding
`mechanism changes above a certain (',ritical concentra(cid:173)
`tion limit.
`Use of L-HPC leads to a reduction in granulate par(cid:173)
`ticle size. This tendency continues as concentrations
`increase due to the small mean particle size (about 50
`~J,ffi (11)) and poor binding properties of L-HPC.
`The particles of paracetamol granulate are more ir(cid:173)
`regular in size (Figure 3) and more oversized, but show
`characteristics similar to those noted for the placebo
`granulate.
`
`transport because powdery. friable granulate has a det(cid:173)
`rimental effect on flow properties and can cause
`demixing. For both types of granulate. and for all binder
`concentrations tested, the highest granule strengths were
`achieved using the tried and trusted binders PVP K30
`PH and Cellulose HP-M 603 (Figure 4).
`The binders Lycatab PGS and Lycatab DSH showed
`different effects in the placebo and paracetamol formu(cid:173)
`lations. The granule strength of the paracetamol granu(cid:173)
`late remained constant for all 3 concentrations of Lyca(cid:173)
`tab PGS but rose in the placebo granulate as
`concentrations of Lycatab PGS increased. However,
`even at a binder concentration of 10%, granule strength
`was lower than in granulate prepared without a binder.
`The maximum granule strength for paracetamol granu(cid:173)
`late is achieved with Lycat.ab DSH at a concentration of
`10% . Granule strength is very dependent on the binder
`Granule Strength
`used. However, the granule strength of the placebo
`granulate is largely unaffected by the concentration of
`Strong granulates are advantageous for subsequent
`Lycatab DSH.
`steps in the production process. such as final mixing and
`!OOr---------------------------------------------------------------------
`
`.,.
`.,.
`JOL---~--------------------------~--~--~--------------~------------~
`"#
`'it
`'it
`"#
`;if,
`'$.
`~
`~
`~
`~
`~
`~
`=:
`l'o
`:};,
`"'
`"'
`~
`~
`"'
`'C
`Q
`>Q
`N
`:t
`i
`~
`~
`a,
`...
`vi
`Q
`c
`...
`~
`Q
`!9
`...
`:r::
`J:>
`"'
`!'!
`>
`"'
`:s
`i'l ...
`~
`
`i
`..J
`
`~
`
`_,
`
`..J
`
`[- .. -
`
`l'arace!amol - - - - Plae.:bol
`
`Figure 4. Strengths of placebo- and paracetamol granulates.
`
`I
`
`I:
`
`. :
`I II
`:I l I
`
`I
`
`: '
`: :
`
`,
`
`I
`
`I i
`
`Par Pharm, Inc.
`Exhibit 1020
`Page 007
`
`
`
`798
`
`Granulate prepared using L-HPC shows low granule
`strength, in many cases even lower than that achieved
`when no binder is used.
`
`Flow Properties
`The flow capacity (expressed as flow angle quotients}
`observed in the analysis of granulate flow are shown in
`Figure 5. The placebo granulates proved to have very
`good flow characteristics with most binders. However,
`use of increasing concentrations of L-HPC leads to an
`increasing inhibition of flow, presumably due to a rise
`in the quantity of fine particles {Figure 2).
`The relationship between flow characteristics and
`binder is even clearer with the paracetamol granulates.
`The flow characteristics of the granulate are poor with(cid:173)
`out binder, but improve when even small amounts (2%)
`of PVP K30 PH or Cellulose HP-M 603 are added. The
`flow characteristics are not affected by further increases
`
`Becker, Rigassi. and Bauer-Brand!
`
`in the concentrations of these binders. Similarly good
`flow characteristics are achieved with Lycatab DSH at
`concentrations above 6%. The optimum binder concen(cid:173)
`tration for Lycatab PGS is 2%. with an further additions
`leading to a rise in the amount of fine particles (Figure
`3) and thus to reduced flow capacity. Paracetamol gran~
`ulates prepared using L-HPC all flow poorly, particu(cid:173)
`larly when the concentration of binder is high (compare
`also particle sizes in Figures 2 and 3).
`
`Bulk and Tapped Densities
`
`Figure 6 shows the bulk and tapped densities for the
`placebo granulates after 1250 taps. Figure 7 shows the
`corresponding results for the paracetamo1 granulates.
`Bulk and tapped densities of both sorts of granulates
`tend to fall as binder concentrations rise (Figures 6 and
`7). This is explained by the fact that, generally speak~
`ing. larger agglomerates form. What is unusual in the
`
`0,&
`
`0,9
`m
`OiJ
`=
`til
`~
`c =
`-
`"S
`= 0,7
`c =
`
`<l.l
`'::.::!
`
`0'
`
`0,6
`
`•--r""' -
`
`(
`
`1
`I
`I
`I
`
`j
`I
`I
`I
`
`I
`I
`I
`I
`
`.. __
`
`-.
`
`\
`\
`\
`
`I
`I
`I
`I
`
`J
`I
`I
`I
`
`\
`\
`\
`
`·-
`
`\
`
`\ • \
`
`\
`\
`
`\
`\
`
`O,S
`....
`" "'='
`:&
`0
`r:::
`
`#.
`\0
`
`<f..
`
`('I
`
`:I:
`P..
`0 ...,
`::.:::
`c.. >
`
`0..
`
`0
`
`('I
`
`0)
`()
`
`<f.. * ~ *
`~
`:E
`rL
`:I:
`
`0
`
`* ~
`
`~
`&
`
`'$.
`N
`
`~ <f..
`
`0
`
`*
`
`\0
`
`:>.l.
`:5
`
`-
`
`<f..
`N
`
`-
`5
`c)
`0..
`j
`
`N
`:I!
`rn
`0
`u}
`ci
`P..
`.0
`.g
`~
`iJ
`j.
`>.
`-l
`J--+- Placebo - .. - Paracetamo! J
`
`·,
`
`Figure 5. Flow properties of granulates (placebo and paracetamol).
`
`Par Pharm, Inc.
`Exhibit 1020
`Page 008
`
`
`
`Effectiveness of Binders in Wet Granulation
`
`<>---o
`
`' '
`'
`
`o---o.
`
`~
`
`'
`
`'\
`
`I
`
`v
`
`'o,
`
`D
`
`I
`
`A
`'\
`'\
`'b..,_
`
`P.
`
`.... "<>'
`
`...-\{
`\
`
`\
`
`J\.
`"
`I
`I
`I
`I
`I 0.
`
`0
`\.
`'\
`\
`
`·o,
`
`(I
`.....
`
`.....
`
`'
`
`'
`
`0
`
`'
`
`J
`I
`.0
`
`\
`'or'
`
`'
`
`0.750
`
`o.700
`
`0.650
`
`01o1Xl
`
`,......
`~
`.._
`~
`'V)
`.,
`,e: 0.550
`
`(1,1
`
`0.500
`
`0.450
`
`799
`
`..
`s
`..,
`~
`1-...
`01
`=
`"'
`"'
`X
`
`l.JOO
`
`1.250
`
`1,200
`
`1,150
`
`uno
`
`CJ
`
`1.050
`
`'>....._
`
`......
`
`.
`0.
`
`0.400 --------------------------------_1 l,OOO
`...
`...
`.,
`#. * ~~
`#
`:,{;
`#.
`;:l!
`=:
`""
`"'
`-
`""
`0
`"'
`" :0
`-
`a
`::I:
`g
`"-
`""
`0 "" ~
`:;;;;
`~
`d.
`"-> "-
`::z::
`
`~
`N
`:i
`vi
`~
`..
`j
`"'
`> •
`...l
`
`N
`
`:X::
`
`::z::
`,..J
`
`"Q
`
`~
`\(>
`
`#
`£
`
`~
`&:,
`
`'#.
`0
`
`~
`~
`
`;i(
`~
`
`'#.
`<"<
`Vl
`0
`0.
`~
`"' ~
`
`...J
`
`'i:i
`v
`• • 0 • · h11lk t.l.;nsity - <> - lappet.l density __..,__ Hausner factor
`
`Figure 6. Bu!k:!tapped densities and Hausner factor of placebo granulates.
`
`case of the L~HPC-paracetamol granulates is that all of
`them, despite increasing looseness as binder concentra(cid:173)
`tions increase, show about the same final tapped density.
`This is probably related to their low granule strength
`(Figure 4),
`The Hausner factor (the quotient of bulk density and
`tapped density) expresses the relative mechanical com(cid:173)
`pression of the granulate (which can occur during trans(cid:173)
`port or as a result of vibrations in the tablet press). It
`thus allows us to make inferences regarding the unifor(cid:173)
`mity of particle size, shape, and crushing strength. With
`the help of the Hausner factor. attempts can be made at
`predicting both the extent of compression, and the flow
`problems which may occur during tabletting. With tbe
`exception of granulates containing high concentrations
`of L-HPC, the granulates tested had a Hausner factor
`
`lower than 1.2, and were thus within a range in which
`no problems were to be expected.
`
`Tablet Properties
`Crushing Strength
`Figures 8 and 9 show profiles for compaction force
`and tablet crushing strength in relation to the binder
`used.
`Even plocefJo granulate without binder could be pro(cid:173)
`cessed into tablets with a good relationship between
`compaction force and crushing strength (Figure 8).
`Above a compaction force of 7.5 kN, an adequate
`crushing strength of 50 N is achieved for tablets with
`the selected diameters. The addition of binders did not
`necessarily lead to an increase in crushing strength. For
`
`'l
`
`I"
`I''
`I
`
`I',
`
`Par Pharm, Inc.
`Exhibit 1020
`Page 009
`
`
`
`800
`
`lHSO
`
`0.700
`
`0,650
`
`0.450
`
`().4()0
`
`Becker, Rigassi, and Bauer-Brand!
`
`!.300
`
`!,250
`
`.. 0
`t;
`..::
`... .,
`"' " "' ::z::
`
`;:
`
`1,200
`
`uso
`
`1,100
`
`0
`
`~.050
`
`"0
`
`. ·o· .. -o .•• -o·
`
`b
`
`_j
`
`..
`.,.
`-:--:--:---:---:---::---:---:--~----:---:-----:-----:------t 1.001)
`,_
`"'
`.,
`.,
`,
`"'
`*
`;!.
`;!.
`'$.
`;[;.
`'$
`'$.
`'$.
`"$
`~
`~
`~
`""
`~
`"'
`""
`"'
`'2
`"'
`;;.:::
`'2
`'0
`,,
`-
`N
`;;.:::
`N
`"'
`:;;;
`:x;
`::t;
`a..
`i
`~
`""
`0
`....
`"'
`VI
`0
`u
`:f
`cl
`..
`~
`;:.,:
`.n
`0...
`,.,
`a..
`::c
`"'
`"-
`>
`~
`"
`::;:;
`"-
`...J
`~·
`v
`
`0
`
`VI
`
`u
`...J
`
`Figure 7. Bu1kltapped densities and Hausner factor of pamcetamol granulates.
`
`example, use of low concentrations of the binders PVP
`K30 PR Cel1ulose HP-M 603, and Lycatab DSH (2%
`and. in the case of PVP K30 PH, 6% as well) led to a
`reduction in crushing strength. Use of Lycatab PGS at
`a concentration of 2% and Lycatab DSH at a concen(cid:173)
`tration of 6% yielded crushing strength comparable to
`that of a granulate lacking a binder (cf. similar results
`in (12}). Only L-HPC at a low concentration improved
`crushing strength (by about 30%) as compared with the
`binder-free placebo formulation. PVP K30 PH does not
`bring about a similar rise in crushing strength until the
`concentration is 10%.
`Cellulose HP-M 603 (6% and 10%). Lycatab DSH
`(10%), and-Lycatab PGS (6%) can be characterised as
`
`differing very little from each other. These binders in(cid:173)
`crease crushing strength by an average of 45% . The
`largest rise in crushing strength. approximately 75%, is
`achieved with a 10% concentration of the binder Lyca(cid:173)
`tab PGS, which is also the concentration recommended
`by the manufacturer for raising crushing strength (13}.
`The paracetamol tablets without binder (Figure 9)
`show poor crushing strength and also are prone to cap(cid:173)
`ping, regardless of the compaction force applied. They
`are improved by the addition of binders (except for 2%
`L-HPC). As expected, differences relating to the type
`and concentration of binder are greater in this problem~
`atic formulation than in placebo. Cellulose HP-M 603
`and PVP K30 PH are best here. Adequate crushing
`
`Par Pharm, Inc.
`Exhibit 1020
`Page 010
`
`
`
`Efii::ctivcne;;s of Binders in Wet Granulation
`
`SOl
`
`350.0
`
`300.0
`
`250Ji
`
`200.0
`
`[
`'§,
`~ .t .,
`"" "' ~ 150.0
`"' "' w
`
`JJ!G.O
`
`50.0
`
`i\0
`
`g
`
`~~
`Q.,
`·.::>
`~ ... ~~
`~
`
`::;,., >
`c:"
`
`35.0
`
`15.0 ft>r~'' !!.:Nl
`
`(/)
`C•
`·' ~
`,D
`
`·~ ti
`
`~~~
`~-=-)
`....:l
`
`',.)"•
`
`""
`·:.::::
`
`,::.
`~--
`N
`:i
`O'l
`0
`A
`·~
`1§
`.~)
`
`"'
`"''
`
`~,.t·;
`
`,.
`~ ~f~
`c
`N
`
`,.
`:§:
`'.:.:::•
`
`,,
`
`~··
`~~
`
`~
`
`~
`....,
`,...-:..
`::;,.,
`~
`,,j
`
`8
`'.0
`~
`rL
`::c
`~
`
`~
`
`~J
`
`Figure 8. Compression t'brce vs. crushing strength profile of placebo mblets.
`
`strength i~ achkved wirh the~e hindJ::r~ even at a concen~
`tratlon as low as 6%. and when tablets are produced
`using a Jow compaction force.
`The hinders Lycatah PGS and Lyz~atab DSH have
`nearly ~:qual binding t.apadty in concentrations up to
`6% .. At higher czmcentrat!oru;, Lycatab PGS shows bet(cid:173)
`ter b indin.g eapadty. The minimum concentratkms of
`either of these binders should not be less than w·;t_, in
`the pamcet;lmol formulation discussed hem. Tt>is reco;n ..
`me:ndadon is based nn the large amount of drug sub(cid:173)
`stance and poor t.;:tbleting properties (capping) of the
`paraceta:mol tablets, With other model formulations
`bastxl on mixtures of lactose and starch, \'-;hicb have
`fewer problematic properties, there is reason to believe
`tha:t an increase in binder cnntem abov<~ the maximum
`
`of 6% investigated here could make possible further
`improvement in crushing strength {14). A binder con"
`c:entnaior: of 5 ~"(, has been described as adequate for
`tablets cont.;:t!ning a filler of anhydrous )actme and <>
`1.0% formulation of drog substance (hydro<.:hlorotbia:tr
`ide) (15).
`Improvement in tablet cnmhing strengtt was cmnpa~
`rabk for L--HPC, at concentrations of 6% and 10%,
`Lycatab PGS (6%), and Lycam.b DSH (6%-:L
`
`friability
`/ul.otber impormnt medmnkal property of mbiets is
`friability (s<~e Figures 1 0 ru:td l I ) _ When regarding the
`figures, keep in mind tllli.t the scale for compaction force
`
`Par Pharm, Inc.
`Exhibit 1020
`Page 011
`
`
`
`Becker, Rigassi, aud Bauer-Brandl
`
`802
`
`350J}
`
`250,0 .i ..... --~·· .·····
`
`200.0
`
`150.0
`
`50/)
`
`(}_{}
`
`~ ..
`'"
`""
`
`.... :.:.
`"'
`·::-
`
`it
`
`-~
`
`... ~
`o'-
`
`""
`
`'$
`c;
`
`:;:;
`--i'
`c;
`""'
`+
`
`..,:~
`
`;i
`: .. ~
`0
`·!§
`"'
`w
`~-
`
`begins at the rear of th.e graph and move:; fonvard. To
`enhance the intelligibility of the chart, no bars were
`drawn i-.'1 for f:riabihty in excess of 2% (a degree of
`ft!ahHlty which would no longer be acceptable in prac(cid:173)
`tice_)-
`For all placebo tormulatiom, friability was low wheu
`(:ompa<::tion fince \V<l5 7.5 kN o:r h:ighex, hut the friabil-(cid:173)
`ity \)f the paracetamol tablets was often high. Suiil(cid:173)
`dently abrasion--resistant tablets (usually understood to
`m{~~:n tablets with a friability of no more than 1 %) could
`{lnly be prepared using PVP a..'1d HP-M 603.
`
`Disintegration Times
`
`The disintegratinn. time of r.tH~ tablets is of critic2l
`impDrtance of their efficacy. Although an effective
`"disintegrant" is added to the OJH(~r phase before
`tahleting, there are considerahk differences in disinte~
`
`g:ratio:n times. Figure 12 compa:res the disint~:gration
`times of placebo tablets produced uuder d:ifferent c.om(cid:173)
`p<Ktion fN·ces, \vhile Figure 13 provides si:mi!ar infM(cid:173)
`mation on pamcetamol tablets. It can be seen that it is
`in t~tct il1e tried and trusled hinders. iPVP and HP~M
`603) which delay disintegration. Lycatah DSH has the
`S<Ull{~ eff(:ct
`
`Ri:nder (~ft'ectiveness refers to the degree to which the
`lowest possible concentration t)f a himkr can contribute
`to the optimi:mti.on of ail granulate and tablet propertk~s.
`For gra.nubtes, these propertie:> include in particular
`f1ow characteristics, crushing strength, and mean pitt··
`tide size. The tablet property most strong1y affected is
`cm.i>hlng strength.
`
`Par Pharm, Inc.
`Exhibit 1020
`Page 012
`
`
`
`2.00
`1.80
`L60
`1.40
`1.20
`LOO
`0.80
`
`~ 0
`.,..,
`Q
`......
`*"'
`,.Q
`~
`"""" ~
`~
`
`"""' -.,..,
`
`2.00
`1.80
`1.60
`lAO
`L20
`1.00
`0.80
`0.60
`0.40
`0.20
`0.00
`
`force
`[kN]
`
`5
`
`<')
`0
`'-0
`~
`'
`r.J...
`::r:
`
`V)
`Q
`P.
`$J
`....,
`~
`ro
`<..>
`>-.
`...l
`
`'"' "'' ('.)
`X
`1:/J
`Q
`$:1
`<>)
`<>)
`u
`>-.
`-'
`
`-
`
`X ,._;;
`' u a.,
`X
`' -'
`
`Figunl HI. Friability of placebo tablets.
`
`'¢~
`0
`
`""'
`......
`(/) a
`.5:;
`t>l
`""'
`d
`Q
`~~'!
`
`-
`
`....
`~ ..,.,.
`~J
`< u ;:s.,
`X
`' -'
`
`(":\
`0
`\,()
`~
`•
`,.,..
`P.
`,.,.,
`
`..D
`~
`~
`;:.-,
`-'
`
`Figure U. Friability of paracemmol tablets.
`
`Par Pharm, Inc.
`Exhibit 1020
`Page 013
`
`
`
`8i
`m
`
`e:
`ttl g
`I
`.....
`ffi r:;
`ttl
`s p.
`"' ~-
`~
`~ .....
`·'"'
`~ rr
`~ ,..,
`
`~
`00
`
`-'-~ -----~-J!~-~.-o< ~ .... , ....... 1
`
`2
`~
`
`10%
`
`/(1
`
`61}(
`
`L·I-IPC-LH! 1 2%
`
`10/n
`; VI
`
`Lycatah DSH 2%
`
`Lycatab PGS 2%,
`
`6%
`
`HP-M 603 2%
`
`PVP K 30 2~{-,
`
`no binder
`
`S
`~
`~
`g;
`"" g
`'E.
`0 ,._.,
`~
`1-'1·
`~< g
`g
`
`Sf, s· ,g
`p
`;;!
`~ !'
`
`~g~
`... ;-.,.
`
`2
`~l
`
`J~:
`.,.. ...
`
`:~}
`
`~~:\:·
`•(,.:;
`
`2 .-;;:-;.
`
`:/~
`
`·:~;
`
`,._..
`
`-~,
`
`~;;;i
`-:~~
`
`:?E
`-~:::
`
`0
`
`disintegration time [sec]
`
`Par Pharm, Inc.
`Exhibit 1020
`Page 014
`
`
`
`Effectiveness of Binders in Wet Granulation
`
`805
`
`4000 .(·-·····-·
`
`l ·~·.
`
`2000 :
`
`i
`:~
`3000 >
`:,'
`2500 !.•
`I
`
`.. -T- ----- ----···············--- ....... ___ -- -·-·······
`
`:[
`
`l·
`
`l
`.· ~
`
`,. ~ ...
`~
`
`force jkN!
`
`!7,5
`
`Flgure 13, Disintegration time of paracetamol tablets.
`
`Par Pharm, Inc.
`Exhibit 1020
`Page 015
`
`
`
`806
`
`Becker, Rigassi. and Bauer-Brandl
`
`Ef
`
`Table 5
`Effect of Binder on Some Properties of Granulates and Tablets: Binder Effectiveness
`
`Binder
`
`PVP K30 PH
`
`Cellulose HP-M 603
`
`Lycatab PGS
`
`Lycatab DSH
`
`L-HPC Typ; LHll
`
`Proportion
`of Binder
`
`Particle
`Size
`(R 50%)
`
`2%
`6%
`10%
`2%
`6%
`lO%
`2%
`6%
`10%
`2%
`6%
`10%
`2%
`6%
`10%
`
`166%
`221%
`221%
`117%
`191%
`183%
`20%
`9%
`16%
`16%
`90%
`174%
`-34%
`-57%
`-58%
`
`Flow-
`ability
`
`33%
`33%
`35%
`35%
`34%
`33%
`28%
`25%
`11%
`12%
`32%
`36%
`-5%
`-20%
`-23%
`
`Granulate
`Strength
`
`Tablet
`Crushing
`Strength
`
`56%
`65%
`62%
`68%
`68%
`68%
`27%
`34%
`40%
`33%
`53%
`7&%
`-4%
`-24%
`-20%
`
`18%
`65%
`121%
`33%
`120%
`163%
`11%
`46%
`83%
`-6%
`26%
`54%
`11%
`35%
`37%
`
`The present assessment has been restricted to these
`features in order to allow conclusions about binder ef(cid:173)
`fectiveness in different formulations to be drawn more
`easily.
`For the graphs, the results obtained with binder-free
`placebo and drug preparations were taken as reference
`values and set at zero. The percentage deviations of the
`properties of the other granulates and tablets were al(cid:173)
`ways calculated in relation to these reference values.
`The arithmetic mean of the values obtained was then
`calculated for both the placebo and paracetamol granu(cid:173)
`lates (Table 5) and plotted (Figure 14).
`PVP K30 PH and Cellulose HP~M 603 show the
`greatest binder effectiveness. A 10% concentration of
`Lycatab DSH can be used as an alternative.
`The main effect of Lycarab PGS is the improvement
`of tablet crushing strength. Although L-HPC worsens
`granulate properties, it nevertheless improves tablet
`crushing strength to some extent.
`
`CONCLUSIONS
`
`Use of the standard binders PVP and Cellulose HP(cid:173)
`M 603 makes possible the manufacture of tablets with
`the best mechanical properties but the worst disintegra(cid:173)
`tion times. Hydroxypropylmethylcellulose in particular
`
`is used in gr