`
`2018-
`
`p
`
`XEomSoC
`
`1
`
`Argentum v Cosmo
`|PR2018—QOO80
`
`Cosmo Ex. 2018-p. 1
`Argentum v Cosmo
`IPR2018-00080
`
`
`
`The Science of Dosage Form
`
`"
`
`Edited by
`Michael E. Aulton BPharm PhD MRPharmS
`
`Professor of Pharmaceutical Technology, De Montfort University, Leicester, UK
`
`CHURCHill
`LIVINGSTONE
`
`EDINBURGH LONDON MELBOURNE AND NEW YORK
`
`Cosmo Ex. 2018-p. 2
`Argentum v Cosmo
`IPR2018-00080
`
`
`
`CHURCHILL LIVINGSTONE
`An imprint of Harcourt Publishers Limited
`A> is a registered trademark of Harcourt Publishers Limited
`
`© Miohael Aulton 1988
`
`All rights reserved. No part of this publication may be
`reproduced, stored in a retrieval system or transmitted in any
`form by any means, electronic, mechanical, photocopying,
`recording or otherwise, without either the prior permission of
`the publishers (Harcourt Publishers Limited, 24-28 Oval
`Road, London NW1 7DX), or a licence permitting restricted
`copying in the United Kingdom issued by the Copyright
`Licensing Agency Ltd, 90 Tottenham Court Road, London
`W1POLP.
`
`First published 1988
`Reprinted 1989
`Reprinted 1990
`Reprinted 1991
`Reprinted 1992
`Reprinted 1994 (twice)
`Reprinted 1995
`Reprinted 1996 (twice)
`Reprinted 1998
`Reprinted 2000
`
`ISBN 0-443-03643-8
`
`British Library Cataloguing in Publication Data
`A catalogue record for this book is available from the British
`Library
`
`Library of Congress Cataloging in Publication Data
`A catalog record is available from the Library of Congress
`
`Printed in China
`
`The
`publisher's
`policy is to use
`paper manufactured
`from sustainable forests
`I
`
`Cosmo Ex. 2018-p. 3
`Argentum v Cosmo
`IPR2018-00080
`
`
`
`32
`
`Mixing
`
`D N Travers
`
`THEORY OF MIXING
`Types of mixtures
`The mixing process
`The scale of scrutiny
`Degree of mixing
`
`THE MIXING OF POWDERED MATERIALS
`Mechanisms of mixing -
`convective, shear
`and diffusive mixing
`Influence of powder aggregates on mixing
`Segregation (de mixing)
`Types of segregation
`Percolation segregation
`Trajectory segregation
`Densification segregation
`Effect of mixing time on segregation
`Ordered mixing
`The practical importance of ordered mixing
`Segregation of ordered mixes
`Ordered unit segregation
`Displacement segregation
`Saturation segregation
`Industrial powder mixers
`Tumbling mixers
`Agitator mixers
`High speed granulators
`Scale:..up of powder mixing
`
`THE MIXING OF MISCIBLE LIQUIDS AND
`SUSPENSIONS
`Propeller mixers
`Turbine mixers
`In-line mixers
`
`THE MIXING OF SEMISOLIDS
`Mixers for semisolids
`T,he planetary mixer
`The sigma blade mixer
`
`550
`
`Mixing may be defined as a process where two or
`more components are treated so as to lie as nearly
`as possible in contact with a particle of each of the
`other components. The aim of the process is to
`produce one of the following:
`
`1 a blend of solid particles (powder mixing),
`2 a suspension of an insoluble solid in a liquid,
`3 a mixture of miscible liquids,
`4 a dispersion of particles in a semisolid as in the
`preparation of ointments or pastes.
`
`A mixing stage is involved at some time in the
`preparation of practically every pharmaceutical
`preparation.
`
`THEORY OF MIXING
`
`Types of mixtures
`
`Miscible liquids, gases and vapours will in time
`completely mix spontaneously by diffusion and no
`energy need be used for this to occur. This positive
`mixing may be contrasted with the negative mixing
`of insoluble solid particles with a liquid where the
`particles will separate out unless work is done by
`stirring to keep them dispersed. A mixture of
`powdered constituents is an example of a neutral
`mix. Work must be done to mix them initially but
`usually there is then no tendency for demixing to
`occur spontaneously; though demixing is possible
`in certain circumstances.
`
`The mixing process
`
`To simplify discussion the principles of mixing
`will be considered by reference to a system
`consisting of equal quantities of two constituents
`
`Cosmo Ex. 2018-p. 4
`Argentum v Cosmo
`IPR2018-00080
`
`
`
`MIXING
`
`551
`
`I
`
`I
`
`(a)
`
`(b)
`
`Fig. 32.1 Powder mixing: (a) segregated particles (b) ideal mixing (c) random mixing
`
`(c)
`
`A and B of the same size and shape though those
`conditions are rarely found in practice.
`The process can be represented by representing
`the materials as
`filled and open squares
`in
`Fig. 32.1. Fig. 32.1(a) shows the particles in a
`completely segregated state prior to mixing. As
`mixing proceeds it is just conceivable that a so(cid:173)
`called perfect mix may be produced where each
`particle lies adjacent to a particle of the other
`
`component (Fig. 32.l(b)). The odds against this
`happening are so great that the best attainable mix
`will actually be a random mix represented by
`Fig. 32.1(c). A random mix may be defined as one
`where the probability of sampling a particular type
`of particle is proportional to the number of such
`particles on the total mix. In a large number of
`trials where single particles were withdrawn there
`would be around the same number of each type,
`
`Cosmo Ex. 2018-p. 5
`Argentum v Cosmo
`IPR2018-00080
`
`
`
`552
`
`PHARMACEUTICAL TECHNOLOGY
`
`drawn from the random mix. In a small number The scale of scrutiny
`of trials a 'run' of a number of black or a number
`of white may be encountered due to chance just
`as a coin may fall 'heads' in three or four consecu(cid:173)
`tive trials but is very unlikely to do so in a
`greater number of trials.
`
`Figure 32.l(c) can also be viewed as a fairly large
`sample taken from a 50/50 bulk mix containing
`many thousands of particles. It contains 196 black
`particles and 204 white which is overall not far
`
`(a)
`
`( 194)
`20 x 20 50% RANDOM MIX
`
`(202)
`20 x 20 50% RANDOM MIX
`
`(c)
`
`( 198)
`
`20 x 20 50% RANDOM MIX
`Fig. 32.2 _Computer generated mixtures of nominal 50% active ingredient. The numbers in parentheses refer to the number of
`'white' particles in the mix, theoretically 200
`
`Cosmo Ex. 2018-p. 6
`Argentum v Cosmo
`IPR2018-00080
`
`
`
`from the required result of 200 of each. Figures
`32.2-32.4 are distribution patterns generated
`using a computer program simulating the taking
`of such samples from a random mix. No 'sample'
`in Fig. 32.2 contains exactly 200 of each but the
`deviation is small and acceptable.
`
`If the proportion of one component is reduced
`then the variation between samples is increased.
`Figure 32.3 is a printout for a nominal 10%
`content. Fig. 32.4 shows the position when the
`computer was
`instructed
`to print
`'samples'
`containing 36 particles of nominal 50% content.
`
`MIXING 553
`
`(a)
`
`(b)
`
`(42)
`
`(35)
`
`20 x 20 10% RANDOM MIX
`
`20 x 20 10% RANDOM MIX
`
`(c)
`
`(43)
`
`20 x 20 10% RANDOM MIX
`
`Fig. 32.3 Computer generated mixtures of nominal 10% active ingredient. The numbers in parentheses refer to the
`number of 'white' particles in each mix, theoretically 40
`
`Cosmo Ex. 2018-p. 7
`Argentum v Cosmo
`IPR2018-00080
`
`
`
`554
`
`PHARMACEUTICAL TECHNOLOGY
`
`(20)
`
`(21)
`
`( 15)
`
`( 17)
`
`6 x 6 50% RANDOM MIX
`Fig. 32.4 Computer generated mixtures of nominal 50%
`active ingredient showing the effect of small sample size.
`The numbers in parentheses refer to the number of 'white'
`particles in each mix, theoretically 18
`
`( 18) of each
`the mean number
`None has
`component and the sample with only IS particles
`is 16% deficient. These diagrams illustrate the
`important fact that 'low dose' preparations with
`scattered particles of the minor ingredient will
`show greater variation in their samples.
`Consider a large weight of powder mix which
`may be subsequently converted into a batch of
`tablets. The number of particles present in each
`tablet will depend on the correct weight and
`particle size of the ingredients and the distribution
`of the drug should be judged on this unit weight
`as the sample weight or 'scale of scrutiny'.
`It is of little consequence how the drug is
`dispersed within this weight but it is important
`that the scale of scrutiny should contain enough
`particles to avoid undue variation between tablets.
`The ingredients should therefore be of small
`particle size though decreasing this unduly may
`lead to aggregation where the fine particles adhere
`together to form agglomerates scattered throughout
`the mix.
`The importance of the scale of scrutiny is well
`illustrated by the results given in Table 32.1
`which were obtained from a computer simulation
`of sampling from a mix containing one part in one
`thousand of the minor constituent. The largest
`
`Table 32.1 Results of computer simulation of sampling from
`a 1: 1000 powder mix
`
`Trial Number
`
`Scale of scrutiny (no. of particles)
`
`1000
`
`10 000
`
`100 000
`
`2
`3
`4
`5
`6
`7
`8
`9
`
`10
`
`0
`
`0
`
`1
`
`0
`
`0
`
`1
`
`0
`0
`
`0
`
`8
`6
`
`10
`5
`4
`
`14
`
`8
`10
`
`9
`
`15
`
`101
`
`89
`
`104
`
`101
`
`96
`
`115
`
`92
`
`104
`
`96
`
`104
`
`The figures are the numbers of particles of minor constituent
`in the samples.
`
`sample size has an acceptable 'composition' with
`the content clustered around the expected mean
`of 100 per sample. The variation in the samples
`containing 10 000 particles would be unacceptably
`large for a real product such as a tablet or capsule.
`
`Degree of mixing
`
`It should be stressed that there is always some
`variation in the composition of samples drawn
`from a random mix and the standard deviation in
`the composition of a large number of such samples
`can be determined if an accurate assay method is
`available. A random mix, which is the best poss(cid:173)
`ible mix, will give samples with a lower standard
`deviation (S)
`than mixes of the same com(cid:173)
`ponents which have not been mixed completely
`to the random state. This fact can be used to
`define a mixing index.
`
`Where Sr is the standard deviation of samples
`drawn from a fully random mix and Sact is the
`actual standard deviation determined on
`the
`partially mixed system. As Sact is larger than Sr
`the mixing index approaches 1 as the mix becomes
`random. Several other more complicated mixing
`indices have been proposed but they all depend
`on similar principles.
`
`Cosmo Ex. 2018-p. 8
`Argentum v Cosmo
`IPR2018-00080
`
`
`
`THE MIXING OF POWDERED MATERIALS
`
`The mixing of powders is often difficult especially
`if a small quantity of a potent substance whose
`dose may be measured in f-Lg is to be mixed with
`a much larger quantity of diluent to form the dose
`weight of a tablet or capsule. The particles of drug
`and diluent may differ in size and shape, adding
`to the difficulty.
`
`Mechanisms of powder mixing -
`shear and diffusive mixing
`
`convective,
`
`If a spatula is inserted into a heap of powder on
`a tile then a small pile may be withdrawn on the
`blade and deposited elsewhere -
`convective
`mixing. At the same time an unstable shear plane
`is created which then collapses and produces some
`further mixing in the heap. Such shear and convec(cid:173)
`tive mixing will occur whenever a blade or ribbon
`moves through a powder mass and several indus(cid:173)
`trial mixers operate on this principle.
`There is a third mixing mechanism which is
`known as diffusive mixing. It can occur, for
`instance, in a drum mixer. In its simplest form
`this is a rotating cylinder in which the mix is lifted
`past its angle of repose (Chapter 36) so that the
`particles tumble over each other. There is a differ(cid:173)
`ence in velocity with the fastest layers near to the
`surface and individual particles can migrate from
`layer to layer in a number which is in some
`respects similar to the diffusion of molecules in
`liquids and gels.
`individual
`involves
`Since diffusive mixing
`particles, it can in time produce a truly random
`mix. Shear and convective mixing can quickly
`produce a rough mix but local groups of particles
`may remain unseparated unless subjected to dif(cid:173)
`fusive mixing. Care must be taken to allow sufficient
`time of mixing but for many purposes a non(cid:173)
`random mix may be quite satisfactory.
`
`I nfiuence of powder aggregates on mixing
`
`Although theory suggests that fine powders should
`form better mixes by increasing the number of
`particles in a given sample weight there is a
`complication caused by the increase in cohesion
`as the particle size decreases. The effect has been
`
`MIXING 555
`
`(1980)
`investigated by Orr et al.
`extensively
`who found marked variation between commercial
`samples of ethinyloestradiol tablets of the same
`nominal dose which was due
`to
`this cause.
`Vigorous convection and shear mixing are
`required to break up aggregates and disperse the
`particles into the bulle
`
`Segregation (de mixing)
`
`in practice will not be
`Powder mixes met
`composed of unisize spherical particles and certain
`factors may then operate when a degree of mixing
`has taken place which causes segregation ( demixing)
`at the same rate as mixing occurs. If that is the
`case then a random mix will not be formed no
`matter how long mixing is continued.
`The main factors promoting segregation are
`differences in particle size, shape and density.
`Smaller particles may fall
`through
`the voids
`between larger particles and thus make their way
`towards the bottom of the mass. The same effect
`can be caused by a difference in density even if
`particles are of equal size but particle size differ(cid:173)
`ence is the inain cause of segregation.
`Segregation may occur on discharge of a mixed
`powder for handling elsewhere. In recent years
`equipment which can. perform both mixing and
`other operations in the same vessel has become
`popular. This avoids handling the mass between
`operations.
`
`Types of segregation
`
`Percolation segregation This is the name
`given to the movement of particles through the
`voids in the powder bed. It will occur in a static
`bed if the percolating particles are so small that
`they can fall into the void spaces between the
`larger particles. The percolation also takes place
`for a short distance on either side of shear planes
`formed during the mixing process and for this to
`occur there need be very little difference in
`particle diameters. This is because of dilatation
`along the planes which separate the particles
`sufficiendy for smaller particles to move downwards.
`Trajectory segregation During mixing par(cid:173)
`ticles are set in motion and kinetic energy is
`imparted to them. Larger particles have larger
`
`Cosmo Ex. 2018-p. 9
`Argentum v Cosmo
`IPR2018-00080
`
`
`
`556
`
`PHARMACEUTICAL TECHNOLOGY
`
`energies and tend to move a greater distance into
`the powder mass before they are brought to rest.
`This may result in preferential separation and it
`can occur in horizontal as well as vertical planes.
`Densification segregation
`Segregation due to
`density differences between particles may cause
`segregation if the mass is subject to vibration.
`This can occur, for instance, in the hopper of a
`tablet machine and it is actively promoted in
`to place
`vibratory sieving (Chapter 35) so as
`smaller particles directly over the mesh. It is
`found that large dense particles move upwards
`through the mass. One explanation for this is that
`the smaller particles beneath a dense large particle
`are slightly compacted, supporting
`the
`large
`particle. On vibration the smaller particles move
`the
`beneath it and raise it upwards in the bed -
`process being repeated as time goes on. For a
`more detailed account of this and other mixing
`phenomena the reader is referred to an excellent
`review by Staniforth (1982).
`
`Effect of mixing time on segregation
`
`Non-segregation mixes will continue to improve
`with an extended mixing time but the reverse may
`be true for segregating mixes. Cook and Hersey
`( 197 4) studied the mixing of fenfluramine and
`quinalbarbitone with various diluents in a Nauta
`mixer and found that the mixes worsened after
`8-10 minutes. That was because
`the factors
`promoting segregation require a longer time to
`establish a segregation mix than is needed to
`produce a reasonable degree of mixing. It is there(cid:173)
`fore disadvantageous to prolong the mixing time
`beyond an optimum point.
`
`Ordered mixing
`
`If one of the constituents of a powder mix is added
`in a fine, often micronized form, then on mixing
`termed carrier particles -
`the larger particles -
`may adsorb some of these very small particles on
`to active sites on their surface and these are held
`tenaciously. The effect was first noticed during the
`mixing of micronized sodium bicarbonate with
`sucrose crystals when the mixture was found to be
`markedly stable towards segregation (Travers and
`White, 1971). Cook and Hersey (1974) termed
`
`such mixes 'ordered' since in contrast to random
`the constituent particles are not inde(cid:173)
`mixes,
`pendent of each other and sampling a host particle
`the adsorbed
`removing some of
`necessitates
`particles with it.
`
`The practical importance of ordered mixing
`
`It appears that the mixing of powders routinely
`met with in pharmaceutical processing is rarely
`in
`that some
`interaction and
`purely random
`cohesion will occur between constituents leading
`to a mix which is partly ordered and partly
`random. That is so regardless of the relative
`particle sizes of the constituents.
`Johnson (1979) for example found that in the
`mixing of tetracyline with lactose the mix was
`partially ordered with some tetracyline absorbed
`on the surface of the crystalline and spray-dried
`lactose which he used as (separate) diluents while
`some remained free and randomly distributed. It
`is now recognized that ordered mixing is an
`important factor preventing the segregation of
`mixes of a drug with one or other of the direct
`compressiOn bases that are discussed in Chapter
`39.
`
`Segregation of ordered mixes
`
`Notwithstanding the above, it is possible for
`segregation to occur in an ordered mix in certain
`circumstances which are discussed below.
`Ordered unit segregation This may occur if
`there are size differences between particles of the
`carrier constituent since the larger particles are
`usually associated with more adsorbed material
`than the smaller particles. Segregation of the
`carrier particles can
`then result in drug-rich
`areas in the mix -
`resulting in ordered unit
`segregation.
`Johnson in the work referred to above found
`that this could occur when coarse crystalline
`lactose was used as diluent and it could be avoided
`with finer grades which acted as earners of
`uniform particle size.
`Displacement segregation The addition of
`another constituent to an ordered mix may result
`in competition for the available sites on the carrier
`particles and some adsorbed particles may be
`
`Cosmo Ex. 2018-p. 10
`Argentum v Cosmo
`IPR2018-00080
`
`
`
`MIXING 557
`
`Fig. 32.5 A Y-cone blender
`
`displaced. Magnesium stearate is commonly added
`as a lubricant to granulates and Lai and Hersey
`(1979) have demonstrated that this can displace
`adsorbed drugs under certain circumstances.
`Saturation segregation As there are only a
`certain number of sites on the carrier particles any
`excess material will rapidly segregate by perco(cid:173)
`lation if too much is added. There is some
`evidence that certain sites are stronger than others
`and particles held on the weaker sites may become
`dislodged if the mix is vibrated vigorously.
`
`Agitator mixers
`
`These mixers depend on the motion of a blade or
`paddle through the charge producing a high
`degree of convective mixing. The ribbon mixer
`(Fig. 32.6) is a common type. The helical blades
`rotate in a hemispherical trough but 'dead spaces'
`are difficult to eliminate and the shearing action
`of the ribbon may be insufficient to break up
`aggregates.
`
`Industrial powder mixers
`
`Tumbling mixers
`
`Free-flowing material can be mixed in tumbling
`mixers which are rotating vessels of various shapes
`so that the charge flows with the vessel when the
`angle of repose is exceeded.
`The Y -cone mixer is a good example of this type
`(Fig. 32.5). On rotation the charge flows into the
`two top arms of theY and then back into the third
`arm. Mixing by shear and diffusion takes place
`when the streams mingle. The mixer is often used
`to blend in lubricant with tablet granules prior to
`compression when the gentle action is an advan(cid:173)
`tage. Time must be allowed for the mix to flow
`into the arms and there is an optimum speed of
`rotation. Other tumbling mixers have the mixer
`vessel in the shape of a cube or double cone and
`prongs may be fixed in the vessel to increase
`convective mixing.
`
`Fig. 32.6 Agitated powder mixer
`
`A more recent type of agitator mixer is the
`Nautamixer (Fig. 32.7). It consists of a conical
`vessel fitted at the base with a rotating screw
`which is fastened to the end of a rotating arm at
`the upper end. The screw conveys material to near
`the top where it cascades back into the mass. The
`mixer thus combines diffusive, shear and convec(cid:173)
`tive mixing with an avoidance of dead spots since
`the screw visits each part of the interior in each
`revolution as the arm rotates.
`
`High speed granulators
`
`As stated earlier, there has been a trend towards
`
`Cosmo Ex. 2018-p. 11
`Argentum v Cosmo
`IPR2018-00080
`
`
`
`558
`
`PHARMACEUTICAL TECHNOLOGY
`
`-+-+- Rotating
`arm
`
`conveyor
`
`Discharge
`
`Fig. 32.7 A Nautamixer. (Courtesy of Nautamixer Ltd)
`
`the use of composite equipment in the pharma(cid:173)
`ceutical industry. A good example is the high
`speed mixer/granulator combining both
`these
`operations (Fig. 32.8).
`One type has a central impeller mounted as
`
`shown in Fig. 32.9. When the mix is rotated, the
`mixer acts to fluidize it without the need to blow
`air through the mass as in a fluidized bed drier
`(Chapter 38). When thoroughly mixed the granu(cid:173)
`lating agent can be added and distributed at a
`
`Cosmo Ex. 2018-p. 12
`Argentum v Cosmo
`IPR2018-00080
`
`
`
`MIXING 559
`
`Fig. 32.8 High speed mixer/granulator. (Courtesy of T K Fielder Ltd)
`
`Fig. 32.9 Diagrammatic section of a high speed mixer/granulator
`
`Cosmo Ex. 2018-p. 13
`Argentum v Cosmo
`IPR2018-00080
`
`
`
`560
`
`PHARMACEUTICAL TECHNOLOGY
`
`slower speed. The damp mass is reduced to gran(cid:173)
`ules by the side entering chopper blade.
`The Diosna mixer/granulator (Chapter 37) is
`another example of this type.
`
`That example illustrates the need for testing and
`validation at every stage in the manufacture of a
`new product.
`
`Scale-up of powder mixing
`
`Development work on a new product must of
`necessity be on a bench scale. There is then no
`guarantee that a mixing process will be satisfactory
`on a manufacturing scale even if a larger mixer of
`the same type can be used. This point was illus(cid:173)
`trated by the work of Delonca et al (1982) who
`examined the homogeneity of a sulphaguanidine/
`lactose mix prepared in plough mixers of different
`size. The bench size (5 litre) mixer yielded a poor
`mix but a 20 litre model gave better mixes
`comparable in quality to those obtained from a
`large ( 130 litre) model. In that instance the mix
`improved in scale-up. In another case drug defi(cid:173)
`ciency in a production scale mix was traced to
`adsorption of the minor constituent on the mixer
`wall which did no.t occur on bench scale working.
`
`THE MIXING OF MISCIBLE LIQUIDS AND
`SUSPENSIONS
`
`Mobile liquids with a low coefficient of viscosity
`are easily mixed with each other. Solid particles
`are also readily suspended though these particles
`will settle rapidly when mixing is discontinued.
`More viscous liquids are difficult to stir but
`suspended particles will settle slowly through at
`a lower rate than predicted by Stokes' law because
`of 'hindering' by other particles.
`
`Propeller mixers
`
`A common arrangement for medium scale fluid
`mixing is a propeller type stirrer which is often
`used clamped to the edge of a vessel. A propeller
`has angled blades which cause the circulation of the
`fluid in both an axial and radial direction. The off-
`
`Flow pattern
`
`--r::l-- Vertical
`
`baffle
`
`Unbaffled tank
`
`(a)
`
`Baffled tank
`
`(b)
`
`Fig. 32.10 Propeller mixer'- formation of vortex (a) unbaffled tank (b) baffled tank
`
`Cosmo Ex. 2018-p. 14
`Argentum v Cosmo
`IPR2018-00080
`
`
`
`centre mounting discourages the formation of a
`the stirrer is
`vortex which may form when
`mounted centrally. A vortex forms when the
`circumferential motion imparted to the liquid by
`the propeller causes it to back up round the sites
`of the vessel and form a depression around the
`shaft. As the speed of rotation is increased air may
`be sucked into the fluid with frothing and possible
`oxidation (Fig. 32.10(a)). Another method of
`suppressing a vortex is to fit vertical baffles into
`the vessel. These divert the rotating fluid from its
`circular path into the centre of the vessel where
`the vortex would otherwise form (Fig. 32.10(b)).
`
`Turbine mixers
`
`The propeller stirrer depends for its action on a
`satisfactory axial and radial flow pattern and this
`will not occur if the fluid is viscous. The ratio of
`the diameter of a propeller stirrer to the diameter
`of the vessel is commonly 1 : 10 to 1 : 20 and it
`operates at speeds of 1-20 rev s- 1• There must
`be a fast flow of fluid towards the propeller which
`can only occur if the fluid is mobile.
`A turbine mixer may be used for more viscous
`fluids and a
`typical construction is given in
`Fig. 32 .11. The impeller has four fiat blades
`surrounded by perforated inner and outer diffuser
`rings. The rotating impeller forces
`the liquid
`through the perforations with considerable radial
`velocity sufficient to overcome the viscous drag of
`the bulk of the fluid. One drawback is the absence
`of an axial component but a different head with
`the perforations pointing upwards can be fitted if
`this is desired.
`The high shear forces as the liquid is forced
`through the orifices also enable such mixers to
`produce emulsions of immisible liquids. Turbine
`type mixes will not deal with liquids of very high
`viscosity which are best treated as semisolids and
`handled in the same equipment as used for such
`materials.
`
`In-line mixers
`
`MIXING 561
`
`J \
`
`+--+-+-- Central turbine
`
`Perforated ~--~<-o
`top
`
`Fig. 32.11 Turbine mixer. (Courtesy of Silverson Mixers
`Ltd)
`
`Central aperture
`
`THE MIXING OF SEMISOLIDS
`
`The problems which arise during the mixing of
`semisolids (ointments and pastes) stem from the
`fact that unlike powders and liquids, semisolids
`will not flow easily. Material that finds its way to
`'dead' spots will remain there. For that reason
`suitable mixers must have rotating elements with
`narrow clearances between the mixing vessel and
`they must produce a high degree of shear mixing
`since diffusion and convective mixing cannot
`occur.
`
`Mixers for semisolids
`
`The planetary mixer
`
`As an alternative to mixing fluids in vessels mobile
`miscible components may be fed through an 'in(cid:173)
`line' mixer designed to create turbulence in the
`stream.
`
`This type of mixer is commonly found in the
`domestic kitchen (Kenwood mixers) and larger
`machines which operate on the same principle are
`used in industry.
`
`Cosmo Ex. 2018-p. 15
`Argentum v Cosmo
`IPR2018-00080
`
`
`
`562
`
`PHARMACEUTICAL TECHNOLOGY
`
`Container
`
`1--1-----'r'r-- Rotational path
`of blade edge
`
`~--1-----H-- Rotational path
`of blade shaft
`
`--H--- Body of mixer
`
`Fig. 32.12 Planetary mixer -
`
`top view showing path of paddle
`
`The m1xmg paddle is set off centre and is
`carried on a rotating arm. It therefore travels
`round the circumference of the mixing bowl
`while simultaneously rotating around its own axis
`(Fig. 32.12). A small clearance between the vessel
`and the paddle gives shear but 'scraping down'
`several times is necessary to mix the contents well
`since some materials are forced to the top of the
`bowl.
`
`The sigma blade mixer
`
`This robust mixer will deal with stiff pastes and
`ointments and depends for its action on the close
`intermeshing of the two blades which resemble the
`Greek letter 2: in shape. The clearance between the
`blades and the mixing trough is kept small by the
`shape shown in Fig. 32.13.
`
`Fig. 32.13 Sigma•arm mixer
`
`Cosmo Ex. 2018-p. 16
`Argentum v Cosmo
`IPR2018-00080
`
`
`
`It is very difficult using primary mixers to
`completely disperse powder particles in a semi(cid:173)
`solid base so that they are invisible to the eye. The
`mix is usually subjected to the further action of
`
`a roller mill or colloid mill so as to 'rub out' these
`particles by the intense shear generated by rollers
`or cones set with a very small clearance between
`them.
`
`MIXING 563
`
`REFERENCES AND BIBLIOGRAPHY
`
`Cook, P. and Hersey, J. A. (1974) Powder mixing in the
`tabletting of fenfturamine hydrochloride; evaluation of a
`mixer. J. Phann. Phannacol., 26, 298-303.
`Delonca, H., Jeannin, C., Joachim, J., Romier, G. and
`Verain, A. (1982) Scaling tests on plough mixers. Int. J.
`phann. Tech prod. Mfr, 3(3), 87-92.
`Johnson, M. C. R. (1979) Powder mixing in a direct
`compression formulation by ordered and random
`processes. J. Phann. Phannacol., 31, 273-276.
`Lai, F. K. and Hersey, J. A. (1979) A cautionary note on
`the use of ordered mixtures in pharmaceutical dosage
`forms. J. Phann. Phamzacol., 31, 800.
`
`Orr, N. A., Hill, E. A. and Smith, J.F. (1980) Dosage
`uniformity in hydrocortisone ointment BP. J. Phann.
`Phannacol., 32, 766-772.
`Rothman, H. (1981) High speed mixing in the
`pharmaceutical industry. Mfg. Chern. Aerosol News, 52(4),
`47-49.
`Staniforth, J. N. ( 1982) advances in powder mixing and
`segregation in relation to pharmaceutical processing. Int.
`J. phann. Tech. prod. Mfr, 3(Suppl), 1-12.
`Travers, D. N. and White, R. C. (1971) The mixing of
`micronized sodium bicarbonate with sucrose crystals. J.
`Phann. Phannacol., 23 Suppl., 260S-261S.
`
`Cosmo Ex. 2018-p. 17
`Argentum v Cosmo
`IPR2018-00080
`
`