A critical view on lactose-based.pdf

7/26/2019 A critical view on lactose-based.pdf

1/18

A critical view on lactose-based drug formulation and device studies for dry powderinhalation: Which are relevant and what interactions to expect?

A.H. de Boer a , , H.K. Chanb , R. Price ca Department of Pharmaceutical Technology and Biopharmacy, University of Groningen, Ant. Deusinglaan 1, 9713AV Groningen, The Netherlandsb Faculty of Pharmacy, University of Sydney, NSW 2006, Australiac Department of Pharmacy and Pharmacology, University of Bath, Bath BA2 7AY, United Kingdom

a b s t r a c ta r t i c l e i n f o

Article history:Received 10 February 2011Accepted 18 April 2011Available online 23 April 2011

Keywords:Adhesive mixturesCarrier lactoseDrug formulationPulmonary drug depositionPowder mixingDispersionInhaler development

Many years of research have not led to a profound knowledge of the mechanisms involved in the formulationand dispersion of carrier based mixtures for inhalation. Although it is well understood that the mixing is a keyprocess in DPI carrier based formulation, there remains a limited understanding of how blending processesaffect in-process material propertiesand the resulting distributionof the drug in the naldosageform.A greatnumber of variables are considered relevant to the interfacial forces in adhesive mixtures, but their effectshave mostly been investigated individually, without taking account of the in uence they may have on eachother. Interactions may be expected and without proper choices made and de nitions given for all thevariables involved, conclusions from studies on adhesive mixtures are of less relevance. By varying any of thevariables that are not subject of the study, an opposite effect may be obtained. Currently, there is a strongfocus on exploring techniques for the characterisation of drug and carrier surface properties that are believedto have an in uence on the interparticulate forces in adhesive mixtures. For a number of surface propertiesit may be questioned whether they are really the key parameters to investigate however. Their orders of magnitude are subordinate to the effects they are supposed to have on the drug-to-carrier forces. Therefore,they seem rather indicators of other variability and their in uence may be dominated by other effects. Finally,the relevance of inhaler design is often ignored. By using powerful inhalers, the effect of many variables of

current concern may become less relevant. Carrier properties that are considered disadvantageous at presentmay even become desirable when a more appropriate type of dispersion force is applied. This can be shownfor the effect of carrier surface rugosity when inertial separation forces are applied instead of the more widelyapplied lift and drag forces. Therefore, inhaler design should be taken into consideration when evaluatingstudies on adhesive mixtures. It should also become an integral part of powder formulation for inhalation.

2011 Elsevier B.V. All rights reserved.

Contents

1. Introduction: adhesive mixtures for inhalation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252. Different research focuses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2.1. Functionality testing of lactose carriers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22.2. Drug and carrier surface properties and interaction studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2592.3. The mixing process: the relevance of carrier bulk properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2612.4. The role of lactose nes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3. The mixing process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4. Interactions between variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

4.1. Linked effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4.2. Conditional effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3. Interacting effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

5. Taking a critical view . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5.1. On the role of carrier surface properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25.2. On the role of amorphous spots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3. On the role of nes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Advanced Drug Delivery Reviews 64 (2012) 257 274

This review is part of the Advanced Drug Delivery Reviews theme issue on Lactose as a Carrier for Inhalation Drug Delivery . Corresponding author. Tel.: +31 503633286; fax: +31 503632500.E-mail addresses: [email protected] (A.H. de Boer), [email protected] (H.K. Chan), [email protected] (R. Price).

0169-409X/$ see front matter. 2011 Elsevier B.V. All rights reserved.

doi: 10.1016/j.addr.2011.04.004

Contents lists available at ScienceDirect

Advanced Drug Delivery Reviews

j o u rn al h o mep ag e : ww w. el sev ie r. co m / lo ca t e / ad d r
http://dx.doi.org/10.1016/j.addr.2011.04.004http://dx.doi.org/10.1016/j.addr.2011.04.004http://dx.doi.org/10.1016/j.addr.2011.04.004mailto:[email protected]:[email protected]:[email protected]://dx.doi.org/10.1016/j.addr.2011.04.004http://www.sciencedirect.com/science/journal/0169409Xhttp://www.sciencedirect.com/science/journal/0169409Xhttp://localhost/var/www/apps/conversion/tmp/scratch_7/Unlabelled%20imagehttp://dx.doi.org/10.1016/j.addr.2011.04.004http://localhost/var/www/apps/conversion/tmp/scratch_7/Unlabelled%20imagemailto:[email protected]:[email protected]:[email protected]://dx.doi.org/10.1016/j.addr.2011.04.004


2/18

6. The relevance of inhaler design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266.1. Effect of grid structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26.2. Effect of mouthpiece length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266.3. Role of the rotating capsule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266.4. In uence of air ow rate in the device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270

7. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction: adhesive mixtures for inhalation

Powder mixing is the most important unit operation in thepreparation of solid dosage forms. It was originally envisaged thatmixing of coarse and small quantities of ne particles would poseextra concern regarding attainable homogeneity and the possibility of segregation of the constituents during handling of the blend [1].However, with the invention of the scanning electron microscope itbecame possible to investigate these mixtures, and it was recognisedthat in such mixtures the coarse excipient particles may act as ahost for adhering drug particles [2]. Fine particles adhere to thesurface of the host crystals by van der Waals, capillary, electrostaticor mechanical forces and as a result, a higher uniformity of drugdistribution in the mixture can be obtained as is theoretically possibleon the basis of homogeneity equations derived for random mixtures[3,4]. These ndings led to the presentation of a new concept of ordered mixtures in powder mixing practise [5]. It was explainedthat the fundamental difference between an ordered and a randommix is the nature of the forces which limit the freedom of migrationfor the ne constituent particles within an ordered mixture [6]. Thein uence of the force of gravity in such mixtures applies to theordered units and not to the nes within these units. The interactionforces between the drug and excipient host particles were consideredbene cial with respect to handling of the powder blend (e.g. for drygranulation, tabletting and capsulation) as it signi cantly reducesthe risk of segregation [7 12]. Therefore, early studies focussed onunderstanding the mixing process [6,13,14] . It was observed thatmixingof thesame constituentsmay have differentoutcomesand thisobservation showed that mixing of particularly ne and coarseparticles is a dynamic process [15]. In an attempt to achieve a betterunderstanding of the different mixture types with their theoreticalvariances, a total mixing concept was introduced [6] . In the sameperiod a debate was started about the correct conception of the termordered mixture and the most appropriate nomenclature for thistype of mixture [6,15 22]. In literature, ordered was used to referto mixtures with a high degree of homogeneity in excess of thatexpectedfor randommixturesas well as to thenew concept describedby Hersey. As a response to this confusion, alternative terms wereintroduced like interactive and adhesive mixtures. A climax to thisdebate was given by Staniforth in his British PharmaceuticalConference Science Award Lecture (1986) in which he explained

that in fact all matter interacts irrespective of its nature and size [15].Therefore, interactive is not the correct term to distinguish randommixtures from mixtures in which ne constituent particles adhereto the surface of much larger coarse constituent particles. Becausethe latter type of mixture neither guarantees a higher degree of homogeneity than random mixtures, the term ordered has to berejected too. What matters is that interaction leads to adhesion (orcohesion) which depends on the balance between the interparticulateforces and the force of gravity. Hence, adhesive mixtures is a bettername. This reasoning by Staniforth has been adapted, and therefore,the name adhesive mixture is used in this manuscript.

Early studies on adhesive mixtures addressed methods to increasethe interparticulate forces in the mixture as a high homogeneitywas the primary objective and segregation was the main concern.

Adhesive mixtures were primarily used for tabletting and granulation

processes. Methods applied to increase interparticulate forcesincluded the use of high energy input mixers, such as ball mills[23,24] . Higher energyinput appeared to result in a more rapid mixingand a higher degree of homogeneity. The positive effect of millingwas explained by the formation of lattice defects which lead to anincrease in surface energy and act as active points for adhesion [25].Mechanical activation of the carrier surface results in a decrease of the degree of order and this determines the mixing rate. An increasein the adhesive tendency was also reported from applying longermixing times [26]. The effect was attributed to an increasedtriboelectri cation of the particles in the blend due to an increasednumber of contacts and collisions between surfaces with longermixing times.

When adhesive mixtures were investigated for pulmonary drugdelivery with dry powder inhalers, the objective changed frompreventing segregation to achieving a high and consistent neparticle dose. For this application, the interparticulate forces need tobe strong enough to facilitate handling but also weak enough toenable separation of drug and excipient during inhalation, using theair ow through the inhaler device as energy source. Controlling,rather than maximising the interparticulate forces became the newchallenge. This required a better understanding of the type of forcesinvolved and knowledge of the factors that in uence these forces.Studies were focussed on drug and carrier surface properties, carrierbulk properties, the presence of naturally occurring nes in carrierproducts and their effects on the distribution of the drug over thecarrier surface and the interaction between the two. Many of theproperties investigated relate strongly to the carrier particle sizedistribution, which also affects the ow properties of the carrier. Thisresulted in functionality testing of commercially available carrierproducts. Drug mixtures with these carriers were tested uponconsistency of delivered dose and ne particle dose which aredetermined by the ef cacies with which the dose compartment isemptied, the powder is dispersed in the air stream and particles arede-agglomerated during inhalation respectively.

Conclusions from many studies undertaken to understand thedrug-to-carrier interactions and the variablesthat control or in uencethese interactions are based on the same end parameters, whichare the in vitro deposition results or ne particle dose. It should berealised that these end parameters are the net result of a series of subsequent processes which comprises selection (including classi -

cation and/or conditioning) or production (including particle engi-neering) of the starting materials (drug and carrier), the mixingprocess, dispersion and de-agglomeration in the inhaler deviceand nally the aerosol characterisation. Each of these operationsin uences the outcome and it is surprising that in many studies onadhesive mixtures for inhalation the role of the mixing process isneglected. Likewise, the in uence of the type of inhaler on the endparameters is often ignored. Different mixer types and inhalers withdifferent dose systems and dispersion principles have been used invarious studies. In some studies mixing times, batch sizes and owrates through the inhaler have not even been mentioned. Not tomention that in most studies the effect of a single variable has beeninvestigated without considering its effect in relation to the in uenceof all other factors or dependence of the speci c properties of

the variable thereon. Therefore, conclusions drawn from a particular

258 A.H. de Boer et al. / Advanced Drug Delivery Reviews 64 (2012) 257 274


3/18

study may apply only for the conditions chosen. Finally, a tremendousamount of energy has been focussed on studying carrier propertiesof which their relevance to the drug-to-carrier interaction may beconsidered doubtful, simply because the order of magnitude of theirin uence is questionable, or because they are linked to respectivelydominated by other variables.

Theaim of this manuscript is to take a critical view on therelevanceand understanding of some of the most extensively investigated

carrier properties and to discuss the importance of some neglectedoperations (mixing and dispersion) in studies on adhesive mixturesfor inhalation. Some examples will be given of variables of whichthe effects on drug particle release from the carrier surface duringinhalation seem well known. However, it can be shown that theseeffects may depend on the choices made for one or more of the othervariables involved, which indicates that interactions between thevariables exist. It can also be shown that the effect of a variable maydepend on its speci c properties, which for instance is the particlesize distribution of lactose nes in relation to that of the drug.Changing the properties of the variable may therefore result in adifferent or even opposite effect.

2. Different research focuses

A large number of studies have been undertaken on the subject of adhesive mixtures for inhalation. They can roughly be divided intofour different research areas which are brie y reviewed to show thevariation in approaches and techniques currently used for drug-to-carrier interaction studies. A critical view on some of these techniqueswill be given in Sections 4 and 5 .

2.1. Functionality testing of lactose carriers

Pharmaceutical companies have an interest in designing orobtaining the rights to inhalation devices which can be used for awide range of different drug formulations [27]. The dry powderinhalation products for these inhalers are developed using eithercatalogue (off-the shelf) carrier products or tailor-made carriermaterial, such as special size fractions which are processed to t thedevice and formulation characteristics. The requiredcarrier propertiesdepend on the type of drug to be processed, the drug concentration(% w/w) in the mixture as determined by the drug dose and theamount of powder to be measured by (or into) the dose system, andthe type of mixing process used. They re ect on the emptying of thedose compartment and the dispersion of the formulation duringinhalation which both determine the consistency of delivered doseand ne particle dose. Carrier excipients furthermore need to beaccepted by regulatory authorities, be pure, stable and available frommore than one supplier, exhibit no batch or supplier variations and bepreferably fully characterisable for parameters known or expected tobe relevant to their performance [27]. However, not knowing theprecise mechanisms of interaction between drug and carrier particles

andall theparameters that in uence these interactions in detail, exactspeci cations for these parameters cannot be given. The relevanceand precise mechanism of action of most parameters is still uncertain.Therefore, in most cases formulation of adhesive mixtures forinhalation in practise is still an empirical process. Different carrierproducts with different properties regarding size distribution, surfacerugosity and anomeric composition are subjected to functionalitytesting in order to decide which one has the greatest potential for thedrug formulation to be used in combination with the inhaler(s)selected for administration [28 31]. Although such focussed-in-housestudies are very important for understanding the performance of marketed carrier products in one particular type of inhaler device,they do not result in widely applicable conclusions. Between twodifferent carrier products many variables regarding particle size and

shape distribution, presence of nes and impurities, rugosity, carrier

surface payload, etc. are different. This makes it impossible to relatethe effect of single physical carrier properties to the aerodynamicbehaviour of the mixtures prepared with these carriers [32,33] .Besides, the performance of the optimal formulation in the study maybe completely different in another inhaler.

2.2. Drug and carrier surface properties and interaction studies

A review on interparticulate (adhesion and cohesion) forces inadhesive mixtures for inhalation has been published previously [34].These interfacial forces have been discussed in conjunction withparticle preparation techniques such as milling, condensation, spraydrying, precipitation and crystallisation which yield different particlesurface properties that may directly affect the drug-to-carrier interac-tion [35]. The recognition of the complex relationship between thephysical lactosecarrier properties and the aerodynamic drugproperties[32] has resulted in a desire to obtain a better understanding of theunderlying mechanism(s). This searchhas stimulated research towardsvariables that are expected to in uence the interparticulate forces andexploring techniques that can be used to qualify and quantify thesevariables. Theexistence of carrier surface areas with high surface energywas described to which drug particles are preferentially attracted[36]. These so-called active sites have been explained in terms of adhering nes [37], amorphous spots and disorders in the crystalstructure [38], impurities [39,40] , water of adsorption [39], clefts orasperities [36] and sites of high surface energy [41]. It could be shownthat some of the surface properties relate to each other [40]. Forexample, the carrier surface impurities characterised with the lightextinction of a 5% aqueous lactose solution at 280 nm (E-280) and thepercent of water of adherence, both per unit calculated carrier surfacearea (CSA), increase with increasing mean carrier diameter forparticles from the same batch of alpha lactose monohydrate(Fig. 1A). This is the result of an increasing size of the carrier surfacediscontinuities with increasing particle diameter ( Fig. 1B) which canbe expressed as Surface Roughness Index (SRI: being the ratio of speci c surface area from nitrogen adsorption to calculated surfacearea, Fig. 1A). There may be a simple explanation for both guressince aws in the crystal lattice at the exterior of the crystal grow withthe diameter of the crystal. This results in a more or less constant ratiobetween the size of the surface discontinuities from lattice faultsand the size of the crystal ( Fig. 1B). When the crystals are removedfrom the crystallisation tank, higher amounts of mother liquor remainin these larger surface discontinuities, which after drying of thecrystals result in higher amounts of impurities (per CSA). Taking intoaccount that peptide and protein like impurities can absorb muchhigher amounts of water than alpha lactose monohydrate, it can beexplained why the SRI, E-280 and % H2O all exhibit the same trendwith increasing carrier diameter.

The understanding that surface properties play a dominant role inthe drug-to-carrier interaction has led to the exploration of a greatvariety of techniques to measure these properties. They have been

reviewed before [42] and include for instance inverse gas chroma-tography (IGC), X-ray microanalysis, atomic force microscopy (AFM),scanning electron microscopy (SEM) in combination with imageanalysis, laser pro lometry, differential scanning calorimetry (DSC),micro-calorimetry, and dynamic vapour sorption (DVS). A disadvan-tage of some of these techniques (AFM and laser pro lometry) is thatonly a very small part of the entire carrier surface area can becharacterised, whereas particles with large surface discontinuities(e.g. granular structures) cannot be measured at all with AFM. On theother hand, AFM directly delivers the force of adhesion between aparticle, attached to the cantilever, and a substrate. This also enablesto measure the separation energy necessary to detach a particle fromthe substrate surface [39]. AFM measurements have been used toshow that the separation energy for a drug particle attached to an

atomically smooth lactose surface increases with increasing relative



4/18

humidity to which a lactose surface has been exposed [39]. The ratherextreme effect of humidity on the adhesion force and separationenergy for salbutamol sulphate lactose and budesonide lactosecombinations is not re ected by the effect of humidity on ne particlefractions for these combinations from inhalers however. With thesame technique, it was shown that increasing the rugosity of a lactosesurface widens the distribution of adhesion forces with a drug particle[43]. One of the practical problems to solve in AFM measurements isthe uncertainty about the real contact area between the probe andthe substrate surface. This makes comparative evaluation of cohesiveand/or adhesive forces between different substances onerous.Therefore, techniques were presented to improve the comparabilitybetween different drug excipient combinations, e.g. by the prepara-tion of small spherical polycrystalline particles in a narrow sizedistribution to reduce the variation in contact area [43]. In anotherapproach a series of different probes with different contact areas

was used to measure the cohesive (X to X) and adhesive (X to Y)force between the same materials repeatedly, yielding different(corresponding) values for both forces from each probe to be plottedin a cohesion adhesion graph [44]. This technique has led to thepresentation of a so-called cohesion adhesion balance (CAB) fromwhich rankings forthe cohesiveand adhesive forcesof different drug

drug and drug excipient combinations were derived. This approachhas been used to predict the dispersion mechanisms and in vitrodeposition performance of various drug-carrier combinations [45,46] ,including ternary mixtures with excipient nes [47]. Drug-carriercombinations in which the cohesive forces are slightly higher thanthe adhesive (CAB N 1) tend to yield higher ne particle fractions [46].This seems logical as a high CAB-value increases the agglomerationtendency of the drug particles and agglomeration increases the ratio

of detachment force to adhesion force. An important conclusion

drawn from CAB studies is that the balance of cohesive and adhesiveforcesis highlydependent on theprocess history of thedrugsfrom thesource of the primary crystals, the energy input during comminutionand the relaxation behaviourof the mechanically activateddrugs. Thisdynamic changein interfacial behaviour of a drug substance has madecontrolling drug-carrier interactions dif cult using current industrialprocessing technologies for the drug. To overcome these limitationsdrugs may need to be conditioned under controlled environmentalconditions or exposed to suitable organic vapours to expedite therate of mechanical relaxation. On the other hand, it can be shownthat variations in the CAB may be overcome by using highly effectiveinhalers.

The understanding that carrier surface heterogeneity existshas increased the desire to characterise the entire carrier surface.Techniques like inverse gas chromatography were introduced tomeasure the distribution of surface energy [41]. Studies presenting

data from inverse gas chromatography measurements are stillrelatively scarce however, particularly those in which the surfaceenergy of carrier materials is related to drug adhesion onto the carriersurface. Values presented for the surface energy of lactose have theorder of magnitude of 40 to 50 mJ/m 2 depending on the lactosepreparation technique and the size fraction [41,48,49] . Milling causesan increase in dispersive surface energy which is attributed to theformation of amorphous regions [50]. When the amorphous fractionin milled lactose is re-crystallised by exposure to a high relativehumidity, the surface energy decreases to the value of the startingmaterial but the energy distribution remains much lower. Bothsieved and milled lactose have broader energy distributions than re-crystallised lactose. For sieved lactose this is explained with aheterogeneous distribution of impurities over the lactose surface

and variations in anomeric composition, whereas for milled lactose

A

B

1

1.5

2

2.5

3

0 50 100 150 200

mean fraction diameter (micron)

r e l a t i v e v a

l u e

E-280 per CSA

SRI

%H2O per CSA

Fig. 1. A. Related carrier surface properties: surface roughness index (SRI) expressed as ratio of speci c surface area from nitrogen adsorption (BET-method) to calculated surfacearea (CSA), UV-absorption of a 5% aqueous lactose solution (E-280) and water content, all per unit CSA, as function of the mean fraction diameter for different size fractions derivedfrom the same batch of lactose. B. Comparison of different carrier size fractions at different magni cations, showing that carrier surface discontinuities and irregularities grow withthe particle diameter.

http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B1%80


5/18

small amorphous regions would be the reason. The addition of lactosenes may result in a more homogeneous surface energy of lactose

particles [49]. Studies presenting correlations between surface energyand dispersion behaviour are scarcely available and they have beengiven for formulations with (co-processed) rifampicin, salbutamolsulphate, salmeterol xinafoate, ipratropium bromide and co-spraydried cromoglycate [48,51 53]. Clear and unique relationships areoften not obtained however [53,54] , unless individual data points are

eliminated from the comparison [48]. This suggests that generalconclusions for the effect of surface energy on drug aerosolisationduring inhalation may not be possible as too many other physicalvariables control or even dominate dispersion [55].

2.3. The mixing process: the relevance of carrier bulk properties

Particle processing, mixing and dispersion are the key operationsin adhesive mixture preparation (and testing) for inhalation. It issurprising that particle processing (and characterisation) has receiveda tremendous amount of attention whereas mixing and particularlydispersion have been rather neglected. This is the reason why theseitems have been given separate chapters in this manuscript to makeup for such omissions. Only one particular aspect of the mixingprocess is reviewed in this Section, which is the effect of inertial andfrictional forces during the mixing process. During mixing, carrierparticles collide with each other and with the walls, blades orimpellersof the mixer. The powder ow in a mixer also causesfrictionforces between the carrier particles. Such inertial and frictionalforces are responsible for breaking up natural drug agglomerates. Ithas been shown that naturally occurring drug agglomerates may bequite strong, at least strong enough to withstand partial break-up inan effective dry powder disperser (RODOS, Sympatec Germany) atrelative high pressures of 50 kPa [56]. Pressures of 300 kPa or moremay be necessary to obtain the primary particles. It has also beenshown that the same drug after mixing with carrier particles may stillpartially exist as agglomerates, but these agglomerates appear to beweaker than the original ones. They can be broken up into almostprimary entities already at a relatively low pressure drop of 4 kPa in aclassi er based test inhaler. The size of the drug agglomerates in theblend appears to depend on particle carrier interactions and thecarrier size fraction they are mixed with [56]. It is not clear whetherthese weakeragglomerates in the mixture are weakened fragmentsof the original drug agglomerates or newly induced ones during themixing process. The inertial and frictional forces are also responsiblefor drug distribution over the entire carrier surface. In the beginningof themixing process drug particlesare randomly distributed over thecarrier particles and they tend to be wiped together in carrier surfacediscontinuities. Spreading over the carrier surface also means a re-distribution from sites with lower adhesion force to sites with higherbinding capacity. The rate of distribution of drug particles over thecarrier surface depends on the drug concentration in the mixture [57].

When the concentration is low, re-distribution is less effective asparticles inside the carrier surface discontinuities nd shelter from theinertial and frictional mixing forces. These forces may also change themagnitude of the interparticulate (adhesion and cohesion) forces inthe mixture. It has been shown with a centrifuge technique that theforce with which particles adhere to a substrate surface depends onthe force with which the particles are pressed against this surface[58,59] . The increase of the adhesion force under applied press-onforce can be attributed to an increasing contact area as the result of plastic deformation and/or local fragmentation at the contact point.Also the distance between the particle and the substrate surface maybe decreased by smoothing out the surface roughness [58]. It couldalso be shown that the ranking of adhesion forces may be changed, asthe increase rate for these forces with increasing press-on force may

be different for different particle

substrate combinations [59]. It may

not be necessary to mention that repeatedly exerting the same force,as during mixing, has the same effect as increasing the press-on force.

The order of magnitude for the inertial and frictional forcesduring mixing depends on a number of different variables, such asthe carrier particle size distribution, the type of mixer used, themixing conditions, the batch size and the lling degree of the mixingcontainer. The ef cacy of the forces depends particularly on the drugconcentration in the mixture. To emphasise the effect of the inertial

and frictional forces coarse carrier fractions can be used. The effectsmentioned in this chapter are elucidated in Fig. 2A C. Fig. 2A showsthe residual drug on carrier (carrier residue) after a dispersion testas percent of the initial carrier payload for three different carrierfractions at 30 L/min as function of the amount of drug in the mixture.

flow rate (l/min)

flow rate (l/min)

2 min mixing time

5 min mixing time

10 min mixing time

30 min mixing time

60 min mixing time

B

C

80

carrrier payload (% w/w)

A

0

20

40

60

80

100

0 10 20 30 40 50 60

2 min mixing time5 min mixing time10 min mixing time30 min mixing time60 min mixing time

0

20

40

60

80

100

0 10 20 30 40 50 60

2 min mixing time5 min mixing time10 min mixing time30 min mixing time60 min mixing time

0

20

40

60

100

0 2 4 6 8

p e r c e n t c a r r

i e r r e s i

d u e

( % )

p e r c e n

t c a r r i e r r e s i

d u e

( % )

p e r c e n

t c a r r i e r r e s i

d u e

( % )

fine carrier

intermediate carrier

coarse carrier

Fig. 2. A. Residual drug on the carrier surface after adispersion test with an air classi erbased test inhaler at 30 L/min as percent of the initial carrier payload (percent carrierresidue: CR) as function of the initial carrier payload for three different carrier sizefractions. B. Percent carrier residue as function of the ow rate for a 0.4% budesonidemixture (carrier size fraction 250 315 m) after different mixing times (Turbula T2Ctumbling mixer at 90 rpm). C. Percent carrier residue as function of the ow rate for a4% budesonide mixture (carrier size fraction 250 315 m) after different mixing times

(Turbula T2C tumbling mixer at 90 rpm).

http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B2%80


6/18

The data were obtained with a classi er based metal test inhalerwhich retains the carrier crystals, thereby making them available forchemical analysis of residual drug on their surface after a dispersiontest [60] . At low carrier payload carrier residues are high for all carrierfractions, as the number of carrier sites with high binding capacity islarge compared to the number of drug particles in the mixture. Whenthe carrier payload is increased, the excess of drug particles relative tothe number of strong binding sites increases and more drug particles

become attached with weaker forces and are thus more easilydetached during the dispersion test. A higher drug concentrationalso increases the agglomeration potential which allows greater drugdetachment from the carrier surface due to an increased detachmentforce [61]. However, at a drug concentration in the mixture of approx.1.0% the carrier residue reaches a minimum value for coarse carrierfractions and a plateau value for intermediate fractions. This is theconcentration at which carrier surface discontinuities are saturatedand drug particles can no longer nd shelter from the inertial andfrictional forces during the mixing process. Abovetheseconcentrationdrug particles become exposed to these mixing (press-on) forceswhich has the consequence that the interparticulate forces in themixture are increased and so is the carrier residue. The end value forthe carrier residue at high carrier payloads depends on the order of magnitude for the press-on forces which obviously is highest for thecoarsest carrier fraction [60]. Fig. 2B and C underline the idea that theef cacy of the inertial and frictional forces depends on the drugconcentration. At a low drug concentration (0.4%) in a coarse carriermixture (250 315 m) there is a relatively great effect of the mixingtime on the carrier residue ( Fig. 2B). The effect is greatest at higher

ow rates ( 40 L/min). Due to the low ef cacy of the inertial andfrictional forces in re-distributing drug particles from sites with lowerbinding force to sites with higher binding force, almost all particlescan be detached from the carrier after short mixing times at 60 L/minor higher. It requires relatively long mixing times to obtain a highdegree of occupation of these sites with higher binding potential,which is expressedin an increased carrier residue at these higher owrates. At low ow rates, when only particles from weaker carrierbinding sites are detached, the effect of mixing time is less noticeable.

A payload of 4% on a coarse carrier fraction is theoreticallysuf cient to expect a multiparticulate (3-fold) drug layer around eachcarrier particle. This complete carrier coverage with drug particlesimplies that most active sites are occupied already in the early phaseof mixing. The effect of the inertial and frictional forces during mixingis therefore primarily con ned to increasing all interparticulate forcesin the blend, which appear to be maximal already after approx.5 minute mixing time. This effect is best noticeable at low ow rates:the carrier residue of 4% mixtures up to 25 L/min is higher than thatfor 0.4% mixtures for all mixing times ( Fig. 2B and C). It should bementioned that the carrier residue is a relative parameter. Its valueat 60 L/min is lower for 4% mixtures than that for 0.4% mixtures.In absolute sense, residual drug per unit carrier surface area (g/m 2)is higher for 4% mixtures [57], and this appears to be independent of

the type of drug investigated. It should also be mentioned thatthe explanation given for Fig. 2B and C has been simpli ed. In reality,the situation is more complex, as drug particle agglomeration playsa role too and a continuous multiparticulate coverage of the carriercrystals with drug particles is not achieved. From scanning electronmicrographs it is known that certain areas of the carrier surface (atthe edges of crystal planes) remain quite clean,whereas thicker layersof drug may be present in depressions or around elevations on thecarrier surface respectively.

2.4. The role of lactose nes

The positive effect obtained from combining micronised drugparticles with a mixture of coarse and ne excipient particles on the

dispersion properties of inhalation powders has been described quite

early [62]. In this patent different weight ratios and different sizeranges for the nes were claimed to increase the in vitro deposition of the drug from blends with coarse excipients. Many studies on theeffect of neshave been completed since andthey were quite recentlyreviewed extensively [63]. Different excipient particles can be used.Either they are of the same material as the carrier particles (mostlyalpha lactose monohydrate), or they are of a different nature, like thesugars glucose, mannitol, sorbitol and trehalose. Different amounts

and different particle sizes for the

nes have been used and alsodifferent mechanisms of action have been proposed, which have beencategorised into two main hypotheses [63]. The nes either occupycarrier surface areas of high adhesion or they tend to co-agglomeratewith the drug particles on the carrier surface. Supporting evidenceexists for both mechanisms but also contradicting conclusions can befound in literature. As support for the occupation of strong carrierbonding sites by added nes, the in uence of the blending order hasbeen mentioned. Mixtures in which the nes are mixed with thecoarse carrier particles rst before the drug particles are added mayshow a better dispersion performance than mixtures for which themixing order is reversed, although the effect depends on the totalmixing time [29]. This so-called corrasion or passivation of active sitesby sequential mixing is a practical application of an invention byStaniforth [36]. Support for the mechanism of co-agglomeration isprimarily obtained from SEM investigation [64]. Due tothe con icting

ndings, many aspects regarding the in uence of nes on dispersionbehaviour remain unclear. In general carriers containing greaterproportions of intrinsic nes seem to have a better performance,and the optimal median diameter for the nes may be in the rangeo f 5 t o 8 m [63]. In contrast to these conclusions drawn, norecommendations can be given for the concentration of nes in themixture, nor for the preferable material they should exist of.

3. The mixing process

Powder mixing is one of the most critical processes in the DPIcarrier-based formulations, as the aerosolisation performance isdependant not only on the formation of an adhesive mixture, butalso on the distribution of the drug onto the carrier and the interfacialforces acting between these contiguous surfaces. Considering itssigni cant role, there remains a limited understanding of howblending processes affect in-process material properties and theresulting distribution of the drug in the nal dosage form.

Within the industry, blending protocols are empirically researchedfor each speci c type of blender and are optimised across the scales,from lab to full production, in creating the desired properties in the

nal blend. The blending parameters at the production scale aretightly constrained to limit batch-to-batch variability and to achievethe requirements of the nal product, such as blend homogeneity,emitted dose and the ne particle dose. Even with these controls inplace, mixing remains a signi cant source of variability within themanufacturing process.

Unlike uid mixing, where molecules randomly diffuse aroundfrom locationsof high to low concentrations, powderparticles requiremotion to be imposed on them to initiate mixing. Thus, all powdermixers need to induce motion either by rotational movement of acontainer or the movement of an impeller within the powder.Historically, two different types of batch processing blenders havebeen used to mix carrier based DPI formulations. These are tumblingbased blenders (e.g. Turbula, V-blenders) and high speed impellermixers. These two mixing processes exhibit a different range of energy inputs, which may have a critical effect on the blendingdynamics and the adhesive properties of the drug to the lactose.

The key mode of operation of a powder mixer is to generatemechanical stress to effectively de-agglomerate the cohesive drug.High stresses are needed to break up the agglomerates so that

individual particles can be liberated to mix and distribute over the



7/18

surface of the carrier. Powder mixing is achieved via a combination of different mechanisms, namely convective, diffusive and shear mixing.Whilst all three mechanisms are likely to occur in a mixing operation,which one predominates will depend on the type of mixer, conditions( ll weight, % loading and speed) and the ow properties of thelactose. The only mechanism capable of generating the level of stresses required for de-agglomeration of cohesive drug particles isshear mixing. Shear can be generated, for example, in a tumbler mixer

along the layer of powderavalanching along a slip failure plane,or in ahigh shear mixer by the stresses created by the impeller rotating athigh speeds within the powder.

For low shear tumbler mixers, the main processing factors aremixing time ( Fig. 2B and C) and mixing speed. Both of these factorshave been shown to in uence the adhesion between drug and lactoseparticles [65]. For high shear mixers, the main process parametersof xed blade geometry blenders is the rotational velocity of theimpeller (n rpm ) and mixing time (t) [66]. In addition to thesevariables, the torque required to turn the impeller in the powdermix can be continually measured to determine the energy input (E in )to the powder by the impeller via:

E in = t n rpm

30 1

where is the torque (Nm).Increasing the energy input in a high shear blender has been

shown to directly impact the particle size distribution of the lactoseand particulate interactions and, thus, may have a pronounced affecton product performance [67]. Bridson et al. have shown that theenergy input and the design of the impeller had a signi cant effect onthe particle size distribution of coarse lactose [67]. They found thatwith increasing input energy there was a loss of lactose nes. Theyalso found that the conditions of storage of lactose prior to blendingdirectly affected the outcomes of the post blending PSD of lactose.These factors will directly affect the ow and uidisation behaviour of the resultant DPI blend. Begat et al. compared the effects of impellerspeed and input energy on the structure and aerosol properties of carrier based DPI formulations [66]. Their data indicated that thedifferent blends exhibited variations in blend structure with respectto the presence of drug agglomerates and the distribution of the drugon the carrier. Furthermore, increasing the input energy of the mixerby either increasing the impeller speed or the mixing time indicatedthat higher input energy decreased the ne particle dose of the drug,suggesting that the increasing the input energy of the mixer mayincrease the adhesive forces in the blend.

There is an increasing trend within the industry to adopt a qualityby design (QbD) approach for innovative process manufacturingand quality assurance. While there remains a paucity of data, on thecorrelation between mixing parameters, formulation structure,interfacial forces and aerosolisation performance, advances in thescience of powder mixing will be limited. This lack of knowledge

limits the move of a paradigm of testing quality in post manufacturingto designing quality during process manufacturing. The introductionof imaging and monitoring technologies within processing mixersmay provide some useful insight into the optimum conditions foref cient de-agglomeration of cohesive drug during powder mixingand formulation structure. However, these chemical based imagingtechniques may not ultimately provide the critical information on theeffects of blending on the functionality of the nal product.

4. Interactions between variables

It is now fully understood that the ne particle fraction obtainedfrom adhesive mixtures during inhalation depends on a great numberof variables relating to the drug and carrier particle properties as

well as to the mixing process. Yet, the relevance of some major

determinants is still often ignored, that of others still unknown. Theyinclude for instance the carrier payload with the drug, the dispersionprocess (particularly, the type of de-agglomeration forces applied)and the aerosol characterisation technique. These parameters canchange the outcome of a study completely or result even in oppositeconclusions when being chosen differently. Some examples are givenin this chapter. Therefore, a major challenge for future research is toinvestigate for which variable(s) the effect(s) depends on:

linkage to other variables ( linked effects ), a speci c quality or property of the variable ( conditional effects ),and/or

the choice made for one or more of the other variables ( interacting effects).

In addition, there may be interplay in more than one way betweenthese linked, conditional and interacting effects,which are categorisedin this chapter only to emphasise the complexity of the relationshipsbetween the variables. When the relevance of certain variables canbe assessed for a particular study, investigating the complex relation-ships may bene t from the use of multivariate analysis of variance(MANOVA). The selection of an appropriate set of variables requiressome knowledge of the conditions under which the variables arerelevant or not and this requires some knowledge of the mechanismsthat are involved too. The practical implementation of multivariatestatistics to such a complex system may include the use of differenttypes of univariate and multivariate analyses however. An additionalmajor challenge for future research in this respect is the search forbetter characterisation methods for many of the variables involved.Having improved qualitative and quantitative expressions for vari-ables like the carrier surface rugosity andimpurity is a prerequisite forcontrolling andstudyingtheireffect in dependence of thetotal systemof variables.

4.1. Linked effects

Fig. 3 shows a scheme of the most relevant variables and the wayin which they can in uence each other. Each of the variablesmentioned represents several properties, like for instance size andshape distribution, water content, hygroscopicity, anomeric compo-sition and surface rugosity of the drug and carrier particles. Fig. 3 alsoshows an example of linked effects (indicated with black arrows).When the size distribution of the carrier particles is changed, thecarrier surface (scale of rugosity and amount of impurities per unitcarrier surface area: see Section 2.2) and bulk properties becomedifferent. Also the carrier surface payload with the drug particles(gramme drug per square metre carrier surface) is changed. This canhave an effect on the degree of saturation with drug particles for the

Fine particle dose (FPD)

Carriersurfacepayload

Mixture properties (drug distribution over and attachment to

the carrier surface)

Drugproperties

Conditioning

Carrier surfaceproperties

Carrier bulkproperties

Inhalationprocess*

Carrier size distribution

Mixingprocess

Storage and conditioningFine particle dose (FPD)

Carriersurfacepayload

Mixture properties (drug distribution over and attachment to

the carrier surface)

Drugproperties

Conditioning

Carrier surfaceproperties

Carrier bulkproperties

Inhalationprocess*

Carrier size distribution

Mixingprocess

Storage and conditioning

Fig. 3. Scheme with variables having an effect on the preparation and dispersionprocesses of carrier-based dry powder formulations for inhalation. The black arrows

indicate the effects linked to the carrier size distribution.

http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B3%80


8/18

carrier surface discontinuities and the strongest carrier bonding sites.A change in the bulk ( ow) properties may cause a change in theorder of magnitude for the inertial and frictional forces during themixing process. This can all in uence the ef cacy with which drugagglomerates are broken down (or softened) and/or newly induced,drug particles are (re-)distributed over the carrier surface andcohesive and adhesive forces in the mixture are increased duringthe mixing process (Section 2.3). Obviously this re ects on the

mixture properties. Changing the carrier

ow properties may alsohave an effect on emptying of the dose system of the inhaler used,whereas circulation and residence time of the powder in whirlchamber, cyclone or classi er based dispersion systems may becomedifferent too. Thenet effect of all these changes maybe a different neparticle fraction but the contribution of each of the linked variablesindividually to this net effect cannot be assessed properly. It istherefore paramount that the effects of variables are investigated in acarefully de ned context of all the other parameters that may have aneffect on the ne particle fraction.

4.2. Conditional effects

An example of a conditional effect is the in uence of carrierparticle surface rugosity on the drug particle attachment to the carriersurface and dislodgment during inhalation. Early studies showed thatthe re-dispersion of drug particles from adhesive mixtures isfacilitated if the rugosity of the carrier particles is reduced [68].Limits to numerical values for the rugosity were set based on surfacecharacterisation methods such as air permeametry (Kozeny Carman)and gas absorption (BET-method) without taking regard of the scaleof the surface discontinuities. In later studies, carrier surfacemorphology was studied more in detail with scanning electronmicroscopes and image analysis techniques for different marketedlactose products with different surface smoothness [69]. Fromdispersion experiments with pranlukast hydrate-carrier mixtureswith these lactose products,it wasconcluded that a surface roughnesson a scale smaller than the diameter of the adhering drug particles(nano- or microrugosity) has a positive effect on the drug fractionreleased from the carrier surface, whereas surface pores, clefts anddiscontinuities larger than the drug particles (macrorugosity) have anegative effect. This was explained by reduced contact area andincreased distance between the drug and carrier particle for carriersurfaces exhibiting nanorugosity, which reduces the magnitude of thevan der Waals force. For the large scale discontinuities mechanicalinterlocking was proposed as mechanism for increasing the drug-to-carrier bond. It is dif cult to draw conclusions from such studieshowever, as differences in surface structure are linked to differencesin polymorphic form. For carrier surface discontinuities which aremuch larger than the drug particles a reduced ef cacy of drugaerosolisation by drag and lift type of removal forces was shown [70],although this may depend on the carrier particle size [71]. Largecarrier pores and clefts are areas where multiple contact points

between drug and carrier may occur and where impurity concentra-tions from the mother liquor are highest as such pores are lled withliquid when lactose crystals are taken from the crystallisation tank(Fig. 1A) [40]. In conclusion, the effect of the variable carrier surfacerugosity on the ne particle fraction obtained seems to depend on thesize of the surface irregularities relative to the size of the drugparticles.

4.3. Interacting effects

Referring to the nal conclusion in the previous Section 4.2 , thegeneral idea exists that carrier surface discontinuities on a scale largerthan the diameter of the adhering drug particles are not bene cial fordrug re-dispersion during inhalation. This idea has been supported by

a series of different studies making use of marketed inhalers like the

Diskhaler [32], Spinhaler [72,73] , Rotahaler [29,33,74] and Pulvinal[75]. Different carrier rugosities for comparative evaluation wereeither obtained by selection or by carrier surface modi cation usingtechniques like re-crystallisation from carbopol, ethanol treatmentor special particle smoothing processes in high-speed mixers. Onlyone example is known in which the bene t of a high rugosity (of roller dried anhydrous beta-lactose) has been mentioned [76]. In thispatent the use of the Miat inhaler is described. This type of inhaler

has a different dispersion principle than classic capsule (Spinhaler,Rotahaler, Cyclohaler) or blister (Diskus) inhalers: it consists of ahelical element in the mouthpiece. This element causes impaction of the particles passing through this mouthpiece. More recently, datahave been presentedthat support this claim andgive reasonto believethat carriers with large surface cavities can be bene cial indeed wheninhalers are used that generate inertial separation forces [77]. Forthis study a classi er based test inhaler was used which retains thecarrier crystals during the dispersion test. The degree of drug particledetachment from the carrier surface was determined by chemicalanalysis of the residual amount of drug on retained carrier crystalsafter the test (carrier residue). Fig. 4 shows the carrier residues forfour different granular carriers with large surface cavities and onecrystalline carrier with muchsmaller surface irregularities in the samesize fraction 250 355 m as function of the carrier payload withmicronised budesonide at 30 L/min from thetest inhaler. Thegranularcarriers were prepared from crystalline products with decreasingmedian particle diameters with the ranking 100 M, 325 M, 200 M and450 M. Fig. 4 shows that the bene t of a granular carrier structure is

rst obtained at a carrier payload of 0.4% and higher. The explanationof this gure may be complex. It is likely that at low carrier payload( b 0.1% drug) a large portion of the drug particles in the mixture iseither attached to strong bonding carrier sites or to sites whereremoval forces are relatively ineffective. Such sites are predominantlywithin the carrier surface irregularities inside which drug particlesarewiped together during the mixing process. When the carrier payloadis increased, the number of drug particles relative to the storagecapacity of the carrier irregularities is increased, and more particlesare attached to weaker binding sites or sites where removal forcesare more effective. In addition, the drug particle concentration onthe carrier surface is increased and this increases the potentialfor drug particle agglomeration. Consequently, drug particles can bedetached more easily during the dispersion test and the carrierresidue decreases. This is more or less the same for all carrier types.When the carrier payload reaches values at which the surfacedepressions become saturated with drug, more drug particles becomeattached to places where they arein reach of the inertial and frictionalmixing forces. This, as has been explained in Section 2.3, increases theinterparticulate forces in the mixture. It occurs at approx. 0.4%

0

20

40

60

80

100

0 1 2 3 4

carrier payload (%)

c a r r

i e r r e s i

d u e

( % )

Crystalline (80 M)

Granular (100 M)

Granular (200 M)

Granular (325 M)

Granular (450 M)

Fig. 4. Effect of the carrier surface structure (pore size and storage volume) on the

percent carrier residue as function of the initial carrier payload.

http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B4%80


9/18


10/18

be as important as they have always been regarded. Only a fraction of the dose is attached to these sites as de ned above and liberation of drug particles from such sites becomes rst of interest at very highdispersion energies. Fig. 6 shows that the ow rates through theclassi er based test inhaler above which no further drug particledetachment takes place (50 70 L/min) correspond with pressuredrops of respectively 8 and 16 kPa. This is far beyond the range of attained pressure drops through marketed inhalers in daily practise.Because these marketed inhalers are also less ef cient in dispersion,the situation at which drug particles have to be detached from thesemost active sites is unlikely to ever be achieved with these devices.This makes the relevance of the so-called active sites in drug-to-carrier interactions questionable when inhalers with low dispersionef cacy are used, but this does not imply that carrier surfaceproperties are irrelevant. The carrier residue decreases over a widerange of ow rates, corresponding with a wide range of separationforces ( Fig. 6) which implies that a wide distribution of adhesionforces has to be overcome. Although this seems to be primarily aconsequence of the distribution of the particle size of the drug [79], itis partially also the result of drug (re-)agglomeration on the carriersurface, the effect exerted by the inertial and frictional mixing forceson the interparticulate forces in the blend, the variation in carrierbinding sitesand the difference between cohesive and adhesiveforcesin the mixture. The contribution of most active binding sites to allthese in uences may be relatively small.

More recently, active sites are expressed in terms of surface freeenergy. As discussed in Section 2.2 , values presented for the surfacefree energy of different types of lactose obtained with inverse gaschromatography measurement typically range from 40 to 50 mJ/m 2

[41,48,49] . This equals 40 to 50 10 9 J/ m2 which is approximatelythe order of magnitude for the contact area of a single drug particle. If this surface free energy is relevant to the interaction forces betweenthe carrier surface and drug particles, there must be a relation withthe separation forces needed to dislodge a drug particle from thecarrier surface. Such separation forces have been measured withatomicforce microscopy (AFM) and they have the order of magnitude

of 1 to 1000 J for a single drug particle in the micron range(depending on the relative humidity) [39]. So, presented orders of magnitude for the surface free energy and the separation energydiffer by a factor of 10 5 to 10 8 . Therefore, it may be questionedwhether differences in drug-to-carrier interaction forces can bepredictive for differences in surface energy for the carrier fraction.It has been proposed that surface energy in itself is not the parameterto consider. Surface energy should rather be used as an indicatorof other variability, like impurity, chemical heterogeneity pro leor surface disorder content at the nanometer level (includingamorphicity, nano-crystallinity, polymorphism, etc.) [42]. This is aninteresting approach. However, if relevant differences in ne particlefractions between different carrier formulations are obtained, whichare likely the effect of carrier surface (and not of mixing) variability,

then ne particle fraction itself is the best indicator particularly

when it is taken into account that surface energy is not a propertythat varies considerably from one batch of material to the next [42].And either way, the relation to carrier surface parameters remainsto be investigated. Interpretation of surface energy data may becomplex as the effects obtained may depend on testing conditions.An example is given in Fig. 7A showing the carrier residue as functionof the ow rate for 0.4% budesonide mixtures and a coarse carrierfraction of 250 315 m. Different mixtures were prepared with

different amounts of lactose

nes having the same particle sizedistribution as the drug. The lactose nes have a higher surfacefree energy ( 301K =44.9 mJ/m 2) than the coarse carrier particles( 301K =38.2 mJ/m 2) which could be the result of a reduction inthe degree of structural order during the micronisation process.When the carrier residues for the drug particles of these mixturesare plotted as function of the surface energy of the carrier blends(Fig. 7B), ne linear correlations are obtained for all ow rates.However, the slopes of the linear curves change from positive at 20and 30 L/min to negative at 40 and 50 L/min and this suggests thatother parameters or effects must be involved.

5.2. On the role of amorphous spots

Milling of particles causes local distortion of the lattice structureof crystalline materials. The distortions are referred to as amorphousspots and it has been shown that the presence of such amorphousspots from milling increases the surface energy [50]. On the basisof this increase in surface energy the expectation of an increasedinteraction force with adhering drug particles seems justi ed.Therefore, amorphous spots are considered as active sites [80] andeffort is put in techniques to quantify small amounts of amorphousmaterial in crystalline carriers [81 84]. It is very dif cult to assess the

Table 1Budesonide carrier residue values after dispersion at 60 L/min in a classi er based testinhaler (CR 60 ) and real residual mg drug per square metre carrier, both as function of the mixing time in a Turbula (T2C) tumbling mixer (90 rpm, batch size 25 mg in a160 cm 3 stainless steel mixing container) for two different carrier payloads (0.4% and4%) on a coarse carrier fraction 250 315 m.

2 min 5 min 10 min 30 min 60 min

A. percent CR600.4% mixture 3.5 8.1 13.0 16.8 24.4

4% mixture 0.9 1.7 1.8 3.2 5.2

B. real residual amount (mg) of drug per square metre carrier surface0.4% mixture 10.14 23.19 37.68 48.55 71.014% mixture 26.09 49.28 52.17 92.75 150.72

A

B

0

20

40

60

80

100

0 20 40 60 80

flow rate (l/min)

c a r r

i e r r e s i

d u e

( % )

c a r r

i e r r e s i

d u e

( % )

no fines

1% lactose fines

2% lactose fines

3% lactose fines

4% lactose fines5% lactose fines

10% lactose fines

0

20

40

60

80

100

25 30 35 40

surface free energy at 333 0K (mJ/m 2)

20 l/min

30 l/min

40 l/min

50 l/min

Fig. 7. A. Percent carrier residue as function of the ow rate for mixtures of 0.4%budesonide on a coarse carrier fraction 250 315 m in the presence of differentamounts of ne lactose particles having the same size distribution as the drug.B. Percent carrier residue at 20 50 L/min plotted as function of the surface free energy

of the carrier blends.

http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%80http://localhost/var/www/apps/conversion/tmp/scratch_7/image%20of%20Fig.%E0%B7%8

Documents

A critical view on lactose-based.pdf