The Influence of Mechanical Processing of Dry Powder.pdf

7/26/2019 The Influence of Mechanical Processing of Dry Powder.pdf

http://slidepdf.com/reader/full/the-influence-of-mechanical-processing-of-dry-powderpdf 1/11

PHARMACEUTICAL TECHNOLOGY

The Influence of Mechanical Processing of Dry PowderInhaler Carriers on Drug Aerosolization Performance

PAUL M. YOUNG, HAK-KIM CHAN, HERBERT CHIOU, STEPHEN EDGE, TERENCE H.S. TEE, DANIELA TRAINI

Advanced Drug Delivery Group, Faculty of Pharmacy, University of Sydney, Sydney, NSW 2006, Australia

Received 10 May 2006; revised 18 June 2006; accepted 10 July 2006

Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/jps.20933

ABSTRACT: The influence of processing on the performance of carrier material used in

dry powderinhalers was investigated.a-Lactose monohydrate crystals were processed by

ball milling for cumulative time durations and their properties evaluated. As expected,

milling reduced the median particle diameter while increasing fine particulate (<10 mm)

and amorphous levels. Recrystallization of these partially amorphous samples resulted

in a reduction in fines, elimination of amorphous material with little change in median

diameter. To study the effects of processing on aerosolization performance, blends of

lactose monohydrate with a model drug (nedocromil sodium trihydrate), were evaluated

using an in vitro multistage liquid impinger (MSLI) model. In general, milling and

storage of the carriers at high humidity (prior to blending) had a significant (ANOVA,

p<0.05) effecton thefine particle fractions (FPF;<6.8 mm).These effects were attributed

predominantly to the fines content, showing a strong correlation between increased fines

and FPF ( R2¼0.974 and 0.982 for milled and recrystallized samples, respectively).

However, this relationship only existed up to 15% fines concentration, after which

agglomerate-carrier segregation was observed and FPF decreased significantly. These

results suggest that, after processing, high-dose drug formulation performance is

dominated by the presence of fines. 2007 Wiley-Liss, Inc. and the American Pharmacists

Association J Pharm Sci 96:1331 –1341, 2007

Keywords: milling; surface activation; fines; aerosolization; dry powder; inhalation;

recrystallization; amorphous

INTRODUCTION

Dry powder inhalers (DPI) are a novel route for

drug delivery, with the capability of targeting disease states both locally (in the case of lung

diseases such as asthma), and systemically (e.g.

in the delivery of proteins and peptides). For

effective deposition in the lower airways and deep

lung, drug particles with aerodynamic particle

sizes of <5 mm are required. However, such

systems are highly cohesive due to the high

surface area to mass ratio of the particulates.

Cohesive systems pose a problem for the desir-

ed deaggregation of particles as uncontrolled

agglomeration occurs naturally. Subsequently,such agglomeration may lead to formulation

variations and a decrease in DPI efficacy. As a

consequence, large inert carrier systems are

employed as one method to overcome this pro-

blem, where the micron sized drug particles are

blended with larger inert material to reduce

agglomeration, improve flow and act as a diluent.

Ideally, during inhalation, the drug particles are

liberated from the carrier to penetrate the lower

airways while the carrier impacts on the ortho-

pharynx and is swallowed. As with all these

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 96 , NO. 5, MAY 2007 1331

Correspondence to: Paul M. Young (Telephone:þ61 2 90367035; Fax: þ61 2 9351 4391; E-mail: [email protected])

Journal of Pharmaceutical Sciences, Vol. 96, 1331–1341 (2007) 2007 Wiley-Liss, Inc. and the American Pharmacists Association



systems, it is important to note, however, that the

energy supplied by the patient during inhalation

needs to be sufficient for liberation and aerosoli-

zation of the drug particulates.

Currently, lactose is the most popular material

approved for use in inhalation as carrier. A

naturally occurring disaccharide sugar, lactose is

found in milk, and is collected as a by-product

of whey produced during the manufacturing of

cheese. Lactose monohydrates for use in the

pharmaceutical industry, are produced through

precipitation from aqueous solution and is avail-

able in many particle size grades.1Furthermore,to

obtain inhalation grade lactose monohydrate, the

crystals are processed through techniques such as

milling and sieving to produce the desired carrier

size characteristics.2 However, due to the indus-

trial scale of such processes, inter-batch, and inter-

supplier variations in the physicochemical proper-ties of the carriers have been observed, leading to

variations in the aerosolization performance of the

final formulations.2 Such batch-to-batch variation

is most likely due to differences in fine particle

content, particle size distribution, surface mor-

phology, and amorphous content.

Intrinsic carrier particle size has been believed

for many years to play an important role in the

performance of DPI.3– 6 Following a reduction in

physical particle diameter, a decrease in mean

mass aerodynamic diameter (MMAD) is observed.

In general, a decrease in MMAD has been asso-ciated with an increase in drug dispersion. How-

ever, a recent study by Islam et al.,7 suggested

that the volume median diameter of carriers had

no significant effect on drug dispersion when

studied without the influence of fines content.

Earlier studies did not specify the fine particle

content, thus with a reduction in particle size, the

increasing concentration of fines may have been

masking the effect carrier size has on aerosol

performance. As fine particles are commonly

introduced inadvertently during the comminution

process, the influence of fines (e.g. carrier particles

with a diameter <15 mm) on DPI performancehas been studied extensively. Lucas et al.,8

Zeng et al.,6,9 Louey et al.,10,11 and Islam

et al.,7,12 have all demonstrated an improvement

in dispersion with the presence of fine particles.

More recently, Steckel et al.,13 indicated similar

findings, suggested that variation in small quan-

tities of fines (<5% below 15 mm diameter) in

different sized sieved fractioned lactose formula-

tions significantly influenced drug aerosolization

performance.

As discussed, processing methods such as

milling may induce variation in surface morphol-

ogy or result in increased amorphous content.

Surface morphology has been demonstrated to

directly influence the contact area between drug

particle and carrier, leading to variations in inter-

particulate adhesion. Several studies have

reported that variations in contact area, as a

result of differing surface structure, could poten-

tially compromise the aerosolization performance

of the drug particles.14–16 In addition, the intro-

duction of amorphous material during high energy

mechanical processes is associated with a higher

surface adhesion energy compared to crystalline

surfaces.17–19 As a consequence of raised adhesion

energy, poor deaggregation of drug particles is

observed.20 No studies have yet been reported in

relation to the direct impact of amorphous content

on drug dispersion from carriers. However, thepresence of amorphous material may cause pro-

blems, for example, due to the fusion of particles,

resulting in poor dispersion.19–21

Milling is commonly used for processing of

powders in the pharmaceutical industry, and in

particular for inhalation drugs and carriers. When

carriers are milled, various changes to the physical

properties are induced. This is important since the

reliability of the DPI product mainly depends on

batch-to-batch consistency of the lactose monohy-

drate carrier. However, as mentioned previously,

variations between batches do occur, and thisstudy was initiated to investigate the effect of any

material changes induced by milling on DPI

performance. Further, as part of an ongoing study,

the influence of storage at high humidity prior to

blending was also investigated. The influence of

these processes on DPI aerosolization efficiency

was investigated using nedocromil sodium tri-

hydrate as a model drug system, and was corre-

lated with the physical properties of the carrier

particles.

MATERIALS AND METHODS

Materials

Micronized nedocromil sodium trihydrate (NST)

was obtained from Sonafi-Aventis (Cheshire,

England). Crystalline a-lactose monohydrate

(Lactochem1 crystals) was obtained from Borculo

Domo (Zwolle, The Netherlands). Water was

purified by reverse osmosis (MilliQ, Millipore,

Molsheim, France). Analytical grade chloroform

1332 YOUNG ET AL.

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 9 6, NO. 5, MAY 2007 DOI 10.1002/ jps



and n-octane were obtained from Biolab (Victoria,

Australia) and Fluka (Germany), respectively.

Preparation of Lactose Monohydrate Samples

Mechanical treatment of lactose monohydrate

samples was achieved by comminution in a

small volume ball mill (approximately 1 L)

containing 60 ceramic balls (mean diameter of

19.30.7 mm). Samples of lactose monohydrate

(approximately 100 g) were weighed into the ball

mill which was rotated at 42 rev min1 for

durations of 10, 20, 30, 40, 50, and 60 min. Each

sample was then collected and stored in tightly

sealed containers over phosphorus pentoxide

prior to sampling or blending.

In addition, to crystallize any amorphous

material, present in the milled samples, approxi-

mately 10 g of each milled lactose monohydratesample was transferred onto glass Petri-dishes

and stored (3 weeks) in tightly sealed containers

with a saturated solution of potassium chloride

(relative humidity, RH, of 85%). The samples were

regularly stirred to ensure moisture penetration

into the powder bed. After 3 weeks, each sample

was removed and transferred into containers with

phosphorous pentoxide (0% RH) for a minimum of

24 h before sampling or blending.

Physical Characterization

Particle Sizing of Processed Lactose Monohydrate Samples and NST

Particle sizing was performed by laser diffraction

(Malvern Mastersizer S, Malvern Instruments

Ltd., Malvern, UK) using a 300RF lens and auto-

mated small volume dispersion sampling unit.

Approximately 200 mg of each lactose monohy-

drate sample or NST was dispersed in about

10 mL of chloroform, and added drop-wise into the

sampling unit containing chloroform until an

obscuration between 10% and 30% was obtained.

Size distributions were based on 2000 sweeps foreach sample, with refractive indices of 1.533,

1.358, and 1.444 for lactose monohydrate, NST

and chloroform, respectively. Each sample was

analysed in triplicate.

Scanning Electron Microscopy

Visualization of lactose monohydrate surface

morphology and the uniformity in blends were

investigated using scanning electron microscopy

(SEM) (XL30, Philips, Japan) at 10 keV. Each

sample was mounted on a carbon sticky tab and

platinum coated (10 nm thickness; Edwards

E306A Sputter Coater, UK), prior to analysis.

Images were obtained at magnifications of 1000

and 5000.

Amorphous Content Quantification of Lactose Monohydrate

Organic Dynamic Vapor Sorption (Organic DVS)

was used to quantitatively determine the amor-

phous content in the milled lactose monohydrate

samples using a method described elsewhere.22

In simple terms, the technique measures the

adsorption of a dispersive molecule (n-octane) into

the surface of a sample as a function of partial

pressure. Since the relative adsorption of the

molecule into amorphous and onto crystalline

samples will vary, a calibration curve may be con-

structed by comparing the relative adsorption inblends of 100% crystalline and 100% amorphous

samples. From this, the amorphous content of an

unknown sample may be determined. Measure-

ments were conducted using a DVS-1 (Surface

Measurement Systems, Alperton, UK), at 258C,

using n-octane as the organic probe. Approxi-

mately 100 mg of lactose monohydrate was

weighed into the sample pan and exposed to a

two-step octane partial pressure ( p / po) cycle of

0–90%. Equilibrium at each step was deter-

mined by a dm /dt of 0.0002% min1. Each milled

sample (n¼ 5), and the amorphous content calcu-lated from the calibration data reported else-

where.22 In addition, a sample of the 60 min

recrystallized sample was analysed to ensure full

recrystallization.

Drug Content Determination

Quantification of NST content uniformity and

in vitro deposition was determined by UV spectro-

photometry (U-2000 spectrophotometer, Hitachi,

Japan) at a wavelength of 376.5 nm. Samples

were prepared and diluted appropriately in water.

The calibration plot for NST was linear over therange 0.5– 50.0 mg mL1 ( R2

¼1.00). Lactose

monohydrate did not interfere with the analysis

at the wavelength used.

Dispersion Studies

Preparation of Blends

The influence of carrier milling on the drug

aerosolization efficiency was evaluated using 5%

w/w blends of NST. Blends of 1 g were prepared

DRUG AEROSOLIZATION PERFORMANCE 1333

DOI 10.1002/jps JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 96, NO. 5, MAY 2007



by geometric hand mixing 50.00 mg NST with

950.00 mg of lactose monohydrate sample in a

glass mortar using a spatula. Blend homogeneity

was performed by analysing 35.001.00 mg

samples of each blend (n¼5, due to the relatively

small blend size) five times for each powder

mixture. In all cases, an acceptable degree of

homogeneity was achieved with the mean drug

content across all blends being within 100.0

3.0% of the theoretical value and each blend

exhibiting a coefficient of variance <5.0%. Ap-

proximately 35.001.00 mg samples of each

blend were manually filled into size 3 hard

gelatine capsules (Capsugel1, NSW, Australia)

and stored at 45% RH and 258C for 24 h prior to

testing.

In Vitro Aerosolization Studies The influence of the carrier humidity conditioning

on the aerosolization of NST from the freshly

milled and milled-recrystallized lactose monohy-

drate carrier was investigated using apparatus C,

the MSLI (Copley Scientific, Nottingham, UK),

according to the method described in the British

Pharmacopoeia 2005. Briefly, the apparatus con-

sisted of a USP throat, four stages, each stage

containing 20 mL of water, and a filter stage,

which, at a flow rate of 60 L min1 produces

MMAD cut-off points of 13.0, 6.8, 3.1, and 1.7 mm

for stages 1, 2, 3, and 4, respectively. The flowrate through the MSLI was controlled by a GAST

rotary vein pump and solenoid valve timer

(Copley Scientific). Prior to testing, a 60 L min1

flow rate through the MSLI was set using a

calibrated flow meter.

The aerosolization performance of each blend

was investigated using a CyclohalerTM DPI

(Novartis, Surrey, UK). Briefly, a capsule of the

formulation was placed into the sample compart-

ment of the CyclohalerTM, inserted into a specially

constructed MSLI mouthpiece adapter, primed

and tested at60 L min1 for 4 s. A 3 s delay prior to

testing was instigated to allow the pump to settle.Deposited drug fractions were collected from the

DPI and MSLI stages using water. In addition,

the extracted solution for the filter stage was

filtered through a 0.22 mm PVDF filter (Millex GV,

Millipore, Billerica, MA) to remove traces of the

glass filter. The amount of active contained in each

aliquot was determined by UV spectrophotometry

using the method described previously. The FPF

was defined as the total amount of NST particles

deposited in stages 3, 4 and filter (corresponding to

particles with an MMAD<6.8 mm) as a percentage

of the total recovered dose. The total recovered

dose (loaded dose) was calculated as the total

amount collected from the inhaler, throat, and all

stages of the MSLI. Emitted dose was calculated as

total recovered from all stages, postdevice.

All samples were evaluated in triplicate and

were randomized for formulation. Temperature

and humidity during the in vitro investigation was

258C and 45% RH, respectively.

RESULTS AND DISCUSSION

Particle Sizing of Processed Lactose MonohydrateSamples and NST

Particle size analysis of NST gave median

diameter of 1.10 mm with 90% particles less than5.40 mm, suggesting the model drug was suitable

for inhalation and DPI studies.

The influence of milling time on the particle size

distributions of both freshly milled and recrystal-

lized lactose monohydrate samples was investi-

gated. As expected, the milling process resulted in

a significant reduction (ANOVA, p<0.05) in the

median particle diameter with respect to time

(Fig. 1). Such observations are in agreement

with previous investigations.23–25 In general, a

decrease in median diameter from 1151 mm for

the untreated lactose monohydrate to 631 and

651 mm for freshly milled and recrystallizedlactose monohydrate samples was observed,

respectively. It interesting to note however, that

incremental increases in mill time resulted in an

Figure 1. Influence of mill time on the median

particle diameter for freshly milled (*) and recrystal-

lized milled (*) lactose monohydrate samples.

1334 YOUNG ET AL.




exponential decrease in median diameter ( R2¼

0.996). Again, such observations are expected

since the probability of individual particles being

involved in fracturing processes diminishes as the

particles become smaller.25,26 The mean energy

required to cause fractures increases with deple-

tion in crystal cleavage planes while the magni-

tude of local stress from contact with the milling

material (in this case, ceramic balls) decreases. In

addition, increasingparticle aggregation following

particle size reduction may occur. As a conse-

quence, energy may be expended in breaking up

the aggregates instead of the particles. In such

cases, particle size reduction will cease to occur at

some practical limit.25 Secondly, as the particles

become smaller and more numerous, friction

diminishes and the sample may behave as a

semi-solid. Larger particles may arch and protect

smaller particles from impact, whilst smallerparticles coat the grinding medium and cushion-

ing the larger particles from impact. This ‘‘protec-

tion’’ may prevent further particle size

reduction.27

Representative particle size distributions for

the freshly milled and recrystallized lactose

monohydrate samples are shown in Figure 2A

and B, respectively. From Figure 2A it can be seen

that large variationsin the size distributionof both

freshly milled and recrystallized lactose monohy-

drate exist, particularly in relation to the small

particle fractions. To further investigate this, thevariation in percentage particles less than 10 mm

(classified as fines for the purpose of this paper)

with milling time was studied. Analysis of the fines

concentration with respect to mill time indicated a

significant (ANOVA, p <0.05) increase from 4.4%

to 18.0% between 0 and 60 min mill time. Again,

such observations are expected, since it has been

suggested that cleavage planes, commonly found

in crystals, fracture into many fine particles. This

results in few relatively large particles, a number

of fine particles and relatively few particles of

intermediate size.27

In comparison, particle size distribution of therecrystallized milled samples suggested a reduced

rate of increase in the fines percentage with

milling time (4.4–10.1%, between 0 and 60 min

mill time). Such observations aremost likelydue to

‘‘fusion’’ between the mill-induced fines and larger

lactose monohydrate particles. Fusion of particles

may be achieved by two possible mechanisms,

solid– liquid bridge formation and/or particle

fusion through amorphous recrystallization. For

example, storage of samples at elevated humid-

ities may allow water vapor to condense in the

capillaries that exist between individual, with

increased levels of fine particles resulting in an

increased number of capillaries.28,29 Furthermore,

highly soluble materials, such as lactose monohy-

drate, may undergo limited dissolution at inter-

particulate contact points with subsequent solidi-

fication, thus resulting in solid– liquid bridge

formation between particles, leading to particu-

late fusion.30 Similarly, such particle fusion couldbe facilitated by the presence of amorphous

content on particulate surfaces recrystallising at

elevated humidity.19,21

Scanning Electron Microscopy

Representative SEM images of the 0, 30, and

60 min mill time freshly milled lactose mono-

hydrate samples are shown in Figure 3A–C,

respectively. As expected, images were in good

Figure 2. Particle size distributions of (A) freshly

milled lactose monohydrate and (B) recrystallizedmilled samples.





correlation with the particle size analysis, con-

firming that increasing the mill time resulted in

both a decrease in median particle diameter with

a concurrent increase in fines material. Repre-

sentative high magnification images of the freshly

milled and recrystallized mill samples are shown

in Figure 4A and B, respectively. From Figure 4A,

it can be seen that the samples from the 60 min

mill process exhibit discrete fine particulates

distributed across the surface of the larger lactose

monohydrate particles. In comparison, analysis of

the recrystallized 60 min mill time samples

(Fig. 4B), suggested many of the fine particulates

had become ‘‘fused’’ to the large lactose monohy-

drate particulate surfaces. Again such observa-tions are in good agreement with particle size

analysis discussed previously.

Amorphous Content

As previously discussed, the degree of amorphous

content present in the milled lactose monohydrate

samples was determined using a novel organic-

DVS technique. As expected, an increase in mill

time resulted in an exponential increase in

Figure 3. Representative SEM images of freshly

milled lactose monohydrate samples after (A) 0 min,

(B) 30 min, and (C) 60 min mill times.

Figure 4. Representative SEM images of (A) 60 min

freshly milled lactose monohydrate blend and(B) 60 min

recrystallized lactose monohydrate blend.

1336 YOUNG ET AL.




amorphous content (Fig. 5) ( R2¼0.999), correlat-

ing with the concurrent decrease in median

diameter observed by size analysis. Again, such

observations may be expected, since amorphous

content was introduced into the sample by surface

molecular modification during the milling pro-

cess.17 As the degree of comminution decreases, it

is logical to conclude that the introduction of

surface molecular damage would follow suit. In

comparison, organic-DVS analysis of the 60 minmill time recrystallized sample suggested a

completely crystalline material (0.0% amorphous

content). Since the milling process, would most

likely induce amorphous domains on the surface

of the lactose, differences in interfacial forces

between NST and the freshly milled or recrystal-

lized system should exist. The freshly milled

sample, under the experimental conditions used

here, would be thermodynamically unstable and

would have surface amorphous domains with a

degree of molecular mobility relative to the

environmental conditions (45% RH). Clearly,

under such conditions, the force of interactionwould be higher and result in a reduced FPF

when compared to the re-crystallized lactose

system.

In Vitro Aerosolization Performance

The aerosolization performance of NST from

blends of milled and recrystallized lactose mono-

hydrate was studied using a MSLI. Analysis of the

deposition data suggested milling resulted in no

significant difference in either loaded or emitted

dose across all mill times and for either freshly

milled or recrystallized lactose monohydrate

samples (ANOVA, p>0.05). Mean loaded doses

of 161479 and 160281 mg and emitted doses of

126458 and 127441 mg were observed for

freshly milled and recrystallized samples, respec-

tively. Since no difference in loaded or emitted

dose was observed between all samples, the

influence of milling and recrystallization of lactose monohydrate carriers on the aerosoliza-

tion performance could be confidently evaluated.

The FPF (MMAD<6.8 mm) of the loaded dose

was used as a measure of DPI performance. The

influence of mill time on the FPF of both freshly

milled and recrystallized lactose monohydrate

samples is shown in Figure 6. In general, the

comminution process had a significant in-

fluence on aerosolization performance of NST in

both freshly milled and recrystallized samples

(ANOVA, p<0.05). Furthermore, with the

changes in the physical properties of the carrier

induced by ball milling, a linear relationship1

was observed between FPF and milling time for

both freshly milled and recrystallized samples

( R2¼0.954 and 0.938 for freshly milled and

recrystallized samples, respectively). It is inter-

esting to note however, a deviation from linearity

for FPF occurred for the freshly milled samples,

after 40 min mill time. In addition, analysis of FPF

Figure 5. Influence of mill time on the degree of

amorphous content in freshly milled lactose monohy-

drate samples. Milled-recrystallized lactose monohy-

drate samples were completely crystalline (0.00%amorphous content; not shown).

Figure 6. Relationship between mill time and fine

particle fraction forfreshly milled(*) and recrystallized

milled (*) lactose monohydrate samples. * R2 relation-

ship for freshly milled samples between 0 and 40 min

mill time.

1Linear analysis for freshly milled samples is only appliedbetween 0 and 40 min mill times.





between freshly milled and recrystallized samples

suggested after 10 min mill time, significant

differences in FPF over the linear region between

20, 30, and 40 min mill times was observed

(ANOVA, p <0.05).

Clearly, the process of milling induces signifi-

cant variation in aerosolization performance with

an apparent linear relationship between mill time

and FPF over the range 0–40 and 0–60 min for

freshly milled and recrystallized lactose mono-

hydrate samples, respectively. As previously dis-

cussed, such observations may be attributed to

many physical characteristics including amor-

phous content, variation in median diameter and

an increase in fines. The relationships between

such factors are discussed in more detail below.

Influence of Median Particle Diameter onIn Vitro Performance

As discussed earlier, multiple changes in the

physical properties of the carrier system were

introduced into the system while milling. The

most obvious change was the significant reduction

in median particle diameter. Since the median

particle diameter for both freshly milled and

recrystallized samples was similar, analysis of

the FPF, with respect to lactose monohydrate

median diameter, suggested a poor relationship

for both samples. In general, R2 values of 0.894

and 0.823 were observed for freshly milled andrecrystallized samples, respectively (Fig. 7).

Thus, a relationship between NST aerosolization

efficiency and median particle size was not as

evident.

A review of the literature revealed conflicting

reports concerning the relationship between med-

ian particle size and FPF. As previously discussed,

many early studies have suggested that reducing

the median particle diameter of carriers signifi-

cantly improves FPF,4,5,31 however, it should be

remembered that with anyenergy induced particle

size reduction processes, fine particles are ‘‘intro-

duced’’. Furthermore, no detailed information

concerning the fines content was given in these

studies. Interestingly, two recent study by Islam

et al.,7,12 and Steckel et al.,13 investigating the

influence of particle size on DPI performance,

reported no significant relationship between FPF

and decreasing median particle diameter, which

correlates with the data presented here.

Influence of Amorphous Content onIn Vitro Performance

The presence of meta-stable, that is amorphous,

material in the surface of DPI carriers may have

implications for product performance. An appar-

ent relationship between amorphous content and

FPF for freshly milled samples was observed

( R2¼0.921). However, since recrystallized sam-

ples contained no detectable amorphous content

there would be no relationship between FPF and

amorphous content (e.g. a plot of % amorphouscontent against FPF in the recrystallized sam-

ples, would result in all data sitting at 0%).

Furthermore, previous reports have suggested

increased amorphous content in DPI systems

resulted in decreased aerosolization perfor-

mance.17,19,21 Clearly such relationships need to

be studied further, in isolation of other physical

factors.

Influence of Fines Content on In Vitro Performance

As previously discussed, a significant increase infines with increased mill time was observed for

both freshly milled and recrystallized lactose

monohydrate samples. The relationship between

fine lactose monohydrate content and FPF is

shown in Figure 8. As with the variation of FPF

and mill time FPF (Fig. 6), a linear relationship

between FPF and the percentage of fines (<10 mm)

present in each sample was observed. However, it

is interesting to note that the relationship was

more significant with R2 values of 0.974 and 0.982

Figure 7. Influence of median particle diameter on

fine particle fraction of freshly milled (*, R2¼ 0.894)

and recrystallized milled (*, R2¼ 0.823) lactose mono-

hydrate samples.

1338 YOUNG ET AL.




being observed for freshly milled and recrystal-

lized lactose monohydrate samples, respectively.

Again, as with the variation of FPF with mill time

data, linearity for the freshly milled samples only

existed up to a certain extent (equivalent to

40 min or 13.9% fines), after which, the FPF

decreased.

More importantly, when compared to the influ-

ence of mill time, the relationship between thefines content of either freshly milled or recrystal-

lized samples and FPF indicated no significant

difference with respect to fines content (ANOVA,

p<0.05). In simple terms, samples with less than

15% fines resulted in no significant difference in

FPF between samples of milled or recrystallized

lactose monohydrate carriers containing similar

fines percentage. Such observations correlate with

previous investigations. Zeng et al.,6,9 and Islam

et al.,7,12 reported a significant reduction in FPF

with a reduction in fines content. The FPF was

returned to the original level when the fines

removed were restored, and further improved withincreasing fines content. Various studies reported

a trend of improved FPF when fine particles

(lactose monohydrate, glucose or PEG 6000) were

introduced to coarse carrier mixtures, irrespective

of coarse carrier particle size.8,10,11,32

In general, two mechanisms have been pro-

posed to explain such observations: the ‘‘active site

theory’’ and ‘‘multiplet/agglomeration’’ theory.

In simple terms, the active site theory suggests

that the increasing carrier fine concentration

occupies high energy-active sites, thus promot-

ing drug adhesion to occur at relatively lower

energy-passive sites.6,8–11,33 Clearly, the result of

such an effect would be the easier detachment

of drug particles and thus increased FPF.

However, although the existences of such sites

have been experimentally verified in recent

publications,16–19,33 it is suggested that at such

high drug loadings (5% w/w) active site effects

would be minimal. Alternatively, adhesion and

redistribution of ingredients, when fines are

present, may produce a mixed agglomerate of

drug and fine material (forming multiplets or

agglomerates). Such agglomeration may result in

improved drug aerosolization, since, due to simple

physics, dispersion of larger agglomerates from

larger carrier particles will be achieved at lower

forces, when compared to typical individual drug

carrier systems.10 In the case of such high drug loadings, the authors proposethat this mechanism

would most likely dominate FPF.

Although such general theories are attractive, a

deviation from linearity at high fine concentra-

tions (>15% fines) still exists. One of the most

likely explanations for the observed decrease in

FPF in samples with fines >15% is a mixed

agglomerate theory, where at high fines concen-

trations drug-lactose monohydrate agglomerates

fail to adhere to the larger lactose monohydrate

carrier particles and become segregated. Such

segregation results in a biphasic system in that thetypical ‘‘ordered’’ mix of micronized drug/lactose

monohydrate becomes a two-component blend:

the carrier/drug agglomerate particles and the

free agglomerates. Such a blend may result in

deviation from an expected agglomerate–carrier

relationship. To further investigate the potential

for such segregation, the particulate structure of

formulations containing different levels of fine

lactose monohydrate were investigated. Represen-

tative SEM images of formulations containing

18.0% (60 min freshly milled lactose monohydrate)

and 10.1% (60 min recrystallized lactose mono-

hydrate) fines are shown in Figure 9A and B,respectively.

When comparing samples containing >15%

fines (Fig. 9A), large drug-fine agglomerates, that

were clearly separated from the coarse carrier,

were present in the system. In contrast, with

moderate fines content (Fig. 9B), a homogenous

blend of mixed drug-fine agglomerates adhering to

the coarse carrier surface was depicted. From such

observations, it is reasonable to assume, that a

critical agglomerate size may exist in drug-fine-

Figure 8. Relationship between percentage lactose

monohydrate fines content (<10 mm), fine particle

fraction for freshly milled (*) and recrystallized milled

(*) lactose monohydrate samples. * R2 relationship forfreshly milled samples where fines are less than 15%.





carrier blends where segregation may result in a

reduction in aerosolization performance. Indeed,

in recent studies similar relationships were

observed.10

Finally, although a direct relationship between

fine concentration and aerosolization performance

was observed (<15% fines), a discrepancy still

exists when comparing the regression slopes of

freshly milled and recrystallized lactose mono-hydrate samples (Fig. 8) (slope¼1.56 and 1.08 for

freshly milled and recrystallized samples, respec-

tively). It is envisaged that this variation in slope

may be due to increased adhesion between the

drug particles and amorphous regions. It is also

important to note, that in this study, drug particles

were blended with the lactose postrecrystalliza-

tion. It is envisaged that, if blended prior to

recrystallization, the drug particles wouldpossibly

fuse to the lactose surface as with the fines,

markedly reducing the FPF. However, these

parameters needed to be studied in isolation (i.e.

without the presence of fines) and are suggested

for future investigation.

CONCLUSION

Industrial processing of carrier materials used in

DPI systems induces changes in the physical

properties of the particles. Here, a simple ball

milling process was used to produce particles

which exhibited a reduced particle size, increased

levels of fines and amorphous material. In addi-

tion, variation in storage conditions of the pro-

cessed excipient was also shown to induce

changes in fines and amorphous content. When

used in a high dose DPI system, significant

changes in the FPF were observed with increasing milling times. The relationships between milling

time, physical property of the carrier and FPF

were investigated. In general, the strongest

relationship between carrier physical property

and FPF was observed when considering the fines

content. Such a relationship was independent of

storage conditions, with increasing fines (<15%)

resulting in a linear increase in FPF. Subse-

quently, the presence of fines was shown to play

the predominating role in influencing DPI perfor-

mance in this system. Furthermore, increasing

fine content above 15% resulted in a deviationfrom linearity and may be related to changes in

overall formulation characteristics. Finally, there

appears to be some evidence that the presence of

amorphous content may contribute to a decreased

FPF. However, like the effect of particle size,

further investigation is required to study the

effect of these two parameters on dry powder

dispersion in isolation, without the masking effect

of fines content.

REFERENCES

1. Edge S, Kibbe A, Kussendrager K. 2006. Lactose,

Monohydrate. In: Rowe RC, Sheskey PJ, Owen

SC, editors. Handbook of pharmaceutical excipi-

ents, 5th Edition. London: Pharmaceutical Press.

pp 389–395.

2. Steckel H, Markefka P, teWierik H, Kammelar R.

2004. Functionality testing of inhalation grade

lactose. Eur J Pharm Biopharm 57:495– 505.

3. Bell JH, Hartley PS, Cox JS. 1971. Dry powder

aerosols. I. A new powder inhalation device.

J Pharm Sci 60:1559– 1564.

Figure 9. Representative SEM images of (A) freshly

milled lactose monohydrate blends containing >15%

fines and (B) recrystallized lactose monohydrate blends

containing <15% fines.

1340 YOUNG ET AL.




4. French DL, Edwards DA, Niven RW. 1996. The

influence of formulation on emission, deaggrega-

tion and deposition of dry powders for inhalation.

J Aerosol Sci 27:769– 783.

5. Steckel H, Muller BW. 1997. In vitro evaluation of

dry powder inhalers II: influence of carrier particle

size and concentration on in vitro deposition. Int JPharm 154:31– 37.

6. Zeng XM, Martin GP, Tee SK, Ghoush AA, Marriott

C. 1999. Effects of particle size and adding

sequence of fine lactose on the deposition of

salbutamol sulphate from a dry powder formula-

tion. Int J Pharm 182:133–144.

7. Islam N, Stewart P, Larson I, Hartley P. 2004.

Effect of carrier size on the dispersion of salmeterol

xinaofoate from interactive mixtures. J Pharm Sci

93:1030–1038.

8. Lucas P, Anderson K, Stanniforth JN. 1998.

Protein deposition from dry powder inhalers: Fine

particle multiplets as performance modifiers.

Pharm Res 15:562.

9. Zeng XM, Martin GP, Tee SK, Marriott C. 1998.

The role of fine particle lactose on the dispersion

and deaggregation of salbutamol sulphate in an air

stream in vitro. Int J Pharm 176:99–110.

10. Louey MD, Stewart PJ. 2002. Particle interactions

involved in aerosol dispersion of ternary interactive

mixtures. Pharm Res 19:1524–1531.

11. Louey MD, Razia S, Stewart PJ. 2003. Influence of

physico-chemical carrier properties on the in vitro

aerosol deposition from interactive mixtures. Int J

Pharm 252:87– 98.

12. Islam N, Stewart P, Larson I, Hartley P. 2004.

Lactose surface modification by decantation: Aredrug-fine lactose ratios the key to better dispersion

of salmeterol xinafoate from lactose-interactive

mixtures? Pharm Res 21:492– 499.

13. Steckel H, Markefka P, teWierik H, Kammelar R.

2006. Effect of milling and sieving on functionality

of dry powder inhalation products. Int J Pharm

309:51–59.

14. Zeng XM, Martin GP, Marriott C, Pritchard J.

2000. The influence of carrier morphology on drug

delivery by dry powder inhalers. Int J Pharm 200:

93–106.

15. Flament MP, Leterme P, Gayot A. 2004. The

influence of carrier roughness on adhesion, content

uniformity and the in vitro deposition of terbutaline

sulphate from dry powder inhalers. Int J Pharm

275:201–209.

16. Young PM, Cocconi D, Colombo P, Bettini R, Price

R, Steele DF, Tobyn MJ. 2002. Characterization of

a surface modified dry powder inhalation carrier

prepared by ‘‘particle smoothing’’. J Pharm Phar-

macol 54:1339– 1344.

17. Price R, Young PM. 2005. On the physical trans-

formations of processed pharmaceutical solids.

Micron 36:519– 524.

18. Buckton G, Darcy P. 1999. Assessment of disorder

in crystalline powders—a review of analytical

techniques and their application. Int J Pharm

179:141–158.

19. Young PM, Price R. 2004. The influence of humidity

on the aerolisation of micronised and SEDS

produced salbutamol sulphate. Eur J Pharm Sci22:235–240.

20. Podczeck F, Newton JM, James MB. 1997. Varia-

tions in the adhesion force between a drug and

carrier particles as a result of changes in the

relative humidity of the air. Int J Pharm 149:151–

160.

21. Ward GH, Schultz RK. 1995. Processed-induced

crystallinity changes in albuterol sulphate and its

effect on powder physical stability. Pharm Res

12:773–779.

22. Young PM, Chiou H, Tee T, Traini D, Chan HK,

Thielmann F, Burnett D. 2007. The use of organic

vapor sorption to determine low levels of amor-

phous content in processed pharmaceutical pow-

ders. Drug Dev Ind Pharm 33:91–97.

23. Chen Y, Ding Y, Papadopoulos DG, Ghadiri M.

2004. Energy-based analysis of milling alpha-

lactose monohydrate. J Pharm Sci 93:886– 895.

24. Kwan CC, Chen YQ, Ding YL, Papadopoulos DG,

Bentham AC, Ghadiri M. 2004. Development of a

novel approach towards predicting the milling

behaviour of pharmaceutical powders. Eur J

Pharm Sci 23:327– 336.

25. Berg TGO, Avis LE. 1970. Exploratory experiments

on kinetics of comminution. Powder Technol 4:27–

31.

26. Piret EL. 1953. Fundamental aspects of grinding.Chem Eng Prog 49:56–62.

27. Parrott EL. 1974. Milling of pharmaceutical solids.

[Review] [95 refs]. J Pharm Sci 63:813–829.

28. Coelho MC, Harnby N. 1978. Moisture bonding in

powders. Powder Technol 2:201– 205.

29. Schubert H. 1984. Capillary forces: Modelling and

application in particulate technology. Powder

Technol 37:105– 116.

30. Young PM, Price R, Tobyn MJ, Buttrum M, Dey

F. 2003. Effect of humidity on aerosolization of

micronized drugs. Drug Dev Ind Pharm 29:959–

966.

31. Bell JH, Hartley PS, Cox JS. 1971. Dry powder

aerosols. I. A new powder inhalation device.

J Pharm Sci 60:1559– 1564.

32. Zeng XM, Pandhal KH, Martin GP. 2000. The

influence of lactose carrier on the content homo-

geneity and dispersibility of beclomethasone dipro-

pionate from dry powder aerosols. Int J Pharm

197:41–52.

33. Young PM, Edge S, Traini D, Jones MD, Price R,

El-Sabawi D, Urry C, Smith C. 2005. The influence

of dose on the performance of dry powder inhalation

systems. Int J Pharm 296:26–33.



The Influence of Mechanical Processing of Dry Powder.pdf

Documents

Transcript of The Influence of Mechanical Processing of Dry Powder.pdf