Dry Powder Inhaler

Pharm Res (2013) 30:458–477
DOI 10.1007/s11095-012-0892-4

RESEARCH PAPER

Freeze-Dried Mannitol for Superior Pulmonary Drug Delivery
via Dry Powder Inhaler
Waseem Kaialy & Ali Nokhodchi

Received: 19 July 2012 / Accepted: 24 September 2012 / Published online: 16 October 2012
# Springer Science+Business Media New York 2012

ABSTRACT
Purpose To show for the first time the superior dry powder
inhaler (DPI) performance of freeze dried mannitol in comparison to spray dried mannitol and commercial mannitol.
Methods Different mannitol powders were sieved to collect
63–90 μm particles and then analyzed in terms of size, shape,
surface morphology, solid state, density, flowability. Salbutamol
sulphate-mannitol aerosol formulations were evaluated in terms
of homogeneity, SS-mannitol adhesion, and in vitro aerosolization performance.
Results Freeze dried mannitol demonstrated superior DPI performance with a fine particle fraction believed to be highest so far
reported in literature for salbutamol sulphate under similar protocols (FPF046.9%). To lesser extent, spray dried mannitol produced better aerosolization performance than commercial
mannitol. Freeze dried mannitol demonstrated elongated morphology, α-+β-+δ- polymorphic forms, and poor flowability
whereas spray dried mannitol demonstrated spherical morphology, α-+β- polymorphic forms, and excellent flowability. Commercial mannitol demonstrated angular morphology, βpolymorphic form, and good flowability. Freeze dried mannitol
demonstrated smoother surface than spray dried mannitol which
in turn demonstrated smoother surface than commercial mannitol. FPF of SS increased as mannitol powder porosity increase.
Conclusions Freeze drying under controlled conditions can be
used as a potential technique to generate aerodynamically light
mannitol particles for superior DPI performance.
W. Kaialy (*) : A. Nokhodchi
Chemistry and Drug Delivery Group, Medway School of Pharmacy
University of Kent
ME4 4TB Kent, UK
e-mail: [email protected]

A. Nokhodchi
e-mail: [email protected]

KEY WORDS aerosol . freeze dried mannitol . morphology .
porosity . spray dried mannitol

INTRODUCTION
Pharmacologically active drugs usually betray poor physicochemical properties and therefore formulation development
is often considered challenging. Milling is the most conventional method to prepare particles in the size range between
1 μm and 10 μm which are widely used in chemical, mineral, and pharmaceutical industries (e.g. preparation of respirable aerosol particles). Many mills have been employed
for drug micronization such as fluid-energy mills (e.g. the jet
mill), high peripheral speed mills (e.g. the pin mill), and ball
mills. Jet milling (or air attribution milling) is the most
commonly used milling technique. Jet milling depends on
introducing pressurized high velocity gas (air or nitrogen)
through nozzles into the milling chamber, which results in
high-speed (sonic velocities) particle-particle collision (interparticle collision) and abrasion. The efficiency of jet milling
is significantly affected by the nature of the particles fed into
the jet mill. For example both large and small particles were
not ideal for jet milling, and the preferred particle size of
most materials for jet milling was in the range of 75 to
100 μm. Also, brittle materials have a tendency to fracture
W. Kaialy
Pharmaceutics and Pharmaceutical Technology Department
University of Damascus
Damascus, Syria
A. Nokhodchi
Drug Applied Research Center and Faculty of Pharmacy
Tabriz University of Medical Sciences
Tabriz, Iran

Inhaled Mannitol: Freeze-Dried vs Spray-Dried vs Commercial

during jet milling whereas ductile materials (or plastic materials) might undergo plastic deformation rather than fracturing. Despite its popularity, jet milling technique suffers
from several disadvantages (extremely inefficient). This
could be attributed to poor performance of jet milled products due to several reasons including the low opportunity to
control particle physical properties (e.g. size, shape, morphology, and surface texture), reduced crystallinity, poor
flowability, high electrostatic charge, high cohesiveness (agglomeration tendency), possible chemical degradation, and
safety concerns due to dust exposure (1).
Spray drying (introduced in the 1980s) has been widely
used as a micronization technique to produce particles for
pulmonary delivery (2). In pharmaceutical industry, spray
drying is the most commonly used technique in preparing
peptides and proteins for inhalation as a dry powder. Nevertheless, spray drying technique suffer from several disadvantages such as reduced crystallinity for spray dried
products, not suitable for substances susceptible to atomization mechanical shear (e.g. biopharmaceutical drugs), not
suitable for substances that are unstable to liquid–air interface or decomposed by oxidation, and very low process
yield. Anti-solvent crystallization is a process where, generally, an organic product can be recovered from aqueous
solutions through the addition of nonsolvent compounds by
which the solute solubility is decreased without creating a
new liquid phase. Recently, anti-solvent crystallisation using
binary nonsolvents was proved to be potential method to
prepare mannitol (3,4) and lactose (5–7) particles with superior dry powder inhaler (DPI) performance. However, antisolvent crystallisation using alcohols suffer from many disadvantages including the requirement of solvent recovery
and the risk associated with the use of flammable solvents at
high reaction temperatures. Also, mechanical stirring used
during crystallization introduces random energy fluctuations
with the solution leading to heterogeneous distribution of
local concentrations and consequently heterogamous crystal
growth. Freeze drying is a technique by which it is possible
to recover dry product from aqueous solutions. Commonly,
freeze drying is used for preparing injectable pharmaceutical products.
After 20 years of using metered dose inhalers (MDIs), the
first DPI was introduced to the market at 1970 (Fisons,
Spinhaler®). In literature, it has been shown that it is
possible to obtain increased respirable drug fraction by
decreasing particle size of carrier particles (8,9). However,
other reports showed that carrier particle mean diameter
has no effect on aerodynamic diameter (10) or on fine
particle fraction of drug (11). Furthermore, other reports
showed that larger carrier particles might outperform
smaller carrier particles (12,13). In comparison to control,
carrier particles with more elongated shape (3,4,14,15) or
less elongated shape (6–8) deposited higher amounts of drug

459

on lower airway regions. Generally, no significant relationship was observed between FPF of drug and flowability of
carrier powder (8). Better aerosolization performance was
obtained from DPI formulations with either better flow
properties (16) or poorer flow properties (3,4). Such apparently dissimilar results could be explained as the physical
properties determinations of carrier particles are dependent
on each other, and to the fact that DPI aerosolization
behaviour is reliant on several interrelated events at the
same time. For example, it was shown that, in determining
DPI performance, carrier polymorphic form and surface
energy dominates over carrier size distribution (17). Also,
type of drug (18), type of inhaler device (19), amounts of fine
carrier particles (20), and carrier surface texture (21) may
have an effect on the preferable carrier size for enhanced
aerosolisation performance. For example, In case of DPI
formulation powders containing coarse carrier particles,
higher respirable fractions were obtained when using highefficiency dispersing systems (high turbulence inhalers).
However, for DPI formulations containing carriers with
large amount of fine particles, effective dispersion was
obtained when using low turbulent inhaler devices.
Lactose has some degree of security when it is used as
inert excipient considering its incompatibility with drugs
that have primary amine moieties making it less suitable
excipient for next generation of inhalable products (i.e.
proteins and peptides). Therefore, using alternative excipients appears to be an attractive option for DPI formulations. Possible excipients for dry powder inhalation
formulations are rather narrow as they have to meet specific
conditions such as being endogenous, able to be metabolised
or cleared, and have no potential to injure or irritate the
lungs. Therefore, in inhalation field, only generally recognized as safe (GRAS) excipients can be used. Mannitol was
attractive alternative since it does not have a reducing effect,
less hygroscopic, and gives high sweet aftertaste which could
be used to observe dose taken by the patient. Also, mannitol
is the most abundant polyol in nature, has been used widely
for commercial pharmaceutical protein formulations due to
its biological stabilizing efficiency properties, is the most
widely used bulking agent in freeze dried formulations,
and is expected to be approved in future for use of DPIs
(22). This is because of mannitol is inert, has good cakesupporting properties, crystallises readily during freeze drying, and allows drying processed at higher product temperatures (23). Mannitol is a polyol cryoprotectant and a
lyoprotectant that lead to crystalline freeze dried systems
(24). The mannitol/ice eutectic mixture has a high melting
temperature (~ −1.5°C) promoting efficient freeze drying
and physical stability of freeze dried mannitol solid (25).
Despite over 40 years of research, low drug delivery
efficiency to the lower airway regions is still a major challenge for dry powder aerosol pharmaceutical dosage forms

460

(8). In pharmaceutical industry, there is increased interest
for pharmaceutical excipients, other than lactose, which
produce efficient drug delivery upon inhalation. The objective of this study was to introduce, for the first time, freeze
dried mannitol as alternative promising carrier in DPI formulations containing salbutamol sulphate as a model drug.
Also, this study was performed with a view to propose
optimal mannitol product (freeze dried mannitol vs spray
dried mannitol vs commercial mannitol) for salbutamol sulphate based drug-carrier dry powder inhaler formulations.
It was intended in this study to show how different mannitol
grades perform inherently under similar protocols including
blending and sieving.

MATERIALS AND METHODS
Materials
Commercial mannitol (CM) (Fisher Scientific, UK) and
spray dried mannitol (SDM) (SPI Pharma, UK) were purchased from the named sources. Micronized salbutamol
sulphate (SS, D10% 00.5 ± 0.0 μm, D50% 01.7 ± 0.1 μm,
D90% 03.1 ± 0.3 μm (3)) was supplied from LB Bohle,
Germany.
Preparation of Freeze-Dried Mannitol (FDM)
Mannitol was freeze dried using a SCANVAC CoolSafe™
freeze-dryer (CoolSafe 110-4, Lynge, Denmark). A 5% w/v
mannitol solution was prepared by dissolving 5 g of mannitol in distilled water such that the final solution volume is
100 mL. 100 mL of mannitol solution was filled into 250 mL
round-bottomed flask and left in freezer overnight after
which it was placed on the shelves of the freeze-dryer.
Samples were freeze dried at −110°C and collected after
48 h after which they were transferred into sealed glass vials
over silica gel until used.
Sieving
In order to limit the influence of mannitol particle size on
aerosolization performance, similar size fraction (63–90 μm)
of each mannitol powder was used. Mechanical sieving was
applied via mechanical shaker (Endecotts Ltd, England) as
described in details elsewhere (3).
Particle Size Measurements
Particle size analysis was carried out using a Sympatec
(Clausthal-Zellerfeld, Germany) laser diffraction particle
size analyser as described in details elsewhere (4). The span
(calculated from Eq. (1)) of the volume distribution was used

Kaialy and Nokhodchi

as a measure of the width of the distribution relative to the
median diameter:
Span ¼

d ½v; 90  d ½v; 10
d ½v; 50

ð1Þ

Image Analysis Optical Microscopy
Quantitative particle shape analysis was conducted using
a computerized morphometric image analyzing system
(Leica DMLA Microscope; Leica Microsystems Wetzlar
GmbH, Wetzlar, Germany; Leica Q Win Standard
Analyzing Software) as explained in details elsewhere
(26,27). In order to get good understanding of particle
morphology, several shape factors have been applied as
none of these descriptors is able to accurately differentiate between geometric shape and surface roughness.
Particle shape were quantified using several descriptors
including aspect ratio (Eq. 2), flakiness ratio (Eq. 3),
sphericity (Eq. 4), compactness (Eq. 5), and simplified
shape factor (Eq. 6) as defined elsewhere (26–28)
Aspect ratio ¼

Length
Breadth

Flakiness ratio ¼

ð2Þ

Breadth
Thickness

ð3Þ

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
3 Width  Thickness
Sphericity ¼
ðLengthÞ2

Compactness ¼

ð4Þ

ðPerimeter Þ2
Area

Perimeter convex

Simplified shape factor ¼
Perimeter

ð5Þ
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi


breadth 2
1
length

ð6Þ

Aspect ratio and flakiness ratio (flatness ratio) are the
fundamental first order shape descriptor of a particle.
Regardless of orientation, a perfect sphere is expected
to have an aspect ratio and flakiness ratio of 1 whereas
non-spherical particle will have an aspect ratio and
flakiness ratio of <1 or >1. Higher aspect ratio indicates more elongated shape and/or rougher surface
whereas higher flakiness ratio indicates flatter particles.
As sphere is the simplest dimensional shape, shape of
solid particles is frequently described by their Sphericity.
Sphericity shape parameter is a descriptor of how a

Inhaled Mannitol: Freeze-Dried vs Spray-Dried vs Commercial

particle is similar to sphere. A typical smooth sphere is
expected to have a sphericity value of unity. A sphere
particle with measurable surface asperities or nonspherical particle will have a sphericity value <1.
Smaller sphericity values indicate higher degree of
shape irregularity and/or higher surface roughness.
Compactness is a measure of how nearly circular an
aggregate cross section is (27).
Scanning Electron Microscope (SEM)
Electron micrographs of different mannitol samples were
obtained using a scanning electron microscope (HITACHI
SU 8036, Japan) operating at 5–15 kV as explained in
Kaialy et al. (26).
Atomic Force Microscope (AFM)
Atomic force microscopy analysis were as performed
using a Veeco MultiMode AFM equipped with an Etype scanner operating via a Veeco Nanoscope IIIa
controller (Bruker AXS Inc., Bruker Nano Surfaces,
Madison, WI, USA) as described in details elsewhere
(4). Roughness analysis was performed using Veeco Nanoscope software (version 5.12b36) on images of 5×5 μm2 and
300×300 nm2 sized.
Differential Scanning Calorimetry (DSC)
A differential scanning calorimeter (DSC7, Mettler Toledo, Switzerland) was used to characterise solid state
nature of different mannitol samples, as described previously
(26).
Fourier Transform Infrared Spectroscopy (FT-IR)
FT-IR was employed to give investigate any chemical
changes at molecular level in freeze dried mannitol and
spray dried mannitol samples in comparison to commercial
mannitol. The method incorporated was adapted from
Kaialy et al. (27).
Powder X-ray Diffraction (PXRD)
The patterns of all mannitol samples were collected on
a Bruker D8 Advance Siemens powder diffractometer
with Cu Kα radiation (1.54056 Å) using the DIFFRACplus software, as explained in Kaialy et al. (14). Quantitative mannitol crystal form analysis (% α-, % β-, or
% δ-mannitol) were performed by Rietveld refinement
using Topas v4 (Bruker). Cif structural models (adapted
from Fronczek et al. (29)) of mannitol polymorphs were
obtained from the Cambridge Structural Database and

461

refined within Topas on pure mannitol samples and then
converted to str files. All refinements were done using fundamental parameters routine based on the configuration of our
diffractometer.

Particle True Density Measurements
True density of all mannitol particles (defined as particle
mass divided by its volume excluding both open pores and
closed pores) was measured using an ultrapycnometer 1,000
according to Kaialy et al. (27).

Characterization of Powder Bulk and Flow Properties
Bulk density, tap density, and porosity (Eq. 7) of each
mannitol powder sample were measured as important
descriptors of powder bulk cohesive properties. Carr’s
index (CI, Eq. 8) was measured for all mannitol powders to characterize flowability of mannitol powders.
The method incorporated was described elsewhere (14)

Porosity ¼

CI ¼

1

Bulk density
True density


 100

Tap density Bulk density
Bulk density

ð7Þ


 100

ð8Þ

Preparation of SS-Mannitol Formulations
Each mannitol powder (3 g) was blended with SS powder at a constant ratio of mannitol: SS 67.5: 1,
w/w which is the same ratio used in commercially
available Ventolin Rotacaps®. This blending was performed in a cylindrical aluminium container (6.5×8 cm)
using a Turbula® mixer (Maschinenfabrik, Basel Switzerland) at a standard mixing conditions (100 rpm mixing speed and 30 min mixing time). Once prepared, all
formulations were stored in firmly sealed vials over silica
gel for minimum 24 h before any investigation to allow
any electrical charge relaxation to occur. In order to
prepare ordered mixtures of two fine particles, a mixer
producing high shear forces would be favorable. However, in this study, a turbulent tumbling mixer (TTM,
Turbula®) was used since TTM mixers are usually recommended to produce drug-carrier ordered mixtures in the
case of formulations containing one coarse particle component (e.g. mannitol) and one fine particle component
(e.g. salbutamol sulphate).

462

Homogeneity Assessment of SS-Mannitol
Formulations
After blending, seven randomly selected samples were
taken from different spots of each formulation powder
for quantification of SS content. Each sample weighs
33±1.5 mg (which is equivalent to a unit dose of SS:
481± 22 μg, in accordance with Ventolin Rotacaps®)
and was dissolved in 100 mL distilled water in a volumetric flask. SS was analysed using an HPLC method
adapted from Kaialy et al. (3). For each formulation, %
potency was calculated as the percent amounts of SS to
the nominal dose while the degree of SS content homogeneity was expressed in terms of percent coefficient of
variation (% CV). % CV of 6% or less was considered as
sufficient uniform for DPI.

Kaialy and Nokhodchi

calculated as the ratio of the RD to the theoretical dose
(481 μg). Theoretical aerodynamic diameter of mannitol
particles were estimated from mannitol volume mean
diameter (VMD) and tapped density (ρ) using Eq. 9
(30–32)


ρ
TAD ¼ VMD
ρ1

 12

ð9Þ

where ρ1 01 g/cm3.
Statistical Analysis
One way analysis of variance (ANOVA) and Tukey’s
Honestly Significant Difference (HSD) (14,26) test was
applied to statistically compare mean results in this
study.

SS-Mannitol Adhesion Assessments
Air jet sieving with a 20 μm sieve was used to indirectly
assess drug-carrier adhesive forces of within all formulations
(7,26,27) as described elsewhere (27).

RESULTS AND DISCUSSION
Particle Size Distribution (PSD)

In Vitro Aerosolisation Study
Each formulation was filled manually into hard gelatine
capsules (size 3) with 33±1.5 mg powder. Prior to any
investigation, all filled capsules were stored in sealed
glass vials for at least 24 h in order to allow any
charge-dissipation to occur. Deposition profiles of all
formulations were assessed in vitro (Aerolizer® inhaler
device with Multi stage Liquid Impinger (MsLI)), as
described in pharmacopoeia and adopted from Kaialy
et al. (3). Every deposition experiment involved the actuation of ten capsules and was repeated three times.
Several parameters were employed to quantify SS deposition profiles from each formulation including recovered dose (RD), emitted dose (ED), mass median
aerodynamic diameter (MMAD), geometric standard deviation (GSD), fine particle dose (FPD≤5μm), impaction
loss (IL), fine particle fraction (FPF≤5μm, % RD), and
dispersibility (DS), as defined elsewhere (3,6). The RD is
the sum of the weights of drug (μg) recovered from
inhaler device with its fitted mouthpiece adaptor (I+M), induction port (IP), and all stages of the impactor. The
emitted dose (ED) is the amount of drug delivered from
the inhaler device, which is collected in the induction
port and all stages of the impactor (i.e. total RD except
for the inhaler device with mouthpiece adaptor). The
impaction loss was calculated as the sum of drug
amounts collected from the induction port and stage 1
of the MSLI expressed as a percentage to the recovered
dose. The percent total recovery (% recovery) was

Laser diffraction analysis for different mannitol powders
showed sigmodial (unimodally distributed) size distributions
with most particles falling into the nominal sieve mesh size
ranges (Fig. 1a). Despite that all mannitol powders were
carefully sieved under similar protocols, CM demonstrated
higher volume mean diameter (VMD0108.1±1.6 μm) than
SDM (81.4±0.9 μm) and FDM (81.0±0.5 μm) (Fig. 1b).
Particle size data were supported by representative photograph images (Fig. 1c–e). In DPI systems, differences in size
of carrier particles could have a considerable effect on DPI
performance. For example, it has been shown that carriers
with smaller size have increased disorder in crystal lattice
and improved aerosolization performance (8,9). Unlike CM
and SDM samples, where sieving was efficient to remove all
particles smaller than 5 μm (FPM <5μm ) and 10 μm
(FPM<10μm) (Fig. 1a, c, d), FDM contained 2.0±0.3% and
3.9±0.6% (v/v) of FPM<5μm and FPM<10μm respectively
(Fig. 1a) which is also confirmed by microscopic image of
FDM (Fig. 1e). Amounts of fines in FDM sample correspond
to intrinsic fine mannitol particles which could not be isolated by sieving. FDM demonstrated wider (more heterogeneous) size distribution (span01.4±0.0, skewness01.7) in
comparison to CM (span 00.8 ± 0.0, skewness 01.4) and
SDM (span00.7±0.1, skewness01.0) (Fig. 1a, b), which
could be also substantiated by the representative photographs for CM, SDM, and FDM (Fig. 1c–e). PSD polydispersity is important in terms of aerosol quality and efficiency
since differences in aerosol particle size might lead to differences in lung deposition regions (9).

Inhaled Mannitol: Freeze-Dried vs Spray-Dried vs Commercial

463

Fig. 1 Particle size distribution
(% undersize) (a); volume mean
diameter (VMD), (Balck Circle)
span (b); and representative
photographs for commercial
mannitol (CM) (c); spray dried
mannitol (SDM) (d); and freeze
dried mannitol (FDM) (e).
(mean±SD, n08). S indicates
statistically similar.

Image Analysis Optical Microscopy
Aspect ratio for different mannitol samples was within the
following rank order: SDM (1.28±0.0)<CM (1.87±0.01)<
FDM (2.54±0.03) (Fig. 2a). In contrary, flakiness ratio for
different mannitol samples was within the following rank order
SDM (0.89±0.00)>CM (0.73±0.00)>FDM (0.64±0.01)
(Fig. 2a). This indicates that, among all mannitol particles,
SDM particles have the least elongated-most flattened particle
shape, whereas FDM particles have the highest degree of
shape-elongation. Such results indicate that, during freeze
drying, mannitol crystals grow faster along their length face
than along their width. Spherecity for different mannitol samples was within the following rank order: SDM (0.93±0.00)>
CM (0.86±0.00)>FDM (0.82±0.00) (Fig. 2b). In contrary,
compactness for different mannitol samples was within the
following rank order SDM (18.8±0.29)<CM (98.8±2.13)<
FDM (103.0±2.73) (Fig. 2a). This indicates that SDM particles
have the most spherical-regular morphology whereas FDM
particles have the highest degree of shape irregularity. A

simplified shape factor was used to estimate the combination
of variation in surface asperities (surface texture factor) and
deviation of shape from a circle (shape factor) (26,27). This
factor is helpful in 2D or 3D shape assessment and can have
values from -1 to 1: the smaller the value the more irregular the
particle morphology. SDM sample exhibited the highest simplified shape factor (0.29±0.00) whereas FDM sample showed
the lowest simplified shape factor (−0.20±0.00).
Such results indicate that SDM proved nearly circular
regular-shaped particles whereas FDM showed the most
elongated- most irregular-shaped particles. Such information
provided by image analysis is compatible with representative
photographs of SDM (Fig. 1d) and FDM (Fig. 1e).
Scanning Electron Microscopy
Since the accuracy of particle image analysis data is hindered by particle orientation and interparticulate contact
area (7), and therefore might not be satisfactory to represent
the influence of particle shape, all mannitol carrier particles

464

Kaialy and Nokhodchi

Fig. 2 Aspect ratio (Black Circle),
flakiness ratio (Yellow Square) (a);
Sphericity (White Square),
Compactness (Green Triangle) (b);
simplified shape factor (c)
(mean±SE, n≥1000); and
representative scanning electron
micrographs of commercial
mannitol (CM) (d); spray dried
mannitol (SDM) (e); and freeze
dried mannitol (FDM); 1: needle
shaped, 2: dendritic shaped, 3:
dragon like shape, 4: fine
particulates (FPM<10μm), 5: fine
particle aggregates (FPA) (f).

were further qualitatively analyzed by SEM (Fig. 2d–f).
SEM images of different mannitol particles revealed crystalline particles with size typically in 63–90 μm range (supporting laser diffraction data (Fig. 1a)). Strike morphological
differences were observed between different mannitol particles (Fig. 2d–f).
CM exhibited the typical angular-subangular shape
reported in previous studies. Representative SEM image of
SDM illustrated spherical (rounded-subrounded, orangelike) particles with relatively uniform (regular) shape, well
curved-plane sides, and well rounded corners and edges
(Fig. 2e). No evidence of particle agglomeration was
observed in case of CM and SDM samples (Figs. 1c, d, 2d,
e). SEM image of FDM revealed irregular-deformed particles with sharp edges and mostly elongated morphology
(Fig. 2f). It was clear that, in comparison to CM and SDM,
FDM particles are less uniform in terms of shape and size,
since several morphologies could be detected: needle shape,
dendritic shape (acicular crystals), and dragon like shape
(Fig. 2f). Also, finer particulates (FPM <10μm) and fine

particle agglomerates (clusters or microcrystalline assemblies) (FPA) could be depicted in case of FDM sample
(Fig. 2f). This supports boarder size distribution (higher span
value) for FDM sample as analysed by laser diffraction
(Fig. 1b). It was assumed that, during freeze drying, several
nucleation points might form resulting in relatively heterogeneous crystal growth and consequently forming particles
with different sizes and morphologies.
Higher amounts of fines in case of FDM could be attributed to their brittle properties as indicated by their
elongated-irregular morphology and the presence of large
number of fractured faces (Fig. 2f). This could promote
particle attrition (abrasion or fracturing of cleavage plans)
upon subjecting the powder to mechanical sieving process
due to inter-particle and particle-sieve wall collisions (27). It
can be assumed that, due to low mannitol concentration
used during crystallisation (5% w/v), crystal growth of FDM
crystals was favoured in direction perpendicular to the c-axis
(polar growth direction) but less dominant on crystal faces
normal to c-axis leading to the formation of elongated crystals.

Inhaled Mannitol: Freeze-Dried vs Spray-Dried vs Commercial

Variations in morphology of carrier particles have a dominating influence on DPI aerosolization behaviour (14). For
example, carrier particles with higher elongation ratio demonstrated smaller surface free energy (33) and improved aerosolization efficiency (3,4). Particles with different shapes will
have different drag forces and terminal velocities during aerosolization which in turn affect particle aerodynamic diameter
and consequently affect particle deposition profiles in the
respiratory airways.
Particle Surface Analysis
Surface morphology of different mannitol particles were
visualised from there SE micrographs (Fig. 3a–c).
All mannitol particles showed unpolished surfaces
(Fig. 3a–c). CM particles showed uneven wrinkled
(rough) surface topography with much fragmentation
and easily visible cavities (Fig. 3a). SDM particle surface
constituted of curved plates of microscopic thickness producing some irregularities in particle surface topography
(Fig. 3b). FDM particles exhibited relatively laminated (conchoidal or waterworn-like) smoother surface morphology
(Fig. 3c).
At the magnification used for SEM images, nanoscopic
indentations might not be visible. Therefore, surface topography of different mannitol particles was analysed by AFM
and the representative topographical, amplitude, and phase
images are shown in Fig. 3d–i. To investigate the nature of
different mannitol surface regions, series of high-resolution
(5 μm×5 μm and 300 nm×300 nm scan size) AFM images
were taken. AFM showed that CM particle has large asperities and protuberances on its surface forming angular edges
with signs of cracking (Fig. 3d, g, j). In comparison to CM
(Fig. 3a, d), the surface of SDM is flatter, (Fig. 3). This is
expected to decrease the number of drug particles remaining in macroscopic surface depressions of mannitol carrier
particles and consequently might facilitate drug detachment
from carrier surface upon inhalation (3,4,14). Observations
of amplitude and phase response suggested the presence of
regions with different physico-mechanical properties and
thus demonstrated that the method used for preparation of
CM sample had a remarkable effect on surface features of
CM particles (Fig. 3g, j). These regions could be related to
clefts and pits (craters) on the surface of CM particle induced by powder preparation technique. The AFM image
of SDM (Fig. 3e, h, k) revealed less corrugated terrace
compared to CM (Fig. 3d, g, j). The amplitude/phase
images of SDM sample suggested a relatively ordered crystalline state where multiple platelets could be seen across the
surface (Fig. 3h, k). The approximate height between platelet steps was around 1–2 μm (Fig. 3e, h, k). Analysis of the
amplitude and phase lag information for the FDM sample
indicated reduced variation in phase suggesting reduced

465

variation in physicochemical property across the surface
(Fig. 3i, l). These results demonstrate that freeze drying
process induced relatively ordered crystalline lattice with
less changes to the surface in comparison to the method
used in preparation of CM sample (and to a lesser extent in
comparison to spray drying) which induced disruption of the
crystalline lattice.
Roughness analysis (5×5 μm) confirmed that CM particles
(Rq0300.1±37.9 nm, Ra0220.0±17.4 nm) have relatively
rougher surfaces than SDM particles (Rq0189.5±17.6 nm,
Ra0157.2±13.9 nm) which in turn demonstrated quantitatively rougher surface than FDM particles (Rq 014.2 ±
10.3 nm, Ra08.7±1.9 nm) (Fig. 3m). Similar conclusions
could be indicated when considering roughness analysis on
300×300 nm mannitol region images (Fig. 3n). It was assumed that the relatively low concentration of mannitol used
during freeze drying (5% w/v) promotes slow kinetics of crystallization and thus slow the crystal growth process resulting in
regular growth pattern and consequently generating nearideal lattices with relatively sufficient time to fill or cover
imperfect lattice layers (less “lattice mistakes”). In DPI systems,
particle surface roughness affects the contact geometry between the drug and carrier particles and consequently might
have a significant impact on drug-carrier adhesion (27).
Particle size and shape determinations are dependent
on each other (14). Laser diffraction takes into account
the lights diffracted from the dispersed particles to measure particle size, but does not take into account apparent
particle density and dynamic shape factors. Therefore,
particle size measurements is dependent on particle morphology and orientation (particle morphology information
is averaged out to provide one-dimensional distribution
during measurement). Angular-corrugated morphology of
CM particles (Figs. 2d, 3a) could promote high-angle
scattering as measured by laser diffraction and consequently might contribute to larger estimated particle size
for CM sample (Fig. 1a) (7).
Solid State Characterization
All mannitol samples produced comparable DSC traces
having one endothermic transition at 168.6±1.2°C (Fig. 4)
corresponding to α-mannitol or β-mannitol melting (fusion)
(Table I). However, FDM showed additional endothermic
event at 153.6±0.5°C (Fig. 4) which is diagnostic to melting
of δ-mannitol phase followed by the solidification of the melt
to form α- or β- polymorph (Table I). Melting enthalpy for
different mannitol samples at 168.6±1.2°C was within the
following rank order: CM (315.3±10.7 J/g)>SDM (292.5±
4.2 J/g)>FDM (257.0±5.6 J/g) (Fig. 4).
All mannitol products did not show any endotherm below 100°C suggesting that they contain a negligible amount
of free water (surface water). Also, this suggests the absence

466

Kaialy and Nokhodchi

Fig. 3 SE micrographs of commercial mannitol (CM) (a); spray dried mannitol (SDM) (b); and freeze dried mannitol (FDM) (c); topographical AFM image of
commercial mannitol (CM) (d); spray dried mannitol (e); and freeze dried mannitol (f); amplitude AFM images of commercial mannitol (CM) (g); spray dried
mannitol (h); and freeze dried mannitol (i); phase AFM images of commercial mannitol (CM) (j); spray dried mannitol (k); and freeze dried mannitol (l); root mean
square average (Rq) and arithmetic mean average (Ra) roughness on a 5 μm region (m); and 300 nm region (n) for commercial mannitol (CM), spray dried
mannitol (SDM), and freeze dried mannitol (FDM) (mean±SE, n≥2).

of detectable amounts of amorphous mannitol and/or mannitol hydrate within all mannitol products, since amorphous
mannitol is usually identified by glass transition at about 13°
C followed by two crystallization exotherms at about 25°C

and 65°C (24). Since α-mannitol and β-mannitol forms are
indistinguishable in DSC traces (Table I), all mannitol samples were further analysed using FT-IR (Fig. 5a) and PXRD
(Fig. 5b).

Inhaled Mannitol: Freeze-Dried vs Spray-Dried vs Commercial

467

Fig. 4 Differential scanning
calorimeter traces and melting
enthalpies of commercial
mannitol (CM), spray dried
mannitol (SDM), and freeze
dried mannitol (FDM).

CM exhibited the typical FT-IR spectrum and PXRD
pattern for the reference β-mannitol (Table I), having the
FT-IR specific diagnostic bands at 929 cm−1, 959 cm−1, and
1,029 cm−1 (Fig. 5a) and PXRD diagnostic peaks at 10.6º,
14.7 º, 23.4 º, and 29.5 º (Fig. 5b). SDM showed the FT-IR
distinctive band of α-mannitol (1,194 cm−1), FT-IR specific

bands of β-mannitol (929 cm−1 and 959 cm−1) (Fig. 5a),
PXRD specific peaks of α-mannitol (9.6º, 13.8º, and 17.2º),
and PXRD specific peaks of β-mannitol (10.6º, 14.7º, 23.4º,
and 29.5º) (Fig. 5b) (Table I). This indicates that SDM consist
of mixtures of α-mannitol (~58%, w/w) and β-mannitol
(~43%, w/w) (Fig. 5b). FDM showed the FT-IR specific band

Table I DSC Melting Points, FT-IR Diagnostic Bands, and PXRD Diagnostic Diffraction Angles for: α-mannitol, β-mannitol, and δ-mannitol Polymorphic
Forms
Technique

Polymorphic from of mannitol
α-

References

β-

References

δ-

References

Melting point
(DSC) (°C)

166.0

Yu et al. (41)

166.5

Yu et al. (41)

150–158

Band (FT-IR)
(cm−1)

1194

1209, 959, 929

9.6, 13.8, 17.2

Burger et al. (34)
Kaialy et al. (3,4,15,27)
Kaialy et al. (4)
Kaialy et al. (15,27)

967

Diffraction angle
(PXRD) (º)

Burger et al. (34)
Kaialy et al. (3,4,27)
Burger et al. (34)
Kaialy et al. (4,27)

Burger et al. (34)
Kaialy et al. (5,15,33)
Burger et al. (34)
Kaialy et al. (4,15,27)
Burger et al. (34)
Kaialy et al. (4,15,27)

10.6, 14.7, 23.4, 29.5

9.74, 22.2

468

Kaialy and Nokhodchi

Fig. 5 FT-IR spectra (a); and
PXRD patterns (b) for
commercial mannitol (CM), spray
dried mannitol (SDM), and freeze
dried mannitol (FDM).

of α-mannitol (1,194 cm−1), FT-IR specific band of βmannitol (929 cm−1 and 959 cm−1), FT-IR specific band of
δ-mannitol (967 cm−1) (Fig. 5a), PXRD specific peaks of αmannitol (9.6º, 13.8º, and 17.2º), PXRD specific peaks of βmannitol (10.6º, 14.7º, 23.4º, and 29.5º), and PXRD specific
peaks of δ-mannitol (9.74º and 22.2º) (Fig. 5b) (Table I). This
indicates that FDM crystallised as mixtures of α-mannitol
(~81%, w/w), β-mannitol (~11%, w/w), and δ-mannitol
(~8%, w/w) (Fig. 5b).

High peak resolution of PXRD patterns (Fig. 5b) in
addition to DSC thermal traces (Fig. 4) and SEM photos
(Fig. 2d–f) confirm the highly crystalline nature of all mannitol products. However, in comparison to CM, decreased
DSC melting enthalpy (Fig. 4) and lower PXRD peak intensities (Fig. 5b) for FDM sample suggest its lower relative
degree of crystallinity. This might be influenced by the
presence of 3.9±0.6% (v/v) of FPM<10μm in FDM sample
(Figs. 1a, 2f), as it is known that finer particles of the same

Inhaled Mannitol: Freeze-Dried vs Spray-Dried vs Commercial

material have smaller relative degree of crystallinity (9).
Also, it is known that δ-mannitol has smaller melting point
since it is enantiotropic toward α- and β- forms whereas βmannitol is monotropic toward α-form (34).
In conclusion, CM product was in β-mannitol form whereas SDM product was mixtures of α- and β-mannitol and FDM
sample crystallized as a mixture of α-, β-, and δ-mannitol
forms (Table II). Such results indicate the suitability of the
applied freeze drying method to prepare crystalline mannitol
product. α-, β-, and δ- mannitol polymorphic forms are stable
for minimum 5 years in dry atmosphere at 25°C.
It is believed that, during freeze drying, the intra- and
inter- hydrogen bonds of D-mannitol are broken by solvents
containing hydroxyl group (such as water) inducing crystal
form conversion. The presence of δ-mannitol in FDM sample might contribute to the elongated morphology of FDM
crystals since both particle shape and unite cell of δmannitol is elongated (oblong) (15). Such differences in
polymorphic form between different mannitol products is
important as it is known that different polymorphs of the
same compound could have different physical and chemical
properties and may lead to pharmaceutical product with
different characteristics (15).

Density and Flowability
CM showed similar true density (particle density) to the
value reported in literature for β-D-mannitol (Table II).
However, SDM and FDM particles showed different true
densities to that of CM (Table II) which could be ascribed to
their different molecular configuration induced by their
different polymorphic form (Table II; ref. 3,4) and different
size (Fig. 1a; ref. 8,9).
Unlike true density (which is particle characteristic), bulk
density and tap density are powder characteristics which are
indicative of powder packing and compaction properties.
Powder porosity refers to the voids within the powder bed
including spaces between agglomerates, between primary
particles, and micro-spaces (micropores) within the particles.
Table II Polymorphic from, True Density, Bulk Density, Tap Density,
Porosity, and Flow Character for Commercial Mannitol, Spray Dried
Mannitol, and Freeze Dried Mannitol (mean±SD, n≥4)
Physical
property

Commercial
mannitol

Spray dried
mannitol

Freeze dried
mannitol

Polymorphic form
True density (g/cm3)
Bulk density (g/cm3)
Tap density (g/cm3)
Porosity (%)
Flow character

β1.52±0.00
0.54±0.01
0.63±0.01
64.5±0.8
Good

α-+β1.45±0.01
0.46±0.00
0.53±0.01
68.4±0.2
Excellent

α-+β-+δ1.47±0.01
0.19±0.01
0.26±0.01
87.1±0.5
Poor

469

SDM powder showed smaller bulk density (apparent density), smaller tap density, and higher porosity than CM
(Table II). This might be due to spherical shape and relatively smaller size (8,9) of SDM particles in comparison to
CM particles (Figs. 1, 2). Higher bulk density, higher tap
density, and smaller porosity for CM powder in comparison
to SDM are indicative of relatively increased number of
interparticulate contacts between CM particles. Among
mannitol powders, FDM powder demonstrated the lowest
bulk density, the lowest tap density, and highest porosity
(Table II) indicating fewer points of physical contact between particles within FDM powder. This could be attributed to pronounced internal friction (interlocking ability)
within FDM powder due to the elongated most irregular
(anisometric) shape of FDM particles (Figs. 1e, 2f) resulting
in additional void space between particles within FDM powder. Such data suggest that interparticulate cohesive forces
between different mannitol powder were in the following rank
order: CM>SDM>FDM.
Among mannitol powders, SDM showed the best flow
properties (excellent flow character, CI014.0±0.8%) whereas
FDM showed the poorest flow properties (poor flow character,
CI027.2±1.2%) (Fig. 6a). It is believed that spherical shape of
SDM particles renders the SDM powder better flowability.
Poorer flowability for FDM powder could be ascribed to its
more irregular particle shape (Figs. 1e, 2f) which induce
pronounced internal friction (geometric interlocking) within
FDM powder. Also, the presence of fine particle mannitol in
FDM powder (Figs. 1a, 2f) could contribute to its poorer
flowability (8,9). Plotting CI of different mannitol powders
against mannitol particle sphericity and FPM<10μm indicated
better flowability in case of mannitol powders with more
higher shape sphericity and smaller fines content (Fig. 6b). It
is believed that higher fines content (higher FPM<10μm) and
poorer flowability (higher CI) of mannitol powder account for
its less homogeneous PSDs (higher span) as evident in Fig. 6c.
In fact, mannitol powders with poorer flowability are more
difficult to pass through mesh opening during sieving process.
Evaluation of Drug-Carrier Formulations
Drug Content Homogeneity
SS- CM and SS-FDM formulations produced similar (P>0.05)
potency with values ranging from 89.9±6.0% to 92.0±8.8%
of the nominal dose which fall within FDA and USP criteria for
content uniformity (85–115%) (Fig. 7a). However, SS-SDM
formulation produced considerably smaller potency (78.1±
2.5%) which exceed the acceptable range (Fig. 7a). This could
be ascribed to spherical shape of SDM particles (Figs. 1d, 2e)
which might facilitate “mechanical disconnection” of SS particles adhering on the mannitol surface during formulation
powder handling processes (e.g. mixing, vial filling, capsule

470

Kaialy and Nokhodchi

affect SS-mannitol blending process. In addition, poor homogeneity of SS-FDM formulation powder could be attributed
to higher PSD polydispersity of FMD (Figs. 1b, 2f), which
might promote percolation segregation leading to the formation of drug-rich areas (higher amounts of drug per unite
mass) within SS-FDM formulation (9). Figure 7b shows that
the wider PSD (higher span) and the poorer the mannitol
powder flowability (higher CI) the poorer the SS content
homogeneity within DPI formulation (higher %CV). Such
data indicate that in DPI mixtures, the efficiency of DPI
formulation blending and thus achieving uniform stable ordered mixture with homogeneous drug content (and therefore
uniform metering doses by the patient) is deeply affected by
the flow/size polydispersity properties of the carrier.
Since blending process might have a significant influence
on interparticulate interactions within a powder formulation
and thus DPI performance (12), no blending-optimization was
conducted for each formulation in order to meet the comparison purpose of this study.
Drug-Carrier Adhesion Assessments. In DPIs, drug inhalation
behaviour depends on the balance between removal forces
(inertial forces: e.g. particle-particle collision and particleinhaler wall collisions; shear forces: e.g. particle-inhaler wall
friction; and lift or drag forces: e.g. forces within turbulent air

Fig. 6 Carr’s index (CI) (a); CI in relation to Sphericity and fine particle
mannitol (FPM<10μm) (b); and Span in relation to CI and FPM<10μm (c); for
commercial mannitol (CM), spray dried mannitol (SDM), and freeze dried
mannitol (FDM) (mean±SD, n≥4).

filling, etc.) and thus might lead to substantial amount of drug
being “lost” on the surface of mixing cylinder, glass vial
container, etc.
CV (%) of SS content obtained from different formulations
was within the following rank order: SS-SDM<SS- CM<SSFDM (Fig. 7a). In case of SS-CM and SS-SDM formulations,
the % CV of SS distribution was <6%, suggesting that the SS
particles were homogeneously distributed throughout the formulation powder mixture. However, FDM generated heterogeneous DPI mixture (least uniform SS content, CV>6%).
Poor flowability of FDM powder (Fig. 6a) might contribute to
poor homogeneity of SS-FDM formulation (7) since it might

Fig. 7 % Potency (Pink Square) and % coefficient of variation (CV) (Black
Circle) of salbutamol sulphate (SS) content (a); and % CV of SS in relation to
and Carr’s index (CI) and span of mannitol (b) for SS-mannitol formulations
containing commercial mannitol (CM), spray dried mannitol (SDM), and
freeze dried mannitol (FDM) (mean±SD, n07) (Asterisk indicates statistically different (P<0.05)).

Inhaled Mannitol: Freeze-Dried vs Spray-Dried vs Commercial

stream) and interaction forces (e.g. Van der wall forces) during
inhalation. Air jet sieving could be better than AFM in characterizing drug-carrier interparticulate forces since AFM has
the drawback of considering only single particle rather than the
overall powder. During air-jet sieving, drug particles are
expected to separate from carrier particles. During air jet
sieving, particles are submitted to both aspiration (generated
by negative pressure) and airflow (generated by the blow
nozzle rotating under the sieve). Assuming particle adhesive
forces are equivalent to particle removal forces (15), less
amount of SS remained in powder on top of the sieve indicate
weaker SS-mannitol adhesion. For all formulations, the
amounts of drug decreased with increasing the functional
sieving time (Fig. 8a). Logarithmic relationships were obtained
when plotting amounts of SS remained against sieving time
(r2 ≥0.98, figures not shown) from which T50% values were
calculated defined elsewhere (27) (Fig. 8a). CM produced the
highest amounts of SS following all sieving times (Fig. 8a)
indicating greatest degree of SS-mannitol adhesive forces. This
Fig. 8 Percent amounts of salbutamol sulphate (SS) remaining on top
of the 20 μm sieve and T50%
obtained from formulations containing commercial mannitol (CM),
spray dried mannitol (SDM), and
freeze dried mannitol (FDM) after
different functional sieving times: 5 s
(Black Diamond), 30 s (Red Square),
60 s (Green Triangle), and 180 s
(Blue Circle) (a); % remained SS in
relation to root mean square average (Rq) roughness on a 5 μm region and 300 nm region of mannitol
particle surface (mean±SD, n≥3)
(b).

471

could be attributed to the larger size of CM particles in
comparison to other mannitol particles (Fig. 1a), which is
expected to generate higher press-on forces (the forces that
press the drug particles onto the carrier particles) during mixing process which acts as adhesive forces (12,15). Also, it can be
assumed that detachment of small SS particles from large
mannitol particles occurs laterally on the mannitol particle
surface (SS particles slip along the mannitol particle surface till
they reach the edge and falls off). Therefore, the distance that
the drug particles have to slip on mannitol carrier surface
increase with particle size of carrier mannitol, thus the greater
the aerodynamic drag force (Fdrag) which plays as adhesive
force between SS and mannitol particles.
After all functional sieving times, FDM produced smaller
amounts of SS than SDM which in turn produced smaller
amounts of SS than CM (Fig. 8a). Such findings indicate that
SS particles attached to FDM carrier particle surfaces detached more easily than SS particles adhered to SDM carriers,
from which SS drug particles separated more easily in

472

comparison to CM particles. T50% obtained from different
formulations was in the following rank order according to
mannitol product: CM (155.5 s)>SDM (52.8 s)>FDM (3.9 s)
(Fig. 8a). This confirms that SS-mannitol adhesive forces within different formulations were within the following hierarchy
according to mannitol product: FCM>FSDM>FFDM (Fig. 8a).
Deceased drug-carrier adhesive forces in case of FDM could
be attributed to the existence of fine mannitol particles
(Figs. 1a, 2f), increased particle shape irregularity (Fig. 2),
and/or the smoother particle surface (Fig. 3) in the case of
FDM, all of which can contribute to reduced contact area
between SS particles and mannitol particles (increased separation distance) and decreased push-on forces (12,27,35). On the
other hand, SS particles adhered in a deep concavity on
mannitol surface, as in case of CM (Fig. 3), would become
entrapped and relatively immobile in the macroscopic
depressions.
Fine mannitol particles are proposed to contribute to weak
drug-carrier adhesion by saturation of “active sites” (high
adhesion sites) on the coarse mannitol particles, which leaves
the passive sites (low adhesion sites) available for SS adhesion.
Also, it can be assumed that fine mannitol particles physically
disrupt the SS-mannitol contacts leading to smaller SSmannitol adhesion. Moreover, based on “agglomeration theory” (7,10), the presence of fine particle mannitol in FDM
sample (Fig. 2f) could promote the formation of SS-fine mannitol mixed agglomerates. These agglomerates are expected
to be e easier to detach from carrier surface upon inhalation
than single SS particles. It is believed that SS-fine mannitol
mixed agglomerates might form at the expense of SS-coarse
mannitol ordered mixtures (interactive mixtures), and this
leads to the reduction of amounts of single SS particles attached to larger mannitol particles and thus smaller SSmannitol total adhesion (Fig. 8a).
The adhesion of particles is a surface phenomenon and
therefore, the drug-carrier adhesion is deeply affected by the
surface morphology of both carrier particles and drug particles. The existence of different surface morphologies (disturbance on the crystal surface) between different mannitol
particles is likely to result in different physicochemical properties including different “adhesive potential”. Fig. 8b shows
that the smoother the mannitol particle surfaces (on both
5 μm and 300 nm mannitol surface region) the weaker the
drug-carrier adhesive forces. This could be ascribed to lower
contact area and thus lower adhesive forces between the
adjacent drug particles and carrier particles with smoother
surface topography (36). Also, it can be assumed that the use
of carrier particles with smoother surfaces could be related
to a less binding sites with multiple contact points (10,37). It
has been suggested that the rougher the carrier particles, the
greater number of active sites which are capable of binding
the drug particles more strongly and as a result the higher
proportion of drug adhering to the carrier particles (38).

Kaialy and Nokhodchi

Decreased SS-mannitol adhesion in case of SS-FDM
(Fig. 8a) formulation might contribute to easier segregation
of SS from mannitol surfaces and thus poor content homogeneity of SS within SS-FDM formulation (Fig. 7a).
In Vitro Aerosolization Performance. Stage by stage mass distribution deposition profiles of SS varied considerably according to the type of mannitol product (Fig. 9a). Amounts of SS
adhered to inhaler wall with mouthpiece adaptor (residual
drug, drug loss, or device retention) ranged from 3.7±0.2%
to 6.2±5.8% (Fig. 9a). Such amounts could be attributed to
electrostatic attraction. Also, it can be assumed that SS particles detach from mannitol surfaces within the inhaler device.
FDM deposited considerably (P>0.05) higher amounts of SS
on throat (9.8±0.6%) in comparison to CM and SDM (3.0±
0.7%) (Fig. 9a). This could be attributed to the presence of fine
particle mannitol (FPM<10μm) (Figs. 1a, 2f) and poor flowability (Fig. 6a) in case of FDM powder, which promote the
formation of FPM <10μm-FPM <10μm and FPM<10μm -SS
agglomerates depositing on throat by inertial impaction
(7,39), especially when using low resistance device such as
Aerolizer® being used in this study (40). Amounts of drug
deposited on I&M and IP are believed to be eventually swallowed and consequently systemically absorbed via the GIT
(Fig. 9a).
In case of SS-CM and SS-SDM aerosol formulations, SS
particles deposited preferentially on stage 1 (cut off diameter0
10.5 μm) representing the upper airways (tracheobronchial
deposition) (Fig. 9a). However, SS-FDM aerosol formulation
exhibited maximal deposition on stages: 2 (representing bronchial airways), 3 (representing peripheral alveolar airways),
and 4 (representing deep lung airways) with no substantial
difference between these stages (19.2±2.3–23.8±2.7%, P>
0.05) (Fig. 9a). All aerosol formulations exhibited minimal
deposition on MsLI-filter with FDM generating significantly
higher amounts of SS than CM (Fig. 9a). The aerodynamic
PSD of SS analyzed by MsLI as obtained from different SSmannitol aerosol formulations is shown in Fig. 9b. All SSmannitol aerosol formulations generated linear log aerodynamic PSD plots (r2 ≥0.904) confirming the absence of particle
bounces and re-entrainment during aerosolization (Fig. 9b).
The slops of these aerodynamic PSD linear regressions was
named as constant K, which could be used as a parameter to
indicate DPI performance (9,14). In determining aerosol PSD,
aerodynamic size is more suitable than geometric size. When
compared to geometric particle size (raw SS powder before
aerosolization, measured by laser diffraction, K092.9), SSmannitol aerosol formulations generated larger aerodynamic
size distributions (after aerosolization, analysed by MsLI, K0
21.2–58.9) (Fig. 9b). This indicates that SS particles were not
sufficiently dispersed during inhalation to recover the primary
(individual) particles. This could be caused by insufficient SSSS deagglomeration (high SS-SS cohesive forces) and/or

Inhaled Mannitol: Freeze-Dried vs Spray-Dried vs Commercial

473

Fig. 9 Amounts of salbutamol
sulphate (SS) deposited on inhaler
plus mouthpiece adaptor (I&M),
induction port (IP), and different
MsLI stages (a); (White Circle)
geometric particle size (before
aerosolization) and aerodynamic
particle size (after aerosolization)
of SS obtained from formulations
containing commercial mannitol
(CM) (Blue Diamond), spray dried
mannitol (SDM) (Red Square), and
freeze dried mannitol (FDM)
(Green Triangle) (b).(mean±SD,
n03). (Asterisk indicates statistically different: P<0.05).

Table III Recovered Dose (RD), Emitted Dose (ED), Mass Median
Aerodynamic Diameter (MMAD), Geometric Standard Deviation (GSD),
Fine Particle Dose (FPD), Impaction Loss (IL), Fine Particle Fraction (FPF),
and Dispersibility (DS) of Salbutamol Sulphate Obtained from Formulations
Containing Commercial Mannitol, Spray Dried Mannitol, and Freeze Dried
Mannitol (mean±SD, n03)
Deposition
parameter

Commercial
mannitol

Spray dried
mannitol

Freeze dried
mannitol

RD (μg)
ED (μg)
MMAD (μm)
GSD
FPD (μg)
IL (%)
FPF (%)
DS (%)

381.5±8.3
367.3±8.8
3.0±0.2
2.2±0.0
64.3±6.5
75.7±0.7
16.8±1.3
17.5±1.4

394.2±23.9
350.6±26.6
2.6±0.1
2.2±0.0
94.6±14.4
62.6±3.6
24.0±2.7
17.5±1.4

465.9±8.2
428.1±4.9
3.2±0.2
2.1±0.1
218.6±20.3
33.6±2.3
46.9±3.6
26.9±2.6

insufficient SS-mannitol deaggregation and/or inadequate
dispersing efficiency of the inhaler device.
The aerodynamic PSD of SS particles generated from SSFDM formulations (K058.9) was closer to the primary PSD of
pure SS powder than that of the aerodynamic PSD generated
from SS-SDM formulations (K029.0) which in turn generated
smaller aerodynamic PSD than SS-CM formulations (K0
21.2) (Fig. 9b). This indicates that FDM generated smaller
aerodynamic size of SS than SDM which in turn generated
smaller aerodynamic size of SS than CM upon aerosolization.
FDM deposited larger proportion of SS dose to the central
and lower airways (or the impactor) than SDM from which SS
particles are expected to reach lower airway regions (which
are the site of action in case of SS) in comparison to SS
particles formulated with CM carrier.
Apart from in vivo lung drug dose or drug safety, RD and
ED are considered important quality control indicators of
pharmaceutical performance. The RD of SS obtained from

474

all formulations was within the range between 381.5±8.3 μg
and 465.9±8.2 μg corresponding to % recovery between
79.3±1.7% and 96.9±1.7%, which is within the acceptable
range of % recovery (75–125%) (Table III), supporting the
highly dispersible nature of all mannitol powders. Also, this
suggests good reproducibility and reliability of the overall
procedures including mixing, sampling, capsule filling, deposition, washing and the analysis of SS were satisfactory accurate. FDM generated higher RD and higher ED of SS than
other mannitol products (Table III) indicating higher dose of
SS available to the patient or the impactor when delivered
from the inhaler device. Upon aerosolisation, the formulation
powder is forced out from the capsules through the pierces
and pulled into the oral cavity by the drag forces exerted upon
the particles generated from the airflow. High RD and ED in
case of SS-FDM formulation reflect the appropriate aerodynamic properties of FDM powder and could be ascribed to
reduced protuberances on FDM particle surface (Fig. 3) which
may reduce the SS-mannitol geometrical interlocking leading
to improved SS-mannitol deaggregation within the inhaler
device during aerosolization. On the other hand, smaller
RD and RD in case of CM and SDM could be ascribed to
their higher bulk density (Table II), since it is known that
cohesive powders are more difficult to fluidize via airflow (lift
as fractures or plugs) than less cohesive powders, such as FDM
(Table II), which fluidize more homogeneously by an erosion
mechanism.
A MMAD in the range of 1–5 μm is very important to
achieve efficient pulmonary drug delivery. In this study,
MMAD of SS ranged between 2.6±0.1 μm and 3.2±0.2 μm
with GSD of 2.2±0.1 indicating polydisperse (or heterodisperse) PSD of aerosolized SS (GSD>1.2) (Table III). Although
the same batch of SS was used in preparation of all formulations, SDM generated smaller MMAD of SS in comparison
than CM and FDM (Table III) indicating smaller drug agglomeration. The presence of fine particle mannitol in case of
FDM (Figs. 1a, 2f) might account for increased MMAD in case
of SS-FDM in comparison to SS-SDM due to the formation of
SS-FDM<10μm aggregates (7). Also, rougher surface of CM
(Fig. 3) might contribute to higher MMAD in case of SS-CM
in comparison to SS-SDM (10). In theory, judging from geometric and aerodynamic particle size of SS (Fig. 9b), all SS
particles are supposed to deposit on lower stages of the MsLI.
However, IL of SS ranged between 33.6±2.3% (SS-FDM) and
75.7±07% (SS-CM) (Table III). This indicates that the operating flow was not sufficient to generate enough energy input
generated by the air stream through the devise and thus not all
of SS particles has detached from mannitol surfaces. By examining only the MMAD values (Table III), it might be thought
that SS-SDM formulation would demonstrate the best aerosolisation efficiency across other formulations. However; the following rank order could be formulated for FPD, FPF, and DS
obtained from different formulations according to mannitol

Kaialy and Nokhodchi

product: CM<SDM<FDM (Table III) whereas the reverse
order could be observed in respect to IL (Table III). Drugcarrier aerosolization performance is, mainly, dependent on
two factors: 1) drug-drug cohesion forces (drug-drug self agglomeration which is reflected by MMAD) and 2) drugcarrier adhesion forces (drug- coarse carrier aggregation). In
fact, smallest MMAD obtained for SS-SDM formulation indicate smallest degree of SS-SS cohesion (SS self agglomeration) but it does not mean necessarily smallest degree of SSSDM adhesion (SS-SDM aggregation), which is proved by
drug-carrier adhesion assessments (Fig. 8). Depending on the
drug-drug cohesion and drug-carrier adhesion balance, formulations with lower MMAD still can produce higher FPF
and vice versa (7–9).
Wider variation in FPF in the case of SS-FDM (SD03.6),
in comparison to SS-CM (SD01.3) and SS-SDM (SD02.7)
formulations (Table III), could be related to poorer drug
content homogeneity of SS-FDM formulation in comparison
to other formulations (Fig. 7a). It is clear that FDM generated
~2.8 fold larger FPF than CM and ~2.0 fold increase in FPF
than SDM. This indicates that SS particles attached to
elongated-FDM were easier to disperse than SS particles
attached to spherical-SDM from which SS particles were
easier to disperse than angular-CM. This confirms superior
DPI performance of FDM since it is known that higher FPF
does not only indicate enhanced therapeutic efficiency and a
decrease in the dosage required for asthma treatment, but also
a minimized side effects and a promoted patient compliance.
It is believed that the SS-FDM formulation presented in this
study showed the best aerosolization performance of SS
reported in literature so far (fine particle fraction of 47%
and dispersibility of 51%, in the case of SS-carrier DPI formations with 63–90 carrier particles, <5 μm drug particles,
Aerolizer® inhaler device, flow rate of 92 L/min, and 67.5:1
(w/w) drug-carrier ratio). This suggests that FDM powder
may even be considered in aerosol formulations for local
and systemic drug delivery.
Since all formulations were prepared under similar protocols, differences in aerosolization performance between different SS-mannitol aerosol formulations may be attributed to
different mannitol products. Regression analysis showed that
the aerosolization performance obtained from different SSmannitol formulations demonstrated poor relationships with
VMD, aspect ratio, flakiness ratio, sphericity, simplified shape
factor, and Carr’s index of mannitol. This suggests that, at
least within data in this study, there is no simple apparent
relationship between size, shape, and/or flowability of mannitol and in vitro aerosolization performance of SS. In fact, the
effect of carrier particle size, shape, and flowability as one
variable on drug aerosolisation efficiency from DPIs was
reported in contrary manner as explained previously. However, Both FPF and IL exhibited good relationship with T50%
(Fig. 10a) indicating that the weaker the SS-mannitol adhesive

Inhaled Mannitol: Freeze-Dried vs Spray-Dried vs Commercial

Fig. 10 Fine particle fraction (FPF) (White Blue Circle), impaction loss (IL)
(Black Circle) of salbutamol sulphate (SS) (a); and FPF of SS in relation to
porosity of mannitol powder (Black Diamond) (b) obtained from formulations containing commercial mannitol (CM), spray dried mannitol (SDM),
and freeze dried mannitol (FDM).

forces the smaller the amounts of SS attached to the carrier
following aerosolization and the increased amounts of drug
delivered to lower airways.
During inhalation, two main mechanisms are believed to
control SS-mannitol detachment : detachment by the flow
stream (flow detachment, fluid forces) and detachment by
impaction (mechanical detachment, mechanical forces). In
the case of SS-FDM formulation, detachment by flow is
expected to be the dominate mechanism, since such detachment dominates in the case of smooth surface–carrier particles
(Fig. 3). However, detachment by mechanical forces is believed to be more relevant is case of SS-CM formulation since
such detachment relies on the abrupt momentum transfer
resulting from particle-inhaler collisions and it is facilitated
for carrier particles with rougher surfaces (Fig. 3) (16).
It should be acknowledged that particles with lower density
are more advantageous for inhalation due to their smaller
aerodynamic diameter in comparison to their geometric diameter. The use of particles with smaller density is likely to lead to
reduced probability of deposition by inertial impaction and
sedimentation (increased residence time). It is believed that, in
determining aerosol PSD, particle aerodynamic size is more

475

suitable than geometric size. Despite that SDM and FDM
particles have similar geometric diameter (Fig. 1b), they demonstrated considerably different aerodynamic diameters: 59.9
±0.8 μm and 41.3±0.6 μm for SDM and FDM respectively.
Superior DPI performance of FDM in comparison to SDM
could be attributed to smaller density and higher porosity of
FDM powder (Table II) which gives rise FDM powder smaller
interparticulate cohesive forces in dry state and reduced setting
velocity in aerosolized state (more airborne) promoting deeper
penetration in lung airways. By comparing different formulations, the higher the mannitol powder porosity the higher are
the amounts of drug delivered to lower airway regions (Fig.
10b). It can be suggested that mannitol particle physical properties (e.g. morphology) manipulate fine particle delivery to
lower airway regions due to their effect on powder bulk properties. Interactions between SS particles, SS agglomerates, SSFDM<10μm mixed agglomerates, and SS-mannitol ordered
units with single- or multi-SS particles could be impacted by
the space size between carrier particles, as supported by percolation theory (40). It is believed that the higher the porosity of
the carrier powder the easier is the powder dispersion
(segregation) upon aerosolization and the smaller the inertial
forces will be generated during aerosolization leading higher
aerosol mass deposited on lower airway regions (since the
inertial impaction is the predominant deposition mechanism
in the lungs). On the other hand, powders with smaller porosity
have higher interparticulate forces due to increased number of
contact points. These forces must be overcome to produce a
dispersed aerosol powder upon inhalation. Therefore, excessive cohesive forces might results in poor dispersion properties
because of enhanced particle aggregation which in turn
decreases particle fluidization. Finally, it was noted that solid
state, size, and aerosolization performance of different mannitol products has not significantly changed following being
stored in ambient conditions (22±1°C, 50% RH) for 6 months
(data not shown). Additional systemic stability studies should be
performed at elevated temperature and humidity conditions to
evaluate variations in physicochemical and inhalation performance of freeze dried mannitol.
Although different mannitol products showed different
aerosolization performance, there remain substantial uncertainties regarding carrier functionality in DPI formulations.
The theories by which carrier physical properties control DPI
performance remain speculative. Full understanding of the
relationship between carrier physical properties and DPI performance is still challenging considering the heterogeneous
nature of the carriers used in different studies and the high
possibility of interaction between different carrier parameters.
Although the benefits of using freeze dried mannitol in dry
powder aerosol formulation is clear, more efforts would be
warranted to comprehensively evaluate the aerosolization
performance of freeze dried mannitol with wider range of
drug particles, inhalers, and inhalation flow rates. Also, in vivo

476

aerosolization performance assessments would be required to
verify if in vitro deposition data correlate well with in vivo
deposition data which is necessary to confirm the use of
freezing drying of mannitol as a robust and reliable means of
improving aerosolization performance of dry powder inhalers.

CONCLUSION
The work proposed is novel in that it demonstrates another
example of the potential of using freeze drying technique in
pharmaceutical industry to prepare freeze dried mannitol
powders that displayed improved aerosolization performance
of salbutamol sulphate from dry powder inhaler formulations.
Freeze dried mannitol product was shown to be virtually in
complete crystalline nature. For the first time, it was shown
that freeze dried mannitol (elongated shape, α-+β-+δ- polymorphic form) produce better aerosolization performance
than spray dried mannitol (spherical shape, α-+β- polymorphic form) which in turn demonstrated better aerosolization performance than commercial mannitol (angular shape,
β- polymorphic form).
In comparison to commercial mannitol and spray dried
mannitol, freeze dried mannitol showed the highest variability
in terms of size, shape, solid state, dose homogeneity, and fine
particle fraction. Freeze dried mannitol did not show smaller
geometric size than spray dried mannitol, however, demonstrated the highest powder porosity. Freeze dried mannitol
demonstrated smoother surface morphology than spray dried
mannitol which in turn demonstrated smoother surface morphology than commercial mannitol. Freeze dried mannitol
generated the weakest salbutamol sulphate-mannitol adhesive
forces whereas commercial mannitol generated the highest
SS-mannitol adhesive forces. It was clear that the smoother
the mannitol surface the weaker the salbutamol sulphatemannitol adhesive forces.
Among angular, spherical, and elongated shaped mannitol
particles, formulators can anticipate better drug delivery to
the lung in case of elongated shape mannitol. No apparent
relationship was obtained between fine particle fraction and
mannitol size, shape, or flowability descriptors. However,
mannitol products with higher powder porosity and weaker
salbutamol sulphate-mannitol adhesive forces produced
higher fine particle fraction of salbutamol sulphate. It was
suggested that porosity of carrier powder is an important
physical property which can be considered as a key optimization parameter which might be predictive of in vitro aerosolization performance of dry powder inhaler formulations.
Mannitol powders with poorer flowability and higher fines
content demonstrated wider particle size distributions following sieving. Mannitol powders with less spherical particle
shape and higher fines content demonstrated poorer flow

Kaialy and Nokhodchi

properties. Better drug content homogeneity was obtained in
case of mannitol powders with better flow properties and
narrower size distributions.
Freeze drying of aqueous mannitol solutions is an attractive
approach to prepare dry powder aerosol formulations due to
its several advantages including enhanced pulmonary drug
delivery, maximal yield, simple, low cost effective, and low
safety risk, since no organic solvents were used. The use of
freeze drying technique can constitute an important step used
in the pharmaceutical industry towards preparing freeze dried
carrier particles which could help to solve some problems
connected to drug-carrier dry powder aerosol formulations.
ACKNOWLEDGMENTS AND DISCLOSURES
Waseem Kaialy thanks Dr. Ian Slipper (University of Greenwich) and Mr. Ian Brown (University of Kent) for help provided with SEM and AFM analysis respectively.

REFERENCES
1. Blagden N, De Matas M, Gavan P, York P. Crystal engineering of
active pharmaceutical ingredients to improve solubility and dissolution rates. Adv Drug Deliv Rev. 2007;59:617–30.
2. Aquino R, Prota L, Auriemma G, Santoro A, Mencherini T,
Colombo G, Russo P. Dry powder inhalers of gentamicin and
leucine: formulation parameters, aerosol performance and in vitro
toxicity on CuFi1 cells. Int J Pharm. 2012;426:100–7.
3. Kaialy W, Martin GP, Ticehurst MD, Momin MN, Nokhodchi A.
The enhanced aerosol performance of salbutamol from dry powders
containing engineered mannitol as excipient. Int J Pharm. 2010;
392:178–88.
4. Kaialy W, Momin MN, Ticehurst MD, Murphy J, Nokhodchi A.
Engineered mannitol as an alternative carrier to enhance deep
lung penetration of salbutamol sulphate from dry powder inhaler.
Colloid Surf B. 2010;79:345–56.
5. Kaialy W, Martin GP, Ticehurst MD, Royall P, Mohammad MA,
Murphy J, Nokhodchi A. Characterisation and deposition studies
of recrystallised lactose from binary mixtures of ethanol/butanol
for improved drug delivery from dry powder inhalers. AAPS J.
2011;13:30–43.
6. Kaialy W, Ticehurst MD, Murphy J, Nokhodchi A. Improved
aerosolization performance of salbutamol sulfate formulated with
lactose crystallized from binary mixtures of ethanol—acetone. J
Pharm Sci. 2011;100:2665–84.
7. Kaialy W, Martin GP, Larhrib H, Ticehurst MD, Kolosionek E,
Nokhodchi A. The influence of physical properties and morphology of crystallised lactose on delivery of salbutamol sulphate from
dry powder inhalers. Colloid Surf B. 2012;89:29–39.
8. Nokhodchi A, Kaialy W, Ticehurst MD. The influence of particle
physicochemical properties on delivery of drugs by dry powder
inhalers to the lung. In: Popescu MA, editor. Drug delivery book.
NY: Hauppauge: Nova; 2011. p. 1–50.
9. Kaialy W, Alhalaweh A, Velaga SP, Nokhodchi A. Influence of
lactose carrier particle size on the aerosol performance of budesonide
from a dry powder inhaler. Powder Technol. 2012;227:74–85.
10. Podczeck F. The relationship between physical properties of lactose
monohydrate and the aerodynamic behaviour of adhered drug particles. Int J Pharm. 1998;160:119–30.

Inhaled Mannitol: Freeze-Dried vs Spray-Dried vs Commercial
11. Steckel H, Markefka P, TeWierik H, Kammelar R. Effect of
milling and sieving on functionality of dry powder inhalation
products. Int J Pharm. 2006;309:51–9.
12. Dickhoff B, De Boer A, Lambregts D, Frijlink H. The effect of
carrier surface and bulk properties on drug particle detachment
from crystalline lactose carrier particles during inhalation, as function of carrier payload and mixing time. Eur J Pharm Biopharm.
2003;56:291–302.
13. Donovan MJ, Smyth HDC. Influence of size and surface roughness of large lactose carrier particles in dry powder inhaler formulations. Int J Pharm. 2010;402:1–9.
14. Kaialy W, Alhalaweh A, Velaga SP, Nokhodchi A. Effect of carrier
particle shape on dry powder inhaler performance. Int J Pharm.
2011;421:12–23.
15. Kaialy W, Larhrib H, Ticehurst MD, Nokhodchi A. Influence of
batch cooling crystallization on mannitol physical properties and
drug dispersion from dry powder inhalers. Cryst Growth Des.
2012;12:300–3017.
16. Le VNP, Thi THH, Robins E, Flament M. Dry powder inhalers:
Study of the parameters influencing adhesion and dispersion of
fluticasone propionate. AAPS PharmSciTech. 2012;13:477–84.
17. Traini D, Young PM, Thielmann F, Acharya M. The influence of
lactose pseudopolymorphic form on salbutamol Sulfate–Lactose interactions in DPI formulations. Drug Dev Ind Pharm. 2008;34:992–
1001.
18. Byron PR, Jashnam R. Efficiency of aerosolization from dry powder blends of terbutaline sulfate and lactose NF with different
particle-size distributions. Pharm Res. 1990;7:881.
19. Donovan MJ, Kim SH, Raman V, Smyth HD. Dry powder inhaler
device influence on carrier particle performance. J Pharm Sci.
2012;101:1097–107.
20. Shur J, Harris H, Jones MD, Kaerger JS, Price R. The role of fines
in the modification of the fluidization and dispersion mechanism
within dry powder inhaler formulations. Pharm Res. 2008;25:
1631–40.
21. Zeng XM, Martin GP, Marriott C, Pritchard J. Lactose as a
carrier in dry powder formulations: the influence of surface characteristics on drug delivery. J Pharm Sci. 2001;90:1424–34.
22. Schneid S, Riegger X, Gieseler H. Influence of common excipients
on the crystalline modification of freeze-dried mannitol. Pharm
Technol. 2008;3:32.
23. Takada A, Nail SL, Yonese M. Influence of ethanol on physical
state of freeze-dried mannitol. Pharm Res. 2009;26:1112–20.
24. Hottot A, Nakagawa K, Andrieu J. Effect of ultrasound-controlled
nucleation on structural and morphological properties of freeze-

477

25.

26.

27.

28.
29.
30.
31.
32.
33.
34.

35.

36.
37.

38.

39.

40.
41.

dried mannitol solutions. Chem Eng Res Design. 2008;86:193–
200.
Kim AI, Akers MJ, Nail SL. The physical state of mannitol after
freeze–drying: effects of mannitol concentration, freezing rate, and
a noncrystallizing cosolute. J Pharm Sci. 1998;87:931–5.
Kaialy W, Ticehurst MD, Nokhodchi A. Dry powder inhalers:
mechanistic evaluation of lactose formulations containing salbutamol sulphate. Int J Pharm. 2012;423:184–94.
Kaialy W, Larhrib H, Martin GP, Nokhodchi A. The effect of
engineered mannitol-lactose mixture on dry powder inhaler performance. Pharm Res. 2012;29:2139–56.
Krumbein WC. Measurement and geological significance of shape and
roundness of sedimentary particles. J Sediment Res. 1941;11:64–72.
Fronczek FR, Kamel HN, Slattery M. Three polymorphs (α, β,
and δ) of D-mannitol at 100 K. Acta Crystallogr. 2003;59:567–70.
Hinds WC. Aerosol technology: properties, behavior and measurement of airborne particles. New York: Jennings: Wiley; 1982.
Edwards DA. Delivery of biological agents by aerosols. AIChE J.
2002;48:2–6.
Hickey AJ. Pharmaceutical inhalation aerosol technology. New
York: Marcel Dekker; 2004.
Ho R, Wilson DA, Heng JYY. Crystal habits and the variation in
surface energy heterogeneity. Cryst Growth Des. 2009;9:4907–11.
Burger A, Henck JO, Hetz S, Rollinger JM, Weissnicht AA, Stöttner
H. Energy/temperature diagram and compression behavior of the
polymorphs of D-mannitol. J Pharm Sci. 2000;89:457–68.
Le VNP, Bierend H, Robins E, Steckel H, Flament MP. Influence of
the lactose grade within dry powder formulations of fluticasone
propionate and terbutaline sulphate. Int J Pharm. 2011;422:75–82.
Ganderton D. The generation of respirable clouds from coarse
powder aggregates. J Biopharm Sci. 1992;3:101–5.
Kassem NM, Ganderton D. The influence of carrier surface on the
characteristics of inspirable powder aerosols. J Pharm Pharmacol.
1990;42(11).
Staniforth JN, Rees JE, Lai FK, Hersey JA. Interparticle forces in
binary and ternary ordered powder mixes. J Pharm Pharmacol.
1982;34(3):141–5.
Le V, Robins E, Flament M. Agglomerate behaviour of fluticasone
propionate within dry powder inhaler formulations. Eur J Pharm
Biopharm. 2012;80:596–603.
Dunbar CA, Hickey AJ, Holzner P. Dispersion and characterization of pharmaceutical dry powder aerosols. Kona. 1998;16:7–45.
Yu L, Mishra DS, Rigsbee DR. Determination of the glass properties of D–mannitol using sorbitol as an impurity. J Pharm Sci.
1998;87:774–7.