Chemie

Ingenieur Research Article 1

Technik

Formation of Mefenamic Acid

Nanocrystals with Improved Dissolution
Characteristics
Christoph Konnerth1,3, Veronika Braig2,3, Atsutoshi Ito1, Jochen Schmidt1, Geoffrey
Lee2,3, and Wolfgang Peukert1,3,*
DOI: 10.1002/cite.201600190

An industrially feasible formulation approach combining media milling and spray-drying was applied to improve
dissolu- tion characteristics of the poorly soluble drug mefenamic acid (MA). The approach was studied for two MA
polymorphs at different stressing and pH conditions. It was found that final MA product particle sizes are rather
determined by the solid-liquid equilibrium than by mechanical fracture. Obtained drug particles are only composed of the
most stable poly- morph. Direct compressed tablets containing MA nanocrystals exhibit a significant improvement of in
vitro dissolution kinetics as compared to tablets with micronized drug particles.
Keywords: Dissolution rate enhancement, Media milling, Nanocrystal, Polymorphism, Poorly water-soluble drug,
Spray drying
Received: December 21, 2016; revised: April 20, 2017; accepted: May 05, 2017

performance, and bio- availability [16, 17].

1 Introduction

The number of potential drug candidates with unfavorable

physicochemical and poor biopharmaceutical properties
increased tremendously over the past decades due to
advan- ces in high-throughput screening methodologies
and com- binatorial chemistry [1, 2]. Especially poor
aqueous solubili- ty is a main hurdle in drug development,
because it is often related to a limited dissolution rate and,
thus, to an insuffi- cient bioavailability after oral
administration [3, 4]. Hence, a wide variety of formulation
strategies such as co-crystal and salt formation, lipid
formulations (e.g., self-emulsifying drug delivery systems,
solid lipid nanoparticles), solubiliz- ing techniques,
complexation (e.g., cyclodextrines), and amorphous solid
dispersions are applied to improve disso- lution and
biopharmaceutical characteristics of problematic drug
compounds [5 – 7].
Whereas comminution processes like micronization of
active pharmaceutical ingredients (APIs) are well estab-
lished unit operations in the pharmaceutical industry since
several decades, the formation of drug nanocrystals via
media milling became a promising formulation approach to
address these needs in recent years [8 – 11]. According to
Noyes and Whitney [12], the dissolution rate of a solid API
is proportional to the specific surface area available for dis-
solution. An increase in specific surface area due to size
reduction down to the nanometer range leads to a signifi-
cant increase in dissolution rate of the drug, which is often
related to an improved oral bioavailability and reduced
gas- tric irritancy [13 – 15]. Therefore, the control of
particle size during drug development and manufacturing
is crucial with respect to handling, shelf-life, drug
Chem. Ing. Tech. 2017, 89, No. 8, 1060– ª 2017 WILEY-VCH Verlag GmbH & Co. KGaA, www.cit-journal.com
1071 Weinheim

www.cit-journal.com
 Chemie
2 Research Article Ingenieur
Technik

storage stability [25 – 28]. For nanosuspensions, a drying

In media milling, particles are stressed between agitated step can avoid agglomeration and, therefore, improve stor-
grinding beads. During collisions the kinetic energy of the
beads is transferred to internal stress and strain in the feed –
particles and, consequently, product formation occurs by 1
Christoph Konnerth, Atsutoshi Ito, Dr. Jochen Schmidt,
fracture. However, also liquid phase and formulation medi- Prof. Dr.-Ing. Wolfgang Peukert
ated growth effects, induced by dissolution, precipitation, [email protected]
and ripening phenomena may influence product particle Friedrich-Alexander-Universita¨t Erlangen-Nu¨rnberg (FAU),
size and shape of APIs during wet comminution [18 – 20]. Insti- tute of Particle Technology, Cauerstraße 4, 91058 Erlangen,
Germany.
Furthermore, mechanical activation of the organic crystal 2
Veronika Braig, Prof. Geoffrey Lee
during the milling process may cause surface near altera-
Friedrich-Alexander-Universita¨t Erlangen-Nu¨rnberg (FAU),
tions of the solid (e.g., crystal defects, local amorphization) Divi- sion of Pharmaceutics, Cauerstraße 4, 91058 Erlangen,
resulting in an activated state of the solid with different
Germany. 3Christoph Konnerth, Veronika Braig, Prof. Geoffrey
physicochemical properties [21 – 24]. Lee,
Spray-drying is commonly used to transform dissolved or Prof. Dr.-Ing. Wolfgang Peukert
dispersed valuable compounds such as proteins and pepti- Friedrich-Alexander-Universita¨t Erlangen-Nu¨rnberg (FAU),
des, DNA for vaccines, or APIs into a solid state to prolong Clus- ter of Excellence – Engineering of Advanced Material
(EAM), Na¨gelsbachstraße 49b, 91058 Erlangen, Germany.

age stability when stabilizing excipients are added [27].

Moreover, spray-dried powders are advantageous Micronised mefenamic acid (MA) form I (Fig. 1, 2-(2,3-di-
regarding further development of an oral solid dosage form methylphenyl)aminobenzoic acid, chemical formula: C 15H15NO2,
and prod- uct handling [29]. purity: > 98 %, physical form: light yellowish powder) was
Mefenamic acid (MA) is a well-known nonsteroidal anti- purchased from TCI EUROPE N.V. (Belgium). Hydroxypropyl
inflammatory drug (NSAID), which is widely used as an cellulose (SSL grade, HPC-SSL)
antipyretic, analgesic, and antirheumatic agent. Moreover, was kindly provided by Nisso Chemicals Europe GmbH
recent studies report on a therapeutic potential of the drug (Germany). Citric acid (monohydrate), disodium
for Alzheimer’s disease [30, 31]. According to the biophar- phosphate, sodium dihydrogen phosphate, sodium
maceutics classification system (BCS), MA is classified as hydroxide, N,N-dimethyl- formamid ( ‡ 99.8 %, DMF)
a BCS Class II compound and it is known to exist in and acetonitrile (ROTISOLV® HPLC Gradient Grade)
different crystalline modifications. Next to the were pur- chased from Carl Roth Chemicals (Germany).
thermodynamic stable form I, two metastable polymorphs D-lactose monohydrate was obtained from Sigma-
(form II and III) were reported in literature [32 – 34]. Its Aldrich Chemie GmbH (Germany). Exci- pients for
physicochemical properties, especially the extremely low direct tableting, namely lactose NF monohydrate
solubility in water (<< 1 mg mL –1 [35]) and distinct SprayDry 315 (Lehmann & Voss
adhesive properties give rise to major challenges during
dosage form development and manufacturing [36]. Hence,
serious effort was made to increase dissolution
characteristics of MA to improve its biopharmaceutical
performance [37 – 40].
Within this contribution an industrially feasible formula-
tion approach for the BCS class II drug MA is presented. A
wet comminution process followed by spray-drying of the
product suspension using lactose as drug carrier was
applied. The formulation approach was studied for
different polymorphs (form I and II) as feed material.
Comminution experiments were performed at different pH
conditions to study acid-base interactions between API and
the liquid phase. MA was characterized before and after
processing to address potential alterations of crystallinity
and polymor- phic state. Moreover, obtained drug
nanoparticles were embedded in a tablet to assess the effect
of the applied for- mulation approach on MA dissolution
characteristics.

2 Materials and Methods

2.1 Materials

ª 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Chem. Ing. Tech. 2017, 89, No. 8, 1060–
Weinheim 1071
 Chemie
Ingenieur Research Article 3
Technik

(MCC, Vivapur® 301, JRS Pharma GmbH & Co. DMF solution of the drug. 20 g of MA raw material (form
KG, Germany) as well as fumed silica (Aerosil I) were dis- solved under stirring in 100 mL DMF at 70 °C.
®
200, Evonik Industries AG, Germany), talcum After the crystals were completely dissolved, the solution
(Talkum Pharma G, was rapidly cooled down to –40 °C and kept at this
C.H. Erbslo¨h GmbH & Co. KG, Germany) and temperature for 1 h until MA was completely recrystallized
magnesium stearate (Baerlocher GmbH, Germany) as form II particles. Subsequently, the solution was heated
were kindly donated by the corresponding up to room tempera- ture and obtained recrystallized form
suppliers. Hydranal® Coulomat AG-Oven analyte II drug particles were filtered, thoroughly rinsed with
was obtained from Honeywell Specialty deionized water and sub- jected to vacuum drying in a
Chemicals (Seelze GmbH, Germany). desiccator.
Demineralised water (18.2 MWcm, total
oxidisable carbon < 5 ppb) produced by a 2.2.2 Preparation of Mefenamic Acid Suspensions
PURELAB® Ultra water system (ELGA LabWater,
Veolia Water Solutions & Technologies, France) MA nanosuspensions were prepared using a media milling
was used through- out the study. All chemicals approach. Prior to all milling experiments, HPC-SSL was
were used as received and with- out further dissolved in MA-saturated water and McIlvaine buffer
purification. solu- tions [41], respectively. Afterwards, MA raw material
was dispersed under mechanical agitation overnight in the
stabi- lizer solution to ensure complete wetting of the
2.2 Experimental Methods hydropho- bic drug particles. The API solids content in all
experiments was set to 5.0 wt % with respect to the mass of
2.2.1 Preparation of Mefenamic Acid Form II suspension. The concentration of the applied polymeric
stabilizer was
Form II crystals of MA were prepared from a 15.0 wt % relative to the amount of MA. Media milling
experiments were conducted using a vertically aligned
& Co. KG, Germany), microcrystalline cellulose Figure 1. SEM micrographs of mefenamic acid raw material (form I).

Chem. Ing. Tech. 2017, 89, No. 8, 1060– ª 2017 WILEY-VCH Verlag GmbH & Co. KGaA, www.cit-journal.com
1071 Weinheim
 Powder X-ray diffraction (PXRD) data was collected
lab-scale stirred media batch mill (PE075, Netzsch-Fein- using the diffractometer D8 Advance Bruker AXS in
mahltechnik GmbH, Germany) equipped with a zirconia- Bragg-Bren-
lined double-walled grinding chamber (0.6 L), which was
connected to an external thermostat FPW80-SL (Julabo
GmbH, Germany) for temperature control. The process
temperature was set to 20.0 ± 1.0 °C. 1.8 kg of wear
resistant yttrium-stabilized zirconium oxide milling beads
(YTZ®, rGM = 6050 kg m–3, dGM = 0.4 – 0.5 mm, Tosho Inc.,
Japan) were filled into the grinding chamber and the
agitator speed of the Al2O3 three-disc-stirrer was varied
between 750 and 2000 rpm corresponding to stirrer tip
speeds v between 2.5 and 6.7 m s–1.

2.2.3 Particle Size Analysis

Particle size distributions were determined by dynamic

light scattering (DLS) and analytical centrifugation (AC).
DLS measurements were conducted using a Honeywell
Ultrafine Particle Analyzer 150 (UPA, Microtrac Inc.,
USA). API-sat- urated demineralized water as well as
buffer solutions were used for dilution of the sample
suspensions prior to size analysis to avoid multiple
scattering and dissolution effects. Average values and
standard deviations obtained from three individual
measurements are reported. AC experiments were carried
out using the LUMiSizer® LS611 (L.U.M. GmbH,
Germany) and polycarbonate cells (L.U.M. GmbH,
Germany) with a path length of 2 mm. A wavelength and
rotor speed of 470 nm and 4000 rpm, respectively, were
used for size analysis. The temperature was set to 20 ± 1
°C.

2.2.4 Spray-Drying and Vacuum Drying

of Mefenamic Acid Suspensions

Spray-drying (SD) was performed using the Bu¨chi

Mini spray-dryer B-290 (Bu¨chi Labortechnik AG,
Switzerland). Prior to SD, MA suspensions were mixed in
equal parts with an API-saturated lactose solution and
buffered solu- tion with suitable pH value, respectively,
resulting in liquid feeds with MA solids concentrations of
2.5 wt % and
10.0 wt % of lactose. The gas inlet temperature was set to
150 °C. The liquid feed was atomized with a two-fluid
noz- zle with a diameter of 0.7 mm at a flow rate of 3 mL
min–1 and an air pressure of 1.5 bar. The aspirator air flow
was set to 80 %. The spray-dryer was pre-conditioned
using dis- tilled water. Prior to vacuum drying (VD), MA
nanosuspen- sions were centrifuged for 30 min with 65
000 g at (20 ± 1.0) °C using the table-top centrifuge
Sigma 3-30KS (Sigma GmbH, Germany) to separate drug
particles and liq- uid phase. Subsequently, the wet
sediment was transferred to a desiccator and subjected to
VD for at least 48 h.

2.2.5 Solid State Characterization

2.2.5.1 Powder X-ray Diffraction

Chem. Ing. Tech. 2017, 89, No. 8, 1060– ª 2017 WILEY-VCH Verlag GmbH & Co. KGaA, www.cit-journal.com
1071 Weinheim
 fied according to Meyer determined with the tablet
tano setup (Bruker Corporation, USA) equipped with a and Zimmermann [43]. hardness tester TBH 28
VÅNTEC-1 detector. Diffraction patterns were collected Experi- ments were (Erweka
using Cu Ka radiation (l = 0.15406 nm) between 10 and performed at room
50° (2q) with a step width of 0.014° and counting times temperature and a load of
of 1 s per step. 153 Pa was applied to all
2.2.5.2 Differential Scanning Calorimetry
powders. In the following,
average values and
The differential scanning calorimeter 822e (Mettler Toledo standard deviations
LLC, USA) was used for thermal analysis. An indium stan- obtained from three single
dard was used for calibration prior to the measurements. measurements are
Standard aluminium pans and covers were used for analy- reported.
sis. Measurements were conducted under inert nitrogen
atmosphere in the temperature range from 30 to 240 °C 2.2.6.2 Tableting and
with a heating rate of 2 °C min –1. All measurements were Tablet Testing
performed in triplicate. Corresponding average values and
Before blending and
standard deviations are reported in the following.
tableting, all excipients
2.2.5.3 Fourier Transform Infrared Spectroscopy were sieved (mesh size
300 mm) to separate
Fourier transform infrared spectroscopy (FTIR) in the agglomerates. 33 wt % of
wavenumber range 4000 – 400 cm–1 was performed using spray-dried lactose
the Excalibur FTS 3100 Spectrometer (Varian Agilent monohydrate and 31 wt %
Tech- nologies, USA). Infrared spectra were collected with of MCC were mixed with
a spec- tral resolution of 2 cm–1. Sample platelets were 6 wt % of a
prepared by mixing 200 mg dry potassium bromide talcum/magnesium
(UVASol, Merck KGaA, Germany) with 1 mg of sample stearate/Aerosil powder
powder. Background and baseline corrections were (ratio 6:3:1 (w/w)) and
performed manually. blended for 15 min in a
TURBULA® mixer T2C
2.2.5.4 Scanning Electron Microscopy (Willy A. Bachofen AG –
Maschi- nenfabrik,
Scanning electron microscopy (SEM) was performed using Switzerland). Then, 30 wt
a Gemini Ultra 55 (Carl Zeiss AG, Germany). Sample pow- % of spray-dried MA
ders were placed on a silicon wafer, which was mounted on particles obtained either
an aluminum pin stub. Measurements were carried out from the coarse suspension
using the SE2 detector. Due to charging effects and to or the nanosuspensions
improve the quality of the SEM micrographs, all samples were added and the
were sputtered with a thin film of gold-palladium (Hummer resulting mixture was
JR Technics, Germany). blended for additional 5
min. Afterwards, the
2.2.5.5 Karl Fischer Titration powder was filled
manually into the die of a
Residual moisture of the spray-dried powders was deter-
Korsch laboratory single-
mined with a Metrohm 831 KF Coulometer combined with
press punch EK0 (Korsch
a KF Thermoprep 832 oven (Deutsche Metrohm GmbH &
AG, Germany) and
Co KG, Germany). 15 to 20 mg of sample powder was
compressed by punches
heated up to 130 °C and the vapoured water was pumped
with a median diameter of
into the titration cell filled with Hydranal® reagent.
13 mm. 20 tablets were
Nitrogen gas was used as carrier gas with a flow rate of
weighted and evaluated
50 mL min–1.
according to the
monograph ‘‘Uniformity
2.2.6 Post Processing of Spray-Dried Suspensions
of Mass of Single-Dose
2.2.6.1 Tensile Strength Measurements Preparations’’ of the
European Pharmacopoeia
Powder flowability was characterized by determining the (Ph. Eur. 8.0, Chapter
powders’ tensile strength using a tensile strength tester as 2.9.5) [44]. For 10
proposed by Schweiger and Zimmermann [42] and modi- randomly chosen tablets
the crushing strength was
www.cit-journal.com ª 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Chem. Ing. Tech. 2017, 89, No. 8, 1060–
Weinheim 1071
flow rate was set mm as deter- mined by International, USA) and GM ð1 — jGM ð1 — eÞÞcv d 2
to 1.0 mL min–1. image analysis) were subjected for analysis.
MA was detected subjected to media High performance liquid where dGM is the
at 275 nm and 25 milling in a batch chromatography (HPLC) grinding media
°C and the stirred media mill. A was used for quantification diameter, rGM corre-
retention time previously identified of MA. The HPLC device sponds to the density
under these condi- formu- lation (Thermo Fisher Scientific of the beads, v is the
tions was 3 min 12 containing the non- Inc.) consists of the stirrer tip speed, jGM is
s. The determined ionic polymer HPC- quaternary analytical the filling ratio of the
limit of detection SSL was selected for pump LPG-3400SD, the grinding media, e is
(LOD) of the drug the preparation of automated sample injector the poros- ity of the
and the used nanosuspensions [45]. ASI-100, the column bulk grinding media,
HPLC system was As introduced by thermostat STH 585 and n, t, and cv are the
4.97 mM as Kwade [46 – 48], the the diode array detector number of revolutions
calculated from a comminution process DAD-3000RS. For MA of the stirrer per min,
linear calibration in stirred media mills analysis, the milling time and vol-
function (N = 10 can be described by chromatographic col- umn ume concentration of
(3-fold two quantities, namely UltraSep ES AMID H the product
replication), f = 8 by the number of RP18 (column size 150 ·3 suspension,
degrees of stress events between mm, SEPSERV Separation respectively. The
freedom, t-factor product particles and Service, Germany) was impact of stirrer tip
(95 %, f = 8) = grinding media and the used as sta- tionary phase. speed on the evolution
2.306). energy transferred The mobile phase was of the vol- ume
during these stress composed of acetoni- trile median particle size
events. Hence, the key and McIlvaine buffer pH x50,3 of MA feed
3 Results and parameters stress 2.8 (60:40 (v/v)) [41] and crystals (form I)
Discussion energy SEGM (Eq. (1)) the during media milling
and stress number SN is given in Fig. 2a.
3.1 Media Milling of (Eq. (2)) can be used During initial stages of
Mefenamic Acid to summarize the the grinding process,
influence of operating size reduction occurs
Crystalline MA parameters on the very fast. As milling
drug particles milling process: proceeds, particle
(xFeret = 2.2 ± 0.9 sizes gradually
GmbH, Germany) become smaller due to
according to the procedure SE ¼ r v2
decreasing breakage
as described / dG G
in ‘‘Resistance to Crushing SE M M rates as a result of
of Tablets‘‘ of the Ph. Eur. GM
decreasing breakage
8.0 (Chapter 2.9.8). probabilities of MA
jGM ð1 — eÞ (2) particles with
SN / n decreasing particle
t
size. According to Eq.
2.2.6.3 In Vitro circulator (Vankel Technology (1), the increase in
Dissolution Study Group, USA). The dissolution stirrer tip speed from
medium was composed of 900 3 2.5 to 6.7 m s–1
Drug release studies were mL pH 6.8 phosphate buffer corresponds to an
performed using the paddle (Ph. Eur. 8.0, Chapter 5.17.1 increase in the applied
method according to Ph. ‘‘Recommandations for stress energy SEGM by
Eur. 8.0 Chapter 2.9.3 Dissolution Studies’’ [44]). more than a factor of 7
‘‘Dissolution Test for Solid Samples of 2 mL were taken at (i.e., SEGM (v = 2.5 m
Dosage Forms’’ certain time intervals, filtered s–1) = 3.5 mJ as
(Apparatus II) in a through a 0.02 mm pore compared to SEGM (v
VK7000 apparatus (Vankel diameter filter (Whatman® = 6.7 m s–1) = 24.8
Technology Group, USA) Anotop®10 syringe filter, mJ). Also SN (cf. Eq.
[44]. Paddle rotation was VWR International, Germany) and (2)) in-
set to 100 rpm and the subsequently diluted with creases by almost the
temperature was set to 37 acetonitrile. Ali- quots were factor of 3 leading to a
± 0.5 °C using the VK then transferred and sealed in higher stress
750D external heater 1.5-mL vials (VWR frequency of the
Chem. Ing. Tech. 2017, 89, No. 8, 1060– ª 2017 WILEY-VCH Verlag GmbH & Co. KGaA, www.cit-journal.com
1071 Weinheim
product particles. particle size around dependent) solid-
Thus, faster x50,3 » 180 nm and x90,3 MA/HPC- process
liquid
process kinetics » 350 nm, could be SSL =
6.7:1,
equilibrium
and increased observed, indicating
MA- than by
apparent breakage that final product saturate pure
rates are observed. particle sizes of MA d
water).
mechanical
However, are not linked to this
fracture.
irrespective of the operating parameter.
result is in good significantly affects
applied stirrer tip This
agreement with product charac- teristics
speed a limit- ing
comparable studies dis- such as particle size
processed at cussed in literature [20, and
lower 49, 50], supporting s
process tem- general observa- tions h
peratures. that in particular high a
Moreover, stress numbers at p
SEM micro- moderate stress energies e
graphs of are advantageous with .
final respect to a fast and
nanosuspensi energy efficient media I
ons milling process of APIs t
processed at and organic crystals,
20 °C (D) respectively. w
and 40 °C The formation of a a
(£) reveal steady state in particle s
some size during media
Figure
irregular milling of organic f
v2.on
speed5.0 Impact of stirrer tip
wtgrinding
%, T kinetics (drug
=
shaped
load: 2
0 crystals (e.g., APIs, o
°C particles pigments) is generally u
,
(grinding observed and frequently n
fragments) discussed in literature. d
as well as The discussion is often
product limited to the dynamic t
particles equilibrium between h
with well- mechanically induced a
defined particle breakage and t
sharp agglomeration processes
crystallo- due to insufficient f
graphic stabilization. However, i
edges (Fig. also liquid phase and n
4). In the formulation mediated a
case of MA growth effects, induced l
processed at by dissolution,
40 °C, nee- precipitation, and p
dle-shaped
ripening phenomena, r
crystals were
which compete with o
found. This
above mentioned d
result
processes will affect u
supports the
product formation c
hypothe- sis
during media mill- ing t
that limiting
as well, although they
product
are only rarely p
particle sizes
discussed. As shown by a
are rather
determined Konnerth et al. [23] and r
by the Steiner et al. [51], the t
(temperature process temperature i
-, solvent-, during wet media c
and milling is a key pro- l
formulation cess parameter, which e

www.cit-journal.com ª 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Chem. Ing. Tech. 2017, 89, No. 8, 1060–
Weinheim 1071
 e s
s n e e
i d l
z l e
e o s v
s w o a
e l t
a r u e
t b d
p i
c r l t
o o i e
m c t m
p e y p
a s e
- s l r
e a
r t v t
a e e u
b m l r
l - s e
e s
p o ,
s e f
t r r
r a t i
e t h p
s u e e
s r n
i e p i
n s r n
g o g
w c
c e e a
o r s n
n e s d
d e
i a d g
t p r
i p o o
o l r w
n i g t
s e a h
d n
w . i e
e c f
r D f
e u c e
e o c
s m t
m t p s
a o o
l u a
l i n r
e n d e
r c s
r m
w e a o
h a t r
Chem. Ing. Tech. 2017, 89, No. 8, 1060– ª 2017 WILEY-VCH Verlag GmbH & Co. KGaA, www.cit-journal.com
1071 Weinheim
e i The influence of
z o formulation, i.e., in
p e f particular the impact of
r . different types of non-
o M ionic polymers and its
- T A applied concentration on
h MA grinding
n e ( performance as well as
o F lim- iting product
u s i particle sizes during
n a g media milling was
c m s already discussed in
e e . detail by Ito et al. [45].
d However, MA exhibits
r 3 polymorphism and the
l e a study was limited to pro-
e s cessing of MA form I. In
a u a principle, it is preferable
d l n to formu- late the
i t d thermodynamic most
n stable form to ensure a
g c b high reproducibility
o ) regarding
t u : manufacturing,
o l bioavailability, and shelf
d P life. In case of MA, the
a r polymorphic form II is
n b o known to exhibit higher
e d saturation solubilities in
i u several solvents,
n o c including aqueous media
c b t [33, 52]. Moreover, the
r s metastable form shows
e e p supersaturation during
a r a dissolution
s v r accompanying an
e e t apparent first-order
d i decrease in the amount
i c of MA dissolved down
n d l to the equilibrium
u e solubility of form I due
p r to polymorphic
r i s transition [32, 33, 52].
o n i Thus, form II seems to
d g z exhibit a greater
u e pharmaceutical benefit.
c m s Furthermore, MA is an
t e ionizable drug
d a compound and shows a
p i n pH-dependent solubility
a a d profile in aqueous media
r [35]. Therefore, milling
t m experiments with
i i different API feed
c l particle stressing and formulation conditions
l l size were smaller, when
e i distribution
n s at con-
s g stant
www.cit-journal.com ª 2017 WILEY-VCH Verlag GmbH & Co. KGaA, Chem. Ing. Tech. 2017, 89, No. 8, 1060–
Weinheim 1071
 Figure 3.
Impact of
process
temperature on
grinding
kinetics (a) and
final product
particle size
distributions (b)
of MA (SEGM =
13.9 mJ, drug
load: 5.0 wt %,
MA/HPC-SSL =
6.7:1, MA-
saturated
water).

Chem. Ing. Tech. 2017, 89, No. 8, 1060– ª 2017 WILEY-VCH Verlag GmbH & Co. KGaA, www.cit-journal.com
1071 Weinheim
 attributed to elevated ripening rates of the drug
during the media milling process (c.f. cMA
(pH 7, 25 °C) / cMA (pH 3, 25 °C) » 300 [35]).

3.2 Post Processing of Mefenamic Acid

Nanosuspension

3.2.1 Physicochemical
Characterization of the Dried
Drug Powder

For oral administration, the transformation of

nanosuspensions into solid dosage forms is
advantageous with respect to physical and
chemical stability of the active compound, prod-
uct handling, manufacturing as well as patient
convenience. Therefore, MA nanosuspensions
were subsequently spray-dried within 24 h after
preparation. Moreover, a small amount of prod-
Figure 4. SEM micrographs of final MA nanosuspensions produced by
media milling at different process temperatures: a) and b) Tprocess = 20 uct suspension was centrifuged directly after
°C, c) and milling and the wet sediment was subjected to
d) Tprocess = 40 °C. vacuum drying to address potential drug altera-
tions due to the drying procedure.
materials (form I and form II) in buffered media as well as SEM micrographs of particles obtained from spray-
in water without pH control were carried out to study the drying of MA nanosuspensions, which were prepared by
influence of polymorphic form and potential acid-base wet grinding of polymorph I and II in an API-saturated
interactions between the API and the liquid phase on grind- aqueous phase (no pH-control) as well as in a buffered sys-
ing kinetics and final MA product particle sizes. pH 3 and tem at pH 3 and pH 7 are depicted in Fig. 5. Irrespective of
pH 7 were chosen with respect to the pKa of the drug the liquid phase and polymorphic form of the feed material,
(pKa 4.54 (25 °C) [35]). Grinding kinetics of MA form I a homogenous and uniform spray-agglomerated product is
and II were practically identical (data not shown) and as obtained. SD powders mainly consist of spherical particles
can be seen in Tab. 1, final product particle sizes are almost of 5 to 10 mm, besides some smaller particles of around
identi- cal using both polymorphic forms as feed material 1 mm size. The surface is smooth, regular and no erosions
and applying the same suspension pH. In contrast, coarser or cracks are visible. Moreover, MA nanoparticles are well
drug product particle sizes are obtained during media em- bedded in the chosen carrier system (Figs. 5b, d, f, h,
milling at pH 7 as compared to comminution experiments and j).
in acidic environment (pH 3), highlighting the influence of The left side of Fig. 6 depicts FTIR spectra (wavenumber
the liquid phase. As outlined above, MA product particle range 4000 to 400 cm–1) of MA feed material (form I and
sizes are strongly influenced by its solid-liquid II) as well as spray-dried product powders using lactose as
equilibrium. Thus, the obtained coarser product particle matrix material. The spray-dried products were obtained
sizes at pH 7 can be from MA nanosuspensions processed at different pH
condi-

Table 1. Final product particle sizes of MA nanosuspensions prepared at different pH conditions and with different feed materials
(n.d. = not determined).

Feed, pHbefore milling pHafter milling DLS AC

pH conditions
x10,3 [mm] x50,3 [mm] x90,3 [mm] x10,3 [mm] x50,3 [mm] x90,3 [mm]

Form I, no pH 5.3 7.0 0.106 ± 0.044 0.188 ± 0.012 0.362 ± 0.214 0.113 0.251 0.743
control
Form II, no pH 5.2 6.9 0.098 ± 0.067 0.189 ± 0.022 0.364 ± 0.128 n.d. n.d. n.d.
control

Form I, pH 3 3.0 3.0 n.d. n.d. n.d. 0.102 0.192 0.493

Form I, pH 7 7.1 7.3 n.d. n.d. n.d. 0.138 0.273 0.760

Form II, pH 3 3.0 3.0 n.d. n.d. n.d. 0.088 0.224 0.523
Form II, pH 7 7.0 7.2 n.d. n.d. n.d. 0.166 0.355 0.657

in all spectra by unfilled triangles A (form I:

bands at 894 cm–1, 756 cm–1, 746 cm–1) and
solid diamonds ^ (form II: bands at 920 cm –
1
, 747 cm–1, 730 cm–1, 575 cm–1), respectively
[32, 33, 53]. Irrespective of the applied pH con-
ditions and the polymorphic form of MA feed
material, spray-dried powders only exhibit spec-
tral features of the thermodynamic most stable
form I. The same result was observed when the
respective nanosuspensions were subjected to
vaccum-drying at ambient conditions (see right
side of Fig. 6): Obtained product powders are
solely composed of polymorph I independent of
the chosen process conditions.
Similar conclusions can be drawn from the X-
ray diffraction patterns of spray-dried and
vacuum-dried API nanosuspensions. Form I and
form II of MA are best discriminated by the
dominating reflexes (Cu Ka, Fig. 7) of 21.3 °
(2q, form I (a)) and 17.9 ° (2q, form II (b)).
Obvi- ously, only form I is found in the dried
product, even though form II was used as feed
material in the milling process, cf. (c) for form I
processed in MA-saturated water (no pH
control), (d), (e) for feed (form I) processed at
pH 3 and pH 7 as well as (f), (g) for feed (form
II) at pH 3 and pH 7, respectively. Moreover,
PXRD patterns of the dried products did not
show any alterations as compared to the
micronised API, indicating that drug particles
preserve most of their crystallinity. However, a
reduction in intensity and a slight broadening of
characteristic Bragg reflections as compared to
the feed could be observed for dried products,
which can be attributed to the smaller size of the
MA nanoparticles and the lower API
concentration in the spray-dried pow- ders [54].
Solid state properties of MA after processing
were also evaluated by differential scanning
calo- rimetry (DSC). In accordance with FTIR
and PXRD data, the presence of only the
thermo- dynamic most stable form was also
proven by thermal analysis of the vacuum-dried
product powders (Tab. 2 and Fig. 8). Due to
decomposi- tion of the lactose carrier at
temperatures below the transition temperature
of the API, DSC mea- surements could not be
Figure 5. SEM images of spray-dried MA nanosuspensions processed at
differ- ent pH conditions and using different polymorphs as feed material:
performed. MA form I exhibits two
a) and b) Form I and no pH control, c) and d) form I and pH 3, e) and f) endothermic events. The first event at 163°C
form I and pH 7.0, (DHtransition = 2.3 kJ mol–1) is related to
g) and h) form II and pH 3.0, and i) and j) form II and pH 7.0. lattice vibrations (wavenumber range < 1000 cm –1)
marked
tions and using different polymorphs as feed material,
cf.
(c) Form I (no pH control), (d) form I (pH 3), (e) form I
(pH 7), (f) form II (pH 3) and (g) form II (pH 7). Different
polymorphs of MA can be distinguished by characteristic
 the thermal transition of form I to form II fol-
lowed by the sharp melting endotherm of form
II at 229.5 °C (DHf = 38.9 kJ mol–1). The experimentally
determined transition temperature and enthalpy of MA
form I differ slightly as compared to literature data. How-
ever, a review of corresponding studies points out, that this
characteristic feature of the drug is strongly dependent on
Figure 6. FTIR spectra of spray-dried (left) and vacuum-dried (right) MA nanosuspensions: a) Form I feed material, b) form II feed
materi- al, c) form I and no pH control, d) form I and pH 3, e) form I and pH 7, f) form II and pH 3, and g) form II and pH 7.

any glass transition event sup-

ports the observation obtained by
PXRD that API crystallinity is
mainly preserved during process-
ing or that the amorphous con-
tent in the sample is rather low to
be detected. Furthermore, a sig-
nificant broadening of the melt-
ing peak of all processed
powders and, as expected, a
melting point depression due to
smaller drug particle sizes and
the presence of the stabilizer as
Figure 7. PXRD diffraction patterns (CuKa radiation) of spray-dried (left) and vacuum-dried
(right) MA nanosuspensions: a) Form I feed material, b) form II feed material, c) form I and compared to the feed material
no pH control, d) form I and pH 3, e) form I and pH 7, f) form II and pH 3, and g) form II and could be observed. Furthermore,
pH 7. the significant melting point
depression can be
the manufacturer, batch, purity, as well as the experimental correlated to a strong interaction between MA and HPC-
conditions (i.e., heating rate) and varies in a broad range SSL, which is elevated, because the polymeric stabilizer is
between 160 and 206 °C [33, 52, 53, 55 – 57]. The absence already molten at the melting temperature of the drug [58].
of

Table 2. Thermal characteristics (from DSC) of vacuum-dried MA particles obtained from nanosuspensions.

Feed, pH conditions Transition form I to form II Melting properties form II

–1
Tonset [°C] Htransition [kJ mol ] Tonset [°C] Hfus [kj mol–1]

Form I 163.2 ± 0.6 2.3 ± 0.1 229.5 ± 0.3 38.9 ± 0.9

Form II – – 228.3 ± 0.5 37.2 ± 0.6

Form I, no pH control 170.2 ± 0.5 1.3 ± 0.4 215.4 ± 0.4 26.2 ± 0.5

Form I, pH 3 167.3 ± 1.0 1.4 ± 0.3 218.8 ± 0.1 24.9 ± 1.0

Form I, pH 7 167.5 ± 0.4 1.6 ± 0.1 213.9 ± 0.7 24.0 ± 0.3

Form II, pH 3 167.0 ± 2.6 1.6 ± 0.0 218.7 ± 0.7 27.1 ± 0.2

Form II, pH 7 169.9 ± 0.4 1.7 ± 0.1 215.5 ± 0.3 25.5 ± 1.1
 tensile strengths of processed powders as well as feed
mate- rials and some additives used for the tableting
process are summarized in Tab. 3. The adhesive character
of MA raw material was confirmed by a high tensile
strength of
19.3 ± 0.9 Pa. However, micronized feed particles show a
lower tensile strength as compared to the corresponding
spray agglomerated product, which is most likely due to
their larger size. Spray-agglomerates of MA nanosuspen-
sions exhibit a moisture content of about 3 %, which may
lead to the formation of capillary bridges in the bulk solid
resulting in a decreased flowability. Although pure spray-
dried powders are still to some extent cohesive, the
flowabil- ity of the tablet composition prepared with spray-
dried MA particles obtained from nanosuspensions is
Figure 8. DSC thermograms of vacuum-dried MA particles ob-
tained from nanosuspensions: a) Form I feed material, b) form
improved due to the spherical shape of the spray
II feed material, c) form I and no pH control, d) form I and pH agglomerates and the effect of added fumed silica as
3, compared to the tablet com- position containing the same
e) form I and pH 7, f) form II and pH 3, and g) form II and pH amount of fumed silica and drug but in the micronized
7. Dashed lines: Zoom into the transition region of form I to
form II. state.
Tablets prepared by the direct compression method con-
taining the spray-dried drug product are circular, flat-faced
Apparently, polymorphic transition of MA form II to and have a diameter of 13 mm (Fig. 9). The average weight
form I occurs irrespective of the applied drying procedure is 511 ± 9 mg and 513 ± 10 mg for tablets containing MA
and its peculiarities, concluding that the transition process nanoparticles and drug particles in the micronized state,
already takes place during the media milling process and respectively. Thus, tablets meet the requirements of the
that the polymorphic form II, which exhibits a greater Ph. Eur. regarding uniformity of mass of single-dose prepa-
phar- maceutical benefit, could not be preserved during rations. With a resistance to crushing of 72 ± 12 N (tablets
process- ing. containing MA nanoparticles) and 62 ± 11 N (tablets con-
taining micronized MA drug particles) the tablets are not
3.2.2 Tableting and In Vitro Dissolution Testing expected to show poor disintegration, but they are stable
against crumbling. The calculated amount of API in all tab-
To evaluate the impact of nanosizing on MA drug dissolu- lets is 30 mg. Comparable market products of MA exhibit
tion after oral administration, spray-dried MA drug par- at least 250 mg as therapeutically effective dose for adults.
ticles were post-processed into tablets. Due to its distinct However, for the API fenofibrate (FF), which is also classi-
adhesive character and tendency to stick on any type of fied as BCS class II drug compound, it could be shown that
sur- face, the flowability of the dried API product powder the drug load in oral dosage forms can be significantly low-
was evaluated in an almost uncompacted state by tensile
strength measurements prior to tableting, which is quite
close to manufacturing conditions in the pharmaceutical
industry. The principle of measurement is outlined in detail
in [42, 43]. Tensile strength TS is strongly dependent on
interparticulate adhesion forces: Low interparticulate inter-
action forces correlate with good powder flowability and Figure 9. Tablets with
low tensile strength [59 – 61]. Experimentally spray-dried MA parti-
cles obtained from
determined nanosuspensions.

Table 3. Powder characteristics of raw materials, spray-dried products and final tablet compositions.

Moisture content [wt %] Tensile strength [Pa]

Micronised drug (form I, raw material) – 19.3 ± 0.9

Lactose (monohydrate) – 8.0 ± 2.0

SD powder micronised drug (form I, no pH control) – 28.8 ± 1.3

SD powder -nanoparticles (form I, no pH control) 2.9 ± 0.1 36.2 ± 4.2

Tablet composition (micronised drug) – 7.6 ± 3.0

Tablet composition (MA nanoparticles) – 3.2 ± 1.1

 ered when nanocrystalline particles of the drug are present
in the formulation. A phase I study with 71 adult subjects practically the total amount of MA is fully dissolved after
proves that the administration of two tablets of Tricor ® 45 min. In comparison, only 45.9 ± 5.0 % of the drug is
(Abbott) with 48 mg of FF as nanoparticles shows bioequi- dis- solved after half an hour and 57.3 ± 1.6 % after 120
valence to a 200 mg capsule of micronized API, which cor- min from tablets containing the micronized drug. Besides
relates with the typical effective dose [62, 63]. Thus, in the in- creased dissolution rate, a positive influence of the
case of MA the same effect is expected. Moreover, the nanopar- ticulate formulation of MA concerning food
above pro- posed formulation approach is scalable, e.g., effect and fol- lowing plasma concentrations is also
wet comminu- tion experiments with drug loads up to 20 expected as already demonstrated in in vivo studies with
wt % could be conducted without significant alterations of beagle dogs for the API aprepitant (Emend ®, Merck Sharp
the final prod- uct particle size [49 – 51, 64, 65]. In & Dohme) [68].
addition, the amount of API in the final tablet can be
adjusted to a certain extent by varying the ratio between
drug nanoparticles and matrix former during the spray-
4 Conclusion
drying process.
An industrially feasible formulation approach combining
MA is a weak monoprotic acid and shows strong pH-
media milling and spray-drying was successfully applied to
dependent solubility behavior. It exhibits an extremely low
–1 improve dissolution characteristics of the poorly water-
intrinsic aqueous solubility of 59 ± 4 ng mL (37 °C) [35].
soluble BCS class II drug compound MA. Wet comminu-
Due to the elevated saturation solubility of the drug in the
–1 tion experiments were performed at different stressing con-
release medium (40.8 ± 0.2 mg mL (37 °C), phosphate
ditions and process temperatures. Furthermore, the effect
buffer, pH 6.8), a fast dissolution of the drug and a steep
of polymorphic form and suspension pH was studied. It
concentration gradient, respectively, will weaken or, in the
was found that final drug product particle sizes are rather
worst case, might mask the size effect of the drug particles
deter- mined by the (temperature-, solvent-, and
on its dissolution characteristics during the in vitro testing
formulation- dependent) solid-liquid equilibrium than by
procedure. Thus, non-sink conditions were chosen to get
pure mechani- cal fracture. Furthermore, obtained drug
more discriminative dissolution results to emphasis the
particles either prepared by spray-drying or vacuum drying
influence of drug particle size on dissolution in the given
were solely composed of the thermodynamic most stable
formulation [66, 67]. Dissolution kinetics of MA micro-
form I of MA irrespective of all peculiarities, indicating
particles and nanoparticles, respectively, embedded in tab-
that the polymor- phic transition already takes place during
lets prepared by direct compression with identical amount
the milling pro- cess. Tablets prepared by direct
of excipients are given in Fig. 10. A slight increase of the
compression which contain nanocrystalline MA particles
dissolved amount of MA due to leaching and light tablet
exhibit a significant improve- ment of in vitro drug
crumbling could be observed for both formulations fol-
dissolution characteristics as com- pared to the dosage
lowed by the complete disintegration of the tablets (cf. the
form containing drug particles in the micronized state,
volatile increase in the released amount of MA between 15
elucidating the great potential of this for- mulation
and 30 min). Thus, the higher amount of dissolved MA
approach for a systematic delivery of this poorly water
observed from tablets with drug nanoparticles at all time
soluble API.
points can be attributed to the larger specific surface area
orschungsgemeinschaft (DFG, grant PE 427/24-1) and the Cluster of Excellence
of the nanoparticles as compared to particles in the micron- - Engineering of Advanced Materials (EAM) for funding this research work. More- ove
ized state. 89.4 ± 3.2 % of MA was released after 30 min
and

respectively.

Figure 10. Dissolution kinetics of 30 mg MA in 900 mL pH 6.8

phosphate buffer at 37 ± 0.5 °C embedded in direct
compressed tablets as nanoparticles (D) and microparticles (£),
 Symbols used

cv [–] volume fraction of the product

suspension
dGM [m] grinding media diameter
n [min–1] number of revolutions of the stirrer
per minute
SE [J] stress energy
SN [–] stress number
Chemie
Ingenieur Research Article 1071
Technik

t [s] time [14] G. G. Liversidge, K. C. Cundy, Int. J. Pharm. 1995, 125

TS [Pa] tensile strength (1), 91 – 97.
[15] E. Merisko-Liversidge, G. G. Liversidge, Adv. Drug Delivery Rev.
v [m s–1] stirrer tip speed
2011, 63 (6), 427 – 440.
x [m] particle diameter [16] M. Juhnke, E. John, Chem. Eng. Technol. 2014, 37 (5), 757 – 764.
[17] M. D. Ticehurst, I. Marziano, J. Pharm. Pharmacol. 2015, 67
(6), 782 – 802.
Greek letters [18] A. Bitterlich, C. Laabs, E. Busmann, A. Grandeury, M. Juhnke,
H. Bunjes, A. Kwade, Chem. Eng. Technol. 2014, 37 (5), 840 –
e [–] porosity of the bulk grinding media 846.
rGM [kg m3] density of the grinding media [19] A. Bitterlich, C. Laabs, I. Krautstrunk, M. Dengler, M. Juhnke,
jGM A. Grandeury, H. Bunjes, A. Kwade, Eur. J. Pharm. Biopharm.
[–] filling ratio of the grinding media
2015, 92, 171 – 179.
[20] C. Konnerth, F. Flach, S. Breitung-Faes, C. Damm, J. Schmidt,
A. Kwade, W. Peukert, Powder Technol. 2016, 294, 71 – 79.
Abbreviations [21] I. Colombo, G. Grassi, M. Grassi, J. Pharm. Sci. 2009, 98
(11), 3961 – 3986.
AC analytical centrifugation [22] P. Bala´zˇ, M. Achimovicˇova´, M. Bala´zˇ, P. Billik, Z.
API active pharmaceutical ingredient Cherkezova- Zheleva, J. M. Criado, F. Delogu, E. Dutkova´, E.
BCS biopharmaceutics classification Gaffet, F. J. Gotor,
system DLS dynamic light scattering R. Kumar, I. Mitov, T. Rojac, M. Senna, A. Streletskii, K.
Wieczo- rek-Ciurowa, Chem. Soc. Rev. 2013, 42 (18), 7571 –
DSC differential scanning 7637.
calorimetry FF fenofibrate [23] C. Konnerth, C. Damm, J. Schmidt, W. Peukert, Adv.
FTIR fourier transform infrared Powder Technol. 2014, 25 (6), 1808 – 1816.
spectroscopy GM grinding media [24] S. Romeis, J. Schmidt, W. Peukert, Int. J. Miner. Process. 2016,
HPLC high performance liquid chromatography 156, 24 – 31.
LOD limit of detection [25] M. Maury, K. Murphy, S. Kumar, L. Shi, G. Lee, Eur. J.
Pharm. Biopharm. 2005, 59 (3), 565 – 573.
MA mefenamic acid
[26] B. van Eerdenbrugh, L. Froyen, J. van Humbeeck, J. A. Martens,
MCC microcrystalline cellulose P. Augustijns, G. van den Mooter, Eur. J. Pharm. Sci. 2008,
NSAID non-steroidal anti-inflammatory drug 35 (1 – 2), 127 – 135.
PXRD powder X-ray diffraction [27] M. V. Chaubal, C. Popescu, Pharm. Res. 2008, 25
SD spray-drying (10), 2302 – 2308.
SEM scanning electron [28] D. A. LeClair, E. D. Cranston, Z. Xing, M. R. Thompson,
microscopy USP United States Pharm. Res. 2016(, 1313) , 2763 – 2776.
[29] B. van Eerdenbrugh, G. van den Mooter, P. Augustijns,
Pharmacopeia VD vacuum drying Int. J. Pharm. 2008, 364 (1), 64 – 75.
[30] Y. Joo, H.-S. Kim, R.-S. Woo, C. H. Park, K.-Y. Shin, J.-P.
Lee, K.-A. Chang, S. Kim, Y.-H. Suh, Mol. Pharmacol.
References 2006, 69 (1), 76 – 84.
[31] M. J. D. Daniels, J. Rivers-Auty, T. Schilling, N. G. Spencer,
[1] C. R. Gardner, C. T. Walsh, O. Almarsson, Nat. Rev. Drug W. Watremez, V. Fasolino, S. J. Booth, C. S. White, A. G.
Discov- ery 2004, 3 (11), 926 – 934. Baldwin,
[2] H. D. Williams, N. L. Trevaskis, S. A. Charman, R. M. Shanker, S. Freeman, R. Wong, C. Latta, S. Yu, J. Jackson, N. Fischer, V.
W. N. Charman, C. W. Pouton, C. J. H. Porter, Pharmacol. Rev. Ko- ziel, T. Pillot, J. Bagnall, S. M. Allan, P. Paszek, J. Galea, M.
2013, 65 (1), 315 – 499. K. Harte, C. Eder, C. B. Lawrence, D. Brough, Nat. Commun.
[3] C. A. Lipinski, J. Pharmacol. Toxicol. Methods 2000, 44 2016, 7, 12504.
(1), 235 – 249. [32] A. J. Aguiar, J. E. Zelmer, J. Pharm. Sci. 1969, 58 (8), 983 – 987.
[4] C. A. Lipinski, Am. Pharm. Rev. 2002, 5 (3), 82 – 85. (https:// [33] F. Kato, M. Otsuka, Y. Matsuda, Int. J. Pharm. 2006, 321 (1 –
www.researchgate.net/publication/279892668_Poor_aqueous_ 2), 18 – 26.
solubility_-_An_industry_wide_problem_in_drug_discovery) [34] S. SeethaLekshmi, T. N. Guru Row, Cryst. Growth Des. 2012,
[5] C. W. Pouton, Adv. Drug Delivery Rev. 1997, 25 (1), 47 – 12 (8), 4283 – 4289.
58. [6] C. W. Pouton, Eur. J. Pharm. Sci. 2006, 39 (3 – 4), 278 [35] A. Avdeef, Adv. Drug Delivery Rev. 2007, 59 (7), 568 – 590.
– 287. [36] A. Adam, L. Schrimpl, P. C. Schmidt, Drug Dev. Ind. Pharm.
[7] A. Fahr, X. Liu, Expert Opin. Drug Delivery 2007, 4 (4), 403 – 2000, 26 (5), 477 – 487.
416. [37] T. Imai, K. Nohdomi, F. Acartu¨rk, M. Otagiri, J. Pharm. Sci. 1991,
[8] E. Merisko-Liversidge, G. G. Liversidge, E. R. Cooper, 80 (5), 484 – 487.
Eur. J. Pharm. Sci. 2003, 18 (2), 113 – 120. [38] G. Owusu-Ababio, N. K. Ebube, R. Reams, M. Habib,
[9] B. E. Rabinow, Nat. Rev. Drug Discovery 2004, 3 (9), 785 – 796. Pharm. Dev. Technol. 1998, 3 (3), 405 – 412.
[10] H. Chen, C. Khemtong, X. Yang, X. Chang, J. Gao, Drug [39] T. Iwasaki, M. Takahara, R. Sonoda, S. Watano, Part. Part.
Discovery Today 2011, 16 (7 – 8), 354 – 360. Syst. Charact. 2007, 24 (3), 236 – 241.
[11] J. P. Mo¨schwitzer, Int. J. Pharm. 2013, 453 (1), 142 – 156. [40] T. Kojima, K. Higashi, T. Suzuki, K. Tomono, K. Moribe, K.
[12] A. A. Noyes, W. R. Whitney, J. Am. Chem. Soc. 1897, 19 Ya- mamoto, Pharm. Res. 2012, 29 (10), 2777 – 2791.
(12), 930 – 934. [41] T. C. McIlvaine, J. Biol. Chem. 1921, 49 (1), 183 – 186.
[13] G. G. Liversidge, P. Conzentino, Int. J. Pharm. 1995, 125 [42] A. Schweiger, I. Zimmermann, Powder Technol. 1999, 101
(2), 309 – 313. (1), 7 – 15.
 Chemie
1070 Research Article Ingenieur
Technik

[43] K. Meyer, I. 2003, 262 (1 – 2), 13 – 22.

Zimmermann, [58] P. J. Marsac, S. L. A. Santos, S. Sistonen, S.
Powder Technol. Shamblin, L. S. Taylor, Hellsten,
Pharm. Res. 2006, T. Laaksonen, L.
2004, 139 (1), 40 –
23 (10), 2417 – 2426. Peltonen, N. Kumar,
54.
[59] J. Schmidt, M. Sachs, C. D. Y. Murzin, M.
[44] European Pharmacopoeia,
Blu¨mel, B. Winzer, F. Louhi- Kultanen, J.
8th ed., Council of Europe,
Toni, K.-E. Wirth, Salonen, J. Hirvonen,
Strasbourg
W. Peukert, Powder V.-P. Lehto, Int. J.
2013.
Technol. 2014, 261, 78 – Pharm. 2011, 416
[45] A. Ito, C. 86. (1), 242 – 251.
Konnerth, J.
[60] C. Blu¨mel, [67] P. Liu, O. de Wulf, J. Laru, T.
Schmidt, W.
Charakterisie Heikkila, B. van Veen, J.
Peukert, Eur. J.
rung der Kiesvaara,
Pharm. Biopharm.
Trockenen J. Hirvonen, L. Peltonen, T.
2016, 98, 98 – 107.
Beschichtung Laaksonen, AAPS
[46] A. Kwade, Powder
zur PharmSciTech
Technol. 1999, 105 (1 – 3), 382
Herstellung 2013, 14 (2), 748 – 756.
– 388.
von [68] Y. Wu, A. Loper, E. Landis, L.
[47] A. Kwade, J.
maßgeschnei Hettrick, L. Novak, K. Lynn,
Schwedes,
derten C. Chen, K. Thompson, R.
Powder Technol.
Kompositpart Higgins, U. Batra, S. Shelukar,
2002, 122 (2 – 3),
ikeln, 1. G. Kwei, D. Storey, Int. J.
109 – 121.
Aufl., Dr. Pharm. 2004, 285 (1 – 2), 135 –
[48] A. Kwade, Chem. Eng. 146.
Technol. 2003, 26 (2), 199 – Hut,
205. Mu¨nchen
[49] A. M. Cerdeira, M. 2015.
Mazzotti, B. Gander, [61] J. Schmidt,
Chem. Eng. Technol. M. Sachs, S.
2011, 34 (9), 1427 – 1438. Fanselow,
[50] A. Afolabi, O. M. Zhao, S.
Akinlabi, E. Bilgili, Romeis, D.
Eur. J. Pharm. Sci. Drummer,
2014, 51, 75 – 86. K.-E. Wirth,
[51] D. Steiner, J. H. Finke, W. Peukert,
S. Breitung-Faes, A. Chem. Eng.
Kwade, Adv. Powder Sci. 2016,
Technol. 2016, 27 (4), 156, 1 – 10.
1700 – 1709. [62] T. Zhu, J.-C. Ansquer, M.
[52] S. Romero, B. Escalera, P. T. Kelly, D. J. Sleep, R. S.
Bustamante, Int. J. Pharm. Pradhan,
1999, J. Clin. Pharmacol. 2010,
178 (2), 193 – 202. 50 (8), 914 – 921.
[53] A. Burger, R. [63] H. Ling,
Ramberger, J. T.
Mikrochim. Acta Luoma,
1980, 73 (1 – 2), 17 D.
– 28. Hilleman,
[54] J. Hecq, M. Deleers, Cardiol.
D. Fanara, H. Res.
Vranckx, K. 2013, 4
Amighi, Int. J. (2), 47 –
Pharm. 2005, 299 55.
(1 – 2), 167 – 177. [64] C. Knieke, C.
[55] T. Umeda, N. Ohnishi, T. Steinborn, S.
Yokoyama, T. Kuroda, Y. Romeis, W.
Kita, K. Kuro- da, E. Peukert, S.
Tatsumi, Y. Matsuda, Breitung-Faes,
Chem. Pharm. Bull. 1985, A. Kwade,
33 (5), 2073 – 2078. Chem. Eng.
[56] R. Panchagnula, P. Technol. 2010,
Sundaramurthy, O. 33 (9), 1401 –
Pillai, S. Agrawal, Y. 1411.
A. Raj, J. Pharm. Sci. [65] A. Monteiro, A. Afolabi,
2004, 93 (4), 1019 – E. Bilgili, Drug Dev. Ind.
Pharm. 2013,
1029.
39 (2), 266 – 283.
[57] L. Fang, S. Numajiri, D.
Kobayashi, Y. Morimoto, [66] T. Limnell, T. Heikkila, H.
Int. J. Pharm.