EVALUATION AND CHARACTERISATION OF NANOSTRUCTURE ...lib.ugent.be/fulltxt/RUG01/001/460/684/RUG01-001460684_2011_000… · In this study, we used a Planetary micro mill “PULVERISETTE

UNIVERSITY GHENT

FACULTY OF PHARMACEUTICAL SCIENCES

Department of Pharmaceutics

Laboratory of Pharmaceutical Technology

Academic Year 2009-2010

EVALUATION AND CHARACTERISATION OF NANOSTRUCTURE PRESERVATION AFTER BALL

MILLING AND SPRAY DRYING

Liesbet D’HAESE

First Master in Drug Development

Promoter Prof. Dr. C. Vervaet

Commissioners Prof. Dr. S. De Smedt Prof. Dr. T. De Beer

COPYRIGHT

”the author and the promoter give the authorization to consult and to copy

parts of this thesis for personal use only. Any other use is limited by the laws

of copyright, especially concerning the obligation to refer to the source

whenever results from this thesis are cited.”

May 27, 2010

Promoter Author

Prof.Dr. C. Vervaet Liesbet D‟Haese

Acknowledgment

First, I wish to thank my promoter from University Ghent, Prof. Dr. Apr. C.

Vervaet, to make this research internship abroad possible.

As an Erasmus exchange student, I did my research and writing of this master

thesis at the University of Helsinki. Therefore, I want to thank all the members

of the Division of Pharmaceutical Technology for helping me and answering

all my questions while performing the laboratory work. In particular Timo

Laaksonen for guiding me and advising me in the writing process.

Also a great thank to J. Heinämäki for welcoming me the first days.

Last but not least, thanks to my parents for supporting me in this big

adventure. This was one of the most wonderful experiences in my life up to

now and it would have never been possible without them.

TABLE OF CONTENTS

1. INTRODUCTION ..................................................................................... 1

1.1. GENERAL ......................................................................................... 1

1.2. COLLOIDAL SUSPENSIONS AND STABILISERS ........................... 2

1.3. PRODUCTION OF NANOPARTICLES ............................................. 7

1.4. TRANSFORMATION INTO A POWDER ........................................... 8

1.5. THEORITICAL APPROACH OF THE USED TECHNIQUES .......... 12

2. OBJECTIVES ......................................................................................... 15

3. MATERIAL AND METHODS .................................................................. 17

3.1. CHEMICALS .................................................................................... 17

3.2. MILLING PROCEDURE .................................................................. 19

3.3. SPRAY DRYING ............................................................................. 20

3.4. SOLUBILITY .................................................................................... 23

3.5. DISSOLUTION TESTS .................................................................... 23

3.6. DSC ................................................................................................. 26

3.7. XRPD............................................................................................... 26

3.8. TEM ................................................................................................. 27

4. RESULTS AND DISCUSSION ............................................................... 28

4.1. MICROSCOPY ................................................................................ 28

4.1.1. TEM of indomethacin ............................................................. 28

4.1.2. TEM of itraconazole ................................................................ 28

4.2. PHYSICOCHEMICAL CHARACTERIZATION OF THE SPRAY

DRIED POWDERS BY DSC AND XRPD ........................................ 31

4.3. EVALUATION OF SOLUBILITY AND DISSOLUTION RATE .......... 36

4.3.1. solubility and dissolution tests of indomethacin ................ 36

4.3.2. solubility and dissolution tests of itraconazole ................... 39

5. CONCLUSION ........................................................................................ 42

6. REFERENCES ....................................................................................... 44

LIST OF USED ABBREVIATIONS

DLVO theory: Derjaguin, Laundau, Verwey, and Overbeek theory

DSC: Differential Scanning Calorimetry

HPLC: High Performance Liquid Chromatography

Ph. Eur.: European Pharmacopoeia

PVDF: polyvinylidenefluoride

rpm: rotations per minute

SDS: sodiumdodecylsulphate or sodiumlaurylsulphate

TEM: Transmission Electron Microscope

UV: ultraviolet

XRPD: X-Ray Powder Diffraction

Introduction

1

1. INTRODUCTION

1.1. GENERAL

During the last decades, the formulation of drugs as nanoparticles has

become a valuable tool in pharmaceutical technology to obtain a high

dissolution rate, and hence increase (oral) bioavailability. As more and more

possible drug compounds have low solubility and/or dissolution rates, this

formulation approach is very interesting. Since 2000, there has been an

exponential growth of scientific and patent publications and at this moment

there are five products on the market that use this nanoscience strategy, more

specifically nanoparticles. They are all produced by using the top-down

approach (see below). Four of them rely on media milling and one on high-

pressure homogenization (see below). These products are all for oral

administration because this route has important advantages e.g. not painful

like injections and the patient compliance is much better. It is also preferred

from a marketing point of view. Notice that four of the five products are in their

solid state. Beside marketing aspects, physical as well as chemical stability

problems of colloidal suspensions are the major reasons for this. (Van

Eerdenbrugh et al., 2008a).

With this master thesis, we wanted to provide more insight in the

appearance and behaviour of the final product that we obtain after the

process of ball milling and spray drying. Following questions have been made:

How do the particles look like? Can we preserve the nanostructure after spray

drying? Are there any changes in their chemical or physical properties? What

about their crystallinity? Do we need excipients to keep a reasonable

dissolution rate after spray drying?

To study this, we worked with two different drugs: indomethacin (a non-

steroidal anti-inflammatory drug) and itraconazole (a triazole antifungal

agent). What they have in common is their poor solubility in water. We used

them purely as model compounds. The compounds themselves don‟t need

any bioavailability improvements but they are easy to obtain and the increase

Introduction

2

in dissolution rate can already be seen in in vitro experiments. Both products

were processed in a similar way. The same measurements were performed in

order to compare their behaviour and properties after ball milling and spray

drying.

1.2. COLLOIDAL SUSPENSIONS AND STABILISERS

A colloidal suspension consists of two separate phases. One phase

has dimensions remarkably smaller than the other. The suspended phase

(internal phase) is dispersed uniformly throughout the suspending medium

(external phase). The size of the suspended particles ranges between 1 nm

and 1 µm and they have a minimal degree of solubility in this system (Nash,

2007).

The goal of developing a colloidal suspension is to improve

bioavailability of poorly water-soluble drugs after oral administration. This

improvement is caused by a higher dissolution velocity, due to a total surface

area that is typically orders of magnitude larger compared to the coarse

suspension. This phenomenon can be explained by the Noyes-Whitney

equation (1.1).

L

CCsDA

dt

dW )( (1.1)

where:dt

dW: rate of dissolution (g/s)

D : diffusion coefficient (cm2/s)

A: surface area of the solid (cm2)

Cs: solubility or concentration of a saturated solution at

the surface of the solid particle (g/ml)

C: concentration of the solid in the bulk dissolution

medium (g/ml)

L: diffusion layer thickness (cm)

Introduction

3

However, the increase of total surface area has one big disadvantage:

this state is thermodynamically unstable because the interface is bigger (the

Gibbs energy (ΔG) is positive). To minimize the total energy (because a

system always wants to be in its lowest energy state), the nanoparticles tend

to agglomerate. Due to this process, the surface area is minimized and a

thermodynamically stable state is obtained. By adding stabilisers, this process

is slower because they increase the activation energy of the agglomeration

process. Figure 1.1. shows the basic theory of thermodynamics and kinetics.

Thermodynamics defines whether a reaction emerges spontaneously or not.

A reaction will occur when ΔG is negative. However, it doesn‟t explain why a

reaction goes fast or slow. Kinetics can offer the answer here: it depends on

the activation energy (Ea). A reaction goes faster when its activation energy is

lower (http://mcat-review.org/rate-kinetics-equilibrium.php).

FIGURE 1.1. ILLUSTRATION OF THE DIFFERENCE BETWEEN

THERMODYNAMICS AND KINETICS: ΔG = GIBBS ENERGY; Ea =

ACTIVATION ENERGY.

(http://mcat-review.org/rate-kinetics-equilibrium.php)

Because of the amphiphilic character of a stabilizing system, it can act

as a boundary between the hydrophobic product and the hydrophilic medium.

The hydrophobic part is orientated to the particles surface, while the

hydrophilic part is orientated to the medium. It is important to find the right

Introduction

4

type of stabiliser in the right concentration. This is the key to a successful

production. Ionic surfactants can be used to obtain colloidal suspensions. Also

non-ionic surfactants or polymers are possible. They fulfil their function

electrostatically and sterically, respectively (Van Eerdenbrugh et al., 2008a;

Van Eerdenbrugh et al., 2009).

FIGURE 1.2. ILLUSTRATION OF THE POTENTIAL ENERGY AS A

FUNCTION OF INTERPARTICLE DISTANCE (CLASSICAL DLVO):

TOTAL POTENTIAL ENERGY (Vtot), ELECTROSTATIC ENERGY (Vel)

AND LIFSHITZ-VAN DER WAALS ENERGY (Vlw) (Van Eerdenbrugh et

al., 2008a).

The Derjaguin, Laundau, Verwey, and Overbeek theory (DLVO theory)

describes the interaction energy between the electric double layers of two

identical charged particles as a function of interparticle distance, illustrated in

figure 1.2. The upper curve shows the electrostatic repulsion, while the lower

curve shows Van der Waals attraction. The sum of those two curves is the

net energy balance. An electrostatic stabiliser influences the electrostatic

curve. Notice that the ζ-potential rather than the Nernst potential rules the

degree of repulsion between identical charged particles, illustrated in figure

1.3. Particles can be stabilised by giving them a maximal ζ-potential. The

energy barrier will be sufficient in suspension, but not when sedimentation

occurs. Then the pressure will cause caking. Adsorption of polymers or non-

ionic surfactants with long chains prevents close approach of the particles

Introduction

5

sterically. When two chains try to come closer, the entropy decreases (partial

loss of conformational freedom), resulting in an increase of Gibbs energy (see

equation 1.2). So, repulsion is preferred from an energetic point of view (Liang

et al., 2007).

ΔG = ΔH – T ΔS (1.2)

Where: ΔG: Gibbs energy (J)

ΔH: enthalpy (J)

T: temperature (K)

ΔS: entropy (J/K)

FIGURE 1.3. DIFFUSE DOUBLE LAYER AND THE ZETAPOTENTIAL

(http://www.malvern.de/LabGer/technology/images/zeta_potential_sch

ematic.png)

Despite of the wide range of stabilizers, only a few are used for

colloidal suspension production. The poloxamers and Tween®80 are the most

used excipients. Also sodiumlaurylsulphate (SDS) as an ionic surfactant is a

common used option. Additionally, we have the lecithins, cellulosics, polyvinyl

alcohol and povidones. You can use them in ratios from 1/3 to 50/1,

representing drug/stabilizer (weight/weight). Their molecular structures can be

seen in figure 1.4. (Van Eerdenbrugh et al., 2008a).

Introduction

6

FIGURE 1.4. MOLECULAR STRUCTURE OF (A) POLOXAMER; (B)

TWEEN®80; (C) SODIUMLAURYLSULPHATE; (D) LECITHINS; (E)

CELLULOSICS; (F) POVIDONES; (G) POLYVINYLALCOHOL.

((A)http://www.medicinescomplete.com/mc/martindale/2009/images/c9003-

11-6.gif;(B)http://ntp.niehs.nih.gov/?objectid=0709A276-0D0E-3EBD-

A3B3CCC2CD707101;(C)http://chemicalland21.com/specialtychem/perchem/

SODIUM%20LAURYL%20SULFATE.htm;(D)http://www.diytrade.com/china/4/

products/489819/Lecithin_Phosphatidylcholine_PC.html;(E)(F)(G)Van

Eerdenbrugh et al., 2009)

Introduction

7

Higher concentrations of stabilisers have, in general, a positive effect

on colloidal suspension production and hence stability. The more hydrophobic

the surface is, the more nanoparticles tend to agglomerate. That is why the

surface hydrophobicity of a particular compound is a decisive factor when

considering the feasibility of obtaining a colloidal suspension (Van

Eerdenbrugh et al., 2009).

1.3. PRODUCTION OF NANOPARTICLES

There are different possibilities to produce nanoparticles. You can

obtain them by building up particles, starting from dissolved molecules that

have to be precipitated (bottom-up approach). They have a great potential to

increase bioavailability but due to some limitations, no products using this

system have been released yet. Some disadvantages of this production

process are: polydispersity, scale-up difficulties, long term process and low

drug loading efficiencies. The most common way is by reducing the particle

size of larger crystals (top-down approach). This includes high pressure

homogenization and media milling. The former technique reduces the particle

size by forcing a suspension through a very thin hole of about 25 µm many

times at high velocity. The latter technique mentioned, grinds the drug

particles and stabiliser using milling media, e.g. glass, zirconium oxide, highly

cross-linked polystyrene resins (Van Eerdenbrugh et al., 2009).

In this study, we used a Planetary micro mill “PULVERISETTE 7

premium line” (Fritsch, Idar-Oberstein, Germany) for wet milling in order to

produce a colloidal suspension. The grinding bowls rotate around their own

axis and around a central axis, though in opposite directions. Consequently,

the grinding material and grinding balls are subject to centrifugal forces which

act alternately in the same and opposite directions. The grinding balls move

along the inner wall of the grinding bowl and hit the opposite wall. Speeds up

to 1100 rotations per minute (rpm) can be reached. Because we can achieve

such high energy values, production of particles below 100 nm are possible

(http://www.fritsch.de/uploads/media/BA_075000_0100_e_02.pdf).

Introduction

8

1.4. TRANSFORMATION INTO A POWDER

As mentioned above, a solid dosage form is preferred over liquid

forms. Patient convenience is one of the reasons. Another reason is the

physical stability: The Ostwald-Freundlich equation (1.3) describes the

process of Ostwald ripening. This is the growth of bigger particles at the

expense of smaller ones. This phenomenon can be understood by a higher

solubility of the smaller particles, so there is a mass transport from high to low

concentration.

212

1 11

303,2

2log

rrRT

MW

S

S (1.3)

Where: S1, S2: solubility of the particles (g/ml) with radius r1 and

r2 (m), respectively

γ: interfacial tension between liquid and solid (N/m)

MW: molecular weight of solid (g/mol)

R: gas constant (J/mol.K)

T: absolute temperature (K)

ρ: density of the solid (g/m3)

Another physical stability issue is sedimentation. Stokes´ law (1.4.) describes

this process.

18

)(2

ss

t

gDV (1.4.)

Where: Vt:sedimentation velocity (m/s)

ρs-ρ :density difference between particle and medium

(kg/m3)

g: gravityconstant (m/s2)

DS: average particle radius (m)

µ: viscosity medium (kg/m.s)

Introduction

9

Agglomeration causes a higher rate of sedimentation. Sedimentation results

in forming a cake at the bottom that is hard to resuspend. It can be explained

by the high interparticular forces in the sediment (Van der Waals attraction at

small distance).

Despite the preference of solid formulations, we have to keep in mind

that the maintenance of the rapid dissolution is the primary goal of our

nanoparticle formulation. Hence, the disintegration of the powder as well as

the redispersion of the nanoparticles cannot be a barrier on the overall

dissolution process (van Eerdenbrugh et al., 2008b).

To transform the colloidal suspensions into a solid state, spray drying

can be used. In the article of Wang et al. (2005), an overview is given of

drying technologies for nanomaterials. They state that spray-drying is often

used because of the low-cost, simplicity, and the ease to scale-up. In our

study, we used the Mini Spray Dryer B-191 (Büchi, Flawil, Switzerland). The

initial step consists of atomizing/spraying suspensions into droplets and their

dispersion into the gas (air). Subsequently, a drying process results in

aggregates of the nanoparticles, from loose flocculates to completely fused

particles (Lee, 2003). They are essentially microparticles, but nanostructured.

The advantage: microparticles are easier to formulate into e.g. tablets,

capsules; while after redispersing in water, the individual nanoparticles are

released and provide an improved therapeutic action (Peltonen et al., 2010).

The article of Cal and Sollohub (2010), gives a clear review about the

spray dryer hardware and process parameters that affect the properties of the

final product. „Trial and Error‟ is the most common way to optimise these

parameters for every single product. The mutual interrelationships among

them can be seen in table 1.1. This tool is very useful when performing the

drying process.

Introduction

10

TABLE 1.1. MUTUAL INTERRELATIONSHIP BETWEEN SPRAY DRY

PARAMETERS (http://www.buchi.com/Mini_Spray_Dryer_B-

290.179.0.html?&no_cache=1&file=308&uid=2283).

Lee (2003) compared particle size distribution of micro- and

nanoparticles before and after spray-drying. In figure 1.5. it can be seen that

nanoparticles show a unimodal peak of 100-500 nm before spray drying, while

microparticles have a bimodal distribution. After spray drying, the

microparticles keep the same distribution. In contrast, the nanoparticles size

distribution has become bimodal and an increase of size can be seen. The

aggregates (microparticles) which have been formed can disintegrate into

nanoparticles after some time. He also noticed that there were no peaks

observed between the main ones during disintegration time. This means that

the aggregates break into primary nanoparticles cooperatively, not one by

one. Due to this possibility of breaking aggregates into nanoparticles, this

nanoformulation will have different dissolution and disintegration

characteristics than a microformulation. Also the dissolution rate will be higher

because primary particles can dissolute before the complete breakup of

aggregates.

Introduction

11

FIGURE 1.5. (a) NANOPARTICLE SIZE DISTRIBUTION BEFORE/AFTER

SPRAY DRYING AND MEASUREMENTS OF CHANGES FOR 25 HOURS

AFTER FIRST MEASUREMENT OF SPRAY DRIED MATERIAL (0 h); (b)

MICROPARTICLE SIZE DISTRIBUTION BEFORE AND AFTER SPRAY

DRYING (Lee, 2003).

Spray drying often requires the addition of a matrix former to prevent

agglomeration. It ensures the maintenance of the rapid dissolution after the

process. Agglomeration results in a bimodal dissolution profile. It consists of

an initial phase of burst release and subsequently a phase that shows slower

release. The former is due to the individually dispersed nanoparticles, while

the latter can be attributed to agglomerated nanoparticles (Van Eerdenbrugh

et al., 2008c). A negligible drying effect on dissolution was obtained for less

Introduction

12

hydrophobic compounds, whereas compounds with a more hydrophobic

surface were subject to agglomeration. The latter compounds mentioned

need a matrix former. There are several mechanisms for explaining the

agglomeration that occurs while drying the colloidal suspension, e.g. capillary

pressure theory, crystal bridge theory, hydrogen bond theory, chemical bond

theory, etc. There is no sole theory that is generally accepted (Wang et al.,

2005).

With the technique of X-ray microanalysis during electron microscopy,

the degree of dispesion of a drug in different matrix formers can be studied

qualitatively. This technique is only possible for drugs that contain a chemical

element that´s not present in the matrix former, e.g. chloride in itraconazole.

The results in the article of van Eerdenbrugh et al. (2008d) suggest that

dispersion of the drug in the matrix former can influence the dissolution

performance.

1.5. THEORITICAL APPROACH OF THE USED TECHNIQUES

Besides performing the dissolution tests of the spray dried powders, we

also made some measurements to investigate their physical properties. In

following paragraphs, some brief descriptions of the used analytical

techniques are given.

Differential Scanning Calorimetry (DSC) is used to perform a thermo

analysis. It measures changes in chemical or physical properties of a sample.

The sample and the reference material are subjected to a controlled

temperature program, while the difference in energy input (enthalpy) is

measured as function of temperature. The temperature increases linear with

time. The basic principle: when a physical transformation occurs, e.g. a phase

transition, the required heat to maintain the same temperature as the

reference will change. When an exothermic process occurs, which means that

heat is released, the flat base line changes into a positive peak. An

endothermic process, on the other hand, results in a negative peak below the

Introduction

13

base line. This technique helped us to determine changes in crystallization

and melting temperatures after ball milling and spray drying (Dean, 1995;

Pungor, 1995).

Another technique, X-ray powder diffraction (XRPD), identifies the

crystallinity of the powders. By comparing the resulting X-ray diffraction

patterns of the powders, changes in crystallinity can be detected. It was used

to confirm the results of the DSC measurements. The basic principle

underlying this technique: when a focused X-ray beam (wavelength 1 Å)

interacts with a crystal, which can be defined as a solid material consisting of

atoms, ions or molecules that are arranged in a characteristic regular pattern

(Holden & Morrison, 1982), a part of the beam will be diffracted (besides a

transmitted, absorbed, refracted and scattered part). Detection of the intensity

of the diffracted light results in a diffraction pattern, also known as a

diffractogram. This is characteristic for each type of crystal and it depends on

the constituents of the crystal lattice, their arrangement and distance. We can

call it the „fingerprint‟ of a crystal (Suryanarayana & Norton, 1998).

Imaging of the colloidal suspension and the spray dried powders was

done by a Transmission Electron Microscope (TEM) (see figure 1.6.). It uses

a high voltage beam of electrons to illuminate the sample. Because the

wavelength of electrons is much shorter than that of photons, the resolution or

resolving power is much bigger. Resolution defines the smallest distance

between two points that can be distinguished. Where glass lenses focus the

light in an optical microscope, electrostatic and electromagnetic lenses are

used to focus the electron beam in an electron microscope. The beam is

transmitted through the sample. The electrons will partially be scattered out of

the beam. This gives information about the structure after magnifying it by the

objective lenses and projecting on the fluorescent screen (Egerton, 2005).

Introduction

14

FIGURE 1.6. SCHEMATIC ILLUSTRATION OF A TRANSMISSION ELECTRON MICROSCOPE (http://en.wikipedia.org/wiki/Electron_microscope).

Objectives

15

2. OBJECTIVES

The formulation of a drug is a very important part in its development. It

is known that decreasing the particle size, increases dissolution rate and

hence the bioavailability of a drug. Production of nanoparticles can be a very

interesting tool in this context, in particular for products with poor water

solubility. Earlier studies have demonstrated that ball milling can produce

nanoparticles when the right type of stabiliser in the right concentration is

used. However, the obtained product is a colloidal suspension which is not

stable enough to preserve for extended periods of time. This is the main

reason why the current products on the market are in their solid state. Spray

drying is a possible way to transform the colloidal suspension into a powder

after ball milling. This process involves some problems: the nanoparticles

aggregate into microparticles. This might form a barrier in the overall

dissolution process while the primary goal of a nanoparticle formulation is to

improve the dissolution process.

This master thesis wants to provide more insight in the appearance

and behaviour of the powder that we obtain after the process of ball milling

and spray drying. In particular, the preservation of the nanostructure of the

particles was the main objective in this study. When this is possible, the

dissolution improvements obtained after ball milling could also be preserved.

The advantage: microparticles are easier to formulate, while after redispersing

in water, the individual nanoparticles are released and provide an improved

therapeutic action.

First, we prepared the colloidal suspension of indomethacin and

itraconazole by ball milling. These drugs were used as model compounds.

Lutrol®F127 was added to both as a stabiliser before ball milling.

In the second step, the colloidal suspensions were transformed into a

powder by spray drying. Because a matrixformer is needed to keep the

dissolution improvements, we added lactose monohydrate before spray drying

Objectives

16

to a part of the obtained colloidal suspensions. Thus, we obtained 4 different

powders: Indomethacin and Lutrol®F127 (ratio 1/0,25); Indomethacin,

Lutrol®F127 and lactose monohydrate (ratio 1/0,25/10); Itraconazole and

Lutrol®F127 (ratio 1/0,25); Itraconazole, Lutrol®F127 and lactose monohydrate

(ratio 1/0,25/10).

We took pictures of the colloidal suspensions and spray dried powders

with a TEM to observe shape and size of the particles. Then, DSC was used

to perform a thermo analysis. It helped us to determine the changes in

crystallization and melting points after spray drying. To confirm the obtained

results of DSC, we used XRPD to study the crystallinity of the powders.

At this stage, the results for spray dried material with and without

lactose monohydrate as a matrixformer were unclear. This led us to perform

the dissolution tests. They allowed us to observe changes in dissolution rate

before and after spray drying. And also the need for a matrixformer could be

examined.

Material and methods

17

3. MATERIAL AND METHODS

3.1. CHEMICALS

Indomethacin (figure 3.1.) was supplied by Hawkins (Minneapolis,

USA).

Molecular weight: 357,79 g/mol

Melting point: 158°C-162°C

FIGURE 3.1. MOLECULAR STRUCTURE OF INDOMETHACIN

(http://upload.wikimedia.org/wikipedia/commons/thumb/a/a5/Indometha

cin.png/200px-Indomethacin.png).

Itraconazole (figure 3.2.) was supplied by Apotecnia (Murcia, Spain).


Melting point: 166 °C

FIGURE 3.2. MOLECULAR STRUCTURE OF ITRACONAZOLE

(http://www.lookchem.com/ITRACONAZOLE/).


18

Poloxamer 407 (=Lutrol® F127) was a gift from Orion (Hollola, Helsinki)

It‟s a polyoxyethylene-polyoxypropylene block polymer. Its general molecular

structure can be seen in figure 3.3.

Molecular weight: between 9840-14600 g/mol

Melting point: 53°C -57°C

FIGURE 3.3. MOLECULAR STRUCTURE OF POLOXAMER 407

WITH a= ca. 98 AND b= ca. 57

(http://www.uspbpep.com/usp28/v28230/usp28nf23s0_m66210.htm).

Lactose monohydrate (figure 3.4.) was supplied by DMV-Fonterra

Excipients (Nörten-Hardenberg, Germany).


Melting point: 214°C

FIGURE 3.4. MOLECULAR STRUCTURE OF LACTOSE

(http://www.edinformatics.com/math_science/science_of_cooking/lacto

se.htm).


19

3.2. MILLING PROCEDURE

We used the Premium Line Pulverisette 7 planetary micro mill (Fritsch,

Idar-Oberstein, Germany) to decrease the particle size of both drugs,

indomethacin and itraconazole. The two vessels were prepared by adding

successively 70,0 g zirconium oxide (ZrO2) milling balls with 1 mm diameter

(Fritsch, Idar-Oberstein, Germany); 0,5 g Lutrol® F127 (this is 25 % of the

amount of the drug), 2,0 g drug and 10,0 ml purified distilled water (according

to the European Pharmacopoeia (Ph. Eur.)). After placing the vessels in the

milling machine, we set the following parameters: speed at 1100 rpm, 6 cycles

of 3 minutes milling and 15 minutes pause with reverse option on. Then the

temperature of the vessels was measured as a quality control. A temperature

of approx. 50°C was reached for both drugs. Another 4 cycles of 3 minutes

milling and 15 minutes pause with reverse option and speed at 1100 rpm

were performed.

At least one hour after cooling down, we poured out the milling balls

and the colloidal suspension of the two vessels on a sieve to separate both.

We added 40,0 ml purified distilled water in several steps to wash off the

indomethacin from the milling balls. For the colloidal suspension of

itraconazole, we carried out the same, but now we used 60,0 ml purified

distilled water, because we noticed that we lost too much of the drug by

washing with only 40,0 ml purified distilled water. As mentioned, there is

always some product that keeps sticking on the milling balls by using this

method. Because we didn‟t exactly know what the real concentration of drug

was at that point, we prepared a dilution of the colloidal suspensions to

quantify with high performance liquid chromatography (HPLC).

For the quantification with HPLC we prepared one sample of

indomethacin by diluting our colloidal suspension (approx. 6,66 weight%)

10 000 times in 100 % ethanol (Altia, Rajamäki, Finland) and another sample

in 50%/50% ethanol/purified distilled water. The ethanol was used to ensure

that indomethacin was dissolved. We achieved a solution with a concentration


20

of approx. 6,66 µg/ml. For itraconazole, we diluted our colloidal suspension of

approx. 5 weight% 2000 times in pure methanol (Rathburn, Walkerburn,

Scotland). We achieved a solution with a concentration of approx. 25 µg/ml.

The HPLC measurements were performed by using Agilent 1100

Series (Agilent technologies, Waldbronn, Germany). We used the Luna 3u

C18 (2) 100 A (Phenomenex, AllerOd, Denmark) column as stationary phase

for indomethacin. 0,2% phosphoric acid (H3PO4) pH 2,0 (Riedel de Häen,

Seelze, Germany) and acetonitrile (VWR Prolabo, Briare, France) with ratio

40%/60% was used as mobile phase. For itraconazole, we used the Gemini

NX3u C18 110 A column (Phenomenex, AllerOd, Denmark) 0,1 % trifluoro

acetic acid (CF3COOH) pH 2,0 (Sigma-Aldrich, Helsinki, Finland) and

acetonitrile with ratio 50%/50% was used as mobile phase. Before starting the

measurements, it was important to filter the mobile phase and the samples to

remove possible particles in the solutions. We used Acrodisc LC 13 mm

syringe filter with 0,2 µm PVDF (polyvinylidenefluoride) membrane (Pall,

Portsmouth, UK). First, we performed a prewash with air locked to avoid

bubbles. Secondly, we carried out a pre-treatment of the column with mobile

phase only, to get a basic line. Finally, we inserted the samples. For

indomethacin, 20 µl of each sample was automatically injected in the system

and the flow rate was 1,5 ml/min. Detection occurred with ultraviolet (UV)

spectrophotometry at 320 nm wavelength. For itraconazole, 10 µl was

automatically injected and flow rate was 1,5 ml/min. Detection occurred with

UV at 261 nm.

3.3. SPRAY DRYING

After measuring the concentration of the colloidal suspension with

HPLC (see above), we were able to dilute exactly to a colloidal suspension

that contained 1 weight% of indomethacin. A first sample was prepared by

diluting 25,0 ml of the 5 weight% colloidal suspension of indomethacin to

125,0 ml with purified distilled water. A second sample was prepared in the

same way, but lactose monohydrate as a matrix former was added in an


21

amount according to 10 times the amount of indomethacin, 12,5 g. The

quantification of the colloidal suspension of itraconazole gave us a

concentration of 4,8 wt%. We diluted 35,0 ml of this to 175,0 ml for a first

sample. A second sample was prepared in the same way but now we added

lactose monohydrate, 10 times the amount of itraconazole, 16,8 g.

These 4 samples were used for spray drying, to obtain a dry powder:

- Indomethacin+Lutrol®F127 ratio: 1/0,25

- Indomethacin+Lutrol®F127+Lactose monohydrate ratio: 1/0,25/10

- Itraconazole+Lutrol®F127 ratio: 1/0,25

- Itraconazole+Lutrol®F127+Lactose monohydrate ratio: 1/0,25/10

We used the Mini Spray Dryer B-191 (Büchi, Flawil, Switzerland), see

figure 3.5. The used parameters can be seen in table 3.1. for indomethacin,

and table 3.2. for itraconazole.

TABLE 3.1. SPRAY DRY PARAMETERS FOR INDOMETHACIN SAMPLES

parameter Setting

aspirator 95 %

Pump 15 %

inlet temperature 160°C

outlet temperature ±95°C

air pressure 6,5 bar

air flow 600-800 ml/min

TABLE 3.2. SPRAY DRY PARAMETERS FOR ITRACONAZOLE SAMPLES

parameter Setting

aspirator 95 %

Pump 15 %

inlet temperature 150°C

outlet temperature ±86°C

air pressure 6,5 bar

air flow 600-800 ml/min


22

FIGURE 3.5. SCHEMATIC PRESENTATION OF A SPRAY DRYER:

1.SAMPLE, 2.FEED PUMP, 3.TWO FLUID NOZZLE, 4.FEED CONNECTION,

5.AIR CONNECTION, 6.INLET COOLING WATER, 7.OUTLET COOLING

WATER, 8.HEATER, 9.DRYING CHAMBER, 10.WASTE COLLECTING

VESSEL, 11.TEMPERATURE PROBE OUTLET, 12. CYCLONE: PRODUCT

SEPERATED FROM AIR, 13.PRODUCT COLLECTING VESSEL, 14.

ASPIRATOR EXHAUST GAS TO FILTER, 15.MAIN SWITCH, 16. AIR FLOW

SWITCH, 17.INLET AND OUTLET AIR TEMPERATURE, 18. ASPIRATOR

IN %, 19. PUMP IN %


23

3.4 SOLUBILITY

For the dissolution tests (described in the next chapter), it is important

to work in sink conditions. When performing them, a correlation can be seen

between the in vitro dissolution rates and in vivo observations (Gibaldi and

Feldman, 1967). According to the Ph. Eur., sink conditions are obtained when

the volume of the dissolution medium is at least 3 to 10 times the saturation

volume. Then, the material that is already in solution does not affect the

dissolution rate of the residual material. To know the saturation volume, we

first performed a saturated solubility test of pure indomethacin in the

phosphate buffer with pH 6,8 (see below). The solubility of pure itraconazole

was tested in the medium containing 0,1 M hydrochloric acid and 0,2% SDS

(see below). The tests were performed in duplicate. 50 mg of drug was put in

50 ml medium. The samples were placed on a horizontal-shaker for 24 hours

at 37°C. We filtrated the suspension by a pumping installation through a 0,2

µm hydrophilic polypropylene membrane filter (Pall corporation, Ann Arbor,

Michigan, USA). The obtained solution of indomethacin was measured with

UV spectrophotometry (see chapter 3.5.). HPLC was used to measure the

obtained solution of itraconazole (see chapter 3.2).

3.5. DISSOLUTION TESTS

To comply with the dissolution requirements for solid dosage forms that

have a conventional release and are administered orally, we followed the Ph.

Eur. We used the paddle apparatus Erweka DT-D6 (Erweka, Espoo, Finland)

to measure the dissolution rate of our samples. Known amounts of the

powders were put in a gelatine capsule and a wire was placed around it to

ensure that the capsule would sink. To compare our results, we also

measured the dissolution rate of the colloidal suspension. We took samples of

0,01 ml of the colloidal suspension that we obtained after the ball milling (this

is approx. equal to 2 mg drug). Because this was a very small amount, we first

mixed it with 5 ml of the buffer and then put it in the medium. All samples were

measured in triplicates. The average was calculated and standard deviation


24

was determined. To give a proper summary of the samples we measured, we

refer to table 3.3.

TABLE 3.3. OVERVIEW OF THE SAMPLES MEASURED FOR THE

DISSOLUTION TESTS OF INDOMETHACIN AND ITRACONAZOLE

Samples amount

pure INDa ±5 mg

physical mixtured of INDa and F127b ±2,5 mg

INDa and F127b after spray drying ±2,5 mg

INDa, F127b and LACc after spray drying ±25 mg

colloidal suspension of INDa and F127b after ball milling

0,01 ml

pure ITRe ±3 mg

physical mixtured of ITR and F127b ±3 mg

ITRe and F127b after spray drying ±3 mg

ITRe, F127b and LACc after spray drying ±25 mg

colloidal suspension of ITRe and F127b after ball milling

0,01 ml

aIND: indomethacin

bF127: Lutrol

® F127

cLAC: lactose monohydrate

dphysical mixture: made by mixing IND and F127 (ratio 1/0,25) with pestle and mortar

eITR: itraconazole

For the dissolution tests of indomethacin we used 900,0 ml of a

phosphate buffer with pH 6,8 as a dissolution medium. Sink conditions were

maintained throughout the experiment. The buffer was made according to the

Ph.Eur: First, 8,0 g sodiumhydroxide (myrkkyä II LK, Sweden) was dissolved

to 1,00 L purified distilled water to obtain a concentration of 0,2 M. Then, 54,4

g potassium dihydrogen phosphate (Riedel-de Häen, Seelze, Germany) was

dissolved to 2,00 L purified distilled water which resulted in a concentration of

0,2 M. 500,0 ml of 0,2 M potassium dihydrogen phosphate was placed in a

2000 ml volumetric flask and 224,0 ml 0,2 M sodiumhydroxide was added. We

diluted this to 2000,0 ml with purified distilled water. We checked the pH of

the buffer with a Schott pH-meter (Mainz, Germany).


25

The temperature of the dissolution apparatus was set at 37°C and

paddle speed at 100 rpm. At specified time points, a 5 ml sample was taken

out of the dissolution medium and 5 ml of fresh buffer, also equilibrated at

37°C, was added. Every sample was withdrawn from the same spot between

the surface of the dissolution medium and the top of the rotating blade.

We quantified dissolved indomethacin with a UV spectrophotometer

(Ultrospec III, Pharmacia LKB, Burladingen, Germany) at 318 nm. Buffer was

used as reference. Before measuring, we filtered the samples using Acrodisc

LC 13 mm syringe filter with 0,2 µm PVDF membrane (Pall, Portsmouth, UK).

We generated a calibration curve by making different dilutions of

indomethacin in ethanol (Altia, Rajamäki, Finland) with known concentration.

Ethanol was used as reference. We made the dilutions in duplicate and the

average of the absorbance was taken. The obtained calibration curve can be

seen in figure 3.6. The correlation coefficient (0.9999) confirms that there was

a linear correlation between concentration and absorbance.

FIGURE 3.6. CALIBRATION CURVE OF INDOMETHACIN IN ETHANOL

MEASURED AT 318 NM.


26

For itraconazole, we used 500,0 ml dissolution medium containing 0,1

M hydrochloric acid (Merck kgaa, Darmstadt, Germany) and 0,2% SDS which

was supplied by Sigma-Aldrich (Helsinki, Finland). This was done to ensure

that this poorly soluble drug could dissolve. Sink conditions were obtained.

We took samples of 3 ml at certain time points and 3 ml of fresh medium was

added. The other conditions were the same as the dissolution tests of

indomethacin. We refer to the second paragraph of this chapter.

Quantification of itraconazole was now done by HPLC (Agilent

technologies, Waldbronn, Germany). The method we used is described in the

forth paragraph of chapter 3.2.

3.6. DSC

The operating conditions of DSC (DSC823e, Mettler Toledo GmbH

analytical, Greifensee , Switzerlan) can be consulted in table 3.4. The

powders we measured are the same as those from XRPD. They can be seen

in table 3.5.

TABLE 3.4. PARAMETER SETTINGS OF DSC MEASUREMENTS

Parameter setting

temperature range 25°C-235°C

heating rate 10°C/min

N2 gas flow rate 50 ml/min

3.7. XRPD

XRPD is used to make a comparison of the crystal structure of the

powders before and after the process of ball milling and spray drying. It was

performed by using the θ-θ diffractometer (Bruker AXS D8, Karlsruhe,

Germany). We used symmetrical reflection mode with Cu Kα radiation (1,54

Å) using Göbel Mirror bent gradient multilayer optics. A scintillation counter

measured the scattered intensities. The range of the angle was from 5° to 40°,

with steps of 0,02° and time was 2s/step. Sharp peaks indicate the presence


27

of crystalline powder, while deflected peaks are an indication of amorphous

material. In table 3.5., an overview is given of the powders that we did

measure.

TABEL 3.5. OVERVIEW OF POWDERS MEASURED BY XRPD AND

DSC

pure INDa

physical mixture of INDa and F127

b

INDa and F127

b after spray drying

INDa, F127

b and LAC

c after spray drying

pure ITRe

physical mixtured of ITR

e and F127

b

ITRe and F127

b after spray drying

ITRe, F127

b and LAC

c after spray drying

pure F127b

pure LACc

LACc after spray drying

aIND: indomethacin

bF127: Lutrol

® F127

cLAC: lactose monohydrate

dphysical mixture: made by mixing IND and F127 (ratio 1/0,25) with

pestle and mortar

eITR: itraconazole

3.8. TEM

We used the Transmission electron microscope (FEI Tecnai F12,

Philips Electron Optics, The Netherlands) to study the size and the

appearance of the particles after spray drying. The samples were made by

dispersing some spray dried powder in purified distilled water. A droplet of

suspension was placed on a Formvar Carbon Film on 300 Square Mesh

Copper Grids. The samples were then dried for at least 3 hours.

Results and discussion

28

4. RESULTS AND DISCUSSION

4.1. MICROSCOPY

4.1.1. TEM of indomethacin

Figure 4.1.a and 4.1.b show TEM images of the suspension after ball

milling that contains indomethacin and Lutrol®F127 (ratio 1/0,25). The size of

the particles was around 500 nm. Figure 4.1.c and 4.1.d show TEM images of

the spray dried powder of indomethacin and Lutrol®F127 (ratio 1/0,25).

Nanoparticles can be seen. Figure 4.1.e and 4.1.f are TEM images of the

spray dried powder of indomethacin, Lutrol®F127 and lactose monohydrate

(ratio 1/0,25/10). The difference with the spray dried formulation without

lactose monohydrate is not really clear. We can only say that some

nanoparticles could still be seen. In general, the morphology of the particles is

similar for the colloidal suspension and the spray dried powders. No sharp

surroundings could be seen, only smooth edges. This can be explained by the

surfactant on the surface of the particles.

4.1.2. TEM of itraconazole

Figure 4.2.a and 4.2.b show TEM images of the suspension obtained

after ball milling of itraconazole and lutrol®F127 (ratio 1/0,25). In contrast to

the suspension of indomethacin, the formed nanoparticles looked like

needles. Their size was around 300 nm. In figures 4.2.c and 4.2.d, we can

observe that nanoparticles maintained their size and morphology after the

spray drying of itraconazole and lutrol®F127 (ratio 1/0,25). Figure 4.2.e and

4.2.f show TEM images of spray dried itraconazole, lutrol®F127 and lactose

monohydrate (ratio 1/0,25/10). The cloudiness in the images is due to the

large amount of lactose monohydrate. In general, the morphology of the

particles was similar for the colloidal suspension and the spray dried powders.

Edges are not sharp because of lutrol®F127 on the surface of the particles

and the needle-like shape was maintained after the spray drying process.


29

FIGURE 4.1. TEM IMAGES OF: (A) and (B) INDOMETHACIN +

LUTROL®F127 AFTER BALL MILLING; (C) and (D) INDOMETHACIN +

LUTROL®F127 AFTER SPRAY DRYING; (E) and (F) INDOMETHACIN +

LUTROL®F127 + LACTOSE MONOHYDRATE AFTER SPRAY DRYING.


30

FIGURE 4.2. TEM IMAGES OF: (A) and (B) ITRACONAZOLE +

LUTROL®F127 AFTER BALL MILLING; (C) and (D) ITRACONAZOLE +

LUTROL®F127 AFTER SPRAY DRYING; (E) and (F) ITRACONAZOLE +

LUTROL®F127 + LACTOSE MONOHYDRATE AFTER SPRAY DRYING.


31

4.2. PHYSICOCHEMICAL CHARACTERIZATION OF THE SPRAY DRIED

POWDERS BY DSC AND XRPD

In figure 4.3. we can see the thermograms of the indomethacin

powders. Figure 4.4. represents the thermograms of the itraconazole

powders. The peaks that are pointing downwards show the melting points of

the materials. The transition from a solid to liquid state requires energy and

thus is an endothermic process. More heat flow is required to maintain the

same temperature increase as compared to the reference material. Normally,

the scale of the thermograms is in mcal/s/g. This means that they are scaled

by the mass of sample used. Here, we scaled the thermograms containing

indomethacin or itraconazole with the mass of indomethacin or itraconazole,

respectively, to get that peak comparable in all the measurements.

exotherm

endotherm

FIGURE 4.3. THERMOGRAM OF, FROM TOP TO BOTTOM:

- PURE LUTROL F127 (GREEN)

- SPRAY DRIED LACTOSE MONOHYDRATE (BROWN)

- PURE LACTOSE MONOHYDRATE (PURPLE)

- SPRAY DRIED INDOMETHACIN + LUTROL F127 + LACTOSE

MONOHYDRATE (LIGHT BLUE)

- SPRAY DRIED INDOMETHACIN + LUTROL F127 (BLACK)

- PHYSICAL MIXTURE OF INDOMETHACIN + LUTROL F127 (PINK)

- PURE INDOMETHACIN (DARK BLUE)


32

exotherm

endotherm

FIGURE 4.4. THERMOGRAM OF, FROM TOP TO BOTTOM:




- SPRAY DRIED ITRACONAZOLE + LUTROL F127 + LACTOSE


- SPRAY DRIED ITRACONAZOLE + LUTROL F127 (BLACK)

- PHYSICAL MIXTURE OF ITRACONAZOLE + LUTROL F127 (PINK)

- PURE ITRACONAZOLE (DARK BLUE)

In figure 4.3. and 4.4., the endothermic peaks (melting points) of pure

drug (dark blue) and pure stabilizer (green) are slightly shifted to the left when

they are in a physical mixture (pink). The peaks of the spray dried drug and

lutrol F®127 (black) are even more shifted to the left. There are mainly three

reasons for this: (1) a smaller crystalline size might lead to a lower melting

point; (2) impurities in the powder decrease the melting point (as in the case

of the physical mixture); (3) the presence of amorphous material. The peak

area is also smaller compared with pure material. There are two main reasons

for this: (1) a smaller crystalline size might lead to a lower melting energy; (2)

a change to an amorphous state of some of the material after spray drying,

which means that a smaller amount was left in crystalline form. This last

possibility can be rejected when comparing with the diffractogram of spray

dried drug and lutrol F®127 (black) in figure 4.5. and 4.6. They show similar


33

peaks as their physical mixture, which means that crystallinity was

maintained. Although the peaks are similar to the physical mixture, their

intensity is lower for spray dried itraconazole and lutrol F®127 (black).

Crystalline nanoparticles are difficult to measure with XRPD because of their

small size. Thus, we can conclude that the nanoparticles of itraconazole have

a smaller size than of indomethacin. This explanation can also be supported

by the TEM results of both drugs (see figure 4.1. and 4.2.). The TEM pictures

of itraconazole represent smaller nanoparticles, which are partially maintained

after spray drying. Notice that the spray dried material shows some peak

broadening in the diffractograms which can occur because of smaller particle

sizes. Another possible reason is from stress and strain caused by the milling

process (Van Eerdenbrugh et al., 2008c).

FIGURE 4.5. DIFFRACTOGRAM OF, FROM TOP TO BOTTOM:




- SPRAY DRIED INDOMETHACIN + LUTROL F127 + LACTOSE


- SPRAY DRIED INDOMETHACINE + LUTROL F127 (BLACK)

- PHYSICAL MIXTURE OF INDOMETHACIN + LUTROL F127 (PINK)



34

FIGURE 4.6. DIFFRACTOGRAM OF, FROM TOP TO BOTTOM:




- SPRAY DRIED ITRACONAZOLE + LUTROL F127 + LACTOSE


- SPRAY DRIED ITRACONAZOLE + LUTROL F127 (BLACK)

- PHYSICAL MIXTURE OF ITRACONAZOLE + LUTROL F127 (PINK)


The thermogram of pure lactose monohydrate (purple) can be seen in

figure 4.3. and 4.4. It shows two endothermic peaks: the peak at 149°C

represents the loss of crystalline water. The second one at 218°C represents

the melting point of the crystalline lactose (Gombás et al., 2002). The

thermogram of spray dried lactose monohydrate (brown), on the other hand,

shows an exothermic peak which represents recrystallisation. Spray drying

changes the state of lactose monohydrate from crystalline to amorphous form.

This explains also why the exothermic peak of loss of crystalline water has

disappeared. The same endothermic peak, the melting point of the crystalline

lactose, is also present here, after recrystallisation occurred. The

diffractogram of spray dried lactose monohydrate (brown) in figure 4.5. and

4.6. confirms that most of the material has changed to its amorphous form

when comparing with the diffractogram of pure lactose monohydrate (purple).


35

The light blue thermograms in figure 4.3. and 4.4., representing spray

dried drug, lutrol®F127 and lactose monohydrate, are different for

indomethacin and itraconazole. For indomethacin, one endothermic peak can

be seen, which is supposed to be the melting point of lactose. It has shifted to

the right. Also the recrystallisation peak has disappeared. It seems that

lactose changes into another crystalline state in presence of indomethacin.

For itraconazole, on the other hand, the same recrystallisation peak of lactose

as pure spray dried lactose monohydrate can be seen. So, also the

endothermic peak, the melting point of lactose, is present. It is slightly shifted

to the left which can be explained by the same reasons as mentioned before.

The bigger area of the endothermic peak is due to the scale of this

thermogram, namely to the mass of drug and not the mass of the whole

sample. The melting peaks of drug and stabilizer have disappeared. This can

be explained by their low mass compared to the mass of lactose in this

sample.

The light blue diffractogram in figure 4.5., representing spray dried

indomethacin, lutrol F®127 and lactose monohydrate, does not show many

peaks. A change to amorphous state of all material can be an explanation for

that. The preservation of the nanoparticle size can also be the reason

because when particles are nanosized, it is difficult to measure their

crystallinity with XRPD. We can reject these statements because the black

curve, representing spray dried indomethacin and lutrol F®127, shows the

same peaks as the physical mixture, which means that the crystalline form

has maintained and that it can be measured. Though, we can conclude that

the lactose monohydrate in this mixture lost its original crystallinity, explaining

the quite flat base line. The other constituents have maintained their

crystalline form, but the amount is so small compared to lactose monohydrate,

that the intensity of the peaks is low. Same conclusions can be made for

spray dried itraconazole and stabiliser in presence of lactose monohydrate.

The only difference here is that the peaks are on the same place as pure

spray dried lactose monohydrate. This is not the case for indomethacin. This

confirms that lactose seems to be in another crystalline state when spray


36

dried in presence of indomethacin. Itraconazole, on the other hand, does not

influence the behaviour of lactose while spray drying.

The results of DSC and XRPD when lactose monohydrate was added

are quite curious. We obtained different thermograms and diffractograms

when another drug was used. In future studies, more research should be

done about how a particular drug and lactose influence each other while spray

drying. At this point, after performing TEM, DSC and XRPD measurements,

we can state that it is possible to partially maintain the nanoparticle structure

after spray drying. And the measurements confirm that the nanoparticle size

of itraconazole is smaller than of indomethacin. The drug keeps its crystallinity

after spray drying which is positive in the context of long-term stability.

Amorphous drugs are quite unpredictable in their behaviour, so a crystalline

form is preferred. The lack of information about the differences between spray

dried material with and without lactose monohydrate, led us to the

performance of some dissolution tests. These tests were essential to give a

final answer to the need of excipients when we want to keep dissolution

improvements after spray drying a colloidal suspension.

4.3. EVALUATION OF SOLUBILITY AND DISSOLUTION RATE

4.3.1. solubility and dissolution tests of indomethacin

The solubility of pure indomethacin in the dissolution medium was 611

µg/ml. This is of course not the solubility of the spray dried material, but it

gives at least an idea of the magnitude of indomethacin‟s solubility. We only

needed the saturation volume to be able to work in sink conditions for the

dissolution tests. By using the known amounts of sample that can be seen in

table 3.3., we were sure that sink conditions were achieved.

By using the calibration curve of indomethacin in ethanol (see figure

3.6.), we obtain the following graphs in figure 4.7. Buffer was set as reference.

We also measured the absorbances of lutrol® F127 and the gelatine capsule

and decreased the measured absorbance value with that amount.


37

FIGURE 4.7. DISSOLUTION RATES OF, FROM TOP TO BOTTOM:

- BALL MILLED COLLOIDAL SUSPENSION OF

INDOMETHACIN AND LUTROL F127 (GREEN)

- INDOMETHACIN, LUTROL F127 AND LACTOSE

MONOHYDRATE AFTER SPRAY DRYING (LIGHT BLUE)

- INDOMETHACIN AND LUTROL F127 AFTER SPRAY DRYING

(BLACK)

- PHYSICAL MIXTURE OF INDOMETHACIN AND LUTROL F127

(PINK)


First, some general comments have to be made. It is impossible to

have a released percentage of more than 100%. Also by taking the standard

deviation into account, we still have a percentage of more than 100% for each

sample that we measured. To be sure that everything was dissolved, we

made the calibration curve of indomethacin in ethanol. The dissolution

samples, on the other hand, were in the phosphate buffer. The polarity of both

media is different and this can influence the absorbance because of a shift of

λmax. An additional reason is the interference with lutrol®F127. The presence

of this component can strengthen the signal. This was confirmed by

performing following experiment: We dissolved a known amount (=2,41 mg) of

the physical mixture of indomethacin and lutrol®F127 (ratio 1/0,25) in ethanol.

Measuring absorbance at 318 nm and ethanol set as reference, resulted in a


38

dissolved mass percentage >100% (= 136%). To avoid this problem in future

studies, we suggest making a new calibration curve where the same amount

of lutrol®F127 as in the samples is added. Note also that the used amounts of

samples for the dissolution tests were very low. Mistakes in actual weightings

could easily be made. We suggest that future studies are performed with

higher amounts without losing the sink conditions. As the shape of the graph

is more important than the absolute data, this issue does not hinder us from

using the dissolution graphs. Taking into account the purpose of the study, the

graphs are of sufficient quality for comparative purposes.

Before discussing the results, note that we consider the highest

percentage released of a sample as 100%. After recalculation we had

following results: The curve of the colloidal suspension (green), obtained by

ball milling, shows that >90% of the material is dissolved in less than one

minute. The curves of pure indomethacin (dark blue) and the physical mixture

(pink) are similar to each other. It takes around one hour to dissolve

everything. The black curve represents the spray dried material without

lactose monohydrate. It takes less than 15 minutes to dissolve 90% of this

material. For the spray dried material with lactose monohydrate (light blue),

the slope of the beginning of the curve is even steeper. It takes less than 5

minutes to dissolve 90% of the drug.

The curves prove that nanosizing the particles improves dissolution

rate. The high dissolution rate of the colloidal suspension can partially be

maintained after spray drying, due to some remained nanoparticles. By

adding lactose monohydrate as a matrixformer, the high dissolution rate can

be preserved even better. The burst release originates from individual

nanoparticles, while the second phase of slower release is due to the

agglomerated fraction.


39

4.3.2. solubility and dissolution tests of itraconazole

The solubility of itraconazole, measured with HPLC, was 104 µg/ml.

Sink conditions were reached with the dissolution method used.

FIGURE 4.8. DISSOLUTION RATES OF, FROM TOP TO BOTTOM:

- BALL MILLED COLLOIDAL SUSPENSION OF

ITRACONAZOLE AND LUTROL F127 (GREEN)

- ITRACONAZOLE, LUTROL F127 AND LACTOSE

MONOHYDRATE AFTER SPRAY DRYING (LIGHT BLUE)

- ITRACONAZOLE AND LUTROL F127 AFTER SPRAY

DRYING (BLACK)

- PHYSICAL MIXTURE OF ITRACONAZOLE AND LUTROL

F127 (PINK)


Several problems occurred during the measurements of the dissolution

samples. First of all, the method we used (see material and methods) was not

good. Two separate peaks appeared at the retention time of itraconazole

when the concentration was small. The limit of detection was not reached for

those samples. Unfortunately, there was no other method available in time.

Secondly, we noticed after all our measurements, that the filter we used –

although it is made of a hydrophilic material - captured the drug somehow. We


40

do not know the reason for this. For better results, we could use a centrifuge

next time to remove possible particles from the samples. Nevertheless, even

with all these problems, we managed to obtain reasonable release curves. Of

course, we can‟t use the absolute data. But this is not really a limitation. For

dissolution tests, it is most important to be able to compare the shape of the

curves of the different formulations.

As we can see in the dissolution graph in figure 4.8., the colloidal

suspension, the physical mixture and the spray dried material without lactose,

only reach around 80% dissolved material while the spray dried material with

lactose reaches 100% dissolved material. The difference is unclear and we

can‟t conclude that for the former ones it is impossible to reach 100%.

Probably, we were not able to integrate the peaks in a sufficiently good way

for those.

As mentioned above, we consider the highest amount that dissolved as

100%. We obtained following results after recalculation: As we can see in

figure 4.8, pure itraconazole (dark blue) has not dissolved in the medium after

two hours. The concentration is 0 at every time point. Probably something

dissolved, but the amount is too low to detect with the HPLC method. The

physical mixture (pink) seems to start dissolving after 10 minutes and in less

than 30 minutes, more than 90% is released. At least we can say that the

stabilizer helps the drug to dissolve. The colloidal suspension (green),

obtained after ball milling, dissolves very fast. The first sample was taken after

1 minute. Already 70% of the drug has dissolved then. All of the drug had

dissolved after 2 minutes. The black curve represents the spray dried material

without lactose as a matrixformer. It had a slower release, which means that

in less then 30 minutes more than 90% of the drug amount had dissolved. On

the other hand, when we added lactose monohydrate before spray drying, the

slope of the beginning of the curve was steeper. We reached 92% dissolved

material after 4 minutes. Everything dissolved within 15 minutes.


41

For itraconazole, the improvement of dissolution rate of spray dried

material with matrixformer compared to the spray dried material without

matrixformer is higher. This might be caused by its higher hydrophobicity. A

drug with a more hydrophobic surface will have more difficulties to break up

into nanoparticles than a less hydrophobic drug. And hence a matrixformer is

more needed in this case.

The graphs might give a wrong impression because we did not

recalculate our results, considering the highest amount dissolved as 100%. In

table 4.1., the percentage of drug dissolved after 10 minutes is given. Here we

do have considered the largest amount dissolved as 100%. This table makes

it easier to compare the different behaviour of indomethacin and itraconazole

when performing the dissolution test.

TABLE 4.1. COMPARISON OF PERCENTAGE DRUG DISSOLVED

AFTER 10 MINUTES

Samples amount of IND

a dissolved

(%) after 10 minutes

amount of ITRb

dissolved (%) after 10 minutes

pure drug 55 0

physical mixture of drug and F127 c 54 71

drug and F127c after spray drying 88 73

drug,F127c and LAC

d after spray

drying 99 99

colloidal suspension of drug and F127

c

99 100

a IND: indomethacin

b ITR: itraconazole

c F127: lutrol®F127

d LAC: lactose monohydrate

Conclusion

42

5. CONCLUSION

TEM results showed that we were able to produce colloidal suspensions

of indomethacin and itraconazole by ball milling, using Lutrol®F127 as a

stabilizer (ratio drug/stabiliser was 1/0,25). The size of the nanoparticles was

approximately 500 nm and 300 nm, respectively. The smaller size of

itraconazole nanoparticles is confirmed by XRPD. The peaks show a lower

intensity.

The TEM pictures also show some remained nanoparticles for both

drugs after spray drying. Unfortunately, the difference between the powders

with and without lactose monohydrate is not clear. The big amount of lactose

(ratio drug/stabiliser/lactose 1/0,25/10) results in cloudiness on the pictures.

Performing a thermo analysis with DSC and measuring crystallinity with

XRPD, gave us some more information about phase transitions and changes

in crystallinity of the powders. Linking both results, allows us to conclude that

both drugs stay in their crystalline form after spray drying which is a positive

result, concerning the long-term stability of the material.

Pure lactose monohydrate changed to an amorphous form after spray

drying. This could also be seen in the presence of itraconazole. For

indomethacin, this behaviour cannot be observed. It seems that it is in another

crystalline state. We suggest performing more studies to examine the different

influence of both drugs on the matrixformer while spray drying.

To examine the effect of the processing variables on the dissolution

rate before and after spray drying, we performed dissolution tests. The shape

of both curves of the colloidal suspensions shows that nanosizing particles

strongly improves dissolution rate. This improvement can partially be

maintained after spray drying, due to remained nanoparticles. When adding

lactose monohydrate, the preservation of dissolution rate is even better.

Especially for itraconazole, the difference with and without matrixformer is

Conclusion

43

more obvious. A matrixformer is more needed here because the drug‟s

surface is more hydrophobic. Then, the aggregates have more difficulties to

break up into nanoparticles. The burst release originates from individual

nanoparticles, while the second phase of slower release is due to the

agglomerated fraction.

It would be interesting to do more research in future with different

amounts of stabiliser and matrixformer. Which concentration of stabiliser

allows us to obtain a colloidal suspension and how do both influence the

preservation of the nanoparticle structure and hence the dissolution rate? Also

using other types of stabiliser and matrixformer could be examined. Another

suggestion is to perform the spray drying again with different process

parameters to examine how they influence the powders‟ behaviour and

appearance. The same measurements could be done as in the master thesis,

because they allowed us to make concrete conclusions about our main

objectives.

In conclusion, it was shown that it is possible to partially preserve the

nanoparticle structure of indomethacin and itraconazole after spray drying

their colloidal suspension. This feature ensures that the increased dissolution

rate is partially maintained, and even more when lactose monohydrate is

added.

References

44

6. REFERENCES

Cal, K.; Sollohub, K. (2010). Spray Drying Technique. I: Hardware and

Process Parameters. J. Pharm. Sci., 99, 575-586.

Dean, J. A. (1995). Dean’s Analytical Chemistry Handbook. The McGraw-Hill

Companies, New York, USA, Chapter 15.2.

Egerton, R. F. (2005). Physical Principles of Electron Microscopy. An

Introduction to TEM, SEM and AEM. Springer Science+Business Media, New

York, USA, Chapter 1.4 & Chapter 3.

European Pharmacopoeia 6.7. (2010).

Gibaldi, M.; Feldman, S. (1967). Establishment of sink conditions in

dissolution rate determinations- theoretical considerations and application to

nondisintegrating dosage forms. J. Pharm. Sci., 56, 1238-1242.

Gombás, Á.; Szabó-Révész, P.; Kata, M.; Regdon Jr. G.; Erõs I. (2002).

Quantitative determination of crystallinity of a-lactose monohydrate by DSC. J.

Therm. Anal. Calorim., 68, 503-510.

Holden, A.; Morrison, P. (1982). Crystals and crystal growing. The MIT-Press,

Massachusetts, USA.

Lee, J. (2003). Drug nano-and microparticles processed into solid dosage

forms: physical properties. J. Pharm. Sci., 92, 2057-2068.

Liang, Y.; Hilal, N.; Langston, P.; Starov, V. (2007). Interaction forces between

colloidal particles in liquid: Theory and experiment. Adv. Colloid. Interfac.,

134-135, 151-166.

References

45

Nash, R. A. (2007). Suspensions. In: Encyclopedia of Pharmaceutical

Technology, Swarbrick, J. (Ed.), Informa Healthcare USA, New York, USA,

pp.3597-3610.

Peltonen, L.; Valo, H.; Kolakovic, R.; Laaksonen, T.; Hirvonen, J. (2010).

Electrospraying, spray drying and related techniques for production and

formulation of drug nanoparticles. Expert. Opin. Drug. Del., published ahead-

of-print as doi: 10.1517/17425241003716802.

Pungor, E. (1995). A practical Guide to Instrumental Analysis. Boca Raton,

Florida, USA, 181-191.

Suryanarayana, C.; Norton, M. G. (1998). X-ray Diffraction: A Practical

Approach. Plenum Press, New York, USA.

Van Eerdenbrugh, B.; Van den Mooter, G.; Augustijns, P. (2008a). Top-down

production of drug nanocrystals: Nanosuspension stabilization, miniaturization

and transformation into solid products. Int. J. Pharm., 364, 64-75.

Van Eerdenbrugh, B.; Froyen, L.; Van Humbeeck, J.; Martens, J. A.;

Augustijns, P.; Van den Mooter, G. (2008b). Drying of crystalline drug

nanosuspensions- The importance of surface hydrophobicity on dissolution

behavior upon redispersion. Eur. J. Pharm. Sci., 35, 127-135.

Van Eerdenbrugh, B.; Vercruysse, S.; Martens, J. A.; Vermant, J.; Froyen, L.;

Van Humbeeck, J.; Van den Mooter, G.; Augustijns, P. (2008c).

Microcrystalline cellulose, a useful alternative for sucrose as a matrix former

during freeze-drying of drug nanosuspensions- A case study with itaconazole.

Eur. J. Pharm. Biopharm., 70, 590-596.

References

46

Van Eerdenbrugh, B.; Froyen, L.; Van Humbeeck, J.; Martens, J. A.;

Augustijns, P.; Van den Mooter, G. (2008d). Alternative matrix formers for

nanosuspension solidification: Dissolution performance and X-ray

microanalysis as an evaluation tool for powder dispersion. Eur. J. Pharm. Sci.,

35, 344-353.

Van Eerdenbrugh, B.; Vermant, J.; Martens, J. A.; Froyen, L.; Van Humbeeck,

J.; Augustijns, P.; Van den Mooter, G. (2009). A screening study of surface

stabilization during the production of drug nanocrystals. J. Pharm. Sci., 98,

2091-2103.

Wang, B.; Zhang, W.; Zhang, W. (2005). Progress in drying technology for

nanomaterials. Dry. Technol., 23, 7-32.

Websites:

http://chemicalland21.com/specialtychem/perchem/SODIUM%20LAURYL%20

SULFATE.htm (04-05-2010)

http://en.wikipedia.org/wiki/Electron_microscope (27-04-2010)

http://mcat-review.org/rate-kinetics-equilibrium.php (20-04-2010)

http://ntp.niehs.nih.gov/?objectid=0709A276-0D0E-3EBD-

A3B3CCC2CD707101 (04-05-2010)

http://upload.wikimedia.org/wikipedia/commons/thumb/a/a5/Indomethacin.png

/200px-Indomethacin.png (20-04-2010)

http://www.buchi.com/Mini_Spray_Dryer_B-

290.179.0.html?&no_cache=1&file=308&uid=2283 (04-05-2010)

References

47

http://www.diytrade.com/china/4/products/489819/Lecithin_Phosphatidylcholin

e_PC.html (04-05-2010)

http://www.edinformatics.com/math_science/science_of_cooking/lactose.htm

(20-04-2010)

http://www.fritsch.de/uploads/media/BA_075000_0100_e_02.pdf (20-04-2010)

http://www.lookchem.com/ITRACONAZOLE (20-04-2010)

http://www.malvern.de/LabGer/technology/images/zeta_potential_schematic.p

ng (18-04-2010)

http://www.medicinescomplete.com/mc/martindale/2009/images/c9003-11-

6.gif (04-05-2010)

http://www.uspbpep.com/usp28/v28230/usp28nf23s0_m66210.htm (20-04-

2010)

EVALUATION AND CHARACTERISATION OF NANOSTRUCTURE ...lib.ugent.be/fulltxt/RUG01/001/460/684/RUG01-001460684_2011_000… · In this study, we used a Planetary micro mill “PULVERISETTE

Documents