Drug Delivery System

POLITECNICO DI TORINO
Repository ISTITUZIONALE
Ordered Mesoporous Silica for Drug Delivery in Topical Applications
Original
Ordered Mesoporous Silica for Drug Delivery in Topical Applications / Gignone, Andrea. - (2016).
Availability:
This version is available at: 11583/2652565 since: 2016-10-10T18:07:22Z
Publisher:
Politecnico di Torino
Published
DOI:10.6092/polito/porto/2652565
Terms of use:
Altro tipo di accesso
This article is made available under terms and conditions as specified in the corresponding bibliographic description in
the repository
Publisher copyright
(Article begins on next page)
04 August 2020
1
Politecnico di Torino
Department of Applied Science and
Technology
XXVIII cycle
PhD thesis
2013-2016
Ordered Mesoporous
Silica for Drug Delivery
in Topical Applications
Andrea Gignone
Tutor: Prof. Barbara Onida
2
ORDERED MESOPOROUS SILICA FOR DRUG DELIVERY IN TOPICAL APPLICATIONS

3
SOMMARIO
Acronyms ............................................................................................................................................. 6
Introduction ........................................................................................................................................ 7
Ordered mesoporous silica as drug delivery system: state of the art ................................. 8
Abstract ........................................................................................................................................... 8
Drug Delivery Systems ................................................................................................................ 8
Drug release profiles .................................................................................................................. 10
Drug Carriers............................................................................................................................... 12
Ordered mesoporous silica ...................................................................................................... 13
OMS for drug delivery ............................................................................................................... 17
Drug loading strategies .............................................................................................................. 18
Supercritical CO2 ........................................................................................................................ 22
Topical Application ..................................................................................................................... 25
A new controlled release technology .................................................................................... 28
Dermatological Disease ............................................................................................................. 29
Aim of the thesis ......................................................................................................................... 32
Ordered mesoporous silica as drug delivery system: Experimental .................................. 34
Materials ........................................................................................................................................ 34
Ordered Mesoporous Silica ................................................................................................. 34
Active Pharmaceutical Ingredient ....................................................................................... 36
Chemicals ................................................................................................................................. 37
Characterization Methods ........................................................................................................ 37
X-ray diffraction ...................................................................................................................... 37
Fourier Transform Infrared spectroscopy ....................................................................... 37
Nitrogen adsorption .............................................................................................................. 38
Termogravimetry.................................................................................................................... 38
Differential scanning calorimetry ........................................................................................ 39
Scanning Electron Microscopy ............................................................................................ 39
Dinamic Light Scattering and Zeta Potential ................................................................... 39
Nuclear magnetic resonance ............................................................................................... 40
Drug Incorporation methods ................................................................................................... 40

4
Solution ..................................................................................................................................... 40
Incipient Wetness Impregnation ......................................................................................... 40
Supercrtical CO2 ..................................................................................................................... 41
Ab-initio and Molecular dynamics simulations ..................................................................... 42
Drug Release ................................................................................................................................ 44
Release Kinetics ...................................................................................................................... 44
Reservoir effect ....................................................................................................................... 44
Ultraviolet-visible spectroscopy .......................................................................................... 45
In vitro release ......................................................................................................................... 45
In-vitro Permeation studies .................................................................................................. 46
High Pressure Liquid Chromatography ............................................................................. 46
Chapter 1: Drug Incorporation .................................................................................................... 48
Abstract ......................................................................................................................................... 48
Introduction .................................................................................................................................. 48
1.1 Solution and scCO2 process comparison ....................................................................... 50
1.2 MCM-41 SCCO2 incorporation ........................................................................................ 52
1.3 Amikacin sulfate incorporation ......................................................................................... 54
Chapter 2: Clotrimazole Characterization in msu-h & mcm-41.......................................... 60
Abstract ......................................................................................................................................... 60
2.1 Drug Distribution Types in OMS ..................................................................................... 60
2.2 Experimental and Theoretical Data ................................................................................. 65
Clotrimazole: molecule and crystal .................................................................................... 65
Clotrimazole adsorption on the silica pore wall ............................................................ 67
Interactions and energetics between clotrimazole and the silica pore wall ............ 68
Model of adsorption: nitrogen adsorption and TG analysis ........................................ 70
Increasing the CTZ adsorbed quantity.............................................................................. 71
Solid state nuclear magnetic resonance ............................................................................ 74
Mobility of adsorbed clotrimazole...................................................................................... 80
Experimental and theoretical FTIR interpretation ......................................................... 82
2.3 MCM-41 filling model .......................................................................................................... 84
Chapter 3: Silicas Determing Factors in the Incorporation Process .................................. 87
Abstract ......................................................................................................................................... 87

5
3.1 OMS characteristics ............................................................................................................. 87

3.2 Characterization of Incorporated OMS ......................................................................... 89
Chapter 4: AKS-OMS Gel ............................................................................................................. 98
Abstract ......................................................................................................................................... 98
4.1 Sustained Release From OMS-AKS in AKS Saturated Solutions ............................. 99
4.2 Composition, Rheological and pH Stability of AKS-OMS Gel ................................ 100
4.3 In-vitro release studies...................................................................................................... 101
4.4 In-vitro permeation studies ............................................................................................. 102
4.5 Topological information on OMS permeation............................................................ 104
CHAPTER 5: CTZ-OMS Gel ...................................................................................................... 106
Abstract ....................................................................................................................................... 106
5.1 Sustained release of OMS-CTZ in CTZ saturated solutions .................................. 107
5.2 Composition and In-Vitro release test ......................................................................... 108
Conclusions ..................................................................................................................................... 110
Bibliography ..................................................................................................................................... 111
Appendix I – List of Pubblications and Congress .................................................................. 118
Publications ............................................................................................................................ 118
Congress ................................................................................................................................. 118
Appendix II – Pubblished Articles .............................................................................................. 119

6
ACRONYMS
Ab-Initio Molecular Dynamics AIMD

Active Pharmaceutical Ingredient API
Amikacin Sulfate AKS
Brauner-Emmet-Teller BET
Clotrimazole CTZ
Controlled Release Technology CRT
Density Functional Theory DFT
Differential Scanning Calorimetry DSC
Dynamic Light Scattering DLS
Drug Delivery DD
Drug Delivery System DDS
Ethanol EtOH
Field Emission Scanning Electron Microscopy FESEM
Fourier Transform Infrared FTIR
High Pressure Liquid Chromatography HPLC
Incipient wetness impregnation IWI
Korea Advanced Institute of Science and Technology KIT
Maximum Toxic Concentration MTC
Michigan State University MSU
Minimal Effective Concentration MEC
Mobil Crystalline of Materials MCM
Non Linear Density Functional Theory NLDFT
Ordered Mesoporous Silica OMS
Pore Size Distribution PSD
Root Mean Square Deviation RMSD
Santa Barbara Institute SBA
Silanol SiOH
Solid State Nuclear Magnetic Resonance SSNMR
Specific Surface Area SSA
Supercritical CO2 scCO2
Termogravimetric TG
Tetraethoxysilane TEOS
Ultaviolet Visible Uv-Vis
X-Ray Diffraction XRD

7
INTRODUCTION
This PhD thesis focuses the development and characterization of all

physicochemical aspects of a new controlled release technology for two
active pharmaceutical ingredient principle (Clotrimazole and Amikacin
Sulfate) using ordered mesoporous silica until the introduction onto the
market.
The first chapter comprehends the characterization of different OMS

synthesized and commercially available; the study of different
incorporation techniques based on hydrophilicity/hydrophobicity of API;
the characterization of the new impregnated OMS.
Chapter 2 is oriented on the interaction details of API on silica surfaces.

A closer look is given to the big questions of OMS-drug phenomena:
mobility, solubility, bioavailability, etc.
Chapter 3 highlight the differences between OMS and the spatial

assembly of drug inside the mesoporous channels.
Chapter 4 describes the development of the new CRT for AKS describing
all the main aspect of the innovative semisolid formulation. In-vitro and
ex-vivo release test has been produced and characterized, revealing the
functionality of the OMS reservoir effect.
In chapter 5 the same DDS have been developed for CTZ. Both the
DDS have been compared with commercially available creams.

8
ORDERED MESOPOROUS SILICA AS DRUG DELIVERY SYSTEM:

STATE OF THE ART
ABSTRACT
Drug delivery (DD) is the method or process of administering a pharmaceutical

compound to achieve a therapeutic effect in humans or animals. For this
purpose, several drug delivery systems (DDSs) have been formulated. These
include liposomes, proliposomes, microspheres, gels, prodrugs, cyclodextrins,
among others. Similar developments with other compounds have produced a
plethora of new devices, concepts and techniques that together have been
termed controlled-release technology (CRT). Amorphous silica plays a key role
due to its favourable characteristics, such as high specific surface,
biocompatibility, etc. In this chapter, starting from an overview on DDSs, the
discussion will be focused on ordered mesoporous silica (OMS). Topics of this
part will be the applications, objects and fields of use of OMS. Numerous
questions will be open and answered regarding the main issues of OMS
literature: firstly, the amorphization of the Active Pharmaceutical Ingredient
(API); secondly, the different types of incorporations; thirdly, the requirement
of DDS for low soluble API and, lastly, the description of a new CRT.
DRUG DELIVERY SYSTEMS
A drug delivery system (DDS) is defined as a formulation or a device that

enables the introduction of a therapeutic substance in the body. It
improves efficacy and safety by controlling the rate, time and place of
release. This process includes the transport of a pharmaceutical
compound in the body with the aim of achieving safely a desired
therapeutic effect at the area where it is needed.[1] Any DDS comprises a
drug formulation, a medical device or dosage form (to carry the drug
inside the body) and a mechanism for the release. These systems have
several criteria, ranging from ease of delivery to effectiveness of the
drugs. Conventional drug delivery (DD) involves the preparation of the
drug into a suitable form, such as a compressed tablet for oral

9
administration or a solution for intravenous administration. These dosage

forms have been found to have serious limitations in terms of higher
dosage requirement, lower effectiveness, toxicity and adverse side
effects.[2] This is the case of many medications, which have unacceptable
side effects due to the drug interacting with parts of the body that are not
the target. Side effects limit the ability to design optimal medications for
many diseases, such as cancer, neurodegenerative and infectious
diseases.[3] It is also important to consider the way in which a drug is
metabolized by the body. For instance, some Active Pharmaceutical
Ingredient (API) are destroyed in the intestinal tract, therefore they cannot
be introduced orally to the body. Others may be dangerous in large
amounts, which means that for patient safety it should be used a time-
release method to deliver the drug.[2] In addition, drug dosage must be
carefully calculated so that the body can use the drug. This requires a
DDS which allows for precise dosing.[2] Hence, it is necessary to develop
suitable dosage forms or controlled DDS to allow the effective, safe and
reliable application of the pharmaceutical compound to the patient.[4] Such
systems are being developed to overcome the limitations of conventional
dosage forms and offer many advantages, which include:[5]
- Improved efficiency by preventing peak-valley fluctuations

- Increased convenience
- Decreased toxicity
- Decreased side effects
- Decreased dosage frequency
- Shorter hospitalizations
- Lower healthcare costs (both short and long term)
- Viable treatments for previously incurable diseases
- Potential for prophylactic applications
- Site specific delivery
- Better patient compliance

10
Several studies have shown that the benefits aforementioned can be

achieved by correct timing of drug administration and controlled kinetics
of drug release. Thus, controlled drug release regulates the rate, the
location and aims at optimizing drug efficiency while simultaneously
reducing adverse collateral effects.[2,5–9]
DRUG RELEASE PROFILE S
The DDS employed plays a vital role in controlling the pharmacokinetic

effect of the drug. An optimal DDS ensures that the active drug is available
at the site of action for the correct time and duration. The drug
concentration at the appropriate site should be above a minimal effective
concentration (MEC) and below a minimal toxic concentration (MTC). This
concentration interval is known as the therapeutic range (Figure I.1).
Dosage forms can be differentiated according to the way the drug is
released.
The immediate release is probably the most used. The drug is released
immediately after administration. These forms usually release the drug in
a single action following a first order kinetics profile. In other words, the
drug is released initially very quickly and then passes through the mucosal
membrane into the body, reaching the highest plasma level in a
comparatively short time. Once taken into the body, the drug is distributed
throughout the body and elimination by metabolism and excretion occurs.
This elimination process also follows a first order kinetics.
The modified release of API can occur in three different ways: delayed,
extended and pulsed release. In the first case, the API release takes place
sometime later the initial administration, after which the release is
immediate. The second one is when the drug release occurs for a
prolonged period after ingestion. This allows a reduction in dosing
frequency compared to a drug presented as a conventional dosage form.

11
Figure I.1: Example of API concentration with zero order release (a) and
pulsed administrations (b).
For immediate release dosage forms the time interval in which the plasma
concentration reach the drug therapeutic range can be quite short.[10,11]
Therefore, frequent dosing or pulsed release, with its associated
compliance problems, is required. As a consequence, there is a
considerable fluctuation in drug concentration level, which often is out of
the therapeutic range.[4] This is especially an issue in chronic diseases
when patients need to take the medication for prolonged periods (Figure
I.1.b).
Extended release can be achieved using a sustained or a controlled

delivery dosage forms. Sustained release systems maintain the rate of
drug release over a sustained period (Figure I.1.a). They achieve this
mostly by the use of suitable polymers. On the other hand, controlled
release systems also offer a sustained release profile but they are
designed to lead to predictably constant plasma concentrations. This
means that they are actually controlling the drug concentration in the
12
body, not just the release of the drug from the dosage form, as is the case
in a sustained release system.
DRUG CARRIERS
Every pharmaceutical compound contains an API, which has a direct

effect in the diagnosis, treatment or prevention of diseases. It is important
to realise that the API is just one part of the medicine and it cannot be
administered to the patient on its own. Therefore, it is necessary to
formulate the drug into a dosage system containing drug carriers, which
are forms that serve as mechanisms to improve the delivery and the
effectiveness of drugs. They can be attached to drug molecules for
targeted delivery, increased efficiency or controlled release. Many of
these can also act as buffers to reduce the toxic effects of medications.
These compounds can also change the way the drug acts in the body.[12,13]
Drug carriers are used in controlled release technology (CRT) to prolong
in-vivo actions, decrease drug metabolism, reduce drug toxicity and
determine where the drug travels and how it behaves when it gets
there.[12] These include synthetic and natural compounds from a variety of
sources, ranging from lipids to nanoparticles. Some of the more common
drug carrier systems are reported in Figure I.2.

13
Figure I.2: Liposomes (I): containers made of a lipid bilayer; their limitation is
the extremely high fragility of the liposomal structure and the low transport
capacity of non-soluble drugs. Microspheres (II-III): synthetic materials, such as
ceramics or polymers, or from natural materials, such as albumin. Dendrimers
(IV): repetitively branched molecules; their properties are dominated by the
functional groups on the molecular surface. Soluble polymers (V): hollow
particles that hold drugs; drugs are loaded into the core hydrophobic block
(yellow); the crosslinking block (green) provides stability to the micelle by
forming pH reversible bonds that allow for triggered drug release; the grey block
gives the micelle aqueous solubility and stealth. Conjugated proteins (VI).[14]
ORDERED MESOPOROUS SILICA
Among all the CRT, ordered mesoporous silica (OMS) are a particular
case of the previous system. Silicon dioxide (SiO2) is one of the most
abundant oxide materials in the Earth’s crust. SiO2 is the essential
components of all silicate materials: crystalline and amorphous. All
silicates are constituted by the SiO4 tetrahedron (Figure I.3.I). The low
energy process of change the siloxane bridge (Si-O-Si) allow the
formation of infinite polymorphisms.[15]
14
Figure I.3: The SiO44- tetrahedron (I); O-Si-O angle: 109°; Si-O-Si angle: 130°-
180°. Phase diagrams of silica (II); most of this crystalline phase are not
reversible.
OMS are amorphous inorganic materials synthesized in the presence of

surfactants as templates for the polycondensation of silicic species.
Synthesis conditions such as source of silica, type of surfactant, ionic
strength, pH and composition of the reaction mixture, temperature and
duration of synthesis determine the characteristics of the porous structure
and the macroscopic morphology.[16,17]
A wide variety of ionic and non-ionic surfactants has been used for
obtaining materials with different porous and morphological
characteristics. The pore sizes of these materials are always very
homogeneous ranging from 2 to 100 nm. The huge surface area which
can easily reach values of 1000 m2·g-1 and the wide void volume (1 cm3·g-
1
) delineate the ability of these materials as CRT. In addition, their thermal,
chemical, mechanical and pH stability defines its superiority to the organic
counterpart.[18–21]
Table I.1: Brief summary of the main silicas used.

15
Name MCM-41 MSU-H SBA-15 KIT-6 Syloid
Surfactant C16TABr P123 None
pH >11 ≈7 <2
Structure Disordered
P6mmm Ia3d
In 1992 a new family of OMS was reported as MCM-X (Mobil Crystalline

of Materials). They are synthesized by using alkyl ammonium surfactants
(i.e. Hexadecyltrimethylammonium bromide (C16TABr)) and tetraethyl
orthosilicate (TEOS) or sodium silicate in basic condition. Their pore size
and wall thickness are between 4 to 2 nm. By varying the synthesis
conditions, different structures of the mesophase can be obtained. For
instance: hexagonal phase (MCM-41), cubic phase (MCM-48) and
lamellar phase (MCM-50).[22]
The first OMS synthesized with non-ionic triblock polymers were reported
by the Santa Barbara centre (Santa Barbara Amorphous).[23] As for MCM-
41, selecting the right surfactant and concentration different structure can
be obtained. For instance, Pluronic P-123,
HO(CH2CH2O)20(CH2CH(CH3)O)70(CH2CH2O)20H, is used mainly for the
silica hexagonal structure (p6mmm); while, Pluronic F-127 for the cubic
ia3d structure. The main characteristics of surfactants are:
 Chain hydrophilicity-hydrophobicity
 Head characteristic:
o Anionic

16
o Cationic
o Amphoteric
o Non-ionic
 Chain elongation
 Chain/Head hindrance volume ratio
These features have to be correlated with other synthesis conditions:
 Temperature
 pH
 Concentrations
 Co-solvents
 Co-surfactant
 Silica source
Other families of OMS are Michigan State University (MSU), Korea

Advanced Institute of Science and Technology (KIT).
After the polycondensation, an hydrothermal treatment can be executed

with the aim of enlarge and modify the pore structure and
connectivity.[21,24] Subsequently, there is the removal of surfactant through
filtration or extraction and calcination.
Surface silanols (SiOH) represent the last characteristic besides highly

organized porosity, SSA, pore volume, chemical stability, etc. Their
content depends on the way the surfactant is removed. It can be
modulated by post synthesis treatments. They are generated as
stabilization of the silica surface. The presence of SiOH on the surface
promotes the adsorption of molecules (from water to proteins) through
hydrogen bonding (H-bond).[25]

17
Figure I.4: The surface isolated SiOH (I), H-bonded (II) and geminal (III).
Furthermore, the presence of geminal and single SiOH allows the

chemical modification grafting different functionalities through covalent
linkages. For instance, they can react with alkylchlorosilanes in order to
obtain an hydrophobic surface.[26]
OMS FOR DRUG DELIVERY
Due to these outstanding features, OMS are excellent candidates in

biomedical systems as local drug delivery systems and bone tissue
regeneration.[13,27–32] For controlled release applications, it has been
shown that silica is able to store and gradually release therapeutically
relevant drugs. Furthermore, silica is used to enhance the biocompatibility
of several DDS, such as magnetic nanoparticles, biopolymers and
micelles.[27] Vallet-Regi et al. were the first to explore the drug release
properties of OMS using ibuprofen. To incorporate API they dissolved it
in hexane, obtaining an incorporated quantity of 30% by mass.[33] They
also have demonstrated the release of erythromycin and
alendronate.[34,35] Other controlled release systems have shown the
delivery of API such as vancomycin and adenosine triphosphate,
fluorescein, β-oestradiol, cholestame and calceine, camptothecin.[31,36–39]
OMS also show ability in the dissolution of poorly water-soluble drugs. An

insufficient dissolution of hydrophobic drugs in the gastrointestinal fluids
strongly limits the oral bioavailability. Mellaerts et al. loaded itraconazole
on SBA-15, an antimycoticum known for its poor aqueous solubility.
Gastrointestinal dissolution tests produced a supersaturated solution
giving rise to enhanced trans epithelial intestinal transport.[40] Ambrogi et
al. using carbamazepine into MCM-41 have verified a remarkable
increase of dissolution rates.[41] This have evidenced that OMS is a

18
promising carrier to achieve enhanced oral bioavailability for drugs with

extremely low water solubility.
At the state of the art, the explanation of these phenomena should be a

consequence of drug amorphization inside OMS. Numerous works have
demonstrated drug amorphization inside OMS just by mechanical mixing,
without any other energy supplement. These could demonstrate that the
most stable condition of drug inside silica mesopores is an amorphous
state. Indeed, NMR studies have evidenced high mobility of drug inside
OMS at ambient temperature and pointed out weak interactions.[42,43]
These arguments will be discussed in chapter 2.
DRUG LOADING STRATEGIES
Drug loading into a host material can be performed with different

techniques.[44]
The solution method is probably the most widely used drug loading
process. In this method, the drug is dissolved in a suitable solvent and the
porous material is dipped in this solution. The drug molecules are
absorbed on the pore walls. The solution method last from one to several
hours, after which OMS is separated from the solution by filtration or
centrifugation. In conclusion, the particles are dried by removing the
remaining solvent.[45] If the drug concentration near to the adsorbent
surface exceeds the saturation concentration, e.g. during the drying step,
the drug may start to crystallize on the external adsorbent surface. Since
this crystalline surface fraction may have different dissolution properties
than the amorphous solids inside the pores, its formation is not
favourable. If crystalline solids are formed on the surface, they can be
removed by washing the loaded particles. Unfortunately, the washing
process is difficult to control.[46] Solution loading is simple to perform and
it gives reproducible results as the properties of the carrier are consistent.

19
High temperature is not required during the process, making it suitable for
loading sensitive molecules. To achieve high loading degrees, relatively
high concentrations of loading solutions have to be used. This can be
challenging, especially with poor soluble drugs. To overcome this
inconvenience, different solvents can be used. Among all the possibilities,
apolar solvent are the most appropriate. Indeed, it has been demonstrated
that water, ethanol and other polar solvents compete in the adsorption
with the drug. In fact, due to the high interaction energy and strong H-
bonds formed with surface silanols, these molecules are strayed from
silica surface only during degas operations. On the other hand, most
apolar solvent are carcinogen or toxic.[8] The main disadvantage of the
immersion method is the large proportion of the drug that is wasted in the
filtration/centrifugation process. It is also difficult to predict the drug
loading degree that will be achieved.
The Incipient wetness impregnation (IWI) method consists in contacting

the OMS with a volume of solvent equal to the silica pore volume. The
capillary action draws the solution into the pores and the drug dissolved
in it. The support can be dried to drive off the volatile components within
the solution, depositing the drug on its surface. While drying, the drug
located in the solution outside the pores is driven inside by diffusion, which
is a much slower process. There are different methodologies
distinguished by the volume of solvent used:[47]
 Impregnation by soaking or with an excess of solution: excess

liquid is eliminated by evaporation or by draining;
 Dry or pore volume impregnation: the amounts of drug are
introduced in the volume corresponding to the pore volume of the
support;
 IWI: it is a procedure similar to dry impregnation, but the volume
of the solution is more empirically determined to correspond to that
beyond which the support begins to look wet.

20
The main advantage of this methods is that it is easy to control the amount
of the drug that is loaded in the carrier. Furthermore, the drug is efficiently
loaded and the method is, therefore, suitable for expensive molecules.
Figure I.5: Comparison between different incorporation process and drug

adsorption.[48]
Covalent grafting method is widely used for payload molecules, which can
be loaded by attaching them covalently on the mesoporous material
surfaces. A commonly used method is to form a peptide bond between
the amine group of a payload molecule and a carboxylic acid linker on the
surface of the material. The benefit of a chemical grafting is the possibility
to control the release of the payload molecules. The release can be
determined by cleavage of the payload molecule from the linker or erosion
of the porous material. In chemical grafting, the drug load is controlled by
the surface area and the density of the linker and the payload molecules.
The maximum loading degree that can be achieved is inevitably lower
than this that can be achieved with non-covalent adsorption methods,
because the payload molecules cannot accommodate the whole pore
volume. With this method, there is a risk that the payload molecule will not
be released from the linker. Nevertheless, this method provides stronger
chemical interactions between the drug molecule and the surface.[16]
It is important to remember that all the methods aforementioned have an

inherent disadvantage: the employment of organic solvents. The

21
subsequent solvent removal steps add processing complexity and

increased cost to these processes. In addition, there are safety issues to
be considered and appropriate environmental protection measures must
be initiated when using organic solvents. To avoid this, other techniques
can also be employed to load drugs onto OMS.
In the hot melt method drug molecules are loaded from a molten phase.
The drug is heated along with the adsorbent to a temperature above the
melting point of the drug. A prerequisite for using this method is that both
the adsorbent and the drug have sufficient thermal stability, which
excludes all of the pharmaceuticals that are known to decompose upon
melting.[46] Some drugs as ibuprofen has been successfully loaded by
using this method. However, this method is not generally appropriate
because it requires the molecules to withstand a temperature above their
melting point and the viscosity must be low enough to allow the molecule
to efficiently enter the pore structure.
Vapour deposition is another technique. In high vacuum condition, the

melting temperature of the drug is reduced. Consequently, it is possible
to avoid degradation and evaporates small amount of drug that will be
deposited in a cold region (OMS). This method allows to control the purity
of the drug avoiding any external parameter: solvent, viscosity, solubility,
etc.
The physical mixing, for instance, requires only that drug and carriers are
blended in desired proportions using spatula for some minutes. A more
scalable process of the physical mixing is the ball milling. Indeed, during
the process the particles are physically broken to favour the drug
entrapment and reconstituted due to the strong interaction energy of silica
surfaces.

22
SUPERCRITICAL CO 2
Organic solvents can be replaced with supercritical carbon dioxide. Drugs

may also be incorporated by dissolving them in compressed high volatile
fluids like supercritical carbon dioxide (scCO2), at temperatures and
pressures above their critical point.[49] A supercritical fluid can be defined
as any substance present at a temperature and pressure higher than its
critical value, and which has a density close to or higher than its critical
density. At the point at which the critical temperature and critical pressure
are reached, the density of both the liquid and vapour are equal and the
supercritical phase is obtained.
The employment of scCO2 for drug loading offers many advantages

compared to traditional organic solvents. This topic will be discussed in
details in chapter 1.
Supercritical fluids have both liquid and gas-like properties. Their liquid-
like nature enables them to act like a solvent, while their gaseous
properties allow quick and easy diffusion through materials.[50] When the
supercritical state is reached, properties like density, viscosity and the
vapour–liquid equilibrium ratio become dependent on temperature at a
certain pressure, which permits the solubility of solutes in the supercritical
fluid to be controlled.

23
Figure I.6: Phase diagram of a CO2 showing the supercritical region and critical
point.[50]
At pressures and temperatures not too far from its critical point, a
supercritical fluid has a high compressibility. Therefore, its density and its
solvent power are easily adjustable over a wide range with a minimal
change in temperature or pressure. This tunability may be used to control
the solubility parameter.[50]
The solubility parameter is a coefficient that indicates a substance

solubilisation in a specific solvent. Materials with similar values are likely
to be miscible, hence a solute presents a complete solubility or miscibility
if it has a solubility parameter as equal as possible to that of the solvent.

24
Figure I.7: CO2 solubility parameter as a function of Temperature, Pressure

and Density.[50]
In a supercritical fluid, the solubility parameter tends to increase with the

fluid density (at constant temperature).[50] Another advantage includes the
reduced processing complexity, as there is no need for solvent removal
steps associated with organic solvent usage.[51–53] Carbon dioxide is the
most widely used supercritical fluid because it presents the advantage of
having easily accessible critical conditions, that are a critical temperature
close to ambient temperature (304.25 K) and a critical pressure which is
not too high (7.39 MPa), it is inert, non-flammable and inexpensive. The
critical temperature is close to a physiological value and so it is safe for
heat-sensitive proteins.
One disadvantage of using supercritical fluids, however, is that

specialised high-pressure equipment and knowledge are required for
25
processing it. Furthermore, some interesting hydrophobic drugs, which

cannot be impregnated by aqueous solution-suspension soaking, can be
incorporated by this method.
ScCO2 is used for many applications like:
 Extraction of desired compound from other products; i.e. caffeine

from coffee bean;[54]
 Chromatography;
 Cleaning;
 Biological applications:[55,56]
o Sterilization
o Virus inactivation
 Particle formation
o Aerogels[57]
 Polymeric processing and foams;[58]
 Impregnation
o Drug loading on different matrixes[59]
In this thesis, it will be used as an incorporation process for drug inside

OMS. This topic will be further discussed in chapter 1.
TOPICAL APPLICATION
Among all the administration routes, topical and transdermal delivery

approaches have unique advantages:
 In case of skin diseases, topical delivery directly carries drugs to

the site;
 Smaller amounts of drugs are needed to produce a therapeutic
effect;
 Plasma level peaking of drugs will be avoided;

26
 Increased bioavailability due to elimination of hepatic first-pass

metabolism;
 Greatly enhanced patient compliance by eliminating frequent
treating.
For these reasons and many others, currently, there are numerous topical
and transdermal products on the market. Many formulations of low
molecular weight drugs and macromolecules have been developed and
some are currently under clinical trial.[60]
At the same time a huge number of CRT have been developed for topical
diseases, such as gels with permeation enhancers, submicron emulsion
vehicle systems, volatile vehicle–antinucleant polymer systems, lecithin
microemulsion gel, oleo-hydrogel systems, deoxycholate hydrogels,
creams containing lipid nanoparticles, solid lipid nanoparticles, liposomes
as drug carriers, etc.[61]
Human skin has a surface area between 1.5 and 2.0 m2 for adults. The
skin thickness varies over the body with the thinnest part of eyelids being
less than 0.1 mm thick and the thickest on the palms, soles and upper
back more than 5.0 mm. Not only is the skin a protective barrier against
toxic substances, pathogens, and organisms, but it is also involved in
many important physiological functions such as fluid homeostasis,
thermoregulation, immune surveillance and sensory detection.[62] These
functions are related to the skin’s complex multiple layers with each layer
associated with highly specified cells and structures.

27
Figure I.8: A schematic image of epidermis, dermis and hypodermis structure.

The appendages such as hair shaft and hair follicle, sweat gland, sebaceous gland,
and arrector pili muscle are illustrated.[63]
The permeation barrier properties of human skin are mostly attributed to

the top layer of the epidermis, the stratum corneum. The barrier function
is related to the unique structure in the stratum corneum layer that is
composed of “bricks (corneocytes) and mortar (intercellular lamellar lipid
bilayers)”.[64]
Approaches that deliver drugs/active compounds through the skin barrier

are referred to the topical administration (as opposed to the enteral and
parenteral route). Passive and active skin penetration enhancement
methods have been successfully used to improve the efficiency of either
the topical delivery (the drugs/active compounds are delivered into skin
strata), or transdermal delivery (drugs/active compounds are delivered
into subcutaneous tissues and are taken up systemically into the body).

28
Topically applied therapies are promising for the treatment of skin

diseases such as psoriasis, contact dermatitis and skin cancers, since the
drugs are delivered directly into skin strata.
A NEW CONTROLLED REL EASE TECHNOLOGY
Despite of all the above descripted benefits of topical administration, there

are some critical drawbacks, among which difficult accurate dose and the
need of frequent reapplications.[11] These frequent reapplications are
required because the post-application efficacy of traditional creams is
limited to a period between 3 to 6 hours. In addition, recurrent treatments
result in considerable inconvenience for the patient and inopportune
amnesia. Moreover, many dermatologic pathologies, grouped under the
generic name of chronic dermatitis, have a cyclic and recurring feature,
creating complex treatment problems over the course of the patient’s life.
What is more, pulsed administration has period of inefficacy and
overdose, due to low/high API concentration on the skin site (Figure I.1).
In order to tackle these issues, zero order release at constant skin

concentration for an extended time interval is required but is utopian.
Sustained controlled release systems aims to simulate as good as
possible the zero order release. A new CRT comprising OMS,
incorporated with different API, blended in a saturated solution of the
same API is here proposed. Indeed, this semisolid formulation let to a
sustained controlled release. After the application on the skin site, the
dissolved API in the saturated solution explicate its function as all the
commercial creams. Subsequently, the concentration of the dissolved API
begins to decrease leading to the end of the cream purpose.
Simultaneously, the API incorporated in OMS begun to dissolve. The API
release preserves the saturated concentration inside the vehicle until the
OMS is empty. As follows, the therapeutic concentration of API is kept

29
constant on the application site for a long and controlled time. The OMS
incorporated with API develop a reservoir effect.
Figure I.9: A schematic image of the patented application.
This new CRT is patented at the WIPO (the global forum for intellectual
property services, policy, information and cooperation) with numbers
WO2012007906 A2-A3. At the European patent office as EP2593083 A2
and on the United States territory as US20130156832 A1.
DERMATOLOGICAL DISEASE
The major challenges in the skin diseases treatment include poor

efficiency of drug delivery into the disease site and risks of increased
toxicity associated with approaches used to improve the drug delivery
efficiency.
In this thesis, two main drugs have been explored for the treatment of
different diseases. The first is clotrimazole (CTZ), a synthetic imidazole
derivative. It is primarily used locally in the treatment of vaginal and skin
infections due to yeasts and dermatophytes. In vitro, it is most active
against Candida spp., Trichophyton spp., Microsporum spp. and
30
Malazzesia furfur (Pityrosporon orbiculare). In addition, it has some in

vitro activity against certain Gram-positive bacteria and at very high
concentrations it has activity against Trichomonas spp. In the treatment
of vaginal candidiasis, CTZ vaginal tablets have produced cure rates
comparable with those of conventional nystatin vaginal tablets. CTZ
topical preparations are generally well tolerated, but local irritation has
required withdrawal of therapy in a few cases. The site of action of CTZ,
like that of miconazole and the polyene antifungal agents, appears to be
the cell membrane to which it is preferentially bound. It has been proposed
that the mechanism of action involves an interaction with the phospholipid
layer of cellular membranes causing alterations in membrane
permeability. Permeability changes result in loss of essential precursors,
metabolites and ions, thus inhibiting macromolecular synthesis.
Absorption of CTZ through intact skin was essentially negligible in
individuals with normal skin. The highest concentration of topically applied
CTZ remained in the epidermis, particularly in the stratum corneum, with
less appearing in the dermis, and very little penetrating subcutaneously.
Consequently, topical administration avoids side effects of oral
administration like itching, nausea, vomiting and abnormal liver activity.[65]

31
Figure I.10: Chemical structure of Clotrimazole: 1-[(2-

Chlorophenyl)(diphenyl)methyl]-1H-imidazole.[66]
The second one, Amikacin sulfate (AKS) is an aminoglycoside antibiotic

used to treat different types of bacterial infections. AKS works by binding
to the bacterial 30S ribosomal subunit, causing misreading of mRNA and
leaving the bacterium unable to synthesize proteins vital to its growth.
AKS is most often used for treating severe, hospital-acquired infections
with multidrug-resistant Gram-negative bacteria, such as Pseudomonas
aeruginosa, Acinetobacter and Enterobacter. AKS can also be used to
treat non-tubercular mycobacterial infections and tuberculosis when first-
line drugs fail to control the infection. Side effects of AKS are similar to
those of other aminoglycosides. Kidney damage and hearing loss are the
most important effects.

32
Figure I.11: Chemical structure of Amikacin: (2S)-4-Amino-N-[(2S,3S,4R,5S)-

5-amino-2-[(2S,3R,4S,5S,6R)-4-amino-3,5-dihydroxy-6-(hydroxymethyl)oxan-2-
yl]oxy-4-[(2R,3R,4S,5R,6R)-6-(aminomethyl)-3,4,5-trihydroxy-oxan-2-yl]oxy-3-
hydroxy-cyclohexyl]-2-hydroxy-butanamide.
AKS may be administered once or twice a day but it must be given by

intravenous or intramuscular route or via nebulization. For the treatment
of dermatological disease, the topical route avoid the majority of side
effects.[67]
CTZ and AKS were chosen for the ability to be topically administrable, low
toxicity, the number of administration during the day (2-5 dose/day) and
their opposite characteristics. The first drug is an antifungal, the second
one is an antibiotic; one have a long half-life, the other short; but more
important, CTZ is hydrophobic, while AKS is hydrophilic. Indeed, the main
incorporation strategies will be scCO2 for CTZ and IWI for AKS.
AIM OF THE THESIS
The aim of this PhD work was to develop and characterize all
physicochemical aspects of this new CRT for CTZ and AKS using OMS
until the introduction onto the market.
The first part comprehends the characterization of different OMS

synthesized and commercially available; the study of different
incorporation techniques based on hydrophilicity/hydrophobicity of API;
the characterization of the new impregnated OMS.
Consequently, the work is oriented on the interaction details of API on

silica surfaces. A closer look is given to the big questions of OMS-drug
phenomena: mobility, solubility, bioavailability, etc.

33
Therefore, all the scCO2 incorporation parameters have been studied,

highlighting the differences between OMS and the spatial assembly of
drug inside the mesoporous channels.
Thus, the patented CRT has been developed for AKS describing all the
main aspect of the innovative semisolid formulation. In-vitro and ex-vivo
release test has been produced and characterized, revealing the
functionality of the OMS reservoir effect.
Finally, the same DDS have been developed for CTZ. Both the DDS have
been compared with commercially available creams.

34
ORDERED MESOPOROUS SILICA AS DRUG DELIVERY SYSTEM:

EXPERIMENTAL
MATERIALS
ORDERED MESOPOROUS SILICA
The synthesis of OMS are here reported.
MCM-41
Four different MCM-41 have been used during this thesis. The first one
named MCM41-1 has been synthesized following the procedure of Jana
et al.[68] Briefly, 25.04 g of hexadecyltrimethylammonium bromide
(C16TMABr) was dissolved in 75.22 g of water at 323 K. Subsequently, 16
g of 1,3,5-trimethylbenzene (TMB) was added and stirred vigorously. In
another solution 1.8 g of H2SO4 and 1.8 g of waterglass (Na2SiO3) were
added to 60 g of H2O. These two solutions were then mixed and stirred
vigorously for 30 min. Then again 30 g of water was added to this mixture,
stirring constantly. The pH of the gel was adjusted close to 10 by adding
dilute sulfuric acid and then the resultant gel was transferred into a teflon-
lined autoclave and heated statically at 373 K for its crystallization under
autogeneous pressure in an oven for 10 days. After crystallization, the
solid product was recovered by filtration, washed with large amounts of
warm deionized water and dried at 373 K and finally calcined in air at 813
K for 6 hours.
MCM41-2, MCM41-3 and MCM41-4 were synthesized following the

procedure of Grun et al.[20] modified in order to change the pore size
distribution of OMS. Briefly, n-Alkyltrimethylammonium bromides of
different alkyl chain lengths were used as template: C16TMABr and TMB
for MCM41-2; C16TMABr for MCM41-3; C12TMABr for MCM41-4. The

35
template was dissolved in 120 g of deionized water to yield a 0.055 mol/L

solution, and 9.5 g of aqueous ammonia (25 % by mass, 0.14 mol) was
added to the solution. While stirring 10 g of Tetraethoxysilane (0.05 mol)
(TEOS) was added slowly to the surfactant solution over a period of 15min
resulting in a gel with the following molar composition: 1 TEOS:0.152
CnTMABr:2.8 NH3:141.2 H2O The mixture was stirred for one hour, then
the white precipitate was filtered and washed with 100 ml of deionized
water. After drying at 363 K for 12 h, the sample was heated to 823 K (1
K/min) in air and kept for 5 h to remove the template.
A fifth sample was a commercially available MCM-41 bought from ACS

Material (MCM41-ACS).
KIT-6
Mesoporous KIT-6 silica materials were obtained following the method

reported Kleitz et al.[69] Briefly, 9.0 g of Pluronic P123 (EO20PO70EO20,
Sigma-Aldrich) was dissolved in 325 g of distilled water and 17.40 g of
HCl (37%) under vigorous stirring. After complete dissolution, 9.0 g of 1-
butanol (BuOH, Aldrich, 99%) was added. The mixture was left under
stirring at 308 K for 1h, after which 19.35 g of TEOS was added to the
homogeneous clear solution. The synthesis is carried out in a closed
polypropylene bottle. This mixture was left under stirring at 308 K for 24
h, followed by an aging step, alternatively at 50, 80, 100, or 130 °C for 24
h under static conditions (this process is referred to as hydrothermal
treatment). The resulting solid products were then filtered and dried for 48
h at 95 °C. For template removal, the as-synthesized silica powders were
first shortly slurried in an ethanol-HCl mixture and subsequently calcined
at 823 K for 2 h. The KIT-6 produced are referred in this thesis as KIT6-
50, KIT6-80, KIT6-100, KIT6-130 in order to report the hydrothermal
temperature treatment.

36
SBA-15
SBA-15 was synthesized according to the method described by Zhao et

al.[23] First, 4.0 g of surfactant, Pluronic P123, were dissolved in 30 g of
water and 120 g of 2 M hydrochloric acid solution. Then, 8.5 g of TEOS
was added. This solution was stirred for 20 h at 308 K and then aged at
353 K overnight, without stirring (hydrothermal treatment). Finally, the
product was collected by filtration, washed, and air-dried at room
temperature. Calcination was carried out at 873 K (1 K/min) for 6 h. These
sample is named as SBA15-C in this work.
Another sample was a commercially available SBA-15 and it was bought

from ACS Material (SBA15-ACS).
MSU-H
OMS of MSU-H type were bought from Sigma Aldritch and kindly provided
by Formac Pharmaceuticals (MSU-H-F)
DISORDERED MESOPOROUS SILICA
Two porous silica, with trade name Syloid AL-1 FP, were donated from
Grace GmbH & Co. These are disordered silica synthesized without the
use of any template but with different pore size distribution centred on 1
and 10 nm. Consequently, they were named as Syloid-1 and Syloid-10.
ACTIVE PHARMACEUTICAL INGREDIENT
Amikacin sulfate (AKS) (MW = 781.76 g/mol) (Ph.Eur 8.2) (purity 99,8%)
and Clotrimazole (CTZ) (MW = 344.837 g/mol) (purity 98.9%) were

37
acquired from Sigma Aldritch. They were used as such, without any
purification procedure.
CHEMICALS
Carbon dioxide with a purity of 99.5% was supplied by SIAD. Bidistilled

water was supplied by J.T.Baker. All the reagents required for OMS
synthesis (C16TMABr, TMB, TEOS, etc.), 1-fluoro-2,4-dinitrobenzene
(FDNB), 1,2-propandiol, glycerol, acetonitrile and hydroxyethyl cellulose
(Natrosol MR) were purchased from Sigma Aldritch. Commercial
formulation of AKS (5% by mass) and CTZ (2% by mass) were purchased
from a community pharmacy.
CHARACTERIZATION METH ODS
X-RAY DIFFRACTION
X-Ray Diffraction (XRD) and wide angles patterns were obtained using a
PANAlytical X’Pert Pro (Cu Kα radiation) diffractometer with a PIXcel1D
photon detector.
FOURIER TRANSFORM INFRARED SPECTROSCOPY
For Fourier Transform InfraRed (FTIR) measurements, powders were

pressed in self-supporting wafers and spectra were recorded at room
temperature with a Bruker Tensor 27 spectrometer operating at 2 cm-1
resolution. If degassing was necessary, the sample were outgassed at
373 K for one hour (residual pressure equal to 0.1 Pa). FTIR spectrum of
crystalline API were recorded on the powder dispersed in potassium
bromide (KBr). FTIR spectrum of API dissolved in solution were recorded
on a diluted carbon tetrachloride (CCl4) solution (1 g/L).

38
NITROGEN ADSORPTION
Nitrogen adsorption isotherms were measured using a Quantachrome

AUTOSORB-1 instrument after degassing at 373 K for 2 hours. Brauner-
Emmet-Teller (BET) specific surface areas (SSA) were calculated in the
relative pressure range 0.04-0.1 and the pore size distribution was
determined through the NLDFT (Non Linear Density Functional Theory)
method, using the NLDFT equilibrium or adsorption model for cylindrical
pores.[22]
TERMOGRAVIMETRY
TermoGravimetric (TG) analyses were carried out between 298 K and

1523 K in air (flow rate 100 mL/min with a heating rate of 10 K/min) using
a SETARAM 92 instrument.
The API desorption analysis were performed using a SETARAM 92

instrument following the procedure explained by Verevkin et al.[70] and
Price et al.[71] Using the Clausius-Clapeyron relation, equation (1), the
enthalpies of vaporization (∆Hvap) at the average temperature of
investigation were obtained.
𝑑𝑚 ∆𝐻𝑣𝑎𝑝
𝑙𝑛 (√𝑇 )=𝐴+ (1)
𝑑𝑡 𝑅𝑇
where dm/dt is the mass loss rate at the specified temperature; R is the
universal gas constant and T is the temperature of the isothermal
experiment. Subsequently, vaporization enthalpies were reported to
298.15 K using general methods of correction, reported by Chickos et
al.[72] (the Sidgwick’s rule):

39
∆𝐻𝑣𝑎𝑝 (298.15) = ∆𝐻𝑣𝑎𝑝 (𝑇) + 0.0545 (𝑇 − 298.15) (2)
In equation (2), T is the temperature of measurement or mean

temperature of measurement if ∆Hvap (T) has been obtained from a
Clausius-Clapeyron treatment of vapor pressures. The experimental
conditions used for the reliable determination of vaporization enthalpies
of low volatile molecular compounds are following the references of
Verevkin at al.[70] A calibration curve with phenol has been done. The
uncertainty of temperature calibration was less than 1 K.
DIFFERENTIAL SCANNING CALORIMETRY
Differential Scanning Calorimetry (DSC) measurements were performed

with a DSC1 STARe (Mettler Toledo) System apparatus of TA Instruments
equipped with a low temperature probe between 298 K and 1073 K under
nitrogen flux (flow rate 60 mL·min-1 with a heating rate of 10 K·min-1).
SCANNING ELECTRON MI CROSCOPY
Field Emission Scanning Electron Microscopy (FESEM) image were

recorded with a FESEM ZEISS MERLIN.
DINAMIC LIGHT SCATTERING AN D ZETA POTENTIAL
Dinamic Light Scattering (DLS) and Zeta Potential analysis were

performed with 90 Plus Instrument (Brookhaven) on a suspended water
mixture of OMS after homogenization with a high shear homogenizer
(Ultraturrax, Ika) for 5 min at maximum velocity.

40
NUCLEAR MAGNETIC RESO NANCE
Solid State Nuclear Magnetic Resonance (SSNMR) spectra were

recorded on a Bruker Avance II 400 instrument operating at 400.23 and
100.65 MHz for 1H and 13C nuclei, respectively. 4 mm o.d. zirconia rotors
13
(sample volume of 80 µL) were employed and spun at 12 kHz for C
CPMAS spectra. A ramp cross-polarization pulse sequence was used with
contact times of 3–5ms, a 1H 90° pulse of 3.8 µs, recycle delays of 1–60
s and 48–4096 transients. 1H MAS spectra were performed on a 2.5 mm
probe with a spinning speed of 32 kHz. A DEPTH sequence (π /2-π-π)
was used to suppress the probe background signal.[73,74] The 1H 90° pulse
length was set to 2.50 πs, the recycle delays to 1-40 s and 32-64
transients were averaged for each sample. 1H and 13
C chemical shifts
were referenced through the resonance of adamantane (1H signal at 1.87
ppm) and hexamethylbenzene (13C methyl signal at 17.4 ppm),
respectively.
DRUG INCORPORATION METHODS
SOLUTION
API incorporation was performed by adsorption from solution. To this

purpose, normally, 300 mg of API and 100 mg of OMS were introduced in
an Erlenmeyer flask with 5 ml of solvent (water or ethanol) and maintained
at room temperature under stirring for 24 hours. Then the powder was
filtered and dried overnight in vacuum (residual pressure 0.1 Pa).
INCIPIENT WETNESS IM PREGNATION
The API loading was performed by Incipient Wetness Impregnation (IWI)

procedure, based on the capillary action [47]. Typically, 1 ml of a saturated
solution of API was added drop by drop to one gram of OMS during
41
mechanical agitation. The homogeneous wet powder obtained was

vacuum dried for 24 hours at 0.1 Pa obtaining a dry powder. For the IWI
water and ethanol were used as solvents.
SUPERCRTICAL CO 2
For the incorporation process by means of scCO2 a homemade device

was used. This consists in a glass cylinder of 1 cm diameter containing a
pellet of CTZ (100 mg) and a pellet of OMS (100 mg) separated by a disc
of filter paper, in order to prevent their contact and to separate the pellets
after the incorporation process.
Figure 1.11: Photograph of the homemade device used for the scCO2
incorporation process.
This device was placed inside a stainless steel vessel, which was put in
an oven that maintained all the system at constant temperature. The
apparatus was also equipped with a volumetric pump and a back pressure

42
regulator. A scheme of the apparatus is reported in Figure 1.12 and further

details can be found elsewhere.[75]
Figure 1.12: scCO2 apparatus details: B – CO2 tank; P1 – CO2 pump; P2 – co-
solvent pump (not used); P3 – hexane pump (not used); Vxx – valves On/Off;
V4/V5/V12/V13 three way valves; VNR – not returning valve; R – reactor; M – Mixer;
S – heating coils; F – oven; BPR1 – manual back pressure regulator; BPR2 –
automatic back pressure regulator; PE – heating pad; Tx – trap; FIx – mass flow
meter.
The incorporation was performed in static condition. The vessel was filled
with liquid CO2 and heated up to 373K. After the heating, additional carbon
dioxide was pumped in the vessel to reach the final desired pressure (25.0
MPa). The system was then maintained in the described conditions for
several hours (from 6 to 18) to allow the drug dissolution and diffusion into
the OMS. At the end of the incorporation process, the temperature was
decreased to room conditions and the apparatus was depressurized.
AB-INITIO AND MOLECULAR DYNAMICS SIMULAT IONS
All the calculations were performed within the Density Functional Theory
(DFT). Concerning static calculations, the developmental version of the

43
CRYSTAL14 code29 in its massively parallel version[77] was adopted and

the computational approach is the same of Ref.[78]. Briefly, the chosen
functional was the Perdew, Burke, and Enzerhof GGA (Generalized
Gradient Approximation) exchange-correlation functional (PBE),31
including the empirical Grimme’s D2 correction,[80] to describe the
dispersive interactions (vdW). In the following, the superscript D means
that Grimme’s correction is included. Split valence double- (for Si atoms)
and triple-ζ (for other atoms) Gaussian type basis sets plus polarization
functions were used to describe the systems.[81,82] Chlorine atoms were
represented with a 86-311G* basis set.[83] Only the atomic coordinates of
the two more superficial layers of each silica slab in the docking
geometries were optimized, to compensate for the reduced thickness of
the models. Starting geometries were generated so to maximize the
interactions between the drug and the surface. Interaction energies, per
unit cell per adsorbate molecule (∆E), were calculated and corrected for
the basis set superposition error (BSSE) according to the counter-poise
methodology described in previous papers by Delle Piane et al.23,25 and
reported in Supporting Information.
Harmonic frequencies were calculated with CRYSTAL14 at Γ point and

the infrared intensity for each normal mode was obtained by computing
the dipole moment variation along the normal mode, adopting the Berry
phase method.[85] For the simulation of the IR spectra of the different
structures, only a fragment consisting of the most interesting chemical
groups has been considered for constructing the Hessian matrix and will
be defined for each case in Results and Discussion.
Enthalpies (ΔH) of adsorption at standard temperature (298 K) were

obtained from the vibrational partition functions, by applying the Zero
Point Energies (ΔZPE) and thermal (ΔET) corrections to the BSSE
corrected electronic adsorption energies (∆EC) as ∆H = ∆EC + ∆ZPE +
∆ET.

44
Ab-initio molecular dynamics (AIMD) simulations were performed using

the CP2K code.[86] The Quickstep technique[87] with a mixed plane wave
and Gaussian basis set methodology (Gaussian and Plane Wave method,
GPW) was employed to calculate the electronic structure. We used the
PBE functional, with the Goedecker−Teter−Hutter pseudopotentials[88]
and a triple-ζ basis set with polarization functions (TZVP)[89] augmented
with the empirical Grimme’s D2 correction.[80] The cutoff for the plane
wave basis was set to 400 Ry. AIMD simulations were run at 300 K in the
NVT ensemble, using the Canonical Sampling through Velocity Rescaling
(CSVR) thermostat.[90] A time step of 0.5 fs was chosen. All simulations
were equilibrated at 300 K with a more stringent thermostat (time
constant: 10 fs) for about 1 ps and then the production phase was run for
at least 10 ps with a more relaxed thermostat (time constant: 50 fs). Since
CP2K requires 3D periodic systems, a value of c = 35 Å was chosen to
separate the slab replicas with enough vacuum. In all cases, only the
superficial layer of the silica slab and the drug molecules were allowed to
move.
DRUG RELEASE
RELEASE KINETICS
The drug release kinetic of API to the saturated concentration were

evaluated in vitro by soaking 100 mg of OMS incorporated by API (OMS-
API) in 5 ml of stirred solvent. At different time intervals, a small amount
of solution (100 μl) was collected, diluted and analysed through UV–Vis
spectrophotometry analysis. To assess the amount of released API a 5-
point calibration curve has been obtained.
RESERVOIR EFFECT

45
The reservoir effect of saturated solution with OMS-API have been

evaluated for AKS and CTZ. In a saturated solution of API a certain
amount of OMS, impregnated with the same API, were added under
agitation. At regular intervals (1h) a small amount of solution (between
500-100 μl) were withdraw and substituted with the same quantity of pure
solvent. The extracted solutions were then diluted and analysed through
UV–Vis spectrophotometry to assess the API concentration. The addition
of pure solvent causes a dilution of the solution and allow us to evaluate
the reservoir effect of impregnated OMS. In order to evaluate the correct
content of drug in the solution a 5-point calibration curves has been done
for CTZ and AKS in the respective solvent. The coefficient of
determination was 0.997 for CTZ and 0.995 for AKS, both in the
concentration range 0.01-0.1 mg·ml-1.
ULTRAVIOLET-VISIBLE SPECTROSCOPY
API concentration in solution were evaluated through UltraViolet Visible

(UV-Vis) spectroscopy using a UV-Vis-NIR spectrometer (Cary 5000,
Agilent technologies). Each time and for each solution, a 5-point
calibration curve has been obtained.
IN VITRO RELEASE
API diffusion from different gels was studied using an apparatus

consisting of a vertical glass diffusion cells (Franz cell) of 6.0 mL volume
equipped by a Spectra/Por (12000-14000 Dalton MWCO) hydrophilic
cellulose membrane (Spectrum Lab). The diffusion area was 2 cm2. Gels
and commercial formulation were employed as a donor phase. The
receiving phase consisted of acetate buffer at pH 5.0 (AKS) or water
solution of HCl (pH 1.0) (CTZ). The apparatus was maintained under
stirring for 72 h at 307 K, during which at scheduled times the receiving
phase was withdrawn and entirely substituted with fresh receiving phase.
46
Each withdrawn was analysed. Each sample was prepared and analysed
in triplicate.
IN-VITRO PERMEATION STUDIES
AKS transepidermal permeation and skin uptake from the different gels
were studied using vertical Franz diffusion cells and porcine ear skin. Skin
slices were isolated using an Acculan dermatome (Aesculap) from the
outer side of pig ears freshly obtained from a local slaughterhouse and
then stored for at least 24 h at 255 K. Prior to each experiment, the
excised skin was rinsed with normal saline solution and pre-hydrated by
floating it in 0.002% (w/v) sodium azide aqueous solution. The skin was
then sandwiched between the two cells with the stratum corneum side
upwards. The diffusion area was 2 cm2. AKS-OMS gel and commercial
formulation were employed as a donor phase. The receiving phase
consisted of acetate buffer at pH 5.0. The apparatus was maintained
under stirring for 168 h at 307 K, during which at scheduled times the
receiving phase was withdrawn and entirely substituted with fresh
receiving phase. Each withdrawn was analysed using HPLC. Each
sample was prepared and analysed in triplicate.
HIGH PRESSURE LIQUID CHROMATOGRAPHY
AKS was derivatized by mixing 100 μl of aqueous solution of the drug with
300 μl of methanol, 40 μl of NaOH (0.05 M) and 50 μl of a methanolic
solution of the derivatizing agent (FDNB) (180 mg·ml-1). The obtained
mixture was heated at 363 K in an air-circulating oven for 10 min, then
cooled and injected in HPLC. Each solution was separately derivatized
prior to injection. The HPLC apparatus consisted of an isocratic pump
(Series 200, Perkin-Elmer instrument) equipped with a μBondapack C18

47
300 mm · 4.6 mm column (particle size: 10 μm; pore size: 125 Å;

endcapped) (Waters, Milford, MA, USA). The mobile phase was a mixture
of acetonitrile–water–acetic acid (47:53:0.1 v/v/v) pumped at 1.5 ml/min.
A spectrophotometric detector (LC 290, Perkin-Elmer) working at 365 nm
was used.

48
CHAPTER 1: DRUG INCORPORATION
ABSTRACT
Clotrimazole (CTZ) and Amikacin Sulfate (AKS), two active principles

widely used in dermatology, were incorporated inside Ordered
Mesoporous Silica (OMS). Three main techniques have been compared:
incorporation by solution, incipient wetness impregnation and supercritical
CO2. At the same time six main OMS have been used: MSU-H, MCM41-
1 to MCM41-4 and MSU-H-F. In all cases the pristine OMS were
characterized by nitrogen adsorption measurement. Numerous
incorporation parameters have been studied. The amount of incorporated
CTZ was higher with the scCO2 process (34% by mass), while AKS
content was higher with the IWI technique (43% by mass). The drug-
containing OMS was characterized by XRD, FESEM, TGA, FT-IR and
nitrogen adsorption analysis. In all cases the drug resulted amorphous
and distributed inside the mesopores.
INTRODUCTION
In the first pioneering work by Vallet-Regi et al.[33], drug incorporation in

MCM-41 was carried out by adsorption from a solution using hexane as
solvent.
Since the long-term toxicity of n-hexane in humans is well known[91,92],

many other solvents have been studied for the incorporation. For
instance, Charnay et al.[93] reported that when highly polar solvents such
as dimethyl sulfoxide (DMSO), dimethylformamide (DMF) and
dimethylacetamide (DMA) were used, the amount of adsorbed drug was
limited, whereas a higher quantity of drug was incorporated when ethanol

49
or hexane were used. The drug incorporation is strongly affected by the

solvent due to the competition between the drug and the solvent in the
adsorption [93].
An answer could be the IWI process. [47] This technique allows the choice
of less dangerous solvents tackling the problems of solvent removal and
adsorption competition. Surely, a high drug solubility must be persecuted
in order to enhance the drug adsorption efficacy. On the other hand, a
repetition of the IWI could be done to enhance the drug content.
Finally, also the incorporation through scCO2 is an alternative way to

adsorption or impregnation from a liquid solution. Unfortunately, not all
drugs can be solubilized by the CO2 flow. Even if pressure and
temperature can modify the solubility parameter of CO2, salts like AKS
cannot be solubilized.
[94]
Belhadj-Ahmed et al. investigated the supercritical impregnation of
vitamine E acetate on silica matrixes by means of a dynamic technique.
Min et al. studied the incorporation of ibuprofen inside SBA-15 through
[95]
scCO2 , showing that the loaded drug quantity increased with a higher
solubility of ibuprofene in scCO2.
The first paragraph of this chapter will discuss the positive aspects of the
scCO2 process compared to the incorporation from solution. For the sake
of brevity, only OMS of MSU-H type will be presented in this part. CTZ is
the model drug and ethanol, for safety reasons, is the selected solvent.
Secondly, the scCO2 process will be replied on the MCM-41 OMSs, 1 to

4.
Thirdly, the IWI method will be studied for AKS in order to compare the
two techniques, using two different OMS (MSU-H and MSU-H-F).

50
1.1 SOLUTION AND SCC O 2 PROCESS COMPARISON
The effect of the scCO2 treatment on the structure and morphology of

OMS was investigated by means of XRD and FESEM prior to the
incorporation of CTZ. The target was to study the possible effects of
scCO2 on OMS. There was no literature on this topic.
As it is reported in “Incorporation of clotrimazole in ordered mesoporous

silica by supercritical CO2” by Gignone et al.[8], wide angle XRD patterns
and FESEM micrographs showed no significant modifications.
ScCO2 was used to incorporate CTZ using the apparatus showed in the
experimental section. Pressure and time of the scCO2 treatment were
varied, in order to investigate the effect on the amount of incorporated
CTZ, whereas the temperature was maintained constant and equal to 373
K. This temperature was chosen significantly higher than the critical
temperature of carbon dioxide, at variance with what was reported by
Belhadj-Ahmed et al. [94] and by Min et al. [95]. Thermal degradation of CTZ
can be ruled out since it was reported to occur at 453 K.[96] The selected
working pressures ensured a good solvent power of the fluid. The density
and solubility parameter were respectively equal to 550 kg·m-3 and 9.1
(MJ·m-3)1/2 at 25.0 MPa, while they were 819 kg·m-3 and 12.5 (MJ·m-3)1/2
at 50.0 MPa.
As a whole, four different incorporation procedures were carried out and

their typical parameters are summarized in Table 1.1, together with the
corresponding loading of incorporated CTZ, which was measured by TG
analyses.

51
Table 1.1: Incorporation process parameters for scCO2 and ethanol solution
(EtOH) and corresponding CTZ loading evaluated by TG analysis.
Drug-
Incorporation Time Temperature Pressure Loading
OMS
Procedure (h) (K) (MPa) (%)
ratio*
scCO2 -1 6 373 25.0 1:1 12
scCO2 -2 12 373 25.0 1:1 30
scCO2 -3 12 373 50.0 1:1 34
scCO2 -4 18 373 25.0 1:1 30
EtOH 24 298 0.1 3:1 9.0
*mass ratio
The maximum amount of incorporated CTZ, corresponding to 34% by

mass, was obtained at 50.0 MPa after a 12-hours-supercritical treatment.
Data in Table 1.1 reveal that time plays a crucial role up to 12 hours. In
fact, the CTZ loading increased from 12% to 30% by mass when time
increased from 6 to 12 hours at 25.0 MPa, whereas no further increase
was observed when the process was carried out for a longer time, i.e. 18
hours, at the same pressure. These data suggest that after 12 hours the
equilibrium between CTZ in scCO2 and CTZ inside the MSU-H was
reached.
At variance, pressure was observed to affect the incorporated amount

only at a minor extent, since an increase from 25.0 MPa to 50.0 MPa
yields a limited increase of the CTZ percentage (from 30% to 34%),
probably due to a higher solubility of CTZ in scCO2 at higher pressure.
It is worth noting that the incorporation via scCO2 was largely more
efficient than that obtained by adsorption from ethanol solution. In fact, in
the latter case the percentage of CTZ inside MSU-H was only 9.0%.

52
Moreover, the amount of CTZ used for the adsorption was three times
higher than that of MSU-H, at variance with the scCO2 treatment for which
the same amounts of CTZ and MSU-H were used.
As already reported[8] CTZ is maintained in an amorphous state. The lack

of crystallization is crucial, because amorphous drugs are widely accepted
to have higher aqueous solubility and dissolution rate than related
crystalline phases. This effect may be particularly important when
molecules poorly soluble in water, such as CTZ, are considered.[41,97,98]
FT-IR spectroscopy, nitrogen adsorption measurements and TG analysis
on this samples showed important details of the CTZ distribution; but this
correlation will be discussed altogether in the following chapters.
1.2 MCM-41 SCCO 2 INCORPORATION
MCM-41 silicas have been widely studied used as drug carrier. During
this research work, through the solution technique, the maximum CTZ
content was about 6% by mass without residue mesoporosity. These data
are different from the MSU-H case, see table 2.1.
In this case four MCM-41 have been synthesized, as already described,

and characterized.

53
Figure 1.1: MCM-41 nitrogen adsorption isotherms. Pore volume, SSA and
PSD are reported in chapter 2.
Table 1.2: MCM-41 nitrogen adsorption results

Mean pore NLDFT SSA NLDFT Volume
OMS
diameter (nm) (m2g-1) (cm3g-1)
MCM41-1 4.88 812 1.286
MCM41-2 4.88 585 1.100
MCM41-3 3.65 930 0.818
MCM41-4 2.94 840 0.598
The nitrogen adsorption analysis showed a good SSA and pore volume.

54
Figure 1.2: MCM-41 FESEM images. In the upper left image MCM41-1 is
showed (mean particles size 360 nm). Upper right, MCM41-2 (210 nm). Lower
left, MCM41-3 (100 nm). Finally, MCM41-4 (1090 nm).
The drug content with the scCO2 was higher than for the MSU-H case.
The higher loading was obtained with MCM41-1 with a CTZ content of
44% by mass. XRD and FTIR analysis reported a complete amorphization
of the drug.
1.3 AMIKACIN SULFATE INCORPORATION
The drug loading was performed by the solution method and the IWI
procedure. In the first case, 300 mg of AKS where solubilized in 5 ml of
H2O and stirred for 24h with 100 mg of MSU-H-F. The water content was
removed trough filtration and drying. For IWI, typically, 1 ml of a water

55
saturated solution of AKS was added drop by drop to one gram of MSU-
H-F (0.898 cm3·g-1) during mechanical agitation. The homogeneous wet
powder obtained was vacuum dried for 24 hours to 0.1 Pa obtaining a dry
powder.
Figure 1.3: (I) XRD spectra of AKS powder (a) and MSU-H-F-AKS (47% by
mass) (b). (II) FTIR spectra of MSU-H-F (a), AKS in KBr (b) and MSU-H-F-AKS
(c). (III-IV) N2 isotherms and PSD of MSU-H-F (a) and MSU-H-F-AKS (47% by
mass) (b).
In order to control the physicochemical properties of our samples, a

complete characterization of all samples has been done. XRD results
shown a complete amorphization of APIs (Figure 1.3). The lack of
crystallization in these materials is a very well reported and crucial
[40,99,100]
phenomenon . Indeed, such application that require a dissolution
and diffusion rate in gels able to overcome the skin absorption it is even
more essential; i.e. crystalline APIs could have a dissolution rate too slow
in order to maintain the saturation point.
56
In figure 1.4 are shown the low angle XRD diffraction pattern presenting
the typical (100), (110) and (200) peaks, related to the ordered hexagonal
(P6mm) network of mesopores. The maintenance of the peaks after the
incorporation process demostrate the mesostructure preservation. The
differences in the peak intensity indicate the drug incorporation: less
density differences from the silica wall to the mesoporous pore core due
to the incorporated AKS reduce the intensity of the diffracted X-ray.
I II
Figure 1.4: (I) Low angle XRD spectra of MSU-H-F (a) and MSU-H-F-AKS (47%
by mass) (b). (II) DLS of MSU-H-F after 5 minute of high shear homogenization
(30 krpm)
FTIR spectroscopy reveals that the APIs interact with the silica surface
through Hydrogen bonds with silanols. The adsorption infrared spectra of
figure 1.3.II describes direct interaction between MSU-H-F and AKS. The
MSU-H-F as such shows two main band: free silanols at 3742 cm-1 and
interacting silanols at 3520 cm-1. The 2800-3000 cm-1 bands are the alkyl
residual chain of the template after the calcination step required to
produce MSU-H-F. AKS in KBr shows low adsorption between 3550 and
3160 cm-1 (-NH2, -NH, -OH groups), 2927 and 2850 cm-1 (CH2, CH
groups) and strong mode at 1640 and 1550 cm-1 (-CONH- group) [67]
.
Other adsorption band of -CO group are under the silica cut-off. In the
MSU-H-F-AKS sample no free silanols are observable (3742 cm-1)
suggesting a direct interaction between surface silanols and AKS.
Moreover, some new broad absorption band are observable in

57
comparison with AKS in KBr. These are attributed to SiOH interacting with
different groups of AKS.
TG analysis reports the API-incorporated quantities. UV-Vis spectroscopy

results and HPLC quantification confirms these results. The TG
measurements show in the best case (IWI) an incorporated quantity of
47% by mass, while with the incorporation by solution it was possible to
reach only a 15% by mass.
Nitrogen adsorption isotherms (Figure 1.3.III) return the specific surface

area, the specific pore volume and the pore size distribution of MSU-H-F
and MSU-H-F-AKS (Figure 1.6.IV). In the case of IWI, NLDFT specific
surface area decreased from 625 m2·g-1 to 69 m2·g-1. The pore volume
decreased from 0.898 cm3·g-1 to 0,121 cm3·g-1. Using the TG analysis
results, the volume excluded from N2 adsorption, due to the incorporation
of AKS in MSU-H-F, can be corrected, as already showed in a previous
paper.[8] Dividing the API incorporated quantity with this corrected
excluded volume the estimation of the amorphous AKS density is: 1.59
g·cm-3. The experimental density of crystalline AKS, evaluated with a
helium pycnometer (Ultrapyc 1200e of Quantachrome) is 1.60 g·cm-3 that
is in agreement with the unique literature data of Bau et al. [101], to the best
of our knowledge. This correlation shows that AKS have almost the same
density in the amorphous and crystalline state. In addition to this, the
NLDFT PSDs show a reduction of the volume without a change in pore
diameter (Figure 1.3.IV). These considerations, compared to other
[99]
literature results , strongly suggest that AKS molecules are
incorporated inside the mesopores but not uniformly distributed on the
MSU-H-F surface. Consequently, with the IWI method, the homogeneity
inside the mesopores is lost but no occlusion is present and the drug still
in an amorphous state. This results are confirmed also with the
incorporation by solution, but the incorporated quantity is much lower. On

58
the other hand, as observable in chapter 2, the scCO2 process is able to

homogenize the adsorbed drug.
Figure 1.5: FESEM images of MSU-H-F aggregates.
Figure 1.5 reports a morphological image of MSU-H-F showing

aggregates of microns composed by particles of about 500-600 nm. The
primary particles resemble short prism of hexagonal base with striping
along the height recalling the ordered hexagonal (P6mm) symmetry. The
DLS measurements of MSU-H-F as such agree with the FESEM image,
resulting in aggregates of 1300 nm mean value. Indeed, after high shear
homogenization (5 min at 30 krpm), MSU-H-F particles results completely
disaggregates (Figure 1.6.II). The mean value is 569±16 nm. The Zeta
Potential measurements show an instability of the mixture (-12 mV) but
the probable re-aggregation should be avoided in a future gel use due to
viscosity.

59
The ions extracted from the MSU-H-F of MSU-H type were Cl-, F-, NO3-
and SO42-. Sulfate is the only one over 2 ppm (18 ppm). This is not a
problem since the API used for in-vitro permeation studies is AKS.

60
CHAPTER 2: CLOTRIMAZOLE CHARACTERIZATION IN MSU-H &

MCM-41
ABSTRACT
The knowledge of the specific interactions between the surface of

mesoporous silica and drugs is necessary to guide development of new
and improved drug delivery systems. However, such knowledge is still
scarce, due to the arduous interpretation of experimental results. This
chapter reports the complete characterization of the interaction of CTZ,
inside OMS by means of a joint computational and experimental
approach. Experimentally the drug was loaded through scCO2 in MSU-H
and MCM-41. Its adsorption was investigated through FTIR, N2 adsorption
measurements, TG analysis, SSNMR spectroscopy. Modelling involved
static and dynamic Density Functional Theory simulations of CTZ
adsorbed on realistic models of amorphous silica surfaces. In the first part,
a more simplified approach is reported. The second part informs about
the agreement between the computational and experimental results,
concerning the energies of adsorption, the IR spectra and the distribution
of drug inside the mesopores. Finally, a tentative description of the
mobility at room temperature of CTZ on the silica surface was done using
molecular dynamics simulations and SSNMR results.
2.1 DRUG DISTRIBUTION TY PES IN OMS
On the same sample of chapter 1 (MSU-H and MCM-41), nitrogen

adsorption isotherms were measured to characterize the SSA, the specific
pore volume and the PSD of OMS as such and OMS-CTZ. In figure 2.1

61
are reported the adsorption and desorption branches for OMS and OMS-
CTZ.
Figure 2.1: Isotherms of MSU-H (I) and MCM41-1 (II) before (a) and after (b)
scCO2.
Table 2.1: NLDFT and TGA elaborated data of the isotherms of figure 2.1
Sample SSA [m2/g] Volume Loading Residual

[cm3/g] [gCTZ/gTOT] Volume
[cm3/g]
MCM41-1 813 1.28 46 % 0.29
MSU-H 597 0.90 34 % 0.32
As already described in a published article[8], from the grams of CTZ

incorporated inside the sample, the volume occupied by CTZ can be
calculated and the residual free mesopore volume in OMS-CTZ can be
estimated. This estimation is strongly affected by the density considered
for CTZ in the calculation. Assuming a density of 1.316 g/cm3,[66] the
mesopore volume in OMS-CTZ resulted equal to 0.336 cm3/g. This value
is in fair agreement with table 2.1. The discrepancy of 0.047 cm3/g may
arise from the value of CTZ density used in the calculation. Indeed, XRD
pattern showed that CTZ in OMS lacks of crystalline form and its density

62
is likely to be lower than that of the crystalline phase. These

considerations can be repeated for the MCM41-1 sample.
Figure 2.2: PSDs of MSU-H (I) and MCM41-1 (II) before (a) and after (b) scCO2.
The red curve (c) reports the mathematical model calculations.
Observing the PSDs reported in figure 2.2, obtained from the nitrogen
sorption isotherms of figure 2.1, it can be said that, for OMS as such (a) a
main family of pores with an average diameter of 8.5 nm and 5.0 nm are
observable. On the other hand, after scCO2 (b), these families of pores
are not anymore present.
In the case of MSU-H, the presence of smaller mesopores, could describe

a phenomenon of internal coverage of mesopores surfaces by CTZ
molecules. Consequently, an attempt has been made to model the
occupation of mesopores as a layer of CTZ molecules. The volume V0 of
a single cylindrical empty mesopore is given by
𝝅
(1) 𝑽𝟎 = ∙ 𝒅𝟎 𝟐 ∙ 𝒉
𝟒
where d0 is the mesopore diameter and h is the mesopore elongation. The

volume V1 of a single cylindrical mesopore occupied by the clotrimazole
layer is given by

63
𝝅
(2) 𝑽𝟏 = 𝟒 ∙ 𝒅𝟏 𝟐 ∙ 𝒉
where d1 is calculated as
(3) 𝒅𝟏 = 𝒅𝟎 − 𝟐𝒅𝒄𝒍𝒐
and dclo is the diameter of a clotrimazole molecule considered as a sphere.

Combining equation 1 and 2, equation 4 is obtained, which, substituting
d1 as in equation 3, gives rise to equation 5.
𝑽𝟎 𝒅𝟎 𝟐
(4) =
𝑽𝟏 𝒅𝟏 𝟐
(𝒅𝟎 − 𝟐𝒅𝒄𝒍𝒐 )𝟐
(5) 𝑽𝟏 = ∙ 𝑽𝟎
𝒅𝟎 𝟐
A new pore size distribution can be calculated from this equation, on the
basis of the values of V0 and d0 of the pore size distribution of OMS as
such (Figure 6a), experimentally obtained by the nitrogen adsorption
isotherm. In order to do so, V0 and V1 in equation 5 have to be normalized
to the same quantity of silica, because V0 is a specific pore volume,
expressed in cm3·g-1. This means that, if gsil is the amount of OMS as such
(silica), the amount of OMS-CTZ is gsil + gclo, where gclo has been
evaluated by TG analysis. Accordingly, equation 5 is rewritten as equation
6.

64
(𝒅𝟎 − 𝟐𝒅𝒄𝒍𝒐 )𝟐 𝒈𝒔𝒊𝒍

(6) 𝑽𝟏 = ∙ 𝑽𝟎 ∙ 𝒈
𝒅𝟎 𝟐 𝒔𝒊𝒍 +𝒈𝒄𝒍𝒐
This calculation is based on a simple modelization of CTZ molecules

equally distributed inside MSU-H pores. Figure 2.3 represents
schematically the above considered type of adsorption.
Figure 2.3: Pictorial representation of CTZ adsorption in mesopores; MSU-H

(I) and MCM41-1 (II).
Using these assumptions, from the PSD of OMS as such a theoretical

PSD, after the adsorption of drug, can be calculated. This is in agreement
with previous data reported for itraconazole by Mellaerts et al.,[48] who
observed a molecular dispersion of the drug in OMS.
On the other hand, these simplified distributions do not take into account
roughness, density profiles and curvature of the channels surfaces.[102]

65
Consequently, a more detailed description have been done later in this

chapter with the theoretical help.
In the case of MCM41-1, due to the fact that there is no residual PSD after
the scCO2 process (Figure 2.2), these considerations are not valid.
Indeed, the connection between nitrogen adsorption data and TG analysis
demonstrate a complete filling of the mesopore channels. A detailed
description will be reported in chapter 3. In any case, figure 2.3.II reports
pictorially these differences.
2.2 EXPERIMENTAL AND THE ORETICAL DATA
In order to describe better the drug-silica interactions the MSU-H and

MCM41-1 incorporated samples were used as models, joining
experimental and theoretical data. This work was done in collaboration
with the Theoretical Chemistry Group of Prof. Piero Ugliengo, with Dr.
Massimo delle Piane and Dr. Marta Corno.
CLOTRIMAZOLE: MOLECULE AND CRYSTAL
CTZ was modeled both in gas phase and as a crystal, before studying its
adsorption on amorphous silica. The starting point was the X-ray
experimental structure by Song et al.[66]

66
Figure 2.4: a) 3D space filling model of the CTZ tetrahedral. b) CTZ

electrostatic potential mapped on the electron density.
The CTZ crystal has been optimized with and without Grimme’s correction
and compared to experimental results (Figure 2.5).
Figure 2.5: CTZ crystal and molecule-faces interactions: a, CTZ crystal with
highlighted faces. b, c and d are the PBE-D2 optimized structures of (100) (010)
and (001) crystal surfaces with one CTZ molecule per cell adsorbed, respectively.
The added molecule is in blue to distinguish it from the clotrimazole crystal slab.

67
The structure is triclinic with two drug molecules per unit cell. The PBE-D2
cohesive energy of crystalline CTZ has been computed as -146.5 kJ·mol-1 (Table
1) and, unsurprisingly, the interactions occurring in the crystal are dominated by
dispersion (78%).
Experimentally, the lack of crystallization of CTZ inside OMS is observed.

This phenomenon has been described in literature both as a confinement
effect[41,98,103,104] and as a competition between crystal cohesion and
adsorption on the silica surface.23 To describe this competition, we have
modelled the interaction of CTZ molecules with three crystalline surfaces
of the same CTZ crystal, figure 2.5. The added CTZ molecule maximizes
the contact with the surface molecules mainly through dispersion and
electrostatics interactions. Unsurprisingly, the interaction energies of CTZ
with its crystal surfaces are comparable to each other with an average
value of -113 kJ·mol-1, lower than the crystal cohesive energy (-132
kJ·mol-1). Experimental TG desorption analysis of crystalline CTZ shows
an experimental enthalpy of vaporization of about 92 kJ·mol-1, reasonably
close to the theoretical values. The comparison between computed and
experimental cohesive energy for the CTZ crystal shows some
overestimation due to the PBE-D2 method.
CLOTRIMAZOLE ADSORPTION ON THE SILICA PORE WALL
Experimentally, CTZ was loaded into MSU-H through scCO2, achieving a

maximum drug loading of 34% by mass.
In the simulations, CTZ was adsorbed on a silica surface model described

in a previous study by some of us.[105] This surface exhibits a silanol
density of 4.5 OH·nm-2, close to the experimentally measured value for
fully hydroxylated surfaces (4.9 OH·nm-2).[106] Of these silanols, only one
is free, while the others are interacting through H-bonds. Particularly,
three SiOHs cooperates in forming a stable H-bonded chain.

68
CTZ was manually docked on the surface, aiming at maximizing the

interactions between exposed silanols and the different drug’s functional
groups. In order to match the experimental conditions, six main starting
geometries have been studied. Four of them are characterized by one
molecule per silica unit cell (drug loading ~13% by mass), while the other
two simulate a molecular layer as observed in previous results[8] with two
and three molecules per unit cell (drug loading ~27% and ~41% by mass,
respectively) (Figure 2.6).
Figure 2.6: View along the z axis of the fully optimized six different geometries
of adsorption. a) imidazole (I). b) imidazole (II). c) phenyls (I). d) phenyls (II). e)
molecular layer (I): two molecules per silica unit cell. f) molecular layer (II): three
molecules per unit cell.
INTERACTIONS AND ENERGETICS BETWEEN CLOTRIMAZOLE AND

THE SILICA PORE WALL
From all the six geometries of figure 2.6, four main types of interaction
between CTZ and silanols are observable, figure 2.7. The imidazole ring
69
can form both two (Figure 2.7.a) and one (Figure 2.7.b) H-bonds with the
surface. H-bonds of the Si-O-H---Cl type (Figure 2.7.c) are weaker than
those with imidazole. Several SiOH-π (surface-CTZ, Figure 2.7.d) have
been observed in almost all the different structures, while π-π edge to
face lateral interactions (CTZ-CTZ) characterize the molecular layer
models.
Figure 2.7: Silica-CTZ types of interactions.
An essential result that these considerations reveal is that all the

computed interaction energies and enthalpies are close to each other,
enlightening a possible competition between crystalline and adsorbed
CTZ. Assuming that the supercritical incorporation is a step route
controlled by dissolution and adsorption of single molecules, this process
is almost isoenergetic, since for each adsorption on silica an average of
110 kJ·mol-1 are gained and for each desorption from the crystal 115
70
kJ·mol-1 are lost. This could be the explanation of the lack of crystallization
in pores smaller than 20 times the molecule diameter explained by
Sliwinska-Bartkowiak et al.[103] and observed by other authors.[41,98,104] In
addition TG desorption analysis of CTZ in MSU-H (34% by mass)
produces an enthalpy of vaporization of 91.6 kJ·mol-1. As
abovementioned, a similar analysis on crystalline CTZ results in a value
of 91.8 kJ·mol-1. Thus, also the experimental data suggest that the two
processes are almost isoenergetic, supporting the hypothesis on silica-
induced drug amorphization.
MODEL OF ADSORPTION: NITROGEN ADSORPTION AND TG

ANALYSIS
As previously described, from the experimental N2 adsorption isotherms,

the PSDs, before and after the scCO2 incorporation of CTZ, can be
obtained. These show that the drug incorporation reduces the mean pore
diameter from 86 Å of bare MSU-H to a lower value of 67 Å (figure 2.2).
Clearly, with the above mentioned approach, the calculated PSDs are
unable to describe correctly the experimental results.[8]
Figure 2.8: Adsorption model. (I) and (II) side views along the a direction of
molecular layers (I) and (II), respectively.
On the other hand, experimental TG analysis reports an incorporated

quantity of 34% by mass, which is in between the loading calculated for
layer (I) and layer (II), which are ~27% and ~41% by mass, respectively.
71
By calculating the experimental planar concentration of CTZ (molecules

per nm2) and comparing it to our theoretical models of molecular layers (2
CTZs/cell, Figure 2.6.e and 3 CTZs/cell, Figure 2.6.f), it is shown that the
real system can be described by a 50:50 mixture of the two molecular
layer geometries. Indeed, the unit cell of the simulated silica surface has
an area of 1.6 nm2 and the experiment reports 2.5 CTZ molecules per 1.6
nm2. As a consequence, a simulated MSU-H-CTZ PSD has to be
calculated, starting from the experimental PSD of bare MSU-H, assuming
thicknesses representative of both the molecular layer (I) and (II) models.
These thicknesses have been evaluated following the Connolly
surfaces[107] before and after CTZ adsorption of the computed models,
with the purpose to take into account the vdW molecular volume and the
roughness generated by the statistical distribution of 2 and 3 CTZ
molecules per 1.6 nm2. Therefore, the molecular layer surfaces have been
discretized in 677 squares (0.25 Å2) in order to evaluate the thicknesses,
for each point, between the starting silica model and the molecular layers.
Subsequently, a new PSD has been calculated for each couple of
evaluated thicknesses and all curves have been combined together in the
final theoretical PSD of figure 2.2.I.c. Such procedure results in an
impressive agreement between simulation and experiment, validating the
data interpretation.
INCREASING THE CTZ A DSORBED QUANTITY
Due to the prefect agreement of this modelization, three different samples

with increasing fillings were prepared using scCO2 process. The aim was
to obtain information on CTZ interaction with silica pore wall at increasing
loading. Through XRD analysis it was possible to evaluate the presence
or not of CTZ crystals. None of the samples show diffraction peaks of
CTZ.

72
Nitrogen adsorption isotherms, figure 2.9, evidence a linear decrease in

comparison to the loaded quantity of SSA and pore volume. In table 2.3
are reported the detailed values. To the best of our knowledge, the only
literature crystalline CTZ density data is a calculated number reported in
the work of H. Song et al.[66] Using a helium picnometer, a value of 1.31
g·cm-3 has been obtained, in perfect agreement with the literature data.
Interestingly the calculated CTZ density inside the mesopore, reported in
Table 2.3, increase with the loading but never reach the crystalline
number. As already demonstrated, the CTZ is amorphous but the linear
increase of density with the loaded quantity could be due to a decreasing
pore occlusion. In addition to this, as evidenced by the position and shape
of the hysteresis loop, the mean pore diameter of the mesoporous
channels decrease with the formation of the CTZ layer.
Figure 2.9: Nitrogen adsorption isotherms of bare MSU-H (a), MSU-H-CTZ-

12 (b), MSU-H-CTZ-18 (c) and MSU-H-CTZ-34 (d). Below the PSD of bare
MSU-H (black) in comparison to MSU-H-CTZ samples (blue); the red curve is

73
the theoretical PSD corresponding to the experimental loading; from left to

right: MSU-H-CTZ-12, MSU-H-CTZ-18 and MSU-H-CTZ-34.
Through NLDFT the PSD of all samples have been calculated.

Hypothesizing the adsorption of CTZ as a step process of single
molecules that lead to the formation of a layer on the silica wall, the
experimental PSD of MSU-H-CTZ can be calculated.[99] In order to
evaluate the theoretical PSD of figure 2.9 (red), the theoretical docking
structure previously described have been used. The same procedure has
been changed with some correction in order to take into account the low
loading configuration. Briefly, a new PSD has been calculated for each
couple of evaluated thicknesses and all curves have been combined
together in the final theoretical PSD representative of each different
loading (red). For instance, to evaluate the theoretical PSD of MSU-H-
CTZ-12, the four single docking geometries (Imidazole (I) and (II), phenyls
(I) and (II)) combined to two clear silica surface (Silica as such) have been
used as starting thicknesses, resulting in 4056 point of thicknesses. Then,
more than 16 million of new PSD, one for each couple of thickness, have
been calculated from the PSD of bare MSU-H and combined together in
the left red curve of figure 2.9. The use of two clear silica surface is
required to obtain the correct planar concentration of CTZ. As it is reported
in table 2.3 a loading of 12% by mass corresponds to a surface coverage
of 0.4 CTZ·Å-2, that can be expressed using the theoretical model in a
0.64 CTZ molecule per silica unit cell.
Table 2.3: Nitrogen adsorption and TG numerical results compared to

theoretical data
CTZ
Pore Loaded CTZ·1.606
Samples experimental
volumea quantityb nm-2 d
densityc
MSU-H 0.900 - - -

74
MSU-H-
0.650 12 1.0 0.64
CTZ-12
MSU-H-
0.529 18 1.0 1.0
CTZ-18
MSU-H-
0.290 34 1.2 2.5
CTZ-34
a(cm 3·g-1). b(% by mass). c(g· cm3) obtained dividing the loss in pore volume for
the loaded CTZ quantity, normalized to the grams of silica. d(Number of CTZ
molecules per theoretical surface area of the silica unit cell). For MSU-H-CTZ-12,
Imidazole (I) and (II), phenyls (I) and (II) and two clear silica surface have been
used, obtaining a theoretical planar concentration of 0.66 CTZ / 1.606 nm 2. For
MSU-H-CTZ-18, Imidazole (I) and (II) and phenyls (I) and (II) have been used.
For MSU-H-CTZ-34, molecular layer (I) and (II) have been used.
SOLID STATE NUCLEAR MAGNETIC RESONANCE
Solid-state NMR measurements were performed in order a) to evaluate

the inclusion of CTZ in the OMS; b) to quantify the amount of included
CTZ; c) to evaluate the mobility of CTZ inside the silica.
The 13C CPMAS spectra of pure CTZ, MSU-H, MSU-H-CTZ-12, MSU-H-

CTZ-18 and MSU-H-CTZ-34 are reported in Figure 2.10, while 1H MAS
spectra are reported in Figure 2.11. Selected 1H and 13
C chemical shift
values are reported in table 2.4.
13
The C CPMAS spectra of MSU-H-CTZ-12, MSU-H-CTZ-18 and MSU-
H-CTZ-34 are characterized by broad peaks with respect to pure CTZ
whose spectrum presents sharp lines typical of highly crystalline systems.
The strong overlapping of the signals in the aromatic region (110-150
ppm) prevents any possible assignment. However, the broadness of the
resonances and the shift of the Csp3 atom on passing from pure CTZ to
75
loaded MSU-H clearly show that CTZ has been included in the cavities of
the silica. Furthermore, the higher width of the peaks for MSU-H-CTZ-12,
MSU-H-CTZ-18 and MSU-H-CTZ-34 with respect to pure CTZ proves the
amorphous character of CTZ inside MSU-H, which is a positive
characteristic since bioavailability usually increases with the
amorphization.[97,108]
Figure 2.10: 13C (100.65 MHz) CPMAS spectra of pure CTZ, MSU-H-CTZ-12,
MSU-H-CTZ-18 and MSU-H-CTZ-34. The symbols ▲ denote residues of CH2-
CH3 groups from surfactants involved in MSU-H preparation processes.
This agrees with the loss of periodicity associated to the inclusion process
which allows for several possible adsorption ways of CTZ on the silica
walls as predicted by NLDFT methods.[99] Moreover, at higher loading, the
predicted formation of amorphous layers still is consistent with the
observed line widths.
The 1H MAS spectra are characterized by three main signals (see Table

76
2.4 and Figure 2.11): from spectrum a, the CTZ hydrogen atoms (HCTZ) at
6-7 ppm (III); from spectrum f, free SiOH groups (I) at 1.8 ppm in
accordance with the literature;[109] finally, from spectrum b, the SiOH
groups in interaction with H2O (SiOH-H2O) at 4-5 ppm (II). The symbols
▲ denote the impurities derived from the silica synthesis (2-5 ppm).
Indeed, as also showed by the FTIR spectra, in accordance with the
literature,[110] these are -CH2 and -CH3 residues of the involved surfactants.
The small differences in the chemical shifts among all samples are related
to the different environment and crystal packing of the CTZ molecules in
the silica, from isolated sites to layers.
Table 2.4: 1H and 13C selected chemical shifts (δ) with assignments for CTZ,
MSU-H, outgassed MSU-H, MSU-H-CTZ-12, MSU-H-CTZ-18 and MSU-H-CTZ-
34.
13 1
C H
Compound SiOH- Free
C-C-Cl C-N Csp3 CHCTZ
H2O SiOH
CTZ 145.1 119.1 75.3 6.1 - -
MSU-
- - - - 4.0 1.8
Houtgassed
4.2/
MSU-H - - - - -
6.0sh
MSU-H- 120.4-
145.0 75.9 6.4 4.3 -
CTZ-12 119.8
MSU-H-
144.0sh 121.6sh 76.0 7.0 5.2 -
CTZ-18
MSU-H- not not

75.6 6.5 3.8sh -
CTZ-34 detectable detectable
sh
shoulder. bb broad band.
77
Figure 2.11: 1H (400.23 MHz) MAS spectra of pure CTZ (a), MSU-H (scCO2
treatment) (b), MSU-H-CTZ-12 (c), MSU-H-CTZ-18 (d), MSU-H-CTZ-34 (e)
and MSU-H (outgassed) (f).
The most important information can be extracted from the variation of the
relative intensities of the signals. Indeed, while the MSU-H spectrum (b)
is almost completely dominated by the SiOH-H2O signal, the spectra of
MSU-H-CTZ-12, MSU-H-CTZ-18 and MSU-H-CTZ-34 are characterized
by a progressively enhancement of the signal around 6-7 ppm attributed

78
to HCTZ atoms. Furthermore, the SiOH peaks reduce their intensity on

passing from pure MSU-H to MSU-H-CTZ-34, from lowest to highest
loading.
As already reported in table 2.4, from the experimental TG results, in

comparison to the N2 adsorption measurements, the number of CTZ per
theoretical surface area of the silica unit cell (1.606 nm2) can be
evaluated. This value, in table 2.5, has been reported as molecules per
square nanometre. With the same approach, also the H2O content has
been calculated.
Table 2.5: Correlation between experimental TG analysis, N2 isotherms,

theoretical molecule distribution and SSNMR.
TGa CTZ·1.606 nm-2 b CTZc H2Od HCTZ e HSiOH-H2O f SSNMR g
0 0.00 0.00 4.5 0.0 13.5 -
12 0.64 0.40 3.0 7.0 10.5 1.0:1.6
18 1.00 0.62 3.0 10.6 10.5 1.0:1.0
34 2.50 1.56 0.0 26.5 4.5 -

a(% by mass) obtained from TG analysis (Figure S1), b(Number of CTZ molecules
per theoretical surface area of the silica unit cell), c(CTZ molecules per nm -2)
obtained from TG results and NLDFT elaboration of experimental nitrogen
adsorption isotherms, d (H2O molecules per nm -2) obtained from TG results and
NLDFT elaboration, e(HCTZ·nm-2), f(HH2O·nm-2), gSSNMR ratio between the peaks
attributed to HCTZ and HSiOH-H2O. As free SiOH surface concentration the value of
4.5 SiOH·nm-2 has been used, in agreement with the theoretical model surface
concentration.[99]
The obtained value of 3.0 H2O·nm-2 from thermogravimetric analysis,

compared to the chemical shift of 4.2 ppm, is in agreement with the
literature.[111] Indeed, Grünberg et al. reported a value of 4.0 ppm with a
surface coverage of 3.6 H2O·nm-2.
The integral values, obtained from the deconvolution analysis of

79
quantitative 1H spectra, of the CTZ and SiOH-H2O signals at 6.1-7.0 and

3.8-6 ppm, respectively result to be consistent with the previously reported
CTZ-silica ratios per unit cell.[99] Indeed, for instance, the 1H spectrum of
MSU-H-CTZ-12 is characterized by a relative intensity of the HSiOH-H2O and
the HCTZ resonances of 1.6 to 1. This is consistent with the reported data
of 0.64 CTZ·1.606 nm-2, experimentally obtained and theoretically
confirmed. Accordingly, in the MSU-H-CTZ-18 situation the HCTZ and
HSiOH-H2O signal intensities are comparable (1:1). This, again, is coherent
with the experimental and theoretical evidenced mode of interaction. On
the other hand, a slight chemical shift of the MSU-H OH peaks is evident.
This small shift has been attributed to an increase of H-bonds between
SiOH and CTZ, as evidenced by FTIR and theoretical calculations.
Indeed, other SSNMR analysis of guest molecules evidenced this
chemical shift. Of particular interest is the work of Shenderovich et al.[112]
reporting the chemical shift of SiOH in interaction with pyridine (9.9 ppm).
For the same reason, hypothesizing a higher chemical shift, in the case
of MSU-H-CTZ-34, no ratio has been evaluated.
It can be concluded that, during the scCO2 process, the CTZ molecules
slowly replace H2O molecules in the H-bond interaction with SiOH.
Indeed, the number of H-bonds formed remains constant, changing the
guest molecule. The reasons should be a stronger interaction of the SiOH-
CTZ, as evidenced by higher FTIR redshift, theoretical energy of
interaction and experimental energy of desorption.[99] On the other hand,
has been evidenced by theoretical calculation that an interaction between
SiOH and guest molecule is favoured by the H2O presence.[113]
Consequently, the cause remains uncertain and could include the
hydrophobicity nature of CTZ that, during scCO2 process, builds up a
hydrophobic molecular layer.
Since, often, the ability of the controlled release of a host molecule inside
a guest is also related to its mobility, the dynamic of CTZ inside the silica

80
has been probe through relaxation measurements. Thus, 1H T1 relaxation

time measurements were performed on all the samples. Indeed, in the
solid state, the T1 is strongly related to the strength of the dipolar
interaction, which directly depends on the presence of fluctuating
magnetic fields generated by the dynamic behaviors of the whole
molecule or of parts of it.
MOBILITY OF ADSORBED CLOTRIMAZOLE
Ab-Initio Molecular Dynamics (AIMD) has been performed on the different

statically optimized geometries in order to evaluate the stability of the local
minimum structures of figure 2.6. Detailed data are reported in a published
article.[99] AIMD simulation on the molecular layer (II) structure shows high
mobility of the three CTZs (per unit cell) on the surface. The RMSD of the
atomic positions along the AIMD simulation shows a large movement
between 4 and 7 ps that results in a new configuration, with a RMSD value
of 4.3 Å with respect to the starting CTZ conformation after 11 ps. The
process is described in Figure 2.12, where the exposed nitrogen atoms of
the imidazole rings of the three adsorbed molecules in each unit cell are
referred as N1, N2 and N3. The graph in Figure 2.12.a, reporting the N1-
SiOH_III and N1-SiOH_I distances, clearly shows a transition state where
the chain is lost and the imidazole’s nitrogen (N1) is equidistant from
SiOH(I) and (III). Looking at the potential energy fluctuations during this
simulation, we estimate, in a very approximate way, the electronic
activation energy of this process as 29.7 kJ·mol-1 that can be considered
an upper limit for the real value.

81
Figure 2.12: AIMD of the molecular layer (II) structure: a) N1-SiOH(I) and N1-
SiOH(III) bond distances in time. b) starting (A) and final (B) positions of N1 and
N2 with respect to the involved SiOHs. c) Top views of the initial (left) and final
(right) configurations in the AIMD simulation, with the corresponding interaction
energies per CTZ molecule (kJ·mol-1); cell borders in pink.
These AIMD results are in accordance with the experimental finding of a

“liquid like” behavior of ibuprofen adsorbed in MCM-41.[42,43] Considering
that CTZ is much more hydrophobic than ibuprofen it is no surprise that
CTZ molecular layers are very mobile, despite the significant underneath
interactions. This mobility can be represented as a walking step process
guided by local changes in the H-bond interactions, helped by a high
flexibility of the silica surface silanols.
From an experimental point of view, the 1H T1 relaxation time of SSNMR

increases from 0.1 s to 80 s by increasing the CTZ loading from MSU-H-
CTZ-12 to MSU-H-CTZ-34. This can be attributed to a reduction of

82
mobility of CTZ inside the silica on passing from the isolated CTZ
molecules (higher mobility for low loading) to the formation of a layer
(lower mobility for high loading) as previously described.[99] Conversely,
pure CTZ is characterized by a 1H T1 of 90 s, typical of very rigid crystal
packing. Summarizing, the evidenced mobility in the molecular dynamics
simulations obtained by some of use[99] follows frankly the SSNMR
relaxation times. Indeed, in the case of single docking (Imidazole (I) and
(II), phenyls (I) and (II)) high energetics phenomena are observed like lose
and acquire of H-bonds due to CTZ movement or rotation in place. On the
other hand, the molecular layers (I) and (II) show higher cooperation
moving as a layer on the silica surface but not among them.
EXPERIMENTAL AND THE ORETICAL FTIR INTERPRETATION
Experimental and theoretical vibrational spectra of CTZ in molecular,

crystalline and adsorbed environments have been obtained[99]
Figure 2.13 reports both the experimental (blue line) and the theoretical
(red line) IR spectra of CTZ in interaction with silica at low and high
coverage. The low coverage experimental spectrum has been acquired
on a MSU-H-CTZ sample with a drug loading of 18% by mass, while the
high coverage corresponds to a drug loading of 34% by mass. As regards
the simulated spectra, the CTZ modes have been obtained in both cases
from a vibrational analysis in the harmonic approximation of the drug in
the four single molecule configurations, combined by Boltzmann
weighting their contribution to the infrared intensity according to the
computed interaction energies.

83
Figure 2.13: Experimental and simulated IR: experimental FTIR spectrum of

MSU-H-CTZ (blue); theoretical IR spectrum of CTZ combined with SiOH
vibrational contribution (red); simulated IR spectrum of SiOH vibrational
contribution of: imidazole (I) (dot-dashed black line); imidazole (II) (dashed black
line); phenyls (I) (dotted black line); phenyls (II) (black line). At the bottom of the
graph, sticks represent the theoretical SiO-H stretching frequencies without
Pimentel’s correction.

84
For both the low and high loading cases, the agreement between theory
and experiment is remarkable and helps the interpretation of the signals.
A more detailed description can be read elsewhere.[99]
As can be deducted from figure 2.13, increasing the CTZ loading results
in an increasing number of H-bonds between SiOH and CTZ molecules.
Indeed, the 3750 cm-1 sharp band, decrease gradually while the 3500-
2600 cm-1 broad band increase. At the highest loading obtained there are
no free silanols. This shift is attributed to the formation of H-bonds
between SiOH and CTZ through the exposed nitrogen of the imidazole
ring. At the same time, other types of H-bonds are formed with the CTZ
molecules (SiOH-π, SiOH-Cl) that are responsible to the change in the
3700-3500 cm-1 band. Starting from SiOH-SiOH interaction in the bare
MSU-H, they evolve gradually in a mixture of the same H-bonds mixed to
interactions with CTZ (SiOH-π and SiOH-Cl). To demonstrate this, the
broad background at 3700-3500 cm-1 responsible to the increase of the
3620 cm-1 peak, more substantial in MSU-H-CTZ-12, almost disappears
in MSU-H-CTZ-34. These conclusions reveal the direct interaction
between CTZ and mesoporous silica wall until the formation of a CTZ
molecular layer.
2.3 MCM-41 FILLING MODEL
TG analysis and nitrogen adsorption measurement comparison applied to

MCM41-1 do not lead to the same conclusions. While FTIR analysis
presents the same adsorption band, the mesopores of MCM41-1 results
filled of CTZ molecules without homogeneity. Indeed, no empty pore are
visible after the scCO2 process. In chapter 3 a more detailed description
is given, but here are reported preliminary theoretical discussion of a new
85
model of incorporation.
As mentioned above, in the case of MSU-H, the pore diameter (8.5 nm)
allows us to consider the silica surface as flat when CTZ is docked. For
MCM-41 this assumption is not anymore satisfied. Consequently, the
surface model has been modified following the MCM-41 model presented
in the article of Delle Piane et al.[84]
Figure 2.14: Longitudinal view of the MCM-41 model (I). (II) and (III) simulation
of the filled CTZ channel.
The MCM41-CTZ model has been optimized at PBE-D2 level of theory.

The basilar information that can be extracted is that 9, out of 11 CTZ
molecules, are interacting with the silica surface. The last 2 are in a
position similar to the molecular layer (I) and (II). In addition, the number
and energies of interactions are higher. This discussion reveals in a

86
simplified way the motivations of the MCM41-1 filling during the scCO2
incorporation process.

87
CHAPTER 3: SILICAS DETERMING FACTORS IN THE

INCORPORATION PROCESS
ABSTRACT
The knowledge of the specific interactions between the surface of

mesoporous silica and drugs took to comprehend the different type of
adsorption geometries depending on the pore size dimension. However,
such knowledge is still preliminary.
In this chapter a complete experimental description on these factor have

been argued. Fourteen mesoporous silica have been incorporated trough
scCO2 process and completely characterized in order to understand the
experimental discriminant factors. In the first part, a characterization of all
OMS as such is reported; after that, the different behaviors are described;
finally, some argumentation are stated.
3.1 OMS CHARACTERISTICS
The 14 mesoporous (Table 3.1) silicas were selected for different

characteristics and considerations. One mesoporous silica is, to be
correct, a disordered microporous silica, Syloid. Four MCM-41, were
synthesized to screen low pore diameter silicas. One commercial MCM-
41 was necessary to avoid laboratory behaviors. Four KIT-6 were
manufactured to control if Ia3d silicas have the same behavior. Two
commercial and one synthetic SBA-15 were used to point out the conduct
with high pore dimension. The MSU-H was reprocessed to control the
supercritical process and the reproducibility of the results.
Table 3.1: NLDFT parameter of all the OMS used. They have been ordered
according to the mean pore diameter that is not ever the maximum of the PSD.

88
Mean pore NLDFT SSA NLDFT Volume

OMS
diameter (nm) (m2g-1) (cm3g-1)
Syloid 1.37 1157 0.378
MCM41-4 2.94 840 0.598
MCM41-3 3.65 930 0.818
MCM41-ACS 4.57 843 0.901
MCM41-1 4.88 812 1.286
MCM41-2 4.88 585 1.100
KIT6-50 5.28 678 0.560
KIT6-80 6.55 603 0.690
SBA15 6.79 763 0.697
KIT6-100 7.86 788 1.157
MSU-H-F 8.14 643 0.836
SBA15-ACS 8.14 424 0.884
MSU-H 8.46 596 0.900
KIT6-130 10.13 525 1.300
From table 3.1 and figure 3.1.I it can be evidenced that the OMS are
mainly divided in two categories: silicas with mean pore diameter lower
than 4.88 nm and silicas higher than this. Indeed, two main straight line
can be verified from the graph 3.1.I. This categorization is due to the wall
thickness. It is well clear that MCM-41 and SBA-15 silicas have this
differences. It is strange that also Syloid and KIT-6 respects this condition.

89
Figure 3.1: Distribution of OMS in function of mean pore diameter and pore
volume (I).
All the OMSs have been characterized by TG, XRD, FESEM analysis. No
important result for this correlation have been noted.
3.2 CHARACTERIZATION OF INCORPORATED OMS
All the OMS have been incorporated through scCO2 at 250 bar and 373K
for 18 hours. No further incorporation was possible increasing the total
time. All the characterization analysis already described were repeated
on the incorporated materials. Absolute data are, first of all, the maximum
incorporated quantity: 46% by mass obtained with MCM41-1. Secondly,
in no circumstances crystalline CTZ was observed. Thirdly, all the N2
adsorption measurements reported a decrease of SSA and pore volume.
In some cases, also the mean pore diameter was decreased.
90
Figure 3.2 reports some of the nitrogen adsorption measurements done.

It can be observed that for Syloid and MCM-41, after the supercritical
process, there is almost no residue porosity (no step or hysteresis) and
very lower surface area (lower inclination). On the other hand, MSU-H and
KIT6-130 still have large empty porosity. A border case is the KIT6-50.
According to chapter 2, it can be argued that the first OMSs are filled of
CTZ as the theoretical modelization of figure 2.14, while in the second
group CTZ is adsorbed as a molecular layer, figure 2.8. Trying to
understand the limiting pore size dimension, a further analysis has been
done. Figure 3.3 reports the incorporated quantity in function of the
NLDFT pore volume. In this case the incorporated quantity is calculated
as the CTZ content on the silica content.

91
Figure 3.2: Nitrogen adsorption measurements before (OMS) and after the
scCO2 process (OMS-CTZ).
This correlation evidences which OMSs follows the filling model. On the
other hand, an R2 of 0.93 is not a good report; but, considering the error
bars, all the data are consistent. The linear equation has been selected
without constant term thinking to the fact that the mesopore could be filled
or empty. On the contrary, in no cases the external SSA was taken into
account. Indeed, this correlation do not consider the external surface
coverage that it is hardly estimable.

92
Figure 3.3: OMS correlations: Incorporated Quantity (gCTZ/gsil) versus NLDFT

Pore Volume (cm3/gsil). The red dot reports a border linearity.
At the same time, the correlation of NLDFT SSA with incorporated

quantity has been done and reported in figure 3.4.
Figure 3.4: OMS correlations: Incorporated Quantity (gCTZ/gsil) versus NLDFT

SSA (m2/gsil).
In this case R2 is 0.95 and the slope (m) of the linear equation is directly
correlated to the number of CTZ·nm-2. The results, reported to the surface
area of the silica theoretical cell (chapter 2), is 1.7 CTZ·cell-1. This result

93
is lower than the presented one of chapter 2 but also the scCO2 conditions
are lower (250 bar).
As it is possible to observe, not all the silica follows perfectly these

relationships. For instance, KIT6-50 and KIT6-80 follow both correlations.
Consequently, in order to correctly distinguish all cases, the PSD
simulations have been applied for all the OMSs, taking into account the
number of CTZ extrapolated from the linear correlation. If it works, it can
be defined as a molecular layer; on the contrary, it could follow the filling
correlation.
The results, reported in figures 3.5 – 3.6 – 3.7, shows three conditions:
 Figure 3.5: OMSs perfectly filled, without residual porosity.

 Figure 3.7: OMSs with a molecular layer of about 1.7 CTZ/cell;
 Figure 3.6: a condition in between.
Therefore, it seems reasonable conclude that from a mean pore radius of

5.2 to 6.8 nm the behavior is mixed. Probably, also the Ia3d structure of
KIT-6 silicas is important. Indeed, the SBA15, which has the bigger mean
pore diameter, seems to follows better the filling model, but it suffers more
than the others of low diffusivity (2D pore structure versus 3D.)

94
Figure 3.5: PSD of OMS as such (black), OMC-CTZ (blue) and the calculated
PSD (red). In brackets the number of CTZ per unit cell used for the model.
These are the silicas that follows a filling theory.

95
These are the silicas border line.

96
These are the silicas that follows the molecular layer theory.
Table 3.2: OMS correlations resume.
OMS Mean Pore Diameter Correlation

Syloid 1.37 Pore Volume
MCM41-4 2.94 Pore Volume
MCM41-ACS 4.57 Pore Volume
KIT6-50 5.28 Both
KIT6-80 6.55 Both
SBA15 6.79 Both
KIT6-100 7.86 SSA
MSU-H-F 8.14 SSA
SBA15-ACS 8.14 SSA
MSU-H 8.46 SSA
KIT6-130 10.13 SSA

97
In conclusion, one more correlation need to be reported. From the NLDFT

pore volume data and the TG analysis the density of CTZ inside the
mesoporous volume can be evaluated. The next equation resumes this
correlation:
𝑔𝐶𝑇𝑍 𝑐𝑚3 𝑐𝑚3 𝑔𝑠𝑖𝑙+𝐶𝑇𝑍

𝐼𝑛𝑐𝑜𝑟𝑝𝑜𝑟𝑎𝑡𝑒𝑑 𝑄𝑢𝑎𝑛𝑡𝑖𝑡𝑦 [ ] = 𝑉′ [ ] − {𝑉′′ [ ]∙( )}
𝑔𝑠𝑖𝑙 𝑔𝑠𝑖𝑙 𝑔𝑠𝑖𝑙+𝐶𝑇𝑍 𝑔𝑠𝑖𝑙
where V’ is the OMS NLDFT pore volume as such and V’’ after the scCO2
𝑔𝑠𝑖𝑙+𝐶𝑇𝑍
process. The incorporated quantity and the ratio ( 𝑔𝑠𝑖𝑙
) are obtained
through TG analysis.
Figure 3.8: Incorporated quantity (gCTZ·gsil-1) versus NLDFT ΔV (cm3·gsil-1): the

density correlation.
The straight line has a slope coefficient (m) of 1.14 gCTZ·cm-3. This result
matches perfectly with all other correlations (chapter 2). In addition to this,
it can be said that all the OMSs that are not on the straight line suffers of
pore occlusion. None of these have large amount of CTZ outside the
mesoporous volume.

98
CHAPTER 4: AKS-OMS GEL
ABSTRACT
A possible and promising area of application for Ordered Mesoporous

Silica (OMS) is the topical therapy of dermatological diseases and
wounds. It is widely known that Active Pharmaceutical Ingredient (API)
incorporated in OMS based materials are adsorbed in the volume in an
amorphous state and in some cases well distributed on the surface as a
single molecular layer. This feature is crucial for many aspects: from the
bioavailability enhancement of molecule poorly soluble in water to the
effects of reservoir that can be tuned in cream for topical application.
Indeed, the use of OMS incorporated of API in a saturated vehicle to the
therapeutic concentration develops a controlled and constant release
system on the skin site.
Figure 4.1: Ideal description of AKS-OMS gel.
This chapter describes the development of topical composition able to

release drug in a sustained and prolonged manner. Incipient Wetness
Impregnation procedure has been used to load AKS in commercial OMS
of MSU-H type. The incorporated OMS characterization (XRD, FESEM,
etc.) are reported in chapter 1. A series of different release test has been
performed to validate the new release technology (UV-Vis spectroscopy).
As a result, a semisolid formulation with AKS has been developed. The

99
stability, pH and rheological properties were investigated overtime. In vitro

release and permeation studies were assessed using an open chamber
diffusion cell system (Franz cell), fitted with semi-permeable membrane
or porcine ear skin. All the results emphasize the reservoir effect of OMS
incorporated of API. In addition, they demonstrate the availability of a new
topical release technology able to reduce the number of administrations.
4.1 SUSTAINED RELEASE FR OM OMS-AKS IN AKS SATURATED

SOLUTIONS
In order to confirm the reservoir effect of the incorporated OMS in a

saturated solution composed by two solvents, AKS, a hydrophilic API, has
been studied. Commercially, the more used topical creams which
contains AKS have a composition at 5% by mass of API, that corresponds
[114,115]
to 6,67% by mass of AKS . Therefore, in order to obtain a solvent
with a saturation point toward AKS equal to the therapeutic concentration,
a solution of glycerol and water has been explored. The best composition
is 45% water and 55% glycerol by mass. The saturation point of this
solution is 71 mg·ml-1 that correspond to 62 mg·g-1 of AKS. Consequently,
the API content is 4.8% by mass. In this case two different AKS saturated
solution has been done. In a solution, 80 mg of OMS-AKS (27.5 % by
mass) has been added under agitation. Every hour, for 4 times, 100 μl
has been withdraw and substituted with the same amount of pure solvent.
Figure 4.2 reports the measured concentration in the four withdraw for
each sample. The saturated concentration is preserved as long as there
is drug that can be solubilized from the OMS. In fact, 22 mg of drug where
presents as reservoir and in the three subsequent dilution 21.3 mg where
withdraw: 100 μl contains 7.1 mg of API.

100
Figure 4.2: AKS sustained release with 21.3 mg of API contained in OMS-AKS
(a) and without reservoir (b).
4.2 COMPOSITION, RHEOLOGICAL AND pH STABILITY OF AKS-OMS

GEL
As previously described an innovative topical semisolid formulation has

been prepared and compared to a commercial formulation. All the results
here presented are referred to the gel named AKS-OMS gel. Among all
the semisolid preparations, a hydrophilic gel was developed, consisting of
a liquid phase within a polymeric matrix composed by a suitable gelling
agent. Firstly, AKS was added in a solution of glycerol (45% by mass) and
water (55% by mass) under magnetic stirring obtaining a saturated
solution of the drug. When AKS was completely dissolved, 15% by mass
OMS-AKS powder was added to the medium and then the liquid
dispersion was purposely homogenized with a high shear homogenizer
for 5 min at maximum velocity. Hydroxyethyl cellulose was selected as
gelling agent, because it is primarily used in topical pharmaceutical

101
formulations and it is generally considered as an essentially non-toxic and

[116]
non-irritant material . With this purpose, 2% by mass of hydroxyethyl
cellulose was added to the homogeneous dispersion and it was gently
stirred for one hour at room temperature.
The gel shows rheological characteristic of shear-thinning fluid. Over one

month the viscosity at low shear rate is halved but preserve is
pseudoplastic behaviour. The Farrow's constant was 3.4 n and remain the
same over the month, going out of linearity only for low shear rate (3.9 n).
During the same period, the pH oscillates from 4.75 to 5.0 without the
addition of any correctors. It should be noted that this pH is equal to the
pH of the commercial formulation.
4.3 IN-VITRO RELEASE STU DIES
Figure 4.3 reports the release curves obtained. In the first 100 minutes,
AKS-OMS gel follows the commercial formulation. After, the commercial
formulation release rate decrease since the AKS dissolved quantity is
ending. In that moment, the OMS-AKS reservoir starts to make a
difference in the AKS-OMS gel. Indeed, the release rate of AKS-OMS gel
remains constant until 200-250 minutes. Finally, also the OMS-AKS
reservoir end and the release rate decrease.
The constant release rate in the first 100 minutes proves that the starting
dissolved concentration of AKS inside AKS-OMS gel and commercial
formulation is the same. The preservation of a constant release rate after
100 minutes of AKS-OMS gel proves that the dissolved concentration of
AKS is sustained to the therapeutics concentration until 200-250 minutes.
These demonstrate the feasibility of this application as a method to reduce
the number of applications during the day.

102
Figure 4.3: In-vitro release test of AKS-OMS gel (a) and the commercial
formulation (b). Each point is the mean value of three different release tests. The
error bars include the maximum variation obtained from all the different trials.
4.4 IN-VITRO PERMEATION STUDIES
In order to evaluate differences in permeation of the two semisolid

formulation on a skin site, a slice of porcine ear skin has been mounted
on the Franz cells (Figure 4.4). Until 100 minutes, as the release test, the
permeated quantities are small and similar. After, the AKS permeated
from AKS-OMS gel exceed the commercial formulation quantity. This is
in agreement with the release studies that show the depletion of the
commercial formulation after 100 minutes. Consequently, the permeation
of AKS is influenced from the depletion of the commercial formulation. On
the other hand, the constant concentration present in the AKS-OMS gel
develop an increase in the permeated quantity. The complete permeation
of AKS from the commercial formulation requires 24 h, while for AKS-
103
OMS gel it exceeds 48 h. After 4 days, no more AKS permeation is

observed.
Figure 4.4: In-vitro permeation test of AKS-OMS gel (a) and the commercial
formulation (b). On the left (I) the first 350 minutes; on the right (II) the

104
subsequent 168 h. Each point is the mean value of three different release tests.
The error bars include the maximum variation obtained from all the different
trials.
These results show that a constant concentration over the skin site for a
prolonged time extend and enhance the permeation of API. The first result
was intended. On the other hand, the enhancement could produce an
overtreatment, solvable reducing the saturation point of the vehicle. This
solution leads to more improvement of the new semisolid formulation.
Indeed, reducing the dissolved starting therapeutic concentration will
increase directly the duration of the stored AKS. Secondly, the same
antibiotic effect could be achieved with less dosage on the application site.
4.5 TOPOLOGICAL INFORMATION ON OMS PERMEATIO N
Numerous works study the possible permeation of OMS through the skin
[117–119]
site of application . To the best of our knowledge, none of these
papers reports a direct permeation of particles greater than 10 nm. On the
other hand, it is well labelled the possibility for these particles to enter skin
[117]
pores and hair follicles, enhancing the drug release . In order to
observe the OMS-Skin interactions, during the permeation studies,
different experiments were stopped at 6, 24, 48 and 72 h (Figure 4.5).
Between 6 to 48 h, the OMS particles are in aggregate on the skin. These
aggregates are formed during samples preparation for FESEM analysis.
However, after 72h, OMS are dispersed in the first 200 μm of thickness
and are hardly to be distinguished. Therefore, OMS-Skin interaction is
weakly and a long period is required to OMS particles to stick to skin pores
and hair follicles.

105
Figure 4.5: FESEM image of the In-vitro permeation test of AKS-OMS gel: 6,
24, 48 and 72h. OMS particles in the form of aggregates are observable very
bright in the upper left part of 6h (a), bright in the upper right of 24h (b), hardly
visible in the upper part of 48h (c). Image (d) reports the OMS dispersion on the
skin after 72h.

106
CHAPTER 5: CTZ-OMS GEL
ABSTRACT
Controlled drug delivery from Ordered Mesoporous Silica (OMS)

platforms represents the possibility to achieve sustained or prolonged
drug release in different administration routes widely used, from the oral
[93]
to the parenteral one . Nevertheless, exploring the literature it is quite
evident that much research work have been addressed to investigate
OMS nanoparticles as drug delivery system for the oral route, in particular
for poorly-water soluble molecules, or for parenteral systems releasing
active pharmaceutical ingredients after a triggered external stimuli
[10,30,97,120]
. To the best of our knowledge, only few papers have been
published about the possibility to use silica-based mesoporous
[121–126]
nanoparticles for topical application . Skin is generally used as
route of delivery for local and systemic drugs.
This chapter presents the development and application of the new

controlled release system with CTZ. The complete physicochemical
characterization of the OMS incorporated with API has been totally
described in the previous chapter. Subsequently, in this section are
reported only the release tests of different saturated vehicle to ensure the
effectiveness of the controlled system. Afterwards, the preparation and
characterization of an innovative topical semisolid formulation for CTZ
comparable to commercial formulations. Finally, sustained release of CTZ
has been developed and outlined with vertical glass diffusion cells.

107
5.1 SUSTAINED RELEASE OF OMS-CTZ IN CTZ SATURATED

SOLUTIONS
In parallel with the sustained release test of AKS, also CTZ has been
tested. For CTZ the typical topical creams are between 1 and 2,5% by
mass of API content [65]. Therefore, a solution of 1,2-propandiol and water
has been studied. The best conditions were with 85% 1,2-propandiol and
15% water by mass. This solution has a saturation point of 25 mg·ml-1 that
corresponds to 2,4% by mass of API. The same procedure used for AKS
has been followed and in figure 5.1 are reported the measured
concentration. The concentration decrease only after the complete
solubilisation of the incorporated API. Each withdraw of 100 μl contains
2.5 mg of API and the reservoir of OMS-CTZ was 7 mg (70 mg of OMS-
CTZ at 10%)

108
Figure 5.1: CTZ sustained release with 7 mg of API contained in OMS-CTZ (a)
and without reservoir (b).
5.2 COMPOSITION AND IN-VITRO RELEASE TES T
A topical semisolid formulation has been prepared and compared to a

commercial formulation. In this case the CTZ gel, named CTZ-OMS gel,
is a hydrophobic semisolid preparation. It consists of 85% 1,2-propandiol
and 15% of bidistilled water. When CTZ is dissolved, 10% by mass OMS-
CTZ powder was added to the medium. In this case MSU-H-CTZ-34 has
been used. Then, the liquid dispersion was purposely homogenized with
a high shear homogenizer for 5 min at maximum velocity. As gelling agent
hydroxyethyl cellulose was selected as gelling agent, 2% by mass. The
gel shows rheological characteristic of shear-thinning fluid.
Figure 5.2: In-vitro release test of CTZ-OMS gel (a) and the commercial

109
formulation (b). Each point is the mean value of three different release tests. The
error bars include the maximum variation obtained from all the different trials.
Figure 5.2 reports the in-vitro release curves obtained. In this case, for the
first 2h, CTZ-OMS gel follows the commercial formulation. The
commercial formulation release rate bends and stops around 5h from the
start. On the other hand, CTZ-OMS gel continues almost linearly until 24h.
The constant release rate in the first 2 hours verifies that the starting free
concentration of CTZ inside CTZ-OMS gel and the commercial
formulation is the same. The preservation of a constant release rate until
5-6 hours proves the capability for CTZ-OMS gel of a sustained release
to the therapeutics concentration. Unfortunately, the release rate seems
to decrease gradually and not uniformly like the commercial formulation.
Probably more tests should be done from 7 to 24 hours in order to control
better the release profile of CTZ-OMS gel.

110
CONCLUSIONS
In this PhD work Ordered Mesoporous Silica has been studied and used
for a new controlled release technology for topical applications: starting
from the physicochemical characterization to the development of real
formulations.
In the first chapter, has been shown the positive and negatives aspects of
different incorporation techniques based on hydrophilicity/hydrophobicity
of Clotrimazole and Amikacin Sulfate. This have been correlated with
dissimilar type of OMS.
With Clotrimazole a theoretical-experimental comparison between results

has been done tackling phenomena like: mobility, solubility,
bioavailability, etc. All the theoretical and experimental results were in line
each other.
Strong of this correlation the differences between types of OMS have

been highlighted trying to explain the spatial assembly of drug inside the
mesoporous channels of OMS in function of pore dimensions.
Finally, the new controlled release technology has been discussed and
demonstrated in its functionality for Clotrimazole and Amikacin Sulfate.
Both systems have demonstrated double release time without leaving the
therapeutic concentration.

111
BIBLIOGRAPHY
[1] K. K. Jain, Ed. , Drug Delivery System, Springer New York, New
York, NY, 2014.
[2] Y. Perrie, T. Rades, Pharmaceutics: Drug Delivery and Targeting,
Pharmaceutical Press, 2012.
[3] N. S. Downing, J. A. Aminawung, N. D. Shah, J. B. Braunstein, H.
M. Krumholz, J. S. Ross, N. Engl. J. Med. 2012, 366, 2284.
[4] M. R. Prausnitz, R. Langer, Nat. Biotechnol. 2008, 26, 1261.
[5] G. Tiwari, R. Tiwari, B. Sriwastawa, L. Bhati, S. Pandey, P.
Pandey, S. K. Bannerjee, Int. J. Pharm. Investig. 2012, 2, 2.
[6] Y. Ozsoy, S. Gungor, E. Cevher, Molecules 2009, 14, 3754.
[7] M. Vallet-Regí, F. Balas, D. Arcos, Angew. Chem. Int. Ed. Engl.
2007, 46, 7548.
[8] A. Gignone, L. Manna, S. Ronchetti, M. Banchero, B. Onida,
Micropor. Mesopor. Mat. 2014, 200, 291.
[9] J. W. Rasmussen, E. Martinez, P. Louka, D. G. Wingett, Expert
Opin. Drug Deliv. 2010, 7, 1063.
[10] R. J. Ahern, J. P. Hanrahan, J. M. Tobin, K. B. Ryan, A. M. Crean,
Eur. J. Pharm. Sci. 2013, 50, 400.
[11] B. A. Lipsky, C. Hoey, Clin. Infect. Dis. 2009, 49, 1541.
[12] S. C. Gad, Pharmaceutical Manufacturing Handbook: Production
and Processes, Wiley, 2008.
[13] F. Hashim, M. El-Ridy, M. Nasr, Y. Abdallah, Drug Deliv. 2010,
17, 282.
[14] A. H. Faraji, P. Wipf, Bioorg. Med. Chem. 2009, 17, 2950.
[15] C. Koike, R. Noguchi, H. Chihara, H. Suto, O. Ohtaka, Y. Imai, T.
Matsumoto, A. Tsuchiyama, Astrophys. J. 2013, 778, 60.
[16] V.-P. Lehto, T. Heikkilä, J. Riikonen, E. Mäkilä, J. Salonen, J.
Hirvonen, T. Laaksonen, L. Peltonen, H. Santos, n.d.
[17] Q. Cai, Z.-S. Luo, W.-Q. Pang, Y.-W. Fan, X.-H. Chen, F.-Z. Cui,
Chem. Mater. 2001, 13, 258.
[18] W. Li, D. Zhao, Chem. Commun. (Camb). 2013, 49, 943.
[19] S. S. Kim, A. Karkamkar, T. J. Pinnavaia, M. Kruk, M. Jaroniec, J.
Phys. Chem. B 2001, 105, 7663.

112
[20] M. Grün, K. K. Unger, A. Matsumoto, K. Tsutsumi, Microporous

Mesoporous Mater. 1999, 27, 207.
[21] F. Kleitz, M. Thommes, V. La, 2010, 19, 9344.
[22] M. Thommes, R. Köhn, M. Fröba, M. Fro, M. Thommes, R. Ko,
Appl. Surf. Sci. 2002, 196, 239.
[23] D. Zhao, J. Feng, Q. Huo, N. Melosh, G. H. Fredrickson, B. F.
Chmelka, G. D. Stucky, Science (80-. ). 1998, 279, 548.
[24] L. Cao, M. Kruk, J. Colloid Interface Sci. 2011, 361, 472.
[25] T. Seki, T. Ikariya, Phys. Chem. Chem. Phys. 2009, 11, 10073.
[26] L. F. Giraldo, B. L. López, L. Pérez, S. Urrego, L. Sierra, M. Mesa,
Macromol. Symp. 2007, 258, 129.
[27] I. I. Slowing, J. L. Vivero-Escoto, C.-W. Wu, V. S.-Y. Lin, Adv.
Drug Deliv. Rev. 2008, 60, 1278.
[28] M. Vallet-Regí, Chem.- Eur. J. 2006, 12, 5934.
[29] I. Izquierdo-Barba, A. Martinez, A. L. Doadrio, J. Pérez-Pariente,
M. Vallet-Regí, Eur. J. Pharm. Sci. 2005, 26, 365.
[30] M. Manzano, V. Aina, C. O. Areán, F. Balas, V. Cauda, M. Colilla,
M. R. Delgado, M. Vallet-Regí, Chem. Eng. J. 2008, 137, 30.
[31] C.-Y. Lai, B. G. Trewyn, D. M. Jeftinija, K. Jeftinija, S. Xu, S.
Jeftinija, V. S.-Y. Lin, J. Am. Chem. Soc. 2003, 125, 4451.
[32] A. Yamamoto, T. Iseki, M. Ochi-Sugiyama, N. Okada, T. Fujita, S.
Muranishi, J. Control. Release 2001, 76, 363.
[33] M. Vallet-Regì, A. Rámila, R. P. Del Real, J. Pérez-Pariente,
Chem. Mater. 2001, 13, 308.
[34] F. Balas, M. Manzano, P. Horcajada, M. Vallet-Regí, J. Am.
Chem. Soc. 2006, 128, 8116.
[35] J. C. Doadrio, E. M. B. Sousa, I. Izquierdo-Barba, A. L. Doadrio,
J. Perez-Pariente, M. Vallet-Regí, J. Mater. Chem. 2006, 16, 462.
[36] S. Giri, B. G. Trewyn, M. P. Stellmaker, V. S.-Y. Lin, Angew.
Chem. Int. Ed. Engl. 2005, 44, 5038.
[37] F. Torney, B. G. Trewyn, V. S.-Y. Lin, K. Wang, Nat. Nanotechnol.
2007, 2, 295.
[38] N. K. Mal, M. Fujiwara, Y. Tanaka, Nature 2003, 421, 350.
[39] J. Lu, M. Liong, J. I. Zink, F. Tamanoi, Small 2007, 3, 1341.
[40] R. Mellaerts, C. a Aerts, J. Van Humbeeck, P. Augustijns, G. Van
113
den Mooter, J. a Martens, Chem. Commun. 2007, 1375.

[41] V. Ambrogi, F. Marmottini, C. Pagano, Micropor. Mesopor. Mat.
2013, 177, 1.
[42] T. Azais, C. Tourne-Peteilh, F. Aussenac, N. Baccile, C. Coelho,
J. M. Devoisselle, F. Babonneau, M. Curie, L. Chimie, D. Matie,
B. Biospin, T. Azais, C. Tourne-Peteilh, F. Aussenac, N. Baccile,
C. Coelho, J. M. Devoisselle, F. Babonneau, Chem. Mater. 2006,
18, 6382.
[43] F. Guenneau, K. Panesar, A. Nossov, M.-A. Springuel-Huet, T.
Azaïs, F. Babonneau, C. Tourné-Péteilh, J.-M. Devoisselle, A.
Gédéon, Phys. Chem. Chem. Phys. 2013, 15, 18805.
[44] J. M. Rosenholm, M. Lindén, J. Control. Release 2008, 128, 157.
[45] H. A. Santos, E. Mäkilä, A. J. Airaksinen, L. M. Bimbo, J.
Hirvonen, Nanomedicine (Lond). 2014, 9, 535.
[46] J. Salonen, A. M. Kaukonen, J. Hirvonen, V.-P. Lehto, J. Pharm.
Sci. 2008, 97, 632.
[47] J. Haber, J. H. Block, B. Delmon, Pure Appl. Chem. 1995, 67,
1257.
[48] R. Mellaerts, J. a G. Jammaer, M. Van Speybroeck, C. Hong, J.
Van Humbeeck, P. Augustijns, G. Van Den Mooter, J. a. Martens,
Langmuir 2008, 24, 8651.
[49] M. E. M. Braga, M. T. V. Pato, H. S. R. C. Silva, E. I. Ferreira, M.
H. Gil, C. M. M. Duarte, H. C. de Sousa, J. Supercrit. Fluids 2008,
44, 245.
[50] R. Span, W. Wagner, J. Phys. Chem. Ref. Data 1996, 25, 1509.
[51] P. York, Pharm. Sci. Technolo. Today 1999, 2, 430.
[52] J. Fages, H. Lochard, J.-J. Letourneau, M. Sauceau, E. Rodier,
Powder Technol. 2004, 141, 219.
[53] I. Pasquali, R. Bettini, Int. J. Pharm. 2008, 364, 176.
[54] N. L. Rozzi, R. K. Singh, Compr. Rev. Food Sci. Food Saf. 2002,
1, 33.
[55] J. L. Silva, A. C. Oliveira, T. C. R. G. Vieira, G. A. P. de Oliveira,
M. C. Suarez, D. Foguel, Chem. Rev. 2014, 114, 7239.
[56] M. D. A. Saldaña, R. S. Mohamed, M. G. Baer, P. Mazzafera, J.
Agric. Food Chem. 1999, 47, 3804.
[57] I. Smirnova, J. Mamic, W. Arlt, 2003, 8521.

114
[58] V. KUMAR, N. P. SUH, Polym. Eng. Sci. n.d., 30, 1323.

[59] M. Bhamidipati, A. M. Scurto, M. S. Detamore, Tissue Eng. Part
B. Rev. 2013, 19, 221.
[60] K. S. Paudel, M. Milewski, C. L. Swadley, N. K. Brogden, P.
Ghosh, A. L. Stinchcomb, Ther. Deliv. 2010, 1, 109.
[61] P. Gupta, S. Garg, Pharm. Technol. 2002.
[62] W. Montagna, The Structure and Function of Skin 3E, Elsevier,
2012.
[63] W. Arnould-Taylor, A Textbook of Anatomy and Physiology,
Nelson Thornes, 1998.
[64] R. J. Scheuplein, I. H. Blank, Physiol. Rev. 1971, 51, 702.
[65] P. R. Sawyer, R. N. Brogden, R. M. Pinder, T. M. Speight, Avery,
Drugs 1975, 9, 424.
[66] H. Song, H.-S. Shin, Acta Crystallogr. Sect. C Cryst. Struct.
Commun. 1998, 54, 1675.
[67] P. M. Monteleone, N. Muhammad, R. D. Brown, J. P. McGrory, S.
A. Hanna, in Anal. Profiles Drug Subst. (Ed: K. Florey), Academic
Press, 1983, pp. 37–71.
[68] S. K. Jana, R. Nishida, K. Shindo, T. Kugita, S. Namba,
Microporous Mesoporous Mater. 2004, 68, 133.
[69] F. Kleitz, S. H. Choi, R. Ryoo, Chem. Commun. (Camb). 2003,
2136.
[70] S. P. Verevkin, R. V. Ralys, D. H. Zaitsau, V. N. Emel’yanenko, C.
Schick, Thermochim. Acta 2012, 538, 55.
[71] D. M. Price, Thermochim. Acta 2001, 367, 253.
[72] J. S. Chickos, S. Hosseini, D. G. Hesse, J. F. Liebman, Struct.
Chem. 1993, 4, 271.
[73] O. N. Antzutkin, S. C. Shekar, M. H. Levitt, J. Magn. Reson. Ser.
A 1995, 115, 7.
[74] O. N. Antzutkin, Y. K. Lee, M. H. Levitt, J. Magn. Reson. 1998,
135, 144.
[75] M. Banchero, L. Manna, S. Ronchetti, P. Campanelli, A. Ferri, J.
Supercrit. Fluids 2009, 49, 271.
[76] R. Dovesi, R. Orlando, A. Erba, C. M. Zicovich-Wilson, B.
Civalleri, S. Casassa, L. Maschio, M. Ferrabone, M. De La Pierre,
P. D’Arco, Y. Noël, M. Causà, M. Rérat, B. Kirtman, Int. J.
115
Quantum Chem. 2014, 114, 1287.

[77] R. Orlando, M. Delle Piane, I. J. Bush, P. Ugliengo, M. Ferrabone,
R. Dovesi, J. Comput. Chem. 2012, 33, 2276.
[78] M. Delle Piane, M. Corno, P. Ugliengo, J. Chem. Theory Comput.
2013, 9, 2404.
[79] J. P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 1996, 77,
3865.
[80] S. Grimme, J. Comput. Chem. 2006, 27, 1787.
[81] A. Schafer, H. Horn, R. Ahlrichs, J. Chem. Phys. 1992, 97, 2571.
[82] R. Nada, J. B. Nicholas, M. I. McCarthy, A. C. Hess, Int. J.
Quantum Chem. 1996, 60, 809.
[83] E. Apra, M. Causa, M. Prencipe, R. Dovesi, V. R. Saunders, J.
Phys. Condens. Matter 1993, 5, 2969.
[84] M. Delle Piane, M. Corno, A. Pedone, R. Dovesi, P. Ugliengo, J.
Phys. Chem. C 2014, 118, 26737.
[85] S. Dall’Olio, R. Dovesi, R. Resta, Phys. Rev. B 1997, 56, 10105.
[86] J. Hutter, M. Iannuzzi, F. Schiffmann, J. VandeVondele, Wiley
Interdiscip. Rev. Comput. Mol. Sci. 2014, 4, 15.
[87] J. VandeVondele, M. Krack, F. Mohamed, M. Parrinello, T.
Chassaing, J. Hutter, Comput. Phys. Commun. 2005, 167, 103.
[88] S. Goedecker, M. Teter, J. Hutter, Phys. Rev. B 1996, 54, 1703.
[89] J. VandeVondele, J. Hutter, J. Chem. Phys. 2007, 127, 114105.
[90] G. Bussi, D. Donadio, M. Parrinello, J. Chem. Phys. 2007, 126,
014101.
[91] A. Herskowitz, N. Ishii, H. Schaumburg, N. Engl. J. Med. 1971,
285, 82.
[92] P. S. S. Spencer, H. H. H. Schaumburg, Proc. R. Soc. Med. 1977,
70, 37.
[93] C. Charnay, S. Bégu, C. Tourné-Péteilh, L. Nicole, D. a. Lerner, J.
M. Devoisselle, Eur. J. Pharm. Biopharm. 2004, 57, 533.
[94] F. Belhadj-Ahmed, E. Badens, P. Llewellyn, R. Denoyel, G.
Charbit, J. Supercrit. Fluids 2009, 51, 278.
[95] M. Ni, Q.-Q. Xu, J.-Z. Yin, J. Mater. Res. 2012, 27, 2902.
[96] J. G. Hoogerheide, B. E. Wyka, in Anal. Profiles Drug Subst.,
1982, pp. 225–255.
116
[97] R. Mellaerts, R. Mols, J. a G. Jammaer, C. A. Aerts, P. Annaert, J.

Van Humbeeck, G. Van den Mooter, P. Augustijns, J. A. Martens,
Eur. J. Pharm. Biopharm. 2008, 69, 223.
[98] V. Ambrogi, L. Perioli, F. Marmottini, O. Accorsi, C. Pagano, M.
Ricci, C. Rossi, Micropor. Mesopor. Mat. 2008, 113, 445.
[99] A. Gignone, M. Delle Piane, M. Corno, P. Ugliengo, B. Onida, J.
Phys. Chem. C 2015, 119, 13068.
[100] K. K. Qian, R. H. Bogner, J. Pharm. Sci. 2012, 101, 444.
[101] R. Bau, I. Tsyba, Tetrahedron 1999, 55, 14839.
[102] M. Impéror-Clerc, P. Davidson, A. Davidson, J. Am. Chem. Soc.
2000, 122, 11925.
[103] M. Sliwinska-Bartkowiak, G. Dudziak, R. Gras, R. Sikorski, R.
Radhakrishnan, K. E. Gubbins, Colloids Surfaces A Physicochem.
Eng. Asp. 2001, 187-188, 523.
[104] K. K. Qian, W. Zhou, X. M. Xu, T. J. Udovic, Pharm. Res. 2012,
29, 2432.
[105] P. Ugliengo, M. Sodupe, F. Musso, I. J. Bush, R. Orlando, R.
Dovesi, Adv. Mater. 2008, 20, 4579.
[106] L. T. Zhuravlev, Langmuir 1987, 3, 316.
[107] M. L. Connolly, J. Am. Chem. Soc. 1985, 107, 1118.
[108] W. S. Cheow, T. Y. Kiew, Y. Yang, K. Hadinoto, Mol. Pharm.
2014, 11, 1611.
[109] Y. K. Bae, O. H. Han, Microporous Mesoporous Mater. 2007, 106,
304.
[110] F. Carniato, G. Paul, C. Bisio, S. Caldarelli, L. Marchese, RSC
Adv. 2012, 2, 1153.
[111] B. Grünberg, T. Emmler, E. Gedat, I. Shenderovich, G. H.
Findenegg, H.-H. Limbach, G. Buntkowsky, Chemistry 2004, 10,
5689.
[112] I. G. Shenderovich, G. Buntkowsky, A. Schreiber, E. Gedat, S.
Sharif, J. Albrecht, N. S. Golubev, G. H. Findenegg, H.-H.
Limbach, J. Phys. Chem. B 2003, 107, 11924.
[113] M. Delle Piane, S. Vaccari, M. Corno, P. Ugliengo, J. Phys.
Chem. A 2014, 118, 5801.
[114] F. R. Hu, I. J. Wang, Curr. Eye Res. 1998, 17, 478.
[115] J. El-On, G. P. Jacobs, E. Witztum, C. L. Greenblatt, Antimicrob.
117
Agents Chemother. 1984, 26, 745.

[116] “Handbook of Pharmaceutical Excipients Seventh edition,” can be
found under
http://www.pharmpress.com/product/9780857110275/excipients,
n.d.
[117] T. W. Prow, J. E. Grice, L. L. Lin, R. Faye, M. Butler, W. Becker,
E. M. T. Wurm, C. Yoong, T. A. Robertson, H. P. Soyer, M. S.
Roberts, Adv. Drug Deliv. Rev. 2011, 63, 470.
[118] C. S. J. Campbell, L. R. Contreras-Rojas, M. B. Delgado-Charro,
R. H. Guy, J. Control. Release 2012, 162, 201.
[119] A. C. Watkinson, A. L. Bunge, J. Hadgraft, M. E. Lane, Pharm.
Res. 2013, 30, 1943.
[120] M. Manzano, M. Colilla, M. Vallet-Regí, Expert Opin. Drug Deliv.
2009, 6, 1383.
[121] S. Sapino, E. Ugazio, L. Gastaldi, I. Miletto, G. Berlier, D. Zonari,
S. Oliaro-Bosso, Eur. J. Pharm. Biopharm. 2015, 89, 116.
[122] C. Dianzani, G. P. Zara, G. Maina, P. Pettazzoni, S. Pizzimenti, F.
Rossi, C. L. Gigliotti, E. S. Ciamporcero, M. Daga, G. Barrera,
Biomed Res. Int. 2014, 2014, 13.
[123] M. Pilloni, G. Ennas, M. Casu, A. M. Fadda, F. Frongia, F.
Marongiu, R. Sanna, A. Scano, D. Valenti, C. Sinico, Pharm. Dev.
Technol. 18, 626.
[124] G. Berlier, L. Gastaldi, E. Ugazio, I. Miletto, P. Iliade, S. Sapino, J.
Colloid Interface Sci. 2013, 393, 109.
[125] Y. Antsiferova, N. Sotnikova, E. Parfenyuk, Biomed Res. Int.
2013, 2013, 10.
[126] L. Gastaldi, E. Ugazio, S. Sapino, P. Iliade, I. Miletto, G. Berlier,
Phys. Chem. Chem. Phys. 2012, 14, 11318.

118
APPENDIX I – LIST OF PUBBLICATIONS AND CONGRESS
PUBLICATIONS
 Gignone, A.; Manna, L.; Ronchetti, S.; Banchero, M.; Onida, B.

Incorporation of clotrimazole in Ordered Mesoporous Silica by
supercritical CO2. Microporous and Mesoporous Materials, 2014,
200, 291-296.
 Gignone, A.; Delle Piane, M.; Corno, M.; Ugliengo, P.; Onida, B.
Simulation and Experiment Reveal a Complex Scenario for the
Adsorption of an Antifungal Drug in Ordered Mesoporous Silica
 Sustained release of Active Pharmaceutical Ingredients from
Ordered Mesoporous Silica (in Preparation)
 Topical administration of Amikacin through OMS (in Preparation)
CONGRESS
 Sustained Release of Amikacin from Ordered Mesoporous Silica.

International Congress: Euro-Asia 2015 Zeolite Conference. Nice
(France) 25-28 January 2015
 Theoretical and experimental investigation of clotrimazole inside
SBA-15. International Congress: 6th International FEZA
Conference. Leipzig (Germany) 08-11 September 2014
 Incorporation of Clotrimazole in Ordered Mesoporous Silica by
Supercritical CO2. National Congress: XVII National Congress of
Catalysis GIC 2013 and XI National Congress of Zeolites Science
and Technology. Riccione (Italy) 15-18 September 2013

119
APPENDIX II – PUBBLISHED ARTICLES

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

Drug Delivery System

Uploaded by

Copyright:

Available Formats

Drug Delivery System

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Drug Delivery System

Uploaded by

Copyright:

Available Formats

POLITECNICO DI TORINO

Ordered Mesoporous Silica for Drug Delivery in Topical Applications

(Article begins on next page)

ORDERED MESOPOROUS SILICA FOR DRUG DELIVERY IN TOPICAL APPLICATIONS

ORDERED MESOPOROUS SILICA FOR DRUG DELIVERY IN TOPICAL APPLICATIONS

ORDERED MESOPOROUS SILICA FOR DRUG DELIVERY IN TOPICAL APPLICATIONS

3.1 OMS characteristics ............................................................................................................. 87

ORDERED MESOPOROUS SILICA FOR DRUG DELIVERY IN TOPICAL APPLICATIONS

Ab-Initio Molecular Dynamics AIMD

ORDERED MESOPOROUS SILICA FOR DRUG DELIVERY IN TOPICAL APPLICATIONS

This PhD thesis focuses the development and characterization of all

The first chapter comprehends the characterization of different OMS

Chapter 2 is oriented on the interaction details of API on silica surfaces.

Chapter 3 highlight the differences between OMS and the spatial

ORDERED MESOPOROUS SILICA FOR DRUG DELIVERY IN TOPICAL APPLICATIONS

ORDERED MESOPOROUS SILICA AS DRUG DELIVERY SYSTEM:

Drug delivery (DD) is the method or process of administering a pharmaceutical