Faculty of Technology and ScienceChemistry

DISSERTATION

Karlstad University Studies2009:33

Anna Hillerström

Effect of solvents during material treatment

applications

Tuning hydrophilicity of silicone rubber and drug loading in mesoporous silica

Karlstad University Studies2009:33

Anna Hillerström

Effect of solvents during material treatment

applications

Tuning hydrophilicity of silicone rubber and drug loading in mesoporous silica

Anna Hillerström. Effect of solvents during material treatment applications - Tuning hydrophilicity of silicone rubber and drug loading in mesoporous silica

DISSERTATION

Karlstad University Studies 2009:33ISSN 1403-8099 ISBN 978-91-7063-257-0

© The Author

Distribution:Faculty of Technology and ScienceChemistry SE-651 88 KarlstadSWEDEN+46 54 700 10 00

www.kau.se

Printed at: Universitetstryckeriet, Karlstad 2009

Abstract Choosing the right solvent is critical for many industrial applications. A useful property for selection of solvents is their solubility parameters. This concept of solubility parameters is central to this thesis and has been used in two different case studies of material treatment applications. Silicone rubber (crosslinked poly(dimethyl siloxane), PDMS) has many favorable material properties making it useful in biomedical devices. However, a limiting aspect of its material properties is a hydrophobic surface. The aim of this work was to prepare a hydrophilic PDMS material while retaining the transparency of the material. To do this, PDMS was combined with a hydrophilic polymer, polyvinylpyrrolidone (PVP) in an interpenetrating polymer network (IPN). A two-step IPN synthesis method was developed and it was found that the solvent used for polymerization of PVP had a significant influence on the water-wettability and the transparency of the PVP/PDMS-IPN. Several different analytical techniques were used for determining the degree of phase separation in the PVP/PDMS-IPN. It was found, by using microscopy techniques, that the PVP phase domains varied between 200 nm up to a few micrometers, and the size of the phase domains was correlated to the solvent used for polymerization of the IPN. The second topic for which solvent effects were explored was for the use of mesoporous silica particles as potential drug delivery devices. In the present work a drug molecule, ibuprofen, was loaded into mesoporous silica particles using different solvents, and in addition adsorption isotherms were established in each solvent. The maximum loading of ibuprofen in the mesoporous material was achieved when using a nonpolar solvent, in particular liquid carbon dioxide was successfully used. One of the advantages of using liquid carbon dioxide is that no solvent residues are left in the final material, which is important for pharmaceutical applications. Furthermore, it was concluded that ibuprofen was stored in an X-ray amorphous form in the mesoporous particles. Release studies in water showed a rapid release of ibuprofen from the mesoporous silica particles, while the dissolution of samples with crystalline ibuprofen was slower. This was verified to be an effect of a larger exposed ibuprofen area in the ibuprofen-loaded mesoporous silica particles, and it was concluded that the intrinsic dissolution rate for the samples were identical.

List of Articles This thesis is based on the following articles which will be referred to in the text by their roman numerals: Article I: Hillerström, A. and Kronberg, B.

A Two-Step Method for the Synthesis of a Hydrophilic PDMS Interpenetrating Polymer Network Journal of Applied Polymer Science, 2008, 110, (5), 3059-3067

Article II: Hillerström, A., Andersson, M., Skov Pedersen, J., Altskär, A.,

Langton, M., van Stam, J. and Kronberg, B. Transparency and Wettability of PVP/PDMS-IPN Synthesized in Different Organic Solvents Journal of Applied Polymer Science, 2009, 114, (3), 1828-1839

Article III: Hillerström, A., van Stam, J. and Andersson, M.

Ibuprofen Loading into Mesostructured Silica using Liquid Carbon Dioxide as a Solvent Green Chemistry, 2009, 11, (5), 662-667

Article IV: Hillerström, A., van Stam, J. and Andersson, M.

Strategies for obtaining a High Enrichment of ibuprofen in Mesoporous Silica, the Effect of Solvent Type, and Release Kinetics Submitted to European Journal of Pharmaceutics and Biopharmaceutics

My contributions: Articles I-IV: I performed all the experimental work included in the thesis, with the exceptions of TEM-images (performed by Annika Altskär at SIK, the Swedish Institute for Food and Biotechnology, Gothenburg, Sweden) and SAXS measurements (performed by Cristiano L.P. Oliveira at Aarhus University, Denmark). I am the main contributor to the interpretation of the results and writing of the articles. Reprints were made with the permission of the publishers.

Patents: Hillerström, A. and Kronberg, B. Method for Modifying Silicone Rubber

European Patent, EP 1 944 328 A1, (Filing date 2007-01-11) PCT International Application (Filing date 2008-01-10), WO 2008/083968 A1 (Publication date 2008-07-17) Hillerström, A. and Wolf S. Method for Loading a Molecule into a Porous Substrate US Patent Application, in the process of being drafted

Article not included in the thesis Andersson, M., Hillerström, A., Svensk, A., Younesi, S.R., Sjöström, E., Blute, I., Kjellin, M., Kizling, J., Kronberg, B., Oldgren, J., Hansson, A. and Sjöstrand, S. A New Class of Labile Surfactants that Break Down to Non-surface Active Products upon Heating or after a Pre-set Time, without the Need for a pH Change Tenside Surfactants Detergents, 2007, 44, (6), 366-372

Table of contents

1. Introduction............................................................................................. 7

1.1 Liquefied gases as solvents................................................................ 8 1.2 The solubility parameter concept.....................................................11 1.3 Synthesis and applications of interpenetrating polymer networks

(IPNs) ............................................................................................... 13 1.4 Synthesis and applications of ordered mesostructured materials .. 19

2. Experimental......................................................................................... 22

2.1 Materials........................................................................................... 22 2.2 Synthesis of PVP/PDMS-IPN ........................................................ 23 2.3 Loading of ibuprofen in mesoporous silica .................................... 24 2.4 Characterization techniques............................................................ 25

2.4.1 Microscopy techniques............................................................. 26 2.4.2 Spectroscopic techniques......................................................... 27 2.4.3 Thermal techniques ................................................................. 29 2.4.4 Miscellaneous techniques ........................................................ 30

3. Results and Discussion......................................................................... 31

3.1 Tuning the hydrophilicity of silicone rubber (Articles I and II) .... 31 3.1.1 One-step synthesis of PVP/PDMS-IPN.................................. 31 3.1.2 Two-step synthesis of PVP/PDMS-IPN ................................. 32 3.1.3 Wettability and transparency of PVP/PDMS-IPN.................. 35 3.1.4 Phase domains of PVP in PVP/PDMS-IPN ........................... 39

3.2 Drug loading in mesoporous silica (Articles III and IV) ............... 46 3.2.1 Loading of ibuprofen into mesoporous silica using different

solvents ..................................................................................... 46 3.2.2 Characterization of amorphous ibuprofen............................... 52 3.2.3 Release studies of ibuprofen .................................................... 54

4. Conclusions........................................................................................... 58

5. Future work........................................................................................... 60

6. Acknowledgements............................................................................... 62

7. References ............................................................................................. 64

7

1. Introduction

Solvents are able to dissolve substances without chemically changing them. They play an important role in the chemical industry but also in our everyday life since the most common solvent is water. In ancient times, Greek philosophers were speculating about the nature of solution and dissolution, but they made no distinction between water and other liquids since the term “water” then referred to anything liquid or dissolved. Later on alchemists considered the role played by solvents when they were searching for a universal solvent [1]. Water is our most important solvent since it is essential for the survival of most forms of life but most other commonly-used solvents are organic substances. Common uses for organic solvents are in solvent-borne paints (mineral turpentines, esters, glycol ethers), glues (acetone, methyl acetate), dry cleaning (tetrachloroethylene), perfumes (ethanol), degreasing products (ketones, chlorinated hydrocarbons) etc. Solvents are also important as dispersion media for pharmaceutical compounds and agrochemicals [2,3]. The reason for using particular solvents vary from application to application depending on the desired outcome or use, such as dissolution and/or viscosity reduction, extraction, separation, a reaction or transport medium or for influencing reaction rates etc [4]. Almost all organic solvents are volatile organic compounds (VOCs) and hazardous air pollutants (HAPs). Evaporation of these solvents are creating severe environmental problems [3]. Apart from the environmental concern of organic solvents, they also cause potential health and safety hazards. Organic solvents are easily absorbed in the body and long-time exposure to these solvents can have detrimental effects on our health. Solvents are generally flammable and different safety measures have to be taken when using them in processes. These drawbacks have caused a general concern to decrease the usage of organic solvents by replacing them with more environmentally friendly alternatives or to use organic solvents with less negative health and environmental impacts. Although there is a strong incentive to replace organic solvents with something more environmentally benign, the alternative solvent should have equal or even better performance than the traditionally used organic solvents for the use of a solvent in a specific application.

8

The aim of this thesis has focused on the solvent effects for two different material treatment applications. The common theme for the two applications is the enrichment of substances in a porous matrix in the presence of solvents. The first application concerns the synthesis of a polymeric material, which can for instance be used for biomedical applications. Specifically, an interpenetrating polymer network was synthesized consisting of silicone rubber and a hydrophilic polymer in order to manipulate the wettability. This requires careful selection of the solvent used for the synthesis, since the solvent needs to have the “right” physical properties for the preparation of a material with the desired properties. The other application that is investigated in this thesis is related to drug delivery devices. The possibility of drug loading into mesostructured silica material has been investigated. For the loading of drug molecules into such silica particles, the selection of solvent has found to be of utmost importance, particularly for maximum drug loading. In addition, for drug applications, solvent residues left in the final material are unacceptable. By using organic solvents, there is a risk that solvent residues are left in the final product. Therefore liquid carbon dioxide was chosen as a solvent for the loading of a model drug molecule, ibuprofen, into mesoporous silica.

1.1 Liquefied gases as solvents

Since there is a shift towards more environmentally benign alternatives than the traditional organic solvents, research has been focused towards the use of supercritical fluids as solvents. Supercritical fluids or liquefied gases have been recognized to have many favorable characteristics for use in the chemical industry and several industrial processes using supercritical fluids have been developed, such as coffee and tea decaffeination, and hops and flavor extraction [5]. Despite the higher cost associated with high pressure processing and operational challenges, the interest in supercritical fluids has been growing steadily since the 1970´s [6]. Supercritical fluids are being evaluated in a diverse range of materials applications, such as the synthesis and processing of porous materials, particle formation and polymer synthesis. There are several reviews covering various aspects of these research areas [7-14].

9

The term “supercritical” refers to the state of matter, where temperature and pressure of a single component are above the critical point at which the phase boundaries disappear. The physical properties in the supercritical region are intermediate to the physical properties of liquids and gases, see Table 1. Table 1. Comparison of physico-chemical properties of a typical fluid in the liquid, gas and supercritical fluid states [15].

Diffusivity (cm2/s)

Viscosity (mN,s/m2)

Density (g/cm3)

Liquid 10-5 1 1 Supercritical fluid 10-3-10-4 10-1-10-2 0.5-1.1 Gas 10-1 10-2 10-3

Supercritical fluids have the mobility of gases but a dissolving power similar to liquid solvents. This implies that supercritical fluids can efficiently penetrate into porous materials due to the low surface tension (5 mN/m for liquid carbon dioxide, [16]) and the mass transfer is high. However, the most interesting feature of supercritical fluids is the ability to tune the solvent power by changing temperature and/or pressure. In particular, density and viscosity change drastically at conditions close to the critical point [5,17]. Liquefied gases at a pressure and temperature close to the critical point (the near-critical region) can be considered as a special case of supercritical fluids because, the properties of the liquefied gas resemble the supercritical fluid, such as reduced density and viscosity. A practical advantage of working with liquid carbon dioxide is that relatively high solvent densities can be achieved at moderate pressures (ρ=0.75 g/cm3 at 20 °C and 55 bar) [18].

10

Pres

sure

Temperature

L

G

Supercritical

regionS

Pres

sure

Temperature

L

G

Supercritical

regionS

Figure 1. Pressure-temperature diagram for carbon dioxide, showing the different phases including the supercritical region (S – solid, L – liquid, G – gas). Carbon dioxide is the most frequently used supercritical fluid (or liquefied gas) due to its relatively mild critical conditions (critical temperature, Tc=31 °C and critical pressure, pc=73.8 bar). In Figure 1, a schematic pressure-temperature diagram for carbon dioxide, showing the different phases, is depicted. Some of the many advantages with CO2 in supercritical or liquid state are listed below (some of the features are general for supercritical fluids, while some of them are specific for CO2) [19,20]: General

− Critical temperature near room temperature, i.e., temperature-sensitive samples can be handled

− No additional drying steps are required since the liquid/supercritical fluid is released as gas when the pressure is released

− Ability to tune the solubility by changing temperature and/or pressure − Ability to tune the solubility by adding cosolvents CO2

− Inert − Relatively nontoxic but CO2 is suffocating and the respiratory limit for CO2

exposure is 5000 ppm (which can be compared with the respiratory limit for hexane of 20 ppm) [21]

− Inexpensive

11

Carbon dioxide has no permanent dipole moment and a low dielectric constant but it has a quadrupole moment. Generally, supercritical carbon dioxide can dissolve substances with nonpolar character, but the quadrupole moment coupled with its Lewis acidity allows supercritical carbon dioxide to participate in interactions absent in hydrocarbons. It is suggested that weak intermolecular interactions exist between carbon dioxide and carbonyl-functional groups of solvated molecules [22,23].

1.2 The solubility parameter concept

The solubility parameter concept is a widely used tool for quantifying intermolecular interactions between a solvent and another component for the selection of a solvent or solvents for a particular application. Solubility parameters give a systematic estimate for the compatibility between two components. The solubility parameter concept was first developed as early as 1916 but was later formalized by Hildebrand and Scott in 1950 [24]. The basic principle was “like dissolves like” since the theory was originally aimed for nonpolar and nonassociating liquids. Today, the concept is extended to other systems and therefore “like dissolves like” can be modified to “like seeks like” since the concept is used in situations where the components actually not dissolves but interacts with each other [25]. The Hildebrand solubility parameter of a substance, δi, is defined as the square root of the cohesive energy density:

1/2

i

ii V

ΔEδ ⎟⎟⎠

⎞⎜⎜⎝

⎛= (Eq.1)

where ΔEi is the molar energy of vaporization (cohesive energy) and Vi is the molar volume of the pure solvent. The cohesive energy is associated with the net attractive interactions in a material.

12

The general rule of thumb for the Hildebrand solubility parameter is that two substances are miscible when their solubility parameters are close to each other or when the difference in solubility parameter is less than 2 MPa1/2 [15]. Values of the solubility parameter range from 7 MPa1/2 for hydrogen to 48 MPa1/2 for water. Materials with a high solubility parameter require more energy for dispersing than what is gained by mixing it with materials with a substantially lower solubility parameter and thus, these materials will be immiscible. On the other hand, two materials with similar solubility parameters will gain free energy when mixed together and they will, as an effect, be miscible [26]. The equation for the solubility parameter is a simple definition but it is not always easy to calculate. The solubility parameters do not take into account geometric aspects, such as size and structure of molecules, but in general compounds with smaller molar volume are better solvents than compounds with larger molar volumes [24]. For example methanol and acetone can dissolve a polymer with larger solubility parameter differences than a solvent with larger molar volume, e.g., n-hexadecane [25]. Later on, different models were developed from the Hildebrand solubility parameter concept taking into account that interactions between different compounds are of different kinds. The most famous method is the Hansen solubility parameter:

2H

2P

2D

2 δδδδ ++= (Eq.2) where the contributions from dispersive/non-polar, (δD), polar (δP) and hydrogen bond (δH) interactions are divided into separate contributions to calculate the total solubility parameter [25]. The solubility parameter concept can be applied for selection of solvents for polymer systems. Polymers usually degrade before the molar heat of vaporization can be measured and Equation 1 is therefore practically inapplicable for polymer solutions. Therefore, two different methods can be applied to determine the solubility parameter for polymer systems. The first method is to treat the polymer with several different solvents and evaluate the polymer swelling. The best solvent for the polymer swells the polymer the most and this solvent has a solubility parameter which resembles the solubility parameter of the polymer most. The other method for determination of the

13

solubility parameter for polymers is to measure the intrinsic viscosity of polymer solutions in several different solvents. The solvent that gives the highest intrinsic viscosity has similar solubility parameter as the polymer. The intrinsic viscosity is independent of polymer concentration but reflects polymer swelling and are thus, dependent on the solvent [27]. Since supercritical fluids or liquefied gases are also used as solvents it is preferable to utilize the solubility parameter concept for them. However, the Hildebrand solubility parameter concept has to be generalized to be valid also for gases and supercritical fluids due to their high nonideality, and the regular solution theory cannot be applied. The following empirical relation for anticipating the solubility parameter of supercritical fluids was derived from studies in liquid chromatography:

rl

rSCF1/2c1.25Pδ

ρρ

= (Eq.3)

where ρrSCF/ρrl is the state effect where ρrSCF denotes the reduced density of the supercritical fluid, ρr=ρ/ρc and ρc is the critical density. The reduced density of the fluid in liquid state is ρrl at its normal boiling point. Pc is the critical pressure and is the chemical effect [28,29]. This relation predicts the trend of solubility qualitatively but the quantitative effects are poorly predicted [30]. Several other studies based on either equations of state or density-dependent solute solubility parameter have followed this empirical relation to obtain a better estimation of the solubility parameters for supercritical fluids and liquefied gases [31-33].

1.3 Synthesis and applications of interpenetrating polymer networks (IPNs)

There is a tremendous variety of different types of polymer materials with widely different physico-chemical properties. In some cases, a certain combination of properties in a polymer is desired, e.g., hydrophilicity and mechanical flexibility, but that specific type of polymer is not available. For those circumstances, there are some different alternatives for how to proceed. One can for instance synthesize a totally new polymeric material with the desired properties but this requires a lot of work effort and is thus an expensive

14

alternative. Another possibility is to make use of already existing polymers and combine them to achieve the desired properties of the material. The easiest way to proceed is to simply mix two polymers to obtain a so called polymer blend. Polymer blending is a promising technique for the improvement of physical and chemical properties of polymer materials. However, there are only a few examples of miscible polymer blends and the great majority of them will appear as phase-separated materials [34]. For some applications, a phase-separated material is not a practical problem but for other systems it is preferable with a miscible polymer system. One method to enhance the compatibility of a polymer blend is to add a compatibilizer, like a block- or graft copolymer consisting of segments that are preferentially miscible with the polymers to be blended [35]. To address the miscibility problem, the synthesis of interpenetrating polymers networks (IPNs) was developed during the late 1960´s. IPNs are found to often possess a lower degree of phase separation in comparison to polymer blends. An IPN is defined as a combination of two (or more) polymers in network form where at least one is synthesized and/or crosslinked in the presence of the other. There are no covalent bonds (chemical crosslinks) between the two pure polymers but the polymer chains are rather entangled together by physical forces, see Figure 2. The term “interpenetrating” indicates that IPNs are miscible on a molecular scale. However, that has proven not be the case for most systems but rather the size of phase domains in the IPN can form finely divided phases of tens of nanometers in size [36].

Figure 2. Schematic drawing of an IPN where each polymer is crosslinked but there are no chemical crosslinks between the different polymers.

15

Contrary to polymer blends, block and grafted-copolymers, an IPN does not dissolve in a solvent but only swelling occurs [37]. IPNs can be synthesized simultaneously where a solution of both monomers and their respective crosslinkers are first mixed and then allowed to polymerize, or alternatively, IPNs are synthesized sequentially where the first polymer is polymerized and crosslinked and then the monomer of the second polymer and crosslinker are swollen into the first polymer by means of a solvent and polymerized in situ. A semi-IPN can also be synthesized where a non-crosslinked polymer is entrapped into another polymer network [38]. During this project, sequential IPNs were synthesized. As mentioned previously, by combining different polymers it is possible to tune the properties of the final material. In the case of a miscible polymer blend or IPN, the material properties of a miscible blend behave similar as a single component system but their properties are a combination of the properties of the individual components [39].

Tg

x 1

Polymer 2

Polymer 1

a

cb

bc

Tg

x 1

Polymer 2

Polymer 1

a

cb

bc

Figure 3. Glass transitions for a) a miscible polymer system (IPN or polymer blend), b) an immiscible polymer system and c) a borderline case of the miscibility of the polymer system. For example, by measuring the glass transition temperature (Tg) of an IPN it can be established whether the components are miscible or not. In a miscible system, the measured Tg will be in between the Tg’s measured for the individual polymers, while in an immiscible polymer system two separate Tg’s will be measured. Of course, there are also borderline cases where the Tg’s of the

16

multicomponent system only show a shift in temperature. The different cases discussed above are also graphically illustrated in Figure 3, showing a miscible, an immiscible and a partly miscible polymer system. The behavior of the glass transition in multicomponent system can also be observed for other properties like, mechanical properties, resistance to chemicals, radiation or heat. It is important to control the morphology of the microphase separation, since the microphase-separated structures directly influence properties and performance of the IPN. The miscibility in IPNs is for instance promoted by matching the rates of the crosslinking reactions of the polymers and minimizing the difference in the solubility parameters and glass transitions of the polymers [40]. Since phase separation in IPNs is a critical parameter, several different techniques are used for characterization of the miscibility between the polymers in the IPN. Measuring the glass transition temperature is an indication of the degree of miscibility but often also different scattering techniques like, small-angle neutron scattering (SANS) and small-angle X-ray scattering (SAXS) are employed for characterization of the miscibility. Microscopy techniques, like scanning electron microscopy (SEM) and transmission electron microscopy (TEM) can also provide detailed information about miscibility and the phase morphology of the IPN [27]. According to the Gibbs free energy of mixing, two polymers will mix if Gibbs free energy of mixing (ΔGMix) is negative:

MixMixMix TΔΔHΔG S−= (Eq.4) where ΔHMix is the enthalpy of mixing, ΔSMix is the entropy of mixing and T is temperature. Equation 4 can be further rewritten using the regular solution theory to express the Gibbs free energy with a combinatorial part (the entropy part) and an interactional part (the enthalpy part):

⎥⎦

⎤⎢⎣

⎡++= 21

s

122

2

21

1

1Mix

Vχln

Vln

VkT

VΔG φφφφφφ (Eq.5)

17

where V is volume, Vi is the molar volume of component i, Vs is the interacting segment volume, k is Boltzmann’s constant, φi is the volume fraction of component i and χ12 is the interaction parameter [35,41]. From a thermodynamically point of view, polymers are rarely miscible since the combinatorial entropy part in Equation 5 makes a negligible contribution to ΔGMix for a polymer-polymer system. However, a polymer-polymer system can be miscible if the heat of mixing is negative or exothermic (the interactional part in Equation 5) but this can only be achieved if there are specific attractive intermolecular interactions like hydrogen bondings or acid-base interactions, between the two types of polymers [27]. The early synthesized IPNs were focused on combining polymers with different mechanical properties, for example in sound and vibration damping applications where stiff and rubbery polymers are combined [36]. Later on, several studies have focused on the effect of other properties to utilize the IPN materials for different biomedical applications and the properties that are of interest in those systems are biocompatibility, barrier properties, hydrophilicity and transparency. The scope of this thesis concerns the combination of the hydrophobic silicone rubber and a hydrophilic polymer to form an IPN with a water-wettable surface and the flexibility of silicone rubber since these materials can be used for biomedical devices, like contact lenses, implants and catheters. For barrier materials, like contact lenses, the permeable phase of the material needs to have continuity to be able to transport for example oxygen and water through the matrix. Silicones are a general category of synthetic polymers, whose backbone consists of repeating units of silicon to oxygen bonds. In addition to these bonds are typically two organic groups connected, where the most applied organic group is the methyl group (CH3) resulting in the poly(dimethyl siloxane) or shortly abbreviated to PDMS, see Figure 4. Due to the low glass transition temperature of PDMS (-120 °C) it is necessary to crosslink the material to make it solid and the thereby obtained material is then generally denoted silicone rubber [42,43]. Hereafter, PDMS will denote silicone rubber/crosslinked PDMS in the following text. PDMS has many advantages, including flexibility, low dielectricity, ultraviolet resistance, non-flammability as well as high thermal and chemical stability [44,45]. Additionally, PDMS has several properties that are interesting for biomedical applications, due to high oxygen permeability, biocompatibility, transparency and the mechanical properties (dimensionally

18

Figure 4. Molecular structure of PDMS and PVP. stable material) [46]. However, the disadvantages of using PDMS for biomedical applications are the low surface free energy and the low water-permeability. There are several examples in the literature where PDMS has been combined with different polymers to form an IPN, like poly(N-isopropyl acrylamide), poly(methacrylic acid), poly(hydroxyl-2-ethylmethacrylate) and poly(vinyl alcohol) [46-52]. Other authors have also synthesized IPNs by combining PDMS with other polymers, for example, polystyrene, and poly(methyl methacrylate) [53-55]. Liu et al. prepared a semi-IPN consisting of silicone rubber and poly(methyl methacrylate) using supercritical carbon dioxide [56]. While supercritical carbon dioxide is a poor solvent for most polymers, it swells many polymers and this offers the possibility to impregnate monomers and other substances in a polymer matrix that subsequently can be polymerized. Hydrogels are hydrophilic polymeric networks where the networks are held together by molecular entanglements and/or secondary forces and hydrogels can absorb substantial amounts of water. Furthermore, hydrogels have been used as carriers for applications such as drug release, but the major problems with hydrogels are their relatively low mechanical strength. Polyvinylpyrrolidone (PVP) was the hydrogel used with PDMS for the synthesis of the IPN in this thesis. PVP finds widespread application in different areas like, medicine, cosmetics and pharmacy [57,58]. PVP is an efficient hydrogel and can absorb up to 60 % of its weight in water depending on the degree of crosslinking [59,60]. The repeating unit of PVP is shown in Figure 4.

19

1.4 Synthesis and applications of ordered mesostructured materials

Considerable synthesis efforts have been devoted to developing frameworks with pore diameters within the mesoporous range (2-50 nm) and in 1992, Kresge et al. synthesized the first mesoporous silica material, by forming the material through a liquid crystalline templating method [61,62]. Several attractive features of mesoporous materials, such as stable mesoporous structures, large surface areas, tunable pore sizes and pore volumes have generated a lot of interest of the mesoporous materials for many different applications. Today, the range of potential applications for these materials are very diverse and include areas within catalysis, adsorbents, sensors, size-selective separation media but also for uses as drug delivery devices and supports for proteins and enzymes in biocatalytic applications [63-71]. For the synthesis of mesostructured silica materials: structure-directing surfactants, a source of silica, a catalyst (typically an acid or base) and a solvent are needed. Liquid crystals form in mixtures of polar solvents (water) and surfactants and the liquid crystals serve as templates for the inorganic silica species for the formation of the mesoporous silica. After templating of the silica species, the organic material can be removed by calcination of the material, yielding a porous material. A schematic drawing of the formation mechanism showing the different steps in the synthesis of mesoporous material is shown in Figure 5. The mesostructured material can be synthesized in the presence of amphiphilic molecules via in principal two different synthesis routes: the precipitation route from dilute systems and the evaporation induced self-assembly route. In the latter method, the morphology of the mesoporous material is controlled by the selective evaporation of solvents [61,72]. The structure and phase behavior of the hybrid assembly depend on complex cooperative interactions between the inorganic and organic species. By varying the type of surfactant used, different mesoporous structure (hexagonal, cubic, lamellar etc) will develop with different pore sizes [73-76].

20

Figure 5. The formation mechanism of mesoporous silica materials with a hexagonal phase structure. Upon synthesis of mesoporous materials, several characterization techniques can be utilized to understand the character of the material. The most commonly used techniques for characterization of mesoporous materials are nitrogen adsorption analysis, small-angle X-ray diffraction and transmission electron microscopy. These techniques are used to establish different parameters of the material like, particle morphology, unit cell size, pore diameter, specific surface area and pore volume. For drug delivery applications, it is desirable that the drug is released with a maintained therapeutic level during the treatment period and preferably also released at a specific site and with a specific rate in the body. To achieve this, the drug molecules have to be encapsulated in another material, i.e., a carrier material. Traditionally, polymer matrices of hydrogels have been used and in those cases, the drug release mechanism is dependent on the rate of erosion of the hydrogel, which may be a process difficult to control. Additionally, there is a problem to distribute the drug homogeneously throughout the polymer matrix and large variations in the drug concentration can affect the release rate [77]. Apart from hydrogels, porous silica has been used as a drug carrier but since the material is amorphous, the same problem as incorporating drug molecules in hydrogels will arise [78]. In order to overcome the limitations encountered when using hydrogels and porous silica as drug carrier, the mesoporous materials have been suggested to be used as a reservoir for drug molecules. Since the mesoporous materials offer several unique features, such as stable mesoporous structure, biocompatibility, large surface areas, tunable pore sizes and volumes, there is a great potential for utilizing the mesoporous materials for controlled release of drugs [79]. In the mesoporous material, the drug release will depend on rate of diffusion of the

21

drug through the pore channel. Because the pores are well-defined and periodically arranged throughout the mesoporous particle, the release of drugs from mesoporous materials can be well-controlled and the release is not dependent on degradation processes as in the case of using hydrogels. Another advantage of using drug-loaded mesoporous particles is that the bioavailability of poorly soluble drugs are expected to be enhanced since the drug is stored in the amorphous form in the mesoporous particles. The amorphous form a drug has higher free energy than their crystalline counterparts and as a result the dissolution rates are generally higher [80,81]. The use of mesoporous silica as a drug delivery system was demonstrated in 2001 when a model drug molecule, ibuprofen, was adsorbed in a MCM-41-type material [82]. Since then several other studies have been done where ibuprofen has been loaded into mesoporous silica materials [83-86]. Ibuprofen (see Figure 6), which has been used within this thesis, is a commonly used model drug molecule for studies of drug delivery applications. It is an analgesic and anti-inflammatory drug with low water solubility (10 ppm at pH 7 but dependent on pH) and pKa ∼ 4.4 [85,87,88].

Figure 6. Molecular structure of ibuprofen. The potential amount of drug loading in the mesoporous material will depend on the solvent used during impregnation of the drug into the silica pores, the type of drug, the mean pore diameter and the specific surface area of the mesoporous material. Drug molecules can be adsorbed into mesoporous silica material through physical interactions (van der Waals interactions, hydrogen bonding or hydrophobic interactions) between the support (containing silanol groups) and the drug molecule. In addition, the pore walls of the mesoporous silica contain silanol groups can be functionalized and this will also affect the drug loading and release [89]. These factors may in turn also affect the release kinetics of the drug but other factors like the pH of the dissolution medium, the pore connectivity and the geometry of the mesoporous particles will also influence the release rate [90].

22

2. Experimental

2.1 Materials

Solvents Table 2. Main solvents used within the thesis and their solubility parameters

Solvents Purity and supplier Solubility parameter (MPa1/2)[26,91-93]

Liquid carbon dioxide 99.7%, AGA Gas 12.0 n-Hexane p.a., Merck 14.9 Cyclohexane p.a., Merck 16.8 Diethyl carbonate 99%, Aldrich 18.0 Toluene p.a., Merck 18.2 Di(ethylene glycol) ethyl ether 99+%, Aldrich 21.9 2-propanol p.a., Merck 23.5 1-propanol p.a., Merck 24.3 Ethanol 99.7%, Solveco 26.0 Materials for tuning hydrophilicity of silicone rubber For synthesis of the PVP/PDMS-IPN, a transparent, crosslinked PDMS elastomer with 0.51 mm thickness and 50 shore A hardness (supplied by Mentor Corporation) and the hydrophilic monomer, N-vinyl-2-pyrrolidone (NVP, 99%, stabilized with NaOH, Aldrich) were used. A crosslinker, triethyleneglycol dimethacrylate (TEGDMA, 95%, stabilized with 80 ppm hydroquinone, Aldrich), and a photoinitiator, Irgacure 2100 (Ciba Specialty Chemicals) were used for the free-radical polymerization reaction of NVP. Materials for loading ibuprofen in mesoporous silica A C16TAB (cetyltrimethylammonium bromide)-templated mesostructured silica material was synthesized in an aerosol-based synthesis procedure developed at YKI, Institute for Surface Chemistry, Stockholm, Sweden [94]. The mesoporous silica material has hexagonal arrays of straight cylindrical mesopores and the BET surface of the particles was 1106 m2/g, pore volume 0.73 cm3/g and mean pore diameter 2.5 nm [94]. Ibuprofen (>98%, Aldrich) was used for loading of the mesoporous particles.

23

2.2 Synthesis of PVP/PDMS-IPN

Initial tests were performed where the synthesis of the PVP/PDMS-IPN was performed in one step. The monomer, crosslinker and photoinitiator were mixed in the solvent/-s and a rectangular disc of PDMS (1 cm x 2 cm) was added to the solution. The sample was impregnated during one hour before the sample was UV-polymerized for at least 60 minutes. The solvents that were used for this synthesis procedure were toluene, cyclohexane, 2-propanol, ethanol or liquid carbon dioxide. In some experiments, mixtures of solvents were used to combine the properties of nonpolar and polar solvents. For the synthesis of the PVP/PDMS-IPN in liquid carbon dioxide, an in-house built stainless steel reactor (1.7 l) equipped with two sapphire windows was used for mixing the components with liquid carbon dioxide, see Figure 7. Experiments performed in liquid carbon dioxide were conducted at room temperature and 55 bar.

Figure 7. The stainless steel reactor used for experiments in liquid carbon dioxide. The two sapphire windows on the reactor are UV-transparent. For the two-step synthesis procedure, see Figure 8, a circular PDMS disc (1 cm diameter) was soaked in a solution containing the crosslinker and the photoinitiator and the soaking time lasted for one hour. Thereafter, the swollen PDMS disc with crosslinker and photoinitiator was placed in another solution containing the monomer, NVP, and the free radical polymerization reaction was immediately initiated by using a 200 W mercury-xenon lamp (365 nm, LC-8, L8868-02, Hamamatsu) using a constant UV-intensity (300 mW/cm2). The sample was exposed to UV-curing for at least 60 minutes to complete the polymerization reaction. After the polymerization reaction, the sample was

24

soaked in distilled water to remove residues and noncrosslinked PVP, and thereafter the sample was stored in distilled water to ensure the hydrophilic behavior of the material.

PDMS

UV

IPN

Solution step 1

Photoinitiator / crosslinker / solvent

Solution step 2

Monomer / solvent

PDMS

UV

IPN

Solution step 1

Photoinitiator / crosslinker / solvent

Solution step 2

Monomer / solvent Figure 8. The two-step method for the synthesis of the PVP/PDMS-IPN.

2.3 Loading of ibuprofen in mesoporous silica

For the loading of ibuprofen in mesoporous silica particles in liquid carbon dioxide with/without cosolvent, the in-house built stainless steel reactor was used. The experiments were performed at room temperature and 55 bar. Ibuprofen was also loaded into mesoporous silica by using cyclohexane or methanol.

25

2.4 Characterization techniques

Several characterization techniques have been used and they are briefly discussed below with a short description of the basic principle for each technique and how the techniques were specifically used within the thesis. A summary of the characterization techniques used in each part of the project is presented in Table 3. Table 3. Characterization techniques and the abbreviations for the used techniques are environmental scanning electron microscopy equipped with an energy dispersive X-ray spectrometer (ESEM-EDAX), transmission electron microscopy (TEM), atomic force microscopy (AFM), small angle X-ray scattering (SAXS), powder X-ray diffraction (XRPD), thermogravimetric analysis (TGA), differential scanning calorimetry (DSC) and dynamic absorption tester (DAT).

Technique PVP/PDMS-IPN Articles I and II

Ibuprofen in mesoporous SiO2 Articles III and IV

Microscopy techniques ESEM-EDAX X X (SEM only) TEM X AFM X Confocal Raman microscopy O X Spectroscopic techniques UV/vis spectrometry X X SAXS X XRPD O X Thermal techniques TGA X DSC X Miscellaneous techniques DAT X Nitrogen sorption X Laser diffraction particle sizing X X – samples were tested with the technique and successful results were obtained. O – samples were tested with the technique but no successful results were obtained.

26

2.4.1 Microscopy techniques

Environmental scanning electron microscopy (ESEM) [95] In the ESEM (XL30 ESEM TMP, FEI/Philips), images of samples are created through scanning an electron beam over the sample and scattered electrons (secondary electrons and primary backscattered electrons) are detected and form an image. Uncoated samples analyzed in the high vacuum mode often have a problem of accumulation of electrical charges on the sample. This can be solved by using the microscope in the environmental-SEM mode (ESEM) where the pressure is higher than in the SEM mode, which helps to reduce the charging of the sample. The microscope was also equipped with an energy dispersive X-ray spectrometer (EDAX) where samples are investigated by analyzing X-rays emitted from the sample and the fundamental principle is that each element has a unique atomic structure and the emitted X-rays are therefore unique for each element. ESEM-EDAX was employed to study the distribution of copper and silicon throughout cross-sections of the PVP/PDMS-IPN samples that had been soaked in 0.1 M CuSO4 (aq). The phase domains of PVP in the dry PVP/PDMS-IPNs were analyzed with SEM by alternating between the secondary electron (SE) detector and the backscattering electron (BSE) detector. The ibuprofen-loaded mesoporous silica particles were also analyzed with the SE detector in SEM-mode to study the morphology of the samples. Transmission electron microscopy (TEM) [96] By using TEM (LEO 906E, LEO Elektronmikroskopie), an electron beam is transmitted through an ultrathin section of the sample (50-100 nm). An image is formed from the interaction of electrons transmitted through the sample. Significantly higher resolution than with SEM can be obtained. TEM was used to study the phase domains of PVP in the water-swollen PVP/PDMS-IPN and the images were compared with TEM-images of PDMS soaked in water. A special sample preparation procedure was used to visualize the PVP-domains in TEM by using freeze etching to fracture the samples (-100 °C for PVP/PDMS-IPN and -130 °C for PDMS). From the cross-section

27

of the sample, a replica consisting of Pt/C corresponding to the water-swollen PVP/PDMS-IPN was formed and the replica was analyzed in TEM. Atomic force microscopy (AFM) [97] The AFM (Multimode Nanoscope IIIa AFM, Digital Instruments) can be used to image sample surfaces by scanning a cantilever with a sharp tip over the surface. When the tip is brought close to or in contact with the surface, the force between the tip and the sample causes a deflection of the cantilever that is measured by a laser spot that is reflected into an array of photodiodes. Different modes can be used with AFM but in this thesis, tapping mode AFM where the tip is made to oscillate close to its resonance frequency in the normal direction was employed. The phase domains of PVP from cross-sections of the dry PVP/PDMS-IPNs were characterized by studying the topography and phase images. Confocal Raman microscopy [98] Raman scattering is an inelastic scattering process and a spectrum that is individual for each molecule is obtained. By combining Raman spectroscopy with confocal microscopy, a depth-analysis through a sample is performed by using a pinhole to eliminate out-of-focus light. The instrument (Alpha 300, WITec) has a lateral resolution of 200 nm and a vertical resolution of 500 nm. Confocal Raman microscopy was used to study the distribution of ibuprofen in the mesoporous silica particles by measuring a Raman spectrum at the surface and in the middle of a particle.

2.4.2 Spectroscopic techniques

UV/vis spectrometry [99] Organic chromophores absorb light in the UV or visible region of the electromagnetic spectrum. The intensity of light that is absorbed or transmitted through a sample is measured with a UV/vis spectrometer (UV Lambda 650, Perkin-Elmer) in the whole range of wavelengths or at one specific wavelength.

28

The transparency of the PVP/PDMS-IPN samples were determined by measuring the transmittance of the water-swollen IPN samples at λ=550 nm by placing a sample in front of the beam in the spectrometer. The transparency of the PVP/PDMS-IPNs was obtained from measurements of the relative transmittance of the IPN and the transmittance of a reference PDMS sample. The release of ibuprofen from the mesoporous particles and the dissolution of crystalline ibuprofen in Milli-Q water were measured at λ=222 nm by using an in-house constructed continuous flow setup. The setup consists of a flow cell cuvette of quartz glass that is mounted in the spectrometer and a peristaltic pump (45 ml/min) is employed to circulate the solution. The absorbance was measured every 0.4 s during 15 minutes and an ultrasonic bath was used to deagglomerate the particles in the solution during the whole measurement period. Small angle X-ray scattering (SAXS) [100] SAXS (modified Nanostar, Bruker AXS) is a nondestructive method used for measuring scattering phenomena in systems occurring at small angles, typically at 2θ less than a few degrees. Scattered X-rays from a monochromatic beam is detected by a detector that is placed perpendicular behind the sample. In a typical experiment, the intensity of the scattered light is measured at different scattering vectors. From the curve, information about the structure of the samples, such as particle size and shape is elucidated. There is no requirement for using crystalline samples for SAXS measurements and domains in the size range 1-100 nm can be resolved. Swollen samples of PVP/PDMS-IPN polymerized in different solvents and PDMS were analyzed with SAXS to study the size of PVP domains in the IPN. Powder X-ray diffraction (XRPD) [101] XRPD (X’Pert Pro, PANalytical) is, similar to SAXS, a nondestructive method and scattering occurs due to variations in electron density in the sample. By measuring the intensity of the scattered X-rays as a function of the scattering angle, a diffraction pattern is obtained. Crystalline domains in a sample appear as sharp peaks in the diffractogram, while broader peaks appear for samples with less degree of crystallinity/smaller crystalline phase domains or particles. Line broadening is detectable when the size of the scattering crystallites/domains gets below 100 nm.

29

PVP/PDMS-IPN samples polymerized in different solvents were analyzed with XRPD in an attempt to calculate the size of the PVP phase domains in the IPN from XRPD line broadening. Ibuprofen-loaded mesoporous silica materials were analyzed with XRPD and the obtained diffraction patterns were compared with the diffraction pattern of ibuprofen and mesoporous silica to confirm that ibuprofen has in fact been loaded into the pores of the mesoporous particles.

2.4.3 Thermal techniques

Thermogravimetric analysis (TGA) [102] TGA (TGA 7, PerkinElmer) determines changes in weight of samples in relation to change in temperature by using a very accurate thermobalance. The sample pan is placed in an insulated oven and a purge gas can be introduced, for example by purging with an inert gas for sensitive samples. The amount of ibuprofen loaded into the mesoporous particles was measured with TGA, using N2 atmosphere to evaporate ibuprofen from the samples. Differential scanning calorimetry (DSC) [103] DSC (DSC1, Mettler-Toledo) is a technique in which the heat-flow rate to the sample is measured against temperature or time, while the temperature of the sample is programmed. The heat-flow rate of the sample is compared with that of a reference sample and the difference in heat-flow rate is reported. It is possible to study many different types of phase transitions, such as melting, crystallization and glass transitions with the technique. The glass transition temperatures of PVP and PDMS in the PVP/PDMS-IPN polymerized in different solvents were studied with DSC.

30

2.4.4 Miscellaneous techniques

Wettability measurements The contact angle of liquids on solid surfaces can be measured with a Dynamic Absorption Tester, DAT (Fibro 1100, Fibro Systems) by depositing a liquid drop and acquiring and analyzing their profile images with a camera. DAT was used to measure the contact angle of water on the surfaces of PVP/PDMS-IPNs (dried and swollen in water). However, the main part of the evaluation of wettability of the PVP/PDMS-IPNs was performed by visual inspection of the spreading of the water film on the IPN when it was removed from water. The results were compared with the wettability behavior of PDMS samples analyzed with a similar procedure. Nitrogen sorption measurements [104] From nitrogen adsorption/desorption measurements, the specific area (BET area) is calculated by measuring the nitrogen adsorption and desorption in a sample as a function of pressure by cooling an evacuated sample tube to cryogenic temperatures and then expose the sample with analysis gas at a series of controlled pressures. Nitrogen sorption measurements (TriSTAR 3000, Micromeritics) were performed on the ibuprofen-loaded mesoporous silica to determine the loading degree of ibuprofen in the pores. Laser diffraction particle sizing [105] The technique of laser diffraction (Mastersizer 2000, Malvern Instruments) is based on the principle that particles passing through a laser beam will scatter light at an angle that is directly related to their size. As the particle size decreases, the observed scattering angle increases logarithmically and a series of detectors are used to measure the scattered light over a range of scattering angles. Particle size distributions of mesoporous silica material and crystalline ibuprofen were measured with this technique and the size distributions obtained were used as input data for the numerical model for calculation of the intrinsic dissolution rate of ibuprofen.

31

3. Results and Discussion

3.1 Tuning the hydrophilicity of silicone rubber (Articles I and II)

3.1.1 One-step synthesis of PVP/PDMS-IPN

Table 4. Comparison of the PVP concentration and water wettability for PVP/PDMS-IPN samples polymerized with a one-step method and a two-step method. The spreading coefficient, S(tot) includes the amount of NVP and the solvent (50/50 monomer/solvent) and is calculated for the samples polymerized with the two-step process.

Solvents S (tot) Concentration PVP in IPN (%)

Water wettable IPN

One-step process Liquid carbon dioxide - 10.3 ± 1.8 No Toluene - 4.5 ± 1.6 No Ethanol - 0.8 No Two-step process n-Hexane -86 32.8 ± 9.7 Yes Cyclohexane -70 36.2 ± 9.1 Yes Diethyl carbonate -60 28.2 ± 3.1 Yes Toluene -58 27.0 ± 5.0 Yes Di(ethylene glycol) ethyl ether

-27 9.7 ± 0.7 Yes

2-propanol -13 17.5 ± 1.3 Yes 1-propanol -7 18.2 ± 2.5 Yes

During the development of a synthesis of a hydrophilic and transparent silicone rubber material, several parameters were varied: the monomer type, the initiator type, the crosslinker, the solvent, impregnation time and polymerization time. In the initial synthesis tests, all components were mixed simultaneously and the PDMS material was soaked into the solution and then polymerized. However, this approach rendered no successful IPNs in terms of altering the surface wettability and a few selected results by using this synthesis procedure are

32

shown in Table 4. The IPN samples using this protocol contained a low amount of the hydrophilic polymer and appeared hydrophobic and white. Liquid carbon dioxide was used as a solvent to synthesize the PVP/PDMS-IPN but these samples also appeared hydrophobic and opaque, which was unfortunate since using a green solvent like liquid carbon dioxide would have been advantageous for the synthesis of a material with a potential use as a biomedical device.

3.1.2 Two-step synthesis of PVP/PDMS-IPN

From the results discussed above, we propose the hypothesis that two requirements have to be fulfilled to form a water-wettable PDMS surface. Firstly, the components used for polymerization of PVP have to be swelled into the PDMS matrix by a solvent and secondly, after the polymerization, the surface of the IPN needs to be covered with an excess PVP layer.

Figure 9. The surface appearances of the PVP/PDMS-IPN after polymerization in solvents with different solubility parameters. In path 1, the surface of the IPN is hydrophobic (polymerized in a solvent with S < 0) while in path 2, a PVP-wettable surface is formed (polymerized in a solvent with S ≥ 0).

Solvent

1)

2)

PDMS

Solvent

Solvent

PVP

PVP/PDMS-IPN

PVP/PDMS-IPN

PDMS

33

To fulfill these two criteria, the selection of the solvent for the IPN synthesis was crucial and in principle, the selection of a solvent for obtaining a water-wettable PVP surface was based on the spreading coefficient defined as: S = γPDMS/solvent – (γPDMS/PVP + γPVP/solvent) (Eq.6) where γij is the interfacial tension between component i and j [106]. If S < 0, PVP will form droplets on the PDMS surface and this corresponds to path 1 in Figure 9, while in path 2 with S ≥ 0 a PVP-surface is formed on the PDMS surface after the IPN synthesis. The interfacial tension of the different phases in Equation 6 was converted into solubility parameters:

)δ)(δδ(δaN

2mVS SolventPVPPVPPDMSa

1 −−= (Eq.7)

which are more convenient to use for calculations than the interfacial tensions (m is a constant, V1 is the molar volume of the solvent, a is the cross-sectional area per molecule and Na is Avogadro´s constant). It was found that the theoretical threshold for wetting of water on a PVP-film was obtained when the solubility parameter of the solvent was δsolvent ≥ δPVP (=23.3 MPa1/2).

Figure 10. Schematic drawing of the degree of swelling of PDMS in solvents with different solubility parameters and the solubility parameter of the solvent (δSolvent ≥ δPVP) that forms a hydrophilic PVP/PDMS-IPN.

34

Apart from fulfilling the conditions for forming a wetting PVP-film on the PDMS-surface, it was necessary to soak the components for polymerization of PVP in an efficient swelling solvent. This was achieved by selecting a solvent with similar solubility parameter as PDMS (δPDMS=14.9 MPa1/2). Consequently, the solvent that should be used for forming a water wettable PDMS surface and swelling PDMS are very different, as illustrated in Figure 10. This led to the use of a two-step process, where the components for polymerization of PVP are first swelled into the PDMS matrix and then the soaked PDMS sample is placed in another solution containing the NVP monomer and the solvent needed to form a water-wettable PVP-surface of the IPN. By using the two-step process for polymerization of the PVP/PDMS-IPN, a high concentration of PVP in the IPN can be reached as shown in Table 4. The samples turned hydrophilic even though the calculated spreading coefficient was negative for all solvents. The reason for this is ascribed to the nonequilibrium conditions that are present for high monomer and photoinitiator concentrations when the polymerization rate is high during the polymerization of the PVP/PDMS-IPN. As soon as some PVP chains are formed at the PDMS surface by polymerization in a nonpolar solvent, these chains will attract more NVP towards the surface and the polymerization of PVP on the surface is accelerated. The reasoning above was supported by calculating the NVP/solvent composition at a PDMS surface (the surface at the initiation of the polymerization reaction) and a PVP surface (the surface as the polymerization of the IPN progresses) by using the regular solution theory and χ-parameters for the different components in the system and using the following equation:

)2x(1χ)2xχ(1RT

)γa(γx1

xlnx1

xln s2

s2

21

2

2s2

s2 −−−+

−=⎟⎟

⎠

⎞⎜⎜⎝

⎛−

−⎟⎟⎠

⎞⎜⎜⎝

⎛−

(Eq.8)

where x2 is the mole fraction of the solvent, x2s is the mole fraction of the solvent at the surface, χ is the interaction parameter in the solution and χs is the interaction parameter at the surface [107]. By using Equation 8 for polymerization in cyclohexane, which predicted a negative spreading coefficient, the NVP concentration was 60 % at the PVP surface while it was 25 % at the PDMS surface. Consequently there is a tendency for PVP

35

polymerization on the PDMS surface in the presence of a nonpolar solvent but the surface excess of PVP increases as the polymerization progresses since the NVP concentration increased from 25 % to 60 %. The arguments using the surface excess calculations suggested that the actual value of the spreading coefficient is not important for the polymerization of the PVP/PDMS-IPN but rather the fact that the process takes place in two steps and the formation of a PVP-surface on the IPN is favored since the initiation of polymerization reaction takes place at the interface of PDMS-surface and the bulk solution. A remark regarding the use of liquid carbon dioxide as a solvent in the two-step method is that liquid carbon dioxide is a suitable swelling solvent of PDMS to soak the oil-soluble components in the PDMS matrix but it would not be a suitable solvent to form a water wettable surface after the polymerization of the PVP/PDMS-IPN since NVP and liquid carbon dioxide are immiscible. The compounds have widely different solubility parameters (δNVP=21.5 MPa1/2 and δCO2=12.0 MPa1/2) and liquid carbon dioxide does not match the criterion in Equation 7 for a PVP-wettable surface of the IPN [108].

3.1.3 Wettability and transparency of PVP/PDMS-IPN

All PVP/PDMS-IPN samples appeared hydrophobic in dry state but changed reversibly to hydrophilic when they were soaked in water, provided that there was a certain amount of PVP in the IPN. The transition from a hydrophobic to a hydrophilic material was correlated to the solvent used for polymerization of the PVP/PDMS-IPN. When the samples were polymerized in solvents with similar solubility parameter as PVP, the PVP concentration in the IPN needed for obtaining a water-wettable surface was lower than for samples polymerized in a nonsolvent for PVP.

36

Figure 11. Contact angle measurements of a) PDMS, b) 7.2 wt-% PVP/PDMS IPN and c) 44.0 wt-% PVP/PDMS-IPN. The IPN samples were polymerized in 30 wt-% toluene/ 70 wt-% ethanol and the samples were soaked in water prior to the contact angle measurements. Generally, the water-wettability of the IPN samples were characterized by visual inspection of the spreading of the water-film when the samples were removed from a water reservoir, but in addition, for some samples, the contact angle of wet IPN samples was measured. As shown in Figure 11, a sample containing a high PVP concentration in the IPN appeared hydrophilic, while a sample with a low PVP concentration was hydrophobic, like the original PDMS sample.

Figure 12. The cross-section of a PVP/PDMS-IPN sample soaked in 0.1 M CuSO4 solution has a homogenous distribution of copper throughout the cross-section.

37

For certain applications, it is advantageous if the surface as well as the bulk is hydrophilic. Figure 12 shows 250 μm of the cross-section (500 μm is the total cross-section of the sample) of a PVP/PDMS-IPN sample in ESEM-EDX for a sample that was soaked in 0.1 M CuSO4 and the copper was homogenously distributed throughout the cross-section. The aqueous copper solution will only penetrate the hydrophilic parts of the IPN, i.e., the PVP-rich regions, and the EDX-analysis proved that the IPN-material is in fact a water-permeable membrane, at least on a micrometer-scale, which is the resolution of the EDX-analysis. Naturally, no penetration of the copper solution would be possible into pure PDMS, which is a well known hydrophobic material. As discussed above, the PVP/PDMS-IPN samples appeared hydrophilic when they were soaked in water and it was demonstrated that the degree of swelling in water of the IPN samples was linearly correlated to the amount of PVP in the IPN samples. The degree of swelling in water for IPN-samples is presented in Figure 13. As can be seen a linear correlation was found for all samples, irrespective of the solvent used for polymerization of the PVP/PDMS-IPN.

0

20

40

60

80

100

0 10 20 30 40 50 60

Swel

ling

in w

ater

[wt-%

]

Concentration of PVP in IPN [wt-%]

Figure 13. The degree of swelling of PVP/PDMS-IPN in water in relation to the concentration of PVP in the IPN samples when polymerized in five different solvents. The symbols are corresponding to = 1-propanol, = toluene, = diethyl carbonate, = cyclohexane and = hexane. The filled symbols represent hydrophilic samples and the unfilled symbols represent hydrophobic samples.

38

The two-step method has shown the possibility to tune the hydrophilicity of silicone rubber but it was desirable to maintain the transparency of the material. Unfortunately, many of the PVP/PDMS-IPN samples showed a substantial decrease in transparency in comparison to the original PDMS material. This is illustrated in Figure 14, where the transparency of water-swollen IPN-samples polymerized in different solvents but with the same PVP concentration in the IPN was analyzed. The transparency of the PVP/PDMS-IPN samples was found to be strongly dependent on the choice of solvent for the synthesis. Especially, the samples polymerized in a solvent with a low solubility parameter were non-transparent, while using a solvent with a higher solubility parameter rendered IPN samples with higher transparency. An optimum in transparency was found in the case where the solvent used for polymerization was di(ethylene glycol) ethyl ether. A final remark about the transparency measurements is that all samples were white in dry state unless they contained a small amount of PVP (below 5 %), which made them maintain the transparency in dry and wet state.

0

20

40

60

80

100

14 16 18 20 22 24

Tran

spar

ency

[%]

δ [MPa1/2]

25 % PVP in IPN

Figure 14. Transparency of PVP/PDMS-IPNs polymerized in solvents with different solubility parameters. The filled symbols were obtained from extrapolated data and the unfilled symbols were obtained from interpolated data taken from Figure 2 in Article II.

39

3.1.4 Phase domains of PVP in PVP/PDMS-IPN

As discussed above, most PVP/PDMS-IPN samples showed a decrease in transparency in comparison to the PDMS material itself, which indicated that the IPN samples were phase separated to some extent. The degree of phase separation between the polymer pairs can be established by measuring the size of the phase domains of PVP in the IPN-samples.

Tg

10-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3[m]

SEMEDX-mapping

Confocal Raman

TEMSAXS

AFM

XRD

Tg

10-10 10-9 10-8 10-7 10-6 10-5 10-4 10-3[m]

SEMEDX-mapping

Confocal Raman

TEMSAXS

AFM

XRDFigure 15. Length scales covered by different analytical techniques for the assessment of the PVP phase domains in PVP/PDMS-IPN. Different analytical techniques can be used for measuring the miscibility in IPNs (and polymer blends) and in Figure 15 the length scale that each technique used within this thesis covers is summarized. When synthesizing a new multicomponent polymer material, the degree of phase separation is unknown and it is recommended to evaluate several techniques covering different length scales to measure the miscibility between the polymer pairs. Moreover, all techniques are not applicable for a specific material. Measurements of the glass transition temperatures of IPNs are frequently employed to establish the degree of miscibility between the polymer pairs and accordingly, the samples were analyzed in DSC to measure the Tg’s of the respective components. From the DSC data in Figure 16 it was found that the Tg of PDMS did not shift in the PVP/PDMS-IPN, which indicates that miscibility between PVP and PDMS was not achieved on a molecular level. In addition, the Tg of PVP could not be identified in the DSC thermogram but

40

-150 -100 -50 0 50 100 150 200

Hea

t flo

w [m

W g

-1]

Temperature [ C]

a) PDMS

b) PVP/PDMS-IPN, 1-propanol

c) PVP/PDMS-IPN, cyclohexane

Figure 16. DSC data for a) PDMS, b) PVP/PDMS-IPN polymerized in 1-propanol (20 % PVP, 79 % transparency) and c) PVP/PDMS-IPN polymerized in cyclohexane (20 % PVP, 13 % transparency) for determination of the miscibility of PVP in the IPN from the glass transition temperatures.

literature values of the Tg of PVP are in the region 50-180 °C, depending on the molecular weight and water content of the hydrogel [109,110]. In the DSC thermograms of the IPN in Figure 16, there is a shift in this temperature range, which may be due to the effect of the water content in the IPN but possibly water in the network is overlapping the glass transition of PVP and no information concerning the miscibility in the IPN was possible to deduce. Scattering of X-ray radiation permits fine structures to be resolved and the miscibility of the polymer pairs can be assessed from the data. For the PVP/PDMS-IPN samples, both SAXS and XRPD were employed to determine the degree of phase separation. The IPN samples analyzed in SAXS are presented in Figure 17 and the samples yielded no differences in scattering behavior in comparison to the original PDMS material. It was suggested that the observed scattering behavior was originating from the silica filler present in the original PDMS material and the phase domains of PVP could not be determined with SAXS.

41

10-4

10-3

10-2

10-1

100

101

102

103

0,01 0,1 1

PDMSPVP/PDMS-IPN, 1-propanolPVP/PDMS-IPN, hexane

I(q)

[cm

-1]

q [Å-1]

0 10 20 30 40 50

PDMSPVP/PDMS-IPN, 1-propanolPVP/PDMS-IPN, hexane

2θ

Figure 17. SAXS data (left graph) and XRPD data (right graph) for PDMS and two different PVP/PDMS-IPN samples, polymerized in 1-propanol (19 % PVP in the IPN, 64 % transparency) or hexane (29 % PVP in the IPN, 10 % transparency). XRPD data were also retrieved for similar kind of samples as with SAXS, see Figure 17. Sharp peaks of crystalline parts in PDMS or PVP/PDMS-IPN samples were, as expected, not detected since both polymers in the IPN are amorphous. Additionally, the broad peaks that were observed were present both in PDMS and in the PVP/PDMS-IPN samples at the same position in the XRPD pattern and the peaks are suggested to appear from a local ordering of the chains in the polymers. If the width of the peaks in the XRPD diffractogram would have varied for PDMS and the PVP/PDMS-IPNs, PVP domain sizes could be calculated from using the Scherrer equation, where smaller domains (crystals) are giving broad peaks in the diffraction pattern, while sharp peaks in the diffraction pattern are corresponding to larger domains (crystals) [111]. But in the present case the position and the width of the peaks for PDMS and PVP/PDMS-IPNS were similar and the Scherrer equation could not be employed. Consequently, XRPD was not possible to use for determination of the size of the PVP phase domains. Neither, the Tg measurements nor the X-ray scattering techniques of the IPN samples could reveal information about the miscibility in the polymers and other analytical techniques for characterization of the PVP phase domains in the PVP/PDMS-IPNs were used. It turned out that different microscopy techniques were more successful to establish information about the degree of phase separation in the PVP/PDMS-IPNs.

42

Figure 18. Tapping mode AFM-images (5 μm x 5 μm) of dried cross-sections of a) PDMS and b) PVP/PDMS-IPN polymerized in hexane (32 wt-% PVP, 0 % transparency). The left images are the height images and the right images are the phase images. By analyzing cross-sections of dried PVP/PDMS-IPNs with tapping mode AFM, it was possible to deduce PVP phase domains in the PVP/PDMS-IPNs both from the topography image (left image, Figure 18) and the phase image (right image, Figure 18). The AFM-images showed a clear distinction between the original PDMS material, where no domains were present in the PDMS material, and in the PVP/PDMS-IPNs with regions of irregularly shaped structures. The IPN sample presented in Figure 18 was polymerized in hexane, which was a nonsolvent for PVP and this sample was white but this sample clearly demonstrates the upper size range of the PVP phase domains (in the micrometer range). Other IPN samples polymerized in solvents with a closer solubility parameter to PVP exhibited smaller phase domains (not shown in the thesis) than the sample presented in Figure 18.

43

Figure 19. Phase domains of PVP of dried cross-sections of PVP/PDMS-IPN samples visualized with SEM using the BSE-detector (left images) and the SE-detector (right images). SEM-images of a) PDMS, b) PVP/PDMS-IPN polymerized in 1-propanol (22 % PVP, 82 % transparency) and c) PVP/PDMS-IPN polymerized in cyclohexane (20 % PVP, 13 % transparency). IPN-samples were also analyzed in SEM using two different detectors and a few selected images are shown in Figure 19. The left images are retrieved from the BSE detector that detects the contrast between areas of different chemical composition, in this case silicon-rich areas (brighter regions in the SEM images) and carbon-rich areas (darker regions in the SEM images). The IPN sample polymerized in cyclohexane exhibited a low transparency and in the SEM image using the BSE detector, small dark spots appear in the image. These spots are less prominent in the PVP/PDMS-IPN polymerized in 1-propanol and they are not at all visible in the reference sample (PDMS). These spots are not detectable with the SE detector, which is sensitive for local surface contours in

44

samples, and this proves that the spots are in fact regions of different chemical composition and not an effect of topography in the sample. The size of the PVP phase domains is 180 ± 7 nm for the sample polymerized in 1-propanol and 350 ± 16 nm for the sample polymerized in cyclohexane (calculated with the program ImageJ, National Institutes of Health), where the smaller domain sizes are obtained for the IPN sample polymerized in 1-propanol.

Figure 20. Phase domains of water-swollen PVP in the PVP/PDMS-IPN visualized with TEM (magnification x4646): a) PDMS (reference) and b) PVP/PDMS-IPN polymerized in toluene (27 % PVP, 58 % transparency). The arrow for the IPN-sample is pointing at a PVP-rich region in the sample. For the AFM and SEM analysis of the PVP/PDMS-IPNs, the samples were in dry state and it has already been confirmed that the IPNs swell to a large extent in water and this would presumable also affect the size of the PVP phase domains. Therefore, a replica of a water-swollen PVP/PDMS-IPN sample polymerized in toluene with a fairly high transparency (58 %) was prepared for TEM analysis. In addition, a PDMS sample was prepared in a similar way for comparison with the IPN sample. The TEM images of the two different samples are shown in Figure 20 and their appearances are different. The PVP/PDMS-IPN sample has regions of a fine network (indicated with an arrow in the image) and this type of structure is not visible in the PDMS sample. In the sample preparation for the TEM analysis, the samples are rapidly frozen, and then fractured prior to sublimation of the amorphous ice. In the next step, sublimation occurs and the PVP domains are dehydrated but still swollen in their expanded state and this causes tensions in

45

the material and wrinkling in those water-rich areas will occur. In the PDMS sample there are no wrinkled areas, which is in agreement that the PDMS sample does not contain any water. From the TEM images it was concluded that the PVP phase domains in the IPN are in the order of 250-350 nm and it should be remarked that only one IPN sample was analyzed with TEM but the obtained phase domain sizes are in good agreement with the size that were determined with SEM (see above). Table 5. Comparison between the transparency of the PVP/PDMS-IPN samples and the measured size of the PVP phase domains.

Technique Transparency of IPN (%) Size of PVP domains (nm)

TEM 58 250-350 SEM 82 180 SEM 13 350 AFM 0 ca 1000 The domain sizes that were obtained using the three different microscopy techniques are summarized in Table 5 and the domain sizes were correlated with the transparency of the analyzed PVP/PDMS-IPN sample, i.e., smaller domain sizes resulted in a higher transparency. Furthermore, there appears to be no difference in the measured domain sizes by analyzing the dry or wet PVP/PDMS-IPN samples, although the measured PVP domain sizes that were obtained for the wet IPN sample with TEM should be slightly overestimated due the swelling of PVP domains in water. In addition, the TEM analysis requires significant more efforts during the sample preparation for analysis of wet samples in comparison to the sample preparation in dry state of the IPNs for the AFM and SEM analysis.

46

3.2 Drug loading in mesoporous silica (Articles III and IV)

3.2.1 Loading of ibuprofen into mesoporous silica using different solvents

Ibuprofen was loaded in mesoporous silica with liquid carbon dioxide as the loading solvent. The solubility limit of ibuprofen in liquid carbon dioxide was determined to 0.20-0.25 wt-% at room temperature and 55 bar. However, in the loading experiments of ibuprofen into mesoporous silica, an ibuprofen amount above the solubility limit (nominal concentration 0.34 wt-%) was used to study the effect of the cosolvent addition for the loading capacity of ibuprofen in mesoporous silica, see Figure 21. The maximum adsorbed amount of ibuprofen in mesoporous silica, 300 mg ibuprofen/g SiO2, was achieved when using liquid carbon dioxide without cosolvents and this amount was in agreement with other studies of ibuprofen loading in a similar type of mesoporous silica material [112]. It was found that the equilibrium time needed to reach the plateau value of ibuprofen adsorption in mesoporous silica was in the order of 8-10 hours, see Figure 21.

0

100

200

300

400

500

0 5 10 15 20

CO2CO2 + 5 mol-% cyclohexaneCO2 + 5 mol-% acetoneCO2 + 5 mol-% methanol

Ads

orbe

d am

ount

of I

bupr

ofen

in S

iO2 /

mg/

g Si

O2

Loading time / h

Figure 21. The adsorbed amount of ibuprofen in mesoporous silica when using CO2 (l) with/without 5 mol-% cosolvent as loading solvent. The nominal ibuprofen concentration in CO2 (l) was 0.34 wt-%.

47

From N2-sorption measurements and calculation of the BET-area it was established that the BET area decreased from 1106 m2/g of the empty mesoporous silica particles down to 62 m2/g for an ibuprofen-loaded sample of mesoporous silica (containing 302 mg ibuprofen/g SiO2, which corresponds to the maximum obtainable ibuprofen loading). Accordingly, the pores of the mesoporous material were not 100 % loaded with ibuprofen, as the expected (outer) geometric surface of totally pore-filled silica spheres would be of the order of 1 m2/g. This finding is likely explained by sterical factors hindering ibuprofen to pack efficiently in the narrow SiO2 pores and some empty space in the pores are expected after the loading procedure of ibuprofen. Adding a cosolvent with similar solubility parameter, i.e., similar polarity as liquid carbon dioxide does not affect the loading capacity of ibuprofen in the mesoporous material to a large extent, only a small decrease in loading capacity was observed, see Figure 21. Moreover, when adding a cosolvent with higher solubility parameter (=higher polarity), like acetone or methanol, the loading capacity of ibuprofen decreased significantly. One explanation for this behavior is the fact that the solubility of ibuprofen is increased by the addition of a polar cosolvent in liquid carbon dioxide, which it becomes less favorable for ibuprofen to adsorb in the silica pores. In addition to the increased solubility in the presence of polar solvents, the molecular structure of ibuprofen contains a carboxyl group and can form hydrogen bonds with silanol groups in the pores of mesoporous silica. However, also the cosolvents with higher solubility parameters, acetone or methanol can also form hydrogen bonds with the silanol groups and there will be a competition between ibuprofen and the cosolvents. Furthermore, the cosolvent molecules are smaller and present in larger amount than the ibuprofen molecules, which favors adsorption of cosolvents in the silica pores, instead of ibuprofen adsorption.

48

CO2

CyclohexaneMethanol

0

50

100

150

200

250

300

0 10 20 30 40 50 60

Adso

rbed

am

ount

of i

bupr

ofen

in S

iO2 (m

g/g)

Concentration (mM)

Figure 22. Adsorption isotherms for ibuprofen/mesoporous silica, loading in CO2 (l), cyclohexane or methanol. A series of adsorption isotherms for ibuprofen loading in mesoporous silica using three different solvents were measured and are shown in Figure 22. Using liquid carbon dioxide or cyclohexane resulted in a high loading capacity of ibuprofen in mesoporous silica but two important differences should be remarked:

i) liquid carbon dioxide reached a higher loading capacity of ibuprofen at a lower ibuprofen concentration than loading with cyclohexane, and

ii) a plateau value of the adsorption isotherm of ibuprofen in liquid carbon dioxide is not reached since the solubility limit is 7.3 mM. Instead, the adsorption isotherm has a more or less linear shape, all the way up to the maximal loading limit.

When loading ibuprofen in mesoporous silica with a polar solvent like methanol, a very high ibuprofen concentration would be necessary to reach the equilibrium plateau value of 300 mg/g and this was in agreement with the results of using liquid carbon dioxide and a polar cosolvent as loading solvent, as discussed above.

49

Interestingly, it can be seen in Figure 22 that a rather low ibuprofen concentration is sufficient for reaching the maximum loading capacity when liquid carbon dioxide or cyclohexane were used as solvent. To quantify this feature, a so called Enrichment Factor (EF) was defined:

evap

ads

AA

=EF (Eq.9)

where Aads (mg/g) is the adsorbed amount of ibuprofen in the mesoporous silica (measured with TGA) and Aevap (mg/g) is the amount of ibuprofen that would be found in the mesoporous silica after condensation of the ibuprofen solution in the absence of any active enrichment of ibuprofen in mesoporous silica and Aevap is calculated by:

SiO2bulkevap cA φ⋅= (Eq.10)

where cbulk is the concentration of ibuprofen in the solution (mg/ml) and φSiO2 is the pore volume of the mesoporous silica particles (cm3/g). The enrichment factor is a distribution coefficient and Aevap is normally small in comparison to Aads in the concentration ranges considered here. In Figure 23, the enrichment factors for the three different solvents were plotted versus the ibuprofen concentration. The upper x-axis in the figure is Aevap, i.e., corresponding to EF=1 and shows that a low concentration of ibuprofen is deposited in the SiO2 pores by evaporation of remaining solvent since the bulk concentration of ibuprofen in the solvent is low. The highest enrichment factor (EF=1111) was obtained for the loading of ibuprofen in mesoporous silica using the lowest concentration of ibuprofen in cyclohexane. However, this system is likely of less important technical relevance since the pores of the silica material are not loaded to the maximum level. From a technical point of view it is more interesting to compare the enrichment factors for the different loading solvents that resulted in maximum loading of ibuprofen in the mesoporous silica. For such comparison, the highest enrichment factor was obtained for loading ibuprofen with liquid carbon dioxide (EF=235) and a lower enrichment factor was obtained for loading in cyclohexane (EF=35). These points are marked with dotted lines in Figure 23.

50

● CO2

□ Cyclohexane

Methanol

Cyclohexane

0

200

400

600

800

1000

1200

0 10 20 30 40 50 60

0 2 4 6 8 10

Enric

hmen

t fac

tor

Concentration (mM)

Aevap

(mg/g)

CO2

● CO2

□ Cyclohexane

Methanol

Cyclohexane

0

200

400

600

800

1000

1200

0 10 20 30 40 50 60

0 2 4 6 8 10

Enric

hmen

t fac

tor

Concentration (mM)

Aevap

(mg/g)

CO2

Figure 23. The calculated enrichment factors for ibuprofen/mesoporous silica samples using three different loading solvents: CO2 (l), cyclohexane or methanol. The upper x-axis is the amount of ibuprofen that will be deposited in the SiO2 without any enrichment, i.e., EF=1. The dotted lines are the threshold for maximally ibuprofen-filled mesoporous silica loaded with CO2 (l) and cyclohexane (see arrows). The enrichment factors in the investigated ibuprofen concentration range for loading ibuprofen with methanol are lower (EF=0-39) than the enrichment factors in liquid carbon dioxide or cyclohexane and this is a consequence of the competitive adsorption of ibuprofen and methanol in the mesoporous silica particles. Consequently, by using the enrichment factor as a measure of the efficiency of ibuprofen loading into mesoporous silica it was shown that liquid carbon dioxide has the largest potential as a loading solvent of the three evaluated solvents. Using the enrichment factors in a wider perspective for other types of systems, a loading solvent should be selected for a particular system that generates a high enrichment factor at the maximum degree of loading in a porous carrier material in order to minimize the amounts of the substance that are needed to be added to the bulk solution.

51

Table 6. Enrichment factors (EF) for experimental results (loading solvents: CO2 (l), cyclohexane or methanol) and calculated from the literature using different loading solvents. The concentration of ibuprofen in the solvent and the adsorbed amount of ibuprofen in mesoporous silica are listed in the table.

Concentration of ibuprofen (mM)

Adsorbed amount of ibuprofen (mg/g)

Enrichment factor

Nonpolar solvents CO2 (l) 7.3 284 235 Cyclohexane 50.6 295 35 Hexane [113] 315.1 590 15 Hexane [82] 160.0 300 10 Cyclohexane [112] 151.0 248 9 Polar solvents Methanol 40.4 63 10 DMA [113] 315.1 0 0 DMF [113] 315.1 47 1 Ethanol [113] 315.1 184 5 DMSO [113] 315.1 26 1

The above results have pointed at the possibility to reduce the amount of ibuprofen in the solution for loading ibuprofen into mesoporous silica and simultaneously reaching the maximum loading level of ibuprofen in the porous material. Several other literature studies have also focused on loading ibuprofen in mesoporous silica with different organic solvents but generally a much higher ibuprofen concentration in the solution was used. The results from the calculated enrichment factors for different loading solvents are presented in Table 6. The enrichment factors (EF) calculated from these literature values were significantly lower than those that were obtained using liquid carbon dioxide as the loading solvent but this is mostly an effect of the higher concentration of ibuprofen that were used in those systems. This strengthens the potential use of liquid carbon dioxide for the present system and liquid carbon dioxide is superior compared to organic solvents since no solvent residues are left in the final material when using liquefied gases/supercritical fluids, which is a primarily important aspect for pharmaceutical applications.

52

3.2.2 Characterization of amorphous ibuprofen

Generally, it is preferred to use the amorphous form of a drug molecule to obtain enhanced bioavailability for the drug in a drug formulation rather than its crystalline form [114]. In the narrow pores (average diameter 2.5 nm) of the mesoporous silica material it is impossible for ibuprofen to be stored in its crystalline state since crystallization is sterically hindered [115]. This was confirmed by analyzing ibuprofen-loaded mesoporous silica material with XRPD and the diffractogram for different samples are summarized in Figure 24.

Figure 24. XRPD data for characterization of the crystallinity of ibuprofen in different samples: a) mesoporous silica, b) 177 mg ibuprofen/g SiO2 loaded in CO2 (l), c) 150 mg ibuprofen/g SiO2, crystalline ibuprofen powder mixed with mesoporous silica, d) ibuprofen treated in CO2 (l) and e) crystalline ibuprofen. Ibuprofen (used as received) in the crystalline form exhibited several crystalline peaks and the treatment of ibuprofen in liquid carbon dioxide did not affect the crystallinity of ibuprofen in the XRPD pattern of ibuprofen. On the other hand, in a sample where ibuprofen was loaded into the mesoporous silica material with liquid carbon dioxide, there were no crystalline peaks in the XRPD pattern apart from those peaks arising from the mesoporous silica material itself. In addition, in a sample with a physical mixture of crystalline ibuprofen powder and mesoporous silica with the same proportions as the

53

sample where ibuprofen was loaded into the silica particles with liquid carbon dioxide was also analyzed. Several crystalline peaks of ibuprofen in the diffractogram were obtained in this sample. This allows us to conclude that ibuprofen was in fact present in the silica pores and that it was stored in an X-ray amorphous form.

Figure 25. Raman spectra of a) crystalline ibuprofen, b) the surface of a mesoporous silica particle loaded with ibuprofen (140 mg/g SiO2) in CO2 (l) and c) the centre of the same particle for investigation of the ibuprofen distribution throughout the mesoporous particles. An additional analytical technique, confocal Raman microscopy, was also employed to establish that ibuprofen was present inside the silica pores. This was performed by measuring a Raman spectrum on the surface of an ibuprofen-loaded mesoporous silica particle and measuring a Raman spectrum in the middle of the same particle, see Figure 25. The Raman spectra from these positions in the particle rendered the same type of spectra as a spectrum for crystalline ibuprofen and confirmed that ibuprofen was distributed throughout the pores of the mesoporous particles. Unfortunately, due to the low lateral and vertical resolution from the confocal Raman microscopy, it was not possible to deduce a more exact position of the distribution of the ibuprofen molecules in the SiO2 particle. In addition, SEM-images (see Figure 26) of empty mesoporous silica particles (for reference) and particles loaded with ibuprofen looked similar and the ibuprofen-loaded particles did not reveal the presence of any large crystalline

54

ibuprofen flakes outside the mesoporous particles. Thereby, the SEM-images further supported the finding that ibuprofen was present inside the pores of the material, as pointed out above also by results obtained from XRPD and confocal Raman microscopy.

Figure 26. SEM images of a) mesoporous silica (reference sample) and b) 177 mg ibuprofen/g SiO2 loaded with CO2 (l) to confirm that no crystalline ibuprofen powder was present outside the SiO2 particles. Both materials display only the spherical particles, with the morphology typical for the mesoporous silica.

3.2.3 Release studies of ibuprofen

The release studies of ibuprofen from the mesoporous silica were performed in water and the release curve of ibuprofen from a sample loaded in liquid carbon dioxide is shown in Figure 27. The samples loaded in cyclohexane or methanol showed the same type of release curve as the ibuprofen/mesoporous silica sample loaded with liquid carbon dioxide presented in Figure 27 but these samples are not included in the graph. There was a marked difference of the release behavior of ibuprofen from the mesoporous particles and dissolution of crystalline ibuprofen, where the complete release of ibuprofen from the mesoporous particles was reached already within a couple of minutes irrespective of the used loading solvent, while the complete dissolution of crystalline ibuprofen was not reached until after 20 minutes. Moreover, the commercial product of ibuprofen, with trademark name Ipren, showed the same release profile in water as the crystalline form of ibuprofen. The results from the release studies demonstrated that it would be favorable to use mesoporous silica as a drug carrier for rapid release of ibuprofen since the dissolution of the same amount of ibuprofen in crystalline form was much slower. A numerical model for the dissolution behavior of ibuprofen was developed to calculate the intrinsic dissolution rate of the different samples.

55

0

20

40

60

80

100

0 200 400 600 800 1000 1200

Ibup

rofe

n (%

)

Time (s)

Ipren

Crystalline ibuprofen

266 mg ibuprofen/g SiO2 (CO2)

0

20

40

60

80

100

0 200 400 600 800 1000 1200

Ibup

rofe

n (%

)

Time (s)

Ipren

Crystalline ibuprofen

266 mg ibuprofen/g SiO2 (CO2)

Figure 27. The dissolution in water of crystalline ibuprofen and the commercial product Ipren and the release of ibuprofen in water from mesoporous silica loaded with CO2 (l). All samples contained 1.8 mg ibuprofen. Briefly, the numerical model assumed release/dissolution of ibuprofen from a spherical continuum and no diffusion restrictions were present in the system. The input data to the model was the particle size distribution of the samples (mesoporous silica and crystalline ibuprofen). From the model, the release or dissolution of ibuprofen versus time could be simulated by calculating the weight of the particles in each size range from the particle size distribution of the material and summing up the total weight of the particles. The linear intrinsic dissolution rate (nm/s) is the linear shrinkage rate of the diameter of a sample and this constant was iterated for the samples until the calculated curve from the model was fitting to the experimental data for the release or dissolution of a sample versus time. The linear intrinsic dissolution rate was converted into the intrinsic dissolution rate (mg/cm2·s), which is commonly used in the literature. Further details about the numerical model for the intrinsic dissolution rate can be found in Article IV. The particle size distribution of crystalline ibuprofen was wide, from the micrometer size up to 150 μm, in comparison to the amorphous ibuprofen in the mesoporous silica. Nevertheless, the different samples were found to possess a similar intrinsic dissolution rate of 55 nm/s or 0.0061 mg/cm2·s, as calculated from the numerical model.

56

Figure 28, shows two samples that were evaluated with the numerical model for the intrinsic dissolution rate of ibuprofen in mesoporous silica and crystalline ibuprofen and the curve fitting of the model to the experimental data was acceptable for both samples. Conclusively, the samples have the same intrinsic dissolution rate and the difference in release rate for samples with amorphous or crystalline ibuprofen is only an effect of the difference in exposed surface area of the samples. In other words, the surface area of the larger ibuprofen crystals was smaller than the surface area exposed for ibuprofen-loaded mesoporous silica.

0

20

40

60

80

100

0 200 400 600 800 1000 1200

Exp.dataCalc. dataExp. dataCalc. data

Ibup

rofe

n (%

)

Time (s)

Figure 28. The calculated data from the numerical model of the intrinsic dissolution rate for the dissolution behavior of ibuprofen was compared with the experimental release profile (266 mg ibuprofen/g SiO2), loaded with CO2 (l) and the dissolution profile of crystalline ibuprofen. Both samples have the same intrinsic dissolution rate (55 nm/s or 0.0061 mg/cm2·s). Although it was confirmed from XRPD that ibuprofen was stored in an amorphous state in the mesoporous silica there are most certainly quite strong physical interactions (hydrogen bonds) between ibuprofen molecules and silanol groups in the SiO2 pores. The above findings that the intrinsic dissolution rates of crystalline ibuprofen and ibuprofen loaded in mesoporous silica are similar suggest that the strength of the physical interactions between silica and the ibuprofen-loaded into the mesoporous silica are of the same order as the bonding strength between individual ibuprofen molecules in bulk crystalline ibuprofen.

57

Besides, strong interactions are likely absent between individual molecules in the bulk amorphous ibuprofen and this is expected to give bulk amorphous ibuprofen a higher intrinsic dissolution rate than crystalline ibuprofen. It shall be pointed out, and concluded, that although the ibuprofen will be present in amorphous form when loaded into mesoporous silica, it does not have to (and does not) behave like a bulk sample of amorphous ibuprofen. This is a conclusion in this thesis, which may also hold true for other drugs. However, technically for the development of tablet formulations, the incorporation of ibuprofen in mesoporous silica may be advantageous as such a formulation strategy will ensure that a high ibuprofen area is exposed for dissolution. This is important since the dissolution is the rate limiting factor for a quick action of ibuprofen in vivo rather than the drug absorption, which approaches 100 % [116]. The incorporation of ibuprofen in mesoporous silica is thereby a promising alternative to the already available techniques for the manufacture of ibuprofen tablets. A consequence of the findings discussed above, is that if the particle size of crystalline and amorphous ibuprofen would be of the same order, then the dissolution profiles also would be similar. Hence, also for crystalline ibuprofen, if finely grounded, the release would reach the plateau level in a couple of minutes. This was verified by using the numerical model that was developed in this study and keeping the linear intrinsic dissolution rate for ibuprofen constant (at 55 nm/s as determined from the experiments) and varying the particle size (monomodal distribution) until the calculated curve fitted the release curve for ibuprofen in mesoporous silica. The outcome of this calculation showed that a particle size of 6 μm for crystalline ibuprofen would give a release of ibuprofen which would be as rapid as that found for ibuprofen loaded in mesoporous silica. Interestingly and logically, 6 μm is just slightly larger than the actual size of the mesoporous silica particles which is typically around 3-5 μm.

58

4. Conclusions

In this thesis, the influence of solvent choice on two different material treatment applications has been investigated. In the first application, a hydrophilic silicone rubber material with tunable transparency has been synthesized. In the other application, liquid carbon dioxide has been used as a loading solvent for adsorption of a poorly soluble drug, ibuprofen, in mesoporous silica. The main conclusions from this thesis work are: − The choice of solvent is decisive if aiming at a process for polymerization of

the PVP/PDMS-IPN in a PDMS matrix. − The selection of the solvent for the two-step method of the

PVP/PDMS-IPN is correlated with the solubility parameter concept and it was found that the solvent has a pronounced effect on the resulting material properties of the IPN, i.e., the wettability and the transparency of the IPN material.

− A certain amount of PVP in the PVP/PDMS-IPN is necessary to obtain a hydrophilic material.

− By using a two-step method for the synthesis of the PVP/PDMS-IPN it is possible to prepare a polymeric material, which in the dried state is hydrophobic but reversibly turns hydrophilic when immersed in water.

− The obtained PVP/PDMS-IPNs were phase separated on a molecular level and the size of the PVP phase domains varied with the solvent used for polymerization of the IPN. Using a solvent with a low solubility parameter, like hexane, resulted in an IPN with PVP phase domains in the micrometer range, while using a solvent with a high solubility parameter, like 1-propanol, resulted in an IPN with phase domains of a few hundred nanometers. Thus, the samples polymerized in a good solvent for PVP, gave smaller scattering units in the IPN, which in turn resulted in higher transparency.

− It has been demonstrated that ibuprofen can be enriched in mesoporous

silica in the presence of a loading solvent and that the adsorbed amount of ibuprofen in the pores can be tuned by the solvent selection. In polar solvents, which can form hydrogen bonds with silanol groups in the mesoporous silica, the amount of ibuprofen loaded into the mesoporous silica is low. The use of nonpolar solvents, on the other hand, yielded a

59

higher amount of adsorbed ibuprofen in the mesoporous silica, eventually leading to a maximum loading plateau.

− Liquid carbon dioxide was successfully used as a loading solvent for ibuprofen in the mesoporous silica and the amount of ibuprofen in the silica pores reached the maximum adsorbed level for filling the mesoporous silica material at a very low concentration (7 mM) of ibuprofen in liquid carbon dioxide.

− It was confirmed that ibuprofen contained in the mesoporous silica material was in an X-ray amorphous state.

− The release of ibuprofen from the mesoporous silica particles in water is rapid. Within three minutes after immersion in water, all ibuprofen from the mesoporous silica was released. This is advantageous for drug delivery applications where a quick dissolution of the drug in vivo is desirable, for example in formulations with analgesic drugs.

− Calculations of the intrinsic dissolution rate of crystalline ibuprofen and ibuprofen loaded in mesoporous silica showed that the rate was similar for the two types of samples. This is explained by the fact that the intermolecular interactions present between ibuprofen and silanol groups in the mesoporous silica are similar to the interactions present in crystalline ibuprofen. Ibuprofen loaded in mesoporous silica does not increase the intrinsic dissolution rate compared to the intrinsic dissolution rate of crystalline ibuprofen but a much larger surface area is exposed when the drug is loaded in the mesoporous silica. This result may be valid for other drugs loaded into mesoporous silica materials.

60

5. Future work

The two-step synthesis of the PVP/PDMS-IPN developed in this work has shown to possess the potential to tune the hydrophilicity of silicone rubber, which is important for biomedical applications. Additionally, the transparency of the IPN can be tuned by the selection of the solvent during the synthesis procedure. Nevertheless, further optimization of the synthesis procedure of PVP/PDMS-IPN is required before the material can be considered to be used in commercial products. The amounts of photoinitiator and crosslinker were used at rather high concentrations in the two-step method and these amounts should be optimized. Rather long UV-exposure time to reach a fully polymerized material was also required for the IPN synthesis and in a time- and cost-effective process, the time for polymerization should be as short as possible and the time for polymerization could possibly be reduced by using an alternative UV-curing lamp with higher UV-intensity or selection of a more effective photoinitiator than the one presently used in the two-step method. It would be very interesting to use the two-step IPN synthesis method for combination of other incompatible polymer pairs that show a large degree of phase separation upon simple mixing the polymers. The easiest starting point is the combination of silicone rubber with other hydrophilic polymers and this type of material would possibly also find potential applications within the biomedical field. The importance of the synthesis procedure for biomedical applications would be even higher if liquid or supercritical carbon dioxide can be used as a solvent for synthesis of the material, because of the health and environmental aspects of liquid carbon dioxide. Drug loading using liquid carbon dioxide of other type of molecules with other surface interactions than ibuprofen through adsorption in mesoporous silica should be highly relevant for the pharmaceutical industry. Carbon dioxide as a solvent has some unique features, making it in some aspects to behave different than hydrocarbons, although liquid carbon dioxide and hydrocarbons have low solubility parameters. The solubility of drug molecules in liquid or supercritical carbon dioxide is generally low and may be the limiting factor for using liquid or supercritical carbon dioxide in a wider aspect for loading drug molecules in mesostructured silica material. However, it should be remarked that solubility studies of drug molecules in liquid or supercritical carbon dioxide is not

61

exhaustive and further studies within this area are required and there may exist several drug candidates with sufficient solubility in liquid or supercritical carbon dioxide, which potentially can be loaded into the mesostructured silica materials. To overcome the issues of the low solubility of drug molecules in liquid or supercritical carbon dioxide, addition of cosolvents to the solution would enhance the solubility of the drug and this would potentially generate a high adsorption of the drug molecules in the mesoporous material. But any addition of cosolvents has to be carefully selected to avoid competitive adsorption between the cosolvent and the drug onto the silanol groups in the mesoporous silica rather than adsorption of the drug molecule in the SiO2 pores, like the behavior that was observed for the cosolvent addition of methanol or acetone in the ibuprofen/mesoporous silica system. Finally, to combine the synthesized PVP/PDMS-IPN and the drug-loaded mesoporous silica by integrating the ibuprofen-loaded mesoporous silica particles into the PVP/PDMS-IPN matrix is an interesting approach and could render the possibility to utilize the IPN loaded with mesoporous particles as a transdermal patch for the release of drugs.

62

6. Acknowledgements

Many people have helped me in different ways throughout my Ph D project. I would especially like to acknowledge the following people: Of course a lot of credit goes to my supervisors. First and foremost, Prof. Bengt Kronberg at YKI, who has always showed a never-ending support during the project. It has been very inspiring with your enthusiasm and deep scientific knowledge. Most importantly you have encouraged me throughout the project when things did not progress well and when I seriously doubted in myself. Dr. Martin Andersson, who has been my supervisor at YKI during the last two years of the project. You managed to quickly get committed to the project and you have always showed a genuine interest and curiosity in the project. Big thanks for all the help regarding patent issues and for being an excellent problem-solver. It has been great fun working with you! Ass. Prof. Jan van Stam, my supervisor at Karlstad University, even though there is some distance between Stockholm and Karlstad, you have been a valuable support with various matters throughout the project. Many thanks for your input to the articles and the thesis during your holiday. Dr. Joachim Karthäuser, formerly at AGA Gas/Linde, is thanked for introducing the project, for stimulating discussions during the IPN-work and for letting me spend the time in the AGA laboratory. Dr. Stefan Wolf at AGA Gas/Linde is also acknowledged for discussions and input to the project. Anders Marcusson is thanked for all the help with various practical matters in the AGA laboratory and especially for designing and building the “Mulder”-reactor and the “Mango”-reactors for the liquid CO2-experiments. I would also like to thank all the other people that I have met during my work in the AGA laboratory: Josefine, Kenneth, Esko, Stefan, Julia and Mahsa. Thanks for your company and all the nice lunches at Lidingö! My two mentors at YKI, Dr. Karin Persson and Dr. Katrin Danerlöv – I am very grateful for all your advise and support not directly linked to scientific matters you have given me. My coauthors, Prof. Jan Skov Pedersen, Ass. Prof. Maud Langton and Annika Altskär, are gratefully acknowledged for the fruitful cooperation.

63

I would like to express my gratitude towards Prof. Per Claesson for discussions about scattering measurements and for the help with the SAXS measurements during your stay at Aarhus University. Dr. Robert Corkery is thanked for valuable scientific discussions covering everything from IPNs to mesoporous materials. Annika Olsson is acknowledged for the help with Photoshop issues that arose when preparing the images for the IPN articles. Britt Nyström is also remembered for quickly finding literature references not available on the internet. Rodrigo Robinson is thanked for his qualified help with the ESEM-EDX measurements on the IPNs. Isabel, Malin, Sarah, Maria, Nina, Andreas, Carina and Christian and the rest of the past and present Ph D students at YKI are thanked for your great friendship and for all the good times together! Especially a big hug to Malin for your great company and for being patient with all my questions related to ibuprofen and mesoporous particles. My present room mates in the office, Eva and Johan - thanks to you I have taken some coffee-breaks (or rather tea-breaks) during the writing of the thesis and you are both cheering up our office although I am not always joining you for the lunches at Quantum... Big thanks to the rest of the people at YKI for making it a very pleasant place to work at. Thanks to all of you who have helped me with all kinds of things since I started working here! The financial support from the Swedish Knowledge Foundation (KK-stiftelsen) and AGA Gas/Linde is greatly acknowledged. Last but not least, thanks to Johan, Mum, Dad, Lillan and Amanda for love and support! Special thanks to my Dad for making the nice illustrations of the IPN and the synthesis of the mesoporous particles in the thesis.

64

7. References

1. C. Reichardt, Solvent and Solvent Effects in Organic Chemistry, Second Edition,

VCH Verlagsgeschellschaft mbH, Weinheim, 1988. 2. D. Stoye and W. Freitag, Eds. Paints, Coatings and Solvents, Wiley-VCH

Verlag GmbH, Weinheim, 1998. 3. G. Wypych, Ed. Handbook of Solvents, ChemTec Publishing, Ontario,

2001. 4. W. H. Nelson, Green Solvents for Chemistry, Perspectives and Practice, Oxford

University Press, New York, 2003. 5. Y. Arai, T. Sako and Y. Takebayashi, Eds. Supercritical Fluids, Molecular

Interactions, Physical Properties and New Applications, Springer-Verlag, Berlin, 2002.

6. C. A. Eckert, B. L. Knutson and P. G. Debenedetti, Nature, 1996, 383, (6598), 313-318.

7. A. I. Cooper, Advanced Materials, 2003, 15, (13), 1049-1059. 8. J. L. Kendall, D. A. Canelas, J. L. Young and J. M. DeSimone, Chemical

Reviews, 1999, 99, (2), 543-563. 9. E. J. Beckman, Journal of Supercritical Fluids, 2004, 28, (2-3), 121-191. 10. J. A. Darr and M. Poliakoff, Chemical Reviews, 1999, 99, (2), 495-541. 11. S. P. Nalawade, F. Picchioni and L. P. B. M. Janssen, Progress in Polymer

Science, 2006, 31, (1), 19-43. 12. D. L. Tomasko, H. Li, D. Liu, X. Han, M. J. Wingert, L. J. Lee and K.

W. Koelling, Industrial & Engineering Chemistry Research, 2003, 42, (25), 6431-6456.

13. N. Ajzenberg, F. Trabelsi and F. Recasens, Chemical Engineering & Technology, 2000, 23, (10), 829-839.

14. E. Kiran, Journal of Supercritical Fluids, 2009, 47, (3), 466-483. 15. M. A. McHugh, Supercritical fluid extraction: principles and practice,

Butterworth-Heinemann, Boston, 1994. 16. D. Fu and Y. Wei, Industrial & Engineering Chemistry Research, 2008, 47,

(13), 4490-4495. 17. T. Clifford, Fundamentals of Supercritical Fluids, Oxford University Press,

New York, 1999. 18. A. I. Cooper, Journal of Materials Chemistry, 2000, 10, (2), 207-234. 19. J. A. Hyatt, Journal of Organic Chemistry, 1984, 49, (26), 5097-5101. 20. H. M. Woods, M. M. C. G. Silva, C. Nouvel, K. M. Shakesheff and S. M.

Howdle, Journal of Materials Chemistry, 2004, 14, (11), 1663-1678. 21. Kemiska ämnen, http://kemi.prevent.se/, version 2009:1 22. Z. Shen, M. A. McHugh, J. Xu, J. Belardi, S. Kilic, A. Mesiano, S. Bane,

C. Karnikas, E. Beckman and R. Enick, Polymer, 2003, 44, (5), 1491-1498.

23. P. Raveendran and S. L. Wallen, Journal of the American Chemical Society, 2002, 124, (42), 12590-12599.

65

24. J. H. Hildebrand and R. L. Scott, The Solubility of Nonelectrolytes, Dover Publications, Inc., New York, 1964.

25. C. M. Hansen, Hansen Solubility Parameters A User´s Handbook, CRC Press, 2007.

26. A. Barton, CRC Handbook of Solubility Parameters and other Cohesion Parameters: second edition, CRC Press, 1991.

27. U. W. Gedde, Polymer Physics, Kluwer Academic Publishers, 1999. 28. J. C. Giddings, M. N. Myers, L. McLaren and R. A. Keller, Science, 1968,

162, (3849), 67. 29. Y. Marcus, Journal of Supercritical Fluids, 2006, 38, (1), 7-12. 30. W. S. E. S. E. W. Egon Stahl, Angewandte Chemie International Edition in

English, 1978, 17, (10), 731-738. 31. J. F. Brennecke and C. A. Eckert, AIChE Journal, 1989, 35, (9), 1409-

1427. 32. S. E. Guigard and W. H. Stiver, Industrial & Engineering Chemistry Research,

1998, 37, (9), 3786-3792. 33. K. P. Johnston, D. G. Peck and S. Kim, Industrial & Engineering Chemistry

Research, 1989, 28, (8), 1115-1125. 34. D. R. Paul and C. B. Bucknall, Polymer Blends Volume 1: Formulation, John

Wiley & Sons, Inc., 2000. 35. O. Olabisi, L. M. Robeson and M. T. Shaw, Polymer - Polymer Miscibility,

Academic Press Inc., New York, 1979. 36. D. Klempner, L. H. Sperling and L. A. Utracki, Eds. Interpenetrating

Polymer Networks, ACS, Washington D.C., 1994. 37. L. H. Sperling, Interpenetrating Polymer Networks and Related Materials,

Plenum Press, New York, 1981. 38. N. Gupta and A. K. Srivastava, Polymer International, 1994, 35, (2), 109-

118. 39. L. A. Utracki, Polymer Blends Handbook, Kluwer Academic, Dordrecht,

2002. 40. H. L. Frisch, Q. Xu and P. Zhou, Polymer Gels and Networks, 1994, 2, (3-

4), 257. 41. D. Patterson, Polymer Engineering & Science, 1982, 22, (2), 64-73. 42. P. Knaub, Y. Camberlin and J. F. Gerard, Polymer, 1988, 29, (8), 1365-

1377. 43. P. Mesquida, A. le Coutre and J. K. Kruger, Thermochimica Acta, 1999,

330, (1-2), 137-144. 44. W. Noll, Chemistry and Technology of Silicones, Academic Press Inc., New

York, 1968. 45. A. L. Smith, The Analytical Chemistry of Silicones, John Wiley & Sons, Inc.,

New York, 1991. 46. L. Liu and H. Sheardown, Biomaterials, 2005, 26, (3), 233-244. 47. C. Erbil, E. Kazancioglu and N. Uyanik, European Polymer Journal, 2004,

40, (6), 1145-1154. 48. J. S. Turner and Y. L. Cheng, Macromolecules, 2000, 33, (10), 3714-3718. 49. J. S. Turner and Y. L. Cheng, Macromolecules, 2003, 36, (6), 1962-1966.

66

50. F. Abbasi, H. Mirzadeh and A. A. Katbab, Journal of Applied Polymer Science, 2002, 85, (9), 1825-1831.

51. F. Abbasi and H. Mirzadeh, International Journal of Adhesion and Adhesives, 2004, 24, (3), 247-257.

52. M.-S. Shin, S. J. Kim, I. Y. Kim, N. G. Kim, C. G. Song and S. I. Kim, Journal of Applied Polymer Science, 2002, 85, (5), 957-964.

53. T. Miyata, J.-I. Higuchi, H. Okuno and T. Uragami, Journal of Applied Polymer Science, 1996, 61, (8), 1315-1324.

54. S. S. Hou and P. L. Kuo, Polymer, 2001, 42, (23), 9505-9511. 55. X. W. He, J. M. Widmaier, J. E. Herz and G. C. Meyer, Polymer, 1992, 33,

(4), 866-871. 56. Z. M. Liu, J. Q. Wang, G. Y. Yang and B. X. Han, Chinese Chemical

Letters, 2002, 13, (7), 683-684. 57. E. S. Barabas, in Encyclopedia of Polymer Science and Engineering, H. F. Mark,

N. M. Bikales, C. G. Overberger and G. Menges, Eds., John Wiley & Sons Inc., New York, 1989, p. 198.

58. Y. E. Kirsh, Water Soluble Poly-N-Vinylamides, Synthesis and Physicochemical Properties, John Wiley & Sons Ltd., West Sussex, 1998.

59. M. F. Refojo, in Kirk-Othmer Encyclopedia of Chemical Technology, 3rd Edition, R. E. Kirk and D. F. Othmer, Eds., Wiley, New York, 1984, p. 720.

60. I. Cifkova, P. Lopour, P. Vondracek and F. Jelinek, Biomaterials, 1990, 11, (6), 393-396.

61. C. T. Kresge, M. E. Leonowicz, W. J. Roth, J. C. Vartuli and J. S. Beck, Nature, 1992, 359, (6397), 710-712.

62. J. S. Beck, J. C. Vartuli, W. J. Roth, M. E. Leonowicz, C. T. Kresge, K. D. Schmitt, C. T. W. Chu, D. H. Olson and E. W. Sheppard, Journal of the American Chemical Society, 1992, 114, (27), 10834-10843.

63. A. Stein, Advanced Materials, 2003, 15, (10), 763-775. 64. S. Hudson, J. Cooney and E. Magner, Angewandte Chemie, International

Edition, 2008, 47, (45), 8582-8594. 65. G. Li and S. Kawi, Sensors and Actuators, B Chemical, 1999, B59, (1), 1-8. 66. A. Bibby and L. Mercier, Green Chemistry, 2003, 5, (1), 15-19. 67. A. Sayari, Chemistry of Materials, 1996, 8, (8), 1840-1852. 68. A. Taguchi and F. Schueth, Microporous and Mesoporous Materials, 2004, 77,

(1), 1-45. 69. A. Corma, Chemical Reviews, 1997, 97, (6), 2373-2420. 70. X. Feng, G. E. Fryxell, L. Q. Wang, A. Y. Kim, J. Liu and K. M.

Kemner, Science, 1997, 276, (5314), 923-926. 71. B. G. Trewyn, S. Giri, I. I. Slowing and V. S. Y. Lin, Chemical

Communications, 2007, (31), 3236-3245. 72. G. S. Attard, J. C. Glyde and C. G. Goltner, Nature, 1995, 378, (6555),

366. 73. D. Zhao, J. Feng, Q. Huo, N. Melosh, G. H. Fredrickson, B. F. Chmelka

and G. D. Stucky, Science, 1998, 279, (5350), 548-552. 74. D. Zhao, Q. Huo, J. Feng, B. F. Chmelka and G. D. Stucky, Journal of the

American Chemical Society, 1998, 120, (24), 6024-6036.

67

75. Q. Huo, D. I. Margolese and G. D. Stucky, Chemistry of Materials, 1996, 8, (5), 1147-1160.

76. C. G. Goltner and M. Antonietti, Advanced Materials, 1997, 9, (5), 431-436.

77. I. F. Uchegbu and A. G. Schätzlein, Eds. Polymers in Drug Delivery, CRC Press, Boca Raton, 2006.

78. M. Manzano, V. Aina, C. O. Areán, F. Balas, V. Cauda, M. Colilla, M. R. Delgado and M. Vallet-Regí, Chemical Engineering Journal, 2008, 137, (1), 30.

79. M. Vallet-Regi, F. Balas and D. Arcos, Angewandte Chemie, International Edition, 2007, 46, (40), 7548-7558.

80. G. Z. Bruno C. Hancock, Journal of Pharmaceutical Sciences, 1997, 86, (1), 1-12.

81. V. P. Lehto, K. Vaehae-Heikkilae, J. Paski and J. Salonen, Journal of Thermal Analysis and Calorimetry, 2005, 80, (2), 393-397.

82. M. Vallet-Regi, A. Ramila, R. P. del Real and J. Perez-Pariente, Chemistry of Materials, 2001, 13, (2), 308-311.

83. J. Andersson, J. Rosenholm, S. Areva and M. Linden, Chemistry of Materials, 2004, 16, (21), 4160-4167.

84. P. Horcajada, A. Ramila, J. Perez-Pariente and M. Vallet-Regi, Microporous and Mesoporous Materials, 2004, 68, (1-3), 105-109.

85. T. Heikkilae, J. Salonen, J. Tuura, M. S. Hamdy, G. Mul, N. Kumar, T. Salmi, D. Y. Murzin, L. Laitinen, A. M. Kaukonen, J. Hirvonen and V. P. Lehto, International Journal of Pharmaceutics, 2007, 331, (1), 133-138.

86. Q. Tang, Y. Xu, D. Wu, Y. Sun, J. Wang, J. Xu and F. Deng, Journal of Controlled Release, 2006, 114, (1), 41-46.

87. Y.-f. Zhu, J.-l. Shi, Y.-s. Li, H.-r. Chen, W.-h. Shen and X.-p. Dong, Microporous and Mesoporous Materials, 2005, 85, (1-2), 75-81.

88. M. Sörensen, Mesostructure Particulate Silica Materials with Tunable Pore Size, Synthesis, Characterization and Applications, Ph D Thesis, Royal Institute of Technology, 2009.

89. S. Wang, Microporous and Mesoporous Materials, 2008, 117, (1-2), 1-9. 90. T. Azaies, C. Tourne-Peteilh, F. Aussenac, N. Baccile, C. Coelho, J.-M.

Devoisselle and F. Babonneau, Chemistry of Materials, 2006, 18, (26), 6382-6390.

91. S. R. Allada, Industrial & Engineering Chemistry Process Design and Development, 1984, 23, (2), 344-348.

92. J. E. Mark, Physical Properties of Polymers Handbook, Americans Institute of Physics, Woodbury, 1996.

93. J. Brandrup and E. H. Immergut, Polymer Handbook, Third Edition, John Wiley & Sons, New York, 1989.

94. N. Andersson, P. C. A. Alberius, J. Skov Pedersen and L. Bergström, Microporous and Mesoporous Materials, 2004, 72, (1-3), 175-183.

95. J. Goldstein, D. Newbury, D. Joy, C. Lyman, P. Echlin, E. Lifshin, L. Sawyer and J. Michael, Scanning Electron Microscopy and X-ray Microanalysis, Third Edition, Springer, New York, 2003.

68

96. D. B. Williams and C. B. Carter, Transmission Electron Microscopy - A Textbook for Materials Science I (Basics), Plenum Press, New York, 1993.

97. S. N. Magonov and M.-H. Whangbo, Surface Analysis with STM and AFM - Experimental and Theoretical Aspects of Image Analysis, VCH Verlagsgeschellschaft mbH, Weinheim, 1996.

98. E. Smith and G. Dent, Modern Raman Spectroscopy - A Practical Approach, John Wiley Sons Ltd., West Sussex, 2005.

99. D. A. Skoog, F. J. Holler and T. A. Nieman, Principles of Instrumental Analysis, Fifth Edition, Saunders College Publishing, Orlando, 1998.

100. O. Glatter and O. Kratky, Small Angle X-ray Scattering, Academic Press, New York, 1982.

101. E. Lifshin, Ed. X-ray Characterization of Materials, Wiley-VCH Verlag GmbH, Weinheim, 1999.

102. P. J. Haines, Ed. Principles of Thermal Analysis and Calorimetry, The Royal Society of Chemistry, Cambridge, 2002.

103. G. Höhne, W. Hemminger and H.-J. Flammersheim, Differential Scanning Calorimetry - An Introduction for Practioners, Springer-Verlag, Heidelberg, 1996.

104. F. Rouquerol, J. Rouquerol and K. Sing, Adsorption by Powders and Porous Solids - Principles, Methodology and Applications, Academic Press, London, 1999.

105. R. J. Hunter, Introduction to Modern Colloid Science, Oxford Science Publications, Oxford, 1993.

106. K. Holmberg, B. Jönsson, B. Kronberg and B. Lindman, Surfactants and Polymers in Aqueous Solution, John Wiley & Sons, Ltd, West Sussex, 2003.

107. B. Kronberg, Journal of Colloid and Interface Science, 1983, 96, (1), 55-68. 108. D. Horak, M. Krystufek and J. Spevacek, Journal of Polymer Science, Part A:

Polymer Chemistry, 2000, 38, (3), 653-663. 109. Y. Y. Tan and G. Challa, Polymer, 1976, 17, (8), 739-740. 110. E. Marsano, C. Costa and E. Bianchi, Polymer, 2001, 43, (3), 1021-1024. 111. C. Whiston, X-ray Methods, John Wiley & Sons, Chichester, 1987. 112. M. H. Sörensen, Y. Samoshina, P. M. Claesson and P. C. A. Alberius,

Journal of Dispersion and Science Technology, 2009, 30, (6), 892-902. 113. C. Charnay, S. Begu, C. Tourne-Peteilh, L. Nicole, D. A. Lerner and J.

M. Devoisselle, European Journal of Pharmaceutics and Biopharmaceutics, 2004, 57, (3), 533-540.

114. C. W. Pouton, European Journal of Pharmaceutical Sciences, 2006, 29, (3-4), 278-287.

115. M. Sliwinska-Bartkowiak, G. Dudziak, R. Gras, R. Sikorski, R. Radhakrishnan and K. E. Gubbins, Colloids and Surfaces, A: Physicochemical and Engineering Aspects, 2001, 187-188, 523-529.

116. M. Klueglich, A. Ring, S. Scheuerer, D. Trommeshauser, C. Schuijt, B. Liepold and G. Berndl, Journal of Clinical Pharmacology, 2005, 45, (9), 1055-1061.

Karlstad University StudiesISSN 1403-8099

ISBN 978-91-7063-257-0

Effect of solvents during material treatment applications

Choosing the right solvent is critical for many industrial applications. A useful property for selection of solvents is their solubility parameters. This concept of solubility parameters is central to this thesis and has been used in two different case studies of material treatment applications. Silicone rubber (crosslinked poly(dimethyl siloxane), PDMS) has many favorable material properties making it useful in biomedical devices. However, a limiting aspect of its material properties is a hydrophobic surface. The aim of this work was to prepare a hydrophilic PDMS material while retaining the transparency of the material. To do this, PDMS was combined with a hydrophilic polymer, polyvinylpyrrolidone (PVP) in an interpenetrating polymer network (IPN). A two-step IPN synthesis method was developed and it was found that the solvent used for polymerization of PVP had a significant influence on the water-wettability and the transparency of the PVP/PDMS-IPN. Several different analytical techniques were used for determining the degree of phase separation in the PVP/PDMS-IPN. The second topic for which solvent effects were explored was for the use of mesoporous silica particles as potential drug delivery devices. In the present work a drug molecule, ibuprofen, was loaded into mesoporous silica particles using different solvents. The maximum loading of ibuprofen in the mesoporous material was achieved when using a nonpolar solvent, in particular liquid carbon dioxide was successfully used. One of the advantages of using liquid carbon dioxide is that no solvent residues are left in the final material, which is important for pharmaceutical applications. Furthermore, it was concluded that ibuprofen was stored in an X-ray amorphous form in the mesoporous particles. Release studies in water showed a rapid release of ibuprofen from the mesoporous silica particles, while the dissolution of samples with crystalline ibuprofen was slower. This was verified to be an effect of a larger exposed ibuprofen area in the ibuprofen-loaded mesoporous silica particles, and it was concluded that the intrinsic dissolution rate for the samples were identical.