Diffusion of proteins inside crowded structures generated using microemulsions

Diffusion of proteins inside crowded structures generated using microemulsions Dissertation zur Erlangung des akademischen Grades eines Doktors der Naturwissenschaften (Dr. rer. nat.) im Fach Chemie an der Fakultät für Biologie, Chemie und Geowissenschaften der Universität Bayreuth vorgelegt von Ralph Neubauer geboren in Eschenbach Bayreuth, der 27. Februar 2013 Die vorliegende Arbeit wurde in der Zeit von August 2009 bis Januar 2013 an der Universität Bayreuth am Lehrstuhl für Physikalische Chemie I unter Betreuung von Herrn Prof. Dr. Thomas Hellweg angefertigt. Vollständiger Abdruck der von der Fakultät für Biologie, Chemie und Geowissenschaften der Universität Bayreuth genehmigten Dissertation zur Erlangung des akademischen Grades eines Doktors der Naturwissenschaften (Dr. rer. nat). Dissertation eingereicht am: 27.02.2013 Zulassung durch die Prüfungskommission: 13.03.2013 Wissenschaftliches Kolloquium: 28.10.2013 Amtierender Dekan: Prof. Dr. Rhett Kempe Prüfungsausschuss: Prof. Dr. Stephan Förster (Erstgutachter) Prof. Dr. Thomas Hellweg (Zweitgutachter) Prof. Dr. Jürgen Senker (Vorsitz) Prof. Dr. Werner Köhler CONTENTS CONTENTS Contents 1 Summary 9 2 Zusammenfassung 11 3 Introduction 15 4 Theoretical background 17 4.1 Microemulsion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 4.1.1 Surfactants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 4.1.2 Micelles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 4.1.3 Structure and dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 4.1.4 The bicontinuous phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 4.1.5 Curvature of the interfacial film . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 4.1.6 Length scales in microemulsions . . . . . . . . . . . . . . . . . . . . . . . . . . 22 4.2 Decontamination applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 4.3 Proteins in microemulsions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 4.4 Fluorescence correlation spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 4.4.1 Fluorescence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 4.4.2 FCS setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 4.4.3 Observation Volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 4.4.4 Diffusion of the fluorophore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 4.4.5 Triplet decay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 4.4.6 Anomalous diffusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 4.4.7 Calibration of the instrument . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 4.5 Scattering methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 4.5.1 Photon correlation spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . 34 4.5.2 Setup of the PCS experiment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 4.5.3 Small angle neutron and X-ray scattering . . . . . . . . . . . . . . . . . . . . . 38 4.5.4 Teubner Strey approximation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 4.5.5 Droplet structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 4.5.6 Neutron-spin echo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 5 CONTENTS CONTENTS 5 Experimental section 45 5.1 Used chemicals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 5.2 Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 5.2.1 Recording of phase diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 5.2.2 Proteins in microemulsions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 5.2.3 Protein fluorescence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 5.3 Experimental details of the FCS measurements . . . . . . . . . . . . . . . . . . . . . . 52 5.4 DLS measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 5.5 SANS/SAXS measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 5.5.1 SAXS setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 5.5.2 SANS setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 6 Results and discussion 57 6.1 Fluorescence measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 6.1.1 Fluorescent impurities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 6.1.2 FCS measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 6.2 The C9G2based microemulsion system . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 6.2.1 Phase behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 6.2.2 Structure sizes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 6.2.3 GFP+dynamics by FCS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 6.2.4 Microemulsion dynamics by DLS . . . . . . . . . . . . . . . . . . . . . . . . . . 74 6.3 The C12G2based microemulsion system . . . . . . . . . . . . . . . . . . . . . . . . . . 75 6.3.1 Phase behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 6.3.2 Size and shape of the microemulsion structures . . . . . . . . . . . . . . . . . 77 6.3.3 GFP+dynamics by FCS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 6.3.4 Microemulsion dynamics (FCS) . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 6.3.5 Microemulsion dynamics (DLS) . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 6.4 Other microemulsion systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 6.5 Other methods for measuring protein dynamics . . . . . . . . . . . . . . . . . . . . . 91 6.5.1 DFPase NSE measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 6.5.2 PFG-NMR measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 6 CONTENTS CONTENTS 7 Conclusion and future prospects 97 8 Danksagung 99 A Appendix 101 A.1 Alternative microemulsion systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 A.2 Solvent sensitive fluorescent dyes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 List of Figures 107 List of Tables 111 References 113 Abbreviations 125 7 1 SUMMARY 1 Summary Microemulsions are thermodynamically stable mixtures of water, oil and surfactant. In the case of microemulsions based on sugar surfactants a cosurfactant as an additional component is necessary, most often a short-chain alcohol. In this thesis mainly pure surfactants are used. The focus is on n-nonyl-β-d-maltosid and n-dodecyl-β-d-maltosid as model systems for microemulsions based on technical grade surfactants. Depending on the ratio of the components microemulsions form different structures. In the bicontinuous phase continuous oil and water domains are present separated by a surfactant/cosurfactant film. At higher surfactant amounts the bicontinuous structure passes over to a droplet phase. The size of these structures is in the scale of up to 100nm, therefore microemulsions look transparent for the human eye. Bicontinuous microemulsions are a promising carrier medium for decontamination applications. Chemical warfare agents are mainly lipophilic, in contrast degradation agents are hydrophilic. A microemulsion is able to solubilize lipophilic and hydrophilic molecules. At the interface the warfare agent is close to the degradation agent and can be eliminated. The protein diisopropyl-fluorophosphatase (DFPase) is a promising candidate for decontamination applications and is able to degrade different warfare agents. Therefore, the knowledge of the dynamics and properties of enzymes inside a bicontinuous structure is of big importance. In this thesis different microemulsion systems based on C9G2or C12G2, water, cyclohexane and 1-pentanol are systematically characterized by X-ray and neutron scattering experiments. By using an improved Green Fluorescent Protein (GFP+) as a counterpart of DFPase with a similar size, the diffusion inside a microemulsion can be studied with the fluorescence correlation spectroscopy (FCS) method. Here, fluorescent impurities in the used components are a problem. By the choice of a suitable concentration of GFP+and regarding the dynamics of the microemulsion structure, it could be shown that the protein is able to move inside the microemulsion structure. Hence, microemulsions are interesting model systems to produce crowding effects in a controlled way. This is probably the most important result of this thesis. In the C9G2system a bicontinuous phase is present. At small oil/water ratios Φwith a high water amount big water domains exist with length scales of approximately 10nm, which allows protein diffusion. With increasing oil/water ratio the protein diffusion is more and more hindered. This results in an 9 3 INTRODUCTION In the 1980s, already the German armed forces used macroemulsions for decontamination based on Marlowet IHF, tetrachlorethylene and calciumhypochloride, but they are not stable, harmful and a danger to the environment [18]. Therefore microemulsions as carrier media for degradating agents are a distinguished alternative for decontamination applications. They offer an oil and a water phase - a lipophilic and hydrophilic part in one system. Moreover, sugar surfactant microemulsions can be composed with biocompatible components which is another big advantage for decontamination purposes [19]. Furthermore, the temperature stability is an important fact for the use in different climatic regions. Studying the structure of these systems requires big efforts in theoretical considerations and in experimental activities, many publications on this research area are a consequence. Due to the nanometer scale and therefore the need of expensive methods like neutron scattering methods the research is sometimes complicated. The present thesis deals with the introduction of another technique for observing microemulsion structure and dynamics, FCS. On the one hand, the diffusion of the microemulsion structure can be measured, namely the collective breathing motion [20], on the other hand the dynamics of particles inside this structure can be investigated. The knowledge of these processes is important: The protein diisopropyl-fluorophosphatase (DFPase) is a famous candidate for decontamination applications, the assignment in a microemulsion system is a promising combination [21]. Moreover, another aim of this paper is the implementation of new microemulsion systems mainly based on pure surfactants that play a role as model systems for technical grade systems and show a higher purity, which is crucial for fluorescence methods. But FCS is not the only method to look inside nano-scale structures. Also small angle scattering (SAS) techniques like small angle X-ray scattering (SAXS) and small angle neutron scattering (SANS) are used to characterize the structure. neutron spin echo (NSE) and pulsedfield-gradient nuclear magnetic resonance (PFG-NMR) provide information on the dynamical behavior of microemulsion systems. 16 4 THEORETICAL BACKGROUND 4 Theoretical background The main concepts for microemulsions and their abilities will be treated in this chapter. Firstly, the important role of surfactants and of the interfacial film will be discussed, further several different methods are used to observe structure and dynamics of microemulsion systems. To understand these applications, the background of these techniques will be explained in the following chapters. 4.1 Microemulsion Microemulsions show interesting structures, although they look transparent for the human eye, which is caused by the small structure sizes about 50 times lower than the visible light wavelength.. But on a closer examination, they feature for example a sponge-like structure in the nanometer-scale for a suitable mass ratio of components. Moreover, the phase behaviour is very rich and shows more different structures [22], which have to be discussed in more detail (see chapter 4.1.3). 4.1.1 Surfactants Amphiphilic molecules with a hydrophobic and a hydrophilic part are called surfactants or detergents and are the base of microemulsions. Normally, the hydrocarbon residue is the hydrophobic part and the hydrophilic part determines the kind of the surfactant. Anionic surfactants, where the hydrophilic part is anionic (for example sulfate, sulfonate and phosphate), are commonly used in washing, and cleaning agents [23],cationic surfactants have a cationic hydrophilic group and are applied in cosmetics, fabric conditioners, and sanitizers [24]. In a zwitterionic surfactants molecule a positive and negative charge is present, because of the skin friendliness this detergents are used in shampoos, body care products, bath additives and household detergents [25, 26]. Furthermore, nonionic surfactants have hydrophilic polar groups without any charge (hydroxyl-, carboxyl-, ester-, amideand polyalkyl-groups). Also alcylpolyglycosides (APG c), known as sugar surfactants are part of the non-ionic detergents and are composed of renewable resources (sugar and fatty alcohol can be used as raw materials), these surfactants are the foundation of the microemulsions used in this work. The APG ccomponents are Glucose and an alkyl-residue with chain lengths of six up to 14 carbon atoms. Longer carbon chains 17 4.1 Microemulsion 4 THEORETICAL BACKGROUND implicate a high hydrophobicity and the formulation of a microemulsion is difficult, because the surfactant is not soluble in water. 4.1.2 Micelles Because surfactants are molecules with an amphiphilic ability, they are able to form micelle in solution, if a critical micell concentration (CMC) is reached. Micelles are a cluster of molecules caused by reversible aggregation. The geometrical form of such particles must not be spherelike, a micelle can also exist as cylindrical, plateor rod-like objects. The CMC is the concentration of surfactant, when the surface or the interface is completely covered with surfactant molecules, below the CMC the surface tension of the fluid is decreased, above the CMC more micelles emerge and the surfactant has no influence on the surface tension anymore. Micelles can also change macroscopic parameters like flow properties [23]. 4.1.3 Structure and dynamics Microemulsions based on sugar surfactants are composed of water, oil, sugar surfactant (CiGj) and co-surfactant (alcohol). Such systems thermodynamically stable. Different phases occur for certain mass ratios of these components. To quantify the amount of constituents, the following definitions will be used: fraction of oil: Φ=moil mw+moil (1) fraction of surfactant: γ=msurf mw+moil +msurf (2) fraction of co-surfactant: δ=mcosurf mw+moil +msurf +mcosurf (3) Due to the number of four components, the phase behavior can be described by a phase tetrahedron (compare fig. 1). When instead of a sugar surfactant a CiGj-surfactant is used, the tetraehedron changes to a phase prism, where the height corresponds to the temperature axis. For CiEj-systems, no co-surfactant is necessary to achieve a 1-phase structure. The temperature plays the role as a tuning parameter and acts like the co-surfactant in a CiGjsystem. For a cut in the phase tetrahedron at a certain oil/water ratio Φ(fig. 1) the phase behavior is illustrated by a two-dimensional graph with the axis γand δ(fig. 2). 18 4 THEORETICAL BACKGROUND 4.1 Microemulsion water oil surfactant co-surfactant Φ δ Figure 1: Phase tetrahedron of a sugar-surfactant based microemulsion with a cut for a water/oil ratio Φand the resulting Kahlweit-fish in reference to [22]. The resulting image looks like a fish, known as "Kahlweit-fish". The one phase structures are located in the fish-tail: bicontinuous, oiland water continuous, and lamellar phase [27]. The phases are also called Winsor-systems [1]: Type I with a water phase where the surfactant and oil is solubilised and an oil phase (located below the fish-body). Type II is an oil phase with solubilised surfactant and water and a water excess phase (region in the phase diagram above the fish body). X γ δ Lα Figure 2: Kahlweit-fish of a sugar surfactant system with the so called X-point, fish-head (3-phase region), fish tail (1-phase) and water excess phase below, oil excess phase above the fish, in reference to [22]. Type III consists of 3 phases, a water phase, a microemulsion phase and a oil phase (in the fishhead). In the case of type IV only one phase is present, were all components are solubilised. The lamellar region is a single phase area indicated by Lαin fig. 2. The Lα-phase and the 19 4.1 Microemulsion 4 THEORETICAL BACKGROUND bicontinuous region are separated by a two phase coexistence zone [22]. •bicontinuous phase (L3phase): Behind the X-point, sponge-like structure with oil and water domains, seperated by the surfactant/co-surfactant interface. •lamellar phase (Lαphase): Oil and water layers interrupted by the interfacial film. The Lαis a smectic type C liquid crystal. •oil in water microemulsion: Oil droplets in water with a radius of some nm, below the lamellar region. •water in oil microemulsion: Water droplets in oil, above the lamellar region. What kind of structure emerges, depends on the composition of the components (Φ,γ) and on the shape of the amphiphilic film, which is mainly determined by the alcohol content δ. surfactant co-surfactant water phase oil phase surfactant co-surfactant (a) (b) Figure 3: Sketch of a (a) bicontinuous microemulsion and (b) lamellar phase A common method to investigate the structure is measuring the conductivity of a sample [28]. In water continuous structures (oil in water) and in the bicontinuous phase the conductivity is relatively high: The water phase is continuous over the whole sample. On the other hand for water in oil phases, the conductivity is low. For lamellar structures the conductivity is lower than in the bicontinuous phase , if the conductivity is measured perpendicular to the interfacial film. However, if the conductivity can be measured in plane of the lamellae it will be higher compared to the bicontinuous phase [29]. The surfactant film and the oil phase hinders the conduction perpendicular to the lamellae. 20 4 THEORETICAL BACKGROUND 4.1 Microemulsion 4.1.4 The bicontinuous phase This phase occurs close to the X-point in the fish tail of a Kahlweit-fish phase diagram (figure 2). The structure can be imagined as a sponge-like pattern and can be visualized by freeze fracture electron microscopy (FFEM) ( [30]). The bicontinuous phase offers a huge internal interface, which is beneficial for decontamination applications. The oil and water component exist as interlaced continuous phases, separated by a surfactant/co-surfactant film. The bicontinuous structure appears preferably at equal oil/water fraction, meaning Φ=0.5 [31]. Bicontinuous phases can be detected by electrical conductivity measurements [32]and length scales can be obtained by SANS and SAXS experiments (chapter 4.1.6). The dynamic of such structures is often called “collective breathing motion” [33,34]and can be measured by NSE, dynamic light scattering (Dynamic Light Scattering (DLS)) and PFG-NMR methods. 4.1.5 Curvature of the interfacial film The interfacial film in the microemulsion systems studied here consists of a CiGjsurfactant (hydrophilic) and a co-surfactant (alcohol, hydrophobic surfactant). The curvature Hof the film depends on the amount of co-surfactant at the interface. Because of the smaller co-surfactant head group compared to the sugar surfactant, Hchanges from positive to negative values when co-surfactant is added. increasing δ H < 0 H = 0 H > 0 non-ionic surfactant co-surfactant Figure 4: Non-ionic surfactant film in a sugar surfactant microemulsion: The curvature H depends on the amount of co-surfactant δ(hydrophobic), in reference to [22]. Hcan be calculated with the local curvatures c1=1 R1and c2=1 R2, located at a point of the 21 4.1 Microemulsion 4 THEORETICAL BACKGROUND surfactant film (with R1and R2: radius of the curvatures): H=1 2(c1+c2)(4) For H>0, oil in water droplets are formed. When Hreaches negative values, the structure changes to water-in-oil droplets. If H=0, there is no curvature of the film, the Gaussian curvature Khas to be considered: K=c1·c2(5) When Kis negative, the bicontinuous structure appears, for positive values the surface is planar and the structure is lamellar. This curvature can be affected with the type and amount of the surfactant and co-surfactant, temperature and addition of polymers. In such a bicontinuous structure, the length scales can be described according to Strey [35]by the characteristical length ξ: ξ=v u t2 c2 1+c2 2 (6) In experiments it was observed that ξis approximately half the periodicity d[36], which is a value for the domain size of the oil and water domain. 4.1.6 Length scales in microemulsions Because of the structural dimensions in the nano-meter scale, a microemulsion looks transparent to the human eye. Hence, the structure cannot be observed by optical techniques. Light scattering methods would reveal the bulk-dynamics of the structure, therefore technologies with smaller wavelengths have to be used: X-ray or neutron scattering. SAXS and SANS experiments yield the dimensions of microemulsion systems (compare scattering theories in chapter 4.5). Another possibility is FFEM, for such measurements the sample has to be frozen very fast. But in a freezing process, the structure of the system might be changing and the appearing structures might not reflect the real state at normal temperatures. Dimensions in a bicontinuous structure are characterized in terms of correlation lengths ξand domain sizes d(equations 44, 45). These dimensions depend on the chain-length of the used surfactant and on the mass ratios Φand γ. For sugar surfactant systems, the correlation length is between several nanometers up to nearly 100nm. For Φvolume =0.5 the oil and water domains show a similar size - at smaller Φ, the water domain size increases and decreases for Φ>0.5. 22 4 THEORETICAL BACKGROUND 4.2 Decontamination applications 4.2 Decontamination applications Chemical warfare agents are a big danger in times when terrorist attacks are a daily issue in several regions on earth. Therefore it gets more and more important that effective decontamination media are available. A couple of features are required for a adequate decontamination system: Ability to penetrate different kinds of surfaces, nontoxic, storable, noncorrosive and cheap to fabricate. Microemulsions can be composed of various different components and offer in the suitable composition all of these preconditions. Most toxic chemicals and warfare agents have hydrophobic abilities. But in contrast the majority of degradation agents are hydrophilic. In a microemulsion, both features are present with a structure on the nanometer scale. Furthermore, a decontamination medium should be able to wet surfaces. The human skin or coatings of military equipment are hydrophobic, therefore a fluid in form of a microemulsion can penetrate such surfaces to bring the decontamination agent close to the warfare substance. Nevertheless, also for hydrophilic surfaces, the media should be capable of wetting these surfaces. Environmental compatibility is another important feature of decontamination media. On the one hand, when media are spread on military vehicles, they come also in contact with the environment. On the other hand, when media are applied on soldiers, it must not be toxic or harmful. A first approach is the composition of biodiesel based microemulsions [37]and systems are currently developed with components approved for cosmetic applications. For a low-cost production it is important, that the necessary amount of surfactant is as small as possible. Therefore the X-point has to be shifted to small γ. Efficiency boosting by block co-polymers plays a major role for existing microemulsion systems, with only small amounts of block copolymers a much bigger amount of surfactant can be replaced [38–40]. Enzymes with decontaminating abilities solubilised in the water phase of a microemulsion are promising candidates for decontamination purposes. Especially the enzyme DFPase (see figure 5) of the squid Loligo vulgaris is able to detoxify nerve agents [18,21,22]. A recently presented decontamination process works as follows: In a first step the warfare agents are solubilised by the oil phase of the microemulsion, where they can get to the interface to come near the degradation agents which are solved in the water phase and the decontamination can 23 4.2 Decontamination applications 4 THEORETICAL BACKGROUND take place [41]. Figure 5: X-ray structure of DFPase of the squid Loligo vulgaris [42], with a molecular weight of 35kDa.[43] Another important ability of microemulsions is the structural stability over a huge temperature range (-20◦C to 50◦C) [44,45], which is a crucial feature for applications in the military field. The length scales of the water and oil domains stay nearly constant determined by SANS and the bending elastic constant is temperature independent measured in NSE experiments. 24 4 THEORETICAL BACKGROUND 4.3 Proteins in microemulsions 4.3 Proteins in microemulsions In the last subsection the use of proteins for decontamination applications was discussed already. Knowing the dynamics of such particles inside a bicontinuous phase of a microemulsion is important for designing efficient decontamination media. One question is, if the enzyme is moving inside e.g. the water domain or if it sticks to the surfactant interface. The dynamics of DFPase with a hydrodynamic radius of 2.34nm [34]are hard to observe with inhouse-methods like photon correlation spectroscopy. Moreover, the microemulsion structure scatters light as well and overlays the signal of the protein. Therefore, we use proteins comparable in size, which are fluorescent and can be directly observed in a more simple way with fluorescence correlation spectroscopy. Green Fluorescent Protein (GFP) is a famous protein used for many applications in science by taking advantage of the fluorescent behavior without the need of labeling (chapter 5.2.3). Besides the questions relate to the use of microemulsions as reaction medium for enzymes, it is also possible to use the microemulsion as model system for the study of sub-diffusive behavior. The domain sizes in a bicontinuous microemulsion or the volume fraction in a droplet system can be easily controlled. Hence, such a system can be used to generate crowding in a well defined way. 4.4 Fluorescence correlation spectroscopy This method exploits the fluorescence ability of molecules and is used to measure particle dynamics down to the nanometer scale [46]. The advantages are that one can decide, what dynamic is measured by specific labeling and that time ranges from microseconds up to 100ms can be covered. But not only diffusion times, also concentration can be measured by FCS. In this section first the effect of fluorescence is discussed and afterwards the theoretical backgrounds will be covered. 4.4.1 Fluorescence When a molecule is excitated into a higher energy state, the energy can be released again in different ways. The fastest process is the vibration relaxation with a half-life period of 10−12 s, then the molecule can fall back to the ground state by emitting light, which is called fluorescence [47], or it can perform inter-system crossing into a triplet state (figure 6), called phospho25 4.4 Fluorescence correlation spectroscopy 4 THEORETICAL BACKGROUND The model can be adapted to more components with different size and diffusion times, which are detected simultaneously: g2(t) = 1+1 N1+T 1−Te−t τT· N X i=1  1 (1+t τi )·Ç1+ ( r0 z0)2·t τi  (19) kiis the ratio of particle i. The diffusion times τihave to differ sufficiently from each other, to be distinguished using a fit according to equation 19, normally by a factor of 2. The higher the number of diffusing components in the equation, the more parameters influence the fit. Therefore a maximum of three components is reasonable, when other boundary conditions like particle ratio are known. 4.4.6 Anomalous diffusion In case of confinement or crowding effects, the diffusion can be anomalous. In this case, the mean square displacement σris not linear in time anymore: <s2>=σr=D·tα(20) The anomalous diffusion exponent αhas to be regarded in equation 16, that leads to [56] g2(t) = 1+1 N1+T 1−Te−t τT·1 1+ ( t τd )α·Ç1+ (S2·t τd )α(21) If α < 1, sub-diffusive behavior is observed. Smaller αmeans more confinement for the diffusing particle. In the bicontinuous phase of a microemulsion, the diffusion of a particle inside the water domains is hindered by the surfactant/co-surfactant interface. 4.4.7 Calibration of the instrument τDis the time a particle needs to pass the observation volume τD=r2 0 4D(22) With a dye, where Dis known from the literature (often fluorescein for the 488nm beampath, rhodamine 6G for 514nm beampath and rhodamine B for 534nm beampath), r0can be calculated and τDobtained by a fit of the autocorrelation curve. With r0and the fit parameters, also z0can be calculated. Then Scan be hold constant for measurements with an unknown sample [51]. 32 4 THEORETICAL BACKGROUND 4.5 Scattering methods 4.5 Scattering methods Structures and dynamics of microemulsions can be investigated using different scattering methods. In principle, scattering is the interaction of a particle with an object. Photons with different energies (or wavelengths) are scattered at objects of various size: In light scattering experiments, differences of the refractive index (for example oil ↔water) and the size of particles influence the scattered light. To resolve structures, the used wavelength has to be in the dimension of the structure size. If the wavelength is smaller (e.g. X-rays) the photon energy is bigger and photon is scattered by the molecule electrons: Therefore, in X-ray scattering applications the distribution of the electrons is crucial. Another commonly used technique is neutron scattering, neutrons have no electric charge and are able to permeate matter and interact with the nucleus of atoms. Light and X-ray scattering techniques can be accomplished at commercially available or self designed equipment. Neutron scattering experiments have to be performed at a research facility, normally the neutron source is a nuclear reactor and the neutron detection is a sophisticated and expensive technology. For X-ray scattering applications, a research facility with a synchroton source is the superior option, because the flux and the quality of the detector is better compared to inhouse-instruments. 33 4.5 Scattering methods 4 THEORETICAL BACKGROUND 4.5.1 Photon correlation spectroscopy This technique, also called Dynamic Light Scattering (DLS), exploits the ability of small particles or structures to scatter visible light. In contrast to static light scattering methods, photon correlation spectroscopy (PCS) yields information about diffusion processes. DLS is particularly suitable to measure the dynamics of macro-molecules, provides information about the diffusion coefficient and also the polydispersity of particles. ki ks ϴq Figure 11: wave vector ~q in a light scattering experiment with incoming wave ~ ks, scattered wave ~ ks and scattering angle Θ Important for light scattering experiments is the wave vector ~q(compare figure 11): q=|~q|=4πn0 λsin Θ 2(23) with: n0: refractive index of the solvent, λ: wavelength of the scattered light, Θ: scattering angle The normalized electric field autocorrelation function is given by: g1(t) = <E∗(~q,t)E(~q,t+t0)> <E(~q,0)>(24) In experiments mostly only the scattered intensity is accessible by a instrument. Thus g1(t)can not be measured directly. The intensity Iis detected timeand qdependent to calculate the autocorrelation function G2(t), which measures the correlation at the time t0with itself at a later time t+t0: G2(~q,t) =<I(~q,t+t0)I(~q,t)>=lim T→∞ 1 TZT 0 I(t+t0)I(t0)dt0(25) The autocorrelation function ca be normalized: g2(~q,t) = G2(~q,t) <I(~q,t0)>2(26) 34 4 THEORETICAL BACKGROUND 4.5 Scattering methods g1 (t) 0 0.2 0.4 0.6 0.8 1 time / s 1e-05 0.0001 0.001 0.01 0.1 1 Figure 12: Autocorrelated measurement data from a photon correlation spectroscopy measurement. Time-axis is logarithmic. The measured intensity provides no information about the electrical field ~ E, but the Siegertrelation connects the intensity autocorrelation and g1(t): g2(t) = 1+β|g1(t)|2(27) βdepends on the used measurement instrument. In a monodisperse system (e.g. spherical particles), g1(t)can be calculated by using a monoexponential function: g1(t) = e−Dq2t=e−Γt(28) The diffusion coefficient Dgives information about the particle radius. For polydisperse systems, a distribution function G(Γ)is implemented, which gives information about polydispersity: g1(t) = Z∞ 0 G(Γ)e−Γtdt (29) G(Γ)can be regarded as a sum of single contributions [57]: G(Γ) = N X i=0 aiδ(Γ−Γi)(30) applied to eq. 28: g1(t) = N X i=1 aie−Γit(31) There are different possibilities to compute the dynamics of a polydisperse system. One is the method of cumulants [58]. Another suitable approach for this problem with more species is the 35 4.5 Scattering methods 4 THEORETICAL BACKGROUND method of the inverse Laplace transformation. The CONTIN program [59],[60]exploits this procedure and can be used to analyze measured light scattering data. The output of CONTIN is a value of Γi, which depends on q. With Γ=Dq2(32) a linear regression when Γis plotted as a function of q2(qcalculated by eq. 23) leads to D. When spherical particles are observed (which is not the case in a bicontinuous structure), the hydrodynamic radius RHcan be estimated by the Stokes-Einstein equation: D=kBT 6πηRH (33) Here, ηis the viscosity. With equation 33 the hydrodynamic radius of e.g. proteins can be estimated (compare chapter 5.2.2). 36 4 THEORETICAL BACKGROUND 4.5 Scattering methods 4.5.2 Setup of the PCS experiment The PCS measurements were performed on an ALV-5000 [61]instrument. In fig. 13 the setup for photon correlations spectroscopy is shown. The light from a Nd:YVO4laser (green, 532nm) or a Helium-Neon laser (red, 633nm) passes a neutral density filter (ND filter, to adjust the laser intensity), a lens and a Glan Thompson Prism and is scattered by the sample. The sample located in a temperature controlled bath filled with toluene scatters the laser light. Nd:YVO4 M1 M2 sample lens GP correlator ALV ND filter θ GP PMT photon counter HeNe 20°C thermostat Figure 13: ALV-5000 Photon Correlation Spectroscopy setup, Mi: mirror, GP: Glan Thompson Prism, PMT: Photo Multiplier Tube Toluene is used for matching the refractive index of the glass cuvette. The scattered light is detected by a photomultiplier mounted on a goniometer, where the scattering angle Θand therefore qcan be adjusted. With a photon counter and a correlator, the signal of the PMT is registered by the computer with the ALV software. 37 4.5 Scattering methods 4 THEORETICAL BACKGROUND 4.5.3 Small angle neutron and X-ray scattering For SAS experiments, the scattering intensity and the scattering vector qare important parameters. The scattering intensity from a single particle depends on the scattered amplitude F(q): I(q) = |F(q)|2(34) F(q)is the Fourier transform of the mass distribution (SANS) or electron distribution (SAXS) ρ(r): F(q) = ZV ρ(r)e−iqr dr (35) ⇒I(q) =  1 VZV ρ(r)e−iqr dr 2 (36) In the case of Nidentical particles, the intensity is given by: I(q) = N V(ρp−ρ0)2V2 p 1 VpZV ρ(r)e−iqr dr 2 (37) with ρ0: mass/electron density of the solvent, ρp: mass/electron density of the particle, VP: particle volume, V: illuminated volume, ρ0−ρ0: scattering contrast and the form factor P(q) =  1 VpRVρ(r)e−iqr dr 2 But also interactions between the particles have to be considered, the total scattering intensity is now calculated by [62](Fk(q): electrical field contribution of particle k): I(q) = 1 V N X k=1¬Fk(q) 2¶+1 V*N X k=1 N X j=1 j6=k Fk(q)F∗ j(q)e−iq(rk−rj)+(38) When the particles are monodisperse and spherical, the intensity can be simplified to: I(q) = n¬Fk(q) 2¶       1+*N X k=1 N X j=1 j6=k e−iq(rk−rj)+       (39) I(q) = nP(q)S(q)(40) The first term P(q)(form factor) reflects the contribution of one particle, the second term, S(q) (structure factor) is based on the contribution of the particle interaction: spatial arrangement 38 4 THEORETICAL BACKGROUND 4.5 Scattering methods of the particles relative to an arbitrary origin. S(q) = 1 in a dilute solution, where the particles do not “feel” their neighbors. 4.5.4 Teubner Strey approximation SANS and SAXS scattering measurements of a bicontinuous phase show a typical broad peak when I(q)is plotted. M. Teubner and R. Strey developed a theory to describe this peak [36], based on an order expansion of the Landau free energy F, which is given by: F=Zf(ψ,∇ψ,∆ψ)d3r(41) free energy density f: f=a0+a1ψ+a2ψ2+a3ψ3+a4ψ4+··· +c1(∇ψ)2+c2(∆ψ)2+... (42) In the case of microemulsions, a2>0, c1<0 and c2>0, all other parameters =0. This yields the scattering intensity distribution: I(q) = 8π/ξ (q−¯ q)2c2/V a2+c1q2+c2q4+bkg (43) with: (q−¯ q)2: mean square fluctuation of the scattering density ρ. domain size (quasi periodic repeat distance of the oil and water domains): d 2π=  1 2a2 c21/2 −c1 4c2  −1/2 (44) and the correlation length ξ(dispersion of d): ξ=  1 2a2 c21/2 +c1 4c2  −1/2 (45) For fitting SANS and SAXS data, eq. 43 will be simplified, the constants in the numerator are summarized in k: k=8π/ξ < n2> /V(46) dividing eq. 43 by kc2: 39 4.5 Scattering methods 4 THEORETICAL BACKGROUND I(q) = 1 a2 kc2+c1q2 kc2+q4 k =1 r2+t1q2+t2q4(47) When ξis calculated similar to eq. 45, k is eliminated and the result is equal to eq. 45: ξ=  1 2r2 t21/2 +t1 4t2  −1/2 =  1 2a2k kc21/2 +c1k 4kc2  −1/2 =  1 2a2 c21/2 +c1 4c2  −1/2 (48) Hence, there are only three parameters left (with an additional background parameter) for fitting SANS and SAXS curves with eq. 47. 4.5.5 Droplet structures Droplet structures can be approximated by a fitting model of polydisperse hard spheres. According to equation 40 the scattering intensity of a sample can be calculated by regarding the structure factor and form factor. In the case of spheres, the form factor can be calculated by (η: scattering length density difference between particle and matrix, R: Radius of the spheres): P(q,R) = 4 3πR3η·3sin(qR)−qRcos(qR) (qR)3(49) In microemulsion systems the droplets are polydisperse, therefore a Schultz-Zimm SZ(R)distribution can be assumed (Ra: scaling parameter - maximum of the distribution for large k, k=1/σ2,σ. variance, Γ(k): Gamma-function): SZ(R) = N Ra R Ra k−1kkexp(−k∗R/Ra) Γ(k)(50) Regarding a monodisperse approximation for the hard sphere structure factor S(q,R)(given by a hard sphere potential which depends on the volume fraction of the spheres and the sphere radius), the intensity can be calculated by: I(q) =<P(q,R)2>S(q)(51) These calculations with the above mentioned assumptions can be done using the SASfit [63]or the GIFT [64,65]software. 40 4 THEORETICAL BACKGROUND 4.5 Scattering methods 4.5.6 Neutron-spin echo With this technique the dynamics of particles or structures can be measured based on neutron scattering [66,67]. The accessible time range is 0.001 to 250ns at a wavelength of 6 to 25 A . selector beam magnetic field 1 magnetic field 2 sample analyser polarizer detector π/2 flipper π flipper π/2 flipper Figure 14: Simplified setup of a NSE experiment in reference to [68]. A polarized neutron beam is wavelength selected and enters a magnetic field B1(length L1) which is orientated perpendicular to the neutron polarization. In the magnetic field 1 the neutrons undergo Larmor precession. After the magnetic field, the neutron spins have a specific phase depending on their velocity v. The Larmor frequency ωLis defined by: ωL=γB(52) and the neutron spin ϕ(v): ϕ(v) = ωLt=ωL1B1 v(53) The polarization Pxin the x-direction is calculated by: Px=cosϕ(54) In the case of elastic scattering in the sample, the neutrons enter the second magnetic field B2 (length L2), which is oriented antiparallel to B1. After B2, the spin phase is: ϕ=γL1B1 v−L2B2 v(55) If L1B1=L2B2, the spin phase is 0 and Px=1 (maximum value), which is called the spin echo point. When the scattering of the sample is inelastic, the kinetic energy of the neutrons changes after 41 5.2 Systems 5 EXPERIMENTAL SECTION taining the protein dynamics is neutron spin echo. But the difficulty here is the need of contrast variation. To measure only the dynamics of the protein, the waterand oil-phase, surfactant and co-surfactant have to be deuterated. Deuterated surfactants are very expensive and not available for every type of sugar surfactant. Hence, the method of choice is FCS. The components of the microemulsion don’t need to be deuterated and compared to other methods, quick measurements are possible. But the main problem withFCS is the need of a fluorescent particle. Normally, a fluorescent label is bound to the tracer particle, which should be observed. However, the ideal way of realizing the FCSexperiment is the usage of a particle, which shows fluorescence itself. This minimizes the problem of an incomplete labeling, where unbound dye-molecules are left and represent an additional dynamical component. This complicates the fitting process of the autocorrelation function. confined diffusion attachement at the interface water phase oil phase surfact. co-surf. protein fluorescent amphiphil Figure 16: Model for particles in the water domain of a bicontinuous microemulsion. Fluorescent surfactants attach at the interface, hydrophilic proteins are dissolved in the water phase. An appropriate choice for this work is the GFP (compare figure 17), particularly GFP+. This protein was found in the jellyfish Aequorea victoria, which was first extracted at 1962 [77]. GFP+ shows a 320 times higher fluorescence intensity than the wild type protein [78]. Compared to DFPase with a molecular weight of 35kDa, GFP+has approximately 27kDa, therefore the size is similar. The excitation wavelength with maximum absorption is λGFPexc =490nm, the emission maximum is λGFPem =510nm. GFP shows quantum yields up to 0.77 [79]and is stable up to 65 ◦C[80]. 48 5 EXPERIMENTAL SECTION 5.2 Systems (a) (b) Figure 17: X-ray structures of (a) Green Fluorescent Protein [75](b) mCherry Protein [76]. These molecules are used as tracer particles. An alternative available protein with similar size is mCherry from the coral “Discosoma sp” [81] with a molecular weight of ≈27 kDa [76], it shows fluorescence at higher wavelengths than GFP+(fig. 18). Therefore it can be used with the 543nm beampath. 49 5.2 Systems 5 EXPERIMENTAL SECTION 5.2.3 Protein fluorescence To characterize the used fluorescing particles, fluorescence spectroscopy measurements were performed on a Jasco FP-6500 spectrometer. In addition to normal I(λ)fluorescence intensity measurements at a fixed excitation wavelength, so called 3D-measurements are possible. At 3D-measurements different excitation wavelengths are used. The result of these measurements are depicted using a contour plot. Thereby, the maximum excitation/emission can be obtained easily. ∆λexc was 5nm in all 3D-measurements. The 3D plots (fig. 19 and 20) of the fluorescence intensity measurements for GFP+and mCherry indicate, that the optimum available FCS laser-excitation-wavelength is the Argon-Ion 488nm line for GFP+and the 543nm HeNe line for mCherry. The fluorescence spectra for the excitation with the laser corresponding wavelength are plotted in figure 18. Both proteins are slightly smaller than the DFPase used for decontamination applications. But their main advantage, the fluorescence, gives the opportunity to observe the dynamics inside the water phase of a bicontinuous phase. normalized intensity 0 0.2 0.4 0.6 0.8 1 emission wavelength / nm 500 550 600 650 700 GFP+ mCherry Figure 18: Fluorescence intensity of GFP+with an excitation wavelength of 490nm and the maximum intensity at 512nm and mCherry with an excitation wavelength of 545nm and the maximum intensity at 604nm. 50 5 EXPERIMENTAL SECTION 5.2 Systems excitation wavelength / nm emission wavelength / nm 620 600 580 560 540 520 500 480 460 normalized intensity 0 0.2 0.4 0.6 0.8 1 350 400 450 500 550 Figure 19: Fluorescence contour plot of the protein GFP+, with an excitation maximum λexc-max = 490nm and emission maximum λem−max =510nm emission wavelength / nm 700 680 660 640 620 600 580 560 540 520 normalized intensity 0 0.2 0.4 0.6 0.8 1 excitation wavelength / nm 500 550 600 650 Figure 20: Fluorescence contour plot of the protein mCherry, with an excitation maximum λexcmax =590nm and emission maximum λemmax =600nm 51 5.3 Experimental details of the FCS measurements 5 EXPERIMENTAL SECTION 5.3 Experimental details of the FCS measurements The FCS measurements were performed with the Zeiss Confocor 2 FCS-system. Usually a 0.5ml sample was prepared and 60µl were transferred into a sample holder, composed of two steele plates (one plate with a cover glass window, thickness 0.15mm), compare figure 21, already introduced from H. Zettl [82]at the same FCS-system. The plates are fixed to each other by screws and an O-ring prevents the sample from evaporating. For measurements with GFP+, the excitation wavelength was 488nm (argon-ion) and for octadecyl-rhodamin B the wavelength was 543nm (helium-neon). steel plate cover glass sample Figure 21: FCS sample chamber, composed of two steel plates with a glued on cover glass. g2(t) 1 2 3 4 5 t / μs 1 100 1e+04 Figure 22: Calibration of the 543nm beampath with the dye rhodamin B The beam path of the FCS instrument has to be adjusted routinely. Therefore, the pinhole is shifted in the x-, yand z-direction. This is done in three steps, first x, then y, then z, while the respective other values are hold constant. The intensity is measured during the pinhole is shifted and the positions yielding the maximum values are used for the measurements (compare fig. 23). 52 5 EXPERIMENTAL SECTION 5.4 DLS measurements (a) (b) (c) Figure 23: Confocor 2 pinhole adjustment with rhodamin B, (a) x-direction, (b) y-direction and (c) z-direction. The value with the maximum intensity is held constant for the measurements. To determine the radius of the observation volume, a measurement is performed with a dye of known diffusion coefficient 22, for the 488nm beampath this yields to r0=254nm and for the 543nm beampath r0=261nm. 5.4 DLS measurements For DLS measurements 1mm samples were prepared and filled in a glass cuvette which is cleaned by ethanol to avoid scattering from dust particles. Measurements are performed at eleven different angles Θin 10◦steps between 40◦and 140◦with a measurement time of three minutes per angle at a laser wavelength of 543nm. The laser intensity was adjusted to reach a sufficient countrate (≈200kHz) over the whole angular range. The obtained autocorrelation curves are treated with the CONTIN algorithm which results in different Γfor every Θ.qcan be calculated from Θand a Γ(q2)plot fitted linearly, leads to the diffusion coefficient D. 53 5.5 SANS/SAXS measurements 5 EXPERIMENTAL SECTION 5.5 SANS/SAXS measurements The structural dimensions in a microemulsion are on the nanometer scale. Therefore, only a few possibilities exist to investigate these structures. With photon correlation spectroscopy the hydrodynamic radius of geometrically simple objects like spheres would be easy to measure. But in the bicontinuous phase the structure is very complex. One possibility to observe such small structures is constituted by electron microscopy. For that the sample has to be frozen. The freezing process happens not instantly, the structure of the sample may change. Hence, a technique is necessary which enables the observation of the structure in situ. SAXS is a scattering method with a wavelength much lower than visible light, the microemulsion structures can be measured. X-ray radiation is scattered by the electron shell of the atoms. In a SANS measurement neutrons are scattered by the atom cores. Due to that, one can play with the deuteration of the sample: Either the oil or water component is deuterated, then the structures of the water/oil domains shine out (bulk dynamics), or the water and oil components are deuterated, then the surfactant interface is dominating the signal (film dynamics). Based on experience, when the water component of a microemulsion is replaced by D2O, the 1-phase boarder shifts to lower δ-values, which was already observed for CiEjmicroemulsions [83]. This circumstance has to be regarded when a sample is composed for neutron scattering. For droplet microemulsions it was already shown by Huang and Wu [84], that the replacement of H2O by D2O does not change the structure sizes of the droplets. The theoretical basis for SANS and SAXS on the bicontinuous phase was introduced by M. Teubner and R. Strey [36]. The model which describes the scattering behavior of a bicontinuous phase is depicted in 4.5.4. The scattering intensity depending on the scattering vector qpictures a broad peak, where the domain size dand correlation length ξcan be obtained by fitting the data with the Teubner-Strey equation. The fits were performed with the Qtiplot [85]software using the Nelder-Mead Simplex, which generates a sequence of simplexes with decreasing diameter and accumulate around the desired minimum [86]. 5.5.1 SAXS setup The SAXS measurements were performed on a SAXSLAB Ganesha instrument [87](fig. 24) at a sample - detector distance of 1.2m. It is an inhouse SAXS system which offers a modern 54 5 EXPERIMENTAL SECTION 5.5 SANS/SAXS measurements high resolution 2D-Pilatus 300k detector, a rotating anode, a pinhole collimation system and a multiple sample holder with automatic alignment. The X-radiation scattered by the sample in an angle of Θenters an evacuated volume, where a detector can be moved to tune the q-range and resolution. The detected 2D intensity “image” is averaged radially to obtain the intensity data depending on q. movable 2D detector x-ray source Θ beamstop sample holder evacuated tube beam Figure 24: Setup of the Ganesha SAXS system. By moving the detector the observed scattering angle Θcan be adjusted. In the sample holder more samples can be measured in a row. 5.5.2 SANS setup A neutron scattering instrument is comparable to a X-ray scattering setup. In most cases the neutron source is a scientific nuclear reactor, where several beamlines are provided with neutrons of different velocities depending on the experiment. In a small angle measurement, the distance between sample and detector is up to 40m (D11 /ILL), the desired q-range can be tuned by the distance. The setup is shown in figure 25. x/y detector vacuum sample collimation slits (adjustable) velocity selector beam shutter chopper neutron beam Figure 25: Principal setup of a SANS experiment using the example of the PAXY instrument (LLB Saclay /France, [88]) 55 5.5 SANS/SAXS measurements 5 EXPERIMENTAL SECTION Normally the intensity is measured for three different distances to obtain a good resolution over the whole q-range. The data evaluation is fundamentally comparable to SAXS scattering, the result is an intensity I(q)plot which can be approximated with an appropriate model. 56 6 RESULTS AND DISCUSSION 6 Results and discussion 6.1 Fluorescence measurements First of all, the problem of fluorescent impurities usingsugar surfactant systems will be discussed. Afterwards, the structures of microemulsion systems will be characterized by SAS measurements and finally, diffusion measurements of GFP+inside the water phase of different bicontinuous structures are shown. 6.1.1 Fluorescent impurities A central complexity in the work with microemulsions based on a sugar surfactant is the existence of fluorescent impurities. Sugar surfactants and co-surfactants exhibit a weak fluorescent activity in the visible range. FCS and fluorescence spectroscopy measurements show, that the concentration of the impurity-molecules is in the range of 10−7−10−8mol/l. This is a problem, because it’s an appropriate concentration for FCS-measurements. emission wavelength / nm 750 700 650 600 550 500 normalized intensity 0 0.2 0.4 0.6 0.8 1 excitation wavelength / nm 450 500 550 600 650 700 Figure 26: Contour plot of the fluorescence intensity of 1-pentanol in the excitation range from 400nm to 800nm without fluorescent label. In fig. 26 the fluorescence spectrum of 1-pentanol is shown. No fluorescent label was added to the sample, but still a broad fluorescent activity nearly over the whole visible range is detectable. The lines in the 3D plot reveal Raman scattering, not fluorescence. 57 6.2 The C9G2based microemulsion system 6 RESULTS AND DISCUSSION 0 0.05 0.10 0.15 0 0.05 0.10 0.15 amount of surfactant (γ) amount of co-surfactant (δ) bicont. Φ = 0.1 Φ = 0.2 Φ = 0.3 Φ = 0.4 Φ = 0.5 Φ = 0.6 Φ = 0.7 Figure 33: Phase diagrams of the C9G2system for different oil/water ratios Φ, the connection of the X-points describe a parabolic trajectory. Table 1: Composition of the C9G2microemulsion samples for SAS and FCS measurements oil/water ratio αsurfactant amount γalcohol content δ 0.1 0.08 0.035 0.2 0.11 0.055 0.3 0.12 0.075 0.4 0.14 0.10 0.5 0.12 0.105 0.6 0.12 0.145 0.7 0.11 0.17 6.2.2 Structure sizes Length scales by SANS The structure of the presented C9G2system was investigated by scattering methods applied to the samples shown in table 1. The SANS measurements were performed in Saclay/France at 64 6 RESULTS AND DISCUSSION 6.2 The C9G2based microemulsion system the LLB (Laboratoire Léon Brillouin) and in Berlin at the HZB (Helmholtz-Zentrum-Berlin). The intensity curves shown in figure 34 exhibit a structure peak for oil/water ratios of 0.3 to 0.5. The Teubner-Strey approximation is only suitable for volume fractions close to 0.5, which applies for the Φ=0.3 to 0.5 samples. Then the correlation length and domain size can be calculated from the fit parameters, discussed in chapter 4.5.4. The results are shown in table 2. I(q) 1 100 10,000 q / ÅÅ-1 0.01 0.1 0.3 0.4 0.5 Figure 34: SANS intensity measurement of C9G2microemulsions with varying Φ. The data was treated with the Teubner-Strey approximation. The fitting results are represented by the solid lines. As a consequence of the deuterated water phase, ξreflects the correlation length of the oil domains. ξincreases for rising Φ, in return, the water domains shrink (fig. 35). The domain size dstays nearly constant around 25nm, which reflects the repeating length of the oil and the water domains together. The water domain size could be estimated by subtracting the correlation length from the domain size, but then the surfactant interface is not regarded. Table 2: Structure sizes for C9G2-microemulsions determined by SANS oil/water ratio Φvolume ratio correlation length ξ/nm domain size d/nm 0.3 0.35 7.7 25.9 0.4 0.46 8.2 24.7 0.5 0.56 9.3 26.7 Also for smaller values of ΦSANS measurements were performed, the intensity curves are shown 65 6.2 The C9G2based microemulsion system 6 RESULTS AND DISCUSSION correlation length /nm 7 7,5 8 8,5 9 9,5 10 oil/water ratio Φ 0,25 0,3 0,35 0,4 0,45 0,5 0,55 Figure 35: Correlation length which results from the Teubner-Strey approximation for the C9G2 system depending on the oil/water ratio Φ. The increase with rising Φmeans growing oil domains, therefore shrinking water domains. in fig. 36. No peak arises in the I(q)plot, therefore Teubner-Strey model is not appropriate. Another way to extract information from the measurements at smaller Φwithout structure peak is to take advantage of the Guinier law: The SANS data in a double logarithmic plot follows a linear decrease at low q-values (Guinier-regime) [94]. Φ = 0.1 I(q) / a.u. 0 100 200 300 400 q / ÅÅ-1 0.01 0.1 Φ = 0.2 I(q) / a.u. 0 200 400 600 800 q / ÅÅ-1 0.01 0.1 Figure 36: SANS intensity measurement of C9G2microemulsions with variing Φ. The data cannot be fitted with the Teubner-Strey model. 66 6 RESULTS AND DISCUSSION 6.2 The C9G2based microemulsion system Table 3: Results from the scaling analysis in fig. 37 for the system C9G2with the resulting slope depending on Φ Φslope mstructure 0.1 -0.64 ellipsoidal 0.2 -0.25 ellipsoidal/spherical 0.7 -0.71 ellipsoidal/cylindrical log(I) -1 0 1 2 3 4 5 log(q) -2.5 -2 -1.5 -1 -0.5 oil/water ratio Φ 0.7 0.2 0.1 Figure 37: Double logarithmic plot of the SANS measurement of C9G2microemulsions with varying Φ. The slope in the low q-range indicates the shape of observed structure. The slope in the Guinier-regime shows, what structure is present in this microemulsion samples. Because of the oil/water ratio oil droplets inside a continuous water phase for small values of Φand water droplets inside an oil continuous phase for high values of Φoccur. The logarithmic plots of the intensity curves are shown in fig. 37. While a slope m=−1 indicates cylindrical objects, m=0 means spherical particles and a slope mof 0 to -1 implies ellipsoidal objects. In the C9G2system mis approx. −1 for Φ=0.1 and 0.7 and near 0 for Φ=0.2, compare table 3. Length scales by SAXS In addition to the SANS observations, also SAXS investigations were performed to verify the SANS results. The electron density is crucial for the scattering process of X-rays. The electron 67 6.2 The C9G2based microemulsion system 6 RESULTS AND DISCUSSION distribution is different for water domains, oil domains and the surfactant film. In contrast to neutron scattering experiments, where the deuteration is important for the scattering contrast, the electron density cannot be varied without changing the solvents and therefore changing the structure of the microemulsion system. I(q) 1 100 1e+04 1e+06 1e+08 q / nm-1 0.1 1 0.1 1 oil/water ratio Φ 0.7 0.6 0.5 0.4 0.3 0.2 0.1 Figure 38: SAXS measurements of series of the different microemulsion C9G2samples with different Φ. The lines represent fits according to the Teubner-Strey approximation, which is not possible for Φ0.6 and 0.7. The intensity of the Φ0.2 ... 0.7 was shifted for clarity reasons. For the measurements samples identical to SANS were used (table 1). The intensity curves shown in figure 38 are again fitted with the Teubner-Strey approximation, the results are summarized in table 4. Compared to the SANS results of the analogous samples, the domain sizes obtained from the SAXS measurements are similar. The correlation lengths differ because the deuteration of the water phase in the SANS samples implies, that the correlation length values apply to the oil domains. However, in a SAXS measurement, the oil or water domains cannot be 68 6 RESULTS AND DISCUSSION 6.2 The C9G2based microemulsion system Table 4: Structure sizes for C9G2-microemulsions determined by fitting the SAXS data with the Teubner-Strey approximation. oil/water ratio Φvolume ratio correlation length ξ/nm domain size d/nm 0.1 0.12 4.55 15.11 0.2 0.24 6.11 18.45 0.3 0.35 9.77 24.21 0.4 0.46 10.27 22.85 0.5 0.56 9.25 23.79 clearly separated. The SAXS-result trend in figure 39 is in good accordance to the SANS results, there is only a small deviation due to the different scattering contrasts. ξ, d / nm 0 5 10 15 20 25 30 oil/water ratio Φ 0 0.1 0.2 0.3 0.4 0.5 ξSAXS ξSANS dSAXS dSANS Figure 39: C9G2microemulsion domain sizes d and correlation lengths ξderived from the TeubnerStrey approximation compared to the SANS results of chapter 6.2.2. 69 6.2 The C9G2based microemulsion system 6 RESULTS AND DISCUSSION 6.2.3 GFP+dynamics by FCS The correlation lengths of nearly 10nm for the oil phase (comp. table 4) imply a similar value for the water domains at Φ=0.5. These lengths are still big enough to allow diffusion of GFP+with a diameter of approximately 5nm. However, the domain size of the water domains is already sufficiently low to produce a notable confinement or crowding effect for the protein. Hence, for the analysis if the data the model introduced by Weiss has to be used [56]. g2 (t) 1 1.2 1.4 1.6 1.8 time / μs 10 100 1,000 1e+04 1e+05 1e+061e+06 GFP+ in buffer GFP+ in Φ=0.2 microemulsion GFP+ in Φ=0.35 microemulsion Φ=0.2 microemulsion dynamics Figure 40: Comparison of FCS autocorrelation curves for GFP+in a C9G2microemulsion. GFP+ shows a slightly slower diffusion inside the water phase of the microemulsion, but faster than the microemulsion dynamics. The protein is non-polar and mainly hydrophilic. Therefore it is confined to the water domains of the microemulsion. Figure 40 shows autocorrelation curves for the C9G2system. The protein is able to move inside the water domains, and the diffusion is slower in a bicontinuous structure (Φ0.3), than in a water continuous structure (Φ0.1). As expected, the diffusion is hindered by the confinement of the water domains. Crowding effects take place for Φ0.1, where oil-in-water structures (see table 3) appear, which already results in a slowed diffusion at this low value of Φcompared to GFP+in buffer. 70 6 RESULTS AND DISCUSSION 6.2 The C9G2based microemulsion system g2 (t) 1 1.2 1.4 1.6 1.8 time / μs 10 100 1,000 1e+04 1e+05 oil/water ratio Φ 0.1 0.2 0.35 0.5 0.6 0.7 Figure 41: GFP+FCS autocorrelation curves inside the water domain of the C9G2microemulsion for different Φ. With rising Φthe curves move to bigger timescales. The lines indicate fits according to equation 21. g2(t) 1 1.5 2 time / μs 1 10 100 1,000 1e+04 1e+05 1e+061e+06 0.1 0.2 0.3 0.4 0.5 Figure 42: Normalized FCS autocorrelation curves of the C9G2microemulsion for different Φlabeled with octadecyl-rhodamin B. With rising Φthe curves move to smaller timescales. Octadecylrhodamin B is amphiphilic and incorporates in the interface. The lines indicate fits according to equation 16. autocorrelation curves shifting to bigger timescales (figure 42). In fig. 43 the diffusion times τsub of GFP+are plotted. The data was fitted with an anomalous diffusion model (equation 21). 71 6.2 The C9G2based microemulsion system 6 RESULTS AND DISCUSSION For small Φthe diffusion is nearly as fast as in dilute solution. τsub / μs 500 1,000 1,500 oil/water ratio Φ 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 GFP+ in ME ME dynamics GFP+ in ME (2 comp) Figure 43: Subdiffusion times τsub of GFP+in the water phase of a C9G2microemulsion, calculated from the fit of the FCS autocorrelation curves. In the range Φ=0.3 to 0.5, the structure changes to bicontinuous and the diffusion is hindered by the sponge like phase. At Φ=0.6 and 0.7 τsub is constant and reflects the dynamics of the microemulsion structure - this is an indicator for the confinement of the protein, it seems to be stuck in the microemulsion and participates in the collective motion of it. When comparing the autocorrelation curves of the microemulsion dynamics and the GFP+autocorrelation curves in a microemulsion, the GFP+curves look more shallow. This is an indicator for anomalous diffusion [95]. The anomalous diffusion exponent αcan also be extracted from the fit. Fig. 44 shows the anomalous diffusion exponent. At small Φαis nearly 1, the protein is hardly confined by the microemulsion structure, which is more droplet-like (slightly elongated drops, see SANS results) than bicontinuous for Φ=0.1 and 0.2. When the structure is bicontinuous starting at Φ=0.3, αdecreases rapidly and indicates, that the protein is confined by the water domains. From Φ=0.6 on, αincreases again, which shows, that the protein gets now stuck in the structure and participates in the diffusive breathing motion of the microemulsion matrix. 72 6 RESULTS AND DISCUSSION 6.2 The C9G2based microemulsion system α 0.5 0.6 0.7 0.8 0.9 1 oil/water ratio Φ 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Figure 44: Anomalous diffusion parameter αfor GFP+in the C9G2system. With an α < 1 subdiffusion is observed, especially for Φ>0.2. 73 6.3 The C12G2based microemulsion system 6 RESULTS AND DISCUSSION of 0.56. At Φ0.6, the radius and the volume fraction suddenly drop to much smaller values. This is due to the inappropriate fit or a structural change to water droplets in an oil continuous phase. R / nm 4 6 8 10 12 14 oil/water ratio Φ 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 SAXS SANS volume fraction 0.2 0.3 0.4 0.5 0.6 oil/water ratio 0 0.2 0.4 0.6 0.8 (a) (b) Figure 52: Results of the GIFT analysis for the C12G2microemulsion samples, (a) sphere radius R (SANS and SAXS values), increases due to the rising amount of oil and decreases for Φ0.6 and 0.7 (b) volume fraction of the spheres (SAXS), increases as well up to Φ=0.5. The SAXS droplet radius results are in good agreement with the SANS outcomes, except for the Φ0.1 value. This could be connected to the diverse scattering contrasts and the different analysis methods. The polydispersity is nearly constant at approximately 20%, which is in good agreement with the results of comparable CiEjsystems [98]. 80 6 RESULTS AND DISCUSSION 6.3 The C12G2based microemulsion system 6.3.3 GFP+dynamics by FCS The SAS results indicate, that the structure is not bicontinuous, the oil component is located as droplets in a water continuous phase, especially for Φ<0.6. This should allow GFP+diffusion inside the water-(buffer-) phase. But with a volume ratio fraction up to 50%, the surrounding of the protein is crowded. Figure 53 shows the situation for a protein diffusing in the water continuous phase of a microemulsion. water phase oil phase surfact. co-surf. protein fluorescent amphiphil hindered diffusion Figure 53: Diffusion model for the C12G2system, where the structure is not bicontinuous but dropletlike. The oil droplets in a continuous water phase are polydisperse and bigger than the protein size. At Φ0.1 the radius is comparable to the GFP+hydrodynamic radius, the diffusion should be nearly unhindered. For rising Φthe droplets grow and merge with other droplets. The system gets crowded and hampers the diffusion of the protein. To observe the GFP+dynamics, FCS measurements were performed. 81 6.3 The C12G2based microemulsion system 6 RESULTS AND DISCUSSION normalized g2 (t) 1 1.2 1.4 1.6 1.8 time / μs 10 100 1,000 1e+04 1e+05 GFP in buffer GFP in Φ 0.1 ME GFP in Φ 0.3 ME Φ 0.1 ME Φ 0.3 ME Figure 54: FCS autocorrelation data of GFP+in a C12G2microemulsion, compared to GFP+in buffer and the pure microemulsion dynamics. The curves were normalized for comparison reasons. The microemulsion shows much slower dynamics than GFP+inside the water domain. Hence, the protein is mobile. normalized g2 (t) 1 1.2 1.4 1.6 1.8 time / μs 10 100 1,000 1e+04 1e+05 oil/water ratio Φ 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Figure 55: GFP+FCS autocorrelation curves inside the water domain of the C12G2microemulsion for different Φ. As seen in figure 54, the GFP+measurements show a hindered diffusion for GFP+inside the microemulsion compared to GFP+in buffer: The diffusion is hindered by crowding and confinement effects. However, under these conditions the protein moves still significantly faster 82 6 RESULTS AND DISCUSSION 6.3 The C12G2based microemulsion system compared to the slow collective mode (breathing mode) of the microemulsion. diffusion time / ms 0,5 1 1,5 oil/water ratio Φ 0 0,2 0,4 0,6 0,8 GFP+ dilute GFP+ in ME ME dynamics Figure 56: Comparison of the FCS diffusion times in the microemulsion system C12G2for GFP+ inside the water domain of the microemulsion, GFP+in buffer and the microemulsion dynamics. In figure 55 the FCS measurements with samples at different Φ(table 5). In general, the autocorrelation curves move to higher times with increasing Φ. This clearly shows that the dynamic is slowed down. The fit results are summarized in figure 56. For small Φ(0.1 and 0.2), GFP+ shows a good mobility because of small oil droplets in a continuous water phase. These droplets hinder the diffusion of the protein not significantly. For high Φ(0.6 and 0.7), the protein dynamics get similar to the microemulsion dynamics. This clearly shows, that GFP+is stuck and can only move with the collective breathing motion of the microemulsion structure. As seen in fig. 57, αis smaller than 1, which indicates subdiffusive behavior. αfor the 1component fit decreases until a minimum at Φ0.5, the protein gets more and more confined by the shrinking water domains. At Φ0.6, αincreases again, which indicates, that the protein is stuck in the bicontinuous structure leading to the recovery of Fickian diffusion. The breathing motion of the microemulsion structure is a diffusive motion., Hence, αconverges to 1. 83 6.3 The C12G2based microemulsion system 6 RESULTS AND DISCUSSION anomalous diff. exp. α 0,6 0,7 0,8 0,9 oil/water ratio Φ 0,2 0,4 0,6 0,8 1-component fit 2-component fit Figure 57: Anomalous diffusion exponent αfor GFP+in the C12G2system. With an α < 1 subdiffusion is observed. The FCS data can also be approximated with a 2 component model, to regard the microemulsion dynamics, which is measured also, due to the fluorescent impurities, αbehaves similar. For small oil/water ratios αis nearly one. This means only minor crowding effects, until Φ=0.4 it decreases to a minimum. Then αrises again because of the diffusive microemulsion dynamics. But the 2-component fit is not really valid from α=0.5 up to higher values. There the protein is stuck and the diffusion shows only one characteristic time decay. These anomalous diffusion exponent results are in the same range compared to investigations of proteins in crowded solutions [99], here the authors found αvalues from 1 to 0.75. 84 6 RESULTS AND DISCUSSION 6.3 The C12G2based microemulsion system 6.3.4 Microemulsion dynamics (FCS) Knowing that the structure of the C12G2samples at Φ=0.1 −0.6 is droplet-like, more can be extracted from the FCS measurements. The diffusion time obtained from the FCS measurements allows to compute the diffusion coefficient. By using the Stokes-Einstein equation also the hydrodynamic radius of the droplets can be calculated. Figure 58 shows the normalized FCS autocorrelation curves. normalized g2 (t) 1 1.2 1.4 1.6 1.8 2 time / μs 1 10 100 1,000 1e+04 1e+05 1e+061e+06 oil/water ratio Φ 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Figure 58: FCS autocorrelation curves of the microemulsion dynamics for different oil/water ratios Φ. The lines represent a 1-component fit. Because of the high polydispersity of the system the obtained values for the radius of the droplets are only approximated results. The data could also be fitted with a 2or 3-component model. But for a convincable fit more parameters like polydispersity or the distribution of diffusion times should be known. The results for the hydrodynamic radius compared to values obtained by other methods are plotted in figure 59. 85 6.3 The C12G2based microemulsion system 6 RESULTS AND DISCUSSION 6.3.5 Microemulsion dynamics (DLS) In the SAS experiments a droplet microemulsion structure was identified. Therefore samples can also be studied by dynamic light scattering. With the resulting apparent diffusion coefficient DDLS, a hydrodynamic radius RHcan be calculated with equation 33. The Stokes-Einstein equation is not fully valid in this case because of polydisperse and interacting spheres. But the hydrodynamic radius can be estimated and therefore it is called apparent hydrodynamic radius. The scattering contrast for photons in the visible range is constituted by the refraction index difference in the system. In microemulsions with a bicontinuous or droplet phase, this difference is mainly caused by the oil and water component. Therefore, the obtained diffusion coefficients represent the collective breathing motion of the domains or the movement of oil-in-water or water-in-oil structures, respectively. R / nm 0 5 10 15 20 25 oil/water ratio Φ 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Rh, app (DLS) Rh, app (FCS) RSAXS Figure 59: Apparent hydrodynamic DLS and FCS radius and radius of the SAXS measurements determined by the GIFT analysis for the C12G2microemulsion system at different oil/water ratios Φ. All three methods show an increase of the droplet radius from Φ=0.1 to 0.5, at 0.6 the radius decreases rapidly to smaller values. Compared to the SAXS measurements, the DLS results show the same trend. Rhexhibits bigger values because of the crowded droplets which move slower in comparison with a dilute solution. Furthermore, the hydrodynamic radius can not be directly compared to a radius of gyration Rgfrom the small angle measurements. For hard spheres the ratio Rg/Rhis expected to be 0.778 [100], the results in fig. 59 yield a ratio of approximately 0.6, which is an indication 86 6 RESULTS AND DISCUSSION 6.3 The C12G2based microemulsion system for spherical droplets. The FCS measurements are in good agreement with the DLS and SAXS results. After Φ=0.5 there is also a similar drop in the radius. 87 6.4 Other microemulsion systems 6 RESULTS AND DISCUSSION 6.4 Other microemulsion systems Further work was done aiming at the study of other systems with surfactants having shorter alkyl chain lengths. Namely the surfactants n-hexyl-β-d-glucoside (C6G1) and n-heptyl-β-d-glucoside (C7G1) were studied using again water, cyclohexane, and 1-pentanol as the other components. It turns out that the phase behavior exhibits no lamellar phase in the γrange, where the systems were scrutinized. The phase diagrams are given in fig. 60. While for C6G1the bicontinuous phase starts at γ=0.16, in the C7G1system the X-point is located γ=0.22, but less alcohol content is needed to form a single-phase microemulsion. In general, the CiG2microemulsions tend more to form a lamellar phase compared to the CiG1systems. 0.05 0.10 0.15 0.20 0.10 0.15 0.20 0.25 0.30 (γ) (δ) 3Φ 1Φ 2Φ 2Φ 0 0.05 0.10 0.15 0.20 0.10 0.15 0.20 0.25 0.30 0.35 (γ) (δ) 3Φ 1Φ 2Φ 2Φ C6G1C7G1 Figure 60: Phase diagrams of C6G1and C7G1microemulsion systems (water, cyclohexane, 1pentanol), no lamellar phase shows up in the studied γrange. SANS measurements were carried out for C7G1samples in Saclay/France at the PAXY instrument. Three samples with different γ(0.25, 0.30, 0.35) and constant Φand δvalues were composed. The intensity curves are plotted in fig. 61. The scattering curves exhibit the typical shape obtained for bicontinuous microemulsions. Hence, the Teubner-Strey approximation was applied to obtain length scales. 88 6 RESULTS AND DISCUSSION 6.4 Other microemulsion systems I (q) 1 10 100 1,000 1 10 100 1,000 q / Å-1 0.1 11 0.1 11 γ 0.35 0.30 0.275 Figure 61: SANS measurements of C7G1microemulsion samples at different γ,Φ0.5 and δ0.09. The data was fitted with a Teubner-Strey approximation (equation 43). ξ, d / nm 20 40 60 80 γ 0.28 0.3 0.32 0.34 0.36 ξ d Figure 62: SANS structure sizes derived from the Teubner Strey approximation, ξ: correlation length, d: domain size. The C7G1based system has a correlation length of 3nm, the domain size decreases from 8nm to 6.5nm for rising γ(compare figure 62). This is due to the rising amount of surfactant in the system which leads to an increase of the interfacial area between water and oil phase, therefore 89 7 CONCLUSION AND FUTURE PROSPECTS 7 Conclusion and future prospects Proteins usually do their work in crowded environments inside the cell and the respective organelles. Hence, an understanding of their diffusion is of fundamental importance for the understanding of the processes of life. Moreover, also in technical processes proteins are often used in confined situations. One example is the use of microemulsions as reaction media for decontamination or in chemical production. Hence, observing the diffusion of a protein inside the water domain of a microemulsion is of major importance for improvement of e.g. decontamination applications realized with a protein like DFPase [18,21,34]. However, this is not an easy task because the structure of the microemulsion is not much bigger than the particles which shall be traced and the microemulsion matrix gives a kind of “background dynamics” in most of the possible experiments. Therefore methods have to be used, where the structure and respectively the dynamics of the microemulsion is faded out and only the desired particle motion can be followed. Methods like dynamic light scattering are not usable, because the refractive index of the system cannot be changed easily. Thus the method of choice was fluorescence correlation spectroscopy where a fluorescent particle can be traced. However, in this technique other difficulties occurred, and e.g. fluorescent impurities disturb the signal of the particle of interest. Even with components of a high purity >99[%]and in 1-pentanol fluorescent impurities were found, in a very low concentration indeed but enough for being detectable by the sensible FCS technique. With technical grade surfactants the fluorescence activity is very high and FCS measurements are not possible. With pure surfactants and a concentration of the fluorescent protein high enough, the detected fluorescent is mainly caused by the protein. Besides the microemulsion dynamics can be taken into account by using a two component model. With a similar size and therefore comparable hydrodynamic behavior, GFP+is a suitable model protein to study dynamics with fluorescence correlation spectroscopy. In this thesis the structure of two different sugar surfactant microemulsion systems was studied systematically with small angle x-ray and neutron scattering measurements. It was shown that the dimensions of this microemulsion system allow diffusion of proteins inside the water domain. The FCS results have yielded that GFP+is mobile depending on the correlation lengths of the water domains. The confinement increases for higher oil volume fraction Φ, meaning smaller water domains. In the present work it could be shown that the diffusion of GFP+is hindered in the different 97 7 CONCLUSION AND FUTURE PROSPECTS microemulsions. This behavior was quantified in terms of the sub-diffusion exponent αand it turns out that αis influenced by the composition of the microemulsion. For smaller water domains αgoes down systematically. Hence, microemulsions are well suited as model systems for the study of sub-diffusion in a controlled way. For too high oil/water ratios the protein gets stuck and reflects the dynamics known as the ”collective breathing motion" of the sponge-like structure. Furthermore, the used system is related to applications. The pure sugar surfactant can be easily replaced by a technical grade surfactant, which leads to cheap microemulsions for commercial use. Studies in the field of skin friendly and environmentally compatible components are already ongoing to provide a system with similar dimensions for wide application possibilities in the decontamination area. It has to be observed in decontamination experiments which microemulsion systems show the best decontamination ability. In some systems the protein is mobile due to bigger water domains, in other cases at high Φor with short sugar surfactant systems the protein is stuck in the water phase but near to the oil phase where the toxic agent is located. Thus it has to be observed which influence the protein mobility and the distance to the toxic compound have on the decontamination efficiency and kinetics. The results related to the sub-diffusion of GFP+inside the different studies are unique and it remains to be clarified whether the sub-diffusive behavior persists on all length scales or not. Works of Schreiber et al. [104]suggest that on very short length scale Fickian diffusion might be recovered. hence, this issue should be addressed using experiments with an appropriate resolution in length. A good choice would be a CiEjmicroemulsion system with a deuterated surfactant in an NSE experiment. Then the phase behavior could be tuned with the temperature. Furthermore, one deuterated component less would be needed and would simplify the contrast matching procedure. 98 8 DANKSAGUNG 8 Danksagung In erster Linie bedanke ich mich bei meinem Doktorvater, Prof. Dr. Thomas Hellweg, für die Überlassung des spannenden Themas, das Vertrauen und die vielfältigen Forchungsmöglichkeiten. Großer Dank gilt auch meinen Kollegen Sebastian Höhn, Christoph Angermann, Bastian Wedel, Christoph Schulreich, Michael Zeiser, Katja von Nessen, Susanne Seibt, Yvonne Hertle, Simone Wagner und Ralf Stehle für interessante Diskussionen, diverse Ratschläge, jegliche Unterstützung und die sehr gute Zusammenarbeit. Herrn Prof. Dr. Stephan Förster danke ich dafür, dass ich meine Arbeit in Bayreuth fertigstellen durfte. Allen “neuen” Kollegen aus der Arbeitsgruppe Förster, insbesondere meinen Bürokameraden Sebastian With und Jan Schröder danke ich für die positive Atmosphäre und die sehr gute Zusammenarbeit. Einen besonderen Dank auch an Karlheinz Lauterbach, der immer ein offenes Ohr hatte und sich unablässig um die einwandfreie Funktion der Lehrstuhlaustattung kümmerte. Mein Dank gilt auch Herrn Prof. Dr. Matthias Weiss und den Mitarbeitern aus der Experimentalphysik I, für die gute Zusammenarbeit und Möglichkeit der Nutzung des Confocor 2 auch nach dem Umzug des Gerätes. Vielen Dank an Uwe Güth für Herstellung und Lieferung der Proteinlösungen. Bei Martin Dulle möchte ich mich für die Unterstützung bei den SAXS-Messungen, Mireia Subinyà für die PFGNMR Messungen und Frank Lüdel für SANS-Messungen bedanken . Elisabeth Düngfelder und Sandra Gericke will ich für die Hilfe bei allen Themen außerhalb der Forschung bedanken. Beim wehrwissenschaftlichen Institut Munster möchte ich mich für die Finanzierung des Projektes bedanken. Vielen Dank den Local Contacts Alain Lapp in Saclay/LLB und Peter Falus in Grenoble/ILL für die Untersützung während der Messzeiten. Von ganzen Herzen möchte ich mich auch bei meiner Familie und bei meiner Freundin Maria bedanken für die Geduld, das Verständnis und dass sie immer an mich geglaubt haben. Meinen Eltern ein herzlicher Dank für die Unterstützung und dass sie mir das Studium überhaupt ermöglicht haben. 99 A APPENDIX A Appendix A.1 Alternative microemulsion systems Due to the problem that 1-pentanol and similar alcohols with other chain lengths contain fluorescent impurities which may disturb FCS measurements, it was searched for alternative fluorescent free co-surfactant. One candidate is 1-propanol, this alcohol is available in high purity and shows nearly no fluorescence activity. Using the components n-dodecyl-β-d-maltoside, cyclohexane and water, the formation of a microemulsion is possible. But the main problem is the high amount of 1-propanol which is needed to reach the 1-phase region, the X-point is located at γ=0.1, δ=0.39. 0.14 0.19 0.24 0.29 0.34 0.39 0 0.05 0.10 0.15 0.20 0.25 (γ) (δ) 1Φ 2Φ 3Φ Figure 68: Phase diagram of the system C12G2-propanol-cyclohexane-water, the 1-phase borders are located at high alcohol amounts. The upper borders of the 1-phase area could not be determined. At such high δvalues, there is more propanol present in the system than water and oil, therefore it is uncertain, what kind of structure is obtained. Another possible explanation for the fact that no upper phase boundary could be formed would be refractive index matching with two phases, then the phases cannot be separated by visible inspection. Two get more information about the structure, SANS measurements were performed at the ILL in Grenoble/France, figure 69. The samples were made at constant δwith varying γstarting close to the X-point for a constant oil/water ratio. 101 A.1 Alternative microemulsion systems A APPENDIX I(q)/cm-1 1 10 q / nm-1 0.01 0.1 γ 0.15 0.20 0.25 Figure 69: C12G2-propanol microemulsion SANS measurements for different γvalues at Φ0.5 and δ0.38. The red lines are fits according to the Teubner-Strey approximation. The scattering curves show structure peaks for higher γvalues moving to higher qfor rising γ meaning decreasing structure sizes. This is common for bicontinuous microemulsion systems, because with increasing surfactant amount more molecules stay at the interface and are able to form smaller water/oil domains. The SANS data can be fitted with the Teubner-Strey approximation, the results are shown in table 7. Table 7: Structure sizes for C12G2-propanol-microemulsions determined by SANS at Φ0.5 and δ 0.38 γ ξ /nm d/nm 0.15 2.1 4.9 0.2 1.8 5.8 0.25 1.4 4.6 The structures are very small compared to other systems and at least for γ=0.15 it has to be clarified, if the structure is really bicontinuous or droplet-like and where the 1-propanol molecules are located, which are the major component in the system. 102 A APPENDIX A.2 Solvent sensitive fluorescent dyes A.2 Solvent sensitive fluorescent dyes Fluorescent dyes can be influenced by the solvent, the consequence is for example a quenched fluorescence or a shift in the emission spectra. nile-red [105,106]is a lipophilic molecule which is sensitive to the polarity of the used solvent. Figure 70 shows fluorescence emission of nile-red in cyclohexane and 1-pentanol. normalized intensity 0 0.2 0.4 0.6 0.8 1 emission wavelength / nm 500 550 600 650 700 1-pentanol cyclohexane Figure 70: Fluorescence of nile-red in different solvents, cyclohexane and 1-pentanol. In water, where nile-red is nearly not soluble, the emission maximum is located at 660nm [105]. The emission maximum is at 535nm for the unpolar solvent cyclohexane and at 630nm for 1-pentanol. This is a big solvochromic shift of 95nm. Nile-red can also be dissolved in a microemulsion, which is plotted in figures 71 (δvariation), 72 (Φvariation) and 73 (γvariation). The influence of a rising content of pentanol in the δvariation shifts the maximum to higher wavelengths. 103 A.2 Solvent sensitive fluorescent dyes A APPENDIX normalized intensity 0 0.2 0.4 0.6 0.8 1 emission wavelength / nm 500 550 600 650 700 δ 0.06 0.08 0.10 Figure 71: Nile-red fluorescence in microemulsions with different δ((components: water, Glucopon 220, cyclohexane, 1-pentanol). normalized intensity 0 0.2 0.4 0.6 0.8 1 emission wavelength / nm 500 550 600 650 700 Φ 0.1 0.3 0.5 0.7 0.9 Φ increases Figure 72: Nile-red fluorescence in microemulsion with different Φ(components: water, Glucopon 220, cyclohexane, 1-pentanol). When Φis varied, the emission maximum should increase with shrinking Φdue to the increasing water amount. But this is not the case, with rising Φthe maximum moves to higher wavelength. Most likely, the necessary higher amount of 1-pentanol influences the nile-red emission in the microemulsion. This is an indication, that the pentanol molecules are located in the surfactant interface and diluted in the cyclohexane phase. Therefore, nile-red could be used as a marker 104 A APPENDIX A.2 Solvent sensitive fluorescent dyes for the distribution of 1-pentanol. A change of the surfactant content γhardly affects the emission wavelength, only in a range of 3 nm. For increasing γ, the microemulsion structure morphs from a bicontinuous phase to a droplet structure. Further γincreasing results in smaller droplets. Thus nile-red is located in the oil domains or droplets, where it is serperated from the water phase by the surfactant interface. Also the distribution of pentanol is not changing when γis increased. normalized intensity 0 0.2 0.4 0.6 0.8 1 emission wavelength / nm 500 550 600 650 700 γ 0.14 0.17 0.20 0.23 Figure 73: Nile-red fluorescence in microemulsion with different γ(components: water, Glucopon 220, cyclohexane, 1-pentanol). 105 REFERENCES REFERENCES References [1]P. Winsor, “Hydrotropy, solubilisation and related emulsification processes,” Transactions of the Faraday Society, vol. 44, pp. 376–398, 1948. [2]J. Schulman, W. Stoeckenius, and L. Prince, “Mechanism of formation and structure of micro emulsions by electron microscopy,” The Journal of Physical Chemistry, vol. 63, no. 10, pp. 1677–1680, 1959. [3]T. Hoar and J. Schulman, “Transparent Water-in-Oil Dispersions: the Oleopathic HydroMicelle,” Nature, vol. 152, pp. 102–103, 1943. [4]S. Friberg, Micellization, solubilization, and microemulsions (Mittal, K.L. and American Chemical Society). Micellization, Solubilization, and Microemulsions, Band 1, Plenum Press, 1977. [5]S. Friberg, Microemulsions: Theory and practice (Prince, L.M.). Academic Press, 1977. [6]G. Gillberg, H. Lehtinen, and S. Friberg, “NMR and IR investigation of the conditions determining the stability of microemulsions,” Journal of Colloid and Interface Science, vol. 33, no. 1, pp. 40 – 53, 1970. [7]K. Shinoda and T. Ogawa, “The solubilization of water in nonaqueous solutions of nonionic surfactants,” Journal of Colloid and Interface Science, vol. 24, no. 1, pp. 56 – 60, 1967. [8]K. Shinoda and H. Saito, “The effect of temperature on the phase equilibria and the types of dispersions of the ternary system composed of water, cyclohexane, and nonionic surfactant,” Journal of Colloid and Interface Science, vol. 26, no. 1, pp. 70 – 74, 1968. [9]K. Shinoda and H. Saito, “The stability of o/w type emulsions as functions of temperature and the HLB of emulsifiers: The emulsification by PIT-method,” Journal of Colloid and Interface Science, vol. 30, no. 2, pp. 258 – 263, 1969. [10]D. Shah, Surface phenomena in enhanced oil recovery. Plenum Press, 1981. [11]C. Solans and H. Kunieda, Industrial Applications of Microemulsions. Surfactant Science Series, M. Dekker, 1997. 113 REFERENCES REFERENCES [12]M.-J. Schwuger, K. Stickdorn, and R. Schomaecker, “Microemulsions in technical processes,” Chemical Reviews, vol. 95, no. 4, pp. 849–864, 1995. [13]R. Schomäcker and Bayer-AG, “Mikroemulsionen als Medium für chemische Reaktionen,” Nachrichten aus Chemie, Technik und Laboratorium, vol. 40, no. 12, pp. 1344–1352, 1992. [14]M. Pileni, “Reverse micelles as microreactors,” The Journal of Physical Chemistry, vol. 97, no. 27, pp. 6961–6973, 1993. [15]J. Sjöblom, R. Lindberg, and S. E. Friberg, “Microemulsions - phase equilibria characterization, structures, applications and chemical reactions,” Advances in Colloid and Interface Science, vol. 65, no. 0, pp. 125 – 287, 1996. [16]M. Höfer, “Chemische Kampfstoffe: Ein Überblick,” Chemie in unserer Zeit, vol. 36, no. 3, pp. 148–155, 2002. [17]H. Murakami, A. Birnbaum, and P. Gabriel, Underground. Vintage International, 2001. [18]M.-M. Blum and A. Richardt, Decontamination of Warfare Agents: Enzymatic Methods for the Removal of B/C Weapons. Wiley-VCH Verlag GmbH & Co. KGaA, 2008. [19]M. Kahlweit, G. Busse, B. Faulhaber, and H. Eibl, “Preparing nontoxic microemulsions,” Langmuir, vol. 11, no. 11, pp. 4185–4187, 1995. [20]S. Wellert, H. Altmann, A. Richardt, A. Lapp, P. Falus, B. Farago, and T. Hellweg, “Dynamics of the interfacial film in bicontinuous microemulsions based on a partly ionic surfactant mixture: A neutron spin-echo study,” The European Physical Journal E: Soft Matter and Biological Physics, vol. 33, pp. 243–250, 2010. 10.1140/epje/i2010-10668-1. [21]A. Richardt, M.-M. Blum, and S. Mitchell, “Was wissen Calamari über Sarin? Enzymatische Dekontamination von Nervenkampfstoffen,” Chemie in unserer Zeit, vol. 40, no. 4, pp. 252–259, 2006. [22]C. Stubenrauch, Microemulsions: Background, New Concepts, Applications, Perspectives. Oxford: WILEY-VCH Verlag, 2009. [23]W. Pritzkow, “Die Tenside, Herausgegeben von K. Kosswig und H. Stache München,” Journal für Praktische Chemie/Chemiker-Zeitung, vol. 336, no. 5, pp. 473–473, 1994. 114 REFERENCES REFERENCES [24]H. Roth and H. Fenner, Arzneistoffe. Stuttgart: Thieme, 1998. [25]Brockhaus ABC Chemie. VEB F. A. Brockhaus Verlag Leipzig, 1965. [26]K. Kosswig and H. Stache, Die Tenside. Hanser, 1993. [27]T. Hellweg, “Phase structures of microemulsions,” Current Opinion in Colloid and Interface Science, vol. 7, no. 1 - 2, pp. 50 – 56, 2002. [28]A. Biasio, C. Cametti, P. Codastefano, P. Tartaglia, J. Rouch, and S. Chen, “Phase behavior of dense three-component ionic microemulsions and electrical conductivity in the lamellar phase,” Physical Review E, vol. 47, pp. 4258–4264, Jun 1993. [29]A. Biasio, C. Cametti, P. Codastefano, P. Tartaglia, J. Rouch, and S. Chen, “Electrical conductivity in the lamellar phase of a water-in-oil microemulsion,” in Trends in Colloid and Interface Science VII, vol. 93 of Progress in Colloid & Polymer Science, pp. 191–193, Steinkopff, 1993. [30]S. Burauer, L. Belkoura, C. Stubenrauch, and R. Strey, “Bicontinuous microemulsions revisited: a new approach to freeze fracture electron microscopy (FFEM),” Colloids and Surfaces A: Physicochemical and Engineering Aspects, vol. 228, no. 1-3, pp. 159 – 170, 2003. Festschrift for Professor Bjorn Lindman. [31]N. Freiberger, C. Moitzi, L. de Campo, and O. Glatter, “An attempt to detect bicontinuity from sans data,” Journal of Colloid and Interface Science, vol. 312, no. 1, pp. 59 – 67, 2007. [32]M. Kahlweit, G. Busse, and J. Winkler, “Electric conductivity in microemulsions,” The Journal of Chemical Physics, vol. 99, no. 7, pp. 5605–5614, 1993. [33]A. Zilman and R. Granek, “Undulations and dynamic structure factor of membranes,” Physical Review Letters, vol. 77, pp. 4788–4791, Dec 1996. [34]S. Wellert, B. Tiersch, J. Koetz, A. Richardt, A. Lapp, O. Holderer, J. Gäb, M.-M. Blum, C. Schulreich, R. Stehle, and T. Hellweg, “The DFPase from loligo vulgaris in sugar surfactant-based bicontinuous microemulsions: Structure, dynamics, and enzyme activity,” European Biophysics Journal, vol. 40, pp. 761–774, 2011. 115 REFERENCES REFERENCES [35]R. Strey, “Microemulsion microstructure and interfacial curvature,” Colloid and Polymer Science, vol. 272, pp. 1005–1019, 1994. [36]M. Teubner and R. Strey, “Origin of the scattering peak in microemulsions,” The Journal of Chemical Physics, vol. 87, pp. 3195–3200, 1987. [37]S. Wellert, M. Karg, H. Imhof, A. Steppin, H.-J. Altmann, M. Dolle, A. Richardt, B. Tiersch, J. Koetz, A. Lapp, and T. Hellweg, “Structure of biodiesel based bicontinuous microemulsions for environmentally compatible decontamination: A small angle neutron scattering and freeze fracture electron microscopy study,” Journal of Colloid and Interface Science, vol. 325, no. 1, pp. 250 – 258, 2008. [38]T. Sottmann, “Solubilization efficiency boosting by amphiphilic block co-polymers in microemulsions,” Current Opinion in Colloid and Interface Science, vol. 7, no. 1, pp. 57–65, 2002. [39]B. Jakobs, T. Sottmann, R. Strey, J. Allgaier, , and D. Richter, “Amphiphilic block copolymers as efficiency boosters for microemulsions,” Langmuir, vol. 15, no. 20, pp. 6707– 6711, 1999. [40]H. Endo, J. Allgaier, M. Mihailescu, M. Monkenbusch, G. Gompper, D. Richter, B. Jakobs, T. Sottmann, and R. Strey, “Amphiphilic block copolymers as efficiency boosters in microemulsions: a SANS investigation of the role of polymers,” Applied Physics A, vol. 74, pp. 392–395, 2002. [41]S. Wellert, H. Imhof, M. Dolle, H.-J. Altmann, A. Richardt, and T. Hellweg, “Decontamination of chemical warfare agents using perchloroethylene - marlowet IHF-H2O-based microemulsions: wetting and extraction properties on realistic surfaces,” Colloid & Polymer Science, vol. 286, pp. 417–426, 2008. [42]E. Scharff, J. Koepke, G. Fritzsch, C. Luecke, and H. Rueterjans, “RCSB protein data bank: 1E1A, crystal structure of diisopropylfluorophosphatase from Loligo vulgaris.” http://www.rcsb.org/pdb/explore/explore.do?structureId=1E1A. (01/2013). 116 REFERENCES REFERENCES [43]E. Scharff, J. Koepke, G. Fritzsch, C. Lücke, and H. Rüterjans, “Crystal structure of diisopropylfluorophosphatase from loligo vulgaris.,” Structure, vol. 9, no. 6, pp. 493–502, 2001. [44]C. Schulreich, C. Angermann, S. Hoehn, R. Neubauer, S. Seibt, R. Stehle, A. Lapp, A. Richardt, A. Diekmann, and T. Hellweg, “Bicontinuous microemulsions with extremely high temperature stability based on skin friendly oil and sugar surfactant,” Colloids and Surfaces A: Physicochemical and Engineering Aspects, vol. 418, pp. 39–46, 2013. [45]S. Wellert, M. Karg, O. Holderer, A. Richardt, and T. Hellweg, “Temperature dependence of the surfactant film bending elasticity in a bicontinuous sugar surfactant based microemulsion: A quasielastic scattering study,” Phys. Chem. Chem. Phys., vol. 13, pp. 3092– 3099, 2011. [46]E. Elson and D. Magde, “Fluorescence correlation spectroscopy. I. Conceptual basis and theory,” Biopolymers, vol. 13, no. 1, pp. 1–27, 1974. [47]J. Lakowicz, Principles of Fluorescence Spectroscopy. Springer London, 2009. [48]P. Pringsheim, Fluorescence and phosphorescence. Interscience Publishers New York and London, 1963. [49]R. Kellogg and R. Bennett, “Radiationless Intermolecular Energy Transfer. III. Determination of Phosphorescence Efficiencies,” The Journal of Chemical Physics, vol. 41, no. 10, pp. 3042–3045, 1964. [50]A. Follenius-Wund, M. Bourotte, M. Schmitt, F. Iyice, H. Lami, J.-J. Bourguignon, J. Haiech, and C. Pigault, “Fluorescent derivatives of the GFP chromophore give a new insight into the GFP fluorescence process,” Biophysical Journal, vol. 85, no. 3, pp. 1839 – 1850, 2003. [51]R. Rigler and E. Elson, Fluorescence Correlation Spectroscopy: Theory and Applications. Berlin: Springer, 2001. [52]S. Keller, Fluoreszenz-Korrelations-Spektroskopie in Polymerlösungen. PhD thesis, München, 2004. 117 REFERENCES REFERENCES [53]J. Ricka and T. Binkert, “Direct measurement of a distinct correlation function by fluorescence cross correlation,” Physical Review A, vol. 39, pp. 2646–2652, Mar 1989. [54]J. Widengren and R. Rigler, “Mechanisms of photobleaching investigated by fluorescence correlation spectroscopy,” Bioimaging, vol. 4, no. 3, pp. 149–157, 1996. [55]J. Widengren, U. Mets, and R. Rigler, “Fluorescence correlation spectroscopy of triplet states in solution: A theoretical and experimental study,” The Journal of Physical Chemistry, vol. 99, no. 36, pp. 13368–13379, 1995. [56]M. Weiss, M. Elsner, F. Kartberg, and T. Nilsson, “Anomalous subdiffusion is a measure for cytoplasmic crowding in living cells.,” Biophys. J., vol. 87, pp. 3518–3524, Nov. 2004. [57]S. Provencher, “Inverse problems in polymer characterization: Direct analysis of polydispersity with photon correlation spectroscopy,” Die Makromolekulare Chemie, vol. 180, no. 1, pp. 201–209, 1979. [58]D. Koppel, “Analysis of macromolecular polydispersity in intensity correlation spectroscopy: The method of cumulants,” The Journal of Chemical Physics, vol. 57, no. 11, pp. 4814–4820, 1972. [59]S. Provencher, “A constrained regularization method for inverting data represented by linear algebraic or integral equations,” Computer Physics Communications, vol. 27, no. 3, pp. 213 – 227, 1982. [60]S. Provencher, “CONTIN: A general purpose constrained regularization program for inverting noisy linear algebraic and integral equations,” Computer Physics Communications, vol. 27, no. 3, pp. 229–242, 1982. [61]A. G. Langen/Germany, “ALV-5000 DLS instrument.” http://www.alvgmbh.de. (11/2012). [62]U. Tübingen, “Small-angle scattering and data analysis.” http://www.soft-matter.unituebingen.de/teaching/SASTutorial.pdf. (12/2012). [63]J. Kohlbrecher and I. Bressler, “SASfit.” https://kur.web.psi.ch/sans1/SANSSoft/sasfit.html. (01/2013). 118 REFERENCES REFERENCES [64]O. Glatter, “GIFT: small angle scattering data evaluation software.” http://physchem.kfunigraz.ac.at/sm/Index.html. (01/2013). [65]A. Bergmann, G. Fritz, and O. Glatter, “Solving the generalized indirect Fourier transformation (GIFT) by Boltzmann simplex simulated annealing (BSSA),” Journal of Applied Crystallography, vol. 33, pp. 1212–1216, Oct 2000. [66]F. Mezei, C. Pappas, and T. Gutberlet, Neutron Spin Echo Spectroscopy: Basics, Trends and Applications. Lecture Notes in Physics, Springer, 2003. [67]T. Munich, “RESEDA-Handout.” http://www2009.ph.tum.de/studium/betrieb/praktika/fopra/versuche/61/. (12/2012). [68]“Institut Laue-Langevin, Grenoble, France - IN15.” http://www.ill.eu/html/instrumentssupport/instruments-groups/instruments/in15/. (12/2012). [69]C. Bargeron, “Measurement of a continuous distribution of spherical particles by intensity correlation spectroscopy: Analysis by cumulants,” The Journal of Chemical Physics, vol. 61, no. 5, pp. 2134–2138, 1974. [70]T. Hellweg, M. Gradzielski, B. Farago, and D. Langevin, “Shape fluctuations of microemulsion droplets: A neutron spin-echo study,” Colloids and Surfaces A: Physicochemical and Engineering Aspects, vol. 183-185, no. 0, pp. 159 – 169, 2001. [71]S. Milner and S. Safran, “Dynamical fluctuations of droplet microemulsions and vesicles,” Physical Review A, vol. 36, pp. 4371–4379, Nov 1987. [72]J. Huang, S. Milner, B. Farago, and D. Richter, “Study of dynamics of microemulsion droplets by neutron spin-echo spectroscopy,” Physical Review Letters, vol. 59, pp. 2600– 2603, Nov 1987. [73]T. Hellweg and D. Langevin, “The dynamics in dodecane/C10E5/water microemulsions determined by time resolved scattering techniques,” Physica A: Statistical Mechanics and its Applications, vol. 264, no. 3-4, pp. 370 – 387, 1999. 119 REFERENCES REFERENCES [74]C. Schulreich, “Bikontinuierliche Mikroemulsionen als neuartige Dekontaminationsmittel: Phasenverhalten und Strukturuntersuchungen in Zuckertensid - Systemen (Zulassungsarbeit),” Bayreuth 2009. [75]R. Wheeler, “RCSB protein data bank: 1EMA, GFP structure.” www.wikipedia.org. (11/2012). [76]X. Shu, N. Shaner, C. Yarbrough, R. Tsien, and S. Remington, “RCSB protein data bank: 2H5Q, crystal structure of mCherry.” http://www.rcsb.org/pdb/explore/explore.do?structureId=2H5Q. (11/2012). [77]O. Shimomura, F. H. Johnson, and Y. Saiga, “Extraction, purification and properties of aequorin, a bioluminescent protein from the luminous hydromedusan, aequorea,” Journal of Cellular and Comparative Physiology, vol. 59, no. 3, pp. 223–239, 1962. [78]O. Scholz, A. Thiel, W. Hillen, and M. Niederweis, “Quantitative analysis of gene expression with an improved green fluorescent protein,” European Journal of Biochemistry, vol. 267, no. 6, pp. 1565–1570, 2000. [79]M. Chalfie and S. Kain, Methods of Biochemical Analysis, Green Fluorescent Protein: Properties, Applications and Protocols. Methods of Biochemical Analysis, Wiley, 2005. [80]S. Bokman and W. Ward, “Renaturation of aequorea green-fluorescent protein,” Biochemical and Biophysical Research Communications, vol. 101, no. 4, pp. 1372 – 1380, 1981. [81]N. Shaner, R. Campbell, P. Steinbach, B. Giepmans, A. Palmer, and R. Tsien, “Improved monomeric red, orange and yellow fluorescent proteins derived from Discosoma sp. red fluorescent protein,” Nature Biotechnology, vol. 22, pp. 1567–1572, Dec. 2004. [82]H. Zettl, New Applications of Fluorescence Correlation Spectroscopy in Materials Science. PhD thesis, Bayreuth, 2006. [83]H. Bagger-Jörgensen, L. Coppola, K. Thuresson, U. Olsson, and K. Mortensen, “Phase behavior, microstructure, and dynamics in a nonionic microemulsion on addition of hydrophobically end-capped poly(ethylene oxide),” Langmuir, vol. 13, no. 16, pp. 4204– 4218, 1997. 120 REFERENCES REFERENCES [84]J. Huang, J. Sung, and X.-L. Wu, “The effect of H2O and D2O on a water-in-oil microemulsion,” Journal of Colloid and Interface Science, vol. 132, no. 1, pp. 34 – 42, 1989. [85]I. Vasilief, “Qtiplot.” http://soft.proindependent.com/qtiplot.html. (10/2012). [86]J. Nelder and R. Mead, “A simplex method for function minimization,” The Computer Journal, vol. 7, no. 4, pp. 308–313, 1965. [87]“Ganesha - Small Angle X-Ray Scattering instrument.” SAXS LAB /Denmark. [88]“PAXY instrument at the LLB Saclay/France.” http://www-llb.cea.fr/en/fr-en/pdf/paxyllb.pdf. (11.2012). [89]G. Burnett, G. Rees, D. Steytler, and B. Robinson, “Fluorescence correlation spectroscopy of water-in-oil microemulsions: An application in specific characterisation of droplets containing biomolecules,” Colloids and Surfaces A: Physicochemical and Engineering Aspects, vol. 250, no. 1-3, pp. 171 – 178, 2004. [90]M. Gradzielski, “Recent developments in the characterisation of microemulsions,” Current Opinion in Colloid & Interface Science, vol. 13, no. 4, pp. 263 – 269, 2008. [91]E. Potma, W. de Boeij, L. Bosgraaf, J. Roelofs, P. van Haastert, and D. Wiersma, “Reduced protein diffusion rate by cytoskeleton in vegetative and polarized dictyostelium cells,” Biophysical Journal, vol. 81, no. 4, pp. 2010 – 2019, 2001. [92]S. Sun, Physical Chemistry of Macromolecules: Basic Principles and Issues. Wiley New Jersey, 2004. [93]T. Sottmann and R. Strey, “Evidence of corresponding states in ternary microemulsions of water - alkane - CiEj,” Journal of Physics: Condensed Matter, vol. 8, no. 25A, p. A39, 1996. [94]M. Summers, J. Eastoe, S. Davis, Z. Du, R. M. Richardson, R. K. Heenan, D. Steytler, and I. Grillo, “Polymerization of cationic surfactant phases,” Langmuir, vol. 17, no. 17, pp. 5388–5397, 2001. 121