Novel microemulsions with an anionic/non-ionic surfactant mixture

Novel microemulsions with an anionic/non-ionic surfactant mixture Dissertation zur Erlangung des akademischen Grades eines Doktors der Naturwissenschaften (Dr. rer. nat.) der Fakultät für Biologie, Chemie und Geowissenschaften an der Universität Bayreuth vorgelegt von Dipl.-Biochem. Univ. Lukas Wolf Bayreuth, Dezember 2011 Die vorliegende Arbeit wurde in den Jahren 2009 – 2011 in Bayreuth unter der Betreuung von Herrn Prof. em. Dr. Heinz Hoffmann in den Laboren der Firma BayColl im „Zentrum für Neue Materialien Bayreuth“, Wolfsbach, durchgeführt. 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.). Promotionsgesuch eingereicht am: 07. Dezember 2011 Zulassung durch die Prüfungskommission: 14. Dezember 2011 Tag des wissenschaftlichen Kolloquiums: 17. April 2012 Amtierender Dekan: Prof. Dr. Beate Lohnert Prüfungsausschuss: Prof. Dr. em. Heinz Hoffmann (Erster Gutachter) Prof. Dr. Yeshayahu Talmon (Zweiter Gutachter) Prof. Dr. Karlheinz Seifert (Vorsitzender) Prof. Dr. Stephan Förster „Grundlagenforschung betreibe ich dann, wenn ich nicht weiß, was ich tue.“ W. von Braun Table of Contents 7 0. Table of Contents Table of Contents 0. TABLE OF CONTENTS....................................................................................................................................7 1. SUMMARY........................................................................................................................................................8 1.1. English Version ...........................................................................................................................................8 1.2. German Version...........................................................................................................................................9 2. INTRODUCTION.............................................................................................................................................10 2.1. Microemulsions and their applications......................................................................................................10 2.2. Phase behaviour of surfactants ..................................................................................................................12 2.3. Microemulsions with non-ionic surfactants...............................................................................................15 2.4. Microemulsions with ionic surfactants......................................................................................................17 2.5. Objectives of this thesis.............................................................................................................................19 3. SYNOPSIS........................................................................................................................................................21 3.1. The surfactant system Ca(DS) 2 /Mg(DS) 2 – IT 3........................................................................................21 3.2. Solubilization of oil into the surfactant mixture ........................................................................................22 3.3. Cryo-TEM imaging of the microemulsion system Ca(DS) 2 /IT 3 – H 2 O/M 2 .............................................24 3.4. Dynamic properties of microemulsions in the single phase channels........................................................28 3.4.1. Introduction of the microemulsion system Mg(DS) 2 /IT 3 – H 2 O/decane ...........................................28 3.4.2. Electric birefringence and rheology measurements...........................................................................30 3.5. Cryo-TEM of microemulsions with a High Internal Phase Microemulsion (HIPME) structure ...............35 3.6. PFG-NMR self diffusion measurements....................................................................................................38 3.6.1. PFG-NMR self diffusion measurements in the single phase channels...............................................38 3.6.2. Influence of excess salt to the microemulsion system.........................................................................40 3.7. Outlook......................................................................................................................................................42 4. REFERENCES..................................................................................................................................................44 4.1. Literature ...................................................................................................................................................44 4.2. List of figures.............................................................................................................................................46 5. PUBLICATIONS..............................................................................................................................................47 5.1. Overview of publications and individual contribution...............................................................................47 5.1.1. Microemulsions from silicone oil with an anionic/nonionic surfactant mixture................................49 5.1.2. Cryo-TEM imaging of a novel microemulsion system of silicone oil with an anionic/nonionic surfactant mixture........................................................................................................................................60 5.1.3. Dynamic Properties of Microemulsions in the Single-Phase channels..............................................69 5.1.4. Microemulsions with a HIPME (High Internal Phase Microemulsion) structure .............................90 5.1.5. PFG-NMR in the Single Phase Channels of Microemulsions with an anionic/non-ionic surfactant mixture.........................................................................................................................................................98 6. ABBREVIATIONS AND SYMBOLS............................................................................................................109 7. PRESENTATIONS AT INTERNATIONAL MEETINGS.............................................................................110 8. ACKNOWLEDGEMENT...............................................................................................................................111 9. ERKLÄRUNG ................................................................................................................................................112 Summary 8 1. Summary 1.1. English Version Microemulsions consist of water, oil and surfactant. In contrast to ordinary emulsions, microemulsions are transparent and thermodynamically stable phases. They appear to be macroscopic single-phase systems but are, however, based on highly complex nanostructures. From the scientific point of view, the so far most studied and best understood microemulsion systems consist of water, oil and either a single non-ionic surfactant or an electrically charged ionic surfactant. Both systems are significantly different, for example in their phase behaviour, thermal stability or their nanostructures. Systems with ionic/non-ionic surfactant mixtures, however, have not yet been investigated intensely. In this work, the phase behaviour of an anionic/non-ionic surfactant mixture with different oils was investigated. The phase diagrams exhibit two optically isotropic microemulsion regions with increasing oil content at constant temperature and surfactant concentration, the so called single phase channels. The two isotropic single phase channels are separated by an optically anisotropic phase region. The microemulsion channel below the anisotropic region extends from the aqueous phase, starting with increasing oil concentration and increasing mass fraction of the non-ionic co-surfactant in the surfactant mixture to the middle of the phase diagram and ends there. The upper single phase channel runs through a steep minimum, with respect to the surfactant/co-surfactant ratio, continuously from the aqueous to the oil-rich side of the phase diagram. In contrast to microemulsions with single non-ionic surfactants, the microemulsion channels are isothermal. The different single-phase regions were examined with various physico-chemical methods. The nanostructures could be identified by measuring electric conductivity, SANS, PFG-NMR and by electron microscopy. While the lower single phase channel consists of small oil droplets in a continuous aqueous phase, which swell with increasing oil content, the nanostructure in the upper channel undergoes a complex structural transition. The oil-free sample, which has a bicontinuous sponge structure, is transformed to a water-in-oil polyhedral foam structure by solubilizing only a few percent of oil. For this so far unknown microemulsion structure, we introduced the term high internal phase microemulsion (HIPME) due to structural similarities to the known high internal phase emulsions (HIPE). This complex structural transition could be observed by transient electric birefringence. The determined structural relaxation times, which also determine the viscosity of the fluids, run through a sharp maximum at the transition point from the bicontinuous to the w/o-foam structures. The observed HIPME structures are probably caused by the presence of the electric charge of the anionic surfactant. The electric charge on the surfactant monolayer leads to a comparably high interfacial tension between the diluted aqueous surfactant phase and the oil. Consequences of this high interfacial tension are oil continuous polyhedral foam structures instead of bicontinuous structures, which are obtained in similar microemulsion systems with single non-ionic surfactants. By shielding the electric charges by the addition of salt, the oil continuous HIPME structures are disturbed what can be concluded from an increased conductivity and mobility of the water fraction, followed by NMR. Summary 9 1.2. German Version Mikroemulsionen bestehen im einfachsten Fall aus Wasser, Öl und Tensid(en). Es handelt sich dabei im Gegensatz zu normalen Emulsionen um transparente, thermodynamisch stabile Phasen. Diesen makroskopisch einphasig erscheinenden Systemen liegen jedoch hoch komplexe Nanostrukturen zu Grunde. Die in wissenschaftlicher Hinsicht bislang am besten untersuchten und verstandenen Mikroemulsionssysteme bestehen entweder aus Wasser, Öl und einem einzigen elektrisch ungeladenen nicht-ionischen Tensid oder einem elektrisch geladenen ionischen Tensid. Beide Systeme unterscheiden sich grundlegend, unter anderem in ihrem Phasenverhalten, ihrer Temperaturstabilität oder ihren Nanostrukturen. Systeme mit Mischungen aus ionischen und nichtionischen Tensiden dagegen wurden bisher kaum untersucht. Im Rahmen dieser Arbeit wurde das Phasenverhalten einer anionischen/nichtionischen Tensidmischung mit verschiedenen Ölen bei konstanter Temperatur und konstantem Tensidgehalt untersucht. Die Phasendiagramme weisen jeweils zwei optisch isotrope Phasengebiete, so genannte Einphasenkanäle, mit steigendem Öl-Gehalt auf. Die beiden Mikroemulsions-Einphasenkanäle sind voneinander durch ein optisch anisotropes Phasengebiet getrennt. Der Mikroemulsionskanal unterhalb des anisotropen Bereichs erstreckt sich von der wässrigen Phase ausgehend mit wachsendem Ölund nichtionischen Co-TensidAnteil bis in die Mitte des Phasendiagramms und endet dort. Der obere Einphasenkanal verläuft durch ein steiles Minimum, in Bezug auf das Tensid/Co-Tensidverhältnis, durchgehend von der wässrigen zur ölreichen Seite des Phasendiagramms. Im Gegensatz zu Mikroemulsionen mit nichtionischen Tensiden handelt es sich um isotherme Einphasenkanäle. Die einphasigen Gebiete wurden mit diversen physikalischchemischen Methoden untersucht. Mittels Leitfähigkeits-, SANS-, PFG-NMR-Messungen und elektronenmikroskopischen cryo-TEM Aufnahmen konnten die Nanostrukturen identifiziert werden. Während im unteren Einphasenkanal die Strukturen aus kleinen Öl-Tröpfchen in einer kontinuierlichen Wasserphase bestehen, welche mit zunehmendem Öl-Gehalt anschwellen, kommt es im oberen Einphasenkanal zu einer komplexen Strukturänderung. Während der ölfreien Probe eine bikontinuierliche Schwammstruktur zu Grunde liegt, wandelt sich diese mit bereits wenigen Prozent an Öl zu einer polyedrischen Wasser-in-Öl Schaumstruktur. Für diese, in Mikroemulsionen bislang unbekannten, Struktur wurde der Begriff High Internal Phase Microemulsion (HIPME) eingeführt, aufgrund ihrer strukturellen Parallelen zu bereits bekannten High Internal Phase Emulsionen (HIPE). Mittels transienter Elektrodoppelbrechung konnte dieser komplexe strukturelle Übergang nachvollzogen werden. Die ermittelten strukturellen Relaxationszeiten, welche zudem die Viskosität der Mikroemulsionen bestimmen, weisen ein deutliches Maximum am Übergangspunkt von der bikontinuierlichen zur HIPME-Struktur auf. Grund für die beobachtete HIPME-Struktur ist vermutlich der Anteil der elektrischen Ladung des anionischen Tensids. Diese sorgt für eine vergleichbar hohe Grenzflächenspannung zwischen der wässrigen verdünnten Tensid-Phase und des Öls. Konsequenz dieser hohen Grenzflächenspannung sind ölkontinuierliche Schaumstrukturen anstatt bikontinuierlicher Strukturen, welche man in vergleichbaren Mikroemulsionen mit rein nichtionischen Tensiden erhält. Durch Abschirmen der elektrischen Ladungen mit Salz werden die HIPME-Strukturen gestört, was sich in einem Ansteigen der Leitfähigkeit und einer erhöhten Mobilität der Wasserphase äußert, welche mit NMR beobachtet wurde. Introduction 16 The changing nanostructures can be easily followed by a steady decrease of the electric conductivity within the channel with increasing mass fraction of oil, when little salt is added to the system. 33 The occurrence of the single-phase channel has to do with the change of the amphiphilic properties of the non-ionic surfactant as non-ionic surfactants of the type C i E j become more lipophilic with increasing temperature. 34 With increasing temperature, the curvature of the amphiphilic monolayer changes from convex to flat and finally to concave (Fig. 2.6). The reason for this lies in the shrinking of the hydrophilic head-groups (EO-groups). At low temperatures, the size of the surfactant head group is larger than that of the hydrophobic chain, leading to an amphiphilic film curved around the oil. By increasing the temperature, the size of the EO-head group is shrinking, whereas the size of hydrophobic chain increases due to the increasing number of chain conformations and the increasing penetration of oil molecules. These trends lead to a gradual change of the interfacial curvature. 35 As a consequence the microemulsion structures change, from the water side, from small oil droplets in water, to bicontinuous structures in the middle of the phase diagram, to w/o droplets on the oil side. Fig. 2.6 Mean curvature H of a non-ionic surfactant film at the oil/water interface with increasing temperature. The size of the hydrophilic EO-head group is shrinking whereas the size of the liphophilic chain is increasing by raising temperature. The interfacial tension between the oil and the water is a sensitive parameter for the change of the interfacial curvature. Optimum solubilisation of oil in a system with non-ionic surfactants is obtained 33 M. Kahlweit, et. al, J. Colloid Interf. Sci 1987, 118, 450. 34 K.. Shinoda, Proceedings of the 5 th International Congress of Surface Activity, Barcelona, Spain, Vol. 2, 1969, 275-283. 35 J. Lyklema, in: Fundamentals of interface and colloid science, Volume V: Soft Colloids, Academic Press Inc, 2005. Introduction 17 at the minimum of the interfacial tension of the diluted aqueous surfactant solution against the oil. 36 For non-ionic surfactants, ultra-low interfacial tensions against the oil-phase are observed that can be as low as 10 -3 mN/m. A typical curve for the change of the interfacial tension is shown in Fig. 2.7. Fig. 2.7 Ultra low interfacial tension of the diluted non-ionic surfactant C 10 E 4 against alkane oils with the chain length k between 8 and 14. Bicontinuous microemulsions are obtained in the region of the minimum of the interfacial tension, as low interfacial tensions allow non-spherical structures due to the equilibrium in the interfacial curvature. 2.4. Microemulsions with ionic surfactants The situation in microemulsion systems with ionic surfactant is very different compared to microemulsions with non-ionic surfactants. The probably most investigated systems are those with AOT (sodium di-2-ethylhexylsulfosuccinate), decane, H 2 O and DDAB (didodecyldimethylammonium bromide), dodecane, H 2 O. 37 , 38 The single phase regions in such systems are usually plotted in triangular phase diagrams (Gibbs phase diagrams). An example for such a phase diagram of the system DDAB, dodecane, H 2 O is shown in figure 2.8. The isotropic phase regions (microemulsions) in these triangle presentations are very different for those of non-ionic surfactants. 39 Systems with ionic 36 T. Sottmann and R. Strey, J. Chem. Phys. 1997, 106, 8606–8615 37 M. Kotlarchyk, S.-H. Chen, J. S. Huang, M. W. Kim, Phys. Rev. Lett. 1984, 53, 941-944. 38 K. Fontell, A. Ceglie, B. Lindman, B. Ninham, Acta Chemica Scandinavica A40 1986, 247-256. 39 M. Kahlweit, R. Strey, Angew. Chern. Int. Ed. Engl. 1988, 24, 654-668. Introduction 18 surfactants, which have been studied usually, contain large isotropic regions in the middle of the triangle while phase diagrams with non-ionic surfactants are very different and contain narrow isotropic channels. However, microemulsion systems with single ionic surfactants channels do not have single phase channels that pass from the aqueous side without crossing a phase boundary continuously to the oil side of the phase diagram. Fig. 2.8 Ternary phase diagram of the system DDAB, water and dodecane. The single phase microemulsion area L 2 is indicated as yellow. The nanostructures of microemulsions with ionic surfactants have been investigated in detail by SANS, SAXS, electrical conductivity and finally imaged by freeze fracture electron microscopy. 40 The results showed that the morphology of the microemulsions for equal amounts of water and oil is completely different to those of non-ionic surfactants. Instead of a bicontinuous microemulsion phase, there was found a water-in-oil droplet structure. In Figure 2.9, two FF-TEM micrographs show the droplet structure of the systems AOT-decane-H 2 O and DDAB-dodecane-H 2 O. It is possible to dilute these phases with oil without the droplets would change in size or structure. The droplet structure is caused by the electric charges of the ionic surfactant that result in higher interfacial tensions compared to microemulsion systems with non-ionic surfactants. However, it was shown by conductivity experiments that the oil-continuous w/o-droplet structures at equal amounts of water and oil can be transformed to samples that have similar bicontinuous features as microemulsions with non-ionic surfactants by shielding the charge on the surfactant layers by excess salt. 41 40 W. Jahn, R. Strey, J. Phys. Chem. 1988, 92, 2294-2301. 41 W. Sager, W. Sun, H.-F. Eicke, Progr. Colloid Polym. Sci. 1992, 89, 284-287. Introduction 19 Fig. 2.9 FF-TEM micrographs of w/o-microemulsions with ionic surfactants at equal amounts of water and oil. Left micrograph: droplet structure of the system AOT (20 wt %), decane and water, right picture: aggregated structure of deformed droplets of the system DDAB (13.9 wt %), dodecane and water. Scale bar = 200 nm. In addition, microemulsions from ionic surfactants are less sensitive to temperature changes and stable over a wider temperature range than microemulsions with non-ionic surfactants, which quickly can drop out of phase when the temperature is not adjusted accurately by a few degrees. 2.5. Objectives of this thesis Microemulsion systems with single non-ionic surfactants and single ionic surfactants are investigated in detail and well understood. However, there are only few studies of properties of microemulsions with mixed surfactants. The behaviour of mixtures with different surfactants is not easy to predict, as the temperature effect on the solubility and the phase behaviour of non-ionic and ionic surfactants are very different. 42 It should be noted that some results on the influence of ionic surfactants on phase diagrams of non-ionic microemulsions have already been published. 43 It was observed that the isotropic channels were widened by the influence of ionic surfactants and that the surfactant efficiency was increased. 44 Nevertheless, there is no information on the structures of the isotropic channel, and no detailed phase diagrams of a four-component system were established. 42 H. Kunieda, K. Hanno, S. Yamaguchi, K. Shinoda, J. Colloid Interf. Sci. 1985, 107, 129-137. 43 M. Kahlweit, B. Faulhaber, G. Busse, Langmuir 1994, 10, 2528-2532. 44 J. A. Silas, E. W. Kaler, Langmuir 2001, 17, 4534-4539. Introduction 20 As shown in 2.3, the evolution of the different microemulsion structures in the single phase channel of non-ionic surfactants is based on the change of the interfacial curvature with temperature. It is known that the interfacial curvature can also be controlled by the interfacial composition, when a hydrophilic surfactant is mixed with a lipophilic co-surfactant. 45 Such a situation is shown in Figure 2.10. Fig. 2.10 Change of the mean curvature H by mixing a hydrophilic with a hydrophobic surfactant. The decrease of H with increasing amount of the hydrophobic surfactant in the composition of the interface δ V,I is caused by the smaller head group of the hydrophobic co-surfactant compared to the hydrophilic surfactant. The aim of this work was to establish a microemulsion system with a surfactant mixture based on an anionic hydrophilic and a lipophilic co-surfactant and to prepare a phase diagram of a four-component system at constant temperature. As it is well known that the increase of temperature makes ionic surfactants more hydrophilic while non-ionic surfactants become more lipophilic 46 , it was hoped that these both effects would compensate in such a microemulsion system and therefore result in a system that is mostly independent to temperature variations. This is of great interest for possible applications, for example in the field of cosmetic industry or other areas like enhanced oil recovery. In addition, the similarities and the differences of such mixed systems compared to the basic non-ionic or ionic microemulsion systems should be investigated by different physicochemical methods, as conductivity, rheology, PFG-NMR and electron microscopy. 45 J. Reimer, O. Södermann, T. Sottmann, K. Kluge, R. Strey, Langmuir 2003, 19, 10692-10702. 46 S. Ajith, A. K. Rakshit, J. Phys. Chem 1995, 99, 14778-14783. Synopsis 21 3. Synopsis 3.1. The surfactant system Ca(DS) 2 /Mg(DS) 2 – IT 3 In the publication „Microemulsions from silicone oil with an anionic/non-ionic surfactant mixture“, we first introduced a new surfactant combination, that turned out to be an interesting choice for the preparation of microemulsions. The main idea was to establish a microemulsion phase diagram in the style of non-ionic microemulsion systems, in which isotropic microemulsion single phase channels exist that run from the aqueous side of the phase diagram to the oil-side without passing a phase boundary. Instead of varying the temperature to influence the interfacial curvature of the surfactant film, we wanted to establish such single phase channels in an isothermal microemulsion system by adjusting the mixing ratio of two different surfactants, namely a hydrophilic and a lipophilic one. Although the idea to use surfactant mixtures for the preparation of microemulsions is not new, no detailed phase diagrams with such a system have been investigated yet. As hydrophilic surfactant, we chose calciumdodecylsulfate, as it was already known that Ca 2+ -salt of SDS can form lamellar phases and even sponge-like phases, when it is mixed with short chain alcohols as co-surfactants. This is not possible for normal SDS. As lipophilic co-surfactant we chose an industrial non-ionic surfactant that is based on a highly branched isotridecanol which is etherified with an average number of three ethylene-oxide groups (isotridecyltriethyleneglycolether, abbreviated as IT 3). By mixing both surfactants, we received a huge variety of different phases, starting by a micellar L 1 phase with the pure Ca(DS) 2 , going over to a large birefringent lamellar area with increasing mass fraction of the lipophilic co-surfactant and finally ending with a two-phase situation in that the IT 3 is forming inverse micelles which are separating from the lower aqueous phase (Fig. 3.1). Rheological results indicated that the lamellar structures change from multilamellar vesicles to planar lamellas with increasing mass fraction of the lipophilic co-surfactant. Fig. 3.1 Surfactant mixtures of Ca(DS) 2 (two upper rows) or Mg(DS) 2 (lower row) with increasing mass fraction x of IT 3. First row shows surfactant mixtures with Ca(DS) 2 in direct light. Second and third rows show samples with Ca(DS) 2 and Mg(DS) 2 between crossed polarizers. Samples prepared with a total surfactant concentration of 15% (w/w), phases observed at T = 40 °C. Synopsis 22 An amazing feature of the surfactant mixture is the existence of an optical isotropic L 3 sponge phase, as L 3 phases with charged surfactants are very rare. When using the Ca(DS) 2 in the surfactant mixture, one has to consider that samples with high Ca(DS) 2 crystallize at room temperature due to its high Krafft Temperature of K T = 60 °C. Therefore we also used the Mg 2+ -salt of SDS, as it has a lower K T of 25 °C. Nevertheless, the surfactant phase sequence as well as the properties of the samples are the same, when replacing the Ca(DS) 2 by Mg(DS) 2 . Another important result in the publication is the observation, that the surfactant mixture itself has a high interfacial tension in diluted aqueous solutions against oil. For microemulsion systems with nonionic surfactants, ultra-low interfacial tensions are reported. The high interfacial tensions of the mixed surfactant system are probably caused by an influence of the electric charges at the surfactant monolayer. This turned out to have consequences for the nanostructure of the microemulsions that were investigated later. 3.2. Solubilisation of oil into the surfactant mixture In a next step, we solubilized oil into the surfactant mixtures to search for isotropic microemulsion areas. As oil we chose the silicone oil hexamethyldisiloxane, as no microemulsion phase diagram with silicone oil has been established before. Nevertheless, we knew from previous solubilisation experiments, that hexamethyldisiloxane (M 2 ) should behave similar as decane. In figure 3.2, the different phases of the surfactant mixtures with increasing amount of oil in the solvent mixture are summarized in a phase diagram. Fig. 3.2 Phase diagram of the system Ca(DS) 2 /IT 3 – H 2 O/M 2 with 15% (w/w) surfactant and 85% (w/w) solvent. Phases observed at 40 °C. Abbreviation ‘‘ME’’ stands for ‘‘microemulsion’’ and indicates area of isotropic microemulsion channels. Synopsis 23 All samples were prepared with a constant surfactant concentration of 15% (w/w) and the phase behaviour was investigated at 40 °C. Most important results are the presence of two isotropic microemulsion single phase channels, which are separated by a birefringent anisotropic area. The upper microemulsion channel extends from the aqueous side of the phase diagram continuously to the oil-side, while the lower single phase channel ends in the middle of the phase diagram at equal amounts of water and oil. As already mentioned, such microemulsion channels also exist in phase diagrams with single nonionic surfactants, where the hydrophilic-lipophilic balance is adjusted by raising temperature. An example of such a phase diagram with C 12 E 5 as surfactant and tetradecane as oil is shown in figure 3.3. 47 Fig. 3.3 Phase diagram of the system C 12 E 5 – H 2 O/Tetradecane with constant 16.6 wt % surfactant. W M = water continuous microemulsion, O M = oil continuous microemulsion, D = bicontinuous microemulsion, L.L.C = lyotropic liquid crystalline. In such systems the nanostructure on the microemulsion changes with increasing temperature and with increasing oil content from oil-droplets in a continuous water phase to bicontinuous microemulsions at equal amounts of water and oil and finally to water-droplets in a continuous oil-phase at the oil-rich 47 U. Olsson, K. Shinoda, B. Lindman, Journal of Physical Chemistry, 1986, 90, 4083. Synopsis 24 side of the phase diagram. The upper channel that starts at the isotropic L 3 sponge phase at higher temperature joins the lower channel in the first half of the phase diagram. An important difference between both shown phase diagrams is the fact, that the microemulsion channels are not connected with each other in the mixed anionic-non-ionic surfactant system. Although microemulsions from both channels are transparent and optical isotropic, samples in the upper single phase channel are somewhat bluish and show some shear induced birefringence with up to 40% of oil in the solvent mixture. This is a first indication, that the nanostructures in both channels must be very different. Conductivity experiments finally gave evidence, that this is indeed the case. While the conductivity values in the lower single phase channel stayed about the same, there was found an abrupt decrease of the electrical conductivity in the beginning of the upper single phase channel with only about 5% of oil in the solvent mixture. At equal amounts of oil and water, no conductivity could be detected any longer. We concluded from the results, that the nanostructure in the upper channel might switch from a bicontinuous morphology to a water-in-oil system, while the oil-inwater structures in the lower channel remain the same with increasing oil-concentration. 3.3. Cryo-TEM imaging of the microemulsion system Ca(DS) 2 /IT 3 – H 2 O/M 2 To verify the different nanostructures which were concluded from the experimental results that were presented in the first publication about the new microemulsion system, we tried to image them directly by cryogenic transmission electron microscopy (cryo-TEM) and discussed the results in a second publication. In the cryo-TEM technique, a small droplet of the sample is pipetted on a perforated carbon film that is supported by a TEM copper grid. With the help of a piece of filter paper, excess liquid is blotted away until a thin liquid film with a desired average thickness of about 200 nm is left in the holes of the perforated carbon. After blotting, the specimen is plunged into a suited cryogen for vitrification. After vitrification, the specimen is transferred under liquid nitrogen into the electron microscope. The specimen preparation can be performed in a so called controlled environment vitrification system (CEVS), in which both the atmosphere as well as the temperature can be controlled. 48 This point is very important for the specimen-preparation of temperature sensitive samples and/or samples that contain volatile compounds. For the preparation of the water-continuous samples, we used liquid ethane as cryogen as it provides cooling-rates that are fast enough to vitrify H 2 O and no ice-crystals should disturb the image. The imaging of the binary surfactant system could be performed without major problems and the results supported the rheological results. The cryo-TEM micrographs show a transition from multilamellar vesicles to planar lamellas with increasing mass fraction of IT 3 in the surfactant mixture. We finally observed the transition from a lamellar structure to a bicontinuous network-structure of the neighbouring L 3 phase (Figure 3.4). As the cryo-TEM method only shows a two-dimensional 48 Y. Talmon, in: “Seeing Giant Micelles by Cryogenic-Temperature Transmission Electron Microscopy (CryoTEM)”, in “Giant Micelles”, chapter 5, R. Zana, E. A. Kaler, Eds., CRC Press, New York, 2007, 163-178. Synopsis 25 projection of the various domains, the sponge-like character of the phase is not as well visible as with other methods as freeze fracture electron microscopy, that can reproduce better the three-dimensional morphology. 49 That is why the structure seen on the cryo-TEM micrograph could also be taken for a network of thread-like micelles instead of a bicontinuous sponge phase. Fig. 3.4 Cryo-TEM micrograph of sample with 15% (w/w) surfactant Ca(DS) 2 /IT 3, x IT 3 = 0.77, prepared at 40 °C. Micrograph shows bi-continuous network-structure of the L 3 phase. The imaging of the microemulsion phases with the silicon oil was more challenging. Although we tried to saturate the atmosphere in the CEVS, it was for example not possible to image clearly the structure of the microemulsions in the lower single phase channel. Instead of tiny microemulsion droplets, small unilamellar vesicles and also multilamellar vesicles were mostly imaged. The problems with the sample preparation were obviously caused by the high volatility of the short chain silicone oil, as relaxation experiments indicated. Hexamethyldisiloxane has evaporation rates which are comparable to acetone. In such relaxation experiments, a certain amount of time is stopped after the blotting-procedure during cryo-TEM specimen preparation, before the sample is plunged into the cryogen. Thereby it can be investigated if the nanostructure changes due to small changes in the sample composition. In figure 3.5, two cryo-TEM micrographs of the same sample from the lower single phase channel with 15% of oil in the solvent mixture are shown. While the micrograph of the quickly prepared specimen shows structures that remind to oil-swollen micelles and some multilamellar vesicles, the micrograph of the specimen that has been prepared with a delay of 30 seconds before the plunging shows collapsed structures and large (multilamellar) vesicles. The 49 H. Hoffmann, C. Thunig, U. Munkert, H. W. Meyer, W. Richter, Langmuir 1992, 8, 2629-2638. Synopsis 32 again. The final process is the longest with a relaxation time of 14 ms. Similar complicated signals have been observed for electric birefringence measurements on dispersion of clay particles. The complicated signals in these systems are due to the fact that a fraction of the particles align parallel to the electric fields, whereas another fraction aligns perpendicular to the field. 55 Fig. 3.11 EB-signal for a microemulsion of the upper single phase channel with 6% decane in the solvent mixture. a) signal shown with time sweep of 2 msec, b) signal with time sweep of 100 msec, longest relaxation τ L = 14 msec. The signals identify three different relaxation processes. Duration of electric pulse = 1 msec. 55 Y. Yamaguchi, H. Hoffmann, Colloids Surf. A 1997, 121, 67-80. 0,00 0,25 0,50 0,75 1,00 1,25 1,50 1,75 2,00 0,0 2,0x10 -8 4,0x10 -8 6,0x10 -8 8,0x10 -8 1,0x10 -7 1,2x10 -7 1,4x10 -7 1,6x10 -7 1,8x10 -7 2,0x10 -7 E-field turned off τ 3 τ 2 E = 273 kVm -1 ∆ ∆ ∆ ∆n t [msec] τ 1 E-field turned on τ 1 τ 2 a) 0 10 20 30 40 50 60 70 80 90 100 -2,0x10 -8 0,0 2,0x10 -8 4,0x10 -8 6,0x10 -8 8,0x10 -8 1,0x10 -7 1,2x10 -7 1,4x10 -7 1,6x10 -7 1,8x10 -7 2,0x10 -7 2,2x10 -7 ∆ ∆ ∆ ∆n t [msec] E = 273 kVm -1 τ L = 14 msec b) Synopsis 33 As the longest relaxation times of the final decay of the signal dominate over the shorter times, they were called main structural relaxation times. Signals were recorded all along the upper channel and the main relaxation times determined by analyzing the exponential decay of the final process. In addition, there was measured the zero-shear viscosity of the microemulsions. An overview of the main structural relaxation times and the zero shear viscosities in the upper single phase channel can be found in figure 3.12. Fig. 3.12 Plot of the longest relaxation time τ and the zero shear viscosity of microemulsions in the upper single-phase channel against the mass fraction of decane in the solvent mixture. The relaxation time increases three orders of magnitude and shows a maximum around 6-10% decane. At the same time, the zero shear viscosity increases two orders of magnitude. With further increasing mass fraction of decane, the relaxation time and the viscosity decrease again. The structural relaxation times around the maximum that were determined by electric birefringence also could be determined by rheology. As conductivity data already implied, there must be a complex transition mechanism from the bicontinuous L 3 phase to a water-in-oil structure with only around 6-10% oil. The complicated signals for microemulsions with low oil content are obviously related with this transition mechanism and show that several microstructures exist in the fluid which are in equilibrium with each other. For higher oil content above 10% oil, EB-Signals become less complicated again. This indicates that the structural transition is complete. The structural relaxation time then is decreasing again due to the 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,1 1 10 0 100 200 300 400 500 Zero-shear viscosity [mPas] x decane |η*| [mPas] τ[msec] Relaxation time τ[msec] 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,1 1 10 0 100 200 300 400 500 Zero-shear viscosity [mPas] x decane |η*| [mPas] τ[msec] Relaxation time τ[msec] Synopsis 34 decreasing droplet size of the w/o droplets with increasing oil content. The viscosity maximum is also the result of a switch in the nanostructure from a low viscous flexible L 3 phase to a densely packed w/o-HIPE structure. Obviously, the structural relaxation time controls the viscosity of the fluid. In the lower single phase channel, simple electric birefringence signals were found that are caused by the deformation of droplets in the electric field. This indicates that the nanostructure doesn’t change much in this channel. With increasing mass fraction of oil, the main structural relaxation time increases with the viscosity, as the droplet size increases (Figure 3.13). The initial decay of the viscosity is due to the transformation of worm-like micelles in binary surfactant mixture to spherical micelles by solubilizing oil. Fig. 3.13 Results of electric birefringence in the lower single phase channel. a) EB-signal of a microemulsion from the lower single phase channel with 20% decane in the solvent mixture. b) Overview of main structural relaxation time and zero shear viscosity with increasing oil content in the lower single phase channel. 0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 -1,0x10 -8 0,0 1,0x10 -8 2,0x10 -8 3,0x10 -8 4,0x10 -8 5,0x10 -8 6,0x10 -8 7,0x10 -8 8,0x10 -8 x IT 3 0.45 x decane 0.2 ∆ ∆ ∆ ∆ n t [msec] a) 0,00 0,05 0,10 0,15 0,20 0,25 0,30 0,35 0,40 0,01 0,1 1 10 x decane τ [msec] η [mPas] η η η η [mPas] τ τ τ τ [msec] b) 10 100 Synopsis 35 3.5. Cryo-TEM of microemulsions with a High Internal Phase Microemulsion (HIPME) structure As the conductivity and electric birefringence results showed the presence of an oil-continuous high internal phase droplet structure with less then 10% of oil, cryo-TEM experiments were performed to image these new microemulsion structures directly. The results were published in a fourth manuscript. Unfortunately, oil continuous phases with linear hydrocarbons like n-decane cannot be investigated by cryo-TEM. Linear hydrocarbons tend to crystallize when they are quenched into the liquid cryogen during the specimen preparation. Branched hydrocarbons, however, can be vitrified and do not crystallize. Therefore, we tried to add some iso-octane to the decane in order to stay as close as possible to the decane system and hoped that this would help to prevent the crystallization process. Good results cryo-TEM results of phases with linear oils that were mixed with branched compounds serving as cryoprotectants are already reported. 56 Cryo-TEM experiments with pure decane/iso-octane mixtures failed, it was not possible to gain pictures of clean vitrified oil-films. Instead, only dark crystal structures of the oil could be imaged, that were sensitive to the electron beam and suffered from irradiation damage. As a consequence, we gave up the idea to use a mixed oil. A cryo-TEM micrograph of such structures is shown in figure 3.14. The ratio of n-decane/iso-octane was 7/3. Fig. 3.14 Cryo-TEM micrograph of n-decane/i-octane, ratio 7/3 (w/w), prepared at RT, cryogen: liquid nitrogen. In another study, the sponge structure of a bicontinuous microemulsion with a single non-ionic surfactant was already successfully imaged by cryo-TEM. The sample was prepared with the 56 D. Danino, R. Gupta, F. Satayavolu, Y. Talmon, J. Colloid Interf. Sci. 2002, 249, 180-168. Synopsis 36 surfactant C 12 E 5 , iso-octane, H 2 O and had a water/oil ratio (w/w) of 7/3. The micrographs clearly show a sponge structure with typical dimensions of water and oil domains around 0.2 µm. Microemulsions of the new mixed surfactant system therefore were prepared with pure iso-octane for cryo-TEM investigations instead of decane. Fortunately, the microemulsion samples from the upper channel showed the same properties as samples with decane. The same abrupt decrease of the conductivity as well as the maximum of the viscosity around 6-10% of oil was found. Moreover, it was not necessary to adjust the surfactant/co-surfactant ratio to obtain single phase microemulsions, the phase composition was about the same. While the cryo-TEM pictures of the L 3 phase without oil showed the well-known bicontinuous network-structures, the situation changes drastically when only 6% of iso-octane are solubilized into the phase. In Figure 3.15, micrographs of the microemulsion from the upper channel with 6% of isooctane in the solvent mixture are shown. Fig. 3.15 Cryo-TEM micrographs of a microemulsion from the upper single phase channel with 15% surfactant Mg(DS) 2 /IT 3, x IT 3 0.62, x iso-octane 0.06, quenched from 25 °C. They clearly show a densely packed w/o-droplet structure, which look like a “polyhedral foam” that has thin films with a thickness of about 3 nm. The diameter of the water domains is about 50 nm. The thin film was obviously made of the surfactants with the hydrocarbon in between the two monolayers, and the water inside the polyhedra. The structures look very much like the structures found in high internal phase emulsions (HIPE). HIPEs are concentrated systems with a large volume of the dispersed phase. Those high volume fractions result in the deformation of droplets into polyhedra, which are separated by thin films of the continuous phase. In such situations the size of the structures are usually in the range of several µm and can easily be seen by light microscopy. Both o/w and w/o-systems of Synopsis 37 HIPEs are known. 57 Based on that, we call the newly found microemulsion structure HIPME (High Internal Phase Micro-Emulsion). Such structures have so far only been observed in real emulsions that are, like any other emulsions, thermodynamically unstable. 58 In the microemulsion sample with the HIPME structure, the water domains are not connected with each other. This can be concluded from the fact that the ice-crystals in dark have the same size as the ice crystals in grey. In spite of their small oil content, the HIMPE phases have a conductivity that is about 3 - 4 orders of magnitude lower than the conductivity of the L 3 phase. The thin surfactant films are practically impenetrable for the transport of ions, explaining the low conductivity of the sample. Microemulsions with a HIPME structure were imaged successfully all along the upper single phase channel (Fig. 3.16). Fig. 3.16 Cryo-TEM micrographs of a microemulsion with a HIPME structure. Left: sample containing 30% of iso-octane, right: sample containing 50% of iso-octane in the solvent mixture. The HIPME structures are obviously the result of the electric charges in the surfactant layers, contributed by the anionic Mg(DS) 2. They cause high interfacial tensions, leading to spherical structures instead of a bicontinuous morphology. It is likely, that transitions to bicontinuous structures can be provoked by shielding the electric double-layers in ionically charged systems by excess salt. 57 N. R. Cameron, D. S. Sherrington, Adv. Polym. Sci. 1996, 126, 163-214. 58 H. Hoffmann, G. Ebert, Angewandte Chemie 1988, 27, 902-912. Synopsis 38 3.6. PFG-NMR self diffusion measurements To underline and to verify the previously obtained results, the microemulsion channels were investigated by pulsed-field gradient nuclear magnetic resonance spectroscopy (PFG-NMR). This method is a special technique, based on nuclear magnetic resonance spectroscopy (NMR). In contrast to normal NMR measurements, inhomogeneous magnetic fields are used. With the help of additional magnetising coils, liner field-gradients are generated and thus locally changing magnetic fields. This allows performing spatially resolving NMR experiments, in which local information can be obtained like the position or the movement of particles, which are visible for NMR. Therefore, PFG-NMR can give detailed information about the structure, fluidity and emulsion type but also indications about the interaction between surfactant and co-surfactant at the interface. More details about the PFG-NMR method can be found in literature. 59 In a fifth publication, the results of PFG-NMR experiment in the microemulsion single phase channels are reported. In addition, the influence of excess salt to the system was investigated by interfacial tension, conductivity and NMR-measurements, as we tried to transform a high internal phase microemulsion to a bicontinuous microemulsion by shielding the electric charges of the anionic surfactant. 3.6.1. PFG-NMR self diffusion measurements in the single phase channels The PFG-NMR analysis is focussed on those spectral regions which can either be clearly assigned to single system constituents (water and decane) or to the mixture of the surfactants (Mg(DS) 2 /IT 3). The integrals of these three spectral regions strongly depend on the strength of the gradient pulse, hereby indicating the average displacement of the corresponding system constituents during the period between the pulses which was set to 50 ms. In a plot of the logarithmic signal intensity vs. the parameter γ²G²δ²(∆-δ/3) (with γ being the gyromagnetic ratio of protons, G the strength of the gradient field, δ and ∆ the duration of and the spacing between the two gradient pulses), the slope is equal to the negative apparent self diffusion coefficient of the given component in the heterogeneous system (Stejskal-Tanner plot). If the component is located in two different environments leading to clearly different self diffusion properties, the plot will show two sections with clearly different slopes. If the component is encapsulated in very small droplets, the motion within the droplets becomes undetectable. In this case, the observed slope reflects the velocity of the Brownian motion of the droplets. In Figure 3.17, a Stejskal-Tanner plot for water and the surfactant in the L 3 phase is shown. The water signal follows a steep decay, corresponding to a self diffusion constant that is just slightly lower than the value for free bulk water. This indicates that water forms a continuous phase that is only slightly affected by phase boundaries of the surfactant layers. In contrast, the signal for MDS/IT3 follows a relatively flat decay. This means there is only a restricted mobility of the Mg(DS) 2 and IT3 molecules. 59 P. Heitjans, J. Kärger, eds., in: Diffusion in condensed matter: methods, materials, models, Springer Germany, 2005, 421. Synopsis 39 Fig. 3.17 Stejskal-Tanner plot for water and the surfactants in the L 3 phase. The situation changes significantly when decane is added to the system. The mobility of water is much reduced. In contrast, the diffusion rates for decane exhibit values which come close to the bulk diffusion rate of decane. In this situation, the self diffusion constant for the water fraction is too small to be assigned to bulk water in a continuous phase. This shows that the observed water molecules are encapsulated in small droplets which undergo Brownian motion in the continuous decane phase. In Figure 3.18, the diffusion constants of H 2 O in the upper single phase channel are summarized with increasing decane content. Fig. 3.18 Self diffusion constants of H 2 O in the upper microemulsion channel as a function of the decane content. 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 1x10 -10 2x10 -10 3x10 -10 4x10 -10 5x10 -10 6x10 -10 7x10 -10 8x10 -10 diff [m 2 /s] % (x decane) diff [m/s] 0,60 0,65 0,70 0,75 0,80 0,85 0,90 sample composition x IT 3 a ) 0,0 5,0x10 10 1,0x10 11 -10 -8 -6 -4 -2 0 water MDS/IT 3 ln I/I 0 γ 2 G 2 δ 2 ( ∆ - δ /3) 0% decane (upper chanel) a ) Synopsis 40 The data for the self diffusion coefficients of water show a clear correlation with the corresponding conductivity plot that was shown in Fig. 3.8. In contrast to the results for the upper channel, the variations between the PFG-NMR results of different positions in the lower channel do not indicate dramatic structural changes. The data resemble those of the upper channel in absence of decane. All in all, the data are clearly in accordance with o/w microemulsions. 3.6.2. Influence of excess salt to the microemulsion system As already mentioned in 3.1, our mixed anionic/non-ionic surfactant system has a very high interfacial tension against the oil phase. In case of the microemulsion system with decane, the minimum interfacial tension of 2.5 mN/m is reached at a surfactant/co-surfactant ratio of 1:1. In non-ionic microemulsion systems, ultra low values as 1 . 10 -3 mN/m are reached. We assumed that by shielding the charge of the anionic surfactant by adding excess salt would lower the interfacial tension. Similar effects were already reported for the anionic surfactant diethylhexyl sodium sulphosuccinate (AOT), where ultra-low interfacial tensions against oil were reached with additional NaCl. 60 Nevertheless, the interfacial tension couldn’t be lowered significantly in our system, no ultra-low interfacial tensions were detected when excess NaCl was added to the surfactant mixtures. However, the upper and lower borders of the single phase regions are shifted to lower x IT 3 values by x IT 3 ~ 0.07 when NaCl is added to the Mg(DS) 2 in a molar ratio of 1:1. The shift to lower x IT 3 values means that the system in total becomes more lipophilic, as less amount of the lipophilic co-surfactant IT 3 in the surfactant mixture is needed to solubilize the oil. By measuring the electric conductivity of a microemulsion with 30% of decane with increasing salt concentration, it was checked if the shift of the phase boundaries might also accompanied by a change in the internal nanostructure of the microemulsion sample. A plot of the conductivity in the single phase region with increasing NaCl concentration is shown in figure 3.19. The conductivity from the NaCl-free to the microemulsion with a molar ratio of Mg(DS) 2 :NaCl = 1:1 increases about three orders of magnitude from a low value of 3 µS/cm to ~ 1000 µS/cm. The conductivity increases in a sigmoid curve with an inflection point around 50% NaCl and not linearly with increasing NaCl concentration. At first look it seems like the nanostructure changes from a w/o-HIPME system to a bicontinuous-like nanostructure. 60 R. Aveyard, B. P. Binks, S. Clark, J. Mead, J. Chem. Soc. Faraday Trans. 1 1986, 82, 125-142. Synopsis 41 Fig. 3.19 Plot of conductivity in the single phase region of a microemulsion with x decane 0.3 and increasing NaCl concentration at 25 °C. 100% NaCl corresponds to a molar ratio of Mg(DS)2:NaCl = 1:1. To proof this, we compared two microemulsions with different salt concentrations by PFG-NMR. The first sample without NaCl had the composition of x IT 3 0.7 and x decane 0.3. The second sample had the composition of x IT 3 0.615, x decane 0.3, and the molar ratio of Mg(DS)2:NaCl = 1:1 (Fig. 3.20). Fig. 3.20 Stejskal-Tanner plots for decane and water of a microemulsion from the upper single phase channel with 30% decane in the solvent mixture. a) plot for microemulsion without excess NaCl, b) plot with excess NaCl. The mobility of water is much reduced in the sample without NaCl. When NaCl is added to the system, the mobility of water is increasing. Nevertheless, the diffusion rate is much slower compared to the bicontinuous L 3 phase (Fig. 3.17) where it is close to bulk water. The diffusion behaviour of decane doesn’t change with NaCl. Therefore we conclude, that the main structure still is present as HIPMEphase. 0 10 20 30 40 50 60 70 80 90 100 1 10 100 1000 conductivity [µS/cm] sample composition NaCl [%] conductivity [µS/cm] 0,61 0,62 0,63 0,64 0,65 0,66 0,67 0,68 0,69 0,70 x IT3 0,0 5,0x10 10 1,0x10 11 -10 -8 -6 -4 -2 0 water peak decane peak ln I/I 0 γ 2 G 2 δ 2 ( ∆ - δ /3) probe without NaCl a ) 0,0 5,0x10 10 1,0x10 11 -10 -8 -6 -4 -2 0 water peak decane peak ln I/I 0 γ2 G 2δ2 ( ∆ - δ /3) probe contains NaCl b ) Publications 48  Microemulsions with a HIPME (High Internal Phase Microemulsion) structure Published in Journal of Physical Chemistry B, 2011 by Lukas Wolf*, Heinz Hoffmann, Takashi Teshigawara, Tohru Okamoto, Yeshayahu Talmon. DOI:. *corresponding author  PFG-NMR Self Diffusion Measurements in the Single Phase Channels of a Microemulsion System with an anionic/non-ionic Surfactant Mixture Submitted to Soft Matter in December 2011 by Lukas Wolf*, Heinz Hoffmann, Jürgen Linders and Christian Mayer. * corresponding author current status: under revision Microemulsions from silicone oil with an anionic/non-ionic surfactant mixture 49 5.1.1. Microemulsions from silicone oil with an anionic/non-ionic surfactant mixture Lukas Wolf*, Heinz Hoffmann*, Kei Watanabe and Tohru Okamoto *corresponding author Published in Physical Chemistry Chemical Physics 2011, 13, 3248-3256. DOI: 10.1039/C0CP00062K. 3248 Phys. Chem. Chem. Phys., 2011, 13, 3248–3256 This journal is cthe Owner Societies 2011 Cite this: Phys. Chem. Chem. Phys ., 2011, 13, 3248–3256 Microemulsions from silicone oil with an anionic/nonionic surfactant mixturew Lukas Wolf,* a Heinz Hoffmann,* a Kei Watanabe b and Tohru Okamoto b Received 30th March 2010, Accepted 2nd December 2010 DOI: 10.1039/c0cp00062k Microemulsion phases have been prepared for the first time from the silicone oil ‘‘M 2 ’’ (hexamethyldisiloxane) and a surfactant mixture of a nonionic surfactant ‘‘IT 3’’ (isotridecyltriethyleneglycolether) and an ionic surfactant Ca(DS) 2 (calciumdodecylsulfate). For such a surfactant mixture the hydrophilicity of the system can be tuned by the mixing ratio of the two components. With increasing IT 3 content, the surfactant mixtures show a L 1 -phase, a wide L a -region and a narrow L 3 sponge phase. For constant temperature, two single phase channels exist in the microemulsion system. The lower channel (low IT 3 content) ends in the middle of the phase diagram with equal amounts of water and oil, the upper channel begins with the L 3 -phase and passes all the way to the oil phase. Conductivity data show that the upper channel has a bi-continuous morphology up to 40% oil while the lower channel consists of oil droplets in water. In contrast to previous studies on nonionic systems, the two single phase channels are not connected and microemulsions with equal amount of oil and water do not have a bicontinuous structure. Introduction Microemulsions (ME) are thermodynamically stable fluids that consist of oil, water and a few percent of surfactant. 1 They usually are low viscous, more or less transparent and optically isotropic liquids. Microemulsions are used in many applications because of their fascinating properties. One of the reasons for their usefulness is that both polar and apolar compounds can be dissolved in microemulsions. They are therefore used for agrochemical, cosmetic, pharmaceutical and other industrial formulations where it is necessary to bring two components in a fluid together which normally are not miscible. Detailed investigations on microemulsions over the last 40 years have shown that the components of the fluids are not dispersed with each other on a molecular level like in miscible solvent mixtures but the fluids are highly structured on a scale of 1–100 nm. Domains of oil and water, with well defined interfaces from surfactants, alternate with each other. These domains are in equilibrium with each other and are very dynamic in behaviour. They constantly change their shape and size on a time scale of microto milliseconds. Today, the basic behaviour of nonionic microemulsions is well understood on the basis of theoretical models in which the mean curvature of interfaces, 1 the bending constants of surfactant monolayers 2 and the interfacial tension between oil and water 3 play a central role. Depending on the composition of the fluids the structures in the fluids can be oil swollen micelles (o/w-microemulsions), bicontinuous structures or water swollen inverse micelles (w/o-microemulsions). Microemulsions can be prepared either from nonionic or from ionic surfactants. 2–6 The single phase regions in systems that consist of oil, water and surfactants are usually plotted in triangular phase diagrams. The isotropic phase regions (microemulsions) in these triangle presentations are very different for ionic surfactants and for nonionic surfactants. 7,8 Systems with ionic surfactants, which have been studied usually, contain large isotropic regions in the middle of the triangle while phase diagrams with nonionic surfactants are very different and contain narrow isotropic channels. 9 For constant surfactant concentration, these channels usually pass from the aqueous side without crossing a phase boundary continuously to the oil side of the phase diagram. Detailed phase diagrams with many different systems in which the oil and the surfactants were varied have shown that two channels exist in microemulsions with nonionic surfactants and with variable temperature. 10 The two channels are connected with each other in the middle of the phase diagram and the phase region in between the channels contains an L a -phase. Small angle neutron scattering measurements have a University of Bayreuth, BZKG/BayColl, Gottlieb-Keim-Str. 60, 95448 Bayreuth, Germany. E-mail: heinz.hoff[email protected], [email protected] b Shiseido Research Center, 2-2-1 Hayabuchi, Tsuzuki-ku, Yokohama, Japan 224-8558. E-mail: [email protected] wElectronic supplementary information (ESI) available: Test tube pictures for the established phase diagram. See DOI: 10.1039/ c0cp00062k PCCP www.rsc.org/pccp PAPER 50 This journal is cthe Owner Societies 2011 Phys. Chem. Chem. Phys., 2011, 13, 3248–3256 3249 shown that the micellar structures in the low temperature channel evolve from globular o/w structures, transform to bicontinuous structures in the cross-over region and change continuously to globular w/o structures in the high temperature channel. 11 The structures in the upper channel are of a bicontinuous nature right from the aqueous side. 12 The goal of this investigation was to prepare microemulsions from the highly volatile silicone oil hexamethyldisiloxane (M 2 ), water and a nonionic/anionic surfactant mixture. Microemulsions from several silicone oils and nonionic surfactants have already been published. 13 But as far as we know, a phase diagram with the silicone oil M 2 had not yet been prepared and investigated in detail. It was known however that M 2 behaves similarly in solubilisation experiments as decane. 14 Viscoelastic solutions with wormlike micelles are transformed to low viscous micellar solutions with the same molar amount of decane or of M 2 . The saturation concentration for both oils is also about the same. It could therefore be expected that microemulsions from M 2 can be formed and the phase diagram with M 2 should look similar for decane for the same surfactant system. To achieve our goal, a mixed nonionic/anionic surfactant system was used. This system had to be optimized for the temperature region at which the microemulsion was to be prepared. Surfactants for the formation of microemulsions can be optimized by interfacial tension measurements. 15 Maximum solubilisation occurs when the interfacial tension of a dilute surfactant solution has its lowest interfacial tension against the oil phase. 16 Furthermore it is known that low interfacial tensions are observed for surfactant systems which form liquid crystalline L a -phases or L 3 -phases at low surfactant concentrations of a few percent. 17 Ideal surfactant systems for the formation of microemulsions can therefore be recognised based on their phase behaviour. 18 Ionic surfactants like sodiumdodecylsulfate are very hydrophilic and do not give L a -phases when mixed with water soluble nonionic surfactants. The situation is improved when Na + -ions are replaced by bivalent Ca 2+ -ions. Previous investigations had shown that in mixtures of SDS and tetradecyldimethylaminoxide the L 1 -phase could be transformed to a L a -phase simply by substituting the Na + -ions by Ca 2+ -ions. 19 Solutions of Ca(DS) 2 on their own are still L 1 -phases in which wormlike micelles are present. 20 If such solutions are mixed with nonionic compounds that are generally used as co-surfactants one obtains L a -phases in a wide composition region. As a co-surfactant the nonionic surfactant IT 3 (isotridecyltriethyleneglycolether) was used. It is commercially produced and is worldwide available under the name ‘‘Marlipal O13/30’’ as other well-known nonionic surfactants like ‘‘Triton X 100’’. 21 The compound is a highly branched isotridecanol, etherified with average 3 EO-groups. Despite its polydispersity, it acts mostly like a pure surfactant for the observed phases. In a general sense the optimum condition for microemulsions for a particular temperature can be adjusted by mixing a hydrophilic surfactant that forms a L 1 -phase and a lipophilic surfactant that is not soluble in the aqueous phase but forms its own L 2 -phase. 22 The used surfactant mixture in our study thus consists of an ionic surfactant that is just a bit too hydrophilic to form a L a -phase and a co-surfactant that is a bit too lipophilic to form a L a -phase on its own. The combination of the two surfactants forms a L a -phase over a wide surfactant composition. For the aim of this project, we used this surfactant-system which could form L a -phases and hopefully would form microemulsions of silicone oils. It should be noted that some results on the influence of ionic surfactants on phase diagrams of nonionic microemulsions have already been published. 23 It was observed that the isotropic channels were widened by the influence of ionic surfactants. There was however no information on the structures of the isotropic channel. As far as we know, our investigations on the described microemulsion results are really the first with a detailed phase diagram that had been established with a nonionic/ionic surfactant mixture. Some of the results turned out to be surprising and could not have been expected on the basis of previous results. The main novel features are: an L 3 -phase with an anionic/nonionic surfactant mixture, a microemulsion with equal amounts of oil and water that cannot have a bicontinuous structure and two isotropic channels that are not connected. Results and discussion The surfactant mixtures In Fig. 1, samples of mixtures of two surfactants are shown that fulfil the criteria described in the Introduction. The surfactants are Caor Mg-dodecyl sulfate (Ca(DS) 2 ,or Mg(DS) 2 ) and isotridecyltriethyleneglycolether IT 3. The Caor Mg-salts can be easily prepared from sodiumdodecylsulfate, and IT 3 is commercially available. Ca(DS) 2 has a high Krafft-temperature of K T =501C and can therefore only be used at higher temperatures or at room temperature, when the Krafft-temperature has been reduced by the second component to room temperature. 24 Mg(DS) 2 has a lower Krafft-temperature of K T =251C. Fig. 1 shows three rows of samples of the surfactant mixtures. The first two rows contain samples of Ca(DS) 2 with increasing mass fraction of IT 3. The total concentration is constant at 15% (w/w) and temperature is at 40 1C. The lower row of the two shows the same samples between crossed polarizers and a third row shows samples with Mg(DS) 2 and IT 3 again with polarizers. The Ca(DS) 2 /IT 3 mixtures contain crystalline precipitates up to a mass fraction of xIT 3 = 0.4. The samples from xIT 3 = 0.5 to 0.75 are birefringent and contain a highly swollen L a -phase. The sample with xIT 3 = 0.8 is optically isotropic and has a low viscosity. It is a L 3 -phase. The L 3 -phase can be recognised based on its microstructure from cryo-TEM measurements. 25 Surprisingly, the samples with an even higher mass ratio of IT 3 contain two phases with a L a -phase as the larger volume fraction. For normal behaviour of hydrophilic and lipophilic mixtures one would have expected to find multiphase regions without an L a -phase. 26 It is conceivable that the two components are no longer miscible in bilayers and phases are formed with different mixing ratios. It is noteworthy to mention that in the combination of Ca(DS) 2 and IT 3 a sample with a two phase L a /L 3 -situation could not 51 3250 Phys. Chem. Chem. Phys., 2011, 13, 3248–3256 This journal is cthe Owner Societies 2011 be observed when the composition of the samples was varied in small steps. This is shown in Fig. 2. Even for the condition when the composition was varied stepwise one percent by one percent the samples were within the single L a -phase region or in the single L 3 -phase region. The samples with the Mg(DS) 2 /IT 3 mixtures form more or less the same phases as the Ca(DS) 2 mixtures. The mixtures with xIT 3 = 0.1 and 0.2 are viscous single phase regions of an optically isotropic L 1 -phase and the sample with xIT 3 = 0.3 is a L 1 /L a two phase sample. The samples with xIT 3 = 0.8 show again a L 3 -phase, and the sample with xIT 3 = 0.9 contains a two phase region with one phase being a L a -phase. It is remarkable that the mixtures with ionic surfactants form a L 3 -phase. 27 Many investigations have shown that neutral L 3 -phases are transformed to L a -phases when small fractions of the surfactants are replaced by ionic surfactants. 28 The existence of the L 3 -phase in the investigated system is probably linked to the fact that the bi-valent metal-ions bind strongly to the dodecylsulfate and the surfactants show only weak dissociation. These surfactants behave therefore very similar like nonionic surfactants. While the existence of the L 3 -phase in the used system at xIT 3 = 0.76 is unexpected, it is noteworthy that such a phase had also been observed in the phase diagram of Ca(DS) 2 with octanol or lower chain length alcohols. 24 In this investigation it was furthermore shown that the L 3 -phase disappeared when 10% of the Ca(DS) 2 was replaced by SDS. What was even more surprising was the appearance of aL a -phase for co-surfactant ratios well above the existence region of the L 3 -phase. These observations show that the observed sequence of phases for the present system cannot be due to the fact that the used co-surfactant IT 3 is not a pure and single component but must be caused by the bivalent counter-ions. The experimental behaviour furthermore points out that the theoretical understanding of multicomponent systems leaves much to be desired. More experimental results in combination with theoretical models are necessary to come to a better understanding of such complex systems. Surface and interfacial tension measurements The samples with the highest solubilisation capacity should be samples with the lowest surface or interfacial tension values. These parameters were therefore measured for a 0.5% surfactant solution as a function of the mixing parameter. Plots of the surface tension measurements are shown in Fig. 3A for the Ca(DS) 2 /IT 3 system, and in Fig. 3B for the interfacial tension values for the Mg(DS) 2 /IT 3 system against the silicone-oil hexamethyldisiloxane. The lowest surface tension (sE26.4 mN m 1 ) is obtained at xIT 3 = 0.55 while the lowest interfacial tension is obtained at xIT 3 = 0.4 with gE4mNm 1 . These measurements confirm that the samples, that have a composition in the range of a L a -phase, have indeed low sand gvalues. It should be noted however that the interfacial tension between the oil and water phases in the presence of nonionic surfactants, that results in the formation of microemulsions, is usually in the range of 1 10 3 mN m 1 . 29 That is three orders of magnitude lower. It is conceivable that this difference comes from the fact that the used surfactant mixture contains an Fig. 1 Surfactant mixtures of Ca(DS) 2 (two upper rows) or Mg(DS) 2 (lower row) with increasing mass fraction xof IT 3. First row shows surfactant mixtures with Ca(DS) 2 in direct light. Second and third rows show samples with Ca(DS) 2 and Mg(DS) 2 between crossed polarizers. Samples prepared with a total surfactant concentration of 15% (w/w), phases observed at T=401C. Fig. 2 Surfactant mixture Ca(DS) 2 /IT 3 with xIT 3 0.75–0.77, 15% surfactant (w/w), shown at 401C between crossed polarizers. 52 This journal is cthe Owner Societies 2011 Phys. Chem. Chem. Phys., 2011, 13, 3248–3256 3251 ionic surfactant and the charges were not shielded by excess salt. Surfactant phases with an ionic surfactant and excess salt can have extremely low interfacial tension against the oil phase as observed by C. A. Miller et al. 30 The interfacial tension of double chain surfactants, which have been used for the preparation of microemulsions, against oil is also in the range of 0.1–1 mN m 1 . 31 It is conceivable that the higher interfacial tensions of the charged surfactant systems are the reason for the different behaviour of the investigated system from microemulsions with nonionic surfactants. Rheological results The phases with different mixing ratios of the two components have different rheological properties. The pure Mg(DS) 2 - solution is already a viscous ‘‘shear thinning solution’’. The L a -phases are viscoelastic phases with a yield stress value. They have already a weak gel-character. The L 3 -phase finally is a low viscous Newtonian solution. Rheograms of samples in the L a -region at 40 1C were measured and are shown in Fig. 4. At a mass fraction of 0.5 (Fig. 4A), the storage modulus G0 runs above the loss modulus G00 at a level of 100 Pa and is independent of the frequency fat a shear stress of t= 5 Pa. With increasing mass fraction of IT 3, the level of G0first keeps constant at 100 Pa at xIT 3 = 0.6, and then drops to B8Paat xIT 3 = 0.7 at tE0.1 Pa, but is still independent of f(Fig. 4C). The decay of the storage modulus G0with increasing mass fraction of IT 3 and the fact that G0breaks in at lower shear stress tshow that the L a -phase looses its gellike character with increasing xIT 3. At xIT 3 = 0.75, measured at tE0.2 Pa (Fig. 4D), G0increases in the double Fig. 3 Surface tension of the surfactant system Ca(DS) 2 /IT 3 (A) and interfacial tension of the surfactant system Mg(DS) 2 /IT 3 against silicone oil hexamethyldisiloxane (B), measured at 25 1C, surfactant concentration 0.5% (w/w). Fig. 4 Rheograms of samples in the L a -region, surfactant system Ca(DS) 2 /IT 3, surfactant concentration 15% (w/w), measured at 40 1C. (A) xIT 3 = 0.5, measured at t= 5 Pa. (B) xIT 3 = 0.6, measured at t= 5 Pa. (C) xIT 3 = 0.7, measured at t= 0.1 Pa. (D) xIT 3 = 0.75, measured at t= 0.2 Pa. 53 3252 Phys. Chem. Chem. Phys., 2011, 13, 3248–3256 This journal is cthe Owner Societies 2011 log plot with a slope of 2 while G00 increases with a slope of one, indicating that the L a -phase lost most of its gel-like character. By reason of the rheological results, we assume a structural change in the L a -phase from multilamellar vesicles to planar lamellas with increasing xIT 3. The change of the rheological properties from the L a to the L 3 -phase by a very small change of the binary surfactant mixture is one of the most startling effects in surfactant science. In the present situation, the difference of the properties of the two phases is especially large because one of the components is ionically charged. But even for systems with nonionic surfactants and co-surfactants the differences between the two phases are remarkable. 18 It should be mentioned that both phases consist of bilayer structures. The origin of the huge change in the rheological and other properties lies in the formation of passages between adjacent bilayers when the L 3 -phase is approached. 32 Solubilisation of hexamethyldisiloxane (M 2 ) in the different surfactant mixtures The surfactant mixtures with 15% (w/w) of surfactant and at different mixing ratios were used to solubilise increasing amounts of oil. Photos of test tubes of samples between crossed polarizers can be seen in the ESI.wThe phase behaviour of the samples was plotted in a phase diagram (Fig. 5). The special features of the diagram are two isotropic phase channels, a lower channel that begins at the water side at xIT 3 = 0.1 and a higher channel that begins around xIT 3 = 0.8, the mass fraction of the L 3 -phase. The upper channel first decreases with increasing oil content and then increases slowly with further increase in oil content to the oil corner. The lower channel increases smoothly and seems to end in the middle phase region. In between the two channels are single phase L a -regions or multiphase regions with one phase being a L a -phase. The phase diagram shows that it is possible in surfactant mixtures to observe single phase channels for a constant temperature to reach from the water side to the oil side. Such channels have not yet been observed for one component surfactant systems. Under such conditions the temperature has to be changed to observe the single phase channel. While the co-surfactant content in the lower phase channel is increasing with increasing oil content, it is the other way around for the upper phase channel. In this case the co-surfactant content passes through a minimum. An interesting consequence of this behaviour is that the single L 3 -phase, where the channel begins, is transformed over a multiphase region with increasing oil solubilisation into the upper phase microemulsion with 30% of oil. It is noteworthy that preliminary results with decane as oil have shown a similar behaviour. The shape of the channel for isothermal conditions is thus a consequence of the used surfactant mixture and not of the oil. Of general interest on the phase diagram is also the fact that with increasing solubilisation of the oil into the L a -phase, the widest extension of the L a -phase is in the middle of its existence region. At this composition the L a -phase can solubilise more than its own weight of oil, before it Fig. 5 Phase diagram of the system Ca(DS) 2 /IT 3–H 2 O/M 2 with 15% (w/w) surfactant and 85% (w/w) solvent. Phases observed at 40 1C. Abbreviation ‘‘ME’’ stands for ‘‘microemulsion’’ and indicates area of isotropic microemulsion channels. 54 This journal is cthe Owner Societies 2011 Phys. Chem. Chem. Phys., 2011, 13, 3248–3256 3253 breaks down. The high content of oil in a L a -phase is however not a special feature of the used oil or the surfactant mixture. Other microemulsion phase diagrams have been reported with L a -phases that contain more oil than surfactant and where the surfactant is a nonionic surfactant. 33 It is however interesting to note that such L a -phases exist that consist of two monolayers of surfactant that encloses the oil in between the monolayers and the whole package is like a sandwich. The multiphase region between the two channels The multiphase region between the two isotropic channels is very large and dominates the phase diagram. It extends to 45% of oil on the oil axis. Other phase diagrams for microemulsions with nonionic surfactants do not have such a large multiphase region. In the surfactant composition region between xIT3=0.2and 0.55 the two phase regions are the L a /L 1 -region. The L a -phases on the water side are transformed with increasing amount of oil into two phase regions and finally into the single phase channel. The amount of oil that is necessary to destroy the L a -phase increases with the mass fraction of IT 3 in the surfactant mixture. It is interesting to note that the surfactant phase with xIT 3 = 0.55 can accommodate more than its own weight of oil before some of the L a -phase is transformed to a microemulsion phase. Close to the upper channel, the situation is more complicated. Most interesting are the sequence of phases in the surfactant composition between xIT 3 = 0.6 and 0.7. With xIT 3 = 0.6, the L a -phase is transformed with 5% of oil into the two phase L a /ME region and with 20% of oil the system re-enters the single L a -phase. For higher oil ratios the system approaches the lower single phase channel in a two phase L a /ME region. With xIT 3 = 0.65, the phase diagram becomes even more complicated. With 5% oil, the L a -phase is transformed into the upper isotropic channel. In the oil region between 20% and 40%, a two phase region with the L a -phase and an isotropic phase exists, in which the L a -phase first increases with the oil content and then decreases again (see tube pictures of samples in ESIw). It is likely that the isotropic phase in the two phase region first is a ME from the upper phase channel while it is a ME from the lower phase channel in the second region. An even different situation is observed with xIT 3 = 0.7. The pure L a -phase is again transformed with little oil into the bi-continuous channel. With more oil, multiphase regions are observed that have no L a -region. The search for a connection of both single phase channels We were surprised not to find a connection between the lower and the upper single phase channel as would have been expected from known phase diagrams with nonionic surfactants. The reason for this could have been that the composition of the investigated samples was not close enough to detect the connection. To search for a possible connection between both single phase channels, more samples between both channels were investigated. Fig. 6 shows a row of samples that have been prepared by mixing directly a sample from the end of the lower single phase channel with a sample from the upper single phase channel. The first mixed sample with a composition of xIT 3 = 0.675 and xM 2 = 0.46 is already a two phase sample with a lower L a -phase and an upper isotropic phase. While we did not analyse the upper phase, it looked like it consists of pure M 2 . This result is a strong indication of existing tie lines between the L a -phase and pure M 2 . It would mean that an isotropic channel cannot exist in between the lower and the upper single phase channel. In the following samples, the volume of the L a -phase decreases and the volume of the upper phase increases, a sign that the distance of the composition of the sample is increasing from the L a -phase and decreasing from the upper phase. With the sixth sample with the composition xIT 3 = 0.775 and xM 2 =0.5,the situation changes completely. The sample clearly consists now of three phases, a lower L a -phase, a middle bluish microemulsion phase and an upper oil phase. This three-phase area is followed by a two phase area with a decreasing lower L a -phase and an upper bluish microemulsion phase. With xIT 3 = 0.85 and xM 2 = 0.53, the sample reached the upper isotropic single phase channel. More samples with this three phase situation were found and drafted as a triangle in the phase diagram (Fig. 5). This three phase area indicates that there cannot be a connection of both single phase channels. The properties of the samples in the single phase channels It is evident from visual inspection of the samples that the structure of two samples with the same amount of oil but at different surfactant ratios is not the same. An overview of Fig. 6 Mixtures of a sample from the lower and upper single phase channels of the system Ca(DS) 2 /IT 3–H 2 O/M 2 with xM 2 = 0.45 and xM 2 = 0.55. Upper row shows samples without crossed polarizers, lower row shows samples in between crossed polarizers at 40 1C. 55 3254 Phys. Chem. Chem. Phys., 2011, 13, 3248–3256 This journal is cthe Owner Societies 2011 samples in the upper and lower single phase channels with increasing amount of oil is shown in Fig. 7. Samples from the upper phase channel (Fig. 7A) are somewhat bluish and the scattering intensity of the samples increases from the water corner to samples with 65% of oil. Samples with larger oil content scatter less. All samples from the lower phase channel scatter much less (Fig. 7B). Samples from the upper channel show flow birefringence up to 40% oil, while the samples from the lower channel do not show flow birefringence. These features are a first indication that the upper channel has a bi-continuous structure while the lower channel consists of oil droplets in a continuous water phase. The structures of the microemulsions in the channels are now already reported 36 and the physicochemical properties like their rheology and their structural relaxation times will be given in another manuscript. It also will be shown that the obtained phase diagram changes only little if M 2 is replaced by decane. The obtained results are thus typical for microemulsion phase diagrams with mixed nonionic/ionic surfactant mixtures. The temperature stability of the samples in both channels hasn’t been investigated in detail, but it was observed that the samples in the lower single phase channel were less sensitive to temperature changes than the samples in the upper single phase channel, especially close to the L 3 -phase. This can be explained by the higher mass fraction of a temperature sensitive nonionic surfactant in the upper channel. Nevertheless, all samples are stable in a wider temperature range than samples with nonionic surfactant mixtures that are only stable within around 11C. Fig. 7 Overview of selected samples in the upper (A) and lower (B) single phase channel of the system Ca(DS) 2 /IT 3–H 2 O/M 2 at 40 1C. Fig. 8 Conductivity measurements in the upper (A) and lower (B) single phase channel of the system Ca(DS) 2 /IT 3–H 2 O/M 2 at 40 1C. 56 This journal is cthe Owner Societies 2011 Phys. Chem. Chem. Phys., 2011, 13, 3248–3256 3255 Conductivity measurements For conductivity measurements, all samples have been prepared with 10 mM NaCl. A plot of the conductivity in the two channels against the weight fraction of oil is given in Fig. 8. Note that the conductivities in the lower channel (Fig. 8B) are much higher at the water corner than those in the upper channel (Fig. 8A). The reason for this is that the Ca(DS) 2 concentration is much higher in the lower channel. If the conductivities are normalized to the same Ca(DS) 2 concentration they are about the same. The conductivities decrease linearly with the oil content to the middle of the phase diagram, which is where the channel ends. The reason for the decrease is mainly the decreasing mass fraction of Ca(DS) 2 . The conductivities therefore indicate that the microstructure in the lower channel does not change and that the channel consists of a continuous water phase in which oil droplets are dispersed. The situation is different in the upper channel. In this channel the conductivity drops for a few percent of oil (5%) to only about 18% of its value in the L 3 -phase even though the fraction of Ca(DS) 2 is increasing from 23% to 35%. For higher mass fraction of oil the conductivities remain about constant and drop to zero for xM 2 = 0.4. The conductivities thus indicate a dramatic change in the microstructure of the bi-continuous channel with solubilisation of small amounts of oil into the L 3 -phase. Obviously the constraints in the microemulsion for the transport of the ions must be much larger than those in the L 3 -phase. This means that the water channels in the ME-phase must be much smaller than those in the L 3 -phase where the conductivity is only about 2/3 of the value of the normal aqueous phase without a surfactant. The complete break-down of the conductivity for oil fractions larger than 0.4 means that the system changes from a bi-continuous structure to a w/o-structure. Conductivities in the isotropic channels of microemulsions from nonionic surfactants have been reported in the literature. 34 In the upper channel the reported conductivities decrease continuously with increasing oil content. These measurements have helped to establish the view which we have today from the structures in the upper channel. With increasing oil content the bicontinuous L 3 -phase swells with the solubilised oil between the bilayers and is finally transformed at high oil content to a w/o system. With equal amount of oil and water both SAXS-data and conductivities show that this phase is still a bicontinuous phase. 35 Our conductivity measurements unambiguously show that the structures in the upper channel of the investigated system are different from the structures of known systems with nonionic surfactants. We find a rather abrupt transition from the L 3 -structure to another bicontinuous structure with swollen aqueous channels at around 10% of oil and another abrupt transition at 40% of oil to a w/o structure. The microemulsion with equal amount of oil and water does not have a bicontinuous structure. In the meantime, this has been confirmed by cryo-TEM images, which show clearly a w/o-structure. 36 Finally, at the end of the discussion we would like to draw attention to several experimental observations which have been made during this investigation and which to our knowledge are not theoretically understood and need further scouting. To mention here are the existence of the L 3 -phase in the anionic nonionic surfactant mixture, the occurrence of a second L a -region after the existence of the L 3 -phase and a structural transition in the bi-continuous upper phase channel. Conclusion The ternary phase diagram of the silicone oil hexamethyldisiloxane M 2 , water and a surfactant mixture of the ionic surfactant Ca(DS) 2 and the nonionic surfactant isotridecyltriethyleneglycolether IT 3 has been established as a function of the mass fraction of IT 3. Two isotropic single phase channels occur in the system with increasing oil content xM 2 . The lower channel, that is the one with the lower xIT 3 value, begins at the phase boundary of the L 1 -phase toward the L a -phase. This channel increases with increasing xM 2 and ends in the middle of the phase diagram at xIT 3 = 0.65. This channel contains oil droplets in a continuous water phase (o/w-system). The upper channel begins at the L 3 -phase and passes with increasing oil content through a shallow minimum to the oil side. It has, like the L 3 -phase, a bi-continuous structure until 40% oil, and then switches to a w/o structure to the oil side. The microemulsion with equal amount of water and oil does not have a bicontinuous structure. Phases of bicontinuous structure show a strong flow birefringence. Samples from the o/w and the w/o channel are transparent and show no flow birefringence. In contrast to phase diagrams with one component, surfactant mixtures with surfactants can form channels that pass from the water to the oil side at constant temperature. The amphiphilic properties for the channels can be adjusted by changing the composition of the surfactant mixture instead of the temperature. The upper channel and the lower channel are not connected with each other. It is likely that the non-connectivity is due to the influence of the ionic surfactant and the high interfacial tension of the surfactant system against oil, which promotes the formation of globular structures instead of a bicontinuous topology. In total, the investigation has shown that microemulsion phase diagrams, that are established with mixtures of an anionic and a nonionic surfactants, differ from the phase diagram that can be produced with ionic or with nonionic surfactants. We propose that the situation in our system is somewhere between and of general importance. Experimental Materials Sodiumdodecylsulfate (SDS, cryst. research grade) was purchased from the Serva Co., Heidelberg. The nonionic surfactant isotridecyltriethyleneglycolether, in the following text abbreviated as IT 3, was obtained from the Sasol Co., Hamburg (product name Marlipal O13/30). MgCl 2 6H 2 O and CaCl 2 2H 2 O were purchased from the Gru ¨ssing Co., Filsum. The silicone oil hexamethyldisiloxane, abbreviated as M 2 , was obtained as a gift from the Wacker Co., Mu ¨nchen. Preparation of Ca(DS) 2 and Mg(DS) 2 For the preparation of Ca(DS) 2 and Mg(DS) 2 , 400 mM SDSsolutions were mixed with either 200 mM CaCl 2 or MgCl 2 57 is not so clearly seen in the cryo-TEM preparation as the three dimensional nature of the bilayer network. The cryo-TEM micrograph in Fig. 5 actually looks somewhat like a two dimensional projection of a three dimensional polyhedral foam. The L 3 -phase is formed only in the very narrow composition of the samples with xIT 3 ¼0.76 to 0.77. In the neighbouring L a -phase with a composition of xIT 3 ¼0.75 the structures are very different. For comparison with the microstructure of the L 3 -phase, micrographs are shown in Fig. 6 of a sample with a slightly different composition, namely with xIT 3 ¼0.75. The micrographs show the typical pattern of a L a -phase in which the bilayers are perpendicular to the surface of the thin film (Fig. 6A). The interlamellar spacing is of the order of 10–15 nm and the bilayer thickness is about 3 nm. The micrographs of the L 3 -phase and the L a -phases are very consistent with each other. On close inspection of the micrographs of the L a -phase one is able to see defects in some regions (Fig. 6B). These defects are indicated on the micrographs of Fig. 6B with white arrows. One notes situations where two adjacent bilayers seem to be connected by bridges. These could be the structures that have been theoretically predicted. 18 It is noteworthy that the micrograph in Fig. 6A looks like a fingerprint with a much larger scale. It is somewhat surprising that the bilayers in the thin films are all perpendicular to the film surface. From optical microscopy, in contrast, it is known that the bilayers are very often aligned parallel to the film surface, what is known under homeotropic alignment. The reason for the different situation that is shown in Fig. 6 may be a result of the film thickness. In the holes of the polymer films the thickness of the surfactant phase varies. On parallel alignment the bilayers would have to form different numbers of bilayers at different positions in the thin film in order to keep the interlamellar distance the same. Such defects cost energy and the film might therefore prefer the perpendicular alignment. The structures in L a -phases with a lower IT 3 content (xIT 3 ¼0.65) are somewhat different (Fig. 7). Large multilamellar vesicles (MLVs) are now seen in which the spacing between the bilayers is much larger than in the micrograph with x¼0.75. The larger spacings are probably the result of applying high shear rates to the MLV during the blotting procedure. 16 Cryo-TEM micrographs of samples with increasing amount of oil are shown in Fig. 8. The first two micrographs (Fig. 8A and B) show similar network-like structures to the micrograph of the L 3 -phase without oil (Fig. 5). It is generally assumed that the L 3 -phase is a symmetric phase, where the inside and outside volume of the three-dimensional tubular structure is the same. It is likely that this symmetry is lost by the solubilisation of oil, and the microemulsion is an asymmetric phase. This can be concluded from the abrupt conductivity change that occurs Fig. 6 Cryo-TEM of samples with 15% (w/w) surfactant Ca(DS) 2 /IT 3, x IT 3 ¼0.75, prepared at 40 C. (A) Typical pattern of L a -phase; (B) L a -phase with defects, shown by white arrows. Fig. 7 Cryo-TEM image of sample with 15% (w/w) surfactant Ca(DS) 2 / IT 3, xIT 3 ¼0.65, prepared at 40 C. Large squeezed-together multilamellar vesicles are seen. 5370 | Soft Matter, 2010, 6, 5367–5374 This journal is ªThe Royal Society of Chemistry 2010 64 around xM 2 ¼0.05. Indeed, the observed bi-continuous structures by the TEM micrographs of samples with solubilised oil look somewhat different from the micrographs of the L 3 -phase without oil. However, some micrographs also show typical structures of L a -phases or MLV phases. It is likely that these structures result from evaporation of M 2 in the thin film. M 2 is an extremely volatile compound. In order to decrease the evaporation as much as possible, the atmosphere in the preparation chamber had to be saturated with the microemulsion solution and the preparation had to be carried out as quickly as possible. The samples with the highest oil content (Fig. 8C and D) have different structures. In both samples globular particles are observed that are in a lighter grey than the background. Obviously, the structure has now changed from the bi-continuous structure to w/o droplets. At 50% oil the droplets are densely packed while in the sample with 75% oil the droplet are more dispersed. The droplets have about the same size in the two samples. The diameter of the droplets varies from 20 to 30 nm. Similar globular particles already have been observed in the system AOT–n-decane–D 2 O by FF-TEM. 19 The micrographs with 50% and 75% oil show some interesting details that are worth emphasizing. In Fig. 8C some of the small globular particles are very dark in comparison to the light grey for most of the other droplets. These particles make up a few percent of the total number of particles. All those particles are crystalline ice, formed during the rapid cooling process. The Fig. 8 Cryo-TEM micrographs of samples with 15% (w/w) surfactant Ca(DS) 2 /IT 3, prepared at 40 C: (A) composition xIT 3 ¼0.65, xM 2 ¼0.05, sponge-like structure; (B) composition xIT 3 ¼0.65, xM 2 ¼0.15, sponge-like structure; (C) composition xIT 3 ¼0.9, xM 2 ¼0.5, densely packed water droplets and (D) composition xIT 3 ¼0.975, xM 2 ¼0.75, less densely packed water droplets. This journal is ªThe Royal Society of Chemistry 2010 Soft Matter, 2010, 6, 5367–5374 | 5371 65 cooling rates achievable by liquid nitrogen, used here to avoid oil dissolution, are insufficient to vitrify water (but are sufficient to vitrify the oil!). Thus, the small water domains freeze into crystalline hexagonal ice. These nano-crystals are randomly oriented with respect to the electron beam, so that only a few satisfy Bragg’s law for electron diffraction, and those appear dark. The ice crystals that do not diffract appear light grey. The micrographs show that the ice particles could only grow to the size of the droplets, because the droplets were not connected to surrounding droplets. Micrographs, Fig. 8A and B, with an oil fraction of 0.05 and 0.15 seem to represent situations that are in between the L 3 -phase and the microemulsion situation with 50% of oil. For both samples the crystalline ice structures are somewhat larger than the water domains. These results are in agreement with conductivity measurements in the channel, which show a transition from the conductivity of the L 3 -phase to a lower conductivity at an oil content of 5%, and finally to loss of conductivity at 40% oil. The development of the structures can therefore at best to be represented by a HIPE like (High Internal Phase Emulsion) structure with varying connectivity between the water domains. In some of the micrographs one can actually observe that some of the droplets are deformed to polygonal structures. Besides the apparent polyhexagon the micrograph also shows that some of the droplets have coalesced with the neighbouring structures, and have formed larger structures. All these details demonstrate that the original droplets are very dynamic species. When the equilibrium conditions are changed, they can quickly adjust to the new conditions. In contrast to the micrograph of the sample with 50% oil, the micrograph with 75% oil shows practically no irregularities. Only very few droplets seem to have formed doublets, while the large majority of the small droplets is randomly distributed in the oil matrix. All in all, clean micrographs of a w/o-microemulsion by cryo-TEM are astonishing and rare, as it had been often assumed, that oil-rich samples cannot be imaged directly, 20 as the oil gets dissolved in the regularly used cryogen liquid ethane, or the alternative cryogen liquid nitrogen would not provide the sufficient coolingrate to vitrify the specimen. 21 Cryo-TEM micrographs in the lower channel The lower single-phase channel has a large slope with increasing oil concentration. In addition, the channel is very narrow. As a result of these two conditions it is difficult to obtain good micrographs from the structures in the channel. As it is obvious from the channel, the structures in the channel could change, and the original structures are no longer in equilibrium with their surroundings, when the samples of the channel lose a few percent of oil by evaporation. This is indeed the case, as is demonstrated in Fig. 9, where two micrographs are shown that were obtained from a sample with xIT 3 ¼0.4 and 15% oil. One micrograph (Fig. 9A) shows only small unilamellar vesicles with diameters ranging from 13 to 40 nm. The other micrograph (Fig. 9B) also shows uniand multilamellar vesicles with interlamellar spacings that are consistent with the surfactant concentration of 15%. In addition, both micrographs show long lines that separate or surround large domains of different shades. Those represent an encapsulation of small vesicles by larger ones. Model for the calculation of the microemulsion droplet size Dimensions for the water or oil droplets in the microemulsions can easily be calculated with the core–shell model and with the help of the oil–surfactant or water–surfactant ratio. With the volume of the core vC¼4 3r3p and the volume of the shell which is equal to the volume of the surfactant, v S ¼4pr 2 d one obtains the simple equation Fig. 9 Cryo-TEM micrographs of sample with 15% (w/w) surfactant Ca(DS) 2 /IT 3, xIT 3 ¼0.4, xM 2 ¼0.15, prepared at 40 C. (A) Small unilamellar vesicles with diameter from 13 to 40 nm. (B) Small unilamellar vesicles and multilamellar vesicles. 5372 | Soft Matter, 2010, 6, 5367–5374 This journal is ªThe Royal Society of Chemistry 2010 66 vC vS ¼4pr3 34pr2d¼R and r¼R3d. In this equation dis the thickness of the surfactant monolayer, namely the thickness of the shell. The thickness dcan experimentally be appreciated from the cryo-TEM micrographs of the L a -phase or can be estimated from the number of CH 2 -groups in the surfactant molecules. In an exact calculation, the different lengths of the two surfactant molecules and their different mole ratios in the upper and lower channel have to be considered. The assumed dvalue, the volume ratio of the core and shell v C /v S , the calculated diameters D cal ¼2r+2dof the droplets and the experimentally determined diameters D exp from the cryoTEM images are given in Table 1. For the calculation of v C /v S , the density of the surfactant mixture was estimated to 1 g cm 3 , for the volume of the oil, we used the density of 0.76 g cm 3 for M 2 . The agreement between the calculated and experimental determined diameters in the lower channel are not very good. The micrograph of the microemulsion with xM 2 ¼0.15 from the lower channel shows a large variety of small vesicles with diameters from 13–40 nm, which does not agree with the model. In contrast to the lower channel, the measured size of the droplets from the upper channel differs only little from the calculated values. Although the droplets show a certain variety in size distribution, the experimental determined diameters are in good agreement with the core–shell model, considering the difficult conditions for sample preparation as the high volatility of the silicone oil and the high temperature. Conclusion The nanostructures in the two isotropic channels of a microemulsion of the silicone oil hexamethyldisiloxane (M 2 ) and a surfactant mixture of the anionic surfactant calcium dodecyl sulfate Ca(DS 2 ) and the nonionic surfactant iso-tridecyl-triethyleneglycolether (IT 3) have been determined by cryo-TEM. Contrary to microemulsions with a pure nonionic surfactant, the channels can be formed in this system at constant temperature by adjusting the mass fraction of the two surfactants. The channel with the higher mass fraction of nonionic surfactant begins with the L 3 -phase on the water side, and runs with a shallow slope to the oil side, when the mass fraction is plotted against the oil content. Conductivity results and cryo-TEM micrographs indicate that bicontinuous structures exist in this channel from the water side to a microemulsion with a 65/35 water oil ratio. The micrographs show that the structure in the bi-continuous region changes from an L 3 -type structure to a HIPE-type structure. Water droplets in the oil matrix exist from the middle of the phase diagram until the oil side. This differs from the nanostructures of microemulsions with nonionic surfactants, where elements of bicontinuous microstructures, even at high oil contents, are reported. 22 The sizes of the w/o-droplets in the upper channel are consistent with simple geometrical considerations. They increase, as expected, with an increasing mass fraction of water. The lower channel begins on the water side with an L 1 -phase and reaches with increasing oil concentration to the middle of the phase diagram. The channel contains o/w droplets in a continuous water matrix. No connection was found between the two channels. Because of the high volatility of the oil, the droplets in this channel were difficult to image with the cryo-TEM method. Small vesicles were obtained in many preparations, instead of the droplets. The vesicles were obviously formed in the thin cryo-film in the time between forming and fixation by temperature quenching. Therefore, we are trying to improve the design of our CEVS to allow us work with more highly volatile solvents such as M 2 , with even smaller concentration changes. We are also looking for a microemulsion system based on a lower-volatility oil, that gives a similar phase diagram. Experimental Materials The nonionic surfactant iso-tridecyl-triethyleneglycolether, abbreviated as IT 3, was obtained from Sasol, Co., Hamburg (Marlipal O13/30). Sodium dodecyl sulfate (SDS, cryst. research grade) was purchased from the Serva Co., Heidelberg. MgCl 2  6H 2 O and CaCl 2 2H 2 O were purchased from the Gr€ ussing Co., Filsum. The silicone oil hexamethyldisiloxane, abbreviated M 2 , was purchased from the Wacker Co., M€ unchen, n-decane was obtained from the Merck Co., Darmstadt. Preparation of Ca(DS) 2 and Mg(DS) 2 For the preparation of Ca(DS) 2 and Mg(DS) 2 , 400 mM SDS-solution were mixed with either 200 mM CaCl 2 or MgCl 2 solution under stirring. The bivalent counterions Ca 2+ and Mg 2+ bind stronger to the dodecyl sulfate than the sodium-ion, leading to a precipitation of Ca(DS) 2 in solution below its Kraffttemperature of 50 C, and Mg(DS) 2 below its Krafft-temperature of 25 C. The solutions were heated up to 60 C for the solution with CaCl 2 , or warmed up above 25 C for the solution with MgCl 2 to obtain a clear solution, and then cooled down to 20 C. After precipitation, Ca(DS) 2 and Mg(DS) 2 were washed several times with de-ionised water to remove excess salt. The purity of the filtered surfactant thus could be checked by measuring the conductivity of its flow through. The washed Ca(DS) 2 and Mg(DS) 2 were dried for several days in a cabinet dryer at 50 C, and later used without further purification. Table 1 Overview of the calculated diameters D cal and experimentally determined diameters D exp of the microemulsion droplets from the lower and upper single-phase channel Size of droplets in lower channel xM 2 d/nm v C /v S r/nm D cal /nm D exp /nm 0.15 1.5 1.12 5 13 13–40 Size of droplets in upper channel xM 2 xH 2 Od/nm v C /v S r/nm D cal /nm D exp /nm 0.5 0.5 1.5 2.83 12.8 29 36 6 0.75 0.25 1.5 1.42 6.4 16 25 6 This journal is ªThe Royal Society of Chemistry 2010 Soft Matter, 2010, 6, 5367–5374 | 5373 67 Preparation of samples All samples were prepared by weighing in directly the components in test tubes, first surfactant and co-surfactant, H 2 O, and, as last component, M 2 , due to its high volatility. The test tubes were sealed with Teflon tape, tempered at 40 C in a water bath, and vortexed several times thoroughly. All samples were incubated at least 3 days at 40 C before being investigated for their phase behaviour. For the phase diagram, the samples below xIT 3 0.3 had to be prepared with Mg(DS) 2 instead of Ca(DS) 2 to avoid problems with precipitation of Ca(DS) 2 . In general, a phase diagram was scanned with a resolution of 5% in the composition of the mass fraction of IT 3 and M 2 . Finer steps were investigated in the beginning of the upper single-phase channel, and in between the two single-phase channels to find a possible connection of both channels. The multiphase samples were viewed and imaged without and in between crossed polarisers, to visualise the birefringence of lamellar regions. Cryo-Transmission Electron Microscopy (Cryo-TEM) The specimens for cryo-TEM were prepared in a controlled environment vitrification system (CEVS) and plunged into liquid ethane at its freezing point. 23 For oil continuous samples, the specimens were plunged into liquid nitrogen in order to overcome problems with solvent dissolution. 24 The CEVS was kept at 40 C and the atmosphere either saturated with H 2 O for the samples without oil, or directly with the investigated microemulsion solution for samples containing oil. Due to high volatility of the silicone oil, the specimens were prepared as quickly as possible. Specimens, kept below 178 C, were examined in an FEI TI2 G 2 transmission electron microscope, operated at 120 kV, using a Gatan 626 cryoholder system. Images were recorded digitally in the minimal electron dose mode by a Gatan US1000 high-resolution CCD camera, with the Digital Micrograph software package. Acknowledgements We want to thank the Russell Berrie Nanotechnology Institute, Technion (RBNI), for providing us the opportunity to investigate our specimen by electron microscopy. For the excellent technical assistance, we especially thank Judith Schmidt and Dr Ellina Kesselman, and all members of the Lab of the Department of Chemical Engineering, Technion-Israel Institute of Technology, Haifa 32000, Israel. Notes and references 1 R. Strey, Curr. Opin. Colloid Interface Sci., 1996, 1, 402–410. 2 S. J. Chen, D. Fennell Evans and B. W. Ninham, J. Phys. Chem., 1984, 88, 1631. 3 S. H. Chen and S. L. Chang, J. Phys. Chem., 1991, 95, 7427–7432. 4 C. Mathew, Z. Saidi, J. Peyrelasse and C. Boned, Phys. Rev. A, 1991, 43(2), 872–882. 5 J. Bergenholtz, A. A. Romagnoli and N. J. Wagner, Langmuir, 1995, 11(5), 1559–1570. 6 T. Sottmann and C. Stubenrauch, Phase Behavior, Interfacial Tension, and Microstructure of Microemulsions, in Microemulsions: Background, New Concepts, Applications, Perspectives, ed. C. Stubenrauch, John Wiley & Sons, Oxford, 2009, ch. 1. 7 T. Sottmann and R. Strey, Microemulsions, in Soft Colloids V— Fundamentals in Interface and Colloid Science, ed. J. Lyklema, Elsevier, Amsterdam, 2005, ch. 5. 8 F. Lichterfeld, T. Schmeling and R. Strey, J. Phys. Chem., 1986, 90, 5762–5766. 9 T. Sottmann and R. Strey, J. Chem. Phys., 1997, 106, 8606–8615. 10 C. Stubenrauch and T. Sottmann, Microemulsions stabilized by Sugar Surfactants, in Sugar-Based Surfactants: Fundamentals and Applications, ed. C. C. Ruiz, Taylor & Francis, CRC Press, Boca Raton, 2009, ch. 12. 11 IT 3 is commercially available under the name ‘‘Marlipal O13/30’’ from Sasol Germany GmbH, Auckelmannsplatz 1, 20537 Hamburg, Germany. 12 L. Wolf, H. Hoffmann, K. Watanabe and T. Okamoto, PCCP, submitted. 13 M. Kahlweit, et al., J. Colloid Interface Sci., 1987, 118, 436–453. 14 R. Strey, Colloid Polym. Sci., 1994, 272, 1005–1019. 15 M. Kahlweit, B. Faulhaber and G. Busse, Langmuir, 1994, 10, 2528– 2532. 16 D. Danino, D. Weihs, R. Zana, G. Or€ add, G. Lindblom, M. Abe and Y. Talmon, J. Colloid Interface Sci., 2003, 259, 382–390. 17 H. Hoffmann, C. Thunig, U. Munkert, H. W. Meyer and W. Richter, Langmuir, 1992, 8, 2629–2638. 18 R. Strey, W. Jahn, G. Porte and P. Bassereau, Langmuir, 1990, 6, 1635–1639. 19 W. Jahn and R. Strey, J. Phys. Chem., 1988, 92, 2294–2301. 20 L. Belkoura, C. Stubenrauch and R. Strey, Langmuir, 2004, 20,4391– 4399. 21 V. Agarwal, M. Singh, G. McPherson, V. John and A. Bose, Langmuir, 2004, 20, 11–15. 22 S. Burauer, L. Belkoura, C. Stubenrauch and R. Strey, Colloids Surf., A, 2003, 228, 159–170. 23 D. Danino, A. Bernheim-Groswasser and Y. Talmon, Colloids Surf., A, 2001, 183, 113–122. 24 Y. Talmon, D. Danino, J. Satayavolu and R. Gupta, J. Colloid Interface Sci., 2002, 249, 180–186. 5374 | Soft Matter, 2010, 6, 5367–5374 This journal is ªThe Royal Society of Chemistry 2010 68 Dynamic Properties of Microemulsions in the Single-Phase channels 69 5.1.3. Dynamic Properties of Microemulsions in the Single-Phase channels Lukas Wolf*, Heinz Hoffmann, Walter Richter, Takashi Teshigawara and Tohru Okamoto. *corresponding author Published in Journal of Physical Chemistry B, 2011, 115(38), 11081-11091. DOI: 10.1021/jp2036789. Published: August 15, 2011 r2011 American Chemical Society 11081 dx.doi.org/10.1021/jp2036789 |J. Phys. Chem. B 2011, 115, 11081–11091 ARTICLE pubs.acs.org/JPCB Dynamic Properties of Microemulsions in the Single-Phase Channels Lukas Wolf,* ,† Heinz Hoffmann, † Walter Richter, ‡ Takashi Teshigawara, § and Tohru Okamoto § † University of Bayreuth, BZKG and BayColl, Gottlieb-Keim-Str. 60, 95448 Bayreuth, Germany ‡ Friedrich-Schiller-University, Centre for Electron Microscopy, Ziegelm€uhlenweg 1, 07743 Jena, Germany § Shiseido Research Center, 2-2-1 Hayabuchi, Tsuzuki-ku, Yokohama, Japan 224-8558 b SSupporting Information ’INTRODUCTION Microemulsions are thermodynamically stable phases from oil, water, and surfactants. 1 The phases contain well-defined structures, such as oil droplets in a continuous water phase, water droplets in a continuous oil phase, and bicontinuous structures. The detailed structures depend on the composition of the system in the ternary phase diagram. Usually, the structures change with a change in the composition of the samples. For many years, microemulsions with ionic surfactants were in the focus of interest. 2 More recently, microemulsions with nonionic surfactants have become the center of interest. 3 As a consequence, we now have a good understanding of microemulsions. It is known, for example, how the phases can be optimized for as little surfactant as possible and how the surfactant can be optimized for a given oil. Of fundamental importance for the understanding of the systems is the value of the interfacial tension of a micellar solution against an oil phase. 4 For a high solubilization of an oil in a micellar solution, the interfacial tension has to be minimized. 5 One of the most fascinating features of nonionic microemulsions is isotropic single-phase channels that pass from the aqueous micellar phase continuously to the oil phase for constant surfactant concentration. 6 Because of SANS and SAXS, we have a good understanding of the structures in these channels. Many systems have been investigated that have two channels from the water to the oil side, one at lower temperature and one at a higher temperature, and a single channel in the middle of the phase diagram that is connected with the two channels at both sides. 7 Today, there is a good theoretical understanding of the structures in the isotropic channels and about the thermodynamics of microemulsions with nonionic surfactant. However, very few investigations have been carried out on the dynamic behavior of the microemulsions, and no systematic investigation has so far been made in a channel from the water side to the oil side. In this investigation, we therefore will study the dynamic properties of microemulsions in the single-phase channels by rheology and the electric birefringence method. The measurements were carried out on a system with an anionic/nonionic surfactant mixture that was similar to one that was previously studied and from which we knew the position of the single phase channels. 8 Received: April 20, 2011 Revised: August 15, 2011 ABSTRACT: We have studied the dynamic and rheological properties in the single-phase channels of a microemulsion system with a mixed anionic/nonionic surfactant system and decane from the aqueous to the oil phase. One isotropic channel, called the “upper”channel, begins at the L 3 phase (sponge-like phase) of the binary surfactant mixture on the water side and passes with a shallow minimum for the surfactant composition to the oil side. The other “lower”single-phase channel begins at the micellar L 1 phase and ends in the middle of the phase diagram. Both isotropic channels are separated by a huge anisotropic single phase L α channel that reaches from the water side to 90% of oil in the solvent mixture. The structural relaxation time of the viscous fluids could be measured with electric birefringence (EB) measurements, where a signal is caused by the deformation of the internal nanostructure of the fluids by an electric field. For the L 3 phase, the EB signal can be fitted with a single time constant. With increasing oil in the upper channel, the main structural relaxation time passes over a maximum and correlates with the viscosity. Obviously, this time constant controls the viscosity of the fluid (η o =G03τ). It is remarkable that the longest structural relaxation time increases three decades, and the viscosity increases two decades when 10% of oil is solubilized into the L 3 phase. Conductivity data imply that the fluid in the upper channel has a bicontinuous structure from the L 3 phase to the microemulsion with only 10% oil. In this oil range, the conductivity decreases three decades, and the electric birefringence signals are complicated because of a superposition of up to three processes. For higher oil ratios, the structure obviously changes to a HIPE (high internal phase emulsion) structure with water droplets in the oil matrix. 70 11082 dx.doi.org/10.1021/jp2036789 |J. Phys. Chem. B 2011, 115, 11081–11091 The Journal of Physical Chemistry B ARTICLE ’EXPERIMENTAL SECTION Materials. The nonionic surfactant iso-tridecyl-triethylenglycolether, abbreviated as IT 3, was obtained from the Sasol Company (Hamburg, Germany) (“Marlipal O13/30”). This compound has a polydisperse distribution of EO groups with an average of three EO units. Sodium dodecyl sulfate (SDS, cryst. research grade) was purchased from the Serva Company (Heidelberg, Germany). MgCl 2 36H 2 O was purchased from the Gr€ussing Company (Filsum, Germany). N-Decane (analytical grade) was obtained from the Merck Company (Darmstadt, Germany). Preparation of Mg(DS) 2 .For the preparation of Mg(DS) 2 , 400 mM SDS solution was mixed with 200 mM MgCl 2 solution under stirring. The bivalent counterion Mg 2+ binds stronger to the dodecyl sulfate than the sodium ion, leading to a precipitation of Mg(DS) 2 in solution below its Krafft temperature around 25 °C. The solution was heated above 25 °C to obtain a clear solution and then cooled to 20 °C. After precipitation overnight, Mg(DS) 2 was filtered and washed several times with deionized water to remove excess salt. The purity of the surfactant thus could be checked by measuring the conductivity of the flow through of the filtered Mg(DS) 2 . The washed Mg(DS) 2 was freeze-dried with the freeze-drying device Alpha 1-4, Christ Company (Osterode, Germany), and used without further purification. Preparation of Samples. All samples were prepared by weighing the components directly in test tubes on an analytical balance. The test tubes were sealed with Teflon tape, tempered at 25 °C in a water bath, and vortexed several times thoroughly. All samples were incubated at least 3 days at 25 °C before being investigated for their phase behavior. In general, a phase diagram was scanned with a resolution of 5% in the composition of the mass fraction of IT 3 and decane. Finer steps were investigated in the beginning of the narrow upper single-phase channel. The multiphase samples were viewed and imaged without and in between crossed polarizers tovisualize the birefringence of lamellar regions. Freeze-Fracture Transmission Electron Microscopy (FF-TEM). The microemulsions, prepared in presence of 20% (w/w) glycerin in the aqueous phase were stored at room temperature (25 °C) and quick-frozen from RT using the sandwich technique. A small amount of the samples was sandwiched between two copper profiles (BAL-TEC/Balzers, Liechtenstein) as used for the double-replica technique and frozen by plunging these sandwiches immediately into a liquefied ethanepropane mixture (v/v 1/1) cooled in liquid nitrogen. Fracturing and replication were performed at 150 °C in a BAF 400T freeze-fracture device (BAL-TEC/Balzers, Liechtenstein) equipped with electron guns and a film sheet thickness monitor. For replication, at first, Pt(C) was evaporated under an angle of 35°(thickness: 2 nm), followed by C under 90°angle (thickness 20 nm). The replicas were placed on electron microscopic copper grids (Mesh 400), cleaned by chloroformmethanol mixture (v/v 2/1), and examined in an EM 900 electron microscope (Zeiss, Oberkochen, Germany). Conductivity and Rheology Measurements. For conductivity measurements, we used the microprocessor conductivity meter LF3000 from the WTW Company (Weilheim, Germany). The rheology was measured with the coneplate rheometer RheoStress 600 from the Haake Thermo Scientific Company (Karlsruhe, Germany). All samples were investigated at 25 °C. Electric Birefringence Measurements. The electric birefringence device is a self-made device with a 632.8 nm HeNe laser, crossed polarizers, aperture plates, a Kerr cell with temperature control, and a photomultiplier. For the generation of an electric pulse, there was used the Cober “high power pulse generator” model 606. The signal of the photomultiplier was recorded with the Voltcraft DSO-2090 USB-oszillograph. Data were processed and evaluated with the computer software “Origin”. SANS Measurements. For SANS measurements, samples were prepared with D 2 O instead of H 2 O. By replacing H 2 Oby D 2 O, the phases in the single-phase channels are slightly shifted. To obtain transparent isotropic phases, we had to adjust the surfactant composition by increasing the mass fraction of IT 3 by ∼2%. SANS data were obtained with the SANS device D11 at the Institue Laure-Langevin in Grenoble, France. Samples were filled in 1 mm quartz cuvette (Helma), and temperature was set to 25 °C. Evaluation of data (background detection, subtraction of signal intensity, scaling of data) was performed with standard computer software. ’RESULTS AND DISCUSSION Phase Diagram Mg(DS) 2 /IT 3H 2 O/n-Decane. On microemulsion systems with a single nonionic surfactant, it is not possible to pass from a single aqueous phase to a single oil phase at constant temperature. The interfacial tension of nonionic surfactants changes with temperature, and thus the single-phase regions of microemulsions are temperature-dependent. The amphiphilic properties of surfactants can also be changed by adding a cosurfactant to the surfactant solution. In the present investigation, we therefore chose a binary mixture of a hydrophilic surfactant with a lipophilic surfactant. For the ionic surfactant, we chose the Mg 2 salt of SDS. Mg(DS) 2 is more lipophilic than SDS, reduces the surface tension more effectively, and is known to form liquid crystalline L α phases when mixed with suited cosurfactants. 8 As lipophilic cosurfactant, we used iso-tridecyl-triethylenglycolether (abbreviation IT 3, = C 13 E 3 ). By changing the mass fraction of the two molecules in the surfactant solution, it turned out to be possible to pass from a micellar L 1 phase with only Mg(DS) 2 over lamellar L α and an L 3 phases (sponge-like phase) with the surfactant mixtures to an L 1 / L 2 (L 2 = inverse micellar phase) two phase situation for the solution with pure IT 3. The phase sequence of the binary surfactant mixture is shown in Figure SI1 of the Supporting Information, and detailed rheological investigations can be found in Figure SI2 of the Supporting Information. Some nonionic surfactants show the same sequence of phases with increasing temperature. The plot of the surfactant mixture against the mass fraction x of the oil decane in the solvent mixture between 0 and 1 is shown in Figure 1. The total surfactant concentration was kept constant at 15% (w/w), the temperature was kept at 25 °C, and samples were prepared with 20% glycerine in H 2 O for possible FF-TEM investigations to prevent freezing artifacts. The phase diagram contains two isotropic channels, a lower one and an upper one. The upper one begins on the surfactant axis at the region of the L 3 phase. With increasing oil, the channel first shifts to a lower IT 3 ratio and then again to a higher IT 3/ Mg(DS) 2 ratio for higher oil ratios. It ends on the oil side at 80% decane and pure IT 3 as surfactant. The lower channel begins at 71 11083 dx.doi.org/10.1021/jp2036789 |J. Phys. Chem. B 2011, 115, 11081–11091 The Journal of Physical Chemistry B ARTICLE the L 1 region and ends in the middle of the phase diagram at an IT 3 ratio of 0.57. Both channels are separated by a huge single-phase birefringent L α region that extends from 0 to 90% decane with slightly increasing mass fraction of IT 3. All samples in the L α channel behave like gels. However, the storage modulus G0is decreasing constantly with increasing oil content, indicating that the gels become softer. Pictures of the L α phases between crossed polarizers at different temperatures and detailed rheology data can be found online in Figure SI3 of the Supporting Information. A significant feature of the L α channel is its high-temperature stability. Samples between xdecane = 0 to 0.9 are stable at least between 10 and 40 °C, and samples with xdecane 0 to 0.4 are even stable at least up to 60 °C. Although the L α channel is quite narrow, its high-temperature stability is amazing. This feature is obvious due to the surfactant mixture, where the hydrophilic lipophilic balance is determined more by the mass fraction of the cosurfactant than by the temperature. L α phases with oil and a single nonionic surfactant of the type C i E j are not so stable. 32 The microemulsions in the lower single phase channel are transparent phases that show no flow birefringence under shear. The samples in the upper phase channel have somewhat different properties. Whereas the L 3 phase without decane is completely transparent, the samples with decane look somewhat bluish and their scattering intensity is most intensive around xdecane 0.03 and 0.1. For higher oil content, the scattering intensity is decreasing again. Pictures of samples from the lower and upper single phase channel are shown in Figure SI4 of the Supporting Information. The different macroscopic properties in the upper and lower phase channels for the same amount of oil are already an indication that the structures in the two single-phase channels are different. In many previous publications of microemulsions, it was shown that the micellar structures in a single-phase channel varied from o/w droplets to bicontinuous structures to w/o structures. 9 The following chapters will show, that the situation in the present system must be different and more complicated. It should, however, be clear that the situation in nonionic and in somewhat ionically charged systems may be somewhat different. Our investigated system is the first system with an ionic surfactant for which the ionic charge was not shielded by excess salt that forms an isotropic channel from the water side to the oil side. Conductivity in the Single-Phase Channels. The plot of the conductivity in the upper and lower single-phase channels against the mass fraction of decane in the solvent mixture is shown in Figure 2. In the upper channel, the conductivity first increases slightly from ∼1000 μS/cm of the sample without decane to 1160 μS/ cm to the sample with 1% decane. The reason for this lies in the change of the composition of the surfactant mixture. In the range from 1 to 10% decane, the conductivity decreases abruptly three orders of magnitude to 1 μS/cm even though the fraction of the anionic Mg(DS) 2 is increasing. For higher mass fractions of decane, the conductivity values decrease continuously to low values as, for example, 0.03 μS/cm for the sample with a water/ oil ratio of 1/1 (w/w). The conductivities thus indicate a dramatic change in the nanostructure of the upper channel with solubilization of small amounts of oil into the L 3 phase. The abrupt collapse of the conductivity indicates that the system Figure 2. Plot of conductivity (red dots) and IT 3 content (gray triangles) against mass fraction of decane in solvent mixture. (a) Conductivity data for the upper single phase channel. (b) Conductivity data for the lower single phase channel. Figure 1. Phase diagram of system Mg(DS) 2 /IT 3H 2 O/decane at 15% (w/w) surfactant and 25 °C, 20% glycerin in H 2 OIT 3 = mass fraction of IT 3 in the surfactant mixture, xdecane = mass fraction of decane in the solvent mixture. “ME”indicates isotropic microemulsion area, and L α indicates area of anisotropic lamellar channel. 72 11084 dx.doi.org/10.1021/jp2036789 |J. Phys. Chem. B 2011, 115, 11081–11091 The Journal of Physical Chemistry B ARTICLE changes from a bicontinuous structure to a w/o structure (waterin-oil). Conductivities in the isotropic channels of microemulsions from nonionic surfactants have been reported in the literature. 10 In such systems, the conductivity in the upper channel decreases continuously with increasing oil content. These measurements have helped to establish the view that we have today from the structures in the upper channel. With increasing oil content, the bicontinuous L 3 phase swells with the solubilized oil between the bilayers and is finally transformed at high oil content to a w/o system. With equal amount of oil and water, SAXS data and conductivities show that this phase is still a bicontinuous phase. 11 Our conductivity data unambiguously show that the structures in the upper channel of the investigated system are different from the structures of known systems with nonionic surfactants. We find a rather abrupt transition from the bicontinuous L 3 structure to a w/o structure with only 10% of oil in the solvent mixture. The conductivity data of the lower channel indicate that the nanostructure in the lower channel does not change much with increasing oil in contrast with the nanostructure in the upper channel. At the water corner, the conductivity in the lower channel with 2900 μS/cm is much higher than the conductivity of the L 3 phase of the upper channel with 1000 μS/cm. The reason for this is that the Mg(DS) 2 concentration is much higher in the lower channel. With increasing oil content, the conductivities decrease slightly to 1500 μS/cm at the middle of the phase diagram, which is where the channel ends. The reason for the decrease is mainly the decreasing mass fraction of Mg(DS) 2 . Obviously, the lower channel consists of a continuous water phase in which oil droplets are dispersed (o/w structure). As already mentioned, the lower and upper single-phase channels are not connected to each other in contrast with the classical microemulsion systems with single nonionic surfactant and hydrocarbons. In the present system, both channels are separated by a large single anisotropic L α channel. A plot of the conductivity in the L α channel can be found in Figure SI4 of the Supporting Information. The conductivity on the water-side is ∼400 μS/cm and decreases with increasing oil content and increasing xIT 3 constantly to a value of 40 μS/cm at the L α phase with 90% decane. It is noteworthy that the sample without decane at the water side has a lower conductivity than the L 3 phase despite the fact that it contains ∼60% more anionic Mg(DS) 2 and also a much lower conductivity than the microemulsion from the lower single-phase channel with the same xIT 3 value of 0.5. This already indicates that the structure consists of densely packed multilamellar vesicles, where the conductivity is low due to the limited movement of the ions. 12 FF-TEM Micrographs in the Upper Channel. An FF-TEM micrograph of the L 3 phase without oil is shown in Figure 3a. In this technique, the replica of a fractured plane across the sample is Figure 3. FF-TEM micrographs with 15% (w/w) surfactant Mg(DS) 2 /IT 3, 20% glycerine in H 2 O, prepared at 25 °C; scale bar = 200 nm. (a) L 3 phase without oil at xIT 3 = 0.79, (b) microemulsion of upper channel with xIT 3 0.67, xdecane 0.2, (c) microemulsion of lower channel with xIT 3 = 0.35, x decane = 0.1, o/w-structure, droplet size ∼7 nm, and (d) microemulsion of lower channel with xIT 3 = 0.45, xdecane = 0.2, droplet size ∼14 nm. 73 11091 dx.doi.org/10.1021/jp2036789 |J. Phys. Chem. B 2011, 115, 11081–11091 The Journal of Physical Chemistry B ARTICLE (16) Schorr, W.; Hoffmann, H. Physics of Amphiphiles: Micelles, Vesicles and Microemulsions: Proceedings of the International School of Physics; Degiorgio, V., Corti, M., Eds.; North-Holland: Amsterdam, 1985; pp 160180. (17) Kr€amer, U.; Hoffmann, H. Macromolecules 1991,24, 256–263. (18) Hoffmann, H.; Kr€amer, U.; Thurn, H. J. Phys. Chem. 1990, 94, 2027–2033. (19) Stellwagen, N. C. Biopolymers 1991,31, 1651–1667. (20) Miller, C. A.; Gradzielski, M.; Hoffmann, H.; Kr€amer, U.; Thunig, C. Prog. Colloid Polym. Sci. 1991,84, 243. (21) Cates, M. E.; Roux, D. J. Phys.: Condens. Matter 1990, 2, 339–346. (22) Strey, R.; Jahn, W.; Skouri, M.; Porte, G.; Marignan, J.; Olsson, U. In Structure and Dynamics of Strongly Interacting Colloids and Supramolecular Aggregates in Solution; Chen, S. H., Huang, J. S., Tartaglia, P., Eds.; Kluwer Academic: Boston, 1992; pp 351363. (23) Porte, G.; Appel, J.; Bassereau, P.; Marignan, M.; Skouri, M.; Billard, J.; Delsanti, M.; Candau, S. J.; Jahn, J.; Nabre, P. Prog. Colloid Polym. Sci. 1991,84, 264. (24) Kahlweit, M.; Strey, R.; Haase, D.; Kunieda, H.; Schmeling, T.; Faulhaber, B.; Borkovec, M.; Eicke, H. F.; Busse, G.; Eggers, F.; Funck, T.; Richmann, H.; Magid, L.; Soderman, O.; Stilbs, P.; Winkler, J.; Dittrich, A.; Jahn, W. J. Colloid Interface Sci. 1987,118, 445. (25) Miller, C.; Gradzielski, M.; Hoffmann, H.; Kr€amer, U.; Thunig, C. Colloid Polym. Sci. 1990,268, 1066–1072. (26) Kr€amer, U.; Hoffmann, H. In The Structure, Dynamics, and Equilibrium Properties of Colloidal Systems; Bloor, D. M., Wyn-Jones, E., Eds.; NATO ASI Series: Series C, Mathematical and Physical Sciences 324;Kluwer Academic Publishers: Boston, 1990; pp 385396. (27) Yamaguchi, Y.; Hoffmann, H. Colloids Surf., A 1997, 121,67–80. (28) Kahlweit, M.; Strey, R.; Haase, D.; Kunieda, H.; Schmeling, T.; Faulhaber, B.; Borkovec, M.; Eicke, H. F.; Busse, G.; Eggers, F.; Funck, T.; Richmann, H.; Magid, L.; Soderman, O.; Stilbs, P.; Winkler, J.; Dittrich, A.; Jahn, W. J. Colloid Interface Sci. 1987,118, 443–444. (29) Uhrmeister,P.SelbstorganisierendeNetzwekeinMikroemulsionen. Ph.D. Thesis, Universit€at K€oln, 2002. (30) Hoffmann, H.; Abdel-Rahem, R. Colloid Polym. Sci 2010, 288, 603–612. (31) Hoffmann, H.; Ulbricht, W. J. Colloid Interface Sci. 1989, 129, 388–405. (32) Lichterfeld, F.; Schmeling, T.; Strey, R. J. Phys. Chem. 1986, 90, 5762–5766. 80 S1 Dynamic Properties of Microemulsions in the Single Phase Channels Supporting Information Lukas Wolf*1, Heinz Hoffmann1, Walter Richter2, Takashi Teshigawara3, Tohru Okamoto3 1University of Bayreuth, BZKG and BayColl, Gottlieb-Keim-Str. 60, 95448 Bayreuth, Germany 2Friedrich-Schiller-University, Centre for Electron Microscopy, Ziegelmühlenweg 1, 07743 Jena, Germany 3Shiseido Research Center, 2-2-1 Hayabuchi, Tsuzuki-ku, Yokohama, Japan 224-8558 *corresponding author, Tel.: +49-921-50736168, e-mail: [email protected] SI1: Phase sequence of binary surfactant mixture Mg(DS)2-IT 3 Pictures are shown without and between crossed polarisers to distinguish between birefringent and non-birefringent phases. The surfactant mixture starts with the pure 15% Mg(DS)2solution that is a micellar L1-Phase. The sample shows a crystalline sediment at 25 °C, whereas the samples with a mass fraction x IT 3 of 0.1 and 0.2 are isotropic. This shows, that the Krafft-Temperature of the Mg(DS)2, that is close to 25 °C, is lowered by the co-surfactant IT 3. The isotropic micellar L1-region is followed by a two-phase L1/Lα-situation from x IT 3 0.3 – 0.4. From x IT 3 0.5 – 0.77, a large single birefringent Lα region is present that turns into an isotropic L3-phase at a narrow composition range of x IT 3 ~ 0.78 – 0.79. The sample with 15% IT 3 (x IT 3 = 1) shows a two phase situation with a lower L1 and an upper L2-Phase (inverse micellar structure). SI1: Phase sequence of the surfactant mixture Mg(DS)2/IT 3 at 15% (w/w) surfactant and 25 °C, 20% glycerine in H2O. x IT 3 indicates the mass fraction of the non-ionic co-surfactant in the surfactant mixture. Lower row: samples shown in between crossed polarisers. 81 S2 SI2: Rheology measurements of the surfactant mixtures without oil Rheograms of the surfactant mixtures are shown when no oil has been solubilised but when the fraction of IT 3 is increased for constant surfactant concentration of 15%. Sample composition and measurement condition is given in each graph. The rheogram for the sample with x IT 3 0.2 (L1-Phase) shows the typical characteristics of a viscous Newtonian solution, where the storage modulus G’ and the loss modulus G’’ are increasing with frequency f and the viscosity η stays independent of the frequency. The high viscosity of around 110 mPas is an indication for the presence of long entangled worm-like micelles. 0,01 0,1 1 1E-5 1E-4 1E-3 0,01 0,1 1 10 G' [Pa] G'' [Pa] η [Pas] G',G'' [Pa] f [Hz] 0,01 0,1 1 η [Pas] x IT 3 0.2 τ = 0.5 Pa 0,01 0,1 1 0,1 1 G' [Pa] G'' [Pa] η [Pas] G',G'' [Pa] f [Hz] 0,01 0,1 1 x IT 3 0.75 τ = 0.03 Pa η [Pas] 0,01 0,1 1 10 0,1 1 10 100 G' [Pa] G'' [Pa] η [Pas] G',G'' [Pa] f [Hz] 1 10 100 1000 x IT 3 0.7 τ = 0.5 Pa η [Pas] 0,01 0,1 1 1E-6 1E-5 1E-4 1E-3 0,01 0,1 G' [Pa] G'' [Pa] η [Pas] G',G'' [Pa] f [Hz] 0,01 0,1 x IT 3 0.79 τ = 0.1 Pa η [Pas] 0,01 0,1 1 10 1 10 100 1000 G' [Pa] G'' [Pa] η [Pas] G',G'' [Pa] f [Hz] 1 10 100 1000 x IT 3 0.5 τ = 0.5 Pa η [Pas] 0,01 0,1 1 1E-3 0,01 0,1 1 G' [Pa] G'' [Pa] η [Pas] G',G'' [Pa] f [Hz] 0,01 0,1 1 x IT 3 0.75 τ = 0.1 Pa η [Pas] 82 S3 Rheograms of the Lα-phases show different features. The stability region for the Lα-phases reaches from a weight fraction of IT 3 from 0.5 to 0.75. It is noteworthy that the rheological properties of the phases vary considerably. At x IT 3 0.5 – 0.7, the phases behave like soft gels. These phases have a rheological yield stress. The storage modulus is frequency independent and about an order of magnitude higher than the loss modulus. Samples of these phases can therefore be used for the preparation of stable emulsions or dispersions. The dispersed particles or oil droplets, or even air bubbles, cannot sediment or up-cream. The systems can be deformed at least 10% before they start flowing. With increasing mass fraction of IT 3, the storage modulus of the phases decreases over a wide region of the composition. The rather high storage modulus is probably a result of the ionic charge of the bilayer phases. It is therefore the more remarkable that it decreases dramatically when the composition of the Lα-phase approaches the phase boundary of the L3phase. With x IT 3 = 0.75, the storage modulus drops to values below 1 Pa and becomes even frequency dependent at higher shear stress and the phase no longer behaves like a weak gel, but more like a Newtonian solution. The reason for this effect lies in a change of the nanostructure from multilamellar vesicles to planar lamellas with increasing mass fraction of IT 3. At x IT 3 0.79, the rheogram for the L3-Phase shows the typical features of a low viscous Newtonian fluid. Measurements with increasing and decreasing shear rates were done to investigate the stability of the Lα-Phases. Viscosity η and shear stress τ in dependency of shear rate γ &of samples at x IT 3 0.5 (left) and x IT 3 0.75 (right). Curves for increasing shear rate indicated with black arrows to the right, for decreasing shear rate with black arrows to the left. Hysteresis effect of sample with x IT 3 0.75 indicated with red arrow. The data of viscosity and shear stress for inand decreasing shear rates of the sample at x IT 3 0.5 lie perfectly together. This shows that the gel, and thus its nanostructure, doesn’t get destroyed or changed when high shear rates are applied to the sample. The situation becomes very different close to the phase border of the L3-Phase with higher mass fraction of IT 3. In contrast to the sample with x IT 3 0.5, the curves for the sample with x IT 3 0.75 show an enormous hysteresis effect. For increasing shear rates, the viscosity first decreases, but then recovers at a higher level with decreasing shear rates. This shear thickening effect could be explained by the transformation of planar lamellas to multilamellar vesicles due to high shear. The structure relaxes to its original state within one or two hours. 0,1 1 10 100 0,1 1 10 100 τ [Pa] η [Pas] τ [Pa] . γ [1/s] . 0,1 1 10 x IT 3 0.75 η [Pas] Hysteresis 0,1 1 10 100 10 100 τ [Pa] η [Pas] γ [1/s] τ [Pa] 0,1 1 10 100 x IT 3 0.5 η [Pas] . 83 S4 SI3: Rheology measurements in the Lα-channel with increasing oil content Rheograms of samples in Lα-channel with increasing mass fraction of oil are shown. Sample composition is given in each graph. All samples along the Lα-channel behave like soft gels. At the sample with x decane 0.9, at very low frequencies the viscous properties are dominant while at frequencies > 0,1 Hz the elastic properties are dominant. A plot of the storage moduli G’ (at a frequency of 1 Hz) of the lamellar phases against the oil content is shown in the next picture. With increasing solubilisation of decane, the level of storage modulus G’ decreases more than two orders of magnitude from about 130 Pa at the sample without decane to 0,4 Pa at the sample with x decane 0.9. 0,01 0,1 1 10 1 10 100 1000 G' [Pa] G'' [Pa] η [Pas] G',G'' [Pa] f [Hz] 1 10 100 1000 η [Pas] x IT 3 0.55 x decane 0 0,01 0,1 1 10 1 10 100 1000 x IT 3 0.55 x decane 0.2 G' [Pa] G'' [Pa] η [Pas] G',G'' [Pa] f [Hz] 0,1 1 10 100 1000 η [Pas] 0,01 0,1 1 10 0,1 1 10 100 G' [Pa] G'' [Pa] η [Pas] G',G'' [Pa] f [Hz] 0,1 1 10 100 x IT 3 0.65 x decane 0.5 η [Pas] 0,01 0,1 1 10 0,01 0,1 1 G' [Pa] G'' [Pa] η [Pas] G',G'' [Pa] f [Hz] 0,01 0,1 1 x IT 3 0.7 x decane 0.9 η [Pas] 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1 10 100 G' [Pa] x decane G' [Pa] 0,45 0,50 0,55 0,60 0,65 0,70 0,75 sample composition x IT 3 84 S5 Measurements with inand decreasing shear rates also where done within the Lα-channel. Two rheograms, one with low, the other with high oil content, are shown. Curves for increasing shear rate are indicated with black arrows to the right, for decreasing shear rate with black arrows to the left. Hysteresis effect indicated with a red arrow. Compared to the rheogram of the sample without oil at low x IT 3 values, all samples with oil in the Lα-channel showed a hysteresis effect. However, the dimension of the hysteresis effect becomes smaller with increasing oil content. It is conceivable that the oil swollen bilayers of the Lα-Phases allow fluctuations that are not possible without oil. A possible candidate for such fluctuations would be peristaltic fluctuations that are fluctuation of the thickness of the bilayers. Another significant feature of the large Lα-channel is its high temperature stability. Samples between 10 °C and 60 °C between crossed polarizers are shown in the next picture. Samples between x decane = 0 – 0.9 are stable at least between 10 °C – 40 °C, samples with x decane 0 - 0.4 are even stable at least up to 60 °C. Although the Lα-channel is quite narrow, its high temperature stability is amazing. This feature is obvious due to the surfactant mixture, where the hydrophilic-lipophilic balance is determined more by the mass fraction of the cosurfactant than by the temperature. 0.1 1 10 0.1 1 10 τ [Pa] η [Pas] γ [1/s] τ [Pa] 0.1 1 10 x IT 3 0.7 x decane 0.9 η [Pas] . Hysteresis 0,1 1 10 1 10 100 1000 τ [Pa] η [Pas] γ [1/s] τ [Pa] 1 10 100 x IT 3 0.55 x decane 0.2 . η [Pas] Hysteresis 85 S6 SI4: Tube pictures of microemulsions of the lower and upper single phase channel Samples with 15% (w/w) surfactant at 25 oC and increasing mass fraction x of IT 3 and decane are shown without polarizers. Upper row: selected samples from lower single phase channel, lower row: selected samples from upper single phase channel. Note that samples from the lower single phase channel are completely transparent while samples from the upper single phase channel are slightly bluish and have a high scattering intensity around x decane 0.03 – 0.1 SI5: Conductivity measurements in the Lα-channel The plot of the conductivities in the Lα-channel is shown. Conductivity values are indicated as red dots, the corresponding IT 3 ratios are indicated as grey triangles. Note that the conductivity decreases more or less constantly with increasing mass fraction of IT 3 and oil. 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1 10 100 1000 conductivity [µS/cm] x decane conductivity [µS/cm] 0,45 0,50 0,55 0,60 0,65 0,70 0,75 sample composition x IT 3 86 S7 SI6: EB signals at different field strenghs EB-signal for a microemulsion from the upper single phase channel with 6% decane in the solvent mixture, recorded at different field strengths E. Note that the signals do not depend on the field strength. The signal identifies three different relaxation processes. Duration of electric pulse = 1 msec. SI7: EB-Signals in the upper single phase channel (x decane 0.005 – 0.1) Sample composition given in each graph. Note the signals become complicated in the range between 1 and 10% decane in the solvent mixture. 0,00 0,25 0,50 0,75 1,00 1,25 1,50 1,75 2,00 -1,0x10 -8 -5,0x10 -9 0,0 5,0x10 -9 1,0x10 -8 1,5x10 -8 2,0x10 -8 2,5x10 -8 3,0x10 -8 ∆ ∆ ∆ ∆ n t [msec] E = 91 kVm -1 0,00 0,25 0,50 0,75 1,00 1,25 1,50 1,75 2,00 0,0 2,0x10 -8 4,0x10 -8 6,0x10 -8 8,0x10 -8 1,0x10 -7 E = 181 kVm -1 ∆ ∆ ∆ ∆ n t [msec] 0,00 0,25 0,50 0,75 1,00 1,25 1,50 1,75 2,00 0,0 2,0x10 -8 4,0x10 -8 6,0x10 -8 8,0x10 -8 1,0x10 -7 1,2x10 -7 1,4x10 -7 1,6x10 -7 1,8x10 -7 2,0x10 -7 E-field turned off τ3 τ2 E = 273 kVm-1 ∆ ∆ ∆ ∆n t [msec] τ1 E-field turned on τ1 τ2 0 2 4 6 8 10 12 14 16 18 20 -2,50x10 -8 0,00 2,50x10 -8 5,00x10 -8 7,50x10 -8 1,00x10 -7 1,25x10 -7 1,50x10 -7 1,75x10 -7 2,00x10 -7 2,25x10 -7 E = 273 kVm-1 ∆ ∆ ∆ ∆ n t [msec] τL = 14 msec 0,0 0,1 0,2 0,3 0,4 0,5 -5,0x10 -8 0,0 5,0x10 -8 1,0x10 -7 1,5x10 -7 2,0x10 -7 ∆ ∆ ∆ ∆n t [msec] x IT 3 0.753 x decane 0.005 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0 -5,0x10-8 0,0 5,0x10-8 1,0x10-7 1,5x10-7 2,0x10-7 2,5x10-7 ∆ ∆ ∆ ∆ n t [msec] x IT 3 0.726 x decane 0.01 87 S8 SI8: EB-Signals in the upper single phase channel (x decane 0.17 – 0.8) EB-Signals of microemulsion of the upper single phase channel with 17 – 80% decane in the solvent mixture. Sample composition given in each graph. Note that the signals become simple again and the main structural relaxation time decreases with increasing oil content. 012345678910 -2,0x10 -7 0,0 2,0x10 -7 4,0x10 -7 6,0x10 -7 8,0x10 -7 1,0x10 -6 1,2x10 -6 x IT 3 0.638 x decane 0.17 ∆ ∆ ∆ ∆ n t [msec] 0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 -5,0x10 -8 0,0 5,0x10 -8 1,0x10 -7 1,5x10 -7 2,0x10 -7 2,5x10 -7 3,0x10 -7 3,5x10 -7 4,0x10 -7 x IT 3 0.7 x decane 0.3 ∆ ∆ ∆ ∆n t [msec] 1 2 3 4 5 6 7 8 9 10 -5,0x10 -8 0,0 5,0x10 -8 1,0x10 -7 1,5x10 -7 2,0x10 -7 2,5x10 -7 ∆ ∆ ∆ ∆ n t [msec] x IT 3 0.684 x decane 0.0225 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 -1,5x10 -7 -1,0x10 -7 -5,0x10 -8 0,0 5,0x10 -8 1,0x10 -7 1,5x10 -7 2,0x10 -7 2,5x10 -7 ∆ ∆ ∆ ∆n t [msec] x IT 3 0.633 x decane 0.04375 0 5 10 15 20 25 30 35 40 45 50 -5,00x10 -8 -2,50x10 -8 0,00 2,50x10 -8 5,00x10 -8 7,50x10 -8 1,00x10 -7 1,25x10 -7 1,50x10 -7 x IT 3 0.613 x decane 0.07875 ∆ ∆ ∆ ∆ n t [msec] 0 20 40 60 80 100 120 140 160 180 200 -5,0x10 -8 0,0 5,0x10 -8 1,0x10 -7 1,5x10 -7 2,0x10 -7 2,5x10 -7 x IT 3 0.61 x decane 0.1 ∆ ∆ ∆ ∆ n t [msec] 88 S9 SI9: EB-Signals in the lower single phase channel (x decane 0.1 – 0.4) EB-signals for the microemulsions of the lower single phase channel. Sample composition given in each graph. Note that the signals are less complicated than signals from the upper single phase channel. Main structural relaxation time increases with increasing oil content. 0,00 0,25 0,50 0,75 1,00 1,25 1,50 1,75 2,00 -5,00x10 -9 -2,50x10 -9 0,00 2,50x10 -9 5,00x10 -9 7,50x10 -9 1,00x10 -8 1,25x10 -8 x IT 3 0.35 x decane 0.1 ∆ ∆ ∆ ∆ n t [msec] 0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 -1,0x10 -8 0,0 1,0x10 -8 2,0x10 -8 3,0x10 -8 4,0x10 -8 5,0x10 -8 6,0x10 -8 7,0x10 -8 8,0x10 -8 x IT 3 0.45 x decane 0.2 ∆ ∆ ∆ ∆ n t [msec] 0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 -5,0x10 -8 0,0 5,0x10 -8 1,0x10 -7 1,5x10 -7 2,0x10 -7 2,5x10 -7 3,0x10 -7 3,5x10 -7 x IT 3 0.5 x decane 0.3 ∆ ∆ ∆ ∆n t [msec] 0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5 4,0 4,5 5,0 5,5 6,0 6,5 7,0 7,5 8,0 -2,0x10 -8 0,0 2,0x10 -8 4,0x10 -8 6,0x10 -8 8,0x10 -8 1,0x10 -7 1,2x10 -7 1,4x10 -7 1,6x10 -7 x IT 3 0.55 x decane 0.4 ∆ ∆ ∆ ∆n t [msec] 0,00 0,25 0,50 0,75 1,00 1,25 1,50 1,75 2,00 -5,0x10-8 0,0 5,0x10-8 1,0x10-7 1,5x10-7 2,0x10-7 2,5x10-7 3,0x10-7 3,5x10-7 4,0x10-7 4,5x10-7 5,0x10-7 5,5x10-7 6,0x10-7 x IT 3 0.8 x decane 0.5 ∆ ∆ ∆ ∆n t [msec] 0,00 0,25 0,50 0,75 1,00 1,25 1,50 1,75 2,00 -2,50x10-8 0,00 2,50x10-8 5,00x10-8 7,50x10-8 1,00x10-7 1,25x10-7 x IT 3 1 x decane 0.8 ∆ ∆ ∆ ∆ n t [msec] 89 ARTICLE The results show a high maximum at around 6 % oil. At higher oil content the viscosities decrease slowly towards the oil side of the microemulsion. The abrupt maximum of the viscosity can be taken as an indication of a structural transition in the channel. The viscosity maximum is reached at about the same oil concentration where the conductivity disappeared. Obviously the transition of the structure is also expressed in the viscosity of the system. The Journal of Physical Chemistry B Fdx.doi.org/XX.XXXX/jpXXXXXXX | J. Phys. Chem. B XXXX, XXX, 000– 000 Figure 8. Zero shear viscosity, η, against the mass fraction of oil of samples from the upper single phase channel: a) Viscosity with increasing mass fraction of decane; b) Viscosity with increasing mass fraction of isooctane. Figure 7. Rheograms of the binary surfactant mixtures at constant 15% surfactant (w/w) with increasing mass fraction x IT 3 at 25 °C. a) Rheogram of Lαphase at x IT 3 0.7, b) Rheogram of L3phase at x IT 3 0.79. We showed that w/o-HIPME structures with dimensions of about 20 - 100 nm can be thermodynamically stable structures in microemulsions. These structures occur in microemulsions formed from surfactant mixtures of anionic and non-ionic surfactants. They are observed in the isotropic channel of microemulsions, when small amounts of oil are solubilized into aqueous L3phases. In spite of their small oil content, the HIMPE phases have a conductivity that is about 3 - 4 orders of magnitude lower than the conductivity of the L3phase. These results thus demonstrate that very thin surfactant films with little oil can practically be impenetrable for the transport of ions. It can be concluded from the measurements that structural transitions can occur in micellar phases, in which bicontinuous structures of non-ionic surfactants and oil are transformed into foam-like structures, when the surfactant layers are charged by ionic surfactants. It is likely that 0,01 0,1 1 10 0,1 1 10 100 G' [Pa] G'' [Pa] η η η η [Pas] G',G'' [Pa] f [Hz] a) 1 10 100 1000 x IT 3 0.7 τ = 0.5 Pa η [Pas] 0,01 0,1 1 1E-6 1E-5 1E-4 1E-3 0,01 0,1 G' [Pa] G'' [Pa] η η η η [Pas] G',G'' [Pa] f [Hz] 0,01 0,1 b) x IT 3 0.79 τ = 0.1 Pa η [Pas] 0,00 0,05 0,10 0,15 0,20 0,25 0,30 0,35 0,40 0,45 0,50 0 50 100 150 200 250 300 350 400 450 500 a) η [mPas] η η η η [mPas] x decane 0,00 0,05 0,10 0,15 0,20 0,25 0,30 0,35 0,40 0,45 0,50 50 100 150 200 250 300 350 400 450 500 b) η [mPas] η η η η [mPas] x iso-octane CONCLUSIONS such transitions can also be produced, when the electric doublelayer in ionically charged systems is shielded by excess salt. AUTHOR INFORMATION Corresponding Author *Tel:+49-921-50736168. Fax+49-921-50736139. E-mail: lukas. [email protected]. ACKNOWLEDGMENT We thank the Technion Russell Berrie Nanotechnology Institute (RBNI) and the Israel Science Foundation (GRANT NO. 962/07) for partial financial support. The electron microscopy was performed at the RBNI Laboratory for Electron Microscopy of Soft Matter. We thank Judith Schmidt and Dr. Ellina Kesselman for their excellent technical assistance, and all members of Prof. Talmon’s group for their help. REFERENCES (1) Stubenrauch, C. In Microemulsions: Background, New Concepts, Applications, Perspectives; Wiley-Blackwell: Chichester, U.K., 2009. 96 ARTICLE The Journal of Physical Chemistry B Gdx.doi.org/XX.XXXX/jpXXXXXXX | J. Phys. Chem. B XXXX, XXX, 000– 000 (1) Stubenrauch, C. In Microemulsions: Background, New Concepts, Applications, Perspectives; Wiley-Blackwell: Chichester, U.K., 2009. (2) Monduzzi, M.; Caboi, F.; Larché, F.; Olsson, U. Langmuir 1997, 13, 2184-2190. (3) Bergenholtz, J.; Romagnoli, A. A.; Wagner, N. J. Langmuir 1995, 11, 1559-1570. (4) Zemb, T. N.; Hyde, S. T.; Derian, P.-J.; Barnes, I. S.; Ninham, B. W. J. Phys. Chem. 1987, 91, 3814-3820. (5) Kotlarchyk, M.; Sheu, E. Y.; Capel, M. Physical Review A 1992, 46, 928. (6) Eastoe, J. Langmuir 1992, 8, 1503-1506. (7) Huang, J. S. J. Surface Sci. Technol. 1989, 5, 83-131. (8) Jahn, W.; Strey, R. J. Phys. Chem. 1988, 92, 2294-2301. (9) Strey, R. Colloid & Polymer Sci. 1994, 272, 1005-1019. (10) Olsson, U.; Würz, U.; Strey, R. J. Phys. Chem. 1993, 97, 4535. (11) Lindman, B.; Shinoda, K.; Olsson, U.; Anderson, D.; Karlström, G.; Wennerström, H. Colloids and Surfaces 1989, 38, 205-224. (12) Kahlweit, M. et al., J. Colloid Interface Sci. 1987, 118, 445. (13) Belkoura, L.; Stubenrauch, C.; Strey, R. Langmuir 2004, 20, 4391-4399. (14) Tlusty, T.; Safran, S. A.; Menes, R.; Strey, R. Physical Review Letters 1997, 78, 2616-2619. (15) Kahlweit, M.; Faulhaber, B.; Busse, G. Langmuir 1994, 10, 2528-2532. (16) Wolf, L.; Hoffmann, H.; Watanabe, K.; Okamoto, T. Phys. Chem. Chem. Phys. 2011, 13, 3248-3256. (17) Talmon, Y. “Seeing Giant Micelles by CryogenicTemperature Transmission Electron Microscopy (Cryo-TEM)”, in “Giant Micelles”, chapter 5, pp. 163-178, Zana, R. and Kaler, E.A., Eds., CRC Press, New York, 2007. (18) Danino, D.; Bernheim-Groswasser, A.; Talmon, Y. Colloid Surf. A: Physicochem. Eng. Asp. 2001, 183, 113-122. (19) Danino, D.; Gupta, R.; Satayavolu, F.; Talmon, Y. J. Colloid Interface Sci. 2002, 249, 180-168. (20) Wolf, L.; Hoffmann, H.; Richter, W.; Teshigawara, T.; Okamoto, T. J. Phys. Chem. B 2011, 115, 11081-11091. (21) Miller, C. A.; Raney, K. H. Colloids and Surfaces A 1993, 74, 169-215. (22) Sottmann, T.; Strey, R. Microemulsions, In: Fundamentals of Interface and Colloid Science, Lyklema J. Ed., Elsevier Academic Press, Heidelberg, Germany, 2005. (23) Wehling, A. The dynamics of L3phases. Ph.D. Thesis, University of Cologne, 2001. (24) Hoffmann, H.; Thunig, C.; Munkert, U.; Meyer, H. W.; Richter, W. Langmuir 1992, 8, 2629-2638. (25) Burauer, S.; Bekoura, L.; Stubenrauch, C.; Strey, R. Colloid Surface A 2003, 228, 159. (26) Choi, S. M.; Chen, S. H.; Sottmann, T.; Strey, R. Physica B 1997, 241, 976. (27) Cameron, N. R.; Sherrington, D. C. Advances in Polymer Science, 1996, 126, 163-214. (28) Hoffmann, H.; Ebert, G. Angewandte Chemie 1988, 27, 902-912. (29) Ninham, B. W.; Chen, S. J.; Evans, D. F. J. Phys. Chem. 1984, 88, 5855-5857. (30) Ben Barak, I.; Talmon Y., Langmuir (submitted). (31) Hoffmann, H., Thunig, C.; Schmiedel, P.; Munkert, U. Langmuir 1994, 10, 3972-3981. (32) Uhrmeister, P. Selbstorganisierende Netzwerke in Mikroemulsionen, Ph.D Thesis, University of Cologne, 2009. (33) Miller, C. A.; Gradzielski, M.; Hoffmann, H.; Krämer U.; Thunig, C. Progr. in Coll. & Polym. Sci. 1991, 84, 243 (34) Faulhaber, B. J. Colloid Interf. Sci. 1987, 118, 443-444. 97 PFG-NMR in the Single Phase Channels of Microemulsions with an anionic/non-ionic surfactant mixture 98 5.1.5. PFG-NMR in the Single Phase Channels of Microemulsions with an anionic/nonionic surfactant mixture Lukas Wolf*, Heinz Hoffmann, Jürgen Linders and Christian Mayer * corresponding author Submitted to Soft Matter in December 2011 current status: under revision DOI:.- - - CREATED USING THE RSC ARTICLE TEMPLATE (VER. 3.1) - SEE WWW.RSC.ORG/ELECTRONICFILES FOR DETAILS Paper www.rsc.org/softmatter | Soft Matter This journal is © The Royal Society of Chemistry [2011] Soft Matter, [2011], [vol] , 00–00 | 1 PFG-NMR Self Diffusion Measurements in the Single Phase Channels of a Microemulsion System with an Anionic/Nonionic Surfactant Mixture Lukas Wolf,*a Heinz Hoffmann,*a Jürgen Linders*b, Christian Mayer*b Received (in XXX, XXX) Xth XXXXXXXXX 200X, Accepted Xth XXXXXXXXX 200X First published on the web Xth XXXXXXXXX 200X DOI: 10.1039/b000000x The single phase channels of a presently reported microemulsion system were investigated by electrical conductivity and pulsed-field gradient nuclear magnetic resonance (PFG-NMR) spectroscopy. The system consists of a mixed anionic/non-ionic surfactant mixture, water and decane. At constant surfactant concentration and temperature, the phase diagram exhibits two single phase microemulsion channels, separated by an anisotropic lamellar channel. The lower microemulsion channel starts from the water side at the phase diagram with a micellar L1 phase and reaches with increasing mass fraction of decane in the solvent mixture and increasing mass fraction of lipophilic co-surfactant in the surfactant mixture the middle of the phase diagram. The upper microemulsion channel passes from the aqueous side with an L3 phase to the oil side of the diagram. Conductivity data and self diffusion coefficients, obtained by PFG-NMR, support the previously made conclusion, that the nanostructure in the upper channel undergoes an abrupt transition from a bicontinuous structure to a water-in-oil High Internal Phase Microemulsion (HIPME) with already less than 10% of oil in the solvent mixture, while the structures in the lower microemulsion channel are oil-in-water droplets. The HIPME structure is a feature of the surfactant mixture and probably due to a high interfacial tension between the aqueous diluted surfactant phase and the oil. By addition of salt, the HIPME structures are obviously disturbed, resulting in an increased conductivity and a faster diffusion rate for the water fraction. Introduction Since their discovery in 1943 by Hoar and Schulman, microemulsions were much in the focus of interest by scientists in the field of colloid and polymer science.1 They defined microemulsions as optically isotropic transparent phases, consisting of oil, water and surfactants.2 In contrast to ordinary emulsions, microemulsions are thermodynamically stable.3 Three different types of nanostructures can be distinguished in microemulsions, namely oil droplets in a continuous water phase (o/w), water droplets in a continuous oil phase (w/o) and bicontinuous structures.4 The type of the used surfactant plays an important role for the emerging nanostructures. The most detailed investigated microemulsion systems are those with a single non-ionic surfactant CiEj, water and oil. In such systems, it is possible to pass from water-rich to oil-rich single phase microemulsions, without crossing a phase boundary in the phase diagram.5 In order to stay in the single phase region, one has to adapt the hydrophilic lipophilic balance (HLB) by changing the temperature, as non-ionic surfactants are very temperature sensitive.6 The behaviour and the nanostructures in these single-phase channels are known and theoretically well understood.7 They have been investigated indirectly by electrical conductivity, small angle neutron scattering (SANS), NMR and directly imaged by freeze fracture transmission electron microscopy (FF-TEM).8-10 With increasing temperature and increasing oil content, the structure undergoes a continuous transition from small oil droplets in water at the aqueous side to a bicontinuous structure at the middle of the phase diagram with equal amounts of oil and water to small water droplets in oil at the oil side.11 The structural transition is caused by the change of the amphiphilic properties of the non-ionic surfactant with rising temperature. Thus, the curvature of the amphiphilic monolayer changes from convex, to flat, to concave. The situation is somewhat different in microemulsions prepared with ionic surfactants. In such systems, it is not possible to pass from the single aqueous phase to the oil phase without crossing phase boundaries at constant surfactant concentration.12 The best known systems with ionic surfactants are probably microemulsions with sodium bis(2ethylhexyl) sulfosuccinate (AOT), decane or di-dodecyldimethylammoniumbromide (DDAB), dodecane and water.1314 In contrast to bicontinuous microemulsions with a single nonionic surfactant, in these systems a w/o droplet structure is present at equal amounts of water and oil.15 We reported recently a new microemulsion system with a mixed anionic/nonionic surfactant mixture.16 In such systems it is possible to pass from the aqueous to the oil side in a single phase microemulsion channel at constant surfactant concentration and constant temperature. This is achieved by changing the HLB not by temperature but by adjusting the surfactant-co-surfactant ratio. Conductivity data, electric birefringence measurements and cryo-TEM pictures indicated that the nanostructure in this single phase channel has a w/ostructure at a water/oil ratio of 1/1 and not a bicontinuous 99 2 | Soft Matter, [2011], [vol] , 00–00 This journal is © The Royal Society of Chemistry [2011] structure as achieved with a single non-ionic surfactant. We called this structure High Internal Phase Microemulsion (HIPME). Furthermore, the transition from a bicontinuous L3 phase to a w/o high internal phase microemulsion seemed to be already completed by solubilising less than 10% oil into the system.17-19 In this investigation, we want to prove by pulsed-field gradient nuclear magnetic resonance (PFG-NMR) that this is in deed the case. Moreover we investigated the influence of the addition of excess salt to the microemulsion system by interfacial tension measurements, conductivity and PFG-NMR, as it was tried to transform the HIPME structures to bicontinuous structures by shielding the charge of the anionic surfactant. Results and Discussion Surface and Interfacial Tension Measurements The binary surfactant mixture of our reported microemulsion system is composed of the hydrophilic anionic surfactant Magnesium Dodecyl Sulfate Mg(DS)2 and the lipophilic nonionic co-surfactant Iso-tridecyl-triethylenglycolether IT 3 (C13E3). We chose the Mg-salt of SDS, as it is known to cause lower surface tension values than SDS and it is possible to form sponge like L3 phases with co-surfactants.20-21 The surface tension and the interfacial tension between the aqueous surfactant and the oil phase play an important role for the formation of microemulsions with non-ionic surfactants.22 Optimal solubilisation of oil should occur when the interfacial tension of the dilute surfactant solution has its lowest interfacial tension against the oil phase.23 Low interfacial tension values are observed for surfactant systems which form liquid crystalline Lα or L3 phases at low surfactant concentrations.24 Before we determined the interfacial tension values between our surfactant mixture and the oil decane, we first investigated the surface tension of the surfactant mixture at a mixing ratio of 1/1 (w/w) with increasing surfactant concentration. Fig. 1 Surface tension of the surfactant mixture Mg(DS)2-IT 3 at a surfactant ratio of 1/1 (w/w) with increasing total surfactant concentration at 25 °C. As it can be seen in Fig. 1, the surfactant mixture reaches its critical micelle concentration (cmc) at a value around 0,025% surfactant and reaches a very low surface tension of ~ 26 mN/m. This is in deed a very low value, if one considers the surface tension of SDS around 35 mN/m above its cmc. In Fig. 2 we show the interfacial tension values of the diluted surfactant mixtures with increasing mass fraction of the cosurfactant IT 3 at constant surfactant concentration of 0,5% surfactant against the oil decane. Fig. 2 Interfacial tension of Mg(DS)2-IT 3 with increasing mass fraction x IT 3 in the surfactant mixture against the oil decane. Surfactant concentration constant at 0,5%, measured at 25 °C. A broad minimum of the interfacial tension is reached between x IT 3 = 0,4 – 0,5 with a value of ~ 2,3 mN/m. In this area, the binary surfactant mixture starts to form single phase liquid crystalline Lα phases at higher surfactant concentrations.18 The data are very similar compared to a previously investigated microemulsion system with a silicone oil.16 In contrast to microemulsions with a single non-ionic surfactant, where ultra-low interfacial tensions in the range of 10-3 mN/m are reached, the values with our surfactant system are very high. The reason for this obviously lies in the charge of the anionic surfactant. In systems with ionic surfactants, it should be possible to lower the interfacial tension by shielding the electric charge with excess salt.25 Phase Diagram of Mg(DS)2/IT 3 – H2O/n-Decane A phase diagram of our investigated microemulsion system is shown in Fig. 3. The total surfactant concentration was kept constant at 15% (w/w) and the temperature at 25 oC. Samples were prepared with 20% glycerine in H2O to prevent freezing artefacts in freeze fracture transmission electron microscopy (FF-TEM) investigations, that were done previously.18 The phase diagram contains two isotropic microemulsion channels, a lower one and an upper one. The upper one begins on the surfactant axis at the region of the L3 phase. With increasing oil, the channel first shifts to a lower IT 3 ratio and then again to a higher IT 3/Mg(DS)2 ratio for higher oil ratios. It ends on the oil side at 80% decane and pure IT 3 as surfactant. The lower channel begins at the L1 region and ends in the middle of the phase diagram at an IT 3 ratio of 0.57. Both channels are separated by a large single phase birefringent Lα region that extends from 0% to 90% decane with slightly increasing mass fraction of IT 3. 1E-3 0,01 0,1 1 25 30 35 40 45 50 55 60 65 surface tension [mN/m] surfactant concentration [%] surface tension 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0 0 1 2 3 4 5 6 7 8 9 10 11 12 interfacial tension [mN/m] x IT 3 interfacial tension 100 This journal is © The Royal Society of Chemistry [2011] Soft Matter, [2011], [vol] , 00–00 | 3 Fig. 3 Phase diagram of system Mg(DS)2/IT 3 – H2O/decane at 15% (w/w) surfactant and 25 oC, 20% glycerin in H2O. x IT 3 = mass fraction of IT 3 in the surfactant mixture, x decane = mass fraction of decane in the solvent mixture. “ME” indicates isotropic microemulsion area, Lα indicates area of anisotropic lamellar channel. The microemulsions in the lower single phase channel are transparent phases that show no flow birefringence under shear. The samples in the upper phase channel have different properties. While the L3 phase without decane is completely transparent, the samples with decane look somewhat bluish and their scattering intensity is most intensive around x decane 0.03 and 0.1. For higher oil content the scattering intensity is decreasing again. A good and quick method that gives first indications for the nanostructures in microemulsions is the measuring of the electric conductivity. It helps to distinguish between conducting water continuous phases and non-conducting oil continuous phases.26 Because we use a surfactant mixture with an anionic surfactant, no additional salt has to be added to follow the conductivity in the phases in contrast to microemulsions with only a single non-ionic surfactant. The plots of the conductivities in the upper and lower single phase channels are shown in Fig. 4a and 4b. In the upper channel, the conductivity first increases slightly from around 1000 µS/cm of the sample without decane to 1160 µS/cm to the sample with 1% decane. The reason for this lies in the change of the composition of the surfactant mixture. In the range from 1% - 10% decane, the conductivity decreases abruptly three orders of magnitude to 1µS/cm even though the fraction of the anionic Mg(DS)2 is increasing. For higher mass fractions of decane, the conductivity values decrease continuously to low values as e.g. 0.03 µS/cm for the sample with a water/oil ratio of 1/1 (w/w). The conductivities thus indicate a dramatic change in the nanostructure of the upper channel with solubilisation of small amounts of oil into the L3 phase. The abrupt collapse of the conductivity indicates that the system changes from a bicontinuous structure to a waterin-oil (w/o) structure. Conductivities in the isotropic channels of microemulsions from non-ionic surfactants have been reported in the literature.27 In such systems, the conductivity in the upper channel decreases continuously with increasing oil content. These measurements have helped to establish the view which we have today from the structures in the upper channel. With increasing oil content, the bicontinuous L3 phase swells with the solubilised oil between the bilayers and is finally transformed at high oil content to a w/o system. Fig. 4 Plot of conductivity (red dots) and IT 3 content (grey triangles) against mass fraction of decane in solvent mixture. a) Conductivity data for the upper single phase channel. b) Conductivity data for the lower single phase channel. With equal amount of oil and water, SANS-data and conductivities show, that this phase is still a bicontinuous phase.28 Our conductivity data unambiguously show that the structures in the upper channel of the presently investigated system are different from the structures of known systems with non-ionic surfactants. We find a rather abrupt transition from the bicontinuous L3 structure to a w/o structure with only 10% of oil in the solvent mixture. Recently published cryo-TEM pictures show a polyhedral w/o foam structure, when 6% of oil was solubilised in the L3 phase.19 These structures were similar to those that are found in so called High Internal Phase Emulsions (HIPE).29 We therefore called the new microemulsion structures High Internal Phase Microemulsions (HIPME). In opposition to the upper channel, the conductivity data of the lower channel indicate that the nanostructure in the lower channel does not change much with increasing oil content. At the water corner, the conductivity in the lower channel with 2900 µS/cm is much higher than the conductivity of the L3 phase of the upper channel with 1000 µS/cm. The reason for this is that the Mg(DS)2 concentration is much higher in the lower channel. With increasing oil content, the conductivities decrease slightly to 1500 µS/cm at the middle of the phase 0,0 0,1 0,2 0,3 0,4 0,5 0,01 0,1 1 10 100 1000 conductivity [µS/cm] x decane conductivity [µS/cm] a) 0,60 0,65 0,70 0,75 0,80 sample composition x IT 3 0,0 0,1 0,2 0,3 0,4 10 100 1000 conductivity [µS/cm] x decane conductivity [µS/cm] 0,10 0,15 0,20 0,25 0,30 0,35 0,40 0,45 0,50 0,55 0,60 sample composition x IT 3 b) 101 4 | Soft Matter, [2011], [vol] , 00–00 This journal is © The Royal Society of Chemistry [2011] diagram, which is where the channel ends. The reason for the decrease is mainly the decreasing mass fraction of Mg(DS)2. Obviously, the lower channel consists of a continuous water phase in which oil droplets are dispersed (o/w-structure). PFG-NMR in the Microemulsion Channels To underline and to verify our results, we investigated the microemulsion channels by PFG-NMR, as this method delivers information about the structure, fluidity and emulsion type. Furthermore, it can give indications about the interaction between surfactant and co-surfactant at the interface. Fig. 5 shows a conventional proton NMR spectrum of the system in the upper channel at x decane 0.7, x IT3 0.85. Fig. 5 Proton NMR spectrum of the system in the upper channel at x decane 0.7, x IT3 0.85. The PFG-NMR analysis is focussed on those spectral regions which can either be clearly assigned to single system constituents (water between 4.8 and 5.3 ppm and decane between 1.3 and 2.0 ppm) or to the mixture of the surfactants (Mg(DS)2/IT3 between 0.2 and 0.4 ppm). The integrals of these three spectral regions strongly depend on the strength of the gradient pulse, hereby indicating the average displacement of the corresponding system constituents during the period between the pulses which was set to 50 ms. In a plot of the logarithmic relative signal intensity ln I/I0 vs. the parameter γ²G²δ²(∆-δ/3) (with γ being the gyromagnetic ratio of protons, G the strength of the gradient field, δ and ∆ the duration of and the spacing between the two gradient pulses), the slope is equal to the negative apparent self diffusion coefficient of the given component in the heterogeneous system (StejskalTanner plot). If the component is located in two different environments leading to clearly different self diffusion properties, the plot will show two sections with clearly different slopes. If the component is encapsulated in very small droplets, the motion within the droplets becomes undetectable. In this case, the observed slope reflects the diffusive dislocation connected to the Brownian motion of the droplets. The resulting Stejskal-Tanner plots for four different states in the upper channel are shown in Fig. 6. The corresponding apparent self diffusion coefficients are listed in Table 1. The first example (Fig. 6 a) refers to the situation in absence of decane (x decane 0). Here, the water signal follows a steep decay, corresponding to a self diffusion constant of Dw = 6.80·10-10 m²/s. This is just slightly lower than the value for bulk water, indicating that water forms a continuous phase only slightly affected by dispersed phase boundaries. In contrast, the signal for Mg(DS)2/IT3 follows a relatively flat decay, pointing to a structure of the surfactant which only allows a restricted mobility of Mg(DS)2 and IT 3 molecules. The situation changes significantly on the addition of 10% decane (x decane 0.1, Fig. 6 b). Now the mobility of water is reduced by a factor of three to Dw = 2.22·10-10 m²/s. All other system constituents, decane as well as the surfactants, exhibit curved decay profiles connected to two distinctly different diffusion constants for each constituent. The largest portion of decane (and a small portion of the surfactants) show a selfdiffusion constant which, with Dd = 3.65·10-10 m²/s, is approximately half of the value for bulk decane. With the relatively small decane content, this indicates that we actually deal with a continuous decane phase. The slower portion of the decane (approximately 3%) seems to be associated with the majority of the surfactant (Dd = 5.68·10-11 m²/s). Altogether, the diffusion profile is compatible with a high internal phase w/o-microemulsion (w/o-HIPME) of 90% water in 10% decane, stabilized by the surfactants. Obviously, a small fraction of the decane is closely associated with the surfactant layer which explains the slow fraction of decane. Correspondingly, some of the surfactant is being dissolved in the decane phase which explains the fast fraction of the Mg(DS)2/IT 3 signal. The surprisingly high diffusion rate of the water indicates significant exchange of water molecules via the thin decane films which separate the water droplets. With increasing decane content (x decane 0.3 and 0.7), the system gradually changes towards a conventional water in oil microemulsion (Figs. 6 c and d). The mobility of water is further reduced by an order of magnitude to Dw = 2.14·10-11 m²/s and Dw = 2.13·10-11 m²/s, respectively. In addition, the mobility of the surfactant as well as the “slow” fraction of the decane is slowed down by a factor of five (Dd = 1.02·10-11 m²/s and Dw = 8.76·1012 m²/s for x decane 0.3 and 0.7). In contrast, the “fast” fraction of the decane exhibits values which now come close to the bulk diffusion rate (Dd = 4.32·10-10 m²/s and Dd = 7.07·10-10 m²/s for x decane 0.3 and 0.7). In this situation, the observed dislocation for water molecules is largely caused by the Brownian motion of small water droplets in the continuous decane phase. With the given viscosity of decane at room temperature, the diameter of the water droplets can be estimated to approximately 20 nm. As before, we assume that part of the surfactant is dissolved in the continuous decane phase, leading to the initial fast decay of the Mg(DS)2/IT 3 signal. Also, again a small fraction of the decane is dissolved in the surfactant layer around the water droplets, leading to the shallow plateau of the decane signal (Dd = 1.02·1011 m²/s and Dd = 8.76·10-12 m²/s for x decane 0.3 and 0.7, respectively). The fact that the self diffusion coefficient for water is still slightly larger than for the droplet wall constituents indicates the exchange of a small fraction of water molecules between the droplets via the hydrophobic phase, an effect which is linked to Oswald ripening. 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 ppm water decane MDS/IT 3 glycerin/MDS/IT 3 decane /MDS /IT 3 102 This journal is © The Royal Society of Chemistry [2011] Soft Matter, [2011], [vol] , 00–00 | 5 Fig. 6 a-d Stejskal-Tanner plots for decane, water and the surfactants in the upper channel. Tab. 1 Apparent self diffusion coefficients of system constituents in the upper channel x water x decane D (water) D (decane) D (decane) plateau [m2/s] [m2/s] [m2/s] 1 0 6.80 10-10 - - 0.9 0.1 2.22 10-10 3.65 10-10 5.68 10-11 0.7 0.3 2.14 10-11 4.32 10-10 1.02 10-11 0.3 0.7 2.13 10-11 7.07 10-10 8.76 10-12 Fig. 7 Stejskal-Tanner plots for decane, water and the surfactants in the lower channel. Tab. 2 Apparent self diffusion coefficients of system constituents in the lower channel x water x decane D (water) D (water) plateau D (decane) [m2/s] [m2/s] [m2/s] 0.9 0.1 1.04 10-9 1.79 10-11 1.37 10-11 0.7 0.3 8.36 10-10 9.04 10-12 2.88 10-13 In contrast to the results for the upper channel, the variations between the PFG-NMR results of different positions in the lower channel do not indicate dramatic structural changes, even though diffusion constants do vary significantly with x decane. An example for a corresponding Stejskal-Tanner plot for the lower channel is shown in Fig. 7. Apparent self diffusion coefficients for two points in the lower channel are listed in Table 2. The data for water and the surfactants resemble those of the upper channel in absence of decane. Again, the water signal shows very steep decays linked to self diffusion coefficients of 1.04 10-9 m²/s for x decane 0.1 and 8.36 10-10 m²/s for x decane 0.3, values which come close to the one in bulk water. In contrast, the decane signal indicates an increasingly slow mobility (1.37 10-11 m²/s and 2.88 10-13 m²/s) which corresponds to Brownian motion of droplets with increasing size. The surfactant seems to be linked to the decane droplets, even though the diffusion rate is slightly larger. All in all, the data are clearly in accordance with an o/w microemulsion with a droplet size significantly growing with the decane content. Parts of the surfactant molecules may 0,0 5,0x10 10 1,0x10 11 -10 -8 -6 -4 -2 0 water MDS/IT 3 ln I/I0 γ2G2δ2(∆-δ/3) 0% decane (upper channel) a ) 0,0 5,0x10 10 1,0x10 11 -10 -8 -6 -4 -2 0 decane water MDS/IT 3 ln I/I 0 γ 2 G 2 δ 2 (∆-δ/3) 10% decane (lower channel) 0,0 5,0x1010 1,0x1011 -10 -8 -6 -4 -2 0 decane water MDS/IT 3 ln I/I0 γ2G2δ2(∆-δ/3) 30% decane (upper channel) c ) 0,0 5,0x1010 1,0x1011 -10 -8 -6 -4 -2 0 decane water MDS/IT 3 ln I/I 0 γ2 G 2δ2 ( ∆ - δ /3) 70% decane (upper channel) d ) 0,0 5,0x10 10 1,0x10 11 -10 -8 -6 -4 -2 0 decane water MDS/IT 3 ln I/I 0 γ 2 G 2 δ 2 (∆-δ/3) 10% decane (upper channel) b ) 103 6 | Soft Matter, [2011], [vol] , 00–00 This journal is © The Royal Society of Chemistry [2011] Fig. 8 Self diffusion constants of system constituents in the upper channel as a function of the decane and IT 3 content. undergo more rapid diffusion via molecular exchange with micelles which would explain for the slight deviation between the slopes for the decane and the surfactant signals. An extremely small fraction of water molecules may be linked to the droplets and explain a possible plateau of the water signal for ln I/I0 < -8. However, with a contribution of only 0.01%, this signal fraction comes close to the noise amplitude and may be insignificant. All apparent self diffusion constants are summarized in Figs. 8 and 9 as functions of the decane content. The data for the water fraction show a clear correlation with the corresponding conductivity plots in Fig. 4. In the upper channel, the water mobility steeply declines with increasing decane concentration (Fig. 8 top). Fig. 9 Self diffusion constants of system constituents in the lower channel as a function of the decane and IT 3 content. This behavior is reproduced by a corresponding decrease of the conductivity (Fig. 4 top) which can be regarded as a direct consequence: with less mobile water molecules, ions in the aqueous solution can be expected to be less mobile as well. However, this effect is far more dramatic on conductivity than on the mobility of individual water molecules: a reduction of the self diffusion coefficient by a factor of 30 results in a loss in conductivity by more than three orders of magnitude. This may be partially explained by a reduced overall ion concentration connected to the decreasing water content. In case of the lower channel, the loss of water mobility under increasing decane content is much smaller (Fig. 9 top). This is again reflected by the conductivity data in Fig. 4 (bottom) 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 1x10-10 2x10-10 3x10-10 4x10-10 5x10-10 6x10-10 7x10-10 8x10-10 diff [m2/s] % (x decane) diff [m/s] 0,60 0,65 0,70 0,75 0,80 0,85 0,90 sample composition x IT 3 a ) water in the upper channel 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 1x10-10 2x10-10 3x10-10 4x10-10 5x10-10 6x10-10 7x10-10 8x10-10 diff [m2/s] % (x decane) diff [m/s] 0,60 0,65 0,70 0,75 0,80 0,85 0,90 sample composition x IT 3 b ) mobile fraction of decane in upper channel 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 1x10 -10 2x10 -10 3x10 -10 4x10 -10 5x10 -10 6x10 -10 7x10 -10 8x10 -10 diff [m2/s] % (x decane) diff [m/s] 0,60 0,65 0,70 0,75 0,80 0,85 0,90 sample composition x IT 3 c ) slow fraction of decane in upper channel 0,1 0,2 0,3 0,4 0,0 2,0x10-10 4,0x10-10 6,0x10-10 8,0x10-10 1,0x10-9 1,2x10-9 1,4x10-9 1,6x10-9 diff [m 2 /s] % (x decane) diff 0,10 0,15 0,20 0,25 0,30 0,35 0,40 0,45 0,50 0,55 0,60 sample composition x IT 3 a ) mobile fraction of water in lower channel 0,0 0,1 0,2 0,3 0,4 0,0 2,0x10 -12 4,0x10 -12 6,0x10 -12 8,0x10 -12 1,0x10 -11 1,2x10 -11 1,4x10 -11 1,6x10 -11 1,8x10 -11 2,0x10 -11 diff [m2/s] % (x decane) diff 0,10 0,15 0,20 0,25 0,30 0,35 0,40 0,45 0,50 0,55 0,60 sample composition x IT 3 b ) slow fraction of water in lower channel 0,0 0,1 0,2 0,3 0,4 0,0 2,0x10 -12 4,0x10 -12 6,0x10 -12 8,0x10 -12 1,0x10 -11 1,2x10 -11 1,4x10 -11 1,6x10 -11 diff [m2/s] % (x decane) diff 0,10 0,15 0,20 0,25 0,30 0,35 0,40 0,45 0,50 0,55 0,60 sample composition c ) decane in lower channel 104 This journal is © The Royal Society of Chemistry [2011] Soft Matter, [2011], [vol] , 00–00 | 7 which show a minor decrease on addition of decane. Here, a decrease of the water mobility by a factor of 1.2 between x decane 0.1 and 0.3 is accompanied by about the same factor of 1.25 in conductiviy. The reason for this lies mainly in the decreasing mass fraction of the anionic Mg(DS)2 in the surfactant mixture. Influence of Salt to the System As already mentioned, our mixed anionic/nonionic surfactant system has a very high interfacial tension against the oil phase compared to the ultra-low interfacial tensions that can be reached with single non-ionic surfactants. We assumed that by shielding the charge of the anionic surfactant by adding excess salt would lower the interfacial tension. Fig. 10 interfacial tension of the surfactant mixtures against decane with increasing amount of NaCl. Surfactant concentration constant at 0,5% in the aqueous phase. 100% NaCl corresponds to a molar ratio of Mg(DS)2:NaCl = 1:1. a) interfacial tension at x IT 3 = 0.5, b) interfacial tension at x IT 3 = 0.8. Similar effects were already reported for the anionic surfactant diethylhexyl sodium sulphosuccinate (AOT), where ultra-low interfacial tensions against oil were reached with additional NaCl.25 To verify our assumption, we measured the interfacial tension at two mixing ratios of the surfactant and co-surfactant with increasing amount of NaCl, namely around the minimum of the observed interfacial tension at x IT 3 = 0.5 and around the mixing ratio of the L3 phase at x IT 3 = 0.8. As it can be seen in Fig. 10, the interfacial tension is lowered only about 0.5 mN/m at the minimum of the interfacial tension at x IT 3 = 0.5 and only about 0.9 mN/m around the L3 phase at x IT 3 = 0.8, when the molar ratio of Mg(DS)2:NaCl in the surfactant mixtures is raised to 1:1. No ultra-low interfacial tensions were detected. As the volume-drop technique can detect low interfacial tensions down to 0.1 mN/m, there shouldn’t be anything wrong with the measurements. We also checked the influence of salt on the phase behaviour of the upper microemulsion channel. Therefore, we had a closer look on the microemulsion with 30% decane in the solvent mixture and investigated how the phase boundaries would shift by adding NaCl to the system. It turned out that the upper and lower borders of the single phase region are shifted to lower x IT 3 values by x IT 3 ~ 0.07 when we added NaCl to the Mg(DS)2 in a molar ratio of 1:1. The shift to lower x IT 3 values means that the system in total becomes more lipophilic, as less amount of the lipophilic co-surfactant IT 3 in the surfactant mixture is needed to solubilise 30% of decane. If the shift of the phase boundaries is accompanied also by a change in the nanostructure was first investigated by measuring the electric conductivity of the microemulsion with increasing salt concentration. A plot of the conductivity in the single phase region with increasing NaCl concentration is shown in Fig. 11. Fig. 11 Plot of conductivity in the single phase region of a microemulsion with x decane 0.3 and increasing NaCl concentration at 25 °C. 100% NaCl corresponds to a molar ratio of Mg(DS)2:NaCl = 1:1. The conductivity from the NaCl-free to the microemulsion with a molar ratio of Mg(DS)2:NaCl = 1:1 increases about three orders of magnitude from a low value of 3 µS/cm to ~ 1000 µS/cm. The conductivity increases in a sigmoid curve with an inflection point around 50% NaCl and not linearly with increasing NaCl concentration. At first sight, the nanostructure seems to change from a w/o-HIPME system to a bicontinuous-like nanostructure. To verify this, we compared two microemulsions with different salt concentrations by PFG-NMR. The first sample without NaCl had the composition of x IT 3 0.7 and x decane 0.3. The second sample had the composition of x IT 3 0.615, x decane 0.3, and the molar ratio of Mg(DS)2:NaCl = 1:1. The resulting Stejskal-Tanner plots are shown in Fig. 12, the corresponding apparent self diffusion constants are listed in Table 3. 0 10 20 30 40 50 60 70 80 90 100 2,0 2,1 2,2 2,3 2,4 2,5 2,6 interfacial tension [mN/m] NaCl [%] interfacial tension a) 0 10 20 30 40 50 60 70 80 90 100 4,50 4,75 5,00 5,25 5,50 5,75 b) interfacial tension [mN/m] NaCl [%] interfacial tension 0 10 20 30 40 50 60 70 80 90 100 1 10 100 1000 conductivity [µS/cm] sample composition NaCl [%] conductivity [µS/cm] 0,61 0,62 0,63 0,64 0,65 0,66 0,67 0,68 0,69 0,70 x IT3 105 Erklärung 112 9. Erklärung Ich erkläre hiermit, dass ich die vorliegende Arbeit selbständig verfasst und keine anderen als die von mir angegebenen Quellen und Hilfsmittel benutzt habe. Ferner erkläre ich, dass ich nicht anderweitig mit oder ohne Erfolg versucht habe, diese Dissertation einzureichen. Ich habe keine gleichartige Doktorprüfung an einer anderen Hochschule endgültig nicht bestanden. Bayreuth, den 5. Dezember 2011 Lukas Wolf