Novel Cholesterol-based Organogelators

Novel Cholesterol-based Organogelators Chapter 6 Novel Cholesterol-based Organogelators In this chapter the synthesis and gelation properties of novel cholesterol-based organogelators are discussed. The design of these compounds is based on the results of studies of the bisamide and bisurea gelators discussed in previous chapters. Gelation by these cholesterol-based structures is governed by self-assembly via van der Waals and hydrogen bonding interactions, although different structural motifs are used compared to those examined in chapter 3 – 5. In these cholesterol-based gelators, the cyclohexane core is preserved, the hydrogen bonding interactions are provided by carbamate groups and van der Waals interactions are provided by two cholesterol moieties coupled at the C3-position. Gelation was observed in a range of solvents of differing polarity for different gelator types, however, gelation by this class of compounds proved to be more restricted in solvent scope. For three of the gelators, crystals were obtained, which were sufficiently large for single crystal X-ray diffraction studies, allowing for elucidation of the packing in the crystalline state. In conjunction with FTIR spectroscopy of the gels, these results give insight into gelation by these compounds: in apolar solvents gelation is governed primarily by hydrogen bonding and van der Waals interactions and in polar solvents gelation is governed primarily by van der Waals interactions. These findings provide guiding principles for the development of “design rules” for gelation by low molecular weight compounds, which showed that tuning of anisotropy in the intermolecular interactions is necessary. Furthermore, the intermolecular interactions have to consist of, at least, two different types of complementary interactions. In general, to design a gelator it is useful to use an already know scaffold, e.g. a cyclohexane ring, and vary the substituents and find a good gelator ultimately. 149 Chapter 6 150 Novel Cholesterol-based Organogelators 6.1 Introduction 6.1.1 System Introduction Cholesterol and cholesteric derivatives are widely recognised for their aggregation behaviour, not only in health care,1 but also in the field of supramolecular chemistry, as they display different types of hierarchical self-assembly processes. It is most famous for its liquid-crystalline phase transition, the cholesteric phase, a type of chiral nematic phase, which was discovered as early as the 19th century.2,3,4 Other examples of self assembly processes by cholesterol or derivatives thereof are the formation of monolayers at the air-water interface,5 incorporation in bilayer membranes in aqueous solutions,6,7 and gelation of solvents.8 The rich supramolecular behaviour of cholesterol is related to the molecular rigidity and aggregation properties of the steroid skeleton. In the field of low molecular weight organogelators, research on steroids, which are capable of gelation, have lead to a wide range of steroid based gelators, in which the so-called ALS, Aromatic Linker Steroid, group is probably the most studied. These gelators consist, as the name already implies, of a steroid, e.g. cholesterol moiety which is connected to an anthracene,9 stilbene,10 or another type of aromatic moiety at the C3 position of the steroid skeleton via different linkers, e.g. ester or carbamate groups. Compounds in this class are able to gelate solvents ranging from apolar to polar by the formation of long gel fibers, thereby trapping the solution.11,12 Well known examples of ALS gelators are those which can switch the helicity of their gel fibers via a cis-trans isomerisation of an azobenzene group in the linker,13 or transcribe the chirality of the gel fibers to silica via calcination as developed by the group of Shinkai (Figure 6-1).14,15 Figure 6-1. ALS gelator in which the helicity of the gel fiber can be switched via cistrans isomerisation. Other notable cholesteric derivatives which are capable of gelating organic solvents include simple esters of cholesterol coupled at the C3-position, e.g. cholesteryl 4-(2anthryloxy)butanoate (CAB), which gelates apolar solvents (e.g. n-dodecane at 4 wt%) or derivatives of cholesterol functionalised at various position of the steroid.16,17 In these studies it was found that the alkyl tail at the C17 position is necessary for 151 Chapter 6 achieving gelation.18 Moreover, many of the cholesterol-based gelators display multiple cholesteric phases upon heating, which makes these systems fascinating compounds for their rich chemical behaviour in self assembly processes.9 In the systems presented, the driving force for the formation of gel fibers arises from self assembly of the compounds guided by anisotropic intermolecular interactions, an attribute of gelators demonstrated to be necessary for achieving gelation.19 The anisotropy can be provided by different types of intermolecular interactions: van der Waals interactions provided by the steroid moiety and/or linker part, π-π stacking from aromatic units and hydrogen bonding interactions if a group capable of hydrogen bond formation is present. From our studies of bisamide and bisurea gelators (Chapters 3 – 5) it was found that the cyclohexane core is an excellent scaffold to introduce anisotropy in the self assembly of the gelator molecules due to its rigidity. Therefore, it is interesting to note that no steroid-based gelators yet employ the cyclohexane as a core unit to tether two cholesterol units as is used for bisamide and bisurea gelators. Also, from previous studies on bisamide and bisurea gelators it was found that both hydrogen bonding interactions and van der Waals interactions need to be present to gelate apolar as well as polar solvents. Our aim in this study is to develop from the design rules developed from this earlier work a new class of cholesterol-based organogelators to demonstrate the generality of these rules. By applying these rules, the cyclohexane unit is used as a scaffold to attach two steroid moieties at the C3 position. This places the steroid moieties in an arrangement to provide anisotropy in the self-assembly process through van der Waals interactions, necessary for the formation of the gel fibers. In addition, carbamate groups are used as hydrogen bonding linker units between the cyclohexane ring and steroid moiety, to provide additional anisotropy via hydrogen bonding interactions. Having two different types of intermolecular interactions present is believed to be a benefit for gelation in solvents of different polarity.20 As an extra benefit the presence of a cholesteric unit may also induce additional cholesteric phases upon heating, which may lead to interesting phase behaviour of the gels, which influences the properties of gels to a great extent as has been shown already for bisurea gelators.21 6.1.2 Chapter Overview In the first part of this chapter, the synthesis of the novel cholesterol-based compounds is discussed. Subsequently, the results of the gelation behaviour of these compounds in a number of solvents are described. If a gel is formed, the critical gelation concentration (cgc) is determined and compared with the cgc’s of the other gelators. By optical microscopy, in conjunction with Differential Scanning 152 Novel Cholesterol-based Organogelators Calorimetry (DSC), the compounds are screened for possible mesogenic behaviour, a property many cholesterol derivatives exhibit. By Scanning Electron Microscopy (SEM) and transmission electron microscopy (TEM) the morphology of gels formed by the gelators is examined. Single crystal X-ray spectroscopy, together with FTIR spectroscopy, allows a comparison of the packing of the compounds in the crystal and gel state. Via these studies, the different intermolecular interactions leading to the formation of the gel fibers are identified. The results, obtained by the different techniques employed, are discussed at the end of this chapter in view of formulating design rules for new low molecular weight organogelators. 6.2 Synthesis and Design A number of structurally different scaffolds based on cyclohexane were used to compare in which manner the different intermolecular interactions in these compounds contribute to gelation. These scaffolds consist of isomers of cyclohexane diamine substituted at the 1,2-(R,R) (1); 1,2-(S,S) (2); 1,2-(R,S) (3); and the 1,4position (4) with L-cholesterol. These variations allow the influence of the position of the attachment of the cholesterol moiety to the cyclohexane ring on the anisotropy and gelation behaviour to be studied. Also, a mono cholesteryl-substituted cyclohexane gelator 5 was synthesised, to study the requirement for multiple steroid groups to be present for gelation. Furthermore, to study the influence of flexibility of the scaffold on gelation properties, a compound possessing an ethylene diamine core disubstituted with cholesteryl groups 6 was synthesised. These different compounds were prepared by coupling an amine carrying scaffold to the steroid moiety via a reaction with L-cholesterol chloroformate using triethylamine as a base.22 This reaction provides a facile route to couple the two groups in good yield and, as an additional benefit, it generates a carbamate moiety as the hydrogen bonding linker unit. 153 Chapter 6 O NH 2 O + 2 Cl H H CH 2Cl2 O Et3N, Reflux NH 2 * * N H H O 1, * = R,R γ = 69 % 2, * = S,S γ = 78 % 3, * = R,S γ = 53 % O N H H HN O + 2 Cl H H 4, γ = 65 % O HN O Cl H H O H O N H Et 3N, Reflux + NH 2 2 Cl H H N H H O O O H O CH2Cl2 H NH 2 H CH2Cl2 Et 3N, Reflux + H O O NH2 NH2 H O O NH2 H H 5, γ = 82 % H O CH 2Cl2 6, γ = 72 % O H Et3N, Reflux H O H N O Figure 6-2. Synthesis route used for the cholesterol based organogelators 1 – 6. The synthesis of compounds 1 – 6 was performed using similar reaction conditions, using dichloromethane as solvent and an excess of triethylamine. After 18 h all cholesterol chloroformate had reacted, as determined by TLC, and the reaction was stopped. Compounds 2, 4 and 6 were purified by multiple washing steps, followed by drying in vacuo. Compounds 1, 3 and 5 were highly soluble in many organic solvents and were purified by column chromatography on silica gel. The yields of the different compounds were good, ranging from 53 % to 72 %. The largest loss of product is due to the purification step as the crude yields were >85 %. 154 Novel Cholesterol-based Organogelators 6.3 Gelation Properties The gelation behaviour for the cholesterol compounds 1 – 6 was examined in a number of solvents ranging from apolar to polar. The solvents used are listed in Table 6-1 according to the commonly used ET(30) scale for solvent polarity,23 together with the critical gelation concentrations (cgcs)24 for the different organogelators. In most of the solvents examined, the compounds are insoluble at room temperature, however upon heating, they dissolve and subsequent cooling to room temperature results in the formation of a gel (above the cgc). To determine the cgcs of the compounds, the gels are diluted, heated and cooled to room temperature repeatedly, until a gel does not form or the gel becomes too weak to withstand gravity. Table 6-1. Gelation properties of the cholesterol based organogelators 1 – 6 in apolar and polar solvents .[a] Solvent Cyclohexane n-Hexadecane Decalin di-n-Butyl ether p-Xylene Toluene n-Butyl acetate Chloroform Cyclohexanone Dimethyl Sulfoxide 1,2-Dichloroethane Acetonitrile 2-Octanol 1-Propanol Ethanol Water ET(30) 30.9 31.1 1 g (7.5) g (10.0) 2 p p 3 s c 4 p g (15.0) 5 s c 31.2 33.0 33.1 33.9 38.5 39.1 39.8 41.3 45.1 45.6 49 50.7 51.9 63.1 s g (10.0) s s p s p p c i s p p i s p p p p s p p p i p g (5.0) g (2.0) i p s s s s s s s s i p p c i g (12.0) p c c g (7.5) s p p s i i i i i p s s s s s s g (5.0) s p s s c i 6 p i vs (25.0) p p p s s p p p i c p c i [a] The following abbreviations are used: microcrystals: c; gelation: g (minimum gelation concentration in mg compound per mL solvent); insoluble: i; precipitate: p; soluble at room temperature (solubility> 20 mg mL-1): s; viscous solution: vs. Table 6-1 already shows that the cholesterol-based organogelators 1 – 6 are only effective gelators for a limited number of solvents. However, if the solvents gelated by the different compounds are compared, it is noted that the diastereomeric compounds 1 and 2 show remarkably different gelation behaviour. Compound 1, which has R,R chirality for the carbamate groups on the cyclohexane ring, gelates 155 Chapter 6 apolar solvents only, e.g. n-hexadecane, whereas its diastereoisomer 2, having S,S chirality, gelates polar solvents only, e.g. ethanol. By changing the configuration of the carbamate group from trans, for compounds 1 and 2, to cis, (R,S), for compound 3 no gelation behaviour is observed. Compound 3 is soluble in many of the solvents tested and in a few cases it does aggregate, but only to form a precipitate or crystals. It was observed already for cyclohexane-based bisamide gelators, that the presence of a cis-configuration for the hydrogen bonding groups is not beneficial to the gelation behaviour, most probably due to the formation of intramolecular hydrogen bonds and the limited ability to form intermolecular hydrogen bonds.25 The position of the hydrogen bonding units on the cyclohexane ring influences the gelation properties also. By changing the position of the carbamate groups from a 1,2to a 1,4-substitution on the cyclohexane ring, as in compound 4, gelation occurs only in apolar solvents and some mildly polar solvents as n-butyl acetate. In more polar solvents the compound precipitates or crystallises, whereas in the various alcohols tested 4 is insoluble. A rationalisation of these results may be that the lipophilic surface of the compound is expanded by the position of the cholesterol groups compared to 1 – 3, leading to a decrease in solubility. By removing one of the cholesterol moieties on the cyclohexane ring, i.e. mono substituted 5, the solubility of the compound is increased to a large extent. Compound 5 is soluble at high concentrations (> 20 mg mL) in most of the solvents tested ranging form apolar to polar. Remarkably, this compound still is able to gelate DMSO, even at low concentrations (5 mg mL-1). Changing the flexibility in the scaffold, by replacing the cyclohexane ring with an ethylene spacer, as in compound 6, the solubility of the compounds is decreased again and in many solvents precipitation/ crystallisation occurs. No gelation is observed, however, 6 forms a viscous solution in decalin. The gels formed by 2 and 5 are not stable at room temperature and crystallisation over time occurs. Especially for the gels formed by 2 this transition is marked, as the gels formed in the various alcohols are clear at first and in time they become turbid. This process is retarded or even stopped when the gels are stored at lower temperatures. 6.4 Thermotropic Behaviour of the Compounds Many cholesteric compounds display a liquid-crystalline phase transition upon heating.3 For example, cholesterol functionalised at the C3 position with an alkylsubstituted carbamate group, e.g. cholesterol N-n-hexadecyl carbamate, has a smectic (79 °C) mesophase as well as a cholesteric mesophase (81 °C).26 Other compounds, which are comparable to the cholesterol-based organogelators 1 – 6, e.g. cholesteryl N-para-substituted-phenyl carbamates, exhibit cholesteric mesophases, and for these systems the mesophase stability is roughly parallel to the increasing molecular 156 Novel Cholesterol-based Organogelators polarisability.26,27 This is found also in a series of cholesteryl carbonates, which exhibit liquid crystalline states.28 In these compounds the molecule can be divided in two parts, one rigid block, the cholesterol moiety, and a more flexible part; a structural feature shown to be effective for the formation of liquid crystalline phases.4 The cholesteryl compounds 1 – 6 were examined for their mesogenic behaviour. The compounds were heated to their respective melting temperature and the process was monitored by optical microscopy. However, no phases other than the solid to isotropic liquid phase transition were observed. Also, by DSC studies no other phases than the normal melting phase transition were observed, which could be obscured for optical microscopy. It is unclear why no liquid crystal mesophase is observed for these cyclohexyl-based cholesterol gelators. A possible reason may be that all the cholesterol compounds which display liquid-crystalline behaviour are linear and can align along a common axis to form a cholesteric mesophase. The cyclohexyl-based compounds in the present study have a bend in the molecular structure, which could prevent proper alignment for the formation of a liquid crystalline phase. The only exception to this are compounds 4 and 6, as their molecular structures allow for the adoption of a linear arrangement. However, these compounds do not exhibit an additional cholesteric mesophase either. A possible explanation can be that for 4 the cyclohexane ring is still too rigid and for 6 the flexibility in the ethylene spacer is too large. 6.5 Gel Fiber Properties 6.5.1 Gel Fiber Morphology Transmission Electron Microscopy (TEM) in conjunction with Scanning Electron Microscopy (SEM) was employed to study the morphology of the gels and viscous solutions formed by the cholesterol-based organogelators. This technique allows for fast determination of the shape and size of the gel fibers formed. It was necessary to coat the gels formed by compounds 1, 2 and 4 with a thin layer of Pd to enhance the contrast in the sample. The morphology of gel fibers formed by 1 in n-dibutyl ether, 2 in 1-propanol, and by 4 in n-butyl acetate were studied in this way. The morphology of the gel formed by 5 in DMSO and the viscous solution formed by 6 in decalin were determined by SEM, and since the material provided good contrast already, no staining was necessary. These can compounds can be polarised to and enough secondary electrons can be produced to provide a stable and reliable image. 157 Chapter 6 Figure 6-3. TEM pictures of gels of cholesterol based organogelator 1 in n-dibutyl ether, 2 in 1-propanol, 4 in n-butyl acetate and SEM pictures of the gel of 5 in DMSO and the viscous solution formed by 6 in decalin. Figure 6-3 shows that the gels of the different gelators exhibit various morphologies. The gels formed by 1, 2, and 4 consist of small fibers. These fibers bundle and split occasionally to form a 3D network, which manifests itself as the gel at macroscopic level. The width of the fibers formed by these compounds is between 55 to 115 nm. Especially for 1 and 4 the width of the fibers is homogenous. In the gel of 2, crystalline material was found. This indicates that this compound not only forms gel fibers, as other phases beside the gel phase can be formed as well. The gel formed by 5 consists of interdigitated long needle-shaped crystals. The viscous solution formed by 6 consists of platelets which grow in a rosette-like structure. The morphologies of the gel fibers of the different compounds showed that only the gelators which posses a rigid central core unit and two cholesterol units form small fibers. The compounds with only one cholesterol unit, 5, or two cholesterol units linked via a flexible spacer form much larger aggregates. This indicates that a rigid scaffold is necessary to introduce anisotropic intermolecular interactions sufficient to form small fibers. 158 Novel Cholesterol-based Organogelators 6.5.2 Phase Diagrams The thermostability of the gels studied by TEM and SEM was assessed with dropping ball experiments.29 In a dropping ball measurement a small steel ball is placed on top of the gel and the gels are heated slowly while the position of the ball is monitored. At a certain temperature the gels are no longer able to withstand the weight of the ball and the ball drops to the bottom of the vial. The temperature, at which the ball reaches the bottom of the vial, is considered the gel-sol phase transition temperature, Tgs, of the gel. If the Tgs’s of a compound are measured over a concentration range, its phase diagram can be constructed. In addition, the presence of different phases can be observed when, upon heating, the gel crystallises. The phase diagrams for 1 in ndibutyl ether, 2 in 1-propanol, 4 in n-butyl acetate and 5 in DMSO were constructed in this manner (Figure 6-4). 1 n-dibutylether 120 110 Temperature (°C) 100 Temperature (°C) 2 1-propanol 120 110 Sol 90 Crystal 80 70 60 50 40 0.01 0.02 0.03 0.04 Sol 90 80 Crystal 70 60 50 Gel 40 Gel 30 30 0.00 100 0.05 0.06 0.00 0.07 0.01 120 4 n-butylacetate 120 Sol 100 0.04 0.05 0.06 0.07 0.06 0.07 5 DMSO 100 90 Crystal 80 70 60 50 40 Sol 90 80 Crystal 70 60 50 Gel 40 Gel 30 0.00 0.03 110 Temperature (°C) Temperature (°C) 110 0.02 Concentation (molar) Concentration (molar) 30 0.01 0.02 0.03 0.04 0.05 Concentration (molar) 0.06 0.07 0.00 0.01 0.02 0.03 0.04 0.05 Concentration (molar) Figure 6-4. Phase diagram for 1 in n-dibutyl ether, 2 in 1-propanol, 4 in n-butyl acetate and 5 in DMSO. From the phase diagrams displayed in Figure 6-4 it can be seen that in all the phase diagrams a gel state, a crystalline state and a sol state is present. The maximum temperature for the gel phase to be stable is defined by the temperature where the ball has reached the bottom of the vial. At this temperature a crystalline phase, formed in the gel phase, can still be stable and the temperature at which the crystals disappear is 159 Chapter 6 considered the crystalline – sol transition temperature. The presence of the crystalline state is easily recognised by a change in the optical appearance of the gel as the turbidity increases rapidly upon formation of the crystals. For 1 the crystalline phase is absent at concentrations lower as 30 mM. The crystalline phase forms at temperatures above 75 °C and is stable over a ca. 20 °C temperature window. For gels formed by 2, 4 and 5 the gel phase is unstable even at room temperature and the gel and crystalline phase are present simultaneously. Upon heating the gel phase converts fully to the crystalline phase, which will be dissolved ultimately. Because the crystallisation process for gels formed by 2, 4 and 5 takes place in the gel phase, the Tgs temperatures cannot be determined reliably for these compounds. 6.5.3 DSC on Gels In order to study the properties of the gels formed by the cholesterol-based organogelators and to determine the nature of the crystallisation process, differential scanning calorimetry (DSC) was used. This technique allows the gel-sol phase transition to be followed and the enthalpy involved in this transition to be determined. The DSC traces of gels of 1 in n-dibutyl ether, 2 in 1-propanol, 4 in n-butyl acetate and 5 in DMSO are displayed in Figure 6-5. 8 Heatflow (mW) 5 6 4 4 2 2 1 0 40 60 80 100 120 140 Temperature (°C) Figure 6-5. DSC traces for gels formed by cholesterol based organogelator 1 in ndibutyl ether, 2 in 1-propanol, 4 in n-butyl acetate and 5 in DMSO. Figure 6-5 shows that in the DSC traces of the gels of the cholesterol based gelators 1, 2, 4, and 5 different phase transitions are present. For example, in the DSC trace of 1, a sharp phase transition at 89 °C is present op top of a broad phase transition from 80 °C – 130 °C. The sharp transition is assigned to a solid-solid phase transition, as 160 Novel Cholesterol-based Organogelators observed for the bisurea gelators (Chapter 4) and relates to a structural reordering in the material. The phase transition at 89 °C is identified by optical microscopy as due to crystallisation occurring in the gel phase. It has been shown already for bisurea and other gelators that crystallisation can occur as the gel phase is not the thermodynamically most stable phase.30 The broad phase transition is related to the dissolution process of the crystal phase. In the DSC traces of gels of 2, 4, and 5 only a broad phase transition is present. Most likely, as in these compounds the crystal phase together with the gel phase is already present at room temperature, the gel-crystal phase transition is not observed and only the broad crystal-sol phase transition is observed in the DSC traces. Due to the presence of additional phases during the gelsol phase transition, it was not possible to determine the melting enthalpy of the gels formed by the compounds using this technique, as crystal and gel phases occur simultaneously. From the DSC and dropping ball experiments it is clear that most of the compounds form only very weak gels, with the exception of 1, whose gels are stable at room temperature. This property might be related to the fact that 1 does not exhibit a crystallisation process at room temperature in contrast to the other gelators. 6.6 Packing of the Compounds 6.6.1 Single Crystal X-ray Spectroscopy Due to the tendency of these cholesterol-based gelators to crystallise, for three of the compounds: 1, 5 and 6, crystals suitable for single crystal X-ray diffraction could be grown. The packing of the compounds in the crystal, in conjunction with FTIR spectroscopy, will give insight in the intermolecular interactions required for gelation. Compounds 2 – 4 also form crystals in some solvents (vide supra), albeit as needles which are too thin to diffract sufficiently. Crystals of 1 were grown from nhexadecane, in which it displays gelation behaviour also. The crystals of 5 were grown from 1-propanol and for 6 the crystals were grown from 1-octanol. In both solvents the compounds do not display gelation behaviour. The packing of the compounds in the crystal will be discussed in conjunction with their intermolecular interactions. (vide infra) 161 Chapter 6 Figure 6-6. Pluto drawing of the dimer formed by 1 in the crystalline state, displayed perpendicular to both cyclohexane rings. Crystals of 1 were grown via a crystallisation process. A gel of 1 was formed in nhexadecane. After heating to 90 °C and prolonged waiting times (6 h), monoclinic crystals began to form in the gel, which, at room temperature, continued to grow. By single crystal X-ray spectroscopy the unit cell of the crystal was obtained (P21 symmetry). Figure 6-6 displays the asymmetric unit cell, which consist of two crystallographically independent molecules linked to form dimers by two intermolecular hydrogen bonds between one of the carbamate groups in each molecule in the dimer. The other carbamate group in the molecule is not hydrogen bonded as no intermolecular hydrogen bond donor is positioned within a distance of 3 Å. Hence, an infinite stack of hydrogen bonds is not present. Most probably, the size of the cholesterol group is too large to allow sufficient space for the formation of hydrogen bonds between each of the carbamate groups. Within each molecule the two cholesterol moieties are bent by 135° with respect to each other. Between two molecules in the dimer the cholesterol moieties make an angle of 45° to each other to adopt a conformation favourable for the methyl groups at the C114, C123 and C142, C151 position of the steroid rings (for atom numbering experimental section).31,32 The alkyl tails on both of the steroid ring are in an all-trans configuration with some thermal disorder, especially in the end part of the tail. This behaviour is not unusual for cholesteric compounds.35 The all-trans configuration of 162 Novel Cholesterol-based Organogelators the alkyl tail part is found in the crystal structures of almost all cholesterol derivatives.34 Figure 6-7. Unit cell of 1 displayed as stick model, visualised in the 0 1 0 direction. The unit cell consist of two dimers, which are aligned next to each other with one dimer rotated 180° along the (0 1 0) axis, having P21 symmetry (Figure 6-7). In this orientation the molecules form a herring-bone motif in the crystal. The distance of the intermolecular hydrogen bond between the carbamate groups was for N11--H101--O21: 2.951(8) Å with an angle of 151(3)° and for N22--H202--O13: 3.015(7) Å with an angle of 141(3)° (experimental section). The length of the hydrogen bonds between the carbamate groups is longer than for the hydrogen bonds between the amide groups for bisamide gelators, 2.85 Å compared to 2.95 Å (Chapter 3). The bond lengths of the hydrogen bonds formed by the carbamate groups indicate that they are weaker compared to those formed by amide groups.33 163 Chapter 6 Figure 6-8. Pluto drawing of 5 in the unit cell. This view shows that there are no intermolecular hydrogen bonds present between the carbamate groups. Triclinic crystals of 5 were grown from a concentrated 1-propanol solution. From single crystal X-ray diffraction it was found that the unit cell of 5 consists of two crystallographically independent molecules with P1 symmetry. In a unit cell with this symmetry the molecules are packed in a lateral fashion with one of the molecules rotated by 180° along the (1 0 0) axis. The cholesterol groups of the two molecules in the unit cell are positioned perpendicular to each other, to adopt a favourable conformation for the two peripheral methyl groups (C114 and C123) of the steroid ring (for atom numbering see experimental section). This position does not allow formation of intermolecular hydrogen bonds between the carbamate groups as the distance between neighbouring carbonyl groups is too large, 10.29 Å, the a-axis length of the unit cell. Most likely, the stabilisation provided by the van der Waals interactions by the steroid moiety is larger than the stabilisation provided by the formation of hydrogen bonds between the carbamate groups. Both of the carbamate groups point in the same direction, leading to a polarisation in the unit cell (Paragraph 6.6.2). Also in this compound, the alkyl tails at the steroid group in both of the molecules adopt an all-trans configuration,34 with some thermal disorder in the packing of the alkyl tail.35 164 Novel Cholesterol-based Organogelators Figure 6-9. Pluton drawing of 3 molecules of 6 linked via intermolecular hydrogen bonds, in which both of the carbamate group adopt a cis-conformation. Platelet-shaped crystals of 6 were grown in 1-octanol at a concentration of 0.02 mM. The concentration of the sample influenced the shape of the crystals to a great extent, as increasing or decreasing the concentration lead to amorphous precipitation or the formation of small needles, respectively. By single crystal X-ray diffraction it was found that 6 forms crystals of P21 symmetry. In the unit cell four molecules of 6 are present, in which the asymmetric unit consists of two molecules of 6 linked by an infinite one-dimensional intermolecular hydrogen bond array along the (0 1 0) axis (Figure 6-9). The carbamate groups in 6 all are arranged in a cis-conformation to form hydrogen bonds, leading to four hydrogen bonds present in the asymmetric unit cell. It is known that the carbamate group can adopt either a cis- or trans-conformation, however, the cis-conformation for the carbamate group in a hydrogen bond arrangement is not common.36 Each hydrogen bond has a different length: N11--H102--O23: 2.945(8) Å with an angle of 175(3)°, N12--H102--O102: 2.843(7) Å with an angle of 175(3)°, N21--H201--O13 2.893(8) Å with an angle of 155(3)°, and N22--H202--O11: 2.929(8) Å with an angle of 177(3)° (for atom numbering see experimental section). The distance and angles of these hydrogen bonds indicate that these hydrogen bonds are stronger compared to that found for 1. The two cholesterol units lay flat with the peripheral methyl groups (C119, C110 and C138, C147) on both steroid groups pointing in opposite direction. Also, the alkyl tail on the steroid groups for 6 adopts in the crystalline state an all-trans conformation. In the unit cell, the ethylene spacer in compound 6 has two angles of 90° leading to a S-turn in the spacer. This conformation leads to the intermolecular interactions 165 Chapter 6 provided by the cholesterol unit and the carbamate groups pointing in different directions leading to an overall low anisotropy. This arrangement favours growth in two dimensions instead of one dimension necessary for gel fiber growth. 6.6.2 FTIR Spectroscopy The remarkable observation that the diastereoisomers 1 and 2 form gels in solvents of opposite polarity is reasoned to be related to a difference in packing of the compounds in the gel fiber. Therefore, FTIR spectroscopy was employed to study the presence and contribution of hydrogen bonding interactions between the carbamate groups in the gels formed in the different solvents. This technique allows the presence of hydrogen bonds to be determined through the N-H stretch and amide I absorption bands, which display distinct shifts upon hydrogen bond formation.37 Upon hydrogen bonding the N-H stretch absorption band shifts form 3430 to 3350-3250 cm-1 and the amide I absorption band shifts from 1710 to 1700-1680 cm-1 depending on the strength of the hydrogen bond. Figure 6-10 displays the FTIR spectra of 1 in different states with characteristic shifts in the spectra upon hydrogen bond formation and Figure 6-11 displays the FTIR spectra of 2 in different states. This will be followed by an overview of the FTIR spectra of the other compounds in different states. 3600 3550 3500 3450 3400 3350 Wavenumber (cm-1) 3300 3250 3200 3600 3550 3500 3450 3400 3350 3300 3250 3200 Wavenumber (cm-1) Figure 6-10. Normalised FTIR spectra of 1 for the N-H stretch absorption band and amide I absorption band in different states. Solid line: 1 dissolved in CH2Cl2 (1 mg mL-1); dashed: dried gel of 1 in n-dibutyl ether; dotted: crystalline 1. The FTIR spectrum of 1 in CH2Cl2 solution (solid line) displayed in Figure 6-10 shows only one absorption band for the N-H stretch vibration at 3429 cm-1. The amide I absorption band appears at 1708 cm-1. Both vibrations are identified as vibrations from non-hydrogen bonded carbamate groups and indicate that no hydrogen bonding interactions are present in solution.11 In the FTIR spectrum of 1 in the crystalline state (dotted line) two N-H stretch absorption bands are found, at 3425 cm-1 and at 3368 cm-1. The crystal structure of 1 showed already that only half of the carbamate groups are hydrogen bonded (vide supra), therefore, the absorption band at 3425 cm-1 is identified as the N-H stretch 166 Novel Cholesterol-based Organogelators absorption band corresponding to a non-hydrogen bonded carbamate group. The second absorption band, at 3368 cm-1, corresponds to a hydrogen bonded carbamate group.37 The amide I absorption band of 1 in the crystalline state is broad, having a maximum at 1703 cm-1 together with a shoulder at 1710 cm-1. The position of the maximum at 1703 cm-1 confirms the presence of hydrogen bonding interactions in the crystalline state between carbamate groups. However, the hydrogen bonds are weak as the shift to lower wavenumbers is small (5 cm-1). The shoulder at 1710 cm-1 corresponds to the amide I absorption band of non-hydrogen bonded carbamate groups. In the FTIR spectrum of 1 in the gel state (dashed line) also two N-H stretch absorption bands are found, at 3454 cm-1 and 3348 cm-1. The first vibration corresponds to non-hydrogen bonded carbamate groups and the second vibration corresponds to hydrogen bonded carbamate groups. Also, two amide I absorption bands are present, at 1715 cm-1 and 1679 cm-1 The first vibration corresponds to a non-hydrogen-bonded amide I vibration and the second vibration corresponds to a hydrogen-bonded amide I vibration. The position of the hydrogen bonded carbonyl band at 1679 cm-1 indicates that the hydrogen bonding interactions in the gel state are stronger compared to those in the crystalline state, as they appear at lower wavenumbers.33 3600 3550 3500 3450 3400 3350 Wavenumber (cm-1) 3300 3250 3200 1750 1700 1650 1600 1550 1500 1450 Wavenumber (cm-1) Figure 6-11. Normalised FTIR spectra of 2 for the N-H stretch absorption band and amide I and II absorption bands in different states. Solid line: 2 dissolved in CH2Cl2 (1 mg mL-1); dashed: dried gel of 2 in 1-propanol; dotted: crystalline 2. The FTIR spectrum of 2 in CH2Cl2 solution (solid line) shows that the compound is not hydrogen bonded as the N-H stretch absorption band appears at 3429 cm-1 and the amide I absorption band at 1708 cm-1. The position of these bands is comparable with those of 1 in CH2Cl2 solution. 167 Chapter 6 The FTIR spectrum of 2 in the gel state (dotted line) shows two N-H stretch absorption bands, at 3385 cm-1 (large) and 3320 cm-1 (small). These bands correspond to N-H stretching absorptions in different chemical environments where one carbamate group is hydrogen bonded and the other non-hydrogen bonded. Also, two amide I absorption bands are present, at 1718 cm-1 (small) and 1702 cm-1 (large). This indicate that in the gel state the majority of 2 is hydrogen bonded and only a small fraction is non-hydrogen bonded. As is shown in paragraph 6.5.2 (vide supra) in the gel state a small amount of 2 is present in the crystalline state, which might explain the presence of non-hydrogen bonded 2. The FTIR spectrum of 2 in the crystalline state (dotted line) 2 shows that the N-H stretch absorption band is present at 3395 cm-1 and a strong amide I absorption band is present at 1719 cm-1. Compared to the FTIR spectrum of 2 in solution, the position of the amide I absorption band is shifted to higher wavenumbers instead to lower wavenumbers as is expected upon hydrogen bonding. A shift in position of the amide I band is caused by a change in the electron density of the carbonyl group, thereby changing the force constant of the vibrational mode.38 In solution the carbonyl groups are solvated by polar CH2Cl2 molecules, ET (30) 41.1,39 and in the solid state it is surrounded by less polar groups, e.g. steroid groups which leads to a shift to higher wavenumbers. In the FTIR spectrum of the 2 in the crystalline state there is a small shoulder present at 1702 cm-1, which is attributed to a small amount of hydrogen bonding between the carbamate groups. This is not unexpected since the crystals are grown via the gel state, in which the compounds are hydrogen bonded. Most likely, a small fraction of 2 is still in the gel state. FTIR spectra of the cholesterol compounds 1 – 6 in different physical states were recorded and the results are listed in Table 6-2. The spectra of the compounds were recorded in solution, in the gel state and, if possible, in the crystalline state. The gel state is recorded as a dried gel, to avoid overlap of solvent absorptions with the absorption bands of interest. 168 Novel Cholesterol-based Organogelators Table 6-2. Selected FTIR bands (ν, cm-1) for cholesterol based compounds 1 – 6 in different states. Compound 1 1 1 2 2 2 3 3 4 4 5 5 5 6 6 Sample CH2Cl2 solution[b] dried n-dibutyl ether gel[c] crystal[c] CH2Cl2 solution[b] dried 1-propanol gel[c] crystal[c] N-H (stretch) 3429 Amide I 1708 Amide II 1518 (broad) 3454 and 3348 1715 and 1679 1532 and 1507 3425 and 3368 1710 (shoulder) and 1703 1530 and 1507 3429 1708 1516 (broad) 3385 and 3320 1718 (small) and 1702 1719 and 1702 (small) 1520 and 1501 3395 1520 and 1501 CH2Cl2 solution[b] solid[c] 3436 1710 1510 (broad) 3335 1720 (broad) 1506 CH2Cl2 solution[b] dried decalin gel[b] 3435 1710 - 3412 and 3342 1718 and 1692 (broad) 1524 (broad) 3436 1710 1506 (broad) 3391 and 3302 1711 and 1680 1543 3263 1709 1504 3447 1710 1511 3386, 3285, 3221, 3146 1732 and 1709 1549 CH2Cl2 solution[b] dried DMSO gel[c] crystal[c] CH2Cl2 solution[b] crystal[c] [a] Uncertainty (± 2 cm-1). [b] Recorded in a liquid cell with CaF2-windows. [c] Recorded as intimate mixture with KBr. 169 Chapter 6 Table 6-2 shows that compounds 1 – 6 all have in solution a N-H stretch absorption band between 3447 cm-1 and 3429 cm-1 and an amide I absorption band between 1708 cm-1 and 1710 cm-1. By comparison of these absorptions with those of 1 in the crystalline state, 3371 cm-1 for the N-H stretch, and 1703 cm-1 for the amide I absorption, it is clear that in solution the different cholesterol-based compounds do not display hydrogen bonding interactions. Compound 3 does not form a gel in any of the solvents examined and it was reasoned that it was due to the formation of intramolecular hydrogen bonds (vide supra). However, by comparing the position of the amide I band in solution to its position in the solid state, it is clear that in the solid state no hydrogen bonding interactions are present. The FTIR spectra of compounds 4 and 5 in the gel state show two absorptions for the N-H stretch and amide I vibration bands also. As was observed for 2, the polymorphism in these samples is most likely caused by a crystallisation process. In the gel state the compounds are hydrogen bonded and in the crystalline state no hydrogen bonding interaction is present, as was shown by single crystal X-ray diffraction on 5. 6.6.3 Gel Formation To study the formation of the gel in more detail, a gelling solution of 1 in n-dibutyl ether was followed in time by FTIR spectroscopy (Figure 6-12). The formation of a gel of 2 in alcohols is more difficult to follow as the solvent blocks absorptions in the regions of interest. Also, the observation that the gels formed by compounds 2, 4 and 5 are metastable due to crystallisation, makes these systems less suitable for study by FTIR spectroscopy. time time 3600 3550 3500 3450 3400 3350 Wavenumber (cm-1) 3300 3250 3200 1750 1700 1650 1600 1550 1500 1450 Wavenumber (cm-1) Figure 6-12. Normalised FTIR spectra of 1 in n-dibutyl ether (25 mg mL-1) following the gel formation after 10, 20 and 80 min (solid) and of 1 as an aerogel (dashed). Visual inspection of the gelling solution of 1 in n-dibutyl ether showed that no gel has formed in the first 10 min, as the solution was still liquid. In the FTIR spectrum peaks related to hydrogen bonding interaction between the carbamate groups are absent also, 170 Novel Cholesterol-based Organogelators as no N-H stretch and amide I vibrations, at 3348 cm-1 and 1679 cm-1 respectively, are present (Figure 6-12). After 20 min visual inspection of the gelling solution showed a gel has formed and in the FTIR spectrum absorption bands related to hydrogen bonding interactions between the carbamate group at 3348 cm-1 (N-H stretch) and 1679 cm-1 (amide I) are present. After 80 min the aggregation process has reached its equilibrium and the intensity of the absorption bands in the FTIR spectrum involved in aggregation remains constant. In the FTIR spectrum there are still two absorption bands corresponding to dissolved 1, at 3500 cm-1 (N-H stretch) and 1730 cm-1 (amide I). This can be accounted for as the aggregates formed are in equilibrium with dissolved 1, as the cgc of 1 in this solvent is 10 mg mL-1 (Table 6-1). Upon removal of the solvent, the equilibrium is pushed towards aggregation and in the FTIR spectrum of the aerogel only two absorption bands for the N-H stretch vibration and two absorption bands for the amide I vibration are observed. Both peaks of the amide I vibration have approximately equal intensity which is a strong indication that in the gel state the two carbamate groups in the molecule adopt a different conformation, as has been found in the crystalline phase for 1 as well (vide supra). One of the carbamate groups is strongly hydrogen bonded and the other carbamate group is not involved in hydrogen bonding. 6.7 Conclusions Based on design rules obtained from the work on bisamide and bisurea gelators, new cholesterol-based compounds were developed of which some showed gelation behaviour. The results show that it is possible to design gelators with a minimum on functional information. However, these gelators obtained are not very effective in solvent scope, as the polarity range in which they are able to form a gel is limited. Nevertheless, they provide surprising results on the different intermolecular interactions necessary to gelate solvents as the diasteromers 1 and 2 gelate solvents of opposite polarity. Unfortunately, many of the gels formed by these compounds are not stable and crystallise in time as was shown by optical microscopy and dropping ball experiments. Due to their tendency to crystallise it was possible to use single-crystal X-ray diffraction to determine the organisation in the crystal for compounds 1, 5 and 6. In conjunction with FTIR spectroscopy on gels it was shown that the compounds have different aggregation modes in the gel state compared to the crystalline state. From these results, for diastereoisomers 1 and 2 the difference in organisation of the molecules upon aggregation in different states is summarised in Figure 6-13, in which the organisation of the hydrogen bonds plays an important role. 171 Chapter 6 Polar Solvents Apolar Solvents 1 (R,R) gelation 2 (S,S) gelation Solution Gel: 1 carbamate group strongly H-bonded 1 carbamate group free crystallisation 90 °C Crystal: 1 carbamate group weakly H-bonded 1 carbamate group free Gel: both carbamate groups H-bonded RT recrystallisation Crystal: Both carbamate groups free Figure 6-13. Proposed states for the diastereoisomers 1 and 2 in different phases. The formation of a gel is governed by kinetics and does not provide the thermodynamically most stable state.40 By FTIR spectroscopy it was shown that 1 forms in apolar solvents a gel in which only one of the carbamate groups is hydrogen bonded strongly and the other carbamate group is not involved in hydrogen bonding. Compound 2 forms a gel in polar solvents in which both carbamate groups are only weakly hydrogen bonded. The gel formed by 2 is not stable in time and a crystallisation process occurs. By FTIR spectroscopy it is found that in the crystalline state both the carbamate groups are not involved in hydrogen bonding. Most probably, the solvent disrupts the hydrogen bonds which leads to the formation of a more stable crystalline state in which the stability is provided by van der Waals interactions between the cholesterol moieties. The gel formed by 1 is stable at room temperature and the crystallisation process takes place only at elevated temperatures (90 °C). After crystallisation of 1, the hydrogen bonding interaction between the carbamate groups has become weaker. Single crystal X-ray analysis shows that 1 in the crystalline state forms dimers in which only one carbamate group of the molecule is involved in hydrogen bonding. From the results on the different contribution of the intermolecular interactions involved in gelation for the different cholesterol gelators, it is shown that it is not yet facile to design new gelators. Nevertheless, with the knowledge gained in recent years from studies on organogelators, designing gelator systems has become more practical. It is apparent, however, that to make a successful gelator all the interaction types that lead to gelation must be correctly balanced with each other. A small mismatch in the intermolecular interactions may lead to a good gelator in one solvent, but leads to crystallisation in another, as was shown for these cholesterol based compounds. 172 Novel Cholesterol-based Organogelators Acknowledgements: Dr. Davide Pantarotto is acknowledged for the SEM pictures of the gels formed by the cholesterol-based gelators. 6.8 Experimental Section General Information For general remarks concerning experimental details, see experimental section of chapter 3. The single crystal X-ray spectrum of 1, 5 and 6 were resolved by drs. Auke Meetsma at the University of Groningen. Trans-(R,R)-1,2-bis(3-cholesteryloxycarbonylamine)cyclohexane (1) A solution of (0.5 g, 4.4 mmol) (1R,2R)-(−)-1,2diaminocyclohexane in 25 mL CH2Cl2 together with (1 g, 10 mmol) triethylamine was added dropwise to a stirred solution of (4.0 g, 8.9 mmol) L-cholesterol chloroformate dissolved in 100 mL of CH2Cl2. After addition, the solution was heated at reflux temperatures for 8 h. After cooling the organic layer was washed twice with a 100 mL 10 % aqueous HCl solution (pH 1), once with 100 mL of brine and was dried finally over Na2SO4. A white amorphous solid was obtained after removing the solvent in vacuo. Column chromatography (CH2Cl2/ MeOH; 100/ 5; Rf = 0.65) on silica gel yielded pure 1 as a white solid (2.9 g, 3.0 mmol, 69.0 %). mp 198 - 200 °C. 1 H-NMR (300 MHz, CDCl3): δ= 0.65 (s, 6H, CH3), 0.80 - 2.00 (m, 84H, Chol), 2.10 – 2.35 (m, 4H, Chol), 3.13 (s, 2H, CHN), 4.42 (mp, 2H, CHO), 4.86 (d, 2H, J= 6.6 Hz, NH), 5.33 (s, 2H, C=CH) ppm. 13C-NMR (75 MHz, CDCl3): δ= 11.8 (p), 18.7 (p), 19.2 (p), 21.0 (s), 22.5 (p), 22.8 (p), 23.8 (s), 31.8 (s), 32.8 (s), 35.8 (t), 36.2 (s), 36.5 (q), 37.0 (s), 38.5 (s), 39.5 (s), 39.7 (s), 42.3 (s), 50.0 (t), 55.4 (t), 56.1 (t), 56.6 (t), 74.4 (t), 122.4 (t), 139.8 (q), 156.6 (q) ppm. FTIR (KBr): ν = 3392, 2939, 1702 cm-1. MS(EI) calcd. for C62H102N2O4: 938.784, found: 939 (M+). Anal. Calcd. for C62H102N2O4: C, 79.25%; H, 10.95%; N, 2.98%; found: C, 79.05%; H, 11.11%; N, 3.02%. 173 Chapter 6 Trans-(S,S)-1,2-bis(3-cholesteryloxycarbonylamine)cyclohexane (2) Compound 2 was synthesised according to the synthesis of 1 starting from (4.3 g, 9.5 mmol) L-cholesterol chloroformate, (0.54 g, 4.7 mmol) (1S, 2S)-(+)-1,2diaminocyclohexane and (1.5 g, 15 mmol) triethylamine. After washing with 200 mL of pet. ether 40 - 60 pure 2 was obtained as a white amorphous solid (3.5 g, 3.7 mmol, 78.0 %). mp. 220 – 226 °C. 1H-NMR (300 MHz, CDCl3): δ= 0.66 (s, 6H, CH3), 0.80 2.00 (m, 84H, Chol), 2.15 – 2.35 (m, 4H, Chol), 3.30 (s, 2H, CHN), 4.42 (mp, 2H, CHO), 4.86 (d, 2H, J= 5.7 Hz, NH), 5.35 (s, 2H, C=CH) ppm. 13C-NMR (75 MHz, CDCl3): δ= 11.9 (p), 18.7 (p), 19.3 (p), 21.0 (s), 22.5 (p), 22.8 (p), 23.8 (s), 24.3 (s), 24.8 (s), 28.0 (s), 28.1 (s), 28.2 (s), 31.8 (s), 32.9 (s), 35.8 (t), 36.2 (s), 36.5 (q), 37.0 (s), 38.5 (s), 39.5 (s), 39.7 (s), 42.3 (s), 50.0 (t), 55.3 (t), 56.1 (t), 56.7 (t), 74.4 (t), 122.4 (t), 139.8 (q), 156.6 (q) ppm. FTIR (KBr): ν = 3392, 2939, 1720, 1702 cm-1. MS(EI) calcd. for C62H102N2O4: 938.784, found: 938 (M+). Anal. Calcd. for C62H102N2O4: C, 79.25%; H, 10.95%; N, 2.98%; found: C, 79.14%; H, 11.02%; N, 2.87%. Cis-(R,S)-1,2-bis(3-cholesteryloxycarbonylamine)cyclohexane (3) Compound 3 was synthesised according to the synthesis of 1 starting from (4.5 g, 10.0 mmol) L-cholesterol chloroformate, (0.55 g, 4.8 mmol) (1R, 2S)-1,2 diaminocyclohexane and (3.0 g, 30 mmol) triethylamine. The crude product was purified by column chromatography over silica gel (CH2Cl2/ MeOH; 100/ 1; Rf = 0.57) and after removing of the solvent pure 3 was obtained as a white amorphous powder (2.4 g, 2.5 mmol, 53.0 %). mp. 162 – 165 °C. 1H-NMR (300 MHz, CDCl3): δ= 0.65 (s, 6H, CH3), 0.80 - 2.00 (m, 84H, Chol), 2.25 – 2.35 (m, 4H, Chol), 3.81 (s, 2H, CHN), 4.47 (mp, 2H, CHO), 4.97 (d, 2H, J= 5.3 Hz, NH), 5.35 (s, 2H, C=CH) ppm. 13C-NMR (75 MHz, CDCl3): δ= 11.8 (p), 18.7 (p), 19.3 (p), 21.0 (s), 22.5 (p), 22.8 (p), 23.8 (s), 24.3 (s), 28.0 (s), 28.2 (s), 29.1 (s), 31.9 (s), 35.8 (t), 36.2 (t), 36.5 (q), 37.0 (s), 38.5 (s), 39.5 (s), 39.7 (s), 42.3 (s), 50.0 (t), 50.8 (s), 56.1 (t), 56.7 (t), 74.5 (t), 122.5 (t), 139.8 (q), 155.9 (q) ppm. FTIR (KBr): ν = 3341, 2933, 1720 cm-1. MS(EI) calcd. for C62H102N2O4: 938.784, found: 939 (M+). Anal. Calcd. for C62H102N2O4: C, 79.25%; H, 10.95%; N, 2.98%; found: C, 79.07%; H, 10.91%; N, 2.99%. 1,4-Bis(3-cholesteryloxycarbonylamine)cyclohexane (4) Compound 4 was synthesised according to the synthesis of 1 starting from (4.8 g, 10.6 mmol) L-cholesterol chloroformate, (0.6 g, 5.3 mmol) trans-1,4-diaminocyclohexane and (2.5 g, 25 mmol) triethylamine. The crude product was purified by washing with 200 mL of ether and pure 4 was obtained as a white amorphous solid (3.2 g, 3.4 mmol, 65.1 %). mp 246 – 249 °C. 1H-NMR (300 MHz, CDCl3): δ= 0.65 (s, 6H, CH3), 0.80 – 2.00 (m, 84H, Chol), 2.15 – 2.35 (m, 4H, Chol), 3.31 (s, 2H, CHN), 3.44 (bs, 2H, NH), 4.46 (mp, 2H, CHO), 5.34 (s, 2H, C=CH) ppm. 13C-NMR (75 MHz, 174 Novel Cholesterol-based Organogelators CDCl3): δ= 11.8 (p), 18.7 (p), 19.3 (p), 21.0 (s), 22.5 (p), 22.8 (p), 23.8 (s), 24.3 (s), 24.8 (s), 25.5 (s), 28.0 (s), 28.2 (s), 31.8 (s), 33.5 (s), 35.8 (t), 36.1 (s), 36.5 (q), 37.0 (s), 38.6 (s), 39.5 (s), 39.7 (s), 42.3 (s), 49.1 (s), 49.6 (s), 50.0 (t), 56.1 (t), 56.7 (t), 74.0 (t), 122.4 (t), 139.9 (q), 155.3 (q) ppm. FTIR (KBr): ν = 3349, 2937, 1699 cm-1. MS(EI) calcd. for C62H102N2O4: 938.784, found: 939 (M+). Anal. Calcd. for C62H102N2O4: C, 79.25%; H, 10.95%; N, 2.98%; found: C, 79.31%; H, 11.09%; N, 3.05%. 3-Cholesteryloxycarbonylaminecyclohexane (5) Compound 5 was synthesised according to the synthesis of 1 starting from (2.1 g, 4.5 mmol) L-Cholesterol chloroformate, (0.4 g, 4.4 mmol) cyclohexylamine and (1.0 g, 10.0 mmol) triethylamine. The crude product was purified by column chromatography on silica gel (CH2Cl2; Rf = 0.90) and after removing of the solvent pure 5 was obtained as a white amorphous powder (1.8 g, 3.6 mmol, 82.2 %). mp 136 - 141 °C. 1 H-NMR (300 MHz, CDCl3): 0.65 (s, 3H, CH3), 0.80 - 2.00 (m, 46 H, Chol), 2.15 – 2.35 (m, 4H, Chol), 3.46 (s, 1H, CHN), 4.46 (mp, 2H, NH + CHO), 5.34 (s, 1H, C=CH) ppm. 13 C-NMR (75 MHz, CDCl3): δ=11.8 (p), 18.7 (p), 19.3 (p), 21.0 (s), 22.5 (s), 22.8 (p), 23.8 (s), 24.3 (s), 24.8 (s), 25.5 (s), 28.0 (s), 28.2 (s), 31.8 (s), 31.9 (s), 33.5 (s), 35.8 (t), 36.1 (s), 36.5 (q), 37.0 (s), 38.6 (s), 39.5 (s), 39.7 (s), 42.3 (s), 49.6 (s), 50.0 (t), 56.1 (t), 56.6 (t), 74.0 (t), 122.4 (t), 139.9 (q), 155.3 (q) ppm. FTIR (KBr): ν = 3262, 2933, 1706 cm-1. MS(EI) calcd. for C34H57NO2: 511.439, found: 512.5 (M+). Anal. Calcd. for C34H57NO2: C, 79.77%; H, 11.23%; N, 2.74%; found: C, 79.61%; H, 11.42%; N, 2.71%. 1,2-Bis(3-cholesteryloxycarbonylamine)ethylene (6) Compound 6 was synthesised according to the synthesis of 1 starting from (4.49 g, 10.0 mmol) L-cholesterol chloroformate, (0.30 g, 5.0 mmol) ethylenediamine and (1.5 g, 15.0 mmol) triethylamine. The crude product was purified by washing with 200 mL of pet-ether 40 - 60 and after drying pure 6 was obtained as a white amorphous powder (2.2 g, 3.6 mmol, 72.3 %). mp 239 – 241 °C. 1H-NMR (300 MHz, CDCl3): δ= 0.65 (s, 6H, CH3), 0.80 - 1.60 (m, 76H, Chol), 2.15 – 2.35 (m, 4H, Chol), 3.28 (s, 4H, 175 Chapter 6 CH2N), 4.46 (mp, 2H, CHO), 5.00 (bs, 2H, NH), 5.36 (s, 2H, C=CH) ppm. 13C-NMR (75 MHz, CDCl3): δ= 11.8 (p), 18.7 (p), 19.3 (p), 21.0 (s), 22.5 (p), 22.8 (p), 23.8 (s), 24.3 (s), 28.0 (s), 28.1 (s), 28.2 (s), 31.9 (s), 35.8 (t), 36.2 (s), 36.6 (q), 37.0 (s), 38.5 (s), 39.5 (s), 39.7 (s), 41.2 (s), 42.3 (s), 50.0 (t), 56.2 (t), 56.7 (t), 74.6 (t), 122.5 (t), 139.8 (q), 156.5 (q) ppm. FTIR (KBr): ν = 3291, 2944, 1733, 1709 cm-1. MS(EI) calcd. for C58H96N2O4: 885.41, found: 886 (M+). Anal. Calcd. for C58H96N2O4: C, 78.67%; H, 10.94%; N, 3.17%; found (%): C, 78.71%; H, 10.97%; N, 3.21%. Scanning Electron Microscopy The SEM measurements (SEI) were performed on a JEOL JSM 7000F field emission SEM operating at 2.0 kV. Gels of 1, 2, and 4 were deposited on a piece of mica on which 20 nm of gold was sputtered. The samples were dried at ambient temperatures and stained subsequently with 5 nm of Pd. For gels of 5 and 6, a small patch of the gel was deposited on a glass slit covered with a 120 nm Al top-layer on which a 1 nm Cr layer was sputtered. These sample where allowed to dry at ambient temperatures and used without additional staining. Single Crystal X-ray Analysis of 141 Colorless parallelepiped-shaped crystals of 1 were obtained by crystallisation from nhexadecane. Although an X-ray structure determination was thwarted by persistent weak scattering power of the crystals, ultimately there was found a crystal, which was ultimately adjusted fit to the X-ray experiment. A crystal fragment, cut to size to fit in the homogeneous part of the X-ray beam, with dimensions of 0.52 x 0.45 x 0.33 mm was mounted on top of a glass fiber and aligned on a Bruker46 SMART APEX CCD diffractometer (Platform with full three-circle goniometer). The diffractometer was equipped with a 4K CCD detector set 60.0 mm from the crystal. The crystal was cooled to 100 (± 1) K using the Bruker KRYOFLEX low-temperature device. Intensity measurements were performed using graphite monochromated Mo-K α radiation from a sealed ceramic diffraction tube (SIEMENS). Generator settings were 50 KV, 40 mA. SMART was used for preliminary determination of the unit cell constants and data collection control. The intensities of reflections of a hemisphere were collected by a combination of 3 sets of exposures (frames). Each set had a different φ angle for the crystal and each exposure covered a range of 0.3° in ω. A total of 1800 frames were collected with an exposure time of 10.0 seconds per frame. The overall data collection time was 8.0 h. Data integration and global cell refinement was performed with the program SAINT. The final unit cell was obtained from the xyz centroids of 7502 reflections after integration. Intensity data were corrected for Lorentz and polarization effects, scale variation, for decay and absorption: a multi-scan absorption correction was applied, based on the intensities of symmetry-related reflections measured at different angular 176 Novel Cholesterol-based Organogelators settings (SADABS),47 and reduced to Fo2. The program suite SHELXTL was used for space group determination (XPREP).46 The intensity data were corrected for decay and absorption: a multi-scan absorption correction was applied, based on the intensities of symmetry-related reflections measured at different angular settings (SADABS)47 and reduced to Fo2. The unit cell48 was identified as monoclinic. Reduced cell calculations did not indicate any higher metric lattice symmetry.49 Space group, P21, was determined from considerations of the unit cell parameters, statistical analyses of intensity distributions: the E-statistics50 were indicative of a non-centrosymmetric space group. Examination of the final atomic coordinates of the structure did not yield extra crystallographic or metric symmetry elements.51,52 The structure was solved by direct methods using the program SIR2002.42 The positional and anisotropic displacement parameters for the non-hydrogen atoms were refined. Hydrogen atoms were constrained to idealized geometries and allowed to ride on their carrier atoms with an isotropic displacement parameter related to the equivalent displacement parameter of their carrier atoms. A few atoms showed unrealistic displacement parameters when allowed to vary anisotropically, suggesting dynamic disorder (dynamic means that the smeared electron density is due to fluctuations of the atomic positions within each unit cell) as a consequence of the rotational-disorder. This is in line with the weak scattering power of the crystals investigated. Final refinement on F2 carried out by full-matrix least-squares techniques converged at wR (F2) = 0.1787 for 10747 reflections and R(F) = 0.0665 for 6967 reflections with Fo ≥ 4.0 σ (Fo) and 1245 parameters and 1 restraints. The final difference Fourier map was essentially featureless: no significant peaks (0.724 e Å-3) having chemical meaning above the general background were observed. In the absence of suitable anomalous scatters, Friedel equivalents could not be used to determine the absolute structure. Therefore, Friedel equivalents were merged before the final refinement and the known configuration (by synthesis route) of the parent molecule was used to define the enantiomer of the final model. The positional and anisotropic displacement parameters for the non-hydrogen atoms and isotropic displacement parameters for hydrogen atoms were refined on F2 with full-matrix least-squares procedures minimizing the function Q = ∑h[w(│(Fo2) k(Fc2)│)2], where w = 1/[σ2(Fo2) + (aP)2 + bP], P = [max(Fo2,0) + 2Fc2] / 3, F0 and Fc are the observed and calculated structure factor amplitudes, respectively; ultimately the suggested a (= 0.0887) and b (= 1.8448) were used in the final refinement. Neutral atom scattering factors and anomalous dispersion corrections were taken from International Tables for Crystallography.55 All refinement calculations and graphics were performed on a HP XW6200 (Intel XEON 3.2 Ghz), Debian-Linux computer at the University of Groningen with the program packages SHELXL56 (least-square refinements), a locally modified version of the program PLUTO57 (preparation of illustrations) and PLATON58 package (checking the final results for missed symmetry 177 Chapter 6 with the MISSYM option, solvent accessible voids with the SOLV option, calculation of geometric data and the ORTEP60 illustrations). Each asymmetric unit contains two formula units (molecule) with no atom setting at special position. The chiral centers of C11, C16, C21 and C26 have all the Rconfiguration,60 as known from the employed synthesis route. The monoclinic unit cell contains four units of the title compound. A search of the distances yielded intermolecular contacts shorter than the sum of the van der Waals radii59 for the atoms: the moieties are linked by hydrogen bonds.43,44 A dimer is formed by the two molecules comprising the asymmetric unit. No missed symmetry (MISSYM) is detected, however, potential solvent-accessible area (66.6 Å3 per unit cell) was detected by procedures implemented in PLATON.60 Single Crystal X-ray Analysis of 545 Colorless needle-shaped crystals of 5 were obtained by crystallisation from 1propanol via slow evaporation of the solvent. Although an X-ray structure determination was thwarted by persistent weak scattering power of the crystals, ultimately there was found a crystal, which was adjusted fit to the X-ray experiment. A crystal fragment, cut to size to fit in the homogeneous part of the X-ray beam, with dimensions of 0.41 x 0.29 x 0.04 mm was mounted on top of a glass fiber and aligned on a Bruker46 SMART APEX CCD diffractometer (Platform with full three-circle goniometer). The diffractometer was equipped with a 4 K CCD detector set 60.0 mm from the crystal. The crystal was cooled to 100 (± 1) K using the Bruker KRYOFLEX low-temperature device. Intensity measurements were performed using graphite monochromated Mo-K α radiation from a sealed ceramic diffraction tube (SIEMENS). Generator settings were 50 KV, 40 mA. SMART was used for preliminary determination of the unit cell constants and data collection control. The intensities of reflections of a hemisphere were collected by a combination of 3 sets of exposures (frames). Each set had a different φ angle for the crystal and each exposure covered a range of 0.3° in ω. A total of 1800 frames were collected with an exposure time of 45.0 seconds per frame. The overall data collection time was 28.0 h. Data integration and global cell refinement was performed with the program SAINT. The final unit cell was obtained from the xyz centroids of 4320 reflections after integration. Intensity data were corrected for Lorentz and polarization effects, scale variation, for decay and absorption: a multi-scan absorption correction was applied, based on the intensities of symmetry-related reflections measured at different angular settings (SADABS),47 and reduced to Fo2. The program suite SHELXTL was used for space group determination (XPREP).46 The intensity data were corrected for decay and absorption: a multi-scan absorption correction was applied, based on the intensities of symmetry-related reflections measured at different angular settings (SADABS)47 and reduced to Fo2. 178 Novel Cholesterol-based Organogelators The unit cell48 was identified as triclinic. Reduced cell calculations did not indicate any higher metric lattice symmetry.49 Space group, P1, was determined from considerations of the unit cell parameters, statistical analyses of intensity distributions: the E-statistics50 were indicative of a non-centrosymmetric space group. Examination of the final atomic coordinates of the structure did not yield extra crystallographic or metric symmetry elements.51,52 The structure was solved by direct methods using the program SIR-97.53 The hydrogen atoms were generated by geometrical considerations and constrained to idealized geometries and allowed to ride on their carrier atoms with an isotropic displacement parameter related to the equivalent displacement parameter of their carrier atoms. Final refinement on F2 carried out by full-matrix least-squares techniques converged at wR(F2) = 0.2090 for 8775 reflections and R(F) = 0.0709 for 5857 reflections with Fo ≥ 4.0 σ (Fo) and 679 parameters and 1 restraints. The final difference Fourier map was essentially featureless: no significant peaks (0.37 (± 8) e Å-3) having chemical meaning above the general background were observed. The absolute configuration of the structure could not be determined reliably: there are only elements in the structure with very small anomalous effects by the used X-ray wave length and the quality of this batch of crystals is not adequate for this purpose (Flack's54 x-refinement gave an ambiguous result (x = -0.5 (± 2.1))), however by synthesis route the absolute exhibited configurations should be the one as stated in this article. The positional and anisotropic displacement parameters for the non-hydrogen atoms and isotropic displacement parameters for hydrogen atoms were refined on F2 with full-matrix least-squares procedures minimizing the function Q = ∑h[w(│(Fo2) k(Fc2)│)2], where w = 1/[σ2(Fo2) + (aP)2 + bP], P = [max(Fo2,0) + 2Fc2] / 3, F0 and Fc are the observed and calculated structure factor amplitudes, respectively; ultimately the suggested a (= 0.1101) and b (= 0.0) were used in the final refinement. Neutral atom scattering factors and anomalous dispersion corrections were taken from International Tables for Crystallography.55 All refinement calculations and graphics were performed on a HP XW6200 (Intel XEON 3.2 Ghz), Debian-Linux computer at the University of Groningen with the program packages SHELXL56 (least-square refinements), a locally modified version of the program PLUTO57 (preparation of illustrations) and PLATON58 package (checking the final results for missed symmetry with the MISSYM option, solvent accessible voids with the SOLV option, calculation of geometric data and the ORTEP60 illustrations). Each asymmetric unit contains two formula units. The triclinic unit cell contains two discrete molecules of the title compound separated by normal van der Waals distances.59 In both residues the chiral centers with S-configuration60 are: C8, C11, C15, C16 and C22 and with the R-configuration60 are C21, C24 and C27. No classic hydrogen bonds, no missed symmetry (MISSYM), however small voids, 8.2 Å3 per unit cell, were detected by procedures implemented in PLATON.60 179 Chapter 6 Single Crystal X-ray Analysis of 661 Colorless platelet-shaped crystals of 6 were obtained by crystallisation from 1-octanol at a concentration of 15 mg mL-1. A crystal with the dimensions of 0.43 x 0.39 x 0.08 mm was mounted on top of a glass fiber, by using inert-atmosphere handling techniques, and aligned on a Bruker46 SMART APEX CCD diffractometer (Platform with full three-circle goniometer). The diffractometer was equipped with a 4 K CCD detector set 60.0 mm from the crystal. The crystal was cooled to 100 (± 1) K using the Bruker KRYOFLEX low-temperature device. Intensity measurements were performed using graphite monochromated Mo-K α radiation from a sealed ceramic diffraction tube (SIEMENS). Generator settings were 50 KV, 40 mA. SMART was used for preliminary determination of the unit cell constants and data collection control. The intensities of reflections of a hemisphere were collected by a combination of 3 sets of exposures (frames). Each set had a different φ angle for the crystal and each exposure covered a range of 0.3° in ω. A total of 1800 frames were collected with an exposure time of 30.0 seconds per frame. The overall data collection time was 18.0 h. Data integration and global cell refinement was performed with the program SAINT. The final unit cell was obtained from the xyz centroids of 8405 reflections after integration. Intensity data were corrected for Lorentz and polarization effects, scale variation, for decay and absorption: a multi-scan absorption correction was applied, based on the intensities of symmetry-related reflections measured at different angular settings (SADABS),47 and reduced to Fo2. The program suite SHELXTL was used for space group determination (XPREP).46 The intensity data were corrected for decay and absorption: a multi-scan absorption correction was applied, based on the intensities of symmetry-related reflections measured at different angular settings (SADABS)47 and reduced to Fo2. The unit cell48 was identified as monoclinic. Reduced cell calculations did not indicate any higher metric lattice symmetry.49 Space group P21 was determined from considerations of the unit cell parameters, statistical analyses of intensity distributions: the E-statistics50 were indicative of a non-centrosymmetric space group. Examination of the final atomic coordinates of the structure did not yield extra crystallographic or metric symmetry elements.51,52 The structure was solved by direct methods using the program SIR2004.62 The hydrogen atoms were generated by geometrical considerations and constrained to idealized geometries and allowed to ride on their carrier atoms with an isotropic displacement parameter related to the equivalent displacement parameter of their carrier atoms. Final refinement on F2 carried out by full-matrix least-squares techniques converged at wR(F2) = 0.1621 for 10041 reflections and R(F) = 0.0609 for 4217 reflections with Fo ≥ 4.0 σ (Fo) and 1174 parameters and 1 restraints. The final difference Fourier map was essentially featureless: no significant peaks (0.31 (± 6) e Å-3) having chemical meaning above the general background were observed. 180 Novel Cholesterol-based Organogelators In the absence of suitable anomalous scatters, Friedel equivalents could not be used to determine the absolute structure. Therefore, Friedel equivalents were merged before the final refinement and the known configuration of the parent molecule was used to define the enantiomer of the final model. The positional and anisotropic displacement parameters for the non-hydrogen atoms and isotropic displacement parameters for hydrogen atoms were refined on F2 with full-matrix least-squares procedures minimizing the function Q = ∑h[w(│(Fo2) k(Fc2)│)2], where w = 1/[σ2(Fo2) + (aP)2 + bP], P = [max(Fo2,0) + 2Fc2] / 3, F0 and Fc are the observed and calculated structure factor amplitudes, respectively; ultimately the suggested a (= 0.0) and b (= 0.0) were used in the final refinement. Neutral atom scattering factors and anomalous dispersion corrections were taken from International Tables for Crystallography.55 All refinement calculations and graphics were performed on a HP XW6200 (Intel XEON 3.2 Ghz), Debian-Linux computer at the University of Groningen with the program packages SHELXL56 (least-square refinements), a locally modified version of the program PLUTO57 (preparation of illustrations) and PLATON58 package (checking the final results for missed symmetry with the MISSYM option, solvent accessible voids with the SOLV option, calculation of geometric data and the ORTEP60 illustrations). Each asymmetric unit contains two formula units (molecule) with no atom setting at special position. The chiral centers of C7, C16, C23, C35, C44 and C51 have all the R-configuration,60 as known by synthesis route. The monoclinic unit cell contains four units of the title compound. A search of the distances yielded intermolecular contacts shorter than the sum of the van der Waals radii59 for the atoms: the moieties are linked by hydrogen bonds,43,44 forming an infinite one-dimensional network along the base vector [0 1 0]. No classic hydrogen bonds, no missed symmetry (MISSYM), however potential solvent-accessible area (voids of 62.3 Å3 per unit cell) was detected by procedures implemented in PLATON.60 181 Chapter 6 Table 6-3. X-ray crystallographic data for cholesteric compounds 1, 5 and 6. Compound Formula fw (g mol-1) crystal dimension (mm) color habit crystal system space group, no. a (Å) b (Å) c (Å) V (Å3) Z ρ (g cm-3) T (K) µ (cm-1) number of reflections number of refined parameters final agreement factors: wR (F2) R (F) GooF 1 C62H102N2O4 939.46 0.52 x 0.45 x 0.33 colorless parallelepiped monoclinic P21 21.742 (± 2) 9.7060 (± 7) 27.727 (± 2) 5731.3 (± 8) 4 1.089 100 (± 1) 0.66 10747 5 6 C34H57NO2 C58H96N2O4 511.83 885.41 0.25 x 0.11 x 0.43 x 0.39 x 0.09 0.08 colorless colorless needle platelet shaped triclinic monoclinic P1 P21 10.291 (± 2) 21.413 (± 3) 10.439 (± 2) 9.8090 (± 10) 14.432 (± 2) 25.647 (± ) 1519.0 (± 5) 5342.8 (± 11) 2 4 1.148 1.101 100 (± 1) 100 (± 1) 0.67 0.67 8775 10041 1245 679 1174 0.1787 0.0665 1.012 0.2090 0.0425 0.997 0.1621 0.1621 0.755 6.9 References 1 Collins, R.; Armitage, J.; Parish, S.; Sleight, P.; Peto, R. The Lancet 2002, 360, 7-22. Reinitzer, F. Monatsh. Chem. 1888, 9, 421-441. 3 Hoffmann, S. Mol. Cryst. Liq. Cryst. 1984, 110, 277-292. 4 Dierking, I. Textures of Liquid Crystals 2003, Wiley, New York. ISBN 3-527-30725-7. 5 Nishi, T.; Shinkai, S. Chem. Express 1993, 8, 173. 6 Echegoyen, L.E.; Portugal, L.; Hernandez, J.C.; Echegoyen, L.; Gokel, G.W. Tetrahedron Lett. 1988, 29, 4065-4068. 7 Fasoli, H.; Echegoyen, L.E.; Hernandez, J.C.; Gokel, G.W.; Echegoyen, L. J. Chem. Soc., Chem. Commun. 1989, 578-580. 8 Lin, Y.; Weiss, R.G. Macromolecules 1987, 20, 414-417. 2 182 Novel Cholesterol-based Organogelators 9 Mukkamala, R.; Weiss, R.G. Langmuir 1996, 12, 1474-1482. Geiger, C.; Stanescu, M.; Chen, L.; Whitten, D.G. Langmuir 1999, 15, 2241-2245. 11 Lu. L.; Cocker, M.; Bachman, R.E.; Weiss, R.G. Langmuir 2000, 16, 20-34. 12 Huang, X.; Terech, P.; Raghavan, S.; Weiss, R.G. J. Am. Chem. Soc. 2005, 127, 4336-4344. 13 Murata, K.; Aoki, M.; Suzuki, T.; Harada, T.; Kawabata, H.; Komori, T.; Ohseto, F.; Udea, K.; Shinkai, S. J. Am. Chem. Soc. 1994, 116, 6664-6676. 14 Ono, Y.; Kanekiyo, Y.; Inoue, K.; Hojo, J.; Nango, M.; Shinkai, S. Chem. Lett. 1999, 475-476. 15 Jung, J.H.; Ono, Y.; Sakurai, K.; Sano, M.; Shinkai, S. J. Am. Chem. Soc. 2000, 122, 8648-8653. 16 Lin, Y.; Kachar, B.; Weiss, R.G. J. Am. Chem. Soc. 1989, 111, 5542-5551. 17 Willemen, H.M.; Vermonden, T.; Marcelis, A.T.M.; Sudhölter, E.J.R. Eur. J. Org. Chem. 2001, 2329-2335. 18 Willemen, H.M.; Marcelis, A.T.M.; Sudhölter, E.J.R.; Bouwman, W.G.; Demé, B.; Terech, P. Langmuir 2004, 20, 2075-2080. 19 Van Esch, J.H.; Feringa, B.L. Angew. Chem., Int. Ed. 2000, 39, 2263-2266. 20 Chapter 3. 21 Chapter 4. 22 Pampín, M.C.; Estévez, J.C.; Estévez, R.J.; Suau, R.; Castedo, L. Tetrahedron 2003, 59, 8057-8066. 23 Dimroth, K.; Reichardt, C.; Siepmann, T.; Bohlmann, F. Liebigs. Ann. Chem. 1963, 661, 1-37. 24 Van Esch, J.; Schoonbeek, F.; De Loos, M.; Kooijman, H.; Spek, A.L.; Kellogg, R.M.; Feringa, B.L. Chem. Eur. J. 1999, 5, 937-950. 25 Hanabusa, K.; Yamada, M.; Kimura, M.; Shirai, H. Angew. Chem., Int. Ed. 1996, 35, 1949-1951. 26 Verbit, L.; Lorenzo, G.A. Mol. Cryst. Liq. Cryst. 1975, 30, 87-99. 27 Verdino, L.; Schadendorff, E. Monatsh. Chem. 1935, 65, 141. 28 Park, Y.A. Acta Cryst. 2005, A61, C274. 29 Tan, H.M.; Moet, A.; Hiltner, A.; Baer, E. Macromolecules 1983, 16, 28-34. 30 Jeong, S.W.; Shinkai, S. Nanotech. 1997, 8, 179-185. 31 Craven, B.M.; DeTitta, G.T. J. Chem. Soc. Perkin II 1976, 7, 814-822. 32 Craven, B.M. Nature 1976, 260, 727-729. 33 Espinosa, E.; Molins, E.; Lecomte, C. Chem. Phys. Lett. 1998, 285, 170-173. 34 Hsu, L.Y.; Kampf, J.W.; Nordman, C.U. Acta Cryst. B 2002, 58, 260-264. 35 Hsu, L.Y.; Nordman, C.U. Science 1983, 220, 604-606. 36 Oku, H.; Yamada, K.; Katakai, R. Acta Cryst. 2003, E59, o1130-o1132. 37 Mido, Y. Spectrochimica Acta 1973, 29A, 431-438. 38 Hesse, M.; Meier, H.; Zeeh, B. Spectroscopic Methods in Organic Chemistry 1997, Tieme Verlag, New York. ISBN 0-86577-667-9. 39 Marcus, Y. The Properties of Solvents 1998, 133-202, Wiley, Chichester. ISBN 0-471-98369-1. 40 Jeong, S.W.; Shinkai, S. Nanotech. 1997, 8, 179-185. 41 Supplementary crystallographic data for this paper are available from the IUCr electronic archives (Reference: CCDCxxxxxxx, as a CIF file). These data can be obtained free of charge via www.ccdc.cam.ac.uk/conts/ retrieving.html (or from the Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax (+44) 1223-336-033; or e-mail: deposit@ccdc.cam.ac.uk). 42 Burla, M.C.; Camalli, M.; Carrozzini, B.; Cascarano, G.L.; Giacovazzo, C.; Polidori, G.; Spagna, R. SIR2002: The program. J. Appl. Cryst. 2003, 36, 1103. 43 Berkovitch-Yellin, Z.; Leiserowitz, L. Acta Cryst. 1984, B40, 159-165. 44 Jeffrey, G.A.; Saenger, W. Hydrogen Bonding in Biological Structures 1991, Springer-Verlag, Berlin. ISBN 0-19-509549-9. 45 Supplementary crystallographic data for this paper are available from the IUCr electronic archives (Reference: CCDCxxxxxxx, as a CIF file). These data can be obtained free of charge via 10 183 Chapter 6 www.ccdc.cam.ac.uk/conts/ retrieving.html (or from the Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax (+44) 1223-336-033; or e-mail: deposit@ccdc.cam.ac.uk). 46 Bruker (2000). SMART, SAINT, SADABS, XPREP and SHELXTL/NT. Area Detector Control and Integration Software. Smart Apex Software Reference Manuals. Bruker Analytical X-ray Instruments. Inc., Madison, Wisconsin, USA. 47 Sheldrick, G.M. 2001, SADABS. Version 2. Multi-Scan Absorption Correction Program. University of Göttingen, Germany. 48 Duisenberg, A.J.M. J. Appl. Cryst. 1992, 25, 92-96. 49 Spek, A.L. J. Appl. Cryst. 1988, 21, 578-579. 50 Snow, M.R.; Tiekink, E.R.T. Acta Cryst. B 1988, 44, 676-677. 51 Le Page, Y. J. Appl. Cryst. 1987, 20, 264-269. 52 Le Page, Y. J. Appl. Cryst. 1988, 21, 983-984. 53 Altomare, A.; Burla, M.C.; Camalli, M.; Cascarano, G.L.; Giacovazzo, C.; Guagliardi, A.; Moliterni, A.G.G.; Polidori, G.; Spagna, R. J. Appl. Cryst. 1999, 32, 115-119. SIR-97. A Package for crystal structure solution by direct methods and refinement. Univ. of Bari, Perugia and Roma, Italy. 54 a) Flack, H.D. Acta. Cryst. 1983, A39, 876-881. b) Flack, H.D.; Bernardinelli, G. Acta Cryst. 1999, A55, 908-915. c) Flack, H.D.; Bernardinelli, G. J. Appl. Cryst. 2000, 33, 1143-1148. 55 Wilson, A.J.C. International Tables for Crystallography vol. C. 1992, Kluwer, Dordrecht. ISBN 07923-1638-X. 56 Sheldrick, G.M. SHELXL-97, Program for the Refinement of Crystal Structures 1997, University of Göttingen, Germany. 57 Meetsma, A. PLUTO, Molecular Graphics Program 2005, University of Groningen, The Netherlands. 58 a) Spek, A.L. PLATON, Program for the Automated Analysis of Molecular Geometry (A Multipurpose Crystallographic Tool) 2005, University of Utrecht, the Netherlands. b) Spek, A.L. J. Appl. Cryst. 2003, 35, 7-13. 59 Bondi, A. J. Phys. Chem. 1964, 68, 441-451. 60 a) Spek, A.L. Acta Cryst. 1990, A46, C-34. b) Spek, A.L. Am. Cryst. Assoc., Abstr. Papers 1994, 22, 66. 61 Supplementary crystallographic data for this paper are available from the IUCr electronic archives (Reference: CCDCxxxxxxx, as a CIF file). These data can be obtained free of charge via www.ccdc.cam.ac.uk/conts/ retrieving.html (or from the Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax (+44) 1223-336-033; or e-mail: deposit@ccdc.cam.ac.uk). 62 Burla, M.C.; Caliandro, R.; Camalli, M.; Carrozzini, B.; Cascarano, G.L.; De Caro, L.; Giacovazzo, C.; Polidori, G.; Spagna, R. SIR2004. An improved tool for crystal structure determination and refinement. J. Appl. Cryst. 2005, 38, 381-388. 184