摘要:

J O U R N A L O F C H E M I S T R Y Materials Feature Article Non-conventional liquid crystals—the importance of micro-segregation for self-organisation Carsten Tschierske* Institute of Organic Chemistry, University Halle, Kurt-Mothes-Str.2, D-06120 Halle, Germany Selected examples of recently synthesised non-conventional liquid crystals are highlighted. These are cyclic and open chain oligoamides, molecules containing tetrahedral or octahedral central cores, dendrimers, polyhydroxy amphiphiles, taper shaped molecules, liquid crystals with perfluorinated or oligosiloxane segments, rod-coil molecules as well as special types of polycatenar and laterally branched calamitic molecules.Their mesomorphic properties are discussed as a consequence of incompatibility, micro-segregation and space filling.The analysis is based on the general concept of amphiphilicity, which describes any chemical or structural contrast within a molecule, such as hydrophilic/lipophilic, polar/non-polar, hydrocarbon/fluorocarbon, oligosiloxane/hydrocarbon or rigid/flexible. These non-conventional liquid crystals will be regarded as block molecules. Depending on the degree of chemical and structural diVerence and the size of the diVerent building blocks micro-segregation can occur with formation of lamellar, columnar or spheroidal aggregates which organise to smectic, columnar and cubic mesophases.The striking analogies between the polymorphism of thermotropic and lyotropic liquid crystals and block-copolymers are pointed out. mers).12 More recently the interest in thermotropic and/or Introduction lyotropic mesophases of rod-like polymers (hairy rods, poly- The transition from the highly ordered crystalline solid to the peptides, polyisocyanates, cellulose derivatives etc.)11a has disordered isotropic liquid state often occurs in a multistep grown.Typical examples of the main types of molecules process via intermediate phases (mesophases) which are desig- forming liquid crystalline phases are shown in Fig. 1. nated as liquid crystalline phases.1 In these phases the order Additionally, the diVerent molecular structures have been of the crystalline state is only partly lost and the individual combined. Polycatenar13 compounds for example have a strucmolecules have already got some degree of mobility.These ture intermediate between rod-like and disc-like molecules and mesophases can occur in pure materials with dependence on amphiphilic molecules have been combined with anisometric the temperature2 (thermotropic mesophases) as well as in structural units to bridge the gap between thermotropic and multicomponent systems with dependence on their composi- lyotropic liquid crystals.11a tion and the temperature (e.g. lyotropic phases).Because liquid Regardless of the molecular structure, liquid crystalline crystals combine order and mobility on a molecular level they phases can be classified according to the degree of long range are important in materials science as well in the life sciences. order of their constituent parts and the symmetry within the Important applications3 of thermotropic liquid crystals are mesophases.1 An overview of the main phase types is given in electrooptical displays, temperature sensors and selective Figs. 2–4. reflecting pigments.4 Lyotropic systems are incorporated in Mesophases in which the positional long range order is cleaning processes5 and are important in the cosmetics indus- maintained, but orientational order is lost are termed distry, they are used as templates for the preparation of meso- ordered crystals or plastic crystals.14 If the orientational order porous materials6 and serve as model systems for biomembranes.7 Furthermore the liquid crystalline state is ubiquitous in living matter.Most important are the biological membranes, DNA can form lyotropic mesophases8 and natural silk is spun by the silkworm Bombyx mori from an aqueous nematic liquid crystalline phase of fibroin.9 1.1 Low molecular weight liquid crystals The examples mentioned above indicate that liquid crystals can have quite diVerent molecular structures.In a classical approach, substances that can produce liquid crystalline phases are considered to belong to one of two distinct classes of materials: the nonamphiphilic but anisometric mesogens and the amphiphilic molecules.The first class includes anisometric, rod- or disc-like molecules or aggregates in most cases giving exclusively thermotropic liquid crystals. The second class includes amphiphilic molecules such as detergents and lipids leading to lyotropic,10 and also thermotropic mesophases.11 Both types of mesogenic molecules can be attached to a polymer backbone (LC-side-chain polymers) or they can be incorporated in the polymer backbone (LC-main-chain poly- Fig. 1 Typical examples of the main types of molecules forming liquid crystalline phases *E-mail: coqfx@mlucom6.urz.uni-halle.de; Fax: +49 (0) 345 55 25664. J. Mater. Chem., 1998, 8(7), 1485–1508 1485Fig. 2 Schematic representation of the diVerent nematic mesophases formed by the organisation of calamitic molecules (N), disc-like molecules (ND), columnar aggregates of disc-like molecules (Ncol) and polymolecular aggregates of amphiphilic molecules (rod-like micelles: Ncol, plate-like micelles; ND) is maintained, but the positional long range order is completely lost the mesophases are termed nematic liquid crystals (Fig. 2). However, the positional order is often only lost in one or two dimensions. These mesophases will be designated as positional ordered mesophases (Fig. 3). Fig. 3 Positional ordered fluid thermotropic mesophases formed The most important positional ordered mesophases are the by calamitic, polycatenar, disc-like and amphiphilic molecules. smectic (S), the columnar (Col) and the cubic (Cub) phases.Abbreviations: SA=smectic A-phase, SC=smectic C-phase (the lamel- Rod-like molecules predominantly form smectic phases. In lar phases are shown as cross sections of a segment of the layers which are extended laterally and in the third dimension), SA~=antiphase, Fig. 3 only smectic phases of the fused type, i.e. without SC~=ribbon phase [ribbon-like segments of the smectic layer which additional positional order in the layers (SA and SC), are are extended in the third dimension and arranged in a centred shown.Some polar calamitic molecules can form mesophases rectangular (SA~, Colr) or oblique (SC~, Colob) 2D-lattice],15 LD= in which the smectic layers are collapsed into ribbons (SA~ lamellar (smectic) mesophase consisting of disc-like molecules;17 Col= and SC~).15 Because the ribbons can be organised in a rectangu- columnar mesophases, CubV=bicontinuous cubic mesophase conlar or in an oblique 2D-lattice, these mesophases could alterna- sisting of mutually interwoven networks of branched cylinders, CubI= discontinuous cubic phase consisting of spheroids.18,77.Space groups tively be described as columnar phases (Colr, Colob).Columnar of the cubic phases are given in italics.19 The arrangement of the phases of the ribbon-type, and SC-phases can also be found molecules in the cubic mesophases of polycatenar compounds is not for polycatenar molecules,13 which have a shape intermediate yet clear. Presumably, they represent mutually interwoven networks between rod-like and disc-like molecules (for an example see of ribbons.16 The thermotropic mesophases of amphiphilic molecules Table 5).In this class of compounds cubic phases (CubV) can usually represent inverted phases (type 2).82 Then, the branched occur as intermediate phases between layer-like and columnar cylinders of the CubV-phase represent the head group areas embedded in the lipophilic continuum of the molten alkyl chains.The CubI- organisation of these molecules. Ia3d and Im3m cubic lattices phases consist of eight nonglobular spheroidal micelles per unit have been found.16 In Fig. 3 two possible structures of these cell,77,78,90 only their position in the unit cell is shown, an explanation lattices are shown; the detailed arrangement of the molecules of a model based on prolate micelles is given in Fig. 4( b). in these cubic lattices however, is not yet clear. Disc-like molecules preferably organise to columnar mesophases. The LD-phase, which is seldom observed, represents a tendency of their hydrophilic and hydrophobic parts to segregate in space into distinct microdomains. Owing to the chemi- layer-like organisation of discotic molecules.17 A wide variety of diVerent mesophases is found for amphi- cal bonding of the amphiphatic parts, segregation does not lead to macroscopic phase separation.Instead, it results in the philic molecules in the pure state,11,18 which is shown in the lower part of Fig. 3. Their mesomorphic properties can be formation of diVerent regions which are separated by interfaces at a molecular scale. These interfaces have characteristic mean influenced by addition of protic solvents, which results in the formation of lyotropic mesophases (see Fig. 4).10 In thermo- interface curvatures, which are largely determined by the mean lateral areas of the competing parts of the amphiphilic mol- tropic and in lyotropic mesophases of these amphiphilic molecules, diVerent types of cubic phases (CubV and CubI) can ecules.10,20 Non-curved planar aggregates organise into lamellar (smectic) mesophases.A finite mean interface curvature also be detected in addition to the smectic and columnar mesophases.18,19 These cubic phases, which can occur in diVer- gives rise to cylindrical aggregates of one component embedded in a continuum of the other component. Extended non- ent classes of compounds, represent a fascinating state of matter, because they are three-dimensionally ordered crystals branched columns can organise to columnar mesophases (Col).Branched columns occur at the transition between lamellar of fluid polymolecular aggregates. The diVerent types of cubic lattices found in lyotropic systems of amphiphilic molecules and columnar mesophases.21 They form three-dimensional interwoven networks which can organise into diVerent cubic are shown in Fig. 4( b). In many cases, however, the structure of the cubic phases is still unclear. lattices [bicontinuous cubic mesophases, CubV-phases, see lower part of Fig. 4(b)]. Further increasing the interface curva- The remarkable feature of amphiphilic molecules is the 1486 J. Mater. Chem., 1998, 8(7), 1485–1508gence of the liquid crystalline phases and typical defect structures, which are influenced by the molecular organisation.Contrary to all other mesophases, the cubic phases are optically isotropic. They can be distinguished from the isotropic liquid state by their high viscosity and the crystal-like shape with cornered phase boundaries of their domains while growing from other phases [see Fig. 5(b)]. Selected examples of the optical textures of smectic, columnar and cubic mesophases are shown in Fig. 5. More detailed investigation of the molecular organisation (diVerentiation of low-temperature smectic phases, diVerentiation between diVerent types of columnar and cubic phases), however, requires X-ray diVraction, NMR methods and further investigations. 1.2 Block-copolymers In recent decades it was found that block copolymers can also produce well-developed liquid crystalline phases (Fig. 6).22 In the most simple case AB-diblock copolymers are made up of two chemically distinct blocks, A and B, covalently bonded to form a single chain. Here, mesomorphic properties are caused by the incompatibility of the diVerent polymer blocks.22 DiVerent polymers are, in general, incompatible. Because of Fig. 4 a: The main lyotropic mesophases that depend on the interface curvature.10 Abbreviations: La=lamellar (smectic A) phase; for the other abbreviations, see Fig. 3. Space groups of the cubic phases found in the diVerent regions of the phase diagram are given in italics.19 Additional non-cubic phases (e.g. tetragonal phases) are possible at the transition from the smectic to the columnar phases.10,19 b: Cubic lattices occurring in lyotropic systems.19 Their position in the lyotropic phase sequence is shown in Fig. 4a. The displayed branched cylinders and the spheroids can represent the headgroup areas (plus solvent) embedded in the lipophilic continuum of the molten alkyl chains (inverted phases, type 2).In the normal phases (type 1) the cylinders and micelles represent the regions of the lipophilic chains surrounded by the head-groups embedded in the polar solvent. The CubI-phase of the Pm3n-type consists of eight nonglobular spheroids per unit cell, only the model assuming prolate micelles is shown19b [i: arrangement of the micelles in the cubic lattice; ii: view upon a face of the cubic lattice built up by prolate micelles, the micelles in the centre and at the corners are rotationally disordered and are represented by dots; adapted from ref. 19(b)]. ture leads to spheroids—closed globular or non-globular micelles—which can also organise into diVerent cubic lattices [discontinuous cubic phases, CubI-phases, see upper part of Fig. 4(b)].18,19 The interfaces between the segregated regions can be curved away from the regions with stronger cohesive forces (the stronger cohesive forces are found in the continuum surrounding the aggregates, normal phases, type 1) or in the direction of these regions (the stronger cohesive forces are found within the aggregates, inverted phases, type 2).The microstructure of the diVerent thermotropic and lyotropic liquid crystalline phases can easily be determined by Fig. 5 Selected examples of typical textures of liquid crystalline phases microscopy between crossed polarisers which allows a reliable as observed between crossed polarizers. a: Oily streaks texture with diVerentiation between nematic, smectic A, smectic C, smectic Myelin-figures of a lyotropic smectic A-phase. b: Bicontinuous cubic low-temperature phases and columnar phases by their typical mesophase (black areas) as growing from a smectic C-phase.c: Columnar mesophase. optical textures. These textures are the result of the birefrin- J. Mater. Chem., 1998, 8(7), 1485–1508 1487diblock copolymer melt is homogeneous. The order–disorder transition occurs when NxAB is increased (low temperature). For symmetric block polymers with fA=fB=0.5, the melt is ordered in a lamellar structure (LAM).As one block (B) is made longer at the expense of the other block (A), the morphology changes to columns (molar fraction of the minor block, fA about 0.3–0.4)25 and then to spheres ( fA<0.3) of component A embedded in a continuum of the major component B. These morphologies represent equilibrium states with the lowest free energy.Because the interaction energy between the segregated regions is mainly localised at their interfacial region, the systems are in equilibrium only when the total area of the interfaces is a minimum. The domains are regular, often forming columnar and cubic mesophases. Besides these classical phases, additional complex structures, such as hexagonally modulated layers (HML), hexagonally perforated layers (HPL)26 and cubic (gyroid) phases with the space group Ia3D27 have been found in a narrow range of f and Nx between lamellar and columnar morphology near the order–disorder transition.Some of these intermediate states may be nonequilibrium states. Above fA=0.5 there is an inversion of the role of the two components.Because the blocks are diVerent, especially due to conformational asymmetry,28 , the Nx versus f diagrams of diblock polymers are not symmetric in respect to f=0.5. As an example the NxAB versus f diagram for a polyisoprene– polystyrene diblock copolymer is shown in Fig. 6.29 The analogy of the mesophases formed by block copolymers with those of amphiphilic molecules is striking and block copolymers can be regarded as polymeric analogues of simple amphiphiles.Like amphiphiles, they can form not only thermotropic phases as pure materials. The addition of selective solvents to one or the other components of the block copolymer can aVect the morphology by changing the eVective volume fractions of the components. A lamellar phase can be changed Fig.6 xN versus fPI (PI=polyisoprene) phase diagram for poly(isovia a cylindrical to a spherical morphology on addition of a prene-b-styrene) diblock copolymers (reprinted with permission from solvent selectively interacting with one of the blocks.If these ref. 29, copyright American Chemical Society) and morphologies of mixed systems are ordered, they can be regarded as lyotropic the ordered structures.Abbreviations: HPL=hexagonal perforated layers; for the other abbreviations, see Fig. 3. mesophases. In a recent example six diVerent mesophases and two micellar solutions (L1, L2) have been realised for a poly(oxybutylene)–b-poly(oxyethylene)–water–xylene system:30 the large size of polymer molecules, the entropy gain on mixing is very small and in the absence of specific favourable inter- L1—CubI1—Colh1—La—CubV2—Colh2—CubI2—L2 actions between the two polymers, it cannot compensate for the enthalpy loss that occurs.With block copolymers, these Thus almost the whole ‘hypothetical lyotropic phase diagram’ (see Fig. 4) was scanned. Also micelles and organised films at incompatibility eVects give rise to a segregation of the diVerent polymer blocks into distinct micro domains.However, the fact interfaces can be formed by block copolymers. The analogy of the mesophases of block copolymers with that the diVerent blocks A and B are chemically connected prevents macroscopic phase separation, leading to microphases the smectic and columnar ordering of nonamphiphilic liquid crystals and the importance of the amphiphilic character31 and with A-rich and B-rich domains separated by internal interfaces.The micro-segregation can be observed below a micro-segregation for mesomorphic self-organisation have been pointed out several years ago.22a–c The calamitic mesogens certain order–disorder temperature, which depends on the size of the blocks and on the degree of chemical and structural for example can be regarded as block compounds composed of a rigid block terminated by flexible alkyl chains forming diVerence between the blocks.More precisely the important parameters which govern phase separation for A–B diblock predominantly lamellar structures due to the segregation of the incompatible molecular parts.31 Furthermore, it was pro- copolymers are the total degree of polymerisation N=NA+NB, the segment interaction parameter (Flory–Huggins parameter posed that thermotropic and lyotropic mesomorphism can be viewed as the consequence of division of space into two media xAB) and the volume fraction of the components f.23,24 The segregation between the two blocks is dictated by the product separated by interfaces, generated by the assemblage of chemical groups with diVerent aYnities.32 Thereby, the curvature of of N and xAB whereas the interfacial curvature is controlled by f.The Flory–Huggins parameter xAB can be estimated the interfaces is largely determined by the molecular structure: rigid molecules favour the formation of flat interfaces, if they according to eqn. (1) are rod-like and they favour cylindrical ones when they are xAB=VR(dA-dB)2/RT (1) disc-like.For flexible molecules, the interface can adapt diVerent shapes with dependence on the temperature and solvent where dA and dB represent the Hildebrand solubility parameter and V R a reference volume. Thus xAB which is a measure of concentration.32 An early attempt to quantitatively describe micro-segre- the incompatibility between A and B depends on the diVerence in the intermolecular interactions in the segregated regions gation in lyotropic systems of amphiphilic molecules was provided by the R-theory of fused micellar phases.14 Biphilic (expressed as the diVerent solubility parameters dA and dB) and on the temperature T .At low NxAB (high temperature) a asymmetry, flexibility asymmetry, electric and steric asymmetry 1488 J.Mater. Chem., 1998, 8(7), 1485–1508have been introduced to describe thermotropic and lyotropic systems in the framework of a generalised molecular asymmetry model.33 In recent years several authors have discussed microsegregation as an additional driving force for columnar mesomorphism in some non-conventional mesogens.34–39 In this way, it becomes more and more evident that the important structural entity of all positional ordered liquid crystalline phases is not the molecule, but the interface built up by the organisation of the molecules.32 Thus, the concepts of thermotropic and lyotropic liquid crystals and block copolymers are closely related and it is often only a diVerence in semantics, like the notation SA, La, G and LAM, that actually denote an ordered array of non-curved ( layer-like) aggregates or the notations Ha, Colh, Dh, Wh, Oh, M, HEX, etc., which all denote mesophases consisting of hexagonally ordered columns. In this paper we will use the descriptor S for lamellar (smectic) layer structure, Col40,41 for columnar mesophases, CubV for bicontinuous cubic mesophases, and CubI for micellar cubic phases,42 irrespective of whether they occur in thermotropic or lyotropic systems or in the phase sequence of block copolymers. 1.3 Scope of this feature article In this feature article, selected examples of recently synthesised non-conventional mesogens will be highlighted and their mesomorphic properties will be discussed as a consequence of incompatibility, micro-segregation and space filling, pointing out the striking analogies between the polymorphism of blockcopolymers, and thermotropic and lyotropic liquid crystals.We will focus our interest especially on materials that can undergo transitions between diVerent phase morphologies by structural variations and/or on addition of solvents. It seems that mesophases consisting of columnar aggregates are ubiquitous in all types of organised soft matter.Therefore columnar mesomorphism will be systematised and discussed in more detail in a separate feature article.43 Here we treat columnar mesophases only if their occurrence is mainly caused by microsegregation. 2 Cyclic and open chain oligoamides Polyacylated azacrowns have been found to exhibit liquid crystalline properties.44 Two selected examples are shown in OCnH2 n+1 O OCnH2 n+1 N N N N OCnH2 n+1 O CnH2 n+1O H H CnH2 n+1O O OCnH2 n+1 CnH2 n+1O O CnH2 n+1O N OCnH2 n+1 O OCnH2 n+1 H CnH2 n+1O O CnH2 n+1O N H CnH2 n+1O O CnH2 n+1O N CnH2 n+1O O OCnH2 n+1 N x 3a n = 10 K 93.5 Colh 104 Iso a 3b n = 16 K 116 (Col 84 Cub 104.5) Iso 4a n = 5 K 86.3 Cub 104.4 Iso 4b n = 10 K 117.5 (M 102.5) Iso b 4c n = 16 K 85.3 Cub 106.9 Iso 5 n = 10 K 61 Colh 120 Iso c C10H21O O C10H21O C10H21O O OC10H21 OC10H21 O C10H21O O N N O OC10H21 O C10H21O 6 g 15 Col 68 Iso Phase transition temperatures (T /°C) of selected open chain oligoamides 3,41 4,46 compound 6,36,39 and the polyamide 5.45 aK 99 (Colh 93) Iso45.bFrom X-ray data it was suggested that M represents a lamellar (smectic) mesophases; however the optical texture is similar to a columnar phase.46 cK 62 Colh 103.3 Iso36.Fig. 7. At first, it was thought that the average disc-like shape of these azamacrocycles is responsible for their columnar Fig. 7 Phase transition temperatures (T /°C) of two selected mesogenic mesomorphism. However, the rather flexible central core allows azacrown derivatives 145 and 244b and schematic sketch of two possible diVerent molecular conformations: an elongated one with a conformations. aFrom X-ray data it was suggested that M represents a lamellar (smectic) mesophase.41 rod-like shape and a radial one with disc-like character (see J.Mater. Chem., 1998, 8(7), 1485–1508 1489Fig. 7).41 Indeed, smectic phases were proposed for some bonding as pure materials also display thermotropic mesophases.This shows that the formation of anisometric core azamacrocycles, such as 1,4,7-triazacyclononanes 2 having 3,4- dialkoxybenzoyl substituents, despite their—at first glance— structures via hydrogen bonding is not necessary for mesophase formation. Furthermore, these compounds do not represent disc-like molecular shape.41 Recently, it also was reported, that the open chain analogues amphiphiles in a classical sense and therefore they do not belong to one of the classical types of liquid crystal forming of these cycles also display thermotropic liquid crystalline behaviour.39,41,45–47 At first it was assumed that the appearance compounds mentioned above.More recently it was proposed that the columnar phases of of columnar mesophases in the case of many open chain compounds such as 3 and 4 is caused by the formation of disc- the polyamide 5 and the ester 6 are solely formed by the segregation of the polar groups from the non-polar aliphatic shaped individual molecules by cyclisation via hydrogen bonding45,47 or by formation of polymolecular aggregates with a chains.36,39 Thus, micro-segregation should be the main driving force for their self organisation with formation of liquid helical backbone.But later it was found that polyamides such as 536,45,47 and even the ester compound 639 which lack a crystalline phases. Interestingly, not only columnar phases were found. The proton donor and thus lack the ability to form hydrogen diethylenetriamine derivative 4b, for example, forms a smectic mesophase.41,46,48 Cubic phases have been found for the short Table 1 Thermotropic transition temperatures (T /°C) of selected liquid chain and the long chain diethylenetriamine derivatives 4a and crystals 7,37 849 and 951 with a tetrahedral central core 4c as well as for the triethylenetetraamine derivatives with higher chain lengths such as 3b.41 This indicates that in strong analogy to lyotropic systems and block copolymers the mesophase type is determined by a delicate balance of the relative space filling of the polar groups and the lipophilic chains. 3 Liquid crystals containing a tetrahedral or octahedral central core Columnar mesomorphism was also found for organic molecules with a tetrahedral central core37,49,50 such as 7,37 849 and 9,51 for octahedral metal complexes such as 13,52,53 C10H21O R2 R1 C10H21O O R2 R1 O O X R1 R2 O O OC10H21 R2 O OC10H21 R1 for diabolo compounds such as 1454a and dendrimers (e.g.compound 15).38,55 Comp. X R1 R2 K Col Iso We will focus our discussion on compounds 7–12. The 7a COOCH2 H H <20 — — formation of these columnar phases was explained by a cylinder 7b COOCH2 OC10H21 H 54 ( 47) model as shown in Fig. 8 which was at first proposed for the 7c COOCH2 OC10H21 OC10H21 41 ( 8) mesophases of compounds 7.37 As evident from CPK-models 8 CONH OC10H21 H 47 66 of the pentaerythritol tetrabenzoates 7, the tetrahedral linking 9 CH2OCH2 OC10H21 H <20 32 unit inhibits the flat disc-like organisation of the taper-shaped Fig. 8 Schematic of the organisation of compounds 639 and 7b37 in their columnar mesophases 1490 J.Mater. Chem., 1998, 8(7), 1485–1508alkyl substituted aromatic units. Instead, these molecules can adapt diVerent conformations with the alkyl chains more or less randomly distributed around the tetrahedral connecting unit and no particular anisometric shape is provided by the central unit. However, the accumulation of the polar groups (consisting of the carboxy groups, the aromatic rings and the ether-oxygens) gives rise to a distinct polar region in the centre of the molecules.The polar regions of diVerent molecules interact preferably with each other and thus segregate from the aliphatic chains into separate domains (see Fig. 8).37 During the process of self organisation the conformation of the individual molecules is influenced.Conformers with a rather flat shape should be favoured over other conformers, because they enable the most eYcient interaction of their polar regions with neighbouring molecules. Thus, the supermolecular organisation of the molecules determines the actual molecular conformation. The relative space filling of the segregated regions and the pre-organisation of the lipophilic chains around the tetrahedral linking unit favour columnar aggregates.The appearance of mesomorphic properties strongly depends on the number of chains. The most stable columnar phases were found for the 3,4-dialkoxybenzoates (e.g. compound 7b). Introduction of an additional alkyl chain at each aromatic ring decreases the mesophase stability. The 12 alkyl chains of compound 7c probably disturb the polar–polar interaction between neighbouring molecules.37 Also the peracylated D-threitol and D-mannitol derivatives 11 and 12 display columnar mesomorphism.49 The increased clearing temperature of the D-mannitol derivative 12 in comparison to the D-threitol derivative 11 indicates that the mesophases can be stabilised by increasing the number of 3,4-dialkoxybenzoyl groups connected with each other.No liquid crystalline phases were found however on cooling the glycerol ester 10.49 It seems that irrespective of the particular shape of the molecules a minimum number of at least four 3,4- dialkoxybenzoyl groups must be connected to achieve liquid crystallinity. The mesophase stability can be enhanced by increasing the intramolecular polarity contrast.For example, the replacement of only one CH2OOC group of compound 7b by an amide group (additional hydrogen bonding in compound 8) raises the clearing temperature of the columnar mesophase49 (a similar comparison can be made between the linear compounds 6 and 3a). Replacement of one carboxy group of compound OC10H21 O OC10H21 O O C10H21O O C10H21O O OC10H21 O OC10H21 O C10H21O O C10H21O C10H21O O O C10H21O O OC10H21 O OC10H21 O OC10H21 O OC10H21 OC10H21 O OC10H21 O O O O C10H21O O C10H21O OC10H21 O OC10H21 O O OC10H21 OC10H21 C10H21O O C10H21O C10H21O O O C10H21O 10 K 98 Iso 11 K 44 (Col 40) Iso 12 K <20 Col 55 Iso 7b by a less polar CH2O group however (compound 9) lowers Transition temperatures (T /°C) of peracylated glycerol, D-threitol and D-mannitol derivatives49 the mesophase stability.51 Also the mesophase formation by dendritic 3,4-dialkoxybenzamides such as 1538 has been described as mainly resulting from micro-segregation.Additionally, the liquid crystallinity tropic mesogens11a forming not only thermotropic, but also lyotropic mesophases on addition of protic solvents.57–60 of the octahedral metallomesogens 13 and diabolo liquid crystals (e.g.compound 14)54a should be mainly due to micro- As a generally accepted rule at least two OH groups are necessary for the occurrence of liquid crystalline phases.58,61 segregation of the polar cores from the lipophilic alkyl chains. Simple primary alcohols (e.g. compound 16)—despite the fact Micro-segregation should also contribute to the special properthat they are amphiphilic and can for example form monomole- ties of main-chain and side-chain liquid crystalline polymers cular films at interfaces68—do not form thermotropic or incorporating rod-like, disc-like and taper-shaped mesogenic lyotropic mesophases.However, they can be turned into meso- units.12,35 gens by introduction of additional hydroxy groups close to the first OH group.As obvious from Table 2, the stability of the thermomesophases continuously increases on increasing 4 Polyhydroxy amphiphiles the number of hydroxy groups (diols<tetraols~monosac- An interesting family of liquid crystals is provided by the charides<disaccharides). This shows that a certain polarity is polyhydroxy amphiphiles, such as diols11a,56–61 (e.g.compound necessary for micro-segregation in these small molecules. It 1758), polyols, such as compound 1862 and the large family of seems that hydroxy groups can give rise to micro-segregation carbohydrate mesogens63 (e.g. compounds 19,64 2065 and 2166). with formation of mesophases only if a suYciently large Here distinct polar regions are formed by the dynamic liquid- number of them are connected side by side forming a distinct like hydrogen bonding networks between the OH groups, polar region (diol, carbohydrate unit) beside the lipophilic which separate from the lipophilic n-alkyl chains.67 A special region of the alkyl chains.In other words, the fixation of feature of these compounds is that the amphiphilicity can be several OH groups close to each other enables them to act covaried over a wide range depending on the number of hydroxy operatively and increases the competition between the antagonistic molecular parts.In this respect the polyhydroxy amphi- groups. Furthermore, these compounds are typically amphi- J. Mater. Chem., 1998, 8(7), 1485–1508 1491OC12H25 OC12H25 C12H25O C12H25O O O C12H25O C12H25O OC12H25 OC12H25 O O Cr C12H25O C12H25O C12H25O C12H25O O O O C12H25O C12H25O O O O C12H25O O N O O OC12H25 O O O OC12H25 C12H25O O C12H25O O O O C12H25O OC12H25 O H O O N O OC12H25 OC12H25 O O O OC12H25 O O O C12H25O O O O OC12H25 OC12H25 O OC12H25 O O O OC12H25 C12H25O O H C10H21O N O H OC10H21 C10H21O N O H C10H21O C10H21O N O H C10H21O C10H21O N O H C10H21O C10H21O N O H C10H21O C10H21O N O H C10H21O C10H21O N O H C10H21O C10H21O N O H C10H21O C10H21O N O H OC10H21 OC10H21 N O H OC10H21 OC10H21 N O H OC10H21 OC10H21 N O H OC10H21 OC10H21 N O H OC10H21 OC10H21 N O H OC10H21 C10H21O N O H OC10H21 OC10H21 N O H C10H21O N N N N N N N N N N N N N N 13 K ? Colr 65 Colh 108 Iso 14 K<20 Col 150-152 Iso 15 K 76 Colh 124 Iso Transition temperatures (T /°C) of representative low-aspect ratio mesogens: octahedral metallomesogens 1353b (approximate values taken from a bar graph), diabolo mesogen 1454 and dendrimer 1538 1492 J.Mater. Chem., 1998, 8(7), 1485–1508Table 2 Transition temperatures (T /C) of the thermotropic mesophases of selected n-alkyl-substituted polyhydroxy amphiphiles (17,58 18,62 19,64 20,65 2166). Abbreviations Lß=lamellar mesophase with rigid hexagonal ordered alkyl chains; for the other abbreviations see Fig. 3. No. Compound Transition temperatures, T/°C 16 K 56 Iso 17 K 68 (Lb 65) Iso 18 K 108 SA 156 Iso 19 K 80.4 SA 143.4 Iso 20 K 102 SA>245 Iso 21 K 111 Cub 205 Iso HO HO OH OH OH OH HO HO O OH HO OH O O OH HO OH O O O OH HO OH HO O OH HO OH HO O C14H29 O OH HO OH HO O philes can also be regarded as block molecules consisting of a hydrophilic oligo(hydroxymethylene) block and a lipophilic oligo(methylene) block.It should be mentioned here that in all polyhydroxy amphiphiles the hydrogen bonding leads to strong cohesive forces in the polar regions which reinforces C10H21O C10H21O COOCH3 22 mp 43 °C micro-segregation and produces temporary cross-links between the individual amphiphile which also contribute to mesophase stability.If the ester group of 22 is replaced by more polar groups, such as a polyhydroxy group, a stronger amphiphilicity is The block molecules described in the preceding sections bridge the gap between classical small mesomorphic (amphi- generated and liquid crystalline properties can be found for these small molecules.The formation of columnar meso- philic and anisometric) molecules and block copolymers. What we can learn is the following: if chemically diVerent units are phases by small taper-shaped amphiphilic compounds was described for several diVerent polyhydroxy compounds. distributed in a molecule in such a way that two or more diVerent regions are formed, then micro segregation can result.Probably the first compound of this type was diisobutylsilanediol 23.70 Selected examples of other taper-shaped polyhydroxy As in block copolymers the micro-segregation can be observed below a certain order–disorder temperature, which corre- amphiphiles61,63,71–74 are shown. In most of these compounds the aliphatic chains were grafted sponds to the the transition from the liquid crystalline phase to the isotropic liquid (clearing temperature).It depends on directly onto the polyhydroxy groups of the polar moieties. In this way the number of attractive hydrogen bonds is reduced the size of the blocks and on the degree of diVerence between the blocks. If the blocks are very large as in copolymers even and therefore a larger number of alkyl chains often leads to the loss of mesogenity. The penta-O-hexyl-myo-inositol 27a75 very small diVerences of the chemical structure give rise to micro-segregation at ambient temperature (e.g.branched and and the tri-O-alkyl-D-glucose derivative 28b72 for example, do not display thermotropic behaviour. non-branched hydrocarbons). If the blocks get smaller, the chemical diVerence between the blocks must be enhanced in In other cases the lipophilic alkyl chains were not directly attached to the polyhydroxy unit, but connected via an order to maintain the ability to segregate.In classical amphiphiles a single ionic head group is suYcient. In polyhydroxy aromatic linking unit.18,60,76–81 This allows the number of hydrogen bonds to remain constant whereas the size of the amphiphiles at least two OH groups are necessary and in nondendritic ester compounds probably at least four MCOOM hydrophobic region can be gradually changed by variation of the number and the length of the grafted alkyl chains.groups must be combined. In this way the same diversity of diVerent mesophases as observed in lyotropic systems and in block copolymers can be 5 Taper-shaped amphiphiles realised in thermotropic mesophases of pure cis-3,5-dihydroxycyclohexanes (e.g.compounds 29, 30),60,80 2,3-dihydroxypro- The pentaerythritol tetrabenzoate 7b can be described as a tetramer composed of four 3,4-dialkoxybenzoate units linked pylamides (compounds 31, 32),76,78 glucamides (e.g. compounds 33–35), and N-methylglucamides.18,76,78 Selected examples are together.The ethyl 3,4-didecyloxybenzoate 2269 which represents a monomeric segment of the liquid crystalline shown. Single chain compounds form exclusively SA-phases.82 Double chain compounds can form smectic (SA), bicontinuous compound 7b is non-mesomorphic. J. Mater. Chem., 1998, 8(7), 1485–1508 1493Fig. 9 Mesophases (transition temperatures, T /°C) of the N-benzoylamino- 1-deoxy-D-glucitols 33–35 showing dependence on the OH Si OH HO OH OH OH OH SC9H19 SC9H19 HO SC9H19 OH HO HO OH SC9H19 HO OC6H13 OC6H13 HO HO OH C6H13O OC6H13 C6H13O C6H13O OC6H13 OH C6H13O OH C6H13O C6H13O OC6H13 OH O OC8H17 HO C8H17O OH OH O OC8H17 C8H17O C8H17O OH OH 23 K 88.4 Col 98.7 Iso 24 K 101.6 Col 125.8 Iso 25 K 143.2 (Cub 129.9) Col 178.6 Iso 26 K 111.5 Colh 167 Iso 27a liquid 27b K 27.7 Colh 35.8 Iso 28a K 97 (Col 87.5) Iso 28b syrup number and length of the alkyl chains18,76 Phase transition temperatures (T /°C) of selected taper shaped polyhydroxy amphiphiles 23,70 24,71 25,74 26,61 27b,73a 28a72 and 28b72 and the alcohol 27a75 Fig. 10 Binary phase diagram of the system 32c–3678 O OH OH CnH2 n+1O CnH2 n+1O O O OH OH CnH2 n+1O CnH2 n+1O O OCnH2 n+1 N CnH2 n+1O CnH2 n+1O O H OH OH N CnH2 n+1O CnH2 n+1O OCnH2 n+1 O H OH OH 29a n = 5 K 73 (SA 66.5) Iso 29b n = 8 K 63 (Cub 63) Colh 95.5 Iso a 29c n = 12 K 79.5 Colh2 124 Iso 30 n = 12 K 55 Colh2 67 Iso 31 n = 12 K 98 Colh2 148 Iso 32a n = 6 K 49 Colh2 91 Iso 32b n = 8 K 59 Colh2 74 Cub12 85 Iso 32c n = 12 K 69 Cub12 126 Iso can very simply be changed by adjusting the alkyl chain length Phase transition temperatures (T /°C) of selected taper shaped poly- and the size of the polar groups.78 Thus, all main types of hydroxy amphiphiles 29,80b 30,60b 3177 and 32.77 aAdditional metastable inverted (type 2) lyotropic mesophases have been realised as liquid crystalline phases occur.thermotropic analogues in the absence of any solvent, simply by tuning the relative size of the lipophilic and the polar regions.cubic (CubV2) and columnar phases (Colh2), depending on the chain length and the size of the polar head group.18,78,80 Triple Interestingly, this adjustment of the interface curvature can also be realised by mixing amphiphiles with diVerent molecular chain compounds can form columnar (Colh2) and in some cases even micellar cubic mesophases consisting of closed shapes.18,76,78 As an example, the binary phase diagram of the single chain compound 36 and the triple chain compound 32c spheroidal micelles (CubI2).18,76–78 Often, the type of mesophase 1494 J.Mater. Chem., 1998, 8(7), 1485–1508Table 3 The influence of alkali metal triflates on the phase transition temperatures (T /°C) of the taper shaped polyether amphiphiles 37,86 3887 and 3989 Transition Transition temperatures of temperatures of the pure the 151 complexes No.Compound materials, T /°C with triflates, T /°C 37 K 54 Iso Li+: K 46.5 Colh 71 Iso 38 K 32 Iso Na+: K 69 (Col 27) Iso 39 K 56 Colh 63 Iso K 52 Colh 134 Iso O O OH O O C12H25O C12H25O C12H25O O O C12H25O C12H25O C12H25O O O O O O C12H25O O O O C12H25O C12H25O OH O O O O is shown in Fig. 10.78 In the contact region between the smectic group is available for hydrogen bonding (see Table 3). Here, a rather large number of aromatic rings with their polar ether phase of compound 36 (non-curved interfaces) and the micellar cubic phase of compound 32c (interfaces are curved in two oxygen atoms are gathered close to each other. Together with the polyether chains, they create a distinct polar region which dimensions) a broad region of a hexagonal columnar mesophase is induced.This can be generalised. For example, it is can segregate from the aliphatic chains. Again, the shape of the polymolecular aggregates can be also possible to obtain columnar mesophases by mixing structurally similar amphiphiles with bicontinuous (CubV2) and influenced by changing the relative size of the polar and lipophilic molecular parts.Thus, the methyl carboxylate 40 micellar cubic (CubI2) mesophases.18,76,78 Additionally, the size of the polar regions can be increased which has nine instead of only three lipophilic chains attached to the 3,4,5-tribenzyloxyphenyl core displays a monotropic by interaction with protic solvents giving rise to a wide variety of lyomesophases.18 cubic phase and the free acid 41 even has an enantiotropic cubic phase consisting of closed spheroidical micelles.90 These Remarkably, the mesomorphic properties of taper shaped amphiphiles are lost if the highly polar polyhydroxy groups molecules can be regarded as second generation dendrimers.The stability of the cubic phases was further increased on are replaced by less polar groups, such as carboxy groups,83 polyether chains or crown ethers.35,84,85 As shown in Table 3, going to the third generation dendrimers 42/43 and decreases again for the fourth generation dendrimers 44/45.90 The the oligo(oxyethylene) derivative 3786 for example is a crystalline solid.However, complexation of compound 37 with LiOTf model of their self-organisation is displayed in Fig. 12.90 The number of dendritic molecules forming the individual micelles results in a destabilisation of the crystalline phase and a spontaneous self-assembly into a cylindrical architecture which depends on the generation of the dendrons: 12, 6 and 2 monodendrons of generation 2, 3 and 4 respectively self displays a hexagonal columnar mesophase.86 The same result was found for the crown compound 38 on addition of sodium assembled to one micelle.The cubic mesophases belong to the Pm3n space group, exactly the same space group as found triflate.87 Introduction of ionic interactions eVected by the complexation of alkali metal ions by the polyether groups for the supermolecular aggregates of the much smaller but more polar triple chain carbohydrate and diol derivatives significantly increases the polarity contrast and probably also produces temporary crosslinks88 between the individual mol- such as 35, 32b and 32c.18,77,78 The dendrimers consist of large blocks and therefore a small segmental polarity contrast ecules.This enables micro-segregation to occur with formation of columnar supermolecular aggregates.35 is suYcient to give rise to the formation of supermolecular aggregates.In the 3,4,5-trialkoxybenzamides 35 and 32b,c the polarity contrast is much larger and therefore segregation can 6Wedge-shaped and cone-shaped dendrimers also occur with these much smaller molecules. In classical amphiphiles, the chemical diVerence represents Replacing the dodecyloxy chains of compound 37 (Table 3) by the 4-dodecyloxybenzyloxy chains increases the size of the a hydrophilic/lipophilic contrast.However in a more general picture amphiphilicity describes any chemical or structural polar regions of the taper shaped molecules and favours their micro-segregation. Compound 3989 for example forms a contrast within a molecule. Typical examples are hydrocarbon/ fluorocarbon systems,91 oligosiloxane/hydrocarbon92 systems columnar mesophase despite the fact that only one hydroxy J.Mater. Chem., 1998, 8(7), 1485–1508 1495Fig. 11 Liquid crystalline properties (T /°C) of the cone shaped dendrimers 40–45.90 aValues taken from the first cooling scan. 1496 J. Mater. Chem., 1998, 8(7), 1485–1508Fig. 12 Schematic presentation of the molecular self-organisation of cone shaped dendritic molecules to micellar cubic mesophases as adapted from ref. 90 (reprinted with permission from ref. 90, copyright American Chemical Society) and also the rigid-core/hydrocarbon systems of calamitic and disc-like liquid crystals. 7 Mesogenic compounds with perfluorinated segments It is possible to use the fluorophobic eVect, which is the incompatibility of fluorinated and non-fluorinated hydrocarbons to induce liquid crystallinity.91 However, there are some special properties of perfluorinated chains, which strongly influence their aggregation behaviour.91 First, the cross-section area of a perfluorinated chain is 27–35 A ° 2, which is large in comparison to alkyl chains (ca. 20 A ° 2) and the cross-section area of biphenyl moieties (ca. 22 A ° 2). Secondly, the perfluoroalkyl chain is more rigid.91,93 Partially fluorinated alkanes can be regarded as amphiphiles and as block compounds (see Fig. 13). The segregation of the COO RO CN OCO RO Br OC12H25 OC12H25 OCO COO OCO CH=C COOR COOR C8H17O OR RO RO OR OR OR 47a R = C10H21 K 40 N 157 Iso 47b R = C10F21 K 141 (Sc 138) SA 167 Iso 48a R = C12H25 K 120 Colh 125 Iso 48b R = (CH2)6C6F13 K 105 Col 137 Cub 171 Iso 49a R = C8H17 K 59 (Sc 53) SA 84 N 94 Iso 49b R = (CH2)2C6F13 K 116 Sc 153 SA 161 Iso 50a R = C5H11 K 69 Colh 122 Iso 50b R = CH2COO(CH2)2C6F13 K 30 Colh 210 Iso fluorinated and the non-fluorinated blocks of these molecules Selected examples for the influence of fluorination on the liquid into diVerent domains is themain driving force for their interesting crystalline properties (T /°C) of calamitic (47a vs. 47b)98 polycatenar (48a, 48b),13 swallow-tailed (49a,102b 49b102a) and disc-like (compound surfactant eVects.94 Thus, they can form normal and reversed 50a105b vs. 50b105) molecules micelles in perfluorinated solvents.95 When mixed with hydrocarbon solvents organised gel-like structure can be found.96 Furthermore, they represent the most simple thermotropic liquid Some calamitic compounds incorporating fluorinated crystals with smectic phases.97 However, the diVerent size of the segments (polyphilic molecules) have been synthesised in fluorocarbon segments and hydrocarbon segments can give rise order to use the micro-segregation to get non-chiral liquid to liquid crystalline phases with partial interdigitation of the alkyl crystals with ferroelectric properties (longitudinal ferrochains (LC1) or even to an antiparallel organisation of the electricity).106,107 molecules (LC2), an arrangement in which hydrocarbon chains Complete segregation was found for calamitic compounds and fluorocarbon chains are not segregated.97 with only one fluorinated chain and without a long alkyl Introduction of perfluorinated segments into classical chain.In the smectic E-phases of compounds 51103 and 52 104 calamitic liquid crystals93,98–104 and also into discotic liquid the rigid cores are completely interdigitated and ordered in crystals105a leads to a significant stabilisation of smectic and an orthorhombic cell, but the perfluorinated chains remain columnar mesophases, respectively. Nematic phases are liquid-like disordered (see Fig. 14). Here, the stress caused by generally suppressed. Also the mesomorphic properties of the diVerent space filling of the intercalated aromatic cores polycatenar compounds are significantly influenced by fluor- (2×22 A ° 2) and the non-intercalated perfluoralkyl chains ination.13,102 As in semifluorinated alkanes, the fluorophobic (27–36 A ° 2) is released by folding the perfluoralkyl chains.eVect is in competition with steric eVects due to the diVerent size of the hydrocarbon and fluorocarbon chains. Thus, in the case of calamitic compounds antiparallel pairs or modulated smectic phases (e.g. SA~, columnar phases) are formed in order to avoid this steric stress.93,99,102a Fig. 14 Mesomorphic properties of calamitic mesogens 51103 and 52104 Fig. 13 Liquid crystalline phases (transition temperatures T /°C) of partly fluorinated alkanes and model of their organisation in meso- incorporating fluorinated segments (transition temperatures T /°C) and model of their arrangement in the smectic layers phases97b (LC1 and LC2 represent smectic phases) J. Mater.Chem., 1998, 8(7), 1485–1508 1497COO—(CH2)2—C10F21 O2N COO—(CH2)2—C8F17 O2N C6H13O O—(CH2)4—C8F17 X 53a K 62 S 92 Iso 53b K 44 S 49 SA 51 Iso 54a K 72 SA 73 Iso X = CN 54b K 91 (SA 81) Iso X = CH2OH 54c K 69 SA 81 Iso X = COOCH3 54d K 67 Iso X = OCH3 Selected examples of fluorinated liquid crystalline compounds 53108b and 54108c incorporating only one phenyl ring (transition temperatures T /°C) Appropriate molecular design using the fluorophobic eVect also allows the formation of smectic liquid crystalline phases for molecules with a single aromatic ring108 which is a nonmesogenic rigid core in related nonamphiphilic hydrocarbon Fig. 15 Induction of columnar mesophases in the nonmesogenic taper compounds. Both micro-segregation and the rigidity of shaped molecules 55,86 57,87a 59,87a 6154b by the fluorophobic eVect perfluoralkyl chains should contribute to this eVect.Increasing (compounds 56,110 58,111 60,111 62;111 transition temperatures, T /°C) the chain-length and the amount of perfluorination increases and schematic presentation of the organisation of the partly fluorinated the mesophase stability. A micro-segregated bilayer arrange- taper-shaped molecules into columnar mesophases, adapted from ref. 110. aIso-Col transition obtained from the cooling scan. ment with a maximal overlap of the perfluorinated segments has been proposed.108c Using this concept mesogenic properties can be realised for simple alcohols (e.g. compound 54b), 8 Liquid crystalline oligosiloxanes which are non-mesogenic if combined with hydrocarbon residues.It seems however, that at least one polar substituent Polysiloxane side chain polymers represent an important family of liquid crystalline polymers.12,112 By appending a mesogenic at the aromatic ring is necessary for mesophase formation. Replacement of the COOCH3 group of compound 54c by the unit to a polysiloxane backbone, smectic phases are generally stabilised.less polar OCH3 group (compound 54d) for example causes the loss of mesogenic properties.108c Obviously, a certain More recently cyclooligosiloxanes,113 oligomeric114 and dimeric liquid crystals with siloxane units115,116 and calamitic degree of intramolecular contrast and/or a certain strength of the cohesive forces are necessary for mesophase formation in molecules with oligosiloxane end groups117 have also been synthesised.Selected examples of these compounds with 4- this class of compounds. Remarkably, no nematic phases have been found for these mesogens, which reveals that cyanobiphenyl rigid cores are collected in Table 4 and compared with the 4-cyano-4¾-hexyloxybiphenyl 63.118 In contrast micro-segregation is the main driving force for their mesogenity. to this conventional calamitic mesogen, which forms a nematic phase, all compounds containing at least two siloxane units A dramatic stabilisation of hexagonal columnar mesophases by semifluorination of the alkyl groups of taper-shaped amphi- exhibit exclusively SA-phases.It is especially remarkable that even the monomeric compound 68 is smectic despite the large philes was recently reported.109–111 In contrast to the hydrocarbon analogues of the same size, which are non-mesogenic, size of the hexamethyldisiloxane end-group.Usually, such large branched moieties give rise to a loss of mesogenic properties the fluorinated tapered molecules can self-assemble into tubular supramolecular architectures. The columns consist of a central or to a transition from smectic layer structures to less ordered nematic phases.The formation of a smectic layer structure in polar segment, the region of the lipophilic melted alkyl chains and the fluorocarbon-rich outer surface (see Fig. 15). the case of the oligosiloxane derivatives 64–68 can be explained on the basis of oligosiloxane/hydrocarbon incompatibility, Using the fluorophobic eVect columnar mesophases could also be obtained by rather small, non-dendritic taper shaped which gives rise to micro-segregation with layer formation.These smectogens can be regarded as triblock compounds crown compounds (compounds 58 and 60)111 and even for non-dendritic carboxylates such as 62.111 Their non-fluori- containing three distinct parts: an aromatic core, a paraYn chain and a siloxane sub-group.Incompatible with each other nated analogues are not liquid crystalline. Thus the fluorophobic eVect can be used in combination with other these parts tend to locate themselves in three separate sublayers superposed in a partially bilayered SA-structure as shown incompatibilities to increase the segregation tendency. In some cases the fluorophobic eVect alone can be suYciently in Fig. 16. A particularly eYcient micro-segregation was found for LC strong to induce molecular self-organisation replacing Hbonding and ion mediated complexation. Therefore it should polysiloxane copolymers with paired mesogens incorporating non-mesogenic dimethylsiloxane groups such as compound 69. be possible that less polar functional moieties (donor– acceptors, NLO compounds) can be used as part of the These copolymers show liquid crystalline behaviour up to a relatively high content of the nonmesogenic dimethylsiloxane cores of the supermolecular columns.111 1498 J.Mater. Chem., 1998, 8(7), 1485–1508Table 4 Comparison of the phase transition temperatures (T /°C) of 4- cyano-4¾-hexyloxybiphenyl 63118 with a related polysiloxane 64,112 the oligosiloxanes 65,116 66,114 67116 and the disiloxane substituted monomeric compound 68117 CN X–(CH2)6O No. X Transition temperatures, T /°C 63 K 57 N 75.5 Iso O O O O O O COO COO OCH3 OCH3 Si O CH3 Si O CH3 CH3 n m 69 64 g 14 SA 166 Iso 9 The importance of micro segregation for calamitic thermotropic mesogens The importance of the distribution of polar and non-polar structural units within calamitic mesogens for the formation 65 g 57 SA 118 Iso of nematic and smectic phases is a long known fact.120 As a general rule polar and aromatic rigid units should not be interrupted by aliphatic or cycloaliphatic segments.Compounds 70 and 71, for example, diVer exclusively in the position of the lipophilic cyclohexane ring.121 In the liquid crystalline compound 70 polar and lipophilic regions are well separated. In the nonmesogenic122 compound 70, however, the lipophilic cyclohexane ring separates the polar nitrile group 66 g-14.7 SA 118.7 Iso from the polar aromatic ring.Besides the molecular geometry and special conformational eVects123 micro-segregation should also contribute to this eVect. Generally, micro-segregation is favoured in molecules with a well-defined intramolecular contrast, i.e.in compounds with distinct polar/rigid and less polar/flexible segments. Here, 67 K 86 SA 91 Iso smectic liquid crystalline phases should be stabilised. If these diVerent regions are mixed up, the stability of the liquid crystalline phases decreases. 68 K 46 SA 53.8 Iso R CH3 O Si Si O Si O Si O CH3 R CH3 R R CH3 R CH3 O Si O O O Si Si Si Si R R CH3 R CH3 CH3 CH3 CH3 CH3 CH3 CH3 R R Si O Si R CH3 CH3 CH3 CH3 CH3 Si O Si R CH3 CH3 CH3 CH3 -(O-Si) n- C7H15 CN C7H15 CN 70 K 30 (S 17) N 59 Iso 71 K 34 [N -25] Iso Mesomorphic properties can also be lost if the rigid core is interrupted by flexible units.For example, no mesomorphic properties have been obtained for low molecular weight liquid crystals incorporating highly flexible mesogens such as diphenylethanes.These molecules can adopt a rod-like conformation and also bent conformations. However, main-chain polymers in which these flexible mesogenic units are connected via aliphatic spacers can form an interesting polymorphism (see Fig. 17).124 A hexagonal columnar phase is observed, with the columns extended in the chain direction and with no positional order along the columns.On lowering the temperature the hexagonal order remains, but segregation of the alkyl chains from the aromatic parts sets in and an additional Fig. 16 Organisation of oligosiloxane containing liquid crystals in the longitudinal order is established, which causes a transition into smectic layers (adapted from ref. 116) a smectic layer structure (SB-phase).It is often argued that the parallel organisation of the rigid cores of rod-like molecules is the main reason for their liquid units in relation to the paired mesogens. For these diluted copolymers a distinct micro biphasic behaviour has been crystallinity. However, only the formation of nematic phases is strictly bound to the rigid anisometric shape of the individual observed with two distinct glass transitions.119 Some of these copolymers can take up significant amounts of decamethyl- molecules.The formation of smectic layers is mainly driven by micro-segregation. Indeed, the rigid calamitic units provide an tetrasiloxane, which can be incorporated in the polysiloxane sublayer119 in a way analogous to the way water is incorporated intramolecular contrast to the flexible aliphatic chains usually attached to them.It is well known that smectic phases are between the polar groups of detergents to give lyotropic mesophases. stabilised on elongation of the terminal aliphatic chains. J. Mater. Chem., 1998, 8(7), 1485–1508 1499Fig. 18 Transition temperatures (T /°C) and model of the smectic Aphase of the triptycene derivative 74127 corresponding taper shaped 1,2-disubstituted molecules (e.g.compound 26). Monocyclic troponoids (compound 73a) which have two alkyl chains, at least one of them connected via a carboxy group with the central tropone core also show monotropic SAphases. 125 Here, the polar tropone unit together with the linking groups form a distinct polar region, which segregates from the aliphatic chains.The SA-phases can be further stabilised by the fluorophobic eVect (compound 73b).126 The recently described triptycene derivative 74127 represents another nice example of a non-classical smectic liquid crystal Fig. 17 Model of the SB-phase and hexagonal columnar mesophase of (‘epitaxygen’). In this compound, five aliphatic chains were main chain polymers with flexible mesogenic units124 grafted laterally to a bulky triptycene core in such a way that they are pre-organised parallel to each other.This compound Starting with a certain length of the flexible aliphatic chains, displays several smectic mesophases with unexpected high the rigid cores and aliphatic chains segregate into diVerent transition temperatures.The segregation of the aromatic tripregions with formation of interfaces. Due to the approximately tycene core from the aliphatic chains and additional lamellar equal diameter of these two units in most cases non-curved pre-organisation of the aliphatic chains due to the topology of aggregates (smectic layers) result. Additionally, the topology their connection with the triptycene unit are probably the of the connection of the flexible chains to the ends of the rigid most important driving forces for mesophase formation.units provides a distinct pre-orientation of these chains Another point which should be discussed here is the fact and in this way the formation of smectic layers is additionally that, due to the preferred parallel arrangement of the rigid, favoured.However, a calamitic rigid core is not necessary for rod-like units, the formation of curved interfaces is strongly a layer-like organisation. The lamellar phases of (single chain) hindered in the rigid calamitic mesogens. Thus, mainly layer polyhydroxy amphiphiles and ionic amphiphiles are well structures can be found. Small diVerences between the space known examples for smectic phases formed by amphiphilic filling of the rigid cores and the pendant terminal chains can molecules without rigid cores.be compensated for by tilting the aromatic cores with respect to the layer normal, which gives rise to tilted smectic phases (e.g. SC-phases).128 In fact, fluid tilted phases (SC-phases) have never been found in amphiphilic liquid crystals without rigid cores.O O C12H25O C11H23 O O O C10F21CH2CH2O C11H23 O 73a K 58 (SA 45) Iso 73b K 101 SA 127 Iso OH HO HO H13C6O OC6H13 OH 72 K 144.8 SA 176.9 I Furthermore, appropriate molecular design using microsegregation also allows the formation of smectic phases for molecules with a single aromatic or aliphatic ring which themselves do not represent mesogenic units. An example of a smectic compound with only one ring is provided by the O N N H H O C8H17O OC8H17 COOH C16H33O O2N O C12H25O O OH OH 75 K 136.7 Cub 161 SC 165 Iso 76 K 126.8 SC 171 Cub 197.2 SA 201.9 Iso 77 K 140 Col 146 SA 169 Iso myo-inositol ether 72.73b Here, the hydroxy groups form a polar region which segregates from the lipophilic chains.In rare cases cubic129,130 and columnar131 phases can also be found for compounds with a rod-like shape (e.g. 75–77), in Additionally, attractive forces are provided by hydrogen bonding which stabilise this arrangement. The 1,4-disubstitu- many cases accompanied by SC-phases. Interestingly, all these compounds incorporate highly polar groups in the core region tion pattern provides a linear pre-organisation of the alkyl chains. Therefore this compound exhibits a smectic A-phase or close to it and they have long alkyl chains. This probably enables a strong segregation and an especially eYcient packing of the monolayer type instead of the columnar phase of the 1500 J.Mater. Chem., 1998, 8(7), 1485–1508Table 5 Mesomorphic properties (T /°C) of representative homologous Table 6 Liquid crystalline properties (T /°C) of rod-coil molecules 81139 and 82137 and the influence of lithium triflate on the mesogenic polycatenar compounds 78a-c13a properties of compound 81139 R COO COOC2H5 No.R Transition temperatures, T /°C No. R Transition temperatures, T /°C 78a OC9H19 K 146.5 SC 268.5 Iso 81 K 135.2 SA 148.1 Iso 78b OC11H23 K 144 SC 146 Cub 163 (Colh 158) Iso 78c OC13H27 K 141 Colh 163 Iso O CH3O 12 +0.2 eq.Li+OTf- K 93.3 Cub 131.5 SA 154.1 Iso of the cores due to cohesive forces in the polar regions. A +0.25 eq. Li+OTf- K 92.8 Cub 137.2 Iso diVerent space filling of the segregated regions may result and +0.3 eq. Li+OTf- K 87.8 Cub 120 Col 126.2 Iso +0.4 eq. Li+OTf- K 77.1 Col 120.5 Iso the smectic layers probably collapse with formation of ribbonlike segments to realise curved interfaces.Thus, the cubic 82 K 21.5 Col 33.5 Iso mesophases could represent interwoven networks of these ribbons (CubV-phases). O CH3O 12 10 Polycatenar compounds One way to obtain curved aggregates in mesophases of nontration. In these cases the layers break up into ribbon-like polar rigid calamitic molecules involves increasing the number aggregates which can organise to bicontinuous cubic and of terminal aliphatic chains.Therefore, molecules with three, columnar mesophases. Thus, smectic C, cubic (Im3m and four, five or even six aliphatic chains attached to the termini Ia3d)16 and columnar phases were found for tetracatenar of an extended rigid core, the so-called polycatenar comcompounds (Table 5), whereas exclusively columnar meso- pounds,13 and molecules with branched terminal chains, the phases were observed for penta- and hexa-catenar molecules.swallow-tailed compounds,132 have been synthesised. However, It should be pointed out that the same diversity of diVerent in this class of compounds mesophases can only be obtained mesophases as known for thermotropic and lyotropic phases if the rigid parts are large enough to allow eYcient segregation of strongly amphiphilic molecules is found for these com- between the rigid units and the aliphatic chains.Furthermore, pounds. Only the formation of a micellar cubic mesophase has extended rigid cores enable suYciently strong attractive internever been proven for polycatenar compounds. Because the actions, which favour a parallel organisation.For example, formation of closed micelles requires a very large interface molecules with five or six chains require a minimum of five curvature, the occurrence of this type of mesophases seems and six rings, respectively. unlikely in this class of compounds. The length of the rigid unit can be reduced, if additional attractive interactions are introduced and the intramolecular contrast is increased.Thus, columnar mesophases have been 11 Rod-coil molecules found for polycatenar tetrone derivatives133 and for poly- It is well known from the literature on low-molecular mass catenar ionic liquid crystals134–136 (as for example compounds calamitic liquid crystals that the replacement of paraYn chains 79133 and 80136) which contain significantly shorter but more by oligo(oxyethylene) segments destabilises the thermotropic polar rigid cores.mesophases.11a The main diVerences with aliphatic chains are primarily the polarity of the oligo(oxyethylene) chains and secondly their higher flexibility and the preferred gauche conformation of the MOCH2CH2OM units. Therefore, these chains can often adopt a helical conformation and occupy a significantly larger mean lateral area per chain than comparable aliphatic chains of the same length.Thus, smectic phases are strongly disturbed. As in polycatenar compounds, mesomorphic properties can only be obtained by molecules with very large rigid segments. Recently, such rod-coil molecules have been synthesised. They consist of a large calamitic unit, connected to an extended flexible oligo(oxyethylene) chain.137–139 As a result of the microsegregation of the lipophilic and rigid cores from the flexible and polar polyether chains, these rod-coil molecules self assemble into well-defined microstructures, such as lamellar and columnar mesophases depending on the block composition N N O O O O OC12H25 OC12H25 OC12H25 C12H25O C12H25O OC12H25 S S O OC11H23 OC11H23 C6H13 CH3 + 79 K 64 Colr 112 Colh 133 Iso 80 K 139 (Colr 135) Colh 154 Ncol 178 Iso BF4 – (see Table 6).Compound 81 for example forms a smectic A-phase which Triple chain polycatenar compounds usually display nematic, smectic C-phases and other smectic low-temperature on addition of lithium triflate turns via an intermediate bicontinuous cubic mesophase into a columnar mesophase.138 These mesophases. Remarkably, no SA-phases have been observed for pure polycatenar compounds with non-polar chains.The phase transitions should be caused by the larger spatial requirement of the complexed coil segments in comparison reason could be that small diVerences in space filling between the rigid cores and the pendant terminal chains can solely be with the non-complexed.Alternatively, the transition from a smectic to a columnar structure can be achieved by enlarging compensated for by tilting the aromatic cores with respect to the layer normal. If, however, a certain diVerence in space the coil segments by introduction of additional substituents in the polyether chain [e.g. the methyl groups in the oligo(propy- filling of the segregated regions is reached, tilting of the molecules alone cannot suYciently reduce this steric frus- lene oxide) derivative 82].137 J.Mater. Chem., 1998, 8(7), 1485–1508 1501Fig. 19 Organisation of rod-coil molecules in the smectic and in the columnar mesophase (induced by LiOTf )124 12 Laterally substituted calamitic mesogens It is well known that the connection of non-polar aliphatic chains to a calamitic rigid core at positions other then the terminal ones decreases the smectic mesophase stability (see Fig. 20).120,140 However, if polar groups are introduced in a lateral position Fig. 21 Comparison of the thermotropic behaviour of methyl 2,5- bis(4-decyloxyphenyl)benzoate 86 with the 2,5-bis(4-decyloxyphenyl)- on a calamitic mesogen, then the smectic phases can be benzamides 87–90 with diVerent sized lateral groups;142 black areas stabilised, because these substituents can cause attractive interrepresent the crystalline state molecular interactions and can increase the intramolecular polarity contrast between a central polar core and the lipophilic chains.The earliest example has been provided by the laterally of possible hydrogen bonds raises the SA-isotropic transition substituted naphthoic acids 84 and 85.141 The smectic tendency temperature despite the fact that the size of the lateral substitof the chloro-substituted naphthoic acid 85 is remarkably high, uent is significantly enlarged.The smectic layers should be whereas the naphthoic acid without the lateral chloro substitu- stabilised by both the attractive hydrogen bonding and the ent 84 is only a nematic liquid crystal.increased intramolecular polarity contrast between the central rigid cores and the terminal lipophilic chains. If the size of these polar lateral groups however is increased without simultaneously increasing the number of attractive hydrogen bonds, the smectic mesophases are destabilised. This is shown in Fig. 22 for the terphenyl derivatives with lateral polyether chains. Surprisingly, however, starting with the C8H17O X COOH 84 X = H K 161.5 N 190 Iso 85 X = Cl K 169 SC 183 N 199 Iso bis(oxyethylene) compound 93 a rectangular columnar mesophase can be observed.143 A ribbon model was proposed for Other examples are given by p-terphenyl derivatives with these columnar mesophases, which can be explained as follows.lateral groups capable of hydrogen bonding.142 The compari- In the series of compounds 91–94, the flexible lateral groups son given in Fig. 21 clearly reveals this. Increasing the number become more and more polar by elongation of the polyether chain. At a certain size and a certain degree of polarity the flexible and polar lateral groups start to segregate from the rigid and lipophilic terphenyl units into separate domains. This causes the smectic layers to collapse incompletely as the layers break up and form ribbons.These ribbons become ordered in a two-dimensional rectangular lattice. A possible model143 is shown in Fig. 23. The ribbons consist of parallel rigid p-terphenyl cores which Fig. 22 Thermotropic liquid crystalline phases of the 2,5-bis(4- Fig. 20 Dependence of the thermotropic clearing temperatures of decyloxyphenyl)benzyl ether 91–94 with diVerent length of the lateral oligo(oxyethylene) chains;143 black areas represent the crystalline state; laterally substituted hydroquinone dibenzoates (T /°C) on the length of the lateral alkyl chain (adapted from ref. 140) the mesophases of compounds 93 and 94 are monotropic 1502 J.Mater. Chem., 1998, 8(7), 1485–1508Fig. 23 Schematic illustration of the SA-phase and the ribbon arrangement for the rectangular columnar phase (Colr) of compound 93.143 The black areas represent the segregated regions of the polar groups. The polar groups of the molecules in the middle of the ribbons cross over the neighbouring calamitic terphenyl units.are laterally separated by the hydrophilic domains of the lateral polyether chains. The alkyl chains are molten and fill up the space between the ribbons in the other dimension. This arrangement seems reasonable because it simultaneously COO OC12H25 C8H17O COO OC12H25 C8H17O O O2N O S N N C8H17O OC8 H17 O O2N O2N NO2 S N N C8H17O OC8H17 + 0.2 TNF 95 N 88.5 Iso 96 K 65 SA 91 Iso 97 K 101 SC 195 N 198 Iso 97 + 0.2 TNF: K ? SA 200 Iso Selected examples of mesophase stabilisation (T /°C) by polar lateral enables the parallel organisation of the rod-like molecules and substituents (nonsubstituted compound 95 vs. 96) and by doping a also the best segregation of polar and lipophilic units into 2,5-diphenylthiadiazole derivative (compound 97) with the nonmeso- separate regions.Thus, the mesophase should be composed of genic electron acceptor TNF an alternating structure of diVerent lamellae. The regions of the molten alkyl chains form one type of lamella. The other ring in the lateral branch (e.g. compound 96) exhibit enant- one is composed of ribbons of rigid p-terphenyl units laterally iotropic smectic A phases. The mesophase thermal stability of separated by cylinders containing the hydrophilic groups.The these strongly branched derivatives can be higher than that of proposed model is related to that suggested for supermolecular the laterally unsubstituted parent compounds (e.g. compound structures which have recently been found in triblock copoly- 95) and nematic phases are replaced by SA-phases.148 mers consisting of three linearly combined blocks.144 Here, Furthermore smectic A-phases can be induced by addition of besides other superstructures, lamellae with cylinders at the the nonmesogenic electron acceptor trinitrofluorenone (TNF) lamellar interfaces have been found.145 Indeed, the molecules to the nematic phases of calamitic molecules such as 97.149 93 and 94 can be regarded as low molecular weight three- Both attractive interactions provided by EDA-interactions and block compounds composed of three diVerent and incompatattractive quadrupole interactions149b,150 and also the increased ible blocks: the rigid cores, the flexible and lipophilic alkyl micro-segregation due to the enhanced polarity in the regions chains and the flexible but polar polyether chains.In contrast of the calamitic parts should contribute co-operatively to the to linear triblock copolymers these Y-shaped block molecules observed mesophase stabilisation.In the systems 97/TNF the have the cylinders located within one of the distinct layers length of the alkyl chains is of unusually great importance, instead of being located at their interface. whereby mesophase stabilisation is only found for long chain Protic solvents such as water and formamide can stabilise compounds.This can be explained by the fact that segregation the columnar mesophases of compounds 93 and 94 and can is increased on increasing the size of the competing parts. induce novel columnar mesophases. For example, with forma- Furthermore, the larger space requirement of the ‘mixed’ mide a columnar phase can be induced in the smectic phase (compound 97+TNF) polar regions necessitates a better space of compound 92.Up to three diVerent rectangular columnar filling of the aliphatic regions by longer or branched149b,151 mesophases have been found for compound 94 with a long alkyl chains. lateral polyether chain on increasing the water concentration. 131b,146 The increase in size and polarity of the lateral groups by hydration should facilitate micro-segregation and 13 Influence of micro-segregation on mesophases of thus should be responsible for the stabilisation and/or induc- disc-like molecules tion of the columnar mesophases. Likewise, columnar lyomesophases have been induced by specific recognition of alkali It seems that micro-segregation with creation of an interface between polar and non-polar regions is also one of the main metal ions by facial amphiphiles consisting of a 4,4-didecyloxy terphenyl rigid core with laterally appended crown ether driving forces for the appearance of liquid crystalline properties in disc-like compounds.units.146,147 Layer-like arrangements of calamitic molecules can also be It has often been observed that columnar mesophases are lost, if polar groups in the centre of disc-like molecules are stabilised by electron donor–acceptor (EDA) interactions between the rigid cores.For example, long chain derivatives replaced by less polar ones.34 For example, the inositol ethers and esters 99, 100152 and the hexakis(alkylsulfono)benzene of two-ring mesogens having an acceptor substituted phenyl J.Mater. Chem., 1998, 8(7), 1485–1508 1503Fig. 24 Selected examples of more complex block copolymers be applied to multiblock copolymers, graft-copolymers, combcopolymers, star-block copolymers etc. In the case of triblock copolymers144 even more complicated supermolecular structures which are a combination of lamellae, cylinders and spheres such as helical morphologies,158 lamellar structures with cylinders at the lamellar interface145 and the ball at the wall structure159 have been found. Interesting new morphologies can also be observed if the X X X X X X X X X X X X 98 X = CH2, m.p. 66 99 X = O K 18.4 Colho 90.8 Iso 100 X = COO K 68.5 Colho 119.5 Iso 101 X = S liquid 102 X = SO2 K 120 Col 138 Iso chain rigidity is increased for one block (rod-coil block copoly- The influence of the polarity in the central core of disc-like molecules mers).160 In poly(hexyl isocyanate-b-styrene) diblock polymers on the columnar mesophase stability.Increasing the polarity smectic C-like and smectic O-like morphologies have been (98<99<100;152 101<102154,155) enhances the clearing temperatures detected.161 The morphology of these materials arises from the competition of the coil and rod blocks into mesoscopically 102153,154 are mesomorphic compounds in contrast to the allordered periodic structures and the tendency of the rod blocks trans-hexaalkylcyclohexane 98155 and the hexakis(alkylthio)- to form anisotropic orientationally ordered structures.Many benzene 101 which are non-mesogenic.Also it is well known classical calamitic liquid crystals can be regarded as block that only acylated azamacrocyles (see section 2) are mesomolecules, i.e. as low molecular weight analogoes of rod-coil morphic44 whereas the corresponding alkyl derivatives are block copolymers. Interesting, new liquid crystalline phases not156 and this should largely be due to the higher polarity of could also be expected for low molecular weight block com- the amide groups in comparison to the amino groups.This pounds, if rigid structures are combined with two or even means that in these cases the micro-segregation also helps the more diVerent and incompatible chains in diVerent positions. columnar supermolecular structures to develop. This eVect is First examples are provided by partly fluorinated liquid crystals especially important if the central core is rather small.A (section 7), polyether amphiphiles with rod-like units (com- suYcient intramolecular polarity contrast and an eYcient pound 77) and calamitic compounds with laterally attached surrounding of the central cores by aliphatic chains are then polyether chains (section 12).necessary to obtain mesomorphic properties. Nevertheless Block and graft copolymers consisting of a coil block and a conformational eVects and the rigidification of the central liquid crystal side chain polymer block have recently been regions may also contribute to this eVect. However, we will reviewed.162 These LC-coil diblock copolymers combine meso- not discuss this topic further because it will be incorporated morphic order with micro phase separation in a single polymer.in a feature article concerning columnar mesomorphism.43 It Here molecular ordering occurs on two diVerent length scales is only pointed out that columnar ordering is not preliminarily within one system. This oVers unique possibilities to obtain caused by the flat disc-like shape of molecules, rather than by new self-organised ordered systems in nanoscopic dimensions.micro-segregation of the rigid cores from the flexible chains, Thus, the LC-subphase can influence the morphology of by attractive core–core interactions and by the topology of the polymer system,163 and on the other hand the block connection of the lipophilic chains with the central unit.Thus morphology can stabilise the liquid crystalline subphases.164 columnar mesomorphism is not restricted to disc-like com- Completely new supermolecular structures, such as flat pounds. Columnar phases can also be found for molecules cylinders and hexagonal shaped cells, are also possible.165 with non-discoid central cores such as tetrahedral pentaerythritol tetrabenzoates (compounds 7) and octahedral metallomesogens (compound 13).However, large flat and rigid cores can 15 Influence of chirality give rise to significantly attractive core–core interactions. It is well known that molecular chirality can give rise to helical Furthermore, they can significantly increase the intramolecular supermolecular structures.166 This is not restricted to nematic contrast and eYciently pre-organise the chains radially around phases which are turned into cholesteric phases.Chirality can this core. also modify positionally ordered mesophases by distortion of the interfaces. Examples are the twisted grain boundary (TGB) 14 Mesophases of more complex block compounds phases, the SQ-phases (tetragonal phases)167 and optically isotropic mesophases which represent distorted layer structures Linear AB-diblock copolymers represent the most simple and best investigated block compounds.Related concepts can also of calamitic compounds.166a Interestingly, a TGB phase has 1504 J. Mater. Chem., 1998, 8(7), 1485–1508Table 7 Mesomorphic properties of complexes between the glycouril derivative 103 and selected resorcinol derivatives171 Fig. 25 The main driving forces of liquid crystallinity N N Ph O O O O O C12H25O C12H25O OC12H25 OC12H25 OC12H25 OC12H25 C12H25O C12H25O OC12H25 OC12H25 OC12H25 OC12H25 O O O O N O Ph N 103 also been found for a chiral block copolymer.168 However, in Host-guest Transition temperatures, systems with strongly competing parts, such as lyotropic and Guest ratio T /°C thermotropic mesophases of amphiphilic molecules no influence of chirality on the structure of fluid positional ordered none K 159–164 (N 161) Iso mesophases has been detected yet.169 For example no influence 1053 K-3 S1 82 S2 131 Iso of chirality on the mesophases of amphiphilic carbohydrates 151 K-4 N 110 Iso and other chiral polyhydroxy amphiphiles could be measured. 63 Due to the strong competition between the antagonistic parts larger forces probably must be applied to aVect their aggregate structure.In systems with a less pronounced intramolecular contrast (mesophases of nonamphiphilic but anisometric mesogens and weakly segregated block copolymers) 151 K 30 Col 141 Iso even small changes can easily disturb the mesophase structure and also molecular chirality can aVect the self organisation. 16 Summary In summary, the unique properties of liquid crystals arise from HO OH COOCH3 HO OH O O OH x y z the combination of order and mobility on a molecular level.The mobility is caused by the thermal motion of the molecules, whereas the order is provided by a combination of molecular anisometry, attractive forces and micro-segregation. Depending copolymers or very hazy (sinusoidal) as for example in the case of SA and SC phases of calamitic compounds and weakly on the particular molecular structure one of these ordering forces can dominate. For example the nematic state is mainly segregated block copolymers. This depends on the degree of competition between the segregated regions.Weakly segregated caused by the anisometry of molecules or aggregates.Lyotropic systems of detergents or lipids are predominante organised by systems can easily be disturbed by chirality, by changing the molecular structure and by addition of solvents. Strongly micro-segregation and attractive forces, whereas micro-segregation alone can be responsible for the mesophase formation segregated systems are significantly less sensitive.In many cases one of the segregated regions is liquid like (aliphatic of block copolymers and also for the mesomorphic properties of low molecular weight block molecules. chains or solvent) and the other one is more or less ordered or even crystalline (parallel arranged discs or calamitic units, In this paper we have focused our interest on positionally ordered mesophases, smectic, columnar and cubic.This paper quasi-crystalline layers of the ionic groups in the ribbon phases of pure soaps, etc.).32,170 This additional order gives rise to a provides facts that establish bridges between the polymorphism of amphiphiles, rod-like molecules, disk-like molecules and large number of diVerent subphases as for example smectic low-temperature phases.1 block copolymers.All molecules dealt with in this article are amphiphilic. The If the molecules are rigid the formation of flat (calamitic molecules) or cylindrical (disc-like molecules) interfaces is segregation of chemically diVerent units (rigid, flexible, aliphatic, polar, perfluorinated, oligosiloxanes, etc.) leads to the strongly favoured. When the molecules are flexible (amphiphiles, block-copolymers), then a wide variety of diVerent formation of interfaces separating two or even more media.This micro-segregation depends on the degree of chemical and mesophases can be found ( lamellar, columnar, cubic and diVerent intermediate phases). In most nonlamellar thermo- structural diVerences and the size of the diVerent building blocks. If the diVerent units are very large as in block copoly- tropic mesophases of low molecular weight compounds the stronger cohesive forces are located inside the aggregates mers even very small diVerences in chemical structure give rise to micro-segregation at ambient temperature.If the blocks are surrounded by lipophilic chains. It would be of interest to synthesise novel molecules which can form aggregates with the small, the chemical diVerences between blocks have to be increased. stronger attractive forces outside, as already realised in block copolymers and as is well known from the type 1 lyotropic At the same time, the molecules are mesogenic because they build up structures intermediate between disordered liquids surfactant–water systems.The first compound of this type is probably the carbohydrate derivative 21 (Table 2).Interesting and highly ordered crystals. The main feature of these mesophases is that they consist of chemically diVerent regions new liquid crystalline phases could also be expected for novel low molecular weight block compounds which combine micro- separated in space by interfaces: planar interfaces in smectic phases, cylindrical interfaces in columnar phases and three- segregation and molecular anisometry.It can be expected that many more mesomorphic compounds dimensionally bent interfaces in the cubic phases and other 3D-mesophases. The interfaces can either be well defined as in with non-conventional molecular structures should be detected in the future. One example is represented by the mesomorphic most lyotropic systems and in strongly segregated block J.Mater. Chem., 1998, 8(7), 1485–1508 1505317; (d) F. S. Bates and G. H. Fredrickson, Annu. Rev. Phys. glycoluril derivatives (e.g. compound 103, see Table 7) and Chem., 1990, 41, 525. (e) S. Fo� rstner, Ber. Bunsenges. Phys. Chem., their complexes with resorcinol derivatives.171 In this case 1997, 101, 1671. micro-segregation should also be a main driving force for self- 23 L.Leibler,Macromolecules, 1980, 13, 1602. organisation. 24 Block Copolymers: Science and T echnology, ed. D. J. Meier, IMM Micro-segregation is also important in other fields. Thus, Press/Harwood Academic, New York, 1983; D. J. Meier, J. Polym. Sci., C, 1969, 26, 81. polymerisation reactions were found to be more facile in 25 These are only approximate values which vary depending on the ordered melts in which the reactive moieties of the monomers chemical structure.are segregated and create a supramolecular reactor of an 26 I. W. Hamley, K. A. Koppi, J. H. Rosedale, F. S. Bates, K. Almdal increased concentration of polymerisable groups.172 and K. Mortensen,Macromolecules, 1993, 26, 5959. As only one example from the life sciences it should be 27 D.A. Hajduk, P. E. Harper, S. M. Gruner, C. C. Honeker, mentioned that micro-segregation represents a fundamental G. Kim, E. L. Thomas and L. J. Fetters, Macromolecules, 1994, 27, 4063. driving force in protein folding. 28 F. S. Bates, M. F. Schulz, A. K. Khandpur, S. Fo� rster, J. H. Rosedale, K. Almdal and K. Mortensen, Faraday Discuss., Thanks are due to G.Lattermann, (University Bayreuth) for Chem. Soc., 1994, 98, 7. helpful criticism of the manuscript. Work was supported by 29 A. K. Khandpur, S. Fo� rster, F. S. Bates, I. W. Hamley, A. J. Ryan, the Kultusministerium Sachsen Anhalt and the Deutsche W. Bras, K. Almdal and K. Mortenson, Macromolecules, 1995, Forschungsgemeirischaft. 28, 8796. 30 P. Alexandridis, U.Olsson and B. Lindman, L angmuir, 1997, 13, 23. References 31 Y. Hendrikx and A. M. Levelut, Mol. Cryst. L iq. Cryst., 1988, 165, 233. 1 Reviews: (a) H. Stegemeyer, T opics in Physical Chemistry, vol. 3, 32 J. Charvolin, J. Chim. Phys., 1983, 80, 15. L iquid Crystals, Steinkopf Darmstadt, Springer, New York, 1994; 33 A. G. Petrov and A. Derzhanski, Mol. Cryst. L iq. Cryst., 1987, (b) G.W. Gray and J. W. Goodby, Smectic L iquid Crystals, 151, 303. Leonard Hill, Glasgow, London, 19 Micro-segregation was considered to contribute to columnar 2 There is also a dependence on pressure: S. Urban and mesophase formation of trans-hexasubstituted cyclohexane A.Wu� rflinger, Adv. Chem. Phys., vol. XCVIII, ed. I. Prigogine and derivatives: K. Praefcke, B.Kohne, P. Psaras and J. Hempel, S. A. Rice, John Wiley, New York 1997, p. 143. J. Carbohydr. Res., 1991, 10, 523. 3 B. Bahadur, L iquid Crystal—Applications and Uses, vol. 1–3, 35 V. Percec, J. Heck, G. Johansson, D. Tomazos, M. Kawasumi World Scientific Publishing Co. Pte. Ltd., Singapore, 1990–1992. and G. Ungar, J. Macromol. Sci., Pure Appl. Chem., 1994, A31, 4 T. J. Bunning and F.-H.Kreuzer, T rends Polym. Sci., 1995, 3, 318. 1031. 5 F. Schambil, S. Paasch and M. J. Schwuger, Fette O� le Fettderivate 36 U. Stebani and G. Lattermann, Macromol. Rep., 1995, A32 Folgeprodukte, 1988, 114, 295. (Suppl. 3) 385. 6 C.G.Go� ltner and M. Antonietti, Adv.Mater., 1997, 9, 431. 37 A. Pegenau, P. Go�ring and C. Tschierske, Chem. Commun., 1996, 7 H. Ringsdorf, B. Schlarb and J.Venzmer, Angew. Chem., 1988, 2563. 100, 117; Angew. Chem., Int. Ed. Engl., 1988, 27, 113. 38 J. H. Cameron, A. Facher, G. Lattermann and S. Diele, Adv. 8 F. Livolant and A. Leforestier, Mol. Cryst. L iq. Cryst., 1992, 215, Mater., 1997, 9, 398. 47. 39 U. Stebani, G. Lattermann, M. Wittenberg and J. H. WendorV, 9 J. D. Bernal and I. Fankuchen, Nature 1937, 139, 923. J.Mater. Chem., 1997, 7, 607. 10 Reviews: (a) G. J. T. Tiddy, Phys. Rep., 1988, 57, 1; (b) 40 K. Praefcke, B. Bilgin, N. Usoltseva, B. Heinrich and D. Guillon, J. M. Seddon and R. H. Templer, in Handbook of Biological J.Mater. Chem., 1995, 5, 2257. Physics, ed. R. Lipowsky and E. Sackmann, vol. 1, Elsevier, 41 U. Stebani, G. Lattermann, R. Festtag, M. Wittenberg, and Amsterdam, p. 1995, 97.J. H. WendorV, J.Mater. Chem., 1995, 5, 2247. 11 Reviews: (a) C. Tschierske, Progr. Polym. Sci., 1996, 21, 775; (b) 42 The descriptors V and I have been introduced by Tiddy10a and M. Paleos, Mol. Cryst. L iq. Cryst., 1994, 243, 159. are widely used in lyotropic systems to diVerentiate between 12 Review: V. Percec and C. Pugh, in Side Chain L iquid Crystals, ed. bicontinuous and discontinuous cubic phases, especially, if the C.B. McArdle, Blackie, Glasgow, 1989, p. 30. cubic lattice is unknown. 13 Review: (a) H. T. Nguyen, C. Destrade and J. Malthete, Adv. 43 G. Lattermann, J.Mater. Chem., in preparation. Mater., 1997, 9, 375; (b) J. Malthete, H. T. Nguyen and 44 (a) J. M. Lehn, J. Malthete and A. M. Levelut, J. Chem. Soc., C. Destrade, L iq. Cryst., 1993, 13, 171.Chem. Commun., 1985, 1794; (b) G. Lattermann, Mol. Cryst. L iq. 14 P. A. Winsor, in L iquid Crystals and Plastic Crystals, ed. Cryst., 1990, 182B, 299; (c) G. Lattermann, L iq. Cryst., 1989, 6, G. W. Gray and P. A. Winsor, Horwood, Chichester, 1974, 619; (d) C. Mertensdorf and H. Ringsdorf, L iq. Cryst., 1989, 5, vol. 1, p. 48. 1757; (e) S. H. J. Idziak, N. C. Maliszewskyj, P.A. Heiney, 15 F. Hardouin, A. M. Levelut, M. F. Achard and G. Sigaud, J. Chim. J. P. McCauley, Jr., P. A. Sprengeler and A. B. Smith, III, J. Am. Phys., 1983, 80, 53. Chem. Soc., 1991, 113, 7666; ( f ) S. H. J. Idziak, N. C. 16 (a) A. M. Levelut and Y. Fang, Coll. Phys. C7, 1990, 51, 229; Maliszewskyj, G. B. M. Vaughan, P. A. Heiney, C. Mertensdorf, (b) B. Donnio, B. Heinrich, T.Gulik-Krzywicki, H. Delo� acroix, H. Ringsdorf, J. P. McCauley, Jr. and A. B. Smith, III, J. Chem. D. Guillon and D. W. Bruce, Chem. Mater., 1997, 9, 2951; Soc, Chem. Commun., 1992, 98; (g) J. Malthete, D. Poupinet, R. (c) A.M. Levelut and M. Clerc, L iq. Cryst., 1998, 24, 105. Vilanove and J. M. Lehn, J. Chem. Soc., Chem. Commun. 1989, 17 H. Sakashita, A. Nishitani, Y. Sumiya, H.Terauchi, K. Ohta and 1016; (h) M. Zhao, W. T. Ford, S. H. J. Idziak, N. C. Maliszewsky I. Yamamoto, Mol. Cryst. L iq. Cryst., 1988, 163, 211. and P. A. Heiney, L iq. Cryst., 1994, 4, 583. 18 K. Borisch, S. Diele, P. Go� ring, H. Mu� ller and C. Tschierske, L iq. 45 H. Fischer, S. S. Ghosh, P. A. Heiny, N. C. Maliszewskyj, Cryst., 1997, 22, 427. T. Plesnivy, H. Ringsdorf and M.Seitz, Angew. Chem., 1995, 107, 19 (a) J. M. Seddon and R. H. Templer, Phil. T rans. R. Soc. L ond., 879; Angew. Chem., Int. Ed. Engl., 1995, 34, 795. 1993, A 344, 377; (b) K. Fontell, Colloid Polym. Sci. 1990, 268, 264; 46 U. Stebani, G. Lattermann, M. Wittenberg, R. Festtag and (c) P. Sakya, J. M. Seddon, R. H. Templer, R. J. Mirkin and J. H. WendorV, Adv.Mater., 1994, 6, 572.G. J. T. Tiddy, L angmuir, 1997, 13, 3706. 47 M. Seitz, T. Plesnivy, K. Schimossek, M. Edelmann, H. Ringsdorf, 20 J. N. Israelachvili, D. J. Mitchell and B. W. Ninham, J. Chem. H. Fischer, H. Uyama and S. Kobayashi, Macromolecules, 1996, Soc., Faraday T rans. 2, 1976, 72, 1525. 29, 6560. 21 Several other non-cubic intermediate phases can be found at the 48 The Designation as smectic phase is based on X-ray data; the transition from the lamellar to the columnar aggregate mor- optical textures are spherulitic textures as typical for columnar phology in lyotropic systems10 and for thermotropic compounds mesophases, see ref. 46. (undulated smectic phases, ribbon phases) in small concentration 49 A. Pegenau, P. Go� ring, S. Diele and C. Tschierske, L iquid and/or temperature regions.Crystals: Chemistry and Structure, ed. M. Tykarska, R. Dabroski 22 Reviews: (a) A. E. Skoulios, in Developments in Block Copolymers, and J. Zielinski, Proc. SPIE, 1998, 3319, 70. ed. I. Goodman, Elsevier, Appl. Sci. Pub., London, 1982, vol. 1, 50 J. Malthete, New. J. Chem., 1996, 20, 925. p. 81; (b) A. Skoulios, in Advances in L iquid Crystals, ed. 51 A. Pegenau and C.Tschierske, manuscript in preparation. G. H. Brown, Academic Press, New York, 1975, p. 169; (c) 52 A.-M. Giroud-Godquin and A. Rassat, C. R. Acad. Sci. Paris, 1982, 294, 241. A. Skoulios and D. Guillon, Mol. Cryst. L iq. Cryst., 1988, 165, 1506 J. Mater. Chem., 1998, 8(7), 1485–150853 (a) H. Zheng and T. M. Swager, J. Am. Chem. Soc., 1994, 111, 761; J. Heck, G.Ungar and S. V. Batty, J. Chem. Soc., Perkin T rans. 1, 1993, 1411. (b) T. M. Swager and H. Zheng, Mol. Cryst. L iq. Cryst., 1995, 88 Remarkably, the same mesophase stabilising eVect can also be 260, 301. obtained by the covalent fixation of the taper-shaped oligo(oxy- 54 (a) J. Malthete and A. M. Levelut, Adv. Mater., 1991, 3, 94; ethylene) derivatives 37 via a polymethacrylate backbone; ref. 86. (b) J. Malthete, A. Collet and A. M. Levelut, L iq. Cryst., 1989, 89 (a) V. Percec, J. Heck, D. Tomazos, F. Falkenberg, H. Blackwell 5, 123. and G. Ungar, J. Chem. Soc., Perkin T rans. 1, 1993, 2799; 55 U. Stebani and G. Lattermann, Adv.Mater., 1995, 7, 578. (b) V. Percec, J. A. Heck, D. Tomazos and G. Ungar, J. Chem. 56 S. Diele, H.-M. Vorbrodt and H. Zaschke, Mol. Cryst. L iq.Cryst. Soc., Perkin T rans. 2, 1993, 2381. L ett., 1984, 102, 181. 90 V. S. K. Balagurusamy, G. Ungar, V. Percec and G. Johansson, 57 A. S. C. Lawrence,Mol. Cryst. L iq. Cryst., 1969, 7, 1. J. Am. Chem. Soc., 1997, 119, 1539. 58 C. Tschierske, G. Brezesinski, F. Kuschel and H. Zaschke, Mol. 91 B. E. Smart, in Organofluorine Chemistry Principles and Cryst. L iq. Cryst., L ett., 1989, 6, 139.Commercial Applications, ed R. E. Banks, B. E. Smart and 59 H. A. van Doren, R. van der Geest, R. M. Kellog and H.Wynberg, J. C. Tatlow, Plenum Press, New York, 1994, p. 57. Recl. T rav. Chim. Pays-Bas, 1990, 109, 197. 92 B. Wang and S. Krause,Macromolecules, 1987, 20, 2201. 60 (a) G. Lattermann and G. Staufer, L iq. Cryst., 1989, 4, 347; 93 (a) T. A. Lobko, B.I. Ostrovskii, A. I. Pavluchenko and (b) G. Lattermann, G. Staufer and G. Brezesinski, L iq. Cryst. S. N. Sulianov, L iq. Cryst., 1993, 15, 361; (b) B. I. Ostrovskii, L iq. 1991, 10, 169. Cryst., 1993, 14, 131. 61 K. Praefcke, P. Marquardt, B. Kohne and W. Stephan, 94 B. P. Binks, P. D. I. Fletcher and R. L. Thompson, Ber. J. Carbohydr. Chem., 1991, 10, 539. Bunsenges. Phys. Chem., 1996, 100, 232. 62 K. Praefcke, B. Kohne, S. Diele, G. Pelzl and A. Kjaer, L iq. Cryst., 95 M P. Turberg and J. E. Brady, J. Am. Chem. Soc., 1988, 110, 7797. 1992, 11, 1. 96 J. F. Rabolt, T. P. Russell and R. J. Twieg, Mac4, 63 Reviews: (a) G. A. JeV rey, Acc. Chem. Res., 1986, 12, 179; 17, 2786. (b) G. A. JeVrey and L. M. Wingert, L iq. Cryst., 1992, 12, 179; (c) 97 (a) C.Viney, R. J. Twieg, T. P. Russell and L. E. Depero, L iq. H. Prade, R. Miethchen and V. Vill, J. Prakt. Chem., 1995, 337, Cryst., 1989, 5, 1783; (b) C. Viney, T. P. Russell, L. E. Depero and 427; and references therein. R. J. Twieg, Mol. Cryst. L iq. Cryst., 1989, 168, 63. 64 J. W. Goodby,Mol. Cryst. L iq. Cryst., 1984, 110, 205. 98 H. T. Nguyen, G. Sigaud, M. F. Achard, F. Hardouin, R.J. Twieg 65 V. Vill, T. Bo� cker, J. Thiem and F. Fischer, L iq. Cryst., 1989, and K. Betterton, L iq. Cryst., 1991, 10, 389. 6, 349. 99 (a) T. P. Rieker and E. P. Janulis, L iq. Cryst., 1994, 17, 681. 66 S. Fischer, H. Fischer, S. Diele, G. Pelzl, K. Jankowski, R. (b) T. P. Rieker and E. P. Janulis, Phys. Rev. E, 1995, 52, 2688. R. Schmidt and V. Vill, L iq. Cryst., 1994, 17, 855. 100 T. Doi, Y. Sakurai, A. Tamatani, S. Takenaka, S. Kusabashi, 67 In most cases the aliphatic chains are liquid-like (SA). Only for Y. Nishihata and H. Terauchi, J.Mater. Chem., 1991, 1, 169. 1,2-diols with long alphatic chains have low temperature meso- 101 S. V. Arehart and C. Pugh, J. Am. Chem. Soc., 1997, 119, 3027. phases with crystalline aliphatic parts been found, ref. 58. 102 (a) G. Pelzl, S. Diele, D. Lose, B. I. Ostrovski and W. Weissflog, 68 G. L. Gaines, Insoluble Monolayers at L iquid-Gas Interfaces, Cryst. Res. T echnol., 1997, 32, 99; (b) W. Weissflog, Dissertation Interscience, New York, 1966. B, Halle 1989. 69 K. Borisch, A. Pegenau and C. Tschierske, unpublished results. 103 S. Diele, D. Lose, H. Kruth, G. Pelzl, F. Guittard and A. Cambon, 70 J.D. Bunning, J. E. Lydon, C. Eaborn, P. M. Jackson, L iq. Cryst., 1996, 21, 603. J. W. Goodby and G. W. Gray, J. Chem. Soc., Faraday T rans. 1, 104 S. Pensec, F.-G. Tournilhac and P. Bassoul, J. Phys. II Fr., 1996, 1982, 78, 713. 6, 1597. 71 (a) W. V. DahlhoV, Z. Naturforsch T eil B, 1987, 42, 661; (b) H. A. 105 (a) U. Dahn, C. Erdelen, H. Ringsdorf, R. Festag, J. H. WendorV, Van Doren, T.J. Buma, R. M. Kellogg and H.Wynberg, J. Chem. P. A. Heiney and N. C. Maliszewskyj, L iq. Cryst., 1995, 19, 759; Soc., Chem. Commun., 1988, 460; (c) A. Eckert, B. Kohne and (b) H. Bengs, F. Closs, F. Frey, D. FunhoV, H. Ringsdorf and K. Praefcke, Z. Naturforsch T eil B, 1988, 43, 878. K. Siemensmeyer, L iq. Cryst., 1993, 15, 565. 72 R. Miethchen, J. Holz and H. Prade, Colloid Polym.Sci., 1993, 106 (a) F. Tournilhac, L. Bosio, J.-F. Nicoud and J. Simon, J. Chem. 271, 404. Phys. L ett., 1988, 145, 452; (b) L. M. Blinov, T. A. Lobko, 73 (a) K. Praefcke, P. Marquardt, B. Kohne, W. Stephan, A.- B. I. Ostrovskii, S. N. Sulianov and F. G. Tournilhac, J. Phys. II M. Levelut and E. Wachtel, Mol. Cryst. L iq. Cryst., 1991, 203, Fr., 1993, 3, 1121. 149; (b) K. Praefcke and D. Blunk, L iq. Cryst., 1993, 14, 1181; 107 E. Dietzmann, W. Weissflog, S. MarkscheVel, A. Jakli, D. Lose (c) K. Praefcke, D. Blunk and J. Hempel, Mol. Cryst. L iq. Cryst., and S. Diele, Ferroelectrics, 1996, 180, 59. 1994, 243, 323. 108 (a) X. Bao and L. R. Dix, Mol. Cryst. L iq. Cryst., 1996, 281, 291; 74 K. Praefcke, B. Kohne, A. Eckert and J. Hempel, Z. Naturforsch (b) S.Takenaka, J. Chem. Soc., Chem. Commun., 1992, 1748; T eil B., 1990, 45, 1084. (c) G. Johansson, V. Percec, G. Ungar and K. Smith, Chem. 75 K. Praefcke, B. Kohne, W. Stephan and P. Marquardt, Chimia, Mater., 1997, 9, 164. 1989, 43, 380. 109 V. Percec, D. Schlueter, Y. K. Kwon, J. Blackwell, M. Mo� ller and 76 K. Borisch, S. Diele, P. Go�ring and C. Tschierske, Chem.P. J. Slangen,Macromolecules, 1995, 28, 8807. Commun., 1996, 237. 110 G. Johansson, V. Percec, G. Ungar and J. Zhou, Macromolecules, 77 K. Borisch, S. Diele, P. Go� ring, H. Kresse and C. Tschierske, 1996, 29, 646. Angew. Chem., 1997, 109, 2188; Angew. Chem., Int. Ed Engl., 1997, 111 V. Percec, G. Johansson, G. Ungar and J. Zhou, J. Am. Chem. 36, 2087. Soc., 1996, 118, 9855. 78 K.Borisch, S. Diele, P. Go� ring, H. Kresse and C. Tschierske, 112 P. A. Gemmell, G. W. Gray and D. Lacey,Mol. Cryst. L iq. Cryst., J.Mater. Chem., 1998, 8, 529. 1985, 122, 205. 79 U. Beginn, S. Keinath and M. Mo� ller, L iq. Cryst., 1997, 23, 35. 113 F.-H. Kreuzer, D. Andrejewski, W. Haas, G. Riepl and P. Spes, 80 (a) G. Lattermann and G. Staufer, Mol. Cryst. L iq. Cryst., 1990, Mol. Cryst.L iq. Cryst., 1990, 199, 345. 191, 199; (b) G. Staufer, M. Schellhorn and G. Lattermann, L iq. 114 G. H. Mehl and J. W. Goodby, Chem. Ber., 1996, 129, 521. Cryst., 1995, 18, 519. 115 F. Hardouin, H. Richard and M. F. Achard, L iq. Cryst., 1993, 81 M. Veber, E. Cheylan, S. Czernecki and J. Xie, L iqu. Cryst., 1996, 14, 971. 21, 197. 116 M. Ibn-Elhaj, A. Skoulios, D. Guillon, J. Newton, P.Hodge and 82 However, cubic phases were found for single chain carbohydrate H. J. Coles, L iq. Cryst., 1995, 19, 373. derivatives with very large hydrophilic groups such as compound 117 J. Newton, H. Coles, P. Hodge and J. Hannington, J. Mater. 26; ref. 66. These phases should be of type 1. Chem., 1994, 4, 869. 83 E.g. ethyl 3,4,5-tridodecyloxybenzoate: mp: 42 °C; H.Meier, 118 J. W. Goodby, G. W. Gray, G. W, K. J. Harrison and J. A. Nash, E. Praß, G. Zerban and F. Kosteyn, Z. Naturforsch T eil B, 1988, Pramana Suppl., 1975, 1, 381. 43, 889. 119 F. Kuschel, A. Ma�dicke, S. Diele, H. Utschick, B. Hisgen and 84 V. Percec, J. Heck, G. Johannson, D. Tomazos, M. Kawasumi, H. Ringsdorf, Polym. Bull., 1990, 23, 373 and references cited. P. Chu and G.Ungar,Mol. Cryst. L iq. Cryst., 1994, 254, 137. 120 Reviews: (a) H. J. Deutscher, R. Frach, C. Tschierske and 85 V. Percec, G. Johannsson, D. Schluter, J. C. Rona and G. Ungar, H. Zaschke, in Selected T opics in L iquid Crystal Research, ed. Makromol. Chem., Macromol. Symp., 1996, 101, 43. H.-D. Koswig, Akademie-Verlag, Berlin, 1990; (b) K. J. Toyne, in 86 V. Percec, D. Tomazos, J.Heck, H. Blackwell and G. Ungar, Critical Reports on Applied Chemistry, vol. 22, T hermotropic J. Chem. Soc., Perkin T rans. 2, 1994, 31. L iquid Crystals, ed. G. W. Gray, Wiley, New York, 1987, p. 28. 87 (a) G. Johannson, V. Percec, G. Ungar and D. Abramic, J. Chem. 121 R. Eidenschink, D. Erdmann, J. Krause and L. Pohl, Angew. Chem., 1977, 89, 103; 1978, 90, 133. Soc., Perkin T rans. 1, 1994, 447; (b) V. Percec, G. Johannson, J. Mater. Chem., 1998, 8(7), 1485–1508 1507122 The phase transition given in square brackets is a ‘virtual’ clear- Angew. Chem., 1997, 109, 1160; Angew. Chem., Int. Ed. Engl., 1997, 36, 1119. ing temperature, which is obtained by extrapolation from mixtures with a nematic host. No mesophase can be detected on 148 A. Jacobi and W.Weissflog, L iq.Cryst., 1997, 22, 107. 149 (a) B. Neumann, D. Joachimi and C. Tschierske, L iq. Cryst., 1997, cooling the pure compound. 123 (a) E. Kleinpeter, H. Ko� hler, C. Tschierske and H. Zaschke, 22, 509. (b) R. Lunkwitz, B. Neumann and C. Tschierske, L iq. Cryst., 1998, in press. J. Prakt. Chem., 1988, 330, 484; (b) C. Tschierske, D. Joachimi, H.-M. Vorbrodt, H. Zaschke, A.Wiegeleben, A. Hauser and 150 M. A. Bates and G. R. Luckhurst, L iq. Cryst., 1998, 24, 229. 151 The columnar phases formed by polycatenar compounds can be D. Demus, L iq. Cryst., 1989, 5, 177; (c) V. Vill and J. Thiem, Mol. Cryst. L iq. Cryst. L ett., 1991, 8, 27. changed into smectic phases via cubic and nematic phases by addition of TNF: I. Letko, S. Diele, G. Pelzl and W.Weissflog, 124 G. Ungar, J. L. Feijoo, V. Percec and R. Yourd, Macromolecules, 1991, 24, 953. L iq. Cryst., 1995, 19, 643. 152 (a) B. Kohne and K. Praefcke, Angew. Chem., 1984, 96, 70; Angew. 125 (a) A. Mori, H. Takeshita, K. Kida and M. Uchida, J. Am. Chem. Soc., 1990, 112, 8635; A. Mori, S. Takematsu, M. Isobe and Chem., Int. Ed. Engl., 1995, 23, 82; (b) K. Praefcke, B. Kohne, W.Stephan and P. Marquart, Chimia, 1989, 43, 380. H. Takeshita, Chem. L ett., 1995, 15. 126 A. Mori, S. Takematsu and H. Takeshita, Chem. L ett., 1997, 1009. 153 K. Praefcke, W. Poules, B. Scheuble, R. Poupko and Z. Luz, Z. Naturforsch. T eil B., 1984, 39, 950. 127 S. Norvez and J. Simon, L iq. Cryst., 1993, 14, 1389. 128 Howeverr quadrupolar interactions and other 154 W. Poules and K. Praefcke, Chemiker-Ztg., 1983, 107, 310. 155 K. Praefcke, P. Psaras and B. Kohne, Chem. Ber., 1991, 124, 2523. influences can lead to tilted phases. 129 (a) H. Schubert, J. Hausschild, D. Demus and S. HoVmann, 156 (a) A. Liebmann, C. Mertesdorf, T. Plesnivy, H. Ringsdorf and J. H.WendorV, Angew. Chem., 1991, 103, 1358; Angew. Chem., Int. Z. Chem., 1978, 18, 256; (b) D. Demus, A. Gloza, H. Hartung, A. Hauser, I. Raphtel and A. Wiegeleben, Krist. T ech., 1981, Ed. Engl., 1991, 30, 1375; (b) G. Lattermann, S. Schmidt, R. Kleppinger and J. H. WendorV, Adv. Mater., 1992, 4, 30. 16, 1445. 130 (a) G. W. Gray, B. Jones and F. Marson, J. Chem. Soc., 1957, 393; 157 B. Kohne, K. Praefcke, W. Stephan and P. Nu�rnberg, Z. Naturforsch. T eil B., 1985, 40, 981. (b) S. Diele, P. Brand and H. Sackmann, Mol. Cryst. L iq. Cryst., 1972, 17, 163. 158 U. Krappe, R. Stadler and I. Voigt-Martin,Macromolecules, 1995, 28, 4558. 131 (a) B. Neumann, C. Sauer, S. Diele and C. Tschierske, J. Mater. Chem., 1996, 6, 1087; (b) C. Tschierske, J. A. Schro� ter, N. Lindner, 159 J. Beckmann, C. Auschra and R. Stadler, Macromol. Rapid Commun., 1994, 15, 67. C. Sauer, S. Diele, R. Festag, M. Wittenberg and J.-H. WendorV, L iquid Crystals: Chemistry and Structure, ed. M. Tykarska, 160 J. T. Chen, E. L. Thomas, C. K. Ober and S. S. Hwang, Macromolecules, 1995, 28, 1688. R. Dabrowski and J. Zielinski, SPIE, 1998, 3319, 8. 132 (a) W. Weissflog, A. Wiegeleben, S. Diele and D. Demus, Cryst. 161 J. T. Chen, E. L. Thomas, C. K. Ober and G. Mao, Science, 1996, 273, 343. Res. T echnol., 1984, 19, 583; (b) W. Weissflog, I. Letko, S. Diele and G. Pelzl, Adv. Mater., 1996, 8, 76. 162 Review: (a) H. Fischer and S. Poser, Acta Polymerica, 1996, 47, 413; (b) E. Chiellini, G. Galli, S. Angelino and M. Laus, T rends 133 U. Nu� tz, S. Diele, G. Pelzl, H. Ringsdorf, W. Paulus and G. Wilson, L iq. Cryst., 1995, 18, 699. Polym. Sci., 1994, 2, 244; (c)M.Walther and H. Finkelmann, Prog. Polym. Sci., 1996, 21, 951. 134 H. Strzelecka, C. Jallabert, M. Veber, P. Davidson, A. M. Levelut, J. Malthete, G. Sigaud, A. Skoulios and P. Weber, Mol. Cryst. 163 H. Fischer, S. Poser and M. Arnold, L iq. Cryst., 1995, 18, 503. 164 V. Percec and M. Lee, Pure Appl. Chem., 1992, A29, 723. L iq. Cryst., 1988, 161, 403. 135 D. W. Bruce, B. Donnio, D. Guillon, B. Heinrich and 165 L. H. Radzilowski and S. I. Stupp, Macromolecules, 1994, 27, 7747. M. Ibn-Elhaj, L iq. Cryst., 1995, 19, 537. 136 F. Artzner, M. Veber, M. Clerc and A.-M. Levelut, L iq. Cryst., 166 (a) J. W. Goodby, D. A. Dunmur and P. J. Collings, L iq. Cryst., 1995, 19, 703; (b) J.W. Goodby, J.Mater. Chem., 1991, 1, 307. 1997, 23, 27. 137 M. Lee, N.-K. Oh and W.-C. Zin, Chem. Commun., 1996, 1787. 167 A. M. Levelut; E. Hallouin, D. Bennemann, G. Heppke and D. Lo� tzsch, J. Phys. II Fr., 1997, 7, 981. 138 (a) M. Lee and N.-K. Oh, Mol. Cryst. L iq. Cryst., 1996, 280, 283; 168 S. P. Gido and E. L. Thomas, Macromolecules, 1997, 30, 3739. (b) M. Lee and N.-K. Oh, J.Mater. Chem., 1996, 6, 1079. 169 For more ordered supermolecular assemblies, however, a signifi- 139 M. Lee, N.-K. Oh, H. K. Lee and W.-C. Zin, Macromolecules., cant influence of chirality was found, review: J.-H. Fuhrhop and 1996, 29, 5567. J. Ko� ning,Membranes andMolecular Assemblies—The Synkinetic 140 W. Weissflog and D. Demus, Cryst. Res. T echnol., 1984, 19, 55. Approach, The Royal Society of Chemistry, Cambridge, 1996. 141 G. W. Gray, J. Chem. Soc., 1955, 236. 170 (a) A. Skoulios and V. Luzzati, Acta Crystallogr., 1961, 14, 278; 142 F. Hildebrandt, J. A. Schro� ter, C. Tschierske, R. Festag, (b) B. Gallot and A. Skoulios, Acta. Crystallogr., 1962, 15, 826. R. Kleppinger and J. H.WendorV, Angew. Chem, 1995, 107, 1780. 171 (a) J. L.M. van Nunen, B. F. B. Folmer and R. J. M. Nolte, J. Am. 143 F. Hildebrandt, J. A. Schro� ter, C. Tschierske, R. Festag, Chem. Soc., 1997, 119, 283; (b) J. L. M. van Nunen and M. Wittenberg and J. H. WendorV, Adv. Mater., 1997, 9, 564. R. J. M. Nolte, J. Chem. Soc., Perkin T rans. 2. 1997, 1473. 144 W. Zheng and Z.-G. Wang,Macromolecules, 1995, 28, 7215. 172 V. Percec, C.-H. Ahn and B. Barboiu, J. Am. Chem. Soc., 1997, 145 C. Auschra and R. Stadler,Macromolecules, 1993, 26, 2171. 199, 12978. 146 J. A. Schro� ter, C. Tschierske, R. Festag, M. Wittenberg and J. H.WendorV, manuscript in preparation. 147 J. A. Schro� ter, C. Tschierske, M. Wittenberg and J. H. WendorV, Paper 8/00946E; Received 3rd February, 1998 1508 J. Mater. Chem., 1998, 8(7

ISSN:0959-9428    

DOI:10.1039/a800946e

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

J O U R N A L O F C H E M I S T R Y Materials Communication Novel synthetic pathways to layered iron(hydro)oxyhydroxide–surfactant composites Gernot Wirnsberger,a Karl Gatterera and Peter Behrens*b† aInstitut fu� r Physikalische und T heoretische Chemie, T echnische Universita� t Graz, Rechbauerstraße 12, A-8010 Graz, Austria bInstitut fu� r Anorganische Chemie, L udwig-Maximilians-Universita�t Mu� nchen,Meiserstraße 1, D-80333 Mu�nchen, Germany this solution, diVerent amounts (5, 8, or 11 ml ) of a 2.5% ammonia solution were added.These solutions of the inorganic Mesostructured lamellar iron(hydro)oxyhydroxide–surfactant composites have been prepared using novel synthetic methods precursor are then aged under stirring for a certain period tA. Afterwards, 10 ml of a 0.07 M solution of CnS- is added.The based on the hydrolysis chemistry of FeIII. By carefully adjusting the reaction conditions, composites with inorganic experiments with C10S- and C12S- were carried out at room temperature. Due to the lower solubilities of C14S-, C16S- walls from ca. 19–26 A ° can be synthesised in a controllable manner. and C18S- at room temperature, the corresponding experiments were carried out at 45, 52 and 62 °C, respectively.In all cases, a brown precipitate was formed immediately after the combination of the solutions. To exemplify the influence of the pH we present the results The discovery of mesoporous materials1,2 with pore openings obtained using diVerent amounts of ammonia solution, but in the range from 20–100 A ° has opened the way to a new class with a constant ageing time of 120 min.By the addition of 5, of inorganic oxides with a wide potential for applications in 8 or 11 ml of a 2.5% ammonia solution, the pH of the precursor gas separation, sorption and catalysis. Generally, the synthesis solution was increased to 1.90, 2.00, or 2.13 A ° , respectively, involves the cooperative assembly of surfactant molecules with after 120 min of stirring.The solids isolated for the C12S- hydrolysable inorganic precursors into a mesostructured pre- composites [X-ray diVraction patterns in Fig. 1(a),(b)] have cipitate. By varying the surfactant head group, the kind of d001 spacings of 42.0, 45.7 and 45.9 A ° , respectively. A compariprecursor and other synthesis parameters like temperature, son of the composites obtained with the diVerent surfactants addition of an acid or base etc., the preparation of an amazing from C10S- to C18S- gives further information: straight line number of (metal) oxide–surfactant composites has been extrapolation of their d001 spacings versus n to n=0 allows a achieved.3 Among these, materials with a lamellar structure rough estimate of the thickness of the inorganic wall (shown have not attracted as much interest as their hexagonal or cubic in Fig. 2, for the synthesis carried out at pH 1.90). The results variants. This is due to the fact that upon removal of the show that this thickness increases from 19.2 to 24.5 A° , when surfactant template, either by oxidative calcination or by the amount of ammonia solution used is increased from 5 to extraction, the lamellar structure collapses and a dense phase 8 ml.A further increase to 11 ml results in materials with even is obtained. However, based on the special two-dimensional thicker inorganic walls of ca. 26.4 A° (in agreement with elemenstructure of the inorganic part, with nanometer-sized dimen- tal analysis showing an increasing Fe5CnS- ratio).The slopes sions in one direction, layered metal oxide–surfactant composites should exhibit special magnetic and optical properties which diVer from those of the corresponding bulk oxides. Therefore, they are attractive from both the theoretical and the application points of view. Recently, Tolbert et al.4 have shown that the magnetic properties of composites based on surfactants carrying a sulfate group (CnH2n+1SO4-, n=10, 12, 14, 16, 18; in the following designated as CnS-) and inorganic walls of hydrated iron(hydro)oxyhydroxide species vary with the thickness of the inorganic part.Their synthetic method involves the oxidation of layered FeII–surfactant compounds with H2O2 resulting in layered FeIII–surfactant compounds with a thickness of the inorganic layer of approximately 16.8 (for C10S-, C12S-, C14S-) or 11A ° (C16S-, C18S-), respectively.We will show that by using diVerent novel synthetic approaches, the thickness of the inorganic part can be increased up to ca. 26.4 A ° and that it can be controlled and varied nearly continuously in a broad range, from ca. 19.2 A ° to that value. Unlike the method of Tolbert et al.4 our synthetic procedures start with Fe3+.All our materials are layered composites, their X-ray diVraction patterns showing mainly the d00l reflections. Fig. 1 Typical powder X-ray diVraction patterns (Cu-Ka radiation) of lamellar mesostructured iron(hydro)oxyhydroxide species obtained by The first two of our novel approaches involve control of the (a) increasing the pH of a 0.1 M FeCl3 solution with NH3 to 1.90 and pH and ageing times.The general procedure starts with a adding 10 ml of 0.07 M C12S- solution after 120 min, (b) using the 0.1 M solution of FeCl3 which has a pH of 1.65. To 50 ml of same method and concentrations as in (a) but at a pH of 2.13, and (c) by mixing 500 ml of a 0.01 M FeCl3 solution with 100 ml of a 0.007 M C12S- solution †Email: pbe@anorg.chemie.uni-muenchen.de J.Mater. Chem., 1998, 8(7), 1509–1510 1509forces deprotonation and hence hydrolysis:6,7 Fe(OH2)63++H2OPFe(OH)(OH2)52++H3O+ Fe(OH)(OH2)52++H2OPFe(OH)2(OH2)4++H3O+ etc. In the case of a pure FeIII solution, the low molecular weight species further condense to give species with higher nuclearity, e.g. the dimer, 2 Fe(OH)(OH2)52+PFe2(OH)2(OH2)84++2H2O and finally a solid precipitate.Hence, we conclude that the condensation of low molecular weight precursors is the first step in the formation of the above described lamellar iron(hydro)oxyhydroxides species. Due to the low pH values, also after addition of ammonia, the rapid formation of bulk ironoxyhydroxide within 2 h is prevented. The condensation of the inorganic precursors into a two-dimensional array is only achieved in the second step by the addition of the Fig. 2 Extrapolation of d001 spacings versus number of carbon atoms, showing an intercept of 19.2 A ° and a slope of 1.99 A ° per carbon atom surfactant solution. In terms of the cooperative templating (starting pH=1.90) mechanism proposed for the formation of mesostructured materials,7,8 this involves the coordination of the negatively charged surfactants to the positively charged inorganic precursor molecules as an initiating reaction and the subsequent ordered self-assembly of these primary composites into solid lamellar composites. This model explains the increase in thickness of the inorganic layer with increasing pH and with increasing ageing time: both factors enhance the degree of condensation of the inorganic precursor species.Correspondingly, when the surfactant is added, larger inorganic precursors are taking part in the self-assembly process. We have shown that novel synthetic methods allow the adjustment and variation of the thickness of iron(hydro)oxyhydroxide layers. With regard to the hydrolytic chemistry of transition metals, the outlined preparative routes should be applicable to the synthesis of similar metal oxyhydroxide– surfactant composites as well as to the synthesis of composites with a mixed-valence inorganic part.References Fig. 3 Increase of the d001 spacing with ageing time of the precursor solution. The dotted line is a guide to the eye. 1 C. T. Kresge, M. E. Leonowicz, W. J. Roth, J. C.Vartuli and J. S. Beck, Nature, 1992, 359, 710. 2 J. S. Beck, J. C. Vartuli, W. J. Roth, M. E. Leonowicz, C. T. Kresge, of the extrapolation lines are in all cases larger than 1.26 A ° K. D. Schmitt, C. T. W. Chu, D. H. Olson, E. W. Sheppard, S. B. McCullen, J. B. Higgins and J. L. Schlenker, J. Am. Chem. Soc., per carbon atom, indicating that the surfactants build up a10834.tilted bilayer or a not fully interdigitated monolayer. 3 See, for example: P. Behrens, Angew. Chem., Int. Ed. Engl., 1996, 35, Fig. 3 exemplifies the influence of varying the ageing time tA 515 and references therein. in the synthesis procedure described above (all these experi- 4 S. H. Tolbert, P. Sieger, G. D. Stucky, S. M. J. Aubin, C.-C. Wu and ments were carried out with the addition of 5 ml of 2.5% D.N. Hendrickson, J. Am. Chem. Soc., 1997, 119, 8652. ammonia solution). The layer spacing d001 increases with 5 R. M. Cornell and U. Schwertmann, T he Iron Oxides, VCH, Weinheim, 1996. longer ageing times; for instance, d001=39.4 A° for tA=50 min 6 C. M. Flynn, Chem. Rev., 1984, 84, 31. and d001=48.0 A ° for tA=1500 min. 7 A. Monnier, F. Schu� th, Q. Huo, D. Kumar, D. Margolese, Another synthesis also starts from Fe3+, but uses highly R. S. Maxwell, G. D. Stucky, M. Krishnamurty, P. PetroV, diluted solutions. For example, adding 100 ml of a 0.007 M A. Firouzi, M. Janicke and B. F. Chmelka, Science, 1993, 261, 1299. solution of C12S- to 500 ml of a 0.01 M FeCl3 solution results 8 Q. Huo, D. I. Margolese, U. Ciesla, D. G. Demuth, P. Feng, in the precipitation of a brown solid with a layer spacing of T. E. Gier, P. Sieger, A. Firouzi, B. F. Chmelka, F. Schu�th and G. D. Stucky, Chem.Mater., 1994, 6, 1176. 34.8 A ° [Fig. 1(c)]. The results can be explained on the basis of the hydrolytic Communication 8/02345J; Received 25thMarch, 1998 chemistry of FeIII solutions. Addition of a base or dilution 1510 J. Mater. Chem., 1998, 8(7), 1509–

ISSN:0959-9428    

DOI:10.1039/a802345j

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

J O U R N A L O F C H E M I S T R Y Materials The synthesis of polyepoxides from unsaturated polymers and their attempted isomerisation to polyketones Mairi Nicol and David J. Cole-Hamilton*† School of Chemistry, University of St. Andrews, St. Andrews, Fife, Scotland, UK KY16 9ST Polybutadienes or polyisoprenes with >98% E-double bonds in the backbone can be epoxidised in quantitative yield to polyepoxides using ButOOH in the presence of dichloro[(1R)-endo-(+)-3-(diethoxyphosphoryl)camphor]dioxomolybdenum(VI) and molecular sieves.The polymers formed can be isolated in a pure form provided that HCl, which otherwise promotes crosslinking, is removed from the reaction solution. Attempts to isomerise the polyepoxides to polyketones using LiBr in the presence or absence of hexamethylphosphoramide (HMPA) or dimethylimidazolidinone (DMI) were partially successful, although the products generally tended to be insoluble as a result of aldol-type crosslinking reactions.For epoxides derived from polyisoprene, soluble isomerisation products can be obtained but these contain, in addition to ketone and unreacted epoxide units, ketals formed from reactions of isomerising epoxides with adjacent intact epoxide units.The thermal properties of the polyepoxides are discussed. Polymers containing epoxide groups have significant uses in on a Perkin-Elmer PE 1710 FTIR spectrometer and thermal analyses on a TA Instruments SDT 2960 simultaneous thermal the coatings and adhesives industries and are of two main types. Prepolymers to epoxy resins1 contain both terminal analyser over the temperature range 25–400 °C with a heating rate of 10 °C min-1.epoxy groups and hydroxy groups pendant from the main chain. They are formed by the reaction of a diol with, for Polybutadienes were obtained from Aldrich [Mw (2–3)×106] or Scientific Polymer Products [Mw (2–3)×105] example, chloromethylethene oxide. Ring opening of the epoxide gives a chlorohydrin, with a remote hydroxy group.A new and polyisoprene (product 18.214-1) from Aldrich. All polymers had >98% cis-1,4 units as confirmed by 1H and 13C epoxide unit is formed from the chlorohydrin by treatment with base. The density of epoxide linkages in these polymers NMR spectroscopy.8 LiBr (Aldrich) was dried before use by pumping in vacuo, whilst hexamethylphosphoramide (HMPA, is generally low.Crosslinking is then eVected by intermolecular ring opening of the epoxide ring by the pendant hydroxy Aldrich), tert-butyl hydroperoxide (TBHP, 3.0 mol dm-3 in 2,2,4-trimethylpentane, Fluka) and dimethylimidazolidinone groups. The other type of epoxide, which has much higher loadings of epoxy rings, is formed by the stoichiometric (using (DMI, Aldrich) were used without purification.The complex, dichloro[(1R)-endo-(+)-3-(diethoxyphosphoryl)camphor]di- peracids)2,3 or catalytic4,5 epoxidation of unsaturated polymers. During the course of studies aimed at the selective functionalis- oxomolybdenum(VI) (MoO2Cl2L) was prepared by a standard literature method.10 Molecular sieves (4 A ° , activated powder, ation of the diVerent kinds of double bonds in polybutadienes5 –9 which contain both backbone and pendant double average particle size 2–3 mm, Aldrich) were predried as described below.bonds, we investigated the epoxidation of these polymers using ButOOH and [MoO2Cl2L] [L=3-(diethoxyphosphoryl)- All manipulations were carried out under dry, oxygen-free nitrogen using standard Schlenk-line and catheter tubing tech- camphor]10–12 and found that this showed very high selectivity for the functionalisation of the backbone double bonds over niques.Solvents were dried by distillation from sodium diphenyl ketyl (toluene, THF, diethyl ether), sodium ( light the pendant double bonds. Small amounts of diol were also formed but these could be reduced if the reaction was carried petroleum, boiling range 40–60 °C) or calcium hydride (CH2Cl2).Benzene (May and Baker) was used as received. out in the presence of molecular sieves.5b,c In that study, we reported on the epoxidation of a polybutadiene containing >98% backbone double bonds in which we obtained up to Synthesis of epoxides 85% conversion of the double bonds to epoxide groups.5b,c (a) In the presence of molecular sieves.11,12 Molecular sieves Because of the commercial potential of these highly epox- (4 A ° , 10 g) were heated in vacuo for 5 min.A solution containing idised polymers and because they might be capable of further polybutadiene (0.54 g) or polyisoprene (0.68 g, 10 mmol of chemical modification, we have now carried out an indepth double bonds) in CH2Cl2 or THF (40 cm3) was added followed study of the epoxidation of polybutadienes as well as of by tert-butyl hydroperoxide (1.25 g, 13.89 mmol in 2,2,4-trime- polyisoprene.We have also studied the isomerisation of the thylpentane). After stirring for 30 min and optionally cooling polyepoxides to polyketones. to 0 °C, [MoO2Cl2L2] (0.1 g, 0.21 mmol) in CH2Cl2 or THF (10 cm3) was added and the resulting suspension was stirred Experimental for 18 h at room temperature. Atomic absorption analyses and microanalyses were performed (b) In the absence of molecular sieves.12 These reactions were at the University of St. Andrews.Solution phase NMR spectra carried out as in (a), above, but omitting the molecular sieves. were recorded on a Bru�ker AM300 spectrometer operating in the Fourier transform mode with, for 13C, broadband proton decoupling, solid state NMR spectra on a Bru�ker MSL 500 (c) Using a preformed MoO2–molecular sieves complex.Molecular sieves (4 A ° , 10 g) were heated in vacuo for 5 min. using magic angle spinning and cross-polarisation, IR spectra [MoO2Cl2L2] (0.1 g, 0.21 mmol) in CH2Cl2 or THF (10 cm3) was added and the resulting suspension was stirred for 15 min.†E-mail: djc@st-and.ac.uk J. Mater. Chem., 1998, 8(7), 1511–1515 1511After the molecular sieves had settled, the solution was (iv) As in (a) (ii) above, but using varying amounts of LiBr and HMPA. decanted and replaced by a solution containing polybutadiene (0.54 g) or polyisoprene (0.68 g, 10 mmol of double bonds) in Methods (a) (i) and (a) (ii) were also attempted using 2,3- epoxybutane (0.5 g) in place of the polymeric epoxide; only CH2Cl2 or THF (40 cm3).After stirring for 15 min, tert-butyl hydroperoxide (1.25 g, 13.89 mmol in 2,2,4-trimethylpentane) method (a) (ii) produced butan-2-one (95% yield by 1H NMR analysis in 3 h, 40% in 1 h). was added and the solution allowed to stir for 24 h at room temperature.Similar high conversions of the double bonds (b) DMI as solubiliser. (i) As in (a) (i) above, but using DMI were obtained using a stoichiometric amount of tert-butyl (0.45 g, 2.5 mmol) in place of HMPA. hydroperoxide. (ii) As in (a) (ii ) above, but using DMI (0.9 g, 5 mmol) in place of HMPA. After refluxing and cooling, the solution was Purification of polyepoxides analysed as in (a) (i) above.c-Butyrolactone, identified by its The product solutions obtained as described above were 1H NMR spectrum, was observed as a product in reactions allowed to settle and the bulk of the molecular sieves removed starting from epoxidised polybutadienes carried out in THF. by filtration. They were then centrifuged to settle the last traces Methods (b) (i) and (b) (ii) were also attempted using 2,3- of molecular sieves, decanted and stirred with CaCO3 (0.45 g, epoxybutane (0.5 g) in place of the polymeric epoxide.Method 0.42 mmol) for 30 min to remove any HCl produced during (b) (ii ) produced butan-2-one (>95% in 1 h). the reaction. The mixture was again centrifuged and decanted into light petroleum (100 cm3). The polymer that separated (c) In the absence of solubiliser.As in (b) (ii) above in THF, was collected and dried in vacuo. Epoxide from polybutadiene: except that solid LiBr was added in place of the solution Found C 65.9, H 8.2; [C4H6O]n requires C 68.6, H 8.5%. containing LiBr and DMI. 13C NMR: product from epoxidised Epoxide from polyisoprene: Found C 68.8, H 8.8; [C5H8O]n polybutadiene: dC 56.6 and 56.9 (epoxide CH), 100, 105, 106, requires C 71.4,.5%. The Mo content of the purified 108 (OCO, tentative assignment) 128, 130 and 131 (CHNCH), polymers formed by method (c) were 0.8 and 1.2% respectively. 203 (CNO); product from epoxidised polyisoprene (a soluble NMR: epoxidised polybutadiene, dH(all broad singlets) 1.7 product was obtained from all reactions using HMPA or DMI (1H), 1.8 (2 H) and 1.9 (1H) (all CH2), 3.0 (2 H) (CH).dC 25.1 in CH2Cl2 or no solubiliser in THF carried out by method and 25.3 (CH2), 56.6 and 56.9 (CH); epoxidised polyisoprene, (b) (ii ), these all gave very similar resonances for the polymer dH (all broad singlets) 1.3 (3 H) (CH3), 1.7 (2 H) and 1.8 (2H) C atoms), dC 60.6 and 60.8 (epoxide CMe), 64.4 and 64.6 (both CH2), 2.8 (1 H) (CH).dC 23.0 (CH3), 26.0 and 31.0 (epoxide CH), 100, 104 and 105 (OCO), 203 and 204 (CNO); (CH2), 60 8 and 60.6 (CMe), 64.4 and 64.6 (CH). resonances from other C atoms are present in the aliphatic region of the spectrum, but the complicated microstructure of Isomerisation of polyepoxide the polymer makes them diYcult to assign with confidence. (a) HMPA as solubiliser.13 (i) HMPA (0.45 g, 2.5 mmol) was added to a stirred solution of epoxidised polybutadiene (0.7 g, Results and Discussion 10 mmol of epoxide) or epoxidised polyisoprene (0.84 g, Epoxidation reactions 10 mmol of epoxide) in benzene, CH2Cl2, or THF (50 cm3).The solution was refluxed for 15 min and cooled to room The results of epoxidation reactions of the polydienes are temperature. LiBr (0.2 g, 2.5 mmol) was added and the solution collected in Table 1.In all cases complete conversion of the stirred for 10 min. Part of this solution (1 cm3) was transferred double bonds can be obtained, as indicated by the absence of into an NMR tube, where it was partially evaporated and characteristic peaks from the double bonds in the 1H (d 5.5) CDCl3 added. Any solid product in the initial solution was and 13C (d 130) NMR spectra.The best conditions for the collected by filtration and analysed by IR and solid state NMR epoxidation reactions (>95% conversion and >95% selecspectroscopy. tivity to the epoxide, even when using a stoichiometric amount (ii) A solution of polybutadiene epoxide (0.7 g, 10 mmol) or of oxidant) involve loading the catalyst onto the molecular polyisoprene epoxide (0.84 g, 10 mmol) in benzene, CH2Cl2, sieves (we have shown that both the b-keto phosphonate ligand or THF was treated with a solution of LiBr (0.4 g, 5 mmol) and HCl are liberated in this step).12 The molybdenumand HMPA (0.9 g, 5 mmol) in benzene, CH2Cl2 or THF containing molecular sieves are then separated from the liquid (30 cm3).The resulting mixture was refluxed for 15 min, cooled phase by filtration before addition of the reaction solvent and washed with water.The organic layer was decanted and (CH2Cl2 or THF), substrate and oxidant. After the reaction, evaporated to dryness to give, in the case of polybutadiene the solutions are filtered,‡ stirred with solid CaCO3 to remove epoxide, only a small amount of residue.Examination of the any HCl present and the polymers precipitated by decanting residue by IR spectroscopy showed only nCNO from the free into light petroleum as white rubbery powders, readily soluble b-keto phosphonate ligand (1740 cm-1) in the carbonyl in CH2Cl2 (polybutadiene) or THF (polyisoprene). Failure to stretching region. The main bulk of the product was a rubbery remove the HCl leads to crosslinking ring-opening reactions solid adhering to the sides of the initial reaction vessel.This such that the product polymers are diYcult to redissolve in was separated from the water washings, dried in vacuo and CH2Cl2 or THF. The reactions can also be carried out using analysed by IR spectroscopy. For polyisoprene epoxide, part a diVerent reaction sequence, in which the molybdenum cataof the organic layer was decanted into an NMR tube and lyst is added to a solution of the substrate, oxidant and treated as described in (a) (i) above.c-Butyrolactone, identified molecular sieves, followed by a similar work-up procedure. by its 1H NMR spectrum, was observed as a product in The main disadvantage of this protocol is that more molybreactions starting from epoxidised polybutadienes carried out denum is present in the final polymer.For reactions carried in THF. out in the absence of molecular sieves, small amounts of diol (iii ) Polybutadiene epoxide (0.7 g, 10 mmol) or polyisoprene residues are obtained (dH 4.0, dC 83) presumably from ringepoxide (0.84 g, 10 mmol), LiBr (0.5 g, 5.7 mmol) and HMPA opening reactions with water catalysed by the Mo complex.(1.0 g, 5.7 mmol) were dissolved in benzene, CH2Cl2 or THF We have shown separately that this catalyst is active, if rather (50 cm3). The resulting mixture was then refluxed for 15 min, slow, for such ring-opening reactions.14 cooled and worked up as in (ii) above. A sample of a polymer prepared in this way from polybutadiene epoxide was analysed ‡For the very high molecular weight polybutadiene (2–3 MD), the by solid state (CPMAS) 13C NMR spectroscopy.dC 21, 23, 39 viscosity of the solution is such that the molecular sieves can be diYcult to remove. and 44 (CH2), 130 (CHNCH), 210 (CNO). 1512 J. Mater. Chem., 1998, 8(7), 1511–1515Table 1 Microstructure of polymers obtained by epoxidation of unsaturated polymers in CH2CI2 under a variety of conditions unreacted substratea MW/amu catalystb epoxide units (%) diol units (%) double bonds (%) PBD (2–3)×106 A 86 5 8 PBD (2–3)×106 B >95 — — PBD (2–3)×105 Ac 80 20 — PBD (2–3)×105 B >95 — — PBD (2–3)×105 C >95 — — PISOP A 70 30 — PISOP B >95 — — PISOP C >95 — — aPBD is cis-polybutadiene, PISOP is cis-polyisoprene.bA=[MoO2Cl2L], B=[MoO2Cl2L] in the presence of molecular sieves, C=MoO2 supported on molecular sieves.cCatalyst concentration is double that usually employed. Spectroscopic properties of the epoxide polymers Physical and thermal properties of polyepoxides The polyepoxides are all white rubbery powders which redis- 1H and 13C NMR analysis of the purified polymers shows that only resonances characteristic of the polyepoxide are present, solve in CH2Cl2 or THF provided they have been purified to remove HCl.In general, they contain appreciable amounts of with no evidence for unreacted double bonds or diol formation. Of particular interest is the observation that the 13C NMR Mo from the catalysts (up to 75% of that used) but the amount of Mo can be greatly reduced by using the preformed resonance of each non-equivalent epoxide C atom in epoxidized polybutadienes appears as two peaks of equal intensity.We Mo–molecular sieves catalyst (see Experimental). In the STA (Table 2), the polymers all show a small weight assign these as arising from the random orientation of the epoxide oxygen atoms at either side of the chain. One resonance loss below 100 °C which we attribute to loss of entrapped solvent.There is then a fairly sharp exotherm peaking above arises from C atoms for which the O atom of the next unit is on the same side of the chain as that in the epoxide of the C 200 °C. Since this is not accompanied by a weight loss, we assume this corresponds to an isomerisation of the epoxide, atom in question, whilst the second resonance is from C atoms where the next O atom is on the opposite side of the chain probably to a polyketone.There is then a weak endotherm below 300 °C which is accompanied by weight loss. We attri- (Fig. 1). This random orientation also gives rise to nonequivalences in the CH2 protons of the chain, which resonate bute this to crosslinking of the polyketones by aldol condensation with the weight loss being caused by loss of water.as a broad triplet. Examination of the structure of the polymer (Fig. 1) shows that, in the conformation shown, the H atoms Finally, there is a more substantial endothermic weight loss above 300 °C presumably arising from complete polymer of each methylene unit are inequivalent but the H atoms of pairs of methylene groups are related by a local centre of degradation. inversion (CHAHB) or a local two-fold axis perpendicular to the plane of the polymer backbone (CHCHD) (assuming that Attempted ring-opening reactions the environment of the H atoms is only aVected by the epoxide The importance of polyketones as photodegradable polymers15 groups on each side of the methylene group in question).It and structural materials led us to investigate the isomerisation seems probable then that the highest field signal is from HD of polyepoxides by ring-opening isomerisation to form polyke- (on the opposite side of the chain from both adjacent epoxide tones.One of the most important polyketones is that derived O atoms) and the lowest field is from HC (on the same side as from the alternating copolymerisation of carbon monoxide both adjacent O atoms).The resonances from HA and HB then and ethene. This polymer has excellent properties in film form, overlap since they have one adjacent O atom on the same side but the regularity of its structure means that it has crystalline of the chain and the other on the opposite side. domains which lead to a high melting point, which is close to For the epoxidised polyisoprene, the two carbon atoms of its decomposition temperature.This, in turn, means that there the double bond are non-equivalent and give rise to resonances are problems with processing the polymer. Generally, these near d 61 (CCH3) and 64 (CH), with both appearing as double are improved by adding propene to the feedstock so as to peaks because of the random orientation of the epoxide introduce some irregularity into the chain and lower the linkages along the chain.This random orientation ensures that melting point. This polymer has just been commercialised16 there are eight diVerent H environments for the methylene for such uses as blades for lawn mowers or gear wheels. protons, but these appear as two broad resonances near d 1.75.In their backbones, these polymers contain two C atoms The CH2 carbon atoms resonate as two singlets at d 26 and between each CNO unit, and we reasoned that further advan- 31. The epoxide protons resonate as a broad singlet, despite tages in terms of processing might be possible if either this the fact that the O atom of the next epoxide in the chain can chain length were variable (2,3 or 4 C atoms between CNO be on the same side or the opposite side of the chain relative units) or it were longer, increasing rotational degrees of freedom to these protons.This suggests that the chemical shift of this by having fewer sp2 hybridised C atoms. proton is relatively insensitive to the stereochemistry of the C In principle, isomerisation of the polyepoxides in the epox- atom in the d position.The methyl resonances of the isoprene idised polybutadienes or polyisoprenes should be able to units appear as a singlet at d 23 in the 13C NMR and a broad provide polyketones of this type (see Fig. 2). Related polymers singlet at d 1.3 (1H). have been synthesised before by the direct oxidation of polybutadienes, but in all cases, the starting materials have contained pendant double bonds which oxidise preferentially to the backbone double bonds to give pendant methyl ketone units.4,17 Oxidation of the backbone double bonds can also be achieved; however, the yields are low.4 Several catalysts are available for isomerisation of epoxides to ketones but one of O O O HA HB HC HC HB HA HD HD the most eVective involves a 151 mixture of LiBr and hexa- Fig. 1 Structure of polybutadiene epoxide showing the various types of H atom present methylphosphoramide (HMPA) in benzene.13 We have studied J. Mater. Chem., 1998, 8(7), 1511–1515 1513Table 2 Simultaneous thermal analysis (STA) data for polyepoxidesa Tonset/°C weight loss (%) theoretical weight loss PBDO PISOPO DHb PBDO PISOPO assignment PBDO PISOPO 50 50 + 5 10 Solvent loss 175 180 - 0 0 Isomerisation 260 230 + 15 10 Crosslinking 12.8 10.7 360 280 + 61 75 Decomposition aPBDO is purified polybutadiene epoxide, PISOPO is purified polyisoprene epoxide.b+ is endotherm, - is exotherm. out in THF in the absence of a solubilising agent. In this case, it is possible to obtain a product which is partially soluble in THF. The 13C NMR spectrum for this polymer shows a resonance at d 203 characteristic of ketone units, showing that the isomerisation has been successful, but the conversion is low, as indicated by the presence of resonances from the epoxide C atoms near d 57 and a resonance near d 130 is consistent with some aldol-type crosslinking having occurred. Using epoxidised polyisoprene as the substrate, a soluble product can be obtained by using HMPA or DMI in CH2Cl2 or THF or by using LiBr alone in THF. 13C NMR studies of the crude reaction solution show the presence of two reson- O O O O O O O O O Me R R R R n R = Me R = H ances at d 203 and 204, characteristic of ketone units, although Fig. 2 Ring opening of polyepoxides showing the random microstruc- not all of the epoxide units have ring-opened as indicated by ture that can in principle be obtained from polybutadiene (R=H), or residual resonances at d 60.6, 60.8, 66.4 and 66.6.In addition, the methyl substituted polymer that can be obtained from polyisoprene there are four very intense resonances near d 100, in the region (R=Me). The arrows on the epoxide indicate the direction of opening in which acetal or ketal carbon atoms resonate. Although of the epoxide in the polybutadiene epoxide.The ring openings are some of these resonances may be associated with products accompanied by a H shift. obtained from ring-opening or oxidation of THF (cbutyrolactone is also a product), others are clearly associated this system for the isomerisation of the epoxidised polymers with the polymer. We assign them as arising from a reaction and find that, for the polybutadiene epoxides, white solids between ketone units and neighbouring epoxides to give brevicwhich are insoluble in all common organic solvents, including omin18,19 type structures along the polymer chain (Fig. 4). m-cresol and perfluoroisopropyl alcohol are obtained. These These type of bicyclic ketals readily form from epoxy ketones solids all show a band at 1710 cm-1 in their infra-red spectra where the ketone is separated from the epoxide by two or and a solid state 13C NMR spectrum of one of the samples three sp3 carbon atoms.Ketals of a diVerent kind can also be (Fig. 3) showed a strong resonance at d 210. Both of these formed during the synthesis of CO–alkene copolymers, especiindicate that ketone functions are formed, and the absence of ally if the alkene contains an a-substituent, but the regular peaks near 905 and 965 cm-1 in the IR spectra or d 55 in the arrangement of CNO groups along the chain ensures that solid state 13C NMR spectrum indicate that little or no epoxide these are monocyclic and that all the ketonic groups form groups remain.Surprisingly, the solid state 13C NMR spectrum ketals.20–22 In these polymers, resonances from the ketal carbon contains a resonance near d 130, suggesting that there is a atoms appear at d 113 in the 13C NMR spectrum.21 substantial amount of double bonds in the final polymer.This The opening of the epoxides presumably occurs in a random suggests that the polymer may have undergone crosslinking fashion along the polymer chain and the brevicomin type units via an aldol condensation reaction.This, in addition to the can only occur when a ketone, or the alkoxide intermediate, high initial molecular weight of the polymer, may perhaps is formed on the unit next to an intact epoxide. In the early account for the low solubility of the product. We have also stages of the reaction, this will normally be the case, but later shown that very similar products are obtained if the reaction on, the random nature of the reaction means that some epoxide is carried out using DMI in place of HMPA in either CH2Cl2 units will become isolated between brevicomin units.These or THF. The role of the HMPA or DMI is to solubilise the will either oxidise to isolated ketone units, which will remain LiBr in an organic solvent, but since LiBr dissolves in THF as ketones, or may be too sterically hindered to be attacked.without a solubilising agent, we find that the same reaction to This perhaps accounts for the fact that very high conversions give a very similar product also occurs if the reaction is carried have not been observed and that the resonances from the brevicomin type units are the major product resonances.The fact that all the resonances are multiple (ketone, epoxide and Fig. 3 Solid state CPMAS 13C NMR spectrum of a polyketone O O O O O O R R R R R R O O R R n n n n Fig. 4 The formation of bicyclic ketals with diVerent ring sizes from prepared by isomerisation of an epoxidised polybutadiene [Mw= (2–3)×106] using LiBr–HMPA epoxidised polybutadiene or polyisoprene 1514 J.Mater. Chem., 1998, 8(7), 1511–15154 M. P. McGrath, E. D. Sall and S. J. Tremont, Chem. Rev., 1995, 95, ketal) suggests that each kind of residue can be in two diVerent 381 and references therein. chemical environments, although the complexity of the polymer 5 (a) M. Gahagan, A. Iraqi, D. C. Cupertino, R. K. Mackie and does not allow us to distinguish what these might be.D. J. Cole-Hamilton, J. Chem. Soc., Chem. Commun., 1989, 1688; (b) A. Iraqi and D. J. Cole-Hamilton, J. Mater. Chem., 1992, 2, 183 and references therein; (c) A. Iraqi and D. J. Cole-Hamilton, Conclusion Polyhedron, 1991, 10, 993. 6 A. Iraqi, S. Seth, C. A. Vincent, D. J. Cole-Hamilton, We conclude that cis-polybutadiene or polyisoprene can be M.D. Watkinson, I. M. Graham and D. JeVrey, J. Mater. Chem., smoothly converted into polyepoxides in which all of the 1992, 2, 1057. double bonds have been epoxidised, the best results being 7 P. Narayanan, B. Kaye and D. J. Cole-Hamilton, J. Mater. Chem., 1993, 3, 19. obtained with a catalyst prepared from [MoO2]2+ adsorbed 8 P. Narayanan, A. Iraqi and D. J. Cole-Hamilton, J.Mater. Chem., on molecular sieves and with tert-butyl hydroperoxide as 1992, 2, 1149. oxidant. These polymers can be isomerised to polyketones 9 X. Li, C. M. Lindall, D. F. Foster and D. J. Cole-Hamilton, using LiBr in THF, but the products are either insoluble, J.Mater. Chem., 1994, 4, 657. probably as a result of extensive crosslinking via aldol type 10 M. Gahagan, R.K. Mackie, D. J. Cole-Hamilton, D. C. Cupertino, M. Harman and M. B. Hursthouse, J. Chem. Soc., Dalton T rans., reactions (polybutadiene epoxides), or contain units obtained 1990, 2195. from intramolecular cyclic ketal formation between ketone and 11 R. Clarke and D. J. Cole-Hamilton, J. Chem. Soc., Dalton T rans., epoxide units (polyisoprene epoxides) in addition to ketonic 1993, 1913. functionality and unreacted epoxides. 12 R. Clarke, M. Gahaghan, R. K. Mackie, D. F. Foster, D. J. Cole- Hamilton, M. Nicol and A. W. Montford, J. Chem. Soc., Dalton T rans., 1995, 1221. We thank Professor M. J. Green and Dr A. D. Poole for very 13 B. Rickborn and R. M. Gerkin, J. Am. Chem. Soc., 1971, 93, 1693. helpful discussions, Dr M. J. Payme for commenting on an 14 R. Clarke, PhD Thesis, University of St. Andrews, 1993. earlier version of the manuscript and BP Chemicals for a 15 A. Sen, Chemtech, 1986, 48. studentship. We are also very grateful to Dr F. G. Riddell for 16 C. E. Ash, J.Mater. Educ., 1994, 16, 1. solid state NMR measurements. 17 A. N. Ajjou and H. Alper, Macromolecules, 1990, 23, 4519. 18 G. T. Pearce, J.Magn. Reson., 1977, 27, 497. 19 W. E. Gove, J. Org. Chem., 1976, 41, 607. 20 J. A. van Doorn, P. K. Wong and O. Sudmeijer, EP 376 364/1991; References Chem. Abstr., 1991, 114, 24 797. 21 A. Batistini and G. Consiglio, Organometallics, 1992, 11, 1766. 1 W. P. Moore, An Introduction to Polymer Chemistry, University of 22 P. K.Wong, J. A. van Doorn, E. Drent, O. Sudmeijer and H. A. Stil, London Press, London, 1963, p. 177. Ind. Eng. Prod. Res. Dev., 1993, 32, 986. 2 D. Zuchowska, Polymer, 1980, 21, 514. 3 C. Pinazzi, J-C. Soutif and J-C. Brosse, Bull. Chem. Soc. Fr., 1973, 1652. Paper 8/00899J; Received 2nd February, 1998 J. Mater. Chem., 1998, 8(7), 1511–1515 1515

ISSN:0959-9428    

DOI:10.1039/a800899j

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

J O U R N A L O F C H E M I S T R Y Materials Fluorinated functional materials possessing biological activities: gel formation of novel fluoroalkylated end-capped 2-acrylamido-2- methylpropanesulfonic acid polymers under non-crosslinked conditions Hideo Sawada,a,b* Shinsuke Katayama,b Yukiko Ariyoshi,a Tokuzo Kawase,c Yoshio Hayakawa,d Toshio Tomitae and Masanori Babaf aDepartment of Chemistry, Nara National College of T echnology, Yata, Yamatokoriyama, Nara 639-1080, Japan bDepartment of Chemistry, Faculty of Advanced Engineering, Nara National College of T echnology, Yamatokoriyama, Nara 639-1080, Japan cFaculty of Human L ife Science, Osaka City University, Sugimoto, Sumiyoshi-ku, Osaka 558-8585, Japan dNational Industrial Research Institute of Nagoya, Kita-ku, Nagoya 462-8510, Japan eFaculty of Agriculture, T ohoku University, T sutsumidori-Amamiya, Aoba-ku, Sendai 981-8555, Japan fDivision of Human Retroviruses, Center for Chronic V iral Diseases, Faculty of Medicine, Kagoshima University, Sakuragaoka, Kagoshima 890-8520, Japan New fluoroalkylated end-capped 2-acrylamido-2-methylpropanesulfonic acid homopolymers were prepared by reaction of fluoroalkanoyl peroxides with 2-acrylamido-2-methylpropanesulfonic acid (AMPS).Similarly, fluoroalkylated end-capped copolymers were prepared by reaction of fluoroalkanoyl peroxides with AMPS and the comonomers such as trimethylvinylsilane and methyl methacrylate. These thus-obtained fluoroalkylated end-capped AMPS polymers were found to form gels not only in water but also in organic polar solvents such as methanol, ethanol, N,N-dimethylformamide and dimethyl sulfoxide under noncrosslinked conditions.On the other hand, AMPS polymer containing fluoroalkylene units {[-RF-(AMPS)q]p-} could cause no gelation under similar conditions. This suggests that fluoroalkylated end-capped AMPS polymers can cause gelation where strong aggregation of the end-capped fluoroalkyl segments is involved sterically in establishing the physical gel network in these media. Interestingly, these fluoroalkylated end-capped AMPS polymer hydrogels had a strong metal ion binding power.Moreover, it was demonstrated that these fluoroalkylated gelling polymers are potent and selective inhibitors of HIV-1 replication in cell culture. In addition, one of these gelling polymers was found to possess antibacterial activity against Staphylococcus aureus.Therefore, these fluorinated gelling polymers are suggested to have high potential for new functional materials through their gelling ability and biological activity. In view of the development of new fluoroalkylated end- Introduction capped polymeric materials, it is very interesting to explore Fluorinated polymeric materials exhibit numerous excellent the fluoroalkylated end-capped polymers containing both catproperties which cannot be achieved by the corresponding ionic and anionic segments by using fluoroalkanoyl peroxides.non-fluorinated materials. However, in general they are very Hitherto, the synthesis of non-fluorinated zwitterionic (betainepoorly soluble in various solvents.1 Therefore, it is interesting type) polysoaps has been reported by Laschewsky and Zerbe;12 to search for highly soluble fluorinated polymeric materials however, there has been so far no report on the synthesis of with excellent properties imparted by fluorine.We have recently fluoroalkylated end-capped betaine-type polymers.In our conreported that a series of fluoroalkylated end-capped silicon co- tinuing eVort to design and develop novel fluoroalkylated oligomers containing carboxy groups, which are prepared by betaine-type polymers, we discovered that novel fluoroalkylusing fluoroalkanoyl peroxides as key intermediates, are highly ated end-capped 2-acrylamido-2-methylpropanesulfonic acid soluble in various solvents, and are eVective in reducing the polymers, which were prepared by the reactions of the corresurface tension of these solvents.2 These fluorosilicon cosponding monomer with fluoroalkanoyl peroxides, could lead oligomers were also found to be potent and selective inhibitors to gelation not only in water but also in polar organic solvents against HIV-1 (human immunodeficiency virus type 1) repliunder non-crosslinking conditions.13 We now give a full cation in vitro.2 Furthermore, we have reported on the synthesis account of the gel formation and properties of these fluoroof fluoroalkylated end-capped oligomers containing trimethylalkylated end-capped polymers, with emphasis on an appli- ammonium segments3 or sulfo segments4 by using fluorocation to new fluorinated gelling functional materials pos- alkanoyl peroxides.These fluoroalkylated oligomers sessing biological activities. containing trimethylammonium or sulfo segments exhibited not only unique properties imparted by fluorine but also biological activities, although they have only two fluoroalkyl- Results and Discussion ated end-caps.3,4 Partially fluoroalkylated polymers were pre- The reactions of fluoroalkanoyl peroxides with 2-acrylamido- pared by a variety of anionic polymerizations5–8 and a 2-methylpropanesulfonic acid (AMPS) were carried out in fluoroalkylated end-capped moiety was introduced through heterogeneous solvent systems [AK-225 (mixed solvents of the ester bond9–11 to perfluoroalkyl-terminated polymers, and the interesting properties of these polymers have been reported. 1,1-dichloro-2,2,3,3,3-pentafluoropropane and 1,3-dichloro- J. Mater. Chem., 1998, 8(7), 1517–1524 1517dihydrochloride. This result also suggests that these fluoroalkylated AMPS polymers form aggregates. Our fluoroalkylated AMPS polymers were found to be easily soluble not only in water but also in polar organic solvents such as methanol, ethanol, N,N-dimethylformamide and dimethyl sulfoxide at under dilute conditions (below ca. 0.5 g dm-3). In order to clarify the solution properties of our fluoroalkylated polymers, we measured the viscosity of an aqueous solution of fluoroalkylated AMPS polymer [C3F7- (AMPS)n-C3F7] under dilute conditions (0.1 g dm-3), and the vicosities of aqueous solutions of perfluoropropylated polymers containing trimethylammonium segments [C3F7-(AETM)n- C3F7] and non-fluorinated AMPS polymer [-(AMPS)n-] were RF-(CH2-CH) n-RF C O N+H2CMe2CH2SO3 – RF = C3F7, CF(CF3)[OCF2CF(CF3)]mOC3F7; m = 0,1,2 nCH2 CH C O N+H2CMe2CH2SO3 – O + RFCOOCRF O [AMPS] 40 °C, 5 h [RF-(AMPS) n-RF also measured under similar conditions for comparison.These Scheme 1 results are shown in Fig. 1. As shown in Fig. 1, the viscosity of an aqueous solution of C3F7-(AMPS)n-C3F7 at 5 °C was higher than that of C3F7- 1,2,2,3,3-pentafluoropropane) and water] by stirring vigorously (AETM)n-C3F7 and -(AMPS)n- under dilute conditions (0.1 g at 40 °C for 5 h under nitrogen.The process is outlined in dm-3). On the other hand, the viscosities of C3F7-(AMPS)n- Scheme 1. C3F7 including both C3F7-(AETM)n-C3F7 and -(AMPS)n- were AMPS was found to react with fluoroalkanoyl peroxides also found to decrease on heating the solutions from 5–50 °C.under mild conditions to aVord fluoroalkylated end-capped However, surprisingly, at concentrations above 1.0 g dm-3 AMPS homopolymers [RF-(AMPS)n-RF] in 35–58% isolated all fluoroalkylated end-capped AMPS polymer-solvent systems yields as shown in Table 1. formed gels.To study this unique gelation, we measured the Similarly, in the copolymerization of AMPS with fluoroviscosity of aqueous solutions of these fluoroalkylated end- alkanoyl peroxides, we succeeded in preparing a series of capped polymers at 30 °C. The results are shown in Fig. 2. fluoroalkylated end-capped AMPS copolymers by using As shown in Fig. 2, the viscosities of -(AMPS)n-, C3F7- comonomers such as trimethylvinylsilane and methyl metha- (AETM)n-C3F7 and perfluoropropylated polymer containing crylate in 30–57% isolated yields as shown in the following sulfo segments [C3F7-(MES)n-C3F7] increased little with Scheme 2 and Table 1. increasing concentrations, and the gel did not form, although As shown in Table 1, not only perfluoropropylated but also these fluoroalkylated polymers were shown to form molecular a series of perfluoro-oxaalkylated end-capped homo- and coaggregates like micelles in aqueous solutions.3,4 On the other polymers were obtained under mild conditions, and the copolyhand, the viscosity of our fluoroalkylated end-capped AMPS merization ratios of these polymers were determined by polymers increased greatly with increasing concentration, and 1H NMR analyses.Under our polymerization conditions, in it became impossible to measure their viscosity owing to the which the concentration of the peroxide was almost the same gelation at concentrations above 0.5 or 1.0 g dm-3. We also as that of AMPS (trimethylvinylsilane or methyl methacrylate) tried to measure the melting temperature of the gel; but, the as shown in Table 1, mainly polymers with two fluoroalkylated gel did not melt either in water or in organic solvents even end-groups would be obtained via primary radical termination when it was heated to around 95 °C.It is suggested that our or radical chain transfer to the peroxide. In fact, we have fluoroalkylated end-capped AMPS polymers could cause a already reported that two fluoroalkylated end-capped acrylic gelation involving a strong aggregation of fluoroalkyl segments acid oligomers [RF-(CH2CHCO2H)n-RF] are obtained by the to a physical gel network in water, methanol, ethanol, N,N- reactions of fluoroalkanoyl peroxides with acrylic acid under dimethylformamide and dimethyl sulfoxide at higher concen- similar conditions.14 The molecular weights (M9 n) of these polymers measured by GPC [gel permeation chromatography calibrated with standard poly(ethylene glycol ) by using 0.5 mol dm-3 Na2HPO4 solution as the eluent] were relatively high (8300–24 500). Considering the fact that water-soluble fluoroalkylated endcapped polymers containing trimethylammonium and sulfo segments easily form molecular aggregates in aqueous solutions, 3,4 this finding suggests that the GPC values indicate the apparent molecular weights.Interestingly, the M9 w/M9 n values of these fluoroalkylated polymers are extremely high (28–726) compared to that of the corresponding non-fluorinated polymer [-(AMPS)n-: M9 n=5400 (M9 w/M9 n=3.25)], which was prepared by using 2,2¾-azobis(2-methylpropionamidine) RF-(CH2-CH) x-(CH2-CR1R2) y-RF C O N+H2CMe2CH2SO3 – R1 = H, R2 = SiMe3 R1 = Me, R2 = CO2Me xCH2 CH C O N+H2CMe2CH2SO3 – O + RFCOOCRF O [AMPS] 40 °C, 5 h + yCH2 CR1R2 [RF-(AMPS) X-(CH2-CR1R2) Y-RF Fig. 1 EVect of temperature on viscosity of aqueous polymer solutions Scheme 2 1518 J. Mater. Chem., 1998, 8(7), 1517–1524Table 1 Homo- and co-polymerizations of AMPS with fluoroalkanoyl peroxides AMPS comonomer RF in peroxide (amt, mmol) (mmol) (amt, mmol) product; yield (%)a M9 n(M9 w/M9 n)b [ x5y]c C3F7 (3) 9 — RF-(AMPS)n-RF; 39 24 500 (28) CF(CF3)OC3F7 (3) 9 — RF-(AMPS)n-RF; 58 20 500 (131) CF(CF3)OCF2CF(CF3)OC3F7 (3) 9 — RF-(AMPS)n-RF; 38 12 000 (282) CF(CF3)OCF2CF(CF3)OCF2CF(CF3)OC3F7 (2) 6 — RF-(AMPS)n-RF; 35 24 000 (76) C3F7 (3) 9 CH2=CHSiMe3 (9) RF-(AMPS)x-(CH2CHSiMe3)y-RF; 57 11 000 (291) [90510] CF(CF3)OC3F7 (3) 9 CH2=CHSiMe3 (9) RF-(AMPS)x-(CH2CHSiMe3)y-RF; 38 10 000 (400) [78522] CF(CF3)OCF2CF(CF3)OC3F7 (3) 9 CH2=CHSiMe3 (9) RF-(AMPS)x-(CH2CHSiMe3)y-RF; 30 17 400 (166) [9852] C3F7 (3) 9 CH2=CMeCO2Me (9) RF-(AMPS)x-(CH2-CMeCO2Me)y-RF; 52 11 300 (216) [9357] CF(CF3)OC3F7 (3) 9 CH2=CMeCO2Me (9) RF-(AMPS)x-(CH2-CMeCO2Me)y-RF; 41 14 300 (211) [9456] CF(CF3)OCF2CF(CF3)OC3F7 (3) 9 CH2=CMeCO2Me (9) RF-(AMPS)x-(CH2-CMeCO2Me)y-RF; 41 8300 (726) [9258] aThe yields are based on the starting materials [AMPS, comonomer and the decarboxylated peroxide unit (RF-RF].bThe molecular weight of each polymer was determined by GPC; however, it is suggested that the obtained values by GPC indicate the apparent molecular weights owing to the strong aggregations of fluoroalkyl segments in aqueous solutions.cCopolymerization ratio was determined by 1H NMR analysis. J. Mater. Chem., 1998, 8(7), 1517–1524 1519likely to promote the gelation sterically compared to the corresponding homopolymers. These fluoroalkylated polymers also exhibited a quite similar property to the common water-swollen crosslinked polymeric hydrogels.That is, these fluoroalkylated AMPS polymers in water exhibited extremely high water adsorption, with the weight of adsorbed water by fluoroalkylated gelling copolymers: C3F7-(AMPS)x-(CH2CHSiMe3)y-C3F7 and C3F7- (AMPS)x-(CH2CMeCO2Me)y-C3F7 201 and 213 times the weight of the dry gel, respectively. These values are similar to that of AMPS polymer hydrogel prepared by radical polymerization of AMPS and N,N¾-methylenebisacrylamide initiated by H2O2.17 The striking characteristic of our AMPS polymers is their gelation both in water and in organic polar solvents under non-crosslinking conditions.This is because fluoroalkyl segments are solvophobic and aggregate in aqueous and organic media. In fact, it was reported that the solvophobic fluorocarbon tails in fluorinated amphiphiles are responsible for the formation of stable bilayer membrances in water and organic solvents.18–22 Semifluorinated alkanes, such as F(CF2)10(CH2)12H, are also known to exhibit gel-like characteristics in hydrocarbon solvents [H(CH2)pH; p=8, 10, 12, 14].23 Such aggregation of fluoroalkyl segments in these media should be enhanced due to the self-organization of polymers which causes gelation in these media.Our AMPS polymers can form gels in both water and organic media due to the synergistic interaction between the aggregations of fluoroalkyl units, and the ionic interactions of the amide cations and the sulfonate anions as shown in the following schematic illustration. On the other hand, in the case of the corresponding non- fluorinated polymer [-(AMPS)n-], only the ionic interactions would operate and the gelation would not occur.Furthermore, Fig. 2 EVect of concentration on viscosity of fluoroalkylated end-capped polymers measured at 30 °C by using a falling-sphere viscometer trations. By contrast, non-fluorinated AMPS polymers were completely soluble in these media, and no gel formed. The gelation abilities of some fluoroalkylated end-capped AMPS polymers were also studied by measuring the minimum concentrations of these polymers necessary for gelation according to the method reported of Hanabusa et al.15,16 The minimum concentrations for gelation (Cmin) in water and dimethyl sulfoxide (DMSO) at 30 °C are listed in Table 2.As shown in Table 2, the gelling ability of fluoroalkylated end-capped copolymers is somewhat superior to that of the homopolymers, with Cmin 2–9 g dm-3 for copolymers and 11–33 g dm-3 for homopolymers. This result is in fair agreement with the values of M9 w/M9 n of polymers in Table 1, and the more polydisperse polymers exhibited the higher gelling Fig. 3 Schematic illustration for gelation of RF-(AMPS)n-RF ability. These findings would suggest that the copolymers are Table 2 Minimum concentrations for gelation (Cmin) of fluoroalkylated AMPS homo- and co-polymers (in g dm-3 solvent) necessary for gelation at 30 °C Cmin/g dm-3 (gelator/medium) polymer water DMSO RF-(AMPS)n-RF; RF=C3F7 25 13 RF-(AMPS)n-RF; RF=CF(CF3)OCF2CF(CF3)OC3F7 31 11 RF-(AMPS)n-RF; RF=CF(CF3)OCF2CF(CF3)OCF2CF(CF3)OC3F7 33 21 RF-(AMPS)x-(CH2CHSiMe3)y-RF; RF=C3F7 6 6 RF-(AMPS)x-(CH2CHSiMe3)y-RF; RF=CF(CF3)OC3F7 6 8 RF-(AMPS)x-(CH2CHSiMe3)y-RF; RF=CF(CF3)OCF2CF(CF3)OC3F7 7 9 RF-(AMPS)x-(CH2CMeCO2Me)y-RF; RF=C3F7 2 5 RF-(AMPS)x-(CH2CMeCO2Me)y-RF; RF=CF(CF3)OC3F7 9 7 RF-(AMPS)x-(CH2CMeCO2Me)y-RF; RF=CF(CF3)OCF2CF(CF3)OC3F7 5 5 1520 J.Mater. Chem., 1998, 8(7), 1517–1524in the case of C3F7-(AETM)n-C3F7 or C3F7-(MES)n-C3F7, only the aggregation of the end-capped fluoroalkyl segments would operate, owing to the electrostatic repulsion of cationic segments (AETM) or anionic segments (MES) between the central polymer chains, and these polymers could not form gels.In general, it is well-known that acrylated and methacrylated polymers containing longer perfluoroalkyl groups are strongly repelled by water or hydrocarbons owing to the strong electronegativity of fluorine.In contrast, the characteristics of our present fluoroalkylated end-capped AMPS polymers are to cause gelation under non-crosslinking conditions. This feature would be due to their unique structure (fluoroalkylated endcapped structure), and the end-capped fluoroalkyl segments in polymers could strongly aggregate rather than being repelled by aqueous or organic media.Therefore, it is suggested that AMPS polymers containing fluoroalkylene units but no fluoroalkyl end-cap units, {[-RF-(AMPS)q]p-}, could not gel since the interaction between the aggregation of internal fluoroalkylene units in the polymers should become weaker than that of the aggregation of end-capped fluoroalkyl units in polymers.Previously, we reported that a polymeric per- fluoro-oxaalkane diacyl peroxide {[-(O=C)RFC(=O)OO]p-} Fig. 4 Relationship between relative amount of metal ions binding is a useful tool for the introduction of the perfluoro-oxaalkylene to [RF-(AMPS)x-(CH2CMeCO2Me)y-RF; RF=CF(CF3)OCF2CF- (CF3)OC3F7] and relative amount of initial metal ions: (#) Co 2 +; unit (-RF-) into acrylic acid polymers.24,25 Thus, we tried to ($) Cr 3 +; (A) theoretical line (corresponds to a 100% binding prepare fluoroalkylene unit-containing AMPS polymers by ratio); (a) Cr 3 + ion (AMPS polymer curved by N,N¾- using a polymeric perfluoro-oxaalkane diacyl peroxide.The methylenebisacrylamide–H2O2 system) (see ref. 17); (b) Co 2 + ion result is shown in Scheme 3. (AMPS polymer cured by N,N¾-methylenebisacrylamide–AIBN system) (see ref. 17). †[AMPS] indicates calculated concentration (mol dm-3) based on polymer monomer unit. towards metal ions on the swelling equilibrium of our fluoroalkylated end-capped AMPS polymer hydrogel. The swelling equilibrium of RF-(AMPS)x- (CH2CMeCO2Me)y-RF[RF=CF(CF3)OCF2CF(CF3)OC3F7] hydrogel was studied in aqueous solutions of metal ions (Cr3+ and Co2+).The metal-ion concentrations of the supernatant liquid after incubation (at 25 °C for 24 h) was spectrophotometrically determined from absorbance at 580 nm (Cr3+) or 510 nm (Co2+). The binding of the metal ions by this fluorinated gel was studied for a wide range of metal-ion concentrations, and the results are shown in Fig. 4. We found a remarkable decrease in the absorbance of Cr3+ or Co2+ after the addition of the fluoroalkylated hydrogel to each metal ion solution. As shown in Fig. 4, the Cr3+ and Co2+ ion bindings increased linearly with an increase in the pq CH2 CH C O N+H2CMe2CH2SO3 – [AMPS] 45 °C, 5 h –(CRFCOO)p- O O -[RF-(CH2-CH) q] p- C O N+H2CMe2CH2SO3 – -[RF-(AMPS) q] p- Yield: 45% b Mn = 120000 (Mw/Mn = 1.46 -RF- = -CF(CF3)[OCF2CF(CF3)] nO(CF2)5O-[CF(CF3)CF2O]mCF(CF3)-; ( n + m = 3) 0.2 mmol a 2 mmol initial concentration of Cr3+ or Co2+ with ca. 60% binding Scheme 3 aCalculated on the basis of the peroxide monomer unit. ratio {based on the relative amount of Cr3+ (or Co2+) binding bYield based on AMPS and decarboxylated peroxide unit (RF-) to gel ([Metal ion]binding/[AMPS]) and relative amount of initial metal ion ([Metal ion]add/[AMPS])}.Furthermore, this As shown in Scheme 3, the reaction of AMPS with a polymeric perfluoro-oxaalkane diacyl peroxide proceeded fluoroalkylated hydrogel was found to have a similar adsorptive property towards Co2+ and Cr3+ ions. This finding would smoothly to give AMPS polymer containing perfluoro-oxaalkylene units {-[RF-(AMPS)q]p-}. The AMPS polymer contain- mean that these metal ions bind ionically to the anionic parts of the oligomer networks in the fluorinated gel, and this ionic ing fluoroalkylene units thus obtained was found to be easily soluble not only in water but also in methanol, ethanol, N,N- interaction is not aVected by the gel structure.On the other hand, bound Co2+ or Cr3+ ion was not released from the dimethylformamide and dimethyl sulfoxide even at higher concentrations above 1.0 g dm-3, and the gel could not form fluoroalkylated gel into water.This result also suggests that the interaction for the binding of the metal ions to the hydrogel in any of these media. This finding suggests that the internal fluoroalkylene units in the polymer are not likely to aggregate is not coordination but ionic.Interestingly, the fluoroalkylated end-capped AMPS poly- sterically with each other compared to the end-capped fluoroalkyl units. Therefore, it is concluded that fluoroalkylated end- mer hydrogel was shown to have a higher metal ion binding power than the corresponding AMPS polymer hydrogels capped AMPS polymers can cause gelation where strong aggregation of the end-capped fluoroalkyl segments is involved crosslinked by N,N¾-methylenebisacrylamide–H2O2 or AIBN (azobisisobutyronitrile) system17 as in Fig. 4. As a result, it can sterically in establishing the physical gel network in aqueous and organic media. be said that the metal ions should interact in part with not only the anionic parts of the fluorinated hydrogel but also the Hitherto, the synthesis of chemically-cross-linked AMPS polymer gels has been reported by Osada et al., and this fluoroalkyl segments in the fluorinated AMPS hydrogel possessing strong electron-withdrawing properties.polymer hydrogel crosslinked by N,N¾-methylenebisacrylamide has a high adsorptive property towards various metal ions.17 In this way, it was verified that the aggregation of fluoroalkyl segments in water and in organic media becomes a new driving Therefore, it is interesting to study the adsorptive property J.Mater. Chem., 1998, 8(7), 1517–1524 1521Table 3 Inhibitory eVect of fluoroalkylated 2-acrylamido-2-methylpropanesulfonic polymers on the replication of HIV-1 in MT-4 cells polymer M9 n: EC50 a/mg ml-1 CC50 b/mg ml-1 RF-(AMPS)n-RF; RF=C3F7 24 500 1.6 >100 RF-(AMPS)n-RF; RF=CF(CF3)OCF2CF(CF3)OC3F7 12 000 1.6 >100 RF-(AMPS)n-RF; RF=CF(CF3)OCF2CF(CF3)OCF2CF(CF3)OC3F7 24 000 1.7 >100 RF-(AMPS)x-(CH2CHSiMe3)y-RF; RF=C3F7 11 000 0.6 >100 RF-(AMPS)x-(CH2CHSiMe3)y-RF; RF=CF(CF3)OC3F7 10 000 2.3 >100 RF-(AMPS)x-(CH2CHSiMe3)y-RF; RF=CF(CF3)OCF2CF(CF3)OC3F7 17 400 1.7 >100 RF-(AMPS)x-(CH2CMeCO2Me)y-RF; RF=RF=C3F7 11 300 1.9 >100 RF-(AMPS)x-(CH2CMeCO2Me)y-RF; RF=CF(CF3)OC3F7 14 300 0.23 >100 -(AMPS)n- 5400 1.6 36 dextran sulfate (MW=5000) 3.5 >100 aFifty percent eVective concentration, based on the inhibition of HIV-1 induced cytopathic eVects in MT-4 cell.bFifty percent cytotoxic concentration, based on the impairment of viability of mock-infected MT-4 cells. factor for gelation as well as the well-known interactions such C3F7-(AMPS)x-(CH2CHSiMe3)y-C3F7 was able to exhibit eVectively not only anti-HIV-1 activity but also antibacterial as hydrogen bonding and ionic interaction.Furthermore, it was clarified that the hydrogels which are built up through activity. In this way, it was demonstrated that our fluorinated AMPS the aggregation of the fluoroalkyl segments have similar metal ion adsorptive properties to the well-known chemically oligomers have not only a gelling ability but also anti-HIV-1 activity or antibacterial activity, although these compounds crosslinked polymer hydrogels.Because our fluoroalkylated AMPS polymers are gelling, it are high molecular mass materials containing only fluoroalkylated end-groups in one oligomeric molecule.Hence, these is of particular interest to investigate their potential as biologically active materials. Thus, a series of fluoroalkylated end- new fluorinated compounds are expected to be widely applicable in various fields as new attractive fluorinated gelling capped AMPS polymers have been evaluated for activity against HIV-1 replication in MT-4 cells (see Table 3).functional materials possessing biological activities. As shown in Table 3, the 50% eVective concentrations of the oligomers were 0.23–2.3 mg ml-1 in MT-4 cells, whereas Experimental they were not toxic at concentrations up to 100 mg ml-1. These NMR spectra were measured using a Varian Unity-plus 500 values are superior to those of dextran sulfate, which has been (500 MHz) spectrometer, while IR spectra were recorded on a considered to be a potent and selective polymeric inhibitor of HORIBA FT-300 FT-IR spectrophotometer.Molecular HIV-1 replication in cell culture to date.26 On the other hand, weights were calculated by using a JASCO-PU-980-Shodex- non-fluorinated polymers were toxic to the host cells. The SE-11 gel permeation chromatography calibrated with stan- mechanism of action of the gelling polymers may also explain dard poly(ethylene glycol) by using 0.5 mol dm-3 Na2HPO4 the inhibition of virus adsorption, as previously demonstrated solution as the eluent. Absorption spectra were recorded on a for fluoroalkylated acrylic acid oligomers.27–29 Interestingly, Shimadzu UV-240 spectrophotometer.Solution viscosities there is some correlation between the activity against HIVwere measured by using a falling-sphere Haake Viscometer 1 and the Cmin values (medium: water). As the activity D1-G.against HIV-1 shown in Table 3 becomes higher, the Cmin values (see Table 2) become in general smaller. Thus, fluoro- Materials alkylated end-capped AMPS copolymers such as C3F7- (AMPS)x-(CH2CHSiMe3)y-C3F7, C3F7OCF(CF3)-(AMPS)x- A series of fluoroalkanoyl peroxides [(RFCOO)2] and a poly- (CH2CMeCO2Me)y-CF(CF3)OC3F7 exhibit a higher antimeric perfluoro-oxaalkane diacyl peroxide were prepared by HIV-1 activity than the corresponding homopolymers, and the method described in the literature.24,25,32 2-Acrylamido-2- these copolymers possess in general a higher gelling ability.methylpropanesulfonic acid was purchased from Tokyo Kasei Therefore, it is suggested that the polymers possessing a higher Kogyo Co., Ltd.Trimethylvinylsilane was purchased from gelling ability (that is, the oligomers which are more Shin-Etsu Co., Ltd. Chromium(III ) nitrate and cobalt(II) chloradsorbable) would interact strongly with the virus leading to ide were purchased from Wako Chemicals. more potent inhibitory eVects against HIV-1 replication.Very recently, we have reported that fluoroalkylated end- General procedure for the synthesis of fluoroalkylated endcapped oligomers containing trimethylammonium segments capped AMPS oligomers possess antibacterial activity against Staphylococcus aureus.3 Hence, our present fluoroalkylated end-capped AMPS poly- Perfluorobutyryl peroxide (3 mmol) in a 151 mixture (AK- 225) of 1,1-dichloro-2,2,3,3,3-pentafluoropropane and 1,3- mers are also expected to show antibacterial activity since these polymers contain the amide segments.These AMPS dichloro-1,2,2,3,3-pentafluoropropane (110 g) was added to an aqueous solution (50%, w/w) of AMPS (9 mmol). The hetero- polymers have been evaluated for their antibacterial activity against S.aureus by viable cell counting method as already geneous mixture was stirred vigorously at 40 °C for 5 h under nitrogen. After evaporation of the solvent, the crude product reported.3 About 108 cells per ml of S. aureus were exposed to 1 mgml-1 of the oligomers in saline. obtained was reprecipitated from water–tetrahydrofuran to give an a,v-bis(perfluoropropylated) 2-acrylamido-2-methyl- Fluoroalkylated end-capped AMPS homo- and co-polymers were in general inactive. However, of these AMPS polymers, propanesulfonic acid polymer (1.13 g). This polymer exhibited the following spectral characteristics: IR (cm-1) 3448 (OH, perfluoropropylated AMPS–trimethylvinylsilane copolymer [C3F7-(AMPS)x-(CH2CHSiMe3)y-C3F7] was found to show NH), 1641 [C(=O)N+H2-], 1310 (CF3), 1259 (SO3-), 1228 (CF2), 1101 (SO3-); 1H NMR (D2O) d 1.51–2.12 (CH2), 1.60 bacterial activity (from 2.2×108 to 2×102 colony forming units levels).In addition, this copolymer was shown to possess (CH3), 2.29–2.71(CH), 3.65–4.22 (CH2); 19F NMR (D2O, ext. CF3CO2H) d -3.45 (6F), -40.60 (4F), -50.20 (4F). a higher anti-HIV-1 activity (see Table 3). Hitherto, the development of antibacterial cationic materials The other products obtained exhibited the following spectral characteristics. possessing fluoroalkyl segments has been limited.30,31 However, 1522 J.Mater. Chem., 1998, 8(7), 1517–1524C3F7OCF(CF3)-(AMPS)n-CF(CF3)OC3F7. IR (cm-1) 3469 (CH2), 1.35 (CH3), 1.82–2.23 (CH), 2.97–3.76 (CH2); 19F NMR (D2O, ext. CF3CO2H) d -5.06 to -10.70 (21F), -48.24 (2F), (OH, NH), 1647 [C(=O)N+H2-], 1304 (CF3), 1228 (CF2); 1H NMR (D2O) d 1.49–2.11 (CH2), 1.60 (CH3), 2.31–2.69 -51.38 (10F), -71.11 (3F).(CH), 3.61–4.15 (CH2); 19F NMR (D2O, ext. CF3CO2H) d Viscosity measurements -1.56 to -5.10 (16F), -48.00 (4F). The viscosities of aqueous solutions of fluoroalkylated end- C3F7OCF(CF3)CF2OCF(CF3)-(AMPS)n-CF(CF3)OCF2 capped AMPS polymers were measured at 5–50 °C using a CF(CF3)OC3F7. IR (cm-1) 3370 (OH, NH), 1647 [C(= falling-sphere viscometer (Haake Viscometer D1-G).O)N+H2-], 1303 (CF3), 1255 (SO3-), 1226 (CF2), 1101 (SO3-); 1H NMR (D2O) d 1.38–1.83 (CH2), 1.53 (CH3), 1.97–2.30 Typical procedure for gelation test (CH), 3.19–3.57 (CH2); 19F NMR (D2O, ext. CF3CO2H) d -4.42 to -8.42 (26F), -54.0 to -56.26 (6F), -66.78 (2F).A procedure for studying the gel-formation ability was based on a method reported by Hanabusa et al.15 Briefly, weighed fluoroalkylated end-capped AMPS polymer was mixed with C3F7OCF(CF3)CF2OCF(CF3)CF2OCF(CF3)-(AMPS )n- CF(CF3)OCF2CF(CF3)OCF2CF(CF3)OC3F7. IR (cm-1) water or organic fluid in a tube. The mixture was treated under ultrasonic conditions until the solid dissolved.The 3465 (OH, NH), 1641 [C(=O)N+H2-], 1238 (CF2); 1H NMR (D2O) d 1.51–2.12 (CH2), 1.54 (CH3), 2.35–2.71 (CH), resulting solution was kept at 30 °C for 1 h, and then the gelation was assessed visually. When it was formed, the gel 3.65–4.17 (CH2); 19F NMR (D2O, ext. CF3CO2H) d -0.98 to -5.50 (36F), -49.10 (6F), -69.10 (4F). was stable and the tube was able to be inverted without changing the shape of the gel.Similarly, a series of fluoroalkylated end-capped AMPS copolymers were prepared by copolymerizations with fluoro- Metal ion binding by fluorinated AMPS oligomer hydrogel alkanoyl peroxides. These exhibited the following spectral characteristics. Fluoroalkylated end-capped AMPS copolymer hydrogel was swelled with water in a measuring flask.After the addition of C3F7-(AMPS)x-(CH2CHSiMe3)y-C3F7. IR (cm-1) 3409 (OH, the required amount of aqueous metal ion solution into the NH), 1649 [C(=O)N+H2-], 1304 (CF3), 1227 (CF2); 1H NMR flask, the flask was allowed to stand for 1 day at 25 °C. The (D2O) d -0.20–0.46 (CH3), 1.51–2.12 (CH2), 1.54 (CH3), metal-ion concentration of supernatant liquid after the incu- 2.29–2.71 (CH), 3.65–4.22 (CH2); 19F NMR (D2O, ext.bation was spectrophotometrically determined. CF3CO2H) d -3.50 (6F), -40.60 (4F), -50.20 (4F). Antiviral assays C3F7OCF (CF3 ) - ( AMPS )y- (CH2CHSiMe3 )y-CF (CF3) - OC3F7. IR (cm-1) 3453 (OH, NH), 1645 [C(=O)N+H2-], Antiviral activity of the compounds against HIV-1 (HTLBIIIb starin) replication was based on the inhibition of the 1304 (CF3), 1232 (CF2); 1H NMR (D2O) d -0.23–0.36 (CH3), 1.49–2.11 (CH2), 1.54 (CH3), 2.31–2.69 (CH), 3.61–4.15 (CH2); virus-induced cytopathic eVect in MT-4 cells as described previously.27 19F NMR (D2O, ext.CF3CO2H) d -1.58 to -5.12 (16F), -48.00 (4F). Antibacterial assessment C3F7OCF(CF3)CF2OCF(CF3)-(AMPS)x-(CH2CHSiMe3)y- The antibacterial activity of the oligomers was evaluated CF(CF3)OCF2CF(CF3)OC3F7. IR (cm-1) 1647 [C(= against Staphylococcus aureus by a viable cell counting method O)N+H2-], 1227 (CF2); 1H NMR (D2O) d -0.18–0.33 (CH3), as described previously.3 1.61–2.31 (CH2), 1.44 (CH3), 2.33–2.72 (CH), 3.59–4.08 (CH2); 19F NMR (D2O, ext.CF3CO2H) d -3.64 to -7.10 (26F), This work was partially supported by a Grant-in-Aid for -48.91 (6F), -67.48 (2F).Scientific Research No. 09650945 from the Ministry of Education, Science, Sports and Culture, Japan, for which the C3F7-(AMPS)x-(CH2CMeCO2Me)y-C3F7. IR (cm-1) 3452 authors are grateful. (OH, NH), 1641 [C(=O)N+H2-], 1228 (CF2); 1H NMR (D2O) d 1.51–2.13 (CH2, CH3 ), 2.29–2.58 (CH), 3.65–4.20 (CH2, References CH3); 19F NMR (D2O, ext. CF3CO2H) d -3.50 (6F), -40.65 (4F), -50.20 (4F). 1 (a) L. A. Wall, Fluoropolymer, Wiley, New York, 1972, Vol. XXV; (b) P. R. Resnick, Polym. Prepr. (Am. Chem. Soc. Div. Polym. Chem.), C3F7OCF (CF3) - (AMPS)x -(CH2CMeCO2Me)y -CF(CF3)- 1990, 31, 312; (c) N. Nakamura, T. Kawasaki, M. Unoki, K. Oharu, N. Sugiyama, I. Kaneko and G. Kojima, Preprints of the First OC3F7. IR (cm-1) 1645 [C(=O)N+H2-], 1315 (CF3), 1238 Pacific Polymer Conference, Am.Chem. Soc. Div. Polym. Chem., (CF2); 1H NMR (D2O) d 1.49–2.11 (CH2, CH3 ), 2.33–2.70 Maui, HI, Dec. 1989; American Chemical Society, Washington, (CH), 3.60–4.15 (CH2, CH3) 19F NMR (D2O, ext. CF3CO2H) DC, 1989, p. 369; (d) Z.-Y. Yang, A. E. Feiring and B. E. Smart, d -1.62 to -5.00 (16F), -48.30 (4F). J. Am. Chem. Soc., 1994, 116, 4135. 2 (a) H. Sawada, N. Itoh, T. Kawase, M.Mitani, H. Nakajima, C3F7OCF(CF3)CF2OCF(CF3)-(AMPS)x-[CH2CMeCO2- M. Nishida and Y. Moriya, L angmuir, 1994, 10, 994; (b) H. Sawada, K. Tanba, N. Itoh, C. Hosoi, M. Oue, M. Baba, T. Kawase, CH3]y-CF(CF3)OCF2CF(CF3)OC3F7. IR (cm-1) 1642 [C(= M. Mitani and H. Nakajima, J. Fluorine Chem., 1996, 77, 51. O)N+H2-], 1236 (CF2); 1H NMR (D2O) d 1.52–2.26 (CH2, 3 H. Sawada, S. Katayama, M.Oue, T. Kawase, Y. Hayakawa, CH3), 2.33–2.78 (CH), 3.59–4.08 (CH2, CH3); 19F NMR (D2O, M. Baba, T. Tomita and M. Mitani, J. Jpn. Oil Chem. Soc., 1996, ext. CF3CO2H) d -3.60 to -7.10 (26F), -48.96 (6F), 45, 161. -67.48 (2F). 4 H. Sawada, H. A. Ohashi, A. M. Baba, T. Kawase and Similarly, fluoroalkylene unit-containing AMPS polymer Y. Hayakawa, J. Fluorine Chem., 1996, 79, 149. 5 J.F. Elman, B. D. Johs, T. E. Long and J. T. Koberstein, {-[RF-(AMPS)q]p-} was prepared by reaction of a polymeric Macromolecules, 1994, 27, 5341. perfluoro-oxaalkane diacyl peroxide with AMPS. This exhib- 6 S. AV rossman, M. Hartshorne, T. KiV, R. A. Pethrick and ited the following spectral characteristics. R. W. Richards, Macromolecules, 1994, 27, 1588. 7 M. O. Hunt, Jr., A. M. Belu, R.W. Linton and J. M. DeSimone, -[CF(CF3)[OCF2CF(CF3)]nO(CF2)5O-[CF(CF3)CF2O]m- Macromolecules, 1993, 26, 4854. CF(CF3)-(AMPS)q]p-; n+m=3. IR (cm-1) 3492 (OH, NH), 8 S.AV rossman, P. Bertrand, M. Hartshorne, T. KiV, D. Leonard, R. A. Pethrick and R. W. Richards,Macromolecules, 1996, 29, 5432. 1643 [C(=O)N+H2-], 1226 (CF2); 1H NMR (D2O) d 0.82–1.80 J. Mater. Chem., 1998, 8(7), 1517–1524 15239 B.Xu, L. Li, A. Yekta, Z. Masoumi, S. Kanagalingam, 22 Y. Ishikawa, H. Kuwahara and T. Kunitake, J. Am. Chem. Soc., 1989, 111, 8530. M. A. Winnik, K. Zhang, P. M. Macdonald and S. Menchen, L angmuir, 1997, 13, 2447. 23 R. J. Twieg, T. P. Russell, R. Siemens and J. F. Rabolt, Macromolecules, 1985, 18, 1361. 10 J.Wang, G. Mao, C. K. Ober and E. J. Kramer, Polym. Prepr.(Am. Chem. Soc., Div. Polym. Chem.), 1997, 38, 953. 24 H. Sawada, E. Sumino, M. Oue, M. Mitani, H. Nakajima, M. Nishida and Y. Moriya, J. Chem. Soc., Chem. Commun., 1994, 11 Z. Su, D. Wu, S. L. Hsu and T. J. McCarthy, Polym. Prepr. (Am. Chem. Soc., Div. Polym. Chem.), 1997, 38, 951. 143. 25 H. Sawada, E. Sumino, M. Oue, M. Baba, T. Kira, S. Shigeta, 12 A. Laschewsky and I.Zerbe, Polymer, 1991, 32, 2070. 13 H. Sawada, S. Katayama, Y. Nakamura, T. Kawase and M. Mitani, H. Nakajima, M. Nishida and Y. Moriya, J. Fluorine Chem., 1995, 74, 21. Y. Hayakawa, Polymer, 1998, 39, 743. 14 (a) H. Sawada, Y.-F. Gong, Y. Minoshima, T. Matsumoto, 26 M. Baba, R. Pauwels, J. Balzarini, J. Arnout, J. Desmyter and E. De Clercq, Proc. Natl. Acad. Sci. USA, 1988, 85, 6132. M. Nakayama, M. Kosugi and T. Migita, J. Chem. Soc., Chem. 27 M. Baba, T. Kira, S. Shigeta, T. Matsumoto and H. Sawada, Commun., 1992, 537; (b) H. Sawada, Y. Minoshima and J. Acquir. Immun. Defic. Syndr., 1994, 7, 24. Y. Nakajima, J. Fluorine Chem., 1992, 65, 169. 28 H. Sawada, A. Ohashi, M. Oue, M. Baba, M. Abe, M. Mitani and 15 (a) K. Hanabusa, R. Tanaka, M. Suzuki, M. Kimura and H. Shirai, H. Nakajima, J. Fluorine Chem., 1995, 75, 121. Adv. Mater., 1997, 9, 1095; (b) K. Hanabusa, K. Okui, K. Karaki, 29 H. Sawada, K. Yamamoto, M. Oue, T. Kawase, Y. Hayakawa, M. Kimura and H. Shirai, J. Colloid Interface Sci., 1997, 195, 86. M. Baba and M. Mitani, J. Jpn. Oil Chem. Soc., 1996, 45, 37. 16 K. Hanabusa, Y. Naka, T. Koyama and H. Shirai, J. Chem. Soc., 30 H. Sawada, A. Wake, T. Maekawa, T. Kawase, Y. Hayakawa, Chem. Commun., 1994, 2683. T. Tomita and M. Baba, J. Fluorine Chem., 1997, 83, 125. 17 Y. Osada and M. Takase, Nippon Kagaku Kaishi, 1983, 439. 31 H. Sawada, K. Tanba, T. Tomita, T. Kawase, M. Baba and T. Ide, 18 T. Kunitake, Y. Okahata and S. Yasunami, J. Am. Chem. Soc., J. Fluorine Chem., 1997, 84, 141. 1987, 104, 5547. 32 (a) H. Sawada and M. Nakayama, J. Fluorine Chem., 1990, 51, 117; 19 T. Kunitake and N. Higashi, J. Am. Chem. Soc., 1985, 107, 692. (b) H. Sawada, M. Yoshida, H. Hagii, K. Aoshima and 20 T. Kunitake and N. Higashi, Macromol. Chem. Phys. Suppl., 1985, M. Kobayashi, Bull. Chem. Soc. Jpn., 1986, 59, 215. No. 14, 81. 21 Y. Ishikawa, K. Kuwahara and T. Kunitake, Chem. L ett., 1989, 1737. Paper 8/02210K; Received 20th March, 1998 1524 J. Mater. Chem., 1998, 8(7), 1517–1524

ISSN:0959-9428    

DOI:10.1039/a802210k

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

J O U R N A L O F C H E M I S T R Y Materials Atom transfer polymerisation of methyl methacrylate: use of chiral aryl/alkyl pyridylmethanimine ligands; with copper(I ) bromide and as structurally characterised chiral copper(I ) complexes David M. Haddleton,*† David J. Duncalf, Dax Kukulj, Alex M. Heming, Andrew J. Shooter and Andrew J. Clark Department of Chemistry, University of Warwick, Coventry, UK CV 4 7AL The use of chiral catalysts in the living radical polymerisation of methyl methacrylate via atom transfer polymerisation (ATP) has been investigated in an eVort to control the stereochemistry of the polymer backbone. Two enantiomerically pure chiral catalysts have been prepared and used in the ATP of methyl methacrylate: the structurally characterised complex, bis[N-(1-phenylethyl)-2- pyridylmethanimine]copper(I) tetrafluoroborate, [Cu(C14H14N2)2][BF4], and the reaction product of copper(I) bromide with N- (1-cyclohexylethyl)-2-pyridylmethanimine, [Cu(C14H20N2)2][Br].Both catalysts were found to be suitable for the ATP of methyl methacrylate in conjunction with either ethyl 2-bromo-2-methylpropanoate in xylene at 90 °C or 4-methoxybenzenesulfonyl chloride in diphenyl ether at 90 °C.The system yields polymer of relatively narrow polydispersity; however, the use of these chiral catalysts did not significantly aVect the stereochemistry of the polymer backbone. This may be due to the chiral centre being too distant from the propagating site to exert any influence over the monomer addition step, or that the reaction proceeds via a completely free-radical mechanism.This is, however, the first time a discrete, structurally characterised copper complex has been used as an eVective ATP catalyst. The bond lengths of the complexed ligand indicate the oxidation state of the copper to be+1 in the complex. Controlled polymerisation of acrylic and methacrylic mon- Atom transfer polymerisation (ATP) has emerged as an eVective, living, transition metal mediated polymerisation for a omers is of continuing interest,1 providing the potential for a wide range of novel materials from currently available mon- wide range of vinyl monomers; ATP has been derived from atom transfer chemistry developed for a range of omers.This has been the objective of many research programs which have produced a number of diVerent systems to achieve organic transformations in more conventional organic synthesis, 6–8 in particular atom transfer cyclisation.9 This was first this aim, viz. living/pseudo-living polymerisation of methacrylates. 1 A major challenge is to avoid complications arising demonstrated, almost simultaneously, by Sawamoto and Matyjaszewski. Sawamoto described the use of RuCl2(PPh3)3 from side reaction with the ester group, or deprotonation of the polymer backbone in the case of poly(acrylics).These in conjunction with strong Lewis acids for the living polymerisation of methyl methacrylate.10–12 In a related system, secondary reactions which destroy the living nature of the polymerisations have proven problematic as most develop- Matyjaszewski has been investigating the use of copper(I) halides with bipyridine and 4,4¾-dialkyl substituted bipyridine ments have centred around anionic or anionic-type polymerisations.For example, group transfer polymerisation (GTP) has ligands for the living polymerisation of styrene, methacrylates and acrylates.13–17 Teyssie has used a well characterised NiII been perhaps the most widely studied system in recent years, and utilises silyl ketene acetals in conjunction with nucleophilic compound for the living polymerisation of alkyl methacrylates. 18 In all of these systems, polymerisation is thought to catalysts to control polymerisation, resulting in polymers with very narrow polydispersity indexes (PDI) in tetrahydrofuran be mediated by a formal Mt n/Mt n+1 redox couple and initiated by homolysis, or at least partial homolysis, of an alkyl halide at temperatures up to 80 °C.2,3 GTP, as with most other systems which rely on anionic type propagation, is highly [Scheme 1(a)].Percec has extended the range of initiators to include sulfonyl halides which, due to the ease of homolysis of sensitive to protic impurities necessitating the use of rigorously dried and purified monomers and solvents. It is for this reason the sulfur–halogen bond, give a fast rate of initiation relative to propagation for virtually all monomers, and are so called that we have been focusing on living free-radical polymerisation, which should be tolerant to many diVerent impurities ‘universal initiators’ for ATP.19–21 We have been concentrating on a copper based ATP pro- and functional monomers/solvents rendering the system more commercially viable.Although stable free radical mediated cess22–24 which uses SchiV bases of type 1 as electron accepting ligands on the metal.25 These ligands stabilise formal CuI, polymerisation using TEMPO and related nitroxide stable organic radicals is relatively successful for styrene and substi- relative to CuII, and promote eVective ATP.Ligands based on tuted styrene,4,5 it is less so for acrylics and methacrylics. Also it is relatively easy to achieve living polymerisation of styrene by conventional anionic polymerisation, which is the basis for commercially available styrene-butadiene block copolymers. We are particularly interested in achieving controlled polymerisation of alkyl methacrylates and methacrylates containing functional groups in the ester side chain which can be used to synthesise polymers with a wide range of properties for diVering applications.The use of transition metal mediated polymerisation is very attractive for this type of monomer. ( a) Rn—X + CuIX Rn CuIIX2 + Monomer Addition Rn—X + CuIX Monomer Addition Rn X CuIIX ( b) • Scheme 1 †E-mail: msrg@csv.warwick.ac.uk J.Mater. Chem., 1998, 8(7), 1525–1532 1525(R)-,(S)- and (±)-N-(1-Phenylethyl )-2-pyridylmethanimine [(R)-2,(S)-2 and (±)-2] A typical preparation was carried out as follows: (R)-1-phenylethylamine (2.5 g, 0.02 mol) was added dropwise to a stirred solution of pyridine-2-carbaldehyde (2.2 g, 0.02 mol) in Et2O (ca. 20 ml).To the reaction mixture was added MgSO4 (1–2 g). The reaction was stirred at room temperature for 4 h after which time the solvent was removed in vacuo yielding (R)-NN N R R = n-alkyl 1 N N 2 * N N * 3 (1-phenylethyl)-2-pyridylmethanimine (3.8 g, 90%) as a clear, slightly yellow, oil, bp 107 °C at 0.5 mmHg; dH(250 MHz, the general structure 1 are extremely easy to synthesize by a 298 K, CDCl3) 8.50 (d, 1H, Py-H), 8.36 (s, 1H, CHNN), 7.96 simple condensation of pyridine-2-carbaldehyde with an appro- (d, 1H, Py-H), 7.57 (t, 1H, Py-H), 7.20 (m, 6H, Py-H/Ph-H), priate primary amine, the reaction proceeding eYciently and 4.51 [q, 1H, Ph(Me)CH], 1.50 (d, 3H, Me); dC(250 MHz, quickly at ambient temperature and yielding the products as 298 K, CDCl3) 159.6 (CHNN), 153.9, 148.4, 135.6, 127.7, 125.9 yellow oils which are readily purified by distillation.The wide (Py), 143.5, 126.1, 123.2, 120.7 (Ph), 68.7 (CH3), 23.8 variety of commercially available primary amines gives a large [CH(Me)Ph]; nmax(neat)/cm-1 1646s, 1587s, 1567s (CNN); family of potential ligands and allows control over catalyst [a]D-1.342[(R)-2], +1.321[(S)-2], 0.000 [(±)-2] (c 1, properties such as solubility in organic media and the CuI/CuII acetone).redox potential. ATP has already been demonstrated to give excellent control (R)- and (S)-N-(1-Cyclohexylethyl-2-pyridylmethanimine [(R)- over molecular weight and polydispersity. In addition it would 3 and (S)-3] be desirable to be able to control the stereochemistry of the A typical preparation was carried out as follows: (R)-1-cyclopolymer backbone.The stereochemistry of the polymer can hexylethylamine (6.5 g, 0.051 mol) was added dropwise to a have significant eVects on materials properties such as the stirred solution of pyridine-2-carbaldehyde (5.47 g, 0.051 mol) glass transition temperature (Tg), as well as other mechanical in Et2O (ca. 30 ml). To the reaction mixture was added MgSO4 properties.Previously we have suggested that ATP of MMA (1–2 g). The reaction was stirred at room temperature for 4 h utilising a CuI catalyst with ligands based on 1 may not after which time the solvent was removed in vacuo yielding proceed via a free-radical mechanism, and we have postulated (R)-N-(1-cyclohexylethyl-2-pyridylmethanimine (9.4 g, 85%) as a concerted propagation mechanism as being a possibility a clear, slightly yellow, oil, bp 96 °C at 0.5 mmHg; dH(250 MHz, where the halogen is not fully abstracted from the propagating 298 K, CDCl3) 8.48 (d, 1H, Py-H), 8.17 (s, 1H, CHNN), 7.85 polymer24 [Scheme 1(b)].Indeed, Teyssie has suggested, in his (d, 1H, Py-H), 7.54 (t, 1H, Py-H), 7.11 (m, 1H, Py-H), 2.97 [q, studies of the ATP of MMA using a NiII complex, that 1H, Cy(Me)CH], 1.65 (m, 4H, Cy-H), 1.36 (m, 1H, Cy-H), propagation may occur via a coordinate mechanism, whereby 1.09 [s, 3H, C(H) (Cy)CH3], 1.07 (m, 4H, Cy-H), 0.81 (m, 4H, the halide from the initiator remains partially bonded to both Cy-H); dC(250 MHz, 298 K, CDCl3) 159.4 (CHNN), 154.7, the growing polymer chain and the metal centre.18 If the 149.0, 136.1, 124.1, 121.0 (Py), 71.4 (CH3), 43.4, 29.6, 26.0, 19.5 catalyst is bound to the growing polymer chain, or even in (Cy), 26.1 [CH(Me)Cy]; nmax(neat)/cm-1 1647s, 1588s, 1568s the vicinity as a ‘radical cage complex’ it may be possible to (CNN); [a]D -2.003 [(R)-3], +2.022 [(S)-3] (c 1, acetone). influence the stereochemistry of the polymer backbone by the use of chiral metal catalysts, produced from enantiomerically (R,R)-, (S,S)- and (±)-Bis[N-(1-phenylethyl )-2- pure, chiral SchiV base ligands.It is noted that the stereochempyridylmethanimine) copper(I ) tetrafluoroborate istry of all poly(methyl methacrylate) produced via ATP to [Cu(C14H14N2)2][BF4] [(R)-4, (S)-4 and (±)-4] date (including CuI, RuII and NiII systems) has been reported to be consistent with free radical polymerisation, i.e.described [Cu(MeCN)4][BF4] was prepared by the method of Kubas.27 by Bernoullian statistics with the persistence ratio close to A typical preparation was carried out as follows: to unity. [Cu(MeCN)4][BF4] (2.0 g, 6.47 mmol) in MeOH (35 ml ) The work outlined in this paper investigates the possibility was added (R)-N-(1-phenylethyl-2-pyridylmethanimine, (R)-2 of using the chiral SchiV bases 2 and 3 with CuI halides as (2.7 g, 12.94 mmol).The reaction immediately became a deep ATP catalysts. Results for both the R and S enantiomers of 2 red–brown colour and was stirred for 4 h. After this time the and 3, as well as for a racemic mixture in the case of 2, are solution was filtered and concentrated to ca. 20 ml and then reported. In all of the studies of copper(I) mediated ATP allowed to cool slowly to -40 °C, whereupon red crystals of published to date, the active copper complex has been formed [Cu(C14H14N2)2][BF4] formed in 70% yield (2.57 g), mp in situ by addition of excess ligand to a suspension of CuI 96–99 °C; dH(250 MHz, 298 K, [2H6]acetone) 8.85 (s, 1H, halide in the reaction solution.We report the use of well CHNN), 7.96 (m, 1H, Py-H), 7.86 (m, 1H, Py-H), 7.42 ( br, defined, fully characterised copper(I) compounds (arising from 1H, Py-H), 7.12, 6.99 ( br, 5H, Ph-H), 4.91 [br, 1H, 2) which were isolated by recrystallisation, characterised and Ph(Me)CH], 1.55 (br, 3H, Me); dC(250 MHz, 298 K, subsequently used as discrete compounds in the polymeris- [2H6]acetone) 161.04 (CHNN), 151.89, 149.98, 139.16, 125.94, ation reaction. 120.36 (Py), 142.98, 129.4, 127.1, 115.72 (Ph), 67.98 [C(CH3)PhH], 38.47 [CCH3(Ph)H]; nmax(Nujol)/cm-1 1612m, 1586s (CNN). Experimental General Crystal structure determinations A suitable crystal of (R)-4 was quickly glued to a quartz fiber, Methyl methacrylate (Aldrich, 99%) was purified by passing through a column of activated basic alumina to remove coated in dry Nujol and cooled in the cold nitrogen gas stream of the diVractometer. The structure was solved by direct inhibitor.Copper(I) bromide (Aldrich, 98%) was purified according to the method of Keller and WycoV.26 Xylene methods. Anisotropic thermal parameters were used for all non-H atoms whilst hydrogen atoms were inserted at calculated (Fisons, 99.8%), ethyl 2-bromo-2-methylpropanoate (Aldrich, 98%), 1-phenylethylamine (Aldrich, 98%), 1-cyclohexylethyl- positions and fixed, with isotropic thermal parameters (U= 0.08 A° 3), riding on the supporting atom.The structure solutions amine (Aldrich, 98%) and 4-methoxybenzenesulfonyl chloride (Avocado, 98%) were used as received. were carried out using SHELXTL28 version 5.0 software on a 1526 J.Mater. Chem., 1998, 8(7), 1525–1532Silicon Graphics Indy workstation, refinements were carried out using SHELXTL29 software, minimising on the weighted R factor wR2. Full crystallographic details, excluding structure factors, have been deposited at the Cambridge Crystallographic Data Centre (CCDC). See Information for Authors, J. Mater. Chem., 1998, Issue 1.Any request to the CCDC for this material should quote the full literature citation and the reference number 1145/96. Typical polymerisation procedures Ethyl 2-bromo-2-methylpropanoate initiator in xylene solvent. Compound 4 (0.25 g, 4.65×10-4 mol)] [or CuIBr (0.066 g, 4.65×10-4 mol) and 3 (0.30 g, 1.40×10-3 mol)] was placed into a nitrogen filled flask and subjected to three vacuum/ nitrogen fill cycles.An aliquot of MMA (5.0 ml 4.65×10-2 mol) together with xylene solvent (5.0 ml) was added to the flask, which was then placed into an oil bath at 90 °C with N N N N Cu * * + N N N N Cu * * + 4 5 O OEt Br 6 7 CH3O SCl O O 8 N N C5H11 stirring. To this mixture was added 6 (0.09 g, 4.65×10-4 mol) to initiate polymerisation. Periodically, 1–2 ml samples were tems23).Together with the high Mn, this suggests ineYcient removed for conversion and molecular weight analysis. initiation and/or side reactions such as termination in the polymerisation. 4-Methoxybenzenesulfonyl chloride initiator in Ph2O solvent. We have previously observed that the substituents on the The procedure detailed above was followed, except that Ph2O SchiV base ligand play an important role in determining the (5.0 ml) was substituted for xylene and 7 (0.095 g, 4.65×10-4 position of the equilibrium between dormant and active polymol) was used as the initiator.mer chains, which is central to the reaction mechanism30 (Scheme 1). The 1-phenylethyl substituent used in this case is more sterically demanding than an n-alkyl group and this may Polymer analysis.Polymer conversion was determined by be the reason for the poor molecular weight control observed gravimetry for polymerisations performed in xylene and by 1H when the reaction is performed in xylene. The steric require- NMR spectroscopy in the case of reactions performed in Ph2O. ments of the ligands are clearly an important factor in the At the end of the reaction the polymer was precipitated into relative stabilities of CuI and CuII, as are electronic factors.light petroleum (bp 40–60 °C) and dried in a vacuum oven at When the initiator is changed to 4-methoxybenzenesulfonyl 70 °C for 8 h, for stereochemical analysis by 13C NMR spec- chloride 7 with diphenyl ether as solvent, linear first order rate troscopy. Molecular weight distributions were measured using plots with an apparent induction period of approximately size exclusion chromatography (SEC) on a system equipped 30 min are obtained (Fig. 3).The rate of reaction is again with a guard column, one 30 cm mixed E column (Polymer independent of the catalyst stereochemistry. Mn increases with Laboratories) and a diVerential refractive index detector, using conversion and shows much better agreement with that tetrahydrofuran at 1 ml min-1 as the eluent.Poly(methyl expected for 100% initiator eYciency (Fig. 4) compared with methacrylate) standards in the range (6×104 to 200 g mol-1) the previous case. Thus, the initiator eYciency is higher with were used to calibrate the SEC analysis. Quantitative 13C 7 than 6. Also, the PDIs are much narrower in all cases (ca.NMR spectra were obtained on a Bruker AM400, using a 1.2) and PDI narrows as conversion increases (Table 2). Both Bruker inverse gated acquisition routine, for the analysis of of the results are consistent with Percec’s20 suggestion that the polymer microstructure. rate of initiation of MMA using 7 is fast relative to the rate of propagation which results in narrower PDIs.The rate of reaction using 7 in diphenyl ether is slightly Results and Discussion lower (ca. 30%) than that using 6 in xylene (Table 3). This diVerence in rate is most likely due to the eVect of the solvent Polymerisations on the position of the equilibrium shown in Scheme 1, resulting in a diVerent concentration of active species and hence a The polymerisation of MMA with the three forms of 4 [(i.e.(R)-4, (S)-4 and (±)-4] and with 6 as initiator in xylene at change in rate. Diphenyl ether may act as a weakly coordinating solvent which will naturally aVect the nature and/or rate 90 °C gave reproducible number average molecular masses, Mn, with PDI between 1.4 and 1.6 (Table 1). In each case the of formation of the active copper species.There may also be a small diVerence due to the diVerent initiators used. A lower first order rate plot (Fig. 1) shows a linear slope, indicating the number of active species remains constant throughout the concentration of active species would result in a lower rate of termination due to normal radical–radical reactions and this, reaction and that the rate of polymerisation is independent of the stereochemistry of the catalyst. Each experiment exhibited coupled with the faster rate of initiation20 from 7, results in better molecular weight control.an apparent induction period of approximately 35 min. These induction periods have been observed previously in similar The polymerisation of MMA in xylene, mediated by copper( I) catalysts made in situ by mixing CuIBr and (R)-3 or (S)- systems and are ascribed to the time required to establish the equilibrium shown in Scheme 1.It should be noted that the 3 [resulting in the catalyst (R)-S or (S)-5] and using 6 as the initiator again gave reproducible rates of polymerisation apparent induction time may actually be S-shaped due to a slowly increasing rate of reaction during the time taken to (Fig. 5). The rate of polymerisation is independent of the catalyst stereochemistry and the first order kinetic plot is linear establish the equilibrium. Although Mn increases with conversion with all three catalysts, as expected for a living polymeris- with an apparent 30 min induction period (Fig. 5).Mn increases linearly with conversion (Fig. 6), but is slightly higher than ation, the actual values are consistently higher than expected for 100% initiator eYciency (Fig. 2). In addition, the PDI is the theoretical Mn, similar to the behaviour seen with 4. The rate of polymerisation is slightly slower than with 4 (Table 3), relatively broad (1.4–1.6 as compared to 1.2 for similar sys- J. Mater. Chem., 1998, 8(7), 1525–1532 1527Table 1 Experimental results for the polymerisation of MMA using ATP with (R)-4, (S)-4 and (±)-4 and ethyl 2-bromo-2-methylpropanoate initiator 6 in xylene at 90 °C (R)-4 (S)-4 (±)-4 t/min conversion Mn PDI conversion Mn PDI conversion Mn PDI 60 0.051 3200 1.40 0.060 4030 1.42 0.056 2290 1.38 120 0.152 4030 1.41 0.179 4780 1.47 0.180 3250 1.38 180 0.242 4310 1.46 0.249 5010 1.65 0.250 4340 1.63 240 0.331 4510 1.52 0.333 5420 1.54 0.370 5320 1.51 300 0.408 5420 1.42 0.412 6200 1.46 360 0.505 6430 1.48 0.503 6570 1.44 Fig. 3 First order rate plot for the polymerisation of MMA by ATP Fig. 1 First order rate plot for the polymerisation of MMA by ATP using (&) (R)-4, (#) (S)-4 and (+) (±)-4 catalysts at 90 °C in xylene using (&) (R)-4, (#) (S)-4 and (+) (±)-4 catalysts at 90 °C in diphenyl ether using 7 as initiator using 6 as an initiator Fig. 4 Dependence of molecular weight on conversion for the Fig. 2 Dependence of molecular weight on conversion for the polymerisation of MMA by ATP using (&) (R)-4, (#) (S)-4 and (+) polymerisation of MMA by ATP using (&) (R)-4, (#) (S)-4 and (+) (±)-4 catalysts at 90 °C in diphenyl ether using 7 as an initiator (±)-4 catalysts at 90 °C in xylene using 6 as initiator Polymer microstructure however the PDI is narrower (Table 4).When 7 is used as the initiator in diphenyl ether solvent a broadening of PDI is The stereochemistry of polymers from all polymerisations was determined by analysis of 13C NMR spectra (triad region, observed at longer reaction times which may to be due to termination reactions. The rate of reaction is significantly 44–46 ppm, corresponding to the backbone quaternary carbon), according to the assignment of Peat and Reynolds.31,32 lower than in the other systems and may suggest that the equilibrium shown in Scheme 1 is not positioned for eVective Table 8 shows the triad fraction, diad fraction and persistence ratio r [eqn.(1)] of PMMA prepared by ATP with (R)-4, (S)- ATP.The first order rate plot shows that the rate of reaction is unaVected by the stereochemistry of the catalyst (Fig. 7). 4, (±)-4, (R)-5, (S)-5 and a non-chiral ATP ligand, and for conventional free-radical polymerisation. The stereoregularity Again Mn increases with conversion as would be expected for a controlled polymerisation (Fig. 8). A diVerence in the poly- is similar in each case, indicating that the use of these chiral SchiV base ligands has no significant influence on the stereo- merisations mediated by 4 and 5 is that the counter ion to the copper is BF4- and Br-, respectively.This may influence the chemistry of monomer addition. The lack of stereochemical control can be explained in several ways: (i) the chiral centre polymerisation reaction since the Br- may coordinate to the copper complex, whereas the BF4- is non coordinating. on the ligand may be too far away from the monomer addition 1528 J. Mater.Chem., 1998, 8(7), 1525–1532Table 2 Experimental results for the polymerisation of MMA using ATP with (R)-4, (S)-4 and (±)-4 and 4-methoxybenzenesulfonyl chloride 7 as initiator in Ph2O at 90°C (R)-4 (S)-4 (±)-4 t/min conversion Mn PDI conversion Mn PDI conversion Mn PDI 60 0.06 1170 1.36 0.04 860 1.36 0.04 640 1.43 120 0.11 1590 1.30 0.10 1220 1.25 0.10 990 1.37 180 0.19 1820 1.27 0.18 1690 1.20 0.20 1930 1.27 240 0.25 2130 1.26 0.26 2570 1.21 0.26 2070 1.16 300 0.28 2320 1.21 0.35 2970 1.22 0.31 2360 1.17 Table 3 Comparison of the slopes from first order rate plots of diVerent Table 4 Experimental results for the polymerisation of MMA using ATP with (R)-5 and (S)-5 and ethyl 2-bromo-2-methylpropanoate systems initiator 6 in xylene at 90 °C kp[R ]/s-1 (R)-5 (S)-5 catalyst 6 7 t/min conversion Mn PDI conversion Mn PDI (R)-4, (S)-4, (±)-4 3.46×10-5 2.35×10-5 (R)-5, (S)-5 2.37×10-5 0.52×10-5 60 0.081 2070 1.15 0.052 1300 1.13 1140 0.784 9480 1.42 0.753 8780 1.44 1440 0.895 11800 1.31 0.857 12500 1.37 Fig. 5 First order rate plot for the polymerisation of MMA by ATP using (&) (R)-5 and (#) (S)-5 catalysts at 90 °C in xylene using 6 as Fig. 7 First order rate plot for the polymerisation of MMA by ATP an initiator using (&) (R)-5 and (#) (S)-5 catalysts at 90 °C in diphenyl ether using 7 as initiator Fig. 6 Dependence of molecular weight on conversion for the Fig. 8 Dependence of molecular weight on conversion for the polymerisation of MMA by ATP using (&) (R)-5 and (#) (S)-5 polymerisation of MMA by ATP using (&) (R)-5 and (#) (S)-5 catalysts at 90 °C in xylene using 6 as initiator catalysts at 90 °C in diphenyl ether using 7 as an initiator site to play a role; (ii) the catalyst may not be rigid enough to anticipate that a more strongly complexed and rigid chiral exert significant stereocontrol on the growing polymer chain SchiV base ligand may be able to provide some degree of since the chirality is in a very flexible ligand (also the complexed stereocontrol in ATP and this is currently under study in our ligand is in dynamic equilibrium with the ligand in solution30); laboratory.(iii ) the ATP reaction may be purely free-radical and so the catalyst would play no part in the monomer addition step. We r=2(m) (r)/(mr) (1) J.Mater. Chem., 1998, 8(7), 1525–1532 1529Table 5 Experimental results for the polymerisation of MMA using Table 7 Summarised crystallographic data for (R)-4 ATP with (R)-5 and (S)-5 and 4-methoxybenzenesulfonyl chloride 7 as initiator in Ph2O at 90°C Crystal parameters Formula C56H56Cu2B2F8N8 M 1141.79 (R)-5 (S)-5 Crystal system Tetragonal Space group P4(3) t/min conversion Mn PDI conversion Mn PDI a/A° 16.9382(4) b/A ° 16.9382(4) 60 0.005 484 1.29 0.004 698 1.16 120 0.011 875 1.24 0.01 1040 1.23 c/A ° 19.5734(5) a(°) 90 180 0.052 943 1.23 — — — 780 — — — 0.221 2420 1.41 b(°) 90 c(°) 90 1320 0.334 3170 1.53 0.313 3890 1.66 reflections 8192 U/A ° 3 5615.7(2) Z 4 dimensions /mm 0.4×0.4×0.3 Dc/g cm-3 1.350 m(Mo-Ka/mm-1 0.828 T /K 180 Data collectiona Data collected h -17 to 21 k -22 to 22 l -26 to 20 total reflections 34133 independent reflections 12021 observed reflections [Foµ4s(Fo)] 6965 h range (°) 1.20 to 28.58 F(000) 2352 Refinement Rb 0.0680 wR2c 0.1999 S 0.914 D/e A° -3 (max, min)d 1.062, -0.340 T (max, min)e 1.000, 0.7479 Flack parameter 0.03(2) Fig. 9 Crystal structure of the ATP catalyst (R)-4 used in polymeris- weighting scheme a,bf 0.087, 12.428 ation studies. The BF4- counter ion is omitted for clarity. aData collected on a Siemens 3 circle diVractometer equipped with a SMART CCD area detector; graphite-monochromated Mo-Ka radi- Crystal structure discussion ation (l=0.71073 A ° ). bR=S|Fo-Fc|SFo [for Foµ4s(Fo)].cwR2= [S[w(Fo2-Fc2)2]/S[w(Fo2)2]]1/2 for all data. dPeaks of unassigned Diazobutadiene (DAB) ligands have been shown to stabilise residual electron density. eBy SADABS. fw-1=s2(Fo2)+aP+bP, low formal oxidation states of the transition metals.33 The where P=[max(Fo2,0)+2Fc2]/3, where max(Fo2,0) indicates that the pyridylmethanimine ligands 2 and 3 used in this work have larger of Fo2 or 0 is taken, a and b are values set by the program.similar electronic characteristics to DAB type ligands, with bipyridine based ligands also thought to be similar. In general, it has been assumed in ATP that the added catalyst contains each ButDAB ligand having a formal charge of -1, YbIII(ButDAB-)3.34 This diVerence in the behaviour of DAB copper with a formal oxidation number of +1 since it is produced from the complexation of CuI and neutral ligands; ligands, on the one hand stabilising low oxidation states and on the other causing one electron oxidation of the metal by however, this may be an oversimplification.Recent studies of complexes of ytterbium and tert-butyldiazabutadiene the ligand, prompted us to investigate the actual oxidation state of copper complexes used in this work by examination (ButDAB) have shown that when zero valent Yb0 is ligated by three ButDAB ligands, although the overall complex is neutral, of the structure of the complex using X-ray crystallography.The crystal structure of (R)-4 was determined and it was the formal oxidation state of the Yb metal centre is +3, with Table 6 Comparison of fractions of diads and persistence ratio of poly(MMA) prepared under diVerent conditions stereochemistryb initiatora catalyst mm mr rr m r r 6 (R)-4 0.0315 0.362 0.607 0.212 0.788 0.925 6 (S)-4 0.0349 0.367 0.598 0.219 0.781 0.930 6 (±)-4 0.0357 0.374 0.590 0.223 0.777 0.926 7 (R)-4 0.0256 0.378 0.597 0.214 0.786 0.892 7 (S)-4 0.0322 0.372 0.596 0.218 0.782 0.917 7 (±)-4 0.0337 0.367 0.600 0.217 0.783 0.927 6 (R)-5 0.0440 0.385 0.572 0.236 0.764 0.939 6 (S)-5 0.0354 0.373 0.592 0.222 0.778 0.926 7 (R)-5 0.0269 0.381 0.593 0.217 0.783 0.893 7 (S)-5 0.0310 0.407 0.562 0.235 0.766 0.882 6 8c 2.77 34.8 62.4 0.202 0.798 0.925 AIBNd — 4.82 36.5 58.7 0.231 0.769 0.973 a6, reaction in xylene; 7, reaction in diphenyl ether; AIBN=2,2¾-azoisobutyronitrile, reaction in toluene.bmm=isotactic triads, mr=atactic triads, rr=syndiotactic triads, m=meso diads, r=racemic diads. cATP with nonchiral catalyst, CuIBr and N-n-pentyl-2-pyridylmethanimine ligand 8. The ratio of MMA5CuIBr5N-n-pentyl-2-pyridylmethanimine: ethyl 2-bromo-2-methylpropanoate was 100515251, reaction temperature=90 °C. dConventional free-radical polymerisation: 25% w/w MMA in xylene, [AIBN]=0.027 mol l-1, 90°C, precipitated from MeOH, Mn=14600, PDI=1.74. 1530 J. Mater. Chem., 1998, 8(7), 1525–1532Table 8 Comparison of average bond length and angle data for [CuI(L)2]+ complexes bond length/A ° angle (°) counter ligand ion CMNacyclic CMNcyclic CMC CuMN NMCuMN dihedral reference N-(1-phenylethyl)-2-pyridylmethanimine (R)-4a,b BF4- 1.282 1.341 1.457 2.040 81.8 90.3 this work 1.273 1.370 1.456 2.030 82.3 94.9 N-tert-butyl-2-pyridylmethanimine BF4- 1.270(4) 1.349(4) 1.467(5) 2.035 81.89(10) 81.9 37 1.263(4) 1.352(4) 1.470(4) 2.035 81.84(10) tert-butyldiazabutadiene Br- 1.268(7) 1.290(7)c 1.453(8) 2.025(4) 82.2(2) 89.5 37 2,2¾-bipyridine ClO4- 1.325(14) 1.385(15) 1.440(15) 2.021(11) 81.5(4) 75.2 35 4,4¾,6,6¾-tetramethyl-2,2¾-bipyridineb Cl- 1.347 1.360 1.488 2.040 81.35 68 36 1,10-phenanthrolineb CuBr2- 1.366 1.352 1.428 2.039 82.2 76.8 38 2,9-dimethyl-1,10-phenanthrolineb NO3- 1.360 1.361 1.450 2.063 83.4 85.7 39 YbIII(ButDAB-)3 c,d — 1.61(6) 1.51(6) 1.39(3) 34 aThere are two molecules in the asymmetric unit.bLengths are averaged (no estimated standard deviations for averaged data). cSecond acyclic bond. dThe Yb complex is included as an example of a DAB ligand with a formal -1 charge. found that the asymmetric unit contains two similar molecules, We would like to thank the EPSRC (D.J. D. and D. K. GR/K90364, GR/L10314, A. M. H. and A. J. S. studentships). one of which is shown in Fig. 9. The important characteristics are summarised in Table 7. The coordinating nitrogen atoms surround the metal centre in a distorted tetrahedral arrangement with the intra-ligand dihedral angles, between the mean References planes defined by the central copper atom and each pair of bidentate nitrogen atoms, measured at 90.3 and 94.9 °.The 1 T. P. Davis, D. M. Haddleton and S. N. Richards, J. Macromol. Sci.-Rev. Chem. Phys., 1994, C34, 243. CuMN distances range from 2.009(5)–2.050(6) A ° , with an 2 O.W. Webster, W. R. Hertler, D. Y. Sogah, W. B. Farnham and average distance of 2.040 A ° . The CNN and CMC bond lengths T. V. Rajan Babu, J. Am. Chem. Soc., 1983, 105, 5706. for (R)-4 are consistent with those found in similar neutral free 3 O. W.Webster, Science, 1991, 887. ligands. These bond lengths indicate that the ligands are 4 M. K.Georges, R. P. N. Veregin and G. K. Hamer, T rends Polym. neutral and the tetrahedral structure of the complex indicates Sci., 1994, 2, 66. that the formal charge of the copper centre is +1. Additional 5 M. K. Georges, R. P. N. Veregin and P. M. Kazmaier, G. K. Homer, Macromolecules, 1993, 26, 2987. material to that shown in Table 7, comprising the atomic 6 D. Bellus, Pure Appl.Chem., 1985, 1827. coordinates, thermal parameters and all bond lengths and 7 J. Iqbal, B. Bhatia and N. K. Nayyar, Chem. Rev., 1994, 94, 519. angles, is available from the Cambridge Crystallographic 8 B. Giese, Radicals in Organic Synthesis: Formation of Carbon- Data Centre. Carbon Bonds, Pergamon, Oxford, 1988. Table 8 compares bond length and angle data for compound 9 K. L. Salazar, M.A. Khan and K. M. Nicholas, J. Am. Chem. Soc., (R)-4 and for a selection of related complexes. It can be seen 1997, 119, 9053. 10 M. Sawamoto and M. Kamigaito, T rends Polym. Sci., 1996, 4, 371. that, for (R)-4, the CuMN bond distances and the NMCuMN 11 M. Kato, M. Kamigaito, M. Sawamoto and T. Higashimura, bond angles are comparable with similar compounds. Macromolecules, 1995, 28, 1721.Interestingly, the CuMN1N2/CuMN3N4 dihedral angles deter- 12 T. Ando, M. Kamigaito and M. Sawamoto, T etrahedron, 1997, mined for (R)-4 are considerably larger than those seen for the 53, 15445. related bipyridine,35,36 N-tert-butyl-2-pyridylmethanimine37 13 K. Matyjaszewski and J.-S.Wang,Macromolecules, 1995, 28, 7901. and phenanthroline38,39 complexes, whilst comparing well with 14 K.Matyjaszewski, T. E. Patten and J. Xia, J. Am. Chem. Soc., 1997, that determined for the similar tert-butyldiazabutadiene 119, 674. 15 J. S. Wang and K. Matyjaszewski, J. Am. Chem. Soc., 1995, 117, (ButDAB) complex.37 Presumably, steric factors are important 5614. due to the large 1-phenylethylamine group and packing forces, 16 J. S. Wang and K. Matyjaszewski, Macromolecules, 1995, 28, 7901.such as ring stacking, may also play a part. The CMN and 17 J. S. Wang, T. Grimaud and K. Matyjaszewski, Macromolecules, CMC bond distances are very similar to those in the free 1997, 30, 6507. ligand (1.283 and 1.496 A ° for ButDAB40) and normal 18 C. Granel, P. Teyssie, P. Dubois and R. Jerome, Macromolecules, C(sp)2MN(sp)2 (1.27 A ° ) and C(sp)2MC(sp)2 (1.48 A ° ) bond 1996, 29, 8576. 19 V. Percec and B. Barboiu, Macromolecules, 1995, 28, 7970. lengths. In contrast, the imine CMN and CMC bond lengths 20 V. Percec, B. Barboiu, A. Neumann, J. C. Ronda and M. Y. Zhao, of ButDAB- in the ytterbium complex are significantly longer Macromolecules, 1996, 29, 3665. and shorter, respectively, since it is a -1 ligand. 21 V. Percec, H.-J.Kim and B. Barboiu, Macromolecules, 1997, 30, 6702. 22 D. M. Haddleton, C. B. Jasieczek, M. J. Hannon and A. J. Shooter, Conclusions Macromolecules, 1997, 30, 2190. 23 D. M. Haddleton, C. Waterson, P. J. Derrick, C. Jasieczek and This paper has shown that atom transfer polymerisation of A. J. Shooter, Chem. Commun., 1997, 683. methyl methacrylate can be eVectively performed using well 24 D.M. Haddleton, A. J. Clark, M. C. Crossman, D. J. Duncalf, defined copper(I) complexes. It was found that the enantiomer- A. M. Heming, S. R. Morsley and A. J. Shooter, Chem. Commun., ically pure chiral ligands used had no eVect on the stereochemi- 1997, 1173. stry of the resulting polymers; however, it may be possible via 25 H. Dieck and L. Stamp, Z. Naturforsch., T eil.B., 1990, 45, 1369. the use of more strongly complexed and rigid chiral SchiV 26 R. N. Keller and H. D. WycoV, Inorg. Synth., 1947, 2, 1. 27 G. J. Kubas, Inorg. Synth., 1989, 28, 68. base ligands. The use of a ligand with a methyl group b to the 28 G. M. Sheldrick, SHEL XL 5.0, Siemens Analytical Instruments, imine nitrogen significantly broadens PDI and reduces control Madison, WI, 1994. over Mn when the reaction is performed in a nonpolar solvent, 29 G. M. Sheldrick, SHELXL 96, University of Gottengen, 1996. such as xylene, compared to SchiV base ligands with n-alkyl 30 D. M. Haddleton and A. J. Shooter, Unpublished results. substituents. The control over Mn and PDI, however, can be 31 I. R. Peat and W. P. Reynolds, T etrahedron L ett., 1972, 14, 1359. improved if the reaction is performed in a more polar/weakly 32 J. S. Wang, R. Jerome, R. Warin and P. Teyssie, Macromolecules, 1993, 26, 5984. coordinating solvent such as diphenyl ether. J. Mater. Chem., 1998, 8(7), 1525–1532 153133 G. van Koten and K. Vrieze, Adv. Organomet. Chem., 1982, 21, 151. D. Kukulj and A. J. Shooter, J. Chem. Soc., Dalton T rans., 1998, 381. 34 M. N. Bochkarev, A. A. Trifonov, F. G. N. Cloke, C. I. Dalby, P. T. Matsunaga, R. A. Anderson, H. Schumann, J. Loebel and 38 P. C. Healy, L. M. Engelhardt, V. A. Patrick and A. White, J. Chem. Soc., Dalton T rans., 1985, 2541. H. Hemling, J. Organomet. Chem., 1995, 486, 177. 35 M. Munakata, S. Kitagawa, A. Asahara and H. Masuda, Bull. 39 M. Munakata, M. Maekawa, S. K. Kitagawa, S. Matsuyama and H. Masuda, Inorg. Chem., 1989, 28, 4300. Chem. Soc. Jpn., 1987, 60, 1927. 36 J. F. Dobson, B. E. Green, P. C. Healy, C. H. L. Kennard, 40 I. Hargittai and R. Seip, Acta Chem. Scand., Ser. A, 1976, 30, 540. C. Pakawatchai and A. H. White, Aust. J. Chem., 1984, 37, 649. 37 D. M. Haddleton, A. J. Clark, D. J. Duncalf, A. M. Heming, Paper 8/00467F; Received 16th January, 1998 1532 J. Mater. Chem., 1998, 8(7), 1525–1532

ISSN:0959-9428    

DOI:10.1039/a800467f

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

J O U R N A L O F C H E M I S T R Y Materials Behaviour of PEO-urethane in solutions of carbonates. Synthesis and electrochemical characterisation of PEO-urethane–coke electrodes Bruno Espanet, Christine Vautrin-Ul, Philippe Gue�gan, Annie Chausse�,* Richard Messina and Herve� Cheradame L aboratoire ‘Mate�riaux Polyme`res aux Interfaces’, EP 109, Universite� d’Evry, Centre CNRS, 2 a� 8 Rue H.Dunant, 94 320 T hiais, France We studied the behaviour of PEO-urethane, resulting from a polycondensation in situ of PEO [poly(ethylene oxide)] with urethane, towards organic carbonate electrolytes. We analysed the weight increase of the PEO-urethane with propylene carbonate (PC) or dimethyl carbonate (DMC) and a lithium salt at diVerent concentrations at diVerent times. Results indicate that the diVusion of the carbonate electrolyte is clearly controlled by the PEO-urethane–lithium salt interactions with PC based solutions and by the PEO-urethane–LiAsF6.(DMC)2 interactions with DMC based solutions. Comparing the transport properties of PEO-urethane with that of Celgard in a liquid electrolyte confirms the results for the weight increase as they suggest a preferential solvation of lithium species by the PEO-urethane network with PC and by DMC when the latter solvent is used.Composite electrodes containing coke and PEO-urethane were prepared using chemical in situ curing. The dependence of the electrical properties of these composites on the coke content was studied. We found an electrical percolation weight loading close to 46%.The influence of the PEO-urethane on the electrochemical behavior of coke was also studied in a carbonate electrolyte using these composite electrodes. A passivation phenomenon always occurs with the first reduction but the stability of the lithiated coke is reduced little with the PEO-urethane. There has been considerable research to develop lithium ion to the properties of hybrid crosslinked electrolytes such as conductivity, electrochemical stability etc.,13–15,19–21 few papers rechargeable batteries based on a carbon (coke or graphite) give information about the swelling of a crosslinked matrix by anode and a transition metal oxide cathode.1–8 Carbon has an organic solvent and how this swelling is aVected by the cycling faradic eYciencies close to 99%, which is suitable for concentration of a lithium salt in the solvent.an industrial application in batteries. It has been demonstrated Little has been reported in the literature on obtaining that the lithium insertion into the carbon matrix is composite electrodes from a polymer and a material by a accompanied, as already reported for metallic lithium, by the chemical in situ crosslinking.Modifications should be obtained formation of a surface passivating layer built up by some as it is well known that the morphology of the electrodes reduction products of the electrolyte.9,10 This layer strongly strongly aVects battery behaviour.22 Ionic–electronic conduc- influences the stability of the lithiated carbon toward the tivity is one of the most basic needs for an electrode material.electrolyte. We have already reported,11,12 that the passivating So, the second objective of this work was to prepare some layer dissolves in an organic electrolyte based on carbonate PEO-urethane–petroleum coke composites with electrical con- solvents so that it does not fully protect the lithiated carbon ductivity. Finally, we investigated their electrochemical behav- from reactions with the electrolyte. This generates a capacity iour in an organic electrolyte in experimental conditions close loss and self discharge phenomena between anode and cathode. to a hybrid electrolyte.To overcome these diYculties, much eVort has been devoted to improving the stability of the passivating layer by modifying the nature of the electrolyte.In this context, use of dry polymer electrolytes appeared beneficial since they should lower solu- Experimental bilities of species and limit their migration, via the electrolyte, Products between the electrodes. However, their low conductivities preclude industrial use and recent research has been targeted PEO with a molecular weight of 2000 was supplied by Merck. toward hybrid electrolytes, i.e.a polymer matrix swollen with The crosslinking agent (Desmodur RE kindly supplied by a liquid electrolyte.13–15 Many hybrid electrolytes reported in Bayer) was a solution of 4,4¾,4¾-methylidynetris(phenyl isocyanthe literature exhibit high ionic conductivities but also ate) in diethyl acetate. deficiencies such as poor mechanical properties so that they The synthesis of the PEO-urethane network was done often have to be hardened by chemical or physical curing.15 A according to an experimental procedure already reported,16–18 few years ago, Cheradame et al.reported that PEO [poly(ethyl- in a glovebox in order to eliminate side reactions due to water. ene oxide)] crosslinked with urethane possesses better proper- The viscous mixture was cast between two glass plates separties compared to PEO.16–18 So, we decided to consider its use ated by a flat spacer (1 mm) to control the thickness of the as a polymer matrix of a hybrid electrolyte.membrane. After one day at room temperature, the membrane In this paper, we have first investigated the swelling of PEO- was then vacuum dried at 80 °C. urethane by an organic carbonate solvent in relation to the Propylene carbonate (PC from Fluka), dimethyl carbonate nature of this solvent (PC or DMC) and the lithium salt (DMC from Merck) or their mixtures with ethylene carbonate concentration.These two carbonates commonly used in batter- (EC from Merck) were stored over molecular sieves, previously ies were chosen because their relative permittivity and donor activated at 200 °C in order to lower their water content.number are diVerent. LiCF3SO3 (from 3M) was dried at 100 °C under vacuum prior to use. LiAsF6 (from Lith Co) was used as received. The Although numerous papers report results of studies relevant J. Mater. Chem., 1998, 8(7), 1533–1539 1533composition of the mixtures is given hereafter in volumes of solvents.Petroleum coke type PC 40 with an average particle size of 14 mm was kindly supplied by Lonza. Apparatus Conductivity measurements (DC and AC) were made using a PAR 273 apparatus from EG&G coupled with a Schlumberger analyser in a frequency range of 105 to 1 Hz or with a multimeter. The membrane was sandwiched between two blocking stainless steel electrodes in a home made cell18 which allows us to work under argon.A ‘Mac Pile’ system (from Bio-Logic Co) that can operate either in a galvanostatic or a potentiostatic mode was used to perform electrochemical experiments. Coin cells were constructed using a working electrode based on coke, a Celgard Fig. 1 Dependence of the weight increase of PEO-urethane samples 2400 microporous film (as separator), a Viledon foil (as electro- on the concentration of LiAsF6 after an immersion time of 2300 min: lyte reservoir) and a lithium foil (from Aldrich) which acts as (a) without drying, in PC based solutions; (b) without drying, in DMC counter electrode and reference electrode.Working electrodes based solutions; (c) with drying, in PC based solutions; (d) with drying, in DMC based solutions were either a mixture of coke and Teflon (respectively 95 and 5% in weight) whose preparation has already been reported in,23 or composite electrodes whose synthesis conditions will be described later. A volume of 300 ml of electrolyte was used during coin cell assembly which was done in an argon filled glovebox.LiCF3SO3 was used as it is less electroactive than LiAsF6. All electrochemical data presented here concern coin cells tested at 45 °C.The potentials are expressed with respect to the Li+/Li redox couple. Cycling tests were done between 2.5 and 0.010 V to avoid the oxidation of the electrolyte and the deposition of the metal lithium. Determination of the transference number was done in the following concentration cell Li in PC–LiAsF6 1 mol l-1//carbonate&ndas/PEO//carbonate–LiAsF6 C2 //Li in PC–LiAsF6 1 mol l-1.Briefly, the PEO-urethane membrane was sandwiched between two carbonate electrolytes with a diVerence in the lithium concentration (C1 and C2). C1 and C2 were varied Fig. 2 Dependence of the weight increase of PEO-urethane samples between 0.1 and 0.9 mol l-1 so that the average concentration on their immersion time in PC based solutions with the following (C1+C2)/2 was always 0.5 mol l-1.Two lithium threads (Li) concentrations of LiAsF6: (a) 0, (b) 0.05, (c) 0.2, (d) 0.5, (e) 1, immersed in a separate compartment containing PC–LiAsF6 (f ) 1.5 mol l-1 1 mol l-1 were used to measure the cell voltage with a Tacussel millivoltmeter. The first and last measurements were done with the same electrolyte on both sides of the membrane to verify that we worked under equilibrium conditions with no deterioration of the membrane.SEM micrographs were recorded with a Philips XL 30 scanning electron microscope. Results Behaviour of PEO-urethane in a carbonate electrolyte Weight increase of PEO-urethane with a carbonate electrolyte. Using PEO network instead of linear PEO gives rise to the problem of salt and solvent incorporation in the material.To understand and control this process, we studied the swelling of PEO based networks dipped in organic carbonate electro- Fig. 3 Dependence of the weight increase of PEO-urethane samples lytes. Samples of the PEO-urethane membranes of 10 mm on their immersion time in DMC based solutions with the following diameter were dipped for diVerent durations in 10 cm3 of concentrations of LiAsF6: (a) 0, (b) 0.05, (c) 0.2, (d) 0.5, (e) 1 mol l-1 solutions based on a lithium salt (LiAsF6) dissolved in PC or DMC solvent.The penetration of the solution in the membranes of PEO-urethane was measured by weighing samples known to slow down the diVusion process of a solvent in a semi crystalline network, was then assumed to be the same in before and after immersion.The results are given in Fig. 1–3 and are expressed in weight increase defined as [(m-m0)/m0] each network and to have a negligible eVect. Considering the low amount of extractible products, one can assume that free with m the weight of the swollen membrane and m0 the weight of the membrane before swelling as reported elsewhere.24 Each chains do not influence the measurements.Fig. 1 shows the weight increase of PEO-urethane mem- polymer chain was assumed to have 44 ethylene oxide units between crosslinks, which allows comparison of the network branes as a function of the concentration of LiAsF6 dissolved in DMC or PC solvents for a dipping time of 2300 min. This behaviour. Furthermore, the thermal history of the networks was assumed to be the same.The influence of the crystallinity, time is suYcient to reach thermodynamic equilibrium as will 1534 J. Mater. Chem., 1998, 8(7), 1533–1539be discussed further. The weight increase of the PEO-urethane We suggest that the weight increase of PEO-urethane membranes immersed in LiAsF6 solutions depends on the nature membranes due to the salt plus solvent increased with the LiAsF6 concentration for DMC solutions (b) but reached a of the carbonate solvent: the diVusion is clearly controlled by the PEO-urethane–lithium salt interactions with PC based maximum for PC solutions (a), at 0.05 mol l-1.The amount of incorporated electrolytes (PC or DMC) is much larger than solutions and by the PEO-urethane–LiAsF6·(DMC)2 interactions with DMC based solutions.would be expected using linear PEO, of adequate size and mechanical properties.15 Drying the PEO-urethane membranes Since the use of LiAsF6 is becoming prohibited in lithium battery applications, we also conducted experiments with leads to the determination of the amount of LiAsF6 incorporated in those membranes. PC based solutions (c) allowed a LiCF3SO3 as lithium salt. The same dependence of the weight uptake in relation to the carbonate and the salt concentration higher salt uptake than DMC based solutions (d) when the LiAsF6 concentrations were high.However, it is worth noting was obtained but the concentration range investigated was narrower because the solubility of LiCF3SO3 is low in DMC that the salt uptake was higher with DMC based solutions when the LiAsF6 concentration was low.Comparison of curves (0.3 mol l-1). Mixtures of solvents are usually used in lithium batteries to (a) and (c) shows that salt uptake in PEO based networks increases with [LiAsF6] in PC solutions, despite the decrease obtain an electrolyte with good conductivity, low reactivity towards the electrode materials and ability to operate over a of the solvent plus salt uptake.This behaviour can be explained by a strong interaction between the salt and the chains of the large temperature range. So, we decided to extend this study to mixtures of carbonates derived from PC–EC–DMC (15153) network: the elasticity of the network decreases with salt uptake, reducing the equilibrium swelling.Curves (b) and (d) which we usually used in coin cells,23 to produce hybrid electrolytes of interest in the field of batteries. EC is solid at dealing with DMC solutions showed a diVerent trend: both uptakes increased with [LiAsF6]. The ratio between the two room temperature and can only be used when mixed with another solvent. The influence of various solvent ratios was uptakes is constant: 2 moles of DMC per mole of salt.The diVusive species, assumed to be a molecule of salt solvated then tested, but the participation of each solvent in the mixture on the weight uptake is diYcult to estimate. by two molecules of solvent [referred to hereafter as LiAsF6·(DMC)2], might have a high interaction with the PEO Table 1 gives the percentage of weight increase measured for PEO-urethane samples after a 48 h immersion in mixtures of network, preventing any measurable incorporation of free solvent molecule.In other words, this behaviour results from carbonates in the absence and in the presence of a lithium salt at 1 mol l-1. The EC–DMC mixture (153 in volume), with or the Flory parameter of interaction between the chains swollen by LiAsF6·(DMC)2 and the free solvent molecules being too without the lithium salt, leads to the lowest percent weight increase.However, the LiCF3SO3 incorporated in the PEO- large. To further understand these phenomena, the weight uptake urethane network does not vary significantly with the mixture compositions but a slight variation is observed when LiAsF6 of PEO-urethane membranes in solutions based on PC and DMC was determined (respectively Fig. 2 and 3) with diVerent is used. The mixture of EC–DMC (153 in volume) is then used for the electrochemical study in order to control the amount LiAsF6 concentrations as a function of the dipping time. Comparing first the weight increase of the PEO-urethane of incorporated salt and to keep the swelling low. membranes immersed in solvents (PC or DMC) without lithium salt gives indications of the solvent–PEO-urethane Conductive properties of hybrid electrolytes based on PEOurethane and a carbonate solvent.The use of PEO-urethane– interactions. The amount of solvent which is incorporated in these membranes is higher for PC than for DMC due to a carbonate membranes as a hybrid electrolyte in lithium batteries assumes that they ensure the transport of charged species higher aYnity of the PEO based macromolecular chains with PC.The initial slope of the solvent uptake is indicative of the including lithium species. Here, we have used the concentration cell method to measure the transference number of the ions; it diVusion kinetics; comparing Fig. 2 and 3 indicates that the diVusion rate of DMC is slightly faster than that of PC.is a potentiometric technique in which the electrolyte undergoes no perturbation because no current is applied. LiAsF6 was Obviously, the strong interaction between PEO units and PC molecules is detrimental to the rate of PC diVusion. chosen because of its good solubility in DMC. The concentration cell method has already been developed Fig. 2 shows the salt plus solvent uptake of some PEOurethane membranes dipped in solutions based on PC with in the literature for polymer electrolytes.25–27 Hypotheses are as follows: (i) the ionic motions are only due to the concen- diVerent LiAsF6 concentrations. For each salt concentration, the equilibrium swelling was reached within less than 2300 min tration gradient, (ii) the activity coeYcients are approximately the same in the two electrolytes.The voltage due to the except for the highest concentration. A slowing of the diVusion process is observed with increased salt concentration: evidence diVerence in concentration is expressed as eqn. (1)27 of the strong interaction between the diVusing salt and the DE=RT /F (tLi+-tAsF6 -) ln(C1/C2) (1) polymer host.The role of the PC solvent seems to be the reduction of the local viscosity of the network since the initial In our experimental cell presented in the experimental section, junction potentials exist due to the separate compartments; rates of diVusion are slightly higher for PC solutions than for DMC at low salt concentration. Kinetics and thermodynamics deviations of the Nernstian law were also observed for the Li+/Li redox system in relation to the salt concentration.show that the weight increase behaviour of the PEO-urethane dipped in PC solutions is controlled by the interactions Elimination of these perturbations was obtained by subtracting the cell voltages obtained with a Celgard membrane and with between the lithium salt and the PEO chains.When the solvent was DMC (Fig. 3), the diVusion into the a PEO-urethane membrane. This consequently allows us to compare the behaviour of the PEO-urethane membrane with PEO-urethane membranes was much faster than with PC and the equilibrium swelling was reached within 1500 min, irrespec- reference to membranes used in commercially available batteries. A plot of (DEPEO-urethane-DECelgard) against log (C1/C2) is tive of the salt concentration. The diVusion slowed with increasing salt concentration [Fig. 3(b–e)]. At the start of the linear and the slope gives access to DtLi (i.e. tLi with PEOurethane membrane-tLi with the Celgard membrane). No cell dipping experiments, both DMC and complexes are diVusing in the PEO-urethane membrane but the diVusion of DMC is voltage is measured when a Celgard membrane is sandwiched between two solutions of PC with diVerent salt concentrations.rapidly stopped due to an increase of the concentration of LiAsF6·(DMC)2 complex in the membrane. Kinetics experi- This result suggests that the Celgard acts as an inert membrane. Table 2 gives DtLi values for diVerent solvents. PEO-urethane ments reveal a good interaction of the complex with the PEO chains.and Celgard are both swollen by the carbonate electrolyte so J. Mater. Chem., 1998, 8(7), 1533–1539 1535Table 1 Weight increase (%) of the PEO-urethane samples after 48 h of immersion in the solution (composition of mixture is expressed as volume of solvent) weight uptake (%) weight uptake (%) in LiAsF6 1 mol l-1– in LiCF3SO3 1 mol l-1– composition of the solvent mixture solvent mixture carbonate mixture weight uptake (%) in solvent mixture before drying after drying before drying after drying 153 400 453 155 210 110 EC–DMC 151 413 547 159 — — 351 393 622 159 — — PC–DMC 153 352 409 143 213 102 EC–PC 151 369 646 177 358 117 PC–EC–DMC 15153 388 — — 273 166 Table 2 Dependence of DtLi on solvent (the composition of the mixture are clearly observed without PEO (a) or when the PEOis expressed as volume of solvent) urethane content is below 50% (c).The two faces of a same composite electrode diVer slightly (b1 and b2, c1 and c2) when Solvent DtLi its bulk structure is homogeneous (b3 and c3). PC -0.2 DMC 0 Electrical conductivity of composite electrodes. Composite PC–EC–DMC (15153) -0.10 electrodes must exhibit some electrical conductivity to be used EC–DMC (153) -0.12 in batteries with good eYciency.This is allowed by adding PC–DMC (153) -0.12 conductive coke particles to the polymer at weight loadings above a threshold called ‘the electrical percolation weight loading’ which corresponds to the formation of a continuous that negative or positive values of DtLi can be attributed to a path of conductive particles where the electrons can flow.32–40 solvation of the species by the polymer.With PC, the DtLi The change in resistivity with coke loading (expressed as value suggests that some interactions exist between the lithium weight or volume fraction) is illustrated in Fig. 5. The electrical species and the polymer. This would assume that the polymer percolation weight loading is close to 46%.The transition interacts more strongly than PC with the lithium species so from non-conductive to conductive composite electrodes with that lithium species move preferentially hopping from one a constant value of resistivity extends over the 46–75% range. oxyethylene unit to another through the breaking and the The theory regarding spherical carbon black systems predicts forming of cation–oxygen bonds.This result agrees well with that the percolation occurs at around 40% by weight of carbon the weight increase results and the results on hybrid electrolytes (34% volume of carbon)39 but it is well established that many reported by other authors28–30 who have observed no Li–PC parameters influence the percolation threshold of composite interactions in the presence of a polymer.With DMC, DtLi is materials such as filler–matrix interactions, filler shape and close to zero, suggesting a preferential solvation of the lithium spatial distribution of the filler. The amorphous and polar species by the carbonate. This would imply that the lithium characteristics of PEO-urethane are detrimental to the percospecies move through the polymer matrix in a solvation shell lation as they encourage a uniform distribution of the coke in of the carbonate molecules.This preferential solvation is the polymer. Moreover, crosslinking is known in the literature consistent with a donor number for DMC higher than the to increase the resistivity and the percolation threshold of donor number for PEO and PC and with the weight increase composites.32 The size of the coke particles in this study is results and the literature relative to hybrid electrolytes.28–30 also detrimental to the percolation as it is at least two orders With EC–DMC (153 in volume) mixture, PC–DMC (153 in of magnitude above that of carbons used in other studies.All volume) mixture or PC–EC–DMC (15153 in volume), DtLi these parameters contribute to increase the electrical percohas an intermediate value.lation weight loading up to 46% by weight of coke. Synthesis and characterisation of composite Electrochemical behaviour of the composite electrodes. PEO-urethane–petroleum coke electrodes Composite electrodes were electrochemically tested finally in the presence of a mixture of carbonates.Our preliminary tests Synthesis of composite electrodes. PEO and coke were first dried at 80 °C under vacuum for 48 h. The appropriate weight have shown that they were electrochemically inactive when PC was used in the mixture. This suggests that the swelling fractions of PEO and coke were added to CH2Cl2 in small excess. PEO dissolves in CH2Cl2 to produce homogeneous with mixtures based on PC causes the separation of the conductive particles into the composite electrodes reducing mixture.This mixture was allowed to stand in contact with air for a few days at room temperature to eliminate the solvent their electrical conductivity, as already reported in.41 This is due to an expansion of the polymer that reduces the coke and it was finally dried under high vacuum for 24 h at 80 °C.The polycondensation reaction was done with the procedure concentration on a volume basis. Our electrochemical results agree well with experimental data given in ref. 1 as large weight reported in the experimental section. Only the ratio of the volume of reactants to the (volume of reactants+volume of increases are obtained with mixtures based on PC.So, in this study, we selected: (i) an EC–DMC mixture as it CH2Cl2) was modified because of the presence of coke; it was fixed at 0.6 to avoid bubble formation and syneresis. gives a smaller weight increase with the PEO-urethane and it is commonly used in lithium batteries and (ii) the composite The PEO weight content was varied between 25 and 75%. The 25% value was chosen according to data available in the electrode with the highest coke loading to maintain electrical conductivity.Electrochemical studies for coke without polymer literature which suggest that with decreasing precursor ratio, the probability of encounter also decreases to the point (gener- (referred to hereafter as latex) are also reported for reference. Fig. 6(a) and (b) show voltammograms obtained for both ally around 25%) where no more reaction (polycondensation or gelation) is possible.31. electrodes. Reduction occurs in two main steps, already attributed in the literature.11–23 Comparing these figures shows Fig. 4 shows SEM micrographs for composite electrodes without (a) or with PEO-urethane (b and c). Coke particles that PEO-urethane has no influence on the potentials of these 1536 J.Mater. Chem., 1998, 8(7), 1533–1539Fig. 4 SEM micrographs of a latex electrode (a) and composite PEO-electrodes with (b) 70% and (c) 40% by weight of coke. SEM micrographs for the two faces are numbered 1 and 2 and that for the cross-section is numbered 3. J. Mater. Chem., 1998, 8(7), 1533–1539 1537a composite electrode.The capacity varies abruptly between the first and second cycle. It fades more slowly during subsequent cycles. The faradic capacity for the first reduction is 290 m Ah g-1 for the composite electrode and 350 m Ah g-1 for the latex electrode. Several causes can explain this diVerence: (i) some coke particles may be isolated from the conductive network and are not involved in the electrochemical processes, (ii) the end of the lithium insertion into the coke occurs at potentials lower than the voltage limit because of the presence of PEOurethane.Cycling results permit the calculation of the capacity loss. A large capacity loss is found during the first cycle and it is lower during the subsequent cycles (respectively 32, 9, 6 and 5%). This confirms that part of the capacity engaged during the first reduction corresponds to some irreversible processes (reduction of the electrolyte or its impurities directly or indirectly with the lithiated coke); the passivation layer on the Fig. 5 Dependence of the composite electrodes onto the coke fraction composite electrode originates from the insoluble products of expressed in weight or volume these processes.These irreversible processes are minor as cycling proceeds because of the passivating layer. Table 3 gives the faradic eYciencies for the first and the fifth cycle as a function of the current density. Contrary to all expectations, the faradic yields decrease when the current density is lowered, i.e. when the contact time between the lithiated composite electrode and the electrolyte increases.This suggests that the passivating layer limits but does not stop the reactions between the lithiated composite electrode and the electrolyte. Lithium stability in the latex electrode or in the composite electrode was also studied using the following experimental procedure: five reduction–reoxidation cycles were performed to generate the passivating layer followed by a final reduction.The reduced electrodes were thererafter stored at open current voltage before electrochemical reoxidation was done. No improvement of the lithium stability was obtained with the composite electrode (Table 4). So, one can assume that the passivating layer is not eVective in the lithium protection due to its heterogeneous nature as PEO-urethane and the products resulting from the reactions between lithium and PEO-urethane or the organic electrolyte help to make up its composition.The less protective properties of a heterogeneous layer have already been reported in the literature for mixtures of organic solvents.42,43 Finally, our results suggest no superiority of hybrid electrolytes based on PEO-urethane over liquid electrolytes when coke is used as anode.Table 3 Faradic eYciency values (%) of the first and the fifth reduction –reoxidation cycle as a function of the current density. EC–DMC (153 v/v); LiCF3SO3, 1 mol l-1. Composite PEO-urethane–petroleum coke (30570) electrode current density/ faradic yield (%) faradic yield (%) mA cm-2 of the first cycle of the fifth cycle Fig. 6 (a) and (b) Voltammograms obtained with a latex electrode (a) and a composite electrode (70% by weight of coke) (b), scan rate of 300 68 95 3.6 mV mm-1.(c) Cycling curves obtained on a composite electrode 200 70 95 (70% by weight of coke), current density of 380 mA cm-2; the letter 160 68 94 corresponds to the cycle number. EC–DMC (153 v/v), LiCF3SO3; 100 67 89 1 mol l-1 60 55 88 electrochemical processes.Reoxidation also occurs in two main steps but their potentials are shifted towards higher values Table 4 Lithium loss obtained during storage in EC–DMC (153 v/v), with the composite electrode. The second reduction scan for LiCF3SO3, 1 mol l-1 of a composite and a latex reduced electrode both electrodes diVers from the first scan; this has already been Storage % of lithium loss % of lithium loss attributed to the formation of a passivating layer on the time/hours with a composite electrode with a latex electrode lithiated coke along with the first lithium insertion.It is obvious that PEO-urethane does not suppress the passivation 250 18 9 phenomenon of the lithiated coke. 500 40 20 Fig. 6(c) gives the first five reduction–reoxidation cycles for 1538 J.Mater. Chem., 1998, 8(7), 1533–153914 R. Koksbang, I. I. Olsen and D. Shackle, Solid State Ionics, 1994, Conclusion 69, 320. 15 J. M. Tarascon, A. S. Gozdz, C. Schmutz, F. Shokoohi and Our results show that the nature of the carbonate governs the P. C. Warren, Solid State Ionics, 1996, 86–88, 49. swelling of the PEO-urethane and the solvation of the lithium 16 H. Cheradame and J.F. Le Nest, Polymer Electrolyte Reviews, ed. species. Results indicate that the diVusion of a liquid electrolyte J. R. Mac Callum and C. A. Vincent, Elsevier Applied Science, in the polymer is clearly controlled by the PEO-urethane– London, 1987, p. 103. 17 J. F Le Nest, A. Gandini, H. Cheradame and J. P. Cohen-Addad, lithium salt interactions with PC based solutions and by the Macromolecules, 1988, 21, 117. PEO-urethane– LiAsF6.(DMC)2 interactions with DMC based 18 J.F Le Nest, Thesis, University of Grenoble, 1985. solutions. 19 R. Frech and S. Chintapalli, Solid State Ionics, 1996, 85, 61. An electrically conductive electrode based on PEO-urethane 20 M. Forsyth, D. R. MacFarlane, M. E. Smith and T. J. Bastow, and petroleum coke calls for a higher carbon loading than Electrochim.Acta, 1995, 40(13–14), 2343. that reported in the literature for carbon–polymer composites. 21 A. Reiche, J. Tu� bke, K. Siury, B. Sandner, G. Fleischer, S. Wartewig and S. Shashkov, Solid State Ionics, 1996, 85, 121. This is due to the particle size of the coke and to the reticulated 22 A. Selvaggi, F. Croce and B. Scrosati, J. Power Sources, 1990, and amorphous structure of the PEO-urethane which enhances 32, 389.a uniform distribution of the carbon particles. PEO-urethane 23 M. Jean, C. Desnoyer, A. Tranchant and R. Messina, has a minor influence on the electrochemical behaviour of J. Electrochem. Soc., 1995, 142(7), 2122. coke in a carbonate mixture but it decreases the stability of 24 S. G. Meibuhr, J. Electrochem.Soc., 1970, 117, 56. 25 J. P. Hoare and C. R. Wiese, J. Electrochem. Soc., 1974, 121, 83. the lithiated coke. An improvement of this stability requires the 26 S. Atchia, J.-P. Petit, J.-Y. Sanchez, M. Armand and D. Deroo, use of a dry PEO-urethane electrolyte. Studies regarding these Electrochim. Acta, 1992, 37(9), 1599. composite electrodes in a solid PEO-urethane electrolyte are 27 M.Clericuzio, W.O. Parker, M. Soprani and M. Andrei, Solid now in progress in our boratory. State Ionics, 1995, 82, 179. 28 Z. Wang, B. Huang, S. Wang, R. Xue, X. Huang and L. Chen, J. Electrochem. Soc., 1997, 144(3), 778. 29 M. Forsyth, D. R. MacFarlane, M. E. Smith and T. J. Bastow, References Electrochim. Acta, 1995, 40(13–14), 2343. 30 X. W. He, J. M. Widmaier, J. E.Herz and G. C. Meyer, Polymer, 1 D. Fauteux and R. Koksbang, J. Appl. Electrochem., 1993, 23. 1989, 30, 364. 2 R. Yazami and Ph. Touzain, J. Power Sources, 1983, 9, 365. 31 J. H. Smuckler and P. Finnerty, Adv. Chem. Ser., 1974, 134, 171. 3 R. Fong, U. von Sacken and J.R. Dahn, J. Electrochem. Soc., 1990, 32 F. Gubbels, R. Jerome, Ph. Teyssie�, E. Vanlathem, R. Deltour, 137, 2009. A. Calderone, V. Parente and J. L. Bredas, Macromolecules, 1994, 4 J. Yamaura, Y. Ozaki, A. Morita and A. Ohata, J. Power Sources, 27, 1972. 1993, 43, 233. 33 J. Navarro-Laboulais, J. Trijueque, J. J. Garcia-Jareno and 5 D. Guyomard and J. M. Tarascon, J. Electrochem. Soc., 1992, F. Vicente, J. Electroanal. Chem., 1997, 422, 91. 139, 937. 34 R. Tchoudakov, O. Breuer, M. Narkis and A. Siegmann, Polym. 6 R. Kanno, Y. Takeda, T. Ichikawa, K. Nakanishi and Networks Blends, 1996, 6(1), 1. O. Yamamoto, J. Power Sources, 1989, 26, 535. 35 C. Klason and J. Kubat, J. Appl. Polymer Science, 1975, 19, 831. 7 T. Ohzuku, Y. Iwakoshi and K. Sawai, J. Electrochem. Soc., 1993, 36 I. Balberg and S. Bozowski, Solid State Commun., 1982, 44(4), 551. 140(9), 2490. 37 J. M. Pernaut and A. L. de Oliveira, Synth. Metals, 1997, 84, 443. 8 T. D. Tran, J. H. Feikert, X. Song and K. Kinoshita, J. Electrochem. 38 K. T. Chung, A. Sabo and A. P. Pica, J. Appl. Phys., 1982, 53(10), Soc., 1995, 142(10), 3297. 6867. 9 D. Aurbach, Y. Gofer, M. Ben-Zion and P. Aped, J. Electroanal. 39 M. Narkis, M. Zilberman and A. Siegmann, Polym. Adv. T echnol., Chem., 1992, 339, 451. 1997, 8, 525. 10 D. Aurbach, A. Zaban, Y. Goser, Y. Ein-Eli, I. Weissman, 40 A. Marquez, J. Uribe and R. Cruz, J. Appl. Polm. Sci., 1997, 66, O. Chusid and O. Abramson, J. Power Sources, 1995, 54, 76. 2221. 11 M. Jean, A. Tranchant and R. Messina, J. Electrochem. Soc., 1996, 41 D. Aurbach, Y. Gofer and J. Langzman, J. Electrochem. Soc., 1989, 136, 3198. 143(2), 391. 42 D. Aurbach, Y. Gofer, M. Ben-Zion and P. Aped, J. Electroanal. 12 M. Jean, A. Chausse� and R. Messina, J. Power Sources, 1997, Chem., 1992, 339, 451. 68, 232. 13 F. Croce, S. Panero, S. Passerini and B. Scrosati, Electrochim. Acta, 1994, 39, 255. Paper 7/08822A; Received 8th December, 1998 J. Mater. Chem., 1998, 8(7), 1533–1539

ISSN:0959-9428    

DOI:10.1039/a708822a

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

J O U R N A L O F C H E M I S T R Y Materials (N-Methylthiocarbamoyl )tetrathiafulvalene derivatives and their radical cations: synthetic and X-ray structural studies Adrian J. Moore,a Martin R. Bryce,a*† Andrei S. Batsanov,a Julie N. Heaton,a Christian W. Lehmann,a Judith A. K. Howard,a Neil Robertson,b Allan E. Underhillb and Igor F. Perepichkac aDepartment of Chemistry, University of Durham, South Road, Durham, UK DH1 3L E bDepartment of Chemistry, UCNW Bangor, Bangor, Gwynedd, UK L L 57 2UW cL .M. L itvinenko Institute of Physical Organic & Coal Chemistry, National Academy of Sciences of Ukraine, Donetsk 340114, Ukraine Lithiation of 4,5-bis(methylsulfanyl)-TTF 9, 4,5-(ethylenedisulfanyl)-TTF 10, 4,5-dimethyl-TTF 11 and 4,5,5¾-trimethyl-TTF 12 (TTF=tetrathiafulvalene) followed by reaction with methyl isothiocyanate aVords the corresponding (N-methylthiocarbamoyl)- TTF derivatives 14–17, respectively, in 54–70% yields. These new TTF derivatives display a broad intramolecular charge-transfer band in their UV–VIS spectra arising from conjugation between the donor TTF ring and the acceptor Nmethylthiocarbamoyl moiety.Steric hindrance between the adjacent N-methylthiocarbamoyl and methyl substituents in 17 causes a marked hyposchromic shift in this band (lmax 395 nm) compared to compounds 14–16 (lmax 435–467 nm).Consistent with the electron-withdrawing properties of the N-methylthiocarbamoyl substituent, its attachment to the TTF ring raises slightly the oxidation potential of the system. Charge transfer complexes of these donors and (N-methylthiocarbamoyl)-TTF 2 with 7,7,8,8- tetracyano-p-quinodimethane (TCNQ) and salts with bromide anions are reported, some of which have high room temperature conductivity values.The X-ray crystal structures are presented for 16, 17 and the salts 2·Br, 14·TCNQ and (17)2·20. The structure of 16 comprises orthogonal dimers (kappa packing) while in the structure of 17 individual molecules are orthogonal to each other.There is weak intermolecular hydrogen bonding in both 16 and 17. In the structure of 2·Br, the radical cations 2+· are almost planar and they form an infinite stair-like stack of dimers, with bromide anions situated between the stacks, and linked with the cation by a strong N–H,Br bond. The structure of 14·TCNQ comprises mixed,DDAADD, stacks; the Nmethylthiocarbamoyl group engages in an interstack N–H,N bond with a TCNQ anion. Analysis of the bond lengths in the structure suggests that there is partial charge transfer from 14 to TCNQ.In the structure of (17)2·20 molecules form mixed ,DDADDA,stacks and analysis of bond lengths suggests that there is only a small degree of charge transfer from donor to acceptor.The geometries of compounds 2, 14, 16, 17 were optimised using the PM3 semi-empirical method and the results compare favourably with the X-ray structural data. It is well established that the structural features of tetrathiaful- one-dimensional metal is unstable to lattice instabilities, valene (TTF, 1) derivatives in the solid state are of crucial notably the Peierls distortion.1e From well established phenomena in solid-state chemistry, the use of I,X (X=CN or halogen)2 and hydrogen-bonding3 interactions are two areas currently being investigated as means of ‘crystal engineering’ in the field of organic conductors.A number of TTF derivatives bearing halogen substituents have been synthesised recently4,5,6 and close I,X (X=CN5,7 or halogen6) cation,anion interactions have been observed. Hydrogen-bonding cation,anion interactions have also been observed recently in a number of conducting and superconducting salts of TTF derivatives,8 fuelling the preparation of TTF systems endowed with substituents capable of participating in intermolecular hydrogen-bond- S S S S S S S S S NH Me S S S S S S S S 1 2 3 ing (e.g.hydroxy, amido groups).Donor,donor hydrogenbonding interactions have been revealed by X-ray analysis importance in obtaining high electrical conductivity (and superof a number of appropriately functionalised neutral TTF conductivity) from their charge-transfer complexes and radical systems.9,10 Notably, it was found that 4-(N-methylthiocarba- ion salts. The control of molecular stacking by means of moyl)tetrathiafulvalene 2 was packed in a kappa fashion10 chemical modification of the TTF molecule remains a major (that is, the p-donor forms orthogonal dimers which are challenge.In this context, the design of TTF-based donor coupled through weak intermolecular forces); such an arrange- molecules bearing substituents capable of eVective intermolecument in a neutral TTF derivative is very rare.9c,11 Currently, lar interactions, which may exert an orientating eVect on the all the TTF-based superconductors with Tc>10 K possess constituent molecules, is an emerging area.An increase in the kappa-phase structures, e.g. k-(BEDT-TTF)2Cu(NCS)212 and dimensionality of the conduction process (most commonly k-(BEDT-TTF)2Cu[N(CN)2]X (X=Br, Cl)13 [BEDT-TTF achieved by close interstack chalcogen,chalcogen inter- 3=bis(ethylenedithio)tetrathiafulvalene]. actions) is known to stabilise the metallic state to low tempera- The synthesis of functionalised TTF derivatives via lithiation ture, and in some cases gives rise to superconductivity.1 A true and subsequent trapping of the intermediate mono-anion with electrophiles was first reported by Green14 and this method †E-mail: m.r.bryce@durham.ac.uk J.Mater. Chem., 1998, 8(7), 1541–1550 1541has subsequently been extended to aVord a wide range of The reaction of compounds 9–12 with 1.1 equivalents of lithium diisopropylamide in diethyl ether (tetrahydrofuran for derivatives.15We employed this approach previously to prepare compound 210 and have now utilised further this methodology compound 1028) at -78 °C aVorded lithiated TTFs of general formula 13, which were trapped with methyl isothiocyanate to to aVord a number of new (N-methylthiocarbamoyl)-TTF derivatives designed to exhibit hydrogen-bonding in the neutral give, after aqueous work-up, the desired products 14–17 (54–70% yields) (Scheme 2).These compounds were isolated state, which may, potentially, be manifested in their chargetransfer complexes and radical ion salts.16 Herein we describe as air-stable crystalline solids, which were notably darker in colour than most TTF derivatives (14–16 are dark purple– (i) the syntheses of compounds 14–17, (ii ) single-crystal X-ray structures of neutral compounds 16 and 17,17 (iii) complexation black; 17 is very dark red) due to intramolecular chargetransfer from the donor TTF moiety to the acceptor N- studies on compounds 2 and 14–17 with 7,7,8,8-tetracyano-pquinodimethane (TCNQ, 19), and reactions with tetrabutylam- methylthiocarbamoyl substituent, as observed previously for compound 2.10 Compared to TTF 1, the electronic spectra of monium tribromide which lead to bromide salts, and (iv) single-crystal X-ray structures of 2·Br and 14·TCNQ salts 2 and 14–17 in acetonitrile present a new, very broad absorption band centred at lmax 435–467 nm (compounds 2, and (both 151 stoichiometry) and complex (17)2·20. 14–16) and lmax 395 nm (compound 17). The marked hypsochromic shift in this band for compound 17 reflects steric Results and Discussion hindrance between the N-methylthiocarbamoyl moiety and the adjacent Me group, causing the former to twist significantly Compounds 9,18 10,18 1119 and 1219,20 were prepared according to literature procedures, utilising the coupling protocol pion- out of the plane of the TTF ring, thereby reducing the intramolecular conjugation (as confirmed by the X-ray struc- eered by Lerstrup and co-workers, which selectively aVords the unsymmetrical products.19,21 As an alternative for the ture, see below).The electrochemistry of compounds 14–17, studied by cyclic voltammetry, shows two reversible one- synthesis of compounds 9 and 10, we have followed a sequence (Scheme 1) based upon a ‘pseudo-Wittig’ condensation22 of electron oxidations typical of the TTF system at potentials consistent with their substitution patterns. 1,3-dithiolium salt 423 and triphenylphosphonium tetrafluoroborate Wittig salt 524 to aVord compound 6:17 slow addition of triethylamine to a solution of compounds 4 and 5 proceeded cleanly aVording a crude reaction mixture from which compound 6 could be easily isolated in 78% yield in ca. 5 g batches. Trimethyl phosphite and triethyl phosphite mediated ‘cross-coupling’25 of 1,3-dithiole-2-thione 7 and 4,5-bis(2¾- cyanoethylthio)-1,3-dithiole-2-thione (or -one) 826 also aVorded compound 6 in moderate yields (20–35%); however, both coupling agents gave a complex mixture of products from which it proved exceptionally diYcult and time-consuming to obtain a pure sample of compound 6.Subsequent removal of the cyanoethyl groups in compound 6 occurred upon reaction with two equivalents of sodium ethoxide in ethanol–tetrahydrofuran, and alkylation of the derived dithiolate anions with either methyl iodide or 1,2-dibromoethane yielded the known products 9 and 10 (85 and 80% yields, respectively).In our hands there is little to choose between the route described herein to compounds 9 and 10 and that championed by Fourmigue� et al.18 in terms of laboratory time, yield and purity of the final products.One merit of Scheme 1 is that it proceeds via compound 6 which is especially versatile as a R R S S S S H R1 R R S S S S R1 R R S S S S R1 S NH Me (ii), (iii) 2 R = R1 = H 14 R = SMe, R1 = H 15 R-R = SCH2CH2S, R1 = H 16 R = Me, R1 = H 17 R = R1 = Me 1 R = R1 = H 9 R = SMe, R1 = H 10 R-R = SCH2CH2S, R1 = H 11 R = Me, R1 = H 12 R = R1 = Me Li (i) 13 building block for analogous TTF systems, using the deprotec- Scheme 2 Reagents and conditions: (i ) LDA, Et2O (THF for compound 1028), -78 °C, 1.5 h; (ii) MeNCS, -78 to 20 °C over 12 h; (iii) H2O tion–dithiolate alkylation protocol of Becher and co-workers.27 S CN S CN S S S S H H SR SR S S S S H H S CN S CN S S S S H H H Ph3P H S CN S CN S S S S S H H X I– BF4 – 4 5 6 9 R = Me 10 R-R = CH2CH2 8 X = O or S 7 (i) (ii) (iii), (iv) Scheme 1 Reagents and conditions: (i ) Et3N, MeCN, 20 °C, 2 h; (ii) P(OR)3 (R=Me or Et), reflux, 4 h; (iii ) NaOEt, EtOH–THF, 20 °C, 4 h; (iv) either MeI or Br(CH2)2Br, 20 °C, 16 h 1542 J.Mater. Chem., 1998, 8(7), 1541–1550CN CN NC NC a b 19 d c NO2 O2N CN NC NO2 20 Compounds 2 and 14–17 all form 151 complexes when mixed with equimolar amounts of TCNQ in hot acetonitrile (isolated as black powders in 35–65% yields).A 251 complex Fig. 3 Donor and acceptor in the structure of 14·TCNQ of compound 17 and 2,4,7-trinitro-9-(dicyanomethylene)fluorene 2029 was isolated from chlorobenzene solution. Similarly, metathesis reactions of compounds 2 and 14–17 with tetra- angles of folding of the dithiole rings along the S,S vectors butylammonium tribromide in hot acetonitrile aVorded 151 (h1 for the thiocarbamoyl-substituted ring and h2 for the bromide salts (isolated as black powders in 62–75% yields). unfunctionalised ring) and the S(1)–C(2)–C(7)–S(5) torsion The conductivity of the complexes, (2 and 14–17)·TCNQ, angle t, defining the orientation of the planar thiocarbamoyl (17)2·20 and (2 and 14–17)·Br, are in the range srt=5×10-8 group.-1.2 S cm-1 (2-probe, compressed pellet measurements, The S(2)–C(3) bond in neutral species 16 and 17 is signifi- 20 °C). The value of 1.2 S cm-1 for 2·TCNQ is a remarkably cantly shorter than other chemically equivalent C–S bonds, high value. by 0.036 A ° in 16 and 0.013 A ° in 17 (cf. 0.03 A ° in 2); this is in agreement with our previous suggestion10a,30 that the meso- X-Ray crystal structures meric eVect of the CNS group in structure 18 gives rise to a significant contribution from the canonical form 18¾, thereby The molecular structures of 16, 17 and the 2·Br and 14·TCNQ increasing the polarity of the TTF moiety (Scheme 3). The salts are shown in Fig. 1–3. The structure of (17)2·20 is shown S(2)–C(3) bond contraction is smaller [and the C(2)–C(7) in Fig. 8.The molecular geometry of the TTF unit (Table 1) bond marginally longer] in 17, wherein the methyl substituent can be compared with that of neutral 2, studied earlier at at C(3) forces an increase of t, interfering with the conjugation. ambient10a and low (150 K)10c temperatures. The central C2S4 Charge transfer from the TTF moiety in 2·Br and 14·TCNQ system of the TTF moiety remains approximately planar in has the usual eVect upon bond distances.31 As the HOMO has all cases; thus, molecular conformations can be described by nodes in the C–S bonds but not in CNC, the latter lengthen (especially the central CNC bond, by 0.04–0.05 A ° ) and the former contract (by 0.02–0.04 A ° ), with the exception of the S(2)–C(3) bond which shortens insignificantly in 14+· and not at all in 2+· .Thus, the non-equivalence of C–S bonds diminishes but persists; this eVect can be easily understood when considering the resonance form 18¾. It is obvious that electron density is withdrawn only from the TTF moiety while the thiocarbamoyl group is practically unaVected.In a TCNQ moiety 19, a and c bonds tend to lengthen and b and d bonds tend to contract with increasing negative charge on the acceptor, b–c and d–c values changing practically linearly32 from 0.069 and 0.062 A ° , respectively, in neutral TCNQ33 to (average) 0.037 and 0.030 A° in a series of A·(TCNQ)2 salts (A=cation with a tetraalkylammonium centre)34 and to zero in a TCNQ radical anion.35 In 14·TCNQ, the average bond lengths are: a=1.367(3), b=1.433(7), c= 1.411(6) and d=1.428(4) A ° (see Table 2) suggesting that the charge transfer (r) is ca. 0.7 electrons per molecule; this is slightly higher than in the parent TTF-TCNQ salt which has segregated stacks of donor and acceptor moieties,36 where similar geometrical estimates give r=0.6 and neutron diVusion methods37a give r=0.59.Similar r values were estimated from Fig. 1 Molecular structures of 16 (a) and 17 (b), showing ‘strong’ IR spectroscopic data.37b The charge r+ on the TTF moiety (dashes) and ‘weak’ (dots) intermolecular hydrogen bonds can be calculated using linear dependence f=1.757–0.0385(r+),31b where f is the average length of the inner C–S bonds. For 14·TCNQ this gives r+ $0.5; the discrepancy with the TCNQ charge may be due to the substituents at the TTF moiety (the equation was derived for unsubstituted TTF).The central CNC bond, slightly shorter than in 2·Br, also suggests partial oxidation of 14. The crystal of 16 is broadly isostructural with that of 2, comprising dimers of molecules (contacting face-to-face) arranged in a kappa mode, i.e.in chessboard fashion with the molecular planes of contacting dimers mutually perpendicular Fig. 2 Hydrogen-bonded cation and anion in the structure of 2·Br and their long axes parallel [Fig. 4(a)]. However, molecules J. Mater. Chem., 1998, 8(7), 1541–1550 1543Table 1 Experimental and calculated selected bond distances (A ° ) and dihedral angles (°) in 2, 14, 16 and 17, and their radical cations X-ray PM3 X-ray PM3 X-ray PM3 X-ray PM3 2 2·Br 2 2+· 14·TCNQ 14 14+· 16 16 16+· 17 (17)2·20a 17 17+· S(1)–C(1) 1.763(3) 1.727(3) 1.762 1.755 1.743(2) 1.763 1.752 1.765(3) 1.763 1.756 1.760(2) 1.758(4) 1.761 1.756 S(2)–C(1) 1.758(3) 1.738(3) 1.764 1.731 1.728(2) 1.764 1.732 1.762(3) 1.764 1.732 1.754(2) 1.748(4) 1.760 1.727 S(1)–C(2) 1.769(2) 1.745(3) 1.771 1.775 1.754(2) 1.771 1.774 1.767(3) 1.771 1.776 1.766(2) 1.777(4) 1.764 1.782 S(2)–C(3) 1.726(3) 1.725(3) 1.726 1.724 1.717(2) 1.726 1.723 1.728(3) 1.726 1.725 1.749(2) 1.736(4) 1.757 1.752 C(2)–C(3) 1.343(3) 1.352(4) 1.352 1.349 1.354(3) 1.352 1.349 1.342(5) 1.352 1.348 1.350(3) 1.357(6) 1.354 1.353 1)–C(4) 1.348(3) 1.396(5) 1.351 1.392 1.386(3) 1.350 1.393 1.346(4) 1.351 1.392 1.344(3) 1.351(5) 1.351 1.392 S(3)–C(4) 1.760(3) 1.718(3) 1.765 1.730 1.743(2) 1.758 1.721 1.749(3) 1.760 1.725 1.760(2) 1.760(4) 1.760 1.725 S(4)–C(4) 1.761(3) 1.728(3) 1.764 1.733 1.735(2) 1.758 1.724 1.759(3) 1.760 1.728 1.759(2) 1.759(4) 1.760 1.729 S(3)–C(5) 1.753(3) 1.718(4) 1.741 1.713 1.754(2) 1.754 1.736 1.762(3) 1.758 1.728 1.760(2) 1.756(4) 1.757 1.728 S(4)–C(6) 1.749(3) 1.727(4) 1.741 1.713 1.741(2) 1.754 1.736 1.764(3) 1.757 1.729 1.760(2) 1.755(4) 1.757 1.729 C(5)–C(6) 1.317(5) 1.341(5) 1.342 1.356 1.361(3) 1.357 1.370 1.335(5) 1.352 1.369 1.333(3) 1.353(6) 1.352 1.369 C(2)–C(7) 1.467(3) 1.490(4) 1.476 1.462 1.483(2) 1.476 1.462 1.465(4) 1.476 1.462 1.477(3) 1.475(6) 1.480 1.463 C(7)–S(5) 1.680(3) 1.680(3) 1.634 1.671 1.667(2) 1.634 1.669 1.678(3) 1.634 1.670 1.675(2) 1.674(5) 1.627 1.675 C(7)–N 1.321(3) 1.312(4) 1.382 1.356 1.328(3) 1.382 1.357 1.322(4) 1.382 1.356 1.326(3) 1.325(6) 1.382 1.354 h1 23.6(1) 3.4(3) 4.8(2) 17.5(2) 4.1(1) 0 h2 14.5(1) 1.8(3) 6.9(2) 4.9(2) 3.9(1) 0 t 12.4(3) 9.6(5) 143.0 0.0 1.1(3) 143.2 0.0 18.1(3) 143.0 0.0 27.7(3) 3.9(6) 101.6 1.2 aFor (17)2·20, carbon atom numbers are C(1n) instead of C(n), N(1) for N.S S S S S NH Me S S S S S– NH Me R R R1 R1 R R 18 18¢ Scheme 3 Table 2 Bond distances (A° ) in the TCNQ radical anion of 14·TCNQ C(11)–C(12) 1.444(3) C(11)–C(16) 1.427(3) C(13)–C(14) 1.427(3) C(14)–C(15) 1.436(3) C(12)–C(13) 1.366(3) C(15)–C(16) 1.367(3) C(11)–C(17) 1.404(3) C(14)–C(18) 1.417(3) C(17)–C(19) 1.433(3) C(17)–C(20) 1.428(3) C(18)–C(21) 1.426(3) C(18)–C(22) 1.423(3) C(19)–N(2) 1.148(3) C(20)–N(3) 1.153(3) C(21)–N(4) 1.153(3) C(22)–N(5) 1.154(3) within a dimer overlap diVerently in 2 and 16.This overlap can be described by relative shifts (slips) of the molecules from a perfectly eclipsed position: longitudinal (d) parallel to the central CNC bond and lateral (e) perpendicular to it. In 2, d= 1.6 and e=0.2 A ° which corresponds to the most common type of overlap observed in structural studies of TTF salts, in which the central CNC bond of one molecule lies over a dithiole ring of the other.In 16 the shifts are much larger, d=3.4 and e=1.6 A ° , the S(4) atom lying over the centre of a dithiole ring (Fig. 5). The interplanar separation (d) between central TTF planes (C2S4) in 16 (3.40 A ° ) is larger than in 2 (3.32 A ° ) and the shortest intradimer S,S contacts in the former (3.70–3.74 A ° ) are slightly longer than twice the van der Waals radius of sulfur (1.81 A ° ).38 Interdimer contacts between the TTF sulfur atoms are even longer (3.88–4.00 A ° ).A short S(5),S(5¾) distance (3.34 A ° ) between thiocarbonyl groups contacting in a nearly linear fashion (C–S,S 172.4°) is in good agreement Fig. 4 kappa-Packing of dimers in 16 (a) showing interplanar separawith the supposed ‘ellipsoidal’ shape of monocoordinate sulfur tions in the dimers, d=3.40 A ° , and of molecules in 17 (b) atoms (with the van der Waals ‘radii’ of 1.60 and 2.03 A ° for the C–S,S angles of 180 and 90°, respectively)39 although this model leaves entirely open the question of the physical nature of 1.03 A ° , obtained from neutron diVraction40), respectively, indicate that the hydrogen bonding is principally with S(5).of this asphericity. The NH group in 16 forms a bifurcated contact with the S(1) and S(5) atoms of a molecule belonging The structure of 17 can be seen as the orthogonal packing of individual molecules rather than dimers [Fig. 4(b)]. to a perpendicularly oriented dimer and symmetry related with the first one via glide plane c.The N,S distances of 3.783(3) Here, the molecule is much closer to a planar conformation. The intermolecular contacts between TTF sulfur atoms and 3.425(3) A° and H,S of 2.97 and 2.51 A° (henceforth all hydrogen bond distances are normalised for N–H bond length (3.95–4.02 A ° ) and the short S(5),S(5¾) contact of 3.59 A ° are 1544 J.Mater. Chem., 1998, 8(7), 1541–1550Fig. 6 Interstack packing in the structure of 2·Br Fig. 5 Molecular overlap in the structures of 16 (a) and 2·Br (b, c) similar to those observed in the structure of 16. The NH group also forms a bifurcated contact with the S(1) and S(5) atoms of a ‘perpendicular’ molecule, but in this case the former atom is more strongly bonded [N,S(1) 3.456(2), N,S(5) 3.697(2), H,S(1) 2.64, H,S(5) 3.06 A° , respectively].The shorter H,S distances thus fall within the normal range for NH,S hydrogen bonding, 2.28–2.72 A ° (average 2.46 A ° ) from neutron diVraction.40 The motif of 2·Br is entirely diVerent to that of neutral 2. Here, 2+· radical cations have almost planar TTF moieties and form an infinite stair-like stack of dimers, running parallel to the y direction. Adjacent cations in a stack are inversionrelated.Within a dimer, TTF moieties are nearly eclipsed (d= 0.6, e=0.1 A ° ), with the interplanar separation d=3.34 A ° and S,S contacts 3.35–3.39 A° . Between these dimers, d=3.46 A° (shortest S,S distances 3.77–3.94 A ° ) and the overlap of TTF moieties is only partial (d=3.1, e=1.4 A ° ).However, this arrangement corresponds to the best possible overlap of the Fig. 7 Mixed stack in the structure of 14·TCNQ entire cations of 2 (i.e. including the N-methylthiocarbamoyl substituent). All TTF planes within a single stack of 2·Br are parallel but hydrogen bond with an anion [H,N 2.00 A ° ], supported by a short C(3)–H,N(5) contact [H,N 2.34 A ° ] with the same inclined by 46° to those of adjacent stacks.The interstack contacts S(1),S(2¾) 3.88, S(5),S(3) 3.47, S(5),S(2¾) 3.54 A ° anion. In the structure of (17)2·20, molecule 17 is in a general and their symmetrical equivalents (Fig. 6) form infinite chains parallel to the z axis. The anion is situated between stacks and position, while that of 2,4,7-trinitro-9-(dicyanomethylene) fluorene (20) is situated on a crystallographic two-fold is linked with the cation by a unique strong N–H,Br hydrogen bond [H,Br 2.22 A ° ] into an ionic pair. Even the N,Br axis, passing through the atoms C(7), C(8) and the midpoint of the C(5)–C(5¾) bond (Fig. 8). One nitro group in 20 is thus distance of 3.210(2) is shorter than the sum of their van der Waals radii (3.51 A ° ).The contacts Br,S(4) of 3.38 and disordered between two positions (4 and 5, in chemical notation) with equal occupancies. The same kind of disorder Br,S(3) of 3.57 A ° with other cations are shorter than the sum of their van der Waals radii (3.68 A ° ).38 This eVect can be (with the same crystallographic symmetry) was displayed by pristine 20;29b the consequent low precision of the molecular explained in terms of specific polar (electrostatic) interactions.2a The structure of 14·TCNQ comprises mixed stacks of a geometry makes comparison with the present structure diYcult.Very elongated thermal ellipsoids of its oxygen atoms, O(41) ,DDAADD, type (Fig. 7) with interplanar separations of TTF,TTF ca. 3.5, TTF,TCNQ ca. 3.2, and TCNQ,TCNQ and O(42), indicate further rotational disorder, probably of a continuous character, since our attempts to refine these atoms ca. 3.25 A ° . The TTF moiety shows only insignificant folding of dithiole rings and a small (3.8°) twist around the central in ‘split’ positions gave no improvement. The intermolecular distance O(41),C(18) 2.83 A° is much shorter than the mean C(1)NC(4) bond. The TCNQ radical anion is planar to within ±0.04 A ° (with a slight boat-like distortion).The TTF moieties van derWaals contact of 3.24 A ° , but the shortening is probably spurious, due to the disorder of the nitro group and probably are laterally slipped (d=0.4, e=1.1 A ° ) and the overlap of TCNQ moieties is of a common ring over CNC bond type. also of C(18), which has large displacement parameters.The conformation of 20 contrasts with those of 2,4,5,7- The NH group is engaged in an N(1)–H,N(5) interstack J. Mater. Chem., 1998, 8(7), 1541–1550 1545favourable stacking, one with sulfur atoms over gaps, corresponds to the minimum of orbital overlap and vice versa, the fully eclipsed stacking is least favourable sterically but most favourable in the sense of HOMO overlap.The former arrangement is observed for neutral TTF derivatives, where HOMOs are fully occupied and their overlap can give no significant bonding contribution. When the TTF system is oxidised to +1, the overlap of these orbitals (now SOMOs) produces the maximum energy gain—hence, the nearly-eclipsed overlap in 2·Br [Fig. 5(b)]. On the other hand, oxidation makes dithiole rings more aromatic in character and simultaneously reduces the repulsion between non-bonding electrons of the sulfur atoms, which causes the boat-like folding of TTF43 (observed, remarkably, in the gas phase44). Therefore, positively charged TTF moieties are invariably planar, while neutral ones adopt both planar and folded conformations (by various degrees).45 Fig. 8 Donor and acceptor molecules in the structure of (17)2·20.Primed atoms are symmetrically related via twofold axis. Molecular orbital calculations To gain a better understanding of bonding in thiocarbamoyl tetranitro analogues.41 In the latter, the fused tricyclic system TTF derivatives, we performed molecular orbital quantum- adopts a boat-like conformation with the dihedral angle of chemical calculations at the Hartree–Fock SCF level of 8–12° between the benzene rings.The nitro groups in positions approximation, for neutral molecules 2, 14, 16 and 17 and 4 and 5 are bent out of the mean molecular plane in opposite their radical cations. The molecular geometries, fully optimised directions and form dihedral angles of 27–42° with this plane. using the PM3 semi-empirical method46 (Table 1), show an In 20, the fused system is essentially planar, the nitro group at average discrepancy with the experimental (X-ray) bond dis- C(4) is inclined to its plane (A) by 77°, but its nitrogen atom tances (for neutral molecules) of 0.006 A ° in the TTF moiety does not deviate from plane A.Thus the steric strain is and 0.012 A ° overall. Particularly, the observed shortening of completely relieved.The nitro group at C(2) and the C(CN)2 the S(2)–C(3) bond as compared to chemically equivalent moiety are inclined to plane A insignificantly, by 5 and 2°. C–S bonds (see above) is well reproduced by the calculations. Molecule 17 in this structure is almost flat. Both dithiole rings The biggest discrepancies are observed for the side-chain are rigorously planar, twist angles around the C(11)–C(14) bonds, C(7)NS(5) (underestimated by 0.044–0.048 A ° ) and and C(12)–C(17) bonds are ca. 4°. C(7)–N (overestimated by 0.056–0.060 A° ). Similar discrepanc- Molecules in the crystal form a mixed ,DDADDA, stack ies were obtained earlier9c for (N,N-dimethylcarbamoyl)-TTF (Fig. 9), with mean interplanar separations D,A 3.35 and and manifest a general shortcoming of the PM3 method in D,D 3.58 A ° .However, bond lengths in the 17 moiety show calculating the structures of various nitrogen-containing only minor and somewhat irregular changes from the neutral compounds.47 species (Table 1), i.e. the overall charge transfer is rather small. Calculated changes in bond lengths upon oxidation of The overlap between donor and acceptor molecules is shown neutral molecules to radical cations do not fully agree with in Fig. 8, the overlap between adjacent molecules of 17 the experimental data for 2A2·Br. In the latter, oxidation resembles that in 2·Br (see above). The distance between the causes lengthening of all CNC bonds and contraction of all amido group and the cyano nitrogen atom N(9) (H,N S–C bonds [except the already short S(2)–C(3) bond] in the 2.45 A ° ) is characteristic of a weak dipole–dipole interaction TTF moiety.The calculated bond lengths behave mostly in rather than a real hydrogen bond. the same way, but the S(1)–C(1), S(1)–C(2) and C(2)–C(3) The overlap between TTF moieties in various structures can bonds remain practically unaltered. Calculations for all other be rationalised as a compromise between two tendencies: to molecule–cation-radical pairs studied herein reproduce the minimise the core (steric) repulsion and to maximise the same pattern.overlap between the HOMOs.42 The main contributors to the The HOMO and LUMO energies for molecules 2, 14, 16, repulsion are sulfur atoms, which are substantially bigger than 17 and their cation radicals were calculated. Although the the carbon atoms.On the other hand, the HOMO is localised PM3 method is rather poor at predicting absolute HOMO mainly on the central C2S4 moiety of the TTF, and especially energies (e.g. giving -7.99 eV for 1, against the experimental on the sulfur atoms. Boat-like bending of a TTF system can ionisation potential in the gas phase of 6.70 eV48), the calcu- only increase this localisation.42d Thus the most sterically lated relative stabilities of the HOMOs, 14 (-8.27 eV)<2 (-8.23 eV)<16 (-8.14 eV)<17 (-8.05 eV), are in good agreement with the succession of E11/2 oxidation potentials for these compounds (0.52, 0.43,9c 0.41, and 0.38 V, respectively).The Mulliken atomic charges in neutral molecules are strongly negative [-(0.19–0.33) e] on carbon atoms and positive (0.25–0.30 e) on sulfur atoms, similar to related TTF derivatives reported earlier.9c The positive charge in the radical cations is delocalised less on the dithiole ring (A) linked to the thiocarbamoyl group (0.2–0.3 e) than on the other ring (ca. 0.5 e). The calculated minimum-energy conformation of neutral molecules is characterised by a planar TTF moiety and a substantial twist around the C(2)–C(7) bond.The torsion angle t (see above) is ca. 143° in 2, 14 and 16; in 17 the 5- methyl substituent induces a stronger twist, t=101.6°. These conformations are quite diVerent from those observed in the Fig. 9 Crystal packing of (17)2·20, projection on the (0 1 0) plane crystal structures. However, the rotation barriers around the 1546 J.Mater. Chem., 1998, 8(7), 1541–1550C(2)–C(7) bonds are very low. Those calculated in rigid-rotor FTIR spectrometer operated from a Grams Analyst 1600. UV Spectra were obtained on a Kontron Uvicon 930 spectrophoto- approximation [i.e. without a relaxation of all other molecular coordinates during the rotation around the C(2)–C(7) bond], meter using quartz cells.Extinction coeYcients (e) are all quoted in M-1 cm-1. Melting points were obtained on a Kofler which always gives an overestimate of the barrier energy, do not exceed 2.6 kcal mol-1. Indeed, calculation of 14 as a hot-stage microscope apparatus and are uncorrected. Cyclic voltammetric data were obtained on a BAS 50W electrochemi- completely optimised rotor gives even smaller barriers of 1.15 kcal mol-1 at t=4.6° (cis-planar conformation) and cal analyser (1×10-5 M solution of donor in acetonitrile under argon, 1×10-1 M tetrabutylammonium perchlorate supporting 0.19 kcal mol-1 at t=179.4° (trans-planar conformation).Therefore, t can easily change under the influence of crystal electrolyte, platinum working and counter electrodes, Ag/AgCl reference electrode, 20 °C).All reagents were of commercial packing and hydrogen bonds. In 17 the adjacent methyl group introduces a much higher barrier to rotation at t=0 or 180°, quality and used as supplied unless otherwise stated; solvents were dried where necessary using standard procedures and comparable with that in the cation radical; however, further from t=0 or 180° the rotation is practically free.For radical distilled for chromatographic use. cations 2+· , 14+· , 16+· and 17+· the favourable conformation 4,5-Bis(2-cyanoethylsulfanyl )tetrathiafulvalene 6 is cis-planar with t close to 0° and the rotation barriers (now defined by electron conjugation) an order of magnitude higher To a stirred solution of salt 423 (10.0 g, 16.08 mmol) and than in neutral 2, 14 and 16.This agrees well with the X-ray phosphonium salt 524 (3.70 g, 16.08 mmol) under argon at structures of 2·Br, 14·TCNQ, and (17)2·20, exhibiting only 20 °C was dropwise added triethylamine (10 cm3, excess) over minor twists. 0.5 h and the mixture stirred overnight. After evaporation, the Whether the favourable conformation of the TTF moiety in residue was chromatographed (silica, 200–400 mesh) eluting neutral molecules is planar is open to discussion. These results with dichloromethane to aVord compound 6 as orange needles seem to be extremely basis-sensitive,43 while electron diVraction (4.69 g, 78%), mp 77 °C (from MeCN) (Analysis found: C, data are indicative of a non-planar conformation in the gas 38.4; H, 2.9; N, 7.5; C12H10N2S6 requires: C, 38.5; H, 2.7; N, phase.44 The ab initio calculations (6–31G* level ) show that 7.5%); m/z (DCI) 375 (M++1, 100%); dH (CDCl3) 6.37 (2H, in any case the potential well is extremely flat-bottomed: s), 3.09 (4H, t, J 7.1), 2.74 (4H, t, J 7.1); lmax (MeCN) (e) 390 folding of both dithiole rings by h=10 and 20° needs only 0.1 (1.3×103), 310 (5.1×103), 294 (4.8×103), 195 (7.2×103) nm; and 1.0 kcal mol-1, respectively.9c The former energy is hardly CV (MeCN) E11/2 0.41, E21/2 0.79 V.of any real significance, and even the latter may be easily furnished by intermolecular interactions. 4,5-Bis(methylsulfanyl )tetrathiafulvalene 9 To a stirred solution of compound 6 (2.00 g, 5.35 mmol) in Conclusions dry ethanol–tetrahydrofuran (151 v/v, 150 cm3) under argon at 20 °C was added sodium ethoxide [from sodium (245 mg, The present study illustrates the fact that recent advances in 10.70 mmol) in ethanol (25 cm3)] and the mixture was stirred synthetic TTF chemistry provide access to a far wider range for 4 h (during this time the solution colour changed from of TTF derivatives than was hitherto available, including orange to deep red).Methyl iodide (2 cm3, excess) was added highly-functionalised systems which possess interesting solid and the mixture stirred at 20 °C for a further 12 h.The mixture state properties. Following our earlier work on (N-methylwas diluted with dichloromethane (250 cm3), washed with thiocarbamoyl)-TTF 2,10 we have eYciently synthesised comwater (3×100 cm3), the organic phase dried (MgSO4) and the pounds 14–17, which were designed to explore the eVects of solvent was evaporated in vacuo.Chromatography (silica, the N-methylthiocarbamoyl substituent on the properties of 200–400 mesh) of the residue eluting with hexane–dichloro- the TTF system. Solution UV–VIS spectroscopic and cyclic methane (151 v/v) aVorded compound 9 (1.44 g, 91%) as a voltammetric studies establish that there is intramolecular red solid, mp 65 °C ( lit.,18 mp 63–65 °C); dH (CDCl3) 6.33 (s, charge-transfer between the donor TTF and the acceptor N- 2H), 2.42 (s, 6H).methylthiocarbamoyl moieties. Within this series of TTF derivatives, X-ray crystal structures reveal unusual packing 4,5-(Ethylenedisulfanyl )tetrathiafulvalene 10 motifs in both the neutral and radical cation species, with intermolecular interactions arising from hydrogen bonding To a stirred solution of compound 6 (1.00 g, 2.67 mmol) in and/or electrostactic interactions of the N-methylthiocarba- dry ethanol–tetrahydrofuran (151 v/v, 150 cm3) under argon moyl group.The observation of distorted kappa packing modes at 20 °C was added sodium ethoxide [from sodium (123 mg, in the structure of 16 is notable.Crystal structures have been 5.35 mmol) in ethanol (25 cm3)] and the mixture was stirred obtained for two 151 salts, viz. 2·Br and 14·TCNQ and for the for 4 h. 1,2-Dibromoethane (0.5 cm3, excess) was added and complex (17)2·20: to date, single crystals of segregated-stack, the mixture stirred at 20 °C for a further 12 h. The mixture mixed-valence salts of donors 2 and 14–17 have not been was concentrated to ca. 30 cm3 and the precipitated solid obtained. The synthesis of new TTF systems bearing substitu- collected by filtration and washed with ethanol (25 cm3). ents which can engage in intermolecular interactions, and the Chromatography (silica, 200–400 mesh) of the residue eluting radical cation salts of these donors, is presently under investi- with dichloromethane aVorded compound 10 (637 mg, 81%) gation, with a view to the ’supramolecular engineering’ of new as an orange solid, mp 204–206 °C ( lit.,18 mp 206–207 °C); dH materials in which intermolecular interactions regulate the (CDCl3) 6.32 (s, 2H), 3.29 (s, 4H).structural and transport properties in a controllable manner.49 General procedure for compounds 14–17 Experimental To a stirred solution of compound 9–12 (3.0 mmol) in dry diethyl ether (tetrahydrofuran for compound 1028) (100 cm3) General methods under argon at -78 °C was added lithium diisopropylamide –tetrahydrofuran complex (2.06 cm3, 3.1 mmol of a 1.5 M 1H NMR Spectra were obtained on a Bruker AC 250 spectrometer operating at 250.134 MHz. 13C NMR Spectra were solution in cyclohexane) and stirring continued for 2 h.Methyl isothiocyanate (330 mg, 4.5 mmol) was added and the mixture obtained on a Varian 400 spectrometer operating at 100.581 MHz. Coupling constants J are given in Hz. Mass allowed to warm to 20 °C over 12 h. Water (100 cm3) was added and the mixture stirred for 2 h (during this time the spectra were recorded on a VG7070E spectrometer operating at 70 eV.IR Spectra were recorded on a Perkin-Elmer 1615 organic phase became burgundy red in colour). The mixture J. Mater. Chem., 1998, 8(7), 1541–1550 1547was extracted with toluene (3×100 cm3), the organic extracts 2.3; N, 14.5%]; nmax (KBr) 2200 cm-1 (CON); srt (2-probe, compressed pellet) 1.2 S cm-1. were combined, washed with water (100 cm3), dried (MgSO4) and evaporated in vacuo. Chromatography (silica 200–400 mesh for compounds 14 and 15; alumina 70–230 mesh for 4-(N-Methylthiocarbamoyl )tetrathiafulvalene 14: TCNQ compounds 16 and 17) of the residue eluting initially with salt. 29 mg, 50% [Analysis found: C, 45.9; H, 2.6; N, 12.4; hexane–toluene (151 v/v) aVorded unreacted compounds 9–12; C22H15N5S7 (151 stoichiometric complex) requires: C, 46.0; H, subsequent elution with toluene aVorded the products.There 2.6; N, 12.2%]; nmax (KBr) 2188 cm-1 (CON); srt (2-probe, were thus obtained: compressed pellet) 5×10-6 S cm-1. 4-(N-Methylthiocarbamoyl )-4¾,5¾-bis(methylsulfanyl )tetra- 4-(N-Methylthiocarbamoyl )tetrathiafulvalene 15: TCNQ thiafulvalene 14. Black needles (697 mg, 63%), mp 134–135 °C salt. 20 mg, 35% [Analysis found: C, 46.5; H, 2.2; N, 12.1; (from MeCN) (Analysis found: C, 32.3; H, 3.2; N, 3.8; C22H13N5S7 (151 stoichiometric complex) requires: C, 46.2; H, C10H11NS7 requires: C, 32.5; H, 3.0; N, 3.8%); m/z (EI) 369 2.3; N, 12.3%]; nmax (KBr) 2212, 2198 cm-1 (CON); srt (2- (M+, 70%), (CI) 370 (M++1, 65); dH (CDCl3) 7.25–7.20 (1H, probe, compressed pellet) 3×10-6 S cm-1.br s), 7.11 (1H, s), 3.24 (3H, d, J 4.9), 2.42 (6H, s); nmax (KBr) 3400–3200 ( br, NH), 1532, 1516, 1362 (CNS), 1205, 1044, 4-(N-Methylthiocarbamoyl )tetrathiafulvalene 16: TCNQ 788 cm-1; lmax (MeCN) (e) 435 (2.8×103), 325 (sh, 2.1×104), salt. 33 mg, 65% [Analysis found: C, 51.4; H, 3.0; N, 14.3; 290 (2.8×104), 198 (2.9×104) nm; CV (MeCN) E11/2 0.52, C22H15N5S5 (151 stoichiometric complex) requires: C, 51.8; H, E21/2 0.83 V. 3.0; N, 13.7%]; nmax (KBr) 2204 cm-1 (CON); srt (2-probe, compressed pellet) 9×10-2 S cm-1. 4-(N-Methylthiocarbamoyl )-4¾,5¾-(ethylenedisulfanyl )tetrathiafulvalene 15. Black needles (595 mg, 54%), mp 185–186 °C 4-(N-Methylthiocarbamoyl )tetrathiafulvalene 17: TCNQ (from CS2) (Analysis found: C, 32.6; H, 2.6; N, 3.9; C10H9NS7 salt. 32 mg, 61% [Analysis found: C, 52.2; H, 3.2; N, 13.4; requires: C, 32.7; H, 2.5; N, 3.8%); m/z (EI) 367 (M+, 45%); C23H17N5S5 (151 stoichiometric complex) requires: C, 52.8; H, (CI) 368 (M++1, 35); dH (CDCl3) 7.30–7.20 (1H, br s), 7.11 3.3; N, 13.4%]; nmax (KBr) 2200 cm-1 (CON); srt (2-probe, (1H, s), 3.30 (4H, s), 3.23 (3H, d, J 4.8); nmax (KBr) 3350–3150 compressed pellet) 7×10-2 S cm-1.(br, NH), 1524, 1518, 1358 (CNS), 1205, 1045, 791 cm-1; lmax (MeCN) (e) 450 (1.2×103), 326 (sh, 1.0×104), 309 (1.3×104), 4-(N-Methylthiocarbamoyl )tetrathiafulvalene 2: bromide salt. 231 (1.1×104), 208 (1.3×104) nm; CV (MeCN) E11/2 0.53, 27 mg, 75% [Analysis found: C, 26.8; H, 2.1; N, 4.0; C8H7BrNS5 E21/2 0.86 V. (151 stoichiometric complex) requires: C, 26.9; H, 2.0; N, 3.9%]; srt (2-probe, compressed pellet) 5×10-5 S cm-1. 4-(N-Methylthiocarbamoyl )-4¾,5¾-dimethyltetrathiafulvalene 16. Black plates (604 mg, 66%), mp 171–172 °C (from toluene) 4-(N-Methylthiocarbamoyl )tetrathiafulvalene 14: bromide (Analysis found: C, 39.5; H, 3.8; N, 4.4; C10H11NS5 requires: salt. 28 mg, 62% [Analysis found: C, 26.9; H, 2.7; N, 3.2; C, 39.3; H, 3.6; N, 4.6%); m/z (EI) 305 (M+, 45%), (CI) 306 C10H11BrNS7 (151 stoichiometric complex) requires: C, 26.7; (M++1, 65); dH (CDCl3) 7.30–7.20 (1H, br s), 7.19 (1H, s), H, 2.5; N, 3.1%]; srt (2-probe, compressed pellet) 1×10-8 3.23 (3H, d, J 4.8), 1.95 (6H, s); dC (CDCl3) 188.5, 133.7, 129.4, S cm-1. 123.0, 122.7, 116.2, 105.2, 32.8, 16.2, 13.7; nmax (KBr) 3400–3250 (br, NH), 1542, 1520, 1346 (CNS), 1216, 1048, 773 cm-1; lmax 4-(N-Methylthiocarbamoyl )tetrathiafulvalene 15: bromide (MeCN) (e) 467 (1.8×103), 323 (sh, 1.9×104), 297 (2.2×104), salt. 31 mg, 70% [Analysis found: C, 26.6; H, 2.1; N, 2.8; 205 (1.7×104) nm; CV (MeCN) E11/2 0.41, E21/2 0.79 V. C10H9BrNS7 (151 stoichiometric complex) requires: C, 26.8; H, 2.0; N, 3.1%]; srt (2-probe, compressed pellet) 3×10-8 S cm-1. 4-(N-Methylthiocarbamoyl )-4¾,5,5¾-trimethyltetrathiafulvalene 17. Dark red needles (670 mg, 70%), mp 174–175 °C (from toluene) (Analysis found: C, 41.3; H, 4.2; N, 4.3; 4-(N-Methylthiocarbamoyl )tetrathiafulvalene 16: bromide C11H13NS5 requires: C, 41.4; H, 4.1; N, 4.4%); m/z (EI) 319 salt. 25 mg, 65% [Analysis found: C, 31.3; H, 2.9; N, 3.8; (M+, 85%), (CI) 320 (M++1, 100); dH (CDCl3) 7.40–7.30 C10H11BrNS5 (151 stoichiometric complex) requires: C, 31.2; (1H, br s), 3.21 (3H, d, J 4.8), 2.25 (3H, s), 1.94 (6H, s); dC H, 2.9; N, 3.6%]; srt (2-probe, compressed pellet) 3×10-6 (CDCl3) 186.6, 137.4, 125.3, 123.0, 122.7, 113.2, 105.4, 33.2, Scm-1. 13.7; nmax (KBr) 3400–3250 (br, NH), 1572, 1524, 1347 (CNS), 1218, 1052, 733 cm-1; lmax (MeCN) (e) 395 (2.2×103), 328 (sh, 4-(N-Methylthiocarbamoyl )tetrathiafulvalene 17: bromide 1.4×104), 297 (2.2×104), 213 (sh, 1.9×104), 205 (2.2×104) nm; salt. 29 mg, 74% [Analysis found: C, 33.2; H, 3.2; N, 3.5; CV (MeCN) E11/2 0.38, E21/2 0.76 V. C11H13BrNS5 (151 stoichiometric complex) requires: C, 33.1; H, 3.3; N, 3.5%]; srt (2-probe, compressed pellet) 1×10-6 S cm-1. General procedure for compounds 2 and 14–17: TCNQ and bromide salts 4-(N-Methylthiocarbamoyl )tetrathiafulvalene 17: 2,4,7-trini- A solution of compound 2 and 14–17 (0.1 mmol) in hot, dry tro-9-(dicyanomethylene)fluorene 20 complex: (17)2·20.acetonitrile (10 cm3) and a solution of (i) 7,7,8,8-tetracyano-p- Compound 2029a (3.1 mg, 0.85 mmol) was dissolved in hot quinodimethane (TCNQ) (20 mg, 0.1 mmol), or (ii) tetrabutylchlorobenzene (1 cm3). The solution was cooled to 35–40 °C, ammonium tribromide (48 mg, 0.1 mmol) in hot, dry acetothen compound 17 was added.After a week at room temperanitrile (10 cm3) were mixed and refluxed for 10 min under ture black crystals (6.2 mg, 73%) suitable for X-ray analysis argon. After cooling, the product precipitated (for 2d·TCNQ were isolated by filtration; srt (2-probe, compressed pellet) slow evaporation of the solvent was necessary before precipi- 5×10-4 S cm-1.tation occurred); the black solid was collected by filtration and washed with ethanol. There were thus obtained: X-Ray crystallography All single-crystal X-ray diVraction experiments were carried 4-(N-Methylthiocarbamoyl )tetrathiafulvalene 2: TCNQ salt. 24 mg, 50% [Analysis found: C, 50.0; H, 2.5; N, 14.3; out at T=150 K (using Cryostream open-flow N2 gas cooling devices), on a Rigaku AFC6S 4-circle diVractometer (for 2·Br, C20H11N5S5 (151 stoichiometric complex) requires: C, 49.9; H, 1548 J.Mater. Chem., 1998, 8(7), 1541–1550Table 3 Crystal data Compound 2·Br 14·TCNQ 16 17 (17)2·20 Formula C8H7NS5+Br- C10H11NS7·C12H4N4 C10H11NS5 C11H13NS5 2C11H13NS5·C16H5N5O6 M 357.36 573.81 305.50 319.52 1002.30 Symmetry monoclinic triclinic monoclinic monoclinic monoclinic a/A° 13.126(4) 7.982(1) 13.199(1) 19.055(1) 18.060(1) b/A ° 7.392(2) 12.876(1) 9.534(1) 6.603(1) 13.566(1) c/A ° 13.570(4) 13.465(1) 10.509(1) 11.544(1) 18.697(1) a (°) 90 108.99(1) 90 90 90 b (°) 111.72(2) 104.44(1) 94.98(1) 102.85(1) 109.25(1) c (°) 90 95.81(1) 90 90 90 U/A ° 3 1223.2(7) 1241.7(2) 1317.4(2) 1416.1(2) 4324.7(5) Space group P21/c P19 P21/c P21/c C2/c Z 4 2 4 4 4 m/cm-1 41.8 6.6 8.5 8.0 5.7 Dx/g cm-3 1.94 1.54 1.54 1.50 1.54 Crystal size/mm 0.44×0.09×0.05 0.4×0.09×0.06 0.38×0.3×0.1 0.4×0.28×0.14 0.52×0.04×0.04 2hmax (°) 55 54 50.5 51.3 50 Data total 2927 8847 5302 6102 12792 Data unique 2803 5299 2138 2423 3719 Data obs., I>2s(I) 2110 4408 2006 2225 2792 Rint 0.067/0.025a 0.061/0.035a 0.037 0.079 0.072 Abs.correction empiricalb integration semiempiricalc — — T min, max 0.4751.00 0.8050.97 0.7650.93 — — No.of variables 164 368 166 207 350 wR(F2), all data 0.075 0.095 0.091 0.090 0.125 Goodness of fit 1.07 1.09 1.28 1.11 1.15 R(F), obs. data 0.031 0.039 0.038 0.032 0.054 Drmax, min/e A° -3 0.63, -0.44 0.32, -0.30 0.27, -0.23 0.32, -0.26 0.31, -0.37 aBefore and after the absorption correction.bOn 108 y-scans of 3 reflections, TEXSAN software.51 cOn Laue equivalents, SHELXTL software. dw-1=s2(F2)+(AP)2+BP, where P=(Fo2+2Fc2)/3. 2h/v scan mode) or a Siemens 3-circle SMART diVractometer We thank EPSRC for funding the work in Durham and Bangor, the Leverhulme Trust for a Visiting Fellowship at with a CCD area detector (v scan mode in frames of 0.3°). Graphite-monochromated Mo-Ka radiation (l : =0.71073 A ° ) Durham University (A.S.B.), the Royal Society for a Leverhulme Senior Research Fellowship (J.A.K.H.) and an was used.The structures were solved by direct methods and refined by full-matrix least squares against F2 of all data, using Exchange Fellowship (I.F.P.). SHELXTL software.50 All non-H atoms were refined with anisotropic displacement parameters, all H atoms in isotropic References approximation (methyl groups in 16 as rigid bodies).For (17)2·20, solution in the space group C2/c implies the crystallo- 1 (a) M. R. Bryce, Chem. Soc. Rev., 1991, 20, 355; (b) A. E. Underhill, graphic symmetry (axis 2) and thus the disorder of 20 and J. Mater. Chem., 1992, 2, 1; (c) J. M. Williams, J. R. Ferraro, gives some unreasonably short intermolecular contacts; there- R.J. Thorn, K. D. Carlson, U. Geiser, H. H.Wang, A. M. Kini and M.-H. Whangbo, Organic Superconductors (Including Fullerenes), fore, we attempted also the refinement in space groups Cc and Prentice Hall, New Jersey, 1992; (d) J. J. Novoa, M. C. Rovira, P21/n (making use of some extremely weak reflections with C. Rovira, J. Veciana and J.Tarre�s, Adv. Mater., 1995, 7, 233; odd h+k, due more likely to l/2 contamination), wherein axis (e) J. R. Ferraro and J. M. Williams, Introduction to Synthetic 2 is absent. Neither model led to a consistent refinement. Electrical Conductors, Academic Press, London, 1987, Attempts to rationalise the rotationalisorder of the nitro pp. 268–271. group O(41)N(4)O(42) in terms of ‘split’ positons of the 2 (a) G.R. Desiraju, Crystal Engineering, T he Design of Organic Solids, Elsevier, Amsterdam, 1989; (b) G. R. Desiraju, Chem. oxygen atoms were also unsuccessful. Commun., 1997, 1475. Crystal data and experimental details are listed in Table 3; 3 (a)M. C. Etter, Acc. Chem. Res., 1990, 23, 120; (b) G.M. Whitesides, full crystallographic details, excluding structure factors, have J.P. Mathias and C. T. Seco, Science, 1991, 254, 1312; (c) E. Fan, been deposited at the Cambridge Crystallographic Data C. Vicent, S. J. Geib and A. D. Hamilton, Chem. Mater., 1994, Centre, see Information for Authors, Issue No. 1. Any request 6, 113. to the CCDC for this material should quote the full literature 4 (a) J. Nakayama, N. Toyoda and M.Hoshino, Heterocycles, 1986, 24, 1145; (b) M. Jørgensen and K. Bechgaard, Synthesis, 1989, 207; citation and the reference number 1145/99. (c)M. R. Bryce and G. Cooke, Synthesis, 1991, 263; (d) J. Y. Becker, J. Bernstein, S. Bittner, L. Shahal and S. S. Shaik, J. Chem. Soc., MO Computational procedure Chem. Commun., 1991, 92; (e) C. S. Wang, A. Ellern, V. Khodorkovsky, J. Y. Becker and J.Bernstein, J. Chem. Soc., The geometries of molecules and cation radicals of 2, 14, 16 Chem. Commun., 1994, 983; ( f ) C. S. Wang, J. Y. Becker, and 17 were optimized using the PM3 semiempirical method J. Bernstein, A. Ellern and V. Khodorkovsky, J. Mater. Chem., as implemented in the HyperChem 4.5 program.52 The 1995, 5, 1559; (g) U. Kux, H. Suzuki, S. Sasaki and M. Iyoda, Polak–Ribiere algorithm was used for the optimization; the Chem.L ett., 1995, 183. gradient norm achieved in all the calculations was less than 5 (a) T. Imakubo, H. Sawa and R. Kato, J. Chem. Soc., Chem. Commun., 1995, 1097; (b) T. Imakubo, H. Sawa and R. Kato, Synth. 0.01 kcal mol-1 A ° -1. Restricted Hartree–Fock (RHF) formal- Met., 1995, 73, 117; (c) T. Imakubo, H. Sawa and R.Kato, J .Chem. ism was used for neutral molecules whereas spin-unrestricted Soc., Chem. Commun., 1995, 1667. Hartree–Fock (UHF) formalism was used for radical cations.53 6 (a) R. Gompper, J. Hock, K. Polborn, E. Dormann and H. Winter, Tables of calculated (PM3) parameters for compounds 2, 14, Adv. Mater., 1995, 7, 41; (b) Y. Kuwatani, E. Ogura, H. Nishikawa, 16 and 17 and their cation radicals are available as supplemen- I.Ikemoto and M. Iyoda, Chem. L ett., 1997, 817; (c) M. Iyoda, tary data (SUP 57385; 2 pp.) deposited with the British Library. Y. Kuwatani, K. Hara, E. Ogura, H. Suzuki, H. Ito and T. Mori, Chem., L ett., 1997, 599. Details are available from the editorial oYce. J. Mater. Chem., 1998, 8(7), 1541–1550 15497 A. S. Batsanov, A. J. Moore, N.Robertson, A. Green, M. R. Bryce, 31 (a) C. Katayama, M. Honda, H. Kumagai, J. Tanaka, G. Saito and H. Inokuchi, Bull. Chem. Soc. Jpn., 1985, 58, 2272; J. A. K. Howard and A. E. Underhill, J.Mater. Chem., 1997, 7, 387. (b) D. A. Clemente and A. Marzotto, J. Mater. Chem., 1996, 6, 941; 8 Review: M. R. Bryce, J.Mater. Chem., 1995, 5, 1481. (c) M. Umeya, S. Kawata, H. Matsuzaka, S.Kitagawa, 9 (a) A. S. Batsanov, N. Svenstrup, J. Lau, J. Becher, M. R. Bryce and H. Nishikawa, K. Kikuchi and I. Ikemoto, J. Mater. Chem., 1998, J. A. K. Howard, J. Chem. Soc., Chem. Commun., 1995, 1201; 8, 295 and references therein. (b) O. Neilands, S. Belayakov, V. Tilika and A. Edzina, J. Chem. 32 S. Flandrois and D. Chasseau, Acta Crystallogr., Sect. B, 1977, Soc., Chem.Commun., 1995, 325; (c) A. S. Batsanov, M. R. Bryce, 33, 2744. J. N. Heaton, A. J. Moore, P. J. Skabara, J. A. K. Howard, E. Ort’, 33 R. E. Long, R. A. Sparks and K. N. Trueblood, Acta Crystallogr., P. M. Viruela and R. Viruela, J. Mater. Chem., 1995, 5, 1689; 1965, 18, 932. (d) P. J. Skabara, M. R. Bryce, A. J. Moore, A. S. Batsanov and 34 (a) R. J. J. Visser, H. J. M. Bouwmeester, J.L. de Boer and A. Vos, J. A. K. Howard, T etrahedron, 1997, 53, 17781. Acta Crystallogr., Sect. C, 1990, 46, 852; (b) R. J. J. Visser, 10 (a) A. S. Batsanov, M. R. Bryce, G. Cooke, J. N. Heaton and H. J. M. Bouwmeester, J. L. de Boer and A. Vos, Acta Crystallogr., J. A. K. Howard, J. Chem. Soc., Chem. Commun., 1993, 1701; (b) Sect. C, 860; (c) R. J. J. Visser, J. L. de Boer and A.Vos, Acta A. S. Batsanov, M. R. Bryce, G. Cooke, A. S. Dhindsa, Crystallogr., Sect. C, 1990, 46, 857; (d) R. J. J. Visser, J. L. de Boer J. N. Heaton, J. A. K. Howard, A. J. Moore and M. C. Petty, Chem. and A. Vos, Acta Crystallogr., Sect. C, 1990, 46, 864; Mater., 1994, 6, 1419; (c) J. A. K. Howard, A. S. Batsanov, (e) R. J. J. Visser, J. L. de Boer and A. Vos, Acta Crystallogr., Sect.M. R. Bryce and J. N. Heaton, Acta Crystallogr., Sect. C, 1994, C, 1990, 46, 869; (f ) R. J. J. Visser, S. van Smaalen, J. L. de Boer 50, 1956. and A. Vos, Acta Crystallogr., Sect. C, 1990, 46, 867. 11 Y. Misaki, T. Ohta, N. Higuchi, H. Fujiwara, T. Yamabe, T. Mori, 35 A. Hoekstra, T. Spoelder and A. Vos, Acta Crystallogr., Sect. B, H. Mori and S. Tanaka, J.Mater.Chem., 1995, 5, 1571. 1972, 28, 14. 12 H. Urayama, H. Yamochi, G. Saito, K. Nozawa, T. Sugano, 36 (a) T. J. Kistenmacher, T. E. Phillips and D. O. Cowan, Acta M. Kinoshita, S. Sato, K. Oshima, K. Kawamoto and J. Tanaka, Crystallogr., Sect. B, 1974, 30, 763; (b) R. H. Blessing and Chem. L ett., 1988, 55. P. Coppens, Solid State Commun., 1974, 15, 215; (c) A. J. Schultz, 13 (a) A.M. Kini, U. Geiser, H.-H. Wang, K. D. Carlson, G. D. Stucky, R. H. Blessing and P. Coppens, J. Am. Chem. Soc., J. M. Williams, W. K. Kwok, K. G. Vandervoort, J. E. Thompson, 1976, 98, 3194. D. L. Stupka, D. Jung and M.-H. Whangbo, Inorg. Chem., 1990, 37 (a) R. Comes, S. M. Shapiro, G. Shirane, A. F. Garito and 29, 2555; (b) J. M. Williams, A. M. Kini, H.-H. Wang, A. J. Heeger, Phys.Rev. L ett., 1975, 35, 1518; (b) J. S. Chappell, K. D. Carlson, U. Geiser, L. K. Montgomery, G. J. Pyrka, A. N. Bloch, W. A. Bryden, M. Maxfield, T. O. Poehler and D. M. Watkins, J. M. Kommers, S. J. Boryschuk, A. V. Strieby D. O. Cowan, J. Am. Chem. Soc., 1981, 103, 2442. Crouch, W. K. Kwok, J. E. Schirber, D. L. Overmyer, D. Jung and 38 R. S. Rowland and R. Taylor, J. Phys.Chem., 1996, 100, 7384. M.-H. Whangbo, Inorg. Chem., 1990, 29, 3262. 39 S. C. Nyburg and C. H. Faerman, Acta Crystallogr., Sect. B, 1985, 14 (a) D. C. Green, J. Chem. Soc., Chem. Commun., 1977, 161; 41, 274. (b) D. C. Green, J. Org. Chem., 1979, 44, 1476. 40 F. H. Allen, C. M. Bird, R. S. Rowland and P. R. Raithby, Acta 15 Review: J. Garý�n, Adv. Heterocycl. Chem., 1995, 62, 249.Crystallogr., Sect. B, 1997, 53, 680. 16 For comments on the retention of the packing of neutral TTF 41 (a) I. F. Perepichka, A. F. Popov, T. V. Artyomova, species in their derived salts see: M. C. Rovira, J. J. Novoa, A. N. Vdovichenko, M. R. Bryce, A. S. Batsanov, J. A. K. Howard J. Tarre�s, C. Rovira, J. Veciana, S. Yang, D. O. Cowan and and J. L. Megson, J. Chem. Soc., Perkin T rans. 2, 1995, 3; E. Canadell, Adv.Mater., 1995, 7, 1023. (b) I. F. Perepichka, L. G. Kuz’mina, D. F. Perepichka, 17 Preliminary report, presented at ICSM ’96, Snowbird, Utah: M. R. Bryce, L. M. Goldenberg, A. F. Popov and J. A. K. Howard, A. J. Moore, M. R. Bryce, A. S. Batsanov, C. W. Lehmann and J. Org. Chem., in press; (c) J. Silverman, N. F. Yannoni and J. A. K. Howard, Synth.Met., 1997, 86, 1901.A. P. Krukonis, Acta Crystallogr., Sect. B, 1974, 30, 1474. 18 M. Fourmigue�, F. C. Krebs and J. Larsen, Synthesis, 1993, 509. 42 (a) B. D. Silverman, J. Chem. Phys., 1979, 71, 3572; (b) J. P. Lowe, 19 K. Lerstrup, I. Johannsen and M. Jørgensen, Synth. Met., 1989, J. Am. Chem. Soc., 1980, 102, 1262; (c) T. Mori, A. Kobayashi, 27, B9. Y. Sasaki, H. Kobayashi, G.Saito and H. Inokuchi, Bull. Chem. 20 (a) A. Dolbecq, M. Fourmigue�, P. Batail and C. Coulon, Chem. Soc., Jpn., 1984, 57, 627; (d) J. D. Martin, E. Canadell, J. Y. Becker Mater., 1994, 6, 1413; (b) A. J. Moore, M. R. Bryce, A. S. Batsanov, and J. Bernstein, Chem. Mater., 1993, 5, 1199. J. C. Cole and J. A. K. Howard, Synthesis, 1995, 675. 43 V. Khodorkovsky, unpublished results. 21 M. Jø K. A. Lerstrup and K. Bechgaard, J. Org. Chem., 44 I. Hargittai, J. Brunvoll, M. Kolonits and V. Khodorkovsky, 1991, 56, 5684. J.Mol. Struct., 1994, 317, 273. 22 N. C. Gonnella and M. P. Cava, J. Org. Chem., 1978, 43, 369. 45 C. Wang, M. R. Bryce, A. S. Batsanov and J. A. K. Howard, Chem. 23 (a) L. R. Mora, J.-M. Fabre, L. Giral and C. Montginoul, Bull. Soc. Eur. J., 1997, 3, 1679 and references therein. Chim. Belg., 1992, 101, 137; (b) A. J. Moore and M. R. Bryce, 46 J. P. Stewart, J. Comput. Chem., 1989, 10, 209, 221. Synthesis, 1997, 407. 47 D. A. Smith, C. W. Ulmar II and M. J. Gilbert, J. Comput. Chem., 24 L. Binet, J.-M. Fabre, C. Montginoul, K. B. Simonsen and 1992, 13, 640. J. Becher, J. Chem. Soc., Perkin T rans. 1, 1996, 783. 48 D. L. Lichtenberger, R. L. Johnson, K. Hinkelman, T. Suzuki and 25 For reviews on the synthesis of TTF derivatives see: (a) A. Krief, F.Wudl, J. Am. Chem. Soc., 1990, 112, 3302. T etrahedron, 1986, 42, 1209; (b) G. Schukat, A. M. Richter and 49 Supramolecular Engineering of Synthetic Metallic Materials: E. Fangha�nel, Sulfur Rep., 1987, 7, 155; (c) G. Schukat and Conductors andMagnets, ed. J. Veciana, NATO-ASI Series, Kluwer E. Fangha�nel, Sulfur Rep., 1993, 33, 245. Academic, Dordrecht, 1998, in press. 26 N. Svenstrup, K. M. Rasmussen, T. K. Hansen and J. Becher, 50 G. M. Sheldrick, SHELXTL, Version 5.03, Siemens X-Ray Synthesis, 1994, 809. Analytical Instruments Inc., Madison, WI, 1995. 27 K. B. Simonsen, N. Svenstrup, J. Lau, O. Simonsen, P. Mørk, 51 TEXSAN: Single Crystal Structure Analysis Software, Version 5.0, G. J. Kristensen and J. Becher, Synthesis, 1996, 407. Molecular Structure Corporation, TX, USA, 1989. 28 K. Ikeda, K. Kawabata, K. Tanaka and M. Mizutani, Synth. Met., 52 HyperChem, Release 4.5 for Windows, 1994-95. Molecular 1993, 55–57, 2007. Modelling System, Hypercube Inc., Waterloo, ON, Canada. 29 (a) T. K. Mukherjee and L. A. Levasseur, J. Org. Chem., 1965, 30, 53 G. Klopman and R. Ivens, in Semiempirical Methods of Electron Structure Calculation, ed. G. A. Segal, Plenum Press, New York, 644; (b) J. Silverman, A. P. Krukonis and N. P. Yannoni, Acta 1977, Chapter 2, Section 1.2. Crystallogr., 1967, 23, 1057. 30 A. S. Dhindsa, Y. P. Song, J. P. Badyal, M. R. Bryce, Y. M. Lvov, M. C. Petty and J. Yarwood, Chem.Mater., 1992, 4, 724. Paper 8/02037J; Received 13th March, 1998 1550 J. Mater. Chem., 1998, 8(7), 1541–

ISSN:0959-9428    

DOI:10.1039/a802037j

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

J O U R N A L O F C H E M I S T R Y Materials Reduction of the transition temperatures in mesomorphic lanthanide complexes by the exchange of counter-ions Koen Binnemans,a Yury G. Galyametdinov,b Simon R. Collinsonc and Duncan W. Brucec*† aK.U.L euven, Department of Chemistry, Coordination Chemistry Division, Celestijnenlaan 200F, B-3001 Heverlee (L euven), Belgium bPhysical-T echnical Institute, Russian Academy of Science, Sibirsky T ract 10/7.420029, Kazan, Russia cDepartment of Chemistry, University of Exeter, Stocker Road, Exeter, UK EX4 4QD Metathesis of the chloride anion in the mesomorphic lanthanide complexes [LnL(LH)2][Cl]2 (LH=salicylaldimine; Ln= lanthanide element) by dodecyl sulfate (DOS) leads to the complexes [LnL(LH)2][DOS]2, which have wider mesomorphic ranges and lower clearing points.Introduction The study of metal-containing liquid crystals (metallomesogens) is a flourishing branch of liquid-crystal research, and the work in this area has been reviewed extensively.1 The majority of the calamitic metallomesogens so far synthesised contain metal ions in a d8–d10 configuration (e.g. AgI, CuII, NiII, PdII), thus exhibiting a geometry which is either linear (CN=2) or planar (CN=4), and the core of these compounds can easily satisfy the basic structural requirements for a calamitic mesogen of a highly anisotropic structure.Because complexes C7H15 O N OH O C12H25 1 CnH2 n+1O N 2 CmH2m+1 OH with coordination numbers higher than four do not form planar, and therefore highly anisotropic, structures, then the design of metallomesogens with a coordination number higher and R¾=C12H25).The corresponding transition temperatures than four is a challenge. However, it is necessary to undertake of the gadolinium complex with nitrate counter-ions are: such a challenge because in this way it is possible to take full Cr 98 SA 192 I. advantage of the rich coordination chemistry of a much wider Subsequently, Galyametdinov reported12 the synthesis of range of transition metal ions.mesomorphic lanthanide complexes with non-mesomorphic Progress has been made in this direction, and mesomorphic SchiV base ligands (2), and Gd, Dy and La were used as the iron and vanadyl complexes with a square pyramidal geometry lanthanide ions with nitrate as the counter-ion. For the gadolin- (CN=5) have been reported by Galyametdinov2 and others,3 ium complex (with R=C12H25 and R¾=C18H37) the transition while Malthe�te4 has reported mesomorphic butadiene iron temperatures are Cr 135 SA 146 I.The work indicated that tricarbonyl complexes, and Deschenaux mesomorphic ferronot only the length of the terminal chains or the type of cenes.5 Calamitic metallomesogens with octahedral geometry lanthanide ions had an influence on the transition tempera- have been reported based on MnI and ReI tetracarbonyl tures, but also the type of counter-ion.If in the above men- fragments complexed to imines,6 or based on bromotioned gadolinium complex, the nitrate ion was replaced by a rheniumtricarbonyl fragments complexed to diazabutadienes7 chloride ion, the transition temperatures increased significantly: and bipyridines.8 On the other hand, during the last few years Cr 164 SA 185 I. Magnetic studies were carried out on deriva- a lot of attention has been paid to ionic metallomesogens due tives of both salicylaldimate13 and enaminoketone,14 and very to the new structures which become possible and because of high magnetic anisotropies were found for the dysprosium an interest in the way ionic interactions can influence mesoderivatives.morphism.9 In this respect, lanthanide-containing liquid crys- Since the high transition temperatures of metallomesogens tals are of interest, because they combine a high coordination are a major drawback for future applications, we asked our- number with the tendency to form ionic compounds.Moreover, selves if it would not be possible to decrease the melting points lanthanide ions have interesting spectroscopic and magnetic of these complexes by an alternative choice of the counter-ion. properties. Although liquid-crystalline lanthanide complexes Previous work on silver complexes of stilbazoles showed that forming columnar phases have been studied for more than 10 when the dodecyl sulfate (DOS) anion was incorporated in years,10 calamitic lanthanide mesogens have only been reported place of the BF4, NO3 or CF3SO3 anions, both melting and more recently.In 1991, Galyametdinov11 reported the first clearing points were reduced.15 Therefore, the DOS anion liquid-crystalline lanthanide complex with SchiV base ligands seemed to be a good candidate for obtaining low-melting, with the stoichiometry [LnL¾(L¾H)2]X2, (L¾H=1), where Ln= liquid-crystalline lanthanide complexes.Pr, Eu, Gd, Dy and X=Cl, NO3 (L¾ is a deprotonated ligand). These complexes form SA mesophases, with a high viscosity. The ligand itself is a mesogen: Cr 43 N 71 I (for R=C7H15 Results and Discussion We prepared the ligand (LH) of type 2 (n=12, m=18), which was readily obtained in two steps from 2,4-dihydroxybenzal- †E-mail: d.bruce@exeter.ac.uk J.Mater. Chem., 1998, 8(7), 1551–1553 1551Table 1 Transition temperatures (2nd heating run) and purification. Hydrated rare earth chlorides were purchased phase behaviour of the lanthanide complexes with general from Aldrich. formula [LnL(LH)2][X]2 [where LH=4-(dodecyloxy)-N-octadecyl- 2-hydroxybenzaldimine] 4-(Dodecyloxy)-2-hydroxybenzaldehyde Ln transition/°C (DH/kJ mol-1) Potassium hydrogen carbonate (0.04 mol, 4.00 g) and 2,4- dihydroxybenzaldehyde (0.036 mol, 5.00 g) were dissolved in X=Cl X=DOS acetone (100 cm3) with stirring. 1-Bromododecane (0.04 mol, 9.96 g) was then added to the mixture before heating at reflux Nd Cr 162 (56.6) SA 187 (9.1) I Cr 61 (59.7) SA 98 (9.7) I Gd Cr 164 (58.7) SA 188 (11.6) I Cr 59 (55.9) SA 112 (9.0) I for 48 h.The solution was allowed to cool and was then Dy Cr 166 (55.6) SA 186 (11.9) I Cr 61 (65.6) SA 89 (13.3) I filtered to remove the solid precipitate of potassium bromide Ho Cr 168 (56.0) SA 185 (11.7) I Cr 59 (71.1) SA 81 (12.0) I and the solvent was then evaporated under reduced pressure.Yb Cr 169 (51.8) SA 185 (11.3) I Cr 59 (65.1) SA 111 (14.5) I The residue was added to acidified water (200 cm3) and extracted with diethyl ether (3×100 cm3). The combined organic layer were washed with brine (100 cm3) and dried with anhydrous sodium sulfate, before the diethyl ether was removed dehyde. Reaction7 of three equivalents of LH with YbCl3·6H2O under reduced pressure.The brown oil was purified by column in absolute ethanol led to [YbL(LH)2]Cl2·3H2O. That the chromatography on silica, with hexane5ethyl acetate (951) as complex is a trihydrate was evidenced both by elemental the eluent (total volume: 750 cm3). The yellowish oil obtained analysis and by the observation of a broad, weak endothermic solidified when placed in an ice bath. The crude aldehyde was peak between 80 and 90 °C in the first DSC heating run; all then recrystallised from hot ethanol (95% v/v).The yellow other chloride salts were also found to be trihydrates. The powder obtained was then washed with a little cold ethanol complex exhibits an enantiotropic smectic SA phase: and dry in vacuo; yield 60%. dH (300 MHz, CDCl3): 0.88 (t, Cr 169 SA 185 I [DH (Cr�SA)=51.8 kJ mol-1; DH(SA�I) 3H, CH3); 1.27–1.47 [m, 18H, (CH2)9]; 1.80 (q, 2H, CH2); 4.01 =5.7 kJ mol-1]. The ytterbium complex was then treated with (t, 2H, CH2O); 6.41 (d, 1H, H-aryl); 6.52 (dd, 1H, H-aryl); 7.40 two equivalents of AgDOS which led to a precipitate of AgCl (d, 1H, H-aryl); 9.70 (s, 1H, CHO); 11.45 (s, 1H, OH); Jo= and the formation of the target complex, [YbL(LH)2][DOS]2. 8.7 Hz, Jm=2.3 Hz. This and the other DOS salts were obtained as anhydrous materials which can be explained by the much higher aYnity 4-(Dodecyloxy)-N-octadecyl-2-hydroxybenzaldimine of lanthanide ions for oxygen donors over chloride donors. The corresponding ytterbium(III) complex with the DOS 4-(Dodecyloxy)-2-hydroxybenzaldehyde (2 mmol, 0.60 g) and octadecylamine (2 mmol, 0.54 g) were dissolved in absonion, [YbL(LH)2][DOS]2, also shows an enantiotropic smectic SA phase, but the transition temperatures have ethanol (50 cm3) with a few drops of glacial acetic acid as a catalyst.The solution was then refluxed for 3 hours and after been reduced dramatically: Cr 59 SA 111 I [DH (Cr�SA)= 65.1 kJ mol-1; DH (SA)�I=14.5 kJ mol-1]. A similar behav- cooling a yellow precipitate was obtained.The precipitate was washed with a little cold absolute ethanol, before recrystallis- iour was found for complexes with other lanthanide ions, and the optical and thermal data for the Nd, Gd, Dy, Ho and Yb ation from hot ethanol and the resulting crystals were dried in vacuo; yield 82%. dH (300 MHz, CDCl3): 0.89 (t, 6H, CH3); complexes are summarised in Table 1.Thus, the exchange of the chloride anion by the dodecyl sulfate anion reduces the 1.2–1.8 (m, 52H, CH2); 3.51 (t, 2H, NCH2); 3.96 (t, 2H, OCH2); 6.33 (dd, 1H, H-aryl); 6.37 (d, 1H, H-aryl); 7.05 (d, 1H, H- melting temperature by about 100 °C, and the clearing point to a lesser extent (about 70–90 °C), so that the dodecyl sulfate aryl ); 8.09 (s, 1H, CHNN); 14.1 (s, br, 1H, OH); Jo=8.5 Hz, Jm=2.5 Hz.salts have a wider mesomorphic range than the corresponding chloride salts. Further, while neither the transition tempera- [Bis(4-dodecyloxy)-N-octadecylbenzaldimino-2-hydroxy][4- tures of the chloride salts nor the melting points of the DOS (dodecyloxy)-N-octadecylbenzaldimino-2-olato]ytterbium(III ) salts are significantly influenced by the lanthanide ion (although dichloride trihydrate there is a small increase in the melting point of the chlorides across the series of lanthanides studied), the clearing tempera- 4-(Dodecyloxy)-N-octadecyl-2-hydroxybenzaldimine (1 g, 1.8 tures of the DOS salts vary over a much broader range.The mmol) was dissolved in warm absolute ethanol and to this origin of this eVect is, for the time being, unclear.was added dropwise a solution of YbCl3·6H2O (1.8 mmol; Whereas the transition temperatures of covalent metallome- 0.70 g) in absolute ethanol. The reaction was stirred at room sogens can be tuned by an appropriate choice of the number temperature for 5 hours, before it was placed in an ice bath. and the length of the terminal chains and by substituents on The yellow powder obtained was collected by filtration, washed the rigid core, we have pointed out the fact that the transition with a little cold ethanol and dried in vacuo.Yield: 76% (0.9 g). temperatures of ionic metallomesogens can be changed by a proper choice of the counter-ion. This principle has been [YbL(LH)2][O3SOC12H25]2 illustrated in this case of lanthanide complexes with SchiV base ligands.The complex [YbL(LH)2][Cl]2·3H2O (250 mg, 0.13 mmol) was dissolved in dichloromethane (10 cm3). This solution was added dropwise to a stirred solution of silver(I) dodecyl sulfate Experimental (97 mg; 0.26 mmol) in dichloromethane (10 cm3) and was stirred for 4 hours with the vessel protected from light. A 1H NMR spectra were recorded on a Bru�ker ACF-300 spectrometer (300 MHz).Elemental analyses (CHN) were per- precipitate of silver(I) chloride was formed and removed by filtration of the mixture through a Celite path and washed formed on a Perkin-Elmer 2400 Elemental Analyser. The mesophases were investigated by optical microscopy and by with dichloromethane (2×20 cm3), the solvent was then removed from the combined organic solutions under reduced diVerential scanning calorimetry (DSC).Microscopy was performed on a Olympus BX50 polarising microscope, equipped pressure. The product was then recrystallised from a ethanol5 dichloromethane mixture (1051) and dried in vacuo. The with a Linkam TMS600 hot stage and a Linkam TMS93 temperature controller. DSC thermograms were recorded on absence of significant amounts of chloride and silver ions was proved by X-ray fluorescence spectrometry.Yield: 81% a Perkin-Elmer DSC-7 or on a Mettler-Toledo DSC-821e. XRay fluorescence spectra were run on a Tracor X-ray Spectrace (0.25 g). All other complexes were prepared as above and on a similar 450. All chemicals were used as received, without further 1552 J. Mater. Chem., 1998, 8(7), 1551–1553Table 2 Yields and analytical data for the new complexes 1996; ( g) D.W. Bruce, in Inorganic Materials, ed. D. W. Bruce and D. O’Hare, Wiley, Chichester, 1996, 2nd edn., ch. 8. 2 Yu. G. Galyametdinov, G. I. Ivanova and I. V. Ovchinnokov, Bull. elemental analyses: calc. (found)% Acad. Sci. USSR, Div. Chem. Sci., 1989, 38, 1776. 3 N. Hoshino, A. Kodama, T. Shibuya, Y. Matsunaga and compound yield/% C H N S.Miyajima, Inorg. Chem., 1991, 30, 3091; J. L. Serrano, P. Romero, M. Marcos and P. J. Alonso, J. Chem. Soc., Chem. Nd-Cl (3H2O) 54 68.7 (69.0) 10.7 (10.9) 2.2 (2.1) Gd-Cl (3H2O) 90 68.2 (68.4) 10.6 (10.9) 2.2 (2.1) Commun., 1990. 859. 4 L. Ziminski and J. Malthe�te, J. Chem. Soc., Chem. Commun., 1990, Dy-Cl (3H2O) 75 68.0 (68.2) 10.6 (10.5) 2.1 (2.1) Ho-Cl (3H2O) 88 67.9 (67.9) 10.6 (10.8) 2.1 (2.1) 1495; P.Jacq and J. Malthe�te, L iq. Cryst., 1996, 21, 291. 5 R. Deschenaux and J. W. Goodby, in Ferrocenes: homogeneous Yb-Cl (3H2O) 76 68.1 (68.2) 10.6 (10.8) 2.2 (2.1) Nd-DOS 76 69.2 (69.6) 10.7 (11.1) 1.8 (2.0) catalysis, organic synthesis, materials science, ed. A. Togni and T. Hayashi, VCH,Weinheim, 1995. Gd-DOS 80 68.7 (68.6) 10.7 (10.9) 1.8 (2.0) Dy-DOS 91 68.5 (68.2) 10.7 (11.1) 1.8 (2.0) 6 D.W. Bruce and X.-H. Liu, J. Chem. Soc., Chem. Commun., 1994, 729; L iq. Cryst., 1995, 18, 165; X.-H. Liu, M. N. Abser and D. W. Ho-DOS 68 68.5 (68.5) 10.6 (11.0) 1.8 (1.9) Yb-DOS 81 68.2 (68.2) 10.6 (11.1) 1.8 (2.0) Bruce, J. Organomet. Chem., 1998, 551, 271; X.-H. Liu, I. Manners and D. W. Bruce, J.Mater.Chem., 1998, 8, following paper. 7 S. Morrone, G. Harrison and D. W. Bruce, Adv. Mater., 1995, 7, 665; S. Morrone, D. Guillon and D. W. Bruce, Inorg. Chem., 1996, 35, 7041. 8 K. E. Rowe and D. W. Bruce, J. Chem. Soc., Dalton T rans., 1996, scale. Yields and analytical data for the complexes are given 3913; Mol. Cryst., L iq. Cryst., 1998, in press. in Table 2. 9 F. Neve, Adv. Mater., 1996, 8, 277. 10 C. Piechocki, J. Simon, J. J. Andre�,D. Guillon, P. Petit, A. Skoulios K. B. is Postdoctoral Fellow of the FWO-Flanders (Belgium) and P.Weber, Chem. Phys. L ett., 1985, 122, 124. and thanks the F.W.O.-Flanders for a travel grant and for a 11 Yu. G. Galyametdinov, G. I. Ivanova and I. V. Ovchinnikov, Bull. Acad. Sci. USSR, Div. Chem. Sci., 1991, 40, 1109. ‘Krediet aan Navorsers’.S.R.C. acknowledges support from 12 Yu. G. Galyametdinov, M. A. Athanassopoulou, K. Griesar, the Leverhulme Trust, while D.W.B. and Yu.G.G. acknowledge O. Kharitonova, E. A. Soto Bustamante, L. Tinchurina, support from INTAS (contract: INTAS 96-1198) . We wish to I. Ovchinnikov and W. Haase, Chem. Mater., 1996, 8, 922. thank W. D’Olieslager (K.U. Leuven, Belgium) for measuring 13 Yu. G. Galyametdinov, G. Ivanova, I. V. Ovchinnikov, the X-ray fluorescence spectra. A. Prosvirin, D. Guillon, B. Heinrich, D. A. Dunmur and D. W. Bruce, L iq. Cryst., 1996, 20, 831. 14 I. Bikchantaev, Yu. G. Galyametdinov, O. Kharitonova, References I. Ovchinnikov, D. W. Bruce, D. A. Dunmur, D. Guillon and B. Heinrich, L iq. Cryst., 1996, 20, 489. 1 (a) A. M. Giroud-Godquin and P. M. Maitlis, Angew. Chem., Int. 15 D. W. Bruce, D. A. Dunmur, S. A. Hudson, E. Lalinde, Ed. Engl., 1991, 30, 375; (b) P. Espinet, M. A. Esteruelas, L. A. Oro, P. M. Maitlis, M. P. McDonald, R. Orr, P. Styring, J. L. Serrano and E. Sola, Coord. Chem. Rev., 1992, 117, 215; A. S. Cherodian, R. M. Richardson, J. L. Feijoo and G. Ungar, (c) S. A. Hudson and P. M. Maitlis, Chem. Rev., 1993, 93, 861; Mol. Cryst., L iq. Cryst., 1991, 206, 79. (d) D. W. Bruce, J. Chem. Soc., Dalton T rans., 1993, 2983; (e) A. P. Polishchuk and T. V. Timofeeva, Russ. Chem. Rev., 1993, Paper 8/03227K; Received 29th April, 1998 62, 291; ( f ) Metallomesogens, ed. J. L. Serrano, VCH, Weinheim, J. Mater. Chem., 1998, 8(7), 1551&nda

ISSN:0959-9428    

DOI:10.1039/a803227k

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

J O U R N A L O F C H E M I S T R Y Materials Mesomorphic di- and tetra-fluorinated imines and their complexes with ReI Xiao-Hua Liu,a,b Ian Manners*b and Duncan W. Bruce*a† aDepartment of Chemistry, University of Exeter, Stocker Road, Exeter, UK EX4 4QD bDepartment of Chemistry, University of T oronto, T oronto, M5S 3H6, Canada The synthesis of some di- and tetra-fluoro imine mesogens is described and their mesomorphism is compared with that of the parent, unfluorinated materials.The comparison reveals that in the fluorinated compounds, more ordered smectic phases were suppressed and crystal phases were destabilised; nematic phases were uniformly destabilised. The imines were then reacted with [ReMe(CO)5] to give related orthometallated complexes which were also mesomorphic. fully illustrated by the work of the Hull group with substituted Introduction terphenyls.5 From all of these studies, it is clear that a subtle In low molar mass liquid crystalline systems, fluoro-substi- combination of steric and electronic eVects acts to determine tution has been known since 1925.1 Fluorine has the smallest the resulting mesomorphism.van der Waals’ radius of any substituent other than hydrogen Lateral fluorination has also been studied in low molar mass (F=1.47 A° , H=1.2 A° )2 and it also has highest electronegativity metallomesogens.For example, the central core of stilbazole of any element. Fluorine can, therefore, exert appreciable silver(I) complexes was modified by introduction of lateral electronic eVects and yet, because of its small size, its eVect on fluoro-substituents into complexes with triflate and dodecyl the close packing of the molecules is less than any other sulfate counter ions.6 Depending on the position of the fluoro substituent.The eVect of fluorination in mesogens has been substituent, the mesomorphic behaviour is aVected in a diVerwell documented3 and several general conclusions can be ent way.When the fluoro substituents occupy position 3 of drawn from investigations carried out on calamitic systems. the aromatic rings (Fig. 2) nematic phases are formed less Lateral fluoro substitution in low molar mass systems has frequently or even suppressed totally, melting points are varied eVects and critically depends on the position at which decreased, and smectic C phases are destabilised, resulting in the fluoro substituent is introduced.4 Fig. 1 shows the transition a pronounced widening of the smectic A phases in comparison temperatures for a parent compound and also for two of its with their non-substituted stilbazole silver(I) complexes. mono-fluorinated derivatives. It can be seen that the 2-fluoro Further, the cubic phase seen in the parent complexes is now derivative (fluorine positioned centrally) has reduced nematic absent.However, when the fluoro substituents occupy position character compared with the parent system, while when the 2, diVerent eVects are observed; the clearing points are signifi- fluoro substituent is positioned near the end of core unit (3- cantly depressed and melting points are also decreased, but to fluoro derivative), nematic character has been eliminated and a lesser extent.The smectic C phase is stabilised at the total smectic phase stability has been enhanced. Both of the fluori- expense of the smectic A phase, the nematic phase is promoted nated compounds have a lower melting point than the parent and for X=C12H25OSO3, cubic phases are retained. Imrie and system.coworkers7 have reported a difluoro-substituted ferrocene metallomesogen (Fig. 3), and here, fluoro substitution led to lower temperature mesogens. Fluorination has also been studied in, for example, complexes of salicylaldimates,8 salen,9 dithiobenzoates10 and b-diketonates.11 As part of our studies on calamitic metallomesogens containing octahedral metal centres,12 we have studied orthometallated O X O Y C6H13O C5H11 Crys • 50 • N • 63 • I Crys • 28 • N • 37 • I Crys • 40 • SA • 49 • I X H F H Y H H F Fig. 1 Example of the dependence of transition temperatures on position of fluorination Predicting the eVect which lateral fluoro-substitution will have is, however, diYcult because steric factors must be N OCnH2 n + 1 N CnH2 n + 1O Ag F F X– + X = OTf, C12H25OSO3 2 2 3 3 considered in addition to molecular polarizability.3b,c In most Fig. 2 Laterally fluorinated stilbazole complexes of silver(I) systems, however, the trends for laterally fluoro-substituted compounds show similar patterns, depending on their number and distribution. Positioned so that they contribute to an ‘outboard’ dipole, they have the possibility to stabilise smectic phases perhaps at the expense of nematic phases, while positioned to create a net lateral dipole, they can reduce clearing points and promote nematic phases, often at the expense of smectic phases (Fig. 1).Some of these eVects are rather beauti- Fe Crys • 115 • N • 129 • I O O OC8 H17 F F Fig. 3 Mesomorphic difluoro-substituted ferrocene †E-mail: d.bruce@exeter.ac.uk J.Mater. Chem., 1998, 8(7), 1555–1560 1555manganese and rhenium complexes of extended imines.13 In isolated by adding aqueous sodium hydrogen carbonate, acidthese systems we typically found that a four-ring imine (e.g. ifying with cold, concentrated hydrochloric acid, and then 10) with a clearing point of around 300 °C and more than one extracting with chloroform. After crystallisation from chlorosmectic phase, could be reacted with [MMe(CO)5] (M=Mn, form, compound 2 (4-alkoxy-2,3-difluorobenzoic acid) was Re) to give the orthometallated, tetracarbonyl derivative (e.g.obtained as colourless crystals in ca. 85% yield. Compound 2 13) of the ligand which would clear at more than 100 °C lower was then esterified with 4-hydroxybenzaldehyde using than the free ligand and would show only a nematic phase.DCC–DMAP to give the aldehyde, 3, which was then con- We were then able to show13c that by using shorter, three-ring densed with various anilines. Crystallisation of the resulting imines it was possible to realise materials with (monotropic) imines from dichloromethane–methanol, and then from toluclearing points as low as about 70 °C.We were therefore keen ene, gave the pure difluorinated imines, 4, which were reacted to look at other strategies which might lead to lower melting with [ReMe(CO)5] in toluene at reflux. The resulting rhepoints and, therefore, we undertook a study of fluorinated nium(I ) complexes (5) were purified by flash column chromaderivatives of the ligands and their metal complexes. Three, tography (neutral alumina), eluting with 20% hexanes–CH2Cl2 four-ringed imines were chosen and both di- and tetra-fluori- (v/v) to obtain the desired product in ca. 80% yield after nated derivatives, and their complexes with rhenium, were crystallisation from dichloromethane and hexanes. obtained. The synthesis of the tetrafluoro-substituted rhenium complexes, 9, (Scheme 2) required 4¾-(4-alkoxy-2,3-difluorobenzoyloxy) aniline derivatives in addition to the difluoro- Results and Discussion benzaldehydes 3.Thus, 4-alkoxy-2,3-difluorobenzoic acid was Synthesis esterified with 4-hydroxynitrobenzene using DCC–DMAP to give the 4-alkoxy-2,3-difluorobenzoyloxy-4¾-nitrobenzene 6 Difluoroimines were synthesized in a four-step reaction which was then reduced to 4-alkoxy-2,3-difluorobenzoyloxy- (Scheme 1).The unsymmetrical ether 1 was first obtained in a 4¾-aniline derivatives using tin(II) chloride. The 4-alkoxy-2,3- Williamson ether synthesis using 2,3-difluorophenol and the difluorobenzoyloxyaniline derivatives 7 were purified by flash related 1-bromoalkane in dimethylformamide. Lithiation was chromatography (silica gel ), eluting with a solution of 1% then eVected with an equimolar amount of butyllithium at triethylamine in 10% MeOH–CH2Cl2 (v5v5v). A brownish -78 °C, followed by quenching with solid carbon dioxide to product, 7, was obtained in ca. 70% yield, which was then give a single product as detected by chromatographic (TLC) analysis of the crude product mixture. The crude product was condensed with the difluorinated aldehyde derivatives 3 and CnH2 n + 1O CnH2 n + 1O CO2 H HO OCmH2 m + 1 F F CnH2 n + 1O F F O O CHO O O H2N F F F F CnH2 n + 1O F F O O OCmH2 m + 1 O O N CnH2 n + 1O F F O O OCmH2 m + 1 O O N Re CO OC CO CO + 1 5a m = 8, n = 8 5b m = 8, n = 12 5c m = 12, n = 12 2 3 i ii iii iv 4a m = 8, n = 8 4b m = 8, n = 12 4c m = 12, n = 12 v Scheme 1 Synthetic route to difluoro-substituted rhenium complexes 5.Reagents and conditions: i, Cn H2n+1Br, K2CO3, DMF; ii, BuLi, CO2(s), THF; iii, DCC, DMAP, DCM; iv, toluene–acetic acid; v, [ReMe(CO)5]–toluene. 1556 J. Mater. Chem., 1998, 8(7), 1555–1560OCmH2 m + 1 CnH2 n + 1O F F O O CHO O O H2N CnH2 n + 1O F F O O OCmH2 m + 1 O O N CnH2 n + 1O F F O O OCmH2 m + 1 O O N Re CO OC CO CO F F F F F F OCmH2 m + 1 O O O2N F F OCmH2 m + 1 O HO F F + 9a m = 8, n = 8 9b m = 8, n = 12 9c m = 12, n = 12 iii 8a m = 8, n = 8 8b m = 8, n = 12 8c m = 12, n = 12 iv 7 6 i ii 2 3 Scheme 2 Synthetic route to tetrafluoro-substituted rhenium complexes 9.Reagents and conditions: i, DCC, DMAP, DCM, 4-nitrophenol; ii, SnCl2·2H2O, ethanol; iii, toluene–acetic acid; iv, [ReMe(CO)5]–toluene. crystallised from CH2Cl2–hexanes and then toluene to give towards the end of the molecules, smectic phases are being stabilised in the shorter chain compound, while nematic phases compound 8 as colourless crystals.The tetrafluorinated imines 8 were reacted with [ReMe(CO)5] in toluene and the product are destabilised in the longer chain compounds. It is also of interest that all of the more highly ordered smectic phases purified by flash column chromatography (neutral alumina), eluting with 10% hexane–CH2Cl2 (v/v) to give the desired have been suppressed.Further, if melting is taken to be the formation of a fluid smectic phase, then melting points are complexes 9 as yellow crystals in ca. 80% yield following crystallisation from CH2Cl2–hexane. systematically lowered in the difluoro compounds and raised in the tetrafluoro compounds.The former may be explained by the unsymmetric nature of the imine, while the latter reflects Ligand mesomorphism increased (and symmetric) intermolecular dipolar associations. Table 1 shows the transition temperatures for the parent, unsubstituted imine ligands 10, 11 and 12,14 and those of their Mesomorphism of the complexes di- and tetra-fluorinated analogues, 4 and 8.Thus, 11 and 12 exhibit crystal J, smectic I, C and nematic phases, while 10 The mesomorphism behaviour of the complexes is collected in Table 2 along with the mesomorphism of the parent unfluori- exhibits crystal G, smectic C and nematic phases. The mesomorphism of these parent ligands and their fluorinated ana- nated complexes; the behaviour is also illustrated graphically in Fig. 6. logues is collected in Fig. 4. The mesomorphism of the di- and tetra-fluorinated deriva- Reaction of the parent imines with [ReMe(CO)5] led to complexes whose melting points (taken as the transition into tives was rather similar in that all showed both a smectic C and a nematic phase, with none of the more ordered smectic a fluid phase) increased by 20–30 °C and whose nematic phase stability decreased by about 120 °C; all other phases were phases being observed.The smectic C phase was characterised by its schlieren texture and the observation of transition bars suppressed. In the case of the fluorinated ligands, complexation again led to the suppression of the SC phase and the destabilis- on its transition to the nematic phase.The largest smectic C ranges were consistently observed for the difluoro imines, ation of the nematic phase. This destabilisation was quite constant in the case of the tetrafluoro systems at 127±1 °C, reaching nearly 150 °C in 4c, but this range was obtained by a destabilisation of the crystal phase. The highest thermal while it varied much more for the difluoro systems.In fact, the drop in TNI on going from 4c to 5c is undetermined (except stability for the SC phase is with the tetrafluoro derivatives, while the highest nematic phase stability was found in the that it is at least 118 °C) as the complex simply melted at 141 °C with no sign of a monotropic phase before crystallisation unsubstituted imines; the reduction in nematic range on increased fluoro substitution is nicely illustrated in Fig. 5. on cooling. Another interesting point is that the stabilisation of the crystal phase on complexation shows a strong depen- Clearly then, because of the position of the fluoro groups J. Mater. Chem., 1998, 8(7), 1555–1560 1557Table 1 Thermal behaviour for the fluoro-substituted imines and their parent compounds CnH2 n + 1O X X O O OCmH2 m + 1 O O N Y Y DH/kJ DS/JK-1 Compound X Y n m transition T/°C mol-1 mol-1 10 H H 8 8 Crys–Crys’ 63 24.7 73 Crys’–G 116 25.2 65 G–SC 124 3.8 10 SC–N 202 1.9 4 N–I 298 2.2 4 4a F H 8 8 Crys–SC 97 27.5 74 SC–N 213 2.6 5 Fig. 5 Plot to show the decrease in the nematic range with increasing N–I 299 1.9 3 fluorine substitution 8a F F 8 8 Crys–SC 127 33.2 83 SC–N 222 4.0 8 Table 2 Thermal behaviour for the rhenium complexes of the fluoro- N–I 288 1.1 2 substituted imines and their parent complexes 11 H H 12 8 Crys–J 91 13.5 38 J–SI 110 0.4 1 SI–SC 118 0.8 2 SC–N 224 1.4 3 N–I 284 1.0 1 4b F H 12 8 Crys–SC 96 32.4 88 SC–N 216 2.1 4 CnH2 n + 1O X X O O OCmH2 m + 1 O O N Re CO OC CO CO Y Y N–I 270 1.7 3 8b F F 12 Crys–SC 120 32.2 82 DH/ DS/J K-1 SC–N 226 4.5 9 Compound X Y n m transition T/°C kJ mol-1 mol-1 N–I 264 1.4 3 12 H H 12 12 Crys–J 99 32.6 88 13 H H 8 8 Crys–Crys’ 135 6.8 17 J–SI 105 0.1 >1 Crys’–N 154 32.2 75 SI–SC 113 1.0 3 N–I 176 1.2 3 SC–N 234 3.3 7 5a F H 8 8 Crys–Crys’ 48 17.1 53 N–I 269 1.8 3 Crys’–N 153 51.1 120 4c F H 12 12 Crys–SC 86 31.0 86 N–I 164 2.6 6 SC–N 234 6.2 12 9a F F 8 8 Crys–N 137 52.1 127 N–I 259 2.4 5 N–I 160 1.6 4 8c F F 12 12 Crys–SC 120 32.6 83 14 H H 12 8 Crys–N 130 48.0 118 SC–N 231 5.2 10 N–I 164 0.9 2 N–I 252 1.2 2 5b F H 12 8 Crys–Crys’ 123 18.5 47 Crys’–N 143 47.0 113 N–I 149 1 2 9b F F 12 8 Crys–N 125 47.6 120 N–I 138 0.7 2 15 H H 12 12 Crys–Crys’ 96 22.8 62 Crys’–N 131 53.4 132 N–I 145 0.9 2 5c F H 12 12 Crys–Crys’ 95 33.2 90 Crys’–I 141 30.6 74 9c F F 12 12 Crys–Crys’ 61 19.1 57 Crys’–I 131 43.2 107 (N–I) (125) 0.5 1 Fig. 4 Schematic representation of the mesomorphism of the non- fluorinated, di- and tetra-fluorinated imines dence on the degree of fluorination, with difluorinated systems being stabilised by between 47 and 55 °C, while tetrafluorinated systems are stabilised by only 5–11 °C. We have no simple explanation for this phenomenon. However, the net eVect is to produce complexes with shorter mesomorphic ranges due to destabilisation of the mesophase and stabilisation of the crystal phase.This behaviour is true for all of the systems examined and indeed, it would appear that the chain length has a larger influence. Fig. 6 Schematic representation of the mesomorphism in rhenium(I) complexes of non-fluorinated, di- and tetra-fluorinated imines These results are perhaps surprising given the more beneficial 1558 J.Mater. Chem., 1998, 8(7), 1555–1560eVects observed by Imrie and coworkers7 with fluorinated ated under vacuum to give a yellow oil. Distillation under vacuum at 69 °C (5×10-3 mm Hg) aVorded a colourless oil. ferrocenes, although it may just be that there is a more advantageous fluorine substitution pattern in these imines Yield 92%, 1H NMR (200 MHz, CDCl3), d 0.90 (t, 3H, CH3), 1.31–1.55 (m, 10H, 5CH2), 1.80 (m, 2H, OCH2CH2), which will oVer the chance of wider ranges and at lower temperatures.This awaits further study. 4.03 (t, 2H, J=6.6 Hz, OCH2), 6.75 (m, 2H, aromatic) and 6.97 (m, 1H, aromatic). MS: m/z 242 (M+). Experimental 2,3-Difluoro-4-octyloxybenzoic acid 2. A solution of the 2,3- difluoro-4-octyloxybenzene 1 (3.0 g, 12.4 mmol) in dry THF Characterisation (100 cm3) was blanketed with nitrogen and then treated drop- Microanalyses were performed at Department of Chemistry, wise at -78 °C with butyllithium in hexane (7.75 cm3 in 1.6 M University of Exeter and Quantitative Technologies Inc., hexanes).When the addition was complete, the mixture was Whitehouse, NJ, Micro-analytical Service. Mass spectra were stirred for ca. 5 h at the same temperature and then poured obtained with the use of a VG 70-250S mass spectrometer into ca. 100 g of crushed, solid carbon dioxide. After 15 h the operating in electron impact (EI) mode. All chemicals were residue was treated with 10% aqueous sodium bicarbonate used as received unless otherwise specified.IR spectra were and then with diethyl ether. The aqueous layer was separated, recorded on a Nicolet Magna 550 IR spectrometer at washed with diethyl ether, acidified with cold concentrated University of Toronto. Spectra were recorded as Nujol mulls hydrochloric acid, and extracted with chloroform. The combetween polythene plates, or as CH2Cl2 solutions, or KBr bined chloroform extracts were dried over anhydrous Na2SO4, disks. 1H NMR spectra were recored on a Varian Gemini 200 and TLC analysis of the extract showed the presence of one spectrometer; chemical shifts are quoted relative to an internal significant product. The solvent was removed in vacuo to give duterium lock. 19F NMR spectra were recorded on a Varian a white residue, which was purified by crystallisation from Gemini 300 spectrometer and CFCl3 was used as reference.chloroform. The melting points and phase transitions of liquid crystal Yield, 2.99 g (84.6%), 1H NMR (200 MHz, CDCl3), d 0.90 materials were observed by diVerential scanning calorimetry (t, 3H, CH3), 1.25–1.65 (m, 10H, 5CH2), 1.86 (m, 2H, (DSC) using a Perkin-Elmer DSC7 system with a Perkin- OCH2CH2), 4.09 (t, 2H, OCH2), 6.8 (m, 1H, aromatic) and 7.8 Elmer 7000 data station using TAS 7 software.Visual charac- (m, 1H, aromatic), 9.0 (br s, CO2H). 19F NMR (300 MHz, terisation of the liquid-crystalline properties of the bulk mate- CD2Cl2), d -132.93 (dd), -158.65 (dd). MS: m/z 286 (M+) rials was performed using a polarising optical microscope (HRMS: found:M+, 286.1381. C15H20F2O3 requires 286.1368).equipped with a Mettler FP82 hot stage and FP80 central processor at University of Toronto. 4¾-(4-Alkoxy-2,3-difluorobenzoyloxy)benzaldehyde 3. All Toluene and tetrahydrofuran were distilled over sodium/ esters were obtained using this general procedure. benzophenone under nitrogen immediately before use. Dicyclohexyl carbodiimide (5.94 g, 28.8 mmol) and 4-(N,NDichloromethane was dried over molecular sieves (4A) for at dimethylamino)pyridine (0.15 g) were added to a stirred soluleast 24 h before use. All other solvents were used as received tion of the relevant acid (24 mmol) and hydroxybenzaldehyde without further purification.(3.0 g, 24.0 mmol) in dry dichloromethane (100 cm3). The reaction mixture was stirred at room temperature for 6 h.The Synthesis precipitated dicyclohexyl urea was removed by filtration and the residue was reduced to dryness on a rotary evaporator. In all cases, the synthesis for one example is given along with The crude product was purified by column chromatography the relevant spectroscopic data. All other homologues were (silica gel, CH2Cl2) to give the product as a colourless solid. the same with analogous spectra.Analytical data and yields n=8; yield: 67.3%, 1H NMR (200 MHz, CDCl3), d 0.90 (t, are collected in Table 3. 3H, CH3), 1.25–1.65 (m, 10H, 5CH2), 1.86 (m, 2H, OCH2CH2), 4.09 (t, 2H, OCH2), 6.8 (m, 1H, aromatic), 7.45 (d, 2H, AA¾XX¾), 2,3-Difluoro-4-octyloxybenzene 1. A mixture of 4-hydroxy- 7.87 (dd, 1H, aromatic), 8.0 (dd, 2H, AA¾XX¾), 10.0 (s, 1H, 2,3-difluorobenzene (5.0 g, 38.4 mmol), 1-bromooctane (8.9 g, CHO). 19F NMR (300 MHz, CD2Cl2), d -132.14 (dd), 46.1 mmol) and potassium carbonate (14.0 g, 96 mmol) in -157,64 (dd). MS: m/z 391 (MH+), 269, 157(100%) (HRMS: DMF (130 cm3) was refluxed for 6 h. After cooling to room found: MH+, 391.1721. C22H24F2O4 requires M, 391.1711). temperature, water (100 cm3) was added in the flask to dissolve salt and then the product was extracted with diethyl ether 4¾-(4-Alkyloxy-2,3-difluorobenzoyloxy)nitrobenzene 6.For three times. The extracts were washed with water and then the compounds 6, the synthesis was performed in an analogous dried over CaCl2 overnight. After filtration, ether was evapormanner as 3 above. m=8; yield: 74.2%, 1H NMR (200 MHz, CDCl3), d 0.90 (t, Table 3 Analytical data for the ligands and complexes 3H, CH3), 1.25–1.65 (m, 10H, 5CH2), 1.86 (m, 2H, OCH2CH2), calc.(found) 4.15 (t, 2H, OCH2), 6.86 (dd, 1H, aromatic), 7.43 (d, 2H, AA¾XX¾), 7.87 (dd, 1H, aromatic), 8.34 (dd, 2H, AA¾XX¾). compound yield (%) C H N 4¾-(4-Alkyloxy-2,3-difluorobenzoyloxy)aniline 7. All anilines ligands were obtained from the parent nitro compound according to 4a 83 72.0 (72.4) 6.9 (6.9) 1.9 (2.0) 4b 75 73.3 (73.2) 7.5 (7.3) 1.8 (1.8) the following general method. 4c 78 74.1 (73.8) 7.9 (7.8) 1.7 (1.7) A mixture of the nitrobenzene (13 mmol) and SnCl2·2H2O 8a 83 68.9 (68.7) 6.3 (6.2) 1.9 (1.8) (15.2 g, 65 mmol) was heated at reflux in ethanol (100 cm3) for 8b 81 70.0 (70.1) 7.0 (7.3) 1.9 (1.8) 6 h. After cooling, the mixture was poured into ice and the pH 8c 76 71.1 (70.9) 7.4 (7.2) 1.6 (1.6) was adjusted to ca. 7–8 using sodium hydroxide. The mixture complexes was then extracted with ethyl acetate. The ethyl acetate solu- 5a 77 55.8 (55.6) 4.8 (4.9) 1.4 (1.4) 5b 82 57.4 (57.1) 5.3 (5.1) 1.3 (1.3) tion was washed three times with brine and was then dried 5c 78 58.8 (58.5) 5.7 (5.6) 1.3 (1.2) over anhydrous MgSO4.The solvent was evaporated under 9a 83 53.9 (53.6) 4.4 (4.3) 1.3 (1.3) reduced pressure. The brownish solid was purified by column 9b 72 55.5 (55.2) 4.9 (5.1) 1.3 (1.3) chromatography (silica gel: 1% Net3–CH2Cl2, v/v) and then 9c 75 57.0 (56.6) 5.4 (5.3) 1.2 (1.2) crystallised from ethanol to give an oV-white solid. J. Mater. Chem., 1998, 8(7), 1555–1560 1559n=8; yield: 71%, 1H NMR (200 MHz, CDCl3), d 0.90 (t, 8.5 (s, 1H, CHNN). 19F NMR (300 MHz, CD2Cl2), d -132.30 (dd, 1F, J=19.6, 7.9 Hz), -132.47 (dd, 1F, J=19.7, 7.1 Hz), 3H, CH3), 1.25–1.65 (m, 10H, 5CH2), 1.86 (m, 2H, OCH2CH2), 3.35 (br s, 2H, NH2), 4.1 (t, 2H, OCH2), 6.7 (dd, 2H, AA¾XX¾), -157.73 (dd, 1F, J=18.7, 8.1 Hz), -157.85 (dd, 1F, J= 19.3, 8.5 Hz). 6.8 (dd, 1H, aromatic), 7.0 (dd, 2H, AA’XX’), 7.82 (dd, 1H, aromatic). 19F NMR (300 MHz, CD2Cl2), d -132.77 (dd), Tetrafluoro-substituted rhenium(I ) complexes 9. Tetrafluoro- -158.03 (dd). MS: m/z 377 (M+), 269, 157 (100%), 108, 57 substituted rhenium(I ) complexes 9 were obtained in a manner (HRMS: found: M+, 377.1803. C21H25F2O3 requires M, analogous to that used in complexes 5. 377.1809). 9a: IR (CH2 Cl2 solution) nmax/cm-1: 2093w (CO), 1992vs (CO), 1934m (CO) and 1735m (CNO), 1623w (CNN). 1H Difluoro-substituted imines 4. All imines were obtained using NMR (200 MHz, CD2Cl2), d 0.9 (t, 6H, 2CH3, overlapped), the following general procedure. 1.25–1.65 (m, 20H, 10CH2), 1.86 (m, 4H, 2OCH2CH2), 4.15 (t, The relevant aniline (19.2 mmol) was dissolved in toluene 4H, J=6.6 Hz, OCH2), 6.9 (ddd, 2H, aromatic, overlapped), (25 cm3) and then acetic acid (2 drops) was added to the 7.08 (dd, 1H, J=8.2 Hz, 2.4), 7.36 (m, 4H, AA¾XX¾), 7.8–7.93 solution.The relevant aldehyde (19.2 mmol) was added to the (m, 4H, 2AMX+2aromatic), 8.65 (s, 1H, CHNN). 19F NMR solution which was stirred for a few minutes then left unstirred (300 MHz, CD2Cl2), d -132.26 (dd, 1F, J=19.3, 8.1 Hz), overnight. The crude product was filtered and crystallised from -132.44 (dd, 1F, J=19.3, 8.1 Hz), -157.72 (dd, 1F, J=19.2, CH2Cl2–MeOH and then from toluene, to give a colourless, 6.6 Hz), -157.83 (dd, 1F, J=19.3, 7.1 Hz). crystalline product. 4a: 1H NMR (200 MHz, CDCl3), d 0.88 (t, 6H, 2CH3), We thank the University of SheYeld and NSERC for funding. 1.25–1.65 (m, 20H, 10CH2), 1.86 (m, 4H, 2OCH2CH2), 4.05 (t, 2H, J=6.6 Hz, OCH2), 4.14 (t, 2H, J=6.5 Hz, OCH2), 6.81 (ddd, 1H, aromatic), 7.0 (dd, 2H, AA¾XX¾), 7.2–7.4 (m, 6H, References AA¾XX¾), 7.82 (ddd, 1H, aromatic), 8.0 (d, 2H, AA¾XX¾), 8.17 1 V.Vill, L iqCryst 3.0, LCI Publishers, Hamburg. (d, 2H, AA¾XX¾), 8.5 (s, 1H, CHNN). 19F NMR (300 MHz, 2 L. D. Field and S. Sternhell, J. Am. Chem. Soc., 1981, 103, 738. CD2Cl2), d -132.30 (dd, 1F, J=19.7, 8.2 Hz), -157.74 (dd, 3 (a) M.A. Osman, Mol. Cryst. L iq. Cryst., 1985, 128, 45; (b) 1F, J=18.7, 8.1 Hz). G. W. Gray, in Molecular Structure and the Properties of L iquid Crystals, Academic Press, London, 1962; (c) G. W. Gray, in L iquid Difluoro-substituted rhenium(I ) complexes 5. All difluoro- Crystals and Plastic Crystals, ed. G. W. Gray and P. A. Winsor, substituted rhenium(I) complexes 5 were synthesised in an Ellis Horwood Limited, Chichester, 1974, vol. 4; (d) K. J. Toyne, in T hermotropic L iquid Crystals, ed. G. W. Gray, Wiley, Chichester, analogous manner using the following method. 1987, pp. 28–63. Ligand 4a (0.15 g, 0.21 mmol) and (pentacarbonylmethyl)- 4 (a) S. M. Kelly, Helv. Chim. Acta, 1989, 72, 594; (b) A. J. Seed, rhenium (0.07 g, 0.21 mmol) were dissolved in dry toluene and K.J. Toyne and J. W. Goodby, J.Mater. Chem., 1995, 5, 2201. heated at reflux for 12–16 h under nitrogen. Solvent was then 5 G. W. Gray, M. Hird and K. J. Toyne, Mol. Cryst., L iq. Cryst., removed in vacuo and passed through a column of neutral 1991, 204, 43. alumina eluting with hexanes–CH2Cl2 (1:4, v/v) to obtain the 6 D. W. Bruce and S.A. Hudson, J.Mater. Chem., 1994, 4, 479. 7 C. Loubser, C. Imrie and P. H. v. Rooyen, Adv. Mater., 1993, 5, 45. desired rhenium(I) complex fraction. Following crystallisation 8 E. Bui, J. P. Bayle, F. Perez, L. Liebert and J. Courtieu, L iq. Cryst., from CH2Cl2–MeOH, a yellow crystalline product was 1991, 8, 5213. obtained. 9 A. B. Blake, J. R. Chipperfield, W. Hussain, R. Paschke and 5a; yield 76.7%, IR (CH2Cl2 solution) n/cm-1: 2093w (CO), E. Sinn, Inorg.Chem., 1995, 34, 1125. 1992vs (CO), 1987s (CO), 1934m (CO) and 1731(CNO), 1606 10 D. W. Bruce, R. Dhillon, D. A. Dunmur and P. M. Maitlis, (CNN). 1H NMR (200 MHz, CD2Cl2), d 0.88 (t, 6H, 2CH3), J.Mater. Chem., 1992, 2, 65. 11 N. J. Thompson, G. W. Gray, J. W. Goodby and K. J. Toyne,Mol. 1.25–1.65 (m, 20H, 10CH2), 1.86 (m, 4H, 2OCH2CH2), 4.05 (t, Cryst., L iq. Cryst., 1991, 200, 109. 2H, J=6.6 Hz, OCH2), 4.14 (t, 2H, J=6.5 Hz, OCH2), 6.9 12 D. W. Bruce, Adv. Mater., 1994, 6, 699; J. P. Rourke, D. W. Bruce (ddd, 1H, aromatic), 7.05 (m, 3H), 7.4 (m, 4H, AA¾XX¾), 7.82 and T. B. Marder, J. Chem. Soc., Dalton T rans., 1995, 317; (dd, 2H), 7.9 (d, 1H, aromatic), 8.15 (d, 2H, AA¾XX¾), 8.67 (s, S. Morrone, G. Harrison and D. W. Bruce, Adv. Mater., 1995, 7, 1H, CHNN). 19F NMR (300 MHz, CD2Cl2), d -132.43 (dd, 665; K. E. Rowe and D. W. Bruce, J. Chem. Soc., Dalton T rans., 1F, J=19.3, 7.8 Hz), -157.82 (dd, 1F, J=19.2, 6.6 Hz). 1996, 3913; S. Morrone, D. Guillon and D. W. Bruce, Inorg. Chem., 1996, 35, 7041. 13 (a) D. W. Bruce and X.-H. Liu, J. Chem. Soc., Chem. Commun., Tetrafluoro substituted imine ligands 8. The imine ligands 8 1994, 729; (b) D.W. Bruce and X.-H. Liu, L iq. Cryst., 1995, 18, 165; were obtained in an analogous manner as those of ligands 4. (c) X.-H. Liu, M. N. Abser and D. W. Bruce, J. Organomet. Chem., 8a: 1H NMR (200 MHz, CDCl3), d 0.88 (t, 6H, 2CH3), 1998, 551, 271. 1.25–1.65 (m, 20H, 10CH2), 1.86 (m, 4H, 2OCH2CH2), 4.15 (t, 14 X.-H. Liu, PhD Thesis, University of SheYeld, 1997. 4H, J=6.5 Hz, 2OCH2), 6.81 (ddd, 2H, aromatic), 7.2–7.4 (m, 6H, AA¾XX¾), 7.87 (ddd, 2H, aromatic), 8.0 (d, 2H, AA¾XX¾), Paper 8/02167H; Received 19thMarch, 1998 1560 J. Mater. Chem., 1998, 8(7), 1555–1560

ISSN:0959-9428    

DOI:10.1039/a802167h

出版商:RSC    

年代:1998

数据来源: RSC

摘要:

J O U R N A L O F C H E M I S T R Y Materials Mesogenic palladium and platinum-diiodide complexes of 4-isocyanobenzylidene-4-alkoxyphenylimines Shuangxi Wang,a Andreas Mayr*a,b† and Kung-Kai Cheunga aDepartment of Chemistry, T he University of Hong Kong, Pokfulam Road, Hong Kong bDepartment of Chemistry, State University of New York at Stony Brook, Stony Brook, New York 11794-3400, USA A series of 4-isocyanobenzylidene-4-alkoxyphenylimines 4-CN-3,5-R2C6H2CHNC6H4-4-OCnH2n+1 (R=Me, Pri), were prepared from 4-isocyanobenzaldehydes and long-chain 4-alkoxyanilines.The palladium and platinum-diiodide complexes of the methylsubstituted ligands, MI2(4-CN-3,5-Me2C6H2CHNC6H4-4-OCnH2n+1)2 exhibit mesomorphic behaviour, forming nematic and smectic C phases (for n>10). The metal centers apparently serve as rigid linear links between the isocyanide ligands, but have little additional influence on the liquid crystal properties, as evidenced by the small diVerences in transition temperatures between the palladium and platinum compounds.The corresponding complexes with lateral isopropyl substituents (R=Pri) do not exhibit mesomorphic properties. On account of their linear shapes and strong binding properties, isocyanides and acetylides are suitable ligands for the formation of metal complexes with mesomorphic properties.1 This has been demonstrated for lyotropic polymeric acetylide palladium complexes more than 15 years ago.2–5 In subsequent studies, a variety of thermotropic liquid crystals based on molecular acetylide complexes of AuI, PdII, PtII and RhIII have been prepared.6–11 The development of liquid crystals based on corresponding isocyanide metal complexes has been initiated more recently, and since the first reports by Takahashi and co-workers12–14 and by Espinet and co-workers,15–17 only a few additional studies have been described.18,19 The relatively small number of investigations on liquid crystals based on isocyanide metal complexes has been suggested to be a possible consequence of the limited commercial availability of isocyanides as well as their low thermal stability and reputation for being dangerous substances.15 Considering their good coordinating properties and favorable molecular shapes, the development of new isocyanides for the synthesis of liquid crystalline metal complexes appears to be a promising undertaking.We have recently employed the formyl-substituted isocyanoarenes 4-CN-3,5-R2C6H2CHO (R=Me, Pri) as versatile precursors for the design of isocyanide metal complexes which are capable of undergoing molecular self-assembly.20 The formyl-substituted isocyanoarenes also lend themselves as precursors for isocyanides bearing long terminal chains.In this contribution, we describe the synthesis of a series of such derivatives by simple SchiV-base formation with long-chain 4- alkoxyanilines. The new isocyanides form linear complexes with PdI2 and PtI2 which exhibit mesomorphic properties (with R=Me). Results and Discussion Synthesis and characterization of the isocyanides and isocyanide N R R C CHO H2N R R OCnH2 n+1 + N R R C C H N OCn H2 n+1 i L5 n = 5 L12 n = 12 L6 n = 6 L14 n = 14 L8 n = 8 L16 n = 16 L10 n = 10 L18 n = 18 L11 n = 11 L4¢ n = 4 L8¢ n = 8 L5¢ n = 5 L11¢ n = 11 R =CH3 R =CH(CH3)2 N R R C C H N OCn H2 n+1 Ml2 + L n or L n ¢ ii Ml2 2 1–9 M = Pd L5–L18 10–18 M = Pt L5–L18 19–21 M = Pd L5¢, L8¢, L11¢ 22–25 M = Pt L4¢–L11¢ metal complexes Scheme 1 Reagents and conditions: i, ethanol, acetic acid, rt, stirring; ii, THF, rt, stirring The SchiV-bases 4-CN-3,5-R2C6H2CHNC6H4-4-OCnH2n+1, Ln (R=CH3, n=5, 6, 8, 10, 11, 12, 14, 16, 18) and Ln¾ (R=Pri, n=4, 5, 8, 11) were obtained by reaction of isocyanobenzal- 2117 cm-1 for the isocyanide functionality.21 The 13C NMR dehydes 4-CN-3,5-R2C6H2CHO with the corresponding ani- resonance for the isocyanide carbon atom is of low intensity lines in ethanol at room temperature (Scheme 1).The new and was found for a few examples near d 169. The imino isocyanides exhibit a characteristic IR absorption at about groups give rise to 1H and 13C NMR resonances at around d 8.4 and 158, respectively. The complexes trans-MI2(Ln)2 and trans-MI2(Ln¾)2 (M=Pt, Pd), 1–25, form upon combination †Currently on leave from SUNY, Stony Brook, USA.J. Mater. Chem., 1998, 8(7), 1561–1565 1561of the isocyanides with PdI2 and PtI2 in THF (Scheme 1). The trans-geometry of the palladium and platinum complexes is evident from the appearance of only single IR absorptions for the coordinated isocyanide groups at ca. 2196 and 2189 cm-1, respectively. The shift of the isocyanide stretching frequencies to higher wavenumbers by 70–80 cm-1 upon coordination to the PdII and PtII metal centers is of the expected magnitude.21 The 13C NMR resonances of the metal-coordinated isocyanide carbon atoms have not been observed due to their low intensity, but would be expected to occur in the same region as for the free ligands.22 The metal centers also exert very little influence on the 1H NMR resonances of the isocyanides.Crystal structures of complexes 22 and 24 Attempts to grow crystals suitable for X-ray crystallographic studies of the mesogenic complexes 1–18 have not been successful. Instead, the molecular and crystal structures of compounds 22 and 24 have been determined. Both compounds crystallize in centrosymmetric space groups (P21/c and P2/c, respectively). The crystal parameters and information on data collection and refinement are summarized in Table 1.The Fig. 1 (a) Molecular structure of compound 22. One of the two symmetry-related isocyanide ligands is truncated to show only the molecular structures of 22 and 24 as well as packing diagrams CN-group. (b) Packing diagram of compound 22. are shown in Fig. 1 and 2 while bond lengths and angles for 22 and 24 are given in Tables 2 and 3, respectively.The metal atoms are located on centers of inversion. The intramolecular structural parameters of the diiodo–bis-isocyanide metal cores as well as the extended organic groups are within the expected range.14 The length of the unsaturated core, as measured by the distance between the two oxygen atoms in the 4-positions of the arylimine portions, is ca. 30 A ° . In the crystal of compound 22, the molecules are stacked along their flat faces. A lateral shift of ca. 4 A° between adjacent molecular units allows the extended p systems to face each other directly despite the sterically demanding iodide ligands and isopropyl groups [Fig. 1(b)]. Nevertheless, the shortest intermolecular atomic contacts between the unsaturated systems are still longer than 3.7 A ° .In the crystal of compound 24, the molecules are also stacked along their flat sides. There is a long lateral shift of ca. 16 A° , as estimated by the PtMPt separation within the Table 1 Crystal and data collection parameters data for complexes 22 and 24 Fig. 2 (a) Molecular structure of compound 24.One of the two symmetry-related isocyanide ligands is truncated to show only the 22 24 CN-group. (b) Packing diagram of compound 24. formula C48H60I2N4O2Pt C56H76I2N4O2Pt Mw 1173.93 1286.14 Table 2 Selected bond lengths (A ° ) and angles (degrees) for complex 22 crystal system monoclinic monoclinic space group P21/c (no.14) P2/c (no.13) Pt(1)MI(1) 2.6013(7) Pt (1)MC(1) 1.979(8) a/A° 14.584(2) 15.827(3) N(1)MC(1) 1.112(9) N(1)MC(2) 1.395(9) b/A ° 6.030(2) 10.085(2) N(2)MC(14) 1.245(10) N(2)MC(15) 1.418(9) c/A ° 27.184(3) 18.826(3) C(2)MC(3) 1.40(1) C(5)MC(14) 1.47(1) b/degrees 91.96(2) 107.54(2) I(1)MPt(1)MI(1) 180.0 I(1)MPt(1)MC(1) 90.5(2) V /A° 3 2389.2(7) 2865.2(9) I(1)MPt(1)MC(1) 89.5(2) C(1)MPt(1)MC(1) 180.0 Z 2 2 C(1)MN(1)MC(2) 178.0(8) C(14)MN(2)MC(15) 120.4(7) T /°C 28 28 Pt(1)MC(1)MN(1) 176.9(7) N(1)MC(2)MC(3) 116.6(8) crystal color yellow yellow N(1)MC(2)MC(7) 119.3(8) N(2)MC(14)MC(5) 122.7(8) Dc/g cm-3 1.632 1.491 N(2)MC(15)MC(16) 115.2(7) N(2)MC(15)MC(20) 126.9(7) m/cm-1 42.57 35.57 diVractometer MAR-IPDS MAR-IPDS 2hmax/degrees 51.2 48.9 Table 3 Selected bond lengths (A ° ) and angles (degrees) for complex 24 unique reflections 4097 3289 reflections used in 2694 1692 Pt(1)MI(1) 2.600(1) Pt (1)MC(1) 1.94(1) least-squares N(1)MC(1) 1.13(1) N(1)MC(2) 1.43(2) refinement N(2)MC(14) 1.23(1) N(2)MC(15) 1.47(2) no.of variables 259 295 C(2)MC(3) 1.38(2) C(5)MC(14) 1.50(2) Ra 0.044 0.041 I(1)MPt(1)MI(1) 180.0 I(1)MPt(1)MC(1) 90.4(4) Rw a 0.054 0.045 I(1)MPt(1)MC(1) 89.6(4) C(1)MPt(1)MC(1) 180.0 (D/s)max 0.02 0.00 C(1)MN(1)MC(2) 178(1) C(14)MN(2)MC(15) 120(1) goodness of fit 2.23 1.72 Pt(1)MC(1)MN(1) 177(1) N(1)MC(2)MC(3) 115(1) Dr/e A° -3 2.00–1.00 0.87–0.40 N(1)MC(2)MC(7) 117(1) N(2)MC(14)MC(5) 122(1) N(2)MC(15)MC(16) 125(1) N(2)MC(15)MC(20) 115(1) aR=.||FO|-|Fc||/.|FO|; Rw=[.w(|FO|-|Fc|)2/.wFO2]1/2; w= 1/s2(FO)=4FO2/s2(FO2). 1562 J. Mater. Chem., 1998, 8(7), 1561–1565Table 4 Mesomorphic properties of the complexes 1–25a Table 4 (continued) DS/J K-1 DS/J K-1 compound transition T /°C DH/kJ mol-1 mol-1 compound transition T /°C DH/kJ mol-1 mol-1 18 (Pt, L18) C1–C2 110 66.5 173.6 1 (Pd, L5) C–N 257 37.0 69.8 N–I 284 —b C2–SC 141 43.8 105.8 SC–I 271 13.8 25.4 2 (Pd, L6) C–N 228 58.8 117.4 19 (Pd, L5 ¾ ) C–I 212 73.9 152.4 N–I 275 —b 20 (Pd, L8 ¾ ) C–I 165 49.9 113.9 3 (Pd, L8) C–N 223 53.7 108.3 N–I 294 —b 21 (Pd, L11¾ ) C–I 172 55.0 123.6 4 (Pd, L10) C1–C2 138 27.7 67.4 22 (Pt, L4¾ ) C–I 187 27.1 58.9 C2–SC 198 36.7 77.9 SC–N 235 1.5 3.0 23 (Pt, L5¾ ) C–I 209 55.3 114.4 N–I 282 —c 24 (Pt, L8¾ ) C–I 168 38.3 86.8 5 (Pd, L11) C1–C2 107 46.9 123.4 C2–SC 190 39.2 84.7 25 (Pt, L11¾ ) C–I 162 44.2 101.6 SC–N 254 1.9 3.6 N–I 290 3.1 5.5 aC=crystal; SC=smectic C; N=nematic; I=isotropic.bPartial decomposition. cNot detected by DSC. dPeaks not resolved. 6 (Pd, L12) C1–C2 94 24.2 65.9 C2–C3 136 9.9 24.2 C3–SC 178 42.4 94.0 SC–N 263 2.9 5.4 rows [Fig. 2(b)]. The shortest intermolecular contacts between N–I 287 4.6 8.2 the p systems are in the range of 3.53–3.69 A ° . Even though the intermolecular distances between adjacent arene rings in 7 (Pd, L14) C1–C2 117 18.6 47.7 C2–SC 174 49.1 109.8 the crystal of 22 are longer than those of 24, the intermolecular SC–N 269 3.7 6.8 contacts of the unsaturated systems in 22 are more extensive.N–I 280 2.9 5.2 This feature may, at least in part, be responsible for the higher melting point of 22 compared to 24. 8 (Pd, L16) C1–C2 104 37.1 98.4 C2–SC 168 56.5 128.1 Liquid crystal properties of the palladium and platinum SC–N 271 12.8d 23.5 N–I 281 —d complexes The palladium and platinum isocyanide complexes 1–25 con- 9 (Pd, L18) C1–C2 76 8.9 C2–C3 108 47.0 123.6 tain organic fragments, which are well established components C2–SC 152 45.2 106.4 of organic liquid crystals,23 and metal complex cores, which SC–I 268 14.1 26.1 have recently been demonstrated to possess favorable mesogenic properties.13,14 Azomethine-bridged aromatic com- 10 (Pt, L5) C–N 244 35.5 68.7 pounds with terminal alkoxy groups have been the subject of N–I 318 —b many studies concerning liquid crystals.23,24 The conjugated 11 (Pt, L6) C–N 206 53.6 111.9 azomethine bridge increases the longitudinal polarizability N–I 301 —b while maintaining a relatively rigid arrangement of connected aromatic rings.Since SchiV-base linkages are easily established, 12 (Pt, L8) C–N 219 59.3 120.5 they provide a convenient means to modify liquid crystal N–I 295 —b properties systematically, even though their sensitivity towards hydrolysis limits practical applications. The palladium and 13 (Pt, L10) C1–C2 136 26.9 65.8 C2–SC 195 36.5 78.0 platinum metal centers extend the length of the linear isocyan- SC–N 249 1.4 2.8 ides and establish the rigid core.Only the diiodo–palladium N–I 284 —c and –platinum fragments were employed. Takahashi and coworkers8 have shown that these possess superior mesomorphic 14 (Pt, L11) C1–C2 104 38.2 100.1 properties over the chloro- and bromo-analogues, primarily C2–SC 179 35.5 78.9 due to the higher stability of the trans configuration.The SC–N 262 2.0 3.8 N–I 290 —c mesomorphic behaviour of the free isocyanides has not been studied. 15 (Pt, L12) C1–C2 106 25.5 67.3 The thermal properties of complexes 1–25 were investigated C2–C3 147 9.2 21.9 by a combination of diVerential scanning calorimetry (DSC) C3–SC 167 30.5 69.3 and polarized optical microscopy. The N phases show marbled SC–N 266 2.7 5.0 and schlieren textures, while the SC phases exhibit fan-shaped N–I 293 3.3 5.8 textures (broken fan) on cooling from the N or I phases.25 The 16 (Pt, L14) C1–C2 124 19.1 48.11 phase transition temperatures and the associated thermal data C2–SC 168 55.9 126.8 are listed in Table 4.A plot of the phase transition temperatures SC–N 277 4.8 8.7 of complexes 1–18 versus the length of the alkyl chains is N–I 289 3.4 6.4 shown in Fig. 3. Only the palladium and platinum complexes with the lateral methyl groups form liquid crystalline phases. 17 (Pt, L16) C1–C2 110 43.3 113.1 C2–SC 161 56.3 129.7 Both the transition temperatures involving the crystalline (C–N SC–N 277 12.5d 22.7 or C–SC) and the isotropic phases (N–I or SC–I) decrease with N–I 280 —d increasing length of the alkyl chains, whereby the melting points decrease faster than the clearing points.Thus the temperature range of the mesophases becomes broader as the J. Mater. Chem., 1998, 8(7), 1561–1565 1563were performed on complexes 5, 9, 14 and 18. At a heating rate of 20 °C min-1, the palladium complexes began to decompose at around 335 °C, the platinum compounds near 350 °C.Below those temperatures, no significant signs of decomposition could be discovered at that heating rate. Experimental Measurements and reagents The 1H and 13C NMR spectra were recorded on Bruker DPX 300 or JEOL 270 spectrometers in CDCl3 with tetramethylsilane as an internal standard. IR spectra in CH2Cl2 were obtained on a Shimadzu FTIR-8201PC IR spectrophotometer.Mass spectra were obtained on a Finnigan MAT 95 mass spectrometer. The textures of the mesophases were studied with a Leica polarizing microscope equipped with a Mettler Fig. 3 Mesomorphic transition temperatures (°C) vs. the length of the FP82 hot stage. The transition temperatures were obtained by alkoxy chain (OCnH2n+1) for complexes 1–18 diVerential scanning calorimetry using a Perkin-Elmer DSC7 with heating rates of 10 or 20 °C min-1 under a nitrogen atmosphere.The data were taken from the first heating scan. length of the alkyl chain increases. The compounds with alkyl Elemental microanalyses were conducted by Butterworth chain lengths of n=5–8 exhibit only a nematic mesophase. Laboratories Ltd. From n=10 on, the appearance of a smectic C phase is 4-Isocyano-3,5-dimethylbenzaldehyde and 4-isocyano-3,5- observed, which reaches its maximum stability at n=16–18.diisopropylbenzaldehyde were obtained as previously For n=18, the nematic phase is no longer observed. This described.20 The 4-alkoxynitrobenzenes and corresponding 4- dependence of the thermal behaviour on the length of the alkyl alkoxyanilines were prepared following literature methods.28,29 chains is fully consistent with models developed for linear All ligands and metal complexes were prepared in the same organic compounds possessing a rigid lath-shaped, elecmanner as outlined in Scheme 1.Typical preparative protronically conjugated core and two flexible terminal cedures and physical data for the ligands and complexes are groups.24,25 For the complexes with very long alkyl chains, given for L8 and 12.The spectroscopic data for 6 are given as untypically high values for the SC–N or SC–I transition enthalpa typical example for the palladium complexes. Most of the ies were found.24 The physical meaning of these data is unclear. new isocyanides were obtained as oils which are diYcult to High transition enthalpy values have also been observed for purify and crystallize.Therefore, no elemental analyses were other liquid crystals with long terminal alkyl chains.26 The performed on these compounds. The palladium and platinum platinum complexes exhibit on average slightly lower melting complexes were all prepared following the procedure described points and higher clearing points than the palladium comfor 12.Satisfactory elemental analyses were obtained for all pounds. Thus the platinum complexes form the more stable isocyanide metal complexes. A complete set of spectroscopic mesophases with the broader temperature ranges. This is in and microanalytical data is deposited as supplementary matecontrast to the results by Takahashi and co-workers14 who rial (SUP 57376).‡ found the reverse situation for the compounds [MI2(4- CNC6H4-4-O2CC6H4OCnH2n+1)2].Nevertheless, since the Selected syntheses and characterisation diVerences in phase transition temperature between the complexes of the two metals are relatively small, our results confirm L8. A mixture of 4-isocyano-3,5-dimethylbenzaldehyde the conclusion reached earlier by Kaharu and Takahashi12 (160 mg, 1 mmol), 4-octyloxyaniline (221 mg, 1 mmol) and 2 that for a given ligand set a change of metal center exerts only drops of acetic acid in ethanol (20 ml ) was stirred at room a minor influence.The fact that the mesophases of complexes temperature. After stirring for 4 h, the solvent was evaporated 1–18 exhibit a broader temperature range than the Takahashi under vacuum to near dryness to give a light yellow solid compounds14 is of interest.Considering that these compounds which was further purified by recrystallization from and the complexes reported here possess the same number of ethanol–THF (551, v/v) (312 mg, 89%). 1H NMR (CDCl3): d aromatic groups, it appears reasonable to speculate that the 8.41 (1 H, s, MCHNN) 7.63 (2 H, s, Ar-H), 7.24 (2 H, d, J= lateral methyl groups in compounds 1–18 may have a favour- 8.8 Hz, Ar-H), 6.93 (2 H, d, J=8.8 Hz, Ar-H), 3.97 (2 H, t, able influence on the mesomorphic properties.The methyl J=6.6 Hz, MOCH2M), 2.51 (6 H, s, Ar-CH3), 1.81–1.91 (12 groups extend laterally about as far as the two iodide ligands, H, m, C6H12), 0.89 (3 H, t, J=6.6 Hz, MCH3). 13C NMR thus giving the central portion of the molecules a more even (CDCl3, 270 MHz): d 169.4, 158.4, 156.3, 144.0, 136.4, 135.4, lath-like shape.23,27 The lack of mesomorphic character of the 127.7, 122.3, 115.1 (CNN, C6H2, C6H4), 68.3 (CH2O), 31.8, metal complexes with the lateral isopropyl groups can be 29.4, 29.3, 26.1, 22.7 (CH2), 18.9 (Ar-CH3), 14.1 (CH3).IR attributed to a disruption of the intermolecular interactions (CH2Cl2, cm-1): n(NOC) 2119 s.MS (EI) m/z: 362 (M). by these bulky substituents. The isopropyl-substituted complexes also have distinctly lower melting points than the PtI2(L8)2 (12). To a solution of L8 (71 mg, 0.2 mmol) in methyl-substituted compounds, by as much as 51 °C for M= 15 ml THF was added PtI2 (45 mg, 0.1 mmol) with stirring Pt, n=8. under N2 atmosphere at room temperature.After stirring for The isocyanide metal complexes, especially the palladium 2 h, the resultant yellow solution was filtered, and the solvent complexes with the shorter side chains, showed signs of was removed under vacuum. The residue was reprecipitated decomposition during the investigation of the phase transitions. from THF–n-hexane to give a yellow microcrystalline solid.Slow heating of the samples during the polarized optical Yield: 68 mg (58%); 1H NMR (CDCl3, 300 MHz): d 8.44 (1 microscopy studies caused a marked darkening in colour. For H, s, MCHNN), 7.68 (2 H, s, Ar-H), 7.26 (2 H, d, J=8.9 Hz, this reason, the DSC scans were performed at a high rate of Ar-H), 6.95 (2 H, d, J=8.9 Hz, Ar-H), 3.98 (2 H, t, J=6.6 Hz, 20 °C min-1.The data for selected examples were also measured at a scan rate of 10 °C min-1, confirming the results ‡Available as supplementary material (SUP 57376; 18 pp.) deposited with the British Library. Details are available from the editorial oYce. obtained at the higher scan rate. Thermogravimetric analyses 1564 J. Mater. Chem., 1998, 8(7), 1561–15652 N. Hagihara, S.Sonogashira and S. Takahashi, Adv. Polym. Sci., MOCH2M), 2.64 (6 H, s, Ar-CH3), 1.82–1.29 (12 H, m, C6H12), 1980, 41, 149. 0.89 (3 H, t, J=6.7 Hz, CH3); 13C NMR (CDCl3, 270 MHz): 3 S. Takahashi, Y. Takai, H. Morimoto, K. Sonogashira and d 158.6, 155.7, 143.7, 137.9, 137.3, 128.1, 122.4, 115.1 (CHNN, N. Hagihara,Mol. Cryst. L iq. Cryst., 1982, 82, 139. C6H2, C6H4), 68.3 (CH2O), 31.8, 29.4, 29.3, 26.1, 22.7 (CH2), 4 S.Takahashi, H. Morimoto, E. Murata, K. Kariya, K. Sonogashira 19.1 (Ar-CH3), 14.1 (CH3). IR (CH2Cl2, cm-1): n(NOC) and N. Hagihara, J. Polym. Sci., 1982, 20, 565. 5 S. Takahashi, Y. Takai, H. Morimoto and K. Sonogashira, 2189 s; Mass (FAB) m/z: 1174 (M+1). Anal. Calc. for J. Chem. Soc., Chem. Commun., 1984, 3. C48H60I2N4O2Pt: C, 49.11; H, 5.12; N, 4.77.Found: C, 49.15; 6 T. Kaharu, H. Matsubara and S. Takahashi, J.Mater. Chem., 1992, H, 5.16; N, 4.84%. 2, 43. 7 T. Kaharu, H. Matsubara and S. Takahashi, Mol. Cryst. L iq. PdI2(L12 )2 (6). Light orange crystals (76% yield); 1H NMR Cryst., 1992, 220, 191. (CDCl3, 300 MHz): d 8.44 (1 H, s, MCHNN), 7.68 (2 H, s, 8 T. Kaharu, R. Ishii, T. Adachi and S. Takahashi, J. Mater.Chem., Ar-H), 7.26 (2 H, d, J=8.8 Hz, Ar-H), 6.95 (2 H, d, J=8.8 Hz, 1995, 5, 687. 9 J. P. Rourke, D. W. Bruce and T. B. Marder, J. Chem. Soc., Dalton Ar-H), 3.98 (2 H, t, J=6.6 Hz, MOCH2M), 2.64 (6 H, s, Ar- T rans., 1995, 317. CH3), 1.80–1.27 (20 H, m, C10H20), 0.88 (3 H, t, J=6.7 Hz, 10 D. W. Bruce, M. S. Lea and J. Marsden, Mol. Cryst. L iq. Cryst., CH3); 13C NMR (CDCl3, 270 MHz): d 158.6, 155.7, 143.7, 1996, 275, 183. 137.9, 137.3, 128.1, 122.4, 115.1, 68.3, 31.9, 29.7, 29.6, 29.4, 29.3, 11 T. Kaharu, H.H. Matsubara and S. Takahashi, Mol. Cryst. L iq. 26.1, 22.7, 19.1, 14.1; IR (CH2Cl2, cm-1): n(NOC) 2196 s; Mass Cryst., 1992, 220, 191. (FAB) m/z: 1196 (M). Anal. Calc. for C56H76I2N4O2Pd: C, 12 T. Kaharu and S. Takahashi, Chem. L ett., 1992, 1515. 13 T. Kaharu, R. Ishii and S. Takahashi, J. Chem. Soc., Chem. 56.17; H, 6.35; N, 4.68. Found: C, 56.33; H, 6.53; N, 4.69%. Commun., 1994, 1349. 14 T. Kaharu, T. Tanaka, M. Sawada and S. Takahashi, J. Mater. Crystal structure determinations Chem., 1994, 4, 859. 15 M. Benouazzane, S. Coco, P. Espinet and J. M. Martin-Alvarez, The crystal parameters of compounds 22 and 24 and infor- J.Mater.Chem., 1995, 5, 441. mation on data collection and refinement are summarized in 16 S. Coco, P. Espinet, S. Flanagan and J. M. Martin-Alvarez, New. Table 1. DiVraction data for compounds 22 and 24 were J. Chem., 1995, 19, 959. collected on a MAR Imaging Plate Detector System using 17 P. Alejos, S. Coco and P. Espinet, New. J. Chem., 1995, 19, 799. graphite-monochromatized Mo-Ka X-ray radiation (l= 18 S.Coco, P. Espinet, J. M. Martin-Alvarez and A.-M. Levelut, J.Mater. Chem., 1997, 7, 19. 0.701 73 A ° ) from an MAR generator (sealed tube 50 kV and 19 R. Bayon, S. Coco, P. Espinet, C. Fernandez-Mayordomo and 50 mA), and processed by DENZO.30 No absorption correc- J. M. Martin-Alvarez, Inorg. Chem., 1997, 36, 2329. tions were made. All structure determinations were done using 20 K.Y. Lau, A. Mayr and K. K. Cheung, Inorg. Chim. Acta, in press. the MSC crystal structure analysis package TeXsan31 and the 21 H. E. Oosthuizen and E. Singleton, Adv. Organomet. Chem., 1983, full-matrix least-squares refinements were on F using reflections 22, 209. with I>3s(I). Hydrogen atoms at calculated positions with 22 J. Guo and A. Mayr, Inorg. Chim.Acta, 1997, 261, 141. 23 T hermotropic L iquid Crystals, Critical Reports on Applied thermal parameters equal to 1.3 times that of the attached C Chemistry, ed. G. W. Gray, John Wiley & Sons, Chichester, 1987, atoms were included in the calculations, but not refined. vol. 22. Full crystallographic details, excluding structure factors, 24 G. W. Gray and J. W. G. Goodby, Smectic L iquid Crystals, have been deposited at the Cambridge Crystallographic Data T extures and Structures, Leonard Hill, Glasgow and London, Centre (CCDC). See Information for Authors, J.Mater. Chem., 1984. 1998, Issue 1. Any request to the CCDC for this material 25 D. Demus and L. Richter, T extures of L iquid Crystals, Verlag Chemie, Weinheim, New York, 1978. should quote the full literature citation and the reference 26 (a) R. Deschenaux, I. Kostics and B. Nicolet, J.Mater. Chem., 1995, number 1145/94. 5, 2291; (b) J. Szydlowska, W. Pyzÿuk, A. Kro�wczyn� ski and I. Bikchantaev, J. Mater. Chem., 1996, 6, 733; (c) D. Pociecha, Support for this work by the Hong Kong Research Grants A. Kro�wczyn� ski, J. Szydlowska, E. Go� recka, M. Glogarova and Council is gratefully acknowledged.We thank Prof. G. Heppke J. Przedmojski, J.Mater. Chem., 1997, 7, 1709. and T. Fuetterer for helpful discussions. 27 S. M. Kelly, J. Chem. Soc., Chem. Commun., 1983, 366. 28 P. R. P. TuYn, K. J. Toyne and J. W. Goodby, J. Mater. Chem., 1996, 6, 1271. References 29 D. L. J. Clive, A. G. Angoh and S. M. Bennet, Inorg. Chem., 1987, 52, 1339. 1 (a) D. W. Bruce, in Inorganic Materials, ed. D. W. Bruce and 30 DENZO: in T he HKL Manual—A description of programs D. O’Hare, John Wiley & Sons, Chichester, UK, 1992; DENZO, XDISPL AYF, and SCAL EPACK, written by D. Gewirth (b) P. Espinet, M. A. Esteruelas, L. A. Oro, J. L. Serrano and with the cooperation of the program authors. E. Sola, Coord. Chem. Rev., 1992, 117, 215; (c) S. A. Hudson and P. Maitlis, Chem. Rev., 1993, 93, 861; (d) Metallomesogens, ed. J.-L. Serrano, VCH, Weinheim, Germany, 1996. Paper 8/02468E; Received 31stMarch, 1998 J. Mater. Chem., 1998, 8(7), 1561&ndash

ISSN:0959-9428    

DOI:10.1039/a802468e

出版商:RSC    

年代:1998

数据来源: RSC